Embed Size (px)
Citation preview
Luís Manuel Valla Teixeira
Regulation of Drosophila metabolism by the transcription factor Slbo and the PAT family
member Lsd2
A thesis submitted for the degree of Doctor of Philosophy in Biomedical Sciences
Faculdade de Medicina Universidade Lisboa
European Molecular Biology Laboratory
2004
Luís Manuel Valla Teixeira
Regulation of Drosophila metabolism by the transcription factor Slbo and the PAT family
member Lsd2
Supervisors:
Doctor Pernille Rørth Developmental Biology Programme
European Molecular Biology Laboratory
Professor Doctor Maria do Carmo-Fonseca
Instituto de Medicina Molecular
Faculdade de Medicina
Universidade de Lisboa
“But it must be said from the outset that a disease is
never a mere loss or excess – that there is always a
reaction, on the part of the affected organism or
individual, to restore, to replace, to compensate for
and to preserve its identity, however strange the
means may be.”
The man who mistook is wife for a hat
(Oliver Sacks 1986)
ACKNOWLEDGEMENTS
First and foremost I am deeply grateful to Pernille Rørth for giving me the liberty
to pursue this project in her lab, for the numerous discussions, suggestions and
advices and for the invaluable support throughout these years. I was fortunate to
do my PhD in her lab.
I also want to acknowledge Carmo-Fonseca for co-supervising this PhD. I want
to thank all the help and support, from the time I did my diploma thesis in her lab
until now.
I thank the members of my EMBL thesis advisory committee, Mathias Treier and
Iain Mattaj, for advice. I would like to thank Steven Cohen for the occasional but
valued support and enthusiasm for what I was doing.
I had a pleasant and fruitful collaboration with Nathalie Vanzo concerning the
Lsd2 story. I thank her, Anne Ephrussi and Catherine Rabouille for that project.
The microarrays experiments would not have been done without the help of
Bélen Miñana and Vladimir Benes and I am grateful to them. I also want to thank
Eileen Furlong and Giorgios Christophides for advices and discussion concerning
the microarrays. The fly embryo sorter was, in the beginning, more of a problem
than a solution, fortunately I faced that challenge together with Anne Atzberg. I
am also grateful to Ann Mari Voie for injecting embryos to generate transgenic
flies. I also thank Andrea Herold and Eliza Izaurralde for having me collaborating
with them. Tiago Marques frequently helped me with statistics. Since you can
prove anything with statistics this has been a priceless help.
I am thankful to all the members of the lab - Simone Beccari, Lodovica Borghese,
Peter Duchek, Tudor Fulga, Andreea Gruia, Gáspár Jékely, Oguz Kanca, Li Lin,
Carlos Luque, Juliette Mathieu, Anne Pacquelet, Kálmán Somogyi, Hsin-Ho
Sung and Gemma Texido – for the excellent working environment, help and
discussions. I would also like to thank Mark Coyne which helped me as a
summer student in the lab.
I had the unique honour of belonging to both the Programa Gulbenkian de
Doutoramento em Biologia e Medicina and the EMBL PhD programme. I am
grateful to both of them for accepting me. I want to thank the PGDBM for the
excellent first year of my PhD. I must also acknowledge the exceptional scientific
environment at EMBL. I was supported with a PhD fellowship from Fundação
para a Ciência e Tecnologia.
I am thankfull to Steve Cohen, Alain Debec, Veit Riechmann and Sérgio Simões
for fly stocks and antibodies.
Many other people helped me and contributed to this work, I am grateful to all.
PREFÁCIO Nesta dissertação são apresentados os resultados do trabalho de investigação
desenvolvido no European Molecular Biology Laboratory (EMBL), em
Heidelberg, na Alemanha, entre Setembro de 1999 e Setembro de 2003. Este
trabalho foi realizado sob a orientação da Doutora Pernille Rørth (EMBL) e da
Professora Doutora Maria do Carmo-Fonseca (Faculdade de Medicina da
Universidade de Lisboa).
A investigação centrou-se na análise da função de dois genes do organismo
modelo Drosophila melanogaster. slbo (slow border cells) codifica um factor de
transcrição homólogo das proteínas C/EBP (CCAAT/enhancer binding protein)
em mamíferos. Lsd2 (Lipid storage droplet 2) codifica uma proteína envolvida no
metabolismo de lípidos. O estudo da função de lsd2 foi feita em colaboração
com a Doutora Nathalie Vanzo do laboratório da Doutora Anne Ephrussi. Na
medida do possível apresentarei a minha parte da investigação. No entanto será
ocasionalmente necessário referir resultados obtidos por Nathalie de modo a
manter a estrutura lógica do trabalho.
As duas partes deste trabalho são suficientemente diferentes para serem
tratadas separadamente. Para cada um dos genes escreverei uma introdução, a
descrição dos resultados e uma discussão. Une os dois projectos o estudo, em
Drosophila, de genes reguladores de processos metabólicos.
Como previsto no Artigo 15º do Regulamento de Doutoramento da Universidade
de Lisboa, a presente dissertação encontra-se redigida em língua inglesa,
contendo um resumo alargado (cerca de 1200 palavras) em língua portuguesa
(Sumário).
As opiniões expressas nesta publicação são da exclusiva responsabilidade do
seu autor.
Parte do trabalho desta tese foi publicado em:
Beccari S., Teixeira L. and Rørth P. (2002). The JAK/STAT pathway is required
for border cell migration during Drosophila oogenesis. Mech Dev. 111(1-2): 115-
23
Contribui para este artigo com a hibridização in situ dos genes Stat92E e
unpaired.
Herold A., Teixeira L. and Izaurralde, E. (2003). Genome-wide analysis of
nuclear mRNA export pathways in Drosophila. EMBO J. 22(10): 2472-83
Contribui para este artigo produzindo os DNA microarrays.
Teixeira L., Rabouille C., Rørth P., Ephrussi A. and Vanzo, N. (2003).
Drosophila Perilipin/ADRP homologue Lsd2 regulates lipid metabolism. Mech.
Dev. 120(9): 1071-1081
A minha contribuição para este artigo é o conteúdo da segunda parte da
dissertação.
TABLE OF CONTENTS
1
TABLE OF CONTENTS
LIST OF FIGURES 4
ABBREVIATIONS 6
1 SUMÁRIO 8
1.1 REGULAÇÃO DE METABOLISMO PELO FACTOR DE TRANSCRIÇÃO SLBO 8 1.2 REGULAÇÃO DO METABOLISMO DE LÍPIDOS POR LSD2 11
2 SUMMARY 13
2.1 METABOLISM REGULATION BY THE TRANSCRIPTION FACTOR SLBO 13 2.2 LIPID METABOLISM REGULATION BY LSD2 14
METABOLISM REGULATION BY THE TRANSCRIPTION FACTOR SLBO 16
3 INTRODUCTION 17
3.1 STRUCTURE AND DNA BINDING OF C/EBPS 18 3.2 MAMMALIAN C/EBPS FUNCTION 21 3.2.1 C/EBPS AND THE ADIPOSE TISSUE 22 3.2.2 C/EBPS AND LIVER FUNCTION 23 3.2.3 C/EBP FUNCTION IN OTHER TISSUES 26 3.3 DROSOPHILA C/EBP 27 3.3.1 SLBO STRUCTURE AND DNA BINDING 27 3.3.2 SLBO FUNCTION 28 3.4 AIM OF THE PROJECT 29
4 RESULTS 30
4.1 FINDING SLBO TARGET GENES - CANDIDATE GENES APPROACH 30 4.1.1 EXPRESSION PATTERN OF SLBO IN EMBRYOS 30 4.1.2 EXPRESSION PATTERN OF THREE CANDIDATE SLBO TARGET GENES 32 4.1.3 SORTING OF HOMOZYGOUS SLBO LOSS-OF-FUNCTION MUTANT EMBRYOS 34 4.1.4 EXPRESSION OF THREE CANDIDATE TARGET GENES IN CONTROL AND SLBO MUTANT EMBRYOS. 37 4.1.5 EXPRESSION PATTERN OF SLBO IN EGG CHAMBERS 39 4.1.6 ANALYSIS OF CANDIDATE SLBO TARGET GENES IN BORDER CELLS 42 4.2 FINDING SLBO TARGET GENES - GENOMICS APPROACH 46 4.2.1 DROSOPHILA CDNAS MICROARRAYS 46 4.2.2 TRIAL MICROARRAY ANALYSIS 48
TABLE OF CONTENTS
2
4.2.3 EMBRYO SORTING WITH THE COPAS SELECT SYSTEM. 51 4.2.4 DNA MICROARRAYS OF CONTROL VS SLBO MUTANT EMBRYOS 53 4.3 SLBO AND METABOLISM REGULATION 62 4.3.1 SLBO MUTANTS HATCHING DEFECT AND LARVAL LETHALITY 62 4.3.2 SLBO MUTANTS AND METHYL P-HYDROXYBENZOATE SENSITIVITY 64 4.3.2.1 slbo mutants are very sensitive to methyl p-hydroxybenzoate 64 4.3.2.2 DmGlcAT-BSII and slbo mutant methyl p-hydroxybenzoate sensitivity 68 4.3.3 SLBO MUTANTS HAVE A GROWTH DEFECT AND AN ALTERED FEEDING BEHAVIOUR 70 4.3.4 SUGAR METABOLISM RELATED GENES MIS-REGULATED IN THE SLBO MUTANT 74 4.3.5 SLBO AND POLYAMINES 77 4.3.5.1 Ornithine decarboxylase and polyamines function 77 4.3.5.2 Odc1 and Odc2 79 4.3.5.3 Odc1 and slbo expression in the fat body 82 4.3.5.4 Genes involved in ornithine metabolism up-regulated in slbo mutants 84 4.3.5.5 Interference with ornithine decarboxylase function 86 4.3.5.6 Odc1 expression and suppression of slbo phenotype 90 4.3.6 SLBO MUTANT AND THE EIF4E HOMOLOGUE CG8023 92 4.4 SLBO AND INNATE IMMUNITY 98 4.4.1 SLBO MUTANTS LARVAE HAVE SPONTANEOUS MELANIZATION 98 4.4.2 INNATE IMMUNITY RELATED GENES MIS-REGULATED IN THE SLBO MUTANT 100 4.5 SLBO FUNCTION IN LARVAL STAGES 105 4.5.1 SLBO EXPRESSION IN LARVAE 105 4.5.2 SLBO REQUIREMENT DURING LARVAL DEVELOPMENT 106
5 DISCUSSION 111
5.1 DNA MICROARRAYS OF CONTROL VS SLBO MUTANT EMBRYOS 111 5.2 SLBO MUTANTS AND METHYL P-HYDROXYBENZOATE SENSITIVITY 113 5.3 GROWTH DEFECT AND ALTERING FEEDING BEHAVIOUR 114 5.4 SLBO AND POLYAMINES METABOLISM 116 5.5 SLBO MUTANT AND THE EIF4E HOMOLOGUE CG8023 119 5.6 SLBO AND INNATE IMMUNITY 120 5.7 OTHER FUNCTIONS OF SLBO 122 5.8 SLBO FUNCTION IN LARVAL STAGES 123 5.9 CONCLUDING REMARKS 125
LIPID METABOLISM REGULATION BY LSD2 127
6 INTRODUTION 128
7 RESULTS 131
7.1 ISOLATION OF A DROSOPHILA LSD2 MUTANT 131 7.2 DROSOPHILA LSD2 PATTERN OF EXPRESSION 131 7.3 ABNORMAL ACCUMULATION OF NEUTRAL LIPIDS IN THE GERMLINE AND EGGS OF LSD21 FEMALES 134 7.4 EMBRYOS LAID BY LSD21 FEMALES HAVE A REDUCED HATCHING RATE 137 7.5 LSD21 ADULTS EXHIBIT IMPAIRED LIPID STORAGE 139
TABLE OF CONTENTS
3
8 DISCUSSION 141
9 EXPERIMENTAL PROCEDURES 144
9.1 FLY HUSBANDRY 144 9.2 FLY EMBRYO SORTING 144 9.3 DROSOPHILA STOCKS USED 145 9.4 GENERATION OF TRANSGENIC FLIES 146 9.5 DNA CONSTRUCTS 146 9.6 CONSTRUCTION OF DNA MICROARRAYS 147 9.7 RNA EXTRACTION 148 9.8 RNA LABELLING AND DNA MICROARRAY HYBRIDIZATION 148 9.9 DNA MICROARRAYS DATA ACQUISITION AND ANALYSIS 149 9.10 RT-PCR 150 9.11 RNA WHOLEMOUNT IN SITU HYBRIDIZATION 152 9.12 SLBO ANTISERUM 153 9.13 IMMUNOCYTOCHEMISTRY 153 9.14 TRIACYLGLYCEROL QUANTIFICATION 154
10 BIBLIOGRAPHY 155
LIST OF FIGURES AND TABLES
4
LIST OF FIGURES Figure 3.1 – Sequence alignment of the basic region and the leucine zipper of mouse C/EBPs, slbo and GCN4._____________________________________________________________________ 18 Figure 3.2 – Structure of the leucine zippers and basic regions of the C/EBPα homodimer bound to DNA.____________________________________________________________________________ 19 Figure 4.1 - slbo expression pattern during embryogenesis. _______________________________ 31 Figure 4.2 – Embryonic expression pattern of slbo, ect, CG9747 and ImpE2. ________________ 33 Figure 4.3 - Positive selection of slbo loss-of-function mutant embryos. _____________________ 36 Figure 4.4 – Comparison of expression pattern of slbo, CG9747, ect and ImpE2 in wild type and slbo mutant embryos.________________________________________________________________ 38 Figure 4.5 – Comparison of expression levels of slbo, CG9747, ect and ImpE2 in control and slbo mutant embryos by RT-PCR. _________________________________________________________ 39 Figure 4.6 – slbo pattern of expression during stage 9 and 10 of oogenesis._________________ 40 Figure 4.7 – Detail of slbo mRNA localization at stage 9 border cells._______________________ 41 Figure 4.8 – slbo expression in wild type and slbory7 mutant egg chambers. _________________ 42 Figure 4.9 – expression pattern of kismet in egg chambers. _______________________________ 43 Figure 4.10 – CG9747 expression pattern in control and slbo hypomorphic mutant stage 10 egg chambers. _________________________________________________________________________ 44 Figure 4.11 - Expression pattern of unpaired in egg chambers_____________________________ 45 Figure 4.12 – Gene expression profile comparison using cDNA microarrays. ________________ 47 Figure 4.13 – RT-PCR confirmation and pattern of expression of genes identified using DNA microarrays.________________________________________________________________________ 50 Figure 4.14 – Scatter plot of wildtype and enGAL4 UAS-GFP flies in the COPAS SELECT embryo sorter. _____________________________________________________________________________ 52 Figure 4.15 - COPAS SELECT system with new collection chamber _______________________ 53 Figure 4.16 – results of slbo versus control embryos microarrays.__________________________ 55 Figure 4.17 – slbo regulation in slbo mutant versus control microarrays. ____________________ 56 Figure 4.18 – Results of DNA microarray analysis of early versus late embryos. _____________ 57 Figure 4.19 – CG11395 expression dependence on slbo _________________________________ 59 Figure 4.20 – RT-PCR results of 12 genes showing downregulation in slbo mutants in DNA microarrays experiments. ____________________________________________________________ 60 Figure 4.21 – slbo mutants have a reduced hatching and survival rate._____________________ 63 Figure 4.22 – slbo mutant are very sensitive to methyl p-hydroxybenzoate __________________ 65 Figure 4.23 – Structural formula of methyl p-hydroxybenzoate. ____________________________ 66 Figure 4.24 – DmGlcAT-BSII embryonic expression is dependent on slbo. __________________ 68 Figure 4.25 – Phylogenetic tree of Drosophila UDP-glycosyltransferases and mammalian UDP-glucoronyltransferases_______________________________________________________________ 70 Figure 4.26 – slbo mutant larvae have a growth defect ___________________________________ 71 Figure 4.27 – slbo mutant larvae have an altered feeding behaviour. _______________________ 73 Figure 4.28 – sugar transporter coding genes up-regulation in the slbo mutant. ______________ 75 Figure 4.29 – sugar metabolism related genes mis-regulated in the slbo mutant. _____________ 76 Figure 4.30 – Odc1 embryonic expression is dependent on slbo. __________________________ 78 Figure 4.31 – Scheme of the polyamines synthesis pathway in animals _____________________ 78 Figure 4.32 – Phylogenetic tree of ornithine decarboxilases from different genus and Drosophila melanogaster homologues.___________________________________________________________ 80 Figure 4.33 – Sequence alignment of a highly conserved region in ornithine decarboxylases and structure of Odc catalytic centre. ______________________________________________________ 81 Figure 4.34 – Odc1 expression pattern in stage 17 embryos. ______________________________ 82 Figure 4.35 – immunostaining of stage 10 egg chambers with rat anti-Slbo antibody. _________ 83 Figure 4.36 – Immunostaining of late embryos with anti-Slbo and anti-Serpent antibodies. ____ 84 Figure 4.37 – Up-regulation of genes involved in ornithine metabolism in the slbo mutant. _____ 86 Figure 4.38 – DMFO treated larvae have a growth defect _________________________________ 87
LIST OF FIGURES AND TABLES
5
Figure 4.39 – Analysis of Odc1 down-regulation by RNA interference. ______________________ 88 Figure 4.40 - Analysis of Odc1 down-regulation effect on hatching and adult eclosion. ________ 89 Figure 4.41 – Effect of Odc1 ectopic expression in the slbo mutant. ________________________ 91 Figure 4.42 - CG8023 embryonic expression is dependent on slbo. ________________________ 93 Figure 4.43 - Phylogenetic tree of eIF4Es and Drosophila melanogaster homologues. ________ 94 Figure 4.44 – Effect of CG8023 ectopic expression in the slbo mutant. _____________________ 97 Figure 4.45 – slbo mutant larvae have high spontaneous melanization _____________________ 99 Figure 4.46 - mis-regulation of genes involved in innate immunity in the slbo mutant. ________ 101 Figure 4.47 – Slbo protein is expressed in larvae._______________________________________ 106 Figure 4.48 – Rescue of slbo lethality with the FRT-slbo-FRT trangene ____________________ 107 Figure 4.49 – requirement of slbo during larval development (I)___________________________ 108 Figure 4.50 - requirement of slbo during larval development (II). __________________________ 110 Figure 7.1 - Expression pattern of Lsd2 during embryogenesis and in 3rd instar larvae._______ 132 Figure 7.2 - Neutral lipid accumulation during oogenesis. ________________________________ 135 Figure 7.3 - TAG quantification in early embryos laid by wild type and Lsd21 homozygous females. __________________________________________________________________________ 136 Figure 7.4 Hatching rate defects in the progeny of Lsd21 homozygous females._____________ 138 Figure 7.5 - Lsd2 is required for normal larval fat body and TAG accumulation in adults. _____ 139 Figure 9.1 – Lowess normalization of DNA microarray data ________ Error! Bookmark not defined.
ABBREVIATIONS
6
Abbreviations ADRP adipocyte differentiation-related protein
Arg arginase
Atta Attacin A
BAT brown fat tissue
bZIP basic region/leucine zipper
C/EBP CCAAT/enhancer-binding protein
cDNA complementary deoxyribonucleic acid
Dig digoxygenin
DNA deoxyribonucleic acid
DFMO difluoromethylornithine
Drs Drosomycin
ER endoplasmic reticulum
eIF4E eukaryotic translation initiation factor 4E
ect ectodermal
emp epithelial membrane protein
FACS fluorescent activated cell sorter
FRT FLP recognition target
G6Pase glucose-6-phosphatase
GFP green fluorescent protein
GS glycogen synthase
hs heat-shock promoter
ImpE2 Ecdysone-inducible gene E2
Lsd Lipid storage droplet
mRNA messenger ribonucleic acid
Mtk Metchnikowin
nec necrotic
Odc ornithine decarboxylase
Oat ornithine aminotransferase
ABBREVIATIONS
7
PAT perilipin adipophilin and TIP47 domain
PCR polymerase chain reaction
PEPCK phosphoenolpyruvate
PGRP peptidoglycan recognition protein
RNA ribonucleic acid
RT reverse transcription
slbo slow border cells
TAG triacylglycerol
UAS upstream activation sequence
UCP uncoupling protein
UGT UDP-glycosylltransferase
WAT white fat tissue
1 SUMÁRIO
8
1 SUMÁRIO
A Drosophila melanogaster é um organismo modelo em investigação biológica,
com especial impacto nas áreas da Genética e da Biologia do Desenvolvimento.
Muitos genes e processos biológicos envolvidos no desenvolvimento da
Drosophila são conservados entre esta e os mamíferos. Tem sido também
demonstrado que alguma regulação metabólica é conservada entre a mosca e
os mamíferos.
Nesta tese descreverei a análise de dois genes, slbo e Lsd2, envolvidos na
regulação do metabolismo da Drosophila. Slbo é um factor de transcrição
homólogo das proteínas CCAAT/enhancer-binding proteins (C/EBPs). Estes
factores de transcrição estão, em mamíferos, envolvidos na regulação de
metabolismo e inflamação. Lsd2 é um homólogo das proteínas da família PAT,
cujos membros, em mamíferos, regulam o metabolismo de lípidos.
1.1 Regulação de metabolismo pelo factor de transcrição Slbo
O homólogo dos genes C/EBP em Drosophila foi descoberto como um gene
necessário para a migração das border cells no ovário da mosca, daí o seu
nome: slow border cells (slbo) (Montell et al. 1992). Este gene é também
expresso e necessário no final da embriogénese; embriões homozigóticos para
mutantes amorfos de slbo morrem imediatamente antes ou após a eclosão do
ovo (Rørth and Montell 1992). No entanto, estes mutantes não apresentam
quaisquer defeitos morfológicos sendo a causa da sua morte, e
consequentemente a função de slbo, desconhecida. Em mamíferos as proteínas
C/EBPs estão envolvidas na diferenciação de diversos tipos celulares (por
exemplo adipócitos e hepatócitos) e na regulação de metabolismo e inflamação
(Lekstrom-Himes and Xanthopoulos 1998; Ramji and Foka 2002). A homologia
1 SUMÁRIO
9
entre slbo e estes factores de transcrição sugere que a sua função em
Drosophila possa ser semelhante à função de C/EBPs em mamíferos. O
objectivo deste trabalho é compreender a função de slbo no desenvolvimento
embrionário e larvar da Drosophila.
O fenótipo dos mutantes slbo foi analisado com mais detalhe depois de
estabelecido um sistema para identificar mutantes homozigóticos. Embora a
maioria dos embriões morra pouco antes ou depois do nascimento, uma
pequena proporção de larvas sobrevive mais tempo. Estas normalmente morrem
antes de chegar à fase de pupa e só muito raramente chegam a adultas. As
larvas mutantes para slbo crescem menos do que as larvas tipo selvagem e têm
um comportamento alterado, em vez de se enterrarem na comida, como é
normal, vagueiam longe dela. Esta combinação de fenótipos também está
presente quando a cascata de sinalização da insulina é sobreactivada (Britton et
al. 2002), ou quando as larvas têm um excesso de aminoácidos na alimentação
ou uma deficiência na degradação da glicina (Zinke et al. 1999). Estes dados
sugerem que no mutante slbo também há uma deficiência na regulação do
metabolismo.
Foi também observado que as larvas mutantes para slbo eram muito sensíveis à
presença de p-hidroxibenzoato de metilo, um antimicrobiano comum na comida
da Drosophila. As moscas tipo selvagem desenvolvem-se sem problemas na
presença deste químico. Em mamíferos o p-hidroxibenzoato de metilo é
rapidamente degradado e excretado do organismo (Soni et al. 2002).
Aparentemente os mutantes slbo têm problemas no metabolismo ou secreção
de xenobiontes.
A expressão de slbo é máxima no final da embriogénese e diminui para níveis
muito baixos durante a fase larvar (Rørth and Montell 1992). Uma questão
interessante levantada pelas observações acima descritas é saber se as larvas
mutantes morrem por não terem slbo durante a sua fase larvar ou por não o
1 SUMÁRIO
10
terem tido no fim da embriogénese. Foi possível detectar expressão da proteína
Slbo no núcleo de células do intestino médio e posterior, dos túbulos de Malpighi
e do corpo adiposo de larvas. No entanto esta expressão é muito variável
sugerindo que a expressão de Slbo é regulada dinamicamente durante o período
larvar. A eliminação do gene slbo durante a fase larvar diminui a sobrevivência
das larvas, demonstrando que este tem uma função após o final da
embriogénese.
Slbo é um factor de transcrição e a sua função passa necessariamente pela
activação da transcrição de outros genes. Para identificar genes cuja expressão
é dependente de Slbo compararam-se embriões mutantes e tipo selvagem
utilizando DNA microarrays. Os microarrays foram produzidos por mim com o
apoio do Genomics Core Facility do EMBL e continham aproximadamente 6000
cDNAs não redudantes, cobrindo cerca de 40% do genoma da Drosophila. Estes
microarrays foram também usados por Andrea Herold e a Doutora Eliza
Izaurralde no estudo da exportação de mRNAs do núcleo, em Drosophila
(Herold, Teixeira and Izaurralde 2003).
Através dos microarrays foram identificados vários genes subexpressos nos
mutantes de slbo. Odc1, além de mostrar, por RT-PCR, uma forte dependência
da presença de slbo, parecia potencialmente importantes na explicação do
fenótipo dos mutantes slbo.
Odc1 codifica uma decarboxilase de ornitina, enzima que catalisa a conversão
de ornitina em putrescina. Esta é a primeira, e um ponto de regulação, das
reacções de síntese de poliaminas. As poliaminas são necessárias para um
vasto número de processos celulares incluindo crescimento, proliferação e
diferenciação celular (Tabor and Tabor 1984). No final da embriogénese Odc1 é
expressa no corpo adiposo, orgão onde é também detectado Slbo.
1 SUMÁRIO
11
A perda de síntese de poliaminas nos mutantes de slbo pode explicar
parcialmente o seu problema de crescimento e letalidade. De facto, larvas
normais alimentadas com DMFO, um inibidor específico da decarboxilase de
ornitina, são menores do que as larvas controlo. Também foi verificado que
Odc1 é um gene essencial pois um transgene que, por RNA de interferência,
perturba a sua expressão é parcialmente letal. A sobre-expressão de Odc1
suprime parcialmente a letalidade do mutant amorfo de slbo. Isto indica que a
subexpressão de Odc1 é uma das razões da letalidade do mutante slbo.
As larvas mutantes em slbo apresentam uma taxa muito elevada de
melanização espôntanea. Melanização é um dos mecanismos de defesa da
Drosophila contra patogénios. Esta melanização espôntanea poderia indicar que
o controlo da imunidade natural dos mutantes está desequilibrado. De facto, os
dados dos microarrays mostram que vários genes relacionados com o sistema
imune são sobre ou subexpressados no mutante slbo.
Os fenótipos observados conjuntamente com a análise dos microarrays indicam
que slbo tem um papel na regulação de metabolismo e na imunidade natural e
que a função das proteinas C/EBPs está conservada entre mamíferos e
insectos.
1.2 Regulação do metabolismo de lípidos por Lsd2
Os lípidos neutros, na forma de triacilgliceróis e ésteres de esteróis, são
armazenados intracelularmente em vesículas lipídicas. A Perilipina e a ADRP,
proteínas de mamíferos da família PAT, associam-se à superfície destas
vesículas e regulam o metabolismo dos lípidos neutros. Ratinhos nulos para a
Perilipina são viáveis e mais magros do que o normal (Martinez-Botas et al.
2000; Tansey et al. 2001). Este fenótipo torna as proteínas PAT alvos
interessantes no estudo e tratamento da obesidade. Em Drosophila existem
dois homólogos das proteínas PAT: Lsd1 e Lsd2 (Lu et al. 2001). Neste trabalho
1 SUMÁRIO
12
analisou-se um mutante amorfo de Lsd2 de modo a investigar o seu papel na
regulação do metabolismo de lípidos.
Lsd2 é predominantemente expressa em tecidos onde há actividade metabólica
de lípidos: o corpo adiposo e a linha germinal das fêmeas adultas. As larvas
mutantes têm um corpo adiposo menos desenvolvido e mutantes adultos
acumulam menos lípidos do que o normal. Estes resultados demonstram que
existe conservação de função entre Lsd2 e as proteínas PAT de mamífero.
Os ovários de fêmeas mutantes apresentam um padrão de distribuição de
lípidos anormal; em vez do padrão punctiforme normal apresentam grandes
agregados de lípidos. A mutação Lsd2 resulta também numa redução do nível
de lípidos nos ovos destas fêmeas. Concomitantemente Lsd2 é um gene de
efeito materno necessário para a embriogénese normal da progénia destas
fêmeas.
Este trabalho mostra que a função de proteínas da família PAT está conservada
entre mamíferos e Drosophila. Deste modo estabelece a Drosophila como um
novo organismo modelo para o estudo de proteínas da família PAT.
2 SUMMARY
13
2 SUMMARY
In this thesis I will describe the analysis of two genes regulating Drosophila
metabolism, slbo and Lsd2. Slbo is a transcription factor homologue to the
mammalian CCAAT/enhancer-binding proteins (C/EBPs). These mammalian
transcription factors are involved in a variety of processes, including metabolism
and inflammation regulation. Lsd2 is a member of the PAT-family of proteins,
known in mammals to regulate lipid metabolism.
2.1 Metabolism regulation by the transcription factor Slbo
Slbo is the Drosophila melanogaster homologue of mammalian C/EBPs. This
transcription factor is expressed and required in late embryogenesis; loss-of-
function homozygous mutants die just before or after hatching. However, null
mutants do not show any obvious morphological defect and the function of slbo
in late embryogenesis has been unknown. Mammalian C/EBPs regulate genes
involved in metabolism and inflammation. The purpose of this work was to
understand the function of slbo in embryonic/larvae development and analyse if
was related to C/EBP functions in mammals.
The approach taken was to compare the expression profiles of wild type and slbo
mutant embryos by DNA microarrays. The slbo mutant phenotype was also
analysed in detail.
slbo mutant larvae have a growth defect and an altering feeding behaviour. This
phenotype indicates that there is a general metabolic problem in these mutant.
The mis-regulation of some sugar metabolism related genes corroborates this
hypothesis.
2 SUMMARY
14
A specific pathway affected in the slbo mutant is the polyamine synthesis
pathway. Polyamines are polycations involved in many different biological
processes and are essential for growth in eukaryotes. Odc1 is severely
downregulated in the slbo mutant. Odc1 encodes the first and a rate-limiting
enzyme in the synthesis of polyamines. Odc1 expression in the slbo background
can partially rescue the slbo mutant lethal phenotype. This indicates that one of
the reasons why the slbo mutant is lethal is the lack of Odc1 expression.
Slbo mutant larvae also showed excessive spontaneous melanization in the
trachea. This could indicate that there is an imbalance in the regulation of innate
immunity genes. This is confirmed in the analysis of the microarray data that
shows that some immune-related genes are mis-regulated in the slbo mutant.
The results of this work indicate that slbo is involved in metabolism and innate
immunity regulation and that C/EBPs function is conserved between mammals
and insects.
2.2 Lipid metabolism regulation by Lsd2 Neutral lipids, as triacylglycerol and sterol esters, are stored in intra-cellular lipid
droplets. Mammalian PAT-family proteins are involved in lipid storage and
regulate lipolysis. In this work an Lsd2 loss-of-function mutant was analysed in
order to investigate its role in lipid metabolism.
Lsd2 is predominantly expressed in tissues engaged in high levels of lipid
metabolism, the fat body and the germline of females. Mutant larvae show a less
developed fat body and mutant adults have a reduced level of neutral lipid
content compared to wild type, showing that Lsd2 is required for normal lipid
storage. This demonstrates conservation of function between Lsd2 and
mammalian PAT family members.
2 SUMMARY
15
In addition, ovaries from Lsd2 mutant females exhibit an abnormal pattern of
accumulation of neutral lipids from mid-oogenesis on. Mutant ovaries show large
patches of neutral lipid accumulation in contrast to the normal punctuated
distribution. The lack of function of Lsd2 in the ovaries also results in a reduced
deposition of lipids in the egg. Consistent with this Lsd2 is a maternal effect gene
that is required for normal embryogenesis.
This work shows conservation of function between mammalian and Drosophila
PAT-family members and establishes Drosophila melanogaster as a new in vivo
model for studying PAT-family proteins.
slbo
16
Metabolism regulation by the transcription factor slbo
3 INTRODUCTION slbo
17
3 INTRODUCTION
For the maintenance and growth of an organism there is a constant requirement
of energy and synthesis of a great variety of molecules. Metabolism can be
defined as the integrated network of reactions responsible for these processes.
Most central metabolic pathways and their enzymes are known and conserved
between organisms. What is less known is how they are regulated and
integrated. This can be done in many ways; levels of enzymes, activity of
enzymes, availability of substrates, etc.
A further level of complexity is present in many multi-cellular organisms. Besides
a basal cellular metabolism there is regulation at the level of the whole
organisms, with specialized tissues controlling specific pathways and storing
nutrients. In mammals the fat tissue and the liver play these pivotal roles. One of
the major functions of these tissues is to insure a constant level of
nutrients/energy for the organism throughout cycles of feeding and fasting and
eventual starvation.
In humans the adipose tissue holds most of the energy reserves of the
organisms in the form of triglycerides. Upon requirement of energy, by other
tissues, it releases free fatty acids. The major place of synthesis of lipids, which
are later transported to the fat tissue, is the liver. The liver also plays a crucial
role in maintaining glucose levels in the blood. It can produce glucose from
glycogen breakdown or through gluconeogenesis or it can store glucose in the
form of glycogen. Finally, the liver is also involved in aminoacid catabolism. The
ammonia produced by this process is toxic and excreted in the form of urea,
synthesized through the urea cycle in the liver.
Members of the CCAAT /Enhancer Binding Protein (C/EBP) family of
transcription factors play an essential role in the differentiation and function of the
3 INTRODUCTION slbo
18
liver and the adipose tissue. C/EBPs directly regulated metabolic genes, hence,
they have a critical role in metabolism regulation, at the transcription level.
During my PhD I have investigated the function of the C/EBP Drosophila
homologue, slbo. In this introduction I will first give a description of mammalian
C/EBPs and their function. Afterwards, I will review what was known about
Drosophila C/EBP.
3.1 Structure and DNA binding of C/EBPs
The first C/EBP to be identified, C/EBPα, was the founding member of the bZIP
class of transcription factors (Landschulz et al. 1988). bZIP stands for basic
region/leucine zipper, two adjacent conserved protein domains. Other members
of this class are Jun, Fos, Myc and GCN4. There are five mammalian C/EBPs,
named C/EBPα to C/EBPε. The sequence of these proteins is highly conserved
in the C-terminal region, the bZIP domain (Figure 3.1). The N-terminal region is is
quite divergent (<20% identity) between C/EBP proteins. The first crystal
structure of a bZIP protein bound to DNA was of GCN4 (Ellenberger et al. 1992).
The structure of C/EBPα has also been described (Miller et al. 2003) (Figure 3.2).
Figure 3.1 – Sequence alignment of the basic region and the leucine zipper of mouse C/EBPs, slbo and GCN4.
The alignment was done using ClustalX.
The bZIP transcription factors function has dimers. The leucine zipper is the
domain of dimerization. This is an α-helix with leucine residues repeated every
3 INTRODUCTION slbo
19
seven residues. The spacing of the leucine residues results in a leucine residue
being present in the same face of the α-helix every other turn (Figure 3.2). Two
leucine zipper domains interact, in parallel, with the aid of hydrophobic
interactions between leucine residues of the two different chains. The specificity
of the dimer formation is determined by the amino acid residues along the
dimerization domain (Vinson et al. 1993). C/EBPs form homodimers but are also
capable of forming heterodimers between themselves in all possible
combinations (Ramji and Foka 2002).
Figure 3.2 – Structure of the leucine zippers and basic regions of the C/EBPα homodimer bound to DNA.
LZ - leucine zipper; BR - Basic region. The leucine residues involved in the dimerization are
shown. Adapted from (Miller et al2003).
Although the leucine zipper does not interact with the DNA, its functionality is
essential for DNA binding (Landschulz et al. 1989). The proteins only binds DNA
3 INTRODUCTION slbo
20
has a dimer and the leucine zippers interaction positions the DNA binding
domains, the basic regions of the two chains, in the correct orientation.
The basic region also forms an α-helix and its amino acid sequence determines
the DNA binding sequence specificity (Johnson 1993). Since the basic region is
very conserved between all C/EBP family proteins, the several homodimers and
heterodimers can interact with the same DNA sequences.
As the crystal structure shows bZIP proteins bind DNA as an inverted Y. The two
basic regions of a dimer are oriented in opposite ways and each binds to half the
DNA recognition sequence. Because of this bZIP homodimers tend to bind
palindromic sequences, each basic region binding half of it. C/EBP dimers also
recognize sequences with a palindromic nature. An in vitro study described the
consensus sequence as RTTGCGYAAY (R=A or G and Y = C or T) (Osada et al.
1996). However, C/EBPs show a “relaxed” binding specificity, allowing much
variation at the binding sequence. C/EBPs bind to sites in different promoters
that share only minimal sequence similarity (Johnson and McKnight 1989). A
good example is that C/EBPα was found twice, independently, in the same
laboratory as a protein that bound to two different DNA elements; CCAAT
pentanucleotides of herpesvirus and murine sarcoma virus promoters and the
enhancer core element (TGTGGWWWG where W = A or T), common to many
animal virus (Graves et al. 1986; Johnson et al. 1987). However, C/EBP binding
to a particular site is specific; single point mutations in a high affinity site may
inhibit binding. This promiscuity in DNA binding specificity is a common feature in
eukaryotic transcriptions factors. Interaction with other transcription factos and
combinatorial specificity contribute to specifically activate target genes.
The “relaxed” binding specificity makes it difficult to predict C/EBP binding sites.
At the same time the consensus sequences generated from binding sites
analysis are so relaxed that a high frequency of potential binding sites will be
3 INTRODUCTION slbo
21
found in a determined DNA sequence. This removes predictive power to this kind
of analysis.
The N-terminal region of C/EBPs is less conserved between them and contaisn
transcription activating domains (Ramji and Foka 2002). There are small
conserved sub-regions that correspond to some transcription activating domains.
C/EBPγ, however, does not contain an activation domain and forms
transcriptional inactive heterodimers with other C/EBPs, being a dominant
negative inhibitor (Cooper et al. 1995).
The complexity of mammalian C/EBPs does not finish in the different dimers
combinations. C/EBPα, C/EBPβ and C/EBPε have alternative isoforms, due to
alternative promoters, alternative splicing, alternative translation initiation site or
regulated proteolysis (in Ramji and Foka 2002). The different isoforms have
different properties. C/EBPβ, for example, can produce a small isoform (LIP) that
lacks the transcription activation domains and acts as a dominant negative
inhibitor of C/EBPs (Descombes and Schibler 1991). C/EBPs can also interact
with other bZIP and non-bZIP proteins (LeClair et al. 1992; Vallejo et al. 1993;
Hsu et al. 1994).
3.2 Mammalian C/EBPs function
C/EBPs are expressed in many tissues (Lekstrom-Himes and Xanthopoulos
1998; Ramji and Foka 2002). C/EBPα is expressed in high levels in the adipose
tissue, liver, intestine, lung, adrenal gland, peripheral blood mononuclear cells,
placenta and ovary. C/EBPβ is enriched in the adipose tissues, liver, intestine,
lung, spleen, kidney and myelomonocytic cells. C/EBPδ is expressed in the
adipose tissue, liver, intestine and lung. C/EBPγ is ubiquitously expressed.
C/EBPε is expressed in myeloid and lymphoid cells.
3 INTRODUCTION slbo
22
3.2.1 C/EBPs and the adipose tissue
There are two types of adipose tissue in mammals; white adipose tissue (WAT)
and brown adipose tissue (BAT). WAT is the place of accumulation of energy in
the form of lipids. BAT is involved in thermogenesis through uncoupled oxidative
phosphorylation. C/EBPs play a role in the differentiation of both kinds of
adipocytes. Upon hormonal induced differentiation of cultured fibroblasts into
adipocytes C/EBPβ and C/EBPδ are first up-regulated (Yeh et al. 1995). At this
stage the preadipocytes undergo two rounds of cell division. C/EBPβ and
C/EBPδ are later down-regulated and C/EBPα up-regulated. C/EBPs have been
shown to bind promoters and activate several adipose-specific genes (Darlington
et al. 1998).
C/EBPβ and C/EBPδ ectopic expression in cell culture preadipocytes stimulates
adipogenesis (Wu et al. 1995; Yeh et al. 1995). C/EBPβ has a stronger effect
and induces the expression of PPARγ, another transcription factor up-regulated
in late adipocyte differentiation and essential for adipogenesis (Wu et al. 1995).
Mice lacking both C/EBPβ and C/EBPδ have a much more severe phenotype
that mice lacking only C/EBPβ or C/EBPδ (Tanaka et al. 1997). This shows that
their function is partially redundant. In double loss-of-function mutants most mice
die within 24h after birth. In these mice there is a block in BAT differentiation.
BAT cells can not accumulate fat droplets and express lower levels of BAT-
specific uncoupling protein (UCP), a marker for BAT differentiation. WAT seems
to have normal differentiation but it is reduced in the double mutant. This
indicates that C/EBPβ and C/EBPδ are required for commitment in the adipocyte
lineage and/or proliferation of preadipocytes. C/EBPβ has been shown, in cell
culture, to be required for mitotic clonal expansion during adipogenesis (Tang et
al. 2003).
Conditional expression of C/EBPα in cell culture preadipocytes triggers adipocyte
differentiation with lipid accumulation and expression of marker genes (e.g.
3 INTRODUCTION slbo
23
422/aP2, GLUT4) (Lin and Lane 1994). The phenotype of the C/EBPα loss-of-
function mutant mouse is very complex (Wang et al. 1995). Mutant mice died
within 8 hours after birth due to hypoglycaemia, a liver related problem (see
below). In these mutant mice, and in mutant mice partially rescued by C/EBPα
expression in the liver, there is a strong reduction of WAT and these cells do not
accumulate lipids (Wang et al. 1995; Linhart et al. 2001). Thus, C/EBPα is
required for proper differentiation of WAT. BAT lipid accumulation and UCP
expression is very reduced in new-born mice but it reaches normal levels by 7
days of age (Linhart et al. 2001). This indicates that C/EBPα function is required
for normal BAT development but can be compensated.
The expression of some adipose-specific genes that in several assays are
downstream of C/EBPs, like 422/aP2, is not severely affected in the C/EBPβ and
C/EBPδ double mutant or the C/EBPα mutant (Wang et al. 1995; Tanaka et al.
1997). This indicates that there could be some redundancy between all these
three C/EBPs. It can also indicate that although normally C/EBPs up-regulate
these genes, in vivo compensatory mechanism mask the loss-of-function effect.
In summary, C/EBPs are required for the differentiation of adipocytes and the
expression of several adipocyte-specific genes.
3.2.2 C/EBPs and liver function
Initial observations, including the fact that C/EBPα was highly expressed in the
liver, lead to the prediction that C/EBPα would have an important role in
metabolism regulation (McKnight et al. 1989). C/EBPs are implicated in
controlling the expression of many metabolism related genes in the liver (in
Darlington et al. 1995).
The analysis of the C/EBPα mutant mice clearly shows its importance in energy
metabolism (Wang et al. 1995). The mice die of hypoglycemia within the first 8
hours. Three crucial genes in glucose metabolism are down-regulated in the
3 INTRODUCTION slbo
24
newborn mutant mouse: glycogen synthase (GS), phosphoenolpyruvate
carboxykase (PEPCK) and glucose-6-phosphatase (G6Pase). GS catalyses the
formation of glycogen from glucose monomers. Its activity increases in late
gestation, coincidentally with C/EBPα expression. The accumulation of glycogen
at this point is important for later providing glucose to the neonate in the first
hours after birth. PEPCK is essential for gluconeogenesis and starts to be
expressed at birth. G6Pase also starts to be expressed at this stage and it is
essential for the release of glucose from the liver to the bloodstream. PEPCK and
G6Pase expression levels, however, come up to almost normal levels 7 hours
after birth. The down-regulation of these genes just before and after birth can
explain the mutant phenotype. The lack of GS before birth prevents accumulation
of mother provided glucose in the form of glycogen. The liver of newborn mutant
mice to not store glycogen. After birth there are no reserves of glucose to release
to the blood stream, furthermore, the newborn mice lack enzymes important for
gluconeogenesis. These problems lead to the hypoglycemic phenotype.
C/EBPβ has also been implicated in liver energy metabolism regulation. C/EBPβ
loss-of-function mutants can have a severe phenotype similar to C/EBPα loss-of-
function mutant (Croniger et al. 1997; Croniger et al. 2001). Although they have
glycogen stores they cannot mobilize it. They also have reduced expression of
PEPCK at birth. These mice die shortly after birth of hypoglycemia. Some mutant
mice do not show this phenotype at birth but later also show strong defects in
glucose homeostasis.
C/EBPs are also important in the urea cycle regulation. There are C/EBP binding
sites in several urea cycle genes (in Kimura et al. 1998). In the C/EBPα mutant
urea cycle genes are down-regulated and the mice have high levels of blood
ammonia (Kimura et al. 1998). C/EBPβ role in urea cycle genes regulation is less
clear. In primary-cultured hepatocytes derived from C/EBPβ loss-of-function
mice, induction of urea cycle genes by glucocorticoids and glucagons is impaired
3 INTRODUCTION slbo
25
(Kimura et al. 2001). However in C/EBPβ loss-of-function mice this response
seems normal, suggesting that there is a compensation mechanism.
C/EBPs also have an important role in the regulation of liver acute phase genes.
These genes encode for proteins up or down-regulated during the acute phase of
inflammation. Their expression in the liver is modulated by cytokines and
hormones produced by tissue injury at a distal site. They are divided in two
classes; class I genes require the cytokines IL-1 and IL-6 to be induced while
class II require IL-6 but not IL-1. Most of the promoters of class I acute phase
genes have functional C/EBP binding sites in their promoters (in Poli 1998). This
DNA element is called type I IL6-response element (IL-6 RE).
C/EBPβ was found for its inducibility by IL-6 or IL1 (Akira et al. 1990; Poli et al.
1990). C/EBPδ is also strongly up-regulated by inflammatory stimuli (Alam et al.
1992; Poli 1998). The activation of C/EBPβ activity by inflammation is done by
increase transcription but also by post-translational modifications induced by IL-6
(Poli et al. 1990). Activation of C/EBPδ is mainly done at the level of transcription
(Ramji et al. 1993). In C/EBPβ mutant loss-of-function mice the induction of some
class I acute phase genes is reduced, however, the induction of others is not
(Cappelletti et al. 1996). Again this could indicate partial redundancy between
C/EBPs. Some acute phase genes initial induction is not affected in C/EBPβ
mutants but their expression, at later stages of the acute response, instead of
increasing, decreases to background levels (Cappelletti et al. 1996). This
suggests that C/EBPβ is required not for the initial induction of their expression
but for its maintenance. C/EBPβ and C/EBPδ up-regulation occurs relatively late
after the inflammatory stimulus (Poli 1998). A sequential model of induction for
acute phases genes (Poli 1998) proposes that the initial induction of these genes
is done by NF-κB and STAT, two transcription factors involved in the acute phase
response. These would also induce C/EBPβ and C/EBPδ that are then recruited
to the promoters of acute phase genes and maintain their expression.
3 INTRODUCTION slbo
26
C/EBPα expression in the liver is down-regulated by inflammatory stimuli (Alam
et al. 1992). However, neonate C/EBPα mutant mice have a defective
inflammatory response (Burgess-Beusse and Darlington 1998). This indicates
that in some conditions C/EBPα is also essential in the inflammatory response
and cannot be compensated by other C/EBPs.
In summary, C/EBPS are required for the expression of several hepatocyte-
specific genes. They are involved in several processes including glucose
metabolism, the urea cycle and the acute phase response.
3.2.3 C/EBP function in other tissues
C/EBPs function is not restricted to the adipose tissue and the liver. They are
also expressed (see above) and required in many other tissues.
C/EBPβ and C/EBPδ are up-regulated in response to inflammatory stimuli in
many other tissues than the liver and probably play a more general role in
inflammation (Akira et al. 1990; Kinoshita et al. 1992). C/EBPα, C/EBPβ,
C/EBPγ, C/EBPδ and C/EBPε are required for differentiation of myelomonocytic
cells and/or their function (Poli 1998).
C/EBPS are also involved in the normal differentiation of keratinocytes, lung and
ovarian (Lekstrom-Himes and Xanthopoulos 1998; Ramji and Foka 2002).
C/EBPs have also been implicated in long-term memory (Alberini et al. 1994;
Ramji and Foka 2002).
C/EBPα is expressed at high levels in terminally differentiated cells (including
adipocytes and hepatocytes). C/EBPα is a strong inhibitor of cell proliferation and
down-regulated during cell proliferation, as for example in liver regeneration
(while C/EBPβ seems to have the opposite role) (Diehl 1998). C/EBPα has also
been shown to be a tumour suppressor in acute myeloid leukaemias (Pabst et al.
3 INTRODUCTION slbo
27
2001). C/EBPα-mediated growth arrest is independent on DNA binding activity
and is done by direct inhibition of Cdk2, Cdk4 and E2F (Muller et al. 1999; Porse
et al. 2001; Wang et al. 2001).
The functions of mammalian C/EBPs are numerous. Two fundamental ones
concern metabolism regulation and the innate immune response.
3.3 Drosophila C/EBP
Drosophila melanogaster has one protein homologue of mammalian C/EBPs. It
was first identified has a gene required for the migration of a specific group of
cells in adult egg chambers (Montell et al. 1992). This group of eight to ten follicle
cells is named border cells, hence the name of the gene: slow border cells (slbo).
3.3.1 slbo structure and DNA binding Slbo homology to mammalian C/EBPs is restricted to the bZIP region (Montell et
al. 1992). The basic region is particularly well conserved, in this domain Slbo is
74% identical, 81% similar to C/EBPα (Figure 3.1). This level of conservation is
the one observed between mammalian C/EBPs in this DNA binding domain.
Accordingly, slbo and mammalian C/EBPα bind to similar DNA sequences (Rørth
and Montell 1992). The leucine zipper is less conserved between Slbo and
mammalian C/EBPs however it can form homodimers with itself and
heterodimers with C/EBPα.
The conservation of function between Slbo and mammalian C/EBPs is better
shown in an in vivo assay. slbo loss-of-function mutants are lethal and do not
reach adulthood (Rørth and Montell 1992). This lethality can be partially rescued
by a slbo genomic construct (Rørth 1994). C/EBPα, C/EBPβ or C/EBPδ,
replacing slbo in this construct, also partially rescue the mutant, although less
3 INTRODUCTION slbo
28
efficiently (Rørth 1994). This shows that mammalian C/EBPs can, to some
extent, replace Slbo function.
In this series of assays it was also shown that Slbo function is defined by the
DNA binding basic region (Rørth 1994). Single amino acids substitutions in the
basic region, that affect in vitro DNA binding, affect in vivo rescue assays.
Moreover, Slbo basic region in a chimerical protein with heterologous
transcription activation domain and leucine zipper also partially rescues the
mutant lethality. This indicates that Slbo functionality is defined by recognizing
the correct DNA binding sequences, as a homodimer, and activating
transcription.
3.3.2 slbo function
Border cells are a group of follicle cells that at stage 9 of oogenesis initiate a well
defined migrationin the egg chamber. They delaminate, as a cluster, from the
anterior follicular epithelium and migrate through the germline derived nurse cells
until they reach the oocyte. slbo is expressed in border cells just prior and during
their migration and is required for this migration (Montell et al. 1992). However,
slbo expression is not sufficient to differentiate other follicle cells into migratory
cells (Rørth et al. 2000). Although slbo is required for normal migration it is not
absolutely required for some border cell differentiation. In mutant clones these
cells still retain some border cells characteristics; they still adhere to each other
and form a distinct cluster.
In general slbo targets in border cells are not well understood. DE-Cadherin,
Myosin VI and focal adhesion kinase are three genes involved in cell adhesion
that have been suggested to be downstream of slbo and could explain its
function in cell migration (Niewiadomska et al. 1999; Bai et al. 2000; Geisbrecht
and Montell 2002). Jing is a transcription factor required for border cell migration
that has also been suggested to be downstream of slbo (Liu and Montell 2001).
3 INTRODUCTION slbo
29
slbo is also expressed and required in late embryogenesis (Rørth and Montell
1992). slbo starts to be expressed 9-10 hours after fertilization (Drosophila
embryoninc development lasts 22h at 25˚C). Its expression peaks at 12-18h of
development when Slbo is expressed in many tissues. At this stages most of
organogenesis is finished and cell are in the last stages of differentiation. slbo
loss-of-function homozygous mutants die just before or after hatching. These
mutants do not show any obvious morphological defect and the function of slbo
in late embryogenesis was unknown. Mammalian C/EBPs also start to be
expressed at late embryogenesis, in differentiating cells and mammalian C/EBP
loss-of-function mutants also do not have strong morphological defects.
3.4 Aim of the project
The objective of this project was to understand the function of slbo in Drosophila
late embryogenesis.
The structural functional conservation, the similar development time of
expression and the end of embryonic development lethality could indicate that
the biological function of C/EBPs could be conserved between mammals and
Drosophila. However, there is a lack of homology outside the bZIP domain
between slbo and mammalian C/EBPs. There is also a big difference between
mammals and insects in physiology and the tissues involved in metabolism and
innate immunity. It would be interesting to know if C/EBPs function in metabolism
regulation and innate immunity is conserved between mammals and Drosophila.
4 RESULTS slbo
30
4 RESULTS
4.1 Finding slbo target genes - Candidate genes approach
4.1.1 Expression pattern of slbo in embryos
As a first approach to study the function of slbo in Drosophila development I have
analysed its expression pattern during embryogenesis. slbo mRNA expression
was detected by in situ hybridization using a digoxygenin (dig) labelled anti-
sense RNA probe. The in situ hybridization shows, essentially, the same pattern
of expression as the antibody staining previously described (Rørth and Montell
1992). slbo is only expressed from stage 11 on (Figure 4.1). It starts to be
expressed in the posterior spiracles, the posterior openings the respiratory
organs of Drosophila, the tracheaa. At stage 14 it starts to be expressed in the
salivary glands and the proventriculus, both part of the digestive system. At stage
15 it is also expressed in the oesophagus, midgut and hindgut, also part of the
digestive system. All the digestive system seems to express slbo at stage 16 and
17. At these stages there is also a strong epidermis staining. In fact many tissues
express slbo at this stage. Because of this widespread expression is difficult to
discern where slbo is or is not expressed. By direct observation it is possible to
conclude that slbo is also expressed in the Malpighian tubules, the excretory
organs of Drosophila, and the trachea. Previous data (Rørth and Montell 1992)
and these in situ hybridization also indicate that slbo is not expressed in the
central nervous system, a tissue that expresses many genes at late
embryogenesis.
The expression or not of slbo in a different particular tissue may be analysed by
using fluorescent antibody staining and a confocal microscope. The confocal
microscope allows the analysis of only one section of the embryo and via the
4 RESULTS slbo
31
structure or co-staining it is possible to identify the tissue. This technique was
used to identify slbo expression in the fat body in section 4.3.5.3.
Figure 4.1 - slbo expression pattern during embryogenesis.
slbo wholemount in situ hybridization of embryos. slbo starts to be expressed at stage 11
(≈5h30m after egg laying) and it is strongly expressed from stage 15/16 on (≈12h a.e.l.). It is first
expressed in the posterior spiracles (ps). At stage 14 it starts to be also expressed in the salivary
glands (sg) and proventriculus (pv). At later stages it is also expressed in the oesophagus (es),
hindgut (hg), midgut (mg) and epidermis (ep). From stage 16 on slbo is highly expressed in a
great number of tissues being difficult to access where is and where is not expressed. All images
with anterior to the left and dorsal up.
4 RESULTS slbo
32
Northern blot analysis shows that slbo expression is greatly reduced at the end of
embryogenesis, from 18h to 22h (Rørth and Montell 1992). It is not possible, by
in situ hybridisation, to verify that or to see if slbo is still expressed in any
particular tissue at this stage because it coincides with cuticle formation, which
prevents the probe penetration into the embryo.
4.1.2 Expression pattern of three candidate slbo target genes
Slbo is a transcription factor and a direct approach to study its function is to
identify its target genes. Previously to this work Pernille Rørth and Irina Orlov
identified three genes dependent on slbo to be expressed in the embryo. A
subtractive hybridization to identify genes expressed only late in embryogenesis
was done. They then compared, by Northern blot, the expression of several
candidate genes in wild type embryos and embryos with a reduced slbo function.
Three genes were identified: CG9747, ectodermal (ect) and Ecdysone-inducible
gene E2 (ImpE2). CG9747 is predicted to be an acyl-CoA δ11 desaturase (The
flybase Consortium2003). ect was found for being AT rich in its 3’ untranslated
region of the transcript in a screen for homologues the mouse α-interferon gene
MuIFNα2 (Nakanishi et al. 1986). However, its predicted protein sequence does
not show similarity to this or any other known gene and its function is unknown.
ImpE2 was found as a gene up-regulated in imaginal disks in response to
ecdysone (Natzle et al. 1986). It appears to be a secreted protein but its function
is unknown (Paine-Saunders et al. 1990).
I analysed the expression pattern of these genes to see if they had any similarity
with slbo expression pattern (Figure 4.2). Similarly to slbo, CG9747 and ect were
only expressed at late embryogenesis. They also share some of the pattern of
expression; they are enriched in the posterior spiracles, the oesophagus and the
epidermis. ect is also expressed in the trachea. This ect expression pattern
confirms a previous description based on in situ hybridization and antibody
4 RESULTS slbo
33
Figure 4.2 – Embryonic expression pattern of slbo, ect, CG9747 and ImpE2.
slbo, ect, CG9747 and ImpE2 wholemount in situ hybridization of embryos. All genes share a late
embryonic expression pattern. All the genes are expressed in the epidermis. emp and CG9747
also have in common with slbo the expression in the posterior spiracles and oesophagus. ImpE2
is also expressed in earlier embryonic stages. This earlier expression lasts until stage 9, it then
stops to be expressed and starts to be expressed again at stage 13. In all images posterior is to
the left and dorsal is up except stage 12 slbo in situ that is a dorsal view. Stages 15 and 16 have
images focused inside and on the surface of the embryo.
4 RESULTS slbo
34
staining in embryos’ sections (Nakanishi et al. 1986; Raha et al. 1990). ImpE2
differs from the other two genes by the fact that it has an early embryonic
expression. However the early expression disappears after stage 9 and
reappears, at stage 13/14. This late expression seems independent from the
early expression because it is mainly at the level of the epidermis and trachea
while the early expression is on the precursor tissues of the gut. ImpE2 late
embryonic expression pattern has been previously described (Paine-Saunders et
al. 1990). ImpE2 and slbo have in common this late embryonic epidermis
expression.
All the three genes expression patterns are compatible with them being
downstream of slbo at the end of embryogenesis. ect and CG9747 expression
patterns are very similar between themselves and share many tissues of
expression with slbo.
4.1.3 Sorting of homozygous slbo loss-of-function mutant embryos
In order to analyse if a gene is down-regulated in the absence of slbo it was
essential to isolate slbo loss-of-function embryos. I have used two slbo loss-of-
function alleles (slboe7b and slbo8ex2) (Rørth 1994). They are both small
chromosomal deletions, resulting from imprecise P-element excisions that affect
only the slbo gene. slboe7b has a deletion of approximately 5kB that removes all
the slbo transcript. slbo8ex2 has a deletion of approximately 1.8kB that removes
most of the slbo open reading frame. slbo loss-of-function homozygous mutant
embryos are non-viable and dye before reaching adulthood (Rørth and Montell
1992). These alleles can only be kept in a heterozygous state.
The slbo gene is on the second chromosome of the fly and the loss-of-function
mutants are kept over the balancer chromosome CyO. In order to get slbo
homozygous embryos it is required to cross together the slbo / CyO flies. From
this cross one quarter of the embryos are homozygous mutant. However, since
4 RESULTS slbo
35
these homozygous embryos are morphologically identical to wild-type embryos it
is impossible to distinguish them from the others three quarters of embryos. In
order to compare slbo mutant embryos with wild-type embryos it was crucial to
devise a way to identify the slbo mutant embryos.
An available method was the use of a balancer chromosome that expressed
green fluorescent protein (GFP) during embryogenesis (Casso et al. 2000).
When crossing together flies that are slbo / CyO GFP the slbo homozygous
mutant embryos were the only ones not expressing GFP. However, the balancer
chromosomes available at the time (CyO kr-GAL4 UAS-GFP) (Casso et al. 2000)
expressed GFP very weakly which made sorting the embryos under a fluorescent
dissection microscope virtually impossible. This difficulty and the fact that it was
based on a negative selection (no GFP) made it not practical or reliable.
To circumvent this problem I chose to have one fly stock with a recombinant
chromosome having the slboe7b allele and the en-GAL4 driver and another stock
with a chromosome with the slbo8ex2 allele and the reporter gene UAS-GFP
(Figure 4.3). When crossing the two recombinant stocks the en-GAL4 driver and
the UAS-GFP will be together in the same embryo only when this embryo has a
heteroallelic combination of slbo loss-of-function alleles. Therefore, the slbo loss-
of-function mutants can be identified by the presence of GFP stripes.
This method has several advantages over the CyO kr-GAL4 UAS-GFP
balancers. 1) The pattern and intensity of the GFP expression make the embryo
selection easy and accurate. 2) This expression starts at stage 10/11, just before
the start of slbo expression, and continues throughout all the rest of the fly life. 3)
The method is based on positive selection so that every embryo picked up for
having striped GFP is a slbo mutant. 4) The slbo mutant analysed is a
heteroallelic combination of two slbo loss-of-function alleles. This later
characteristic is important considering that each slbo allele is kept heterozygous
over a balancer chromosome. Since the chromosomes carrying these mutations
4 RESULTS slbo
36
do not recombine for several generations, they may accumulate background
mutations. When analysing a homozygous embryo for one slbo allele it is in fact
possible that we do not only observe the phenotype of slbo loss-of-function but
also of the other mutations in a homozygous state. When analysing the
heteroallelic combination of slbo loss-of-function alleles the probability of having
the same random background mutation in the two independent chromosomes is
virtually zero. This minimizes the problem of background mutations.
Figure 4.3 - Positive selection of slbo loss-of-function mutant embryos.
slbo loss-of-function mutant embryos are obtained from the cross of two slbo heterozygous
stocks. One stock carries the loss-of-function allele slboe7b recombined, in the same
chromosome, with en-GAL4. The other stock has the loss-of-function allele slbo8ex2 recombined
with UAS-GFP. Only the embryos with the heteroallelic combination of slbo loss-of-function
alleles will have the en-GAL4 driver and the UAS-GFP and therefore only these will express GFP.
The expression of GFP starts when the germ band is extended (embryo on the left, at stage 10)
and is continuously on throughout the rest of embryogenesis (embryo on the right is at stage 17,
the last stage of embryogenesis), larval stages, pupal stages and adulthood.
4 RESULTS slbo
37
The control embryos used in experiments concerning the slbo loss-of-function
are the progeny of an en-GAL4 stock crossed with an UAS-GFP stock and are
handled and selected exactly the same way as the slbo mutants.
4.1.4 Expression of three candidate target genes in control and slbo mutant embryos.
The expression pattern of slbo, ect, CG9747 and ImpE2 was compared, by
wholemount in situ hybridisation, between wild type and slbo mutant embryos
(Figure 4.4). As expected slbo expression is completely absent in slbo loss-of-
function mutants. However, none of the other genes’ expression patterns are
visibly affected in the slbo mutant.
The in situ hybridization protocol used is not quantitative; the last step of the
protocol is based on an enzymatic reaction to produce the coloured dye. That
means that only big differences on expression levels can be detected with it. In
order to better compare the levels of expression a semi-quantitative method was
used. Total RNA was extracted from control and slbo mutant embryos, a first-
strand reverse transcription reaction was performed and fragments
corresponding to the transcripts of interest were amplified by PCR. The PCR
reaction was done with different number of cycles in order to be able to compare
the control and slbo mutant reactions at the exponential phase of the
amplification. A reproducible difference in the intensity of the PCR band between
two samples can be interpreted as a real difference of quantity of transcript in the
original samples. This technique will be referred throughout this thesis as RT-
PCR (Reverse transcription – polymerase chain reaction). This semi-quantitative
technique was chosen over a Northern blot because it is more sensitive and
requires less starting material. It is also faster and easier to test several different
genes.
4 RESULTS slbo
38
Figure 4.4 – Comparison of expression pattern of slbo, CG9747, ect and ImpE2 in wild type and slbo mutant embryos.
slbo, ect, CG9747 and ImpE2 wholemount in situ hybridization of control and slbo mutant
embryos. Same stage embryos are compared for each gene. Each pair was done simultaneously.
Except for slbo itself no difference is detected in the expression patterns of these genes in control
and slbo mutant embryos.
Stage 16 control and slbo mutant embryos were compared by RT-PCR (Figure
4.5). As expected the slbo transcript was absent from the slbo loss-of-function
mutants. CG9747, ect and ImpE2 levels were reproducibly slightly reduced in the
slbo mutant. Although the expression pattern is not affected, the levels of
4 RESULTS slbo
39
expression of these genes are dependent on slbo presence. These genes could
be direct or indirect targets of Slbo.
Figure 4.5 – Comparison of expression levels of slbo, CG9747, ect and ImpE2 in control and slbo mutant embryos by RT-PCR.
Total RNA was collected from stage 16 embryos and a first strand cDNA was synthesized for
each genotype. A PCR reaction was set up for each tested gene in each genotype with the gene
specific primers and primers for rp49. The gene specific primers were designed to give an
amplified product of approximately 600bp. The rp49 primers gave an amplified product of
approximately 300bp. rp49 encodes for a ribosomal gene and its levels should not change in the
slbo mutant. In these experiments it serves as a loading control. slbo transcript is absent in the
slbo mutant. CG9747, ect and ImpE2 levels are slightly reduced in the slbo mutant.
4.1.5 Expression pattern of slbo in egg chambers
Another place of slbo expression is the egg chamber where it is required for
border cell migration (Montell et al. 1992) (Figure 4.6). slbo starts to be
expressed at stage 9 in the border cells, just before their migration. slbo
expression in these cells increases during and after migration. At stage 10 it is
also expressed in the centripetal cells, a group of follicular cells that migrates
inwards the egg chamber at stage 10. However, slbo is not essential for this
migration. slbo is also expressed at the posterior polar cells. Polar cells are two
groups of two cells that exist at the anterior and the posterior of the egg chamber.
They differentiate early in the egg chamber development and function as
organizer centres for the patterning of the egg chamber. The anterior polar cells
constitute the centre of the migrating border cells cluster and also expresse slbo.
4 RESULTS slbo
40
Figure 4.6 – slbo pattern of expression during stage 9 and 10 of oogenesis.
In situ hybridization of stage 9 and 10 egg chambers with slbo anti-sense probe. During stage 9
and 10 slbo is expressed in the border cells (bc). At stage 10 slbo is also expressed in the
centripetal cells (cp). The posterior polar cells (ppc) also express slbo. Images are not to scale.
All images with anterior to the left.
A curious observation from these in situs is that slbo transcript localization is
polarized when it starts to be expressed, in the beginning of stage 9 and before
the border cell cluster rounds up (Figure 4.7). Most of the transcript is localized in
the apical (inside in relation to the egg chamber) side of cells. It can even be
detected at some cytoplasmic extensions that grow in-between the nurse cells. It
is also possible to observe this slbo mRNA asymmetric localization in the
posterior polar cells, where it is also enriched in the apical side (Figure 4.6). The
4 RESULTS slbo
41
asymmetric slbo mRNA localization in border cells is loss once the border cells
cluster rounds up and starts to migrate. It is tempting to speculate that the
asymmetric mRNA localization plays a functional role and may be associated
with signalling events originated from inside the egg chamber. I do not know if
this a particular property of the slbo mRNA or if its shared with other mRNAs.
The function of slbo in border cell migration was found in a hypomorph slbo
mutant with viable adults (Montell et al. 1992). This slbo allele is caused by a P-
element insertion that specifically interferes with slbo expression in border cells.
It is possible to see, by in situ hybridization, the difference in slbo mRNA levels
between the wild type and mutant egg chambers (Figure 4.8). When processed
in the same conditions as the wild type slbory7 mutant borders cells barely show
any staining. However, when the development step is longer it is possible to see
that slbo is still expressed in these mutant border cells albeit at a much lower
level. The same was observed for the alleles slbory8and slbo1310.
Figure 4.7 – Detail of slbo mRNA localization at stage 9 border cells.
In situ hybridization of early stage 9 chambers with slbo anti-sense probe. The slbo mRNA
localization is polarized and enriched in the apical side of the border cells.
4 RESULTS slbo
42
Figure 4.8 – slbo expression in wild type and slbory7 mutant egg chambers.
Left panels: slbo in situ hybridization of wild type and slbory7 stage 9 egg chambers. The
hybridization and image capture were done simultaneous and in equal conditions. Arrow points to
border cells. slbo expression is strongly reduced in the slbory7egg chambers. Right panel: slbo in
situ hybridization of slbory7 stage9/10 egg chamber developed longer in order to show that slbo is
still expressed in these border cells. All images with anterior to the left.
4.1.6 Analysis of candidate slbo target genes in border cells
In Pernille Rørth’s laboratory the main topic of research is guided cell migration
with the border cells as a model system. One approach taken to find genes
required for border cell migration was to screen for suppressors of the slbo
hypomorph phenotype. That is, to find genes that, when expressed in slbo
mutant border cells, partially suppress the migration phenotype. The screen was
done using the EP-element, a transposon based vector that randomly integrates
in the fly genome (Rørth 1996; Rørth et al. 1998). The EP-element contains UAS
sites which promote expression of genes nearby the integration place when
GAL4 is present. In this case GAL4 was specifically expressed in border cells in
a slbo hypomorph background. 2300 independent lines of flies were first
screened (Rørth et al. 1998) and later more 3500 liners by Pernille Rørth,
Gemma Texido and Kálman Somogyi. Each line, in principle, carries a different
random insertion. The flies that, in this context, showed some degree of border
cell migration were selected and the place of the EP-element insertion mapped in
order to identify the up-regulated gene.
4 RESULTS slbo
43
The genes that come out of this screen were good candidates for being
downstream of slbo since they could suppress the partial loss-of-function of slbo
itself. They could potentially also be downstream of slbo in the embryo and
consequently interesting for this work. To this list other genes studied in the lab
and related to border cell migration were added. The first approach to analyse if
they could be downstream of slbo was to check if they were expressed in border
cells. In situ hybridizations in egg chambers were done for the following genes:
tramtrack, Lk6, rutabaga, CG3149, Hsp27, α-Adaptin, kismet, tribbles, CG12342,
l(3)05822, CG6513, hyperplastic discs, calreticulin, CTP:phosphocholine
cytidylyltransferase 1, neuralized, CG5802, gilgamesh, widerborst, capricious,
CG5874, cortactin, CG13917, CG5874, CG8600, Malic enzyme, CG14995, 26-
29kD-proteinase, pontin, CG5508, CG7879, Hsp70Ab, CG6896, VhaPPA1-1,
CG7066, puckered, CG32103, Ubiquitin-specific protease 64E.
Figure 4.9 – expression pattern of kismet in egg chambers.
In situ hybridization of egg chambers with kismet anti-sense probe. Arrow points to border cells
cluster in a stage 10 egg chamber (the localization of the border cells was done using Nomarski
optics).
None of these genes showed a clear expression pattern in egg chambers. Egg
chambers are a difficult tissue to perform in situ hybridizations. The main reason
is that over 70% of the fly genes are thought to be expressed in the germline.
The germline is metabolically very active and a large number of transcripts is
required for this activity. On the other hand the nurse cells are also producing
high levels of mRNAs to be deposit in the oocyte. These mRNAs are responsible
4 RESULTS slbo
44
for virtually all protein synthesis in the first stages of embryonic development.
This high load of mRNAs in the germline has two consequences: most of the
genes will be expressed in the nurse cells and it increases the non-specific
hybridization of the probes. Both cases could mask expression in the border
cells.
A typical in situ hybridization of ovaries can be seen in Figure 4.9. In none of the
above genes was possible to see a stronger staining in the border cells than the
rest of the egg chamber. Consequently, there were no strong reasons to pursue
their analysis.
Figure 4.10 – CG9747 expression pattern in control and slbo hypomorphic mutant stage 10 egg chambers.
In situ hybridization of egg chambers with CG9747 anti-sense probe. Two wild type stage 10 egg
chambers are shown, one focused in the centre of the egg chamber, in order to show the border
cells staining, and one focused on the surface, in order to show the follicular staining. Arrow
points to border cells. Note that CG9747 is still expressed in the non-migrating slbory7 border
cells.
4 RESULTS slbo
45
However, it is possible to find by in situ hybridization genes expressed in border
cells. In situ hybridizations in egg chambers were also done for CG9747, ect and
ImpE2. Of these CG9747 showed expression in all follicle cells, including border
cells, from stage 10 on (Figure 4.10). CG9747 was still expressed in border cells
of slbo hypomorphic mutants. Because the in situ hybridization is not quantitative
it was not possible to know if the levels of expression were affected.
One other gene that showed expression in some border cells by in situ
hybridization was unpaired (Figure 4.11). unpaired is expressed only in the polar
cells, both anterior and posterior. The expression pattern of this gene was
studied in collaboration with Simone Beccari and for a different reason (Beccari
et al. 2002). unpaired codes for the ligand of the drosophila JAK/STAT pathway.
The activation of this pathway induces, and is required for, slbo expression in the
border cells (Silver and Montell 2001; Beccari et al. 2002).
Figure 4.11 - Expression pattern of unpaired in egg chambers
In situ hybridization, with unpaired anti-sense probe, of stages 5, 7, 8 and 10 egg chambers. The
labelled cells are polar cells.
Although it is possible to find some genes expressed in the border cells (slbo,
CG9747 and unpaired) it was not possible to find slbo candidate target genes.
This in situ screen of suppressors of slbo mutant cell migration defect did not
seem the good approach.
4 RESULTS slbo
46
4.2 Finding slbo target genes - Genomics approach
4.2.1 Drosophila cDNAs microarrays
As written before, a direct approach to understand slbo function is to identify its
target genes. The candidate gene approach only produced three genes whose
expression was partially dependent on slbo: CG9747, ect and ImpE2. This
information was very limited and I was interested in a broader view of slbo
function. To achieve this the expression profile of control and slbo mutant
embryos was compared. The genes that would be downregulated in the slbo
mutant embryos, in relation to control embryos, would be potential targets of
slbo. Moreover, to know which genes were down-regulated or up-regulated in
mutant, directly or indirectly, provides information on slbo function.
Late embryos were compared because slbo is highly expressed in many tissues
at this stage. Using all embryos, in opposition to isolated tissues or cells, has the
advantage of reducing manipulation of the samples and potential production of
artefacts. It is also much easier and simpler to obtain enough RNA for the
expression profiling. It has however the disadvantage of comparing many
different tissues simultaneously. Some tissues do not express slbo and slbo
function can be different between different slbo expressing tissues. Differences
between the slbo mutant and the control in specific tissues could be under-
estimated by comparing the whole embryo. However, the advantages of using
whole embryos prevailed.
The expression profiles were compared using DNA microarrays. This technique
provides a fast way of analysing the relative expression of thousands of genes
simultaneously. It is also relatively easy to probe different samples. The DNA
microarrays proper are glass slides with many (normally thousands) orderly
arranged spots of DNA (Figure 4.12). Each spot in the array has one type of DNA
4 RESULTS slbo
47
Figure 4.12 – Gene expression profile comparison using cDNA microarrays.
DNA microarrays are used to compare the expression profiles of two samples. Same amount of
mRNA from both samples is independently reversed transcripted and labelled with Cy3 or Cy5.
The two probes are then mixed together and hybridized to a CDNA microarray. The cDNA
microarray contains thousands of DNA spots. Each spot corresponds to a different cDNA. After
the hybridization the microarray is scanned and the fluorescence on each channel on each spot is
read. The intensities of the different channels, in the same spot, reflect the corresponding mRNA
abundance in the different samples. The results of the scanning are present as the ratio or the log
ratio of the two channels. Adapted from Cummings and Relman (2000).
molecules attached to the glass surface and serve as a probe for a specific gene.
The number of spots corresponds to the number of hybridisations simultaneously
done in an array. In each hybridization two samples are labelled with different
dyes and simultaneously probed. The proportion of labelled cDNA from one
sample in comparison to the labelled cDNA from the other sample, in one spot,
reflects the relative proportion of mRNAs of the corresponding gene in the
4 RESULTS slbo
48
original samples. I have used a cDNA based microarray with approximately 6000
non-redundant cDNAs, covering about 40% of the Drosophila predicted genes.
Because the microarrays technique is based on RNA hybridization onto DNA
spots there can be a problem of non-specific hybridization. This can lead to
under-estimation of real differences. However, the high-through output of the
technique makes it the best approach.
4.2.2 Trial microarray analysis
In order to test the efficiency and suitability of the technique I collaborated with
George Dimoupolis and Valerie Schaeefer in the production of Drosophila cDNA
microarrays at EMBL.
I initially compared two samples of 14-20-stage hours control and slbo mutant
embryos. The labelling protocol required a minimum of 5µg of polyA RNA, which
corresponds to 100 to 200µg of total RNA, equivalent to 1000 to 2000 embryos
for each sample. The embryo collection, based on GFP presence, was done by
hand which made the sample collection a limiting step in this trial phase. So that
10µg of total RNA would be sufficient for one probe an RNA amplification step
was introduced (Van Gelder et al. 1990). Two replicates of the DNA microarray
hybridization were done from the same amplified RNA. The data was analysed in
Excel and the fold-regulation for each gene was calculated has an average of the
four data points in the two slides (with 2 replicated spots each). Only genes that
showed down-regulation or up-regulation in both slides were considered.
The results indicated that there were only small differences between the
expressions profiles of control and slbo mutant embryos. The most
downregulated gene in slbo mutants, comparing with control, showed only 3.1
fold regulation. slbo itself showed only 2.5 fold regulation when it should show a
very strong regulation since is gone in the mutant. This could be due to non-
4 RESULTS slbo
49
specific hybridization and it will be discussed in section 4.2.4. One other gene
showed 2.1 fold down-regulation and all the other genes showed less then 2 fold
regulation. This general low-fold regulation could have several explanations and
will also be discussed in section 4.2.4.
To test if the DNA microarray results were reflected real differences in the RNA
samples a small subset of genes was selected for checking by RT-PCR. Twelve
genes that showed a fold change around 1.5 or more were selected (Figure
4.13). Three genes showed up-regulation in the mutant and nine showed
downregulation. These genes were not the top twelve regulated genes. ken for
example was the 49th most down-regulated gene. The purpose was to test if in
general the microarray results were valid. In the PCR reactions two primer pairs
did not work. Of the other 10 RT-PCRs, eight gave the expected up or down-
regulation while two showed no difference in the expression level between
control and slbo mutant sample. Thus, from one sample (two hybridizations) I
could get an 80% true positives rate. This was quite promising, since the false
positives could potentially be reduced by using replicates. Moreover, all four of
the downregulated genes where it was possible to get an wholemount in situ
hybridization (CG6347, Spn5, CG7675 and ken) showed an expression pattern
overall similar to ect and CG9747 (Figure 4.2 and Figure 4.13). All these genes
share with slbo the fact that they are expressed in the foregut and the posterior
spiracles. In the in situ database of the Berkeley Drosophila Gene Collection
(Tomancak et al. 2002) only 34 out of 1711 characterized genes are expressed
in these two tissues simultaneously giving some significance to this common
expression pattern. Cyp28d1, an up-regulated gene in the slbo mutant, is
expressed in the trachea. This is also a place of slbo expression. The analysis of
sample genes identified in the test DNA microarrays indicated that this technique
could be used to identify genes whose expression is dependent on slbo.
4 RESULTS slbo
50
Figure 4.13 – RT-PCR confirmation and pattern of expression of genes identified using DNA microarrays.
RT-PCR and wholemount in situ hybridization of genes misregulated in the slbo mutant, identified
by DNA microarrays. Twelve genes were select for analysis. The primers for CG10872 and
CG17930 did not work. The RT-PCR of CG17052 and CG9057 gave bands of equal intensity in
control and slbo sample and are not shown. Each RT-PCR presents a band corresponding to the
analysed gene and an rp49 band as control. Gene function is according to flybase (The flybase
Consortium2003). The slbo/control ratio is the average of the four data points in the two analysed
DNA microarrays. CG11395, CG8023 and CG15096 in situ hybridization did not work. All
embryos have anterior to the left. All embryos have dorsal up, except Cyp28d1 which is a dorsal
view.
4 RESULTS slbo
51
Unfortunately, my share of DNA microarrays finished abruptly and it was not
possible to spot new arrays from the PCR products because these were
inadequately kept. I had to re-do the cDNA microarrays from start. I was helped
by Belén Miñana from Vladimir Benes' EMBL Genomics Facility. As a start
replicates of the cDNA library were done. This insured that an untouched master
copy was kept. This also allowed the production of a "dirty" collection from where
individual interesting EST could be picked. This served me in my future work as
well as many Drosophila labs at EMBL. The PCR reactions and purifications
were done by Belén Miñana. These microarrays were also used by Andrea
Herold, in Elisa Izaurralde's laboratory, to study nuclear mRNA export (Herold et
al. 2003).
4.2.3 Embryo sorting with the COPAS SELECT system.
In order to reduce artifacts in the expression profiles I decided to label isolated
mRNA instead of doing an amplification step. I also planned to compare two
different embryonic development time points with several replicates. This meant
that several thousand embryos had to be sorted. It would be advantageous to
have some sort of automatic device.
A fly embryo sorting machine was first reported by Furlong et al (2001). I have
used a commercial embryo sorter from Union Biometricas. The working principle
is very similar to a Fluorescent Activated Cell Sorter (FACS). A stream of sample
containing the Drosophila embryos is passed through a laser and the
characteristics of each embryo (size, density and fluorescence in two channels)
are measured (Figure 4.14 and Figure 4.15). Each embryo is then, according to
setup, either collected or trashed.
4 RESULTS slbo
52
Figure 4.14 – Scatter plot of wildtype and enGAL4 UAS-GFP flies in the COPAS SELECT embryo sorter.
Scatter plots of the fluorescence values in the red and green channel. Wild-type embryos present
some background fluorescence. In the sample with the enGAL4 driver and the UAS-GFP it is
obvious the presence of a population expressing GFP.
Despite the publicized 95% sorted purity, the original COPAS SELECT system
sorted embryos with a purity that ranged from 60 to 90%, in an unpredictable
way. Anne Atzberger (running EMBL's FACS facility) and I identified the problem
as a bad design of the waste collection chamber. Because of that the sorted
embryos were contaminated with waste embryos. We designed a new waste
collection chamber for the machine that prevented this problem (Figure 4.15).
The machine sorts now with at least 95% purity and with 98-99% purity in ideal
conditions. Embryos were sorted at a speed of 15-25 embryos/second. Control
and slbo mutant embryos were collected from mixed parent strains (see section
4.1.3). This meant that theoretically 12.5% of the embryos would be GFP positive
but in practice only around 8-10% were positive. An average of 25.000 staged
embryos had to be screened per 2000 embryos sample.
4 RESULTS slbo
53
Figure 4.15 - COPAS SELECT system with new collection chamber
The scheme on the left shows the basics of the embryo sorter. The sheet fluid with the mixed
embryos passes through the flow cell and detector where the characteristics of the passing
embryos are analysed. By default the diverter blows the sample to the waste container (blue line).
When an embryo of interest is detected the flow of air stops for milliseconds and a droplet of
sheet fluid with the sorted embryo (in red) is allowed to pass through and be collected. In the
central photograph the column of sheet fluid is being blown to the waste collection chamber. On
the right the sheet fluid column is being “collected” (when really sorting there is no fluid column
but droplets). The design of the waste collector minimizes the accumulation of “waste” embryos at
the edge of the waste collection chamber.
4.2.4 DNA microarrays of control vs slbo mutant embryos
Wild-type and slbo mutant embryos were compared in two different embryonic
development time points. Embryos 12-14h and 16-18h after egg laying were
collected and sorted. The early time point has embryos mainly at stage 16. This
is the time point when slbo is widely and strongly expressed. The late time point
is at stage 17, when slbo, normally has already been expressed for some hours.
It is expected to find direct target genes of slbo on both time points. Indirect
effects of slbo loss-of-function are also expected in both time points however,
they should be stronger in the later sample.
Several replicates for each time point were planned but, unfortunately, only 2
replicates were done for the first time point and 3 for the late. This was due to
loss of samples through the mRNA purification procedure. Although it was
4 RESULTS slbo
54
possible to isolate mRNA from all of them, many did not label properly at the
following step. The labelling problem was, through an unknown mechanism, due
to the re-use of the polyT purification beads (Dynabeads). The re-use was part of
the kit protocol and the problem was only detected at the labelling step.
Nevertheless, two replicates are considered sufficient in many studies and much
information could still be extracted from the five hybridisations.
The scatter plots of the normalized data of the 5 hybridisations are shown in
Figure 4.16. One very obvious result from this analysis is that the differences
between control and slbo mutant samples are small. Over 95% of the genes
show an up- or down-regulation of less than twofold in all the different
microarrays, except in one of the 12-14h comparisons that shows more variation
than the others. More problematic is the low regulation shown by the slbo spots
(Figure 4.17). As discussed above these small differences between the two
samples could be because I used mRNA from whole embryos. This can be
further aggravated by the fact that non-specific hybridizations in the DNA
microarrays can lead to under-estimation of expression differences.
To exclude the possibility that there was technical problem two very different
samples, 0-2 hours and 14-16 hours embryos, were compared (Figure 4.18).
The results were the expected. There were huge differences between the
samples, 2000 genes (≈33% of the analysed genes) showed differences superior
to two fold. Oskar for example, a gene maternally deposited in the embryo but
absent from late embryos, showed a 29 fold difference between early and late
embryos. This control indicates that the technique is working and the low
regulation seen on the slbo mutant analysis must have a biological significance.
The data of the slbo spots are harder to explain. It would be expected a high fold
difference because slbo mRNA is absent in the mutant. In the early versus late
embryonic analysis slbo also does not show a very high regulation (Figure 4.18),
the ratios are actually very similar to the 12-14h experiments (Figure 4.17). In this
case it would be also expected a very high regulation because slbo is not
4 RESULTS slbo
55
Figure 4.16 – results of slbo versus control embryos microarrays.
A – scatter plots of 12-14h and 16-18h slbo versus control embryos microarrays. Each dot represents one spot in the microarray (one gene). The straight line where the ratios equal to one is shown. Genes below this line, in the graph, are downregulated in the slbo mutant, genes above this line are upregulated in the slbo mutant. The lines corresponding to a two fold change are also shown. B - histogram of the intensity ratios of 12-14h and 16-18h slbo versus control embryos microarrays. Each dot represents one spot in the microarray (one gene).
4 RESULTS slbo
56
expressed in the early sample but is highly expressed in the late sample. An
explanation could be that the slbo DNA sequence spotted on the arrays
hybridises non-specifically, masking any real regulation.
Figure 4.17 – slbo regulation in slbo mutant versus control microarrays.
The slbo mutant over control ratio of slbo expression is shown for the five microarray experiments
(logarithmic scale). In each microarray there were 15 spots of slbo DNA, each represented by a
square.
The fact that the 16-18h experiments show less slbo regulation than the 12-14h
experiments may have in fact a biological explanation. slbo expression is
downregulated at the end of embryogenesis. The 16-18h samples contain stage
17 embryos at a development stage where slbo is already downregulated hence
diminishing the differences between control and mutant embryos.
There are several reasons why there would be little regulation, overall, when
comparing the expression profiles of control and slbo mutants. One would be that
Slbo mainly or only regulates the relative levels of its target genes. When Slbo is
not present the level of expression of these genes only decreases slightly and
that is translated into these microarray results. Another aspect to consider is that
4 RESULTS slbo
57
whole embryo extracts are being compared. Slbo could completely control the
expression of one gene in a particular tissue but, at the embryo level, the mRNA
levels of that gene not be greatly affected. Also, Slbo could completely control
the expression level of a gene expressed only in a particular tissue but, at the
embryo level this expression is relatively low and thus the noise to signal ratio
high, affecting the microarray analysis.
Figure 4.18 – Results of DNA microarray analysis of early versus late embryos.
(A) Scatter plot of DNA microarray data of early (0-2h) versus late (14-16h) embryos.
Approximately 2000 genes have expression levels differences of more than 2-fold between
samples. Arrow points to Oskar’s datapoint. Oskar is expressed 29 times higher in early embryos
than late embryos, according to the microarray data.
(B) Ratio of slbo expression in early over late embryos in the DNA microarray hybridization
(logarithmic scale). Each square represents a slbo DNA spot in the microarray.
A different control available was to look at the genes identified in the first trial
microarray analysis (Section 4.2.2) and see how they behaved in these new
microarrays. The genes showed the expected slight downregulation in some
4 RESULTS slbo
58
hybridizations but not in others (Figure 4.19 A). This was not a defect of the
microoarrays. The expression of one gene, CG11395, was tested by RT-PCR in
the same RNA samples used for a microarray which gave no difference between
slbo mutants and control (first 16-18h sample from the left of Figure 4.19 A) and
a sample collected when the first trial microarrays were done (from Figure 4.13).
The RT-PCRs gave the same results as the microarrays (Figure 4.19 B). The
conclusion to draw is that the differences between control and slbo were
changing from sample to sample. This change had a temporal component to it.
The two samples that showed less regulation of these genes (12-14h I and 16-
18h I) were the last samples to be collected. Something changed in the fourty
days between the first samples to be analysed and the last. This time
corresponds to more or less three generations when the flies were being kept in
high numbers.
It was not possible to determine what the cause of this change was. A posterior
attempt to address this was done; the recombinant mutant stocks were re-
established to get rid of potential backgroung compensatory mutations. However
CG11395 and other genes’ regulation were not recovered. Nonetheless, the
important fact to consider is that the lethal phenotype of the slbo mutant was
always present. This indicates that crucial functions of slbo are always absent, in
all samples.
The small differences between the two samples made the analysis difficult
because with fold changes so low the noise to is very high. Only with 4 to 6
replicates per sample could the truly slightly up- or down-regulated genes be
extracted with statistical methods. Since it was not possible to make a broad
perspective list of genes affected by slbo, a pragmatic approach was taken. To
find genes dependent on slbo it was only considered genes that showed a down-
regulation of 1.5 fold or more in all the hybridizations. This 1.5 fold limit seemed
reasonable and had worked in the trial hybridisations analysis. The list was
comprised of twelve genes (Figure 4.20). If the genes showing a down-regulation
4 RESULTS slbo
59
of more than 1.5 fold in each microarray were just the result of noise (whith this
low regulation many may be) no gene (0.006 genes) would be expected to be
down-regulated in all samples. Thus, finding twelve genes being down-regulated
in all samples is not just the result of chance.
Figure 4.19 – CG11395 expression dependence on slbo
(A) Ratio of CG11395 expression in early over late embryos in the DNA microarray hybridizations
(logarithmic scale). The EST corresponding to CG11395 in the DNA microarrays is GH14572.
(B) RT-PCR with CG11395 specific primers of two different samples (see text for details). rp49
primers were used as control.
These genes were then analysed by RT-PCR, from new RNA samples, to verify
the microarray data. Three genes showed absolute dependence on slbo to be
expressed. Three were always downregulated in the slbo mutant but still present.
The expression levels of three genes only showed, by RT-PCR, to be dependent
of slbo in the first time point. One gene did not show any regulation by slbo and
the PCR reaction did not work for two genes. The possible role of some of these
4 RESULTS slbo
60
genes (particularly the ones completely dependent on slbo) in the slbo phenotype
and function will be discussed later.
EST Gene RT-PCR Function
GH04024 CG8023 gone Translation initiation factor
GH13851 Odc1 gone Ornithine decarboxylase
LD11648 GlcAT-P gone UDP-glucuronosyltransferase
GH11618 CG7149 downregulated Diacylglycerol cholinephosphotransferase
LD46766 CG8505 downregulated Cuticle protein
LD18835 CG7080 downregulated -
LD40136 CG10657 downregulated (12-14h sample)/
not regulated (16-18h sample)
Retinal binding
LD23868 CG16721 downregulated (12-14h sample)/
not regulated (16-18h sample)
-
LD19727 Tsp96F downregulated (12-14h sample)/
not regulated (16-18h sample)
(Membrane protein)
GH13704 CG15281 not regulated -
GH21171 CG8677 n.d. (PHD domain)
LD21452 CG4060 n.d. -
Figure 4.20 – RT-PCR results of 12 genes showing downregulation in slbo mutants in DNA microarrays experiments.
The 12 genes selected in the DNA microarrays experiments were tested by RT-PCR in 12-14h
and 16-18h samples. Results in the RT-PCR column refer to comparison of slbo mutant sample
to control sample. CG8677 and CG4060 primers did not work. Function is according to Flybase
(The flybase Consortium2003) except the two notes between brackets.
Apart from this more solid set of genes the rest of the data was not ignored.
Although it was not correct to present, as data, lists of genes affected in the slbo
mutant in one or the other development stages, these list were looked upon. I
also looked at all genes that showed some regulation in 4 out the 5
hybridisations. The genes picked up in this way and thought to be relevant will
also be discussed further on. One word of precaution is however required. With
this type of analysis (as well as with most microarray analysis) all kind of genes
4 RESULTS slbo
61
are found to be affected. Since I will present only the genes that I find relevant I
am logically being biased. There is the risk that many lists of different genes
could be extracted from this fuzzy analysis and therefore many different
conclusions and theories presented. However, I will only discuss mis-regulated
genes that could be relevant to explain the phenotypes of the slbo mutant.
4 RESULTS slbo
62
4.3 slbo and metabolism regulation
4.3.1 slbo mutants hatching defect and larval lethality Aided by the possibility of directly selecting, by GFP fluorescence, slbo mutants
the late embryonic/early larval lethal phenotype was re-examined. Mutant
embryos and control embryos were selected 16-20 hours after egg laying and
followed up in order to check hatching and larvae survival rates (Figure 4.21).
While all control embryos hatched and were alive one day after hatching, slbo
mutants presented problems in hatching and in larvae survival. Many times slbo
mutants dye literally upon hatching; half inside, half outside the eggshell. It is
also possible to observe many slbo mutant larvae dead just next to the eggshell.
The hatching rate of slbo mutants is quite variable. In the presented data varies
from an average of 40 to 87% (Figure 4.21 B and C) but any value from 0 to
100% can be observed. There is also variability at the level of larvae survival one
day after hatching. However, this survival rate, in contrast to the hatching rates,
is always low (16.7% and 1.7% in Figure 4.21 B and C).
This variation has, probably, many causes, some of them being extrinsic to the
embryos. One example can be seen on Figure 4.21. The two data sets shown for
slbo mutant were done simultaneously and crossing the same two stocks of
progenitors flies. This means that the genetic background and the external
conditions of the analysed embryos in the two graphs were exactly the same.
The difference between them lies only on which stock provided the females and
which provided the males. That implies that there is some sort of maternal
contribution affecting the hatching phenotype. This maternal contribution is not
due to the presence of slbo RNA or protein in the egg since these are not
detected in the embryo before stage 11, 5h30m after the beginning of
embryogenesis (Figure 4.1). Other maternal provided factors must modulate slbo
function.
4 RESULTS slbo
63
A - Control
0
20
40
60
80
100
Hatching 1 2 3 4 5 6
Days after hatching
Perc
enta
ge o
f hat
chin
g or
sur
viva
l
B - slbo (I)
0
20
40
60
80
100
Hatching 1 2 3 4 5 6
Days after hatching
Perc
enta
ge o
f hat
chin
g or
sur
viva
l
C - slbo (II)
0
20
40
60
80
100
Hatching 1 2 3 4 5 6
Days after hatching
Perc
enta
ge o
f hat
chin
g or
sur
viva
l
Figure 4.21 – slbo mutants have a reduced hatching and survival rate.
Hatching and survival rate of control (A) and slbo mutants from two different crosses (B and C).
For each panel three replicates of twenty eggs were placed in new apple plates with fresh yeast
and monitored. For control flies (A) survival rate was only checked on the first and fifth day after
hatching. Fifth day value in control flies (A) corresponds to number of pupae. No pupae were
detected in the other two experiments. In (B) progenitors are: ♂ slboe7b, enG4/ CyO and ♀
slbo8ex2, UAS-GFP / CyO. In (C) progenitors are: ♂ slbo8ex2, UAS-GFP / CyO and ♀ slboe7b,
enG4/ CyO. Bar height represents mean and error bars standard deviation.
Variability in hatching and larvae survival rates is also observed when using the
same stocks of females and males progenitors but from different generations.
However, when using exactly the same conditions at the same time the values
observed are acceptably reproducible from replicate to replicate. So, although
values can not be compared between experiments they can be compared within
an experiment. Care was taken, in what concerns slbo mutant analysis, to always
compare samples handled simultaneously and with mothers from the same
stock.
4 RESULTS slbo
64
The novelty in this re-analysis is the observation that some slbo mutant larvae
are alive one day after hatching. These larvae do not develop normally (see
below, section 4.3.3) and usually die before reaching pupariation. From the first
day after hatching on the survival rate is fifty percent or more per day, which
contrasts with the very low survival rate before that time point. This indicates that
there is a first episode of death, just around hatching, distinct from the death of
the escaper larvae.
Very rarely larvae reach pupal stages and even more rarely adulthood. External
morphology in the adults seems overall normal but for some rare small small
defects in the legs. Adults are also very feeble, sometimes uncoordinated, dying
within one day after eclosion. They present a tendency to stick to the food or glue
their legs upon each other or to the rest of the body.
4.3.2 slbo mutants and methyl p-hydroxybenzoate sensitivity
4.3.2.1 slbo mutants are very sensitive to methyl p-hydroxybenzoate
Serendipitously, I observed that slbo larvae survived much better in filter papers
soaked in PBS than in the apple-agar plates normally used. Apple-agar plates
are constituted of water, agar, sucrose, apple juice and methyl p-
hydroxybenzoate, an anti-mold reagent. slbo larvae were grown on agar plates
with sugar and apple juice or only methyl p-hydroxybenzoate in order to test
these components for toxicity. While larvae grew a little bit better with sugar than
without sugar (a known fact) they were severely affected by the presence of
methyl p-hydroxybenzoate, at the concentration normally used in fly food (0.15%)
(Figure 4.22 C and D). slbo larvae grew dramatically better in simple agar plates.
Not only many more larvae survived (85% in the first day after hatching
compared with 10% in the presence of methyl p-hydroxybenzoate) but more
larvae pupariated (25% compared with 3%) and became adult (10% compared
with 0%).
4 RESULTS slbo
65
One interesting aspect is that the very high lethality around the hatching period is
strongly diminished when slbo larvae are grown on simple agar plates which
indicates that these initial deaths were mainly due to the methyl p-
hydroxybenzoate. The high number of non-hatching embryos is gone in the
absence of it. Generally, slbo mutant larvae seem healthier and more active in
the absence of methyl p-hydroxybenzoate. Adults were also healthier and
survived more than one day after eclosion, surviving nonetheless less than ten
days (in optimal conditions adult lifespan is around 50 days (Ashburner and
Thompson 1978)).
Figure 4.22 – slbo mutant are very sensitive to methyl p-hydroxybenzoate
control and slbo mutant 16-20h embryos were placed on agar plates with or without methyl p-
hydroxybenzoate. Larvae were fed with fresh yeast. The percentage of surviving larvae was
determined from one day after hatching on. Pupae and adult flies’ numbers were also monitored.
Values present for pupae and adult flies are cumulative. Three samples of twenty embryos were
scored for each condition. Bars represent standard deviations.
4 RESULTS slbo
66
Methyl p-hydroxybenzoate is a common preservative present in food,
pharmaceutical and cosmetic products (for a comprehensive review see Soni et
al. 2002). It is the most used member of a family of p-hydroxybenzoic acid esters
(Figure 4.23) and is also known as methyl-paraben, Nipagin or E-218. Soni et al.
(2002) estimate that about 50 mg of methyl p-hydroxybenzoate is consumed per
person per day. This is the result of its widespread use in products for human
consumption. Consequently there is a large body of data concerning toxicity in
mammals. Methyl p-hydroxybenzoate is rapidly and completely absorbed from
the gastrointestinal tract, metabolized and excreted. None of the original
compound or its metabolites accumulates in the body. Methyl p-hydroxybenzoate
is excreted in the form of p-hydroxybenzoic acid and glucoronic acid and sulfuric
acid conjugates of p-hydroxybenzoic acid. Most of the xenobiotic substance is
metabolized in the liver and kidneys. As expected, methyl p-hydroxybenzoate
has a low acute toxicity in tested mammals (e.g. oral LD50 dose vary between 2-
8g/kg). Fatal doses result in rapid loss of muscular control (ataxia), paralysis,
depression of the central nervous system and rapid death. The death of slbo
mutant larvae grown on the presence of methyl p-hydroxybenzoate could be
related with these ataxia and paralysis. In fact many of the larvae die just next to
or very near the egg.
Figure 4.23 – Structural formula of methyl p-hydroxybenzoate.
Studies on methyl p-hydroxybenzoate do not show a single major mode of action
against microorganisms or in its effects, at high doses, on mammals.
Biochemical studies indicate that p-hydroxybenzoate can affect some enzymes
HO C
O
O CH3
4 RESULTS slbo
67
of microorganisms and mammals (e.g. oxidative enzymes in E. coli). It has
cytotoxyc effects related with mitochondria failure dependent on membrane
permeability transition and depolarization (Nakagawa and Moore 1999). Methyl
p-hydroxybenzoate has been also shown to have some local anaesthetic activity
and some cases of contact sensitivity associated with cutaneous exposure have
been reported in humans. Overall it is impossible to hypothesize how the drug
might be killing the slbo mutants.
Methyl p-hydroxybenzoate is extensively used in fly food preparation (most
Drosophila stocks in the world grow with it) and reported not to affect Drosophila
at the used concentration (Ashburner and Thompson 1978). However, methyl p-
hydroxybenzoate has adverse effects on the fitness of other insects (Rojas et al.
1990; Alverson and Cohen 2002). A prenylated form of it was isolated from a
pepper family plant as its major insecticidal principle against Aedes atropalpus
mosquito larvae (Pereda-Miranda et al. 1997).
Curiously, in my experiments, there was around 10% lethality when control
larvae were raised in agar plates with methyl p-hydroxybenzoate, more than the
2% lethality observed on control larvae raised in simple agar plates (Figure 4.22
A and B). Pupariation and eclosion was also slightly delayed in the presence of
methyl p-hydroxybenzoate. Normally wild type larvae grow without problems on
apple-agar plates, which have the same concentration of methyl p-
hydroxybenzoate (see Figure 4.21 A). The reason control sensibility in my
experiments could be the absence of sucrose in the medium. It could be simply
that larvae are less healthy when their diet does not have high levels of sugar
and consequently more sensitive to the drug. It could also be that sugars are
required for the metabolism of methyl p-hydroxybenzoate by the larvae. A final
hypothesis would be that the methyl p-hydroxybenzoate effective concentration is
lower when the medium is prepared with sugar. Methyl p-hydroxybenzoate has
been reported to react with sugars at around 50˚C (Ma et al. 2002) and fly foods
are normally heated at this temperature level when prepared.
4 RESULTS slbo
68
4.3.2.2 DmGlcAT-BSII and slbo mutant methyl p-hydroxybenzoate sensitivity
Since control flies have a slight sensitivity to methyl p-hydroxybenzoate and it is
known that in mammals the drug is rapidly metabolized and excreted. The
sensitivity of slbo mutants to methyl p-hydroxybenzoate could be related to
inefficient elimination of this compound from the organism.
Glycoside formation is a common mechanism of detoxification of endogenous
and exogenous compounds, increasing their solubility. These reactions are
catalyzed by the protein superfamily of UDP-glycosyltransferases (UGTs). As
mentioned above the glucoronic acid conjugate of p-hydroxybenzoic acid is a
common secretion product in mammals methyl p-hydroxybenzoate metabolism
(Soni et al. 2002). More interestingly, in the insect Heliothis virescens a metabolic
product of methyl p-hydroxybenzoate was identified as methyl benzoate-4-β-
glucoside, the product of methyl p-hydroxybenzoate glucosidation (Rojas et al.
1990). This concurs with the preferential use of UDP-glucose rather than UDP-
glucoronic acid as sugar donor by insect UGTs (Luque et al. 2002).
Figure 4.24 – DmGlcAT-BSII embryonic expression is dependent on slbo.
RT-PCR analysis of control and slbo mutant embryonic samples with DmGlcAT-BSII specific
primers. rp49 primers were used as control.
4 RESULTS slbo
69
One of the three genes whose expression is completely dependent on slbo,
DmGlcAT-BSII (GlcAT-P in Flybase) (Figure 4.24 and see section 4.2.4),
encodes a UTP-glycosyltransferase (Kim et al. 2003). Its embryonic expression
pattern has been determined by the in situ database of the Berkeley Drosophila
Gene Collection (Tomancak et al. 2002). It is enriched in the salivary glands but
also expressed in other non-specified tissues of the embryo.
Drosophila melanogaster genome encodes over 30 predicted UGTs (Luque and
O'Reilly 2002). It is not possible to predict the substrate specificity of a
determined UGT. It may have an UGT activity towards some members of a
compound class but not others and it may have specific activity towards
members of very different classes (Luque et al. 2002). The mammalian proteins
most similar to DmGlcAT-BSII are GlcAT-I, GlcAT-S and GlcAT-P. These
enzymes are UDP-glucoronyltransferases involved in the formation of
proteoglycans and glycoproteins (Kitagawa et al. 1998; Terayama et al. 1998;
Shimoda et al. 1999). A phylogenetic tree with DmGlAT-BSII, most Drosophila
UGTs and the mammalian GlcAT-I, GlcAT-S and GlcAT-P was done (Figure
4.25). DmGlAT-BSII clusters with GlcAT-I, GlcAT-S and GlcAT-P, together with
DmGlcAT-I and DmGlcAT-BSI. This indicates that these Drosophila genes may
also be involved in sugar modifications of proteins. In vitro data show that these
Drosophila proteins do indeed have glucoronyltransferase activity towards the
substrate of GlcAT-I (Kim et al. 2003). DmGlcAT-I is not only the one more close
to these mammalian proteins in the phylogenetic tree but also the Drosophila
protein that shows more specificity. DmGlAT-BSI and DmGlAT-BSII also added
UTP-gluycoronic acid to other two similar substrates.
The biochemical analysis of DmGlAT-BSII shows that it is a UTP-
glucoryltransferase (Kim et al. 2003), so unlikely a UTP-glucosyltransferase. It
also indicates that its biological function is probably more related to post-
translational protein modification than to detoxification. However, the above
4 RESULTS slbo
70
experiments were done in vitro and DmGlAT-BSII showed some lack of
specificity. It is still possible that it may be related with methyl p-hydroxybenzoate
sensitivity. Unfortunately, this sensitivity was only found at the end of the PhD
research so the relation between it and DmGlAT-BSII was not properly
addressed.
Figure 4.25 – Phylogenetic tree of Drosophila UDP-glycosyltransferases and mammalian UDP-glucoronyltransferases
DmGlAT-BSII is underlined. Phylogenetic tree done by Neighbour-Joining method using Clustal
X. Number in tree branches indicate bootstrap values (1000 bootstrap trials). Branch size is
proportional to phylogenetic distance. The tree is unrooted.
4.3.3 slbo mutants have a growth defect and an altered feeding behaviour
slbo mutant escaper larvae present several abnormal characteristics. One
obvious initial phenotype is that they move slower than control larvae. It is also
possible to observe that slbo escaper larvae have a growth defect (Figure 4.26).
4 RESULTS slbo
71
slbo mutant larvae are always smaller than control larvae, with their size varying
from individual to individual. Most larvae grow very little, as shown in the bottom
panel of Figure 4.26. These larvae can pass the normal time of pupariation
without starting pupariation and live much longer as larvae than normal. slbo
mutant larvae also seem more translucent the control larvae, even if comparing
same size (different age) larvae (compare top panel control day 4 larva with
middle panel slbo mutant day 5 larva in Figure 4.26). Larval tissues are in
general translucent allowing the visualization of internal organs. One of these
organs, the fat body, is opaque in third instar larvae as it accumulates
metabolites in the form of triacylglycerols, glycogen and proteins (Britton et al.
2002). The translucency of slbo mutants may reflect a less opaque and
developed fat body.
Figure 4.26 – slbo mutant larvae have a growth defect
Control and slbo larvae were followed up through development. Control representative larvae are
on the top panels. largest slbo mutant larva detected on a specific day is depicted on middle
panels, smallest slbo larvae of that day is depicted on the lower panel. Most slbo larvae grow very
little over time.
The growth defect phenotype could indicate a problem in metabolism of slbo
mutants. However, it can also be a symptom of any general problem.
4 RESULTS slbo
72
Interestingly, slbo mutants present two other phenotypes that could be related to
metabolism problems: they wander away from the food and eat less than normal
larvae (Figure 4.27). This is in striking contrast with the normal larval behaviour.
Larvae are essentially eating machines. As they hatch they dive into the food and
continuously eat for four/five days, growing and accumulating nutrients, until they
wander away from the food and pupariate.
When put into an agar plate with fresh yeast normal larvae disappear into the
yeast and are difficult to observe. slbo mutant larvae are normally easy to spot
since they remain on the surface of the agar plate. They also have the tendency
to leave the agar and go to the lid of the Petri dish. To better analyse this
abnormal behaviour larvae were placed in an agar plate with fresh yeast in the
middle and the number of larvae outside the yeast was quantified one day later
(Figure 4.27 A). The number of control larvae outside the yeast was very low
(1.7%) while most of the slbo mutant larvae (82%) were wandering in the plate.
slbo mutant larvae could be outside the food simply because they move less and
do not find it. To discard this possibility the larvae were actually placed inside a
ring of yeast so that wandering larvae were necessarily larvae that moved away
from the food. This wandering behaviour could be either a preference to be out of
the food or incapacity to recognize where the food is. Anecdotal evidence
indicates that slbo mutants are “aware” of food localization and chose to be out of
the food. For instance, slbo mutant larvae seem to linger around the fringe of the
fresh yeast. Also, when mutant larvae are transferred from a plate without fresh
yeast to a plate with fresh yeast many move towards the yeast. In this kind of
setup the main difference between slbo mutant and control larvae is that the
former seldom dive into the yeast. This movement towards the food is later
reversed as posterior observations of the same plate show that the mutant larvae
are mainly out of the yeast.
slbo mutant larvae were fed with dyed fresh yeast in order to analyse their food
intake (Figure 4.27 B). While all control larvae had the gut filled with food, slbo
4 RESULTS slbo
73
mutant showed, again, a feeding defect. Most mutant larvae had only some food
in the gut, although, a few had the gut full of food and a few had none. This result
shows that slbo mutant larvae can eat but do not do it normally. A lower intake of
food could be an explanation for the growth defect.
Figure 4.27 – slbo mutant larvae have an altered feeding behaviour.
(A) One day after hatching control and mutant larvae were placed in the center of a ring of fresh
yeast in an apple-plate and scored the next day for wandering larvae. Note the slbo mutant
wandering larvae. The graph shows average of three samples of twenty larvae for each
genotype. Error bars represent standard deviation.
(B) One day after hatching control and mutant larvae were fed with fresh yeast dyed with
bromophenol blue and the quantity of ingested food was checked the next day. All control larvae
had the gut full of food. slbo mutant displayed a range of phenotypes, from no food in the gut to
the gut full of food. Most slbo larvae had an intermediate phenotype.
The above observations were done in the presence of methyl p-
hydroxybenzoate. In its absence mutant larvae still present growth defects,
although slightly less severe; while many slbo larvae still persist without almost
any growth throughout time some tend to develop at a rate only slightly slower
than control. The feeding behaviour also follows the same trend, it is still present
4 RESULTS slbo
74
but less severe. It is common to find larvae on the yeast and outside it, still very
different from control larvae which are always in the yeast.
The combination of these phenotypes (growth defect, wandering behaviour and
low food intake) has also been described for larvae with the insulin/PI3 signalling
pathway constitutively activated (Britton et al. 2002) or with an excess of
aminoacids in the food or an aminoacid catabolism deficiency (Zinke et al. 1999).
As in mammals, the insulin/PI3 signalling pathway in Drosophila is a sensor of
the nutritional status of the organism (Britton et al. 2002; Ikeya et al. 2002).
Aminoacids levels are also a signal of the nutritional status of an organism
(Oldham and Hafen 2003)). The altered feeding behaviour is thus present when
a high nutritional level is mimicked. This behaviour is similar to the behaviour of
larvae just before pupariation, in the wandering stage. An interpretation proposed
for these observations was that a high nutritional level reading, normally only
achieved at the end of larval stages, triggers the wandering behaviour and
cessation of feeding (Zinke et al. 1999; Britton et al. 2002). It is possible that in
the slbo mutant a similar deregulation of metabolism and mimicry of a high
nutritional level is present.
4.3.4 sugar metabolism related genes mis-regulated in the slbo mutant
One of the first observations of the DNA microarray results was that there was a
high number of sugar transporters being up-regulated in the slbo mutant. When
all the sugar transporters coding genes present in the microarrays are observed
it is clear that overall there is an up-regulation of their expression (Figure 4.28).
This up-regulation of sugar transporters genes is consistent with the high
nutritional status mimicry suggested in section 4.3.3. The perception of high
levels of circulatory sugars could trigger this response. In mammals, insulin (a
signal of high glucose concentration in the blood) stimulates sugar uptake in
target tissues (reviewed inCzech 1995). One of the main mechanisms of insulin
4 RESULTS slbo
75
action is the redistribution of the glucose transporter GLUT4 from intracellular
membranes to the cell surface. However GLUT4 mRNA levels are also regulated
by the nutritional status and insulin levels (Sivitz et al. 1989). Fasted or insulin
deficient rats have a much lower level of expression than normal while insulin
treated or reefed fasted (with a fast increase of blood sugar) have a higher level
of expression than normal.
Figure 4.28 – sugar transporter coding genes up-regulation in the slbo mutant.
Ratio of sugar transporter coding genes expression in early over late embryos in the DNA
microarray hybridizations (logarithmic scale). All the sugar transporter encoding genes present in
the DNA microarrays are shown. Each is represented by a square, in each sample. The ESTs in
the microarray are: GH25507, GM13904, LD44652, GH07001, SD10060, GH09052, LP08815
and GH19118.
Another gene related with sugar metabolism mis-expressed in the slbo mutant is
CG5171 (Figure 4.29). This gene is down-regulated and encodes a trehalose
phosphatase, which converts α,α-trehalose-6P to α,α-trehalose. In insects
trehalose, a disaccharide, is the main circulatory carbohydrate and is produced in
the fat body (Keeley 1985). Trehalase which converts trehalose to glucose
4 RESULTS slbo
76
seems to be up-regulated in the late stages slbo mutant embryos (Figure 4.29).
In the insect Aedes aegypti infected with the parasite Plagiorchis elegans the
opposite regulation happens, trehalose phosphatase is up-regulated and
trehalase is down-regulated, resulting in accumulation of trehalose (Wallage et
al. 2001). In the slbo mutant the combination of these two gene mis-regulation
may result in a lower level of trehalose in the hemolymph. Within the same high
nutritional status mimicking hypothesis the perception of high levels of circulating
sugars would prompt a reaction from the slbo mutant to reduce them.
slbo/control
EST Gene 12-14h I 12-14h II 16-18h I 16-18h II 16-18h III
LD21828 CG5171
(trehalose phosphatase) 0.892 0.507 0.594 0.405 0.176
GH13461 Trehalase 1.146 1.05 1.238 1.914 2.492
LD36528 sugarbabe 0.961 0.844 0.657 0.67 0.7
Figure 4.29 – sugar metabolism related genes mis-regulated in the slbo mutant.
Ratio of slbo mutant over control embryos expression of some sugar metabolism related genes in
the DNA microarray hybridizations.
sugarbabe was found as a gene fast and strongly induced by sugar feeding
conditions (Zinke et al. 2002). It encodes a transcription factor and it is postulated
to be involved in the response to high levels of sugar. In late embryonic stages
sugarbabe is downregulated in the slbo mutant (Figure 4.29). In this case, in
opposition to the other genes discussed in this section, a gene known to be
induced by high sugar levels in actually downregulated in the slbo mutant. Based
on the presence of C/EBP binding sites in the upstream regulatory region of
sugarbabe, Zinke et al (2002) suggested that sugarbabe could be a target gene
of slbo. C/EBP binding sites are short and very variable. Therefore finding C/EBP
binding sites in a determined DNA sequence is relatively frequent and not very
informative. However, my microarray results corroborate the suggestion.
Contrary to what was suggested for the other genes in this section, sugarbabe
4 RESULTS slbo
77
could be a direct target of slbo. Its downregulation could be partially responsible
for the general metabolism mis-regulation.
The transcription profile of the sugar transporters and trehalose metabolism
genes is compatible with a high nutritional status mimickry. On the other hand,
sugarbabe downregulation in the slbo mutant could partially contribute to this
metabolism mis-regulation.
4.3.5 slbo and polyamines
4.3.5.1 Ornithine decarboxylase and polyamines function
Another gene whose expression is completely dependent on slbo is Ornithine
decarboxylase 1 (Odc1) (Figure 4.30 and see section 4.2.4). The Odc enzyme is
the first and a rate-limiting enzyme in the synthesis of polyamines in eukaryotes
(bacteria can also synthesize polyamines through other pathways) (Figure 4.31)
(For reviews seeTabor and Tabor 1984; Igarashi and Kashiwagi 2000; Thomas
and Thomas 2003). Polyamines are organic polycations, their function is very
diverse and many different processes are influenced by polyamines. They have
electrostatic interactions through their amino groups and hydrophobic
interactions through the carbon chains. Polyamines can interact with DNA,
RNAs, proteins, phospholipids and nucleotide triphosphates. The binding of
polyamines frequently change the structure of the bound molecule and
consequently its properties. Sometimes their action can be replaced by other
simple inorganic cations, sometimes their binding and role is very specific.
Mutants in the polyamine pathway have defects in growth and cell proliferation.
In Saccharomyces cerevisiae, Neurospora crassa and Schizosaccharomyces
pombe polyamines are essential for both this processes (in Tabor and Tabor
1984; Chattopadhyay et al. 2002). Escherichia coli mutants with no polyamines
proliferate slower. In mammals Odc (and polyamines) is upregulated in
proliferating cells and can induce cell proliferation (Thomas and Thomas 2003).
4 RESULTS slbo
78
These characteristics make Odc a marker of carcinogenesis and tumor
progression. The relation of polyamines and growth is, most probably, related
with these compounds requirement and association with DNA replication, RNA
transcription and protein translation.
Figure 4.30 – Odc1 embryonic expression is dependent on slbo.
RT-PCR analysis of control and slbo mutant embryonic samples with Odc1 specific primers. rp49
primers were used as control. rp49 band in the control sample is less intense than in the slbo
sample because of competition with the Odc1 fragment in the PCR reaction.
Figure 4.31 – Scheme of the polyamines synthesis pathway in animals
In insects polyamines are also associated with growth and cell proliferation. In
crickets Odc activity is required for juvenile hormone induced neuroblast
4 RESULTS slbo
79
proliferation and putrescine feeding induces this same proliferation (Cayre et al.
1997). In the silkmoth Odc activity is induced by the hormone ecdysone and
correlates with the growth response to this hormone (Wyatt et al. 1973). In
Drosophila adult females Odc activity is induced by juvenile hormone and
correlates with juvenile hormone induced vitellogenesis (the female ovaries are
the main site of growth in the adult fly) (Birnbaum and Gilbert 1990). The three
examples have in common a hormone induced Odc activity correlated with a
growth response. In Drosophila it is also known that administration of spermidine
to larvae increases ribosomal RNA production (Byus and Herbst 1976).
The enzymes of the polyamine synthesis pathway are tightly regulated (seeTabor
and Tabor 1984; Igarashi and Kashiwagi 2000; Thomas and Thomas 2003). The
turn-over rate of Odc is one of the fastest known in mammals, about 10 to 30
minutes. Odc expression is also controlled at the levels of transcription and
translation. The enzymatic activity is also controlled by antizymes, proteins that
act as inhibitors by binding to Odc.
The connection of Odc and growth as well as the basic fact that polyamines play
essential roles in life made looking more closely to its relation with the slbo loss-
of-function phenotype interesting.
4.3.5.2 Odc1 and Odc2 Drosophila melanogaster has two genes very similar to mammalian Odc: Odc1
and Odc2 (Rom and Kahana 1993). They are located on chromosome 2 just next
to each other, having the same exon/intron structure (Rom and Kahana 1993).
That and their position in a phylogenetic tree of several Odcs (Figure 4.32)
indicate that these two genes are a result of recent gene duplication. From the
phylogenetic tree is also possible to observe that Odc2 is slightly more derived
because (branch size relates to phylogenetic distance).
4 RESULTS slbo
80
Figure 4.32 – Phylogenetic tree of ornithine decarboxilases from different genus and Drosophila melanogaster homologues.
Phylogenetic tree done by Neighbour-Joining method using Clustal X. Number in tree branches
indicate bootstrap values (1000 bootstrap trials). Branch size is proportional to phylogenetic
distance. An alignment of Odc1, Odc2 and Odcs from several other organisms shows that
aminoacid residues conserved between all other Odcs are not conserved in
Odc2. On Figure 4.33 A, it is shown one stretch of the proteins highly conserved
between all Odcs and Odc1 but is not conserved in Odc2. This stretch
corresponds to an alpha-helix near the catalytic centre in the determined
structure of Trypanosoma’s Odc (Kern et al. 1999). An aspartate residue in this
alpha-helix makes hydrogen bonds with the co-crystalized product of the
catalysed reaction and, potentially, with the substrate of the enzyme (Figure 4.33
B). This aspartate also makes a hydrogen bond with a conserved arginine
residue in a different domain of the protein, also contacting the catalytic centre of
the protein. In Odc2 this negatively charged aspartate residue is replaced by the
bulkier uncharged tyrosine. This substitution probably disrupts the structure of
the catalytic centre and the enzyme activity. This indicates that Odc2 is unlikely
4 RESULTS slbo
81
Figure 4.33 – Sequence alignment of a highly conserved region in ornithine decarboxylases and structure of Odc catalytic centre.
A) Sequence alignment of a highly conserved region in ornithine decarboxylases but not in Odc2.
Alignment was done with Clustal X. Bar on top indicates the region of interest. Asterisk indicates
aspartate residue discussed in the text and not conserved in Odc2.
B) Structure of the catalytic centre of Trypanosoma Odc (based onKern et al. 1999). Odc is a
homodimer, blue and white indicate the two chains. In red is putrescine (product of the enzymatic
reaction) and piridoxal-5’-phosphate (coenzyme). Bottom yellow aminoacid residue is aspartate
(indicated in A), top yellow is an arginine residue. The aspartate residue makes hydrogen bonds
with the arginine residue and putrescine. Replacement of the aspartate residue with a tyrosine
residue in Odc2 should alter the structure of the catalytic centre.
4 RESULTS slbo
82
to encode a functional ornithine decarboxylase. Odc2 is also lowly expressed
since there are no ESTs for it and the authors of the paper describing it had
difficulties in amplifying it from a cDNA collection (Rom and Kahana 1993). This
suggests that Odc1 is the only functioning ornithine decarboxylase in Drosophila.
4.3.5.3 Odc1 and slbo expression in the fat body
One of the first approachs to study Odc1 was to check its embryonic expression
pattern (Figure 4.34). Odc1 is only detected at the end of embryogenesis, in the
fat body of stage 17 embryos. It has been previously reported that embryonic odc
activity was low during almost all embryogenesis and peaked at the end of
embryogenesis (Birnbaum and Gilbert 1990), which confirms this result. The fat
body is the Drosophila analogous to the liver and the adipose tissue, having an
essential function in metabolism and metabolites storage. This expression
pattern indicates that the fat body is the main organ of polyamines synthesis. It
has also been previously reported that the fat body of adult females is the main
place of odc activity (Birnbaum and Gilbert 1990).
Figure 4.34 – Odc1 expression pattern in stage 17 embryos.
Odc1 wholemount in situ hybridization of stage 17 embryos. Odc1 is expressed in the fat body. It
was not possible to detect Odc1 in earlier embryos. Odc1 was also detected in the fat body in first
instar larvae (not shown). Left image has the anterior to the left and dorsal up. Right image is a
dorsal view with anterior to the left.
stg17stg17
4 RESULTS slbo
83
To access how direct is Odc1 expression dependence on slbo it was important to
check if slbo is also expressed in the fat body at this stage. As mentioned above,
in section 4.1.1, at stage 16 and 17 it is difficult, by in situ hybridization, to
discern where slbo is and is not expressed. To circumvent this problem it was
necessary to use a confocal microscope and fluorescent immuno-staining. For
that a polyclonal rat antibody against Slbo was raised. The specificity of this
antibody was checked by antibody staining of egg chambers (Figure 4.35). The
antibody only recognised the nuclei of border cells and centripetal cells, as
expected, showing that the antibody is very specific.
Figure 4.35 – immunostaining of stage 10 egg chambers with rat anti-Slbo antibody.
Polyclonal rat anti-Slbo antibody of stage 10 egg chamber detected with a secondary fluorescent
antibody. Only the nuclei of border cells (arrow) and centripetal cells (arrow heads) are stained,
showing that the antibody is very specific. Anterior is to the left/bottom corner.
An embryo immunostaining with the rat anti-Slbo antibody and mouse anti-
Serpent, a fat body marker at this stage (Rehorn et al. 1996), was done (Figure
4.36). Slbo is present in many cells of the embryo, including Serpent expressing
cells. This shows that Slbo is expressed in the fat body at the end of
embryogenesis.
4 RESULTS slbo
84
Figure 4.36 – Immunostaining of late embryos with anti-Slbo and anti-Serpent antibodies.
DAPI and antibody staining with polyclonal rat anti-Slbo and polyclonal mouse anti-Srp. Srp at
this stage is expressed in the fat body and hemocytes. DAPI labels all nuclei. Slbo is expressed
in many cells, including Srp expressing cells.
Although there is no proof that Odc1 is directly regulated by slbo, their co-
expression in the fat body at the end of embryogenesis makes it a possibility.
4.3.5.4 Genes involved in ornithine metabolism up-regulated in slbo mutants
Other genes involved in ornithine metabolism are mis-regulated in the slbo
mutant (Figure 4.37). From the microarray data it was possible to see that
Ornithine aminotransferase (Oat), CG5675 (which encodes an ornithine
transporter) and arginase were up-regulated in the mutant. Although only Oat is
up-regulated in all samples, according to the previous criteria (ratio slbo/control
bigger than 1.5), the other two genes also clearly show that tendency.
Importantly note that these genes were picked up in the microarray analysis
before any significance was given to the Odc1 down-regulation. That is, they
were not handpicked a posteriori from a vast list of regulated genes. These three
genes together with Odc1 formed the tighter cluster of mis-regulated genes
involved in a metabolic pathway.
4 RESULTS slbo
85
Arginase catalyses the irreversible conversion of arginine into ornithine. Oat is
the regulatory enzyme of the pathway that converts glutamate or proline into
ornithine and vice-versa. Both arginase and Oat are thought to be important, in
mammals, for the synthesis of ornithine and consequently polyamines (Yu et al.
2003). The three pathways controlled by arginase, Oat and Odc1 are the only
ornithine related metabolic pathways in Drosophila. Ornithine transcarbamoylase
(and Argininosuccinate synthetase) are not present in the Drosophila genome
and there is no urea cycle in Drosophila. Nitrogen is mainly excreted in the form
of uric acid.
Interestingly the three enzyme coding genes are mainly expressed in the fat body
(Figure 4.37) (Samson 2000; Tomancak et al. 2002). Oat activity in third instar
larvae is also reported to be enriched in the fat body (Yoshida et al. 1997). This
suggests that the fat body is the main organ where ornithine related and
polyamines metabolism occurs.
CG1628 is homologous to ORNT1, an ornithine transporter that shuffles ornithine
between mitochondria and cytoplasm. ORNT1 expression regulation is also
thought to be a mean of controlling ornithine metabolism in mammals (Morris
2002). Drosophila Oat is predicted to be a mitochondrial enzyme (Yoshida et al.
1997), as the mammalian homologue is, while Odcs are cytoplasmic proteins.
The mis-regulation of these genes indicates that the ornithine metabolism is
overall disturbed in the slbo mutant. The up-regulation of Oat, arginase and
CG1628 may be a sign of a compensatory mechanism of the organism due to
the lack of Odc activity. As the product of Odc1, polyamines, is reduced the
system tries to increase the substrate of the reaction - ornithine - in order to
increase the reaction rate.
4 RESULTS slbo
86
slbo/control
EST Gene 12-14h I 12-14h II 16-18h I 16-18h II 16-18h III
LD16544 CG1628
(ornithine transporter) 1.05 1.46 1.58 1.40 1.66
GH01984 Oat 1.93 2.60 1.92 2.09 1.87
GH02581 arginase 1.40 2.50 1.12 1.54 1.51
Figure 4.37 – Up-regulation of genes involved in ornithine metabolism in the slbo mutant.
A) Scheme of ornithine related metabolism pathways in Drosophila. Genes up-regulated in the
slbo mutant are indicated with a green arrow. Odc1 is downregulated in the mutant. In situ
hybridizations of Ornithine aminotransferase (Oat) (from Tomancak et al. 2002), arginase
(fromSamson 2000) and Odc1 are shown. All of them are enriched in the embryonic fat body.
B) Ratio of slbo mutant over control embryos expression of ornithine metabolism related genes in
the DNA microarray hybridizations.
4.3.5.5 Interference with ornithine decarboxylase function
An Odc inhibitor was used in order to test if loss of Odc activity could partially
explain the slbo mutant phenotype. Difluoromethyl-ornithine (DFMO) is a specific
irreversible inhibitor of Odc (Weeks et al. 1982). This drug has been previously
administered by feeding to mammals and insects (Weeks et al. 1982; Cayre et al.
1997). In crickets this treatment only partially affects the levels of polyamines
4 RESULTS slbo
87
(Cayre et al. 1997). The larvae treated with DFMO showed a reduced size
comparing to control larvae (Figure 4.38). This indicates that odc activity is
required for normal growth in Drosophila larvae. Thus, loss of Odc1 expression
could partially explain the growth phenotype in slbo mutant larvae.
Figure 4.38 – DMFO treated larvae have a growth defect
One day old wild type larvae were treated with 2% DFMO in PBS or just PBS for 24h. Treated
larvae have grown less than control. Pictures have the same scale.
In order to pursue this line of analysis a transgenic fly which expresses a double
stranded RNA of Odc1 was established. Double-stranded RNA is known to knock
down the expression of endogenous genes by RNA interference in Drosophila
(Kennerdell and Carthew 1998). The construct synthesised was based on the
pWIZ vector (Lee and Carthew 2003). This construct allowed synthesis of an
Odc1 double stranded RNA molecule under the control of the GAL4-UAS system
(Figure 4.39 A). The phenotype of the knockdown of Odc1 was analysed by
expressing the transgene ubiquitously, from the embryo to the adult, with the
tubulin-GAL4 driver.
A primary analysis of several transgenic flies identified only one line where there
was a lethality associated with the transgene expression. Upon further analysis it
was clear that that specific line actually contained two insertions of the
transgene, one on the second chromosome and one on the third. Expression of
Control 2% DFMO
4 RESULTS slbo
88
Genotype Replicate
Tub-GAL4 + I II III Sum
- 29 47 22 98
pWIZ-Odc1B10B 28 37 25 90
pWIZ-Odc1B10BB 32 28 15 75
pWIZ-Odc1B10B pWIZ-Odc1B10BB 9 3 1 13
Figure 4.39 – Analysis of Odc1 down-regulation by RNA interference.
A) Scheme of the pWIZ-Odc1 vector used. One fragments (≈300bp) of the Odc1 is inserted twice
in the vector, with opposing directions. The white gene intron 2 is localized between the two Odc1
fragments. The all construct is expressed under the GAL4/UAS system. Once the construct is
transcribed the intron is spliced out and a double stranded RNA hairpin structure with no loop is
formed. The insertion of the intron sequence facilitates cloning. Based on Lee (2003).
B) Scheme of cross and results of the Odc1 knockdown. The markers of the balancers allowed
following each transgene independently. Tubulin-GAL4 is a ubiquitous driver. The pWIZ – Odc1
transgene is inserted in the second chromosome (B10B) and the third (B10BB). In the table the
number of emerging adults for the different genotypes are shown. Only Tub-GAL4 containing flies
were considered. Values are shown for three replicates and their sum. Each of the genotypes is
theoretically expected to occur with equal frequency. One copy of pWIZ-Odc1 has a small effect
in the flies’ viability however only 15% of the expected flies emerge when two copies are present.
4 RESULTS slbo
89
Figure 4.40 - Analysis of Odc1 down-regulation effect on hatching and adult eclosion.
On top is shown the scheme of the crosses used for this analysis. The markers on the TM3
balancer allow the selection of only Tub-GAL4 containing embryos for analysis. The pWIZ-Odc1
transgene carrying line is homozygous for both insertions which implies that all its progeny will
have one copy of each insertion. The control line used was w118. Three replicates of twenty
embryos were analysed for hatching rate and adult eclosion. Bar height represents mean and
error bars standard deviation. The control presents a lower hatching and eclosion rate than
normal. That is due to the presence of the Tub-GAL4 driver which is know to be, by itself,
deleterious to the fly. pWIZ-Odc1 expression has a strong effect on embryo viability but all the
hatched embryos survived to adulthood.
either of the copies alone had no or a small effect on adult viability while the
simultaneous expression of the two transgenes reduced viability (Figure 4.39 B).
Having to combine several independent insertions in order to obtain a phenotype
in an RNAi experiment is common. The results indicate that Odc1 is required for
normal fly survival.
4 RESULTS slbo
90
To further characterize the phenotype of Odc1 knockdown, embryo hatching
rates were analysed (Figure 4.40). Embryos expressing pWIZ-Odc1 have a much
lower hatching rate than control embryos (20% compared with 65%). This
indicates that Odc1 expression is required for normal embryo development.
All the embryos that hatched survived to adulthood. However, none showed any
clear sign of a growth defect. This would indicate that Odc1 would not be
required for larval growth. The DFMO results indicate otherwise. There can be
several explanations for these discrepancies. The DFMO is used in a relatively
high concentration (2% in PBS) and could have non-specific effects. On the other
hand, most of the times RNAi, in flies, reduces the level of the expression of the
target gene, it does not eliminate it. In this case maybe Odc1 is not being
sufficiently knocked down to show the growth phenotype. I have not confirmed if
Odc1 is in fact being knocked down. With this data is not possible to conclude if
Odc1 is required for normal growth.
4.3.5.6 Odc1 expression and suppression of slbo phenotype
In order to analyse to what extent lack of Odc1 expression is responsible for the
slbo mutant phenotype the Odc1 gene was expressed in the slbo background.
The objective was to see if Odc1 re-expression could rescue the slbo mutant.
Transgenic flies with the Odc1 open reading frame downstream of a UAS
promoter were generated. The driver used to express the transgene was
armadillo-GAL4 (arm-GAL4) a ubiquitous driver with a low level of expression. In
the first assays no clear difference was observed between slbo mutants with or
without ectopic Odc1 expression. These assays were done in the presence of
methyl p-hydroxybenzoate and consequently the main cause of death in both
genetic backgrounds was the drug. Once that factor was eliminated an obvious
observation was that from day 2 on the survival rates of slbo mutants
overexpressing Odc1 are lower than the ones not expressing it (Figure 4.41 A).
At the long term overexpressing Odc1 was deleterious. This is not totally
4 RESULTS slbo
91
surprising since, from what is known in mammals, Odc expression is highly
regulated and in this system it is constantly being expressed in all the tissues.
The arm-GAL4 driver is regarded as a weak driver. However these flies
fluorescence, due to the presence of arm-GAL4 and UAS-GFP, is quite strong.
This means that the arm-GAL4 driver is relatively strong and may be driving
aconsiderable over-expression of Odc1.
Figure 4.41 – Effect of Odc1 ectopic expression in the slbo mutant.
A) Survival rates throughout larval and pupae development of slbo mutant without or with Odc1
ectopic expression. The genotypes analysed were slboe7b,arm-GAL4/ slbo8ex2, UAS-GFP and
slboe7b,arm-GAL4/ slbo8ex2, UAS-GFP; UAS-Odc1/ +. Three replicates of twenty embryos were
scored for each sample. Bars represent standard deviations. First day survival rates are
significantly different (t-test, P=0.023).
B) Larval survival rates on the first day after hatching of slbo mutant without or with Odc1 ectopic
expression. Five replicates of twenty embryos were scored for each sample. Bars indicate
standard deviations. The two samples are significantly different (t-test, P=0.004).
4 RESULTS slbo
92
Odc1 over-expression in the slbo mutant background did not seem suppress the
growth phenotype. However, it is difficult to take any conclusions from this
experiment. The growth defect presents always a lot of variation in the same
sample so a partial suppression would be difficult to detect. There could also be
several causes for the growth defect in the slbo mutant and restoring Odc1
expression may be not sufficient to rescue the phenotype.
Odc1 ectopic expression in slbo mutants induces, however, a small improvement
in 1st day survival (93% survival compared with 87%). It is not a big improvement
but the values are significantly different in a t-test (P=0.023). In order to confirm
this effect the experiments were repeated (Figure 4.41 B). Again there is a small
but significant difference between the two samples (t-test P=0.004). These
results indicate that lack of Odc1 expression in the slbo mutant is partially
responsible for the low early lethality.
4.3.6 slbo mutant and the eIF4E homologue CG8023
The third gene picked in the microarray analysis and that showed complete
dependence on slbo to be expressed was CG8023 (Figure 4.42). The predicted
protein has a high similarity to the mammalian eukaryotic translation initiation
factor 4E (eIF4E) (reviewed inGingras et al. 1999). CG8023 is one of six
Drosophila genes homologues to mammalian eIF4E (Figure 4.43) (Lasko 2000).
eIF4E recognizes and binds to the 5’ terminus structure of mRNAs, the cap.
Through its interaction with eIF4G, eIF4E connects the mRNA with the
translation machinery. One of the main regulatory steps in translation is
translation initiation. The rate-limiting step in translation initiation is normally the
binding of 43S pre-initiation complex to the mRNA and eIF4E is the limiting factor
of this process.
4 RESULTS slbo
93
Figure 4.42 - CG8023 embryonic expression is dependent on slbo.
RT-PCR analysis of control and slbo mutant embryonic samples with CG8023 specific primers.
rp49 primers were used as control.
Cell growth and proliferation require synthesis of new proteins and therefore
require up-regulation of translation. Signaling pathways that promote growth and
proliferation, like the insulin pathway and the TOR pathway, regulate translation
initiation by modulating eIF4E activity (reviewed inGingras et al. 1999; Oldham
and Hafen 2003). A major pathway for this regulation involves eIF4E-binding
protein (4E-BP) phosphorylation. The hypophosphorylated 4E-BP binds eIF4E
and inhibits its binding to eIF4G. Upon the activation of the signaling pathways
4E-BP is phosphorylated, no longer inhibits eIF4E and translation is initiated.
As mentioned above Drosophila has six eIF4E homologues. The phylogenetic
tree of these genes indicates that all Drosophila homologues are equally related
to the mammalian eIF4E (Figure 4.43) (confusingly one of the Drosophila genes
is called just eIF4E). Other organisms, including mammals, have several eIF4E-
like genes. To which extent these are redundant or have the same function is just
starting to be addressed. For example, in zebrafish two proteins very similar to
the human eIF4E and possessing all the residues previously thought to be
4 RESULTS slbo
94
important for its function behaved very differently from each other (Robalino et al.
2004). eIF4E-1A could complement a S. cerevisiae eIF4E mutant, bind the
m7GTP cap of mRNAs and interact with eIF4GI and 4E-BP. eIF4E-1B failed in all
these assays suggesting that it has a different unknown function. In
Schizosaccharomyces pombe one of the eIF4E isoform is a stress-response
factor with low affinity to eIF4G (Ptushkina et al. 2004). The Drosophila genes
are not completely redundant since mutants for Drosophila gene named eIF4E
are lethal and present growth defects (Lachance et al. 2002).
Figure 4.43 - Phylogenetic tree of eIF4Es and Drosophila melanogaster homologues.
Phylogenetic tree done by Neighbour-Joining method using Clustal X. Number in tree branches
indicate bootstrap values (1000 bootstrap trials). Branch size is proportional to phylogenetic
distance.
The relationship between eIF4Es, growth and proliferation made it interesting to
look more closely at the role of CG8023 in the slbo mutant phenotype.
4 RESULTS slbo
95
One characteristic of CG8023 predicted protein is that a conserved serine
residue (ser209 in mouse’s eIF4E) is substituted by a proline residue.
Phosphorylation of this serine residue is thought to be correlated with increased
eIF4E activity and can be mediated by extracellular stimuli (Gingras et al. 1999;
Scheper and Proud 2002). The exact function of this phosphorylation is not clear
but this residue has been shown to be required for the proper function of
Drosophila eIF4E (Lachance et al. 2002). It is not known what is the effect of
replacing this residue by a proline and therefore predict exactly what are the
particular properties of CG8023.
It was not possible to find CG8023 expression pattern by in situ hybridization.
Several attempts were made but no clear pattern was obtained. This probably
indicates that CG8023 is lowly expressed in the embryo. Although the RT-PCR is
not a quantitative technique the same was inferred from its results. The CG8023
band was always weak.
A fly strain with a P-element insertion in the 5’ end of the CG8023 transcript was
available from a public stock (P{w+mC=lacW}l(3)L0139). This fly strain could
potentially be a CG8023 mutant. Homozygous flies with this insertion were not
viable. However, trans-heterozygous flies with this insertion and a chromosome
deletion that uncovered CG8023 were viable. This indicated that the
homozygous lethality of the original stock was not due to a mutation in CG8023
but due to a secondary mutation in the same chromosome. The original stock
was crossed back with the control line w118 in order to produce recombinants
between the original chromosome and the control chromosomes. Twenty
potential recombinant lines carrying the P-element insertion were followed up. In
one line the chromosome with the P-element insertion was homozygous viable
indicating that the secondary mutation was not present. I have further
characterized this clean line.
4 RESULTS slbo
96
The homozygous and trans-heterozygous with the chromosome deletion were
male sterile. The sterility was confirmed to be due to the P-element insertion
since excision of the P-element reversed this phenotype. This sterility is
compatible with the results of two microarray studies that find CG8023 to be
enriched in males (Jin et al. 2001; Arbeitman et al. 2002). However, the P-
element insertion does not downregulate the expression of CG8023 at late
embryonic stages. It was still possible to detect its band in an RT-PCR of P-
element homozygous embryos. The male sterility phenotype is probably due to a
tissue/stage-specific problem in CG8023 expression.
As it does not affect CG8023 expression in the embryo I could not use this
hypomorph mutant to see if embryonic CG8023 loss-of-function relates to slbo
loss-of-function phenotype. It should be possible and relatively easy to make a
complete loss-of-function of this gene by imprecise P-element excision. That was
not done due to time constrains.
Similarly to the Odc1 experiments rescue transgenic flies were generated with
CG8023 downstream of an UAS or heat-shock promoter. The presence of the
heat-shock CG8023 construct was, by itself, deleterious at 25˚C. The transgenic
lines carrying these constructs were generally sick. This indicates that ectopic
expression of CG8023 has deleterious effects.
Again most of the rescue assays were done in the presence of methyl p-
hydroxybenzoate with negative results. The ability of the UAS-CG8023 construct
to rescue the slbo mutant was also studied in the absence of methyl p-
hydroxybenzoate (Figure 4.44). Again the ectopic expression of the transgene
was overall deleterious. I could also not detect a suppression of the growth
defect when CG8023 is over-expressed in the slbo mutant background.
4 RESULTS slbo
97
Figure 4.44 – Effect of CG8023 ectopic expression in the slbo mutant.
Survival rates throughout larval and pupae development of slbo mutant without or with CG8023
ectopic expression. The genotypes analysed were slboe7b,arm-GAL4/ slbo8ex2, UAS-GFP and
slboe7b,arm-GAL4/ slbo8ex2, UAS-GFP; UAS-CG8023/ +. Three replicates of twenty embryos were
scored for each sample. Bars represent standard deviations. First day survival rates are at the
border line of being statistically significant (t-test, P=0.051).
Similar to the Odc1 ectopic expression, the first day survival rates of slbo
mutants with CG8023 ectopic expression were slightly better. However, this
difference is at the border line of being statistically significant (t-test P=0.051). It
would be interesting to repeat this experiment with more replicates. Most
probably lack of CG8023 expression in the slbo mutant is partially responsible for
the low early lethality.
4 RESULTS slbo
98
4.4 slbo and innate immunity
4.4.1 slbo mutants larvae have spontaneous melanization
Raising the larvae in medium without methyl p-hydroxybenzoate allowed the
observation of more and older slbo mutant larvae. One phenotype observed was
that slbo mutant larvae showed spontaneous melanization. Melanization is part
of the innate immune response that Drosophila has against microorganisms (for
reviews on melanization see Ashida and Brey 1997; Söderhäll and Cerenius
1998; Meister and Lagueux 2003). This process is triggered by injury or
recognition of microorganisms and results in the deposition of melanin. This
deposition sequesters and kills the invading microorganisms. The process is the
product of a cascade of serine proteases that culminates in the activation of
prophenoloxidase. Phenoloxidase is an oxireductase that catalyses the
conversion of phenols to quinines, these polymerase and form the melanin. The
product of these reactions may also be toxic to the microorganisms.
When larvae or adults are pricked with a needle, melanization occurs at the injury
site and it is easily observable for its black/brownish colour. slbo mutant
spontaneously (without injury) showed melanization (Figure 4.45 C). Five days
after hatching fifty percent of the surviving larvae showed some melanization and
nine days after hatching all of the surviving larvae showed it. Spontaneous
melanization can also be observed in normal larvae, however, it is much rarer. In
the control experiment, done simultaneously, no melanization was observed. The
degree of melanization in the slbo larvae somehow correlates with the
healthiness of the larvae. The ones that grew faster and eventually pupariated
had none or very little melanization. The ones that had extensive melanization
looked sicker and were more prone to die. It is difficult to assert if melanization
affects the fitness of the organism or vice-versa.
4 RESULTS slbo
99
Figure 4.45 – slbo mutant larvae have high spontaneous melanization
(A) slbo larva showing melanization of right trachea and the anterior left trachea trunk (anterior is
to the left)
(B) Detail of melanized trachea cell (anterior is to the left)
(C) Percentage of larvae alive and showing melanization. Number of larvae reduces with time
due to death and pupariation. From day 9 on all present larvae showed melanization. Three
replicates of twenty embryos were scored for each sample. Bars represent standard deviations.
Control larvae in same experiment showed zero percent melanization at any stage.
The melanization was mainly observed at the level of the trachea (Figure 4.45 A
and B) although not exclusively; it was also observed in other tissues throughout
the body. The tracheae are the respiratory organs of insects. They consist on two
dorsal branches parallel to the body axis. They are hollow and open to the
4 RESULTS slbo
100
exterior through the spiracles at the anterior and the posterior extremities. The air
enters and exits the organism through these openings and is distributed directly
to the tissues through tracheal ramifications that penetrate the body. The
observed melanization of the trachea is much more frequent in the extremities
and seem to evolve from the spiracles to the interior of the trachea.
Because tracheae are in direct contact with the exterior they are a place of entry
for microorganisms. Together with the gut and the epidermis they are considered
barrier epithelia. The trachea (as well as the epidermis and parts of the gut) are
coated with cuticle which constitutes a physical barrier for the microorganisms.
The cuticle is secreted by the epithelial cells. In the cuticle there is also
prophenoloxidase, probably produced by haemocytes (homologous to
mammalian leukocytes) and deposited there (Ashida and Brey 1995). The
trachea can also respond to infection by local synthesis of anti-microbial peptides
(in opposition to a systemic response in which the peptides are produced by the
fat body) (Ferrandon et al. 1998; Tzou et al. 2000).
The spontaneous melanization phenotype suggests that there is an imbalance
related to innate immunity in the slbo mutant.
4.4.2 Innate immunity related genes mis-regulated in the slbo mutant
Many innate immunity related genes are mis-regulated in slbo mutant embryos
(Figure 4.46).
Gasp and CG17052, which encode peritrophin-like proteins, are down-regulated
in the slbo mutant (Figure 4.46 A). Peritrophin-like proteins, with chitin-binding
domains, were originally found as constituents of the peritrophic matrix. This
matrix is constituted of chitin, proteins and proteoglycans (in Barry et al. 1999)
and serves as a physical barrier between the ingested food and the midgut.
Peritrophin-like proteins are not, in a strict sense, part of a specific innate
4 RESULTS slbo
101
immune response to pathogens but are believed to have a protective role against
them. There are seven peritrophin-like encoding genes in the microarray with
these two showing mis-regulation. Gasp is not part of the peritrophic matrix; it is
actually expressed in the tracheae (Barry et al. 1999). The down-regulation of
Gasp expression in the tracheae could be one of the causes of their spontaneous
melanization.
A slbo/control EST Gene 12-14h I 12-14h II 16-18h I 16-18h II 16-18h III LD05259 Gasp 0.50 0.51 0.85 0.48 0.20 LD43683 CG17052 0.89 0.65 0.84 0.53 0.18 LP05763 AttA 1.417 n.d. 1.61 2.60 3.09 SD04493 PGRP-LB 1.32 1.91 1.34 1.79 1.97 GH21008 PGRP-LB 1.37 n.d. 1.49 1.63 1.75 GH07464 PGRP-SC1b 1.72 n.d. 1.62 n.d. 1.35 GH14535 PGRP-LD 0.42 0.29 0.95 0.28 0.42 GH21896 Spn4 1.45 1.02 1.37 1.42 1.83 LP03106 CG1342 0.92 0.38 0.69 0.66 0.42 LP08647 nec 1.06 1.02 0.81 0.58 0.49
B
Figure 4.46 - mis-regulation of genes involved in innate immunity in the slbo mutant. A - Ratio of slbo mutant over control embryos expression of innate immunity related genes in the
DNA microarray hybridizations. Two different ESts in the DNA microarrays correspond to PGRP-
LB. N.d. – not determined because of bad quality of DNA spot.
B – RT-PCR analysis of control and slbo mutant embryonic samples with rp49, Drs and Mtk
specific primers. The RT-PCR was done on polyA purified RNA, of 18-20h embryos. The rp49
reaction is used as a control.
A very brief summary of what is known about Drosophila humoral innate
immunity is required at this point. Most studies focus on the systemic humoral
4 RESULTS slbo
102
innate immunity (as discussed above there is also a local epithelium response).
Upon infection of the fly with microorganisms the fat body synthesizes and
secretes anti-microbial peptides. The activation of this response is done through
two main pathways. The Imd pathway is mainly responsive to Gram-negative
bacteria while the Toll pathway is responsive to Gram-positive bacteria and fungi.
These two pathways differentially induce antimicrobial peptides with different
spectrums of action. Some antimicrobial peptides require or are induced by the
two pathways.
Attacin A (AttA) is up-regulated in the slbo mutant (Figure 4.46 A). AttA is a anti-
microbial peptide mainly active against Gram-negative bacteria (Hultmark et al.
1983) but can be induced by some Gram-positive bacteria, Gram-negative
bacteria and fungi and is dependent on both the Toll and the Imd pathway for full
induction (Irving et al. 2001; De Gregorio et al. 2002). I have also tested, by RT-
PCR, the mis-regulation of several genes encoding anti-microbial peptides
(Drosomycin, Diptericin, Drosocin, Cecropin A, Defensin and Metchnikowin), not
present in the DNA microarrays, in slbo mutant embryos. Drosomycin (Drs) and
Metchnikowin (Mtk) show up-regulation while the others did not show regulation
(Figure 4.46 B). Drs is mainly induced by the Toll pathway (Lemaitre et al. 1996)
while Mtk is induced by both the pathways (Levashina et al. 1998).These results
show that the basal level of expression (without induction) of some anti-microbial
peptides is increased in the slbo mutant embryo.
Three genes encoding members of the peptidoglycan recognition proteins
(PGRP) family, out of five present in the DNA microarrays, are mis-regulated in
the slbo mutant: PGRP-LB, PGRP-SC1b and PGRP-LD (Figure 4.46). The first
member of this family was isolated from the silk moth for its ability to bind
peptidoglycan and to restore the reactivity of the melanization cascade to
peptidoglycan in an in vitro assay (Yoshida et al. 1996). In Drosophila there are
13 members of this family (Werner et al. 2000; De Gregorio et al. 2001). PGRP-
SA is required for the Toll pathway activation by Gram-positive bacteria (Michel
4 RESULTS slbo
103
et al. 2001) and PGRP-LC is involved in the recognition of Gram-negative
bacteria and activation of the Imd pathway (Choe et al. 2002; Gottar et al. 2002;
Ramet et al. 2002). The function of these two Drosophila genes and the above
mentioned silk moth PGRP show that members of this family have a crucial role
in identifying non-self invading microorganisms and activating the appropriate
innate immunity response.
Of the PGRP family members mis-regulated in the slbo mutant only PGRP-SC1B
function has been studied (Mellroth et al. 2003). It is actually an enzyme capable
of degrading peptidoglycans, however, it does not show antibacterial activity. The
authors of the study suggest that PGRP-SC1B has a scavenger function.
Finally, Spn4, CG1342 and necrotic (nec) are also mis-regulated in the slbo
mutant embryo (Figure 4.46). These genes encode proteins of the serpin family,
suicide substrate serine proteases inhibitors. There are a total of nine serpin
encoding genes present in the DNA microarrays. Many serpins regulate serine
proteases cascades that activate immune and immune-related responses in
mammals and insects (some serpins regulate processes not related to immunity).
Spn27A, for example, inhibits the terminal protease of the melanization cascade
(De Gregorio et al. 2002; Ligoxygakis et al. 2002).
Of the serpins mis-regulated in the slbo mutant only nec (also known as
spn43Ac) function has been studied. nec mutants have a constitutive activation
of the Toll pathway (Levashina et al. 1999). Drs and, to some degree, Mtk are
constitutively up-regulated in this mutant. The down-regulation of nec in the slbo
mutant embryos (at late stages) may therefore partially explain the up-regulation
of Drs, Mtk and AttA (which is also partially downstream of the Toll pathway). The
nec mutant also presents spontaneous melanization (Green et al. 2000). Its
down-regulation could also be a cause of the slbo mutant spontaneous
melanization.
4 RESULTS slbo
104
The above results show that in slbo mutants there is altered basal expression of
several innate immunity related genes. This, together with the spontaneous
melanization phenotype, suggests that slbo regulates fly innate immunity.
4 RESULTS slbo
105
4.5 slbo function in larval stages
4.5.1 slbo expression in larvae
slbo expression peaks at the end of embryogenesis and almost disappears
during larval stages (Rørth and Montell 1992). However, throughout this thesis I
have described many larval phenotypes. Moreover, if slbo mutant are raised
without methyl p-hydroxybenzoate there is a small embryonic lethality and most
mutants dye at larval stages. This raises an important question: is slbo required
during larval stages or is the absence of slbo at the end of embryogenesis
responsible for the later lethality?
A first approach to this question was to analyse if slbo was expressed in some
tissues during larval stages, although overall its expression was much lower
when compared with embryonic stages. The antibody staining of third instar
larvae gave a surprising result; although Slbo was not detected in the majority of
larvae, it was present in some, sometimes with strong staining (Figure 4.47). The
number of Slbo positive larvae varied from sample to sample. Overall it was
present in about one larva out of eight. Slbo was mainly detected in the nuclei of
midgut and Malpighian tubules and sometimes in the fat body.
The fact that only a few larvae express Slbo and most do not explains the overall
very low expression previously reported. This result also suggests that slbo
expression in the larvae is regulated. I have tried to access if this was a
developmental regulation; I have stained late third instar wandering and non-
wandering larvae but could get Slbo staining or complete absence of it in both.
Slbo expression in larvae indicates that slbo has a function during larval
development.
4 RESULTS slbo
106
Figure 4.47 – Slbo protein is expressed in larvae.
Antibody staining of w118 third instar larvae with rat anti-Slbo.
(A) Midgut, hindgut and Malpighian tubules; (B) Detail of midgut staining; (C) Detail of Malpighian
tubules staining; (D) Detail of fat body staining.
4.5.2 slbo requirement during larval development
To specifically test if slbo is required during larval development a mechanism to
remove slbo function only after hatching had to be developed. slbo mutants can
be partially rescued by a 15kB genomic fragment containing the slbo gene (Rørth
1994). This fragment was re-engineered so that the slbo transcript (which is just
one exon) was flanked by two FLP recognition target (FRT) sites. The strategy
used was to rescue the slbo mutant with the FRT-slbo-FRT construct and then,
4 RESULTS slbo
107
after hatching, express FLP (from a transgene with FLP downstream of a heat-
shock promoter). The FLP then recombines the two FRT sites and flips-out the
slbo ORF, hence producing a slbo mutant larva.
The first test was to see if the FRT-slbo-FRT transgene rescued slbo loss-of-
function. The FRT-slbo-FRT trangene was working as expected and rescued
36% of the slbo homozygous mutants up to adulthood (Figure 4.48).
Figure 4.48 – Rescue of slbo lethality with the FRT-slbo-FRT trangene
The parent strains for the rescue assay are shown on the top. The three potentially viable
progenies’ genotypes are shown on the bottom. If the FRT-slbo-FRT transgene completely
rescued the slbo loss-of-function than the number of flies of the left genotype should be equal to
one fourth of all slbo heterozygous flies. Based on this, the rescue percentage was calculated
dividing the number of adult flies with the left genotype by one fourth of the number of slbo
heterozygous adult flies. In three different assays the rescue percentage was 26%, 29% and
52%. 0% of homozygous slbo mutant flies without the FRT-slbo-FRT transgene reached
adulthood. Note that the cross was done in normal fly food and the slbo mutants were competing
with their siblings, contrary to other assays previously shown.
The cross to test slbo requirement during larval development are shown in Figure
4.49. From this cross one day old slbo mutant larvae could be selected, by its
fluorescence, and all of them contained one copy of FRT-slbo-FRT. Batches of
30 larvae were selected and placed in a food vial. Half the food vials were heat-
shocked and half of them not. Each batch of larvae contained, on average, 50%
larvae with the hs-FLP transgene and 50% larvae without it. When adults these
4 RESULTS slbo
108
two genotypes could be distinguished by the presence or absence of the FM7
balancer. The objective was to compare the number of hs-FLP; enGal4
slboe7b/UAS-GFP slbo8ex2; FRT-slbo-FRT adults obtained from a non-heat-
shocked vial (where the FRT-slbo-FRT trangene was kept throughout
development) and an heat-shocked vial (where the slbo ORF of the transgene
was lost one day after hatching).
Figure 4.49 – requirement of slbo during larval development (I)
The cross to test slbo requirement during larval development is shown on the top. slbo mutants
are selected by the enGal4 UAS-GFP system. All the progeny will carry one copy of the FRT-
slbo-FRT transgene. Half the progeny will carry the hs-FLP transgene, half the FM7 balancer.
Upon heat-shock the hs-FLP is expressed flipping out the FRT-slbo-FRT transgene and thus
generating a slbo null. At the time of selection larvae with the FM7 chromosome or the hs-FLP
are undistinguishable, only as adults the FM7 balancer’s marker allows the distinction. Each
replicate contained 30 larvae picked up one day after hatching. Half the vials were heat-shocked
at 37˚C for one hour, the other half not. On the table the number of adults of each genotype in
heat-shocked and non heat-shocked vials is shown. 6 replicates were done for each condition.
The 6 replicates were done in three different days; each line corresponds to one day. The mean
and standard deviation of each class is shown. The heat-sock reduces only the viability of the
flies carrying the hs-FLP transgene.
4 RESULTS slbo
109
The FM7; enGal4 slboe7b/UAS-GFP slbo8ex2; FRT-slbo-FRT flies were an internal
control for the effect of the heat-shock per se. The number of adult flies with this
genotype did not change much with or without heat-shock indicating that by itself
the heat-shock did not affect these larvae (Figure 4.49).
The average number of hs-FLP; enGal4 slboe7b/UAS-GFP slbo8ex2; FRT-slbo-
FRT adult flies was bigger in the non-heat-shocked vials than in the heat-shock
vials (Figure 4.49). The loss of slbo gene in the heat-shocked adult flies was
confirmed by genomic PCR. This indicates that slbo is required for normal larval
development. However, the standard deviations on this assay are very big and it
is not possible to make a statistical analysis. There are several sources for this
high noise in the results that can be dealt with. The 6 assays were done in three
different days which contributes to the variability of slbo mutants’ survival rates
(see section 4.3.1). The second source of noise is that although the number of
hs-FLP; enGal4 slboe7b/UAS-GFP slbo8ex2; FRT-slbo-FRT larvae in each batch
should be on average 15 there is some variation associated with this too.
The same experiment was also done with four simultaneous replicates of
samples only containing larvae of the genotype hs-FLP; enGal4 slboe7b/UAS-
GFP slbo8ex2; FRT-slbo-FRT (Figure 4.50). This setup does not have the two
sources of variation discussed above. 49% of the non heat-shocked larvae reach
adulthood while only 29% of the heat-shocked ones do. The difference is
statistically significant. A statistically significant difference already exists at the
level of larvae that pupariate. This is noteworthy because the slbo could be only
required after pupariation and not in larval stages. These results show that lack
of slbo function during larval life affects larval survival.
4 RESULTS slbo
110
Figure 4.50 - requirement of slbo during larval development (II).
On the top it is indicated the cross to generate hs-FLP; enGal4 slboe7b/UAS-GFP slbo8ex2; FRT-
slbo-FRT larvae for analysis. The experimental procedure was the same as for Figure 4.49 with
the exception that 4 replicates of 20 embryos each were done simultaneously. The graph shows
the percentage of larvae that pupariated and reached adulthood in heat-shocked and non-heat-
shocked vials. Bar height represents mean and error bars standard deviation. With t-test analysis
the number of pupae of the two samples are different with P=0.031 and the number of adults of
the two samples are different with P=0.017.
5 DISCUSSION slbo
111
5 DISCUSSION
The purpose of this work was to understand the slbo function in embryogenesis
and, later on, in larval stages. There were two reasons for this interest: a) the
slbo loss-of-function is lethal which clearly indicates that slbo has an essential
function; b) The function of the mammalian homologues is related to metabolism
regulation and immunity, two topics I thought interesting and relevant to study in
Drosophila.
5.1 DNA microarrays of control vs slbo mutant embryos
Part of the strategy was to compare expression profiles of slbo mutant embryos
with control embryos. This was a logical approach since Slbo is a transcription
factor and its function necessarily its function necessarily involves transcription
regulation. This approach required the setup of Drosophila microarrays and the
fine-tuning of the Drosophila embryos sorter. These setups were technically
challenging and required a fair amount of time.
The DNA results of the DNA microarrays, comparing slbo mutant with control
embryos, had two “problems”. One was that the dependence on slbo for the
expression of some genes changed with time. In the space of fourty days, from
the first sample collection to the last, the expression of some genes stopped to
be dependent on slbo. I could not understand why this happened. However, the
important factor to consider here is that the slbo mutation was always a lethal
mutation. That means that in all samples analysed the core problem of the
mutants remained; the transcription of some essential genes was always
misregulated. Therefore, the results obtained were valid to analyse slbo function
in the embryo.
5 DISCUSSION slbo
112
The second “problem” was that a low regulation was observed. Importantly, this
low regulation was not a technical problem of the microarrays. However it makes
analysing and understanding the results difficult. For one hand, low regulation
implies a high noise to signal ratio. To present a reliable “list of genes
misregulated in the slbo mutant” a bigger number of replicates than I have would
be required.
On the other hand the importance of a partial regulation is more difficult to
access and understand than all or none situations. This does not mean that it has
no biological significance. Slbo could be responsible for only regulating the level
of expression its targets, which is the general case with mammalian C/EBPs. It
could also be a problem of using mRNA from whole embryos. This procedure
could mask tissue specific complete dependence on slbo.
Although three genes which expression is completely dependent on slbo were
identified, most of the genes seem to be only partially regulated. A partial
reduction in the expression of a particular important gene or the accumulation of
many small down-regulations in a particular process or pathway may translate
into severe consequences.
Another aspect to consider is that the expression profiles show direct and indirect
effects of slbo loss-of-function. With this data it is not possible to distinguish
between them. A possible approach would have been to identify genes rapidly
up-regulated when slbo is over-expressed in embryos. Nonetheless, the direct
and indirect effects of slbo loss-of-function are all useful to understand slbo
function.
I was interested in relating the information of the expression profiles with the
phenotype of the slbo phenotype. A better analysis of the mutant phenotype was
possible once a method to isolate slbo mutant was established. First, it was
5 DISCUSSION slbo
113
possible to observe that some larvae survived a few days. Secondly, these
larvae presented themselves more revealing phenotypes.
5.2 slbo mutants and methyl p-hydroxybenzoate sensitivity The early high lethality of the slbo mutant was shown to be due to the presence
of methyl p-hydroxybenzoate in the fly food. This substance is rapidly
metabolized and eliminated in animals. In another insect methyl p-
hydroxybenzoate is mainly metabolized by glucosilation (Rojas et al. 1990).
Interestingly, one of the three genes whose expression is completely dependent
on slbo is an UDP-glycosyltransferase (DmGlAT-BSII). Its phylogenetic tree
analysis and in vitro data indicate that it is probably involved in protein
modifications and not detoxification (Kim et al. 2003). The in vitro assay also
showed that it lacked complete specificity which leaves the possibility that its
down-regulation and slbo methyl p-hydroxybenzoate could be related.
Unfortunately, the slbo sensitivity to methyl p-hydroxybenzoate was only found at
the end of the PhD research. The role of DmGlAT-BSII in this phenotype could
have been tested by expressing it in the slbo mutant background to see if it can
suppress the sensitivity. It would be also possible to make an excision mutant
from an existing Drosophila line with a P-element upstream of the gene. The
mutant would then be analysed for methyl p-hydroxybenzoate sensitivity.
Even if DmGlAT-BSII is not the UGT responsible for methyl p-hydroxybenzoate
glycosylation it would be worth to find if such one exists. The first step would be
to test if methyl p-hydroxybenzoate is glycosylated and, if so, if it ceases to be
glycosylated in the slbo mutant. If the answer to both questions would be yes
then all the UGTs, a big but limited number, could be screened.
It is not known the identity of any UGT responsible for methyl p-hydroxybenzoate
glycosylation. There would interest in its identification mainly because of the vast
5 DISCUSSION slbo
114
use of methyl p-hydroxybenzoate in products for human consumption. Esters of
p-hydroxybenzoic acid also have oestrogenic activity in human cells and mice
(Darbre et al. 2003) and there has been proposed a link between it and breast
cancer (Darbre 2003).
The sensitivity to methyl p-hydroxybenzoate of the slbo mutant seems to be
developmentally regulated. The massive death caused by this drug is restricted
to the first 24 hours after hatching. Larvae that survive this critical period dye,
afterwards, at lower rates. It has been reported that the UDP-glucosyltransferase
activity changes with Drosophila development (Rausell et al. 1997). The
hypothetic glycosilation of methyl p-hydroxybenzoate could be carried out, on
older larvae, by a different UGT, independently of slbo. This could be tested by
transfering one day old slbo mutant larvae from drug-free to drug-containing
plates and see if they were resistant to drug.
To find an UGT involved in detoxification dependent on slbo would be particularly
interesting. Mammalian C/EBPα regulates the expression of UGTs in the liver, an
important organ in detoxification (Lee et al. 1997; Hansen et al. 1998). Induction
of UGTs involved in detoxification could be an ancestral function of the C/EBP
family.
5.3 Growth defect and altering feeding behaviour
slbo mutant larvae present a growth defect and an altered feeding behaviour.
Both phenotypes, although milder, are still present in the absence of methyl p-
hydroxybenzoate. The phenotypes indicate that slbo mutants have a general
metabolism defect. The altered feeding behaviour phenotype is very specific and
has only been previously reported in two cases (Zinke et al. 1999; Britton et al.
2002). In both of them it was suggested that a high nutritional level was sensed
by the larvae. The microarray data showing that, in general, sugar transporters
5 DISCUSSION slbo
115
are up-regulated in the slbo mutant support the hypothesis that the same is
happening in this mutant.
In the slbo mutant the balance of synthesis and degradation of carbon and
energy reserve metabolites, like glycogen and triacylglycerols, could be altered.
Maybe in the slbo mutants there is less accumulation of these molecules and a
higher concentration of small metabolites. This hypothesis can be tested by
measuring the several metabolites concentrations.
The down-regulation of sugarbabe in the slbo mutant could be part of the
explanation. Sugarbabe seems to be a transcription repressor that inhibits lipid
catabolism genes (Zinke et al. 2002). Its down-regulation could lead to an up-
regulation of lipid catabolism genes which would decrease lipids content and
increase small metabolites concentrations.
There is a similar example in mice that have no white adipose tissue and
reduced brown adipose tissue (Moitra et al. 1998). These mice were created by
interfering with bZIP transcription factors, including C/EBPs, in adipocytes
percursors. The disruption of energy storage (and communication between the
white adipose tissue and other organs) leads to several defects including high
levels of blood glucose and insulin.
C/EBPα loss-of-function mutants main reason of death is hypoglycemia (Wang et
al. 1995). This exactly the opposite of what I am suggesting is happening in the
slbo mutant. However, one of the main reasons C/EBPα are hypoglycemic is
because they lack glycogen synthase expression at late gestation and do not
accumulate glucose reserves in the form of glycogen. So, in C/EBPα mutants
there is, among others, a problem of energy reserve metabolites synthesis.
In the slbo versus control DNA microarrays the glycogen synthase spot did not
give a clear result (slight up and down-regulation in different arrays). This could
5 DISCUSSION slbo
116
be due to technical reasons and it may be worth to analyse its regulation by
Northern blot. The Drosophila PEPCK, another gene involved in energy
metabolism and a target of C/EBPα, is not present in the microarray and its
regulation could also be analysed by Northern blot.
In Drosophila over-activation of the insulin pathway is one of the causes of the
altered feeding behaviour (Britton et al. 2002). An early attempt was done to
check if the insulin pathway was over-activated in the slbo mutants. The
experiment was based on expressing the pleckstrin homology domain of GRP1
fused to GFP (Britton et al. 2002) in control and slbo mutant larvae. This fusion
protein is recruited to the cell membrane when the insulin pathway is activated.
The purpose was to check if there was more membrane associated GFP signal
in the slbo mutant than in control larvae which would indicate that the insulin
pathway is more active in the slbo mutant. The technique is difficult and better
done in 2nd and 3rd instar larvae. The fact that almost all larvae died very small,
due the methyl p-hydroxybenzoate, made it impossible to practically do the
experiment. This experiment can now be repeated. Other more direct genetic
experiments, like tinkering with mutants in the insulin pathway in slbo mutant
background and checking its behaviour, can also be done.
5.4 slbo and polyamines metabolism
Ornithine related metabolism pathways are mis-regulated in the slbo mutant.
Odc1 expression is absent while Oat, CG5675 (which encodes an ornithine
transporter) and arginase were up-regulated in the mutant. The three enzyme
encoding genes are enriched in the fat body at the end of embryogenesis. They
probably act as a unit for the synthesis of polyamines. Odc1 initiates polyamines
synthesis from ornithine. arginase makes the substrate of this reaction, ornithine,
from arginine. Oat in theory could regulate the glutamate/proline to ornithine
pathway as well as the inverse reactions. However, most probably, the pathway
works in the direction of ornithine synthesis from glutamine. Drosophila is
5 DISCUSSION slbo
117
auxotrophic for arginine while it can synthesize glutamate (Borror and DeLong
1971). It would not make sense to synthesize ornithine from the essential
aminoacid arginine to produce glutamine, a non-essential aminoacid. The
concentration of polyamines synthesis in one insect organ, the fat body,
contrasts with the mammalian situation where polyamines can be synthesized in
many different tissues (Schipper and Verhofstad 2002; Yu et al. 2003).
Odc2 lacks many aminoacids absolutely conserved in all other Odcs, which
indicates that this gene does not encode a functional Odc. Lack of Odc1
expression most probably results in low polyamines levels. The up-regulation of
Oat, CG5675 and arginase is, in my perspective, a mechanism to increase
ornithine concentration, in order to drive polyamines synthesis.
Another important question was: how did the lack of Odc1 expression contributed
to the slbo phenotype? The straight forward experiment was to over-express
Odc1 in the slbo mutant background and see if it rescued any of the defects. At
the long term over-expressing Odc1 is deleterious, which is not completely
surprising. In mammals Odc expression and activity are tightly regulated
(seeTabor and Tabor 1984; Igarashi and Kashiwagi 2000; Thomas and Thomas
2003). At the short term there is a small but statistically significant decrease in
lethality.
A small decrease in lethality, in the first day after hatching, may look insignificant.
However, two points must be considered. First, since long-term over-expression
of Odc1 is deleterious, it is not possible to know what would be the long-term
effect of reinstating normal Odc1 expression. Second, and more importantly, slbo
lethality is most likely multi-factorial. In my results I have shown that it is at least
due to methyl p-hydroxybenzoate sensitivity and lack of Odc1 expression and
most probably due to lack of CG8023 expression too. Many other genes are also
down-regulated in the slbo mutant and that also likely contributes to its lethality. It
5 DISCUSSION slbo
118
would be very difficult to completely rescue the slbo mutant lethality by restoring
the expression of just one gene.
The Odc1 result identifies one of the genes whose down-regulation is a cause of
slbo lethality. Regardless of the rescue assay, the data indicate that the
ubiquitous biochemical pathway of polyamines synthesis, essential in eukaryotes
for growth and cell proliferation, is controlled at the end of embryogenesis by
slbo.
Unfortunately there is no data in the literature, to my knowledge, indicating if
mammalian C/EBPs control or not Odc expression. However polyamines
synthesis seems to be regulated differently in Drosophila and mammals. As
discussed above, while in Drosophila appears to be centrally produced in the fat
body, in mammals it is produced in several tissues.
In C/EBPα-deficient mice genes of the ornithine cycle, including arginase, are
down-regulated (Kimura et al. 1998). In the Drosophila slbo mutant arginase is
up-regulated. This difference may be explained by the fact that in mammals the
ornithine cycle in the liver is essential for ammonia excretion while in Drosophila
it does not exist. So arginase regulation in mammals is in a different context than
in Drosophila.
Besides the lethality, the growth defect phenotype could also be due to the lack
of Odc1 expression. The data showed conflicting results. On one hand the Odc
specific inhibitor DFMO induces a growth defect. On the other hand Odc1 RNAi
ubiquitious expression and slbo RNAi expression in the fat body do not cause a
growth phenotype. There can be several explanations for these discrepancies.
The DFMO is used in a relatively high concentration (2% in PBS) and could have
non-specific effects. The Odc1 RNAi may not be down-regulating Odc1
expression sufficiently, in the relevant tissue. The bottom line is that the
contribution of Odc1 down-regulation for the growth defect is not clear.
5 DISCUSSION slbo
119
One factor to also have in mind is that in all these assays larvae were fed with
live yeast, a very rich diet, from which they could assimilate polyamines and
partially compensate for their block in the polyamine pathway. It would be
interesting to repeat some of these assays in synthetic medium without
polyamines.
5.5 slbo mutant and the eIF4E homologue CG8023 Another gene whose expression is completely dependent on slbo is CG8023.
This is a homologue of the translation initiation factor eIF4E. However it is not
possible to predict exactly what is its function. Over-expression of CG8023 in the
slbo mutant background seems to partially suppress the lethality phenotype.
More assays would have to be done to have a significant statistical value.
I have briefly examined a potential mutant in this gene. However, if a mutant, it is
a hypomorph and does not affect embryonic expression. A CG8023 loss-of-
function mutant can be generated by an excision screen of the P-element
insertion line P{w+mC=lacW}l(3)L0139. It would be interesting to analyse its
phenotype and relation with the slbo mutant phenotype.
It is possible that there may be a link between Odc1 and CG8023. Mammalian
Odc expression is regulate at many levels, one important level is translation
initiation. Odc translation is dependent and correlates with eIF4E activity (Shantz
and Pegg 1999). CG8023 expression could be required for proper Odc1
expression.
5 DISCUSSION slbo
120
5.6 slbo and innate immunity
An interesting phenotype of the slbo mutant is a very high rate of spontaneous
melanization; 50% of five days old larvae showed it. Normally melanization is
only trigged upon infection or injury, as an innate immunity defence mechanism.
The process is regulated by a cascade of serine proteases. The last step of the
melanization cascade is inhibited by Spn27A and requires the Toll pathway
activation (De Gregorio et al. 2002; Ligoxygakis et al. 2002). The spontaneous
melanization phenotype indicates that the regulation of this process is abnormal
in the slbo mutant.
The main tissue showing spontaneous melanization was the trachea. One gene
expressed in the trachea and down-regulated in the slbo mutant is Gasp. This
gene encodes a peritrophin-like protein and could have a protective role against
microorganisms invading the trachea.
The trachea is a barrier epithelium in direct contact with the exterior and prone to
infection. Using a GFP reporter gene under the control of the drosomycin (an
anti-microbial peptide) promoter Ferrandon et al. (1998) report some
spontaneous expression of this gene in the absence of experimental challenge.
The spontaneous expression is especially detected in the trachea and most
frequently in the anterior and posterior spiracles. This is exactly the pattern
observed of spontaneous melanization in the slbo mutant. Ferrandon et al.
(1998) show that this is due to natural infection since animals reared in axenic
conditions have almost no reporter gene expression.
The slbo mutant melanization could be explained by two similar but diametrical
opposed hypotheses; the slbo mutant innate immunity response could be
hampered and consequently a natural infection progresses to a point that triggers
melanization or the slbo mutant could be hypersensitive to microorganisms and
low levels of infection trigger the melanization. The pattern of melanization would
5 DISCUSSION slbo
121
be due to the fact that the entrance for microorganisms is at the extremities of the
trachea. A third hypothesis would be that the structure of the tracheae is deficient
and leads to physical damage and melanization (as sterile pricking induces
melanization). The entries of the tracheae are also more exposed to external
conditions and probably more sensitive. An easy way to distinguish between the
first two and the third hypothesis would be to raise slbo mutant larvae in axenic
conditions and check if the spontaneous melanization is present or absent. The
first two hypothesis could be distinguished by infecting slbo mutant larvae and
analyse their immune response.
A series of genes involved in innate immunity or related to genes involved in
innate immunity were mis-regulated in the slbo mutant. Among other genes,
three encoding anti-microbial peptides are up-regulated. One of them, Drs, is
specifically downstream of the Toll pathway, the other two, AttA and Mtk, are also
regulated by this pathway (Lemaitre et al. 1996; Levashina et al. 1998; Irving et
al. 2001; De Gregorio et al. 2002). An up-regulation of the Toll pathway could
explain these results. It could also, partially explain, the spontaneous
melanization. Interestingly nec is down-regulated in the slbo mutant and this
could lead to the hypothetical up-regulation of the Toll pathway; nec is a serpin
that inhibits the Toll pathway (Levashina et al. 1999).
It is important to note that these results concern non-challenged embryos and are
basically descriptive of an imbalanced basal status. It would have been very
interesting to see if slbo mutant have a defective immune response when
challenged with pathogens.
Other cues point to a role of slbo in innate immunity. In mammals the acute
phase response involves three transcription factors: NF-κB, Stat and C/EBPs
(Poli 1998). NF-κB Drosophila homologues have been known, for some time, to
be important for the innate immune response activation (Ip et al. 1993; Dushay et
al. 1996). Recently it has been reported that Drosophila Stat has a role in the
5 DISCUSSION slbo
122
response to septic injury (Boutros et al. 2002; Agaisse et al. 2003). The activation
of the JAK/STAT pathway, in border cells (non-related to immunity), induces slbo
expression (Silver and Montell 2001; Beccari et al. 2002). In the mammalian
acute phase response the JAK/STAT pathway also activates the expression of
C/EBP genes that then maintain the induced state of acute phase genes. It is
tempting to speculate that slbo would also be up-regulated and required for the
Drosophila innate immune response.
κB-RE and IL6-RE motifs were found in the analysis of the promoter of diptericin,
which encodes an anti-microbial peptide (Georgel et al. 1993). The κB-RE motif
is the binding site for the NF-κB homologues and is required for tissue-specific
induction of the gene upon immune challenge (Meister et al. 1994). IL6-RE motif
is present in many mammalian acute phase genes and is a C/EBP binding motif
(C/EBPβ is also called NF-IL6) (Poli 1998). In the diptericin promoter it is
required not for the induction of the gene upon immune challenge but for a higher
level of induction (Meister et al. 1994). Again it is tempting to speculate that slbo
could be required for proper levels of diptericin induction. In my results diptericin
basal expression was not dependent on slbo. However, it is possible that the up-
regulation upon induction is dependent on slbo.
This is a clear case where it was unfortunate to find the methyl p-
hydroxybenzoate toxicity to slbo only at the end of the PhD research. Drosophila
innate immunity is a flourishing field and many genetic tools and protocols are
accessible. With 3rd instar larvae and adult slbo mutants available it would have
been relatively easy to start analysing this problem.
5.7 Other functions of slbo
Many other genes not discussed in detail are down-regulated in the slbo mutant.
Some like CG9747, predicted to be an acyl-CoA δ11 desaturase, and CG7149, a
diacylglycerol cholinephosphotransferase, are enzymes involved in different
5 DISCUSSION slbo
123
aspects of lipids metabolism. The role of slbo in metabolism regulation is
probably broader than the discussed here. A transcription factor related to
C/EBPs, but less conserved, has also been shown to be involved in lipid
metabolism in Caenorhabditis elegans (Bai et al. 2000)
slbo and many genes dependent on slbo are expressed in the epidermis and
other epithelial tissues. slbo could also have a role in epithelial cell differentiation
analogous to C/EBPs role in keranocyte differentiation. More specifically slbo
could be involved in cuticle formation. Cuticle is synthesized at the end of
embryogenesis in the epidermis, trachea, anterior gut and posterior gut. All
places of expression of slbo at the right development stage. Moreover, one of the
confirmed genes down-regulated in the mutant, CG8505, is a cuticle protein
encoding gene. However, the cuticle pattern is not affected in the slbo mutant
(my own results and Rørth and Montell 1992). This particular role of slbo could
be verified by a more detailed analysis of the cuticle structure.
The function of slbo is probably very broad. It could have many different functions
in the different tissues where it is expressed. An example of this is its function
during border cell migration which seems completely unrelated to the functions
discussed in this thesis.
5.8 slbo function in larval stages
Although slbo expression in larvae has been described as almost absent in larval
stages (Rørth and Montell 1992), it can be detected, in some larvae, in the nuclei
of midgut, Malpighian tubules and fat body. This expression in the larvae has a
functional role since losing slbo function at larval stages is deleterious for
survival.
In a given sample only a small fraction of the larvae have Slbo expression, most
of them do not. This data suggests that Slbo expression is somehow regulated
5 DISCUSSION slbo
124
during larval stages. It does not seem to be developmentally regulated which
raises the hypothesis that it could be regulated by other cues.
Screening for conditions that up-regulate slbo expression in larvae would be an
interesting approach to understand slbo function in Drosophila. One possibility is
that it is related with the nutrition of the larvae. The altered feeding behaviour in
the slbo mutant larvae points in that direction. The down-regulation, in the slbo
mutant, of sugarbabe, a gene strongly induced by sugar feeding conditions
(Zinke et al. 2002), also suggests that. Interestingly sugarbabe, in larvae, is
expressed in the exact same tissues as slbo (Zinke et al. 2002).
Another possibility would be that slbo expression is induced upon an immune
challenge. This would fit with the suggestion that slbo has a role in the innate
immune response.
I have only observed Slbo expression in a fraction of larvae and only a fraction of
larvae that loose the slbo gene die. It is possible that these two results are
related and slbo is only expressed and required, during larval stages, in a fraction
of the larvae.
Throughout the thesis I have tried to relate the DNA microarray data with the slbo
phenotypes. An underlying problem in this is that the expression profiles concern
late embryos and the phenotypes are, mainly, of larvae. It would be important to
know to each extent the mis-regulation seen in the slbo mutant embryos is
present in mutant larvae. However, the tissues at the end of embryogenesis,
when slbo is expressed, are basically the larval tissues so the context of the end
of embryogenesis is very similar to the context of larval development.
5 DISCUSSION slbo
125
5.9 Concluding remarks
A problem in a gene expression profiles analysis is to relate the list of genes
differentially regulated with the problem under analysis. I have tried to relate the
DNA microarray data with the phenotypes of the slbo mutant. Admittedly, many
of the relationships are speculative and there is a lack of functional data on how
the mis-regulation of a particular gene is directly connected to a particular
phenotype of the mutant. The best connection is the partial rescue of the slbo
mutant lethality by over-expressing Odc1, and possibly CG8023. However, the
analysis allowed me to draw some conclusions on what the function of slbo is.
The DNA microarrays data and larval phenotypes indicate that slbo has a role in
Drosophila metabolism. The methyl p-hydroxybenzoate sensitivity could be
related with the drug metabolism and excretion. The altered feeding behaviour
clearly suggests that slbo mutants have a general metabolic problem,
corroborated by some of the mis-regulated genes. The mis-regulation of ornithine
metabolism genes is the best evidence of one pathway affected in the mutant.
The data also suggests that slbo is involved in innate immunity. Several innate
immune related genes are mis-regulated in slbo mutant, though, this was
analysed only in non immune challenged animals. Nonetheless, it shows a basal
imbalance of the system. One of the results of this imbalance is, most probably,
the excessive melanization in larval stages. The other result could be an
ineffective or disproportioned response to infection. The organism has to find a
fine balance where the immune response is normally off but a pathogen is rapidly
detected and specifically dealt with.
The suggested involvement of slbo in metabolism regulation and innate immunity
shows a conservation of function between Drosophila and mammalian C/EBPs. It
is also interesting the fact that slbo function partially relates to fat body function.
Odc1 and other ornithine metabolism related genes are specifically expressed in
5 DISCUSSION slbo
126
the fat body. The fat body is involved in metabolism regulation and innate
immunity in Drosophila. This insect organ is analogous to the liver and adipose
tissue, main organs of C/EBPs expression and function in mammals. The liver, in
particular, is a central organ in metabolism regulation at the organism level and
the place of expression of acute phase genes. This indicates that C/EBPs, in
both organisms, operate in tissues that have a central role in metabolism and
innate immunity, at the level of the whole organism, even if these two tissues are
not homologous structures.
Interestingly it is also common in both organisms that C/EBPs are expressed in
late embryogenesis, in differentiating cells. In both systems C/EBPs probably
activate genes required for the transition from embryonic development to
independent life. In mammals C/EBPα is required just before birth for
accumulation, in the form of glycogen, of maternally provided glucose. These
reserves are crucial for survival in the first hours after birth. In Drosophila,
embryo development is supported by the extra embryonic yolk. At the end of
embryogenesis larval tissues, like the gut and the fat body, have to take over the
metabolic functions of the yolk. slbo expression in these tissues could be
essential for the establishment of these functions.
The physiology and the tissues with a central role in metabolism and
inflammatory response are different between mammals and insects. The
particular genes expressed in these tissues and with a role in these processes
will be in some cases different between the two organisms. However, C/EBP
core function in regulating genes important for the organism homeostasis seems
to be conserved.
Lsd2
127
Lipid metabolism regulation by Lsd2
6 INTRODUTION Lsd2
128
6 INTRODUTION Lipids are a major form of energy storage in animals. They are stored in the
intracellular neutral lipid droplets of specialized tissues such as adipose tissue in
mammals and the fat body in insects. Although initially found in fat-related
tissues, lipid droplets are organelles present in many, if not all, cell types
(reviewed in Murphy 2001). Hence understanding the role of lipid metabolism at
the level of the cell and whole organisms requires identification of the molecular
mechanisms governing the biogenesis, trafficking and turnover of lipid droplets.
Lipid droplets are formed by a unique monolayer of amphipatic phospholipids
surrounding a central hydrophobic core of neutral lipids, mainly consisting of
triacylglycerol (TAG) and sterol esters. Two mammalian proteins have been
studied for their property to specifically localize at the surface of these
organelles: Perilipin and ADRP (adipocyte differentiation-related protein also
known as adipophilin) (Greenberg et al. 1991; Blanchette-Mackie et al. 1995;
Brasaemle et al. 1997). Besides this particular property, ADRP and Perilipin also
show sequence similarity, especially in their N-terminal region where they are
~40% identical (Lu et al. 2001). This N-terminal region, also present in another
mammalian protein - TIP47, has been termed PAT domain (Lu et al. 2001).
TIP47 was originally identified as a protein required for the transport of mannose
6-phosphate receptors from endosomes to the trans-Golgi network (Diaz and
Pfeffer 1998). Although initially controversial (Barbero et al. 2001; Wolins et al.
2001), the association of TIP47 to lipid droplets was recently verified (Miura et al.
2002). The presence of a PAT domain correlates with the ability of proteins to
localize to lipid droplets, although it has been recently shown not to be absolutely
required (Garcia et al. 2003; McManaman et al. 2003; Targett-Adams et al.
2003). Proteins with a PAT domain have been found in a wide variety of species,
including Drosophila, and together form the PAT family (Lu et al. 2001).
6 INTRODUTION Lsd2
129
Perilipin and ADRP were originally identified as genes highly expressed in
adipose tissue (Greenberg et al. 1991; Jiang and Serrero 1992). Further studies
showed that Perilipin expression is restricted to differentiated adipocytes and
steroidogenic cells (Greenberg et al. 1993; Servetnick et al. 1995), whereas
ADRP is more ubiquitously expressed (Brasaemle et al. 1997; Heid et al. 1998).
In cultured cells, it has been shown that the ectopic expression of Perilipin or
ADRP increases the capacity of cells to take up long fatty-acids from the medium
and to accumulate neutral lipids (Gao and Serrero 1999; Brasaemle et al. 2000;
Imamura et al. 2002; Souza et al. 2002). Reciprocally, the addition of fatty acids
to the culture medium stimulates neutral lipid accumulation in cells and increases
intracellular levels of Perilipin or ADRP (Brasaemle et al. 1997; Gao et al. 2000;
Souza et al. 2002). The reciprocal regulation of perilipin/ADRP and neutral lipids
suggests that these two proteins have a role in lipid metabolism regulation.
The function of Perilipin in vivo has been analyzed in Perilipin-deficient mice
(Martinez-Botas et al. 2000; Tansey et al. 2001). These mice are viable, fertile
and have normal size and weight. However, they have a reduced adipose tissue
mass and are more muscular than controls. They are resistant to induced obesity
and have a higher metabolic rate. These phenotypes are explained by the
observed increase in basal lipolysis activity. This analysis together with data
from cell culture experiments (Souza et al. 1998; Brasaemle et al. 2000; Souza et
al. 2002) led to the proposal that Perilipin has a protective role against lipases.
The viability of Perilipin mutant mice could be explained by a partial
compensation by other mammalian PAT-members. The identification of the in
vivo role of the other members of this family waits the generation of mutants.
Two Drosophila melanogaster members of the PAT family, Lsd1 and Lsd2, have
been identified by BLAST search (Lu et al. 2001). Both are equally related to any
of the three mammalian members of the PAT family (Miura et al. 2002).
Ectopically expressed GFP-tagged Lsd1 or Lsd2 localize to lipid droplets, both in
mammalian cell culture and in the fat body of Drosophila (Miura et al. 2002). This
6 INTRODUTION Lsd2
130
shows that the targeting to lipid droplets is a feature conserved between
Drosophila and mammalian PAT family members.
Nathalie Vanzo, working in Anne Ephrussi’s laboratory, and I have collaborated
on a project concerning Lsd2. The objective of this work was to find if this
Drosophila PAT-family member had a conserved function in lipid metabolism.
7 RESULTS Lsd2
131
7 RESULTS
7.1 Isolation of a Drosophila Lsd2 mutant
Nathalie Vanzo generated an Lsd2 mutant by imprecise P-element excision
(Teixeira et al. 2003). The Lsd21 mutant has a small deletion of ~500bp
immediately upstream of the gene transcription start. This deletion affects Lsd2
expression and Lsd21 is a protein null mutant. Nathalie verified this by Western
blot, using a polyclonal antiserum raised against Lsd2. The Lsd2 protein is
present in wild type flies but absent in Lsd21 homozygous mutants.
7.2 Drosophila Lsd2 pattern of expression As a first approach towards the investigation of Drosophila Lsd2, I examined its
pattern of expression during embryo development by in situ hybridization (Figure
7.1, panels A-D). A high level of uniformly distributed Lsd2 transcript was found
in the early stages of embryogenesis (Figure 7.1 A). The mRNA present at these
stages is provided maternally (Edgar and Schubiger 1986). Upon cellularization
of the embryo, at stage 5, Lsd2 mRNA disappears except in the pole cells, the
germline precursors at the posterior pole of the embryo (Figure 7.1 B). Later, at
stage 11, zygotic expression begins in the amnioserosa (data not shown). At
stage 14, Lsd2 is relatively broadly expressed with an enrichment in the fat body
and the midgut (Figure 7.1 C). At the end of embryogenesis, these tissues,
together with the hindgut, are the main sites of Lsd2 expression (Figure 7.1 D).
The enrichment of Lsd2 in the fat body is particularly interesting because it is the
main organ for lipid storage in insects (Canavoso et al. 2001).
The enrichment of Lsd2 mRNA in the fat body at the end of embryogenesis
prompted me to determine if Lsd2 is also expressed in larval fat body. The larval
fat body is a site of active synthesis and storage of lipids while the larva rapidly
7 RESULTS Lsd2
132
Figure 7.1 - Expression pattern of Lsd2 during embryogenesis and in 3rd instar larvae.
(A-D) Whole-mount in situ hybridization of wild type embryos with digoxygenin-labeled antisense
Lsd2 probe. Embryos are oriented with anterior to the left. A-C are lateral views with dorsal up, D
is a dorsal view. The pole cells (pc), amnioserosa (as), fat body (fb), midgut (mg) and hindgut (hg)
are indicated. In C the dashed line delimits fat body and midgut staining, small arrows indicate
putative haemocytes.
(E-F) Immunostaining of wild-type (E) and Lsd21 homozygous (F) 3rd instar larvae with the rat
anti-Lsd2 serum. The fat body is indicated with arrowheads. The cuticle is non-specifically stained
in both wild-type and mutant larvae.
7 RESULTS Lsd2
133
grows and prepares for metamorphosis. I performed immunodetection of Lsd2
protein in wild type and Lsd21 homozygous 3rd instar larvae, using the Lsd2
antiserum (Figure 7.1, panels E and F, respectively). At this stage, Lsd2 is mainly
found in the fat body of wild-type larvae. Mutant larvae present no staining in this
tissue, showing the specificity of the detection. The enrichment of Lsd2 protein in
the larval fat body supports a role for this protein in lipid metabolism.
Because intense deposition of lipids is known to occur during oogenesis in the
female germline (Mahowald and Kambysellis 1980), the expression of Lsd2 in
this tissue was investigated by Nathalie Vanzo (Teixeira et al. 2003). Whole-
mount in situ hybridisation on ovaries showed that Lsd2 mRNA is detected from
mid-oogenesis on (stages 7/8), where it accumulates in the oocyte . From stage
10 on, Lsd2 mRNA expression was greatly increased throughout the germline,
with strong cytoplasmic staining in both the nurse cells and the oocyte. This
accumulation is consistent with the previous detection of a high level of Lsd2
mRNA during the first stages of embryogenesis. The distribution of Lsd2 protein,
in ovaries, was examined by immunofluorescence. The temporal expression
pattern of the protein was similar to the mRNA. Lsd2 was detected in the
cytoplasm of the germline. However, Lsd2 was detected to a lower level in the
oocyte despite the abundance of mRNA.
The subcellular localization of Lsd2 was analysed by electronic microscopy by
Catherine Rabouille, in collaboration with us (Teixeira et al. 2003). Endogenous
Lsd2 was detected at the surface of neutral lipid droplets. This shows that the
subcellular localization of PAT-family members is conserved between mammals
and Drosophila.
The expression of Lsd2 in the female germline and its localization to neutral lipid
droplets prompted us to further examine the process of lipid accumulation in the
germline. To visualize neutral lipids we stained wild type ovaries with Nile red, a
fluorescent probe known to label neutral lipids (Brown et al. 1992). Neutral lipids
7 RESULTS Lsd2
134
could be detected at a low level in both nurse cells and in the oocyte from stages
7/8 (Figure 7.2 A). Whereas evenly dispersed in the oocyte, neutral lipids are
enriched in a network surrounding the nuclei in the nurse cells. This pattern of
accumulation in nurse cells is similar to the organization of the endoplasmic
reticulum (ER) at these stages (Bobinnec et al. 2003). Indeed, in a co-detection
with a marker of the ER lumen, neutral lipids appear distributed similarly to the
ER network (Figure 7.2 A), although they do not perfectly co-localize.
At stage 10, a higher level of punctuate and evenly-distributed lipid droplets was
visible in the cytoplasm of nurse cells and, to a lower extent, in the oocyte (Figure
7.2 B). At the onset of stage 11, the cytoplasmic content of the nurse cells is
progressively delivered into the oocyte through a process called dumping,
causing massive growth of the oocyte and resulting in higher fluorescent Nile red
staining of the ooplasm compared to previous stages. At the end of stage 12,
dumping is complete and nurse cells have transferred their cytoplasm into the
oocyte. At this stage, traces of neutral lipid droplets were still visible, surrounding
the nucleus, of the apoptotic nurse cells. At stage 14, the intense fluorescence
visible throughout the oocyte cytoplasm revealed the high content of lipid
droplets deposited at the end of oogenesis in the mature egg. Taken together,
these results show that Lsd2 expression coincides with the accumulation of
neutral lipid droplets during oogenesis
7.3 Abnormal accumulation of neutral lipids in the germline and eggs of Lsd21 females
Lsd21 mutant nurse cells revealed a different pattern of neutral lipid accumulation from stage 10 on (Figure 7.2 C). In contrast to the punctuated distribution in wild
type, prominent patches of brightly stained neutral lipids were detected in the
cytoplasm of mutant nurse cells. They often distribute in a radial pattern
surrounding the nucleus of the nurse cells (see insets in Figure 7.2 C). During
stages 11/12, neutral lipids aggregate in enlarged lipid structures, sequestered in
7 RESULTS Lsd2
135
Figure 7.2 - Neutral lipid accumulation during oogenesis.
(A) Early stage of neutral lipid accumulation (left panel) and ER distribution (middle panel) in a
wild-type stage 7/8 egg chamber. The overlay between the two signals is shown in the right
panel. (B-D) Nile red staining of wild type (B), homozygous Lsd21 (C) and Lsd21/P{SUPor-
P}Lsd2KG00149 (D) egg chambers. Insets in (B) and (C) are enlarged views of the region inside the
broken line, except the lower inset in C, which is from another egg chamber. Arrowhead in panel
B indicates the degenerating nurse cells.
7 RESULTS Lsd2
136
the apoptotic nurse cells. At the end of oogenesis, these structures persist,
confined near the respiratory appendages at the antero-dorsal side of the oocyte.
This genetic analysis shows that Lsd2 is required for normal neutral lipid droplet
accumulation in the nurse cells.
Figure 7.3 - TAG quantification in early embryos laid by wild type and Lsd21 homozygous females.
Five independent collections of 0-1 hour embryos were analyzed for each mother genotype
indicated in the figure. TAG level in each sample was normalized for protein concentration. Wild
type is used as the reference value. Bar heights correspond to the average values, error bars
indicate standard deviations. Using a pairwise Wilcoxon test, early embryos from Lsd21 females
have a significantly different TAG level from wild type (P<0.02).
While the Lsd21 mutant was generated, an independent P-insertion called
P{SUPor-P}Lsd2KG00149 was isolated in the Lsd2 5’UTR (generated by the
Drosophila gene disruption Project, 2001). Genetic complementation analysis
between Lsd21 and P{SUPor-P}Lsd2KG00149 revealed a defect in the pattern of
7 RESULTS Lsd2
137
neutral lipid accumulation similar to Lsd21 homozygous nurse cells at stage 10
(compare Figure 7.2 C and D). This demonstrates that Lsd21 and P{SUPor-
P}Lsd2KG00149 mutations are allelic and the phenotype is specific for the Lsd2
gene.
Surprisingly, the aberrant pattern of neutral lipid accumulation observed in nurse
cells was not seen in the oocyte. In addition, despite the obvious retention of
neutral lipids in the nurse cells, the oocyte accumulates lipid droplets, as
revealed by the increase of its fluorescence from stage 10 to 14. We concluded
that Lsd2 is not strictly required for neutral lipid droplet accumulation in the
oocyte. However, to test whether the aberrant accumulation of lipids in the
mutant nurse cells could impair normal deposition into the oocyte, I have
quantified neutral lipids in early embryos (0-1 hour) (Figure 7.3). 50% less TAG is
detected in embryos from mutant mothers compared to wild type, demonstrating
that Lsd2 is required for normal deposition of neutral lipids in the oocyte.
7.4 Embryos laid by Lsd21 females have a reduced hatching rate
I have observed that embryos in the Lsd21 homozygous stock have a
significantly lower hatching rate than those of a wild type control stock (Figure
7.4, compare nearly ~95% in control (a) and ~63% in mutant stocks (c)). This
defect was also seen in the progeny of homozygous mutant females crossed with
wild-type males (Figure 7.4 (d)). This shows that the hatching defect was not
suppressed by providing a wild type Lsd2 copy from males and indicated that it is
dependent on the genotype of the mother, rather than of the zygote. Consistent
with this, no defect was observed in the progeny of Lsd21 heterozygous females
crossed with hemizygous mutant males in spite of the fact that half of the
progeny was mutant (Figure 7.4 (b)). This demonstrates that Lsd2 is a maternal
effect gene. To further investigate the hatching defect, I collected embryos from
wild-type and Lsd21 homozygous mothers and examined their development at
7 RESULTS Lsd2
138
Figure 7.4 Hatching rate defects in the progeny of Lsd21 homozygous females.
Hatching rates were determined in the progeny of w118 parents (a), Lsd21/ FM7C KrGAL4 UAS-
GFP females crossed with Lsd21/Y males (b), Lsd21/ Lsd21 females crossed with Lsd21/Y males
(c), Lsd21/ Lsd21 females crossed with w118 males (d). Eggs were collected for 90 minutes and
the number of hatched eggs counted after 36 hours. Six independent samples of 100 embryos
were analyzed for each cross. Bar heights are average values, error bars indicate standard
deviations. Using pairwise Wilcoxon tests corrected for multiple tests (Holm 1979) (c) or (d) have
a significantly different hatching rate from (a) or (b) (P<0.05).
two time points. No significant difference was visible between the two populations
after ~7 hours of development, most embryos being at stage 11. However, after
~21 hours, just before hatching, there were clear differences between the wild
type and the embryos of Lsd21 mothers. In the wild-type population, 99% of
embryos presented an elongated larva-like morphology (late stage 17). In
contrast, 78% of embryos from Lsd21 mothers ranged from a stage 17 embryo-
shaped morphology to the wild type elongated larva-like morphology. Moreover,
the remaining 22% of embryos from Lsd21 mothers appeared to have
7 RESULTS Lsd2
139
degenerated. Thus, the loss of viability of ~37% in the progeny of Lsd21 mothers
results from developmental defects occurring after stage 11.
Figure 7.5 - Lsd2 is required for normal larval fat body and TAG accumulation in adults.
(A) 3rd instar Lsd21 homozygous larvae have less developed fat body than wild type. Lsd21 and
wild-type larvae were fed under similar conditions for 4 days after hatching. Note the difference in
opacity.
(B) TAG quantification in wild type and Lsd21 3 days old adult males. Five independent samples
of 8 males were analyzed for each genotype. TAG level in each sample was normalized for
protein concentration. Wild type is used as the reference value. Bar heights correspond to
average values, error bars indicate standard deviations. Using a pairwise Wilcoxon test, Lsd21
adult males have a significantly different TAG level from wild type (P<0.02).
7.5 Lsd21 adults exhibit impaired lipid storage
A further examination of the Lsd21 homozygous stock revealed that whereas
larvae develop normally in a rich diet, they were less opaque than wild type
larvae (Figure 7.5A). This seemed to be due to the fact that the fat body of these
larvae, easy to visualize because of the transparency of the body wall, is less
developed than that of control larvae. To test whether the lipid storage function of
the fat body is impaired in the mutant, I quantified the TAG content in adult flies
(Figure 7.5 B). The level of TAG was 27% lower in the mutant than in the wild
7 RESULTS Lsd2
140
type. This demonstrates that Lsd2, similarly to Perilipin in the mouse, is required
for normal storage of lipids in the fly.
8 DISCUSSION Lsd2
141
8 Discussion
We showed that Lsd2 is predominantly expressed in the fat body and in the
female germline of Drosophila melanogaster. These two tissues are engaged in
high levels of lipid metabolism. The former is the principal site of lipid storage and
mobilization (Canavoso et al. 2001). The latter is the site of neutral lipid
deposition during oogenesis, for later use during embryogenesis (Mahowald and
Kambysellis 1980). A GFP-tagged version of Lsd2 was previously reported to
localize to lipid storage droplets when expressed in Drosophila (Miura et al.
2002). In our collaboration with Catherine Rabouille we showed that the
endogenous Lsd2 localizes to the surface of lipid droplets in the germ-line of
Drosophila females. This sub-cellular localization and its concentration in lipid-
enriched tissues support a function of Lsd2 in the lipid metabolism of the fly.
Confirming this, we described phenotypes related to defects in this process in
both the fat body and in the germline of the Lsd21 mutant.
The observation that, in 3rd instar larvae, the main organ of Lsd2 expression, the
fat body, is less developed in the mutant than in the wild type suggested that
there was a problem in lipid storage capacity. Lsd2 mutant adults present a
reduced TAG content. An independent study on Lsd2 has also reached the same
conclusion (Gronke et al. 2003). The Lsd2 phenotype is reminiscent of the
phenotype observed in Perilipin-deficient mice (Martinez-Botas et al. 2000;
Tansey et al. 2001). These mice are also viable but are leaner than the wild type,
due to a defect in neutral lipid accumulation in adipocyte tissues. This similarity of
phenotype supports a conservation of function in lipid storage between
mammalian PAT-family members and Drosophila Lsd2. Also, in both organisms,
Lsd2 and perilipin (or ADRP) are predominantly expressed in tissues involved in
fat storage.
Besides the defect in fat storage in adults, there is an aberrant pattern of neutral
lipid accumulation in the nurse cells of Lsd2 mutant ovaries at stage 10. From this
8 DISCUSSION Lsd2
142
stage on, most neutral lipids do not appear as evenly dispersed discrete droplets,
as they do in wild type, but form larger aggregates often radially distributed
around the nurse cell nuclei. This is the first time that a PAT-family member has
been described to affect the structure of lipid droplets. This pattern is similar to
the distribution of neutral lipids and ER at earlier stages (compare insets in
Figure 7.2 C and Figure 7.2 A). According to the model in which lipid droplets
form from the ER (Murphy 2001), these large aggregates could result from the
inefficient formation of lipid droplets and accumulation of neutral lipids in the ER.
However, we do not observe complete co-localization of neutral lipids and an ER
lumen marker either in the wild-type at early stages (Figure 7.2 A) or in the
mutant at later stages, when the ER marker is diffusely distributed throughout the
cell (data not shown). This could be due to the localization of neutral lipids and
the ER marker in different sub-domains of the ER. However, it cannot be exclude
that the aberrant lipid structures might be not a part of the ER but localize to a
different compartment. Further analysis of this phenotype may further light on the
molecular function of Lsd2.
A defect in lipid droplet formation in nurse cells would be expected to interfere
with lipid droplet deposition into the oocyte. Indeed, the neutral lipid level is
reduced in early embryos of Lsd2 mutant females. There is also a reduced
hatching rate in the progeny of these females, revealing that Lsd21 has a
maternal effect. One hypothesis is that maternal Lsd2 is required, during
oogenesis, to ensure optimal deposition in the mature egg of the neutral lipid
pool required later on for proper embryogenesis. In such a model, the lipid
reserve would be consumed before the end of embryogenesis, in some of Lsd2
mutant progeny, and would result in a development arrest and hatching defect.
Another possibility is that maternal expression of Lsd2 might be required in the
embryo to guarantee the proper use of the lipid pool during its development.
In mammals, the accumulation of Perilipin or ADRP and formation of neutral
lipids were shown to occur concomitantly and to be mutually stimulatory. The
8 DISCUSSION Lsd2
143
analysis in the female germline also showed that Lsd2 and neutral lipid droplet
accumulation coincide after mid-oogenesis. In addition, the lack of Lsd2 affects
neutral lipid accumulation in the adult fly. It would be interesting to investigate if
lipid levels also affect Lsd2 expression. This could be addressed in mutants for
genes involved in lipid metabolism, such as midway, in which the enzyme
converting diacylglycerol into TAG is inactivated (Buszczak et al. 2002).
Although there are defects on lipid content in the Lsd2 mutant there is still 73%
of the normal content and the flies are viable. Maybe there is some redundancy
in function between Lsd2 and Lsd1 and the latter can sustain a sufficient level of
neutral lipid metabolism for survival.
In the absence of Lsd2 there are defects consistent with a role in lipid droplet
biogenesis and storage. This work provides new evidence for a function of Lsd2
in Drosophila melanogaster lipid metabolism and supports the conservation of
function between PAT-family members from invertebrates to vertebrates.
9 EXPERIMENTAL PROCEDURES
144
9 Experimental procedures
9.1 Fly husbandry
Flies were grown on standard corn meal molasso agar (for 1l of water: 12g agar,
18g dry yeast, 10g soja powder, 22g molasso, 80g malt extract, 80g corn
powder, 6.2ml propionic acid, 2.4g methyl p-hydroxybenzoate). All crosses,
embryo collections and larvae growth experiments were done at 25˚C, unless
otherwise indicated.
Staged embryos were collected in apple juice agar plates (for 1l: 750ml water,
250ml apple juice, 22.5g agar, 25g sucrose, 1.5g methyl p-hydroxybenzoate)
with fresh yeast. Adults were kept in cages and at least two plates were changed
before start collecting, in order to optimize synchronization of the embryos.
Embryo survival rates and observation of larval development were done in apple
juice agar plates with fresh yeast. For feeding behaviour experiments fresh yeast
was dyed with bromophenol blue.
For heat-shock experiments vials were submerged in a 37˚C water bath for one
hour.
9.2 Fly embryo sorting
When small numbers of embryos were required they were sorted by hand under
a fluorescent dissection scope.
When large numbers of embryos were required they were sorted using the
COPAS SELECT system from Union Biometrica, modified by Anne Atzenberg
and I. Embryos were dechorionated in sodium hypochlorite (50% solution of
9 EXPERIMENTAL PROCEDURES
145
original 6-14% stock solution). Embryos were then placed in a solution of 1x
PBS, 0.8% Triton-X and sorted. This solution was also used as sheet fluid during
the sorting. Sorting was done with at least 95% purity, at a speed of 15-25
embryos/second.
The staged embryos’ collection for RNA extraction for the DNA microarrays
analysis where sorted with a standardised timing: 30 minutes for collecting
embryos from plates, dechorionation and start of sorting; 60 minutes of sorting;
30 minutes for washes and homogenization in Trizol. A small sample of embryos,
from each sorted embryo collection, was put apart and fixed. This sample was
used to verify by in situ hybridization the staging and purity of the collection. slbo
mutant and control embryos were treated equally throughout the all process.
9.3 Drosophila stocks used The following table lists the fly stocks used:
Allele Comment Source slboe7b slbo loss-of-function Pernille Rørth
slbo8ex2 slbo loss-of-function Pernille Rørth
slbory7 slbo hypomorph Pernille Rørth
slbory8 slbo hypomorph Pernille Rørth
Slbo1310 slbo hypomorph Pernille Rørth
P{SUPor-P}Lsd2KG00149 Lsd2 loss-of-function mutant Bloomington
enGAL4 Embryonic GAL4 driver Steven Cohen
Tub-GAL4 Ubiquitous GAL4 driver Bloomington
arm-GAL4 Ubiquitous GAL4 driver Steven Cohen
UAS-GFP Inducible GFP Bloomington
GFP-Pdi ER marker Alain Debec
hs-FLP Heat shock induced Flipase Bloomington
TM3 Twist-GAL4 UAS-GFP TM3 fluorescent balancer Bloomington
The transgenes generated by me are described in DNA constructs.
9 EXPERIMENTAL PROCEDURES
146
9.4 Generation of transgenic flies
Transgenic flies were generated by co-injecting a P-element transformation
vector with the transgene together with a helper transposon that encodes
transposase but is transposition-defective (∆2-3) into w118 embryos. 6µg of
transformation vector and 1µg of ∆2-3 were ressuspendend in 20 µl of injection
buffer (5mM KCl, 0.05mM K2HPO4, 0.05mM KH2PO4, pH 6.8). The injections
were done by Ann Mari Voie as described (Voie and Cohen 1998).
Transformants were detected on the second generation by their orange eye
colour. The insertions were mapped to a chromosome and balanced.
9.5 DNA constructs The preparation of plasmid DNA, purification of PCR products and isolation of
DNA fragments from agarose gels was done with Qiagen kits according to the
manufacturer’s protocols. DH5α competent cells (GibcoBRL) were used for
transformation of plasmid DNA.
UAS-CG8023 was done by cloning a CG8023 cDNA EcoRI/XhoI fragment from
the GH4024 EST into pUAST.
UAS-Odc1 was done by cloning an Odc1 cDNA EcoRI/XhoI fragment from the
GH13851 EST (incomplete XhoI digestion) into pUAST.
The FRT-slbo-FRT construct was done in several steps. An Asp718/Asp718
FRT-lacZ-FRT fragment was cut from a nanos-FRT-lacZ-FRT-G4VP16 vector
(from Pernille Rørth) and inserted into the pBluescript vector with the BamH1 and
SalI sites inactivated. Using BamH1 and SalI, the lacZ was removed form this
vector and replaced by slbo ORF, from pBS-slbo (from Pernille Rørth). The
9 EXPERIMENTAL PROCEDURES
147
Asp718/Asp718 FRT-slbo-FRT fragment was cut from this construct and inserted
in the Asp718 site in pw15bknx (Rørth 1994).
The pWIZ-Odc1 construct was done according to Lee (2003). A 300bp fragment
was amplified by PCR, using GH13851 as template and the primers
AAAATCTAGACTCACTCAGTTGAGATCTTTCC and AAAATCTAGAGACATCG
AAGCCCAAGACC. PCR was done in a 50µl reaction containing 0.5µl of 5 U/µl
AmpliTaq (Roche), 5µl of 10x supplied buffer, 1µl of 10µM DNTP mix and 2.5µl of
each primer at 10µM. This fragment was inserted in the pWIZ vector twice, in
opposing directions and with the white gene intron 2 in between.
9.6 Construction of DNA microarrays
The DNA microarrays used are based on cDNAs. The cDNAs collection was
from a public source, Drosophila Gene Collection 1 (DGC 1) (Rubin et al. 2000),
and is constituted of ESTs from genes expressed in embryos, larvae, pupae and
adult heads and ovaries. It is composed of approximately 6000 non-redundant
cDNAs, covering about 40% of the Drosophila predicted genes. To these genes
cDNAs collected from EMBL Drosophila laboratories were added.
The cDNAs were amplified by PCR, purified and spotted according to (Richter et
al2002). Universal primers for the vectors used in the cDNA collection, pOT and
pBS, were used to amplify cDNAs. The PCR reactions and product purifications
were done by Belén Miñana, in Vladimir Benes laboratory.
The purified PCR products were dehydrated by speed-vacuum, re-suspended in
2X SSC and aliquoted. All the reactions and handling were done in microtitter
plates and with the help of a liquid phase handling robot. The spotting was done
on aminosilane-coated glass slides. Each slide contains two replicate spots, in
different locations, for each EST. This allowed normalisation of the values and
9 EXPERIMENTAL PROCEDURES
148
safeguarding against imperfections in an area of the slide. If one area of the
microarray was lost there was still information in the replicated area.
9.7 RNA extraction Embryos were homogenized in TRIzol reagent (GibcoBRL) with a pestle and a
27G needle in a plastic syringe. Samples were either immediately processed or
kept at -80˚C.
Samples were processed according to TRIzol reagent (GibcoBRL) instructions.
RNA was extracted, twice, with chloroform, precipitated with isopropanol and re-
suspended in water.
9.8 RNA labelling and DNA microarray hybridization
On the preliminary DNA microarrays hybridizations a polyA RNA amplification
protocol was used. This was a modified Eberwine protocol as described in
(Dimopoulos et al 2002). This amplification step is linear and is based on a
reverse transcription reaction where the first-strand primer has a T7 promoter
fused to a polyT sequence (GGCCAGTGAATTGTAATACGACTCACTATAGGGA
GGCGG(T)24). The polyT sequence served as the primer for the reverse
transcription done with a Superscript kit (GibcoBRL). After the double stranded
cDNA was produced, RNA was in vitro transcribed with a T7 MEGAscript kit
(Ambion).
PolyA purification was done with Dynabeads (Dynal), according to their protocol
(however beads should not be re-used since it inhibits the posterior cDNA
labelling). On average 2µg of polyA RNA was isolated from 100µg of total RNA.
9 EXPERIMENTAL PROCEDURES
149
DNA microarrays based on cDNAs are simultaneously probed with the two
samples to be compared. The two RNA samples were independently labelled.
dUTP-Cy3 or dUTP-Cy5 was incorporated in a first-strand reverse-transcription
reaction using a Superscript kit (GibcoBRL). After removal of unincorporated
dNTPs with a Qiagen PCR purification kit (Qiagen), the probes were combined,
lyophilized, and resuspended in hybridization buffer containing 50% formamide,
6× SSC, 0.5% SDS, 5 × Denhardt's reagent, and 0.5 mg/ml poly(A) DNA. Arrays
were prehybridized in 6 × SSC, 0.5% SDS, and 1% BSA at 42°C for 90 min,
hybridized overnight at 42°C in humidified hybridization chambers, washed twice
in 0.1 × SSC, 0.1% SDS (30 min), twice in 0.1 × SSC (15 min), rinsed with de-
ionized H2O, and dried.
9.9 DNA microarrays data acquisition and analysis
The slides were scanned on a GenePix 4000B Microarray Scanner (Axon
Instruments). The intensities values for each spot were extracted using GenePix
3.0. The median of the pixels intensity was the value chosen for the signal. The
median is a statistical value robust to variation at the extremes of the sample.
The median of the fluorescence of the area immediately outside the spot circle
was taken as the background.
The data was then further analysed using GeneSpring (Silicon Genetics). The
background values were deducted from the signal. Data was normalized using
an intensity-dependent normalization (Lowess). The Lowess normalization
assumes that overall most genes have the same expression levels in the two
samples, which is the case in these experiments.
9 EXPERIMENTAL PROCEDURES
150
9.10 RT-PCR
First-strand reverse-transcription reaction was done from a polyT primer, using
the Superscript kit (GibcoBRL), according to the manufacturer’s protocols. The
product of this reaction serves as a template for the following PCR reaction. 50µl
PCR reaction mixes were prepared with 1µl of first-strand reverse-transcription
reaction, 1µl of 5 U/µl AmpliTaq (Roche), 5µl of 10x supplied buffer, 1µl of 10µM
DNTP mix, 3µl of DMSO and 2.5µl of each primer at 10µM. In each reaction
there were primers for the rp49 gene, as a control, and primers for the gene
being analysed. The 50µl mix was divided in three samples that were amplified
22, 25 and 28 cycles. The objective of doing this is to increase the chances of
stopping the reaction in the exponential phase. A touch-down PCR programme
was used in which the extension reaction temperature drops gradually from 68ºC
to 54ºC in ten cycles and stays at 54ºC the rest of the cycles.
For each gene, primers were designed to amplify a fragment of approximately
600bp, except for rp49 and some anti-microbial peptides encoding genes. The
primers for rp49 amplify a 300bp fragment, distinguishable of the 600bp fragment
of the gene being analysed. Some of anti-microbial peptides encoding genes are
very small and their primers were designed to amplify the biggest possible
fragment. The sequence of the primers used for RT-PCR is listed on the
following table:
Name Sequence rp49-f TCCTACCAGCTTCAAGATGAC rp49-r CACGTTGTGCACCAGGAACT slbo-f AATGCTTAACATGGAGTCGCC slbo-r GTCTGCTGCTGCAGGAC CG9747-f ATGGGTCACCTGTCCACC CG9747-r CTTGTAGCTCTTGTGGGTCC ect-f GAATCGAAAGGCGGAAGAGC ect-r GTTCGGAAAATTATGCGAGCC impE2-f AAGCCCGTTGCCTTGATCC impE2-r AGAATGTTCTGAGCAGGTTCC CG6347-f ACTTTGTGACTGCCAGGAATC CG6347-r CTACAGAATGGGGTAGCTGC
9 EXPERIMENTAL PROCEDURES
151
CG11395-f GCCAAAATTCTGGAGGATTCC CG11395-r AAGAGACTGCGCTGATTTACC Spn5-f TCGAACTACTCTCTCGTGTCC Spn5-r AAGCCCGCCATCTTTTACTGC CG8023-f ACTTTGTGGCACTTGGAGAAC CG8023-r TCTCGAAAAGCATTGGTGCGC CG7675-f CCGCTGACATTGTGAAGACC CG7675-r TGGGGCGTAAGCTTGACAATC ken-f TGTCGCACATCAGCCTAAACC Ken-r AATATGAACGAGCGGAGTCGC Cyp28d1-f CTGACCTCGCACACAATGAC Cyp28d1-r AACAAAATCCAGCCAAACGCC CG15096-f GCCATGTCCTTTGTATTCTC CG15096-r AATGCGAGAATGCATTGCACC CG10872-f ATCATTGGTCTGGGATTCTCC CG10872-r AAAACACATCGCCGAACACCC CG17930-f CATGACCATCAATCGAGTAC CG17930-r TGCCTGTCTTAGCGTCAATCC CG17052-f AAGTTCTACGTGTGCCTGAAC CG17052-r TGGGAAGTGAGTTTGGGTGTC CG9057-f CCTCCTTGAATACTACTTCCC CG9057-r GTGGTAGGTAGAGGGGTATC Odc1-f AGTTTCCACGTCGGCTCC Odc1-r TATAGCTTGGAAGTACAGGGTC GlcAT-P-f GCCCTGCTCTAATGGGAC GlcAT-P-r GAAGTTCCTTGCTATTAGCCC CG7149-f GTTCTCTTCATCTTTTGGGGC CG7149-r ACGACAGTCATGAGAAAGCC CG8505-f TGCAGTGTGCTGAGGACC CG8505-r GAGGCGAAGGTTACAGCC CG7080-f TGCAGCACCAGTCACAGC CG7080-r GGGTATCTGGAGAGGACC CG10657-f CACGTTCACATCGTCGCC CG10655-r GAGAAGATGACCTGACGTCC CG16721-f CAGGTGCTTAAGCTGGCC CG16721-r GTTATCACAGACGCTCCAGC Tsp96F-f CCATCCAAAGACCCAGCC Tsp96F-r ACAAAGTGTGACGTTCAGGC CG15281-f GAACGCTTTCGACAGTGTAAAC CG15281-r CGCTATTGTAAATAGTAATTTCTGTAC CG8677-f GAGCGGAGTAAACAGGAGC CG8677-r GGGTTAATGATCACTGCTCC CG4060-f TCCGCGTGTCAGGATTCC CG4060-r CTAGATCTGGATGTGACGAGC Drs-f AGCTCCGTGAGAACCTTTTCC Drs-r AATATGTGTAAGTAGTGGAGAGC Dpt-f TCAATTGAGAACAACTGAGATGC Dpt-r TATAATAGCTAGACTCGGATACC Drosocin-f ATGAAGTTCACCATCGTTTTCC Drosocin-r GCAGAATGGGTGGTGGC Cecropin A-f AACTTCTACAACATCTTCGTTTTC Cecropin A1-r TCAACCTCGGGCAGTTGC Cecropin A2-r AGATAGTCATCGTGGTTAACC
9 EXPERIMENTAL PROCEDURES
152
Defensin-f TGAAGTTCTTCGTTCTCGTGGC Defensin-r TCAATTGCGGCAAACGCAGAC Mtk-f GCCACCGAGCTAAGATGC Mtk-r TATCAATGTGTTAACGACATCAGC
9.11 RNA wholemount in situ hybridization
Anti-sense probes were labelled with dig-UTP by in vitro transcription from
cDNAs, using MEGAscript kits (Ambion). Reaction was done according to
manufacturer’s protocol with a mix of 2µl each of ATP, CTP, GTP (Megascript),
1,5µl of 75mM UTP (Megascript) and 4µl dig-UTP 10mM (Boehringer stock).
Embryos were dechorionated in sodium hypochlorite (50% solution of original 6-
14% stock solution). Embryos were fixed in 50% Heptane, 50% (4%
paraformaldehyde in PBS) for 30 minutes and devittelinized with
heptane/methanol. Embryos were stored in methanol, at –20°C.
Embryos were re-fixed in 4% paraformaldehyde in PBS for 15min and pre-
hybridized in hybridization solution (50% Formamide, 5x SSC, 0.1% Tween-20)
with salmon sperm DNA. Hybridization was done overnight at 65ºC with probe
diluted in hybridization solution. After several washes at 65ºC, embryos were
incubated at room temperature with anti-dig alkaline phosphatase conjugated
antibody (Roche) in PBT (1x PBS, 0,1% Tween-20) for one and half hour. The
staining of the embryos was done with NBT (Roche) and BCIP (Roche) in NBT-
buffer (100mM NaCl, 50mM MgCl2, 10mM Tris pH 9.5 and 0,1% Tween-20).
Egg chamber in situ hybridizations were done similarly to embryos’ hybridizations
but were fixed only once in 4%paraformaldehyde in PBS, for 30min. To prevent a
high background the probe concentration was lower in these hybridizations (the
concentration was determined experimentally).
Samples were photographed in a Zeiss axiophote.
9 EXPERIMENTAL PROCEDURES
153
9.12 Slbo antiserum
Recombinant Slbo without the polyglutamine repeat (slboo7) was expressed in
Escherichia coli and purified in large scale by FPLC by Pernille Rørth as
described in Rørth and Montell (1992). 250µ~Protein was mixed with Ribi
adjuvant (Corixa) and injected in rats by the Laboratory Animal Resources at
EMBL. Three rat’s antisera were generated. I have used a mix of the three sera.
9.13 Immunocytochemistry For immunocytochemistry samples were fixed in 4% paraformaldehyde in PBS.
Embryos were pre-fixed and dechorionated as described in section 9.11. Care
was taken to expose the embryos to methanol the minimum amount of time
possible.
Samples were pre-incubated 30min in PT-NGS (1x PBS, 0.1% Triton-X and 5%
normal goat serum). Primary antibody was diluted in PT-NGS and incubation was
done overnight. Washes were done for 2h in PT-BSA (1x PBS, 0.1% Triton-X
and 0.2% BSA). Secondary antibody was diluted in PT-NGS and incubated for
two hours. Optionally, before washes, samples were incubated with 1µg/ml DAPI
in 1x PBS, 0.1% Triton-X, for 10 minutes. Samples are then washed in 1x PBS,
0.1% Triton-X for two hours.
Samples were observed in a Zeiss Axiophot or with a Leica laser scanning
confocal microscope.
Primary antibodies used were: rat anti-Slbo diluted 1:1000, rabbit anti-Serpent
diluted 1:1000 (a gift from Veit Riechmann), and rat anti-Lsd2 serum diluted
1:2000.
9 EXPERIMENTAL PROCEDURES
154
9.14 Triacylglycerol quantification The protocol for TAG quantification was kindly provided by Ronald Kühnlein
(Gronke et al. 2003). In order to quantify the triacylglycerol (TAG) levels in
embryos and adult males samples were homogenized in 1ml PBS 0.05% tween
20. Samples were incubated 5 minutes at 70˚C in a water bath and centrifuged
for 1 minute at 5000 rpm in a tabletop micro centrifugator. 500 µl of supernatant
were transferred to a new tube and centrifuged for 3 minutes at 14000 rpm, in a
tabletop micro centrifugator. From this supernatant 50µl were added to 1ml of
used of Triglycerides solution (ThermoTrace) and incubated for 5 minutes in a
shaker at 37˚C and 150 rpm. Absorbance was then measured at 520nm. In order
to normalize the samples the protein concentration was also determined from
50µl of the last supernatant using the Bio-Rad Protein Assay, according to the
manufacturer’s protocols.
155
10 Bibliography Agaisse, H., U. M. Petersen, M. Boutros, B. Mathey-Prevot and N. Perrimon (2003). "Signaling role of hemocytes in Drosophila JAK/STAT-dependent response to septic injury." Dev Cell 5(3): 441-50. Akira, S., H. Isshiki, T. Sugita, O. Tanabe, S. Kinoshita, Y. Nishio, T. Nakajima, T. Hirano and T. Kishimoto (1990). "A nuclear factor for IL-6 expression (NF-IL6) is a member of a C/EBP family." Embo J 9(6): 1897-906. Alam, T., M. R. An and J. Papaconstantinou (1992). "Differential expression of three C/EBP isoforms in multiple tissues during the acute phase response." J Biol Chem 267(8): 5021-4. Alberini, C. M., M. Ghirardi, R. Metz and E. R. Kandel (1994). "C/EBP is an immediate-early gene required for the consolidation of long-term facilitation in Aplysia." Cell 76(6): 1099-114. Alverson, J. and A. C. Cohen (2002). "Effect of antifungal agents on biological fitness of Lygus hesperus (Heteroptera: Miridae)." J Econ Entomol 95(2): 256-60. Arbeitman, M. N., E. E. Furlong, F. Imam, E. Johnson, B. H. Null, B. S. Baker, M. A. Krasnow, M. P. Scott, R. W. Davis and K. P. White (2002). "Gene expression during the life cycle of Drosophila melanogaster." Science 297(5590): 2270-5. Ashburner, M. and J. Thompson, J. N. (1978). The laboratory culture of Drosophila. The Genetics and Biology of Drosophila. M. Ashburner and T. R. F. Wright. New York, Academic Press. 2a: 2-109. Ashida, M. and P. T. Brey (1995). "Role of the integument in insect defense: pro-phenol oxidase cascade in the cuticular matrix." Proc Natl Acad Sci U S A 92(23): 10698-702. Ashida, M. and P. T. Brey (1997). Recent advances in research on the insect prophenoloxidase cascade. Molecular Mechanisms of Immune Responses in Insects. P. T. Brey and D. Hultmark. London, Chapman and Hall: 135-172. Bai, J., Y. Uehara and D. J. Montell (2000). "Regulation of invasive cell behavior by taiman, a Drosophila protein related to AIB1, a steroid receptor coactivator amplified in breast cancer." Cell 103(7): 1047-58. Barbero, P., E. Buell, S. Zulley and S. R. Pfeffer (2001). "TIP47 is not a component of lipid droplets." J Biol Chem 276(26): 24348-51. Barry, M. K., A. A. Triplett and A. C. Christensen (1999). "A peritrophin-like protein expressed in the embryonic tracheae of Drosophila melanogaster." Insect Biochem Mol Biol 29(4): 319-27. Beccari, S., L. Teixeira and P. Rørth (2002). "The JAK/STAT pathway is required for border cell migration during Drosophila oogenesis." Mech Dev 111(1-2): 115-23. Birnbaum, M. J. and L. I. Gilbert (1990). "Juvenile hormone stimulation of ornithine decarboxylase activity during vitellogenesis in Drosophila melanogaster." J Comp Physiol [B] 160(2): 145-51. Blanchette-Mackie, E. J., N. K. Dwyer, T. Barber, R. A. Coxey, T. Takeda, C. M. Rondinone, J. L. Theodorakis, A. S. Greenberg and C. Londos (1995). "Perilipin is located on the surface layer of intracellular lipid droplets in adipocytes." J Lipid Res 36(6): 1211-26.
156
Bobinnec, Y., C. Marcaillou, X. Morin and A. Debec (2003). "Dynamics of the endoplasmic reticulum during early development of Drosophila melanogaster." Cell Motil Cytoskeleton 54(3): 217-25. Borror, D. J. and D. M. DeLong (1971). An introduction to the study of insects, Holt, Rinehart and Winston, Inc. Boutros, M., H. Agaisse and N. Perrimon (2002). "Sequential activation of signaling pathways during innate immune responses in Drosophila." Dev Cell 3(5): 711-22. Brasaemle, D. L., T. Barber, A. R. Kimmel and C. Londos (1997). "Post-translational regulation of perilipin expression. Stabilization by stored intracellular neutral lipids." J Biol Chem 272(14): 9378-87. Brasaemle, D. L., T. Barber, N. E. Wolins, G. Serrero, E. J. Blanchette-Mackie and C. Londos (1997). "Adipose differentiation-related protein is an ubiquitously expressed lipid storage droplet-associated protein." J Lipid Res 38(11): 2249-63. Brasaemle, D. L., B. Rubin, I. A. Harten, J. Gruia-Gray, A. R. Kimmel and C. Londos (2000). "Perilipin A increases triacylglycerol storage by decreasing the rate of triacylglycerol hydrolysis." J Biol Chem 275(49): 38486-93. Britton, J. S., W. K. Lockwood, L. Li, S. M. Cohen and B. A. Edgar (2002). "Drosophila's insulin/PI3-kinase pathway coordinates cellular metabolism with nutritional conditions." Dev Cell 2(2): 239-49. Brown, W. J., T. R. Sullivan and P. Greenspan (1992). "Nile red staining of lysosomal phospholipid inclusions." Histochemistry 97(4): 349-54. Burgess-Beusse, B. L. and G. J. Darlington (1998). "C/EBPalpha is critical for the neonatal acute-phase response to inflammation." Mol Cell Biol 18(12): 7269-77. Buszczak, M., X. Lu, W. A. Segraves, T. Y. Chang and L. Cooley (2002). "Mutations in the midway gene disrupt a Drosophila acyl coenzyme A: diacylglycerol acyltransferase." Genetics 160(4): 1511-8. Byus, C. V. and E. J. Herbst (1976). "The effect of polyamines on the synthesis of ribonucleic acid by Drosophila melanogaster larvae." Biochem J 154(1): 23-9. Canavoso, L. E., Z. E. Jouni, K. J. Karnas, J. E. Pennington and M. A. Wells (2001). "Fat metabolism in insects." Annu Rev Nutr 21: 23-46. Cappelletti, M., T. Alonzi, E. Fattori, C. Libert and V. Poli (1996). "C/EBP beta is required for the late phases of acute phase genes induction in the liver and for tumour necrosis factor-alpha, but not Interleukin-6, regulation." CELL DEATH AND DIFFERENTIATION 3(1): 29-35. Casso, D., F. Ramirez-Weber and T. B. Kornberg (2000). "GFP-tagged balancer chromosomes for Drosophila melanogaster." Mech Dev 91(1-2): 451-4. Cayre, M., C. Strambi, P. Charpin, R. Augier and A. Strambi (1997). "Specific requirement of putrescine for the mitogenic action of juvenile hormone on adult insect neuroblasts." Proc Natl Acad Sci U S A 94(15): 8238-42.
157
Chattopadhyay, M. K., C. W. Tabor and H. Tabor (2002). "Absolute requirement of spermidine for growth and cell cycle progression of fission yeast (Schizosaccharomyces pombe)." Proc Natl Acad Sci U S A 99(16): 10330-4. Choe, K. M., T. Werner, S. Stoven, D. Hultmark and K. V. Anderson (2002). "Requirement for a peptidoglycan recognition protein (PGRP) in Relish activation and antibacterial immune responses in Drosophila." Science 296(5566): 359-62. Consortium, T. F. (2003). "The FlyBase database of the Drosophila genome projects and community literature." Nucleic Acids Research 31: 172-175 http://flybase.org/. Cooper, C., A. Henderson, S. Artandi, N. Avitahl and K. Calame (1995). "Ig/EBP (C/EBP gamma) is a transdominant negative inhibitor of C/EBP family transcriptional activators." Nucleic Acids Res 23(21): 4371-7. Croniger, C., M. Trus, K. Lysek-Stupp, H. Cohen, Y. Liu, G. J. Darlington, V. Poli, R. W. Hanson and L. Reshef (1997). "Role of the isoforms of CCAAT/enhancer-binding protein in the initiation of phosphoenolpyruvate carboxykinase (GTP) gene transcription at birth." J Biol Chem 272(42): 26306-12. Croniger, C. M., C. Millward, J. Yang, Y. Kawai, I. J. Arinze, S. Liu, M. Harada-Shiba, K. Chakravarty, J. E. Friedman, V. Poli and R. W. Hanson (2001). "Mice with a deletion in the gene for CCAAT/enhancer-binding protein beta have an attenuated response to cAMP and impaired carbohydrate metabolism." J Biol Chem 276(1): 629-38. Cummings, C. A. and D. A. Relman (2000). "Using DNA microarrays to study host-microbe interactions." Emerg Infect Dis 6(5): 513-25. Czech, M. P. (1995). "Molecular actions of insulin on glucose transport." Annu Rev Nutr 15: 441-71. Darbre, P. D. (2003). "Underarm cosmetics and breast cancer." J Appl Toxicol 23(2): 89-95. Darbre, P. D., J. R. Byford, L. E. Shaw, S. Hall, N. G. Coldham, G. S. Pope and M. J. Sauer (2003). "Oestrogenic activity of benzylparaben." J Appl Toxicol 23(1): 43-51. Darlington, G. J., S. E. Ross and O. A. MacDougald (1998). "The role of C/EBP genes in adipocyte differentiation." J Biol Chem 273(46): 30057-60. Darlington, G. J., N. Wang and R. W. Hanson (1995). "C/EBP alpha: a critical regulator of genes governing integrative metabolic processes." Curr Opin Genet Dev 5(5): 565-70. De Gregorio, E., S. J. Han, W. J. Lee, M. J. Baek, T. Osaki, S. Kawabata, B. L. Lee, S. Iwanaga, B. Lemaitre and P. T. Brey (2002). "An immune-responsive Serpin regulates the melanization cascade in Drosophila." Dev Cell 3(4): 581-92. De Gregorio, E., P. T. Spellman, G. M. Rubin and B. Lemaitre (2001). "Genome-wide analysis of the Drosophila immune response by using oligonucleotide microarrays." Proc Natl Acad Sci U S A 98(22): 12590-5. De Gregorio, E., P. T. Spellman, P. Tzou, G. M. Rubin and B. Lemaitre (2002). "The Toll and Imd pathways are the major regulators of the immune response in Drosophila." Embo J 21(11): 2568-79.
158
Descombes, P. and U. Schibler (1991). "A liver-enriched transcriptional activator protein, LAP, and a transcriptional inhibitory protein, LIP, are translated from the same mRNA." Cell 67(3): 569-79. Diaz, E. and S. R. Pfeffer (1998). "TIP47: a cargo selection device for mannose 6-phosphate receptor trafficking." Cell 93(3): 433-43. Diehl, A. M. (1998). "Roles of CCAAT/enhancer-binding proteins in regulation of liver regenerative growth." J Biol Chem 273(47): 30843-6. Dimopoulos, G., G. K. Christophides, S. Meister, J. Schultz, K. P. White, C. Barillas-Mury and F. C. Kafatos (2002). "Genome expression analysis of Anopheles gambiae: responses to injury, bacterial challenge, and malaria infection." Proc Natl Acad Sci U S A 99(13): 8814-9. Dushay, M. S., B. Asling and D. Hultmark (1996). "Origins of immunity: Relish, a compound Rel-like gene in the antibacterial defense of Drosophila." Proc Natl Acad Sci U S A 93(19): 10343-7. Edgar, B. A. and G. Schubiger (1986). "Parameters controlling transcriptional activation during early Drosophila development." Cell 44(6): 871-7. Ellenberger, T. E., C. J. Brandl, K. Struhl and S. C. Harrison (1992). "The GCN4 basic region leucine zipper binds DNA as a dimer of uninterrupted alpha helices: crystal structure of the protein-DNA complex." Cell 71(7): 1223-37. Ferrandon, D., A. C. Jung, M. Criqui, B. Lemaitre, S. Uttenweiler-Joseph, L. Michaut, J. Reichhart and J. A. Hoffmann (1998). "A drosomycin-GFP reporter transgene reveals a local immune response in Drosophila that is not dependent on the Toll pathway." Embo J 17(5): 1217-27. Furlong, E. E., D. Profitt and M. P. Scott (2001). "Automated sorting of live transgenic embryos." Nat Biotechnol 19(2): 153-6. Gao, J. and G. Serrero (1999). "Adipose differentiation related protein (ADRP) expressed in transfected COS-7 cells selectively stimulates long chain fatty acid uptake." J Biol Chem 274(24): 16825-30. Gao, J., H. Ye and G. Serrero (2000). "Stimulation of adipose differentiation related protein (ADRP) expression in adipocyte precursors by long-chain fatty acids." J Cell Physiol 182(2): 297-302. Garcia, A., A. Sekowski, V. Subramanian and D. L. Brasaemle (2003). "The central domain is required to target and anchor perilipin A to lipid droplets." J Biol Chem 278(1): 625-35. Geisbrecht, E. R. and D. J. Montell (2002). "Myosin VI is required for E-cadherin-mediated border cell migration." Nat Cell Biol 4(8): 616-20. Georgel, P., M. Meister, C. Kappler, B. Lemaitre, J. M. Reichhart and J. A. Hoffmann (1993). "Insect immunity: the diptericin promoter contains multiple functional regulatory sequences homologous to mammalian acute-phase response elements." Biochem Biophys Res Commun 197(2): 508-17. Gingras, A. C., B. Raught and N. Sonenberg (1999). "eIF4 initiation factors: effectors of mRNA recruitment to ribosomes and regulators of translation." Annu Rev Biochem 68: 913-63. Gottar, M., V. Gobert, T. Michel, M. Belvin, G. Duyk, J. A. Hoffmann, D. Ferrandon and J. Royet (2002). "The Drosophila immune response against Gram-negative bacteria is mediated by a peptidoglycan recognition protein." Nature 416(6881): 640-4.
159
Graves, B. J., P. F. Johnson and S. L. McKnight (1986). "Homologous recognition of a promoter domain common to the MSV LTR and the HSV tk gene." Cell 44(4): 565-76. Green, C., E. Levashina, C. McKimmie, T. Dafforn, J. M. Reichhart and D. Gubb (2000). "The necrotic gene in Drosophila corresponds to one of a cluster of three serpin transcripts mapping at 43A1.2." Genetics 156(3): 1117-27. Greenberg, A. S., J. J. Egan, S. A. Wek, N. B. Garty, E. J. Blanchette-Mackie and C. Londos (1991). "Perilipin, a major hormonally regulated adipocyte-specific phosphoprotein associated with the periphery of lipid storage droplets." J Biol Chem 266(17): 11341-6. Greenberg, A. S., J. J. Egan, S. A. Wek, M. C. Moos, Jr., C. Londos and A. R. Kimmel (1993). "Isolation of cDNAs for perilipins A and B: sequence and expression of lipid droplet-associated proteins of adipocytes." Proc Natl Acad Sci U S A 90(24): 12035-9. Gronke, S., M. Beller, S. Fellert, H. Ramakrishnan, H. Jackle and R. P. Kuhnlein (2003). "Control of Fat Storage by a Drosophila PAT Domain Protein." Curr Biol 13(7): 603-6. Hansen, A. J., Y. H. Lee, E. Sterneck, F. J. Gonzalez and P. I. Mackenzie (1998). "C/EBPalpha is a regulator of the UDP glucuronosyltransferase UGT2B1 gene." Mol Pharmacol 53(6): 1027-33. Heid, H. W., R. Moll, I. Schwetlick, H. R. Rackwitz and T. W. Keenan (1998). "Adipophilin is a specific marker of lipid accumulation in diverse cell types and diseases." Cell Tissue Res 294(2): 309-21. Herold, A., L. Teixeira and E. Izaurralde (2003). "Genome-wide analysis of nuclear mRNA export pathways in Drosophila." Embo J 22(10): 2472-83. Holm, S. (1979). "A simple sequentially rejective multiple test procedure." Scandinavian Journal of Statistics 6: 65-70. Hsu, W., T. K. Kerppola, P. L. Chen, T. Curran and S. Chen-Kiang (1994). "Fos and Jun repress transcription activation by NF-IL6 through association at the basic zipper region." Mol Cell Biol 14(1): 268-76. Hultmark, D., A. Engstrom, K. Andersson, H. Steiner, H. Bennich and H. G. Boman (1983). "Insect immunity. Attacins, a family of antibacterial proteins from Hyalophora cecropia." Embo J 2(4): 571-6. Igarashi, K. and K. Kashiwagi (2000). "Polyamines: mysterious modulators of cellular functions." Biochem Biophys Res Commun 271(3): 559-64. Ikeya, T., M. Galic, P. Belawat, K. Nairz and E. Hafen (2002). "Nutrient-dependent expression of insulin-like peptides from neuroendocrine cells in the CNS contributes to growth regulation in Drosophila." Curr Biol 12(15): 1293-300. Imamura, M., T. Inoguchi, S. Ikuyama, S. Taniguchi, K. Kobayashi, N. Nakashima and H. Nawata (2002). "ADRP stimulates lipid accumulation and lipid droplet formation in murine fibroblasts." Am J Physiol Endocrinol Metab 283(4): E775-83. Ip, Y. T., M. Reach, Y. Engstrom, L. Kadalayil, H. Cai, S. Gonzalez-Crespo, K. Tatei and M. Levine (1993). "Dif, a dorsal-related gene that mediates an immune response in Drosophila." Cell 75(4): 753-63.
160
Irving, P., L. Troxler, T. S. Heuer, M. Belvin, C. Kopczynski, J. M. Reichhart, J. A. Hoffmann and C. Hetru (2001). "A genome-wide analysis of immune responses in Drosophila." Proc Natl Acad Sci U S A 98(26): 15119-24. Jiang, H. P. and G. Serrero (1992). "Isolation and characterization of a full-length cDNA coding for an adipose differentiation-related protein." Proc Natl Acad Sci U S A 89(17): 7856-60. Jin, W., R. M. Riley, R. D. Wolfinger, K. P. White, G. Passador-Gurgel and G. Gibson (2001). "The contributions of sex, genotype and age to transcriptional variance in Drosophila melanogaster." Nat Genet 29(4): 389-95. Johnson, P. F. (1993). "Identification of C/EBP basic region residues involved in DNA sequence recognition and half-site spacing preference." Mol Cell Biol 13(11): 6919-30. Johnson, P. F., W. H. Landschulz, B. J. Graves and S. L. McKnight (1987). "Identification of a rat liver nuclear protein that binds to the enhancer core element of three animal viruses." Genes Dev 1(2): 133-46. Johnson, P. F. and S. L. McKnight (1989). "Eukaryotic transcriptional regulatory proteins." Annu Rev Biochem 58: 799-839. Keeley, L. L. (1985). Physiology and Biochemistry of the Fat Body. Comprehensive insect physiology, biochemistry, and pharmacology. G. A. Kerkut and L. I. Gilbert. Oxford Oxfordshire ; New York, Pergamon Press. 3: 211-248. Kennerdell, J. R. and R. W. Carthew (1998). "Use of dsRNA-mediated genetic interference to demonstrate that frizzled and frizzled 2 act in the wingless pathway." Cell 95(7): 1017-26. Kern, A. D., M. A. Oliveira, P. Coffino and M. L. Hackert (1999). "Structure of mammalian ornithine decarboxylase at 1.6 A resolution: stereochemical implications of PLP-dependent amino acid decarboxylases." Structure Fold Des 7(5): 567-81. Kim, B. T., K. Tsuchida, J. Lincecum, H. Kitagawa, M. Bernfield and K. Sugahara (2003). "Identification and characterization of three Drosophila melanogaster glucuronyltransferases responsible for the synthesis of the conserved glycosaminoglycan-protein linkage region of proteoglycans. Two novel homologs exhibit broad specificity toward oligosaccharides from proteoglycans, glycoproteins, and glycosphingolipids." J Biol Chem 278(11): 9116-24. Kimura, T., S. Chowdhury, T. Tanaka, A. Shimizu, K. Iwase, S. Oyadomari, T. Gotoh, H. Matsuzaki, M. Mori, S. Akira and M. Takiguchi (2001). "CCAAT/enhancer-binding protein beta is required for activation of genes for ornithine cycle enzymes by glucocorticoids and glucagon in primary-cultured hepatocytes." FEBS Lett 494(1-2): 105-11. Kimura, T., V. M. Christoffels, S. Chowdhury, K. Iwase, H. Matsuzaki, M. Mori, W. H. Lamers, G. J. Darlington and M. Takiguchi (1998). "Hypoglycemia-associated hyperammonemia caused by impaired expression of ornithine cycle enzyme genes in C/EBPalpha knockout mice." J Biol Chem 273(42): 27505-10. Kinoshita, S., S. Akira and T. Kishimoto (1992). "A member of the C/EBP family, NF-IL6 beta, forms a heterodimer and transcriptionally synergizes with NF-IL6." Proc Natl Acad Sci U S A 89(4): 1473-6. Kitagawa, H., Y. Tone, J. Tamura, K. W. Neumann, T. Ogawa, S. Oka, T. Kawasaki and K. Sugahara (1998). "Molecular cloning and expression of glucuronyltransferase I involved in the biosynthesis of the glycosaminoglycan-protein linkage region of proteoglycans." J Biol Chem 273(12): 6615-8.
161
Lachance, P. E., M. Miron, B. Raught, N. Sonenberg and P. Lasko (2002). "Phosphorylation of eukaryotic translation initiation factor 4E is critical for growth." Mol Cell Biol 22(6): 1656-63. Landschulz, W. H., P. F. Johnson and S. L. McKnight (1988). "The leucine zipper: a hypothetical structure common to a new class of DNA binding proteins." Science 240(4860): 1759-64. Landschulz, W. H., P. F. Johnson and S. L. McKnight (1989). "The DNA binding domain of the rat liver nuclear protein C/EBP is bipartite." Science 243(4899): 1681-8. Lasko, P. (2000). "The drosophila melanogaster genome: translation factors and RNA binding proteins." J Cell Biol 150(2): F51-6. LeClair, K. P., M. A. Blanar and P. A. Sharp (1992). "The p50 subunit of NF-kappa B associates with the NF-IL6 transcription factor." Proc Natl Acad Sci U S A 89(17): 8145-9. Lee, Y. H., B. Sauer, P. F. Johnson and F. J. Gonzalez (1997). "Disruption of the c/ebp alpha gene in adult mouse liver." Mol Cell Biol 17(10): 6014-22. Lee, Y. S. and R. W. Carthew (2003). "Making a better RNAi vector for Drosophila: use of intron spacers." Methods 30(4): 322-9. Lekstrom-Himes, J. and K. G. Xanthopoulos (1998). "Biological role of the CCAAT/enhancer-binding protein family of transcription factors." J Biol Chem 273(44): 28545-8. Lemaitre, B., E. Nicolas, L. Michaut, J. M. Reichhart and J. A. Hoffmann (1996). "The dorsoventral regulatory gene cassette spatzle/Toll/cactus controls the potent antifungal response in Drosophila adults." Cell 86(6): 973-83. Levashina, E. A., E. Langley, C. Green, D. Gubb, M. Ashburner, J. A. Hoffmann and J. M. Reichhart (1999). "Constitutive activation of toll-mediated antifungal defense in serpin-deficient Drosophila." Science 285(5435): 1917-9. Levashina, E. A., S. Ohresser, B. Lemaitre and J. L. Imler (1998). "Two distinct pathways can control expression of the gene encoding the Drosophila antimicrobial peptide metchnikowin." J Mol Biol 278(3): 515-27. Ligoxygakis, P., N. Pelte, C. Ji, V. Leclerc, B. Duvic, M. Belvin, H. Jiang, J. A. Hoffmann and J. M. Reichhart (2002). "A serpin mutant links Toll activation to melanization in the host defence of Drosophila." Embo J 21(23): 6330-7. Lin, F. T. and M. D. Lane (1994). "CCAAT/enhancer binding protein alpha is sufficient to initiate the 3T3-L1 adipocyte differentiation program." Proc Natl Acad Sci U S A 91(19): 8757-61. Linhart, H. G., K. Ishimura-Oka, F. DeMayo, T. Kibe, D. Repka, B. Poindexter, R. J. Bick and G. J. Darlington (2001). "C/EBPalpha is required for differentiation of white, but not brown, adipose tissue." Proc Natl Acad Sci U S A 98(22): 12532-7. Liu, Y. and D. J. Montell (2001). "Jing: a downstream target of slbo required for developmental control of border cell migration." Development 128(3): 321-30. Lu, X., J. Gruia-Gray, N. G. Copeland, D. J. Gilbert, N. A. Jenkins, C. Londos and A. R. Kimmel (2001). "The murine perilipin gene: the lipid droplet-associated perilipins derive from tissue-specific, mRNA splice variants and define a gene family of ancient origin." Mamm Genome 12(9): 741-9.
162
Luque, T., K. Okano and D. R. O'Reilly (2002). "Characterization of a novel silkworm (Bombyx mori) phenol UDP-glucosyltransferase." Eur J Biochem 269(3): 819-25. Luque, T. and D. R. O'Reilly (2002). "Functional and phylogenetic analyses of a putative Drosophila melanogaster UDP-glycosyltransferase gene." Insect Biochem Mol Biol 32(12): 1597-604. Ma, M., T. Lee and E. Kwong (2002). "Interaction of methylparaben preservative with selected sugars and sugar alcohols." J Pharm Sci 91(7): 1715-23. Mahowald, A. P. and M. P. Kambysellis (1980). Oogenesis. The Genetics and Biology of Drosophila. M. Ashburner and T. R. F. Wright. New York, Academic Press. 2D: 141-224. Martinez-Botas, J., J. B. Anderson, D. Tessier, A. Lapillonne, B. H. Chang, M. J. Quast, D. Gorenstein, K. H. Chen and L. Chan (2000). "Absence of perilipin results in leanness and reverses obesity in Lepr(db/db) mice." Nat Genet 26(4): 474-9. McKnight, S. L., M. D. Lane and S. Gluecksohn-Waelsch (1989). "Is CCAAT/enhancer-binding protein a central regulator of energy metabolism?" Genes Dev 3(12B): 2021-4. McManaman, J. L., W. Zabaronick, J. Schaack and D. J. Orlicky (2003). "Lipid droplet targeting domains of adipophilin." J Lipid Res 44(4): 668-73. Meister, M., A. Braun, C. Kappler, J. M. Reichhart and J. A. Hoffmann (1994). "Insect immunity. A transgenic analysis in Drosophila defines several functional domains in the diptericin promoter." Embo J 13(24): 5958-66. Meister, M. and M. Lagueux (2003). "Drosophila blood cells." Cell Microbiol 5(9): 573-80. Mellroth, P., J. Karlsson and H. Steiner (2003). "A scavenger function for a Drosophila peptidoglycan recognition protein." J Biol Chem 278(9): 7059-64. Michel, T., J. M. Reichhart, J. A. Hoffmann and J. Royet (2001). "Drosophila Toll is activated by Gram-positive bacteria through a circulating peptidoglycan recognition protein." Nature 414(6865): 756-9. Miller, M., J. D. Shuman, T. Sebastian, Z. Dauter and P. F. Johnson (2003). "Structural basis for DNA recognition by the basic region leucine zipper transcription factor CCAAT/enhancer-binding protein alpha." J Biol Chem 278(17): 15178-84. Miura, S., J. W. Gan, J. Brzostowski, M. J. Parisi, C. J. Schultz, C. Londos, B. Oliver and A. R. Kimmel (2002). "Functional conservation for lipid storage droplet association among Perilipin, ADRP, and TIP47 (PAT)-related proteins in mammals, Drosophila, and Dictyostelium." J Biol Chem 277(35): 32253-7. Moitra, J., M. M. Mason, M. Olive, D. Krylov, O. Gavrilova, B. Marcus-Samuels, L. Feigenbaum, E. Lee, T. Aoyama, M. Eckhaus, M. L. Reitman and C. Vinson (1998). "Life without white fat: a transgenic mouse." Genes Dev 12(20): 3168-81. Montell, D. J., P. Rørth and A. C. Spradling (1992). "slow border cells, a locus required for a developmentally regulated cell migration during oogenesis, encodes Drosophila C/EBP." Cell 71(1): 51-62. Morris, S. M., Jr. (2002). "Regulation of enzymes of the urea cycle and arginine metabolism." Annu Rev Nutr 22: 87-105.
163
Muller, C., M. Alunni-Fabbroni, E. Kowenz-Leutz, X. Mo, M. Tommasino and A. Leutz (1999). "Separation of C/EBPalpha-mediated proliferation arrest and differentiation pathways." Proc Natl Acad Sci U S A 96(13): 7276-81. Murphy, D. J. (2001). "The biogenesis and functions of lipid bodies in animals, plants and microorganisms." Prog Lipid Res 40(5): 325-438. Nakagawa, Y. and G. Moore (1999). "Role of mitochondrial membrane permeability transition in p-hydroxybenzoate ester-induced cytotoxicity in rat hepatocytes." Biochem Pharmacol 58(5): 811-6. Nakanishi, Y., M. L. Paco-Larson and A. Garen (1986). "DNA homology between the 3'-untranslated regions of a developmentally regulated Drosophila gene and a mouse alpha-interferon gene." Gene 46(1): 79-88. Natzle, J. E., A. S. Hammonds and J. W. Fristrom (1986). "Isolation of genes active during hormone-induced morphogenesis in Drosophila imaginal discs." J Biol Chem 261(12): 5575-83. Niewiadomska, P., D. Godt and U. Tepass (1999). "DE-Cadherin is required for intercellular motility during Drosophila oogenesis." J Cell Biol 144(3): 533-47. Oldham, S. and E. Hafen (2003). "Insulin/IGF and target of rapamycin signaling: a TOR de force in growth control." Trends Cell Biol 13(2): 79-85. Osada, S., H. Yamamoto, T. Nishihara and M. Imagawa (1996). "DNA binding specificity of the CCAAT/enhancer-binding protein transcription factor family." J Biol Chem 271(7): 3891-6. Pabst, T., B. U. Mueller, P. Zhang, H. S. Radomska, S. Narravula, S. Schnittger, G. Behre, W. Hiddemann and D. G. Tenen (2001). "Dominant-negative mutations of CEBPA, encoding CCAAT/enhancer binding protein-alpha (C/EBPalpha), in acute myeloid leukemia." Nat Genet 27(3): 263-70. Paine-Saunders, S., D. Fristrom and J. W. Fristrom (1990). "The Drosophila IMP-E2 gene encodes an apically secreted protein expressed during imaginal disc morphogenesis." Dev Biol 140(2): 337-51. Pereda-Miranda, R., C. B. Bernard, T. Durst, J. T. Arnason, P. Sanchez-Vindas, L. Poveda and L. San Roman (1997). "Methyl 4-hydroxy-3-(3'-methyl-2'-butenyl)benzoate, major insecticidal principle from Piper guanacastensis." J Nat Prod 60(3): 282-4. Poli, V. (1998). "The role of C/EBP isoforms in the control of inflammatory and native immunity functions." J Biol Chem 273(45): 29279-82. Poli, V., F. P. Mancini and R. Cortese (1990). "IL-6DBP, a nuclear protein involved in interleukin-6 signal transduction, defines a new family of leucine zipper proteins related to C/EBP." Cell 63(3): 643-53. Porse, B. T., T. A. Pedersen, X. Xu, B. Lindberg, U. M. Wewer, L. Friis-Hansen and C. Nerlov (2001). "E2F repression by C/EBPalpha is required for adipogenesis and granulopoiesis in vivo." Cell 107(2): 247-58. Ptushkina, M., N. Malys and J. E. McCarthy (2004). "eIF4E isoform 2 in Schizosaccharomyces pombe is a novel stress-response factor." EMBO Rep 5(3): 311-316. Raha, D., Q. D. Nguyen and A. Garen (1990). "Molecular and developmental analyses of the protein encoded by the Drosophila gene ectodermal." Dev Genet 11(4): 310-7.
164
Ramet, M., P. Manfruelli, A. Pearson, B. Mathey-Prevot and R. A. Ezekowitz (2002). "Functional genomic analysis of phagocytosis and identification of a Drosophila receptor for E. coli." Nature 416(6881): 644-8. Ramji, D. P. and P. Foka (2002). "CCAAT/enhancer-binding proteins: structure, function and regulation." Biochem J 365(Pt 3): 561-75. Ramji, D. P., A. Vitelli, F. Tronche, R. Cortese and G. Ciliberto (1993). "The two C/EBP isoforms, IL-6DBP/NF-IL6 and C/EBP delta/NF-IL6 beta, are induced by IL-6 to promote acute phase gene transcription via different mechanisms." Nucleic Acids Res 21(2): 289-94. Rausell, C., J. Llorca and M. D. Real (1997). "Separation by FPLC chromatofocusing of UDP-glucosyltransferases from three developmental stages of Drosophila melanogaster." Arch Insect Biochem Physiol 34(3): 347-58. Rehorn, K. P., H. Thelen, A. M. Michelson and R. Reuter (1996). "A molecular aspect of hematopoiesis and endoderm development common to vertebrates and Drosophila." Development 122(12): 4023-31. Richter, A., C. Schwager, S. Hentze, W. Ansorge, M. W. Hentze and M. Muckenthaler (2002). "Comparison of fluorescent tag DNA labeling methods used for expression analysis by DNA microarrays." Biotechniques 33(3): 620-8, 630. Robalino, J., B. Joshi, S. C. Fahrenkrug and R. Jagus (2004). "Two Zebrafish eIF4E Family Members Are Differentially Expressed and Functionally Divergent." J Biol Chem 279(11): 10532-41. Rojas, M. G., H. J. Williams, R. D. Stipanovic, S. B. Vinson and J. Reibenspies (1990). "A detoxification product of the xenobiotic methyl parasept by Heliothis virescens (F) (Lepidoptera: Noctuidae)." Comp. Biochem. Physiol. 96C(2): 281-285. Rom, E. and C. Kahana (1993). "Isolation and characterization of the Drosophila ornithine decarboxylase locus: evidence for the presence of two transcribed ODC genes in the Drosophila genome." DNA Cell Biol 12(6): 499-508. Rørth, P. (1994). "Specification of C/EBP function during Drosophila development by the bZIP basic region." Science 266(5192): 1878-81. Rørth, P. (1996). "A modular misexpression screen in Drosophila detecting tissue-specific phenotypes." Proc Natl Acad Sci U S A 93(22): 12418-22. Rørth, P. and D. J. Montell (1992). "Drosophila C/EBP: a tissue-specific DNA-binding protein required for embryonic development." Genes Dev 6(12A): 2299-311. Rørth, P., K. Szabo, A. Bailey, T. Laverty, J. Rehm, G. M. Rubin, K. Weigmann, M. Milan, V. Benes, W. Ansorge and S. M. Cohen (1998). "Systematic gain-of-function genetics in Drosophila." Development 125(6): 1049-57. Rørth, P., K. Szabo and G. Texido (2000). "The level of C/EBP protein is critical for cell migration during Drosophila oogenesis and is tightly controlled by regulated degradation." Mol Cell 6(1): 23-30. Rubin, G. M., L. Hong, P. Brokstein, M. Evans-Holm, E. Frise, M. Stapleton and D. A. Harvey (2000). "A Drosophila complementary DNA resource." Science 287(5461): 2222-4.
165
Sacks, O. W. (1986). The man who mistook his wife for a hat. London, Picador. Samson, M. L. (2000). "Drosophila arginase is produced from a nonvital gene that contains the elav locus within its third intron." J Biol Chem 275(40): 31107-14. Scheper, G. C. and C. G. Proud (2002). "Does phosphorylation of the cap-binding protein eIF4E play a role in translation initiation?" Eur J Biochem 269(22): 5350-9. Schipper, R. G. and A. A. Verhofstad (2002). "Distribution patterns of ornithine decarboxylase in cells and tissues: facts, problems, and postulates." J Histochem Cytochem 50(9): 1143-60. Servetnick, D. A., D. L. Brasaemle, J. Gruia-Gray, A. R. Kimmel, J. Wolff and C. Londos (1995). "Perilipins are associated with cholesteryl ester droplets in steroidogenic adrenal cortical and Leydig cells." J Biol Chem 270(28): 16970-3. Shantz, L. M. and A. E. Pegg (1999). "Translational regulation of ornithine decarboxylase and other enzymes of the polyamine pathway." Int J Biochem Cell Biol 31(1): 107-22. Shimoda, Y., Y. Tajima, T. Nagase, K. Harii, N. Osumi and Y. Sanai (1999). "Cloning and expression of a novel galactoside beta1, 3-glucuronyltransferase involved in the biosynthesis of HNK-1 epitope." J Biol Chem 274(24): 17115-22. Silver, D. L. and D. J. Montell (2001). "Paracrine signaling through the JAK/STAT pathway activates invasive behavior of ovarian epithelial cells in Drosophila." Cell 107(7): 831-41. Sivitz, W. I., S. L. DeSautel, T. Kayano, G. I. Bell and J. E. Pessin (1989). "Regulation of glucose transporter messenger RNA in insulin-deficient states." Nature 340(6228): 72-4. Söderhäll, K. and L. Cerenius (1998). "Role of the prophenoloxidase-activating system in invertebrate immunity." Curr Opin Immunol 10(1): 23-8. Soni, M. G., S. L. Taylor, N. A. Greenberg and G. A. Burdock (2002). "Evaluation of the health aspects of methyl paraben: a review of the published literature." Food Chem Toxicol 40(10): 1335-73. Souza, S. C., L. M. de Vargas, M. T. Yamamoto, P. Lien, M. D. Franciosa, L. G. Moss and A. S. Greenberg (1998). "Overexpression of perilipin A and B blocks the ability of tumor necrosis factor alpha to increase lipolysis in 3T3-L1 adipocytes." J Biol Chem 273(38): 24665-9. Souza, S. C., K. V. Muliro, L. Liscum, P. Lien, M. T. Yamamoto, J. E. Schaffer, G. E. Dallal, X. Wang, F. B. Kraemer, M. Obin and A. S. Greenberg (2002). "Modulation of hormone-sensitive lipase and protein kinase A-mediated lipolysis by perilipin A in an adenoviral reconstituted system." J Biol Chem 277(10): 8267-72. Tabor, C. W. and H. Tabor (1984). "Polyamines." Annu Rev Biochem 53: 749-90. Tanaka, T., N. Yoshida, T. Kishimoto and S. Akira (1997). "Defective adipocyte differentiation in mice lacking the C/EBPbeta and/or C/EBPdelta gene." Embo J 16(24): 7432-43. Tang, Q. Q., T. C. Otto and M. D. Lane (2003). "CCAAT/enhancer-binding protein beta is required for mitotic clonal expansion during adipogenesis." Proc Natl Acad Sci U S A 100(3): 850-5. Tansey, J. T., C. Sztalryd, J. Gruia-Gray, D. L. Roush, J. V. Zee, O. Gavrilova, M. L. Reitman, C. X. Deng, C. Li, A. R. Kimmel and C. Londos (2001). "Perilipin ablation results in a lean mouse
166
with aberrant adipocyte lipolysis, enhanced leptin production, and resistance to diet-induced obesity." Proc Natl Acad Sci U S A 98(11): 6494-9. Targett-Adams, P., D. Chambers, S. Gledhill, G. Hope, J. F. Coy, A. Girod and J. McLauchlan (2003). "Live cell analysis and targeting of the lipid droplet binding protein ADRP." J Biol Chem. Teixeira, L., C. Rabouille, P. Rorth, A. Ephrussi and N. F. Vanzo (2003). "Drosophila Perilipin/ADRP homologue Lsd2 regulates lipid metabolism." Mech Dev 120(9): 1071-1081. Terayama, K., T. Seiki, A. Nakamura, K. Matsumori, S. Ohta, S. Oka, M. Sugita and T. Kawasaki (1998). "Purification and characterization of a glucuronyltransferase involved in the biosynthesis of the HNK-1 epitope on glycoproteins from rat brain." J Biol Chem 273(46): 30295-300. Thomas, T. and T. J. Thomas (2003). "Polyamine metabolism and cancer." J Cell Mol Med 7(2): 113-26. Tomancak, P., A. Beaton, R. Weiszmann, E. Kwan, S. Shu, S. E. Lewis, S. Richards, M. Ashburner, V. Hartenstein, S. E. Celniker and G. M. Rubin (2002). "Systematic determination of patterns of gene expression during Drosophila embryogenesis." Genome Biol 3(12): RESEARCH0088. Tzou, P., S. Ohresser, D. Ferrandon, M. Capovilla, J. M. Reichhart, B. Lemaitre, J. A. Hoffmann and J. L. Imler (2000). "Tissue-specific inducible expression of antimicrobial peptide genes in Drosophila surface epithelia." Immunity 13(5): 737-48. Vallejo, M., D. Ron, C. P. Miller and J. F. Habener (1993). "C/ATF, a member of the activating transcription factor family of DNA-binding proteins, dimerizes with CAAT/enhancer-binding proteins and directs their binding to cAMP response elements." Proc Natl Acad Sci U S A 90(10): 4679-83. Van Gelder, R. N., M. E. von Zastrow, A. Yool, W. C. Dement, J. D. Barchas and J. H. Eberwine (1990). "Amplified RNA synthesized from limited quantities of heterogeneous cDNA." Proc Natl Acad Sci U S A 87(5): 1663-7. Vinson, C. R., T. Hai and S. M. Boyd (1993). "Dimerization specificity of the leucine zipper-containing bZIP motif on DNA binding: prediction and rational design." Genes Dev 7(6): 1047-58. Voie, A. M. and S. M. Cohen (1998). Germline transformation in Drosophila melanogaster. Cell biology: a laboratory handbook. San Diego, Academic Press: 510-517. Wallage, H. R., D. F. Niven and M. E. Rau (2001). "Effects of Plagiorchis elegans (Trematoda: Plagiorchiidae) infection on the carbohydrate metabolism of fourth instar Aedes aegypti (Diptera: Culicidae)." J Med Entomol 38(2): 312-7. Wang, H., P. Iakova, M. Wilde, A. Welm, T. Goode, W. J. Roesler and N. A. Timchenko (2001). "C/EBPalpha arrests cell proliferation through direct inhibition of Cdk2 and Cdk4." Mol Cell 8(4): 817-28. Wang, N. D., M. J. Finegold, A. Bradley, C. N. Ou, S. V. Abdelsayed, M. D. Wilde, L. R. Taylor, D. R. Wilson and G. J. Darlington (1995). "Impaired energy homeostasis in C/EBP alpha knockout mice." Science 269(5227): 1108-12. Weeks, C. E., A. L. Herrmann, F. R. Nelson and T. J. Slaga (1982). "alpha-Difluoromethylornithine, an irreversible inhibitor of ornithine decarboxylase, inhibits tumor promoter-induced polyamine accumulation and carcinogenesis in mouse skin." Proc Natl Acad Sci U S A 79(19): 6028-32.
167
Werner, T., G. Liu, D. Kang, S. Ekengren, H. Steiner and D. Hultmark (2000). "A family of peptidoglycan recognition proteins in the fruit fly Drosophila melanogaster." Proc Natl Acad Sci U S A 97(25): 13772-7. Wolins, N. E., B. Rubin and D. L. Brasaemle (2001). "TIP47 associates with lipid droplets." J Biol Chem 276(7): 5101-8. Wu, Z., Y. Xie, N. L. Bucher and S. R. Farmer (1995). "Conditional ectopic expression of C/EBP beta in NIH-3T3 cells induces PPAR gamma and stimulates adipogenesis." Genes Dev 9(19): 2350-63. Wyatt, G. R., K. Rothaus, D. Lawler and E. J. Herbst (1973). "Ornithine decarboxylase and polyamines in silkmoth pupal tissues: effects of ecdysone and injury." Biochim Biophys Acta 304(2): 482-94. Yeh, W. C., Z. Cao, M. Classon and S. L. McKnight (1995). "Cascade regulation of terminal adipocyte differentiation by three members of the C/EBP family of leucine zipper proteins." Genes Dev 9(2): 168-81. Yoshida, H., K. Kinoshita and M. Ashida (1996). "Purification of a peptidoglycan recognition protein from hemolymph of the silkworm, Bombyx mori." J Biol Chem 271(23): 13854-60. Yoshida, K. M., N. Juni and S. H. Hori (1997). "Molecular cloning and characterization of Drosophila ornithine aminotransferase gene." Genes Genet Syst 72(1): 9-17. Yu, H., P. K. Yoo, C. C. Aguirre, R. W. Tsoa, R. M. Kern, W. W. Grody, S. D. Cederbaum and R. K. Iyer (2003). "Widespread expression of arginase I in mouse tissues. Biochemical and physiological implications." J Histochem Cytochem 51(9): 1151-60. Zinke, I., C. Kirchner, L. C. Chao, M. T. Tetzlaff and M. J. Pankratz (1999). "Suppression of food intake and growth by amino acids in Drosophila: the role of pumpless, a fat body expressed gene with homology to vertebrate glycine cleavage system." Development 126(23): 5275-84. Zinke, I., C. S. Schutz, J. D. Katzenberger, M. Bauer and M. J. Pankratz (2002). "Nutrient control of gene expression in Drosophila: microarray analysis of starvation and sugar-dependent response." Embo J 21(22): 6162-6173.
![Carbohydrate Metabolism in Drosophila: Reliance on the ... · In Drosophila, there is a glucose transport system with similar kinetics to vertebrate Glut transporters [18]. A Drosophila](https://img.dokumen.tips/doc/110x75/5f73231376e389099a64a306/carbohydrate-metabolism-in-drosophila-reliance-on-the-in-drosophila-there.jpg)
