15
CENTER FOR BIOLOGICAL SEQUENCE ANALYSIS Technical University of Denmark - DTU Department of systems biology EVOLUTION in immunity Sunday, 14 June 2009

EVOLUTION - CBS · 2009. 6. 16. · past genome-wide evolution. Globins and lysozyme are representative of the first class; not being functionally constrained as much as fundamental

  • Upload
    others

  • View
    1

  • Download
    0

Embed Size (px)

Citation preview

Page 1: EVOLUTION - CBS · 2009. 6. 16. · past genome-wide evolution. Globins and lysozyme are representative of the first class; not being functionally constrained as much as fundamental

CENTER FOR BIOLOGICAL SEQUENCE ANALYSIS

Technical University of Denmark - DTUDepartment of systems biology

EVOLUTIONin immunity

Sunday, 14 June 2009

Page 2: EVOLUTION - CBS · 2009. 6. 16. · past genome-wide evolution. Globins and lysozyme are representative of the first class; not being functionally constrained as much as fundamental

Technical University of Denmark - DTUDepartment of systems biology

CE

NT

ER

FOR

BIO

LOG

ICA

L SE

QU

EN

CE

AN

ALY

SIS

EVOLUTION OF THE IMMUNE SYSTEM

716 Genomes and evolution

Fiaure 1

Protostomes

Echinoderms

Sea urchin

Urochordates

Tunicate

Cephalochordates

Amphioxus

lawless fish

Lamprey, Hagfish

Cartilaginous fish

Shark

Zebrafish, PufferfIsh

Mammals

Mouse

I------ Reptiles

L Birds

Chick

I I I ( MIllIon years ago

800 600 400 200 0

C 1996 Current Oplnton I” Genetics & Development

Phylogenetic relationships among extant deuterostome classes and

phyla. (The exact branching order of lampreys and hagfish is a matter

of debate.) Note that the time span in which the advanced vertebrate

body plan evolved - between the divergence of amphioxus and

cartilaginous fish - is relatively short. Known phylogenies such as this

one are essential for inferring duplications in gene families and other

evolutionary events.

which is error-prone because independent gene loss or

gain cannot be inferred-demonstrates that one set of

duplications probably occurred in a common ancestor

of all jawed and jawless vertebrates after the lineage

leading to present day amphioxus diverged. The second

one probably occurred in the common ancestor of jawed

vertebrates after the jawless ones diverged (Fig. 2).

Many terminal target genes follow different rules

Terminal target genes are poorer indicators generally

of past genome-wide events, as there are at least four

classes of such genes which do not allow inferences of

past genome-wide evolution. Globins and lysozyme are

representative of the first class; not being functionally

constrained as much as fundamental regulators, they have

accumulated mutations of all kinds-point substitutions,

regulatory mutations, additional duplications, and gene

losses-in the 500 million years that have elapsed

since the last common ancestor of all jawed vertebrates

(reviewed in [18]). The historical record of genome-wide

events that is retained in fundamental regulators because

of stringent selective constraints is thus erased in this class

of terminal target gene.

Tubulins and actins-representatives of the second class

of terminal target gene-are encoded by multiple genes

because they need to be expressed at high levels but

undergo periodic homogenization by gene conversion or

cycles of duplication and loss. Homogenization among

paralogs, by whichever mechanism, erases their molecular

evolutionary record which renders inference of any events

before the last round of homogenization impossible [26].

A third class of terminal target gene comprises the

ubiquitously expressed, single-copy housekeeping genes

such as manganese superoxide dismutase and the large

subunits of RNA polymerases. As these genes are not

required in several copies to maintain a high protein

expression level, they have survived genome duplications

merely as single copy genes. In this class of terminal target

gene, records of past duplications are thus also lost.

Finally, the fourth class of terminal target gene is

exemplified by the olfactory receptor and immunoglobulin

gene families. These proteins interact with ligands of

environmental origin and potentially unlimited diversity.

Structural alterations in the absence of regulatory changes

allow new duplicates to interact with novel ligands. Newly

duplicated paralogs that interact with new ligands are

positively selected, which has caused expansion of these

gene families since the origin of vertebrates and erased

their records of early genome-wide events.

Other genome-wide events

Genome duplications at the origin of vertebrates in

evolution do not preclude the possibility that other events

may have occurred at or around the same time, such

as the de IIO’V’O generation of vertebrate-specific genes or

individual gene duplications or losses. The very tight

linkage of 1Cr!lf-5 and ;11&6 and of insulin and Igf-2. for

example, may be a remnant of ancient tandem duplications

[ 181 which could be an alternative to the earlier of the two

proposed genome duplications. A more comprehensive

description of the genomic events during the early

evolution of vertebrates will be achieved once more

sequences from cephalochordates (amphioxus), jawless

vertebrates (hagfish and lamprey), and cartilaginous fish

(sharks) become available. If, however, we accept the

evidence for two large-scale gene duplications at the origin

of vertebrates (Fig. Z), the question of what happened to

the genes since then presents itself.

Developmental software evolution: regulatory

changes After the genome duplications, some of the newly

duplicated structural genes-the genetic ‘hardware’ of

Innate immunityToll like Receptor (TLR)

Sunday, 14 June 2009

Page 3: EVOLUTION - CBS · 2009. 6. 16. · past genome-wide evolution. Globins and lysozyme are representative of the first class; not being functionally constrained as much as fundamental

Technical University of Denmark - DTUDepartment of systems biology

CE

NT

ER

FOR

BIO

LOG

ICA

L SE

QU

EN

CE

AN

ALY

SIS

TOLL LIKE RECEPTORS

Sunday, 14 June 2009

Page 4: EVOLUTION - CBS · 2009. 6. 16. · past genome-wide evolution. Globins and lysozyme are representative of the first class; not being functionally constrained as much as fundamental

Technical University of Denmark - DTUDepartment of systems biology

CE

NT

ER

FOR

BIO

LOG

ICA

L SE

QU

EN

CE

AN

ALY

SIS

TOLL LIKE RECEPTORSCAN RECOGNIZE PATTERNS

François Leulier & Bruno LemaitreNature Reviews Genetics 9, 165-178 (March 2008)

TLRs share a prototypical organization of N-terminal (N) extracellular leucine-rich repeat (LRR) motifs.

TLRs are dimerized and the ectodomain forms a horseshoe-shaped solenoid.

Sunday, 14 June 2009

Page 5: EVOLUTION - CBS · 2009. 6. 16. · past genome-wide evolution. Globins and lysozyme are representative of the first class; not being functionally constrained as much as fundamental

Technical University of Denmark - DTUDepartment of systems biology

CE

NT

ER

FOR

BIO

LOG

ICA

L SE

QU

EN

CE

AN

ALY

SIS

EVOLUTION OF THE IMMUNE SYSTEM

716 Genomes and evolution

Fiaure 1

Protostomes

Echinoderms

Sea urchin

Urochordates

Tunicate

Cephalochordates

Amphioxus

lawless fish

Lamprey, Hagfish

Cartilaginous fish

Shark

Zebrafish, PufferfIsh

Mammals

Mouse

I------ Reptiles

L Birds

Chick

I I I ( MIllIon years ago

800 600 400 200 0

C 1996 Current Oplnton I” Genetics & Development

Phylogenetic relationships among extant deuterostome classes and

phyla. (The exact branching order of lampreys and hagfish is a matter

of debate.) Note that the time span in which the advanced vertebrate

body plan evolved - between the divergence of amphioxus and

cartilaginous fish - is relatively short. Known phylogenies such as this

one are essential for inferring duplications in gene families and other

evolutionary events.

which is error-prone because independent gene loss or

gain cannot be inferred-demonstrates that one set of

duplications probably occurred in a common ancestor

of all jawed and jawless vertebrates after the lineage

leading to present day amphioxus diverged. The second

one probably occurred in the common ancestor of jawed

vertebrates after the jawless ones diverged (Fig. 2).

Many terminal target genes follow different rules

Terminal target genes are poorer indicators generally

of past genome-wide events, as there are at least four

classes of such genes which do not allow inferences of

past genome-wide evolution. Globins and lysozyme are

representative of the first class; not being functionally

constrained as much as fundamental regulators, they have

accumulated mutations of all kinds-point substitutions,

regulatory mutations, additional duplications, and gene

losses-in the 500 million years that have elapsed

since the last common ancestor of all jawed vertebrates

(reviewed in [18]). The historical record of genome-wide

events that is retained in fundamental regulators because

of stringent selective constraints is thus erased in this class

of terminal target gene.

Tubulins and actins-representatives of the second class

of terminal target gene-are encoded by multiple genes

because they need to be expressed at high levels but

undergo periodic homogenization by gene conversion or

cycles of duplication and loss. Homogenization among

paralogs, by whichever mechanism, erases their molecular

evolutionary record which renders inference of any events

before the last round of homogenization impossible [26].

A third class of terminal target gene comprises the

ubiquitously expressed, single-copy housekeeping genes

such as manganese superoxide dismutase and the large

subunits of RNA polymerases. As these genes are not

required in several copies to maintain a high protein

expression level, they have survived genome duplications

merely as single copy genes. In this class of terminal target

gene, records of past duplications are thus also lost.

Finally, the fourth class of terminal target gene is

exemplified by the olfactory receptor and immunoglobulin

gene families. These proteins interact with ligands of

environmental origin and potentially unlimited diversity.

Structural alterations in the absence of regulatory changes

allow new duplicates to interact with novel ligands. Newly

duplicated paralogs that interact with new ligands are

positively selected, which has caused expansion of these

gene families since the origin of vertebrates and erased

their records of early genome-wide events.

Other genome-wide events

Genome duplications at the origin of vertebrates in

evolution do not preclude the possibility that other events

may have occurred at or around the same time, such

as the de IIO’V’O generation of vertebrate-specific genes or

individual gene duplications or losses. The very tight

linkage of 1Cr!lf-5 and ;11&6 and of insulin and Igf-2. for

example, may be a remnant of ancient tandem duplications

[ 181 which could be an alternative to the earlier of the two

proposed genome duplications. A more comprehensive

description of the genomic events during the early

evolution of vertebrates will be achieved once more

sequences from cephalochordates (amphioxus), jawless

vertebrates (hagfish and lamprey), and cartilaginous fish

(sharks) become available. If, however, we accept the

evidence for two large-scale gene duplications at the origin

of vertebrates (Fig. Z), the question of what happened to

the genes since then presents itself.

Developmental software evolution: regulatory

changes After the genome duplications, some of the newly

duplicated structural genes-the genetic ‘hardware’ of

Adaptive immune system

Immunoglobulins(Abs and TCRs)

Sunday, 14 June 2009

Page 6: EVOLUTION - CBS · 2009. 6. 16. · past genome-wide evolution. Globins and lysozyme are representative of the first class; not being functionally constrained as much as fundamental

Technical University of Denmark - DTUDepartment of systems biology

CE

NT

ER

FOR

BIO

LOG

ICA

L SE

QU

EN

CE

AN

ALY

SIS

EVOLUTION OF THE IMMUNE SYSTEM

716 Genomes and evolution

Fiaure 1

Protostomes

Echinoderms

Sea urchin

Urochordates

Tunicate

Cephalochordates

Amphioxus

lawless fish

Lamprey, Hagfish

Cartilaginous fish

Shark

Zebrafish, PufferfIsh

Mammals

Mouse

I------ Reptiles

L Birds

Chick

I I I ( MIllIon years ago

800 600 400 200 0

C 1996 Current Oplnton I” Genetics & Development

Phylogenetic relationships among extant deuterostome classes and

phyla. (The exact branching order of lampreys and hagfish is a matter

of debate.) Note that the time span in which the advanced vertebrate

body plan evolved - between the divergence of amphioxus and

cartilaginous fish - is relatively short. Known phylogenies such as this

one are essential for inferring duplications in gene families and other

evolutionary events.

which is error-prone because independent gene loss or

gain cannot be inferred-demonstrates that one set of

duplications probably occurred in a common ancestor

of all jawed and jawless vertebrates after the lineage

leading to present day amphioxus diverged. The second

one probably occurred in the common ancestor of jawed

vertebrates after the jawless ones diverged (Fig. 2).

Many terminal target genes follow different rules

Terminal target genes are poorer indicators generally

of past genome-wide events, as there are at least four

classes of such genes which do not allow inferences of

past genome-wide evolution. Globins and lysozyme are

representative of the first class; not being functionally

constrained as much as fundamental regulators, they have

accumulated mutations of all kinds-point substitutions,

regulatory mutations, additional duplications, and gene

losses-in the 500 million years that have elapsed

since the last common ancestor of all jawed vertebrates

(reviewed in [18]). The historical record of genome-wide

events that is retained in fundamental regulators because

of stringent selective constraints is thus erased in this class

of terminal target gene.

Tubulins and actins-representatives of the second class

of terminal target gene-are encoded by multiple genes

because they need to be expressed at high levels but

undergo periodic homogenization by gene conversion or

cycles of duplication and loss. Homogenization among

paralogs, by whichever mechanism, erases their molecular

evolutionary record which renders inference of any events

before the last round of homogenization impossible [26].

A third class of terminal target gene comprises the

ubiquitously expressed, single-copy housekeeping genes

such as manganese superoxide dismutase and the large

subunits of RNA polymerases. As these genes are not

required in several copies to maintain a high protein

expression level, they have survived genome duplications

merely as single copy genes. In this class of terminal target

gene, records of past duplications are thus also lost.

Finally, the fourth class of terminal target gene is

exemplified by the olfactory receptor and immunoglobulin

gene families. These proteins interact with ligands of

environmental origin and potentially unlimited diversity.

Structural alterations in the absence of regulatory changes

allow new duplicates to interact with novel ligands. Newly

duplicated paralogs that interact with new ligands are

positively selected, which has caused expansion of these

gene families since the origin of vertebrates and erased

their records of early genome-wide events.

Other genome-wide events

Genome duplications at the origin of vertebrates in

evolution do not preclude the possibility that other events

may have occurred at or around the same time, such

as the de IIO’V’O generation of vertebrate-specific genes or

individual gene duplications or losses. The very tight

linkage of 1Cr!lf-5 and ;11&6 and of insulin and Igf-2. for

example, may be a remnant of ancient tandem duplications

[ 181 which could be an alternative to the earlier of the two

proposed genome duplications. A more comprehensive

description of the genomic events during the early

evolution of vertebrates will be achieved once more

sequences from cephalochordates (amphioxus), jawless

vertebrates (hagfish and lamprey), and cartilaginous fish

(sharks) become available. If, however, we accept the

evidence for two large-scale gene duplications at the origin

of vertebrates (Fig. Z), the question of what happened to

the genes since then presents itself.

Developmental software evolution: regulatory

changes After the genome duplications, some of the newly

duplicated structural genes-the genetic ‘hardware’ of

No Immunoglobulins

Sunday, 14 June 2009

Page 7: EVOLUTION - CBS · 2009. 6. 16. · past genome-wide evolution. Globins and lysozyme are representative of the first class; not being functionally constrained as much as fundamental

Technical University of Denmark - DTUDepartment of systems biology

CE

NT

ER

FOR

BIO

LOG

ICA

L SE

QU

EN

CE

AN

ALY

SIS

ADAPTIVE IMMUNITY?

Hagfish Lamprey

Sunday, 14 June 2009

Page 8: EVOLUTION - CBS · 2009. 6. 16. · past genome-wide evolution. Globins and lysozyme are representative of the first class; not being functionally constrained as much as fundamental

Technical University of Denmark - DTUDepartment of systems biology

CE

NT

ER

FOR

BIO

LOG

ICA

L SE

QU

EN

CE

AN

ALY

SIS

ADAPTIVE IMMUNITY

•Capacity of transplantation rejection

•The responses to skin allografts is specific

•Accelerated second-set rejections indicates immunological memory.

Sunday, 14 June 2009

Page 9: EVOLUTION - CBS · 2009. 6. 16. · past genome-wide evolution. Globins and lysozyme are representative of the first class; not being functionally constrained as much as fundamental

Technical University of Denmark - DTUDepartment of systems biology

CE

NT

ER

FOR

BIO

LOG

ICA

L SE

QU

EN

CE

AN

ALY

SIS

CHALLENGE WITH BACTERIA

•Able to produce serum antibodies.

•Responses have been connected with cells that are morphologically similar to the lymphocytes from jawed vertebrates

Sunday, 14 June 2009

Page 10: EVOLUTION - CBS · 2009. 6. 16. · past genome-wide evolution. Globins and lysozyme are representative of the first class; not being functionally constrained as much as fundamental

Technical University of Denmark - DTUDepartment of systems biology

CE

NT

ER

FOR

BIO

LOG

ICA

L SE

QU

EN

CE

AN

ALY

SIS

Somatic diversification of variablelymphocyte receptors in the agnathansea lampreyZeev Pancer1,2, Chris T. Amemiya6, Gotz R. A. Ehrhardt1,5, Jill Ceitlin7, G. Larry Gartland1,4 & Max D. Cooper1,2,3,4,5

1Division of Developmental and Clinical Immunology, Departments of 2Medicine, 3Pediatrics and 4Microbiology, and the 5Howard Hughes Medical Institute,University of Alabama at Birmingham, Birmingham, Alabama 35294, USA6Molecular Genetics Program, Benaroya Research Institute at Virginia Mason, Seattle, Washington, 98101, USA7University of Michigan, Ann Arbor, Michigan 48109, USA

...........................................................................................................................................................................................................................

Although jawless vertebrates are apparently capable of adaptive immune responses, they have not been found to possess therecombinatorial antigen receptors shared by all jawed vertebrates. Our search for the phylogenetic roots of adaptive immunity inthe lamprey has instead identified a new type of variable lymphocyte receptors (VLRs) composed of highly diverse leucine-richrepeats (LRR) sandwiched between amino- and carboxy-terminal LRRs. An invariant stalk region tethers the VLRs to the cellsurface by means of a glycosyl-phosphatidyl-inositol anchor. To generate rearranged VLR genes of the diversity necessary for ananticipatory immune system, the single lamprey VLR locus contains a large bank of diverse LRR cassettes, available for insertioninto an incomplete germline VLR gene. Individual lymphocytes express a uniquely rearranged VLR gene in monoallelic fashion.Different evolutionary strategies were thus used to generate highly diverse lymphocyte receptors through rearrangement of LRRmodules in agnathans (jawless fish) and of immunoglobulin gene segments in gnathostomes (jawed vertebrates).

Adaptive immune responses in jawed vertebrates are initiated whenantigens are recognized by specific lymphocyte receptors. Antigenreceptor diversity is generated through recombination of variable,diversity and joining gene segments in the immunoglobulin andT-cell receptor (TCR) gene loci. This combinatorial rearrangementgenerates vast repertoires of antibodies against unprocessed anti-gens and of TCRs that recognize antigen fragments presented bymajor histocompatibility complex (MHC) class I and II molecules.Clonally diverse lymphocytes thus form the cornerstone of verte-brate adaptive immunity in the form of immunoglobulin-bearing Bcells and TCR-bearing T cells that differentiate from stem-cellprecursors within primary haematopoietic tissues and the thymus.The principal elements of this recombinatorial immune system areconserved in all jawed vertebrates, and the multigene TCR andimmunoglobulin loci are remarkably complex even in the most basalgnathostome representatives, such as sharks, skates and rays1–3.Adaptive immune responses have also been reported in the

jawless vertebrates lamprey and hagfish, the only surviving descen-dents from the early vertebrate radiation4. For example, lampreysproduce specific circulating agglutinins in response to immuniz-ation5–10, reject second set skin allografts at an accelerated rate9,11

and exhibit delayed type hypersensitivity reactions5. Theseresponses have been attributed to agnathan cells that morphologi-cally resemble the lymphocytes found in jawed vertebrates5,9,11–16. Aswith theirmammalian counterparts, lamprey lymphocytes aremoreirradiation-sensitive than other blood cell types9, aggregate andproliferate in response to antigenic stimulation5,12, and expresstranscription factors such as PU.1/Spi-B and Ikaros that areinvolved in mammalian lymphocyte differentiation17–20. However,immunoglobulin, TCR and MHC genes have not been identified injawless vertebrates nor in the draft genome sequence of theinvertebrate urochordate Ciona intestinalis21.With this in mind, we initiated a search for primordial molecular

elements of the vertebrate immune system in the sea lampreyPetromyzon marinus. In an earlier analysis of transcripts expressedby lymphocyte-like cells from lamprey haematopoietic tissues weidentified several homologues of immune system molecules, but

none of the immunoglobulin superfamily receptor elements used byjawed vertebrates for specific adaptive immunity16,22,23. Reasoningthat immunocompetent cells in the circulationwould bemore likelyto express the genes involved in adaptive responses, we began with asurvey of the transcriptome of activated lymphocytes from theblood of immunostimulated lamprey larvae. This search revealed anew type of highly variable lymphocyte receptors.

Immunostimulated blood lymphocyte transcriptsTo survey the transcriptome of activated lymphocytes, lampreylarvae were stimulated by intraperitoneal injections of an antigen/mitogen cocktail two–four times at weekly intervals. The fraction oflarge lymphocytes among peripheral blood leukocytes 3 days afterthe second booster stimulation was 13-fold greater than in unsti-mulated individuals, and the fraction of myeloid cells was sixfoldgreater (Fig. 1a). Compared with the small lymphocyte-like cells,the activated lymphocytes were nearly double in size, had extensiveazurophilic cytoplasm and featured prominent nucleoli (Fig. 1b).These cells were sorted on the basis of their light scatter character-istics and used to construct subtracted complementary DNAlibraries enriched in messages of activated lymphocytes.

The most abundant group of sequences identified among 1,507subtracted clones consisted of 319 proteins featuring variablenumbers of diverse LRR motifs, which clustered with a set of 52LRR-containing expressed sequence tags (ESTs) from our databaseof unstimulated lymphocyte-like transcripts. A set of 239 uniquelydiverse LRR proteins was obtained after purging sequences withonly the 3 0 ends, and 22 ESTs from this set encoded most or all ofthe open reading frames (ORFs) of 239–304 residues in length(Supplementary Fig. 1). These proteins were provisionally namedvariable lymphocyte receptors (VLR) because each of these 239sequences was unique and their transcripts were found to beexpressed predominantly or exclusively by lymphocytes (Fig. 1c).Unstimulated animals showed highest VLR levels in lymphocytesfrom haematopoietic tissues, whereas immune stimulation resultedin enhanced VLR transcription by the large blood lymphocytes. TheVLR transcripts observed in the myeloid lane of Fig. 1c may reflect

articles

NATURE | VOL 430 | 8 JULY 2004 | www.nature.com/nature174 © 2004 Nature Publishing Group

Nature 430, 174-180 (8 July 2004)

Sunday, 14 June 2009

Page 11: EVOLUTION - CBS · 2009. 6. 16. · past genome-wide evolution. Globins and lysozyme are representative of the first class; not being functionally constrained as much as fundamental

Technical University of Denmark - DTUDepartment of systems biology

CE

NT

ER

FOR

BIO

LOG

ICA

L SE

QU

EN

CE

AN

ALY

SIS

NATURE BIOTECHNOLOGY VOLUME 26 NUMBER 4 APRIL 2008 403

self-tolerance. The fact that VLRs are mod-ular single-chain polypeptides makes them amenable to molecular engineering, and the presence of usually two or three hypervari-able solvent-exposed sites per module should facilitate targeting of these residues by ran-dom mutagenesis in vitro to increase affinity or alter specificity, as previously shown for engineered synthetic LRR proteins1. Finally, VLRs in lamprey lymphocytes are assembled by a process of gene conversion through homologous regions that can span just a few nucleotides. To increase the combinatorial diversity of a VLR library, it might there-fore be possible to recapitulate this process by shuffling domains through homologous recombination in vivo using yeast surface display10. The most ancient antigen receptors may soon take a central place in the biotech-nology arena.

1. Binz, H.K., Amstutz, P. & Plückthun, A. Nat. Biotechnol. 23, 1257–1268 (2005).

2. Pancer, Z. et al. Nature 430, 174–180 (2004).3. Herrin, B.R. et al. Proc. Natl. Acad. Sci. USA 105,

2040–2045 (2008).4. Kim, H.M. et al. J. Biol. Chem. 282, 6726–6732

(2007).5. Pancer, Z. & Cooper, M.D. Annu. Rev. Immunol. 24,

497–518 (2006).6. Alder, M.N. et al. Science 310, 1970–1973 (2005).

Mammalian antibody genes

Light chain Heavy chainPreassembly

germline genes

Mature lymphocyte

genes

Cell-bound antigen receptors

Humoral antibodies

RAG-mediatedrearrangement

Lamprey VLR gene

Assemblyfrom flanking

LRR cassettes

LRR CT

C C C C

C C C C

C C C C C C C C

SP

LRR

1

LRR

1

LRR

1

LRR

V

LRR

V

LRR

V

LRR

V

LRR

V

LRR

V

LRR

V

LRR

V

LRR

V

LRR

V

LRR

V

GPI Hydrophobic LRR NT CP

LRR NT

LRR NT SP Stalk

D JV JV C C

JV V V J J V D JV V D J JC C

Stalk LRR CT

LRR CT

LRR CT

Figure 1 Assembly of lamprey variable lymphocyte receptors (VLRs), compared with the V(D)J-rearrangement that gives rise to mammalian antibodies. VLR genes of jawless vertebrates are assembled by sequential insertion of LRR cassettes from flanking arrays into the incomplete germline gene via gene conversion. Jawed vertebrate antibody genes are assembled via recombination activating gene (RAG)-mediated joining of immunoglobulin gene fragments consisting of variable (V), diversity (D) and joining (J) elements, as well as constant (C) exons. The mature antigen receptors in both cases are expressed on the surfaces of lymphocytes and can be secreted to the plasma. A VLR comprises a set of highly diverse LRR modules capped by disulfide-bondedN-terminal LRR (LRRNT, 24-32 amino acids) and C-terminal LRR (LRRCT, 45-62 amino acids) modules. The 25-residue LRR1 is followed by one to eight 24-residue LRRVs and then a 16-residue LRR, the connecting peptide (CP). There are two or three hypervariable sites in solvent-exposed positions in each of these modules. The invariant portions of VLRs include an N-terminal secretion peptide (SP) and an 81-residue C terminus that contains a threonine/proline-rich stalk (33 amino acids) and a glycosyl phosphatidylinositol (GPI) membrane anchorage motif, which tethers the VLR to the lymphocyte surface. Seven cysteines in the 22-residue hydrophobic C-terminal domain can participate in VLR oligomerization.

siRNAs with gutsIan MacLachlan

Leukocyte-directed, small interfering (si)RNA against cyclin D1 shows promise for treatment of inflammatory bowel disease.

Ian MacLachlan is at Protiva Biotherapeutics Inc., 100-3480 Gilmore Way, Burnaby, BC, Canada, V5G 4Y1. e-mail: [email protected]

7. Rogozin, I.B. et al. Nat. Immunol. 8, 647–656 (2007).

8. Alder, M.N. et al. Nat. Immunol. 9, 319–327 (2008).

9. Flajnik, M.F. Nat. Rev. Immunol. 2, 688–698 (2002).

10. Gai, S.A. & Wittrup, K.D. Curr. Opin. Struct. Biol. 17, 467–473 (2007).

Colitis and Crohn’s disease are debilitating inflammatory bowel diseases in which the digestive tract becomes inflamed, causing severe diarrhea and abdominal pain that can lead to life-threatening complications. Although anti-inflammatory and immu-nosuppressive drugs—the two main classes

of treatment—can bring about long-term remission, they are encumbered by dose-lim-iting side effects. A recent paper by Shimaoka and colleagues1 in Science describes a more targeted approach in which siRNA against cyclin D1 (CyD1) is directed to leukocytes expressing β7 integrin.

CyD1, a cell cycle regulator, is upregulated locally, in association with inflammatory bowel disease, in both epithelial and immune cells. Shimaoka and colleagues1 show that leukocyte-directed CyD1 siRNA inhibits the intestinal inflammatory response in a mouse

NEWS AND V IEWS

©2

008 N

atu

re P

ub

lis

hin

g G

rou

p

htt

p:/

/ww

w.n

atu

re.c

om

/natu

reb

iote

ch

no

log

y

NATURE BIOTECHNOLOGY VOLUME 26 NUMBER 4 APRIL 2008 403

self-tolerance. The fact that VLRs are mod-ular single-chain polypeptides makes them amenable to molecular engineering, and the presence of usually two or three hypervari-able solvent-exposed sites per module should facilitate targeting of these residues by ran-dom mutagenesis in vitro to increase affinity or alter specificity, as previously shown for engineered synthetic LRR proteins1. Finally, VLRs in lamprey lymphocytes are assembled by a process of gene conversion through homologous regions that can span just a few nucleotides. To increase the combinatorial diversity of a VLR library, it might there-fore be possible to recapitulate this process by shuffling domains through homologous recombination in vivo using yeast surface display10. The most ancient antigen receptors may soon take a central place in the biotech-nology arena.

1. Binz, H.K., Amstutz, P. & Plückthun, A. Nat. Biotechnol. 23, 1257–1268 (2005).

2. Pancer, Z. et al. Nature 430, 174–180 (2004).3. Herrin, B.R. et al. Proc. Natl. Acad. Sci. USA 105,

2040–2045 (2008).4. Kim, H.M. et al. J. Biol. Chem. 282, 6726–6732

(2007).5. Pancer, Z. & Cooper, M.D. Annu. Rev. Immunol. 24,

497–518 (2006).6. Alder, M.N. et al. Science 310, 1970–1973 (2005).

Mammalian antibody genes

Light chain Heavy chainPreassembly

germline genes

Mature lymphocyte

genes

Cell-bound antigen receptors

Humoral antibodies

RAG-mediatedrearrangement

Lamprey VLR gene

Assemblyfrom flanking

LRR cassettes

LRR CT

C C C C

C C C C

C C C C C C C C

SP

LRR

1

LRR

1

LRR

1

LRR

V

LRR

V

LRR

V

LRR

V

LRR

V

LRR

V

LRR

V

LRR

V

LRR

V

LRR

V

LRR

V

GPI Hydrophobic LRR NT CP

LRR NT

LRR NT SP Stalk

D JV JV C C

JV V V J J V D JV V D J JC C

Stalk LRR CT

LRR CT

LRR CT

Figure 1 Assembly of lamprey variable lymphocyte receptors (VLRs), compared with the V(D)J-rearrangement that gives rise to mammalian antibodies. VLR genes of jawless vertebrates are assembled by sequential insertion of LRR cassettes from flanking arrays into the incomplete germline gene via gene conversion. Jawed vertebrate antibody genes are assembled via recombination activating gene (RAG)-mediated joining of immunoglobulin gene fragments consisting of variable (V), diversity (D) and joining (J) elements, as well as constant (C) exons. The mature antigen receptors in both cases are expressed on the surfaces of lymphocytes and can be secreted to the plasma. A VLR comprises a set of highly diverse LRR modules capped by disulfide-bondedN-terminal LRR (LRRNT, 24-32 amino acids) and C-terminal LRR (LRRCT, 45-62 amino acids) modules. The 25-residue LRR1 is followed by one to eight 24-residue LRRVs and then a 16-residue LRR, the connecting peptide (CP). There are two or three hypervariable sites in solvent-exposed positions in each of these modules. The invariant portions of VLRs include an N-terminal secretion peptide (SP) and an 81-residue C terminus that contains a threonine/proline-rich stalk (33 amino acids) and a glycosyl phosphatidylinositol (GPI) membrane anchorage motif, which tethers the VLR to the lymphocyte surface. Seven cysteines in the 22-residue hydrophobic C-terminal domain can participate in VLR oligomerization.

siRNAs with gutsIan MacLachlan

Leukocyte-directed, small interfering (si)RNA against cyclin D1 shows promise for treatment of inflammatory bowel disease.

Ian MacLachlan is at Protiva Biotherapeutics Inc., 100-3480 Gilmore Way, Burnaby, BC, Canada, V5G 4Y1. e-mail: [email protected]

7. Rogozin, I.B. et al. Nat. Immunol. 8, 647–656 (2007).

8. Alder, M.N. et al. Nat. Immunol. 9, 319–327 (2008).

9. Flajnik, M.F. Nat. Rev. Immunol. 2, 688–698 (2002).

10. Gai, S.A. & Wittrup, K.D. Curr. Opin. Struct. Biol. 17, 467–473 (2007).

Colitis and Crohn’s disease are debilitating inflammatory bowel diseases in which the digestive tract becomes inflamed, causing severe diarrhea and abdominal pain that can lead to life-threatening complications. Although anti-inflammatory and immu-nosuppressive drugs—the two main classes

of treatment—can bring about long-term remission, they are encumbered by dose-lim-iting side effects. A recent paper by Shimaoka and colleagues1 in Science describes a more targeted approach in which siRNA against cyclin D1 (CyD1) is directed to leukocytes expressing β7 integrin.

CyD1, a cell cycle regulator, is upregulated locally, in association with inflammatory bowel disease, in both epithelial and immune cells. Shimaoka and colleagues1 show that leukocyte-directed CyD1 siRNA inhibits the intestinal inflammatory response in a mouse

NEWS AND V IEWS

©2

008 N

atu

re P

ub

lis

hin

g G

rou

p

htt

p:/

/ww

w.n

atu

re.c

om

/natu

reb

iote

ch

no

log

y

402 VOLUME 26 NUMBER 4 APRIL 2008 NATURE BIOTECHNOLOGY

The oldest antibodies newly discoveredZeev Pancer & Roy A Mariuzza

Lamprey monoclonal antibodies provide stable and high-avidity alternatives to immunoglobulins.

VLR6,7 (Fig. 1). The selected cassettes are sequentially incorporated into the germ-line gene framework to form the functional mature VLR. This process can generate a vast repertoire of VLRs, estimated at over 1014 unique receptors, which is sufficiently diverse to recognize the molecular determinants of most, if not all, potential pathogens.

Lamprey were previously shown to respond vigorously to immunization with bacteria and bacterial antigens such as the anthrax spore coat6. Within several weeks after repeated intraperitoneal injection of spore coats, the lamprey plasma accumulates VLRs that react specifically with the anthrax spores. Interestingly, the BclA glycoprotein component of the spore, which is the major antigenic determinant in mice, was also recognized by the anti-anthrax VLRs. Now, Cooper and colleagues3 have cloned and produced monoclonal VLRs in a human embryonic kidney cell line (HEK-293T) to characterize the specific VLRs acquired by naive lamprey after exposure to the anthrax spore coat. Altogether, 212 VLR transfectants were assayed for antigen recognition by the secreted proteins. Of these, 14 (6.6%) were identified as positive for BclA.

Further analysis revealed that seven mono-clonal VLRs could discriminate between the C-terminal domain of BclA of Bacillus anthracis and that of Bacillus cereus, which differ by only 14 of 134 residues (89.5% identity). The nucleotide sequences of these VLRs were closely related, with the few vari-able residues clustered at surface-exposed positions. These positions mapped to the LRR β-strands that form a β-sheet at the concave surface of the VLR, and to a loop in each of the capping modules at the N- and C-termini of the diversity region (Fig. 1). This suggests that these modules may also contribute to diversity in the capacity for binding antigen6,7.

The secreted VLRs were oligomeric, com-prising four or five disulfide-linked dimeric subunits, somewhat akin to vertebrate IgM antibodies. They were extremely stable, retaining antigen-binding activity after elu-tion from a BclA affinity column at pH > 11 and after storage for one year in the refrig-erator, one month at room temperature or 36 h at 56 °C. Remarkably, one of the VLRs

agglutinated anthrax spores 1,000-fold more efficiently than an equivalent amount of a mouse monoclonal antibody, demonstrating its high avidity for the antigen. By contrast, a monomeric form of this VLR that lacked the C-terminal cysteine-rich domain possessed a much lower affinity for the spores.

In a related recent paper in Nature Immunology8, the same group characterized the immune response of lamprey to anti-gens including bovine serum albumin and keyhole limpet hemocyanin. They failed to detect VLRs that could recognize these pro-teins (injected with or without adjuvants, in soluble form or denatured, or conjugated on beads) and concluded that the lamprey immune response is biased to particulate antigens composed of repetitive epitopes. Although this may be true, it is equally pos-sible that much remains to be learned about how to evoke a robust immune response in jawless fish.

For example, low-affinity membrane-bound and secreted IgM antibodies are found in all jawed vertebrates, arranged as monomers, tetramers and pentamers9. Upon repeated exposure to antigen, these IgM immunoglobulins are replaced by high-affinity antibody isotypes. It is tempting to speculate that a similar process of VLR affin-ity maturation may occur in the lamprey7. Furthermore, surveys of much larger librar-ies of recombinant VLRs may be required to unravel the true potential of LRR-based rearranging antigen receptors. This could be achieved using phage or yeast surface display technologies, which should allow the con-struction and screening of VLR libraries that are orders of magnitude larger than is fea-sible using transfected HEK-293T cells.

VLRs may be excellent protein scaffolds for molecular recognition. Although their likely immunogenicity in humans would hinder therapeutic applications, they could be used for diagnostic applications (e.g., protein chip, flow cytometry, immunohis-tochemistry, enzyme-linked immunosorbent assays), for immunoaffinity purification or for the creation of novel fusion proteins. In addition, the lamprey immune system may prove a rich source of reagents that recognize mammalian antigens invisible to immunoglobulin-based antibodies due to

Zeev Pancer is at the Center of Marine Biotechnology and Roy A. Mariuzza is at the Center for Advanced Research in Biotechnology, both Centers of the University of Maryland Biotechnology Institute, 701 East Pratt St., Baltimore, Maryland 21202, USA.e-mail: [email protected]

The search for alternatives to immunoglob-ulin antibodies has led to binding scaffolds such as lipocalins, fibronectins, ankyrin repeats and src homology domains1. But none of these are natural antigen receptors. Indeed, the only natural adaptive immune system that is not based on immunoglobu-lins is found in lamprey and hagfish—jawless vertebrates that produce variable lymphocyte receptors (VLRs)2 (Fig. 1). In work published recently in the Proceedings of the National Academy of Sciences, Cooper and colleagues3 describe the production of monoclonal VLRs from lamprey, opening the way to biotech-nological applications of the evolutionarily oldest antibodies. Besides offering an excit-ing demonstration of a bona fide adaptive immune response mediated by nonimmu-noglobulin receptors, this study provides the first functional characterization of antigen-specific VLRs.

Jawless fish are members of the ancestral vertebrate taxon and the only remaining sis-ter group of jawed vertebrates such as shark, birds and mammals. Like all other verte-brates, lamprey and hagfish have circulating lymphocytes equipped with highly diverse cell surface and secreted antigen receptors. However, VLRs consist not of immuno-globulin domains, but of leucine-rich repeat (LRR) modules2—structural motifs that form a horseshoe-shaped fold with an inte-rior parallel β-sheet and an exterior array of helices4. The prevalence of LRRs in pattern recognition receptors, such as animal Toll-like receptors and plant pathogen resistance proteins, underscores their extraordinary competence for microbial recognition5.

Much like our own antibodies and T-cell receptors, VLRs are generated in lympho-cytes by DNA rearrangement. Each VLR is assembled from several LRR cassettes, selected from an array of several hundred that flank the gene encoding the germline

NEWS AND V IEWS

©2008 N

atu

re P

ub

lish

ing

Gro

up

h

ttp

://w

ww

.natu

re.c

om

/natu

reb

iote

ch

no

log

y

Sunday, 14 June 2009

Page 12: EVOLUTION - CBS · 2009. 6. 16. · past genome-wide evolution. Globins and lysozyme are representative of the first class; not being functionally constrained as much as fundamental

Technical University of Denmark - DTUDepartment of systems biology

CE

NT

ER

FOR

BIO

LOG

ICA

L SE

QU

EN

CE

AN

ALY

SIS

TOLL LIKE RECEPTORSCAN RECOGNIZE PATTERNS

François Leulier & Bruno LemaitreNature Reviews Genetics 9, 165-178 (March 2008)

TLRs share a prototypical organization of N-terminal (N) extracellular leucine-rich repeat (LRR) motifs.

TLRs are dimerized and the ectodomain forms a horseshoe-shaped solenoid.

Sunday, 14 June 2009

Page 13: EVOLUTION - CBS · 2009. 6. 16. · past genome-wide evolution. Globins and lysozyme are representative of the first class; not being functionally constrained as much as fundamental

Technical University of Denmark - DTUDepartment of systems biology

CE

NT

ER

FOR

BIO

LOG

ICA

L SE

QU

EN

CE

AN

ALY

SIS

GENOME DUPLICATIONS716 Genomes and evolution

Fiaure 1

Protostomes

Echinoderms

Sea urchin

Urochordates

Tunicate

Cephalochordates

Amphioxus

lawless fish

Lamprey, Hagfish

Cartilaginous fish

Shark

Zebrafish, PufferfIsh

Mammals

Mouse

I------ Reptiles

L Birds

Chick

I I I ( MIllIon years ago

800 600 400 200 0

C 1996 Current Oplnton I” Genetics & Development

Phylogenetic relationships among extant deuterostome classes and

phyla. (The exact branching order of lampreys and hagfish is a matter

of debate.) Note that the time span in which the advanced vertebrate

body plan evolved - between the divergence of amphioxus and

cartilaginous fish - is relatively short. Known phylogenies such as this

one are essential for inferring duplications in gene families and other

evolutionary events.

which is error-prone because independent gene loss or

gain cannot be inferred-demonstrates that one set of

duplications probably occurred in a common ancestor

of all jawed and jawless vertebrates after the lineage

leading to present day amphioxus diverged. The second

one probably occurred in the common ancestor of jawed

vertebrates after the jawless ones diverged (Fig. 2).

Many terminal target genes follow different rules

Terminal target genes are poorer indicators generally

of past genome-wide events, as there are at least four

classes of such genes which do not allow inferences of

past genome-wide evolution. Globins and lysozyme are

representative of the first class; not being functionally

constrained as much as fundamental regulators, they have

accumulated mutations of all kinds-point substitutions,

regulatory mutations, additional duplications, and gene

losses-in the 500 million years that have elapsed

since the last common ancestor of all jawed vertebrates

(reviewed in [18]). The historical record of genome-wide

events that is retained in fundamental regulators because

of stringent selective constraints is thus erased in this class

of terminal target gene.

Tubulins and actins-representatives of the second class

of terminal target gene-are encoded by multiple genes

because they need to be expressed at high levels but

undergo periodic homogenization by gene conversion or

cycles of duplication and loss. Homogenization among

paralogs, by whichever mechanism, erases their molecular

evolutionary record which renders inference of any events

before the last round of homogenization impossible [26].

A third class of terminal target gene comprises the

ubiquitously expressed, single-copy housekeeping genes

such as manganese superoxide dismutase and the large

subunits of RNA polymerases. As these genes are not

required in several copies to maintain a high protein

expression level, they have survived genome duplications

merely as single copy genes. In this class of terminal target

gene, records of past duplications are thus also lost.

Finally, the fourth class of terminal target gene is

exemplified by the olfactory receptor and immunoglobulin

gene families. These proteins interact with ligands of

environmental origin and potentially unlimited diversity.

Structural alterations in the absence of regulatory changes

allow new duplicates to interact with novel ligands. Newly

duplicated paralogs that interact with new ligands are

positively selected, which has caused expansion of these

gene families since the origin of vertebrates and erased

their records of early genome-wide events.

Other genome-wide events

Genome duplications at the origin of vertebrates in

evolution do not preclude the possibility that other events

may have occurred at or around the same time, such

as the de IIO’V’O generation of vertebrate-specific genes or

individual gene duplications or losses. The very tight

linkage of 1Cr!lf-5 and ;11&6 and of insulin and Igf-2. for

example, may be a remnant of ancient tandem duplications

[ 181 which could be an alternative to the earlier of the two

proposed genome duplications. A more comprehensive

description of the genomic events during the early

evolution of vertebrates will be achieved once more

sequences from cephalochordates (amphioxus), jawless

vertebrates (hagfish and lamprey), and cartilaginous fish

(sharks) become available. If, however, we accept the

evidence for two large-scale gene duplications at the origin

of vertebrates (Fig. Z), the question of what happened to

the genes since then presents itself.

Developmental software evolution: regulatory

changes After the genome duplications, some of the newly

duplicated structural genes-the genetic ‘hardware’ of

Sunday, 14 June 2009

Page 14: EVOLUTION - CBS · 2009. 6. 16. · past genome-wide evolution. Globins and lysozyme are representative of the first class; not being functionally constrained as much as fundamental

Technical University of Denmark - DTUDepartment of systems biology

CE

NT

ER

FOR

BIO

LOG

ICA

L SE

QU

EN

CE

AN

ALY

SIS

GENOME DUPLICATIONS716 Genomes and evolution

Fiaure 1

Protostomes

Echinoderms

Sea urchin

Urochordates

Tunicate

Cephalochordates

Amphioxus

lawless fish

Lamprey, Hagfish

Cartilaginous fish

Shark

Zebrafish, PufferfIsh

Mammals

Mouse

I------ Reptiles

L Birds

Chick

I I I ( MIllIon years ago

800 600 400 200 0

C 1996 Current Oplnton I” Genetics & Development

Phylogenetic relationships among extant deuterostome classes and

phyla. (The exact branching order of lampreys and hagfish is a matter

of debate.) Note that the time span in which the advanced vertebrate

body plan evolved - between the divergence of amphioxus and

cartilaginous fish - is relatively short. Known phylogenies such as this

one are essential for inferring duplications in gene families and other

evolutionary events.

which is error-prone because independent gene loss or

gain cannot be inferred-demonstrates that one set of

duplications probably occurred in a common ancestor

of all jawed and jawless vertebrates after the lineage

leading to present day amphioxus diverged. The second

one probably occurred in the common ancestor of jawed

vertebrates after the jawless ones diverged (Fig. 2).

Many terminal target genes follow different rules

Terminal target genes are poorer indicators generally

of past genome-wide events, as there are at least four

classes of such genes which do not allow inferences of

past genome-wide evolution. Globins and lysozyme are

representative of the first class; not being functionally

constrained as much as fundamental regulators, they have

accumulated mutations of all kinds-point substitutions,

regulatory mutations, additional duplications, and gene

losses-in the 500 million years that have elapsed

since the last common ancestor of all jawed vertebrates

(reviewed in [18]). The historical record of genome-wide

events that is retained in fundamental regulators because

of stringent selective constraints is thus erased in this class

of terminal target gene.

Tubulins and actins-representatives of the second class

of terminal target gene-are encoded by multiple genes

because they need to be expressed at high levels but

undergo periodic homogenization by gene conversion or

cycles of duplication and loss. Homogenization among

paralogs, by whichever mechanism, erases their molecular

evolutionary record which renders inference of any events

before the last round of homogenization impossible [26].

A third class of terminal target gene comprises the

ubiquitously expressed, single-copy housekeeping genes

such as manganese superoxide dismutase and the large

subunits of RNA polymerases. As these genes are not

required in several copies to maintain a high protein

expression level, they have survived genome duplications

merely as single copy genes. In this class of terminal target

gene, records of past duplications are thus also lost.

Finally, the fourth class of terminal target gene is

exemplified by the olfactory receptor and immunoglobulin

gene families. These proteins interact with ligands of

environmental origin and potentially unlimited diversity.

Structural alterations in the absence of regulatory changes

allow new duplicates to interact with novel ligands. Newly

duplicated paralogs that interact with new ligands are

positively selected, which has caused expansion of these

gene families since the origin of vertebrates and erased

their records of early genome-wide events.

Other genome-wide events

Genome duplications at the origin of vertebrates in

evolution do not preclude the possibility that other events

may have occurred at or around the same time, such

as the de IIO’V’O generation of vertebrate-specific genes or

individual gene duplications or losses. The very tight

linkage of 1Cr!lf-5 and ;11&6 and of insulin and Igf-2. for

example, may be a remnant of ancient tandem duplications

[ 181 which could be an alternative to the earlier of the two

proposed genome duplications. A more comprehensive

description of the genomic events during the early

evolution of vertebrates will be achieved once more

sequences from cephalochordates (amphioxus), jawless

vertebrates (hagfish and lamprey), and cartilaginous fish

(sharks) become available. If, however, we accept the

evidence for two large-scale gene duplications at the origin

of vertebrates (Fig. Z), the question of what happened to

the genes since then presents itself.

Developmental software evolution: regulatory

changes After the genome duplications, some of the newly

duplicated structural genes-the genetic ‘hardware’ of

Sunday, 14 June 2009

Page 15: EVOLUTION - CBS · 2009. 6. 16. · past genome-wide evolution. Globins and lysozyme are representative of the first class; not being functionally constrained as much as fundamental

Technical University of Denmark - DTUDepartment of systems biology

CE

NT

ER

FOR

BIO

LOG

ICA

L SE

QU

EN

CE

AN

ALY

SIS

It has recently been demonstrated that this organism contains histocompatibility-relevant genes and lymphocyte immune signaling-relevant genes, i.e., components that may be recruited by adaptive immune processes.Fish & Shellfish Immunology, Volume 26, Issue 6, June 2009, Pages 843-849Zhenhui Liu, Lei Li, Hongyan Li, Shicui Zhang, Guangdong Ji and Yanling Sun

Sunday, 14 June 2009