Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
Zn vitro study of bone marrow derived progenitor
ceiis in liver-like microenvironments
Aüreza Khademhosseini B.A.Sc., Department of Chernical Engineering, University of Toronto
A thesis submitted in conformity wirb the requirements for the degree of Masters of Applied Sciences and Engineering
Department of Chernical Engineering and Applied Chemistry institute of BiornateriaIs and Biomedical Engineering
University of Toronto
AÉquisitiocls and Acquisitions et Bibliugraphïc Services seniicés bWiographiques
The author has granted a non- exclusive licence allowing the National Library of Canada to reprodpce, Ioan, disûi'bute or seII copies of this tfiesis in microform, papa or electronic formats.
The author reretains ownership of the copyrighî in this thesis. Neither the thesis nor substantial extracts h m it may be printed or otherwise reproduced without the author's permission,
L'auteur a accordé une licence non exchisive pemet~int ii la Bibliothèque nationale du Canada de reproduire, prêter* distribuer ou vendre des copies de cette thèse sous la forme de microfiche/film, de reproduction sur papier ou sur format électronique.
L'auteur comme la propriété du droit d'auteur qui protège cette thèse. Ni ia thése ni des extraits substantieis de celleci ne doivent être imprimés ou autrement reproduits sans son autorisation.
"ln vitro study of bone mamow derived progenitor tells in hqatic micmenvironments"
Alireza Khademhasseini, M.A.Sc. (2001) Department of Chernical Engineering and Institute of Biomaterials and Biomedical
Engineering, University of Toronto
Abstract
Published observations of the in vivo differentiation of h n e m m w (BM) derived ceiis
inio hepatocytes led us to study the mechanisms that regulate this differentiation in vitro.
Phenotypically defined (Lin-) BM populations were cultured in niicroenvironments that
mirnic parenchymal and nonparenchymai liver components. A significant difference in the
spectnim of ceUs generated fiom BM under various conditions was observed. Specifically,
emîched BM ceUs gave nse to endoderm-like celts in CO-culture with fibroblasts.
Lin- cells were shown to geneate liver-like ceUs that expressed CKS and albumin upon
exposure to high HGF concentrations. Also, individual hematopoietic stem cells (HSC - c-
kit' Lin* Sca-l+) were show to give rise to aibumin-secreiing ceIIs, in presence of HGF
whiie the addition of hematopoietic cytokines decreased the frequency of hepatic pmgeny.
These results suggest chat HGF is an important replator of HSC, which may act in inducing
HSC to give rise to hematopoietic or hepatic pmgeny.
"lferror is correcred whenever it is recogni'zed as mch, the path of error is the path of the tnuk "
- Hans Reichenbach (189'-1953)
"And ye shall h o w the tmh, and the cruth shsrll make you free. " - The Bible John 8:32
"Whut is now proved was once only UMgined" William Blake (1751-1827)
" The tragedy of the world is that those whn ure imaginative have but slight expetience, Md those who are expenenced have feeble imagination. Fools act on imagümion wirhout
knowledge, pedanrs act on knowledge withaut imagination The fask of a universùy is to weld togefher imagination rmd experience. " Alfred North Whitehead (1 86 1- 1947)
" When I was a boy of 14 my f&r was so ignorant I CU& hardly s t d to have the old man
arounà B u w k n I got tu be 2 1.1 was astonished by how much he had l e m in 7 years". This
quotation by Mark Twain seems accurate in descniing my nraniration and development as an
undergraduate and graduate snident at U of T.
For these experiencs 1 am indebted to my m p e ~ s o r s P d P.W. Zandsaa and R o t M.V. Sefton. 1
would like to thank Peter for taking on an arrogant and naive snident and m i n g him into a more
knowledgeable and mature (and still somewhat arrogant) cesearcher. 1 also would like to thank Peter for
allowing me to work on such a fascinating scientific area as stem cel1 bioengineering. Furthermom, 1
would like to thank my original mentor. Rof. M.V. Sefton for his guidance and mchings over the past
few years as well as providing me with resources to conduct independent research, and freedom to
explore my ideas and broaden my imagination. 1 have been blessed with excellent teachers and the goai
of having the same positive impact on my students will be a lifelong challenge.
1 would like to thank Dr. Bill Stanford for his guidance, and for giving me access to "glowing mice",
which was a key component of this project. Also, 1 am grateful to my thesis comminee members Drs.
Grant Allen and Jane A u b i for their time and critical assessrnent of this work.
1 have benefited greatly h m rny association with al1 past and present members of Sefton and
Zandstra lab, in particuiar Shahab Lahwti, Michael May. Kim Jones, Jennifer Vallabacka, Alison
McGuigan and Eric Tsang (in Sefton's lab) as well as Ger& MadIambayan, Sowmya Viswanathan,
Karen Chang. Celine Bauwens, EIaine Fok and Roni Dattani (in Zandstra's lab). A simple thank you is
insufficient for Ting Yin for her technical assistance rhroughout my stay at Zandstra's laboratory.
My graduate work has also been a learning experience in tearnwork and leadership in organizations
such as BESA, CEGSA and tissuetng.net. 1 wodd like CO chank ail staff and sntdents whose interactions
aided in my development into becoming a more complete individuai. h padcular, 1 would likc to
acknowledge Nanthan Yogachandran, leff Karp, Paul Dalton, and XuDong Cao for theü suppoh
1 ais0 wodd like to acknowledge the funding sources for Prof. Stfton and Prof. Zandstra as weU as
NSERC-PGSA for ptoviding the suppon for my graduate studies. Finally, 1 would like to thank my
family, friends and Yu San for their unconditional support and love.
TABLE OF CONTENTS
. . ABSTRACr ........................................................................................................................................ u
.............................................................................................................. ACKNORZEDGEMENTS iii
................................................................................... ..... *......-. ........ TABLE OF CONTENTS .. .. ,. iv ... ......*.*........*...... .............................................................................................. LIST OF TABLES ..- vu1
LIST OF FIGURES ................................ ,.,, ........................................................................................ ix .. ABBREVIATIONS ........................................................................................................................... xu
..................................................................................................... THESIS SCOPE AND FORMAT xv
....... 1 WTRODUC'IION AND BACKGROUND ...-,..... ....-.,....... -2
1.1 Stem ce11 therupy in rissue engineering ........................................................................................ 2
1.2 Sources of m m ceils ....................... ... ....................................................................................... 3
1.2.1 Ernbryonic stem cells ............................................................................................................. 3
................................................................................. ......... 1.2.2 Adult stem cells and plasticity .. 4
............................................... 1.3 Motivations for the use ojcell therupy in liver tissue engineering 6
.............................................................................. 1.4 Liver architecture. jûnctian Md regetterarion 7
1.4.1 Liver function and architecm .............................................................................................. 7
1 .4.2 Liver and bone rnarrow interactions ....................................................................................... 8
1.4.3 Liver regeneration .................................................................................................................. 9
1.4.4 Liver progenitor hierarchy .................................................................................................... 10
1.4.4.1 Srnall hepatocytes ....................................................................................................... 10
1.4.4.2 Oval cells ...................... ,. ...................... ....................................................... 11
1.4.4.3 Bone marrow denved stem cells ............................................................................... 11
1.4.5 Liver growth factors ............................................................................................................. 13
1.4.5.1 Hepatocyte growth factor (HGF) ................................................................................ 13
1.45.2 Transforming growth factor* RGFQ) ................................................................ 14
1.4.5.3 Thrombopoietin ('PO1 .................................................................................................. 15
1.5 Hematopoiesis a d hematopoietic stem cells .......................................................................... 15
............... .....*..... .......... ........*...............................................-...-.-..... 1.5.1 Hernatopoiesis .... , , ....,., 15 . . 1.5.2 Hematopoieuc ce11 culture ........................ ,.,..... ...................................................... 17
. 1.5.2.1 Stroma mediateci Iong-mm cdhues ................................+....................................... 17
1.5.2.2 Aiternative cultures ..- ............................. .. ..................................................................... 17
2 OBJECTIVES AND H Y P O ~ , , . - - - ~ m . z i c r r . r - ~ , ~ 1 9
.......................................................................... 3.1 Creananon of an in vitro liver microenvironment 20
.................*. ............................................................... 3.1.1 Hepatocyte isolation and culhue ........ 20
3.1.2 Ce11 Iines ............................................................................................................................... 21
3.1.2.1 AMLl2 .......................................................................................................................... 21
3.1.2.2 MMH and other murine iiver ce11 iines ......................................................................... 21
3.1.2.3 NM-3T3 ...................................................................................................................... 2 2
3.1.3 ECM Related Factors .. ........................................................................................................ î2
3.1.4 Medium requirernents .......................................................................................................... 23
3.2 Cell t r a c h g merhodr in vitro .................................................................................................... 23
3.2.1 YFP ...................................................................................................................................... 23
3.2.2 Membrane dyes and other techniques .................................................................................. 23
3.3 Analysis ....................................................................................................................................... 24 . . . .
3.3.1 Idenuficauon of hematopoieuc stem cells ............................................................................ 24
3.3.1.1 Functionai assays ........................................................................................................... 25
3.3.1.2 Phenotypic markers ....................................................................................................... 25
............................................................................................... 3.3.2 Identification of hepatic cells 27
3.3.2.1 Functional assays ....................................................................................................... 27
....................................................................................................... 3.3.2.2 Phenotypic markers 27
4.1 Bone mamw stem ceil extraction anà enrichmeru .................................................................... 29
4.1.1 Bone marrow isolation ......................................................................................................... 29
4.1.2 Stem ce11 e~chment ............................................................................................... 30
. 4.2 Primaq heparocyre isolaiion and culture ................................................................................ 30
4.3 ceil c h r e ............................................................................................................................ 31
4.3.L Ce11 lines ....................... ... .... ,. ................................................................. - ......--.. 31
43.2 Bone mamw cultures ...................................................................................... ......... 3k
43.3 Bone marrow ct~cultures with feeder cells .. ................................................................-...... 32
.......................................................................................................... 433.1 ~ a t r i ~ e l @ .....-....... 33
................................................................................ 43.3.2 Cytokines 33
43.3.3 CFSE staining ............................................................................................ .. ...............-. 33
* .. 4.4 Hemaroporenc d y s i s ............................................................................................................... 33 . . 4.4.1 Colony-forming ability ......................................................................................................... 33
4.4.2 Hematopoietic morphoiogical markers and anaiysis ............................................................ 34
4.5 Heparic analysis .................................................................................................................. ....... 34
4.5.1 Senun albumin detection ...................................................................................................... 34 45.2 Mïl'..................................................................................................................................... 35
4.5.3 Hepatic morphological markers and analysis ....................................................................... 35
............................................................................................................................ 4.6 Inwge analysis 36
.............................................................................................................................. 4.7 Data anaiysis 36
........................................................................ 5.1 Bone marrow Purificanon ond characrerizdon 37
5.2 Developmeni of the CO-culture systems ....................................................................................... 40
5.2.1 Rimary hepatocytes isolation and culture ............................................................................ 40
5.2.2 Hepatocyte and fibroblast characterizaiion .......................................................................... 42
5.23 Tracking of bone marrow derived tells in vitro ................................................................... 46
5.2.3.1 CFSE .............................................................................................................................. 46
5.2.3.2 Yellow fluorescent protein (YFP) ............................................................................ 47
........... ............ 5.3 Effecrs of in vitro cultures on hematopoieric pmgeniror ce11 fate .. .. ............ 49
5.3.1 Effects of AMLl2 md NIH-3T3 frbmblasrs on hematopoietic differentiation of BM ddved
ceils ...................................................................................................................................... 49
5.3.2 TPO and HGF alter bone manow progenitor ceIl fate ..................................................... 50 ........................................ 9.33 Effects of HGF supplementation on hematopoietic devetopment 52
5.4 Bone marrow derived endodenn differentianon ......................................................................... S4
5.4.1 Coculture of Lin cells with AML12 or 37'3 œlls ............................................................... 54
5.4.2 Effects of hepatocyte growih factor (HGF) on BM progeniton ......................................... 58
5.4.2.1 Exogenous HGF supplementation to cocuttures .......................................................... 58
5.4.22 HGF induces hepatocyte differentiation in a dose dependent manner .......................... -60 5.4.23 Frequency of differentiation and effects of hematopoietic cytokines on HGF mediated
generation of endodenn-like celts ................................................................................ 64
6.1 Development of the comlnrre system ...................... .., ...................................................... 68
6.2 Effects of in vitro culture conditions on hematopoietic pmgenitor ce11 fae ............................... 69
6.3 Bone marrow derived endodenn differentiation ........................................................................ 70
7 CONCLUSIONS AND RECOMMENDATIONS *m--m-".--.--n---nn7S
7.1 Conclusions .......................................................................................................................... 7 5
..... 7.2 Recommendations andjiture work ..................................................... .... 7 5
LIST OF TABLES
Table 1: Donor ce11 markers and their difiïculties in detection by flow cytomtry. .................................. 24
Table 2: Percentage of various hematopoietic markers in mononucleated WBM cells ............................. 37
Table 3: Perceniage of various Irematopoietic markers in mononucleated Lin- cells, after column
separation ....................................... , ....................... , .......... ...., ...... . ................................................ -38
Table 4: Expression of various hernatopietic and hepatic markers for primary hepatocytes. AML12,
NM-3T3 cells ................................................................................................................................ 45
Table 5: Percentage of CK8' YFP' ceIls detected in various CO-cultures using fluorescent microscopy
after 7 and 14 days. ............................................................................................................... 5 9
Table 6: Albumin secretion of KLS cetls plated individually in % well plates after 17 days in various
rnediums. ..............................................-.....-....-..................................,..................................-..... 65
viii
LIST OF FIGURES
Figure 1-1 A mode1 for stem cell differentiation .......................................................................................... 5
Figure 1-2 The surface of the iiver ................................................................................................................ 7
Figure 1-3 A single liver lobule .................................................................................................................... 7
Figure 1 4 The ontogeny of hematopoirtic progenitors ................................. .. ............................................. 9
Figure 1-5 Interactions between hepatic stem ceiis .................................................................................... 10
Figure 1-6 Hematopoietic differentiation hierarchy .............. ,., ................................................................... 16
................... Figure 2-1 Possible factors that regulate BM denved cell fate in hepatic microenvironments 19
Figure 3-1 Markers for murine HSC ........................................................................................................... 26
Figure 3-2 Markers for murine hepatocytes ................................................................................................ 26
Figure 4-1 Schematic of BM isolation process ........................................................................................... 29
Figure 5-1 CFC micrographs: CFU-GEMM (a), CFU-GM (b) and BFU-E (c) ......................................... 38
Figure 5-2 Phenotypic analysis for isolation of KLS cells and the ~stricted gates used to sort HSC ........ 39
... Figure 5-3 c-Met and CK8 expression of isotype control (a), L i (b), WBM (c) and AMLL2 (d) cells 40
Figure 5-4 Micrographs of prirnary hepatocyte cultures cultured on collagen after O (a). 7 (b) and 15 (b)
days ............................................................................................................................................. 41
Figure 5-5 Cell specific h4Ti' activity for isolated primary hepatocytes in various cuiture conditions ...... 41
Figure 5-6 Albumin secretion rate in various primary hepatocyte cultures ......................... ,. .................. 42
Figure 5-7 Micrograph of NiH-3T3 fibroblasts . .......................................
Figure 5-8 Micmgraph of AMLl2 hepatocytes .......................................................................................... 43
Figure 5-9 Typical CK8 expression of AML12 (a) . NIHOT3 (b), pnmary hepatocytes (c) and WBM cells
(d) ................................................................................................................................................ 44
Figure 5-10 Typical CK18 expression of AML12 (a) . NIH/3T3 (b) and freshly isalated WBM wlIs (c) . 44
Figure 5-1 1 Expression of Ecadhenn (a) and c-Met (b) on AMLL2 cells ................................................. 45
Figure 5-12 CFSE stained WBM (center) and unstained AMLl2 (left) cells on day O were distinguishable
based on CFSE expression; however. after 7 days in coculture, the IWO popdations were . . . . tndistingushable (nght) ............................................................................................................. 46
Figure 5-13 YFP expression of CRI-GM h m YFP* BM cells in CFC assay .......................................... 47
Figure 5-14 BM denved cells from YFP* mice had distinguishably higher YEP fluorescence than AMLI2
............... ceiis .......................................................................................................................... ... 48
Figure 5-15 BM denved ceU numbers in various cocultures chat were initiated with 5xid b ceiis .... -50
Figure 5-16 Effects of CO-culture on the frequency of CFCs to total cells in cdtures that were initiateci . . with 5 x l d Lin celis . ................................................................................................................. 50
Figure 5-17 Effects of CO-culture condition on CFC types, 7 days after culture initiation ......................... 50
F i p 5-18 Effects of TPO or HGF supplementation on BM ce11 counts in cultures initiated wiih 5 x 1d . - Lm cells ...................................................................................................................................... 51
Figure 5-19 Effects of TPO and HGF on the hquency of CFCs compared to total cells in BM cultures
initiated with 5 x i d Li cells. ................................ . .... , ................................................. 5 1
Figure 5-20 Effects of HGF and TPO supplementation on BM derived CFC Types in cuItures initiated
with 5 x l d Li- cells after 7 days .......................................................................................... 52
Figure 5-21 Effects of HGF supplementation on YFP' ceIl counts in cultures initiated with 5 x l d Lin'
celis. .................................................................................................................... .. ..................... 53
Figure 5-22 Effects of HGF suppiementation on ihe frequency of CFC colonies to total cells at any point
in time for culture^ initiated with 5 x id Lin' celis ....................................... ,. ............. 53
Figure 5-23 Effects of HGF supplementation on CFC types in cultures initiated with 5 x l d ceUs at day 7.
................... ......... ..*+......................................................+............. ** ................ . - . - . - . 5 3
Figure 5-24 Expression of CK8KK18 and YFP on NM-3T3 (a), AMLl2 (b), and Lin' cells (d) ............. 54
Figure 5-25 Doublet discrirninator was used to minirnize aggregates in flow cytomecric analysis ............ 55
Figure 5-26 CK8 and YFP expression of Lin' cells a k r two days in CO-culture with AMLl2 cells ....,.... 55
Figure 5-27 CK8 and YFP expression of LiF ceil in coculture with AML12 cells, 2 days (a), 7 days (6) * *
and 15 days (c) after culture iniuauon. ....................................................................................... 56
Figure 5-28 CK8 and YFP expression of Lin' cells cultured without direct ce11 contact with AML12
feeder layers, 2 days (a), 7 days (b) and 15 days (c) after culture initiation ....................... . ....... 56
Figure 5-29 Flow cytomeuic expression of CK8 and YFP oii BM derived Lin*(a-b) and Lin- (cd] ceils in
CO-culture with NIH-3T3 cells. ..+...............................-..--....--.-.-......-............... . .........-...-..-... 58
Figure 5-30 The presence of individual cells expressing CK8, after 7 days of CO-culture with NM-3T3
and AhlL12 cultures supplemented with HGF (lhdrnl) was c o n f d using fluorescent
rnicrographs. ...........................................................,................................................................,.. 59
Figure 5-3 1 CK 8 expression of Lin' cells cultured with 3T3 fibroblasts in absence (a) or presence (b) of
IO ng/mI HGF seven days after initiation of the cultures. ........................... ,. ........................ 6û Figure 5-32 Aibumin secretion h m Lin cells coculNted with NM-3T3 feeder cells ............................. 6û Figure 5-33 Culture conditions used in expeciments to determine effects of HGF concenaatioa .............. 61
Figure 5-34 Flow cytomtric analysis of CKS and c-met expression on Lin' BM cultures supplemented
with exogenous HGF after6 and 16 days ...-...............-...--W..-.-... ... ............. . ...... - .... .......-.A2
Figure 5-35 Effects of HGF concentration on albirmin expression in BM cultures initiated with 5 x ld . - Lin celIs ................... .. ...............................--.----.-----.-----..--.....+-........+....-.---...-..............-.--.------- 63
Figure 5-36 Morphologicai feanrres of AMLI2 hepatocytes cultured on ~ a t r i ~ e l ' in cornparison with
(a)HGF (LW ndmi) supplemented BM cdnrres (b). ........................................................... 63
Figure 5-37 Albumin secretion h m various controis including various media and Li- CD45ceiis. ...... 65
Figure 5-38 Albumin senetion after 17 days in 1000 nglml HGF (HGF) with addition of 1000 LinXDlS
(HGFCD45), or 1 (HGFL), 10 (HGF10) or 100 (HGF100) KLS cells to each well ................... 66 Figure 5-39 Albumin secretion a k r 17 days in cultures initiated with HM and 1000 ngimi HGF
(HGFHM) with addition of 1 (HGFHML), 10 (HGFHMLO) or 100 (HGFHM100) KLS cells CO
each weiI. .................................................................................................................................... 67
Figure 6-1 Schematic of soluble factors that may influence HSC ceIl fate in liver-like microenvironments.
2 - M F
7-AAD
AFP
AS cells
$-gai
BEC
3M
BS A
CO?,
CFC
CFSE
CFU-E
CFU-GEMM
OU-GM
CFUS
CK
C'LE'
CMP
CSF
D DMSO
E
ECM
EGF
ELISA
Epo
ES
FAH
F ACS
FBS
FlSH
2-acety laminofluorene
7-aminoactinomycin D
aipha-fetoprotein
aduIt stem cells
p-galactosidase
biliary epirhelial ce11
bone marrow
bovine serum albumia
carbon dioxide
colony forming ce11
carboxyfluorescein diacetate succinimidyl ester
cotony forming unit - erythroid
colony forming unit - granulocyte, erythroid, monocyu,
J=gakaryow colony forming unit - pulocyte, monocytes
colony forming unit - spleen cytokeratin
common Iymphoid prog~oitor
commoa myeloid pcogenitor
coIony stimuiating factor
daltons
dimethyi sulfoxide
embryonic day
exaaceIlular rnaaix
epidermal growth factor
enzyme linked immmo(~ss~~y
eryttuopoietin
embryonic stem
fumarylacetoacetate hydrolase
fluorescence actiwed ceIl sortiag
ferai bovine serum
fluorescent in situ hybridization
mc FGF
FL
GGT
GI
HBSS
HGF
HM
HRP
HSC
IF
EMDM
IL
KLS
LIF
Lin'
LT-HSC
LTBMC
LTC
LTC-IC
MEM
m InLd
NM
NSC
OSM
MW
PBS
P a
PDGF
PE
PEG
PH
PI
PP='
fluorescein
fibmblast gcowth factor
FLT-3 ligand
gamma-glutamyl traaspeptidase
gastrointestinal
Hank's balanced saIt soiution
heptocyte growth factor
hematopoietic medium
horseradish pmxidase
bernatopoietic stem cells
intermediate Filament
kove's Modified Dulbecco's Medium
interleukin
Lin' Sca-l* c-kit'
leukernia inhibitory factor
Lincage depIeted
long-tenn hematopoietic stem cells
long-tenn bone m w culture
long-term culture
long-tenn culture initiating ce11
minimum essentid medium
3-(45dimethyi-thiazol-2-y1}-2~diphenyt bromide
milliliter
Nationai hstitute of Health
neural stem ceils
oncosutin M
molecular weight
phosphate-buffered saline
cyanine 5
pIateIet denved growth factor
phycoerithin
polyethylene giycol
partial hepatectomy
pmpedium idide
parts per million
RBC
rh
ml
Sca-1
SCF
SP cells
sr-HSC
TPO
TGF
TNF
TMB
USA
WBM
YFP
red blood ceiis
recombinant human
recombinant mitrine
stem ce11 antigen - 1
stem ceil factor @Ca. steel factor or c-kit Ligand)
side popdation cells
short-temi hematopoietic stem cells
thrombopoietin
transforming growth factor
tumor necrosis factor
33JS-tetramethyl benzidine
United States of America
whole bone marrow
yellow fluorescent pmtein
xiv
THESIS SCOPE AMI FORMAT
The behavior of bone marrow (BM) derived progenitor cells in hepatic microenvironrnents was
investigated. With the underlying premise that BM derived cells exist in liver and have the capability to
reconstitute hematopoiesis and generate functional iiver cells, the objective of this project was to
develop a culture system to invesügate the e fk î s of various microenvironmental parameters on
hematopoietic and hepatic differentiation of SM derived progenitors under controiied conditions.
As a first approach to study this behavior, liver-like microenvironments were developed using primary
hepatocytes as well as hepatocyte and fibroblastic ce11 lines. The ability of various BM populations in
these cultures was evaluated in tenns of their viabiIity, proliferation, and hematopoietic and hepatic gene
expression. The results from these experiments provided the basis for a more mechanistic study of
specific signais which regulate BM derived ceIl fate in liver microenvironments. More specifically, the
effects of TPO and HGF were investigated.
The thesis is divided into the following sections. Chapter 1 provides the background for the work
and introduces the previous research regarding various aspects of the project. Chapter 2 outlines the
hypotheses and objectives of this thesis. Chapter 3 provides the rationale and the approach taken in
ordet to achieve the objectives of the work. Chppter 4 provides a detailed account of the procedutes and
rnethods used in developing the cocuiture system in addition. a full description of the analytical
procedures is provided. These detailed instructions are intended to assist readers in performing the
experiments described in this thesis. Chapter 5 outIines the results obtained in this work in four parts. In
pan 1, different populations of BM derived progenitors are characterized. In part 2, the results regarding
the developrnent of the culture system and the dificulties encountered with their senip are explained. In
part 3, the results regarding the hematopoietic maincenance in various cuitures are andyzed, while in part
4, the endoderm differentiation of BM denved ceils is mdied. Chapter 6 provides a discussion of the
results of the project Chapter 7 provides an outiine of the conclusions and recommends d i t i o n s for
future work.
1 iNTRODUCTION AND BACKGROUND
1.1 Stem ceii therapy in tissue engineering
Despite attempts to encourage organ donationsI'2, there is a shortage of transplantable human
tissues such as bone marrow (BM), hearts. kidneys. Livers and lungs. For example, cunently more than
74000 patients in the United States (USA) are awaiting organ transplantation, while oniy 21000 people
receive transplants annually3. This gap between eligible patients and availabie organs is growing
constantlfl. It is from these motivations that tissue engineering has emerged.
Tissue engineering is an interdisciplinary fietd that applies the principles of biology and
engineering to the development of viable substitutes, typically composed of biological and synthetic
components which restore, maintain, or improve the function of human tissues5.6. This generally
differs h m standard drug thenipy in that the engineered tissue becomes intepted within the patient,
providing a potentially permanent and specific cure for the diseased state.
One approach currently under devebpment to alleviate human tissue shortage is
xenotransplantation. Xenomsplantation. despite its ethical and immunological considerations2 may
overcome the cunent shortage of organ donors. To prevent rejection of xenotransplants, immune
reaction to transplanted organs has been minimized using inimunasuppressive drugs (Le. cyclosporine)
or genetic modification of donor specific antigens. However, significant challenges such as
carcinogenic and toxic effects of immunosuppressive drugs7 and cross-species viral infections8 still
remain. These side effects have encouraged mearchers to search for alternative methods of ceIl
delivery.
An emerging therapeutic approach to overcome the short supply of transplantable cells is the use of
stem cells. Stem cells are cells that are capable of exrended self-renewai as weii as the ability to give
rise to many functionai tissue types (i.e. differentiate into multiple lineages). Because stem cells can
differentiate and give rise, typicalty in a cascade of progressive maturation steps to progeny with the
functional and genetic pmperties of mature cells, they reprwent a poiential source for tissue and ceUular
engineering. For example, adult or embryonic stem ceils may one day be used to generate
cardiomyocytes for cardiac muscle regeneration, pancfeatic islet ceils for diabetes, liver cells for
hepatitis, and neural cells for Parkinson's or klzheimer's diseases.
It is known that highly differentiated ceIls can ody divide a limiteci number of ~ i e s ~ , ~ o . Stem
ceiis are advantageous in regenerative medicine because of their capabiity to self-renew (Le. give rise
to cells with the same potential and genetic expression as the original ceiis). As evident by the
developmental potential of embryoderived cells or by the abüity of hematopoietic stem celIs (HSC) to
reconstitute primary and secondary hosts' hematopoietic system in BM transplants, a smaii number of
stem cells can produce a therapeutically signifiant amount of tissue. Due ro these properties. stem
cells can be applied to regenerate tissues directly, such as in BM transplantation, or in coajunctian with
tissue engineering consmcts. thus providing a method for in vivo tissue engineeringll. Frathermore,
stem cells are suitable gene therapy vehicles (since the gene product may be expresse. in ail cheiu
progeny), making it possible to achieve a life long correction for a s p i f i c disorder.
It has k n estimateci that in the USA aione, over 128 million people cm be helped by the resulting
benefits fmm stem ce11 researchlz. However, despis this promise. therapeutic use of stem cells has
been limited because of our limited knowledge of the microenvironmental factors which affect stem cc11
differentiation and self-renewal. Clearly, in order to utilize stem cells in clinically relevant
applications, an improved understanding of the miroenvironmentai cues that govern their behavior is
required. Therefore, M e r research into understanding stem ce11 microenvironments will be of grwit
thecapeutic value by providing an ample supply of cells for transplantation.
1.2 Sources of stem cells
Stem celis can be divided into two distinct groups: embryonic stem (ES) cells and adult stem (AS)
ceilsL3. ES cetls are pluripotent (Le. capable of generating ail differentiated celis of the body) cells that
can be denved from the inner ce11 mass of the developing blastocysts14*L5. In contras& AS cells are
multipotent cells that reside in adult tissues and are believed to be important in tissue regeneration and
maintenance.
1.2.1 Embryonic stem cells
Although rodent ES cells have existed for many years14*15, the interest in the clinical use of
embryo-derived cells was ignited by the denvation of human ES c e 1 l s 1 ~ ~ ~ . ES ~ l l s can be geaetically
modifie& produced in large numbers, and differentiated prior to use in regenerative medicine.
Excitement over the use of ES cells has increased with their demoastcated ability to give rise to
hemat0~oie t ic~~~~0, end~thel ial~~, c a r d i a c 2 ~ . neud4 , osteogenic2S, hepatic26 and pancrrotiS7
tissues in vitro. However, despite their therapeutic potential, ES cells present a number of chaIIenges
associated with their clinical application. For example, ES œlls may have immunologid
incompaa3ility with recipients. To overcome immune rejection, ES c e k wouid rquire a cloning step
in which the nucleus of a somatic ceU h m the cecipient is t ransfed ta the donor ceUs in order to
overcome the possibiiity of tissue rejection. However, this step may be ümited by the availabüity of
human eggs and the possible generation of abnormd embcyos2s h m the inerinsic genetic instabüity of
cloned cel~s*~. Furthemore. transplanteci ES cells may form nimors; thus, great care must be taken KI
ensure that ail cells are fully differentiated. Also, more resesrch is qu ired to uaderstaad the signais
thac direct ES ceIl differentiation into various lineages.
One of the major challenges in the wide spread use of h m ES cells is due to bio-ethical
coniroversy in using cells that q u k e the destniction of embryos. Furthermore, despite enormous
patient support12, ES ceIl research has encountered significant mistance30131 by people who oppose
their derivation h m human embryos for fear chat their use will lead to the neation of geneticdly
engineered hurnan beings32. Due to these difficulties. dong with h e increased porentiai of AS cell
research described in sections 1.4.2 and 1.22. some legislators oppose ES ceIl research. However, the
rnajoity of scientists advocate ES ce11 as weH as AS ceIl research33.
1.2.2 Adult stem cells and plasticisy
in conmt to theu embryonic counterparts. AS cell therapy has k e n used in the form of BM
transplantation for many years. AS ce11 research dœs not have the ethical issues associated with che use
of embryos since they can be derived from the patient and thus have the potential to be histocompatible
with the patient. Also, AS cells may not form tumars as readily in patients, as BM aaospIautations
have demonstraced. However, the use of AS cells was rhought to be lirnited since &y have been
isolated h m relativeiy few tissues and have been considend to generate only a m w spectrum of cell
types-
Traditional dopas regarding stem ceII differentiation held that organ specific stem cells are
resuicted to generate differentiated cell types h m the tissues in which they reside. Thus. they wen
tfiought to be much more restricied in their developmental capabilities than ES cells. This dogma was
supported by embryological expiments in which cells of undifferentiatcd tissue transplanted h m onc
region of the embryo to another, quicidy lost cheir ability to becorne part of the new tissue13. However,
new evidence suggests t h t AS cells have much greater regenerative capabiiity than previously
believed.
The f i t demonsuation chat AS cells rnay be las lineage restricted chan previously thought was
show with the in vivo derivation of rnyogenic pmgenitors h m whole bone marrow ( w B M ) ~ ~ celis;
as weli as the in vitro differentiation of BM smma cells into conmctile myotubes35 and
cardiomyocytes36. In addition, muscle ceih were shown tu give rise to hematopoietic ce11s~~. Further
experhents revea1e.d rhat enriched HSC were capable of migraihg to muscle38 and differentiating into
muscle-tike ce1ls3~ (however. despite initiai opthni&, cecent resuIo have questioned the fuactid
capability of BM derived musde fibers41). Furthemore, Bjomson et al demomtntted that nemice11
cultures containhg neural stem cells. which were thought to be Iimited in potential to the development
of neuroas, oligodendrocytes and asmytes, could differentiate into blood c e ~ l s ~ ~ . Moreover, BM
denved cells differentiate into astroglia and microglia following injection into the brain 43 and adherent
BM stroma have been shown to give rise to neurons and glia44.
The most convincing evidence regardmg the mdti-tissue regenerating potential of AS ce& is in the
ability of HSC to give rise to ceIIs of other tissues such as hepatocytes or other epithelial ceUs (see
Section 1.4.4.3). Definitive proof of the multiorgan generating capability of BM denved stem cells
was recently demonstrated by Krause et fi, who showed that a single cell, which homed to the BM.
codd reconstitute hematopoiesis in primary and secondary recipients as well as differentiate into
epithelial ceHs of the liver, lung, intestine and slrin.
Researchers have thus begun to question the classical germ layer ongin of tissues. An emerging
mode146 for the hierarchal structure of stem cells is shown in Figure 1-146. in this model, early
embryonic development is initiated h m undifferentiated totipotent cells. These totipocent stem ceUs
give cise to pluripotent ES cells, which generate the three germ layers and perhaps putative somatic
stem cells. The somatic stem ceils are mdtiporent stem cells, and tissue-restricted stem cells may derive
from multipotent stem cells. Hematopoietic stem cells (HSC), muscle stem cells (MuscleSC), and
neurai stem (CNS-SC) cells may be multipotent stem cells or tissue-restricted stem ceils h t
dedifferentiate into muItipotent stem cells.
Figure 1-1 A mode1 for stem ecll ciUterentiPtion
Despite numerous scientific Papen on the subject, AS cell transdiffe~entiation or mdti-otgan
regenerating capabdity bas not yet been clearly unders td Most papers have failed to quantify the
5
degree of transdifferentiation or to evaluate functionaiicy of the hanspIanted cells in the h t .
Furthemore, most literattm has failed to identifj the progenitor or the importance of Ut vitro c u i w g
prior to transpIantation. It is not cIear whether AS cells transdifferentiate through gene repmgramming
when exposed to the novel microenvironment or if there exists a primitive adult cell ihai has signifiant
multi-tissue regenerating capability. Studies that demonstrate transdifferentiation without cdture
manipulation suggest that the tissue environment is a critical factor. Therefore. the muon for the
inability of specific AS cells to give rise to certain tissues m y be because of their failure to access them
iiz vivo47. if this is the case, then U1 vitro culture systems which mimic in vivo conditions (as developed
in this thesis) as well as genetic modifications leading to receptor expression which preferentiaily lead
to Iocalization in a specific microenvironment may be critical in developing future ceIl based
regenerative therapies.
However, despite uncertainties regarding the inmnsic potential of AS cells, the ability of cells of
one tissue to give rise to differentiated cells of another tissue will be of great clinical significance. This
ability rnay allow for the expansion of AS cells of one tissue pnor to differentiation into cells of a
different tissue, which may overcome much of the shortcornhg and cnticism associami with
difficulties in expansion of certain AS cells in vitro. Furthennoce, this ability may provide alternative
AS ceil sources that are easier to access (i.e. peripheral blood or fatderived stem cetls) as substitum
for traditional sources (i.e. BM derived).
1 3 Motivations for the use of ceii therapy in iiver tissue engineering
The physiologiml functions of the liver include bile production, de-amination of amino acids and
maintenance of constant blood glucose concenmtion as well as metabolism of proteins (such as ures).
The Iivcr also synthesizes a number of plasma proteins such as albumin and coagulation factors; it
detoxifies toxic substances using enzymes (such as alcohol dehydrogenase or cytochrome P (CYP)
enzymes) and provides storage for iron and vitamins A, D and B12. In addition. Kupffer ceIls (liver
macrophages) eliminate damaged erythrocytes. With these properties, the iiver is critical to health and
its functional 1 0 s wilI quickly lead to multiple organ failure.
There are over 200,000 patients in the US who are suffenng h m fiver diseases such as liver
cancer, cirrtiosis and metabolic disorders6. Histoncaiiy, technologies that have ainied to replace iiver
function have focused on removal of toxins through means such as dialysis, immobilized enymes, or
exchange transfusions48. However, these methods do not provide the patient with the full tixnctionality
of a healthy liver. In order to regain the fuII potential of a fimctionat iiver, hepatoqes mut replace the
Iost organ. AIthough a number of ceil transplantation devices have been developed to mat Iiver failine,
such as hepatocyte microencapsulation and artifidal tiver de~ices6*48-5~, the shorcage of tninspiantablt
hepatocytes remains a significant barrier to the therapeutic applications of Liver tissue engbeered
dences.
The difficulty in obtaining transplantable hepatocytes has been enhanced by the obstacles
encountered in expandïng hepatocytes in vim. Stem ceils have been proposed as an ideal source fw
generating functiooal hepatocytes. The use of AS cells is particularly inviting for liver therapy since,
for chronic or acute liver failure, a temporary aeamient with artificiai liver devices provides sufficient
lag time to grow transplantable liver tissue in vino.
1.4 Liver architecture, function and regeneration
1.4.1 Liver function and architecture
The liver, which lies in the upper right abdominal cavity beneath the diaphragm, is the largest
viscerd organ and the largest gland in the body. It weighs rpproximately 1.5 kg in humans and 1-2 g in
mice. Because it is functionaily located between the gaswintestinal (GI) tract and the heart, the liver
has a dud blood supply. It receives oxygenated blood from the general circulation via the hepatic
artery (25%) and a Iarger voiume of poorly oxygenated blood draining the GI tract, stomach, pancreas,
spleen. and small and large intestines via the portal vein (75%). The hepatic tissue is a highly
vasculari2ed tissue in which hepatocytes (which maice up the parenchymal portion of the Iiver) provide
the body with diverse metabolic functions, whiie nonparenchymal cells usuaily serve structural and
regdatory prirposes. The liver is cornprisecl of 3 major lobes (see Figure 1-252). which consist of
hexagonal columns of hepatocytes arranged radiaiiy h m the cenual vein (set Figure 1-352). The
hepatocyte columns are covered by sinusoids on eoch side.
Figure 1-2 The sprtpce of the liver Figure 1-3 A sin& Uver bbuk
The liver contains two major diffcrentiaied endodem ceU types: hepatocytes located in the hepatic
parenchymal plates and biiiary epittieiiai ceiis locattd in the bile ducts. The most abundant and
functionally important cells within the liver are the hepatocytes, which account for 95% of the liver
mass and 60% of the total liver ce11 numbed3. Hepatocytes are large (18-30 pm). polyhedrai cells
that are rich in organelles such as rough endoplnsmic reticulum, smooth endoplasmic reticulum,
mitochondria and Golgi complexes which are reflective of their functional capabilities54.
Besides hepatocytes. the manunalian liver is composai of biliary epithelial cetIs (BEC or
cholangiocytes), as well as severai celi types presumed to be of mesenchymal origin including
endothelial celis, Kupffer cetls (macrophages in hepasic sinusoids) and stellate cells (or Ito celh - unique to the liver and located under the sinusoids where they store vitamin A, synthesize comective
tissue proteins and secrete several growth factors). The liver also contains nerve ceUs and cells
entrained in periusing blood. The majority of these cells are associated with hepatic sinusoids, which
are large, irregular shaped capillaries. These cells are separated h m the hepatocytes by a n m w space
within which stellate cells lie (thus controiling sinusoid lumen aperture). Cholangiocytes are the most
abundant cells next to hepatocytes- They constitute abaut 5% of the normal liver population and
occupy no more than 1% of the total Iiver mass and are typically much smaller than hepatocytes (-10
um)55- Kuppfer cells reside within the lumen of the sinusoids. The sinusoirial Iining is comprised of
endothelial cells (-48% of total sinusoid cells), Kupffer cells (-29%) and scellate cells (-20%)56-
1.4.2 Liver and bone m w o w interactions
Throughout embryonic devclopment, the= is a very close association between the fetal liver and
hematopoiesis (see Figure 1 .4~) . In mice. primitive hematopoiesis occurs within the blood islands of
the yolk sac at embryonic day 7 0. Blood islands fuse and establish blood vessels and circulation,
and with it hematopoiesis shifts h m embryonic yok sac ta fetal hersa at El 1 after the emrgence of
the liver bud h m the endoderm lining of the v e n d fongutS9. The feral liver contains hepatoblasts,
the bipotential progenitors60*61 that express aIpha-fetoprotein (AFP), cytokeratin 8 (CK8). cytokeratin
18 (CK18) and gamma-gIutamyl transpeptidase (GGT - a protein associated with most fetal epitheliai
cells in the liver), and subsequendy become nsuicted to choIangiocytes or hepaiocytes pas mat al^#^.
Figure 1-4 The onhgeny of hemtopoktk pmgeniînrs
in feral liver development, hemaropoietic ceUs induce the differentiation of hepatoblasts into
hepatocytes by secreting Oncostath M (osM)~~. Aiso Fibroblast growh factor 8 (FGF8), ptoduced by
the liver mesoderm tissue. conmbuces to the marphogeaic outgrowth of the hepatic endoderm@. The
fetai liver remains the major hematopoietic organ during nid and late gestation. At Eï5, shortiy pnor to
birth, hematopoiesis begins CO shift to spleen and BM. The hepatocytes assume k i r polarizect definitive
differentiation oniy just before birth65.
The Liver and BM interaction continues throughout adulthood. It has been shown that rare Lin c-
kit' Sfa-L' murine cells with hematopoietic repopulating ability reside in the l i v e # ~ ~ ~ . Furthmore.
adult Iiver can becorne the site of hematopoiesis during myelopmliferative disease such as
myelofibrosisa8.
In the pmcess of normal assue renewai in the liver. only one in tûtiNMWû hepatocyus are
dividing at any one tirrse69-71. Although the liver is a quiescent organ, t e m h i l y differentiated
hepatocytes are able to undergo active prdiferation a b ce11 loss72. In fact, division of mature ceik is
the mon common pathway utilized in the 'neofonnuion' of liver parenchymal cellJ3. For -p[e in
2l3 partiai hepatectomy. a mjor portion of the Iiver mass can be surgicaily removecl without pemaimt
damage- In this process. the mature hepatocytes are first activatecl, foiiowed by a delayeci proliferation
of cholangiacytes and sinusoiciai ceiis. This regmerative capability has even bccn mentiand in Gteek
mythalogy Ïn which Prometheus's liver was subjected b daily parrial hepatectomy by a bird after he
siole the secret of tüe h m the gods and passed it to humans74*75-
In n o r d Iiver qenerarion, 'unipotential" hepatocytes as weii as al1 0 t h cellular populations
within the tiver can protifmte to replenish îhe iost tissue massz76. Usuaiiy. these ceils have the
capabiiity CO regenerate the liver. For exampIe, hepatocytes have been s h to be capable of
9
undergoing 12-16 ce11 divisions77, which would normaiiy suff~ce for regenerating the lost tissue.
However, uader conditions of severe damage or impaired hepatocyte division due to toxic injury (e.g.
chernical carcinogenesis), the stem ce11 cornpartment of the liver is activated.
1.4.4 Liver progenitor hierarchy
The hematopoietic cells. or other ceUs such as skin, rely on the participation of stem cells in theu
cenewai ptocess. Although the liver is different h m these systems, in that it is composed of highiy
differeatiated cells which retain replicative potentiai, the existence of liver stem cells has been
reported78t79. However. until tecently, the purification of liver stem cells capable of multilineage
differentiation into hepatic and biliary cells had eluded researchen. With recent evidence, a picture of
liver stem ce11 hierarchy is emerging, compnsed of BM decived stem cells, ovai cells and small
hepatocytes 68.
Figure 1-5 Interactiom between hepatic stem œlk
Pathways for derivation of various hepatic ccllo are indicated in embryonic (bhe m w s ) aad lrdult tissues (orange), (dotted m m indkaîe pmthways w b k h bave aot beeo crinduslveiy proven)68. Ei-CFU-C (c-kitïCD45/TERI19') are dis lsoiated fmm murine fetai Kver rhich express kpatic and biïe duct markers and were recentiy introduced 9s hepatic stem cellsaO.
1.4.4.1 SmaU hepatocytes
Small hepatocytes are a population of celis initiaily isolateci fmm rat liver81 and mbsequently
shown to cause clona1 expansion of hepatocytes in humans82 Tbese cells comprise 1-296 of
hepatocytes and are able to undergo 5 6 ce11 divisions within a few days and yet retain hepatocyte
phenotype in vitro. The precise origin or location within the Iiver of these cells has not been definai. ii
has been proposeci chat ovai cells (the more primitive tiver cells) give rise to d l hepatocytes 68.
Fitrthermore, since these cells have not been shown to give rise to cholangiocytes UI vivo, they seem to
be the direct precursors of hepatocytes81*83. However, a ce11 fraction of smaii hepatocytes, grown
clonally in culture, has b e n shown to give rise to biliary epithelial cel1s84. which leaves k passib'ity
that these cells may be able to give rise to cholangiocytes under certain conditions.
Oval ceils
Until recently, oval cells were thought to be the most likely candidate for iiver stem ce11s83*85.
Oval celIs appear in the liver of duit animais when the liver damage is so severe that hepatocytes
cannot proliferate or are sornehow prevented h m proliferating (by either exposure to hepatotoxins or
carifinogens alone or combined with other surgical or dietary regimens) 60*79,86*87. They reside in a
heterogeneous population of nonparenchymal cells and are known by the morphologicd descriptor
"oval celis". Ovai cells are believed to originate h m the canais of ~errin$**6 (that is the region
where cells are transitional between the periportal hepatocytes and the biliary cells Iining the smalIest
terminal bile ducts) or blast Iike ceiis Iocated next to bile ductsg8.
Oval cells are thought to have both clonogenic and bipotential capacity to proliferate and
differentiate into both hepatocytes and biliary epithelial cells89*90. Interestingly, there is evidence that
under certain conditions oval cells can be induced to differentiate into non-hepatic lineages including
intestinal and pancreatic epithelium79 and mature cardiomyocytes in adult hem tissue9 l.
Oval cells share numerous markers with fetai hepatocytes, hcluding the expression of AFP,
albumin. GGT, and the pattern of cytokeratin expression (CK8, CK18, CK19). Oval ceils also express
rnany phenotypicai rnarkers that are usuaily associated with HSC In rodents. ovaI celis express
C D ~ @ ~ , 'hy-193 (a marker that is present on rat HSC) as well as c-kitg4 mRNA and proteins. In
humans, oval cells and HSC share CD3495 and c-kitg6. The presence of such markers was the first
evidence t h t hepatic stem cells may in fact originate h m the BM.
1.4.4.3 Bone rnarrow derived stem ceils
Recent papers offer a new explanation for the location of liver stem cells. B a d on evidence that a
number of surface markers are shared between HSC and oval ceiis, Peterson et a l dernonsaaied that BM
derived ceils have the potential to give rise to ovaI cells and to M e r differentiate into hepatocytes
andlor cho1angiocytes8*. By transplantmg rat BM into IethaIIy irradiated recipients and faIIowing the
fate of syngeneic ceUs using various markers, they showed saiking changes in the Iivers mduced ta
regenerate using 2-acetylaminofluorene (2-AAF) creatment (to block hepatocyte differentiation)
foliowed by partial hepatectomy (to induce oval ceii regeneration). In these tivers, evidence was
presented to suggest chat the donor ceIls migrated into the liver of cecipient animaln and subsequently
undement differentiation to become hepatocytes, though it was less clear whether bile duct ceih
developed.
A second published study used a similar approach but without a liver injury stepg7. They observeci
that after lethally irradiating fernale mice and transpianting male donor BM cells, up to 2.2% of toial
hepatocytes were identified, (by simuitaneousiy expressing Y chromosome and albumin mRNA) as
donor derived. Since these irradiated mice had no demonsaable acute hepatic injury such as
inflammation, necrosis, oval cell proliferation or scarring, it was concluded that BM derived hepatic
differentiation could occur in absence of Iiver injury. In both studies, the number of cells which
undergo the transition appears to be relativeiy small. In addition to rats and mice, generation of BM
derived hepatocytes has also been dernonstrated in hurnan patients. In these results, the number of BM
derived hepatocytes and cholangiocytes ranges h m 443% and 448% ~ t i v e ~ ~ ~ ~ ~ ~ ~ .
The experiments rnentioned above mostly used WBM populations and left many questions
unanswered regarding the identity of the BM population capable of differentiating into hepatic lineages.
The identity of this ce11 was revealed in a series of expehents. Theisc et al dernonstrated that 200
CD34Zin- cells were capable of giving rise to functional hepatocytes (-0.2-058 of total hepatocytes)
after 8 months97. However, the mort definitive demonstration of BM to liver mosplanmion was
perfonned by Lagasse et ai who showed that when a small number of c-lric!'@ thy" Lin Sca-l* ceils
(believed to be greatly enriched in HSC) was injected intravenously into IethalIy irradiami
fumacylacetoacetate hydrolase (FAH?) deficient mice (a mode1 animal which shows progressive liver
failure), the mice were rescued and the biochemicai hutction of liver was restored. At seven months
p s t transplantation, more than 30% of the hepatocytes were shown to be donor derivcdi~.
Furthemore, the ability of cells to perfonn non-bematopoietic reconstitution was liited CO EBC. In
experiments in which a single HSC was aanspIruited into irradiateci hosts, donor derived cells which
morphologically resembled hepatocytes and cholangiocyte& have been obscrved. Unfortunateiy, the
antiaodies that were used do not stain Iiver c y t o ~ l O 1 , thus a definitive demonsmtion of this
differentiation was not possible.
It is nat yet detennined whether HSC or their progeny seed and en@ to the liverlOO. In vivo
experiments have dernonstrated two distinct modes of hcpatocyte engrafmient9? which may suggest
distinct mechanisms of growth. In the fim, BM denved ce& dispersecl throughout the liver, suggesting
randorn integration into the parenchpu foiiowed by clonai growth and differentiation, m a manner
similar to other liver ce11 transplantation studies, cesulting in a cluster of hepatocytes98,100,1m.
However, in the presence of Liver injury, BM derived stem cells appear to become an ovai ceil-like
i n t e d i a t e with subsequent expansion and differentiati0n8~. Therefore based on these results, oval
cells are derived h m HSC, which then differentiate to small hepatocytes More tùlly differentiating
into functional hepatocytes.
Evidence regarding engrahn t of HSC into hepatocytes indicates that liver regenerative
envimnments have higher frequencies of BM derived h e p a t o ~ ~ t e s ~ ~ . In addition to demonsnating the
importance of the liver microenvironment, this observation may provide important insight into
generating in vitro conditions that allow for expansion and clinical use of such cells in regenerative
medicine103.
1.4.5 Liver growth factors
The mechanism by which hepatocytes and their pmgenitors exit h m quiescence (GO) and enter
into the proliferation cycle (GIS) is triggered by the surmunding extracellular matrix (KM), and
growth factors such as hepatocyte growth factor (HGF), transforming growth factor - a (TGF-a),
epiderrnal growth factor (EGF). tumor n m s i s factor - a (TNF-a) and interleukin 6 (iL-6) derived
from non-parenchymal cells acting in paracrine or jutacrine f a ~ h i o n l ~ * ~ o ~ . The results here focused
on three cytokines, two of which (HGF and TGF-a) are directly involved in hepatic regeneration and
one that is important in hematopoiesis m).
Hepatocyte growth factor (HGF)
Hepatocyte growth factor (HGF) is a 90kD heparin binding gtycopmtein that is secreted as a single
polypeptide chah and cleaved extracellularly into a biologically active heterodimer lM. HGF is also
called scatter factor because of its abiiity to spread coherentiy growing epitheiial cells, a propeny which
is attributed to the disruption of cell-cell junctionslo7. Furthermore, HGF induces variety of orhcr
responses in a variety of cells ranging from mitogenesis, cell motility and matrix in~asion108.10~.
HGF activates a transmembrane tyrosine kinase receptor encoded by the Met pmtwncogene (c-Met).
c-Met receptor is expressed in aduIt epitheiiai tissues including liverI1O, intestine, kidneyLll,
hernatopoietic precursors112-113, endotheliai ce1lsl1~. BM stroma, osteogenic and
oste0sarcomasl13*115,~~6- A similar expression pattern is found during development, where c-Met is
expressed on embryonic ecythroblasts, pancreatic c e l ~ s l ~ ~ , humanL l8 and rat119 fetal liver and muse
yoik sac120.
HGF is produceci by a number of including adult1=l23 and embryonic fibroblastsl24,
osteoblastslu, monoc~macrophages, Bcells, and smooth muscle ceUs 107*126. in the liver. HGF
is produceci by the non-parenchymal ~ornpartmentll~~12~ in particulsr by stellate ce~s11~.128. In vivo
extracts h m pancreas. brain. thyroid, salivary gland and s d l intestine have been shown to contain
HGF. Amounts per pancreas, brain, thymid and small intestine are shown to be 3300, 30.76 and 250
n& tissue respectiveiy.
HGF is essential for normal liver d e ~ e l o p m e n t ~ ~ ~ . HGF is expressed in murine fetai liver thmugh
2 weeks neonatal afier which time its expression is no longer detectedl3O. HGF or c-Met hwkout
mice die in utero between E13-E16.5. In these animais the Iivers show rerarded growth and contain
large *emptyr sinusoidal spaces and the parenchymai ceiis appear dissociated and apoptotic131-133-
HGF also mediates mesenchymaicpitheliaI ce11 interactions in vivo and plays an important rote in
organ ~~eneration13~+13*. HGF has potent mitogenic effects on hepatocytes 136 and can rapidly
induce the proliferation of aduit rat hepatocytes in primary cu1nues~3~. Although little HGF is present
in adult liver, during Liver regeneration pancreatic HGF can & directly transported ta the liver through
the portai circulation 138. It has also been proposed h t HGF is constantiy made available to the liver
from exocrine pancreas and small intestine, and its continuai mitogenic stimulation is balanced by
mitotic inhibitors such as TGF-p. Very small amotuits of TGF4 mRNA are present in the normal
liver139. TGF-$, an inhibitor of hepatocyte proiiferation both in vivo and in vitro, inhibits hepatocyte
proliferation induced by HGF in as weli as HGF expression in stellate CEUS thus regulating
in part the balance between Iiver fibrosis and regenencion 128.
HGF is an important reguiator of hematopoietic progenitor cells, derived h m CD34' cells in
human BM, peripherai blood and umbilical c d blwd130*14i*142. HGF has a stimulatory effrct on
colony forming cells (CFCs) and iis addition to BM cultues enhances non-adhcrent ceIl
n~rnbenL12*113-~~~. h particular, erythroid pmgenitors are stimuiated to grow and differentiate by
the release of his growth factor in BM cultures112. These effefts are synergistic with the activity of
grruiuiocytecotony stimulating factor (G-CSF), granuIocytdmmophage-colony stimdating factor
(GMCSF), Interleukin-3 a - 3 ) and SCF 12*1309142-144-
1 -4.5.2 Transforming growth factor-a (TGF-a)
Transfonning pwtt i factor - a (TGF-a). is a ligand that binds wîth the epiderrnal p w h factor
(EGF) receptor. TGF-a is beiieved CO reguiate normal growth in epithelial tissues, and its
overproduction in celIs possessing EGF receptor is often comlated wich malignant transf'tion.
TGF-a is aiso involved in cell gmwth accompanying the healing process in multiple argan systems, and
it influences fetal liver development as well as Iiver repair foliowing hepatotoxic damage or
regeneration following partial hepatectomy. TGF-or is show to be particulariy important for eady
stages of liver regeneration, at which time hepatocytes are signaled through TNF-a to
produce and respond to T G F - ( ~ ~ ~ ~ in an autocrine mannec
1.4.5.3 Thrombopoietin (TPO)
Tiuombopoietin 0) (a.k.a. megakaryocyte colony stirnulating factor) and its receptor c-mpl
proto-oncogene act as stimulators of megakaqocyte differentiati0nl~~-~50* TPû is produceci mainiy
by hepatocytes151t152 but also by kidney cells. However, factors regdating TPO expression are not
well understood and do not seern to change with respect to liver injury152.
The TPO receptor (c-mpl) is found on a large proportion of HSC h m fetal liver as well as
~ ~ 1 5 3 , 1 5 4 hrn which TPû enhances megakaryocyte formationls. In the liver, the TPû receptor is
found in endothelid cells and has been show to induce the proliferation and cytokine regdation of
these c e l ~ s ~ ~ ~ . No study of their effect on liver resident HSC has been performed.
TPû modulates the biologicd response of HSC both in vitro and in vivo. The addition of TPO to
BM cultures has been show to result in the generation of both long and short-term repopulating HSC,
as detected by in vivo competitive repopulating a s ~ a ~ s l ~ ~ . Furthermre, it has been show that
targeted disruption of the mpl gene greatly reduces the nurnber of spleen colony fonning units after 12
days ( C F U & I ~ ~ ~ ~ . It has been obsewed that TPû alone supports survival and modest proliferation of
highly enriched HSC in vitro158*159- However, the direct effect of TPû on HSC expansion is under
debate with some views expressing that TPO induccs megakaryocyte differentiation, which in turn
secretes other factors that alter the fate of HSC.
1.5 Hematopoiesis and hematopoietic stem cells
Hematopoiesis is an active process in which biood cells are regenerated. For example, in addt
humans, the red bIood ce11 (RBC) pool of approximately 25 trillion celk is tumed over every 120 days,
which necessicates the production of approximately 200 billion ceils per day to maintain hematopoiesis.
This production is maintainai by the continuous production of a large number of blood ceIls h m r
relatively mal1 nurnber of HSC that reside in the BM.
Great pmgress bas been made over the past four decades to understand blwd cell hierarchy since
Ti1 and McCullach h t reported the abity of the injecteci BM ceiis to fom macroscopic colonies in
the spleen of micela. It is believed haî the process of blood production is arrangeci in an irreversible
descending hierarchy (see Figure 1-@). In this scheme, a hematopoietic stem celi passes through
severaI stages of differentiation to produce functionai blood cells. Rimitive HSC are divided into
long-term 0 and short-term (ST) HSC based on theù functional capabiIity CO recOElStitute
hematopoiesis in mice. LT-HSC are slow cycling yet highly self-renewing cells that have the potential
to reconstitute hematopoiesis in primary and secondary recipients. As these cells become more
differentiated they give rise to ST-HSC and multipotent pmgenitors. These progenitors then become
committed to be either common lymphoid or myeloid precursors (CLPs and CMPs respectively). CLPs
migrate to lymphoid tissue where they divide and differentiate into B and T lymphocytes. CMPs
remah in the BM and divide further until theù progeny becornes coinmitted to producing one type of
blood cell. These comrnitted unipotencial progenitors pmliferate significantly as they differentiate
towards the mature form of the b l d cell. Upon differentiation. these restricted pmgenitors give rise
to eryrhrocytes, megakaryacytes, granulocytes and monocytesimacrophages (see Figure 1-61.
1.5.2 Remcatopoietic cell cuiture
In vivo, hematopoiesis resides in a welldefined microenvironment characterizai by local geomeay
(structurai and vasculahire). by stroruai cells (accessory cells of mixed ongin), and by an extracellular
rnaaix @CM) composed of collagen-like molecules and proteogiycans (produced by stroma1
cells)161,162. Stroma1 cells and the ECM aIso provide surfaces for the pcesentation of adhesion
molecules and cytokines (bound cytokines. which cannot be internaiid by the recipient cells, act
different than soluble cytokines). Previous studies into the maintenance of BM denved hematopoietic
cultures have involved the use of numemus methods in order to rnimic the ut vivo micrœnvironment.
However, despite advancernents in the longienn culture (LTC) of hematopoietic cells in vitro, cultures
of HSC usually result in dramatic pmliferation coupled with a signifiant functional loss of HSC. This
suggests that HS cell diiferentiation (not self-renewai) is the predominant fate of HSC in vitro.
1 S.2.1 Stroma mediated long-term cultures
One of the best in vitro models of hematopoiesis is the cuIture system developed by Dexter and
coLleagues163. in these cultures. hematopoiesis is maintained for penods of several months in the
form of nonadherent hematopoietic cells generated frorn initiai cultures of whole bone marrow (WBM),
which develop an adherent stroma Iayer containing a Large variety of ce11 types including fibroblast,
endothelial cells and adipocytes. The use of such stroma layers and the signals they provide to the
HSC has b e n used as one method of initiating long-tenn bone mmow cultures (LTBMC).
The role of BM stroma is believed to be the creation of a microenvironment that is suitable for
hematopoietic ce11 grnwth164-~67. in particular, the use of fibmblast feeder cells has been devetopl
as a method of maintaining HSC cultures and meastaring the frequency of long-tem culture initiating
cells (LTC-[Cs - the LTC-IC assay tests for the presence of cells capable of initiating LTBMC and
giving rise to pmgenitors chat can be detected by replat* into CFC rissay after 5 to 8 weeks 168*164. Stroma1 ceils condition the media by secreting cytokines and extracellular maaix molecules. Studies
have show that direct contact between stroma cells and the HSC is not requtred170. By using cultures
in which HSC are separated by a membrane h m stroma cells, it has k e n demonsmted that LTC-ICs
are better rnaintained in non-contact condition^^^? This indiutes that the hematopoiesis obtaiaed in
these cultures is mediated by solubie factors and not d i t ceIl contact,
1 S.2.2 Alternative cultures
With the increased knowIedge of the molecules that are deemed as essentid in rcgdating HSC
fate, stroma free cuIture system have been developed. These cultures provide more control in studying
the effects of specific signais on HSC behavior, and they may be desirable in c l i ca l applications of ex
vivo HSC expansion. Thmugh rescarch, a "cacktaii" of cytokines bas bcen developed which contains
SCF, FL and IL4 1 famiIy of cytokines172.
Stem cell factor (SCF - aka. Steel factor and c-kit Ligand) is a hematopoietic growth factor that is
important in regulating eady stages of hematopoiesis. SCF has been shown to be critical in in vitro
maintenance and expansion of HSC. SCF or c-kit elimination in mice, leads to a profound reduction in
the number of repopulating ~ ~ ~ 1 7 3 . SCF administration has also been shown to expand the number of
nansplantable ~ ~ ~ 1 7 4 . However, SCF alone cannot maintain HSC ùi vitro175. Thus, the interaction
of other cytokines is important in maintaining undifferentiated HSC phenotype in culture. SCF has
ken show to act synergistically with various growth factors including TPO (described previously), FL
and IL-1 1 to induce proliferation and maintenance of myeloid, erythroid and lymphoid progenitors.
Flt-3 ligand (FL) is a hematopoietic cytokine that biids and signais through the tmnsmembrane
receptor Fit-3Eik-2 Flt-3 receptor is expressed on ~ ~ 3 4 ' ~ ~ ~ and c - I~ i t ! "@~~~ fetal liver and BM HSC.
FL is very important in hematopoiesis and Flt-3 receptor knockouts have a five-fold reduction in the
number of long term repopulating stem c e ~ l s ~ ~ ~ . In vitro, FL has been shown to synergize with a wide
variety of hematopoietic cytokines (in particular SCF and the IL41 family of ~ ~ t o k i n e s l ~ ~ ) to
stimulate the proliferation, self-renewal and differentiation of ~ ~ ~ 1 8 0 .
L-f 1 and other family members such as IL-3, iL-6 and hyper iL-6 (HIL-6) an pleiompic
cytokines chat have multiple effects on HSC. Although L I 1 does not seem to be necessary in vivo, as
demonstrated by the lack of hematopoietic defects in L I 1 knockouts r n i ~ e l ~ ~ , its administration in
vitro in combination with SCF and FL are sufficient for maintaining the number of HSC in semm fiite
culuires182-183. In our expriments, HIL-6 was used as a substitute far IL-II. Hyper iLd also
interacts with GP130 receptor and has been has been shown to have the self-renewal effects on
hematopoietic cu~tures~~o.
2 OBJECTIVES AND HYPOTHESES
Although numemus in vivo have suggested the existence of BM derived cells
capable of hematopoietic and endoderm differentiation. the mechanisms that govern this behavior have
not been elucidated. The overall goal of this work was to develop an in vitro culture system to
investigate the hepatic differentiation fmm BM denved pmgenitors. The objective of the th& wos to
study the effects of cellular and microenvfromental parameters that mimic the regenerathg
tver on hematopoietic and hepatic differentiation of BM derived ceiis. In particular. the endoderm
differentiation of BM derived cells into hepatic lineages was studied.
It was hypothesized that culture conditions designed to mimic Iiver regeneration will allow for the
expression of markers and function associateci with hepatic derived cells by hematopoietic stem cells.
Fibr
Hematopoietic 7 Endoderm differentiation differentiation
Figure 2-1 Possible factors that reguiate BM derived ceii Lîe in hepatic microeavironments
3 APPROACH, AND DEVELOPMENT OF METHODOLOGIES
3.1 Creation of an in vitro iiver mhenvironment
The in vivo micmenvironment is very cornplex and rnechanisms that govem ceU behavior are
difficult to detennine. Despite this difficulty, much progress has been made in determining important
parameters that conml ceIl fate decisions. While the specific parameters differ depending on the ceII
type involved, in vitro systems attempting to maintain or induce differentiation include exnaceliular
matrices, appropriate cell-ceil interactions and the presence of hormones, growth factors or
'differentiation inducen'.
To clarify the mechanisrns of hepatic differentiation. we established culture systems in which the
effects of various liver factors could be investigated. Our system was based on the CO-culture of
defied populations of yellow fluorescent protein (YFP.) BM denved stem cells with hepatocytes,
fibroblasts. soluble factors and exa;tcelluIar mamces. At various tirnes after initiation of the cultures,
cells were harvested and analyzed using flow cytomtry and fluorescent microscopy for the expression
of endodenn intermediate filament markers cytokeratin 8 (CK8) and 18 (CKLS), semm albumin as well
as clonogenic CFC assays.
3.1.1 Hepaîocyte isolation and c u k e
To establish a liver microenvirunment, the feasibility of using primary hepatocyte cultwes was
investigated, However, culture and genetic manipulation of hepatocytes require a high degree of ski11
and long-term culture of hepatocytes is yet impossiblel~.
To isolate hepatocytes, a common f e a m is the use of collagenase to promote tissue dissociation.
The most efficient technique in isolating hepatocytes involves the use of a twestep perfusion
method185.186. This method involves perfusion through the heptic portal vein by a cdcium removing
solution followed by collagenase containing medium. However, the use of a single sîep isolation
method was adoptai in this thesis for its ease of use as well as its relatively good yields.
Many attempts have been made to establish a primary hepatocyte culture for extended periods of
time. such as the use of spheroid culhires. culture in welldetined 1nedia~8~, the addition of
extracellular mamces (such as ~ a m ~ e P ~ ~ ~ and colIagen) and coculture with non-parenchymal
c e ~ s ~ ~ ~ . In this smdy, hepatacyte cuirures were established by modifying the ECM and the culture
media.
3.1.2 CeU lines
Due to the difficuities in maintainhg hepatocytes, the use of hepatocyte and fibrobiast ce11 lines
were considececi. Fibroblast ce11 lines were used as a substitute for nonparenchymai cells while
hepatocyte ce11 lines were used to test the effects of pacenchymai iiver cells on BM derived cells.
The ce11 line selected to mimic hepatocyte hction in vitro was AMLl2 (a mouse liver 12). which
is a murine derived hepatocyte ceil linelm. The AMLl2 cell Lie was established h m hepatocytes
h m mice (CD 1 strain. line MT42) transgenic for human T G F U ~ ~ ~ . Elecuon microscopy shows that
these cells exhibit typical hepatocyte features such as peroxisomes and bile canaiicular-like structure.
They retain the capacity to express high levels of mRNA for serum (albumin. AFP, glyceraldehydes-3-
phosphate dehydmgenase and transferrin) and gap junction (connexins 26 and 32) pmteinsi91~192.
AML12 cells express hi& b e l s of human TGFa and lower levels of mouse T G F a As previously
discussed, TGF-a is an important regenerative liver growth factor and the use of this ce11 line was
airned to enhance the formation of a regenerating liver tnicroenvimnment in the cultures. Expression of
liver specific proteins decreases with tUne in AMLl2 cultures, but it is reactivated by growing the cells
in semm-fke medium.
MMH and other murine liver ceil lines
Other murine hepatocyte cet1 limes considered were the MMH, HepG2 and the L2039 ce11 lines.
M M ' cells were generated by Amicone et ai193 h m liver explants at different developmental stages
of transgenic mice expressing a truncated and constitutively active form of the c-Met MMH cells,
display characteristics typical of hepatocytes, as indicated by morphologicai, phenotypicai and
functional criteria. MMH cells fail to grow on soi? agar and are unable to give rise to tumors in nude
mice; they retain epithelial ceIl polanty, express hepatocyte enrichecl transcriptional factors and
differentiated liver pr0ducts~~3. Th- ceIIs also contain bipotentiai precursots that are capable of
differentiating into hepatocytes and bile duct ceus m rcsponse to EMC and soluble factors and do not
express the liver-enriched uansciption factorslg4. However, AMLl2 cells were preferred kause of
heir expression of TGF-a and rhe formation of conditions which mirnic liver regnemion.
HepG2 cells derived h m human hepatowcinoina were aiso considered to mimic in vivo iiver
microenvironment because they express -y üver pteins. However, HepG2 cells are t r a n s f i o d
nunor ceils, and for this reason they may lack c ~ a c s typical of hepatocytes. Furthemore, they
are human hepatocytes and their secreted products or ceilceII signais may not react with murine celis.
Another liver ce11 line considered was the L2039, which was derived from El4 fetai muse liver
after transformation with temperature sensitive S V 4 large T antigen. These ceils have epithefid
morphologies and express the gens of both biliary and hepatocytic genes such as albumin, AFP, CK8,
CK18 and laminin. This ceIl üne was rejected because it requires abnormal temperature (39 OC) to
express hepatocyte phenotypeLg5-
The NiH-3T3 is a continuous ce11 Iine of highIy contact-inhibited cells, established h m NM
(National Institute of Health) Swiss mouse embryo cultures. It was used to mimic the scellate cells
(which produce various p w t h factors including H G F ~ ~ ~ * ~ ~ ~ and s C F ~ ~ ~ ~ ~ ) in the liver. in
addition to mimicking stellate cells, NM-37'3 cells have been shown to induce differentiation of oval
cells into hepatocytes in vitro ai mes superior [O stellate ce11 Lines 199-2m-
NM-37'3 cells secrete biologicaily active HGF as evidenced by the ability of their conditioned
media to stimulate HGF receptor phosphorylation in epitheliai c e ~ l s ~ ~ ~ . Autocrine-mediated Met-HGF
signal transduction in met transfected 3T3 cells causes hem to becorne t ~ ~ r n o r ~ e n i c ~ ~ ~ ~ ~ ~ ; further
evidenced in their ability to secrete HGF. NIH-3T3 cells also secrece hematopoieticaily important
factors such as S C F ~ ~ . Due to these pmpenies. they were chosen as the ceil line representing non-
parenchymal liver cells.
3.1.3 ECM Related Factors
ECM is a dynamic assembly of interacting molecules that recognizes and regulates ce11 hct ion in
response to endogenous and exogenous stimuli2W. ECM is produced by cells and consists of
collagens, proteogiycans, adhesive glycopmteùis and glycoasaminoglycans and associated bound
pmtein modulators of ceIl function. Along with providing a fnünework withii which cells fonn tissues,
ECM modulates ce11 attachment, shape, morphology, migration, orientation and pmiiferation. For
example, changes in mauix and mutations in genes for ECM proteins cause co~ezt ive tissue disordes.
ECM also serves as a reservoir for various growth factors. It has been proposed thai the existence of
rnanix is essentiai for the activity of p w t h factors (such as HGF, TGFg and acidic and basic FC;F)
since bound growth factors are more stable205-
ECM is very important in maintainhg differentiated hepatocyte phenotype. Pnmary isolates of
hepatocytes plated on plastic attach pouriy. However, when cens arc cultureci on m r derived
basertient ma& (Le. ~ a a i ~ e l * or Englbreth-HolmSwarm sarcoma matrix) thac contaius laminin, type
N collagen. heparan-sulfate pniteogiycans and other proteins. hepatocyte morphoIogy and Iiver
specific function, as measure by albrmiin and cytocbrome P450 (CYP450) expression, are
presewed2~.
3.1.4 Medium requùements
The basa1 nutrient medium used can greatiy affect culture behavior. The medium used in the
majonty of the experiments was a combination of F12 and DMEM medium that has been previously
optimized to maintain differentiared function of hepatocytes in c u l t d û I . This medium has been
found particularly useful for serum-free snidie~2~8. in later experiments, a combiiation of SCF. TPO,
FL and H L 6 were added to regulate HS ceIl face in vitro (se Section 15.2.2).
The ability to distinguish BM derived hepatocytes h m other cells in the CO-culture (such as
hepatocytes) requires the use of ce11 tracking systems. To establish such cultures, several ce11 tracking
systems were considered.
3.2.1 YFP
The green (G) and yellow (Y) fluorescent proteins (FPs) have proven to be excellent markers for
mcking living cells209. GFP is a srnaII cytoplasmic protein that is cloned h m jellyfish Aequirea
victoriaalo. GFP molecule absorbs blue light and emits green without substrates, cofactors or other
gene pmducts211~212. Although GFP was the original protein used for this purpose, severai variants
have been engineered with more efficient translation and brighter fluorescence 211,213314. YFP is a
modified form of G F P ~ ~ ~ with a longer fluorescence l i f d U and better overall expression.
GFP/YFP expression on cultured cells or in tissues of transgenic mice does not cause any observable
deletenous effects2l6. in particular, no effect on the in vivo proliferation and differentiation of the
hematopoietic progenitors has been observed2I7. YFP mice were created by ged ine transmission of
YFP' ES cells218. in these mice, the YFP gene is ubiquitously expressed in all ceUuIar progeny
throughout development and adulthood. The SM h m green to yeltowish green emission is not gceat
enough to make changes in flow c y t o d c pattern thus YFP can be analyzed using the sanie filter as
GFP during microscopy and FACS ana1~sis2~~.
3.2.2 Membrane dyes and other techniques
Cek Iabeled with membrane bound dye have dso been used to detect transplanteci ceiis or certain
populations in coculnires. In particuiar, ~ m - 2 ~ and CESE dyes have been used. CFSE b i i
23
itRversibIy to the intemal constituents of the cytoplasm, and it is equally partitioned to the daughter
çells~l-=.
Table 1: Donor ceii markers and their dlnlcuities h detection by tlow cytocytometry.
1 Marker 1 Expressionsite 1 Dicuities 1 1
CFSE 1 Stained cells 1 Expression lost with ceU division 1 1
I YFP 1 Transfeeted cells 1 Difficult to transfect cells I 1 1
Y-chmmorome 1 Male cells Difficult to detect 1 1 1
Ly-5 (CD45) 1 All leukocytes 1 Only on hematopoietic a l l s
Other methods of tracking cells in culture include the usage of ROSA26 mice, which have been
derived h m transgenic ES cell lines expressing the $-gaiactosidase ($-Gd) p r ~ t e i n ~ ~ ~ . These mice
have a ubiquitous expression of $-Ga1 that allows for their utility in chimera, transplantation and CO-
cultures studies. However. the flow cytometric analysis of ROSA-26 mice is considenbly more
dificult since it requires fluorescent staining of the $-gal rnolecule prior to a n a ~ ~ s i s ~ ~ ~ .
Fluorescent in situ hybridization (FISH) for Y-chmmosome can aiso be used to track transplanted
cells using Ythromosome specific repetitive DNA probe that has been applied to implanted
hepacocytes~6 and BM cells227. This methoci seerns to have a suisitivity in ihc odcr of 0.1%.
However, detection using flow cytornetry is tedious due ta low fluorescence emission h m a single
chromosome. Another widely used technique that has been used to track hematopoietic cells in BM
transplant studies is the Ly5.115.2 system However, this method was not an option in tracking of non-
hematopoietic cells since they do not express Cil45 antigen.
As Table 1 indicates, the use of YFP marker was the most attractive option. This was due to the
expected difficulties associated with some techniques and the failure of 0 t h techniques that were
tested (see Section 5.23)
1 1
Various markers can be used to identifj various hematopoietic and hepatic ceil types. Thse
markers include phenotypical marken and functionai assays.
PKH-2
3.3.1 Identr&ation of hematopoieîk stem ce f i
In c o n m to Iiver stem cells, the identity of HSC bas been much more rigorousty a n a l y ~ e d ~ ~ * ~ ~ .
Despite this, a i i stem cells are difficult to identij. because they comprise a small ceU popuIation in the
Stained cells Expression lost with ceU division
organ in which they mide (cg. HSC reptesent a rare population of O.Ol4.OFlb of W B M ~ . A h ,
since stem cells can undergo both self-renewal and differentiation Hito functionally specialized maturr
c e ~ l s ~ ~ , tbeu definitive identification requires their ability to 'function' as stem cells be demormatd.
A wide variety of hinctional and phenotypical assays have been developed to detect HS and progenitor
cells. in functional assays, ceils are recognized based on their functional propenies of their progeny. In
phenotypic assays, combiiations of surface receptors are used to enrich for the stem ceII campartment.
3.3.1.1 Functional assays
Stem cells are most ngorously defuied by theu in vivo repopulating capability. In MC. this is
usually demonsuated using assays such as the competitive repopulating unit (CRU) assay. This assay
provides a relatively reproducible and specific method for detennining the fresuency of transplantable
celIs with long-tenn in vivo reconstituting capacity. However, most U, vivo tests are Iimited because of
the extended period of time required to perform the test as weli ;is high costs. Therefore, in order to
identiS. the functiond propemes of stem cells, in vitro methods have been developed One system to
detect stem cells is by the use of colony forming assays. Since HS and progenitor ceIls can be activated
to proliferate, a smegy in identiSing putative stem cells is to stimulate the cells CO proliferate and then
focus on their essentiai ability to generate differentiated ce11 types. CFC assays can be used to quanti@
h e frequency of hematopoietic progenitors that form discrete coionies in semisotid gets (aga or metbyl
celidose) in the presence of colony stimulating f a c t ~ r s ~ ~ ~ a ~ . Lincage-ceaicted commiaed
hematopoietic pmgenitors can be evaluated as CFüGM, BFU-E, CFU-G/M. On the other band
multipotent hematopoietic precursors cm be distinguished as CF~-GEMM colonyn3.
Phenotypic markers
isolation of BM derived HSC has been performed by ceIl surface antigen e ~ ~ r e s s i o n ~ ~ ~ ~ .
Numemus ce11 surface markers have been used to identify and isolate HSC h m murinc BM, though a
definitive set of markers, which clearly identifies a HSC, has so far eluded machers- Despite this,
occasionai transplants with singie purifieci stem cells have been successiùi in iî bas kcn
pointed out that the functionaiity and the purity of HSC maybe iiited in reppulating assays since ody
20% of the injected cells home to the BM, the site best suited for hematopoietic e n g r a f b d % thus,
the enrichrnents observed by phenotypicai assays may acnially be an underestimation of tbe achral HSC
content
sa:m@ CD45
Thy- 1 'ow
Positive for: *kit, b t Negative for: 8220, Mac-1, Gr-1 CD3,4,5,8 Tefi19
Positive for: CK8, CK18, Albumin, CyPW
Negative for: C M , -1, okit, CKl9
The ma commonly used markers for identification of murine HSC are CD34 Sca-L, Lin and c-kit.
CD34 is a monomeric msmembrane phosphoglycopmtein that is present in capiIlary endothelid cells
and on some BM stroma1 progenitorsU7. In hematopaietic Iineage. CD34 expression is genedly last
as cells m a m . In mice, it has k e n reported that CD34 is expressed on hematopoietic progenitors
capable of short-term but not long-tenu reconstitution of lethally irradiated m i ~ e ~ 3 ~ . Recent results
suggest that CD34 is not necessarily expressed on H S C ~ - 2 4 0 ; for example, cbnd ~ ~ 3 4 " can long-
terni ceppulate & c g 6 and low expression of CD34 has been shown on HSC*~~. C w n t loiowledge
associates CD34 with M C ce11 cycle activation242 which induces CD34 expression in a revenible
manner243.
Stem ce11 antigen-1 (Sca-1 - also known as Ly6AIE) is a ce11 surface protein that is expressed on
rnultipotent HSC in miceL78~w247. Sca-1' HSC are found in the aduk ~ ~ ~ ~ ~ - ~ ~ * ~ ~ 8 - 2 5 1 and
fetai l i v e p m 3 but not in die early embryonic yolk sac or interaernbryonic hemopoietic sites2S4.
Sca-1' W C can aiso be mobiiked to the peripheral blood and spleen in adu1is255. O k than HSC,
Sca-1 has aIso been found on B and T lymphocytes256~7, myeloid c e ~ l s ~ ~ ~ , and severai non-
hematopoietic tissues 248.
C-Kit (CD L 17) is a transmembrane cyrosine kinase receptor which mets with its Iigand SCF. C-kit
is expressed in adult BM on he~mtopoietic progenitors, including on the Sca-l 'E H S C ~ ~ ~ - * ~ I . It k
also expressed on myeloidlerythroid progenitors2m and lymphoid erecursors262.
One of the most successful mirrine phenotypical enrichment protocols has been developed by
Weissman and coiieagues who e ~ c h e d for M C witfiin c-kit+ Sca-1' Lin- (KLS - where l i g e
markers CDZ CD3, CD4, CDS. CD8, NKI.1, B220, Teri19, GR4 and Mac4 are expressed on
differentiated hematopoietic cells) c e H p0~dation*~~263-*65 and demusaaied their abiity to
clonally reconstitute hemaîopoicsis in lethally inadiiued hosts235S9. HSC have a h been iscilotod
using a 48-bour homing assay in which quiescent ceüs are segregated and have been scen UI repopuiate
other rnice266.
In this work, BM ceUs were enricheci for HSC to varying degrees. In the majority of experiments,
Lin- populations were tested and cornpareci wiih Lii and WBM populations. In some experiments that
required the use of purified HSC. KLS celIs were used
Functional assays
Albumin is the most abundant serum protein accounting for 50% of serum proteias. It is
synthesized exclusively by hepatocytes and it is one of the most comrnonly used markers to detect
hepatocyte function. By adhering to specific substances, albumin ai& in the aansportation of m y
substances such as dmgs, lipids, hormones, and toxins wi th i i the bloodstream Once the dmg ceaches
the liver, it is detached h m the albumin and rnetabdized to a water-soluble form that can be excretd.
Albumin has a half-life of 21 day&' and cm be easily detected in senun; therefore. it is very sensitive
in detecting differentiated hepatocyte functioas.
A nurnber of other liver secreted proteins can be used to detect hepatocyte function. The most
popular of these proteins is CYP450. CIP450 encompasses a highly diverse 'superfamily' of enzymes
responsible for the metabolism of toxic ~arbohydrates~ Also, many other secreted pmteins and
metaboiic enzymes have been used to adyze hepatacyte fitnction. However, these markers were not
analyzed due to the unavailability of mwùw specific antibodies or anticipateci difficulties with low
expression.
3.3.2.2 Phenotypic markers
Hepatocytes and other liver cells can be identifiai with monoclonal antibodies that react
specifically with their specific cytoskeletal prweins. Each ce11 type has a specific and stable pattern of
expression of cytoskeled subunits such as cytokeratins268. Cytokcratins (Ch) are a family of over
20 polypeptides expressed in different epithclid cdls during differentiation269~2~O. Therefare,
different endoderm tissues (such as Iiver) can be disthguished and characterized by theii specific
cytokeratin expression patterns. CKs are divided into 2 groups (acidic and basic) which are
differentiaily expressed as pairs271. Most ceIIs express one acidic and one basic of the 21 different
cytokeratin protein~2~2.
CK8 is a basic intermediate filament that is expresseci in simple epitheliai ceils such as trachea
smail intestine, bladder, pancreas. colon and mammary gland ducts. Furthemore, CK8 and CKI8 are
the first cytokeratin expresseci in embry0s2~3. Diffctentiated hepaiocytes express the pair CK8 and
~ 1 8 2 7 2 , while cholangiacytes express CK7 and CK19 in addition to hepatocyte C K S ~ ~ ~ . These
expression patterns are especiaiiy strong in hepatocytes and cultured hepatocytes have been sbown to
continue the expression of hepatocyte Iike cytokeratins long after they lose expression of other
functionai markers275.
Very little is known about the role of CKs in hepatocytes27l. It is believed that CKs provide ceils
with mechanical integrity and may be very important for endodenn differentiation. in CK8 knackouts,
50% of mice die before ~12-132~6 due to growth retardation and Liver dysfunction; interestingly,
CK18 knockout mice appear n o d and functiona127~- Based on these observations. it appears that
CKWCK18 intermediate filaments are important for liver integrity.
ûifferent reactivities of cytokeratin antibodies with the specific cytokeratins of different tissues
have been noted with conventional a n t i ~ e n u n ~ ~ ~ - ~ 8 0 and monoclonal antibodies281. Thus, great care
was taken in chwsing the antibodies that reacted with the infrequently studied murine b e r systea In
mice. TROMA-1.2.3 stain CK8. CK18 and CK19 respectively28~83. These ancibodies have been
show to react with murine liver cytokeratins although they react differently depending on the
rnolecular weight 0 of the p ~ o t e i n ~ ~ ~ .
E-cadherin was also used to characterize murine hepatocyta. Ecadherin is a regulator of junctions
in epithelial cells. in hepatocytes, Ecadherïn is expnssed on ceIl surface where it ai& in regulating
cellçell contact with other hepatacytes and endotheliai cells100.
Flow cytometry was used as the preferred analytical method for the detection of Iiver specific
phenotypical marken due to its capability to quanti@ population dynarnics and detect simdtanwius
markers. Flow cytornetry is a fast and accurate technique in which individual ceUs are analyzcd (as
supposed to celI populations). Although flow cytometry has been used for isolation of +tic
progenitor ce1ls80285 and characterizauon of hepatocytc suqensions268, its widespred use has nat
been adopted for adyzing hepatocytts. Thus, the development of this technique for measuring cell
surface and innacellular niaricers was one of the goah of the pmject.
4 M A T E U AND METHODS
4.1 Bone marrow stem ceIl extraction and enrichment
4.1.1 Bone murrow isolation
Ali procedures involving mice were perfomied accordiig to the University of Toronto approved
protocols. BM cells were collected h m the femur and tibia of 5 weeks or oIder male C57BV6 mice
(Charles River laboratories. WiImington, MA), as well as YFP' mice or Wu YFF parent lm
(generous gifts from Dr. W. Stanford). The process of BM isolation is illustrateci in Figure 4-1. Mice
were euthanized using cervical dislocation, and their femurs and tibias were then dissected and placed
within precwled Dulbecco's Modified Eagle Medium (DMEM - Gibco BRL, Life Technologies,
Grand Island, NY) containing 1000 Ulm1 penicillin and 1000 pghl smptomycin (10X-DMEM - Gibco BRL). The bones and isolated cells were maintained on ice to prevent any deleterious effects on
celb.
@ +,- ENfchment for Urrcslb
W e CSIB I (Stemsep !of m u ~ e
YFPl YFP œüs, stem ceü Tech.) iri Figure 4-1 Schemadc of BM isolation p m e a
The bones were then clenned of any attached muscle tissue. To flush out the BM, scissors were
used to rernove a small portion of the bone fmm each end of the femur or tibia. A 22 H gauge d e
{Becton Dickinson (BD). Franklin Lakes, NJ) attached to a 3 ml syringe (BD) was then inserted
through the cartilage plate at the discal end of the femur and the mamw was fiuskd out with 10X-
DMEM into a 15 ml centrifuge tube (Corning Glas Works, Corning, iîhica, W. The bones were
Bushed two more rimes to ensure maximum recovery of the marrow cells. Typically. the cells h m 24
mice were flushed with the same medium and thus pooled into a single batch and clumps were broken
up by aspiration through the needle.
Any large b n e fragments were then removed by running the c e h through a 70 pm cell saainer
(Fiiicon, BD, Ftaaldin Lakes, NI). In some experiments erythrocytes, were lysed with a solution
containing 8.3 @L ammonium chioride in 0.01 M Tris-HCi buffer (Red BIood Celi Lyshg Buffer,
Sigma, SL Louis, MO) for 8 minutes. The nucleated c e b were then washed with phoûpbate buffered
soIution (PBS - Gibco BRL) and counted prior to further e ~ c h m e n t or analysis using a
hemocytometer. Viable ceii numbers were determineci by mypan bIue (Sigma) dye exclusion. if BM
preparation and soriing couid not be performed on the same day, the BM was kept in the refigeratur
overnight prior to staining and sorting.
4.1.2 Stem ceU enrichment
Lineage negative (Lin') BM cells were isolated using the commercially available negative selectian
system ( ~ i e m ~ e ~ @ , StemCell Technologies hc. (STI), Vancouver, BC, Canada). in this system,
unwanted celis were labeled with multiple antibodies that were atîached to magnetic particles. The
celis bindiig the magnetic particles on their membrane were then separated within a high gradient
magnetic column. StemSep depletion cocktail contains antibodies to murine CDS (T and B celis),
CID45R.A (B and T cells and lyrnphoid progenitors but not HSC), CDllb (macrophages and
monocytes), GR-1 (granulocytes) and TER1 19 (erythroid cells and erythrablasts). Cells chat do not
express lineage markers are not labeled with antibodies and are thus ready for furtber manipuiation.
To isolate Lin' cells, WBM cells were suspended at 5 x 10' cellsiml in PBS containiag 2% feral
bovine s e m (FBS - Gibco BRL) with 5% normal rat senim (STI) and incubated for 15 min.
Subsequently, they were incubated with 10 pUmL of the enrichment cocktail for 15 min at 4 T. The
celts were then washed and resuspnded at 5 x 10' cells/mi in DMEM containkg 2% FBS and
incubated with 100 pUml of anti-biotin tetrameric antibody complex for 15 min at 4 T. Afterwards,
the ceils were mixed with 60 pUmI of magnetic colloid solution for 15 min and then dincdy passed
through the colurnn.
To Wolate c-kit* Lin Sca-1' (KU) populations. column enriched Un- cells were stained with 2
pglmL of direcdy labeled FiTC-CD45, PE-Sca-1 and APCc-kit or appropriate IgG isotype antibodies
(BD Pharrningen, Franklin Lakes. NJ) for 1 hour. Desired cells were then isolated using fluorescence
activated ceIl sorter (FACS), prfotmed at Princes Margaret Hospital by a MoFlow FACS
(Cytomation. Fort Collins, COL).
4.2 Primary hepatocyte isolation and culture
Hepatocytes were isolated h m livers of male C57BU6 mice. To enhance the yield of viable cells,
al1 media were pre-warmed to 37 T. To isolate priniary tivcr cells, the abdominal region was cut and
ail liver lobes were dissected h m the mice and placed on a 60 mm pem dish (Falcon, BD) containmg 2
mL of 0.1% (wlv) collagenase (Sigma). The lobcs were then minced with scissors into s d pieces.
The chopped tiver sections were then placed into 2û mi of 0.1% (wlv) coiiagenase in a 50 mL
centrifuge tube and were vigorousIy pipetted. The solution was kept at 37 "C for 25 minutes with
frequent mixing. The resuiting ce11 mixture was then washed with PBS and re-suJpended in 15 mL of
ceIl dissociation buffer (Gibco BRL) for 20 minutes with m u e n t mixing. F d y , to remove
dissociated cells h m nondigested tissue aggregates, the cells were filtered through a 70 pm coarse
filter and then washed with PHs.
Collagen (Type IV, Sigma) coated dishes were used to enhance the viability of pcimary hepatocytes
in cuiture. To coat dishes, 10 pg/cm2 of a solution containing 0.01% (wlv) of collagen in 0.1 M aceac
acid (Sigma) was allowed to bind with the dish ovemight at 4 T. Excess fluid was then m v e d and
the coated dishes were sterilized by ovemight exposure to W light in a sterile tissue culture hood
(Forma Scientific Inc., OH). Fmally the dihes were rinsed with PBS prior to initiation of the cultures.
isolated hepatocytes h m each mouse were cultured in a 25 cm' tissue cdture dish (Sarstedt hc.,
Newton, NC) with or without collagen coatings and theu metabolic functionaiity and aibumin
expression was measured at regular intervals. The media used for hepatocyte cultures were either ihe
HepatoZYME - semm free medium (HepatoZYMESFM - Gibco BRL, which has been specifically
designed for maintenance of metabolic activity of hepatocyte culture) or a 1: 1 mixntre of DMEM and
Ham's F12 (Gibco BRL) containing 10% FBS and other supplements (see Section 43.1).
All ce11 lines were purchad h m Amencan Tissue Type Collection (ATCC, Rockville, MD) and
maintained in 25 cm' tissue culture flasks in a 95% airis% C a hunidified incubacor at 37 T. To
prevent bacterial infection, 100 U/mL penicillin and LOO p g l d sueptomycin (Gibco BRL) were added
to all cube media. NiH-3T3 fibroblasts were maintained in DMEM medium containing 10% (vlv)
FBS. These cells were subcultured every 5-7 days by the addition of 1 mL of 0.25% Trypsin- 1 rnM
EDTA (Gibco BRL). The flask was then incubated for a few minutes until the celIs were detached
fmm the bottom of the dish. Detached cells were then suspended in IO mL of fresh mdium and
centrifugeci for 5 minutes at 1000 rpm Following centrifugation, the medium was aspirated and the
celIs were then resuspended in normal medium and subculnued using a 1:LO subcuIture ratio.
AMLI2 cells were maintained in a medium comprised 90% of 1: 1 (vlv) mixture of DMEM and
Ham's F12 medium O F medium) with 0.005 mg/ml i d i n (Gibco BK) , 0.005 mgh i cransienin
(Sigma), 5 ng/d selenium (Sigma), and 40 nghl dexameihasone (Sigma). and 10% FBS (Gibco BRL).
AMLl2 ceIls were passaged every week (using a 1:4 subculture ratio) and fed every 3 4 days.
4.3.2 Bone marrow cultures
To initiate BM cultures, 2 x 10' WBM or 5 x ld Lin ceiis wece culaued on to each weU of a six-
weIl plate (Faicon, BD) precoated with h4atrigela (Collaborative Biomedical Products, Bedford, MA),
each containiag 5 ml. of DF medium dong with 10% FBS and other supplements, UOIess otherwise
indicatecl The medium was routinely chaaged every 3 4 days (beginning after week one during which
no feeding is done) by removing half of the volume (2.5 ml), without taking out any of the ceils, and
replacing it with an equivalent volume of ûesh medium. The ceiis of a dish were harvested. counted
and analyzed using the CFC assay or flow cytomeay. The medium supernatant was frozen at -2û OC
for subsequent andysis for albumin content analysis (see Section 4.5.1).
For experiments designed to d e t e d e the clonal endoderm differentiation ability of the KLS cells,
1 to 100 K U or 1OOO Lin- CD45 cells were placed in wells (using FACS) that had been pre-coated
with hiamgel@ and pre-fiiled with 75 pl of either 1 pglml HGF or hematopoietic medium (HM 100
ng/mt SCF, 100 ng/ml FL, LOO ng/ml HIL6 and 20 ng/ml TPO) or the combination of two media in HF
without semm The presence of single cells in each welI was confiimed one day after culture initiation.
These cultures were maintained for 6 days after which, an additional 75 pl of the original medium
suppIeniented with 10% FBS was added to each well. After 17 days the number of viable cclls in each
positive well was then recounted for clone size determination and medium supernatant was frozen at - 20 T for subsequent albumin content analysis. To ensure that the secreted aibuxnin was derived from
cells, wells treated with various media ware used as conmls.
4.3.3 Bone marrow CO-cuhres wifh feeder cells
To initiate cocultures. 1 x 105 3T3 or AML12 feeder cells were cultured on to each welI of a six-
well plate precoated with ~ a t r i ~ e l @ (see Section 4.3.3.1) for 24 hours to obtain an adherent layer. BM
cells were then isolated h m YFP mice or stained with CFSE and culnired directiy or after enrichment
ont0 the cultures at 2 x 10' WBM, 5 x l d Lin' or 1 x 10' Lin' cells. For control cultures, dishes without
feeder layers were used (siinilar to BM cultures). Unless otherwise indicated, the medium used for the
cocultures was the DF medium plus supplemnts (see Section 4.3.2). Half of the medium was changed
every 3-4 days without removing non-adherent cells. The ceIls of each dish were harvested, counted
and analyzed using CFC assay or flow cytometry. The medium supernatant was h z e n at -20 T fw
subsequent anaiysis for albumin content (see Section 45.1).
In coîultures in which d i t ceIl contact was prevented, 0.2 pm tissue culture insens (Nunc
Pducts) were used to sepatate the feeder ceI1s from the BM cells. In these cultures, the feeder ceils
were cultured as previously describeci. However, h4atrigel@ coating was performed on the surface of
the tissue culture mserts and BM populations were cultured dirtctiy on the membranes. Al1 teagents
and analysis procedures were similar to the other cultures.
~ a t r i g e p was used to establish 3 dimensional ( 3 4 ) cultures or to coat dishes. Frozen (-20 OC)
aiiquots were thawed initially at 4°C overnight and then &ed in a 1:l (vlv) ratio with cold DMEM
tissueculture medium (4 T).
To coat dishes, wells were pretreated with ~ a t r i g e l ~ solution (diluted 1:10 with DMEM) or with
15 p~crn' of 1:1 diluted ~atrigel@ and incubated overnight at 37 T. To establish 3-D cultures, 150
Cr~cm2 of ~ a t r i g e l ~ solution that containeci cells was added to each dish and maintained at 37 OC. To
recover ceils from 3-D ~ a t r i ~ e l @ cultures, dispase (Gibco BRL) was used. However, in coated dishes.
the use of dispase was not necessary as trypsin was suficient to recover the cells.
Cytokines
Al1 cytokines were stored at 4 0 OC, they were obtained as gifts or purchased from companies as
foIlows: recombinant mouse (nn) TPO (Sigma); recombinant human (rh) FL (immunex, Seattle, WA);
rh HIL-6 (Gift from Dr. Stefan ~ose-lohn*a6), rh HGF (R Bi D Sysrems. Minneapolis, MN) and rm
SCF (Calbiochem, La Jolla. CA).
CFSE staining
Carboxfluorescein diacetate succinimidyl ester (CFSE - Sigma) stock solution was prepared at 5
mM in dimethyl sulfoxide (DMSO -Sigma) in small aliquots that were stored at -20 O C To label, cells
were incubated at a concentration of 5 x 106 cells/ml in a solution of 10-50 pM CFSE in Hanlr's basal
salt solution (HBSS - Gibco BRL) for IO min at 37 T. Further uptake of dye was ceased by the
addition of 10 fold volume of cold HBSS with 10% FBS- The cells were then washed three times with
HBSS containing 5% FBS and incubated overnight in regular medium. To analyze the ceils under flow
cytomeny, 106 celIs were suspended in 100 pL of PBS dong with 2 pg pmpidium iodide (PI - Sigma)
to stain dead cells.
Cells were assayed to determine theu ability to form eryrhropoietic (BFU-E), granulopoietic 1
monocytic (CFü-GM) and multi-lineage (CFLT-GEMM) colony forming cells (CFC). CFC assays were
perfonraed by plating an appropriate number of celis (to give < 100 coionies per 1 mL culture - for
WBM this represented about 2 x 10' cells) in FCS-supplemented methylceliulose containing Iscove's
medium (h4ethaCult m. !3ïï). The folIowiag growth factors were included as part of the mdium to
induce colony formation: Epo 3 unitslrnL, 10 nglmL rh IL-3.10 ng/mL rh I L 4 and 50 ng/mL rm SCF.
Methylceiidose culnrres were aliquotted using a 3 mL syringe (BD) and 16g blunt-end needle (STI) in
1.1 mL volumes into 35 mm peni dishes (BD) and then incubated at 37 OC in humidifiai ammsphete of
5% a in air. Colonies were then scored in sinr using well establisteci aiteria; 35 mm dishes were
ptaced withïn 60 mm grideci dishes (Nunc Products) after 7-12 days of incubation. AU CFC assays
were done in triplicate.
4.4.2 Hernatopoietic morphologicd markers and unalysis
The following ce11 surface an t iMes were used to detect various antigens found on hematopoietic
cells: FiTC or PE conjugated rat anti-murine CD45. J3TC conjugated rat anti-murine CD34. PE
conjugated rat ami-rnurine Sca-1 and APC or biotiii conjugated rat anti-murine c-kit (BD Phnrmingen,
Mississauga, ON. Canada). For al1 diredy labeled antibodies, appropriately IabeIed rat IgG isotype
antibodies (BD Phamtingen. Mississauga, ON, Canada) were used as controI. For antibodies
conjugated with biotin. the appropriate strepiavidin-fluorochrom construct was used as contrd.
To stain. test samples were washed once in PBS containing 2% FBS. The samples were then
resuspended in HBSS solution containing 2% FBS at I 1 x 10' cells/ml with 2 p g l d of specific
monoclonal antibctdies at 4T for one hour. For biotin conjugated samples, the cells were washed with
HBSS + 2% FBS and resuspended at S 1 x 10' celldml with 10 pg/mL of streptavidin-fluomchrome
consmct for 45 minutes. 7-aminoactinomycin D (7-AAD - Molecular Probes, Eugene, OR) or PI was
added at 2 pg/mL to stain dead cells.
Stained cells were anaiyzed using the EFKS XL flow cytometer (Beckman Coulter Inc., Fullenon,
CA). The 488 nm argon ion laser was used for excitation with emission king measured at appropriate
band pass fileers. Data was collected and analyzed using the EXPû32 software (Beckman Coulter hc.).
4.5.1 Serum albumin detection
Aibumin enzyme iinked immunoassay (ELISA - Bethyl Laboratones, Montgomery, TX) was
perfarmed at roam temperature using double ana'body sandwich metticxi. The Iowa l i t of sensitivity
of the mouse aibumin ELISA was about 100 pg/ml. The weIis of a microtiter plate were pretreated with
LOO p l of a coaring buffer comfised of punfied goat anti-mouse albumin antibody at 10 pdmi in a
diluent containing 0.05 M sodium bicarbonate (Sigma) adjusted to pH 9.6 and kubated for 60 miautes.
The plaie was washed twice with PBS cootaining 0.05% Tween 20 (Sigma) and then blocked with 200
of a 1% (wtv) bovine semm aibirmin @SA - Sigma) sotution m PBS for 30 minutes. After washing.
100 pl of purifieci murine albumin staadards and sampIes (diIuted in 1% BSA dissolvecl in PBS with
0.05% Tween 20) were added and incubated for 60 minutes. After 2 washes. 100 pl of diluted (1t5000)
detection anîibodyfenzyxne conjugate solution was added to each well which was then incubated for 60
minutes. The wells were then washed 3 more times and 100 pl of 3.3'53'-teeramethyl benzidine
(TMB - Bio-Rad Laboratones, Hercules, CA) substrate solution (prepared by 1:9 (vh) mixing of the
two substrate reagents) was added to each well. The plate was incubated for 5-30 minutes based on the
intensity of the colour change.
After the incubation, the TMB reaction was then stopped by adding 100 pl of 2 M &SO, (Sigma) to
each well. A microplate spectmphotometer (VERSA-w, Molecular Devices, Sunnyvale, CA) was
then used to anaiyze the absorbante intensity of each plate at 450 nm.
The rnetabolic activity of the isolated hepatocytes was measured by using the hflT (3445-
dimethy~thiazol-2-yl)-2,5-diphenylteÉlilPolium bromide - Sigma) assay. This assay is based on the
ability of the cells to conven MïT, a yeilow teaazolium salt, to purple forrnazan crystals via the
lnitochondriai activity of the cells. These sait crystals are insoluble in aqueous solution but may be
dissolved by adding a solubiiiition solution and it is then read with a spectmphotometer.
Cells were incubated with 1.1 pM MTT in normal medium. After 4 hours, the supematant w;is
carefully aspirated and DMSO was added to dissolve the MïT crystals. The intensity of the color of
the solution was detennined using a micropIate specuophotomter (VERSA-max, Molecular Devices.
Sunnyvaie, CA) at an excitation of 5 19 nm and analyzed using Softmax Pro V3.1 (Molecular Devices).
4.5.3 alepatic morphologicd markers and anaiysis
Two diffennt approaches were used to rinalyze the ce11 surface as well as inmcellular markers of
hepatic cells. The ce11 surface markers used were c-Met (tyrosine kinase ceil surface protein287
activnted by ~ ~ ~ 2 8 8 that is expressed in many tis~uesl*~) and E~adherin. The following antibodies
were used: c-Met antibody (sc-162, Santa C m Biotechnology Inc.) and secondary PE conjugated
rabbit anti-goat IgG (Biomeda, Foster City, CA), rat anti-mouse Ecadherin (Si*) and mC-anti-rat
IgG (Sigma). To stain the cells aiready established protocols were u~ed28~2m. Briefly, cells were
resuspended in HBSS solution containing 2% FBS at I 1 x 10' cells/ml for 1 hour with 2 p g h L of
specifrc monoclonal antibodies. For secondary antibodies the cells were washed and resuspended with
appropriate dilution of the secondary a n a i y for 1 hou.
Rac ana rnurine CK8, 18 and 19 antibodies (TROMA 1.2.3 - markers for mouse endaden&)
were obtained h m DeveIopmental Hybridoma Bank (TROMA-1) and as generous gifts from Dr. R
Kemler (Max-Planck Institute, Germany for TROMA 123). These anti'bodies were used to stain cclls
for intracellular antibodies. using standard protocols (In-p permeabilization kit, Emmunotech-
B e c b Coulter Co., Marseille. France). Briefly, the ceUs were suspended at 5 1x10' c e h /ml and
diquo& in 100 ul samples into 5 mL test tubes (Falcon BD). Aliquots were then diluted at 1:l ratio
with aqueous fixation medium containing 5.5% (vlv) foddehyde and incubacd for 15 min at m m
temperature. Cells were then washed with PBS, centrifuged and the supernatant was discarded. To
permeabiiii, cells were incubated for 5 minutes in 100 pl of saponin rich solution, which w u added to
the cells without mixing. The antibodies was added dicectly to the saponin solution at a 1 5 dilution
and incubated for 3 hours at room temperature. The samples were then washed once in PBS, and
resuspended in 100 pl of PBS containing 5% rat s e m and 10 pg/mL of the PEconjugated secondary
antibody (Immunotech-Beckman Coulter Co.) or biotinylated goat anti-rat IgG (Jackson
immunoresearch Laboratones, West Grove, PA) and incubated for 1 hour. For biotinylated antibodies,
the ceIIs were subsequently washed in PBS and resuspended in LOO pl of LOO ~glmL streptavidb
fluorochorm substrate. Stained cells were analyzed using EPICS XL flow cytometer as previously
describeci (see Section 4.4.2). For negative controls, cells were stained with isotype-matcW contds
for irrelevant specificity.
4.6 Image anaiysis
AI1 samples were viewed using an Olympus CK40 RFL epifluoresçent light microscope {Carsen
Group Inc. (CGI), Markham, ON. Canada) with an attached digital carnera (CGI). The microscope was
equipped with a laser and filters for FiTC and PE (CGI).
4.7 Data analysis
Data from experiments chat were performed in duplicate or triplicates were analyzed as the mean f
SD. Student t-test was used to determine the statisticd significance between paired data. Group means
were compared using single factor analysis of vzuiance (ANOVA) to determine staiistical
~ i ~ n i f i c i u i c e ~ ~ ~ . P c 0.05 was used to detect statistical difference ktween samples.
5 RESULTS
5.1 Bone marrow Purification and çhacacterization
Whole bone marrow (WBhf) and enriched ceIl populations were anaiyzed using phenotypic
markers and CFC assays. The percentages of various hematopoietic markers on viable Pr WBM cells
are show in Table 2. The majority of the BM derived cells were CD45'. indicating char they were
hematopoietic. The expression of hematopoietic pmgenitor markers CD34, c-kit and Sca-1 were also
andyzed. The results obtained were similar to previously reported data10m2. For example, the
typical expression of CD34', c-kit', Lin and Sca-1' in BM has been previously reportai at 1.796292,
7.7%. 6.6% and 4-28 respective~ylOO.
Table 2 Percentage of various hematopoietic markers in mownucieated WBM celis
Different types of CFC pmgenitors were measured based on the phenotypic properties of the
derived colonies. Example figures of various colony types are presented in Figure 5-1. Typicaiiy,
CFU-GM colonies contained small, uniformly round cells with distinct cytopiasrnic membranes thar
often fonned grape-like clusters. CFU-GM colonies were f o d in a wide range of sizes. BFU-E
colonies contained predominantIy erythroid cells with a minium of three cIusters (usually three to
eight clusters), each containing 8 to 32 erythroblasts. The cells were s d l and cefractüe, and the
clusters had irregular shapes. Unlike their human counterparts, some murine BFü-E colonies were not
reddish-brown. CFü-GEMM colonies represent the most primitive of the clonogenic pmgenitor types.
CFü-GEMM colonies were recognized by their ability to form large colonies that contained erythroid
clusters. granulocytes, monocyte 1 macrophages as well as megakaryocytes. Most CFüGEMM
colonies contained dense cotes with erythroid clusters visible usuaiiy at the edges of the colony ai
h i e r magnifications.
In gened, the rate of colony focmation in unpurifieci BM cells was very low. For example, less
than 1 in 200 WBM cells had the abiiity ta fom colonies in the CFC assay. The majority of these
-Y
Total CFC
CFü-GM
BN-E
CFU-GEMM
% Purity*SD
0.4 I O. 1
0.31 0.05
0.06 I 0.01
0.035 I 0.005
Marker
CD%'
CD49
C-kit*
Sca-1'
c-kit' Sca-1'
(% Positive)
1.6
99.0
9.3
7.5
1.6
CFCs were comprised of W G M colonies (-80%). fotiowed by BFü-E (-16%) and CFü-GEMM
(-4%).
Figure 5-1 CFC micrographs: CFU-GEMM (a), CFU-CM (b) a d BFU-E (c). (bar = 50 pm)
As indicated. HSC and pmgenitor ceüs compriscd a small fraction of the overall BM content. To
obtain more purified ceIl populations. celis were enriched either as Lin- or c-kit' Lin Sca-1' (KU)
populations. To remove cells chat expressed hematopoietic Iineage markers (Lin), cells were stained
with CDS, CD45RA, CD 1 lb, GR-1 and TER1 19, which stain differentiated hematopoietic cells. Table
3 shows the recovery and enrichment of various hematopoietic progeaitors after Lin- isolation.
As expected, column e~chmen t enhanced the proportion of cells that expressed various
hematopoietic pmgenitor markers. The enrichment of these markers v a r i d greatly depending on their
expression on differentiated hematopoietic cells. For example, since CD45 is ubiquitously expressed
on hematopoietic ceiis (icluding HSC), the Li separation process did not tead to its enrichment. For
hematopoietic pmgenitor markers, the cesulting enrichment varieci b m 3.7 fold for Sca-1 (wbich is
also expcessed on TceUs) to 17 foId for CD34 CoIumn enrichmnt also enbanccd the frequency of
various types of CFCs fmm 6 to 9 fold.
Table 3: Percentage of various hematopoietic mrkers in mononucieated Lin' ce& afier column separation b
Seri-1'
c-kit'
CD34'
CrM!?
CFC
CFU-GM
BFU-E
CFU-GEMM
% Positive t SD
22.0
44.8
39.3
99.8
% Purity t SD
3.8 * 1
3 i 0.9
0.5 î 03
0.3 I 0.2
96 Recovery * SD 44.1
33.1
89.7
11.8
% Rccovery * SD
36 i7
44 i 8
2815
33 &
Fold Enrichment t SD
3.7
5.5
17
1
Fold Ehcichment î SD
9i2
9 î 2
611
811
CL c-kit CD45
Figure 5-2 Phenotypic adysis for isolation of KLS celis and the restricîed ptes used to sort HSC.
in some experiments. phenotypidly defined KLS cells were isolated from Lin' e ~ c h e d
populations. These cells expressed CD45, and were thetefore hematopoietic in origin [see Figure
5-2(d)}. Furthemore, KLS cells are bigMy enriched in W C activity, and contain -20% CFCs and
-5% CRU c e l ~ s ~ ~ ~ . Figure 5-2 illusmes the approach used to isolate these cells. Column e ~ c h e d
Lin cells were gated on fiow cytornetry bastd on their cellular morphology, their ability to exclude PI
dye and express c-kit and Sca-1. The c-kit' Sca-1' cells constituted about 15% of the Lin- population,
uanslating to about 1 in 1000-5000 of mononucleated BM cells.
To examine HGF receptor expression on hematopoietic progenitor cells, c-Met (HGF recepmr)
expression on WBM and Lm- cells was analyzed. As show in Figure 5-3, a small fraction of WBM
cells express c-Met (-3%). while this popdation was edched in the Lui' cells (- 6 fold to 18%). These
results correlate with previously published reports regardiig c-Met expression on hematopoictic
progenitors, which showed that about 30-5096 of the human CD34' cells as well as a high percentage of
CFCs simuitaneously expcessed c - ~ e t l l ~ . Furthemore, as e ~ ~ e c c e d 2 ~ ~ , freshly isolated WBM cells
did not express CK8 or CK18, markers chat are associated with simple epithelial cells (see Figure 5-3).
in contcast, the staining panern of AMLl2 hepatocytes revealed that the majority of CK8+ cells
simultaneously cwxpress c-Met. This may indicate that celis which express the highat levcls of CK8
expression also express the c-Met receptor and hus have the capability to respond to HGF protein.
Figure 5-3 c-Met d C M expression of Isotype control (a), Lin' (b), WBM (cl and AMLl2 (dl c e k
5.2 Development of the co-culture systems
To devetop in viiro iiver microenvironmenu, significant effort was dedicated to the design of co-
culture systems. To establish such cultures, the use of primary hepatocyees as weU as, hepatocyte and
fibroblast ce11 lines was investigated Also. the effectiveness of various ceii tracking techniques for
distinguishing BM derived ceh h m feeder cells was examined.
5.2.1 Prinrtuy hepatocytes isolrdion and cuhre
To evahate the feasibility of using primary hepatocytes in coculture with BM cells, hepatocyte
cuitures were estabtished. ïsolated liver cells were culnued either on m t e d or non-mted collagen
coated petri dishes, and monitored for the expression of hepatocyte marlrws and overaii metabolic
activity.
Figure 5 4 Micrographs of prùniry hepatocyte cuihves culturecl on collogen after O (a), 7 (b) end 15 (b) fm-
As shown in Figure 5-4, freshly isolated hepatocytes were typically found in ce11 aggregates chat
contained approximately 85-98% viable cells. Despite numerous attempts with various enzymatic
digestions ranging h m dispase to collagenase, a complete single ce11 suspension of hepatocytes was
not obtained. Isalated hepatocytes did not proliferate in culture but instead after 7 and 15 days,
fibroblastic cells overgrew the primary hepatocyte cultures. This culture overgrowth occurred
regardles of the propenies of the coated dishes. The majority of hepatocyte aggregates disappeared
during the two-week culture period. However, based on M T ï and albumin secretion results (see Figure
5-5 and Figure 5-6), it was clear that the loss of hepatocyte viability and function o c c d w i h the
fmt few days of culture.
Time (days)
Figure 5-5 Ceü spceilir MlT acîivity for hiated prinmry hepatocytes in various raltore coMUtl0~0 (SFM = HepatomMESFM and FI2 represenb DF medium)
At different cimes after Iiver culture initiation, ceII specific metabolic activity was analyzed using
M'ï ï assay. As indicated in Figure 5-5. the rate of formazan aborbance per c e U d d with time
from - 2 x IO*^ abs unitdceli for freshly isoIated hepatocytes to - 4 x IO^ abs unit5 /ceIl after a wmk of
culture. This represents a signifiant dectease in ce11 specific rnetaboiic activity. This lais cm be
amibuted to a decrease in hepatocyte metabolic function as weii as changes in population dynamics due
Figure 5-6 Albumin secretion rate in various primary hepatocyte cuIturrs. (SFM = HepatoZYMESFM and FI2 represeats DF d u m )
To test for hepatocyte function within these cultures, the extent of albumin secretion from
hepatocyte cultures in vnrious conditions were measured. As demonstnted in Figure 5-6, despite the
use of collagen coated flasks or HepatoZYME-SFM, the rate of dbumin semetion in al1 cultures
decreased to nearly undetectabIe Ievels within one week after culture initiation. These results are strong
evidence that the use of primary hepatocytes, at least under the conditions tested, was not feasibIe in
order to facilitate the long-term co-culture of BM denved celis with hepatocytes.
To overcome the difficuities faced in maintaining primary hepatocytes in virro, A m 1 2 hepatocytes
and NM-3T3 fibroblasts were used. These ce11 lies were also used for the dewlopment of analytical
methods to characteriz mocphological and functional ptopenies of primary hepatocytes and hepatic
ceil lines.
Figun 5-7 Micrograph of NM-3T3 fibroblpptE. Figure 5-8 MIcrogiliph of AMLl2 bepatocytes (bar = 50 cm) (bar = 50 pm)
In culture, MH-3T3 and AMLl2 ceU ünes demonsuateci phenotypical f e a m that were
characteristic of their ptimary fibroblastic and hepatic counterparts. As showa in Figure 5-7, NM-3T3
ceUs were adherent and spread to form fibroblastic cells which resembled the fibroblasts h m primary
BM or liver cultures- On the other hand, AMLl2 cells had polygond morphology with distinct
inuacellular features such as nucleus and endoplasmic reticulum.
The rate of albumin secretion from NIH-37'3 and AMLl2 cells was also measured. Albumin was
not detected h m NM-3T3 cells (c 2 x 104 ngi100 cellhiay); however, AML12 ceils secreced aibumin
at approxirnately 0.044.02 ngI100 cells/day. This rate is lower than the secretion rates obtained h m
primary hepatocyte spheroid cultures or in reactors, which has been reported to be on the order of L Co
10 ngllûû celldday respe~t ive l~2~~. Furthemore, no significant difference in the rate of aibumin
secretion from AMLl2 cells was observed with time (data not show). Despite a documnted decrease
in hep;itic funcuon with timelgl, the strict use of AML12 cells with passage numbers between 37 to 47
facilitateci consistent hepatic properties.
The ability of TROMA-1 antibody to stain CK8 in AMi.12, NJH-3T3. prînwy hepatocytes and
isolated BM cells was tested using flow cytornetry. As it can be seen in Figure 5-9, bdh primary
hepatocytes and AML12 cells expressed signif~cantly higher leveis of CKB compiued ta theu rrspective
isotype controls. interestingly, human anti-cytokeratin anui ies that were tested did not stain liver
cytokeratins. though they have been shown to reaa with non-iiver murine endoderm cellslol. As
expected, NM-3T3 or freshly isolated WBM ceUs did not express CK8. The expression of CK18 as
measured using the TROMA-2 antibody was also examined The staining pattern of CKl8 was similar
to that of CK8. AMLl2 ce& expressed CK18, while CK18 expression was not detected in WBM and
NIEI-3T3 celIs (see Figure 5-10).
CKS - PE
c-met -PE
Figure 5-11 Eitp&n of k d h e r i n (a) and c-Met tb) on AMLl2 c e k Red and btack represent btype and Etained taunpies wqecdvely.
To M e r characterize AMLI2 and NM-3T3 celIs. the expression of other niarphological markers
such as E-cadherin and c-Met was also snidied. As iIlusuated in Figure 5-11, a l 2 cells expressed
Eciidherin and c-Met. Table 4 summarizes the expression of various phenotypic marlrers on primary
hepatocytes, AML12 and NM-3T3 cells. As indicated, AMLIZ and NM-3T3 cells did not express the
hemtopoietic markers CD45 and c-kit. In addition, W 1 2 cells expressed al1 hepatic markers that
were examined. whiIe NM-37'3 stained negatively for al1 hese markers.
The use of surface markers was particularly inmguing in the smdy of hepatocytes simie it ailows for
the andysis of celh without the need for fixation or penneabilization. hterestiogiy, although CK8 or
CK8-Lice p a i n expression has k e n reported on the siaface of hepasacytes 2 9 5 a . mitber CK8 nor
CK18 couId be derected on AMLl2 ceus by surface staining. Attempts were aiso made to rinalyze celi
specific aIbumin expression h m hepatocytes. hwever, resuits from tbese expechmts were
inconsistent due to low albumin secretion or heterogeneity of albumin ~ e c c e t i o n ~ ~ ~ h m hepatocytes
(data not shown].
5.2.3 Tracking of bone marrow derived ce& in vitro
in order to distinguish BM derived cells h m f i e r layers, CFSE staining and YFP expression
were used to track cells in cocultures.
5.2.3.1 CFSE
The feasibility of CFSE dye to track BM derived cells was anaiyzed by staining freshly isolated
BM derived cells prior to CO-culture with unstained f i e r ceus. Despite pubiished accounts on the
ability to track cell division dynamics, in our cultures CFSE sraineci BM cells did not maintain high
level of expression after one week. Figure 5-12 shows the flow cytomeuic results from CO-culture
experirnents between CFSE srained WBM and AMLl2 cells. As it cm be seen. despite clear distinction
between initial ce11 populations. the CFSE expression h m BM cells w u not sufficient to dlow for
their tracking after 7 days.
CFSE CFSE
This was surprising considering that co-culnires with labeled and unlabeied NM-3T3 cells seemed
to maintain their CFSE expression for longer periods than BM cells (data not shown). This cari be
amibuted to the significant ce11 death in BM cuItures which may result in the retease of CFSE content
in the form of Iysed ceIi membranes. These lyseci membranes or soluble CFSE may in tum stain the
f i e r cetls. Furthemore. primitive BM progenitors quickly piif' in culture. rwdting in the rapid
decrease of dye expression in theiu progeny.
UT a
Y lu
a' a' d d
CFSE
5.2.3.2 Yeiiow fluorescent protein (YFP)
To overcome the dificulties encountered in tracking OSE labeled BM cells in cocultures, BM
ceiis derived h m YFP+ mice were utilized. These mice had normal hematopoietic systems and their
BM ceils were capable of giving rise to vananous types of CFCs (see Figure 5-13). Furthemore, YFP
expression was smng enough to be demted h m individual ceiis that fonned CFC colonies. This
property was used to distinguisti BM derived CFC colonies h m ocher colonies during analysis.
Figure 5-13 YFP expression of CF'ü-CM h m Y W SM cdh in CM: m y .
(Bar idkates 50 pm-rultPrrs wtre bPrvntcd 742 days dlcr iaftttka)
Flow cytometric analysis confirmeci that nucleaced BM ceiis expresseci significantiy higher levels of
YFP in cornparison with background feeder cetIs (such as AMLl2 ceus -Figure 5-14 (a*)). However,
since the YFP gene shuts down at later stages of erythroid differentiatim, the anaiysis h m WBM
(which included erythnicytes), indicated the presence of a peak with low YFP expression which was
elitninaced after erythmcytes were lysed. Lin- subpopulation expresseci high levels of YFP. and
maintained its expression after 7 days in culture with or without the presence of background feeder cells
(Figure 5-14 (d-0).
GFP
GFP
GFP
Figure 5-14 BM dedveâ cek h m Y W mke hd dipElagulpbobly higber YFP Iluo- thn m l 2 celk
Baseci on the ability to dishguish YFP+ BM derived cells h m YFF feeder ceiis, aii subsequent
coculture experiments utilized YFP+ œlls as the desired tracking tool.
53 Effects of in vitro cultures on hematopoietic progenitor celi fate
Although the primary focus of tbis project was to study the endoderm differentiation of BM derived
celis, the creation of these cultures revealed imptant in fmt ion about the behavior of hemaiopoietic
cells in liver-li enviromnts.
5.3.1 E m s of AMLL2 itnù MH13T3 fibrobkrsts on hernoropoietk
dtgerenîùation of BM derived cells
As an initial approach to understand the effects of the Liver microenviro~aent on BM derived
progenitor ceIl fate, Lin- cells were cocultured with AML12 hepatocytes and NIH-3T3 fibmblasts. The
heuutopoietic differentiation of BM derived pmgenitors in various conditions was examind based oa
YFP+ ceIl counts as well as CFC assays. As demonstrateci in Figure 5-15, a significantly higher number
of cells were derived h m cocultures with fibroblasts in cornparison with AMLl2 or without f&r
layers. NIH-3T3 cocultures led to >3 fold expansion of the total number of BM cells within a week
after culture initiation, and they maintained their YFP' ce11 numbers after two weeks of culture.
Surprisingly, cocultures with hepatocytes also allowed for moderate expansion of BM derived cells
after a week of culture. However, the total nurnber of YFP* cells in these cultures greatiy diminished
after two weeks.
With respect to hematopoietic progenitor maintenance, cocultures with fibmblast and hepatocyte
feeders maintained similar frequencies of CFCs CO totd BM derived cells throughout the experiment
(see Figure 5-16). However, cultures without addition of exogenous cytokines or fecder Iayen lost
their abiiity to give rise to CFCs wirhin the fUst week of culture. Despite the similarity in the
proportion of CFC in both hepatic and fibroblastic cc~iultures, higher ce11 counts yieIded significantiy
higher overall progenitor numbers in fibroblastic cultures.
F'ïgun 546 Eitecîs of coalîure on tûe tugncy of CFCs îo total in culairs t were InltiatedwithSxl 9 LY& Control repmenk cultures without feeder cl&
Figure 5-17 Eûerts of conilture aindltbn on CFC types, 7 &ys after cultuit initiation.
Cultures pritb ~108e l~us HGF hi dgnkantly Mgbr proportion of BFU-E progeahors io cornparison to the input (pgulpiioo WW tbt hepatic c d t i u a &d Signurcaatly reduœd ratio of BFU-E progeailors Ip c 0.05). No CFCs were dctcded In eomtrol ailQutp aithout focder layen.
Co-culnue conditions affected the types of hematopoietic progenitors producd as wasured by
CFC assay. C o a i ~ e s with fibroblasr ceils gave cise to a significantly higher proportion of BFU-E
colonies campareci to co-cultures with AMLl2. The values for cultures without feeder cells could not
be determiued due CO iite inabiIity to recover suffkient number of cells.
5.3.2 TPO and HGF alter bone rnarrow progenitor ceU frrte
To elucidate the effects of liver growth factors on hematopoietic differentiatiaa of BM progenitors,
two specific cytokines. TW and HGF (lûng/ml) were investigated These growth factors were
patticularly interesMg since they were secreted by hepatic and fibroblastic cells (hepatocyte ce11 i i i
secrece 'PO152 and NIH-3T3 cells secrete HGF~OI). Cultures supplemented with HGF or TPû
yielded higher overail ceil and CFC counts in comprison with cultures without exogenous cytokines
(see Figure 5-18 and Figure 5-19). in gened, cultures supplemented with exogenous TPO generated
more cells. Cultures supplemented oniy with HGF generated lower ceU numbers when compared to
TPO supplemented cultures; however, they generated significantly higher total ce11 numbes tban
cultures without feeder layers. The fresuency of CFCs compareci to cotai number of BM derived cells
ptesent in a coculture at any point in cime decreased for al1 coculntre conditions in a logarithmic
fashion. However, TPO supplemented cultures maintained higher proportions of CFCs during the WC+
week trial in comparison with HGF supplemented cultures. CFC colonies disappeared within the h t
week in BM cultures without exogenous addition of TPO or HGF. Also. HGF supplemented cultures
had a higher proportion of BFU-E progenitors thaa non-HGF supplemented culnues (see Figure 5-20),
+ Conboi
1 .ox1 on - + TPO
- - +HGF O
O 2 4 6 8 10 12 14 16 Time (days)
O 2 4 6 8 10 12 14 16 Time (days)
1 Input 1 HGF 1 TPO
BFU-E CF U-GM CFU-GEMM
Day 7 controt
Figure 5-20 Eûects of HGF snd TPO mppkmntatko on BM derived CFC Types in dtures inithîed with 5 x td~i1ice11safter7t1ays
HCF supplQmeated rulturos bPd a dgdbndy h@er proportion of BFU-E progeaiiors in cornparison wiîh otbcr culhue coadiaoir9 (pdW).
5.3.3 Effects of HGF supplententaiion on hematopoieti. development
The emergence of a higher proportion of BFü-E colonies in CO-culnires with NM-3T3 led us IO
investigate the effects of exogenous HGF on cocultures. To examine the effects of HGF on
hematopoietic differentiation of BM derived cells, HGF was added CO ML12 and MH-3T3 c*
cultures. As shown in Figure 5-21. HGF (10 ng/rni) supplemenration to CO-cultures resulted in an
incre3se in the t o d cells numbers (- 2 fold increase) versus untreated cultures.
Figure 5-21 of HGF suppbmentation on YFP* Figun 5-22 EIIcets of HGF suppkmcalition on ceii couiib in cultures inittted with 5 x ld the frequencg of CFC cobnies to Lii'œlls. tohi œlb at sny p&t in for
cuihves initiated Witb 5 x 10s Lin- cells.
BFU-E CFU-GM
HGF HGF
Figure 5-23 mec& of HCF Suppkmentatbn on CFC types m cultures inithîed wi(b 5 x ld cdk 8t day 7.
As üiustrated the ratio of BFU-E colonies was micb bwer b eoailbucs 4th AMLlZ allP hi cornparbon with HGF rrlls. Furüermirr, exqpous oddihbn d EGF enband the proportion of BFU-E c o b n k Tbc exogewus addition of HGF to tlbrobl.stir dûuci did mt dgdkmtiy enbance tbe proportion of BFü-E cobaia whüe its uidîthn to bcpotic colhvcs kceased the pmportion of BIW-E cnïtnres (p ~0.0s) .
Also, CO-cultures supplemented with HGF eahaaced the total number of CFCs and expressed higher
BM derived ceil counts (see Figure 5-23 and Figure 5-22). Interestingly, the addition of HGF to
hepatocyte feeder layers also enhanced the proportion as well as the total number of BFU-E progenitocs
compared to cultures with no HGF supplementation.
5.4 Bone marrow derived endoderm differentiation
5.4.1 Co-culhrre of lin' c e h with AMLl2 or 3T3 celh
As a fmt approach to dissect the Iiver micrœnWonment in studying the factors that influence BM
denved hepatic differentiation. YFP' BM cells were cpcultured wich AML12 and 3T3 cells. in initial
experiments, Lin* cells were directly co-cuitured with feeder cells and monitored for CK8 and CK18
expression as well as albumin secretion. Unfortunarely, due to the secretion of albumin by hepatocytes
and AML12 cells, detection of serum albumin was not used in cocultures with AMLL2 cells.
Figure 5-24 Expredon of CKWCKi8 and YFP on MH-3T3 (a), AMLl2 (b), and Lin' alls (d)
Figure 5-24 illustrates the expression of CK8 and YFP on cells that initiated the cocultures. As
shown previously, NM-3'M and AMLl2 cells were distinguishable from YFP+ BM derived cells.
Also, AML12 hepatocytes expressed CK8 and CKl8 which was not expressed on fibroblast or BM
derived cells. In addition, YFP+ CK8'cells were not observed in any of the cultures, thus indicating
that the cells that were used to initiate such cocuInnes did not express both YFP and CK8
simultaneously.
FS PI Peak
It utiiized FSSS ptes (a) toibwed by d&aimiuatlon of evenis in wbich PI-PI peak distribution dkl not increpst Linesrfy (b).
Figue 5-26 CKû and YFP exprcsslon of LW alls afîer two Qys in CO-culîwe with AML12 c e k
In co-cultures containing Lin' and AML12 cells. there was no detectable amount of endoderm
differentiation based on microscopic analysis (see Table 5). However, initial flow cytomemc analysis
yielded results conuaciicting those of the microscopic analysis (see Figure 5-26). These results
indicated that Lin' BM cells (i.e. mature BM populations, which do not have the capability to give rise
to non-hematopoieac cellsLoO) gained CK8 expression whi1e in c ~ u I t u r e with AML12 ceUs. As
c o n f d by fluorescent mimscopy, the presence of CKS'YFP' cells were due to ce11 aggregates that
contained CK8' AML12 and W BM cdls. These aggregates were fonned because cells were not
completely dissociateci pnor to flow cytomeaic anaiysis. Thus, despite FSSC discrunination. the
presence of BM derived cells that were attached to 'sticky' hepatocytes yielded an aggregate co-
expressing YFP and CK8-
By utilizing the doublet discriminator (see Figure 5-25), the percentage YFJ? events that expressed
CK8 due to clumping significantiy decreased. However. even the use of doublet discriminat~ did not
elimùiate this phenomenon (see Figure 5-26).
F i i 5-27 CKS and YFP e x p d n of Lh- di in co.cuIture with AMLl2 œJk, 2 dags (a), 7 days (b) siid IS days (c) Pfkr culture ioitintion.
Figure 5-27 illustrates the flow cytomemc results h m cocultures with AML12 cells, in which,
despite the use of the doublet discrirninator. a srnall population of ce11 aggregates remained. Evidence
that showed no individuai cells co-expressed YFP and CK8 was observed using fluorescent rnicroscopy
(see Table 5). Further evidence of the inability of AMLl2 cells to give rise to CK8' YFP* BM ceils
was dernonstrated by coculturing Lin cells with AML12 cells separaml by a penneable M e r .
in these cultures, a membrane prevented direct ce11 contact between the feeder and BM derived cells.
As shown in Figure 5-28, the flow cytomemc analysis of the BM denved ceils cultured in this manner
did not show the emergence of a W CK8' ce11 population.
Figue 5-28 CK8 and YFP expression of L i n cdls culQrted wiîbout direct cd1 Pr#h AMLIt fccdcr Isyers, 2 dnys (a), 7 days (b) and 15 days (c) after dtum inithtioa,
These experiments demonstrateci the importance of single cell suspensions in the analysis of ca-
cultures that contain hepatocytes. Also, the inabiiity to detect endodenn markers on BM derived cells
indicated that CO-culture with hepatocyte-lile cells was not sufficieut in inducing BM derived hepatic
differentiation.
To investigate the nonpatenchymal (Le. steiiate ceii) component of the liver, BM derived
progenitors were cocultured with NM-3T3 cells, Thse co-cultures were much easier to analyze due
to the absence of hepatic markers (such as albumin, CK8 and CK18) in freshly isolated WBM and
fibroblastic cells. Figure 5-29 indicates the CKS and YFP expression of Li* cells that were CO-
cultured with NIH-3T3 cells through a NO-week period. After 7 days, a population of Lin ceils
expressed CK8 (and CK18) markers. The expression of CK8/CKl8 in the BM cultures pealred on day
7, and then it became undetectable by day 15. In Lin' cells, no CK8+ YFP' cells emerged throughout
the culture period
Albumin was also detected in these ccxuitures over the fmt week of culture (see Figure 5-32).
However, its expression quickly diminished over the culture period. This coincideci with the presence
of YFP'CK8' cells within the culture. These resulis indicated the emergence of a transiently appearing
C K S + W population in cocultures with NM-3T3 cells, which is only present in the Lin*
subpopulation.
Figure 5-29 Flow cytomtric erpression of CK8 oad YFP on BM derivd Lin+(a-b) ami Lin- (cd) celis in co- culture wiîb NIH-3T3 c e k
In Lia* eo-culhtres witb NEL3T3 ails, YFP' CW reIIs did not a p p r after 7 (a) or 15 &ys (b). However, in Lin- caitures YFP+ C W œik ap- after 7 days (cl ami bst tbek CKll expreskn nftbla two weeks olcuiture (d).
5.4.2.1 Exogenous HGF supplementation to CO-cultures
Revious reports. in which HGF has &en shown to induce transition of mesenchymal cells to
epittielia129* and end0derrn2~~ tissues. dong with our findings h m CO-cultures with NM-3T3
fibroblasts (which have been shown to secrete EIGF~~I), ted us to question the effets of HGF in
inducing endcdem differentiarion of murine BM denved ceiis, Aside h m king a hepatic mitogen
that is secreted by mesenchymai œlls (such as €ibmbIasts), HGF has been shown to play a regdatory
d e in murine BM pgenitor ceIl fate decisions. In fact, it was demonstrated that c-Met (HGF
receptor) is expresseci on a large popdation of hematopoietic progwitors (see Sectioa 5.1). Since the
murine c-Met Thus, rWGF was used in subsequent experhents in order to test the effects of HGF on
BM derived endoderm differentiation.
To examine the exogenous effects of HGF, CO-cultures were set up as in previous experiments with
the addition of medium supplemented with HGF. in these cultures, the presence of individual BM
derived cells (i.e. YFP*) expressing endodenn market (Le. CK8) was observed (see Figure 5-30). in JI
cocultures, it was observed chat HGF enhanced the proportion of BM derived cetls that express
endodem markers CK8 and CK18. Table 5 represents the microscopie analysis of the co-culture
systems in which exogenous HGF was added.
Table 5: Pemntage of C U Y W cek detected in various co-cuituros using fluorescent mirtoseopy after 7 and 14 dayr
Day 7
HF medium HF medium + 10 nglml
Lightmicroscopy C K ~
-
HGF Day 14 HF
HF + IO ng/ml HGF
YFP
AMLl2 1 3T3 (96 - number of (+)ve ceh / total number d YFP* ce& andyzed)
Flgure 5-JO The presence of individuai orlls upredmg Cg% rttcr 7 dryd d apniftivc uith MH-3T3 rad AMLl2 cuiûues suppkmiited witô HGF (10adml) was coiiiinned Ihiarraccnt damgmphr
(Bar iadicnîes5ûpm)
in nonnal MH-3T3 cocultures (Le. not initiateci with HGF supplenieated medium), rare BM
denved cells that expressed CK8 (-0.6%) were observed after 7 days of cuitura However, the CKS'
population was no longer detectable after 14 days. In these CO-cultures, the addition of exogenous
HGF increased the overail number and îkquency of CK8' YFP+ cells generated after 7 days of
co-cdtures (see Figure 5-31) and nearIy doubled the rate of aibumin secretion (see Figure 5-32).
59
(Not detected - O / 457) (1.5 % - 8 / 544)
(Not detected - O 1 362) (0.9 % - 5 / 507)
(0.6 8 - 4 / 687) (1.2 8 - 10 1789)
(Not detected - O 1 523) (0.2 % - 1 / 423)
in cocultures with AMLl2 cells, the addition of HGF ted to the generation of rare CKr YFP' ceIIs that
remrrined present after two weeks of culture.
Figue 5-31 CK 8 expression of L k d i s c u i t d with 3T3 fibroMPals in obseaee (a) or preseaee (b) of 10 ng/ml HGF sevea days after initiation of tbt cuibire9
-1 -0-1 2-7 7-9 9-1 5 15-19
Tirne (days)
5.4,2.2 HGF induces hepatocyte differentiation in a dose dependent manner
The dose response of BM derived progenitors to HGF was analyzed by culturing 5 x 1d Lin cells
on dishes pnxoated with hfaaigela and chen rneasrinng the rate of alburnin secretion and ceiiular
proliferation. SM derived Li- cells were cultured m HF medium without s e m for 6 days in a range
of HGF coacentrations. The cells were then cultured in HF medium containing 1û% FBS in the
presence of 10 nghl HGF (see Figure 5-33).
Figure 5-33 Colture conditions wd in experfmcats to determine dl&& of HGF conceatiition.
The percentage of viable cells in these cultures droppeâ significantly within the fmt week Due to
insuficient number of cells retrieved h m each well, ceIl counts could n a be performed. However.
despite this difficulty, no qualitative difference in the number of viable cells in different HGF
concentrations was detected. Based on the limited number of cells retrieved from these cuitures it is
estimated that only about 5 - 10% of the initial Ki' populations remaineci viabte past one week Also,
due to the inability to recover a significant number of cells, only a Limiteci number of cells were used in
fiow cytometnc anaiysis. These results further indicate that c-Met' CK8' cells may be responsible for
the albumin detected in cultures, since their emergence correlateci with the expression of albumin (see
Figure 5-34). Typically, the flow cytometric analysis in these cultures was limiteci to d O O cells, thus
large scale experiments may be required to dernonstrate this phenornenon in a statisticaily sipificant
m e r . As shown in Figure 5-34, the emergence of CKS' cells was detectable on &y 6 in HGF
cultures and was maintaineci at least to day 16. Interestingly, the majority of CKS' celIs coexpress c-
Mec, which funher strengthens the hypothesis that HGF induces their hepatic differentiation behavior.
Day 6 Day 16
c-met - PE c-met - PE
F m 5-34 Flow cytometric naplysis of CKâ ami c-nœt expression on Lin' BM cuitures suppkmented witb exogewus HGF after6 and 16 Qys
Results indicaie tbe of CK8+ ecllP in cuituns initiateci with 1OOO qhü HGF (a-b) wbüt no CKS+ edb were dttPcrcd ia cPlbvcs tht were wt smppkmtntcd with CICF (cd).
Figure 5-35 illustrates the rate of albumin secretion h m various conditions over the 2iday
experïrnent. As indicated. albumin secretion rate was regulated in a dose dependant manner. with
sustained expression (after 21 days) only observed in cultures that were initiateci with 1 pghi HGF.
Interestingiy, al1 cultures containing HGF seemed to secrete detectable levels of albumin above controis
within the first week of culture. However. soon after most cultures initiami with low HGF
concentration quickly last their ability to secrete aibumia. This behavior was similar Co NM-3T3 CO-
cultures in which albumin-secrethg cells were deiected in culture at early t h e points.
D a y 7-1 2 D a y 15-16 D a y t7-21
35 - - * 0
1000 nglm l 100 nglrn l 1 0 ng/ m I O nglm I Matrigel
H G F Concent ra t ion Figure 5-35 Etrects of HGF conuntrrition on albumin expression in BM ailturcs Ytiatrd wUh 5 x 1# Uri ceiln
(* indicates pdI.05 in compPrison with non-HGF suppkmooted c o a W
Interestingiy, in cultures that continuously secreted albumin, spheroids containhg -5-20 ceUs were
observed These spheroids were morphologically similar to the aggregates f d when AML12 ceUs
were culnued on ~auigel@ (see Figure 5-36). The presence of non-viable cetls was detected in these
cuinires (dark and granular cells visible in Figure 5-36) indicates chat only a small fraction of Li cclls
swived.
5.4.2.3 Frequency of differentiation and effects of hematopoietic cytokines
on HGF mediateci generation of endoderm-like ceUs
To funher investigate the mechanisms that regdate HGF on inducing BM derived hepatic
differentiation, ttie addition of cytokines that have show to be imporiant in expansion of HSC in
culture was anaiyzed. Also, to define whether BM derived hepatic pmgenitors were hematopoietic.
HSC phenotypically defineci by c-kit' Sca-1' Lin CD49 (KU) markers were tested against Lin' Cil45
ce11 populations. These populations were plated at 1 to 100 KLS cells or 1000 Lin CW5 cells in
medium that was supplemented with eirher 1 pgiml HGF, or HM or combinarion of HGF and HM.
After 17 days, cultures containhg HM or 1 pg/ml HGF plus HM indu& the proliferation of Li
cells while no significant proliferation was detected in cultures that contained HGF aione. in these
cultures, only a Fraction of the wells that were initiated with 1 KLS ce11 contained live cells regardiess
of the type of medium used. However, majority of wells contained cellular debris and dead cells, which
indicated that viable cells had been present previously. Despite this, it was clear h m wells initiated
with 1 10 U S cells that ce11 proliferation was greater in cultures that contained hematopoietic
cytokines. Cultures that contained HM plus HGF formed the largest colonies, providing m e r
evidence that HGF works synergistically with hematopoietic cytokines to enhance hematopoietic ceIl
proliferation.
Albumin secretion was mureci to determine the frequency of albumin secreting cells as well as
the rate of secretion. Figure 5-37 illustrates the arnount of aibumin detected after 17 days h m weils
initiated by 1000 CD45 Lin- cells. No statisticai difference between the cultures was observai (p <
0.05). A background level of albwnin was detected in al1 wells due to the hfatrigela coatiag. The fact
that the various d i a did not significantly modify albumin secretion me indicates that the different
cytokines did not cross-react with EUSA antiùodies. Furthermore, as Figure 5-37 indicates, CWT Lia-
cells did not give rise to progeny which secreted albumin regardiess of the microenvitonmntai crics
provided.
K U in BM BM Lin-HGF Lin-HM Lni-HGF+HM HGF HM HGF+HM
To detetmine the ftwluency of KLS ceils chat gave rise to aIbumin secreting ceIls, albumin
measurements per each well that was initiated h m one KLS ce11 were compared wiîh background
medium (HM medium alone was chosen since ir had the highest background levei). The vafues were
then compared using srringent criteria (p < 0.01) to eliminate false positives. Table 6 indicates the
obtained îi-equencies of wells which secrete albumin and thus contain BM deriveci hepatic cells. As
indicated about 10.1% of wells containing 1 KLS cells in HGF, and only 4.2% af the cells that
combined HGF with HM gave rise to detectable levels of aibumin. No significant albumin was
detected fmm any of the weh containing medium rich only in hematopoietic cytokines.
Tabk 6: Albumin wcretion of KLS ce& phtod individueüy io % weii plates aller 17 dPys hi w k u s
HCF
M O v e weik are determibtd by siringent criteria @ c û.01) fn cornparhm to htghst b.çLgroPnd b&
Funhennore, the average levels of deiectable albumin in various positive wells were higbest in
cuInues containing only exogenous HGF compared to conditions in which HCiF and HM were added
This couid be interpreted either as a proliferative or functional effect of HGF on the resdting celIs.
Thus either the endoderm-like celis ggeaerared in HGF cultures secreted albumin at higher rates than
Hoiirpiopietk medhim
HGF + Hemiîopoietic m d h
(#of (+)w welis /#of total weUs) 1 Irom (+Ive a& (numi) t SD 10.1 46 - (7 / 69)
0 % -(0/58)
4.2 % - (3 1 72)
1718
Noc applicable
9 & 2
HGF+HM conditions or chat they generated more endoderm-like cells. Based on typical albumin
expression of cdtured hepatocytes, the values obtained in these resuits suggest similar values as would
be obtained by the secretion of 1 to 10 hepatocytes (depending on albumin secretion 1evels2~~) in
cultures for the duration of the experiment. Thereiore the observeci proiiferation data in each welI
correlated with the albumin secretion results.
Figure 5-38 Albumin e n after 17 days in 1000 nghi HCF (HCF) with addition of 1000 Lia- (HGFCWS), o r 1 (HCFl), 10 (HCF10) o r 100 (HCF100) KIS c e h to each weü.
(* indicotes @.O1 in comparimn to 1000 @ml HCF abne witbout cens).
Figure 5-38 represents the average albumin secretion h m al1 the plates comprised of 1, 10 and IO0
KLS cells. as well as 1000 Lin' CD4S cells in HGF medium As expectd al1 conditions in which
KLS cells were added to the medium resulted in significantly higher rneasured albumin values.
Surpnsingly, the albumin secretion per well dœs not d i i t l y inmase based on the number of KLS
cells present in each well. This could be due to cellular interactions at early stages of cell fate
determination of KLS cells. The presence of a large number of KLS cells or ttieir hematopoietic
progeny may in fact prevent endodem differentiation. These results are also dernonstrateci in culnues
with medium contaking HGF and hematopoietic cytokiues (see Figure 5-39). In these cultures. the
addition of 10 or LOO KLS cells did n a dramatically in~ease the albumin secreted. However, more
e x p e k n t s are required to determine the exact reasons for these observations.
F'igure 5-39 Albumin seccetion d'ter If days in cultures initiated with HM and 1000 ng/mi HCF (HCFHM) with addition of 1 (HCFHMI), 10 (HGFHM10) or 100 (HCFHM100) K U tells to mcb w&
6 DISCUSSION
6.1 Devetopment of thé CO-culture systerns
A co-cuiture syçtem was developed to analyze the fate of BM derived cells in hepatic
micmenWonments. This was accomplished by fmding techniques to distinguish BM derived ceils
kom feeder Iayers. to cfiaracterize murine hepatocytes and to select proper feeder cells that would
m i i c various aspects of the liver microenvironment.
Despite preliminary attempts to maintain prirnary hepatocytes using various extracellular matrices
and media, these cells did not sustain normal metabolic activity or hepatocyte-specific markers such as
albumin. in addition, the overgrowth of fibroblastic cells in primary hepatic cultures was undesired as
it skewed the cell population dynamics. This overgrowth also prevented the separation of parenchyrnal
h m nonparenchymal cells in the attempt to distlnguish the ce11 type that was responsible for endoderm
différentiation of BM derived cells. As expected, the yield of isolatecl hepatocyte was much lower than
the published values for the two-step collagenase method301; howwer, the main reason for abandoning
the use of primary hepatocytes as the desired feeder layer was their pooriy maintained phenotype in
culture.
As discussed previously, the test experimentd system utilizes AMLIS and MH3T3 ce11 lines.
ïhese ce11 lines were ideal for these studies because they were representative ofhepatocytes and steliate
cells. The AML12 cell Iine expressed al1 functional. phenotypical, and morphoIogica1 characteristics of
hepatocytes that were tested. Furthemore. these cells secreted the hepatic mitogen TGF-a in an
autocrine Iwp. Therefore, signals provided by hbiLl2 ceils provide a suitable microenvironment for
the maintenance of hepatic functiona1 and morphologtcaI feannw.
To mimic the mesenchymai component of the liver micr~e~vironrnent, the NM-3ï3 celi lme was
used. It is believed that the factors produced by liver mesenchymai cells conml the diffefentiation and
growth of hepatocytes. For example in liver development, the cardiac mesoderm is required to induce
endoderut cells to proliferate and form hepatic prirnordium302. Also, a second induction by the
mesenchpal cens is needed to enable celis to express hepatic markers302 The mesenchymal cells
fomd m close proximity with hepatocytes are the stellate celis. These cells are believed to be the niain
source of ECM (collagen, nbronectm, laminin, hep& hepam sulphate) and growth factors.
Furthmore, these ceus are believed to be important U1 controlling the devehpmental fate of oval
cells303. Based on previous researcflO, NiH-3T3 ceIIs have been s h o w to be a good representative
of liver mesenchymal cells. ui fact NIH-3T3 cells induce differentiation of oval cells into hepatocytes
at higher fkquencies than stellate ce11 ~ Ï n e s ~ ~ * . Aithough this is surprishg since steilate ceils teside in
the liver, this behavior may be explained by the inactivation of stellate cells in culture, as they may no
longer produce the growth factors and matrix components necessary to elicit differentiation.
6.2 Effects of in vitro culture conditions on hematopoietic progenitor ceU fate
Although the focus of this work was to study the factors which induce hepatic differentiation of
BM dcrived stem cells, it was shown that CO-cultures with firoblasts increase the nurnber of CFCs and
total cells in comparison to CO-cultures with M I 2 cells. This is not surprising since t?broblast feeder
cells have been used as stroma1 cells for in vitro BM cultures in order to maintain CFCs and LTC-[Cs.
MH-3T3 fibmbiasts secrete numerous cytokines (such as SCF and F L ) ~ ~ ~ that are shown to be
important in HSC maintenance. Furthermore, NM-3T3 cells sccrete biologically active ~ ~ ~ 3 ~ ~ 9 3 0 6 .
which bas been shown to be an important hematopoietic regulatorl 12. Interestingly, NM-3T3 co-
cultures greatIy enhanced the fiequency of BFU-E progeniton. This trend was similar to cultures m
which HGF was added exogenously, whch may suggest that BFU-E enhancing effects of firoblasts
are attributed to HGF. AIso, since HGF acts synergistically with SCF, GM-CSF and IW1 1391309142, it
may combine with the other NiH-3T3 secreted cytokines in enhancing the total hematopoietic ce11
nmbers fiom these CO-cuItures2O~ 130.
Surprisingly, AML12 cells supported moderate proliferation of hematopoietic cells and
maintenance of CFCs. This was the fint demonsuahon that AML.12 cells were capable of supporthg
hematopoiesis. Traditionally, hepatocytes have not been associated with adult hematopoietic
maintenance, mainly due to the difficulties in maintaining primary hepatocytes cultures. However.
embryo derived hepatocyte ce11 lines (such as MMH) have been shown to support in vitro
hematopoiesis. These cells express cytokines involved in the survival and self-renewal of early
progenitor cells (SCF and Fi.), as well as those acting at various stages of progenitor differentiation
(LE, IL-6, GM-CSF, M-CSF, G-CSF and TPO)~~'. AIthough the expression of some of these
cytokines is shut off during development. the expression of TPO and GM-CSF, G-CSF. and FL
continues in adult hepatocytes307 ïhis cytokine profXe may expiain the reason why CO-cuItures with
AMLl2 cells gave rise to considerably higher proportion of CFU-GM colonies m comparison to
erytbroid progenitors. As stated previously, TPO induces megakaryocyte differentiation, whereas GM-
CSF and G-CSF induce formation of granulocytes and monocytes (which can be detected by CFU-GM
Furthermore. TGF-a has been shown to induce granulocytic and monocytic
differentiation of human leukemia cells308- ïherefore. its presence may M e r enhance the formation
of CFU-GM colonies m AMLl2 CO-cultures.
Although membrane bound TGF-a bas been shown to be present on erythroblasts and 0 t h
erythoid ce~s309, AML12 co-cultms did oot enhance erythropoiesis. This could be due to the fact
that TGF-a may require the presence of TGF-P tç enhance the self-renewal of erythrocytic
progenitors310. Since TGF-P is not secreted by hepatocytes3 11, the presence of TGF-a alone may not
be sufncient to enhance BFU-E generation.
Co-cultures with supplemented HGF, contained significantly higher overall ce11 numbers,
probably due to synergistic effects of HGF with other hematopoietic cytokines (SCF or IL-
311291309142,143). Furthemore. this effect is shown to be concentration dependent, as indicated by
the increase in overall ce11 numbers in co-cultures with NIH-3T3 cells (since HGF is already present in
these cultures - see Figure 5-21). Based on our analysis, the ratio of the progenitor cells to the total
number of cells within these cultures did not change significantly in the presence of exogenous HGF;
however, the resulting change in the overall number of CFC was due to total population expansion.
Aiso, HGF enhanced the total nurnber of BFü-E cells in both co-cultures, ttrther demonsmting its
importance in erythroid differentiationl12. The BFU-E stimulating activity of HGF is sa11 under
investigation. However, recent data suggests that HGF induces the activation of signal msduction
pathways linked to Epo receptor312 and therefore plays a cmcial role in the commitrnent of rnultipotent
myeloid progrnitors into erythroid lineage.
6.3 Bone marrow derived endoderm differentiation
There is sîrong evidence to support the notion that adult BM contains stem cells that are capable of
differentiating into al1 hematopoietic and hepatic lineages46~100. The goal of this project was to design
in vitro systems to differentiate BM denved cells into hepatocytes, and by doing so. study the factors
which regdate BM denved hepatic differentiation.
BM derived cells were induced to differentiate mto CKü or CK18' endoderm-like cells that
serreted aibumin in certain culture conditions. The CO-cuh..es of Lin- cells with AMLI2 hepatocytes
without HGF did not induce hepatic differentiation fiom BM derived cells, as detected by the
expression of CK8 or CK18. This suggests that soluble or bound signals by the AMLIS ceII Iine are
not sufiicient to induce differentiation of BM cells into hepatocytes-like cells. Furthemore, these
results suggest that TGF-a secreted m these cultures does not induce BM derived endodm
differentiation. However. the addition of HGF to these CO-cultures resulted in the formation of CK8'
YFP* cells. Therefore, even though TGF-a and other signals tiom AMLl2 cells did not induce BM
deriveci endodenn differentiation, they may be helpfii in rnaintaining differentiated phenotype of BM
derived hepatocyte-like ceils.
The CO-cultures with NiH-3T3 cells gave rise to a rare ce11 population ( 4 % ) which transiently
expressed CK8 / CKI8 as well as secreted albumin. However, even though these cultures secreted
aibumin, the resulting aIbumin expression was much Iess than if al1 the CK8' cells were secreting
aibumin at typicai hepatocyte levels. For example, it was estimated that the albumin secretion rate in
cultured hepatocytes ranged from I to IO ng albumin I day /IO0 hepatocyte294. Based on these
secretion rates, the number of functional1y active BM derived hepatocytes in NM-3T3 co-cultures was
very low. This could be mterpreted in two ways, either a small fiaction of CK8- YFP' cells expressed
albumin or a larger portion of the cells secreted albumin at lower levels. The ability to obtain ceIl
specific albumin measurernents, as detennined by flow ~ ~ t o m e t r y ~ ~ ~ or irnrnunohistochemistry. could
be used to answer this question. If only a mal1 fraction of these cells secreted albumin, then the other
cells rnay bc dedifferentiated hepatocytes or other types of BM derived epithelial tissues such as cells of
lung, liver (cholangiocytes), skin. stomach and gastrointestinai tracf15.
The addition of exogenous HGF to NIH-3T3 CO-cultures increased the rate of albumin secretion as
well as the proportion of YFP' CK8' cells. Therefore. the increased albumin secretion rates may be due
to the increased number of CK8' cells.
In experirnents in which exogenous HGF was added to Lin BM cultures (without feeder cells),
hepatic differentiation was induced in a dose dependent manner. In these cultures, hepatic marken
were maintained only in cultures that were initiated at high HGF concentrations. The need for high
HGF concentrations could be attributed to seved factors. For example, the quick intemalization and
degradation of surface bound HGF/c-Met compIex on hepatocytes3 13. rapidly depletes soluble HGF at
low concentrations. Therefore. the use of heparin (ancüor heparan sulphate p ro teog lycans )~~4~3~~. or
immobilization to (bioreactors or tissue engineering scaffolds) surfaces rnay increase the 'effective'
concentration of HGF. Another reason for hi& HGF concentrations could be due to hgh physiological
concenuations required to induce BM derived endoderm differentiation. For example, it has been
shown that extremely high HGF concentrations were present in blood der liver injury in vivo141.
AIso, perhaps high HGF concentrations are required in vitro. due to de-differentiation of hepatocyte-
iike cells m cuIture at Iow HGF concentrations. This is further supported smce aibumi? was detected
a e r 6 days in ail cultures which were initiated with exogenous HGF. Thus, the use signals that
enhance hepatocyte maintenance m culture may lower the minimum required HGF concentration.
The morphology of the 'candidate' hepatic cells in these cultures greatly resembled those of
AML12 cells on ~ a t r i ~ e l ~ . in both cases. the ceils p w in a clonai marner and gzve rise to spheroids.
Hepatocyte spheroids have been show to enhance the maintenance of hepatic genes in culture, while
reducing the proliferative capacity of the celIs3[6. Based on the aibumin secretion rate of hepatocytes
in primary spheroid cultures, the rate of aibumin secreted kom 5 x 10' BM derived L m cells m 1 pg/ml
of HGF, correlated to about -3000 fimctional hepatocytes. Siace KLS cells constituted about 1% of
Lin' ceiis, this transtates to approximately 5 x 103 KLS cells present within these cultures. Based on the
KLS hepatic efficiency obtained in these results (-IO%), it can be deduced that each KLS ce11
differentiated to hepahc Iineage gave rise to approximately 6-8 hctionaIly active hepatocytes (these
values also correspond to the albumin secretion rates obtained From cional anaiysis). However, some
spberoid colonies had more than 6-8 cells which may indicate the presence of other ceIls (such as bile
duct cells) or hepatocytes with secretion levels below those of primary hepatic cuItures. in any case, it
appears that KLS celb in culture do not, under the conditions tested, have a high proliferative capacity.
Therefore, M e r research may be required to expand the hepatic yeld from BM derived cells. as well
as to identify other ce11 types that are generated from their differentiation in vitro.
It was demonstrated that rare KLS cells have the capability to give rise CO albumin producing cells
in a clonai manner. ïhese results support recent observations by a numbet of groups45*100, who
suggest that KLS are the population responsible for liver engraftment in BM ûansplants. Furthemore,
it was demonstrated that these cells are CD45 (or hematopoietic in origin), which further supports that
HSC (and not other BM derived cells) are capable of gtving rise to both blwd and liver Lineages. The
kquency of KLS cells that gave rise CO albumin secreting cells was -10% in medium containhg HGF.
This indicates that the KLS cells that are highiy enriched for HSC may dso be highIy enriched for cells
capable of giving rise to hepatic cells. The inability to detect albumin in medium that contained only
bernatopoietic cytokines also points to the importance of HGF.
Fibroblasts : Stellate cells Hepatocytes TPO TGF-alpha
Multipotent progenitors
ûîbr diierentiated hematopoietic cells
CKB*/CKl û+ albumin
Hematopoietic differentiation Endoderm differentiation
F@re 6 1 Schemtic of soluble fadors that mey inlluence HSC ceU fate in iiver-ULe microenvlmaments.
Phenotypidy deflaed c-kit* Lia' Sca-1* CD& (which may plpo co-expm C-Met) BM stem nb arc Innuenced in the iiver mkraenironment by dgnaL9 from hepatucytes and stellste-like celk HGF and hemstopjctic cytokinui (SCF, TPO, FL and KL6) may act in cornpethg pathways ta determine early ceU fate dec&ioap bto hematopokdc or endoderm i k q e . HGF gives rise ta CKVlCKlV aibumin secrethg eeils w W bemptopoktic cytakino9 give rise to hemtopaktic cells. Furthemore, UGF enbances the proportion of BFU-E progenitors and may act synecgistidy wiîh hematopiedc cytokines to eohanœ ihe toîai number ofeelln
The frequency of KLS cefls that gave rise to albumin secrethg ceUs in medium comprised of HGF
plus hematopoietic cytokines ( S e , TPO, FL, HL61 was iower than cultures with HGF supplemented
medium. Furthemore, the rate of albumin secretion in positive wells was also significantly lower.
These results may suggest that hematopoietic and hepatic differeatiations are conmlled by cytokine
networks that regulate their behavior in competing pathways (see Figure 6-1). Therefore.
hematopoietic cytokines may induce the hematopoietic differentiation of BM derived ceus, while Liver
regenerating cytokines may induce hepatic differentiation of BM denved cells. Although the absence
of hematopoietic cytokines resuIts in apoptosis or necrosis of BM derived ceiis, it is possible chat HGF
mduces HSC to differentiate into endodenn ceiis independent of the action of hematopoietic cytokines;
therefore, the presence of hemaropoietic cytokines may compete in directing the differentiation of BM
derived HSC into endodermal or mesenchymal fates. These explanations are M e r mpported in NM-
3T3 cocuitures (that seccete hematopoietic cytokiws), in which Iower amounts of albumin is detected
m cornparison to cultures that are only supplernented with HGF. Furthennoce, the expennients in
which more ttian one KLS ce11 was plated per well may also be explamed by this 'competing pathway'
hypothesis. In these cultures. some celfs may have genented progeny that secreted hematopoietic
cytokines that may W e r inhibit the hepatic differentiation of BM derived cells. In addition, this
hypothesis rnay be used to explain the reason that hepatic differentiation of BM derived cells has not
been dernonstrated in any other tissues.
The explanation above assumes a deterministic (or instructive) effect of cytokines on HS ceIl fate
decisions251. in this scheme, cytokines and gowth factors induce the differentiation of stem cells. For
example, HGF induces BM derived cells to differentiate into hepatic lineage. .4n alternative theory
suggests a stochastic r n e ~ h a n i s m ~ ~ ~ in which HSC fate is determined inninsically. Therefore, a stem
cell will have a pre-detennined probability of giving rise to a progenitor ceIl of a specific Imeage whose
survival and proliferation is determined by microenvironmental factors. Based on this scheme. HGF
allows for the ntrvival and prohferation of independently committed hepatic progenitors. The fact that
no BM derived hepatocytes-like ceils were detected in liver regenerating conditions (Le. CO-culture with
TGFir semting AML 12 cells) rnay suggest that the effect of HGF is deterministic. However, because
of the uncminties with the other signals that AML 12 cells secrete this alone cannot exclude the
possibility that the endoderm differentiation h m HSC is stochastic. The evidence regarding
embryonic lethality of HGF/c-Met knockout mice (due to retarded liver developtnent12~ further
supports that HGF is essential in liver development and may induce hepatic differentiation in a
deterministic manner. However. this could also be expiained in that only HGF (and no other liver
regenerating cytokine) allows for the sumval of the BM derived liver progenitor. Clearly. more
research is required to d e t m n e the stochtic or deterministic nature of HGF on HSC differentiation.
7 CONCLUSIONS AND RECOMMENDATIONS
7.1 Conclusions
The behavior of BM derived cells in liver-like environments was stiidied W V BM derived ceiis
were easily tracked in CO-cultures with fibroblasdc and hepatic cells which did nut express the
fluorescent geae. The ability to m k BM derived popdation r e v d e d significant information
regardiig the frequeacy and the type of progeny generated by these cells in CO-cultures (see Figure 6-1).
It was observed that hepatocyte ce11 lines eahanced bernatopoiesis in vitro compated to BM cuihrres
without feeder layers. However, this maintenance was not as profound as the degree of maintenance
provideci in cocultures with fibroblasts. Furthennorie. fibcoblastic cultures seemed to eahauce the
generation of erythroid progenitors, wMe hepatocytes did not support the formation of BFü-E colonies
in a nianaer chat may suggest the importance of HGF.
The ability of BM derived celIs to generate endoderm-lï cells in vitro was anaiyzed. It was
observed chat YFP* CK8* endodena-like cells were not generated in co-cdturts with hepatocytes.
However, in fibroblast cultures, transiently appearing YFP* CKS* cells emerged within the fmc week of
culture. To M e r study tbese mchanisms, the role of HGF on BM derived cells was analyzed. It
was observed dia[ c-Mec (HGF receptor) expression is enricheci in BM progenitors. Funkrmore, HGF
supplementation enhanced the frequency of BM derived endodenn cells. The effects of HGF
concentration on BM derived ce11 indicated rhat hepatocyte-like cells appeared in these cultures in a
dose dependant m e r with high concentrations (1 pgld HGF) inducing long-tem differentiation of
BM derived cells into hepatocytes. The clonal ability of purifieci HSC to give rise to aibumin secrethg
cells was also analyzed. Ir was observesi chat -10% of KLS cells gave rise to albumin secrctiog ccUs
when exposed to HGF rich mdium, whiie no albumin was detected in cultures without «togenous
HGF. Also. the addition of hematopoietic cytokines (SCFI TPO. Hao and FL) reduced tht kquency
as well as the amount of albumin secreted in these celis. Togethet, these results ïndicate that HGF
secreted by mesenchycrral celis may reguiate the differentiation of BM ceils to hepatocytes in culture.
This understanding may Iead CO the deveIopment of U1 virm cultures in order to expand hepatocytes or
pre-treat cells prior to uansplantation.
73 Recommendations and future work
Future w o k should be aimed ar funher development a d optimization of the culture cwditiotls
developed in this thesis. To fuaher demonsuate the hepatic nature of the ceils derived in thse cultures,
the ability of individuai cells to ceexpress cytokerarin Ifs aImg with albumia and otber eady (nich as
AFP and GGT) and late iiver specific markers (mdi as W450 and urea) shauId k tramiaari. This is
i m p o m t in deterrnining the functionality of tbe genenued bepatocytes, since it bas kcn shown ihat
primary hepatocyte cultures which express alb-Jmin and cytokeratin markers, quickiy lose their ability
to express other functionai hepatic markers such as ~ 1 ~ 4 5 0 2 9 4 .
Also, albumin production reporteci in this thesis is calculated based on tirne-averaged albumin
production. These results do not necessarily reflect the profile of secretion within tbe specified pend
For example, the aibumin secretion over a week of culture can be achieved in a day within the specified
time as supposed to being evenly secreted over the entire length of the test. Due to these factors, an
aliquot of the serum should be taken each &y rather than at medium exchanges.
The need for extremely high concentrations of HGF must be e l i i a t e d in order to enhance t&e
commercial viability of this approach in clinical applications. A number of appmaches can be iaLen to
enhance the functional effectiveness of HGF. For example, if one of the fatures of hi& HGF
concentration is to allow for in vitro maintenance of hepatocytes, then the addition of other factors that
enbance hepatocyte maintenance may decrease the need for HGF (such as the presenœ of floating
~ a t r i ~ e P 3 l 7 which has been shown to stabilize the differentiated phenotype on cultured hepatocytes).
Aiso, as previoudy mentioned, HGF can be covdently bound to culture dishes or tissue engineering
scaffold surfaces in order to decrease the internalition of HGFIc-Met cornplex. Another approach is
to add factors that may enhance the stability of HGF. For example, HGF binâs not only to c-Met but
also to heparan sulphate proteoglycans on ceIl surface and soluble heparin in se~m31~-315$18319.
Heparin in low concentrations has been shown to enhance HGF potency in the activation of c-Met
receptor. most likely through the stabilization of the molecules and enhanced receptor
dimerization314,315. in addition. the basai HGF production by fibroblasts may be upcegulated by the
addition of heparin or phorbol ester125, IL-1, ~ ~ ~ - a ~ ~ ~ * ~ ~ ~ * ~ ~ 0 * 3 ~ ~ , EGF, PDGF, aFGF. ~FGFL~S
and agents such as forskolin, cholera toxin and ~~~322-324- Therefore, the addition of fihblasrs u,
cu1ture.s that are supplemented with appropriate stimulatory factors rnay upreguiate HGF Icvels to
sufficiently high values without exogenous addition of HGF.
It was demonstrated that the addition of cytokines found to be important in cr vivo expansion of
HSC was not effective in increasing the production of albumin or the frequency of ceIIs which gave rise
to albumin secrethg cells. Therefore the use of other cytokines which are deemed to k hportant in
liver regeneration. such as EGF, FGF, IL+, TNF-a, insuiii, TGF-a and 0 ~ ~ 3 2 5 , shodd be cvaluated
in determinhg the optimum 'soup' required for differentiating and expanding BM derived hepaiocytes
in vitro*
To utiIize this cuIture system to enhance Lver therapy, future studies couid utilize in w b modeh
(such as the F& or partial hepatectomy or CCL treated mice), to determine the f~a~lii l i ty of
'priming' HSC in cuIture prior to implantation. Various studies can be perfanned chat invotve prc-
culturing celb with HGF, EGF or other hepatic regenerating cytokines priw to transplantatia
However, since these cytokines may induce downstream endoderm diffmntiation in cuIture (whicti
may decrease their capability to engraft the host and expand). the use of cytokine coclaails known to
expand LT-HSC may be the p r e f e d approach in pre-freating BM ceils prior CO tramplanration. It
would be interesthg CO study the relationship between the BM and hepatic engraftment under similar
liver regmerathg conditions. If long-wm hematopoietic engraftment cornlates with bver en&raftment
then primitive cells dong with regenerative liver microenvironment are the key componenw in restocing
iiver function. Thus future research hto co-mlanting geneticaiiy engineered ceUs or other factors
rbat modify the existing Iiver microeavironmtnt may be ultiniately critical to successful Iiver therapies.
It is aiso important to verify th& hematopoietic reconstitution and liver engcafmient occur in the c-Met'
popuIation of MC.
8 REFERENCES NiWason LE, Langer R Prospects for organ and tissue rrplacemenr Jama. 2001;285(5):573-6.
Plan JL. New directions for organ aansplantatiou. Nature. 1998;392(6679 Supp1):ll-7.
Petit-Zeman S. Regenerative medicine. Nat Biorechnol. 2001;19(3):201-6.
Gage FH. CeU therapy. Nature. 1998392(6679 Suppl):18-24.
Nerem RM. CeUular engineering. Ann Biorned Eng. 199 1 ; l9(5):529-45.
Langer R, Vacanti P. Tie engineering. Science. 1993;260(51 I0):920-6.
Nabel GJ. A transformed view of cyclosporine. Nature. 1999;397(6719):471-2.
PERV &ta renew xeno debate. Nat Biotechnol. 2000;18(10):1032-3.
Hayûick L, The ümited in viko üfierimc of human diploid ce11 sirains. &ut. Ce11 Res.
1965;37:614-636.
Hascgawa N, Yamamoto M, Imamura T. Mitsui Y, Yamamoto K. Evaluation of long-tcrm
culnucd endothelid c e k as a mode1 system for siudying v m l a r ageing. Mech Ageing Dev.
l988;46( 1-3):111-U.
Noishiki Y. Dreams for the fume m the field of in vivo tissue engineering. Amr Orgm.
200 1 ;25(3): 159-63.
Peny D. Patients' voices: the pwcrfid sound in the stem ceU debate. Science.
2000;287(5457): 1423.
Anderson DJ, Gage M, Weissman IL. Can stem celis cross limage boundaries? Nat Med.
200 1;7(4)393-5.
Manin GR isolation of a pluïpomit ceii h e €rom early moue embryos cultured in medium
conditioned by teratocarchoma stcm celis. froc Nat1 Acad Sci USA. 1981;78(12):7634-8.
Evans MJ, K a u h MH. Establishent in culture of plwipotential c e k h m mouse embryos.
Narure. 1981;292(58 19): 154-6.
Shamblott MI, AxcIman I, Wang S. Bugg EM, LittlcfieId JW, Donovan PI, Blumenthal PD,
Huggins GR. Gearhart SD. Derivation of pluripotent stem cclls h m culturrd human primordial
g m ceiis. Proc Nad Acnd Sci U S A. I998;95(23):13726-3 1.
Thomson JA, Ifskovik-Eldor l, Shapùo SS, Waknih MA, Swicrgiel JJ, Matshaii VS. Joncs JM. Embryonic stem ceU lines d&ed fhm human blastocym. Science. 1998;282(5391):1145-7.
Reubinoff BE, Pesa MF, Fong CY, Troumon k Bongso A. Embryonic stem ceU Iines h m
human blastocysts: somatic diffcrentiatian in vitm. Nat Biotechnol. 2000;18(4):399-404.
Paiacios R Goluaski E, Samaridis J. In vitro generation of hematopoietic stem celis h m an
embryonic stem ceU üne. h c N d A d %i US A. 1995;92(16):7530-4.
Bigas A, Martin Di, Bcmstein ID. Gaieration of hcmatopoietic colony-forming cek h m
embryonic stem tek. synergy between a sohtbk Fdctor h m NIH-3ï3 c e k and hematopoietic
growth factors. Blood. t995;85( 1 I):3 127-33.
Hnashmia M, Kataoka Fi, Nshilrawa S. Marsuyoshi N. Maturation of embryonic stem c e k into
endotheiiai celis m an in vitro d e 1 of vasdogencsis. Blood. 1999;93(4):1253-63.
Rohwedcl J, MaItsev V, Bober F, Arnold HH, Heschelcr J, Wobus AM. Muscle cell
differeniiation of embryonic stem cc& reflects myogaicsis in vivo: developmentalIy reguiaad
expression of myogenïc determination genes and fiuictional expression of ionic currents. Da,
Biot. 1994; l64(1):87-lO 1.
Fleischmann M, Bloch W. Kolossov E, Andrcssen C, Muller M, Brem G, Hcsckla I. Addicks
K, Reischmann BK. Cardiac specific expression of the green fluorescent protein dinmg early
muiine embryonic development. FEBS Leu. 1998;440(3)370-6.
Brusde O, Joncs KN, Learish RD, Karram K, Choudhary K, W i d e r OD, Duncan ID, McKay
RD. Embryonic stem celiderived giiai prccursoa: a source of myeharing aansplants. Science.
1999;285(5428):754-6.
Buttery LD, Bourne S. Xynos ID, Wood H, Hughes FJ, Hughes SP. Episkopou V, Polak JM.
DBerentiation of ostcoblasts and in vitro bone formation h m murine embryonic stem celis.
nsue Eng. 200 1;7(1):89-99.
Hamazaki T, iïboshi Y, Oka M, Papa PJ, Meacham AM, Zon LI, Terada N. Hepatic mawation
in differentiating embryonic stem cells in vitro. FEBS Lea 2001;497(1): 15-9.
Lumtlsky N, Blondel O, Laeng P. Velasco i+ Ravin R McKay R DiEtèrentiation of embryonic
stem cells to msulm-secrethg s m i m siniiiar to panmatic islets. Sn'ence.
2OOl;292(5StO): 1389-94.
Wiaston R Embryonic stem ceU rcsearcb The case for. Nat Med. 200 L;ï(4)%JW.
Himiphtrys D, Eggan K. Alnitsu H, Hochediingcr K. Rideout WM, 3rd Biniszkiewicz D,
Yanagimachi R, Jaenisch R. Epigenetic mstability in ES ce& and cloned micc. Science.
200 1;293(5527):95-7.
Denka H. Embryonic stem cck. An cxciting field for basic mcarch and tisnte engineering. but
ais0 an ethical dilemma? Celh ûrgam. 1999;165(3-4):246-9.
Edwards BE, Gtarhart ID, Wailach EE. The h m pluripotent stem ceII: impact on medicine
and socicty. Ferri7 Sterit. 2000,74(1):1-7.
Antoniou M- Embryonic stem ceil mcarch. The case agaînst Nat Med. 22001;7(4):397-9.
Iacnisch R W i 1. Devclopmental biology. Don? clone htnna~s! Science.
200 I ;29 1(55 l3):2552.
Fmari G, Cuseiia-De Angeiis G, Coletta U, Paolucci E Stomaiuolo A, Cossu G, Maviiio F.
Musde regendon by boue marrowdcrivcd myogeaic progcnitots. Seience.
1998;279(5356): 1528-30.
Wakitani S, Sait0 T, Caplan AI. Myogmic ceiis derived h m rat bane marrow mtsenchymai
stem ceIIs urposed to 5-azacytidiue. Murcle Nme. 1995; l8(l2): 14 17-26.
Makino S, Fnkada K, Miyoshi S, Konishi F. Kodama H, Pan l, Sano M, TakahaShi T, Hori S,
Abc H, Hata J, U m m A, Ogawa S. Cardiomyocytes can be generated h m m o w srromal
ceiis m vitro. J Clin invert. 1999;103(5):697-705.
Jackson KA, Mi T, Goodd MA Hematopoietic poteatial of stem ceüs isolaml h m mrnint
S c l d muscle [sec commu1tr]. Proc N d AcadSn' U S A. 1999;96(25):14482-6.
Bittncr RE, Schofer C, WeipoItshannner K, Ivauova S, Strcubcl B, Hauser E Fniiingcr M.
Hoga H, EIbc-Bmger A, Wachtier F. Recraitmcnt of bone-marrow-dcrivcd ceUs by s k e I d and
cardiac muscle m aduit dymaphic mdx mice. Anar Embryol (Bero. 1999; l99(5):39I-6.
Gussani E, Soneoka Y, Süickland CD, B m e y EA, Khaa MK, Fünt AF, KimLel LM, Muiiigan
RC. Dystrophm expression in the mdx mouse restorcd by nem celi transplantation. Nanue.
1999;401(6751):390-4.
Sede P. Rudnidti MA. A new look at the ongin, fimaion, and "Stem-CeUn staw of muscle
satcilite c e k Rocess Citation]. Da, Biol. 2000.21 8(2): 1 15-24.
Ferraxi G, Stornaiuolo A, Mavilio F. Boue-manow aanspfantationFaiim to correct murine
muscuiar dystrophy. N a m . 2001;41 l(684I): 1014-5.
Bjornson CR, Rieee RL, Reynolds BA, Magli MC, Vescovi AL. Tuming brain into blood: a
hanatopoietic fate adqted by adult neural stem c c k in vivo [sec comments]. Science.
1999;283(S401):534-7.
Eglitis MA. Maey E. Hematopoietic tells dinerentiate mto both microglia and macroglia m the
brai^^^ of adult mice. h c Na1 Acad Sci U S A. 1997;94(8):4080-5.
Kopen GC, Rockop DJ, Phimiey DG. Marrow mmal ceüs migrate throughout forcbrain and
ccrcbeiium, and they diffmntiatc into astrocytes a& mjection into neonatai mow brains.
Roc N d Acad Sci U S A. 1999;96(19):10711-6.
K r a w DS, ïheise ND, Coiîector MI, Henegarni O, Hwang S. Gardner R Ncuecl S. !ihdàs
SJ. Muiti-organ, multi-lineage mgraftment by a siagIe boue mnrrow-dmivcd stem MU. Ceil.
2001;105(3)369-77.
Lagassc E, ShUuni JA, Uchida N. Tsukamoto A, Weissman II. Toward regaierative mcdicinc.
Immwiify. 2001: 14(4):425-36.
Morrison SI. Stem ceIl potentiak can anything makc anything? Curr Biol. 200 1; 1 1( 1):R7-9.
Yaxmush ML, Dunn JC. Tompich RG. AssesSnait of amficiai livcr support tcchnology. Cell
T i p l m t . 1992; l(5):323-4 1.
Uludag Et Sefton MV. Microencapsulami human hcpatoma (HcpG2) cek. in vitro growài and
protein rcleax. J Biomed Mater Ra. l993;2ï(lO): 1213-24.
Demeuiou AA, Rcisner A, Sanchez J, kenson SM, Moscioni AD, Chowdimry IR Tr;msplantation of microcarrier-artachecl hepatocytes into 90?h pamaiiy hepatecmrnized rats.
Hepafology. 1988;8(5): 1 W 9 .
Demeaiou AA, Whamg JF, FeIdman D, Levenson SM, Chowdhary NIL Moscioni AD, Kram
M, Chowdhiiry J R Replacement of iiver b t i o n in nts by miqlanQtion of micmcmicr-
attacheci hepatocyts. Science. 1986;U3(4769):1190-2
Gray Et Gass CM. A ~ t o m y of the human body. 29Th Amcrican ed PhiIadelphia: Lea &
Febigcr, 1973.
Faust0 K Bepatocyhe d i f f ~ o u and hcr progrnitor c e k C m Opin Ceil Biol.
l99O;2(6): 1036-42,
Qrlandi F, JezegueI AM. L w mpd h g s ; edîted by F. Orlundi und A M Jaequel. Londori:
Academic Rss; 1972.
PmIa M, Chctseman KH, Biocca ME, D;amani MU, Siater TF. Biochtmical studies on bile
duct cpithclial ccüs isolatcd from rat hm. J Hepatuf. l99O;lq3):M 1-5.
Laskin DL. Siousoiclai Iining ceils and hepatotoxicity. Tdncol Puthol. 1996;24(l):ll2-8.
Ikuta K, Uchida N, Friedman J, Weissman K. Lymphocyte developmmt h m stem cells. Annu
Rev Immwl. 1992;10:759-83.
Moore MA, Metcaif D. Ontogeny of îhe haanopoietic system: yoik sac origin of m vivo and in
vitro colony forming cciis in the developing mause cmbryo. Br J Humatol. 1970; 18(3):279-96.
GuaIdi & Bossard P, Zhcng M. Hamada Y, Coleman IR. Zam KS. Hepatic specification of the
gut mdodcrm in vitro: ctlI signaimg and Wnscnptiod confrol. Cenes Dev. 1996; 10( 13): 1670-
82.
Germain L, BIouin MJ, Marceau N. Biliary epitheliai and hcpatocytic ce11 tineage nlationsh~ps
in embryooic rat livcr as detexmincd by the di€femtial expression of cytokcratins, alpha-
fctoproteia aibumin, aud ceil surface-acposed componmts. Cancer Res. t 988;48( l7):4909-18.
Btouin MJ, Lamy I, Laranger k NoeI M, Coriu A, Gugum-Guiilouzo C, Marceau N.
Speciaiization switch in diffmtiating embryonic rat her progenitor ce& m rrspanse to
sodium butyrate. Exp CefI Res. 1995;217(1):22-30.
Shiojin N, Lemire JM, Fausto N. CeIl üneagts and oval ccII pmgcnimrs in rat Liva
dcveIopmcnt Cancer Ra. 1991 ;51(10)26 t 1-20,
Kinadhita T, Sekignchi T. Xu MT, Ito Y, Kamiya A, Tsuji K, Nakahata T, Miyajima A. Hepatic
diflcmitiation mduccd by oncosah M amuates f d livcr hematopoiûis. Pmc Na11 Acad Sci
US A. 1999;96( L3):7265-70.
Itmg J, Zhcng M, Goldfxb M. Zarct KS. hh!ion of mammalian liver devclopment h m
cndoderm by fibroblast growth factors. Science. 1999;284(5422): 1998-2003.
Stamatoglou SC. Ennc6 C, Manson MM, Hughes RC. Temporai changes in the cxpICSSion and
disaibution of adhesion molecuIcs duriug liva devclpmeat and qencration. J Ce11 Biol.
I992;l M(6): 1507-15.
Taniguchi H, Toyodha T, Fukao K, Nabuchi EL h m c e of hematopoiaic stcm celis m the
adult liver. Nm Med. 19%-2(2):198-203.
WaQnabe H, Miyaji C, Seki S. Ab0 T. c-kit+ mm ccüs and thymocytc prrcursors in the iivcrs
of aduit mice. J EV, M d 19%; 1 %4(2):687-93.
Mitaka T. Heptic stem c e k h m bone marrow ceils to hepatocyas. Biochem Biophys Res
COmmtn. 2001;281(1):I-5.
Selt S. Lnru stem cclls. Mod P&I- I994,7( 1):105-12
Fausto N, Mead JE. Regdation of liver growth: protooncogenes and transforming growth
factors. iub I n v a t989;60(1):4-13.
Vessey CJ, de Ia Haii PM. Hepatic stem ceik a review. Pathology. 2001;33(2):130-41.
Michalopouios GK, DeFrances MC, Livcr ngeneration. Science. 1997$276(5309):60-6.
Grisbarn JW, Coleman WB. Neoformation of iiver epithelial cells: progcnitor celis, stem celis,
and phenotypic transitions [editoriat., comment]. Gamoenterology. 1996; 1 lO(4): 13 1 1-3.
Aeschylus. Prometheus bound; 430 BC.
Michalopoulos GK. Liver ceg~~tation: molecular mechanisms of growth controL Faseb J.
l99O;4(2): 176-87.
AIison MR Regdation of hepatic growttL Physiol Rev. 1986;66(3):499-541.
Rhim JA, Sandgrcn EP, Degen I l , Palmiter RD. Brinster RL. Replacement of diseased mouse
Iiver by hepatic ceU transplantation. Science. 1994;263(5 150): 1149-52.
Seil S. Disaibution of alpha-fc~pain- and aibumin-containing ceüs in the livers of Fischer
rats fed four cycles oCN-2-fluoreny Iacecamidc Cancer R u . 197838(9):3 107-13.
Thorgeirsson SS. Hcpatic stem ceüs in livcr rcgeneration. Fmeb J. 1996; 1û( 1 1): 1249-56.
Taniguchi H, Suzriki A, Bcng Y. Konda R Takada Y. Fukunaga Ki, Seino K. Ywwa K,
Otsuka M, Fdcao K., Nakauchi H. Usefulncss of flow-cytomctric ce11 sorthg for cnrkhmtnt of
hepatic stem and progcnitor ceifs in the livcr. Transplant Roc. 200032(2):249-5 1.
Mitaka T, Mikami M. Sanla GL, Pitot HC, Mochizuici Y. Small cc11 colonies appear in the
primary culture of addt rat hepatocytes m the presence of nicotinamide and epidermal growth
factor. Heparology. 1992; l6(2):4%4@7.
Hho H, Tatcno C, Sato H, Yamasaki C, Karayama S, Kohashi TT Aratani -4, Asahara T. Dohi
K, Yoshizato K. A long-tenn culture of human hepatocytes which show a high growrh potcntial
and express their dinkrentiated phenotypcs. Biochem Biophys Res Commun. 1999;256( 1): 184-
91.
SÛam AI, Crosby HA. Hepatic stcm ceiis. Gtcrt 2000,46(6):743-5.
Tateno C, Yoshizato K. Growth and diffmntiation in culture of clonogcnic hcpatocytes thar
express bath phnotypes of hepatocyns and biliary epithcliai ccUs. Am J Parhol.
1996;149(5):1593-605.
Petersni BE, Bowcn WC, Patrene KD, Mars WM. Sullivan AK, Murase N, Boggs SS,
Greenberger JS, Goff JP. Bone m m w as a potmtial source of hepatic oval ceüs. Science.
1999$284(5417):1168-70.
Fausto N, Lemirt IM, Shiojiri N. C d hcages m hepatic dcvelopment and the identification of
progcnitor c e k m normal and injured tivcr. Proc Soc Erp Bi01 Med. 1993;204(3):237-41.
Lanirt IM, Shiojin N, Faosto N. Oval celt proîifciation and t h origin of smaü hcpatocytes in
i iva injury mduccd by D-galacmsamiae. Am J Pathol. 1991;139(3)535-52.
Novikoff P M , Yam A, Oikawa L Bb-lïke ceII comparanent m carcinogcn-induccd
proliferating bile ductules. Am J Parhot. 1996;148(5):1473-92.
Dabcva MD, M c ? D k Activation, proliferation, and diffcrentiation of progenitor cc& mto
hepatocytcs m the D-@mSamme d l of liver rcgeneration Am J Pafhol.
1993;143(6):1606-20.
Lemi R, Liu MH, Tarsetti F, SIott PA, Atpini G, Zhai WR, Paronetto F, Lerizen R Tavoloni N.
Histogenesis of bile &a-Ure ceIIs prolifcrating during ethionine hepato~arciaogenesis~
Evidence for a biliary cpithelial mant of oval ah. Lab Invest. 1992;66(3):390-402.
Malouf NN, Coleman WB, Grisharn JW, Lininga RA, Madden VJ. Sproul M. Anderson PA.
kdult-derived stem ceils h m the iiva becorne myocytes in the heart m vivo. Am J Przrhol.
2001;15(3t6):1929-35.
Omori N, Omori M. Evarts RP, Teramoto T, Milia UI, Hoang RI, Thorgeinson SS. Partial
cloning of rat CD34 cDNA and e x p d o n during stem ccll- dependent liver rcgeneration m the
adult rat. Heparology. 1997;26(3):720-7.
Petersen BE, Goff JP, Greenbtrger IS, Michalopoulos GK. Hepatic aval cells express the
hematopoietic srcm ceH marker Thy-1 m the rat. Heparology. 1998;27(2):433-45.
Fujio K, Erarts RP, Hu 2, Marsden ER, Thorgeitsson SS. Expression of stem ceU factor and its
rcccptor, c-kit, during liver rcgeneration fiom putative stem cells in addt rat, Lab Invesr.
1994;70(4):5 11-6.
Lemmer ER Shepard EG, BIaicolmcr K, Kirsch RE, Robson SC. Isolation fiorn human fetai
livn of cells C O - e m g CD34 bacmatopoietic stem ccil and CAM 5 2 pancytokeratin
marlrm. J Hepatol. 1998;29(3):450-4.
Baumann U, Crosby HA. Ramani P. Kelly DA, Smin AJ. Expression of the stem ccll factor
meptor c-kit in nomai and diseascd pediatnc livcr: identification of a human hepatic
progenitor c d ? Hepclrology. 19993q 1): 1 12-7.
ïheise ND, Badve S, Saxcna R, Haiegariu O, SeU S. Crawford JM. Krause DS. Derivation of
hepatocytcs fiom bone mamw tells in mice d e r radiation-induced myeloablation. Hepatology.
ZOOO;3 l(I):U5-40.
AIison MR Poulsom R Jeffzry R, DWon AP, Quaglia A, Jacob J. Noveili M, Prentice G,
Williamson J, Wright NA. Hcpatocps fiom non-hcpatic adult stem celis. Nonue.
2000,406(6793):257.
ïheise ND, Nifimalcayaiu M, Gardner R, iiici PB, Morgan G, Tcpcrman L, Hnicgariu O.
Krause DS. Livcr h m bone mamw m humans. Heparology- 200032(1): 1 1-6.
Lagassc E, Cornors H, Ai-Dhaiimy M, Rcitmia M, Do& M, Osbamc i, Wang X, Fmegold M.
Weisman K., Grompe M. Airincd hmtopoietic stem c e k can differentiatc mto hcpatocytes
in vivo. Nat Med 2000,6( 1 1):1229-34.
Martin CA, Salornoni PD, &dm AF. Cytokerah immunorraCtiYity in mouse tissues: siudy of
riiffernt m i e s with a ntw dcttction system, Appl Immunohistochem Molecul Morphol.
2001;9(1):70-3.
Gupta S, Rajvansbi P. Irani AN, PaIcstro CJ, Bbargava KK. integration and proliferation of
trausplanted ceIls m hcpatic pmchyma foilowing D-galactosamine-induccd acutc mm in
F344 rats. J Pathol. 2000,190(2):203-10.
Crosby EIA, Strain M. AduIt ha stem ceiis: bone manow, blood, or iïver derived? Guf.
200 1 ;48(2): 153-4.
Stuart KA, Riordan SM, Lidder S, Crosteiia L, Wüliams R Skouteris GG. Hepatocytt growth
factorlscatter factor-induccd inûactllular signalling. Int J &xp Pathol. 2OOO;8 1( 1): 17-30.
bfort RI, Cody DB, Stuard SB, Randaii CI, Miller C, Ridder GM, Doersen CJ, Richards WG.
Yodcr BK, Wilkinson JE, Woychik RP. The combiaation of epidermd growth factor and
tmnsforming growth factor-beta induces novel phenatypic changes m mouse liver stem ceil
lmes. J Ce11 Sci. 1997; 1 1 O(Pt 24):3 1 17-29.
Gak E, Taylor WG, Chan AM, Rubin JS. Rocessing of hepatocyte growth factor to the
heterodimcric form is r q d for biological activity. FEBS Len. 19923 1 l(1): 17-21.
Stokcr M, Ghaardi E. Perrymaa M, Gray J. Scancr factor is a fibroblast-derivcd modulator of
epithelial ceil mobility. Namre. 1987;327(6119):23942.
Stcb MC, Comogho PM. HGF: a multihinctional growth factor conmlling ceil scattering. h i J
Biochem Cell Biol. 1999;3 I(12): 1357-62.
Ghcrardi E, Stokcr M. Hcpatocytc growth factor-scatter factor: mitogcn, motogcn, and met.
Cancer Celh. 1991;3(6):227-32.
Noji S, Tashiro K, Koyama E, Nohno T, Ohyama K, Taniguchi S, Nakamura T. Expression of
hepatocytc growth factor gent m endothelid and Kupffer ceils of damagcd rat iivcrs, as
nvealed by in situ hybridhtion. Biuchem Biophys Res Commun. 1990;173( 1):42-7.
Prat M. Crepaldi T, Gandino L, Giordano S. Longati P. Comoglio P. C-ttmiinal mcated f o m
of Met, the hepatocytc growth factor rcceptor. Mol Cell Biol, 1991;11(12):5954-62.
Gaümi F. Bagnara GP, Bousi L, Comnc E, Foiienzi A, Simeone A, Comoglio PM. Hepatocytc
growth factor mduccs proWmtion and diffcftntiation of muitipotent and erythroid hcmopoietic
progeniton J Cell Biol. 1994;127(6 Pt 1):1743-54.
Kmiecik TE, Keiia JR, Roscn E, Vande Woude GF. Hepatocytc growth factor is a synagistic
factor for the growth of hcmatopoietic progcnitor celis. Blood. 1992;80(10):2454-7.
So~enberg E, Meyer D, Weidner KM, Bkcbmcier C. Scatter factouhepatocytc growth factor
and its receptor, thc c-met tyrosme kinase, can mcdiaa a signal exchange between mesenchyme
and epithelia during mouse dcvclopwnt JCeli Biol. 1993;123(1):223-35.
Gohda E, Maminaga T, Kataoka R Takebe T, Yamamoto 1. Induction of hepatocyte growth
factor in human skïn fibmbIasts by epidexrnal growth factor, platclet-dcrived growth factor and
fibroblast growth factor- Cytokine- 1994;6(6):633-40.
Takai K, Hara f, Matsumoto K, Hosoi G, Osugi Y, Tawa A, Okada S, Nakamma T. Hepatocyte
growth factor is constihItiveIy produccd by humau bone marrow smmai celis and mdircctly
pmmotcs hcmatopoitsis. Blood. 1997;89(5):I560-5.
Defiauces MC, Wolf HK, Michalopoulos GK, Zamcgar R ïk prtscnce of hepatocytc p w t h
factor m the &veIopmg rat Development. 1992;116(2):387-95.
Sctdcn C, Joncs M. Wade D, Hodgson H. Hcpatotropm d W A cxprwpion m human foetal i i v a
development and m Iiver regcacration. FEES i.ezf- l99O270( 1-2):s 1-4.
Hu 2, Evara RP, Fujio K Marsden Thorgeiman SS. Expression of hcpatocytc growth
fmor and c-met gcncs during hepatic differwtiation and livcr dcvelopmtnt in the rat. Am J
Pathol. 1993; 142(6): 1823-30.
Chan AM, King HW, Deakin EA, Tempest PR Hikens J. Kroezm V, Edwards DR WilIs AJ,
Bmokes f, Cooper CS. Characterization of the mouse met pmtooncogeac. Oltcogene.
1988;2(0):593-9.
Weidner KM, Amkaki N, Bartmaaa G. Vandekerckhove J, Weingart S, Rieder H, Fonatscb C,
Tsubouchi H, khida T, Dailnihara Y, et ai. Evidence for the identity of human scatter factor
and human hepatocyte p w t h factor. h o c Nad Rcad Sci US A. 1991;88(16):7001-5.
Bhargava M, Joseph A, Kricsel J, Halaban R Li Y, Pang S. Goldkrg [, Setter E, Donovan Mk Zamegar R, et al. Scatter factor and hepatocyte growth factor: activitics. propcrties. and
mtchanism. Cell Gmwrh Difer. 1992;3(1):11-20.
Nakamura T, Nishizawa T, Hagiya M. Seki T, Shimonishi M. SugSmua A, Tashiro K, Shmiini
S. MoIccular cloning and cxprcssion of human hcpatocyte growih factor. Nohire.
1989;342(6248):440-3.
Stokcr M, Pcrryman M. An epithclial scam factor rcIeased by embryo firoblasts. J Cell Sci.
1985;77:209-23.
Taichman R ReilIy M, Vtmÿi R Ehrmman K, Emerson S. Hcpatocyte growth factor is
secntcd by ostcoblasts and cooperatnrely pmnits the survival of haematopoietic progaiton. Br
J Hamafol. 200 1 ; 1 12(2):438-48.
Rosen EM, GoIdberg ID, Kacmski BM, Buckhoiz T, Vinter DW. Smooth muscie rclcases an
cpithetial ce11 scatter factor which binds to heparia In Vim Cell Dev Biof. l989;25(2): t63-73.
Zaraegar R Muga S, Rîhija R, Mictialopouios G. T i e dimibution of hcpatopoietm-A: a
heparin-bmdïng polypeptide growth factor for hepatocytes. froc Nad Acud Sci U S A.
1990;87(3):1252-6.
Ramadori G, Ncubauer K, Cldenthal M, Nakamura T, KnitttI T, Schwogler S, Meya zum
Buschenfclde KH. n e genc of hepaiocytc growth factor is expresscd m fat-storing cth of rat
Iiva and is downreguhtcd during c d growtb and by transformhg growd~ factor-bcta Biochem
Biophys Ra Comrmnt. 1992;i83(2):739-42.
Schmidt C, Bladt F, Gocdecke S, Brinlmÿuin V, Zschiesche W, Sharpe M, Ghcrardi E Birchmeier C. Scatter ficiorfi- growth factor is essenaal for hcr dtvelopmcnt. N a m
1995;373(6516):699-702.
N i o T, Ebb El, Nistiino N, Adachi M, ikehara S. Hep- p w t h factor as a
hematopoiaic reguIaror. Bluod. 1995;85(1 1):3093-100.
Uebara Y, Minowa O, Mon C, Shiota EL, Kuno J, Noda T. Kitamiua N. Placentai dcfcct and
embryonic lethality m mice lacking hepatocyte p w t h factorlscatter factor. Nature.
1995;373(65 16):702-5.
Schmidt C, Bladt F, Goedcckc S, Brinianann V, &hiesche W. Sharpe M Gheardi E,
Birchmeia C. Scatter factor/hepatocyte growth factor is essentiai for iivcr development. Nature.
1995;373(65 16):699-702.
Maina F, Casagranda F, Audero E, Simeone A, Comoglio PM, Klein R, Pometto C. Uncouphg
of Grb2 h m the Met receptor in vivo rcveals cornplex mles in rnuscIe devetoprntnt. C d .
1996;87(3):53 1-42.
NaLamura S. Gohda E, Matsunaga T, Yamamoto 1, Minowada J. Production of hepatocyte
growth factor by human haernatopoietic ceii Iines. Cytokine. 1994,6(3):285-94.
Kawaida K. Matsumoto K, Shmÿrzu H, Nakamura T. Hepatocyte growth factor prevcnts acutc
renal faiiure and accelerates m a i rcgeneration in mice. Proc Nad Acad Sci U S A.
l994;g I(lO):4357-6 1.
Nakamura T, Yoshimoto K, Nakayama Y, Tomita Y, Ichihara A. ReqrocaI modulation of
growth and diffemtiated functions of manne rat hepatocytes in ptimary culturc by ceif-ceil
contact and ceil membranes. Proc Nat1 Acad Sci U S A. 1983;80(23):7229-33.
Kimura M. Ogihara M. holiferation of adult rat hepatocytcs by hepatocyte growth faftor is
potentiated by both pheaylephrinc and metaprotercnol. J Phannacol Exp fier.
1997;282(3): 1 146-54.
Tsutsumi O, Kmchi H, Oka T. A physiological role of epidermal growth factor in male
reproductive fitnction. Science. 1986;233(4767):975-7.
Nakamura T, Nowa K, Ichihara A, Kaisc Nt Nishino T. Purification and subunit stnicm of
hepatocyte growth factor from rat platelets. FEBS L m . 1987;224(2):3 1 1-6.
Zarncgar R, Michalopoulos G. Purification and biological characterizarion of human
hepaiopoiefin A, a polypeptide growth factor for hepatocytcs. Cancer R a . 1989;49(12):33 14-
20.
Zarnegar Et, Michalopoulos GK. ïhe many faces of hepamcytc growth factor: fiom
hepatopoiesis to hematopoiesis. J Ce11 Biol. 1995; l29(5): 1 177-80.
Goff JP, Shields DS, Petersen BE, Zajac VF, Michalopouios GK, Greenberger SS. Syncrgistic
effects of hepatocyte growth factor on human cord blood CD34+ progenitor ceh are the d t
of c-met nceptor expression. Stem Ce&. 1996;14(5):592602.
Goff JP , Shields DS, Boggs SS, Gmnberga JS. Effcco of fe~~mbÏnaat cytokines on colmy
formation by irradiated human cord blood CD34+ haImOp0ictic progenitor cells. Radiirt Res.
1997;147(1):61-9.
Yu CZ, Hisha H, Li Y, Lian 2, N i i o T, Toki J, Adachi Y, inaba M, Fan TX, Sm T, Igachi T,
Sogo S, Hosaka N, Song ïH, Xmg J, ikchara S. Stimulatory effects of hepamsyte growth factor
on hemopoiesis of SCFlc- kit systemdeficient mice- Stem Cells. 1998;16( l):667ï.
Gallucci RM, S~meonova PP, TOnnrni W. Luster ML TNF-alpha rtguiates d o r m i n g p w t h
factor-aipha in regentrathtg rmrrine h e r and isolatcd hepatocytes. J Immmol.
2000; 164(2):872-8.
Mead JE, Fausto N. Transfounhg growth factor @ha may be a physiological rrgulator of b e r
regeneration by means of an aumcrhe mechanism. Proc Nat1 Acad Sei U S A. 1989;86(5):I558-
62.
Kaushansky K. 'Ibrombopoiain: the primary rcgulator of platelet production. Bfood.
1995;86(2):419-31.
Cramer EM, Noml F, Guichad 1, Breton-Gorius J, Vainchenkn W. Masse JM, Debiii N.
Ultrastructuce of platekt formation by human megakaryocytes cultured with the Mp1 ligand
Blood. 1997;89(7):233646.
Lok S, Foster DC. The smicturc, biology and pomtiai therapeutic applications of recombinant
thrombopoietin. Stem Ce&. i994; 12(6):586-98.
Kato T, Mamunom A, Ogami K, Tahara T, Monta H, Miyazaki H. Native thrombopoietirc
structure and fimction. S m CelLr. 1998; l6(5):322-8.
Nomura S, Ogami K., Kawamum K, Tsukamoto L Kudo Y, Kanalcura Y, Kitamuta Y. Miyazaki
H, Kato T. CeUular localization of thrombopoietin mRNA in the livn by in situ hybridizaaou
E q Hemarol. 1997-;25(7):565-72.
Shimada Y, Kato T, Ogami K, Horie K, Kokubo A, Kudo Y. Maeda E, Sohma Y, Akahon K
Kawamm K, et al. Roduction of thrombopoietin (TPO) bg rat hepatocytes and hcpatoma ce11
lines. Ejcp Hemarol. 1995;23(13):1388-96.
Vigon ï, Mornon P. Cocault L, Mitjavüa MT, Tambourin P. GisseIbrecht S. Souyri M.
Molecular cloning and characterization of MPL, the human homolog of the v-mpl oncogcne:
identitication of a member of the hematopoietic growth factor receptor superfamily. Froc Nad
Acad &i U S A. 1992;89(12):5640-4.
Solar GP, Kerr WG, Zeiglcr FC, Hess D, D o d u e C, de Sauvage FJ, Eaton DL. Role of c-mpl
in early hanatopoiesis. Blmd. 1998;92(1):440.
Zeigier FC, de Sauvage F, W~dmcr HR, Kciier GA, Donahue C, Schretk RD, Maiioy B. Hass
P. Eaton D, Manhews W. In v i w mcgakaryocytopoietic and thrombopoietic activity of c-mpl
ligand (TPO) on purificd hanatopoietic stem ccüs. Blood. 1994;84(12):4045-52.
Vagi M, Ritchie KA, Shicka E, Storcy C, Roth GJ, Barteimez S. Sustained ex vivo expansion
of hcmatopoietic stem ceils mcdiattd by thrombopoieh Roc Nat1 Acad Sci U S A.
1999;96(14):812&31.
Kimura S. R o b AW, MetcalE D, Aiemndm WS. Hematopoietic stem ccU dtficïmïes m
mice lacking c-Mpi, the reccptor for thrombopoietk Froc Nat1 Acad Sei U S A.
I998;95(3):1195-200.
R m c U V, Borge OJ, Veiby OP. Cardier J, Murphy MI, Jr., Lyman SD, Lok S, Jacobsen SE.
Thrombopoietin, but not erythropaictin, dinctly stimuiatcs muidimage growtù of ptimitive
murine boue marrow progrnitor c c k m synergy with early acting cytokines: distinct
intnaaions with tht ligands for c-kiî and FLT3. Blood. I996;88(12):448 1-92.
Sitnicka E, Lm N, Priestley GV. Fox N, Broudy VC, Wolf NS, Kaushansky K. The eff' of
thrombopoietin on the proliferatiw and diffmntiation of murine hematopoietic stem cek.
Blood. 1996;87(12):4998-5005,
Till JE, McCuUoch EA, RadUrrion Resemch. 196 1; l4:2 13.
Koiier MR, Manchel 1, Palsson 00. Cmportance of parenchymal:stromal ceU ratio for the ex
vivo reconstitution of human hematopoiesis. Stem Celk. 1997; 15(4):305-13.
Koiier h4R Bender IG, Miller WM, Fapoursakis ET. Expansion of primitive human
hematopoietic progenitors in a m i o n biomctor system with IL-3, K.-6, and stem cell factor.
Biotechnologv (N Y ) . 1993;I 1(3)38-63.
Dexter TM, Simmons P, Pumeli RA, Spooncer E, Schofield R. The regdation of hemopoietic
ceU developmect by the s m d ccli enviromnent and diffustIble regdatory molecules. Pmg
Clin Bi01 Res. 1984;148:13-33.
Gupta P, McCarthy JB, Verfaillie CM. Smd fibroblast hcparan sulfate is required for
cytokine-mediated ex vivo mainunancc of human long-tcrm culture-initiating ceh. Blood.
1996;87(8):3229-36.
Gupta P, Oegema TR, Ir., Brazil JJ, Dudek AZ Slungaard A. Verfaillie CM. S m c d y
q x d c heparan d a t e s support primitive human hematopoicsis by fomtion of a
muitimolecular stem ccii niche. Biood. 1998;92(12):4641-51.
Verfaille CM, Bais A, Iida I, McGlavt PB, McCanhy JB. Adhesion of committed human
hernatopoietic pmgcnitors to synthctic peptides fxom the C-terminal hcparin-binding domain of
fibroncctin: cooperation bctwcen the mtcgrin aipha 4 beta 1 and the CD44 adhesion nccptor.
Blood. 1994;84(6): 1802-1 1.
Muller-Sicburg CE. Deryugina E. Tite stroma1 tek' guide IO the stem cell universe. Stem Ceik.
1995; Ki(S):4lï-86,
Sutherland HJ, Eaves CJ, Eaves AC, Dragowska W, Lansdorp PM. Characterization and partial
purification of human m m w cc& capabie of mitiating long-tcrm hematopoicsis in v i n .
Biood. 1989;74(5): 1563-70.
Motrison SJ, Uchida N, Weissman IL. Tfie bioIogy of hcmatopoictic stem ce&. Annu Rev CeiZ
Dev Biol. 1995;1135-71.
VertâiUie CM. Direct coutact betwem human primitive hematopoietic progenitors and bone
marrow stroma is not rrqaired for long-tam m vitro hematopoiesis. Blood. 1992;79(11):2821-6.
Verfaillie C, Hurley R, Bhatia R, McCarthy JB. Role of bone marmw matrix in normal and
ahormai hematopoicsis. Cnt Rev Oncol Hemmol. 1994; 16(3):201-24.
Audet J, Zaadsua PW, Eaves CJ, Piret JM. Advances m hematopoietic stem ceU cuInrre. Curr
Opin Biotechnol. 1998;9(2): 1465 1.
Brondy VC. StexnceU factor and hemaropoiesis. Blood. 1997;90(4):1345-64.
Bodme DM, Seidel NE, Zscbo KM, Otlic D. in vivo admmisaation of stem ceU fnaor to mice
mcreases the absohrtt d e r of phnipotent hcmatopoictic stan cells. Blood. 1993;82(2):445-
55.
Li CL, Johnson GR Stem ceii &or eahaaces the survivai but not the self-renewai of murine
hematopoietic long-tenn rcpopulaciag ceUs. Blood. 1994;84(2):408-14.
Zeigler FC, Beiincn BD, Jordau CT, Spencer SD, Baumhueter S. Canoli KI, Hoolcy J, Bauer K,
Manhews W. Cellular and mo1ecuIar charactcrization of the rok of the fk-21flt-3 recqtor
tyrosine kinase in hematopoietic stcm cth. Blood. 1994;84(8):2422-30.
Rosnet O, Buhring HI, Marchetto S. Rappold T, Lavagna C, Saiuty D, ArnouIet C, Chabannon
C, Kanz L, Hannum C. Bimbaum D. Human FLT3mK2 receptor tyrosine kinase is expresscd
at the surface of nonna1 and malignaat hernatopoietic celis. Leukemia. 1996;10(2):23848.
Yonemura Y, Ku H, Lyman SD, Ogawa M. In vitro expartsion of hematopoietic progenitors md
maintenance of stem ceb: coniparison between R.ï3/FLK-2 Ligand and iUï ligand Blood.
l99?;89(6): 1915-2 1.
Petzer AL, Z a n h PW, P k r IM, Eaves CJ. Differeatial cytokine tffects on primitive
(CD34+CD38-) human hematopoictic ce&: novel rcsponses to FM-ligand and thrombopoietin.
J EXp Med. 1996; 183(6):255l-8.
Audet J. Miller Ci, Rose-John S. kt JM, Eavts CJ. Discinct role of gp130 activation in
promociag seif-reaewal divisions by mitogenically stimuiated murine hematopoietic stem celb.
Proc Nat1 Acad Sci US A. 200 1 ;98(4): 1757-62.
Nandurkar tM, Robb L, Tariiaton D, Bamen ï, Kontgen F, Begiey CG. AduIt mict with
targeted mutation of the UiterIeukia- 1 1 receptor (IL1 IRa) display no& hernatopoiesis. Blood.
199'7;90(6):1148-59.
Miller CL, Eaves U. E--ion m vitro of adult murine hematopoietic stem ce& with
uaasplantable Iympho-myeloid reconstituiing ab-. Proc Nad Acad Sci U S A.
1997;94(2S): 13648-53.
Mackarchtschian K, Hardm SD, Maan KA, Boast S, Goff SP, Lemischka IR. Targeted
disruption of the W f l t 3 gene i d to deficiencies in primitive hematopoietic progcnitors.
Immwrity. 1995;3(1): 147-61.
Berry MN, GriveU AR, Griveil MB, PhilIips SW. Isolatcd hepatocytes-pst, prcsent and hm.
Ce11 Biot Toxi'cot. 1997; l3(4-5):U3-33.
SegIen PO. Pteparation of nt Liver cc&. 3. Enzymatic requiremcnts for tissue dispersion. E q
Ce11 RB. l973;82(2):39 1-8.
Segica PO. PnpamÏon of rat Liva c e k II. EEects of ions and chclators on tissue dispasion.
Enp Cefi Res. l973;76( l):U-30.
Block GD, Locker J, Bowen WC, Petaxn BE, Katyd S. Smm SC, Riley T, Howard TA,
Michalopouias GK. Popdation expansion, clonai growth, and spccifïc différcatiation pamrps m
primaxy cuitutes of hcpamym induced by HGFJSF, EGF and TGF alpha in a chcmidy
&&mi (HGM) Mdium J Ce11 Biol. 1996;132(6):1133-49.
Chen Hi., Wu HL., Fon CC, Chm PJ, Lai MY, Chen DS. Long-term cuIturt of hepatocytes h m
human aduits. J Biomed Sci. 1998;5(6):435-40.
Morin O, Noxmand C. Long-term rnaintcnancc of hcpatocyte hctioaai activity in co-cuitun:
rrquHcmaus for simsoidai endothehi cells and dcxamcthasone. J Cet1 Physiol.
1986;129(1):103-10.
Dumenco L, Ogucy D, Wu J, Mcssier N, Fausto N. Inaoduction of a mufine p53 mutation
comesponding to human codon 249 into a rrmrine hepatocytc ce11 linc rcsults in gmwth
advantagc, but not in transformation. Heporotogy. 1995;22(4 Pt 1): 1279-88.
Wu JC, Meriino G, Fausto N. Establishment and characterization of difiermtiated,
nonaansformcd hepatocyte ceii Lincs dcrived h m mice m g e n i c for tfatlsforming growth
factor aIpha. froc N d Acad Sci US A. L994;9 1(2):674-8.
Tarn C, Bilodeau ML, Bullingcr R i , Adrkani OM. DBercntial immcdiate early gcnc
expression in conditional hcpatitis B vmis PX-aansforrning versus nontransforming hepatocytc
ceii lines. J Bi01 Chem. 1999$274(4):2327-36.
Amicone L, Galimi MA, Spagnoli FM, Tommasini C, De Luca V, Tnpodi M. Temporal and
tissue-specific expression of the MET ORF driven by the complete uausct-iptionai unît of human
MAT gene m aansgenic micc. Gene. !995;162(2).323-8.
Spagnoii FM, Amicone L, Tripodi M. Weiss MC. Identification of a bipotential prccunor ceii in
hcpatic ceU lines dcrived h m transgcnic mice mpmsmg cyto-Met m the iiver. J Ce11 Biol.
1998;143(4):1101-12.
Fiorino AS, Diehl AM. Lin HZ, Lemisch ïR, Reid LM. Maturation-dependent gene
expression in a conditionaily transfonned liver progcnitor cell h e . In Vitro Ce11 Dev Biol Anim.
1998;34(3):247-58.
Brito JM, MermeIstcin CS. Tempone AJ, Barojcvic R Mast cells can tevert dexamcthasonc-
mcdiated dom-ngulation of smn ceii factor. Eur J Phumacol. 200 1 ;4 14( 1): 105- 12,
Gaca MD, Pickering SA, Arthur MT, Bayon RC. Human and rat hcpatic stciiate ce& produce
stem ceii factor: a possible rnechanism for mast ce11 rrcniimient in Iiva fibrosis. J Heparol.
1999;30(5):850-8.
Bnto lM, Borojevic R Livn grdomas in schistosomiasix mast ceIldependent induction of
SCF expression in hepatic stchtc ceIls is mediated by TNF-alpha, J Leuikoc Biol.
1997;62(3):389-96.
Finotto S, Buetke Ad, Lingnau K. Schmia E. Galle PR, N m t h MF. Local adtniriistration of
antisensc phosphmthioate oiigonncltotidts to the c-kit ligand, stem ceU factor, supptesses
airway idammation and IL-4 pmduction in a murine mode1 of asthma. J AIlergy Clin Inununol.
2001; 107(2):27946.
Lazaro CA, Rhim JA, Yamada Y, Fausto N, Genaation of hepatocytes h m ovaI c d
precunon in culture. Cancer Res. 1998;58(23):55 14-22.
Kochhar KS, Johnson ME, Volpert O, Iym M. Evidcace for aumaine basis of mrd"rmati011
m NïH-3T3 celis nansfected with metRIGF receptor genc. Growrh Faciors, 1995;12(4):303-13.
Rong S, Bodescot M, Blair D, Dunn J, NaLamura T, Mizuno K, Park M, Chan A, Aaronson S.
Vande Woude GF. Tumorigenicity of the met proto-oncogene and the gene for hepatocytc
growth factor. Mol Ce11 Biol. 1992;12(11):5152-8.
Rong S. Oskarsson M, Faleno D, Tsarfaty ï, Resau IH, Nakamura T, Rosen E, Hoplans RF, 3rd.
Vande Wou& GF. Tumorigenesis mduced by coexpression of human hepatocyte p w l h factor
and the human met protooncogene leads to high levels of expression of the Ligand and meptor.
Ce11 Growrh D @ i . 1993;4(7):563-9.
Slatcr M. Dynamic interactions of the extracellular ma& Hktol Hisroparhol. 1996; 1 1( 1): 175-
80.
Fiaumenhaft R, Rifkin DB. ExtraceUular matrix rcgulation of yowth factor and protease
activity. Curr Oput Ce11 Biol. l991;3(5):8 17-23.
Schuek EG, Li D, Ornieciusici CJ, Muller-Eberhard U, Kleinman K Elswick B. Guzeiian PS.
Regdation of gene expression in adult rat hcpatocytcs culturcd on a basment membrane
matrix. J Cell Physioi. 1988; 134(3)309-23.
Mitaka T, Sattler GL, Pitot HC. Amho acid-tich medium (Leihvitz L-15) enhances and
prolongs proliferation of prirnary cultumi rat hepatocytcs in thc absence of scrum J CeiI
Physiol, 199 1: 147(3):495-504.
Mather JP, Sato GH. The use of hormone-supplemented serum-fret media in pnmary cultures.
Erp Ce11 Res. 1979;124(1):215-21.
Girottï M. Banting G. TGN38-green fluorcsccnt protcm hybrid proteins cxprcssed in stably
transfcaed cukaryotic cells provide a tool for the ml-tirne, in vivo study of membrane aaffic
pathways and suggcst a possible rok for rafiGN38. J Ceil Sci. L996;109(Pt 12):2915-26.
Cody CW, Prashcr UC, Westler WM, Prendcrgast FG, Ward W. Chernical structure of the
hexapeptide chromophore of the Aequom green- fiuorcsccnt protein. Biochemisny.
1993;32(5):1212$.
Cubin AB, Heim R Adams SR Boyd AE, Gros LA, Tsien RY. Undastanding, improving and
using green fluorcsccnt proteins. Trends Biochm Sci. I99F$l( 1 1):448-55.
Chalfie M, Tu Y, Euskidm G, Ward WW, Rashcr M. Gmn fluorescent pmtem as a marker
for gcne expression. Science. 1994;263(5148):802-5.
Heim R, Cuba AB, Tsien RY. improved green fluorescence. Nime. 1995;373(6516):663-4..
Cormack BP, Valdivia RH, FaIkow S. FACS-optimizcd matanfs of the gmn &tomcent protein
(GR). Gerte. 1996; I73(1)33-8.
Pcppcrkok R, Sguire A, Geley S, Baaiaens PI. Sirmiltvleous detection of multiple green
flnorrsccnt proteins in livt celis by flumccnce iifktimc h g h g mimscopy. Cum Biol.
1999;9(5):269-72.
C h g L, Fu I, Tsukamoto A, HawIey RG. Use of gtcni fluorescent pmtcm variants to manitor
geae uansfer and expnssion m mammai;nn ceh. Nat Biorechnol. 1996;14(5):6069.
Kirnie A, Hasbiyama M. Suda T, ûzawa K. Gceen fiuorescent protein as a selectablc markcr of
retrovirally aansduced hematopoietic progenitors. Stem Celk. 1999;17(4):22632.
Hadjantouakk AK, Gertscnsteiri M, Ikawa M, Okabe M, Nagy A. Genetaring green fluomcmt
micc by gcrmlint transmission of green fluorescent ES cek. Mech Dm. 1998;76(1-2):79-90.
Lybsrger L, Dempsey D, Partcc.son GEX, Piston DW, Kain SR, Cbmcnak R Dual-color flow
cywmcaic detection of fluorescent protcins using smgk-laser (488-nm) excitation. Cyromeq.
1998;3 l(3): 147-52.
Teare GF, H o m PK, Slnak SE. Smith C, Hay JB. Long-tcrm m h g of lymphocytes in vivo:
the migration of PKH-labeled lymphocyas. Cell Immrutol. 199 1; I34( 1): 157-70.
Lyons AB. Analyshg ce11 division in vivo and in vitro using flow cytomctric mcasurcment of
CFSE dye dilution. J Immunul MethadS. 2000;243(1-2): 147-54.
Lyons AB. Parish CR Detemination of Iymphocytc division by fiow cytomctry. J lmunol
Merho&. 1994;t71(t):131-7.
Oostcndorp RA, Audtt I, Eaves CJ. Hi@-remlution tracking of c d division suggests similar
ceU cycle kinetics of hematopoictic stem cells stjmuIatcd m vim and in vivo. Blood.
2000,95(3):855-62.
Friedrich G, Soriauo P. Promotcr Uaps in embryonic stem ceik a genctic s c m n to identify and
mutate developmcntal genes in mice. Genes Dw. 1991;5(9): 15 13-23.
Zambrowia BP, Imamoto A. Ficring S, Hmcnbcrg LA, Kerr WG, Sonan0 P. Disrupiion of
ovalapping tmscripts in the ROSA beta geo 26 gme trap sn;iin Ieads to widcspread expression
of ha-galactosidase m moux d r y o s and hematopoictic celIs. h c Nad Acad Sei U S A.
t997;94(8)3789-94.
Krishna Vauaja D. Sivahimar 8, Jesudasan RA, Smgh L, Janatduiasanna MK, Habibuilah CM.
in vivo identification, strrvivaL and hctional cfnfacy of aanspiantcd hepatocytes in acutc liver
failm micc madcl by FISH using Y-chmmosome ph. Cen Trrmïpht. 1998;7(3)267-73.
van Dekken H, Hagcnbeek k Bauman IG. Detcctiou of host ccüs foUowing sex-mismatched
boue mamw ùanspIanfation by fluorescent m situ hybridization with a Y- chromosome specific
ptobe. hhmia . 1989;3(10):724-8.
Eaw C, Müla C, Colmdy E, Audct J, Oostendorp R Cashman J, Zandstra P. Rosc-John S,
k t J, Eaves A. umoduction to stem ccII biobgy in vitro. îhrshoId to ihe hurefimrrr Am N Y
Acad Sci t999;872:1-8.
Morrïson SJ, Shah NM, Anderson DJ. Rcgulatory mechanimis m stem ceii biology. Cell.
1997;88(3)287-98.
Harrison DE, Stone M, Astte CM. Effects of nansplantation on the primitive
imrmmohematopoietic srcm celL JErp Med. 1990;172(2):13 1-7.
Pl& DH, S a d L. The ctoning of normal "mastn celis in tissue culture. J Ce11 Physiol.
1965;66(3):3 19-24.
Bradley TR, Metcaif D. The growth of mouse bone mamw celis in vitro. Aut J Exp Biol Med
Sci. 1966;44(3)287-99.
Metcalf D. The molecuiar control of ceii divisiou, diffmtiation cornmitment and maturation in
haemopoietic ceiis. N a m . 1989;339(62 19)27-30.
Watt FM. Epidermal stem celis: markers, paaerning and the control of stem ceU fate. Philos
Tram R Soc Lond B Biol Sci. 1998;353(1370):83 1-7.
Smith LG, Weissman IL, Heunftld S. Clonal analysis of hematopoietic srem-ce11 differentianon
in vivo. Roc Nat1 Acad Sci US A. 199 1 ;88(7):2788-92.
Osawa M, Hanada K, Hamada H, Nakauchi H. Long-tm lymphohematopoietic reconstitution
by a single CD34 lowinegative hematopoietic stem ccU. Science. 1996;273(5272):242-5.
Baumhctcr S. Singer MS, Henze1 W. H ~ ~ ~ l e r î c h S. Renz M. Rosen SD. Lasky LA, Binding of
L-selectin to the vascular sialomucin 0 3 4 . Science. 1993;262(5132):436-8.
Bioscicnce B. Research Pmduct Catalog. .2000:42.
GoodeU MA, Brosc K, Paradis G, Canner AS, Mulligan RC. Isolation and functionai prOpemes
of murine hernatopoietic stem celis that are rcpliciating in vivo. J Erp Med. 1996;183(4):1797-
806.
Zanjani ED, Almeida-Porada G, Livingston AG, FIake AW, Ogawa M. Human bone m w
CD34 celis engraft in vivo and undago multilineage expression that includes giving Nc to
CD341 celis. Exp Hemaf~l. 1998;26(4):353-60.
Jones RJ, CoUcctor MI, Barber JP, Vala MS, Facklcr MJ, May WS, Griffin CA. Hawkins AL,
Zehnbauu BA, Hiton I, CoIvin OM, Sharkis SJ. Characterization of mouse
lymphohematopoietic stcm ceih lacking spleen colony-forming activity. Blood. 1996;88(2):487-
91.
Sato T, Lava JH, Ogawa M. Rtyefslile e x p d o n of CD34 by murine hematopoietic stem
cclls. Blood. 1999;94(8):2548-54.
Nakamura Y. Ando K, C b r p J, Kawada H, Sato T. Tmji T, Hotta T, Kato S. Ex vivo
generation of CD34(+) ceils h m CD34(-) hmiatopoietic celis. Blood. l999;94( l2):4OS3-9.
Spangrudc GJ, HeimfeId S, Wcismm K. eurifcation and charactaization of moue
hernatopoietic stem cch. Science. f988;î4l(486I):5862.
LmcCic R, V z 'IT, k!zxn: nV. EnrichnArr; a d f.=io.onal charactcrization of Sm-l+WGA+,
Lm-WGA+, Lm- Sca-l+, and LinSca-I+WGA+ bone marrow c e k h m mice with an Ly-6%
haplotype. Blood. 1993;82(9)3673-83.
S p g d e GJ, Brooks D U Mouse suairi variability m the expression of the hematopoietic stem
ceU antigcn Ly-6A/E by boue marmw ce&. Blood. 1993;82(I 1):3327-32.
Ito M. APan K, Misawa M, Kai S, Haca H. In vitro diffmtiation of murint Sca-l+Lin- ccUs
imo myeIoid, B c d 3ud T cc11 lintagcs. Stem Celh. 19%;14(4):412-8.
van de Rijn M, Hemifeld S, Spangrude GT, Wcissman IL. Mouse hanatopaietic aem-ceil
h g c n Sca-1 is a mcmbcr of tf» Ly-6 antigen famiiy. h c Nutl Acad Sci U S A.
1989;86(12):4634-8.
Morrison SJ, Wantiycz Ah& Hcmmati fID, Wright DE, Weissman IL. Idenafication of a lmeage
af muitipotent hematopoietic progenitots. Develapment. 1997;124(10): 1929-39.
Osawa M. Nalcamura K, Nishi N, Talcahasi N. Toimomoto Y, [noue H, Nakauchi H. in vivo
self-renewal of c-Kir+ Sca-l+ Lin(low1-) hemopoietic stem cells. J Immtlml. 1996;l!i6(9):3207-
14.
Morrison SJ, Weissman IL. The Iang-tmn rcpopulating subset of htmatopoieac stem cclls is
detctministic and isoIatable by phcrrotypc. f rnmf~~i ty . 1994;1(8):661-73,
Macrison SJ, Hcmmati HD, Wandyn AM, Weissman IL. The purification and characterimion
of fctal livcr hcmatopoietic stem cclIs. Proc Natl Acod Sci USA. 1995;92(22): 10302-6.
Kawamoto H. Qhunua K. Katsura Y. Dirrct cvidepcc for the cornmitment of hematopoietic
stem ceUs to T, B and myeIoid Lincages in murine feial Iivcr, Int Immunol. 1997:9(7): 10 1 1-9.
Huang H, Aucrbach R Identification and charactcrïzation of htmatopoittic stem ccüs nom the
yok sac of thc carly mouse embryo. fruc Natl Acad Sei US A. 1993;90(21):101104.
Yaniamoto Y, Yasrrmini R Amou Y, Watmbc N, Nihio N, Toki 3, Fuinihara S, Ikehara S.
Characterization o f p u i p h d blood stem c t h m mice. Blood. 1996;88(2):445-51.
Aihara Y, B u h g HJ, Aihara M, Klein J. An actempt to producc "prc-T ccU hybridomas and m
idennfy k i r antigens. Eur J Immunol. 1986;lq 11): 1391-9.
Spangrudt GJ. Aihara Y. Wcissman IL, KIcm I. The stem ccU antigcns Sca-1 and Sca-2
subdividt thymic and pcriphcral T Iymphocytes uito unique subsea. J Immunol.
1988; 14 11 1 115697-707.
Ogawa M, Matsuzaki Y, NÏÏhikaw S, Hayashi S. Rimisada T, Sudo T, iüna T, Nakauchi K.
Expression and firnction of c-kit in hemupoictic progeaitot ceiis. J Exp Med. 1991;174(1):63-71.
Orlic D, Fischer R Nishilrawa S, Nicnhuis AW+ Bodinc DM. Rirification and charactcrization
of heterogcacous pluripotent htmatopoictic stem ceU populations exprrSSmg high IeveIs of c-kit
r r ~ e p t o t . Blood. 1993;82[3):762-70.
Katayama N, Shih JP, N ï i w a S. Kina T, Clark SC, Ogawa NI. Stage-spcdc m o n of
C-kit potein by mutine hcmatopoietic progcnitocs. Blwd. 1993;82(8)2353-60.
Moore T, Huang S. T m m p p LW, Bamen Y Kumar V. Expsion of CD93 on murine and
human phiripotent hanatopoietic stan ceh. Jlnummol. IW,153(11):4978-87,
Ric+Vargas SA, Weiskopf B, Nrshikawa S, Ommnd DG. c-ld expression by B ceIl prrcarsors
in mouse bone mamw. Stimulation of B ceII gmcsis by in vivo trtatment with anti-c-kit
m t i i . J I~ununol. lW,l52(6)L!84S-Z
h t a K, Weissman IL. Evidence that hematopuidc s î m ceüs express mouse c-kit but do not
depend on factor for hcir g m d o u P m Nuri Acad Sei U S -4. l992;8!l(4): 1502-6.
Uchida N, WeissmPl L Searchg for hcmatopaietic stem c e k cvibct that Thy-1.110 Lin-
Sca-l+ ceüs are the ody stem cells in C5'IBL/Kn-Thy-1.1 bouc marniw. J @ Med.
t992; l is( 1): 175-84.
Wcisjman K. Stem cck. clouai progenito~~, and cummitment ta the thret lymphocyte iincages:
T, B. and NK cck. Immunity. 1(7}:529-3 1.
ïmzkmn SM, Collecter MI, Sharias SI. Hematopoietic stcm celI nacking in vivo: a compatison
of short-ttnn and loag-wrm rcpopulating ceüs. BCod. l999;93(6): 19 16-21.
Coulter B. Personai c~mmimicatioa -2001.
Pan J. Na& S, Sanwgini H. T W a D. Jamgui H. Flow cy<omemc charactcrization of
isolated porcine bepatocytc suspuisions for Iiver support. A r n ~ O r g ~ l ~ ~ . 1996;10( 1 1 ): 1 173-80.
Fuchs E. Coulombe PA. Of micc and men: genttic sicin discases of kcratiu. Ceil.
1992;69(6):899-902.
Fuchs E, Weber K. Intermediate fiiaments: m~~rn, dyi.amics, function, and disease. Annu Rev
Biochem. 1994;63:345-82.
Marceau N, Laranger A. Cytokcratin expression, fibrlllar organization, and mbtlt function in
Iivcr cck. Bi~chem CeII Biol. 1995;73(9-10):619-25.
MOU R F d c WW, Schiller D L Geiger B. KrcpIcr R Tht catalog of human cytokcratins:
patterns of expression in normal epithelia, nimon and cuIturrd cells. Celi. t982;3 l(1): 11-21.
Magin TM. Jorcam Ji. Franke WW. Cytokeracin expression in simple epithciia. U. cDNA
cloning and squencc ctiaracerimcs of bovine cytokcratin A lao. 8). Diferenriarion.
1986J0(3):254-64.
Van Eyken P, Dmntt VI. Cytokcratins and the iivcr. Liver. t!W;l3(3): 113-22.
Frankc W. Mayer D, Schrnid E, Dcnk H. Borcnticund E. Diaermcm of expression of
cytoskeietal proteins in culntmI rat hepafocyts and hcpatoma tek. Exp Cell Res.
198 l;l34(2)245-65.
Banhaoit H, Rice J, Miyai K, Oshima RG. Mid-gestationai Iethality in mice hckmg kcratin 8.
Gents Dev. 1993;7(7A): 1191-202.
Magin TM, Schroder R Lcitgeb S. Wambger F, Zatioukd K., Grund C, MeIton DW. Lessons
hm keratin 18 hockout uticc: formatioa of novei keratin niamcnts, secondary Ioss of kcratm 7
and accumulation of iiver- Specific b t i n &positive aggregates. JCelI Biol. 1998;140(6):1441-
51.
Franke WW, Scbmid E, Osborn M, Weber Ii Da-t mtcnnediate-skd filamnits
distmgurshed by immunofluorcscmcc microscopy. Aw: Narl Acad Sci U S A.
t978;75(10):50344.
Franke WW, Denk H, Kait R, Sehmid E. Biochemical and imrnunological identification of
eytakcratït~ prottins prcscnt in hcpatoc).tcs of rnammalian üva tissue. & Ce11 Res.
1981;131(2)~2!l9-318.
Frankc WW, SchilIa DL, MOU R, Wmîcr S. Schmid E. Engeibrecht I, Denk H, K q l a R,
Phtzer B. Diversity of cytoktratins. Diffmntiation spccitïc expression of cytolerami
polypeptides m epitheiiai cells and Cissucs. J Mol Biol. 1981;153(4):933-59.
Laue EB. Monoclonal anti'bodies prwide specific i n ~ o k c u l a r markm for the study of
cpithtiiai tonofdament organization. J Ce11 Biol. 1982;92(3):665-73.
K d e r R Enilet P, Schncbclm MT, Gaillard J, Jacob F. Rcactivity of monoclonal a n a i e s
agamst intermediate filament proteins during embryonic dcvelopmmt. J Embryol 4 Morphol.
198 1;64:45-60.
Bruiet P. Babmet C, Kemler R Jacob F. Monoclonal anabodies against trophectodem-specific
mark= during mouse blastocya formation. h c Narl Acad Sci U S A. 1980;77(7):4113-7.
Cadrin M, Marceau N, French SW. Cytoketatin of apparent high molecular weight in liven
h m griseofidvin-fed mice. J Hepatol. 1992;14(2-3):226-3 1.
Suzuki A, Zhaig Y, Kondo R Kusakabe M, Takada Y, Fulcao K, Nakauchi H, Taniguchi H.
FIow-cytometric separation and enrichment of hepatic progenitor ceüs m the developing mouse
liver. Hepatology. 2000;32(6):1230-9.
Fischer M. Goldschmin J, Pcschel C, Brakenhoff JP, Kalien KJ, Wollmer A, Grotzingcr J,
Rosdohn S. 1. A bioactive designer cytokiue for human hanatopoietic pmgenitor cell
cpnC;oz ?:ut Biorechnol. l997;15(2): 142-5.
Park M, Dean M. Cooper CS. Schmidt M. O'Brien SJ, Blair DG, Vande Woude GE Mechanisrn
of met oncogene activatioa. Cell. 1986;45(6):895-904.
Bottaro DP, Rubm JS. Faleno DL, Chau AM. Kmi& ïE, Vande Woude GF, Ammon SA.
Identification of the hepatocyte growth factor rcceptor as the c-met proto-oncogcne produn
Scimce. 1991;251(4995):802-4.
Badie B. Scharma J, Klavcr J, Vorpahl J. In vitro moddation of mimglia moaiity by &orna
ceils is mediated by hcpatocyte growth factor/scatter factor. Neurusurgety. 1999;44(5):1077-82;
discussion 1082-3.
Okano J, Shiota G, Kawasaki H. Expression of hcpatocyte gmwth factor (HGF) and HGF
rrceptor (c-met) proteins m hcr diseases: an immnnohistochanicai shidy. Liver.
1999;19(2):151-9.
Mongomcry DC. Design and Anaipis of Experiments. New Y o k John Wiicy & Sans; 1991.
McAdam TA. Milla WM, Papoutsakis ET. Variations in culture pH affect the cionhg
efficicncy and diffemtiation of progenitor cells in ut vivo haernopoicsis. Br J Huenratol.
1%7;97(4):889-95.
A& I. Personal communications. -2001.
Powns M. Pusonai c o d c a t i o n s . -2001.
Hcmbrough TA, Kralovich KR, Li L, Gonias SL. Cytokmtin 8 released by breast canmoma
c c k in vitro binds plasminogen and tissue-type plasminogen advator and promotes
pIasminogcn activation. Bbd tmJ . 1 9 9 6 ; 3 t t 3):763-9.
Hemhugh TA, Vasudevan J, Allicaa MM, GIass WF, 214 Gonias SL. A cytoicaatin 8-liLe
proicm with pIasminogen-bmding activity is prrscot on the a t m a i surfaces of hcpatocytes,
HepG2 cells and b m s t CarCrnoma ceU lines. JCell &i- 199S;108(Pt 3):1071-82.
Moorman AF, De Botr PA, Evans D, CharIts R Lamm W. Expression pattcms of mRNAs
for aipha-fetopmtein and &Umm in the dtvelophg nt: the oniogenwis of hcpatocyte
hetcfo~encity. Hkzochem J. 1990;22(12):6536(3.
Tsarfaty i, Rong S. Resau IH, Meng S. da S h PP, Vande Woudt GF. The Met prota-
oncogeue mcsenchymal to cpithelial ceil convasion, Science. f 994J63(5143):98-tOl.
Oh SH, Miyazalti M. Kouchi H, houe Y, Sakaguchi W Tsuji T. Shima N, Higashio K, Namba
M. Hepatocytc growrh factor induces diffmtiation of adult rat bont mamw cciis iuto a
hcpatocytc Lincage in vitro. Biochem Biophys Res Commun. 7000;279(2):500-4,
Zaniegar R Pacrsen B. DeFrances MC, ~chalapoulos G. Locaiization of hcpatocytc gmwth
factor (HGF) gcne on human chromosome 7. Genomicr. t 992;12(t): 147-50.
Kiaunig JE, Goldblan PI. Hintou DE, Lipsky MM, Chacko J. Tnimp BF. Mousc tivtf ceil
cuIturt. 1. Hepatocytc isohtion. In Viao. 1981;17(10):913-25.
Houssaint E. Diffmntiatiou of the m o u heptic prinmrdium. 1. An analysis of &c
mmctions m h e p a m diffmtiatim CelI D i f i . 1980;9(5):269-79.
AIison MR Golding MH. S a d CE. Pluripotattial l iva stcm c c k f a d t a t k stem cdls
locatcd m the biliary m. Cet1 Profif: 1996&29(7):373-402
P m M. SoIcm F, Goldscfmudt J, S c h h a c k P. Rose-John S. InttfItukin-6 and the soluble
mterieukin-6 ccceptor mhce stem ccn factor and Fit-3L expression in vivo and in vim. Erp
Hematol. 2001 t29(2): 146-55.
Rong S. Scgai S, Anver M, Resau M, Vande Wou& GF. Invasivcncss and metastasis of NM
3T3 celIs induccd by Met-hepamcytc growth hctor/scatm factor autocrine StinniIation. Ruc
N d Acad Sct U S A . l994>l(I l):473 1-5.
Cooper CS, Tempest PR, Bedmÿui MP, HeIdm CH, Brookes P. Amplificatioa and
ovcrntprcssion of the mct gcne m spontanmusIy transf01IlBed MH3T3 mouse fibrobiasts. Embu
J- 1986;5(10):2623-8.
Aiua A, Cicchini C, Banardini S, FedcIt G, Amicm L, Fantoui A, Tripodi M. Hmiatopoictic
nippart and cytokine arprcssion of mraine-stable heparocyte ceil Iincs (MlkW). Hepcrrology.
I998;28(6): 1645-X
Wak TM, Malm C, W a s ~ o l l A. Exprrssron of the ü a n s f i i g g r o d factor aipha
Waiz ïM, Malm C, Nisi~ikawa BK, Wasteson A. T d o r m i a g p w t h fictor-alpha (TGF-
alpha) in human bone mamw: demonsuation of TGF-alpha in erybiroblasts and eosinophilic
prccursor ceh and of epidennal growth factor mcptors in b h s t i h ç c U ~ of ~ t lomonocy t i~
origh Blwd. 1995;85(9):2385-92.
Ganddion O, Schmidt U, Beug H, Samarut J. TGF-bcta coopcrates with TGF-alpha to induce
the self-renewal of normal erythtocytic pmgmitors: evidencc for an autocrine mechanism.
Embo J. 199% 18(lO):2704-8 1.
Btaun L, Mead JE, Panzica M, Milrumo R, Bell GI, Fausto N. Transforming g r o h factor beta
mRNA incrases during iivcr regeueration: a possibk paracrine mechanism af g r o d
regdation. Proc Nat1 Acad Sci U S A. 1988;85(5):153943.
Iguchi T, Sogo S. Hisha H, Taketani S, Adachi Y, Miyazaki R Ogaa H, Masuda S. Sasaki R Ito M. Fukuhara S. ikehara S. HGF activates signal transduction fiom €PO nceptor on buman
cord bIood CD34+/CD45+ cells. Stem Celis. 1999;17(2):82-91.
Naka D, Shimomura T, Yoshiyama Y, Sato M, Ishii T, H m H. iutemalization and degradation
of hepatocyte growth fador in hepatocytes with dowu-regdation of the receptodc-Met. FEES
Lez 1993;329(1-2): 147-52.
Mamimoto K, Tajima H. Okazaki H, Nakamura T. Heparin as an induccr of hepatocyte p w t h
factor. J Biochem (Tokyo). 1993;114(6):820-6.
Zioncheck TF, Richardson L, Liu J, Chang L, King KL, Bennett GL, Fugd P, Chamow SM,
Schwall RH, Stack RI. Sulfatcd oiigosaccharides promotc hepatocyte growth factor amxiatiou
and govem its mitogenic activity. J Biot Chm. 1995-270(28): 1687 1-8.
Koidc N, Shinji T, Tanabe T, Asano K, Kawaguchi M. Sakaguchi K, Koide Y, MorÎ M. Tsuji T.
Continucd bgh aibumin production by multiceiiular spheroids of adult rat hepatocytes formed
m the pmence of iiver-derived protcoglycans. Biochem Biophys Res Commun.
1989:161(1):385-91.
Michaiopoulos G, Sanler GL, Pitot HC. Homona1 regdation and the effects of glucose on
tyrosine aminotransferase activity m adult rat hcpatocytcs d m e d on floating collagai
membranes. Cancer Res. 1978;38(6): 1550-5.
Naka D, Ishii T, Shimomura T, Hishida T, Hara H, Heparia modulates thc receptor-binding and
mitogenic acaivity of hepatocyte growth factor on hepatocytes, Eip CeiZ Res. l993;209(2)3 lï-
24.
Schwall RH, Chang LY, Godowski PI, Kahn DW, nillan KJ, Baucr KD, Zioncheck TF.
Heparin mduces dimerization and c o d m prolif'erative actMty onto the hcpatoqte growth
factor antagonists NKl and NK2. J Cen Ebl. 1996;133(3):709-18.
Maamioto K, Okazaki H, Nakamura T. Up-regdation of hepatucyte growîh factor gene
expression by mtcrieakm-1 m human skm îïïmbiasts. Biochm Biophys R a Comnrwr.
1992;188(1)235-43.
321. M a M, Koyama H, Hmo M, O b S, Terada M, N i i w a Y, Nishino T. Morü H.
Reguiation of release of hepatocyk p a n h factor Emm human promyeIocytic leukemia ceUs,
HL-60, by 1.25-dihydmxyvitamin D3, 12-0- tetndecanoylphorbol 13-acetatc, and di'butyryl
cyclic adenobe monophosphare. Blood. 1993;82(1):53-9.
322. Tamura M, AraLalci N. Tsubouchi H, Takada H, Daihihara Y. Enhancement of human
hepatocyte growth factor production by mterleulcin- 1 alpha and -1 beta and nuwr necrosis
factor-alpha by fibmblasts in cuIture. JBiol Chem. 1993;268(11):8140-5.
323. Matsunaga T, Gohda E, Takebe T, Wu YL, Iwao M. Kataoka H, Yamamoto 1. Expnssron of
hqatocyte growth factor is up-ngulatcd through activation of a CAMP-mediated pathway. Exp
Ce11 Res. 1994;210(2)326-35.
324. Sugiyama A, Arakaki R Ohnishi T, Arakaki N, Daikuhara Y. Takada H. Lipoteichoic acid and
intcrleukin 1 stimulate synergistically production of hepatocyte gmwth factor (scatter factor) in
human gingival fibroblasts in culture. Infcr Immun. 1996;64(4): 1426-3 1.
325. Darlington GJ. Molecular mcchanisms of iivcr devtIoprntnt and diffacntiation. Cun Opin CeIl
Biol. 1999;11(6):678-82.