CA2113555A1 - Yolk sac stem cells - Google Patents

Abstract

The present invention is directed to yolk sac stem cells. In particular, it relates to the characterization, culturing, and uses of yolk sac stem cells for hematopoietic reconstitution and therapy. Yolk sac stem cells isolated from the early embryonic yolk sac prior to blood island formation exhibit a homogeneous morphology and primitive cell surface phenotype without the expression of mature leukocyte markers and major histocompatibility complex encoded antigens. The cells can be cultured and expanded long-term without alteration of their pluripotency. Therefore, yolk sac stem cells may have a wide range of applications including but not limited to the reconstitution of a destroyed or deficient human hematopoietic system, and the construction of large and small animal models for the production of human blood cells, human antibodies, and testing of human diseases, immune function, vaccines, drugs and immunotherapy.

Description

WO93/~2182 2 ~ 1 3 ;~ ~ ~ PCT/~S92/0591~

YOLK SAC STEM CELLS

l. INTRODUCTION
The present invention is directed to yolk sac stem cell In particular, it relates to the characterization, culturing, and uses of yolk sac stem cell~ ~or hematopoietic recon~titution and therapy.
Yolk sac st~m cells isolated from the early embryonic yolk sa prior to blood island formation exhibit a homogeneous morphology and a primitive cell surface phenotype without the expression of mature leukocyte markers and major histocompatibility complex encoded antigens. The cells can be cultured and expanded long-term without alteration of their pluripotency.
lS Therefore, yolk sac stem cells may have a wide range -~
of applications including but not limited to the : reconstlt~tion of a destroyed or deficient human : hematopoietic system, and the construction of large and smalI animal models for the production of human 2 0 blood cells, human antibsdies, and testing of human : disPases~ immune function, Yaccines~ drugs and ;~ immunotherapy. ~ :

2. BACKGROUND~OF THE INVENTION
~ : ~ multip~ten~ial stem cell population is capable~of giving~ riee to blood cells of diverse : :morphology and function ~Golde, l99l, Scientific : ~ .
E American,~December:~6). Since blood cell formation is first detectable ~in~the~:embryonic yolk sac Parly in 30~ embryogen~sis, it has been hypothesized that pluripotent hematopoietic stem cells may be present within the yolk sac, but the charact~ristics of such cells are still:poorly understood and such cells have not hereto~ore been identified (Moore and ~etcalf, : 35~ l970, 18:279). Durin~ fetal dev~lopment, the stem ' :.

SUI~ST~T~T~ S~E~T

WO93/02182 2 ~ 1 3 i 5 5 - 2 - PCT/US92/05918 cells migrate to the fetal liver where they reside temporarily, and eventually move to give rise to the bone marrow which is the permanent site of blood cell formation in the adult. Studies on the development of blood cells have led to the identification of a variety of important growth and differentiation factors that regulate hematopoissis. Further, tissue typing technology has ushered in dramatic advances in the use of hematopoietic stem cells as a form of therapy in patients with deficient or abnormal - hematopoiesis.

2.l. HEMATOPOIETIC STEM CELLS
A pluripotent stem cell is believed to be lS capa~le of self-renewal and differentiation into blood cells of various lineages inc7uding lymphocytes, granulocytes, macrophages/monocytas, erythrocytes and ~ - megakaryocyte (Ikuta et al., lg92, AnnO Rev. Immunol.
:~ 10:759). The mechanism by which a stem cell commits :~ 20 to a specific cell lineage has not been fully elucidated. The mech~nisms involv~d in stem cell replication without differen~iation are also unknown.
; However, it is Glear that such eYents must, in part, :be influenced by a variety of growth and 25 :~differentiation fac~ors that specifically regulate ~ hema~opoiesis. Other factors which are not yet :; ~ : identified may also be involved (Me~calf, 1~89, Nature E : 339:27~. The :commonly known hematopoietic factors include :erythropoietin (EPO), granulocyte~macrophage 30 colony stimulating factor (G/M-C5F), granulocyte colony-~timulating factor (G-CSF), macrophage :colony-stimulating ~M-CSF), interleukin l-l~ (IL-l to IL-12~, and stem cell factor (5CF~.
An~understanding of hematopoie~is is : 35 critical to the therapy of h~matopoietic disorders.

~ESTiTIi~L ~ ET

W~93/02182 - 3 - ~ ~ 1 3 ~ ~ t ~ PCT/US92/05918 Neoplastic transformation, immunodeficiency, genetic abnormalities, and even viral infections all can affect blood cells of different lineages and at different stages of development. For ex~mple, basic knowledge of blood cell development has contributed to the success of bone marrow transplantation in the treatment of certain ~orms of hematopoietic malignancies and anemi~s.
ConYantional therapy utilizes whole bone marrow harvested from the iliac crest but this approach has certain limitations. ~one marrow stem cells are present at extremely low concentrations, and .they may not be at the earliest stage of differentiation. An impediment in bone marrow transplantation is the need for matching the major histocompatibility complex ~C) between donors and : recipi~nts through H~A tissue typing techniques~
~ Matching at major loci within the MHC class I and : ¢las~ II genes is critical to ~he prevention of rejection responses by the recipi~nt against the engrafted cells, and more importantly, donor cells may also mediate an immunological reaction to the host : tissues referred to as graft versus host disea e. In ordPr to facilitate graft acceptance by the host, 2S immunosuppr~ssive agents often have been employed, ~:~ which render the patients susceptible to a wide range of opportunistic infections.
Hollands~examined the in vivo potential of embryonic cells, and ~ound that day 7 embryonic mou~e cells could colonize the hematopoietic system of normal non-irradiated allogeneic mice (Hollands, 1988, British J. Haematol. 69:437). However, it was not clear which embr~o~ic cell population actually contributed to thi~ result, a~ total ambryonic cells were used for in vivo transfsr. In a study on the .

S~ST~T~TE Sr~EET

WO~3/~21~2 _ 4 _ PCT/US92/0~
~ ~ 1 3 . j ~ j effects Qf ~n utero cell transfer, d~y 9 yolk sac cells were injected into syngeneic fetuses which differed from the donor cells only at the ~-globin 1QCUS (Toles et al., 1989, Proc. N~tl. Acad. Sci.
U.S.A. 8~:7456). The donor cells were shown to induce hematopoiesis. Both of these ln ViYo studies utilized freshly isolated cells from mouse embryos, and there was no suggestion that long-term cultured and expanded embryonic yolk sac cells could retain their pluripot~ncy. Such methods involved diffusion chambers ~mbedded within the species o~ origin of the yolk sac tissue (Symann et al., Exp. Hemat. 6:749, .1978~ or method~ that led to ln vitro malignant transformation of the yolk sac cells. For example, long-term yolk sac cell lines were es~ablished from day 10-13 mouse embryos, and they were shown to give rise to tumor cells at high frequency ~Globerson et : al., 1987, Differentiation 36:~85). Therefore, the potential of tumor formation renders such long term cultured cells undesirable for use in reconstitution ~: :
~: therapy.

2.2. N~ ISTOCOMPATIBILITY_COMPLEX
: : The MHC is a highly pol~morphic complex of ~:: : 25 genes (Bach and Sachs, 1987, New Eng. J. Med.
317:489). It was first discovered by its ~lose association with the phensmenon of transplantation rejection of tis~ue grafts. Subse~uent studies conclusi~ely demonstrated that antigens encoded by MHC
class~I gene~ are the major targets of transplantation rejection responses. Such antigens are expressed by all somatic cells.
MHC class II genes encode molecules on a limited array of cells, most of ~hich are re~ated to S~D~TITUTE S~ET

WO93/02182 2 1 ~ S~ r 5 r the hematopoietic system. They can also elicit reactions by allogeneic immune cells.
Studies on the expression of MHC antigens by embryonic yolk sac cells yielded inconsistent results.
S Billington and ~enkinson (Transplantation 18:286, 1974), working with cells of the yolk sac of 10-14 day mouse embryos, found that these cells expres~ed both H-2 and non-H2 ~murine major and minor histocompatibility) antigens. The work of Patthey &
Edidîn (Transplantation 15:211, 1973), cited by Billington and Jenkinson, reported that H-2 antigens : firs~ appeared on day 7 embryos which could provoke a .strong immune reaction, but ~he latter suggested that these antigen~ did not make an appearance in utero until day 9 or la~er. See THE EARLY DEVELOPMENT OF
~AM~ALS 21g (Balls and Wild, eds., Cambridge U.:1975).
.:~ Heyner reported that H-2 antigens were detectable in ~: day 7 mouse embryos (Heyner, 1973, Transplantation Ç75). Further, mous~ yolk sac cells obtained at 2~: day 9 of gestation were shown to be capable of generatin~ a graft-versus-host response in vitro H~fman and Globerson, ~973, Eur. J. Immunol. 3:1793.
However, Parr~et ~al~ demonstra~ed that H-2 antigens were absent on the apical or the laterobasal membrane Z5~of thé mouse yolk~sac~endoderm even at day 20 of pregnancy (Parr:et~;al., 1980, J. Exp. Med. 152:945).
Thus,:no consensu has been established in regard to : ;the an~igenicity of yolk sac cells.
' : 30 3. UMHARY OF THE INVENTION
: The present in~ention relate~ to yolk sac stem cells, a~method of isolating and culturing yolk : ~ sac stem cellR ~ and a method of using the cultur~d yolk sac calls for recsns~ituting an allogen~ic or x~nogen~ic hematopoietic ~yste~.
' S~ T~TUTL S~d~E~ ~

WO93/0218? 6 PcT/us92/o5~l8 21 13 ~
. The invention is based, in part, on Applicants' discovery that the murine yolk sac, isolated from mouse embryos prior to visible blood island formation, contains a homogeneous population of cells that are CD34+, Thy-1 , MHC class I and ~lass II . Such cells can be e~panded in number by long-term in vitro culture with minimal differentiation, and can give rise to mature blood cell~ of diverse lineages when subsequently treated 0 with the appxopriate hematopoietic growth and differentiation factors. Further, the long-term cul~ured cells also can mature into functionally : ~.competent blood cells ln vivo, capable of mediating antigen-specific immune responses, repopulating lympho-hematopoietic organs, and prolonging survival o~ animal~ with a destroyed hematopoietic system. The ~: yolk sar cells of the invention can be successfully transplanted into allogeneic ~etuses in utero and into : non-immunosuppressed xenogeneic hosts ~nd since these 20~ cells do not induce graft-versus-host an~
host-versus graft reactions, tran~plantation will : ~ ~resuIt:in tissue chimerism.
;~ ~ The i~vention is described by way of examples in which murine yolk sac cells are isolated, 2~5~ and their cell surface phenotype is characterized.
The h~mogeneous population:of yolk sac cells is : expanded in long-term culture, and shown to retain plurip~tency in vitro and in vivo. A wide variety of ùs~s for the yolk sac cells are encompassed by the invention described herein.

4. BRTEF DESCRIPTION OF T~E DRAWINGS
FIG. 1. A schematic drawing Qf ~he appearanc~ of mouse embryos around day 7 and day 8.5 of gestation.

SU5~TITUTE S~I~ET

W~93J021~2 _ 7 2113 ~ ;~ i PCT/US92/05918 FIG. 2. ~urine yo!k sac cells from a day 7 embryo are more homogeneous in appearancP than cells from a day 8.5 embryo by flow cytometry analysis.

FIG. 3. Murine yolk ~ac cells from a day 7 embryo express CD34 but not Thy-l, M~C class I and class II antigens.

FIG, 4. Cultured yolk sac cells can differentiate into mature blood cells in vitro, including ~;
(4A) monocytes, (4B~ megakaryocytes, (4C) erythrocytes, and (4D) lymphocytes.

15 FIG, 5. Yolk sac cells recovereà from recipient mou~e spleens following in YiVC) transfer demonstrate the expression of mature laukocyte antigens by donor ce~ls.

FI~. 6. : Hemagglu~ination of red blood cells coated with antigens (FIG. 6A, lipopolysaccharide, and FIG . 613, ~ human serum albumin~ by sera of inununodef icient mice trea l:ed with yolk sac cells, demonstrating restoration of immune ~: ~ 2s: ~ ~ function by yolk sac ~ells in vivo.
: : :
: FIG.::7. Cultured~yolk sac cells repopulate the : spleens of chemi~ally-ablated mice and give ~: rise to~colony-forming units in vivo; (7A) A
comparison betw~en a ch~mo-ablated mouse : ~pleen and a: fully repopulated spleen; (7B~
A repopulated spleen at day 7 post-yolk ac treatment; (7C) A populat2d ~pleen at day 14 post y~lk-sac treatment.

3~

T~TUTE ~HeET

WO93/02182 ~ 8 PCT/US92/0S91~
~1 13 ' ~ 5 FIG. .8. In utero injection of yolk sac cells into allogenic mice leads to tissue chimerism in new born mice.

FIG. 9. Survival and differentiation of long-term cultured murine yolk sac cells in a sheep and a goat which had received multiple high dos~s of yolk ~ac cells.

105. PETAILED DESCRIPTION OF THE_INVENTION
The present invention relates to yolk sac stem cells, to methods of isolating and culturing the .yolk sac stem cells, and to methods of using the yolk ~ac stem cells~
15~lthough the specific procedures and methods described herein are exemplified using murine yolk sac cells~ they are merely illustrative for the practice of the inVention. Analogous procedures and techniques are equally ap~licable to all mammalian species, 20 iIlcluding human sub~ects. Therefore, human yolk sac stem cells may be isolated from the embryonic yolk sac prior to blood island formation. The cell~ having the phenc)~:ype of CD34~, Thy-l , and MHC class I and II
may be cultured under the same conditions described 25 herein, infra.
Mammalian development may be divided into three distinct stages: the ~t from fertilization t~ cleavage; the embryo, from cleava~e to the formation of all somites; and the fetus, frc3m the 30 formation of the last somite urltil birth~ This inventiorl takes advantage of the unique properties of embryonic yolk ~sac cells after their course of devalop~,lent is dst~rmined, but beiEore l:h~y have los~
either immuno- incomp~tency or the ability to 35 proliferate rapidly.

uT~T~T~ S~ET ~`

21~3 ~
W~93~0218~ _ 9 _ PCT/VS92/05918 . It is known that when completely undifferentiated cells of the blastula or morula are transplanted into a developed animal, they produce tumors. These totipotent, tumorigenic cells are of no S value for in vivo recon5titution ~herapy. However, in accordance wit~ the invention, it is advantageous to transplant ~ells which have reached a stage of specialization at which they have become committed to a particular sequence of development, or lineage.
Such cells may be used alone or to deliver genetic material, or it5 expression products, into a particular tissue of the body, including blood cells.
The cells can be transplanted into a host before or after transformation with an exogenous gene of interest, and allowed to de~elop into the tarset tissue.
While it is necess~ry to use cells which have matured to the point of losing totipotency, fully mature cells will be rejected by a histoincompatible ~O host, Consequentlyl it is desirable to use cells which have just lost totipotency, but still xetain : pluripotency for a particular tissue typ~. Such cells : also may retain the ability to colonize, thus facilitating their delivery to the target tissue.
: 25 Stem cells of the embryonic yolk sac offer ~ particular advantages for hematopoietic .
re~onstitution. Unlike,the cells of the embryo, the cells of the~yolk sac develop into only a small number of dif~erent ~i~sues. Among those ~i sues is the hematopoietic system, which includes the red and white blood cells, and the tissue of the vein~, arteries and capil~aries. Thu~, by day 8 in the development of the ~ use embryo, mesodermal cells in the yolk sac begin - : to ~orm blood islands. The cells of the blood i~lands diff~rentiat~, the peripheral cells becoming the SÇ~ TIT~T~ E~

WO93~02182 l0 PCT/~'S92/05918 ~ r ~

endothelium of the future blood vessels, and the central cells becoming first mesenchymal cells and then the red and white blood cells. The blood islands establish communications to form a circulatory network, which is extended into the embryo proper.
The yolk sac cells of the subject invention do not express MHC antigens, and can mature in allogeneic and xenogeneic hosts, demonstrating their ability to escape immune rejection. By contrast, research with bone marrow cells has depended on the use of immunocompromised hosts. The culture methods described herein maintain the yolk sac in their .undifferentiated state, and are applicable to mass culture of yolk sac cells, providing donor cells for larqe numbers of recipients.

5.1. ISOLATION OF YOLK SAC CELLS
The embryonic yolk sac is ~he first dentifiable site of blood cell formation in ontogeny.
The~yolk sac cells travel to the fetal liver during embryogenesis~and eventually migrate to the bone marrow~where they~reside~and differentiate into mature blood cells throughout the entire adult life.
The~embryonic development of the mammalian 2s yolk sac is rapid and occurs within a narrow time frame. Thé murine yolk sac is fully formed by day 7 of gestation, and~;the formation of blood is detectable in the mesenchyme~;of the body stalk and in neighboring ;
areas `of the yolk sac. Shortly therea~ter, masses of mesenchymal cells round up and become aggregated to form blood islands. By day 8.5, extensive blood island formation in the murine yolk sac is readily ~isible microscop~ically. At this stage, embryonic development has~reached a level where fetal liver is formed and yolk sac cells begin to migrate to the "

~51~T''~T~ S~ET

W~93/0218~ 3 ~ ~ S PCT/US92/05918 fetal liver. Upon the departure of the yolk sac stem cells, the yolk sac begins to atrophy. Similar events also occur in embryonic development of other species, but the timing of developmental events varies between S different species. In humans, the yolk sac is formed by day lQ of gestation, and bl~od island formation occurs shortly thereafter. Thus, human yolk sac cells isolated at day lO may be comparable to the murine cells at day 7.
Since the yolk sac is where blood cell for~ation is first established in development and the yolk sac cells eventually reach the bone marrow to become the bone marrow hematopoietic cells, it is reasoned that the~yolk s~c represents the earliest ;15 site for the generation of primordial hematopoietic ~- cell precursors~ The cells have committed to the hematopoietic differentiative pathway so that they are no longer totipotent.~ However, the yolk sac cells are still pluripotent,~since~they have not yet committed 20~ to~a~particul~ar blood cel1 lineage as seen by their ability to make cel1s~of lymphoid, myeloid, and erythroid~lineages~.~ Hence, yolk sac cells may be the ideal cell population~for use in reconstitution therapy including,~but not limited to, bone marrow 25~transplantation~. ~In;;addition, the primitive nature of these cells, as~evidenced~by the absence of cell surface expression~of various mature markers and MHC
transplantation~rejection~antigens, may render these --cells uniquely~capable o~f being used as a universal ~30; donor cell population in allogeneic and even xenogeneic hosts.~
The isolation of the embryonic yolk sac may be achieved using~a variety of surgical methods.
Traditionally, the;~yolk sac of a ~ouse embryo is .-.
~ 35- disaggregated by the use of enzymatic digestion and :::
T~T~T~ S~r~T

WO93/021g2 - 12 - PCTtUS92/059lX
2 ~
mechanical separation upon surgical removal. A
gentler method of detaching the cells from the yolk sac membrane and separating them from each other is described in Section 6.~.1. in which a yolk sac is immersed in an EDTA solution which causes the cells to segregate and form a single cell suspension. This method minimizes cell lysis due to physical force and cell surface protein alteration due to enzymatic treatment.
ln Since the establishment of blood islands in the yolk sac marks the beginning of csllular differentiation and blood cell ~ormation, it is preferable that y~lk sac cells be isolated prior to extensive blood island formation. Large numbers of highly ho~ogeneous yolk sac cells of day 7 murine embryos (or similar stage human yolk sac cells), can be isolated using the method described herein, and cells obtained at this stage should in principle : contain the least committed and least differentiated p~uripotent stem cells suitable for long~term in vitro culture, or:use in immediate in vivo therapy or as ca~rriers of specific exogenous genes for use in gene therapy. : ~
For:long-term maintenance of the yolk sac Z:S cells, the cells are grown in medium containing a : relatively high concentration of serum supplement, between 15-20%. Various cytokines may be added to : suppress differentiation of the stem cells, including but not limited to, leukemia inhibitory factor (LIF) or stem cellifactor/the c-kit ligand (SCF) or SCF in combination with other cytokines such as IL-3. Such factors accelerate the multiplication of cultured ~ cells, while inhibiting cellular differentiation in : : vitro. The examples presented in Section 6, infra, were ~11 performed using yolk sac cells grown in the .

~U~T~TLiTL S~ET ~

W~ 93/û2182 -- 13 ~

presence of lO-lOO U/ml of LIF. However, higher LIF
concentrations may be used to achieve stronger suppression of differentiation. The growth of cells using SCF could produce similar results.
Alternatively, a number of other known hematopoietic factors such as I~-3, CSF~s and EPO also may be used in combination depending on the need to select for a particular cell type. For example, the combined use of IL-3 and EPO may assist in driving cultured yolk sac cells towards the erythroid pathway. The maintenance of cells at the appropriate temperature, CO2 concentration, humidity level and the frequency of changing the culture media are within the ordinary skill of the art~
5.2. CHARACTE IZATION OF YOLK SAC CELLS
As sh~wn by the examples de~cribed herein, : yolk sac cells obtained from mous~ embryos prior to blood island formation are more homogeneous in ~O app~arance ~an ce}ls obtained at a later stage.
Freshly isQlated yolk sac c~lls from day 7 and day 8.5 :.
murine embryos were compared by light scattering using : : flow cytometry analysis, see Section 6.2.1., infra.
: It is apparent that yolk sac cells of day 7 mouse embryos are extremely uniform with respect to both cell size and cell shape. By day 8.5, distinct populations of cells are clearly visible, suggesting that the earlier~stage yolk sac cells may be clonally ~:
derived and the difference of 1 day in development may be cri~ical to the nature of th~ yolk ~ac cells.
Another indication of the primitive nature : of the early yolk sac cells is their cell surface.
phenotype in regard to the e~pression of various lineaqe-~ecific blood cell markers. Thi~ form of analysis may be most conveniently carried out by the TlJT~ S~EET

WO93/02182 - 14 - PCT/US9~/0~91~

2 1 1 ~3 ~'J; ~ J
use of a panel of marker-specific monoclonal antibodies. When the day 7 yolk sac cells were reacted with antibodies, the results show~d that they lacked expression of all mature blood cell markers.
S In addition, such cells did not express MHC-encoded products which are the major targets of transplantation rejection responsesO Thus, yolk sac stem cells can be characterized as CD34~, Thy-l , MHC
class I and MHC class I$ . Similarly, human yolk sac stem cells obtained from day 10 embryos should display an îdentical cell suxface phenotype.
The CD34 and Thy-1 markers previously have been demonstrated to be associated with bone marrow hematopoietic stem cells (Spangrude et al., 1988, t5 Science 2~1:58). While CD34 expression declines as stem cells differentiate and mature, the presence of Thy~ retained and its density increa~ed in certain .
mature blood cells, particularly T lympho~yt~s. The finding that yolk sac stem cells are positive for CD34 Z0 expression is consistent with the~e cells being stem cells. However, the absence of Thy-~ expression suggests that yolk sac cells may represent an ~arlier cell popula icn than the bone marrow stem cells which e~press low levels of Thy-1 in the bone marrow micro~nviro~ment, In fact, when yolk sac cells ~re cultured in vitro, a:small percentage of the cells :~ esrape the effect of LIF and begin to express Thy~
further suggesting that Thy 1 expression is a later event of stem cell development.
MH~-encoded class I and class II molecules are involved in immune regulation between T, B, and antigen presenting cells. ~hese highly polymorphic molecules also serve as targets in major transplantation rejection responses between 3S geneti~ally mismatched individuals. Therefore, HL~

.

f~TUTE S~IEET

Wn93/0~182 2 ~ 5 ~

tissue typing is currently a routine clinical procedure in ensuring graft acceptance in human transplant patients by matching the donors and recipients ~t the major MHC genetic loci. The absence of MHC antigens on the yolk sac cell surface strongly suggests the possibility of using such cells as universal donors in hematopoietic reconstitution therapy, alleviating the need of tissue typing and the restrictiv~ use of only MHC-matched tissues as donor cells. ~he development of adoptively transferred yolk sac cells in the environment of the host may lead to specific tolerance between the host and donor cells ~or each other, causing a diminution of the potential for inducing graft-versus-host and host-versus-graft lS reactions.
The ~bove-described yolk sac phenotype i5 seen with the va t majority of cells isolated fr~m day 7 murine embryos. Therefore, early isolation of yolk sac cells pr~vi~es for a highly homogeneous and : 20 enriched population of stem cells. This is in ; ~ contradistinction to the purification procedure needed for murine bone marrow hematopoietic stem cells which are of CD34~ and Thy-1+ phenotype. Such cells must be isolated and enriched by a sPries of ~election steps, as they constitute only less than 0.1% of the total cells in the bone marrow (Spangrude et al., 1991, Blood 78:1395). On the other hand, yolk sac stem c~lls can be obtained in an essentially homogeneous s~ate without requiring additional puri~ica~ion, and such cells retain their phenotype and functional act}vity during long-term in itro growth.
..
5.3. FUNCTIONAL ACTIVIT~ES OF Y~OLK SAC CELLS
The pluripotency of yolk sac stem cells to -~5 differentiate and mature into functionally competent SUB~TiTUT S'~EET

WO93/02182 - 16 - PCT/US92/0~9l~
2 ~
blood cells of various hematopoietic lineages was tested by a number of in vitro and in vivo me~hods described herein. The presence of a pluripotent population in long-term cultured yolk sac cells was 5 first demonstrated as follows. After 10 passages of in itro growth, yolk sac cells were washed from LIF
and exposed to a combination of cytokines including IL-3, CSF's, and EPO at previQusly determined optimal concentrations for an additional three weeks in culture. ~t the end of the period, the stimulated yolk sac cells were prepared as blood smears and stained with hematoxylin. The result of this analysis .reveals the appearance of blood cells that can be identified as erythrocytes, ~ranulocytes, megakaryocytes, and l~mphocytes.
~ similar study also was carried out in VlVO
by rec~v~ring donor cells four weeks af~er ia v vo injection into allogeneic SCID mice. The yolk sac :;
cells used in this study had been expanded in culture : 20 for o~er 40 passages. Double-staining of the spleen, bone marrow, and thymus cells of the SCID mice was performed using antibodi~s ~pecific for the donor cell haplotype of H-2d and antibodies against mature blood :
: cell markers such as B220 for B cells, CD3 and Thy-1 for T cells, and Mac-l for macrophages~ The results of this in vivo study confirm the in vitro study that long-term cultured yolk ~ac cells are capable of giving rise to mature T cells, B cells and macrophages/monocytes.
In addition to morphologic evidence of blood cell maturation f rom yolk sac cells, the adoptively transferred yolk sac ceIls were tested for functional activities in the form of specific antibody production. One month after re~eiYing an infusion of yolk sac cells, the mice were immunized with either S~t~ i ~TUT~ ~EET

W093/02182 17 - 2 1 1 3 ~; ~ P~T/US92/05918 lipopolysaccharide (LPS) or human serum albumin (HSA).
Sera of mice were diluted serially, reacted with the two antigens, and compared with normal mouse sera as controls. LPS is a T cell-independent antigen which activates polyclonal ~ cells directly. The high titer of LPS specific antibodies in the sera of yolk sac cell-bearing beige nude xid mice after LPS
immunization indicates the presen~e of functiona~ly competent antibody producing cells, i.e., B
lympho~ytes and plasma cells. Additionally, HSA, which is a T cell dependent antigen, elicited a weaker yet detectable specific antibody production in mice.
Since the a~ti-HSA antibody response requires T cell help which, in turn, is first activated by antigen-presenting cells such as macrophages, this resultprovides evidence for the presence of mature and functional T cells, B c~lls, and macrophag~s which co operate and interact in the generation of antibodies.
As a corollary, this also suggests that other T cell 20 :and macrophage-mediated functions such as : cytotoxicity, lymphokine and ~ytokine secretion, phagocytosis, antigen processing and pres~ntation may all develop from the transferred yolk ~ac stem cells.
: ~ The ln Yiy~ transfer of yolk sac cells also ~S repopulated ~he spleens of mice whose hematopoietic system had been previously destroyed by ch~mical abla-~ion or lethal doses of irradiation. This resembles situations in which a patient's lymphohematopoietic system is deficient du~ to a genetic disorder or an acquired viral in~ection, or a patient's system is intentionally destroyed by ch~motherapy or radiotherapy in order to eradicate tumor cells in the bone marrow. The admini~tration of - yolk sac cells induc~d colony forming units-sple~n ~U-S) in l~thally irradiated or chemo-ablated mic~

SU~'TITUTE S~E~T
, . . . . . . , . . . .... ...... ..... ....... , ... , .. , .. ..... .. . ~ . .. ... , . ~ .. .. . . .

WO93/02182 - 18 - PCT/~'S92/0591X
2 ~ 1 3 i ri S
whose. spleens, othPrwise, frequently exhibited a necrotic appearance. On the other hand, expansion of the yolk sac cells over a period of time in vi~o supported repopulation and restoration of spleens S completely normal in appearance. Further, the yolk :
sac cell-treated mice experienced a prolongation of survival time when compared with the untreated control group~ Therefor~, long-term cultured yolk sac cells may be useful in a variety of settings in which bone marrow reconstitution can be applie~ as an effective means of therap~
Transplantation of murine yolk sac cells .into allogeneic fetuses in utero and xenogeneic new born animal~ did not induce graft rejection reactions.
lS The yolk ~ac cel-s persisted ln vi~o and established hema~opoietic chimerism in the spleen, li~er, and pexipheral blood of the host. Thus, yolk sac cell5 may be useful as universal donor cells in various mammalian species, including humans.

: 5.4. USES OF YOLK SA STEM CELLS
The absence of MHC antigen expreqsion by ~: yolk sac stem cells provides for a source of donor : cells for in v transplantation and reconstitution ; 25 therapy. The cellB ~ay be used immediately after isolation from the yolk sac or after long-term expansion in Yi~ in order to procure larger numbers for more effective doses. Introduction of ~xogenous genes into the yolk sac cells may be achi~ved by conventional method~ during in yitr~ culture and/or in ivo gene therapy. Long-term cultured c~lls may be used as a mixed population or progenitors can be : pre-selected based on the primitive phenotype of CD34~, Thy-l9 MHC cla~s I and class II , or ~y limitin~ dilution cloning, prior to in Ya~ use.

TI ~ ~TE S~ET

WO93/021~ - l9 2 ~ ~ 3 J~ 5 PCT~US92/~5918 5.4.l. HUMAN YOLK SAC CELLS IN MICE
Human yolk sac cells may be obtained, grown in vitro and transferred into immunodeficient or immunocompromised mice. Such mice contain a human hematopoietic system and may be used for the study of human blood cell development in vivo, the identific~tion of novel hematopoietic growth and differentiation factors, and testing for cytotoxic and/or inhibitory compvunds that affect various stages of blood cell formation as well as anti-cancer drugs.
Such a chimeric mouse referred to as HumatoMo~se~
herein would be superior to the conventional SCID/Hu mouse model in which mice are reconstituted with human bone marrow stem cells because HumatoMouse~ wou~d permit studies in the delineation of the e~rliest events in hematopoiesis. Furthermore, yolk sac cells may be implanted in utero in~o nor~al mouse fetuses ~or engraftment of human blood cells in a normal mouse environment. Such yolk sac cells may be trans~ected 20 with a drug-resistance ~ene so as to allow subsequent selestive ablation o~ only the host cells using th~
corresponding dru~.
It has been observed that SCID mice are not totally i~munodeficient and that a small amount of restoration of immune function is correlated wit~ the age of he mi~e. SCID mice possess detectable natural killer cell and macrophage activities. A small percentage of mice even re-acquire T and ~ cell function as they mature. Thus, conventiQnal SCID mice may not be the most appropriate host~ ~or the construction o~ the HumatoMouse~ as their immune function may interfere with the analysis of the d~nor yolk sac cells. The steel mice possess a ~utation at the steel locus which encodes SCF, a ligand for the proto-oncogene c-kit cell surf~ce receptor. ~ouse ~IJ~T~F~T~ ~It~

~ ~ 1 3~1~J~` - 20 -fetuses that are homozygous for this mutation liveonly to about day 15 of gestation before they are aborted due to the absence of a hematopoietic system and blood cell formation. Hence, human yolk sac cells may be injected into the developing homozygous fetuses in utero prior to abortion, e.g., at day 8, to reconstitute their hematopoietic function. The resulting neonates should have a fully humanized system with no contribution by the host as they would 10~ not normally have lived to birth.
Studies described herein demonstrate that cultured yolk sac cells can develop into mature blood .cells in vivo, suggesting that the cells secrete the necessary growth and differentiation factors for supporting their own development. A further i~provement of the Humatomouse~ modPl includes the introduction o~ human growth and di~ferentiation factor genes in the mice. In the event that certain of the~critical cytokines for human blood cell 20 ~formation are species-specific, such as SCF, and mouse molecules do not~act effectively to promote growth and differentiation of~human cells, transgenic SCID or steel~mice may be~constructed to result in endogenous production of human~cytokines of interest such as 25~ IL-3, CSF's,~and~SCF.~ ~Alternatively, human yolk sac cells may be transfected with murine receptor genes.
he subsequent transfer of human y~lk sac cells to 5 ~ these~mice should~give rise to a more complete and efficient human hematopoietic system in mire.
5.4.2. TRANSPLANTATION USING YOLK SAC CELLS
The repeated transfer of high doses of long-term cultured mouse~yolk sac cells into a foreign species, i.e. sheep, has shown that the cells persist ~ vivo, differentiate into mature lymphocytes, and do SU~ i 5TaTL S!~EET

2~ 1~5~
W~93/02182 - 21 - PCT/US92/05918 not mediate graft versus host disease. Although the mature donor mouse cells eventually express MHC
antigens in vivo, the donor ce~ls are present in high quantities in the peripheral blood of the xenogeneic host. The absence of graft rejection (ho~t versus gra~t) and graft versus host reactions may be attributed to the primitive nature of the yolk sac cells, particularly the lack of MHC antigen expression, allowing the cells and the ho~t immune ~0 system to 'llearn" each other as self prior to MHC
expression and thusl induce a state of specific tolerance.
Xenogeneic transplants of solid organs have been carried out in humans in situations where there :~
is a shortage of ~LA-matched organs. With respect to xenogeneic transplant o~ primitive hematopoietic stem cells r yolk sac cells may be used to reconstitute the hematopoietic system of any mammalian ~pecies, for exa~ple, in a human patient with HIV infection. Since non-human T cells cannot be infected by human HIV, : this approach may serve as a means of limiting HIV
- infection in humans. Yolk sac cells may also be trans~ected with genes which are designed to disrupt HIV gene sequences involved in HIV replication prior to in ivo~administration. Such exogenously : introduced genes may encode anti-sense RNA or ribozyme : : :molecules that specifically interfere with HIV
replication. Further, the induction of tolerance by the transfer of xenogsneic yolk sac cells may allow subsequent transplantation of sol id organs, including but not limited to heart, liver and kidn~y from donor : animals sharing the same genetic makeup of the y~lk sac donors. This raises the po~sibility of using ~HC-~ismatched yolk sac cells not only for reconstitution purpos~s, but also as first step T5TUTE ~ ET

WO93/02182 - 22 - PCT/US92/0~918 2 1. 3 .` ., .~

tolerogens for inducing specific tolerance in a recipient for subsequent organ transplants.
In additi~n, this form of yolk sac cell transplantation may be applied in situations where a genetic defect has been det2cted in a fe~us. Human or sther mammalian yolk sac cells carrying a normal wild type gene or an exogenously introduced gene may be injected into the developing f etus in a routine pro~edure similar to that of ~mniocentesis ln utero.
lQ The gen~tic ~isorders for which this approach may be : applicablP include, but are not limited ~o, sickle cell anemia, thalassemia, and adenosine deaminase ~eficiency. Alternatively, yolk sac cells may be used in ~ettings where a pregnant mother is dia~nosed to carry ~IV, and reconstitution of ths fetus with yolk a~ cells may prevent:viral infection of the fetus.
The ability of yolk sac cells to grow in ;~ xsnogeneic animals with no irradiation or chemical treatment allows for large scale production o~ human 20 ~h~matopoietic ~ells and their secreted factors in YL~- Human yolk sac cells may be injec~ed in a large fa~rm animal, the blo~d collected, and large quantities of~human proteins~or cells such as red blood cells, : lymphocytes, granulocytes, platelets, monoclonal antibodies and cy~okines purified for clinical use.

5.~5. BONE M~RROW REPLACEMENT THERAPY_IN HUMANS
: A protocol for the replacement of bone marrow cell in human patients requiring bone marrow transplantatlon may be devised using cultured human or : ~ xenogeneic yolk sac cells. Yolk sac cells obtained fr~m human yolk sac at day 10 of gestation may be ola~ed using the procedures described herein, expanded in culture, and cryogenically preserved as donor cells for t~e transplant.
.

SII~T~TUTr SH~T

W~93/02182 2~ ~3~ J ~ pCT/US92/0591~

. Ablation of recipient patient bone marrow cells may not be required, but if it is used, it can be accomplished by standard total body irradiation (Kim, et al., Radiology, 122:S23, 1977) or by 5 chemotherapy with a variety of commonly used compounds including, but not limi~ed to Busulfan (Tutschka, et al., 31Ood, 70:1382-1388, 1987), following the conventional methods. Yolk sac cells can be introduced into the recipient, using similar methods 0 for bone marrow cells. Prior to ln vivo transfer, yolk sac cells may be transformed with a drug-resistan~e gene, such as the methotrexate resistance gene. This allows the subsequent use of high doses of the corresponding chemotherapeutic drug to eradicate ~5 the less resistant host cells in a pa~ient, without damage to the transferred yolk s~c cells. Post-operative care would be the same as with transplantation using bone marrow cells from a donor.
High doses of yolk sac cells obtained from allogeneic or xenogeneic sources may be conti~uously infused into a bone marrow transplant recipient in the absence of prior chemotherapy or radiotherapy. This presents a novel approach to bone marrow transplantation without immunosuppressing the recipient~

~ 5.6. IDENTIFICATION O~ NEW
: MARKERS ON YO~K SAC CELL5 Murine yolk sac cells express CD34 but none of the other known leukocyte markers. It is possible that yolk sac cells express other early markers which have not yet been identified. If so, previous ~ailure in identifying these unique molecules might be due to their d~cr~ased expre~sion in more mature cells or ~ven stem cells after migration to other sit~ out of the yolk sac. There*ore, yvlk sac cells may be used ~U~Ti,I~TE S~ET

WO93/0218. 2 ~ ~ 3; ~j 5 - 24 - PCT/US92/05918 to generate antibodies against their cell surface antigens in order to identify and characterizs such unknown markers.
Also within the scope of the invention is the production of polyclonal and monoclonal antibodies which recognize novel antigenic markers expressed by yolk sac cells. Various procedures known in the art may be u-~ed for the production of antibodies to yolk sac cells;. For the production of antibodies, various host animals can be i~munized by injection with viable yolk sac cells, fixed cells or membrane preparations, including1 but not limited to, tho~e of rabbits, hamsters, mice, rats, etc. Various adjuvants may be used to increase the immunological response, depending on the host species, includin~ but not limited to Freund's (complete and incomplete), mineral gels such : as aluminum hydroxid&, surface aotive substances such as lysolecithin, pluronic polyols, polyanions, peptides, oil mulsions, keyhole limp.et hemocyanin, dinitrophenol, and potentially useful human adjuvants such a~ BCG (bacille Calmette Guerin) and ~ Corynebacterium ~arvum.
: Monoclonal antibodies to novel antigens on yolk sac cells may be prepared by using any technique which provid s for the production of antibody molecules by continuous cell lines in cultur ~ These include, but are not limited to, the hybridoma techni~ue originally described by Kohler and Milstein (1975, Nature 256, 495-497), the more recent human B-cell hybr~doma technique (Kosbor et al., 1983,Immunology Today 4:72; Cote et al., l983; ProcO Natl.
Acad. Sci. ~0:2026-2030) and the EBV-hybridoma technique ~Cole et al., 1985, Monoclonal Antibodies and Canc~r Therapy, A~an R. Liss, Inc., pp. 77-96).
Techni~ues developed for the producti6n o~ ~'chimeric TE S~E~

21 ' 3 ~
~93tO2182 - 25 - PCT/~'S92/05918 antibodies" by splicing the genes from a mouse antibody molecule of appropriate antigen specificity together with genes from a human antibody molecule can be used (e.~., Morrison et al., 1984, Proc. NatlO
Acad. Sci., 81:6851-6~55; Neuberger et al~, 1984, Nature, 312:604-608; Takeda et al., 1985, Nature 314 452-454)o In addition, technique~ described for the production of single chain antibodies (U.S. Patent 4,946,778) can be adapted to produce single chain 10 antibodies.
Syngeneic, allogeneic, and xenogeneic hosts may be immunized with yolk sac cells which can be prepared in viable form, or in fixed form, or as extracted membrane fragments. Monoclonal antibodies can be screened differentially by selective binding to yolk sac cells, but not to mature macrophages, granulvcytes~ T, and B ells.
Antibody fragments which contain the binding sit~ o~ the molecule may be generated by known technigues~ For example, such fragments include but are n~t limited to: ~he F(ab' )2 fragments which can be produced by pepsin digestion of the a~tibody molecule and the Fab ~ragments which can be generated by : ~ reducing the disulfide bridge~ of the ~(ab') 2 : ~ 25 ~ragments.

6. EXAMPLE: GENERATION OF YOLK SAC
STEM CELLS FOR IN VIVO
HEMATOPOIETIC RECONSTITUTION

6.1. MATERIALS AND METHODS

6.1.1. ANIMA~S
BALB/c, C57BL/6 t beige nude X-linked im~unodeficient (BNX), and C3H/SCID mice were purchased from Jackson Laboratories (Bar Harbor, ME) T~T~TE SaE~T

WO93/~2182 - 26 - PCT/US9~/05918 2 1 ~ 3 ~ ;J

and kept in the animal facility of Edison Animal Biotechnology Center.

6.l.2. ISOLATION OF THE YOLK SAC
On d2y 7 of gestation (day of plug was counted as day 0, female mice were sacrificed by cervical dislocation, and uteri containing embryos were placed in petri dishes with Dulbecco's Phosphate Buffered ~aline (PBS) plus penicillin and streptomycin antibiotics (~inal concentration :l000 units potassium penicillin G and l000 ~g streptomycin sulfate/ml).
Under a laminar air-flow bench ,each uterine segment containing an embryo was aseptically removed by dissection with the aid of a dissenting microscope.
Each embryo surrounded by decidua capsularis was transferred to another petri dish containing PBS plus penicillin-streptomycin. The decidua capsularis was opened with watchmakeris forceps and each embryo transferred into an individual petri dish where yolk sac tissue was excised from the amnion, placenta, embryo, and Reichert's membrane in 0.02% EDT~ in PBS
at ~C for 15-30 minutes. The yolk sac cells in single cell ~uspension were then washed in PBS before culturing.
~5 6.l.3. CULTURE CONDITIONS
Disaggregated yolk sac cells were grswn in alpha medium (Sigma) supplemented with l8~ :
heat-inactivated fetal calf serum, 0.2 ~m ~-mercaptoe~hanol, 50.~g/ml of gentamicin and 10% LIF
conditioned medium (medium of a LIF-producing cell line, Cho LIFD at l00-lO00 u/ml). Cells were grown without feeder layQrs on collagen or gelatin coaked ~ishe-~ and incubated at 37C in 5~ C2 in air. ~edia were changed every other day.

SU~STITUT~ SHET

W~9~/02182 - 27 ~ 3 ~ PCT/US92/0591 6.l.4. FLOW CYTOMETRY ANALYSIS
l06 yolk sac cells were washed twice in cold PBS containing O.l BSA and sodium azide. The cell pellets were suspended in the same buffer containing the test antibodies at 4C for 30 minutes. Cells were then wash~d in cold PBS twice and analyzed by flow cytometry. Antibodies specific for Thy-l, Ly-l, Ly 2, Mac~ C class I and class II were purchased from Boehringer Mannheim. Anti-Ml/70, anti-H2d and anti-H2b antibodies w~re purchased from Pharmingen (San Diego, CA~. Anti-CD34 was used as hybridoma supernatant.

Ç.l.5. NDUCTIO~ OF YOLK SAC DIFFERENTIATION
BALB/c yolk sac cells were grown to l approximately 50% confluency in medium containing LIF.
The cell~ were harvested, washed and medium containing growth fact~rs was a~ded. Growth factors used were:
LIF (l00-l000 U/ml), SCF (50 U/~13, EPO (1-25 U/ml);
IL-2 (l0-200 U/ml), an~ IL 3 (10-200 U/ml) in various combinations. The medium wa~ changed every 2 days until confluency was reached, at which time the yolk : ~ sa~ cells were passed 1:4 into new gelatinized 35 mm : culture dishes. At day 5, and 21, cells were prepared for blood staining. Day 0, 5, and 2l cells were ~: 25 analyzed by flow cyto~etry for the appearance of differentiated blood cells.
:: :
: 6.l.6. HEMAGGLUTINATION ASSAY
: Lipopoly~accharide (LPS) conjugated to trinitrophenol (TNP) and human serum albumin (HSA) c~njugated to TNP ~ere injected at 20 ~g/mouse int~aperitoneally into BNX mice and SCID mice J
resp~ctively, both of which had previously received l06 ~urine yolk sac cells intraperitoneally a month earlier. A ~econd injection of the antigens waæ

T!T~T~ SH~ET

WO93/02182 - 28 - PCT/US92/0~g18 ~113 i~, performed one week later, animals were bled after seven days and sera assayed for the presence of specific antibodies.
A two~fold serial dilution of the mouse sera was made in microtiter plates. Sheep red blood cells (SRBC) coated with dinitrophenol (DNP) were added to ~ach well. The plates were incubated at room temperat~re for one hour. The results of the assay were a~sessed visually. A dif~used pattern of S~BC
indicated a positive TNP-specific antibody response.
Negative wells had a small, tight pellet o~ SRBC.

. 6.2. RESULTS

6O2.1. ISOLATION OF MURINE_YOLK SAC CELLS
In the:mouse, the yolk sac is fully f~rmed by day 7 and blood island formation usually appears by day 8.5 of gestation. Therefore, in order to isolate ~: ::; homogeneous and undifferentiated yolk sac cells, mouse embryos were surgically removed prior to visible blood island formation, preferably at day 7 of gestation.
: The yolk sac region of the embryos was separated by excision, and the external surface of the yolk sac was , .
immersed in cold EDTA~which caused the detachment of 25 ~:the yolk sac cells from the membrane into a single cel~l suspension ~FIG.~ 1). When the physical :; ~ appearance of yolk sac cells obtained from day 7 and : day 8.5 embryos was compared by flow cytometry analysis, freshly isolated day 7 cells clearly displayed a much more uniform cell shape and cell size than ~he day 8.5 cell:s, suggesting that yolk sac cells ~ were a homogeneou population at day 7 but by day 8.5, : ~ifferentiative activities had already occurred to generate a mixed population of cells in the yolk sac (FIG. 2). There~ors, day 7 yolk sac cells were used ~ r ~EEl W~93/02182 ~1~3 ~ Pcr/~ls9~/o59l~

for all in vitro and in vivo studies described herein, infra~

6.2.2. CELL SURFACE PHENOTYPE
OF YOLK SAC CELLS
The freshly isolated yolk sac cells from day 7 mouse embryo-~ were immediately examined for their cell surface expression of a number of known leukocyte markers by reactivity with monoclonal antibodies.
~0 Such uncultured yolk sac cells express CD34 but not Thy-l, MHC class I and class II antigens (FIG. 3).
Th~ expr~ssion of C~34 by yolk sac cell~ is consistent with t~em being primitive stem cells as CD34 is currently the earliest detectable marker on bone marrow hematopoietic stem cells. The absence of MHC
antigen expression at this stage is significant in that the likelih~od of rejection of these cells by a : genetically di~parate host upon ln vivo transfer is greatly reduced. ~urther, the lack o~ Thy-l expression indicates that the yolk ~ac cells of the invention represent an earlier cell population in ~ ontogeny than the Thy-l+ hematopoie~ic stem cells : found~in bone marrow, thus should contain a pluripotent population that is less committed to any specific ~ell lineages.

6.2.3. LONG-TERM MAINTENANCE OF YOLK SAC CELLS
The yolk sac cells isolated from day 7 embryos were established in culture in the presence o~ -3~ leukemia inhibitory factor (LIF) at l0-l00 U~ml without a feeder layer. rhe cells expanded in number, .-having a doubling time of about 18 hours. Such cultured cells have been grown in vitro for over 4l passages covexing a period of time over nine months in 35 c:on~inuous clllture. Alternatively"~olk sac cells S~T~TUT~ SH~ET

WO93/021~2 30 _ PCT/US92/05918 2113'jC` i could also be grown in stem cell factor with similar results.
Although LIF is capable of suppressing di~ferentiation of the yolk sac cells over an extended period of in vitrQ yrowth, the effect of LIF is incomplete because a small fraction of the cultured cells beyan to express certain differentiation markers including Thy-l and MHC-encoded molecules. However, the majority o~ the long-term cultured cells retained their original cell surfac~ phenotype. Further, such cells con~inued to be pluripotent as evidenced by their ability to give rise to mature blood cells in vitro and in y~y~/ infra. The cells with the original phenotype in long-term cultures may be obtained by cel1 sorting or by repeated limiting dilution cloning.

: 6.2.4. DIF ERENTIATION OF YOLK SAC CEL~S IN VITRO
~ After one month of in vitro culture in the :~ : presence o~ LIF, the yolk sac cells were tested for their ability to differentiate into mature blood cells of:all lineag~s in response to various known hematopoietic growth factors including IL-3, IL-2, and EPO.~ When cultured in IL-3 and EPO, the appearance of : red blood cells:was readily detectable in the yolk sac 2~ cultures. In responce to CSF's and IL-3, the yolk sac : cells matured into megakaryocytes and granulocytes.
:Fig. 4 is a blood stain of a yolk sac culture grown in : ~ the:presence of:a~ombination of cytokines and the : appearance of various blood cell lineages can be identified, iIn addition, the expression of various leukocyte surface markers by these cells became detectable, including CD34, CD45, LFA, MAC-l, Ly-l and Ly-2.

:

~ T~TUTE SHE~

3 ~
W~93/02182 - 31 - PCT/US92/05918 6.2.5. DIFFERENTIATION OF YOLK SAC CELLS IN VIVO
A long-term culture of yolk sac cells of BALB/c origin was injected into allogeneic C3H/SCID
mice after 22 passages in vitro. Four weeks later, spleens and livers of the treated animals were analyzed for the presance of donor cells by monoclonal antibodies.
The donor cells were identified by antibodies specific for the donor H-2d haplotype.
Double staining eXperiments utilizing two antibodies ~ further demonstrated that certain subpopulations of the donor cells expressed CD3, Thy-~, B220 and Ml/70 .(FXG. 5). Therefore, these results indicated that the long-term cultured mouse yolk sac cells were capable of di~ferentiating na~urally in vivo into T cells, B
cells and macrophages .

67 :2 . 6. GENERATION OF INMUNOCOMPETENT
CELLS BY YOLK SAC CELLS IN VIVO
2 0 In order to examine whe~he~ the yolk sac cells could give rise to functionally mature blood cells, the long term cultured yolk sac line was transferred in vivo into allogeneic SCID or BNX mice, : ~ and tested for spPcific antibody production. When the 25 ~B~X mice received yolk sac cells and subsequently were : immunized with LPS a month later, specific antibody tit~rs were detected in the sera (FIG. 6). As LPS is a polyclonal B~cell activating agent, this result shows the presence of functionally active 3~ antibody-producing B cells. Additionally, when SCID
mice were injected with yolk sac cells followed by HSA
im~unization, which is a T cell dependent antigen, an antibody reeponse was again detectable, suggesting that lony term cultured yolk sac cells could diff~rentiate to become immuno-competent T and B cells in vivo.

TITUTE S'~EET

9~

6.2.7. YOLK SAC CELLS REPOPULATE
CHEMO-ABLATED MOUSE SPLEENS
Certain classes of chemotherapeutic drugs are effective, and have been used, as ablative agents for bone-marrow in bone-marrow transplantation procedures (Floersheim and Ruszkiewicz, 1969, Nature 222:854). One o~ the most effectiYe agents used t~
replace whole body irradiation in bone-marrow transplantation procedures is the drug Busulfan (~utschka et al., 1987, Blood 70:1382)~ Through careful tItration of the dose of Busulfan and the use of inbred lines of mice ~C57BL/6) of a defined age and weigh~ (3-4 weeks of age~, doses o~ Busulfan have been det~rmined which fully ablate the bone-marrow of these mice but do not directly kill them. These doses of Busulf an result in the sventual death of the treated mice between 11 and 14 days if they do not receiYe tran~p}anted bone-marrQw. This dose is 65 mg of Busulfan/g:of body weigh~ administered in a single dose by I.P. in3ection. When C57BL/6 mice were treated:with this dose of Busulfan and then received an I.Pv injection o~ 106 syngeneic cultured yolk-sac cells 24 hrs. following Busulfan treatment, the transplant recipients revealed spleen repopulation at day 7 and:14 post Busulfan treatment (FIG. 7). On day 7,:spleen colony formation within the recipient was observe~, indicative~of the initial stages of splenic repopulation ~y~the~transplant. Additionally, comparison at day 12 post treatment; of the spleens of control ~usulfan treated mice not receiving yolk-sac transplants and those~ animals receiving transplants howed a~marked dif:ference in splenic viability.
~hile the spleens of control animals were dark, aimost black in color, and appeared necrotic, the spleens of transplant recipients displayed a red/pink color and appeared normal and healthy. Further, the survival SU~Ti ~ ~'TE S~E~r ~93/021X~ 2113 ~ ~ 5 PCT/US~2/0591~

time of the yolk sac cell-treated mice was extended to between l8 and 20 days.

6.2.8. IN UTERO ADMINISTRATION OF
YOLK SAC CELLS RESULTS IN
TISSUE CH.IMERISM ____ A long-term cultured yolk sac cell line was te~ted for its ability to survive in an allogeneic ho~t~ l0,000-50,000 B~LB/c yolk sac cell~ after 13-20 passages in ~itro were injected in utero in day 8 embryos of C57BL/6 mics. At birth, the spleens and livers of the neonates were harvested and analyzed for the presence of donor cells.
Since the donor cells were of the H-2d haplotype, a monoclonal antibody specific for H-2d antigens was used to id~ntify the donor cells by flow cytometry analysis. FIG. 8 presents the results from two neonates examined and it clearly shows that donor : cells were present in both he liver and spleen of the : recipient mice in su~stantial number~. Therefore, in terQ administration of yolk sac cells into : MHC-mismatched mice resulted in tissue chimerism, and : survi~al and homing of the cells to the lymphohematopoietic organs. Tissue chimerism was retained when the mouse tissues were examined even one month after birth.

: 6.2.9. XENOGENEIC TRANSPLANTATION
: ~ OF YOLK SAC CELLS RE5ULTS
IN LONG-T~RM PERSISTENCE OF
~ELLS IN VIVO _ _ In order to test the feasibility of using yolk sac ceIls in xenogeneic transplantation and reconstitution, long-term cultured BALB/c yolk sac cel1s were injected into a newborn Hampshire sheep and a Nubian goat. The sheep receiv2d 40xl06 murine yolk sac cells intravenously at day 3 after birth and the S~TITUT~ SHEET

WO93/02182 _ 34 _ PCT/US92/0591~
2 1 ~ ~ 'J ~ j ;3 goat received the same cell dose at day 7 after birth.
Four days later, both animals received a second dose of 200xl06 cells. After four additional days, a final injection of 60xlO6 cells was given, the peripheral 5 blood mononuclear cells were harvested for antibody staining and flow cytometry analysis about one and a half month later.
FIG. 9 demonstrates that a substantial n D ber of blood cells obtained from the sheep were :~ lO reactive with anti-H-2d antibody. While there were : ~ lower numbers of donors in the peripheral blood of the goat, donor cells were nonetheless detectable. In addition, cells expressing the murine T cell marker Ly-l were also present from both animals. However, 15 neither animal had cells that were positive for the :;.
mu~ine macrophage marker Mac-l, consistent with the :fact that macrophages are not normally present in the peripheral blood.
The results~ of~this experiment are revealing in a number of ways.; It illus~rates the possibility of xenogene;ic reconstitution using murine yolk sac cells:. Neither:animal was pre-treated with i:rradiation or oytotoxic~drug. The high cell doses and~the repetitive injections did not induce graft 25 ~rejection~. Both:~animals~also appeared normal and heàlthy, having~;no~indication of graft versus host reaction.~ The~consis~tent finding of a high number of donor cells~recoverable~;from the sheep than the goat : may be a result ~of~the goat being of an older age before recei~ing:the first cell injection. The yvunger age of the sheep when it was given the first cell dose might have resulted in a more efficient.
~; induction of tolerance. However, there was still : acceptance of the donor cells in the goat in the : :35 absence of any prior immunosuppressive treatment. If :

~U3~T~TUTE ~ET

2 1~ ~ 3 i~~
~'~ 93/0~182 PCT~'S92/0591 induction of tolerance is the mechanism underlying this observation, this further sugyests that the tolerized hosts may also accept other solid organs including the heart, liver a~d kidney from xenogeneic donors sharing the same haplotype of the original donors. Finally, the expression of a T cell marker indicates normal di~ferentiation and maturation in yivo, and the absance of macrophages in the p~ripheral blood suggests the appropriate homing of the right cell lineages in the host upon intravenous administration of yolk sac cells.

7. EXAMPLE: IN VIVO TRANSFORMATION
OF YOLK SAC_CELLS . _ :
In this example, both untransformed yolk sac cells and a retroviral vector containing the exogenous gene of interest are injected into the target animal.
The exogenous gene used is the growth hormone gene ~: (bGH~.
2~ Yolk sac cells harvested from day 8 C57~5JL
mouse embryos are cultured on an STO feeder layer system until approximately 20 x l06 cells per culture flask (l50 cm) were generated. Cells were passed to new ~lasks when the cell density became greater than 80%. AlI experiments were performed with cells at passage l0 or greater.
Newborn mice were injected I.P. with yolk sac cells (2 4 x 106) at 3 to 5 days of age. Two months [positive results have been obser~e~ with infections as early as two weeks ~ollowing yolk sac injection) after I.P. injection of yolk sac cells, ~ animals received lxl0~ viral particles of a replication : deficient retroviral vector produced from the Moloney murine L~kemia Virus based Mulligan ~2 packaging cell line after tr~nsformation with the plasmid pLJPCKbGH
by ~.V. injection into the tail vein.

SUEST~T~T~ S~I~E~

W093/~2182 36 - PCT/~IS92/0591~
~113- ~rj . Of 145 animals treat~d by this procedure, 112 were positive for bGH in the serum by ELISA assay.
The following is a breakdown of the positive bGH
levels of these test animals:
20-59 ng/ml = 56 mice 60-100 ng/ml = 20 mice 100-200 ng/ml - 14 mice 200-S00 ng/ml = 13 mice > 500 ng/ml = 9 mice .~
An equal number of control animals were .injected with the retroviral vector but not with the yolk sac cells. None of these were positive for bGH.
It is believ~d that either the retroviral vector specifically can transfect the injected yolk sa~ cells, and/or the yolk sac cells secrete f~ctors having an effeot on the nearby cells, rendering them susceptible to transfection.
::~ 20 ~ 25 ~ :

:
, :

~ ::

:: :

~l~d~ T~ LE~ ~

Claims (24)

WHAT IS CLAIMED IS:

1. A cellular composition comprising a substantially homogeneous population of mammalian yolk sac stem cells displaying a phenotype of CD34+, Thy-1-, MHC class I- and MHC class II- which are capable of differentiating into mature blood cells in vivo.

2. The composition of Claim 1 wherein the yolk sac stem cells are isolated from a yolk sac prior to blood island formation.

3. The composition of Claim 2 wherein the yolk sac stem cells are isolated from a mouse yolk sac at day 7 of gestation.

4. The composition of Claim 2 wherein the yolk sac stem cells are isolated from a human yolk sac at day 10 of gestation.

5. A method of preparing a cellular composition of mammalian yolk sac stem cells comprising:
(a) excising a yolk sac from a mammalian embryo;
(b) detaching mammalian yolk sac stem cells from yolk sac membrane; and (c) isolating a substantially homogeneous population of mammalian yolk sac stem cells displaying a phenotype of CD34+, Thy-1-, MHC
class I- and MHC class II-.

6. A method of expanding a cellular composition of mammalian yolk sac stem cells comprising culturing, in the presence of an agent which suppresses cellular differentiation, a substantially homogeneous population of mammalian yolk sac stem cells displaying a phenotype of CD34+, Thy-1-, MHC class I- and MHC class II-.

7. The method of Claim 6 wherein the agent is leukemia inhibiting factor.

8. The method of Claim 6 wherein the agent is stem cell factor.

9. A method of hematopoietic reconstitution comprising administering the yolk sac stem cells of Claim 1 to an animal.

10. The method of Claim 9 wherein the cellular composition is administered intravenously.

11. The method of Claim 9 wherein the cellular composition is administered in utero.

12. The method of Claim 9 wherein the animal is a mouse.

13. The method of Claim 9 wherein the animal is a sheep.

14. The method of Claim 9 wherein the animal is a goat.

15. The method of Claim 9 wherein the animal is a human.

16. The method of Claim 15 wherein the human is infected with the human immunodeficiency virus.

17. A non-human animal having a hematopoietic system reconstituted with mammalian yolk sac stem cells.

18. The animal of Claim 17 wherein the animal is a mouse.

19. The animal of Claim 17 wherein the animal is a sheep.

20. The animal of Claim 17 wherein the animal is a goat.

21. A method of producing blood cells in vitro, comprising culturing, in the presence of a growth factor, a substantially homogeneous population of mammalian yolk sac stem cells displaying a phenotype of CD34+, Thy-1-, MHC class I- and MHC class II- which are capable of differentiating into mature blood cells in vivo.

22, The method of Claim 21 wherein the growth factor is EPO, IL-2, IL-3, G-CSF, M-CSF, GM-CSF, or a combination thereof.

23. A method of producing blood cells in an animal comprising:
(a) injecting into an animal a substantially homogeneous population of mammalian yolk sac stem cells displaying a phenotype of CD34+, Thy-1-, MHC class I- and MHC class II which are capable of differentiating into mature blood cells in vivo; and (b) collecting blood cells from the animal.

24. A method of tolerizing an animal comprising administering a substantially homogeneous population of allogeneic or xenogeneic mammalian yolk sac stem cells displaying a phenotype of CD34+, Thy-1-, MHC class I- and MHC
class II-.