ICAPB University of Edinburgh West Mains Road Edinburgh EH9 3JT Scotland lCRC Beatson Laboratories,
Garscube Estate, Bearsden, Glasgow G61 1 BD, Scotland

Introduction

The hematopoietic system can be considered as a spectrum of differentiation and self renewal with at one extreme terminally differentiated cells such as erythrocytes or macrophages, and at the other, pluripotential hematopoietic stem cells (HSC) , which can give rise to all the different lineages of the hematopoietic system, and due to their capacity for self renewal, persist throughout adult life (Spangrude et al. 1991 ). Introduction of limiting numbers of these cells into recipients whose own hematopoietic systems have been ablated (by, for example, irradiation) results in long-term reconstitution by the donor cells. These properties of HSC have make them tempting targets for gene manipulation, both for investigating the molecular and cellular biology of genes that are expressed in different lineages of the hematopoietic system and in somatic gene therapy. In the former case, gene manipulation of HSC would have particular benefit where the mutations, if transmitted through the germ line, would result in embryonic lethality. Gene manipulation in HSC has proved difficult, at least in part due to an inadequate understanding of the biology of these cells; addition of genes (transgenesis) is dependent on retroviral infection (Fairbairn and Spooncer 1993). Gene targeting by homologous recombination in HSC has not yet been shown to be practical. In contrast, there is considerable experience in specific and facile manipulation of the mouse genome using embryonal stem (ES) cells, which as totipotential cells, can give rise to all the different cellular lineages in the organism (Evans and Kaufman 1981, Robertson 1986) .In particular, ES cells have proved to be amenable to gene targeting by homologous recombination. Early work demonstrated the appearance of erythrocytes following differentiation of ES cell under specific conditions in vitro (Doetschman et al. 1985). Under conventional circumstances, these cells could only have arisen from differentiation of pluripotent hematopoietic precursors, themselves derived from the ES cells. ES cells are competent to produce HSC; using HSC isolated from fetal livers of mice derived in their entirety from ES cells are capable of rescuing and reconstituting secondary recipient mice (Forrester et al. 1991 ). The question of whether HSC can be derived from ES cell differentiation in vitro is currently an active area of research (Hole et al. 1994, Muller and Dzierzak 1993). However, most mature hematopoietic lineages can be detected following in vitro differentiation of ES cells (Chen et al. 1993, GutierrezRamos and Palacios 1992, Keller et al. 1993, Nakano et al. 1994, Wiles 1993). As part of our study of the hematopoietic potential of ES cells, we have examined the lymphoid commitment of ES cells differentiated in vitro. The ability to isolate T lymphocytes from ES cells in vitro may offer an alternative strategy in the study of thymocyte development and positive and negative T cell selection.

Materials and Methods

ES cell differentiation.

This was carried out essentially as described (Hole and Smith 1994). In brief, ES cells were cultured as hanging drops in the presence of LIF for 48 hours prior to harvesting (day O of differentiation) and suspension culture. During this suspension culture, the ES cell aggregates differentiate into cystic structures known as embryoid bodies (Doetschman, et al. 1985).


Flow cytometry

Embryoid bodies of varying periods of differentiation were disaggregated into single cell suspension by digestion with Dispase (1 ulml). Primary rat monoclonal antibody (Pharmingen) binding was detected with either goat anti-rat FITC conjugated antibody (T AGO tissue culture services) or streptavidin-FITCIPE (Sigma) as appropriate. Fluorescence was determined on FACScan flow cytometer (Becton Dickenson).


RT-PCR.

Embryoid bodies from various time points were harvested and washed. mRNA was isolated using RNAzol (Biotecs Lab) and cDNA reverse transcribed by AMV reverse transcriptase (Promega) according to the respective manufacturers instructions. PCR reactions were as described (Schmidt et al. 1991 ). In brief, denaturation (95°C), anneal and extension (72°C) times were 20, 60 and 60 seconds respectively for thirty cycles. Anneal temperatures and primers were as noted below.

PCR products from both reverse transcribed and mock-transcribed reactions were run on agarose gels
and detected by ethidium bromide staining.

Results

Flow cytometry The expression of surface antigens by single cells liberated from the differentiating embryoid bodies is shown in figure 1 a-c. HSC from mice are believed to have a characteristic phenotype, namely Ly-6E+ve, Thy-1 +ye (Spangrude, et al. 1991 ). Expression of these antigens was assessed individually (fig.1 a and b) and together (data not shown). Thy-1 was found on undifferentiated ES cells, the numbers of cells expressing the antigen decreasing over the first 5 days of differentiation and apparently increasing thereafter to approximately 30% of embryoid body cells over the time course studied. The numbers of cells expressing both antigens were less than 0.1% of total cells at any time point examined. Thy-1 and Ly6E are not HSCrestricted, and are expressed on other cell types, including T cells. In order to examine the total hematopoietic commitment and T -cell commitment, expression of CD45, CD3 and alfaßTCR were examined (Figs 1 band 1 c). CD45+ve hematopoietic cells increase in number over the time course, from about 1% of cells at day 7 to approximately 7% of cells after 20 days of differentiation. However , a relatively small number of these cells are likely to be T cells; CD3 or af3 TCR +ve cells are first detected around day 10, again increasing over the time course to about 10% of the numbers of CD45+ve hematopoietic population. The small numbers of TCA +ve cells recovered precluded two-colour flow cytometry. However, CD4+ve and CD2+ve cells were also found around the time point when these T cell markers could be detected.

Discussion

This works suggests that lymphoid commitment of ES cells can be demonstrated in vitro. Detectable numbers of cells bearing Iymphoblastoid characteristic markers can be detected at or around 10 days of differentiation in vitro, albeit at a low level. The presence of markers for CD5, CD2, CD3 all indicate that T cell development is detected. Productive expression of alfa ßTCR is dependent on the presence of the product of RAG-1 gene (Wayne et al. 1994). It is not surprising therefore to find that RAG-1 expression in these differentiating cells precedes the appearance of cells bearing alfa ßTCR (fig1 ). The numbers of cells expressing the T cell markers Thy-1 and Ly6E are dramatically higher than those expressing the markers listed above. This is indicative of one of the problems in this system; embryoid bodies contain cells of many different lineages, not just hematopoietic. Indeed undifferentiated ES cells contain already contain transcripts for Thy-1. Embryonal stem cells are clearly competent to produce the entire range of hematopoietic lineages; ES-derived fetal livers contain pluripotent repopulating HSC capable of repopulating the entire hematopoietic system of recipient mice (Forrester, et al. 1991). However, inducing ES cells to progress down all hematopoietic lineages in vitro has proved more elusive. Myeloid lineage progenitors can be detected following in vitro differentiation of ES cells (Wiles 1993), yet reconstitution of the myeloid lineages in lethally irradiated mice by adoptive transfer of ES-derived progeny has proved more elusive. In contrast, in at least one report the lymphoid lineages of mice have been reconstituted with the in vitro differentiated progeny of ES cells (Muller and Dzierzak 1993). Part of the problem with this system may the relatively low numbers of hematopoietic cells arising as a result of in vitro differentiation. In this study, although percentages varied between experiments, the numbers of alfa ßTCR+ve cells were typically less than 5% at best. One approach to enhance the numbers produced may be to exploit cytokines to enhance hematopoietic development. One cytokine that is thought to playa role in early hematopoietic development is FLK2 ligand (Zeigler et al. 1994). At least in this system, transcripts for FLK2 ligand can be detected within the developing embryoid bodies, allowing the possibly of paracrine regulation of hematopoiesis. However, the failure to detect its receptor at this level of analysis suggests that the numbers of cells that may express the receptor may be very low. An alternative, less well defined system would be to use stromal cell lines to support hematopoietic development. This approach has been used by others (Gutierrez-Ramos and Palacios 1992, Nakano, et al. 1994). The presence of certain cytokines may interfere with hematopoietic differentiation in vitro, as has been suggested for M-CSF (Nakano, et al. 1994). Of interest is that transcripts for M-CSF cannot be detected in the early differentiating embryoid bodies in our system (data not shown). The approaches of cytokine enhancement and stromal cell support are currently under study. The detection of cells expressing productive alfa ßTCR suggests that these cells may be capable of positive and negative selection. The relatively facile nature of transgenesis of ES cells may allow us the opportunity to address issues not just of the control of primitive hematopoietic commitment, but of T cell development. The development of these transgenic cells endogenously in embryoid bodies, in vitra in fetal thymus organ culture and in viva in rescued and heamtopoietically reconstituted mice is currently under study.

Acknowledgements

This work was supported by grants from The Wellcome Trust,
The Leukaemia Research Fund and The Royal Society.

References

Chen, J., Gorman, J. R., Stewart, V., Williams, B., Jacks, T. and Alt, F. W. (1993) Generation of normal lymphocyte populations by Rb-deficient embryonic stem cells Curr. Bioi. 3: 405-413

Doetschman, T., Eisetter, H., Katz, M., Schmidt, W. and Kemler, R. (1985) The in vitro development of blastocyst-derived embryonic stem cell lines: formation of visceral yolk sac, blood islands and myocardium. J.Embryol.exp.morphol. 87: 27-45

Evans, M. J. and Kaufman, M. (1981) Establishment in culture of pluripotential cells from mouse embryos Nature 292: 154-156

Fairbairn, L. J. and Spooncer, E. (1993) Retroviral gene transfer into haemopoietic cells In: Haemopoiesis N. G. Testa and I. Molyneux (eds.) Oxford University Press Oxford Forrester, L. M., Bernstein, A., Rossant, J. and Nagy, A. (1991) Long-term reconstitution of the mouse hematopoietic system by embryonic stem cell derived fetalliver Proc.Natl.Acad.Sci.(USA) 89: 7514-7517

Gutierrez-Ramos, J. C. and Palacios, R. (1992) In vitro differentiation of embryonic stem cells into lymphocyte precursors able to generate T and B lymphocytes in vivo Proc.Natl.Acad.Sci.(USA) 89: 9171-9175

Hole, N., Graham, G. and Ansell, J. D. (1994) Self renewing and repopulating hematopoietic progenitors can be derived from embryonal stem cells in vitro J.Cell.Biochem. 188: 181

Hole, N. and Smith, A. G. (1994) Embryonic stem cells and hematopoiesis In: Culture of hematopoietic cells R. I. Freshney, I. B. Pragnell and M. G. Freshney (eds.) Wiley-Liss New York

Keller, G., Kennedy, M., Papayannopoulou, T. and Wiles, M. V. (1993) Hematopoietic commitment during embryonic stem cell differentiation in culture Mol.Cell.Biol. 13: 473-486

Müller, A. M. and Dzierzak, E. A. (1993) ES cells have only a limited lymphopoietic potential after adoptive transfer into mouse recipients Development 118: 1343-1351

Nakano, T., Kodama, H. and Honjo, T. (1994) Generation of Iymphohematopoietic cells from embryonic cells in culture Science 265: 1098-1101

Robertson, E. J. (1986) Pluripotential stem celllines as a route into the mouse germ line Trends. Genet. 2: 9-13

Schmidt, R. M., Bruyns, E. and Snodgrass, H. R. (1991) Hematopoietic development of embryonic stem cells in vitro: cytokine and receptor expression Genes Develop. 5: 728-740

Spangrude, G. J., Smith, L., Uchida, N., Ikuta, K., Heimfiled, S., Friedman, J. and Weissman, I. L. (1991) Mouse hematopoietic stem cells Blood 78: 1395-1402

Wayne, J., Suh, H, Misulovin, Z, Sokoi, K A, Inaba, K. and Nussenzwig, M C (1994) A regulatory role for recombinase activating genes RAG-1 and RAG-2 in T cell development Immunity 1 95-107

Wiles, M V (1993) Embryonic stem cell differentiation in vitro Meth.Enzymol 225900-918

Zeigler, F C., Bennett, B D, Jordan, C. T., Spencer, S D., Baumhueter, S., Carroll,
K. J, Hooley, J., Bauer, K. and Matthews, W. (1994) Cellular and molecular characterisation of the role of the FLK-2/FL T -3 receptor tyrosine kinase in hematopoietic stem cells Blood 84 2422-2430