R802 Dispatch

Cell therapy: In search of pluripotencyAustin Smith

Cell replacement as a route to treat cellular diseaseand injury is currently limited by the availability of suitable donor cell populations, but recent results withmouse embryonic stem cells suggest that isolatedhuman pluripotent cells could provide a source of cellsfor transplantation and gene therapy.

Address: Centre for Genome Research, University of Edinburgh,King’s Buildings, West Mains Road, Edinburgh EH9 3JQ, UK.

Current Biology 1998, 8:R802–R804http://biomednet.com/elecref/09609822008R0802

© Current Biology Ltd ISSN 0960-9822

Injury and disease states are often associated with loss ordeath of cells or with cellular dysfunction. In many suchcases, cell replacement is a desirable option for restorationof tissue function. Bone marrow transfer to replaceleukaemic blood cells is perhaps the best known exampleof this approach. Cell transplantation might also provide auseful way of delivering molecules for genetic therapy,either gene correction or gene addition. Stem cells derivedfrom the early embryo might be expandable in culture tosupply specific cell types for regenerative transplants. Thispossibility is considered here from both practical and sci-entific standpoints.

Successful cellular transplantation in general requires theuse of stem cells. A molecular definition of a stem cellremains elusive. At the cellular level, however, a stem cellcan be defined as a cell that has the capacity both to pro-duce identical daughter cells — ‘self-renewal’ — and toproduce daughters that are fated to differentiate — ‘com-mitment’. The potential for self-renewal means that in

principal a stem cell population can be maintained orexpanded indefinitely (Figure 1).

Self-renewal is central to transplantation for two differentreasons. The first is that it makes possible the long-termmaintenance of grafts of renewing tissues. In tissues thatundergo continuous turnover, such as the haematopoieticsystem, functionally mature cells survive for only finiteperiods. A red blood cell, for example, has an average lifespan of around 100 days. Effective transplantation there-fore requires both initial reconstitution and ongoingreplenishment of the entire tissue. This can only beachieved from a transferred stem cell population with thecapacity for long-term self-renewal.

The second reason why self-renewal is so important is thatit enables the large-scale production of cells for transplan-tation. In the case of ‘cytostatic’ tissues that do not under-go continuous turnover, such as the brain, transfer of stemcells should not be necessary for long-term graft mainte-nance. The mature neurons found in the adult brain arenot readily isolated, however, and in any case they have lit-tle or no capacity to integrate into a new environment.Adaptable precursor cells can be obtained from foetal tis-sue, but this supply is limiting. The isolation and ex vivoexpansion of stem cells is likely to prove crucial, therefore,for production of immature cells that are competent forfunctional incorporation into the adult tissue.

Efforts to isolate, expand and genetically manipulate stemcells from adult tissues have to date met with only partialsuccess. An alternative to the primary derivation of tissue-specific stem cells would be the development of an in vitrosystem in which lineage-restricted stem cells could be gen-erated from a founder stem cell population of humanpluripotent stem cells (HPCs). Pluripotent cells with thecapacity to generate all foetal and adult cell types existonly in the early embryo and in a particular type of germcell tumour, teratocarcinoma.

HPCs from teratocarcinomas have been established inculture, but their tumour origin makes them unsuitablefor therapeutic exploitation. In mice, however, the isola-tion of teratocarcinoma stem cells was followed by thederivation of non-transformed pluripotent stem cellsdirectly from embryos [1]. These embryonic stem (ES)cells are genetically and phenotypically stable if culturedappropriately, and can be expanded indefinitely. Theycan be induced to differentiate in vitro into multiple lin-eages, including haematopoietic, myogenic, cardiac andneural cell types.

Figure 1

Alternative stem cell fates.

Stem cell

ApoptosisCurrent Biology

DifferentiationSelf-renewal

If undifferentiated ES cells are transplanted back into amouse embryo they will contribute extensively to the devel-oping foetus. The resulting chimaeric mice contain ES-cell-derived functionally differentiated cell types in all tissues.This paradigm can be extended to transplantation intoadults following in vitro differentiation. Haematopoietic,cardiac and neuronal derivatives of ES cells have recentlybeen transferred directly into adult mice with evidence offunctional engraftment [2–4].

The isolation of HPCs analogous to mouse ES cellswould thus create the possibility of generating humancells for transplantation. To make such a strategy possi-ble, there are a number of key features that the HPCsshould exhibit. Firstly, they should be pluripotent so thatany desired cell type can be produced in vitro or in vivo.Secondly, they should be immortal, so that, unlike pri-mary cells, they undergo unlimited expansion. Thirdly,they should be phenotypically stable, expressing noimmortalising or transforming genes so as to minimisethe risk of deregulated differentiation and tumour devel-opment. And fourthly, they should be geneticallytractable so as to facilitate precise genome modifications,such as the introduction of therapeutic genes or engi-neering of immunocompatibility.

So what are the prospects of isolating such cells? Untilfairly recently, proven ES cells had only been demon-strated in rodents. In a very significant development,however, Thomson and colleagues [5] have succeeded inestablishing several pluripotent cell lines from embryosof both marmosets and rhesus macaques. Taken in con-junction with the occurrence in humans of teratocarcino-mas containing pluripotent stem cells, the evidence sug-gests that the biological framework may be permissivefor the isolation of HPCs. That this has not beenachieved to date is in part due to the legislative and

logistical issues involved in research with human embry-onic material.

The conventional route to isolation of mouse ES cells isthe culture of pre-implantation blastocysts, which containa pluripotential stem cell population, the epiblast. Inmany countries, research using pre-implantation stages ofhuman development is either not permitted or, as in theUnited States, is legal but may not be supported by gov-ernment funds. In other countries, such as the UnitedKingdom, legislation allows for certain types of researchunder statutory guidelines. Eggs may be donated forresearch by couples receiving infertility treatment, or cou-ples who have completed such treatment and have eggsremaining in frozen storage. However, a high proportionof the eggs tend to be abnormal and fail to develop to theblastocyst stage. The overall availability of human blasto-cysts is thus a limiting factor.

An alternative route is to attempt to derive HPCs at laterstages using aborted foetal tissue. In the mouse, pluripo-tent cells with all the essential features of ES cells can be established by culturing primordial germ cells from thedeveloping foetus [6]. The mechanism by which germcells convert to ES cells is obscure, but recent work inpigs [7] indicates that this phenomenon may be repro-duced in different mammals. Preliminary studies inhumans suggest that pluripotent cells may be isolated bythis route, although these cells have yet to be charac-terised in detail [8].

Progress in establishing HPCs is also limited by ourignorance of the molecular basis of the pluripotent pheno-type. The ‘POU’ domain transcription factor Oct-4 seemsto be an essential hallmark of a pluripotent stem cell [9],but the mechanisms that regulate Oct-4 expression are notknown. The continuous propagation of mouse ES cells can

Dispatch R803

Figure 2

Stem cell therapy via somatic cellreprogramming. Biopsy somatic cells

Transfer nucleus toenucleated oocyte

Reprogram anddevelop to

blastocyst stage

Isolate and propagatepluripotent stem cells

Transplant

Current Biology

Differentiatedcell types:HaematopoieticNeuronal

CardiacMyogenic

R804 Current Biology, Vol 8 No 22

be sustained by leukaemia inhibitory factor (LIF) or relat-ed cytokines that activate a second transcription factor,STAT3 [10]. With the exception of embryos from strain129 mice, however, provision of LIF alone does not appearsufficient for the initial generation of ES cells from eitherepiblast or germ cells. This may be because additional sig-nals are required [6,11], and/or because signals derivedfrom differentiated cell types may induce differentiation orapoptosis of the stem cells [12]. This is an area in need ofsystematic research if we are to develop rational and robustmethods for isolating HPCs.

A second area in which our present knowledge is inade-quate is in the generation of pure populations of definedcell types. Using current protocols, the differentiation ofES cells is heterogeneous and disorganised, producing amixture of cell types. Mixed cell populations are not suit-able for transplantation, because they can provoke inap-propriate host responses and may impede access of cells ofinterest to the host microenvironment. Furthermore, ifundifferentiated HPCs persist, they may initiate teratomadevelopment. The characterisation of inductive pathwaysthat mediate lineage and cell-type-specific differentiationin the mammalian embryo holds the key to progress. Thismay be complemented by techniques for purifying specif-ic cell types, for example by applying a genetic selectionfor expression of a marker gene [4,13].

Non-identity with the recipient would be a problem fortransplantation using HPCs from embryonic sources.Immunosuppression would be required to avoid allogene-ic rejection. It may prove possible to reduce this problemby genetic manipulation of histocompatibility loci. A pre-ferred solution, however, would be to derive HPCs direct-ly from the patient. Following the demonstration thatnuclei of adult somatic cells can be reprogrammed to allowcloning — as in the production of the cloned sheep ‘Dolly’— it is realistic to conceive of generating material for HPCderivation by nuclear transfer. Indeed in the future it maybecome routine for HPCs to be derived from individualsearly in life and stored frozen for later use in regenerativeand repair therapies (Figure 2).

Finally, ethical concerns have been expressed about efforts toestablish HPCs. Such concerns derive from the requirementto use human embryonic material, as discussed above, andfrom the formal possibility that HPCs could be exploited forhuman germline modification. With regard to the latter con-cern, there are already various methods available by whichthe human germline could in theory be manipulated, and theadvent of HPCs would not seem to present any significantnew issues. The potential for transforming clinical practiceshould therefore outweigh hypothetical scenarios of abuse.

AcknowledgementsThe author’s research is supported by the Biotechnology and BiologicalSciences Research Council of the United Kingdom.

References1. Robertson EJ: Teratocarcinoma and Embryo-derived Stem Cells: a

Practical Approach. Oxford: IRL Press;1987.2. Hole N, Graham GJ, Menzel U, Ansell JD: A limited temporal win-

dow for the derivation of multilineage hematopoietic progenitorsduring embryonal stem cell differentiation in vitro. Blood 1996,88:1266-1276.

3. Deacon T, Dinsmore J, Costantini LC, Ratliff J, Isacson O: Blastula-stage stem cells can differentiate into dopaminergic and seroton-ergic neurons after transplantation. Exp Neurol 1998, 149:28-41.

4. Klug MG, Soonpaa MH, Koh GY, Field LJ: Genetically selected car-diomyocytes from differentiating embryonic stem cells form sta-ble intracardiac grafts. J Clin Invest 1996, 98:216-224.

5. Thomson JA, Marshall VS: Primate embryonic stem cells. Curr TopDev Biol 1998, 38:133-65.

6. Matsui Y, Zsebo K, Hogan BLM: Derivation of pluripotential embry-onic stem cells from murine primordial germ cells in culture. Cell1992, 70:841-847.

7. Piedrahita JA, Moore K, Oetama B, Lee C, Scales N, Ramsoondar J,Bazer FW, Ott T: Generation of transgenic porcine chimaerasusing primordial germ cell-derived colonies. Biol Reprod 1998,58:1321-1329.

8. Shamblott MJ, Bugg EM, Blumenthal P, Huggins G, Gearhart JD:Properties of cell lines derived from human primordial germcells. Am Soc Cell Biol 1997, Abstract:01094.

9. Nichols J, Zevnik B, Anastassiadis K, Niwa H, Klewe-Nebenius D,Chambers I, Schöler H, Smith AG: Formation of pluripotent stemcells in the mammalian embryo depends on the POU transcrip-tion factor Oct-4. Cell 1998, in press.

10. Niwa H, Burdon T, Chambers I, Smith AG: Self-renewal of pluripo-tent embryonic stem cells is mediated via activation of STAT3.Genes Dev 1998, 12:2048-2060.

11. Dani C Chambers I, Johnstone S, Robertson M, Ebrahimi B, Saito M,Taga T, Li M, Burdon T, Nichols J, Smith AJ: Paracrine induction ofstem cell renewal by LIF-deficient cells: a new ES cell regulatorypathway. Dev Biol 1998, in press.

12. Brook FA, Gardner RL: The origin and efficient derivation ofembryonic stem cells in the mouse. Proc Natl Acad Sci USA1997, 94:5709-5712.

13. Li M, Pevny L, Lovell-Badge R, Smith AG: Generation of purifiedneural precursors from embryonic stem cells by lineage selec-tion. Curr Biol 1998, 8:971-974.