COMMENTARY

http://biotech.nature.com • MAY 2002 • VOLUME 20 • nature biotechnology

The word very much on the minds of thestem cell community during the past severalyears has been “plasticity.” Several publica-tions have appeared with provocative“alchemical” titles such as “Turning [yourfavorite tissue] into [some other tissue].”Many of these reports claim that somaticstem cells previously thought to be restrictedto producing mature cells specific to theirtissue of origin may in fact be capable of amuch wider spectrum of differentiation.Collectively, these reports have led manyinvestigators to rethink traditional somaticstem cell dogmas, and a minority to claimthat these dogmas are outdated and shouldbe revised. In addition to biological interest,reports of stem cell plasticity also raisenumerous new and exciting therapeutic pos-sibilities. For example, it has been suggestedthat the bone marrow may be a source oftransplantable muscle, liver, and even neu-ronal tissues. Traditionally, it has been con-sidered to be a source of stem cells restrictedto the production of blood cells orhematopoietic populations. In the extreme,some have begun to consider the bone mar-row as a source of stem cells that resembleembryonic stem (ES) cells in their develop-mental capacities. Such arguments are beingemployed by groups opposed to humanembryonic stem cell research.

So what is the real situation regardingsomatic stem cell plasticity? In my opinion,while the numerous studies are intriguing andcertainly interesting, it is not yet time to abandon traditional notions established (atleast in the most extensively analyzedhematopoietic system) by four decades ofresearch. In fact, it may be argued that settingup an oppositional situation—“Is it or is it notplastic?”—is overly narrow and not construc-tive. The main reason I argue these pointsstems from the definition of the word “plas-tic,” which implies the ability of one entity tochange into another. In the case of somaticstem cells, this could be, for example, the abili-ty of a hematopoietic stem cell to change itsfate to become capable of producing liver cells.The lesson to be learned from the field ofhematopoiesis is the value of clonal analysis. Itwas just such an emphasis on clonal analysis

that led to the establishment of the basic prop-erties of stem cells in the mid-1960s1. That is, itwas clearly shown that a single stem cell canboth self-renew and give rise to robust popu-lations of different blood cell types. Thus, toestablish plasticity, it must be clearly anddirectly demonstrated that a single somaticstem cell can give rise to both its “expected”tis-sue and to an“unexpected”tissue.

Given that early hematopoietic studies wereperformed by combining considerablethought and experimental elegance with min-imal,“bare bones” technology, surely the samedegree of clonal precision should be insistedon in this technologically rich postgenomicera. It is surprising and a little sad how fewinvestigators have stressed this point, which inmy opinion is of central importance. A secondpoint involves the ability of a stem cell to giverise to large clones of mature progeny. Forexample, in a situation where a single, tradi-tionally defined hematopoietic stem cellreconstitutes an entire blood cell system andyields a small number of, say, epithelial celltypes, is it fair to claim plasticity? Or is this anaberrant, low-probability phenomenonreflecting the consequences of the extremeproliferative activity of the transplanted cell?In such a case the alleged “plasticity” may beformally equivalent to a few primary cellsescaping senescence during extensive in vitroculture, resulting from accumulated genetic orepigenetic alterations.

Several reports have come close to achiev-ing the necessary level of clonal precisionand satisfying the robust contribution crite-ria outlined above. Examples have includedthe production of hepatic tissue by highlypurified hematopoietic stem cells in thebone marrow2, and the substantial reconsti-tution of cardiac tissue by similarly purifiedbone marrow stem cells3. Although bothstudies are extremely interesting and werewell performed, it is not yet possible to con-clude from them that a hematopoietic stemcell has expanded non-hematopoietic differ-entiation abilities. It seems clear nonethelessthat substantial liver and cardiac differentia-tion potential can be found in some popula-tions of bone marrow cells. Regardless ofwhether this reflects stem cell plasticity ornot, these observations are certainly ofpotential interest from a clinical perspective.An interesting report has suggested that asingle transplanted bone marrow stem cellcan significantly contribute to epithelial cellpopulations, in addition to all blood cell lin-

eages4. All of these studies are extremelypromising; as with all sound science, howev-er, they must be independently verified.

Recently, several publications haveappeared that highlight the sometimes unex-pected complications that can arise in studiesof stem cell plasticity. In short, these studiessupport the notion that extraordinary claimswill require an extraordinary degree of exper-imental rigor. In one case, it has been demon-strated that the hematopoietic activityobserved in muscle cell populations resultsfrom bone marrow–derived cells that are pre-sent in the muscle tissue5. Although the bio-logical significance of this is not clear, it seemsunlikely that muscle stem cells could “transd-ifferentiate” into blood-forming cells5.Indeed, it has been demonstrated that bonemarrow stem cells circulate through theperipheral blood at a surprising rate.Therefore, any reports of blood-formingactivity by non-hematopoietic tissues must becarefully evaluated. A second publicationreports the failure to reproduce an earlierreport that nervous tissue can give rise toblood, raising the possibility that the originalobservations may have been due to changesaccrued during the extensive in vitro culturesthat were employed6. Finally, two studies7,8

have shown that coculturing nervous tissue orbone marrow cells with ES cells can result inrare fusion events that yield tetraploid cells.These cells can acquire at least some of theundifferentiated potential characteristic ofembryonic stem cells. This complicates anyinterpretations regarding the reprogrammingor transdifferentiation of somatic cell nuclei.

The above discussion is not meant todiminish the interest in the possibility ofsomatic stem cell plasticity. Clearly, interestingresults will be obtained from the many ongo-ing studies in this area. Rather, I wish to stressthe need for experimental rigor and cautionbefore old dogmas are ripe for revision.Indeed, the paradigms provided by the experi-ments of the mid-1960s are still very muchalive, and in need of mechanistic explanation.

1. Wu, A.M., Till, J.E., Siminovitch, L. & McCulloch,E.A. J. Cell Physiol. 69, 177–184 (1967).

2. Lagasse, E. et al. Nat. Med. 6, 1229–1234 (2000).3. Orlic, D. et al. Nature 410, 701–705 (2001).4. Krause, D.S. et al. Cell 105, 369–377 (2001).5. McKinney-Freeman, S.L. et al. Proc. Natl. Acad. Sci.

USA 99, 1341–1346 (2002).6. Morshead, C.M., Benveniste, P., Iscove, N.N. & van

Der Kooy, D. Nat. Med. 8, 268–273 (2002).7. Ying, Q.-L., Nichols, J., Evans, E.P., & Smith, A.G.

Nature 416, 545--548 (2002)8. Terado, N. et al. Nature 416, 542--545 (2002)

425

Rethinking somatic stem cell plasticityIhor Lemischka

Ihor Lemischka is an associate professor in thedepartment of molecular biology, PrincetonUniversity, Princeton, NJ 08544(ilemischka@molbio.princeton.edu).

©20

02 N

atu

re P

ub

lish

ing

Gro

up

h

ttp

://b

iote

ch.n

atu

re.c

om