[讨论]Stem cells: Lost in transl

Stem cells: Lost in translation

KENNETH R. CHIEN

Kenneth R. Chien is at the Institute of Molecular Medicine, University of California, San Diego, La Jolla, California 92093, USA.
e-mail: kchien@ucsd.edu


The potential use of stem cells as agents of repair in human disease makes them the subject of high-profile studies. But we should be wary of prematurely pushing laboratory research into clinical practice.


On the road map for 'translational medicine' — often referred to as bench-to-bedside research — stem-cell therapy is a prime destination. Stem cells are cells that have not taken on the identity of any specific cell type and are not yet committed to any dedicated function; they can divide indefinitely and may be induced to give rise to one or more specialized cell types. They not only offer scientists a tool to study the early molecular events in organ development, they also offer hope for tissue repair and regeneration to patients suffering from a spectrum of degenerative diseases.

Given this level of excitement, it is hardly surprising that the theory-to-therapy approach to the use of stem cells would be thrust forward into translational human studies at the earliest possible stage. A prompt for action came in 2001 with the publication of a paper in Nature by Orlic and colleagues1 suggesting that stem cells derived from bone marrow can replace heart muscle lost as a result of heart attack, and can improve cardiac function. Injecting bone-marrow stem cells into an injured heart potentially represented a new therapy, triggering the launch of numerous clinical studies to investigate the effect of directly injecting these cells into the damaged heart muscle of patients following a heart attack. In new papers published online and accompanying this article, independent studies by Murry et al.2 and Balsam et al.3 seriously challenge Orlic and colleagues' initial observations and the scientific underpinnings of the ongoing human studies.

Both sets of authors used state-of-the-art genetic tools to examine whether bone-marrow stem cells transplanted into damaged hearts (Fig. 1) could take on the role of heart muscle cells and improve heart function. Murry et al.2 used bone-marrow stem cells that included a foreign gene — LacZ — in their genetic complement and that only expressed it when they were in the heart muscle, creating a specific molecular tag. Little or no activity of this 'reporter' could be seen after the stem cells were injected directly into damaged hearts in mice. Balsam et al.3 found that bone-marrow stem cells labelled with a different tag — green fluorescent protein (GFP) — also showed little evidence of becoming cardiac muscle cells after they were transplanted into damaged mouse hearts. Not only that, the stem cells developed into different blood-cell types, despite being in the heart.


Figure 1 Two strategies used to show that bone-marrow stem cells do not take on the role of damaged heart cells. Full legend

High resolution image and legend (71k)



Although Orlic et al.4 have also previously suggested that it is normal for bone-marrow stem cells to circulate in the blood, continually homing, repopulating and regenerating heart muscle cells, these two new studies indicate that this is probably not the case. GFP-tagged bone-marrow stem cells that were continually transfused into the bloodstream of a mouse with a large scar resulting from a heart attack showed no signs of becoming heart muscle cells3. The new reports raise serious concerns regarding the feasibility of using stem cells derived from the bone marrow to drive cardiac regeneration.

So scientists are asking why there are wide discrepancies between the earlier report and the current investigations. As Murry et al.2 suggest, the differences may arise from the difficulty of tracking the in vivo fate of transplanted cells within an intact organ. Orlic et al.1 mainly relied on detecting unique protein constituents of bone-marrow stem cells using fluorescently tagged antibodies. Murry et al.2 and Balsam et al.3, however, created intrinsic genetic markers that can be easily recognized without antibody staining. Owing to its high density of muscle-specific contractile proteins, intact heart muscle tends to have a high inherent background fluorescence, and can also display nonspecific antibody binding to the abundant muscle proteins. This makes it difficult, even for the most experienced labs using the most specialized microscopes, to track cell fate by simply using techniques that rely on fluorescent antibody staining of cardiac proteins.

For exactly this reason, genetic markers have long been used as the most reliable way to trace cell lineages during heart development (for example, see ref. 5). Similarly, specific genetic techniques that can determine whether stem cells have taken on muscle-cell-specific behaviour because they have fused with native muscle cells, rather than because their genes have genuinely become reprogrammed to assume the new cell type (transdifferentiation), have become routine in studies of transplanted stem cells in the heart6, 7 and many other organs6, 8, 9.

Over the past few years, various experiments using several types of stem cells have supported the view that transdifferentiation occurs rarely, if at all, in many organ systems, including heart muscle6, 7. Less than 2% of the transplanted or injected cells take on the in vivo fate of heart cells. If this is true, then the improvement in cardiac function seen by Orlic et al.1 might have arisen not because the stem cells transdifferentiated, but because new blood vessels were encouraged to grow around the injected area. Such growth of new blood vessels has been consistently found in transplantation studies of diverse cell types in the heart (for a review, see ref. 10). Recently, studies in large-animal models of transplanted bone-marrow-derived stem cells in the injured heart also failed to document cardiac regeneration. Again, the implication is that any functional improvement seen may not be related to an increase in functioning heart muscle per se11. In addition, a recent clinical study, in which bone-marrow stem cells were transplanted into injured hearts, has been terminated because of serious cardiac side effects that threatened the blood flow to the heart12. This again suggests that further experimental testing is warranted in large-animal model systems.

For physicians, the use of stem-cell therapy in treating cardiac-muscle diseases remains a worthy, but perhaps longer-term, goal. A glance at what formed the basis for the use of another cell type — myoblasts (progenitors of non-cardiac muscle cells) — in clinical trials for transplanting to damaged hearts is highly instructive (Fig. 2; for a review, see ref. 13). These clinical studies, which used patients' own myoblasts in the hope of increasing the amount of viable muscle after a heart attack, were the culmination of more than seven years of painstaking basic research. They showed that the transplanted myoblasts could adopt a muscle fate in vivo and improve heart function. A rigorous system was implemented to generate sufficient quantities of clinical-grade myoblasts from each patient to allow these cells to be directly injected into their own injured hearts.


Figure 2 From experiment to therapy. Full legend

High resolution image and legend (125k)



The early signs were cautiously encouraging, although there were reports of the patients developing an abnormal heartbeat. This was probably due to electrical disorganization, caused by the failure of the transplanted cells to couple electrically with their neighbours to maintain synchronous contraction and impulse propagation during each heartbeat. Nonetheless, an overriding improvement in heart function was seen in the remaining patients. In short, we have had proof-of-concept that transplantation of muscle-cell progenitors may improve function after a heart attack — the task is to find the ideal source of cells.

Previous results suggest that fetal heart cells might be the optimal cell type for heart-cell transplantation therapy10. They graft easily, adopt the identity of an adult cardiac cell type without fusion, and are electrically coupled. But the controversy that surrounds their use in scientific research means that they cannot be used in clinical studies, at least for the foreseeable future.

Perhaps now is the time to search for the presence of naturally occurring, authentic heart progenitor cells, and to identify and dissect the signals that guide their migration, renewal and differentiation, as opposed to running the risk of becoming lost in translation with cell types that can rarely become functional heart muscle cells. This will require a rigorous approach to defining the molecular pathways that maintain the pool of endogenous cardiac progenitors during the process of heart formation itself. The question will then be whether remnants of these cells can be isolated from the fully developed adult heart and expanded in sufficient quantities to allow transplantation. In short, the key to unlocking the potential of cardiac stem-cell therapy may lie in a firm understanding of the biology of embryonic cardiac progenitors during development (see ref. 14 for a review).

For scientists, physicians and patients, the long-term future of cardiac stem-cell therapy has never been brighter. But we first need to commit the necessary time and resources to identify the best cardiac stem-cell story to translate.



References 1. Orlic, D. et al. Nature 410, 701–705 (2001). | Article | PubMed | ISI | ChemPort |
2. Murry, C. et al. Nature doi:10.1038/nature02446 (2004). | PubMed |
3. Balsam, L. B. et al. Nature doi:10.1038/nature02460 (2004).
4. Orlic, D. et al. Proc. Natl Acad. Sci. USA 98, 10344–10349 (2001). | Article | PubMed | ChemPort |
5. Cai, C. L. et al. Dev. Cell 5, 877–889 (2003). | PubMed | ISI | ChemPort |
6. Alvarez-Dolado, M. et al. Nature 425, 968–973 (2003). | Article | PubMed | ISI | ChemPort |
7. Oh, H. et al. Proc. Natl Acad. Sci. USA 100, 12313–12318 (2003). | Article | PubMed | ChemPort |
8. Vassilopoulos, G., Wang, P. R. & Russell, D. W. Nature 422, 901–904 (2003). | Article | PubMed | ISI | ChemPort |
9. Wang, X. et al. Nature 422, 897–901 (2003). | Article | PubMed | ISI | ChemPort |
10. Dowell, J. D., Rubart, M., Pasumarthi, K. B., Soonpaa, M. H. & Field, L. J. Cardiovasc. Res. 58, 336–350 (2003). | Article | PubMed | ISI | ChemPort |
11. Bel, A. et al. Circulation 108, Suppl. 1, II247–II252 (2003). | Article | PubMed |
12. Kang, H. -J. et al. Lancet 363, 751–756 (2004). | Article | PubMed | ISI | ChemPort |
13. Menasche, P. Heart Fail. Rev. 8, 221–227 (2003). | Article | PubMed | ISI |
14. Chien, K. R. & Olson, E. N. Cell 110, 153–162 (2002). | PubMed | ISI | ChemPort |