Advertisement
Lost in transdifferentiation
Mark H. Hoofnagle, Brian R. Wamhoff, and Gary K. Owens
Department of Molecular Physiology and Biological Physics, University of Virginia, Charlottesville, Virginia, USA.
Address correspondence to: Gary K. Owens, Department of Molecular Physiology and Biological Physics, University of Virginia, MR5 Room 1220, 415 Lane Road, PO Box 801394, Charlottesville, Virginia 22908, USA. Phone: (434) 924-2652; Fax: (434) 982-0055; E-mail: gko@virginia.edu.
Find articles by Hoofnagle, M. in: JCI | PubMed | Google Scholar
Department of Molecular Physiology and Biological Physics, University of Virginia, Charlottesville, Virginia, USA.
Address correspondence to: Gary K. Owens, Department of Molecular Physiology and Biological Physics, University of Virginia, MR5 Room 1220, 415 Lane Road, PO Box 801394, Charlottesville, Virginia 22908, USA. Phone: (434) 924-2652; Fax: (434) 982-0055; E-mail: gko@virginia.edu.
Find articles by Wamhoff, B. in: JCI | PubMed | Google Scholar
Department of Molecular Physiology and Biological Physics, University of Virginia, Charlottesville, Virginia, USA.
Address correspondence to: Gary K. Owens, Department of Molecular Physiology and Biological Physics, University of Virginia, MR5 Room 1220, 415 Lane Road, PO Box 801394, Charlottesville, Virginia 22908, USA. Phone: (434) 924-2652; Fax: (434) 982-0055; E-mail: gko@virginia.edu.
Find articles by Owens, G. in: JCI | PubMed | Google Scholar
Published May 1, 2004 - More info
J Clin Invest. 2004;113(9):1249–1251. https://doi.org/10.1172/JCI21761.
© 2004 The American Society for Clinical Investigation
The prevailing theory of smooth muscle cell (SMC) contribution to vessel lesions is that in pathological states, such as injury and atherosclerosis, SMCs migrate to the intima from the media of the vessel (1). This theory, which has persisted for three decades, is now being challenged by results from models of vessel injury, transplant arteriosclerosis (TA) models, and human allograft studies indicating that a portion of the cells bearing SMC differentiation markers in intimal lesions may have originated from the hematopoietic system and/or circulating progenitor cells (2–5). However, these studies show variable contributions of marrow-derived cells to lesions, with increased frequency correlated to severity of medial injury or to degree of donor/allograft mismatch (6). It appears that only with necrosis of medial SMCs are bone marrow cell (BMC) investment frequencies extremely high (7), indicating that the marrow is not solely responsible for populating the intimal lesion but may represent a default pathway for SMC regeneration and vessel repair in circumstances of severe vessel wall damage.
Two seminal studies by Hu et al. provide new insights into the origin of cells populating the developing neointima in a mouse model of TA (8, 9). Using a combination of bone marrow reconstitution and vein allografts (Figure 1), they demonstrated that donor BMCs give rise to neointimal cells and smooth muscle (SM) α-actin–positive cells (2, 3), although evidence for this is somewhat controversial due to the lack of high-resolution confocal microscopic analyses showing definitive colocalization of bone marrow lineage marker genes and SM α-actin immunostaining (6). However, bone marrow reconstitution studies using donor bone marrow from SM22 promoter-LacZ–transgenic mice have shown that donor BMCs do not give rise to SM22-expressing neointimal cells (2, 3) — a surprising result, since SM22 is an early SMC differentiation marker gene (6). Donor vein graft cells also do not contribute SM22-positive neointimal cells, but host vessel–derived cells can give rise to the population of SM22-positive graft neointimal cells. However, it must be noted that the failure of donor vein graft cells to contribute to the intima appears to be unique to this allograft model, wherein there is virtual destruction of all donor vessel cells, including medial SMCs. Thus, neointimal cells that express SM22 do not appear to be bone marrow–derived but rather to have some other host-based source.
Figure 1What are the true origins of the SMCs present in the intimal lesions of TA? Hu et al. previously demonstrated that donor BMCs give rise to neointimal cells (I) and that a low percentage costained with smooth muscle α-actin, suggesting transdifferentiation of BMCs to SMCs. (II) However, when SM22-LacZ bone marrow was used for lineage tracing, the neointimal cells were SM22-LacZ negative, suggesting that the BMC-derived SM α-actin_positive intimal cells were some cell type other than smooth muscle (e.g., macrophages or myofibroblasts) or were not fully differentiated SMCs (III). Interestingly, cells from the host vessel (V) but not the donor vessel (IV) can give rise to intimal cells that are SM22-positive. However, it must be noted that the failure of donor vessel cells to contribute to the intima appears to be unique to this allograft model, wherein there is virtual destruction of all donor vessel cells, including medial SMCs. Thus, in this allograft model, a key question is, what is the source of the SM22-positive intimal cells derived from the host vessel? In this issue of the JCI, Hu et al. show that Sca-1+ cells purified from the adventitia of the aortic root (VI) can migrate through an irradiated vein graft to the neointima of the vessel and express SM22. These results are of interest in that they identify a potential new source of cells that contribute to neointimal formation in TA. However, several outstanding questions remain: Is this an artifact of exogenous application of the Sca-1+ cells? Does it represent fusion of these cells to SMCs? Or does this represent bona fide transdifferentiation of Sca-1+ cells from the adventitia into SMC lineages?
In this issue of the JCI, Hu and colleagues hypothesize that in addition to circulating progenitor cells, Sca-1+ progenitor cells that reside in the adventitia may transdifferentiate into SMC-like neointimal cells (10). These results provide fresh arguments in a key controversy in adult stem cell and progenitor cell research — i.e., that of defining the relative roles of cell fusion versus transdifferentiation.
The role of adult stem cell contribution to tissues has recently undergone a revolution and counterrevolution. Initial optimism about transdifferentiation of neuronal stem cells (11), hematopoietic stem cells (HSCs) (12), and other adult stem cell types has diminished in the face of reports of cell fusion (13) and minimal adult stem cell plasticity (14). Surprisingly, the pair of papers that began the furor over cell fusion throughout the adult stem cell field had little relevance to most cell types being studied. Terada et al. and Ying et al. demonstrated rare cell fusion events when embryonic stem cells are extensively cocultured with BMCs or neural cells, respectively (15, 16). These results, however, did lead to a healthy skepticism about the transdifferentiation potential of adult stem cells. Ianus et al. developed a Cre-lox model to specifically test for fusion in their cell systems, and found none, despite the apparent transdifferentiation of BMCs into insulin-secreting pancreatic β cells (17). Some groups examining BMC transdifferentiation into hepatocytes found that their previous results were mainly due to cell fusion (13). Still others, alerted to the fusion problem, found cell fusion occurred in their systems, but at too low a frequency to account for their transdifferentiation results (18, 19).
Thus, a confusing picture has emerged for most scientists curious about the role of adult stem cells in human disease. We are left with several questions: Is transdifferentiation real or just an artifact? Is fusion a problem or is it a potential physiological mechanism to exploit? Are some tissues more susceptible to fusion than others? The liver, for instance, has proven to be highly susceptible to fusion, but, interestingly, in the fumarylacetoacetate hydrolase–/– mouse models studied, fusion of HSCs to hepatocytes corrects the enzymatic deficiency, raising the possibility that this phenomenon has therapeutic applications (12). Of note in this issue of the JCI is the work of Camargo et al., who explored the fusibility of hepatocytes with HSCs and discovered that it is specifically the myeloid lineage of these cells, rather than the stem cells themselves, that is responsible for fusion, thus raising the intriguing possibility of “fusion therapy” (20). These groundbreaking new results also emphasize the importance of understanding fusion and transdifferentiation, as they appear to play a potentially important role in several disease states that biomedical scientists have been trying to understand for decades.
Adding to the controversy over the true transdifferentiation potential of adult stem cells are the exaggerated claims made by both the scientific and the lay press regarding the relative merits of adult versus embryonic stem cells and their potential for therapeutic applications. Both avenues of research would be better served by more accurate portrayals of their science in the press, which might help to lessen the impact of partisan politics on serious consideration of the merits of either stem cell type and their potential contributions to science and human therapeutics.
In this issue of the JCI, Hu et al. provide novel evidence that may result in a more coherent picture of different cell contributions to the lesions of TA. Of major interest is the fact that Hu et al. isolated a putative SMC progenitor population from the adventitia of the aortic root with a surface phenotype of Sca-1+/c-kit+/lin– characteristic of HSCs. Further, through a series of lineage tracing studies, they demonstrated that these cells were not of hematopoietic origin, suggesting they are not HSCs that have lodged or fused in the adventitia. In culture, PDGF-BB induces expression of multiple markers of SMC differentiation, and these cells, when purified from an SM22-LacZ/ApoE–/– transgenic mouse and applied to the outside of an irradiated vein allograft, migrate to the neointima of the vessel and appear to activate de novo SM22 transcription in the neointima (Figure 1).
These results are exciting and significant in that they identify a potential new origin of cells that contribute to neointimal formation in TA. However, caution must be taken, as critical experiments are required to identify the physiological relevance of these cells and eliminate two limitations of this study. First, application of exogenous Sca-1+ cells to an irradiated vein graft represents an artificial situation in which migration may have been an experimental artifact due to lack of medial SMC contribution. To determine if such migration is physiologically relevant, a Sca-1 promoter-Cre mouse could be crossed to a ROSA-stoplox-LacZ/ApoE–/– mouse to determine if Sca-1 was ever activated in neointimal SMCs. Second, fusion of Sca-1+ cells with cells from another source, such as irradiated SMCs of the media, must be eliminated, although other groups suggest that this is an unlikely eventuality (19). A Cre-lox fusion detection system (17) could be used to quell concerns that Sca-1+ cells adopt the SMC phenotype by fusing to other SMCs rather than by transdifferentiating.
From all studies so far we are still left with the fundamental question: Do non-SMCs from the adventitia, host vessel, or bone marrow transdifferentiate to become bona fide SMCs? It is possible that these cells are expressing only a few SMC differentiation markers and/or still expressing non-SMC genes. Finally, and most importantly, we ask, are the results in these animal models relevant to the development of atherosclerosis in humans? Can these same cells be identified in humans by Sca-1 antigen? The novel and exciting studies by Hu et al. in this issue of the JCI implicate potential contributions by a putative adventitial progenitor cell population in TA and might open yet another chapter in the long quest to define origins of cells in intimal lesions and their mechanistic contributions to the pathogenesis of atherosclerosis.
Copyright © 2024 American Society for Clinical Investigation
ISSN: 0021-9738 (print), 1558-8238 (online)