The Workshop on Stem Cell Plasticity was held on March 21, 2000 in the Rockledge ll Building in Bethesda, Maryland. The Working Group was constituted to assess the state-of-science in the stem cell plasticity research area and identify future research directions to be explored in this area. Brief highlights from the scientific presentations, group discussions, and suggestions of potential research directions in this area are presented below.

Dr. Makio Ogawa presented murine data which clearly demonstrated that CD34 expression of adult mouse stem cells is driven by the activation state of the stem cells. Various manipulations were used to "tip" stem cells into cycle such as, 5-fluorouracil injection, culture with c-kit ligand and interleukin-11 and G-CSF mobilization. In all of the above circumstances, stem cells in a "resting state" were shown to be predominately CD34- whereas upon activation, the stem cells transitioned to become CD34+. As to developmental changes of CD34 expression, stem cells from 5 week old and younger mice are all CD34+; 7 week old mice showed both CD34- and CD34+ stem cells.  The majority of stem cells (~80%) in 10 week old mice are CD34-. Thus, CD34 expression by adult stem cells appears to be a mirror of the activation state of the stem cell.

In the ensuing discussion, it was pointed out that AA4.1 appeared to be another marker associated with activated but not quiescent reconstituting cells. AA4.1 is present on stem cells in murine fetal liver, where stem cells actively cycle, and on marrow stem cells cycling during regeneration following exposure to 5-FU, but not on stem cells in normal marrow. The Working Group agreed that it is important to make the link between the presence or absence of the CD34 marker and global change or biological potential (e.g., trancriptional, phoshphorylation) of processes within the cell. It was also noted that correlation between phenotype and biological potential is present but less than absolute. The Working Group acknowledged that the molecules that are tracked in phenotypic studies are not likely to be closely related to the mechanisms that specify lineage or self-renewal.

Dr. Sharkis gave an excellent review of the current literature and his own work regarding CD34expression in murine transplant models. Transplant studies were conducted primarily with mouse bone marrow cells collected at a flow rate of 25ml/minute by elutriation (FR25 cells) This cell population has been further enriched by selecting cells that are negative for lineage antigen expression (Lin- cells). An intracellular enzyme was used to positively select FR25 cells which were further separated cells into 34- and 34+ cells. The 34+ cells represented approximately 30- 40% of the above FR25 cell population. FR25 34+ but not 34- cells from this population reconstituted animals. CD34 negative and positive cell separation data was presented using a rat anti-mouse CD34 monoclonal and polyclonal antibodies. Transplant results indicated that the frequency of CD34- stem cells is significantly lower than CD34+.

To extend the previous studies and to diminish the concern of CD34 contamination in the cell population, CD34 knockout mice were obtained from individuals in Toronto (Mak) and Genetech (Lasky). Studies were conducted which include stem cell isolation, transplantation, long-term marrow culture and homing. Transplant studies conducted with (Mak) mice (previously reported to have no stem cell phenotype defect) marrow and normal mice marrow indicate no difference in reconstituting capability in a normal recipient. However, transplant studies with the Lasky mice marrow (reported phenotype - significant reduction in progenitor cell number) showed a reduction in reconstituting capability at each dose of cells.

These studies and others have suggested that CD34 may be an important adhesion molecule and important in stem cell homing. CD34- FR25 lin- cells (Mak mice) were isolated, PKH labeled, and injected into irradiated wild type mice. Forty-eight hours later marrow and spleen cells were harvested and analyzed for the frequency of PKH-positive cells. Using PKH bright cells from normal mice transplanted into secondary recipients, 80% of mice (using 100 cells) are protected with cells from the marrow whereas cells recovered from spleen fail at 100,000 cells to short-term reconstitute lethally irradiated animals. Thus, 2 days following transplant with normal cells it appears that long-term reconstituting cells have a selective advantage for going to the bone marrow compared to the spleen. However, using the homing assay with the Mak mice, the recovery is predominately in spleen cells and not in marrow as seen with normal cells. These data indicate that cells from the knockout mice may have a difference in homing capability. Thus, a primary function of CD34 may be the regulation of the localization and differentiation of hematopoietic stem cells within the hematopoietic microenvironment.

Dr. Lemischka presented the utility of his Stem Cell Database (SCDb), an interactive World Wide Web resource designed to provide insights into the molecular nature of stem cells and their microenvironment. An important objective is to precisely correlate global gene expression profiles with quantitatively measured biological properties such as multipotentiality, self-renewal ability, and the potential for in vivo engraftment. An important goal of these studies is to construct a comprehensive genetic program primarily for long-term repopulating cells and secondarily for various, putatively distinct stages of the hematopoietic developmental hierarchy. The characterization of such comprehensive genetic programs will facilitate the elucidation of potentially complex regulatory pathways and networks. The SCDb (a joint project between the Lemischka laboratory and the laboratory of Dr. G. Christian Overton, the director of the Center for Computational Biology and Informatics at Penn) contains the results of numerous global computational analyses which in many cases have assigned a protein family membership and/or a candidate biological function to the novel gene-products. Most of the analyzed gene-products in SCDb are from the mouse. There are also gene-products from a smaller human project in which a high quality, representative cDNA library from bone marrow CD34+Lin-CD38- cells was subtractively depleted of sequences expressed in the closely related but more committed CD34+Lin-CD38+ cell population. Collectively, SCDb has approximately 34,000 individual sequences. The sequence data have been analyzed using sophisticated bioinformatics.  omparisons to the public databases and peptide-motif analyses have identified numerous known and novel gene-products with structural features that are consistent with regulatory functions. These include transcription factors, cell surface molecules, and intracellular signaling proteins. These molecules have been identified by performing subtractive hybridization, high-throughput sequencing, sophisticated bioinformatic analyses, and high-density array hybridization.

The SCDb will eventually be made available to the public. It will contain information from functional assays, sequences, transplant competitive repopulation results down to molecular biology (tutorials). The Working Group was pleased to hear that the database would be made public and felt strongly that this would significantly facilitate and encourage new work. It is anticipated that other investigators will take advantage of SCDb to deposit and analyze sequences from their own molecular efforts. In this way, the SCDb will become a central resource for the entire stem cell research community.

Discussion arose regarding the subtraction hybridization strategy used in the above studies. Fractions that were being subtracted were not homogeneous and early stem cells might constitute only a small percentage of the fractions. There was uncertainty expressed whether stem cell-specific genes would be accessible by subtraction if stem cells were only a small subset. Dr. Iscove described how this issue drove the conception of his approach, in which cDNA is obtained from single cells whose exact biological potential is known. The approach resolves transcript expression to specific stages and is able to discriminate transcripts expressed specifically in very early stages. Prospective work is directed at capturing cDNA from long-term reconstituting cells directly confirmed by in vivo assay of sibling cells. The work has identified transcriptional "previewing" of lineage maturation transcripts in multipotent cells. Previewed transcripts are subsequently down-regulated downstream and then eventually up-regulated again as typical transcripts in maturing cells (Ig, Ly6A/Sca-1, lysozyme). He is attempting to define the breadth of the previewed repertoire. Significantly, only hematopoietic transcripts have been detected in multipotent cells, while transcripts that characterize a spectrum of non-hematopoietic lineages including neural, connective and epithelial tissues have not thus far been detected.

Currently, the issue of stem cell homogeneity is a complication in all enrichment strategies, and is very difficult to rigorously address; given that in vivo transplantation is the only reliable assay for the most primitive stem cell population.

Dr. Mulligan summarized several of his lab's recent studies which have focused on further characterizing the surface phenotype of murine bone marrow-derived cell populations capable of the reconstitution of lethally irradiated recipients, and determining the potential of bone marrow and organ tissue or organ-derived populations to give rise to specific non-hematopoietic cell types after bone marrow transplantation (BMT). He first reviewed studies which indicate that, using a FACS-based procedure involving dual-wavelength analysis of Hoechst-33342 stained bone marrow or muscle cells, it is possible to isolate 'stem cell' populations (termed SP cells) which are capable of giving rise to both hematopoietic cells and muscle cells after BMT in a mouse model of muscular dystrophy (mdx mice). Although the proportion of muscle fibers showing evidence of donor cell derived engraftment was variable, in one animal approximately 10% of all muscle fibers showed contribution from donor cells. Interestingly, in the case of animals engrafted with bone marrow SP cells, it was not possible to detect the presence of donor derived 'satellite' cells, the stem cell population in muscle utilized for the normal regeneration of muscle, while in the case of animals engrafted with muscle SP cells, there appeared to be some engraftment of this 'compartment'. Dr. Mulligan indicated that these results may suggest that the engraftment of muscle via SP cells may involve a novel mechanism distinct from that utilized by satellite cells.

Dr. Mulligan also presented data that bone marrow derived cells can give rise to endothelial cells after BMT.  Interestingly, although in unperturbed BMT recipients, no evidence for endothelial contribution was observed, donor derived endothelial cells were readily detected in BMT recipients which subsequently received organ transplants, or in recipients in which a myocardial infarction was induced. These results suggest that the ability of transplanted stem cells to give rise to non-hematopoietic stem cells may be dependent upon local environmental stimuli, induced by organ or tissue injury.

Dr. Mulligan then reviewed the results from his laboratory which suggest that, using the FACS methods his laboratory had used to purify bone marrow hematopoietic stem cells, it may be also be possible to isolate stem cell populations from a variety of adult tissues, including muscle, brain, heart, kidney, and liver. Specifically, experiments with muscle show that muscle SP cells exist, and give rise to both hematopoietic and muscle cells after BMT.  Preliminary experiments with CNS derived SP cells suggest that they are enriched for engraftment of the brain. Dr. Mulligan indicated that current efforts are being directed towards further evaluating the developmental potentials of both bone marrow and other putative adult stem cell populations, and the biological circumstances necessary to reveal those potentials. Dr. Mulligan summarized his studies by suggesting that a variety of adult stem cell populations may exist and have common functional properties, and that the ability of the different stem cell populations to give rise to particular cell types may be dependent upon local environmental stimuli. Accordingly, it may be critically important to understand the nature of those biological cues for differentiation, and how to manipulate the cells and the host in order to facilitate production of desired cell types after stem cell transplantation. Lastly, Dr. Mulligan presented data which suggests that only CD34- SP cells are capable of the complete and long term reconstitution of lethally irradiated murine recipients. These results was discussed in detail, in light of the results presented by Drs Ogawa and Sharkis, and the possibility that different transplantation models may reveal the ability of different cell types to provide for hematopoietic cell engraftment was raised.

Dr. Verfaillie described the isolation, cultivation, and differentiation of adult bone marrow derived mesodermal stem cells (MSC). Mesodermal progenitor cells were isolated by depletion of CD45 and GlyA positive cells from mononuclear bone marrow cells. The differentiation of MSC into three types of muscle (skeletal, heart, and smooth muscle) was discussed. Regarding cultivation of MSC, after plating approximately 5000 cells, single cells grow out in about every third well. It is almost 3 weeks before anything can be seen growing. Currently, the phenotype of the cells is unknown and will have to wait until enough cells can be isolated. The MSC cells are essentially negative for most markers except cytokine and adhesion receptors. If cultured "correctly," (very low density/confluency) cells do not change phenotype after numerous cell divisions. It takes 14-21 days get them going and they divide every 40 hours. If the cells become confluent, they stop growing, start to differentiate, and express CD44, HLA antigens and other markers. The cells have been taken out to 65-70 doubles. Telomere length (in cells from adults and children) does not change much before and after culture and no major cytogenetic abnormalities have been seen.

The frequency of the MSC with massive multi-potentially is approximately 1 in 10,000,000 whole bone marrow mononuclear cells. Cells are plated with limited dilution. One cell per well does not grow well at all and it maybe that cells need to talk to each other. It is not clear at this time whether a single cell can generate multiple cell types.  Specific culture conditions (developed by trial and error) to differentiate MSC into endothelial cells, adipocytes, skeletal muscle, and neural cells were discussed. These cells also differentiate into osteoblasts and chondroblasts.

Dr. Verfaillie has attempted to coax the undifferentiated MSC to hematopoiesis by changing culture conditions (using a combination of hematopoietic cytokines) and culture the cells on the AFT024 feeder layer. No blood cells have been generated at this time although the cells that emerged were GATA-2 and c-kit positive. Three types of CNS cells were generated in this culture system; astrocytes, oligodendrocytes, and neurons. Testing is now ongoing using clonal analysis of retrovirally marked MSC to demonstrate that undifferentiated MSC generate a neural stem cell that then gives rise to the three types of CNS cells. These cells have been assessed in an infarcted rat model.  Undifferentiated MSC can differentiate in situ into cells that express early and more mature neuronal marker including astrocyte and oligodendritic markers. After implantation, the infarcted rat model appears to be functionally better. Characterizing the molecular mechanisms underlying neural stem cell differentiation will be extremely important for the development of therapies for traumatic or degenerative neural defects.

The challenge of doing these experiments on single cell level was discussed in detail. Being able to clone these cells and recreate in vitro results in vivo are among the next major challenges facing this research.

Dr. Iscove's presentation outlined a framework in which findings that have been interpreted as "plasticity" could be explainable alternatively in terms of our current state of knowledge. The long prevailing view was that stem cells resident within each tissue (e.g. hematopoietic, epithelial, liver, skeletal muscle, neural, testis) were restricted in their repertoire to maintaining only that kind of tissue. Recent findings have been interpreted as challenging that view.  His comments focused on repertoire, site, and origin of individual nuclei in multinucleate cells. Regarding repertoire, the definitive way to determine the potential of a cell is to address it "clonally." The nature of a cell can not be established definitively simply on the basis of origin from a particular location, a defined population or possession of a particular phenotype. Proof of "plasticity" is likely to require some attempt (cloning, limiting dilution, insertion marking) to ask "what can one cell do?" Clonal demonstrations have been lacking in most of the recent findings suggestive of plasticity. Location of a cell may not be a reliable indicator of its origin. Marrow stem cells can circulate and we know how to enhance their frequency of doing so. Throughout the body a large number of sites are potentially hematopoietic. For example, 0.5-1% of the normal thymus is hematopoietic, and most of the organs including skeletal muscle can become infiltrated with actively hematopoietic cells in various circumstances. These observations suggest that hematopoietic stem cells may be widely distributed in non-medullary tissues, although normally quiescent.

With interest in potential plasticity recently kindled, some non-hematopoietic tissues are being measured for hematopoietic potential perhaps for the first time. Like the bone marrow, other systems may also occasionally export stem cells to the circulation and to diverse tissues. Reservoirs throughout the body could contain mixed populations of stem cells individually restricted to their own kind of repertoire, but mixed together in heterotopic sites and possible sharing very similar phenotypic features. Additionally, macrophages of marrow origin extensively populate most tissues including the musculature. Experiments designed to probe hematopoietic cells for potential to differentiate into non-hematopoietic lineages will have to prove quite conclusively that cells, or even fused nuclei, within a non-hematopoietic tissue do not simply represent or derive from differentiated hematopoietic bystanders. Clearly such alternative explanations would have to be ruled out in future studies, ideally by clonal methods, for the concept of "stem cell plasticity" to acquire a firmer footing.

Dr. Darwin Prockop was unable to present his data due to time constraints but called our attention to a PNAS article that would shortly be published. I have attached the reference and abstract of the article.

AUTHORS: Colter DC; Class R; DiGirolamo CM; Prockop DJ; Center for Gene Therapy, MCP Hahnemann University, 10118 New College Building, 245 North 15 Street, Philadelphia, PA 19102-1192, USA. Proc Natl Acad Sci U S A 2000 Mar 28;97(7):3213-8.

ABSTRACT: Cultures of plastic-adherent cells from bone marrow have attracted interest because of their ability to support growth of hematopoietic stem cells, their multipotentiality for differentiation, and their possible use for cell and gene therapy. Here we found that the cells grew most rapidly when they were initially plated at low densities (1.5 or 3.0 cells/cm(2)) to generate single-cell derived colonies. The cultures displayed a lag phase of about 5 days, a log phase of rapid growth of about 5 days, and then a stationary phase. FACS analysis demonstrated that stationary cultures contained a major population of large and moderately granular cells and a minor population of small and agranular cells here referred to as recycling stem cells or RS-1 cells. During the lag phase, the RS-1 cells gave rise to a new population of small and densely granular cells (RS-2 cells). During the late log phase, the RS-2 cells decreased in number and regenerated the pool of RS-1 cells found in stationary cultures. In repeated passages in which the cells were plated at low density, they were amplified about 10(9)-fold in 6 wk. The cells retained their ability to generate single-cell derived colonies and therefore apparently retained their  ultipotentiality for differentiation.

Recommendations

The Working Group discussed a number of important research needs and several areas of focus are identified as follows: (1) substantiate the stem cell plasticity concept (e.g., defining the repertoire of stem cells in various anatomical locations); (2) determine whether cells with hematopoietic potential can express other potentials;  (3) whether cells with other potentials (muscle, CNS, etc.,) in other sites of the body also have hematopoietic potentials; (4) definition of relevant cell populations and develop approaches to address the issue of clonality; and  (5) develop clinically relevant models to make use of the concept of stem cell 'plasticity.'

The Working Group acknowledged that much work is needed to substantiate the stem cell 'plasticity' concept which clearly has important therapeutic implications. Collaboration among research groups was identified as the key element in our ability to better substantiate this concept and to characterize the biological mechanisms that might drive the process. Thus, the Working Group strongly recommended that research in this area should be supported by using a collaborative/interactive research grant mechanism which allows investigators from various locations to share information and resources. They strongly emphasized a 'no walls concept' was needed to foster efforts in this important research area.