Stem cell self-renewal: lessons from bone marrow, gut and iPS toward clinical applications.

The hematopoietic stem cell (HSC) is the prototype organ-regenerating stem cell (SC), and by far the most studied type of SC in the body. Currently, HSC-based therapy is the only routinely used SC therapy; however, advances in the field of embryonic SCs and induced pluripotent SCs may change this situation. Interest into in vitro generation of HSCs, including signals for HSC expansion and differentiation from these more primitive SCs, as well as advances in other organ-specific SCs, in particular the intestine, provide promising new applications for SC therapies. Here, we review the basic principles of different SC systems, and on the basis of the experience with HSC-based SC therapy, provide recommendations for clinical application of emerging SC technologies.

Toward stem cell systems biology: from molecules to networks and landscapes.

The last few years have seen significant advances in our understanding of the molecular mechanisms of stem-cell-fate specification. New and emerging high-throughput techniques, as well as increasingly accurate loss-of-function perturbation techniques, are allowing us to dissect the interplay among genetic, epigenetic, proteomic, and signaling mechanisms in stem-cell-fate determination with ever-increasing fidelity (Boyer et al. 2005, 2006; Ivanova et al. 2006; Loh et al. 2006; Cole et al. 2008; Jiang et al. 2008; Johnson et al. 2008; Kim et al. 2008; Liu et al. 2008; Marson et al. 2008; Mathur et al. 2008). Taken together, recent reports using these new techniques demonstrate that stem-cell-fate specification is an extremely complex process, regulated by multiple mutually interacting molecular mechanisms involving multiple regulatory feedback loops. Given this complexity and the sensitive dependence of stem cell differentiation on signaling cues from the extracellular environment, how are we best to develop a coherent quantitative understanding of stem cell fate at the systems level? One approach that we and other researchers have begun to investigate is the application of techniques derived in the computational disciplines (mathematics, physics, computer science, etc.) to problems in stem cell biology. Here, we briefly sketch a few pertinent results from the literature in this area and discuss future potential applications of computational techniques to stem cell systems biology.

The AFT024 stromal cell line supports long-term ex vivo maintenance of engrafting multipotent human hematopoietic progenitors.

The immortalized murine stromal cell line AFT024 has been reported to maintain human hematopoietic progenitors in an undifferentiated state in vitro. In the current studies the beige/nude/xid (bnx) mouse in vivo xenograft model was used to examine the engraftment and multilineage generative potential of human hematopoietic progenitors after 2-3 weeks growth on AFT024 stroma, in comparison to primary stromal monolayers derived from post-natal human bone marrow. Eight to 12 months after transplantation of human CD34+CD38- cells from umbilical cord blood, cultured on AFT024 vs human stroma for 2-3 weeks, the murine bone marrow was harvested and analyzed for the presence of human myeloid and lymphoid cells. The mean percent engraftment of total human hematopoietic cells in the murine marrow was significantly higher after co-cultivation on AFT024 than on human stroma. Human myeloid and lymphoid lineage cells were detected in all mice. However, engraftment of myeloid lineage cells (CD33+), B lymphoid (CD19+), and T lymphoid cells (CD4+and CD8+) were significantly higher after co-cultivation of the human cells on AFT024 than on human stroma, prior to transplantation. Interestingly, the length of time in culture did not significantly affect the engraftment of the myeloid and T lymphoid lineage progenitors, but the percentage of B lymphoid lineage engraftment decreased significantly between 2 and 3 weeks of co-cultivation on both types of stroma. Cells with a primitive phenotype (CD45+/CD34-/CD38- and CD45+/CD34-/lin-) and cells with the capacity to generate secondary human CFU after recovery from the bnx bone marrow were maintained at significantly higher levels during culture on AFT024 stroma than on human stroma. The current studies demonstrate that the AFT024 murine stromal cell line supports the ex vivo survival and maintenance of human hematopoietic progenitors that are capable of long-term multilineage reconstitution for 2-3 weeks ex vivo, to levels superior to those that can be obtained using human stromal cells.

Stem cell dogmas in the genomics era.

Recently, much excitement has been generated by strong suggestions that stem cells isolated from diverse somatic tissues may have a previously unsuspected degree of developmental or differentiation plasticity. For example, a hematopoietic stem cell may be capable of producing mature liver cells, muscle tissue or even neurons. Similarly, central nervous system stem cells or muscle stem cells may be capable of producing mature blood cell populations. These observations have called into question several fundamental dogmas of developmental biology. In addition, these observations offer extraordinary promise in the clinical setting. It is of paramount importance to rigorously assess the suggested plasticity phenomena using precise clonal analysis. In order to explore the plasticity phenomena in more direct ways, it is necessary to develop in vitro systems where such behavior can be recapitulated in a well-defined setting. Finally, stem cell plasticity will be governed, at least in part, by cell-autonomous mechanisms: that is, those mediated by the panel of gene products expressed in stem cells. Therefore, it is necessary to identify the complete gene expression profile that defines the stem cell.

Umbilical cord blood cells capable of engrafting in primary, secondary, and tertiary xenogeneic hosts are preserved after ex vivo culture in a noncontact system.

This report describes stroma-based and stroma-free cultures that maintain long-term engrafting hematopoietic cells for at least 14 days ex vivo. Umbilical cord blood (UCB) CD34(+) cells were cultured in transwells above AFT024 feeders with fetal-liver-tyrosine-kinase (FL) + stem cell factor (SCF) + interleukin 7 (IL-7), or FL + thrombopoietin (Tpo). CD34(+) progeny were transplanted into nonobese diabetic-severe combined immunodeficiency (NOD-SCID) mice or preimmune fetal sheep. SCID repopulating cells (SRC) with multilineage differentiation potential were maintained in FL-SCF-IL-7 or FL-Tpo containing cultures for up to 28 days. Marrow from mice highly engrafted with uncultured or expanded cells induced multilineage human hematopoiesis in 50% of secondary but not tertiary recipients. Day 7 expanded cells engrafted primary, secondary, and tertiary fetal sheep. Day 14 expanded cells, although engrafting primary and to a lesser degree secondary fetal sheep, failed to engraft tertiary recipients. SRC that can be transferred to secondary recipients were maintained for at least 14 days in medium containing glycosaminoglycans and cytokines found in stromal supernatants. This is the first demonstration that ex vivo culture in stroma-noncontact and stroma-free cultures maintains "long-term" engrafting cells, defined by their capacity to engraft secondary or tertiary hosts. (Blood. 2001;97:3441-3449)

Complementation of Myc-dependent cell proliferation by cDNA expression library screening.

The targeted knockout of the c-myc gene from rat fibroblasts leads to a stable defect in cell proliferation. We used complex cDNA libraries expressed from retroviral vectors and an efficient sorting procedure to rapidly select for cDNAs that can restore the growth rate of c-myc deficient cells. All of the biologically active cDNAs contained either c-myc or N-myc, suggesting that no other cellular genes can effectively bypass the requirement for c-myc in fibroblast proliferation. This approach provides a powerful screening method for cell cycle changes in genetically defined systems.

The genetic program of hematopoietic stem cells.

Blood cell production originates from a rare population of multipotent, self-renewing stem cells. A genome-wide gene expression analysis was performed in order to define regulatory pathways in stem cells as well as their global genetic program. Subtracted complementary DNA libraries from highly purified murine fetal liver stem cells were analyzed with bioinformatic and array hybridization strategies. A large percentage of the several thousand gene products that have been characterized correspond to previously undescribed molecules with properties suggestive of regulatory functions. The complete data, available in a biological process-oriented database, represent the molecular phenotype of the hematopoietic stem cell.

Expression of CD27 on murine hematopoietic stem and progenitor cells.

Hematopoietic stem cells (HSC) are defined by self-renewal and multilineage differentiation potentials. In order to uncover the genetic program of HSC, we utilized high-density arrays to compare gene expression in highly purified mouse HSC and their mature progeny. One molecule specifically expressed in immature cells is CD27, a member of the TNF receptor family previously shown to play roles in lymphoid proliferation, differentiation, and apoptosis. We show here that the CD27 protein is expressed by about 90% of cells in a purified HSC population. Interestingly, the CD27pos cells are enriched for cells with short-term hematopoietic activities (colony forming potential in vivo and in vitro), while the minority CD27neg population is more effective in clonal long-term transplantation.

The molecular characterization of the fetal stem cell marker AA4.

We have identified and characterized the stem cell antigen AA4. This molecule is a type I transmembrane protein whose overall structure suggests a role in cell adhesion. During fetal ontogeny (days 9-14 of development), AA4 is expressed in three major cell types: vascular endothelial cells, aorta-associated hematopoietic clusters, and primitive fetal liver hematopoietic progenitors. In the adult, AA4 is abundant in lung, heart, and whole bone marrow. In the adult hematopoietic compartment, aa4 transcripts are present in bone marrow CD34(-/lo) Lin- Sca-1+ c-Kit+ and CD34hi Lin- Sca-1+ c-Kit+ stem and progenitor cell subsets. Our observations suggest that AA4 plays a role in cell-cell interactions during hematopoietic and vascular development.

Searching for stem cell regulatory molecules. Some general thoughts and possible approaches.

Hematopoietic development in the mammal can be represented as a numerically expanding hierarchy of cell populations that are progressively restricted in their self-renewal and differentiation abilities. Classical functional studies have now been extended to provide exact physical descriptions of various stages in the hematopoietic hierarchy. In particular, much information is available that defines the properties of the most primitive stem cell compartment. In addition, a number of in vitro culture systems suggest the possibility of maintaining and expanding these cells in a defined context. In all developmental systems, unique profiles of expressed genes define distinct differentiation stages. Within these profiles are gene products that play crucial roles in the regulation of cell-fate decisions. Recent progress in hematopoietic biology provides the framework within which to define molecular phenotypes for hematopoietic stem cells and their immediate clonal progeny. Identifying novel gene products expressed predominantly in uncommitted stem cells together with functional loss and gain-of-function approaches should begin to unravel the molecular mechanisms that govern biological phenomena such as self-renewal, commitment, and proliferation in the hematopoietic system.

The myeloid-lymphoid initiating cell (ML-IC) assay assesses the fate of multipotent human progenitors in vitro.

Hematopoietic stem cells (HSC) are cells with self-renewing multilineage differentiation potential. Although engraftment in xenogeneic recipients can be used to measure human HSC, these assays do not allow assessment of individual progenitors. We developed an in vitro assay that allows the identification of a single human bone marrow progenitor closely related to HSC, which we termed "Myeloid-Lymphoid Initiating Cell," or ML-IC, because it is capable of generating multiple secondary progenitors that can reinitiate long-term myeloid and lymphoid hematopoiesis in vitro. The assay is done in contact with murine AFT024 fetal liver stromal cells and with Flt3-Ligand, stem cell factor, and interleukin-7. In this assay, 0.2% to 1.7% of Lin -/34(+)/DRdim cells could generate 1 to 3 long-term culture initiating cells (LTC-IC) as well as 1 to 4 NK-IC after 4 to 6 weeks. In addition, this assay measures contribution of net-progenitor conservation and net-progenitor proliferation over time, providing insight in the fate of individual LTC-IC and NK-IC. This assay will prove useful to enumerate the number of very primitive human progenitors with multilineage differentiation potential, as well as to evaluate future ex vivo culture conditions.

The type of stromal feeder used in limiting dilution assays influences frequency and maintenance assessment of human long-term culture initiating cells.

The goal of this study was to evaluate if differences in culture conditions used in long-term culture assays affect enumeration of LTC-IC in freshly sorted or ex vivo expanded CD34+/HLA-DRdim/CD2-/CD7- (34+/Lin-) cells. The variables examined included different stromal feeders (murine bone marrow fibroblast cell line, M2-10B4 and murine fetal liver cell line, AFT024) and presence or absence of cytokines (MIP-1alpha + IL-3). The absolute LTC-IC frequency in 34+/Lin- cells measured in limiting dilution assays (LDA) on AFT024 (4.45 +/- 0.69%) was significantly higher than in M2-10B4 (1.45 +/- 0.20%) LDA. Addition of MIP-1alpha and IL-3 to AFT024 LDA increased the measured LTC-IC frequency to 6.8 +/- 0.9%. We also determined the fraction of LTC-IC that persisted after 34+/Lin cells were cultured for 5 weeks by replating progeny in the three LDA readout systems. The measured LTC-IC maintenance was significantly lower when M2-10B4 LDA (13.1 +/- 3.5%, P < 0.05) were used compared with AFT024 LDA (36.6 +/- 5.5%) or AFT024 LDA supplemented with MIP-1alpha and IL-3 (29.1 +/- 6.3%). Thus, the number of LTC-IC measured in freshly sorted 34+ cells depends on the stromal feeder used in LDA assays. Furthermore, and most important, assessment of LTC-IC expansion or maintenance may vary significantly depending on the type of stromal feeder used to enumerate LTC-IC.

Single adult human CD34(+)/Lin-/CD38(-) progenitors give rise to natural killer cells, B-lineage cells, dendritic cells, and myeloid cells.

Marrow stromal cultures support adult CD34(+)/Lin-/HLA-DR- or CD34(+)/Lin-/CD38(-) cell differentiation into natural killer (NK) or myeloid cells, but unlike committed lymphoid progenitors (CD34(+)/Lin-/CD45RA+/CD10(+)), no B cells are generated. We tested whether different microenvironments could establish a developmental link between the NK and B-cell lineages. Progenitors were cultured in limiting dilutions with interleukin-7 (IL-7), flt3 ligand (FL), c-kit ligand (KL), IL-3, IL-2, and AFT024, a murine fetal liver line, which supports culture of transplantable murine stem cells. NK cells, CD10(+)/CD19(+) B-lineage cells and dendritic cells (DC) developed from the same starting population and IL-7, FL, and KL were required in this process. Single cell deposition of 3,872 CD34(+)/Lin-/CD38(-) cells onto AFT024 with IL-7, FL, KL, IL-2, and IL-3 showed that a one time addition of IL-3 at culture initiation was essential for multilineage differentiation from single cells. Single and double lineage progeny were frequently detected, but more importantly, 2% of single cells could give rise to at least three lineages (NK cells, B-lineage cells, and DC or myeloid cells) providing direct evidence that NK and B-lineage differentiation derive from a common lymphomyeloid hematopoietic progenitor under the same conditions. This study provides new insights into the role of the microenvironment niche, which governs the earliest events in lymphoid development.

The murine stromal cell line AFT024 acts specifically on human CD34+CD38- progenitors to maintain primitive function and immunophenotype in vitro.

A stromal cell line derived from murine fetal liver (AFT024) has been demonstrated to maintain long-term repopulating murine stem cells for up to 7 weeks in vitro. We evaluated the ability of AFT024 to maintain the immunophenotype and function of primitive human progenitors in vitro by comparing the cocultivation of CD34+CD38 cells on AFT024 with that on primary human stroma (HS). We have previously reported that within the CD34+CD38- population of bone marrow and cord blood, a highly primitive progenitor subpopulation can be identified functionally by its ability to generate colony forming unit-cells (CFU-Cs) in extended long-term culture (ELTC), that is, beyond 60 days of stromal cocultivation. Cocultivation of bone marrow and cord blood CD34+CD38-cells on AFT024 produced significantly greater cell expansion (p=0.0002) and CFU-C output (p=0.0007) during the ELTC period compared with culturing on HS. CFU-C production continued up to 9 weeks longer on AFT024 stroma. After 3 to 4 weeks of bulk culture on either AFT024 or HS, cells were replated in a limiting dilution to measure the number of cobblestone area-forming cells (CAFCs) maintained on each stroma. AFT024 maintained significantly more CAFCs than did HS (n=3, p=0.002). Fluorescence-activated cell sorter analysis of AFT024 and HS cocultures showed that both the frequency (p=0.018) and absolute number (p=0.027) of CD34+CD38- cells were significantly higher in cultures on AFT024 than in those on HS (n=9). The effects of AFT024 on preservation of primitive progenitors were not seen in transwell (noncontact) cultures. Thus, AFT024 acts by direct contact to maintain the phenotype and function of the most primitive and quiescent human progenitors currently identifiable by in vitro assays.

Maturation-dependent gene expression in a conditionally transformed liver progenitor cell line.

We have isolated a conditionally transformed liver progenitor cell line with phenotypic similarities to both hepatoblasts (bipotent embryonic liver cells that give rise to hepatocytes and intrahepatic biliary epithelial cells) and liver epithelial cells (primitive hepatic cells isolated from adult livers capable of generating both hepatocytic and biliary lineages). Cell line L2039 was derived from E14 fetal mouse liver after transformation with temperature-sensitive SV-40 large T antigen. At 33 degrees C, these cells have an epithelial morphology with a high nucleocytoplasmic ratio and express both hepatocytic and biliary genes, including albumin, alpha-fetoprotein, glutamine synthetase, insulinlike growth factor II receptor, fibronectin and laminin, and cytokeratins 8 and 19, a set of markers characteristic for hepatoblasts. The presence of cytokeratin 14, vimentin, and several oval-cell antigens link cell line L2039 to nonparenchymal liver epithelial cell populations thought to contain progenitor cells. Serum-free, hormonally defined media conditions and extracellular matrix requirements were determined for growth and differentiation of this cell line. During culture on type IV collagen at 39 degrees C, L2039 cells cease dividing and demonstrate hepatocytic differentiation with the assumption of a hepatocytelike morphology and glucocorticoid-dependent regulation of liver-specific genes, including albumin, alpha-fetoprotein, phosphoenolpyruvate carboxykinase, and liver-enriched transcription factors. The number of albumin-positive cells increases during culture at 39 degrees C, indicating that L2039 cells convert from a prehepatocytic to a hepatocytic phenotype. Under conditions specific for hepatocytic differentiation, C/EBPs were expressed and differentially regulated, with C/EBPbeta and C/EBPdelta upregulated early and C/EBPalpha only slightly expressed after 7 d, indicating that C/EBPalpha may not be a crucial factor in commitment to the hepatocytic phenotype.

In vitro maintenance of highly purified, transplantable hematopoietic stem cells.

The cellular and molecular mechanisms that regulate the most primitive hematopoietic stem cell are not well understood. We have undertaken a systematic dissection of the complex hematopoietic microenvironment to define some of these mechanisms. An extensive panel of immortalized stromal cell lines from murine fetal liver were established and characterized. Collectively, these cell lines display extensive heterogeneity in their in vitro hematopoietic supportive capacity. In the current studies, we describe a long-term in vitro culture system using a single stromal cell clone (AFT024) that qualitatively and quantitatively supports transplantable stem cell activity present in highly purified populations. We show multilineage reconstitution in mice that received the equivalent of as few as 100 purified bone marrow and fetal liver stem cells cultured for 4 to 7 weeks on AFT024. The cultured stem cells meet all functional criteria currently ascribed to the most primitive stem cell population. The levels of stem cell activity present after 5 weeks of coculture with AFT024 far exceed those present in short-term cytokine-supported cultures. In addition, maintenance of input levels of transplantable stem cell activity is accompanied by expansion of other classes of stem/progenitor cells. This suggests that the stem/progenitor cell population is actively proliferating in culture and that the AFT024 cell line provides a milieu that stimulates progenitor cell proliferation while maintaining in vivo repopulating activity.

Hematopoietic activity of a stromal cell transmembrane protein containing epidermal growth factor-like repeat motifs.

Primitive hematopoietic stem cells are closely associated with discrete in vivo microenvironments. These "niches" are thought to provide the molecular signals that mediate stem cell differentiation and self-renewal. We have dissected the fetal liver microenvironment into distinct cellular components by establishing an extensive panel of stromal cell lines. One particular cell line maintains repopulating stem cells for prolonged in vitro culture periods. A subtraction cloning strategy has yielded a cDNA that encodes a cell surface glycoprotein with a restricted pattern of expression among stromal cell lines. This molecule, previously identified as delta-like/preadipocyte factor-1, contains epidermal growth factor-like repeats that are related to those in the notch/delta/serrate family of proteins. We have investigated the potential role of this molecule in hematopoietic stem/progenitor cell regulation. We show that the delta-like protein displays activity on purified stem cells by promoting the formation of "cobblestone areas" of proliferation. These cobblestone areas contain both primitive high-proliferative potential progenitors and in vivo repopulating stem cells.

Relative to adult marrow, fetal liver repopulates nearly five times more effectively long-term than short-term.

Multilineage precursor cells from 14-day B6 (C57B1/6J) mouse fetal liver and adult bone marrow that repopulate both the lymphoid and myeloid systems were compared by competitive repopulation. Cells were assayed in normally functioning populations, and enrichment, tissue culture, and induced marking were avoided since these manipulations might affect cell function. Fetal or adult donor cells were mixed with marked adult competitor cells and transplanted into irradiated recipients whose blood was tested at short (25-33-day) or long (105-245-day) time periods after transplantation. Proportions of lymphocytes, granulocytes, and platelets descended from donor precursors were measured by GPI (glucosephosphate isomerase) isozyme genetic markers in congenic mice, and represent the repopulating abilities of these precursors relative to the standard competitor. For short-term repopulation 25-33 days after transplantation, fetal and adult donor cells were similar; in three studies, fetal liver contributed 0.8, 1.1, and 1.4 times as much as adult marrow per 10(5) cells transplanted. However, when long-term (105-245-day) repopulation was tested in the same recipients, fetal liver contributed 3.5, 5.0, and 7.1 times as much as adult marrow. Ratios of long-term/short-term repopulating abilities in fetal liver relative to standard adult marrow competitors were 2.5, 8.9, and 4.7, while in marrow controls, these ratios remained approximately one (1.14 and 0.80). Thus, 14-day fetal liver contains several times more long-term repopulating cells relative to short-term repopulating cells than does adult marrow. Ratios of long-term/short-term fetal cells were unchanged by precursor enrichment. The AA4.1+, Ly-6A/E+, lineage low fraction had a ratio of 4.4, although it repopulated 276 times better than unenriched fetal cells whose ratio was 4.7. There are two hypotheses that explain these data most simply: 1) There may be only a single multilineage precursor, but after transplantation cells seed in different microenvironments that support either long-term or short-term function. 2) Conversely, the difference may be at the stem cell level rather than the microenvironmental level, so that there are tow types of stem cells with multilineage differentiating ability, but only one functions over the long-term. The current report defines new conditions required by each hypothesis. If functional life spans are defined by seeding sites, as in hypothesis 1, fetal cells seed much higher proportions of long-term sites than adult cells. If different types of stem cells function short-term and long-term, as in hypothesis 2, they are not distinguished by markers allowing a 276-fold enrichment to 1367 times the repopulating ability of fresh marrow.