A novel insight into aging: are there pluripotent very small embryonic-like stem cells (VSELs) in adult tissues overtime depleted in an Igf-1-dependent manner?

Molecular evidence

PSCs, according to their definition, should, at the molecular level, i) express acknowledged markers of pluripotency (e.g., Oct4 and Nanog), ii) possess bivalent domains in promoters of develop-mentally crucial transcription factors (TFs), and iii) reactivate the X chromosome (in female PSCs), which becomes silenced at early stages of embryogenesis [15-16].

To support the notion that VSELs are pluripotent, our recent research has demonstrated that VSELs, not only express the transcription factor Oct4 (which is a characteristic marker of PSCs) at the mRNA and protein levels, but, as mentioned above, the Oct4 promoter in VSELs is in an active/open state [5]. Furthermore, our data indicate that murine VSELs display bivalent domains in promoters of several homeodomain-containing developmental TFs (ie.g., Sox21, Nkx2.2, Dlx1, Lbx14, Hlx9, and Pax5). The presence of transcriptionally active histone codes, such as H3K4me3, physically coexisting with repressive histone codes, such as H3K27me3, was confirmed by employing the carrier CHIP assay (submitted for publication). Finally, studies on female murine VSELs show that these cells partially re-activate the inactivated X chromosome. Thus, VSELs possess several molecular features expected from PSCs.

In vitro evidence

PSCs, according to their definition, should also differentiate into cells from all three germ layers (endo-, meso- and ectoderm) in vitro [7]. To support the concept of pluripotentiality of VSELs in vitro, we have also successfully developed several co-culture condi-tions where VSELs, if plated over supportive cell lines, may expand/differentiate into a variety of somatic cell types [17]. One of these is a co-culture system over OP9 stromal cells that allows hematopoietic specification of VSELs [18]. Another is a co-culture of VSELs over the myoblastic cell line C2C12, which allows some VSELs to differentiate and form spherical structures that resemble embryoid bodies and stain positively for the fetal isoform of alkaline phosphatase [19]. More importantly, cells isolated from these murine VSEL-derived spheres, if plated into cultures promoting tissue differentiation, expand into cells from all three germ-cell layers [3]. Interestingly, the formation of spheres by VSELs was associated with a young age in mice, and no sphere formation was observed with cells isolated from older mice (>2 years old). Unfortunately, we have so far not been able to reproduce this phenomenon using human VSELs, which suggests the involvement of murine cell-specific factors. In summary, while VSELs do not display all the molecular markers characteristic of PSCs, they are able to differentiate into cells from all three germ layers in vitro.

In vivo evidence

Our recent collaborative work in vivo indicates that VSELs may also be specified in vivo into MSCs [20] and cardiomyocytes [21-22]. Accordingly, in the first study by Taichman et al, VSELs isolated from GFP⁺ mice were implanted into SCID mice and 4 weeks later the formation of bone-like tissues was observed. To further validate that VSELs exhibit true MSC activity (bone formation), stromal cells were harvested from Col2.3ΔTK mice and implanted into SCID mice to generate thymidine kinase-sensitive ossicles. At 1.5 months, the ossicles were injected with 2000 GFP⁺VSELs. At harvest, there was observed to be a co-localization of GFP-expressing cells with an antibody to the osteoblast-specific marker Runx-2, the endothelial marker CD31, and the adipocyte marker PPARγ. Based upon the ability of uncultured VSELs to (i) differentiate in vivo into mutiple mesenchymal lineages and (ii) to generate osseous tissues at low density, Taichman et al. proposed that this population of cells fulfills many of the required characteristics of MSCs [20].

In another collaborative study, BM-derived VSELs freshly isolated from GFP⁺ mice were injected into the hearts of mice that had undergone ischemia/reper- fusion injury [21]. After 35 days of follow-up, VSEL-treated mice exhibited improved global and regional left ventricular (LV) systolic function (by echocardio-graphy) and attenuated myocyte hypertrophy (by histology and echocardiography) in surviving tissue when compared with vehicle- treated controls. Since VSELs isolated from GFP⁺ transgenic mice were employed for transplantation, we could track the fate of injected cells in the myocardium. Although VSEL transplantation resulted in isolated new myocytes and capillaries in the infarct region, their numbers were too small to account for all of the observed benefits [22]. Thus, it is likely that, in this particular tissue injury model, some paracrine effects by transplanted VSELs played an additionally important role and we are now analyzing the profile of growth factors and cytokines secreted by these cells.

However, unlike PSCs from established embryonic stem cell lines, VSELs neither complete blastocyst development after injection into the developing blastocyst nor form teratomas after injection into immunodeficient mice. The explanation for this is, as we have reported, epigenetic modification of crucial imprinted genes, from which some (i.e., Igf2 and RasGRF1) are important for senescence and longevity [5]. This will be discussed in more details in the following paragraph.

Epigenetic changes in imprinted genes related to insulin factor signalin keep VSELs quiescent in adult tissues. It is well known that changes in Ins/Igf signaling have important implications for aging. Accordingly, i) Igf-1 signaling negatively regulates lifespan in worms, flies, and mammals [23] and ii) Igf-1 and insulin level in blood is regulated positively by caloric uptake [24].

What is crucial for our hypothesis and is the main topic of this review is that we have recently identified an important epigenetic mechanism that governs the VSEL quiescent state, preventing them from unleashing proliferation and the spontaneous growth of teratomas [5]. This mechanism is based on the epigenetic changes in selected somatic-imprinted genes (i.e., Igf2-H19, Ig2R, and RasGRF1) that are involved in insulin factor signaling (Igf-1, Igf-2, and Insulin) (Figure 2).

Accordingly, VSELs downregulate expression of Igf2 that, at the embryonic stage of development, is an autocrine growth factor for these cells, and in addition, upregulate expression of Igf2R, which serves as a mole- cular sink for Igf2 [25]. In contrast Igf2 may signal by binding to Igf-1R. Additionally, Ins/Igf signaling is impaired due to erasure of the imprint on RasGRF1, which is a small GTP exchange factor (GEF) for both Igf-1R and the insulin receptor (InsR) [26]. The involvement of these genes in signaling from activated insulin factor receptors is depicted schematically in Figure 3. Thus, as we believe, impaired signaling from Igf-1R and InsR is responsible for keeping these cells quiescent in adult tissues. We hypothesize also that chronic stimulation over time with insulin and Igf-1 could lead to accelerated depletion of these cells. Since the number of VSELs decreases with age in mice and correlates with their senescence, this could explain the aging process at the level of the PSCs that are deposited in adult tissues during development (VSELs) [4]. A decrease in these cells will impair the regenerative function of their progeny - the pool of TCSCs.

In fact, our studies performed on normal young (4-week-old) and old (2-year-old) mice revealed that the number of VSELs and their pluripotentiality decreases during ageing. Accordingly, VSELs from old mice show lower expression of the pluripotentiality master-regulators such as Oct4, Nanog, Sox2, Klf4, and cMyc and, at the molecular level, the Oct4 promoter in VSELs becomes hypermethylated with age and shows a closed chromatin structure. Furthermore, VSELs from old mice show the somatic type of methylation at both Igf2-H19 and RasGRF1 loci, which suggests that VSELs from old mice have increased sensitivity to insulin factor signaling. This suggests that chronic Ins/Igf signaling in VSELs may contribute to, age-related depletion of these cells.

Lessons from Laron dwarf mice

To test the hypothesis that chronic Ins/Igf signaling may accelerate depletion of the pool of VSELs in adult tissues, we became interested in Laron dwarf mice. Due to a deficiency of growth hormone (GH) receptor, these animals display a severe reduction in the Igf-1 plasma level and do not display increase in GH-mediated Igf-1 plasma level in response to caloric uptake [27]. Interestingly, these animals live 30-40% longer than their normal littermates [28]. Several possible explanations for this effect have been proposed, but we became interested in whether a reduced level of Igf-1 in the plasma of these animals may directly impact the survival of VSELs in adult murine tissues.

Thus, to explain better the role of Igf-1 signaling in VSELs, we measured the number of VSELs in BM of murine Laron dwarfs by FACS. Interestingly, we noticed that the number of VSELs in the BM of plasma Igf-1-deficient Laron dwarfs is maintained at a 3-4-fold higher level than normal wild type littermates during aging. Our molecular analysis studies have additionally demonstrated that the Oct4 promoter in these animals shows also a higher level of demethylation [29].

Furthermore, while analysis of PB cell counts did not reveal any differences in the number of erythrocytes, platelets, and leucocytes between Laron dwarf mice and wild type controls, we observed that Laron dwarf mice have in BM i) a ~4-5-fold increase in the number of Sca-1⁺c-kit⁺lineage^- (SKL) hematopoietic stem cells and ii) a >4-fold higher number of clonogenic CFU-Mix, CFU-GM, BFU-E, and CFU-Meg cells.

Since the plasma Igf-1 level is regulated in mice by caloric uptake and in humans by calorie uptake if in particular protein content in the diet is low [30], these data shed new light on caloric restriction, senescence, and the size of hematopoietic stem cell compartment. Based on this, we propose a new paradigm in which chronic Igf-1 deficiency somehow protects VSELs from age-related elimination from BM. This mimics a situation seen in chronic caloric restriction in mice, where the Igf-1 level is low and results in longevity. Since the long-living Laron dwarf mice that maintain low levels of Igf-1 have higher numbers of VSELs and HSCs in BM, we postulate that chronically elevated levels of Igf-1, resulting, for example, from high caloric uptake, may lead to premature depletion of the stem cell pool, including VSELs and HSCs, and thus be responsible for premature aging in mice. This hypothesis is currently being tested in mice that overexpress GH and thus have high plasma Igf-1 level, and, interestingly, in contrast to Igf-1-deficient Laron dwarf mice, show a much shorter life span.

Further studies are also needed to link the effect of chronic high Igf-1 signaling in VSELs with the development of cancer. Of note, both murine Laron dwarfs [28] and human Laron dwarfs [31] with chronically low Igf-1 levels are significantly protected from developing cancer [28,31]. Since many human malignancies are activated by Igf -1 and Igf-2 signaling (e.g., due to loss of heterozygosity or loss of imprinting at the Igf-2-H19 locus) [32], we hypothesize that excessive activation of VSELs in an Ins/Igf-dependent manner could promote malignant transformation of these cells. However, the hypothetical reverse effect of low Ins/Igf signaling on a pool of VSELs that could become cancer initiating cells [33] requires further experimental study.