PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of hhmipaAbout Author manuscriptsSubmit a manuscriptHHMI Howard Hughes Medical Institute; Author Manuscript; Accepted for publication in peer reviewed journal
 
Cell. Author manuscript; available in PMC 2009 April 17.
Published in final edited form as:
PMCID: PMC2582221
HHMIMSID: HHMIMS75293

Hmga2 promotes neural stem cell self-renewal in young, but not old, mice by reducing p16Ink4a and p19Arf expression

SUMMARY

Stem cells persist throughout life in diverse tissues by undergoing self-renewing divisions. Self-renewal capacity declines with age, partly due to increasing p16Ink4a expression, but little is known about the mechanisms responsible for these changes. We discovered the Hmga2 transcriptional regulator was highly expressed in fetal neural stem cells but expression declined with age, partly due to increasing let-7b microRNA expression. Hmga2 deficiency reduced stem cell frequency and self-renewal throughout the central and peripheral nervous systems of fetal and young adult mice, but not old adult mice where it was no longer expressed. Hmga2 deficiency did not affect restricted neural progenitor proliferation. Hmga2 deficient fetal and young adult stem cells exhibited increased p16Ink4a and p19Arf expression, and deletion of p16Ink4a and/or p19Arf partially rescued their self-renewal. let-7b over-expression reduced Hmga2 expression and increased p16Ink4a/p19Arf expression. Changes in let-7b and Hmga2 expression during stem cell aging increase p16Ink4a/p19Arf expression and reduce self-renewal.

INTRODUCTION

Stem cells self-renew throughout life in numerous tissues including the central (CNS) and peripheral (PNS) nervous systems (Kruger et al., 2002; Maslov et al., 2004; Molofsky et al., 2006). Nonetheless, the properties of stem cells change between the fetal, young adult, and old adult stages ((Kruger et al., 2002; Kim et al., 2007) and references therein). Most notably, self-renewal capacity declines with age: fetal and young adult stem cells have more self-renewal potential more than old adult stem cells (Kruger et al., 2002; Janzen et al., 2006; Molofsky et al., 2006). Neural stem cell frequency and the rate of neurogenesis also decline with age (Enwere et al., 2004; Maslov et al., 2004; Molofsky et al., 2006).

The reduced stem cell frequency, self-renewal potential, and mitotic activity in aging tissues is partially caused by increasing p16Ink4a expression. p16Ink4a generally cannot be detected in fetal and young adult tissues but is induced in a variety of aging tissues including the nervous system (Zindy et al., 1997; Krishnamurthy et al., 2004; Molofsky et al., 2006). p16Ink4a deficiency partially rescues the decline in stem/progenitor cell function in the aging CNS and other tissues without affecting stem/progenitor cell function in young adult tissues (Janzen et al., 2006; Krishnamurthy et al., 2006; Molofsky et al., 2006). However, it remains uncertain what regulates this change in p16Ink4a expression in old tissues and what mechanisms act in fetal and young adult stem cells to increase their self-renewal relative to old adult stem cells.

Consistent with the changes in stem cell function with age, different transcriptional programs regulate stem cell self-renewal at embryonic, fetal/neonatal, and adult stages. Fetal and neonatal hematopoietic stem cells (HSCs) depend on Sox17 for their maintenance while adult HSCs do not, demonstrating the existence of a distinct fetal/neonatal self-renewal program (Kim et al., 2007). Bmi-1-deficient mice are born with normal numbers of stem cells but exhibit postnatal self-renewal defects that lead to the depletion of adult stem cells (Lessard and Sauvageau, 2003; Molofsky et al., 2003; Park et al., 2003). Other transcriptional regulators are also required by postnatal but not fetal stem cells (Hock et al., 2004a; Hock et al., 2004b; Shi et al., 2004), demonstrating the existence of a postnatal/adult self-renewal program. This raises the question of whether all stage-specific transcriptional regulators of self-renewal cluster into temporally distinct embryonic, fetal/neonatal, and postnatal/adult programs.

Bmi-1 promotes the self-renewal of postnatal stem cells largely by repressing the expression of p16Ink4a and p19Arf (Bruggeman et al., 2005; Molofsky et al., 2005). p16Ink4a is a cyclin-dependent kinase inhibitor that promotes Rb activation, slowing cell cycle progression or inducing cellular senescence (Lowe and Sherr, 2003). p19Arf promotes p53 activity, also slowing cell cycle progression or inducing cellular senescence. Bmi-1 represses p16Ink4a and p19Arf expression within postnatal stem cells (Molofsky et al., 2003). However, p16Ink4a and p19Arf expression increase in aging stem cells despite ongoing Bmi-1 expression (Molofsky et al., 2006). In addition to contributing to the decline in stem cell function with age, p16Ink4a, and perhaps p19Arf, may also contribute to diseases of aging (Sharpless and DePinho, 2007). These studies demonstrate that adult stem cell maintenance critically depends upon transcriptional mechanisms that repress p16Ink4a and p19Arf; however, these studies also raise important questions. Do fetal stem cells have different mechanisms to repress p16Ink4a/p19Arf? Why do old adult stem cells express higher levels of p16Ink4a and p19Arf than young adult stem cells?

Hmga2 is a member of the high mobility group A (HMGA) family that encodes a small, chromatin-associated protein that has no intrinsic transcriptional activity but can modulate transcription by altering chromatin structure (Reeves, 2001). Hmga2 is widely expressed in undifferentiated cells during embryogenesis but expression becomes more restricted as fetal development progresses (Zhou et al., 1995; Hirning-Folz et al., 1998). Hmga2 is rarely detected within normal adult tissues, but is expressed by a variety of benign and malignant tumors (Schoenmakers et al., 1995; Fusco and Fedele, 2007). Recent studies have demonstrated that let-7 family microRNAs are important negative regulators of Hmga2 expression: loss of the let-7 binding sites in Hmga2 increases Hmga2 expression and promotes tumorigenesis (Lee and Dutta, 2007; Mayr et al., 2007; Yu et al., 2007). Hmga2−/− mice exhibit a dwarf phenotype (Zhou et al., 1995). In contrast, over-expression of truncated (let-7 insensitive) HMGA2 in mice increases body size and the incidence of certain tumors (Arlotta et al., 2000; Fedele et al., 2006). It has not yet been tested whether Hmga2 regulates somatic stem cell function.

In a full genome analysis, Hmga2 was the only gene that we found to be preferentially expressed by stem cells and that showed a progressive decline in expression with age. This decline in Hmga2 expression inversely correlated with increasing let-7b expression during aging, and deletion of the let-7 binding sites in the 3′ untranslated region of Hmga2 was required to over-express Hmga2 in stem cells from old mice. Hmga2 increased the self-renewal of fetal and young adult stem cells relative to old adult stem cells by negatively regulating p16Ink4a and p19Arf expression in fetal and young adult, but not old adult, stages. The requirement for Hmga2 in fetal and young adult stem cells, along with the requirement for Bmi-1 in adult stem cells, demonstrates there are overlapping transcriptional mechanisms to prevent the expression of p16Ink4a and p19Arf in stem cells. Such pathways are temporally regulated by changes in let-7b and Hmga2 expression to allow p16Ink4a and p19Arf expression in aging stem cells.

RESULTS

Hmga2 expression declines with age in stem cells

We reasoned that genes that regulate changes in stem cell function with age might show progressive changes in their expression in stem cells with age. In a prior study we identified 26 genes that were more highly expressed by fetal HSCs as compared to young adult HSCs (Kiel et al., 2005). In this study we examined these genes by quantitative (real-time) PCR (qPCR) and with additional microarray data sets (data not shown) to identify genes that fit two additional criteria. First, we identified genes that remained detectably expressed by young adult HSCs but declined significantly further in expression in old adult HSCs (>3-fold, P<0.01). Second, we identified genes that were more highly expressed in HSCs as compared to all CD45+ hematopoietic cells. Only one gene in a full genome analysis met all of these criteria: Hmga2 (Fig. 1A,B).

Figure 1
Hmga2 expression is elevated in stem cells in the hematopoietic and nervous systems but declines with age, in contrast to p16Ink4a and p19Arf

To test whether this unusual expression pattern is observed in other tissues, we examined Hmga2 expression in CNS stem/progenitor cells by performing qPCR on cells from embryonic day (E)14.5 telencephalon, postnatal day (P)0 lateral ventricle ventricular zone (VZ), P30 lateral ventricle subventricular zone (SVZ), P360 SVZ, and P720 SVZ (locations in which neural stem/progenitor cells reside within the forebrain). We again observed declining Hmga2 expression with age (Fig. 1C). Hmga1 did not exhibit a similar change in expression with age by either microarray analysis in HSCs (data not shown) or by qPCR in CNS progenitors (Fig. 1C). Hmga2 was also expressed at higher levels in CNS stem/progenitor cells as compared to differentiated cells by qPCR (Fig. 1D), Western blot (Fig. 1E), and in situ hybridization (Fig. 1F). We confirmed the decline in Hmga2 expression with age in neural stem/progenitor cells by both Western blot (Fig. 1E) and in situ hybridization (Fig. 1F): Hmga2 expression was extinguished in the SVZ of old mice by every technique. The declining Hmga2 expression with age contrasted with the increasing expression of p16Ink4a and p19Arf with age in stem/progenitor cells (Fig. 1C). These data raised the possibility that Hmga2 might contribute to the increased stem cell function in fetal and young adult mice as compared to old adult mice.

Hmga2 promotes fetal and adult neural stem cell self-renewal

To assess Hmga2 function we examined Hmga2 deficient mice (Zhou et al., 1995). Consistent with earlier reports (Benson and Chada, 1994; Zhou et al., 1995), Hmga2 deficiency led to significant (P<0.01) growth retardation that was evident by birth and which remained evident throughout adult life (Suppl. Fig. 1A–F). The brains of adult Hmga2−/− mice were also significantly (P<0.01) smaller than the brains of littermate controls (Suppl. Fig. 1G,H). Despite this growth retardation, we detected no premature death among Hmga2−/− mice.

To assess whether Hmga2 regulates changes in stem cell function during aging we elected to focus on neural stem cells from the CNS and PNS because these stem cell populations exhibit clear declines in frequency, mitotic activity, and/or neurogenesis in vivo during aging (Kruger et al., 2002; Molofsky et al., 2006). We cultured CNS stem cells from the forebrain of Hmga2−/− mice and littermate controls to test whether Hmga2 was required for normal neural stem cell function (Fig. 2A). The percentage of dissociated cells that was capable of forming multipotent neurospheres did not differ between Hmga2−/− and wild-type telencephalon at E11.5 or E14.5, but was significantly reduced (P<0.05) in the Hmga2−/− P0 VZ and P49-56 SVZ as compared to littermate controls (Fig. 2B). Hmga2+/− cells performed similarly to wild-type cells in this assay (data not shown) so cells from wild-type littermates were used as controls in subsequent experiments. To minimize fusion between neurospheres, cells were cultured at very low density in non-adherent cultures (<1 cell/μl of medium), then transferred to adherent secondary cultures to determine the fraction of neurospheres capable of multilineage differentiation. Fetal and adult Hmga2−/− neurospheres were significantly smaller (P<0.05) than wild-type neurospheres and gave rise to significantly fewer multipotent secondary neurospheres (P<0.01) upon subcloning (Fig. 2A, B). These data suggest Hmga2 is not required for the formation of stem cells during embryonic development but that Hmga2−/− deficiency reduces self-renewal potential throughout fetal development and young adulthood, reducing stem cell frequency by birth and even further by adulthood.

Figure 2
Hmga2 promotes the self-renewal of fetal and young adult CNS stem cells

To test whether Hmga2 is required for stem cell function in other tissues, we examined neural crest stem cells (NCSCs) from the guts of Hmga2−/− mice and littermate controls. NCSCs give rise to the enteric nervous system within the gut during fetal development and then persist throughout adult life in the gut wall (Kruger et al., 2002). In the PNS, the frequency of cells capable of forming multipotent neurospheres also did not differ between Hmga2−/− guts and littermate controls at E14.5, but was significantly reduced in Hmga2−/− guts at P0 and P49-56 compared to littermate controls (P<0.05; Suppl. Fig. 2B). The Hmga2−/− neurospheres were also significantly smaller (P<0.01) and gave rise to fewer secondary multipotent neurospheres (P<0.05) upon subcloning at all stages examined (Suppl. Fig. 2B). Hmga2 was thus required to maintain normal neural stem cell self-renewal and frequency throughout the PNS and CNS.

The reduced self-renewal of Hmga2-deficient CNS stem cells was associated with significantly reduced proliferation within stem cell colonies (P<0.05; Fig. 2C, 2F) but no detectable effect on cell death, whether this was assessed based on activated caspase-3 staining (Fig. 2D) or nuclear morphology (Fig. 2E; cell death was rare in all colonies). In the PNS, Hmga2 deficiency also led to the formation of smaller colonies with a lower frequency of dividing cells (Suppl. Fig. 2C–E). These data suggest that Hmga2 promotes neural stem cell self-renewal by promoting division.

To assess whether the reduced self-renewal of Hmga2−/− neural stem cells could be rescued by Hmga2 over-expression, we infected P0 VZ cells from an Hmga2−/− mouse with a dual promoter lentivirus expressing Hmga2 and GFP, or control virus lacking Hmga2 (GFP-only). After infection, the cells were subcloned into secondary cultures to form neurospheres. GFP-only vector did not affect the diameter or self-renewal of neurospheres as compared to uninfected neurospheres within the same cultures (Fig. 2G). In contrast, Hmga2 over-expression significantly increased the diameter and self-renewal of neurospheres as compared to both uninfected neurospheres within the same cultures as well as compared to GFP-only neurospheres (P<0.05; Fig. 2G). Hmga2 over-expression can thus rescue the self-renewal defect in Hmga2−/− neural stem cells.

To assess the effect of Hmga2 deficiency on neural stem cell differentiation we adherently cultured cells from the P0 VZ at clonal density. We first examined multilineage colonies 7 days after plating. Hmga2−/− colonies showed no signs of premature differentiation as only rare cells expressed neuronal or astrocytic markers in either Hmga2−/− or wild-type colonies and no cells expressed the oligodendrocyte marker O4 (Fig. 3A). We cultured for another 7 days in medium that promotes differentiation and observed similar multilineage differentiation to neurons, astrocytes, and oligodendrocytes in Hmga2−/− and wild-type CNS stem cell colonies (Fig. 3B). Similar results were observed in NCSC colonies (data not shown). Indeed, the vast majority of neurospheres cultured from the CNS and PNS at all stages of development underwent multilineage differentiation irrespective of Hmga2 genotype (Suppl. Fig. 3). Hmga2 deficiency therefore does not affect the ability of neural stem cells to undergo multilineage differentiation. However, adherent stem cell colonies, like neurospheres, were significantly smaller in the absence of Hmga2 (Fig. 3A).

Figure 3
Hmga2 deficiency does not affect the ability of CNS stem cells to undergo multilineage differentiation in culture but does reduce proliferation in vivo

Hmga2 deficiency leads to defects in the CNS and PNS in vivo

We examined proliferation in the SVZ of young adult mice by administering a pulse of BrdU. Hmga2−/− mice exhibited a significantly lower overall rate of proliferation in the SVZ in vivo (P<0.01; Fig. 3C). This was consistent with the reduced proliferation within Hmga2−/− CNS stem cell colonies in culture (Fig. 2F), and the reduced brain size of Hmga2−/− mice (Suppl. Fig. 1G, H). The P0 VZ was also slightly but significantly thinner in Hmga2−/− mice as compared to littermate controls (Fig. 3D), consistent with the reduced stem cell frequency (Fig. 2B) and overall mitotic activity (Fig. 3C) in these mice. In the PNS we counted the number of myenteric plexus neurons per transverse section through the distal ileum. Defects in NCSC self-renewal, such as those observed in the absence of Bmi-1, lead to a reduction in the number of neurons per section (Molofsky et al., 2003; Molofsky et al., 2005). Young adult Hmga2−/− mice also had significantly fewer neurons per section than littermate controls (P<0.01; Fig. 3E,F). Hmga2−/− mice exhibit phenotypes in vivo consistent with reduced proliferation by CNS and PNS neural stem cells.

Hmga2 is not required for the proliferation of many restricted neural progenitors

To test whether Hmga2 promotes the proliferation of all progenitors, we examined the effect of Hmga2 deficiency on the proliferation of restricted neuronal progenitors (that make neuron-only colonies) and restricted glial progenitors (that make glia-only colonies) in adherent cultures at clonal density. Whether these restricted progenitors were cultured from the E14.5 telencephalon, the P0 VZ, or the P49-56 SVZ, the size and frequency of CNS neuron-only colonies and glia-only colonies were unaffected by Hmga2 deficiency (Fig. 4A–D). The same was true in the PNS (Fig. 4E–H). These results contrast with the reduced size of multilineage stem cell colonies from the CNS (Fig. 2) and PNS (Suppl. Fig. 2) in the absence of Hmga2 in both adherent and non-adherent cultures. Hmga2 therefore does not generically promote the proliferation of all cells but rather is preferentially required for the self-renewal of stem cells.

Figure 4
Restricted neural progenitors from the CNS and PNS proliferate normally in the absence of Hmga2

Hmga2 negatively regulates p16Ink4a and p19Arf expression

Since Hmga2 expression declines in SVZ cells as p16Ink4a and p19Arf expression increase during aging (Fig. 1C), we wondered whether Hmga2 might promote neural stem cell self-renewal by negatively regulating p16Ink4a and p19Arf expression. We examined p16Ink4a and p19Arf expression by qPCR in CNS and PNS neurospheres cultured from Hmga2−/− mice and littermate controls. p16Ink4a and p19Arf expression increased significantly in CNS (Fig. 5A) and PNS (Fig. 5B) neurospheres from E14.5, P0, and P49-56 Hmga2−/− mice as compared to littermate controls. Fetal neurospheres showed the greatest increase in p16Ink4a and p19Arf expression in the absence of Hmga2. Importantly, there was no effect of Hmga2 deficiency on p16Ink4a or p19Arf expression by neurospheres cultured from old (P570-600) mice, consistent with our failure to detect Hmga2 expression in neural stem/progenitor cells from old mice (Fig. 1). The same trends were evident at the protein level by Western blot in CNS (Fig. 5C) and PNS (Fig. 5D) neurospheres. p16Ink4a and p19Arf expression levels therefore increase in Hmga2−/− CNS stem/progenitor cells from fetal and young adult, but not old adult mice.

Figure 5
Hmga2 negatively regulates p16Ink4a and p19Arf expression in CNS and PNS stem/progenitor cells from fetal and young adult mice but not from old adult mice

To ensure that Hmga2 also negatively regulates p16Ink4a and p19Arf expression in CNS stem/progenitor cells in vivo, we performed Western blots on E14.5 telencephalon cells, P49-56 SVZ cells, and P570-600 SVZ cells from Hmga2−/− mice and littermate controls. In wild-type mice, p16Ink4a and p19Arf were not detectable in cells from fetal telencephalon or young adult SVZ, but became detectable in old adult SVZ (Fig. 5E), consistent with a prior study (Molofsky et al., 2006). In contrast, p16Ink4a and p19Arf were expressed in fetal telencephalon and young adult SVZ of Hmga2−/− mice, though the levels in old adult SVZ were not affected by Hmga2 deficiency (Fig. 5E). Hmga2 is thus required to repress p16Ink4a and p19Arf expression by CNS stem/progenitor cells from fetal and young adult mice, but not old adult mice, in vitro and in vivo.

Over-expression of Hmga2 in neurospheres cultured from the P0 VZ of Hmga2−/− mice also rescued the p16Ink4a expression phenotype (p19Arf was not assessed in this experiment due to the limited protein available). Over-expression of Hmga2 in wild-type neurospheres only slightly increased Hmga2 protein levels and did not affect p16Ink4a levels (Fig. 5F); however, over-expression of Hmga2 in Hmga2−/− neurospheres restored approximately normal Hmga2 and p16Ink4a protein levels (Fig. 5F).

These results were also confirmed using an shRNA against Hmga2, which partially knocked down Hmga2 protein expression within CNS neurospheres. This led to increased p16Ink4a expression within the neurospheres and reduced self-renewal potential (Suppl. Fig. 4).

Hmga2 promotes self-renewal by negatively regulating p16Ink4a and p19Arf expression

To test whether Hmga2 promotes neural stem cell self-renewal by negatively regulating p16Ink4a and p19Arf expression we cultured neurospheres from young adult Hmga2/p16Ink4a/p19Arf compound mutant mice. p16Ink4a/p19Arf deficiency did not significantly affect the percentage of SVZ cells that gave rise to multipotent neurospheres in culture, the diameter of these neurospheres, or their self-renewal potential (number or percentage of cells from individual primary neurospheres that gave rise to multipotent secondary neurospheres upon subcloning; Fig. 6A). In contrast, p16Ink4a/p19Arf deficiency significantly (P<0.05) but partially rescued the reductions in the percentage of Hmga2−/− SVZ cells that gave rise to multipotent neurospheres in culture, the diameter of Hmga2−/− neurospheres, and their self-renewal potential (Fig. 6A). p16Ink4a/p19Arf deficiency also significantly but partially restored the frequency and self-renewal potential of Hmga2−/− NCSCs (Suppl. Fig. 5A).

Figure 6
Deletion of p16Ink4a and/or p19Arf partially rescues the defects in neural stem cell frequency and self-renewal potential as well as SVZ proliferation in Hmga2−/− mice

We also cultured neurospheres from young adult Hmga2−/−p16Ink4a−/− mice and Hmga2−/−p19Arf−/− mice and littermate controls. p16Ink4a deficiency (Fig. 6B) or p19Arf deficiency (Fig. 6C) did not affect the percentage of wild-type SVZ cells that gave rise to multipotent neurospheres in culture, the diameter of these neurospheres, or their self-renewal potential. In contrast, p16Ink4a deficiency (Fig. 6B) or p19Arf deficiency (Fig. 6C) significantly but partially rescued the reductions in the percentage of Hmga2−/− SVZ cells that gave rise to multipotent neurospheres in culture, the diameter of Hmga2−/− neurospheres, and their self-renewal potential in the CNS. Each of p16Ink4a and p19Arf also partially rescued NCSC frequency and self-renewal potential in Hmga2−/− mice (Suppl. Fig. 5B, C). Therefore, p16Ink4a and p19Arf both contribute to the reduced neural stem cell frequency and self-renewal in the CNS and PNS of Hmga2−/− mice.

To test whether p16Ink4a/p19Arf deficiency can also rescue neural stem/progenitor defects observed in Hmga2−/− mice in vivo we examined the rate of proliferation in the SVZ of young adult mice by administering a two-hour pulse of BrdU immediately before sacrifice. p16Ink4a/p19Arf deficiency did not affect the percentage of SVZ cells that incorporated BrdU in an otherwise wild-type background (Fig. 6D,E). However, the percentage of SVZ cells that incorporated BrdU was significantly (P<0.01) reduced in the absence of Hmga2, and this reduction was partially rescued by p16Ink4a/p19Arf deficiency (Fig. 6D,E). In the PNS, p16Ink4a/p19Arf deficiency did not affect the number of myenteric neurons per transverse section through the distal ileum in wild-type mice (Suppl. Fig. 5D,E). However, the number of myenteric neurons per transverse section through the distal ileum was significantly (P<0.01) reduced in the absence of Hmga2, and this reduction was significantly but partially rescued by p16Ink4a/p19Arf deficiency (Suppl. Fig. 5D,E). p16Ink4a and/or p19Arf deficiency also partially rescued the reduction in brain mass within Hmga2−/− mice but not the reduction in overall body mass in Hmga2−/− mice (Suppl. Fig. 6). p16Ink4a/p19Arf deficiency therefore partially rescues the neural stem/progenitor defects in the CNS and PNS of Hmga2−/− mice in vivo.

Hmga2 does not regulate the self-renewal of neural stem cells from old adult mice

Hmga2 expression declined with age and no longer affected p16Ink4a or p19Arf expression in the SVZ of old adult mice (Fig. 5E). To test whether Hmga2 regulates neural stem cell function in old mice we aged Hmga2−/− mice and cultured neurospheres from P570-600 mice and littermate controls. The frequency of SVZ or gut cells that formed multipotent neurospheres in culture was significantly (P<0.05) reduced with age in wild type mice (compare Fig. 7A to Fig. 2B, and Suppl. Fig. 2B to Suppl. Fig. 7A) as previously reported (Molofsky et al., 2006). Hmga2 deficiency did not significantly affect the percentage of SVZ cells or outer muscle/plexus layer gut cells from old mice that formed multipotent neurospheres in culture, the diameter of these neurospheres, or their self-renewal potential (Fig. 7A; Suppl. Fig. 7A). Interestingly, the frequencies of CNS and PNS cells that could form stem cell colonies in culture recovered to near wild-type levels in old Hmga2−/− mice, suggesting that homeostatic mechanisms are able to restore normal neural stem cell frequencies in old mice, when Hmga2 is no longer required for self-renewal. Hmga2 deficiency also did not significantly affect the number of cells per colony or the rate of BrdU incorporation within CNS (Suppl. Fig. 7B, C) or PNS (Suppl. Fig. 7D, E) stem cell colonies cultured from old mice. Finally, Hmga2 deficiency did not significantly affect the overall rate of proliferation within the SVZ of old mice (Fig. 7B). These data demonstrate that Hmga2 is not required for the self-renewal of neural stem cells in old mice, in contrast to its role in fetal and young adult mice.

Figure 7
Hmga2 is not required for the self-renewal of CNS stem cells from old mice, and increasing let-7 microRNA expression with age contributes to the decline in Hmga2 expression

Increasing let-7 expression contributes to the decline in Hmga2 expression with age

To test whether let-7 microRNAs contribute to the declining expression of Hmga2 within aging neural stem/progenitor cells we compared the expression of 7 members of the let-7 family by qPCR in E14.5 telencephalon, P0 VZ, P360 SVZ, and P720 SVZ. Only let-7b showed progressive increases in expression with age that inversely correlated with the declines in Hmga2 expression (Fig. 7C). To test whether elevated expression of let-7b in neural stem cells could reduce their self-renewal potential we over-expressed let-7b and GFP in wild-type cells cultured from P0 VZ using a dual promoter lentivirus (Rubinson et al., 2003). The GFP-only control virus did not significantly affect the diameter of neurospheres or the self-renewal of multipotent neurospheres as compared to uninfected neurospheres within the same cultures (Fig. 7D). However, the let-7b+GFP virus did significantly (P<0.01) reduce the diameter of neurospheres and the self-renewal of multipotent neurospheres as compared to uninfected neurospheres within the same cultures as well as compared to GFP-only infected neurospheres in sister cultures (Fig. 7D). let-7b over-expression in these neurospheres also reduced Hmga2 expression and increased p16Ink4a expression (p19Arf was not assessed in these experiments due to the limited protein) (Fig. 7F, first panel). These data indicate that increasing let-7b expression can reduce Hmga2 expression, increase p16Ink4a expression, and reduce self-renewal, potentially explaining the changes that are observed in aging neural stem cells.

We also tested whether Hmga2 over-expression could increase the self-renewal of neural stem cells from old mice. Interestingly, over-expression of wild-type Hmga2 in neural stem cells did not significantly affect neurosphere diameter or the self-renewal of multipotent neurospheres from old wild-type (Fig. 7E) or Hmga2−/− mice (Suppl. Fig. 7F). Moreover, in contrast to what was observed in neurospheres from the P0 VZ (Fig. 5F), over-expression of wild-type Hmga2 in neurospheres from the P600 SVZ only marginally increased Hmga2 protein levels, irrespective of whether the cells were from wild-type or Hmga2−/− mice (Fig. 7F). Since we had observed significantly increased let-7b expression in P720 SVZ as compared to P0 VZ (Fig. 7C) we hypothesized that Hmga2 expression in old neural stem cells is limited by increased let-7b expression in these cells. To test this we over-expressed a form of Hmga2 with a wild-type coding sequence but a truncated 3′ untranslated region that lacks the let-7 binding sites. This let-7 insensitive form of Hmga2 increased Hmga2 protein levels and reduced p16Ink4a protein levels in neurospheres from old mice to levels similar to those observed in neurospheres from newborn mice (Fig. 7F). Over-expression of let-7 insensitive Hmga2 also significantly (P<0.05) increased the diameter and self-renewal potential of multipotent neurospheres from P600 wild-type mice (Fig. 7E) and Hmga2−/− mice (Suppl. Fig. 7F). These data suggest that increasing let-7 expression within aging neural stem/progenitor cells contributes to the decline in Hmga2 expression in these cells.

Hmga2 may indirectly regulate p16Ink4a/p19Arf expression

We have not so far been able to detect Hmga2 binding to the p16Ink4a/p19Arf locus by chromatin immunoprecipitation (Suppl. Fig. 8A). In contrast, we have been able to detect Hmga2 binding to the JunB locus (Suppl. Fig. 8A). JunB promotes p16Ink4a/p19Arf expression in stem cells (Passegue et al., 2004). To test whether Hmga2 regulates JunB expression we compared JunB transcript levels by qPCR in neurospheres cultured from E14.5 telencephalon, P0 VZ, P49-56 SVZ, and P600 SVZ from Hmga2−/− mice and littermate controls. We observed significantly increased JunB expression in Hmga2−/− neurospheres as compared to wild-type neurospheres cultured from the fetal and young adult, but not the old adult stage, in both the CNS (Suppl. Fig. 8B) and PNS (Suppl. Fig. 8C). Finally, JunB expression levels increased with age in wild-type SVZ cells (Suppl. Fig. 8D). These data are all consistent with the possibility that Hmga2 negatively regulates JunB expression by a mechanism that involves binding to the JunB locus and that JunB promotes p16Ink4a/p19Arf expression in the absence of Hmga2. While Hmga2 has generally appeared to be a transcriptional activator, there are some contexts in which it has appeared to have repressive activity (Reeves, 2001). We consider this model for Hmga2 function preliminary as it will be necessary to generate and age Hmga2/JunB compound mutant mice to functionally evaluate the model.

DISCUSSION

Neural stem cell frequency, self-renewal potential, mitotic activity, and neurogenesis all decline with age (Kruger et al., 2002; Enwere et al., 2004; Maslov et al., 2004; Molofsky et al., 2006). Yet the mechanisms that increase the frequency and function of fetal and young adult stem cells relative to old adult stem cells have remained unclear. One important insight was the discovery that p16Ink4a expression is induced in aging tissues (Zindy et al., 1997; Krishnamurthy et al., 2004) and that this contributes to the decline in stem/progenitor cell function with age (Janzen et al., 2006; Krishnamurthy et al., 2006; Molofsky et al., 2006). p19Arf expression also increases in aging tissues, though it has not yet been tested whether this contributes to the decline in stem cell function or tissue regenerative capacity with age. The recent discovery of polymorphisms near the p16Ink4a/p19Arf locus that are associated with age-related diseases suggests that this locus may be broadly involved in age-related morbidity ((Sharpless and DePinho, 2007) and references therein). For these reasons it is important to understand the mechanisms that regulate changes in p16Ink4a/p19Arf expression and stem cell function with age.

In this study we demonstrate that Hmga2 increases the frequency and self-renewal of fetal and young adult stem cells as compared to old adult stem cells (Fig. 2 and Suppl. Fig. 2B), partly by negatively regulating the expression of p16Ink4a and p19Arf in fetal and young adult, but not old adult stem cells (Fig. 5). The defects in Hmga2−/− stem cells could be significantly but partially rescued by deleting p16Ink4a and/or p19Arf (Fig. 6). This demonstrates that p16Ink4a/p19Arf can be expressed by fetal stem cells in vivo and that specific mechanisms are required to prevent this. Together with the fact that Bmi-1 represses p16Ink4a and p19Arf postnatally in stem cells (Jacobs et al., 1999; Molofsky et al., 2003; Bruggeman et al., 2005; Molofsky et al., 2005), our results demonstrate there are overlapping transcriptional mechanisms that maintain stem cells throughout life by preventing the expression of p16Ink4a and p19Arf.

Our results are consistent with a number of prior studies that found Hmga2 to be a proto-oncogene (Fusco and Fedele, 2007) and raise the possibility that the tumorigenic effects of Hmga2 may partly reflect its ability to negatively regulate the expression of the p16Ink4a and p19Arf tumor suppressors. Another recent study found that HMGA1 and HMGA2 can promote cellular senescence by cooperating with p16Ink4a to promote the formation of senescence-associated heterochromatin foci (SAHF) in cultured human fibroblasts (Narita et al., 2006). Our results are not inconsistent with those of Narita et al. for two reasons. First, Narita et al. found HMGA1 was necessary for the formation of SAHF but that HMGA2 had much less effect (Narita et al., 2006). This raises the possibility that the main physiological function of Hmga2 may be to promote stem cell function in fetal and young adult mice while Hmga1 may be the main regulator of cellular senescence. Second, the ability of HMGAs to promote senescence is context dependent: HMGA function may be different in mouse neural stem cells during normal development/aging as compared to human fibroblasts in culture.

Additional mechanisms likely play important roles in regulating the change in p16Ink4a/p19Arf expression with age, either by acting upstream or downstream of Hmga2 or by acting in parallel. For example, it is unclear whether polycomb complexes function as part of the same pathway or as part of a different pathway as compared to Hmga2 to regulate p16Ink4a/p19Arf expression. Recent data raised the possibility that loss of Ezh2 function in aging stem cells may contribute to increasing p16Ink4a/p19Arf expression (Bracken et al., 2007).

Together, our data suggest the existence of a novel pathway that regulates stem cell aging: let-7b expression increases with age in neural stem cells, decreasing Hmga2 expression, which in turn increases JunB and p16Ink4a/p19Arf expression, reducing stem cell frequency and function. Future studies will be required to further evaluate the roles of let-7 family members and JunB in this pathway, but the data in this study clearly demonstrate an important role for Hmga2 in regulating age-related changes in stem cell function and p16Ink4a/p19Arf expression.

EXPERIMENTAL PROCEDURES

Hmga2+/− (Zhou et al., 1995), p16Ink4a/p19Arf +/− (Serrano et al., 1996), p16Ink4a+/− (Sharpless et al., 2001), and p19Arf+/− (Kamijo et al., 1997) mice were housed at the University of Michigan Unit for Laboratory Animal Medicine and backcrossed at least six times onto C57BL/Ka background. All mice were genotyped by PCR as described in Suppl. Material.

Cell culture and self-renewal assay

CNS and PNS progenitors were isolated as described in prior studies (Molofsky et al., 2003; Molofsky et al., 2005) (see Suppl. Materials for details). For adherent cultures, CNS and PNS progenitors were plated at a clonal density of 0.33 cells/μl (500 cells per 35mm well), in 6 well plates (Corning) that had been sequentially coated with 150 μg/ml poly-d-lysine (Biomedical Technologies, Stoughton, MA) and 20 μg/ml laminin (Sigma). For the non-adherent culture of neurospheres, CNS and PNS progenitors were plated at a density of 0.67–1.33 cells/μl (1000–2000 cells per 35mm well) in ultra-low binding 6-well plates (Corning). The medium was a 5:3 mixture of DMEM-low:neurobasal medium, supplemented with 20 ng/ml recombinant human bFGF (R&D Systems, Minneapolis, MN), 1% N2 supplement (Gibco), 2% B27 supplement (Gibco), 50 mM 2-mercaptoethanol, and penicillin/streptomycin (Biowhittaker). ‘Self-renewal medium’ for CNS cultures (designed to promote the expansion of undifferentiated cells) also contained 20 ng/ml EGF (R&D Systems), and 10% chick embryo extract (CEE). ‘Self-renewal medium’ for PNS cultures contained 15% CEE, 35 mg/ml (110 nM) retinoic acid (Sigma), and 20 ng/ml IGF1 (R&D Systems). ‘Differentiation medium’ for CNS cultures contained 10 ng/ml (instead of 20 ng/ml) bFGF, no EGF, no CEE, and 5% fetal bovine serum (Gibco). Differentiation medium for PNS cultures was the same except that retinoic acid was also added. After being grown in self-renewal medium, neurospheres were routinely transferred to adherent cultures containing differentiation medium before being stained to assess multilineage differentiation (see below). All cultures were maintained at 37°C in 6% CO2/balance air.

To measure self-renewal, individual primary CNS neurospheres were dissociated by trituration, then replated at clonal density in non-adherent secondary cultures. Secondary neurospheres were counted 7–9 d later and then transferred to adherent cultures containing differentiation medium to assess the percentage of secondary neurospheres that could undergo multilineage differentiation. Individual PNS neurospheres were replated for 72 h into adherent plates to allow the spheres to spread out over the culture dish, facilitating dissociation. The adherent colonies were then treated with trypsin and collagenase (four parts 0.05% trypsin-EDTA plus one part 10 mg/mL collagenase IV) for 3 min at 37°C followed by trituration. Two thousand dissociated cells were replated per well of a six-well plate and secondary neurospheres were cultured and counted as described for CNS neurospheres.

For viral infection experiments, CNS progenitors were plated at a high density of 10 cells/μl and cultured adherently in CNS self-renewal medium. After 48 hours, viral supernatant was added for 24 hours, then switched to CNS self-renewal medium for a further 24 hours. Cells were harvested by incubating for 1.5 min at 37°C in trypsin/EDTA and transferred to non-adherent cultures to form neurospheres for 6–7 days.

Immunocytochemistry, immunohistochemistry, western blots, qRT-PCR, and ChIP

See Supplementary Materials.

Supplementary Material

01

SUPPLEMENTARY MATERIAL:

Supplementary figure 1: Hmga2-deficient mice exhibit progressive growth retardation and reduced brain mass. Hmga2+/+ and Hmga2−/− mice appeared grossly indistinguishable at E14.5 (A), and body masses were not significantly different (B; N=23 wild-type mice and 25 Hmga2−/− mice). At P0 however, Hmga2−/− mice were noticeably smaller (C), and their masses were significantly reduced relative to wild-type littermates (+/+) (D; N=21 wild-type and 16 Hmga2−/−, *P<0.01). By P49-56, this growth retardation became more apparent (E), and the masses of Hmga2−/− mice (−/−) were approximately one third of wild-type littermates (F; N=18 wild-type and 16 Hmga2−/−, *P<0.01). This reduced body mass remained evident at P570-600 (F; N=6 wild-type and 6 Hmga2−/−, *P<0.01). Brain masses were not significantly different between wild-type and Hmga2−/− mice at P0 (H; N=14 wild-type and 12 Hmga2−/−). However, at P49-56, the brains of Hmga2−/− mice were noticeably smaller (G), and their masses were significantly reduced compared to wild-type at both P48-56 (H; N=18 wild-type and 16 Hmga2−/−, *P<0.01) and at P570-600 (H; N=6 wild-type and 6 Hmga2−/−, *P<0.01). All statistical comparisons were done using unpaired T-tests in this figure.

Supplementary figure 2: Gut neural crest stem cell (NCSC) frequency and self-renewal potential are reduced in the absence of Hmga2. A) Typical neurospheres that formed after 10 days in non-adherent cultures from E14.5 gut cells. B) The percentage of cells from E14.5 gut, P0 plexus/outer muscle layer, and P49-56 plexus/outer muscle layer of Hmga2−/− mice and littermate controls that gave rise to multipotent neurospheres, the diameter of these neurospheres, and their self-renewal potential (number and percentage of cells from individual primary neurospheres that gave rise to multipotent secondary neurospheres upon subcloning). Hmga2 deficiency significantly reduced neural stem cell self-renewal at all stages and frequency at P0 and P49-56 (5–7 independent experiments per stage, all statistics represent mean±SD, *P<0.01, **P<0.05). C) P0 gut cells were dissociated and plated in adherent cultures at clonal density and the numbers of cells per colony were counted after 6 and 9 days. Only colonies with the appearance of multilineage NCSC colonies were counted. At each time point, Hmga2−/− (−/−) colonies contained significantly fewer cells than wild-type (+/+) colonies (3 independent experiments, *P<0.01). The percentage of cells that incorporated a pulse of BrdU in P0 NCSC colonies (D) was significantly reduced within Hmga2−/− colonies as compared to wild-type colonies (E; 3 independent experiments, *P<0.01). All T-tests were paired.

Supplementary figure 3: Hmga2−/− neurospheres from the CNS and PNS undergo multilineage differentiation in a manner indistinguishable from wild-type neurospheres. CNS neurospheres were routinely cultured from dissociated E14.5 telencephalon, or P0 VZ (A), or P49-56 lateral ventricle SVZ at very low cell density and then transferred individually to adherent cultures before being stained for markers of differentiation. At E14.5, 95% of Hmga2+/+ neurospheres and 94% of Hmga2−/− neurospheres formed Tuj1+ neurons, GFAP+ astrocytes and O4+ oligodendrocytes. At P0, 94% of Hmga2+/+ and 95% of Hmga2−/− neurospheres formed neurons, astrocytes and oligodendrocytes. At P49-56, 94% of Hmga2+/+ and 93% of Hmga2−/− neurospheres formed neurons, astrocytes and oligodendrocytes. PNS neurospheres were routinely cultured from dissociated E14.5 gut, P0 gut plexus/outer muscle layers (B), or P49-56 gut plexus/outer muscle layers at very low cell density and then transferred individually to adherent cultures before being stained for markers of differentiation. At E14.5, 95% of Hmga2+/+ and 93% of Hmga2−/− neurospheres formed peripherin+ neurons, GFAP+ glia, and smooth muscle actin+ myofibroblasts. At P0, 93% of Hmga2+/+ and 92% of Hmga2−/− neurospheres formed neurons, glia, and myofibroblasts. At P49-56, 94% of Hmga2+/+ and 92% of Hmga2−/− neurospheres formed neurons, glia and myofibroblasts.

Supplementary figure 4: Hmga2 knockdown in CNS stem cells using shRNA leads to increased p16Ink4a expression and reduced self-renewal potential. A) P0 VZ cells were dissociated from wild-type mice and infected with either GFP-only or Hmga2 shRNA+GFP lentiviral vectors. Multipotent neurospheres were then allowed to develop from these cells at clonal density. Acute knockdown of Hmga2 by shRNA significantly decreased their self-renewal relative to uninfected neurospheres in the same cultures (3 experiments: **P<0.05) as well as relative to GFP-only infected neurospheres in control cultures. B) Protein was extracted from the primary CNS neurospheres examined in (A) and subjected to Western blot for Hmga2, p16Ink4a, and β-actin (loading control). Hmga2 shRNA decreased Hmga2 protein expression and increased p16Ink4a protein expression in P0 CNS neurospheres. Un: uninfected, In: infected.

Supplementary figure 5: Deletion of p16Ink4a/p19Arf, p16Ink4a alone, or p19Arf alone, partially rescues the defects in NCSC frequency and self-renewal potential as well as gut neurogenesis in Hmga2−/− mice. Images show typical primary PNS neurospheres formed after 10 days culture of P49-56 gut plexus/outer muscle layer cells. p16Ink4a/p19Arf deficiency (A; 4–6 mice per genotype in 4 independent experiments), p16Ink4a deficiency (B; 4–5 mice per genotype in 3 independent experiments), or p19Arf deficiency (C; 3–5 mice per genotype in 3 independent experiments) did not affect the percentage of wild-type gut cells that formed multipotent neurospheres or their self-renewal potential (absolute number or percentage of primary neurosphere cells that gave rise to multipotent secondary neurospheres upon subcloning of individual neurospheres) but did significantly increase the percentage of Hmga2−/− gut cells that formed multipotent neurospheres, the diameter of these neurospheres, and their self-renewal potential. All data represent mean±SD (*, significantly different (P<0.05) from wild-type; §, significantly different from Hmga2+/+p16Ink4a/p19Arf−/− mice (A) or Hmga2+/+p16Ink4a−/− mice (B) or Hmga2+/+p19Arf−/− mice (C); #, significantly different from Hmga2−/−p16Ink4a/p19Arf+/+ mice (A) or Hmga2−/−p16Ink4a+/+ mice (B) or Hmga2−/−p19Arf+/+ mice (C)). D) Gut sections from Hmga2/p16Ink4a/p19Arf mutant mice in which myenteric plexus neurons are indicated with brackets. E) p16Ink4a/p19Arf deficiency partially rescued the reduction in HuC/D+ neurons per transverse section through the distal ileum in young adult Hmga2−/− mice without affecting the numbers of neurons in wild-type littermates (3 mice per genotype, 6–10 sections per mice). All T-tests were paired.

Supplementary figure 6: p16Ink4a/p19Arf deficiency, or p16Ink4a deficiency, or p19Arf deficiency increases the brain mass but not the overall body mass of Hmga2−/− mice. Body masses (A,C,E) or brain masses (B,D,F) of Hmga2/p16Ink4a/p19Arf (A,B; 8–10 mice per genotype), Hmga2/p16Ink4a (C,D; 7–9 mice per genotype), or Hmga2/p19Arf (E,F; 9–11 mice per genotype) compound mutant mice were examined at P49-56. In each case, Hmga2 deficiency significantly reduced body mass. p16Ink4a/p19Arf deficiency, p16Ink4a deficiency, or p19Arf deficiency did not affect the body mass of wild-type or Hmga2−/− mice. p16Ink4a/p19Arf deficiency or p16Ink4a deficiency did not affect the brain mass of wild-type mice but did partially rescue the brain mass reduction observed in Hmga2−/− mice. p19Arf deficiency showed a trend toward rescuing brain mass but the effect was not statistically significant. All error bars represent SD (*, significantly different (P<0.05) from wild-type; §, significantly different from Hmga2+/+p16Ink4a/p19Arf−/− mice (A,B), Hmga2+/+p16Ink4a−/− mice (C,D) or Hmga2+/+p19Arf−/− mice (E,F); #, significantly different from Hmga2−/−p16Ink4a/p19Arf+/+ mice (A,B), Hmga2−/−p16Ink4a+/+ mice (C,D) or Hmga2−/−p19Arf+/+ mice (E,F)). Statistical comparisons were done using unpaired T-tests in this figure.

Supplementary figure 7: Hmga2 is not required for the proliferation or self-renewal of gut NCSCs or CNS stem cells from old mice, and Hmga2 protein expression is regulated post-transcriptionally in CNS neurospheres from old Hmga2−/− mice. A) Cells were isolated from wild-type or Hmga2−/− gut plexus/outer muscle layers at P570-600, and cultured to generate PNS neurospheres (A; 10 days in culture). The frequency of PNS cells that formed multipotent neurospheres, the diameter of these neurospheres, and their self-renewal potential upon subcloning were unaffected by Hmga2 deficiency (A; 3 independent experiments). (B–E) Hmga2 deficiency did not affect the numbers of cells per colony within adherent cultures of CNS SVZ cells (B) or gut cells (D) from P570-600 mice. Only colonies with the appearance of stem cell colonies were counted (3 independent experiments). Hmga2 deficiency did not affect the percentage of cells within adherent colonies formed by SVZ cells (C) or gut cells (E) from P570-600 mice that incorporated a pulse of BrdU (3 independent experiments). F) P600 SVZ cells from Hmga2−/− animals were infected with GFP-only control lentivirus, Hmga2+GFP lentivirus, or 3′-UTR truncated Hmga2 (lacking let-7 binding sites)+GFP lentivirus, and allowed to form neurospheres. Neither over-expression of GFP nor wild-type Hmga2 altered the size or self-renewal of neurospheres. In contrast, over-expression of 3′-UTR truncated Hmga2 significantly increased the size and self-renewal of neurospheres (3 experiments: **P<0.05). All T-tests were paired.

Supplementary figure 8: Hmga2 protein binds to the junB locus in CNS neurospheres, and junB expression is increased within neurospheres in the absence of Hmga2 or within wild-type SVZ cells in vivo as Hmga2 expression declines during aging. A) Chromatin immunoprecipitation (ChIP) of Hmga2 protein in P0 CNS neurospheres. P0 SVZ cells from wild-type animals were infected with Hmga2-Flagx3 retrovirus and allowed to form neurospheres. Genomic DNA was then extracted from the neurospheres and subjected to ChIP with anti-FLAG or with anti-mouse IgG (control) antibody. junB locus amplification was detected in the FLAG pull-down fraction (FLAG), but not in the IgG pull-down fraction (IgG). Neither p16Ink4a nor p19Arf locus amplification were detected after FLAG pull-down. We also did not detect Hmga2 binding at other loci that encode proteins that can regulate p16Ink4a or p19Arf expression, including Bmi-1, tbx2, cbx8, or E2F3a. Input is the fraction before immunoprecipitation. (B) CNS neurospheres were cultured from wild-type or Hmga2−/− E14.5 telencephalon, P0 VZ, P49-56 SVZ, or P570-600 SVZ. (C) PNS neurospheres were cultured from wild-type or Hmga2−/− E14.5 gut, or P0, P49-56, or P570-600 plexus/outer muscle layer. RNA was extracted from primary CNS (B) or PNS (C) neurospheres and the levels of junB, bmi-1, and cbx8 were determined by qPCR. Each bar shows the fold-increase in Hmga2−/− as compared to wild-type neurospheres (error bars represent SD, 3–4 independent experiments per stage; **P<0.05). junB expression was increased in CNS and PNS neurospheres, from fetal but not from old mice, in the absence of Hmga2. D) Bmi-1, junB and Pten expression were compared by qPCR in freshly dissected E14.5 telencephalon, P0 VZ, P30 SVZ, P360 SVZ, and P720 SVZ (expressed as fold change relative to P0 SVZ; each bar represents mean±SD for 3–4 mice per stage). junB expression significantly increased with age (*P<0.01,**P<0.05), as Hmga2 expression declines and p16Ink4a/p19Arf expression increase. These data are consistent with the possibility that JunB may mediate the effect of Hmga2 on p16Ink4a/p19Arf expression. All T-tests were unpaired.

Acknowledgments

This work was supported by the Howard Hughes Medical Institute, the National Institute of Neurological Disorder and Stroke (R01 NS040750), and the National Institute on Aging (R01 AG024945-01). Flow-cytometry was partially supported by the UM-Comprehensive Cancer NIH CA46592. Antibody production was supported in part by NIDDK Grant NIH5P60-DK20572 to the Michigan Diabetes Research and Training Center. JN was supported by a postdoctoral fellowship from the Japan Society for the Promotion of Science. Thanks to David Adams and Martin White for flow-cytometry, and to Elizabeth Smith (UM Hybridoma Core) for antibody production. Thanks to Charles Sherr for providing p19Arf−/−f mice, and to Ron DePinho for providing p16Ink4a/p19Arf−/− and p16Ink4a−/− mice. Thanks to Scott Lowe and to Masashi Narita for providing anti-Hmga2 antibody and to Tom Lanigan in the UM Vector Core for generating virus.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  • Arlotta P, Tai AK, Manfioletti G, Clifford C, Jay G, Ono SJ. Transgenic mice expressing a truncated form of the high mobility group I-C protein develop adiposity and an abnormally high prevalence of lipomas. J Biol Chem. 2000;275:14394–14400. [PubMed]
  • Benson KF, Chada K. Mini-mouse: phenotypic characterization of a transgenic insertional mutant allelic to pygmy. Genet Res. 1994;64:27–33. [PubMed]
  • Bracken AP, Kleine-Kohlbrecher D, Dietrich N, Pasini D, Gargiulo G, Beekman C, Theilgaard-Monch K, Minucci S, Porse BT, Marine JC, et al. The Polycomb group proteins bind throughout the INK4A-ARF locus and are disassociated in senescent cells. Genes Dev. 2007;21:525–530. [PubMed]
  • Bruggeman SW, Valk-Lingbeek ME, van der Stoop PP, Jacobs JJ, Kieboom K, Tanger E, Hulsman D, Leung C, Arsenijevic Y, Marino S, et al. Ink4a and Arf differentially affect cell proliferation and neural stem cell self-renewal in Bmi1-deficient mice. Genes Dev. 2005;19:1438–1443. [PubMed]
  • Enwere E, Shingo T, Gregg C, Fujikawa H, Ohta S, Weiss S. Aging results in reduced epidermal growth factor receptor signaling, diminished olfactory neurogenesis, and deficits in fine olfactory discrimination. J Neurosci. 2004;24:8354–8365. [PubMed]
  • Fedele M, Visone R, De Martino I, Troncone G, Palmieri D, Battista S, Ciarmiello A, Pallante P, Arra C, Melillo RM, et al. HMGA2 induces pituitary tumorigenesis by enhancing E2F1 activity. Cancer Cell. 2006;9:459–471. [PubMed]
  • Fusco A, Fedele M. Roles of HMGA proteins in cancer. Nat Rev Cancer. 2007;7:899–910. [PubMed]
  • Hirning-Folz U, Wilda M, Rippe V, Bullerdiek J, Hameister H. The expression pattern of the Hmgic gene during development. Genes Chromosomes Cancer. 1998;23:350–357. [PubMed]
  • Hock H, Hamblen MJ, Rooke HM, Schindler JW, Saleque S, Fujiwara Y, Orkin SH. Gfi-1 restricts proliferation and preserves functional integrity of haematopoietic stem cells. Nature. 2004a;431:1002–1007. [PubMed]
  • Hock H, Meade E, Medeiros S, Schindler JW, Valk PJ, Fujiwara Y, Orkin SH. Tel/Etv6 is an essential and selective regulator of adult hematopoietic stem cell survival. Genes Dev. 2004b;18:2336–2341. [PubMed]
  • Jacobs JJL, Kieboom K, Marino S, DePinho RA, Lohuizen MV. The oncogene and polycomb-group gene bmi-1 regulates cell proliferation and senescence through the ink4a locus. Nature. 1999;397:164–168. [PubMed]
  • Janzen V, Forkert R, Fleming HE, Saito Y, Waring MT, Dombkowski DM, Cheng T, DePinho RA, Sharpless NE, Scadden DT. Stem-cell ageing modified by the cyclin-dependent kinase inhibitor p16INK4a. Nature. 2006;443:421–426. [PubMed]
  • Kamijo T, Zindy F, Roussel MF, Quelle DE, Downing JR, Ashmun RA, Grosveld G, Sherr CJ. Tumor suppression at the mouse INK4a locus mediated by the alternative reading frame product p19ARF. Cell. 1997;91:649–659. [PubMed]
  • Kiel MJ, Iwashita T, Yilmaz OH, Morrison SJ. Spatial differences in hematopoiesis but not in stem cells indicate a lack of regional patterning in definitive hematopoietic stem cells. Dev Biol. 2005;283:29–39. [PubMed]
  • Kim I, Saunders TL, Morrison SJ. Sox17 dependence distinguishes the transcriptional regulation of fetal from adult hematopoietic stem cells. Cell. 2007;130:470–483. [PMC free article] [PubMed]
  • Krishnamurthy J, Ramsey MR, Ligon KL, Torrice C, Koh A, Bonner-Weir S, Sharpless NE. p16INK4a induces an age-dependent decline in islet regenerative potential. Nature. 2006;443:453–457. [PubMed]
  • Krishnamurthy J, Torrice C, Ramsey MR, Kovalev GI, Al-Regaiey K, Su L, Sharpless NE. Ink4a/Arf expression is a biomarker of aging. J Clin Invest. 2004;114:1299–1307. [PMC free article] [PubMed]
  • Kruger GM, Mosher JT, Bixby S, Joseph N, Iwashita T, Morrison SJ. Neural crest stem cells persist in the adult gut but undergo changes in self-renewal, neuronal subtype potential, and factor responsiveness. Neuron. 2002;35:657–669. [PMC free article] [PubMed]
  • Lee YS, Dutta A. The tumor suppressor microRNA let-7 represses the HMGA2 oncogene. Genes Dev. 2007;21:1025–1030. [PubMed]
  • Lessard J, Sauvageau G. Bmi-1 determines the proliferative capacity of normal and leukaemic stem cells. Nature. 2003;423:255–260. [PubMed]
  • Lowe SW, Sherr CJ. Tumor suppression by Ink4a-Arf: progress and puzzles. Current Opinion in Genetics & Development. 2003;13:77–83. [PubMed]
  • Maslov AY, Barone TA, Plunkett RJ, Pruitt SC. Neural stem cell detection, characterization, and age-related changes in the subventricular zone of mice. J Neurosci. 2004;24:1726–1733. [PubMed]
  • Mayr C, Hemann MT, Bartel DP. Disrupting the pairing between let-7 and Hmga2 enhances oncogenic transformation. Science. 2007;315:1576–1579. [PMC free article] [PubMed]
  • Molofsky AV, He S, Bydon M, Morrison SJ, Pardal R. Bmi-1 promotes neural stem cell self-renewal and neural development but not mouse growth and survival by repressing the p16Ink4a and p19Arf senescence pathways. Genes Dev. 2005;19:1432–1437. [PubMed]
  • Molofsky AV, Pardal R, Iwashita T, Park IK, Clarke MF, Morrison SJ. Bmi-1 dependence distinguishes neural stem cell self-renewal from progenitor proliferation. Nature. 2003;425:962–967. [PMC free article] [PubMed]
  • Molofsky AV, Slutsky SG, Joseph NM, He S, Pardal R, Krishnamurthy J, Sharpless NE, Morrison SJ. Increasing p16INK4a expression decreases forebrain progenitors and neurogenesis during ageing. Nature. 2006;443:448–452. [PMC free article] [PubMed]
  • Narita M, Narita M, Krizhanovsky V, Nunez S, Chicas A, Hearn SA, Myers MP, Lowe SW. A novel role for high-mobility group a proteins in cellular senescence and heterochromatin formation. Cell. 2006;126:503–514. [PubMed]
  • Park IK, Qian D, Kiel M, Becker M, Pihalja M, Weissman IL, Morrison SJ, Clarke M. Bmi-1 is required for the maintenance of adult self-renewing hematopoietic stem cells. Nature. 2003;423:302–305. [PubMed]
  • Passegue E, Wagner EF, Weissman IL. JunB deficiency leads to a myeloproliferative disorder arising from hematopoietic stem cells. Cell. 2004;119:431–443. [PubMed]
  • Reeves R. Molecular biology of HMGA proteins: hubs of nuclear function. Gene. 2001;277:63–81. [PubMed]
  • Rubinson DA, Dillon CP, Kwiatkowski AV, Sievers C, Yang L, Kopinja J, Rooney DL, Zhang M, Ihrig MM, McManus MT, et al. A lentivirus-based system to functionally silence genes in primary mammalian cells, stem cells and transgenic mice by RNA interference. Nat Genet. 2003;33:401–406. [PubMed]
  • Schoenmakers EF, Wanschura S, Mols R, Bullerdiek J, Van den Berghe H, Van de Ven WJ. Recurrent rearrangements in the high mobility group protein gene, HMGI-C, in benign mesenchymal tumours. Nat Genet. 1995;10:436–444. [PubMed]
  • Serrano M, Lee H, Chin L, Cordon-Cardo C, Beach D, DePinho RA. Role of the INK4a locus in tumor suppression and cell mortality. Cell. 1996;85:27–37. [PubMed]
  • Sharpless NE, Bardeesy N, Lee KH, Carrasco D, Castrillon DH, Aguirre AJ, Wu EA, Horner JW, DePinho RA. Loss of p16Ink4a with retention of p19Arf predisposes mice to tumorigenesis. Nature. 2001;413:86–91. [PubMed]
  • Sharpless NE, DePinho RA. How stem cells age and why this makes us grow old. Nat Rev Mol Cell Biol. 2007;8:703–713. [PubMed]
  • Shi Y, Chichung Lie D, Taupin P, Nakashima K, Ray J, Yu RT, Gage FH, Evans RM. Expression and function of orphan nuclear receptor TLX in adult neural stem cells. Nature. 2004;427:78–83. [PubMed]
  • Yu F, Yao H, Zhu P, Zhang X, Pan Q, Gong C, Huang Y, Hu X, Su F, Lieberman J, et al. let-7 regulates self renewal and tumorigenicity of breast cancer cells. Cell. 2007;131:1109–1123. [PubMed]
  • Zhou X, Benson KF, Ashar HR, Chada K. Mutation responsible for the mouse pygmy phenotype in the developmentally regulated factor HMGI-C. Nature. 1995;376:771–774. [PubMed]
  • Zindy F, Quelle DE, Roussel MF, Sherr CJ. Expression of the p16(INK4a) tumor suppressor versus other INK4 family members during mouse development and aging. Oncogene. 1997;15:203–211. [PubMed]