p14Arf acts as an antagonist of HMGA2 in senescence of mesenchymal stem cells-implications for benign tumorigenesis.

HMGA2 is a major regulator of benign tumorigenesis from mesenchyme-derived tissues and stem-cell self-renewal. It has been postulated that HMGA2 mediates its critical function by decreasing p16(Ink4a)/p14(Arf) expression and cellular senescence. To repress the oncogenic activity of HMGA2, the lin-28-let-7 axis is thought to increasingly repress the expression of HMGA2 with age. To understand the HMGA2-p14(Arf) -relationship in benign tumorigenesis, we performed a series of experiments on mesenchymal stem-cells, i.e., the proposed cells of origin of lipomas and uterine leiomyomas. The expression of both genes was inversely correlated during senescence in vitro but contrary to the expectations in adipose tissue derived stem cells (ADSCs) stimulation of HMGA2 by FGF1 increased the expression of p14(Arf) . Based on the assumption that in ADSCs p14(Arf) is repressing HMGA2, siRNA silencing of p14(Arf) was performed resulting in a significant upregulation of HMGA2. To see if p14(Arf) can repress HMGA2 by a TP53-dependent mechanism, nutlin-3, a known MDM2 antagonist, was used which not only increased the activity of the senescence, associated markers p21 and beta-galactosidase, but also decreased the expression of HMGA2, suggesting that p14(Arf) indeed influences HMGA2 by a p53-dependent mechanism because nutlin-3 stabilizes p53. Accordingly, the HMGA2 response triggered by serum was reduced by treatment of ADSCs with nutlin-3. As to the interaction between HMGA2 and p14(Arf) in benign tumorigenesis, we propose a model where akin to MSC self-renewal during tissue repair the simultaneous increase of p14(Arf) with HMGA2 ensures genomic stability, whereas in turn p14(Arf) can repress HMGA2 via TP53.

BMP4 increases expression of HMGA2 in mesenchymal stem cells.

BMP4 has been linked to early steps of adipocyte lineage differentiation but only little is known about its corresponding downstream pathways. Herein, we have investigated whether or not the expression of high mobility group protein HMGA2, another protein linked to proliferation and differentiation within the process of adipogenesis, may be influenced by BMP4 signaling in adipose tissue derived stem cells. Compared to FGF1, a strong inducer of HMGA2 in immortalized pre-adipocytes, BMP4 was found moderately to induce the HMGA2 mRNA expression in serum starved adipose tissue derived stem cells and myometrial cells. In contrast, no such activity was noted in canine bone marrow derived mesenchymal stem cells. As to adipocyte lineage differentiation the functions of BMP4 and HMGA2 mechanistically overlap. Thus, we propose that in adipose tissue BMP4 acts in part by activating HMGA2 making this architectural transcription factor one of the major downstream players in that system.

HMGA2 and the p19Arf-TP53-CDKN1A axis: a delicate balance in the growth of uterine leiomyomas.

Pathogenetically, uterine leiomyomas (ULs) can be interpreted as the result of a monoclonal abnormal proliferation of myometrial cells. Oncogene-induced senescence (OIS) is a frequent phenomenon in premalignant lesions that leads to a growth arrest mainly by the activation of two potent growth-inhibitory pathways as represented by p16(Ink4a) and p19(Arf). The relevance of OIS for the development of UL has not been addressed, but HMGA2, encoded by a major target gene of recurrent chromosomal abnormalities in UL, has been implicated in the repression of the Ink4a/Arf (CDKN2A) locus. This prompted us to examine if HMGA2 contributes to the growth of leiomyomas by repressing this locus. Contrary to the expectations, we were able to show that generally ULs express significantly higher levels of p19(Arf) mRNA than myometrium and that UL with 12q14 approximately 15 rearrangements showed higher expression levels than UL with other cytogenetic aberrations. Furthermore, the finding of a significant correlation between the expressions of p19(Arf) and CDKN1A shows that p19(Arf) triggers senescence rather than apoptosis in UL. Furthermore, the expression levels of HMGA2, p19(Arf), and CDKN1A were found to be correlated with the size of the tumors, indicating that an enhanced growth potential is counterbalanced by the p19(Arf) pathway. Mechanistically, the UL may thus execute a program already present in their cell of origin, where it is activated to protect the genome, for example, in the case of enhanced proliferation. In summary, the results identify the p19(Arf)-TP53-CDKN1A pathway as a major player in the growth control and genomic stability of uterine fibroids.

Germ line mutations of the HMGA2 gene are rare among the general population.

Proteins encoded by the HMGA family are architectural transcription factors, which induce conformational changes in the DNA and thus influence gene expression. Despite the obvious association of the expression of high mobility group protein genes with human cancer, very little is known about the variation of the HMGA proteins within human populations. Therefore, the coding regions of HMGA2 from 87 normal healthy donors were sequenced with the aim of detecting single nucleotide polymorphisms. There was only one sequence divergence leading to an amino acid change in coding regions of HMGA2. Thus, HMGA2 is not only well conserved between species but there is also a high intra-individual conservation of HMGA2, further supporting the important role of HMGA proteins in cellular processes. This analysis clearly demonstrates that as a rule, germ line mutations of HMGA2 are not the cause for benign tumors, e.g. uterine leiomyomas, or human malignant solid tumors.

Expression patterns of the LPP-HMGA2 fusion transcript in pulmonary chondroid hamartomas with t(3;12)(q27 approximately 28;q14 approximately 15).

The high frequency of the t(3;12)(q27 approximately 28; q14 approximately 15) in lipomas and pulmonary chondroid hamartomas (PCHs) makes the HMGA2-LPP fusion gene the most frequent fusion gene in human tumors. We analyzed 11 PCHs with a t(3;12)(q27 approximately 28;q14 approximately 15) for the expression of the LPP-HMGA2 fusion transcript. In a previous study, all of these tumors were shown to express the HMGA2-LPP fusion transcript, composed of exons 1-3 of HMGA2 and exons 9-11 of LPP. In the present study, reverse transcriptase-polymerase chain reaction revealed the expression of the reciprocal fusion transcript in 8 of 11 cases. In all positive tumors, the reciprocal fusion transcripts had the same structure, namely, exons 1-8 of LPP and exons 4-5 of HMGA2 encoding a protein composed of the proline-rich region and the first LIM-domain of LPP and the acidic tail of HMGA2. To our knowledge, this is the first report of the expression of the LPP-HMGA2 fusion transcript in a series of PCHs. Its frequent occurrence in PCHs indicates the absence of a larger deletion of the LPP locus accompanying the translocation, such as has been described in a lipoma. Thus, based on this one finding, a role of LPP-HMGA2 in PCH should be considered.

A microRNA encoded in a highly conserved part of the mammalian HMGA2 gene.

The high mobility group protein HMGA2 plays an important role as a chromatin component of stem cells and as a protein causally related to the development of a variety of benign tumors (e.g., uterine leiomyomas, lipomas, and pleomorphic adenomas of the salivary glands). Herein, the existence of a highly conserved region within intron 3 of HMGA2 encoding a microRNA is described. The co-expression with HMGA2 suggests that as an intronic microRNA, this microRNA may cooperate with HMGA2 in its physiological and/or aberrant functions.