High mobility group A1 (HMGA1) is highly expressed during embryogenesis and in poorly differentiated cancers, and high levels portend a poor prognosis in some tumors. HMGA1 induces oncogenic transformation in cultured cells and causes aggressive cancers in transgenic mice, whereas blocking it interferes with transformation in experimental models. These findings suggest a pivotal role for HMGA1 in cancer. This review focuses on two recently described HMGA1 transcriptional targets that mediate inflammatory signals and drive malignant transformation because they could serve as biomarkers or therapeutic targets. Further elucidation of HMGA1 function in transformation promises to have a major impact on our war on cancer. Cancer Res; 70(2); 436–9
The high mobility group A1 (HMGA1, formerly HMG-I/Y) proteins participate in diverse, fundamental cellular processes, including transcriptional regulation, embryogenesis, transformation, cell cycle regulation, differentiation, senescence, viral integration, and DNA repair (1–22). They are members of a superfamily of nonhistone chromatin-binding proteins whose name reflects their rapid electrophoretic mobility in polyacrilamide gels (high mobility group). All HMG proteins share an acidic carboxyl terminus and associate with chromatin, but are distinguished by unique functional motifs that confer distinct DNA binding domains and biologic activities (1–3). The AT-hook DNA binding motif defines the HMGA family, which consists of HMGA1 and HMGA2. Here, we focus on the HMGA1 subfamily, comprised of HMGA1a and HMGA1b protein isoforms. These isoforms result from alternatively spliced mRNA and differ by 11 internal amino acids present only in HMGA1a (4). Three AT-hook domains mediate binding to the minor groove of chromosomal DNA at AT-rich regions, which enables HMGA1 to recruit additional transcription factors and alter chromatin structure as an essential component of a higher order transcriptional complex or “enhanceosome” (3). In concert with other factors, HMGA1 modulates gene expression. Although first described in the context of the interferon-β promoter, the enhanceosome function likely applies to many promoters. HMGA1 proteins also displace histones and relieve histone-mediated repression of transcription, thereby globally facilitating transcription (5). Because HMGA1 proteins induce conformational changes and alter gene expression, they are called “architectural transcription factors.” Although their role in transcription is well established, the repertoire of genes targeted by HMGA1 is only beginning to emerge.
The first evidence linking HMGA1 proteins to cancer was their discovery in cervical carcinoma cells 25 years ago (1). Subsequent studies showed high HMGA1 expression in rapidly dividing cells, neoplastic cells (1–18), and during embryogenesis (6), but not in differentiated, adult tissues. Their evolutionary conservation and widespread occurrence in development suggested an important biologic function. HMGA1 is induced in quiescent fibroblasts by serum or growth factors and displays delayed-early response kinetics (7). Increasing evidence shows that HMGA1 is widely overexpressed in aggressive malignancies arising from diverse tissues of different embryonic origins. Although the list is expanding, these include cancers of the thyroid, head and neck, colon, lung, breast, pancreas, hematopoietic system, cervix, uterine corpus, prostate, and central nervous system. Moreover, large scale gene expression studies show that high expression portends a poor prognosis in some tumors (8, 18). HMGA1 is also enriched in embryonic stem cells and high-grade (poorly differentiated) cancers, including breast, bladder, and brain cancers (18). Tissue microarray studies also show that increasing protein levels correlate with poor differentiation status and more advanced disease in diverse cancers. These findings indicate that HMGA1 could be a useful biomarker and therapeutic target for high-grade cancers and suggest that HMGA1 orchestrates the molecular underpinnings of a more primitive, poorly differentiated state, both in cancer and development.
HMGA1 Function in Cancer
Following the discovery that HMGA1a is linked to cancer, investigators began to look for its function in transformation. Our laboratory provided early evidence that HMGA1 contributes directly to neoplastic transformation (9–11). Forced overexpression of HMGA1a or HMGA1b induces a transformed phenotype with anchorage-independent cell growth and xenograft tumorigenesis in immortalized cells. HMGA1a is also directly activated by the c-Myc proto-oncoprotein and up-regulated in Burkitt's lymphoma, an aggressive tumor with constitutive c-myc expression. Moreover, inhibiting HMGA1 expression in Burkitt's and other cancer cells blocks anchorage-independent cell growth and proliferation. Knock-down of HMGA1 also causes apoptosis in lung, breast, and thyroid cancer cells, indicating its central role in diverse malignancies (12). In MCF-7 breast cancer cells, HMGA1b induces an epithelial-to-mesenchymal transition (13), further linking these proteins to a primitive, de-differentiated state in cancer.
The subsequent development of mouse models misexpressing HMGA1 provided the most compelling evidence that HMGA1 causes cancer. As reported here 5 years ago, HMGA1a transgenic mice succumb to aggressive T-cell lymphoid malignancy by 2 to 10 months (11). The females also develop uterine sarcomas (15). In this model, murine HMGA1a is driven by the H-2K promoter and immunoglobulin μ enhancer with expression in major histocompatibility complex class I-expressing cells and B cells (11). Another group reported a transgenic model with murine HMGA1b driven by a CMV promoter, and these mice develop natural killer lymphomas and pituitary adenomas (14). These tumors occur later and with a decreased penetrance compared with the HMGA1a mouse model. The variation in phenotypes could relate to the different isoforms expressed or to different levels of tissue-specific transgene expression. More recently, HMGA1 knock-out mice were made and both homozygous and heterozygous null mice develop a decrease in T cells with a relative increase in B cells, erythroid cells, and myeloid cells, and a phenotype described as lympho-myeloproliferative disease (19). The proliferative disease progresses to frank malignancy in some cases, which led to the speculation that HMGA1 has tumor suppressor activity. The hematologic findings also resemble lymphoproliferative disease seen in humans with T-cell immunosuppression, suggesting that T-cell deficiency is the cause of the hematologic abnormalities.
Despite recent progress, the molecular mechanisms that mediate HMGA1 function are only beginning to emerge. Its diverse activities have been attributed to its role as a chromatin remodeling transcription factor, which could lead to induction of different molecular pathways depending on the cellular context. Accordingly, promoter analyses and gene expression profile studies have uncovered downstream gene targets and an HMGA1 “oncogenic transcriptome” is emerging. These studies suggest that HMGA1 orchestrates diverse cellular processes through modulating gene expression, including pathways involved in inflammation, proliferation, transformation, metastatic progression, angiogenesis, and DNA repair (Fig. 1). Although only about 50 HMGA1 transcriptional targets have been reported, most include nuclear factor-κB (NF-κB) regulatory elements in their promoter regions, and many participate in mediating inflammatory pathways. Given the increasing evidence that inflammation is a precursor lesion in some cancers, the link between HMGA1 and NF-κB suggests that these proteins cooperate to induce inflammatory signals and drive transformation. Here, we highlight two recently identified gene targets related to inflammation and oncogenesis because their functions could be blocked in cancer therapy.
HMGA1a in Cancer: New Insights
Cyclooxygenase-2 (COX-2) was originally identified as a putative HMGA1 target gene in hypoxic vascular endothelial cells (20). Under low oxygen tension, HMGA1 binds to an AT-rich region near an NF-κB binding site in the COX-2 promoter in vitro and up-regulates COX-2 expression in transfection experiments. We also found that HMGA1 activates COX-2 expression during tumorigenesis (15, 16). Uterine sarcomas from the HMGA1a transgenics have increased COX-2 expression, and treatment with sulindac (a COX-1/COX-2 inhibitor) inhibits tumor growth (15). Both HMGA1 and COX-2 are also overexpressed in leiomyosarcomas, a highly lethal, human uterine sarcoma. Sulindac or celecoxib (a more selective COX-2 inhibitor) results in blockade of anchorage-independent cell growth and xenograft tumorigenesis in human uterine sarcoma cells, but only in cells with high HMGA1a levels (16). This finding indicates that targeting COX-2 is effective in selected tumors with dysregulation of the HMGA1-COX-2 pathway. COX-2 is induced by inflammatory signals, and, like HMGA1, is thought to elicit a number of cellular pathways involved in neoplastic transformation, including proliferation, angiogenesis, metastasis, and inhibition of apoptosis. It is not clear if hypoxia is a prerequisite for COX-2 induction by HMGA1, and further studies are needed to elucidate the role of this pathway in other cancers.
Signal transducer and activator of transcription 3
Recent studies also uncovered signal transducer and activator of transcription 3 (STAT3) as a critical downstream target of HMGA1 (17). STAT3 is up-regulated in cells and transgenic mice engineered to overexpress HMGA1a. Similar to the COX-2 promoter, HMGA1 binds to an AT-rich region near an NF-κB binding site and activates STAT3 expression. HMGA1 occupies this promoter in cultured cells from hematopoietic malignancies. Moreover, a STAT3 small molecule inhibitor induces apoptosis in leukemic cells from HMGA1 transgenics, but not in normal lymphoid cells, indicating that this pathway could be a rational therapeutic target in some malignancies. In ongoing studies, we are testing additional STAT3 small molecule inhibitors in our transgenic models in vivo, and preliminary data show decreases in the tumor burdens.1
1Resar L.M.S. Unpublished data.
Although recent studies have identified downstream gene targets, it is likely that these results provide only a glimpse of HMGA1 transcriptional networks in cancer (Fig. 1). A global investigation of both mRNAs and microRNAs should lead to the discovery of rational therapeutic targets relevant to diverse cancers. Interestingly, both COX-2 and STAT3 are important transducers of inflammatory signals with NF-κB regulatory elements in their promoter regions near the HMGA1 binding sites. The presence of NF-κB sites in HMGA1 transcriptional targets suggests that these proteins function together as reported in the regulation of interferon-β expression. Although not directly proven, HMGA1, together with NF-κB and other transcription factors, could form an enhanceosome that constitutes a major “hub” activated by inflammation. This hub, in turn, could induce cellular pathways that lead to cancer initiation and progression. In some settings, overexpression of HMGA1 could activate inflammatory pathways and circumvent the need for inflammation to promote oncogenic transformation. Our preliminary data from gene expression analysis support the role of HMGA1 in inducing multiple inflammatory pathways in transformation (Resar L.M.S. Unpublished data). Further dissection of these pathways will reveal how HMGA1 promotes transformation and tumor progression, and could identify rational interventions to treat, or even prevent, cancer. Our studies have focused on pathways downstream of HMGA1 because the ability to target nuclear oncogenes therapeutically is only emerging. Additional studies to identify molecular circuitry induced by HMGA1 as well as functional analysis in preclinical mouse models are needed before these findings can be translated to the clinic.
Although it is clear that growth factor signaling, oncogenic transcription factors, viruses, hypoxia, and inflammatory signals induce HMGA1 expression, the molecular mechanisms mediating HMGA1 overexpression in most tumors is unknown. Amplification has not been detected to date. Epigenetic changes or microRNA alterations could play a role and studies to uncover the molecular mechanisms inducing HMGA1 should advance our ability to block its activity in cancer therapy. Recent studies also implicate HMGA1 in maintaining “stemness” and understanding its role in this setting should elucidate therapeutic targets relevant to cancer and cancer stem cells (18). Emerging evidence also indicate that HMGA proteins participate in cellular senescence (21), and this activity could potentially be exploited in therapy. Finally, HMGA1 activity in DNA repair could play a central role in transformation and is worthy of further study (22). It is likely that there are, as yet, undiscovered HMGA1 cellular activities that will provide insight for novel therapeutic approaches in cancer, given their importance in embryogenesis and cancer. Further investigation of HMGA1 proteins in development and malignant transformation promises to bolster our war on cancer-clearly a war worth waging.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
This work is dedicated to Dr. Donald Coffey, the consummate teacher, mentor, and scientist.
Grant Support: Grants R01 CA092339, R03 CA139331, R21 CA18343, and P50 CA096888 from the National Institutes of Health, the Leukemia & Lymphoma Scholar Award, Alex's Lemonade Stand Foundation Innovation Award, and the J.P. McCarthy Foundation.