HAF: the new player in oxygen-independent HIF-1α degradation

Abstract

Adaptation to hypoxia is primarily mediated by the hypoxia-inducible transcription factor, HIF. The regulation of HIF activity by the oxygen-dependent degradation of the HIF-1α and HIF-2α subunits by the pVHL E3 ligase complex has been well characterized. We have recently described the hypoxia-associated factor, HAF, as an E3 ligase for HIF-1α that does not degrade HIF-2α. Here we summarize the mechanism of HAF-mediated HIF-1α degradation and the importance of oxygen-independent HIF-1α regulation in cancer. We also discuss the implications of the new HAF: HIF-1α degradation pathway with respect to other novel mediators of oxygen-independent HIF-α degradation. Finally, we review the significance of HAF as an isoform-specific E3 ligase in light of new information on the non-overlapping functions of HIF-1α and HIF-2α in cancer.

Introduction

Hypoxia is a state of reduced oxygen availability that restricts the function of organs, tissues, or cells in the adult mammal ¹. However, hypoxia plays an important and beneficial role in mammalian physiology; and is critical for proper embryogenesis. Many solid tumors also contain hypoxic regions, due to the inability of the local vasculature to supply sufficient oxygen to the rapidly growing tumor and because of the severe structural abnormality of tumor microvessels ². Tumor hypoxia is of major clinical significance since it can promote both tumor progression, and resistance to radiation and chemotherapy ³.

The cellular adaptation to hypoxia includes the induction of angiogenesis, a switch from aerobic metabolism to anaerobic glycolysis, and the expression of a variety of stress proteins regulating cell death or survival. The hypoxia-inducible factor (HIF) is recognized as the master regulator of the hypoxia response, activating the transcription of a large number of genes critical for the adaptation to hypoxia ⁴. HIF also contributes to tumor growth and its increased expression has been correlated with poor patient prognosis ⁵. The involvement of HIF in pathophysiological conditions such as ischemia and cancer and its value as a therapeutic target have generated considerable interest in the understanding of HIF regulation ⁶.

The HIF transcription factor is a heterodimer of the oxygen-regulated HIF-α subunit, and the constitutively expressed HIF-1β subunit (also known as the aryl hydrocarbon receptor nuclear translocator (ARNT) ⁷. Under aerobic conditions, HIF-α is degraded by the von Hippel Lindau protein (pVHL) through the ubiquitin-proteasomal pathway. Under hypoxic conditions, HIF-α is stabilized, enters the nucleus and heterodimerizes with HIF-1β and binds to a conserved DNA sequence known as the hypoxia responsive element (HRE) to transactivate a variety of hypoxia-responsive genes ⁸. Recruitment of the coactivator p300/CBP to the C-terminal transactivation (CTAD) domain of HIF-α, is required for transcription of HIF downstream targets. Under normoxic conditions, the ability of HIF-α to activate transcription is further prevented by factor inhibiting HIF (FIH), which hydroxylates an asparagine residue within the C-TAD of HIF-α, hence disrupting its interaction with p300/CBP ⁹.

To date, three HIF-α isoforms (HIF-1/2/3α) have been described (Figure 1), of which HIF-1α and HIF-2α are the best characterized. HIF-1α is expressed ubiquitously whereas HIF-2α displays more tissue-specific expression ¹⁰. Although HIF-1α and HIF-2α display 48% amino acid identity and have similar protein structure and mechanism of regulation, they regulate both common and unique target genes ^{11, 12}, and may be differentially regulated depending on the duration and severity of hypoxia exposure ¹³. The existence of distinct HIF-1α/HIF-2α downstream targets can be at least partly explained by their different affinities for specific regulators of HIF-1/2α activity and stability such as the prolyl hydroxylases (described in greater detail below) and FIH , and differential recruitment of additional isoform-specific transcriptional co-activators or the repressors ^{12, 14, 15}. HIF-1α^-/- mice exhibit midgestation lethality with severe developmental defects such as in the neural tubes and cardiovascular malformations ^{16, 17}. HIF-2α^-/- mice manifest a variety of defects depending on the study, possibly due to the different genetic backgrounds of the mice studied, including embryonic lethality due to defective vascular remodeling during development ¹⁸, embryonic lethality due to impaired cathecholamine synthesis ¹⁹, and death either midgestation or soon after birth from severe respiratory failure due to insufficient surfactant production ²⁰. In cancer, while both HIF-α isoforms have been associated with poor patient prognosis, differences in their target genes has led to a proposed association of HIF-1α with the regulation of glycolysis, and HIF-2α with increased proliferation and tumor metastasis ^21-23.

Figure 1. Domain structures of HIF-1α, HIF-2α, HIF-3α and HIF-1β/ARNT.

The HIF-α and HIF-1β/ARNT subunits are basic helix-loop-helix/Per-ARNT-Sim homology (bHLH/PAS) transcription factors. HIF-1α and HIF-2α contain two transactivation (TAD) domains, one at the amino terminal and the other at the carboxy terminal. Six splice versions of HIF-3α have been identified (only two are depicted here) and those that have been mapped contain only an N-TAD. HIF-1β/ARNT, the constitutive member of the HIF heterodimer contains a C-TAD, which is not required for HIF transcriptional activity.

In addition to altering cellular energy metabolism and angiogenesis, hypoxia also influences the proliferation and differentiation of various stem/progenitor cell populations. An emerging theme relating to HIF and cancer has been the implication of HIF in the maintenance of cancer stem cells (or tumor initiating/progenitor cells) ²⁴. Cancer stem cells are defined as a specific minority of tumor cells which are, like their normal counterparts, endowed with extensive potential for self-renewal and for generating the diverse tumor cell population ^{25, 26}. Quiescent cancer stem cells are believed to be more resistant to chemotherapy and to be the main source of tumor recurrence ²⁷. Furthermore, low stages of differentiation in certain tumor types that can be induced by hypoxia, are associated with poor outcome and disseminated disease ^{28, 29}. In this respect, HIF-1α has been shown to interact with and stabilize the intracellular domain (ICD) of Notch, a factor important for the maintenance of the stem/progenitor cell state, and to activate transcription of Notch downstream targets such as Hey-2 and Hes-1 ³⁰. The hypoxia-induced activation of the Notch pathway was also observed in A-498 renal carcinoma cells that exclusively express HIF-2α ³¹, suggesting that HIF-2α may play a similar role as HIF-1α in Notch activation ³⁰. On the other hand, HIF-2α but not HIF-1α, selectively induces the expression of OCT-4, an important transcription factor for the maintenance of the stem cell phenotype and embryonic pluripotency ³². The targeted replacement of HIF-1α by a HIF-2α knock-in allele in murine embryonic stem cells generated teratomas with increased growth and an increased population of undifferentiated cells compared to their controls through a mechanism dependent upon the HIF-2α-induced activation of OCT-4 ^{32, 33}.

Oxygen-dependent HIF-α degradation

Under aerobic conditions, HIF-1/2α is hydroxylated by specific prolyl hydroxylases (PHD1, PHD2, PHD3 and, recently, PHD4) at conserved proline residues within its oxygen-dependent degradation (ODD) domain in a reaction requiring molecular oxygen, 2-oxoglutarate, and ascorbate ^{34, 35}. While PHD2 is a primary hydroxylase for both HIF-1α and HIF-2α ³⁶, genetic studies suggest that PHD3 is mainly responsible for HIF-2α hydroxylation ³⁷. Hydroxylated HIF-α is recognized by pVHL, which forms the substrate recognition module of an E3 ubiquitin ligase complex comprising elongin C, elongin B, cullin-2, and ring-box 1 (Rbx1), which directs HIF-α poly-ubiquitination and proteasomal degradation ^{38, 39}. The central role of pVHL in HIF-α regulation is manifested in VHL disease where the inactivation of the VHL gene results in inappropriate overexpression of HIF-target genes including those involved in angiogenesis and invasion, and the development of highly vascularized tumors in the kidney, retina and central nervous system ⁴⁰.

The oxygen-dependent degradation of HIF-1α by pVHL can also be mediated by hypoxia-induced HIF-1α small ubiquitin-like modifier (SUMO)ylation. In this regard, hypoxia induces the SUMOylation of HIF-1α through an unknown mechanism, which allows the binding of the pVHL E3 ligase complex to HIF-1α without requiring HIF-1α prolyl hydroxylation, leading to HIF-1α degradation ⁴¹. The pVHL-mediated degradation of SUMOylated HIF-1α in hypoxia is prevented by SENP1 (SUMO1/sentrin specific peptidase 1), a nuclear SUMO protease that deconjugates SUMOylated HIF-1α, allowing it to escape pVHL-mediated degradation. Further work will be necessary to demonstrate whether HIF-2α is subject to the same SUMOylation/deSUMOylation regulation as HIF-1α.

Oxygen-independent HIF-α degradation

While pVHL has been well established as the key player in the regulation of the oxygen-dependent degradation of HIF-1α, there has been increasing evidence of pVHL-independent pathways of HIF-1α degradation ⁴². Indeed, although primarily regulated by changes in cellular oxygen tension, it has now become clear that HIF-1α may be also activated by a wide range of growth-promoting stimuli and oncogenic pathways such as insulin, insulin-like growth factor-1, epidermal growth factor (EGF) and mutant Ras, Src and/or PI-3K pathways ^43-46. This is particularly important in cancer where deregulation of growth factors and/or their cognate receptors can result in constitutive activation of HIF-1α. The main mechanism implicated for HIF-1α activation by the PI-3K/mTOR pathway is an increase in HIF-1α protein translation, without a change in HIF-1α degradation ^{9, 47, 48}. The increase in HIF-1α translation appears sufficient to shift the balance between synthesis and degradation towards an accumulation of HIF-1α in normoxia. Additionally, oncogene activation such as by Ras mutation or by Src activation results in increased HIF-1α activity through p42/p44 MAPK phosphorylation ^{49, 50}.

Although HIF-1α promotes the cellular adaptation to hypoxia, continuous HIF-1α activation has, under certain settings, been reported to inhibit cell-cycle progression ^{51, 52} and to promote apoptosis ⁵³. It is thus likely that other mechanisms contribute to the maintenance of HIF activity at a level conducive for continued cell growth and survival. We propose that HIF activity requires tight regulation by both oxygen-dependent and oxygen-independent mechanisms to cope with diverse growth conditions and cellular stimuli.

Compared to the pVHL pathway, the regulation of these new oxygen-independent pathways may be less affected by oxygen availability but more dependent upon specific cellular conditions such as calcium or the presence of growth factors. pVHL-and oxygen-independent degradation of HIF-1α has been observed in overexpression studies, such as by the overexpression of GSK3 or FOXO4; or in drug studies, such as by treatment with histone deactylase (HDAC) inhibitors or heat shock protein 90 (Hsp90) inhibitors ^54-57. Additionally, HIF-1α has been shown to be subject to transcriptional-dependent proteasomal degradation, independent of the PHDs and pVHL, which may limit the accumulation of HIF-1α under severe hypoxia ⁵⁸. Furthermore, constitutive, ubiquitin-independent degradation of HIF-1α and HIF-2α via the proteasomal pathway has also been proposed ⁵⁹. We have recently identified a new oxygen-independent modulator of HIF-1α degradation termed the Hypoxia-Associated Factor (HAF) ⁶⁰. Here we will discuss the role of HAF as an important new mediator of HIF-1α degradation and the potential ramifications of oxygen-independent HIF-α regulation and of modulation of the balance between HIF-1α and HIF-2α activity

HAF mediates HIF-1α-specific oxygen-independent degradation

HAF, also known as SART1₈₀₀ (squamous cell carcinoma antigen recognized by T-cells), was originally identified as a nuclear protein expressed in proliferating cells ⁶¹. HAF is a component of the U4/U6•U5 tri-snRNP (small nuclear ribonucleoprotein) complex and is essential for the assembly of mature spliceosomes ⁶². The murine homolog of SART1₈₀₀, HAF, was given its name due to its ability to bind to the promoter of the erythropoietin (EPO) gene and to contribute to its hypoxia inducibility ⁶³. In our study, we showed that HAF promotes HIF-1α degradation independently of cellular oxygen tension ⁶⁴. HAF overexpression decreased HIF-1α levels, whereas HAF knockdown increased HIF-1α levels independently of pVHL or oxygen. Intriguingly, neither HAF knockdown nor overexpression caused any change in the levels of HIF-2α. This suggests that HAF is a novel HIF-1α isoform specific E3 ligase. Further characterization showed that HAF bound to HIF-1α causing HIF-1α ubiquitination and proteasomal degradation in an oxygen- and pVHL-independent manner. In this respect, we found that recombinant HAF ubiquitinated non-hydroxylated recombinant HIF-1α in in vitro ubiquitination assays in an E2-dependent manner. These studies demonstrated that HAF is a bona fide E3 ligase and not a member of a larger multiprotein E3 ligase complex like pVHL. Hence, our study identified HAF as a novel E3 ligase for HIF-1α and the mediator of a new mechanism of HIF-1α degradation that is independent of pVHL and oxygen.

Consistent with HAF’s ability to degrade HIF-1α irrespective of cellular oxygen tension, we found that HAF knockdown using siRNA increased endogenous levels of HIF-1α in normoxia, hypoxia and under EGF stimulation in a panel of cell lines. Furthermore, HAF siRNA increased the half life of HIF-1α in pVHL-deficient RCC cells, whereas HAF overexpression decreased the half life of hypoxia-induced HIF-1α in HT29 cells expressing wild-type pVHL. In contrast, knockdown of pVHL using siRNA increased the levels of HIF-1α only under normoxic conditions but did not affect HIF-lα levels induced by hypoxia or EGF. The effect of HAF under a variety of conditions is an important demonstration of a critical role for HAF in the regulation of HIF-1α turnover/degradation beyond that of pVHL, and that it is functional under conditions in which pVHL is inactive or insufficient, such as during hypoxia or in cells with elevated HIF-1α levels due to growth factor activation. Hence, HAF mediates an oxygen-independent HIF-1α degradation pathway that is complementary to that of the oxygen-dependent pathway mediated by pVHL, providing an additional layer of control that allows the precise regulation of HIF-1α levels under diverse conditions. Additionally, the unique ability of HAF to discriminate between HIF-1α and HIF-2α for degradation provides a novel mechanism for a switch from HIF-1α to HIF-2α signaling. This is significant in view of recent findings on the non-redundancy of HIF-1α and HIF-2α target genes and may facilitate tumor progression to a more aggressive, undifferentiated phenotype.

The role of HAF as a multifunctional protein

The N-terminus of HAF has previously been shown to have DNA binding activity and to induce the transcription of EPO and VEGF ⁶³. This indicates that HAF may be a modular protein with DNA binding activity within its N-terminus and E3 ligase activity within its C-terminus, in addition to its previously described role in pre-mRNA splicing. The dual-functionality of DNA binding and E3 ligase activity is not unique to HAF but has also been demonstrated by other key regulatory proteins such as BRCA1, which can bind DNA independently of its E3 ubiquitin ligase domain ⁶⁵, and p300, which acts as both a transcriptional coactivator and ubiquitin ligase for p53 ⁶⁶. The multi-functionality of these proteins may be indicative of their significance in specific cellular processes where mutation or loss may be highly detrimental and potentially oncogenic. Indeed, HAF has been identified in a genomic RNAi screen as a gene essential for cell division, the silencing of which results in mitotic arrest ⁶⁷. However, the dual functionality of HAF appears contradictory in that it inhibits HIF-1α signaling by degradation, and yet promotes the transcription of some of its key downstream targets. In this regard, HAF may provide a measure of selectivity by promoting the activation of only a subset of HIF downstream targets while degrading HIF-1α. Additionally, HAF transcriptional activity may contribute to HIF-2α mediated transcription. Hence, the dual-functionality of HAF may provide avenues for regulation by varied inputs or opportunities for therapeutic modulation.

Our mapping studies have shown that the C-terminus of HAF is responsible for both the HAF E3 ligase activity and its binding to HIF-1α. Specifically, we found that the C-terminus of HAF (amino acids 654-800) bound to HIF-1α within amino acids 296-400, a region flanking the HIF-1α ODD domain. Intriguingly, HAF 654-800 did not contain a proto-typical E3 ligase motif (HECT, RING or U-Box), whereas mutational analyses showed that HAF did not require catalytic cysteines for its ubiquitin ligase activity. This suggests that HAF is one of a growing list of atypical E3 ligases ^{68, 69}. Moreover, we observed that overexpressed HAF is mono-ubiquitinated which, unlike poly-ubiquitination, does not serve as a proteasome targeting signal. Mono-ubiquitination has also been reported in tumor suppressors such as p53, Forkhead Box (FOXO) and phosphatase and tensin homolog (PTEN) and can be induced by certain cellular stimuli (such as oxidative stress in the case of FOXO), and can affect substrate protein activity, stability, cellular localization or interaction with other proteins ⁷⁰. Additionally, HAF is also a target for SUMOylation ⁷¹. Further work will reveal whether the post-translational modification of HAF by mono-ubiquitination and/or SUMOylation by certain physiological stimuli plays a role in regulating HAF-mediated HIF-1α degradation.

The role of HAF in cancer

HAF has been detected in all cell lines (both normal and tumor-derived) and in proliferating tissues but is undetectable in normal, non-proliferating tissue ⁶¹. Additionally, HAF has been detected at higher levels in some tumors compared with the surrounding, non-proliferating normal tissue ^{72, 73}. As tumors are almost invariably highly proliferative, it is difficult to determine whether HAF expression plays a significant role in tumorigenesis, or is simply a reflection of the increased growth rate of tumor cells. In our study, we found that HAF overexpression did not affect cell proliferation in vitro. However, HAF overexpression was associated with marked inhibition of xenograft tumor growth in vivo, decreased HIF-1α levels and transactivation, and reduced tumor angiogenesis ⁶⁴. This is consistent with a HIF-1α dependent effect on proliferation as much of the effect of HIF-1α on tumor growth can be attributed to VEGF inhibition, which is not required for cell growth in vitro. This suggests that under certain conditions, HAF, by degrading HIF-1α, may act as a tumor suppressor, but this appears to conflict with the observations of increased HAF expression in cancer. Intriguingly, although HIF-1α has been associated with aggressive tumors and poor patient prognosis, high levels of HIF-1α in some cases such as in renal cell carcinoma (RCC), has been related to increased patient survival ⁷⁴. Additionally, overexpression of the HIF-2α subunit has been shown to enhance the growth of pVHL-deficient RCC xenografts, whereas overexpression of the HIF-1α subunit has been shown to retard xenograft growth ⁷⁵. This suggests that at least in some settings, HIF-1α and HIF-2α have opposing roles in tumorigenesis whereby HIF-1α inhibits, whereas HIF-2α promotes tumor growth. In support of this notion, it has been reported that HIF-1α expression gradually decreases, whereas HIF-2α expression increases as RCCs develop ⁷⁶. This trend has also been observed in colorectal cancer where tumor aggressiveness has been significantly correlated with overexpression of HIF-2α but not with HIF-1α ⁷⁷, and in neuroblastoma where high levels of HIF-2α have been proposed to promote an aggressive tumor phenotype ¹³. Nevertheless, although the precise pro/anti-tumorigenic properties of HIF-1α in comparison to HIF-2α are still under debate ⁷⁸, they do highlight the complex interaction between HIF-1α and HIF-2α and their different roles in tumorigenesis. Hence, we propose that HAF promotes tumor progression in cells expressing both HIF-1α and HIF-2α isoforms by degrading HIF-1α whilst allowing HIF-2α to activate downstream transcription (Figure 2). This may facilitate the propensity of HIF-2α to activate specific pro-tumorigenic factors such as OCT-4 and c-Myc that are not activated, or even antagonized (in the case of c-Myc) by HIF-1α ^{22, 23}. Similarly, the selection of the HIF-2α in favor of HIF-1α by HAF may allow the hypoxic adaptation of tumor cells without the activation of the HIF-1α specific downstream targets such as BNIP3 and BNIP3L that may promote apoptosis. On the other hand, although HIF-2α has received increased attention as a potent pro-tumorigenic factor, the case for HIF-1α as an indicator of morbidity has also been well established ^79-81. Indeed, the pro-apoptotic activity of HIF-1α through BNIP3 and BNIP3L may only be manifest under conditions of severe hypoxia, necrosis and/or acidosis and these may be counteracted by a number of other anti-apoptotic factors induced by HIF-1α ⁸². Hence, the differential outcomes of HIF-1α versus HIF-2α activation in tumors are certainly not clear cut. Consequently, in addition to promoting tumor growth in cells expressing both HIF-α isoforms by facilitating HIF-2α signaling, we propose that HAF may play more of a tumor suppressor role in tumors in which HIF-1α is the driving factor.

Figure 2. HAF, the new player in HIF-1α degradation.

Under aerobic conditions, the hydroxylation of both HIF-1α and HIF-2α by prolyl hydroxylases facilitates the binding of the pVHL E3 ligase complex (pVHL E3L) to HIF-1α/HIF-2α. This results in the poly-ubiquitination of HIF-1α/HIF-2α and their proteasomal degradation. Under hypoxic conditions, HIF-1α/HIF-α2 are stabilized and enter the nucleus where they heterodimerize with HIF-1β and initiate downstream transcription. The presence of growth factors or specific extracellular stimuli may promote the binding of HAF to HIF-1α by a yet to be identified oxygen-independent mechanism. This promotes the poly-ubiquitination and degradation of HIF-1α while HIF-2α is allowed remain to activate downstream transcription.

Other oxygen-independent HIF-α regulators

In addition to HAF, other mediators of oxygen-independent HIF-α degradation have been reported. For the most part, although oxygen-dependent HIF-1α degradation has been observed, investigators have been unable to identify the E3 ligase or precise mechanism involved ^{55, 56, 58}. Additionally, the situations under which the constitutive, proteasomal dependent, ubiquitin-independent degradation of HIF-1α occurs remain unclear ^{59, 83}. However, two new mediators of HIF-α degradation — RACK1 and Int6 have been characterized in greater detail and are summarized below.

RACK1

HIF-1α degradation mediated by Hsp90 inhibitors has been attributed to receptor of activated protein kinase C (RACK1) ⁸⁴. RACK1 competes with Hsp90 for binding to the PAS-A domain of HIF-1α. Consequently, inhibition of Hsp90 by Hsp90 inhibitors such as 17-(allylamino)-17-demethoxygeldanamycin (17-AAG), promote the oxygen-independent degradation of HIF-1α through increased binding of RACK1 ⁸⁴. RACK1 binds HIF-1α and recruits elongin C and other components of the E3 ligase complex to HIF-1α, in a manner mechanistically similar to pVHL, leading to HIF-1α ubiquitination and proteasomal degradation. Intriguingly, a RACK1-dependent mechanism has also been implicated in HIF-1α degradation mediated by HDAC inhibitors as treatment with trichostatin A increases Hsp90 acetylation, which destabilizes the HIF-1α-Hsp90 interaction and promotes HIF-1α degradation ⁸⁵. However, the involvement of RACK1 in HDAC inhibitor mediated HIF-1α degradation remains to be demonstrated.

RACK1-mediated HIF-1α degradation can be inhibited by calcium through the activity of calcineurin, a calcium/calmodulin dependent and serine/threonine-specific protein phosphatase ⁸⁶. Hence, increasing intracellular calcium concentration by treatment with ionomycin protects HIF-1α from RACK1-mediated degradation. Paradoxically, the chelation of intracellular calcium in normoxia may cause the accumulation of HIF-1α protein through inhibition of HIF PHDs, hence preventing pVHL-induced HIF-1α degradation ⁸⁷. Thus, modulation of intracellular calcium concentration can affect the stability of HIF-1α protein through both the pVHL and RACK1 pathways and the final outcome may well be context dependent. Certainly the fact that both pVHL and RACK require the E3 ligase complex (Elongin B/C, Cul 2, Rbx1) for HIF-1α ubiquitination and subsequent degradation lends the possibility of co-regulation of both pVHL and RACK pathways by fluctuations in the availability or activities of the various E3 ligase components. In contrast, the HAF C-terminus displays E3 ligase activity without requiring any ancillary factors and so degrades HIF-1α even in the absence of Elongin B/C and hence may be regulated in a manner entirely independently of the pVHL pathway ⁶⁴.

Int6

The murine mammary tumor integration site 6 gene (Int6), also known as EIF3E, has been described as a HIF-2α-specific E3 ligase ⁸⁸. The Int6 gene was first identified from a screen in which the mouse mammary tumor virus (MMTV) was employed as an insertional mutagen to identify genes whose functions are critical for breast tumor formation ⁸⁹. The MMTV insertion into Int6 generates a C-terminal truncated protein (Int6ΔC) that is a dominant negative, transforms cells, which, when injected into nude mice, induces tumor formation ⁹⁰. Int6 interacts specifically with HIF-2α (but not with HIF-1α or HIF-3α) and mediates the proteasomal dependent degradation of HIF-2α in an oxygen- and pVHL-independent manner ⁸⁸. Overexpression of Int6ΔC induces HIF-2α expression even during normoxia. While the mechanism of Int6-induced HIF-2α degradation remains unclear, overexpression of Int6 induces the degradation of endogenous and overexpressed HIF-2α, whereas Int6 knockdown increased the expression of endogenous HIF-2α in even under normoxia. In this regard, although Int6 has no clear homology with the ubiquitin ligase family members, it has been known to associate with the 26S proteasome ⁹¹, and it has thus been suggested that this association between Int6, HIF-2α, the proteasome, and potentially an unknown E3 ligase, promotes HIF-2α degradation ⁸⁸.

The ability of Int6 to selectively degrade HIF-2α resembles the HIF-1α-selective activity of HAF. These isoform-specific mediators of HIF-α degradation may provide a mechanism for the selective recruitment of either HIF-1α or HIF-2α to HIF-1β to form the HIF heterodimer in order to fulfill precise cellular requirements.

Summary

Since the identification of the HIF transcription factor in 1992, the understanding of the regulation of the hypoxic response by HIF has dramatically increased. From the identification and characterization of HIF-1α/HIF2α/HIF-3α and HIF-1β subunits, to the elucidation of pVHL-mediated oxygen dependent regulation of HIF, and the involvement of HIF in normal development and in cancer morbidity, the field of HIF regulation has come a long way. Our recent identification of HAF, a ubiquitin ligase specific for HIF-1α that promotes its oxygen-independent degradation is yet another piece to fit into the complex HIF puzzle. HAF and other recent oxygen-independent modulators of HIF-α degradation highlight an until recently, unexpected role for HIF in non-hypoxic gene regulation and the need for HIF-α isoform-specific degradation. However, the work with HAF raises further questions as to how HAF activity is regulated and in what context it is beneficial to the cell to degrade HIF-1α in favor of HIF-2α. Certainly, it raises the question of the role of HAF in cancer, whether HAF is present due to the high proliferative activity of cancer cells or whether it plays a more active role in tumorigenesis/cancer progression. Further work should provide new insights into the HAF: HIF connection that may be utilized for therapeutic benefit.