TNF receptor-associated protein 1 (TRAP1) was firstly identified by a yeast-based two-hybrid system as a novel protein binding the intracellular domain of the type 1 receptor for tumor necrosis factor (TNFR-1IC), hence its name. The protein immediately showed strong sequence homology with members of the heat shock proteins 90-kDa (HSP90) family and variable expression levels in human skeletal muscle, liver, heart, brain, kidney, pancreas, lung, and placenta and in eight different transformed cell lines (Song et al. 1995). A second, independent yeast two-hybrid screening led to the identification of a new heat shock protein functioning as a molecular chaperone for the retinoblastoma protein (Rb). Antibodies prepared against fusion protein between glutathione S-transferase and domains of this new heat shock protein specifically recognized a 75-kDa cellular protein, therefore designated HSP75. In mammalian cells, Rb formed complexes with HSP75 under two special physiological conditions: during mitosis and after heat shock; moreover, HSP75 showed the ability to refold denatured Rb into its native conformation in vitro (Chen et al. 1996). Further analyses of their sequences revealed that the two studies had actually identified the same protein (Felts et al. 2000). Although none of these two studies led to new insights into the role of this protein in modulating TNFR signaling and its downstream transduction pathways or in regulating Rb in cell cycle progression, especially in response to external stimuli, the identification of this new member of the HSP90 family raised interest into the specificity of its functions. In 2000, Felts and colleagues performed an initial characterization of TRAP1, showing that it is mainly localized in mitochondria, as confirmed by the presence of a mitochondrial targeting sequence within the N-terminal domain. Consistently with its striking homology to HSP90, TRAP1 ATPase activity was inhibited by both geldanamycin and radicicol, two specific inhibitors of HSP90 function; however, it did not bind and fold HSP90 client proteins, nor form stable complexes with the classic HSP90 co-chaperones p23 and Hop, suggesting distinct functional properties (Felts et al. 2000). Contemporarily, immunogold electron microscopy in rat tissue sections showed strong labeling of TRAP1 in mitochondria of all the tissues examined and, additionally, in a number of non-mitochondrial locations in certain tissues (Cechetto and Gupta 2000).
The first insights into TRAP1 function were obtained into 2007, when it was demonstrated that TRAP1 was upregulated in tumor cells and adapted to mild oxidative stress conditions and that such overexpression led to increased resistance to multiple anticancer drugs and to toxic effects of oxidants (Amoroso et al. 2014). At the same time, a comparative immunohistochemistry analysis of primary mouse tumor specimens and their matched normal tissues in vivo showed that TRAP1 was strongly expressed in tumor cells of adenocarcinoma of the pancreas, breast, colon, and lung, whereas normal matched epithelia contained very low levels of this protein (Kang et al. 2007). Moreover, the authors showed that TRAP1 regulates mitochondrial permeability transition in response to apoptotic stimuli, through the interaction with cyclophilin D, an immunophilin that induces mitochondrial cell death (Kang et al. 2007). These authors demonstrated that TRAP1, along with cytosolic HSP90, is mainly located inside the mitochondria of cancerous tissues and can antagonize cyclophilin D function and, strikingly, that disabling this pathway using HSP90 ATPase antagonists directed to mitochondria caused sudden collapse of mitochondrial function and apoptosis. Since these seminal studies, a growing body of evidence raised the interest toward TRAP1 as an important player in cancer biology.
TRAP1: The Gene and the Protein
The human TRAP1 gene spans a distance of almost 60 Kbp on chromosome 16 and is splitted in 18 exons and 17 introns, potentially producing 17 alternative splice variants. The main transcript is 2296-bp-long and codes for a 704-amino acid protein – with a predicted molecular mass of 80.11 kDa – containing a 59 amino acid-long N-terminal presequence that is responsible for targeting to mitochondria and that is putatively cleaved off within the organelle by the mitochondrial processing peptidase. It has been shown that the TRAP1 gene is a MYC target; however, its regulation has never been investigated so far. Beside the abovementioned N-terminal mitochondrial targeting sequence, the protein contains three main domains: (1) a histidine kinase-like ATPase domain, found in several ATP-binding proteins including heat shock protein HSP90; (2) a heat shock protein HSP90 domain, with a highly conserved N-terminal domain separated from a conserved, acidic C-terminal domain by a highly acidic, flexible linker region; (3) a ribosomal protein S5 domain two-type fold that is found in numerous RNA/DNA-binding proteins. TRAP1 is susceptible to different amino acid modifications, including several phosphoserine, phosphothreonine, and acetyllysine. Phosphorylation of TRAP1 by the mitochondrial serine/threonine-protein kinase PINK1 is pivotal for its cytoprotective function within mitochondria (see below). Crystallographic studies (Lavery et al. 2014) have demonstrated that TRAP1 protein adopts a homodimeric, asymmetric conformation that provides the potential for sequential ATP hydrolysis steps to drive both client remodeling and client release.
TRAP1 in Cancer and Other Diseases
Since the first formal demonstration of an asymmetrical distribution of TRAP1 in normal versus malignant tissues (Kang et al. 2007), several studies have addressed this issue, strongly indicating that this chaperone plays a relevant role in cancer biology. TRAP1 has been found dysregulated in several human cancers, but its best characterized contributions concern prostate, colorectal, breast, and ovarian carcinomas, glioblastoma, and non-small cell lung cancer. TRAP1 involvement in other pathological conditions has been successfully demonstrated, with particular emphasis on Parkinson disease and congenital anomalies such as CAKUT with or without VACTERL association.
It has been shown that TRAP1, along with a mitochondria-localized HSP90, is abundantly and ubiquitously expressed in both localized and metastatic prostate cancers, but largely undetectable in normal prostate or benign prostatic hyperplasia in vivo. Its silencing in androgen-independent prostate cancer cells enhanced apoptosis, as well as targeting TRAP1 by mitochondria-directed HSP90 inhibitors (gamitrinibs) selectively caused death of prostate cancer cells (Leav et al. 2010). Transgenic mice expressing TRAP1 in the prostate develop epithelial hyperplasia and cellular atypia, and when examined in a Pten+/− background (a common alteration in human prostate cancer) show accelerated incidence of invasive prostatic adenocarcinoma, whereas deletion of TRAP1 delays prostatic tumorigenesis. Accordingly, global profiling of Pten+/− – TRAP1 transgenic mice by RNA sequencing and reverse phase protein array reveals modulation of oncogenic networks of cell proliferation, apoptosis, cell motility, and DNA damage (Lisanti et al. 2016), further supporting a role as a driver of prostate cancer for TRAP1, could be potentially targeted for therapy.
Initial evaluation of TRAP1 expression in a limited number of human colorectal carcinomas showed upregulation in comparison to normal matched peritumoral mucosa in 65% of cases (Amoroso et al. 2014). Accordingly, TRAP1 levels are increased in colorectal carcinoma cells resistant to 5-fluorouracil, oxaliplatin, and irinotecan and, in turn, its overexpression leads to 5-fluorouracil, oxaliplatin, and irinotecan resistance. Consistently, it has been later shown that the increase of TRAP1 expression level in colorectal cancer is significantly correlated to lymphonode metastases, advanced tumor stage, and reduced overall survival (Amoroso et al. 2014; Agliarulo et al. 2015). Mechanistically, TRAP1 regulates two signaling pathways responsible for colorectal carcinogenesis, i.e., BRAF and Wnt/β-catenin signaling. Indeed, TRAP1 modulates the expression of the oncogene BRAF, and the two proteins interact in colorectal cancer cells and are frequently co-expressed in human colorectal carcinomas (Condelli et al. 2014). Reciprocally, BRAF cytoprotective signaling involves TRAP1-dependent inhibition of mitochondrial apoptotic pathway and TRAP1 targeting by the mitochondria-directed HSP90 inhibitor. Gamitrinib induces apoptosis and inhibits colony formation in BRAF-driven colorectal carcinoma cells, which are known to be poorly responsive to anticancer therapies (Condelli et al. 2015). Furthermore, TRAP1 expression is enriched in stem cells located at the bottom of intestinal crypts and, accordingly, in cancer stem cells sorted from colorectal cancer cell lines, and TRAP1 silencing in HCT116 colorectal cancer cells results in loss of the stemlike signature (Lettini et al. 2016). Noteworthy, TRAP1 regulates the phosphorylation and the degradation of β-catenin, thus playing a positive control on Wnt/β-catenin pathway that is activated in colorectal cancers with high TRAP1 expression (Lettini et al. 2016). Consistently, the proteomic analysis of TRAP1 client protein network in human colorectal carcinomas suggest that the upregulation of TRAP1 and six of its client proteins provides a protein signature predictive of poor prognosis in advanced disease (Maddalena et al., unpublished observation).
TRAP1 is aberrantly upregulated in breast tumors compared to control tissues, and its expression in human breast cancer specimens inversely correlates with tumor grade. TRAP1 knockdown in breast cancer cells downregulates mitochondrial aerobic respiratory capacity, sensitizes cells to lethal stimuli, and inhibits tumor growth (Zhang et al. 2015). TRAP1 expression in breast cancer cells also confers resistance to paclitaxel, a microtubule stabilizing/ER stress inducer agent widely used in breast cancer therapy, and to genotoxic agents, as anthracyclines. Accordingly, paclitaxel- and anthracycline-resistant cell lines show higher levels of TRAP1 compared to nonresistant counterparts, but its knockdown or inhibition by gamitrinib restores drug sensitivity (Amoroso et al. 2014).
Unexpectedly, initial immunohistochemical analyses on 208 patients affected by ovarian cancer showed that high TRAP1 expression correlates significantly with favorable chemotherapy response and has a significantly positive impact on overall survival (Amoroso et al. 2014). Subsequent studies have indeed shown that TRAP1 expression inversely correlates with grade and stage and directly correlates with better overall survival in a large cohort of ovarian cancer patients. Accordingly, TRAP1 knockdown induces resistance to cisplatin in ovarian cancer cells (Matassa et al. 2016). TRAP1 gene is often deleted in high-grade serous ovarian cancer patients, and its expression is reduced in tumor metastases, while being associated with an epithelial phenotype both in vitro and ex vivo (Amoroso et al. 2016). In keeping with these data, it has also been reported an inverse correlation between TRAP1 expression and tumor stage in cervical, bladder, and clear cell renal cell carcinoma. Remarkably, among them, cervical carcinoma relies mostly on oxidative phosphorylation for its energetic metabolism, which is particularly relevant, considering that TRAP1 is an important determinant of tumor metabolism and has been proposed as a pivotal driver of cancer cell’s shift from oxidative phosphorylation toward aerobic glycolysis, the so-called Warburg effect (see below).
Immunohistochemical analysis of human grade IV glioblastoma specimens revealed that TRAP1 was highly expressed in the tumor cell population; adjacent normal astrocytes did not contain TRAP1, whereas a low level of TRAP1 expression was detected in neurons. Treatment of glioblastoma cells with shepherdin, a mitochondrial-permeable inhibitor of HSP90 ATPase activity, triggered glioblastoma cell death and suppressed intracranial glioma growth in mouse models (Siegelin et al. 2010).
Non-small Cell Lung Cancer
TRAP1 knockdown reduces cell growth and clonogenic cell survival and impairs mitochondrial functions in non-small cell lung cancer cells. A moderate TRAP1 staining by immunohistochemical analysis was found in normal bronchial mucosa, as opposed to adjacent tumor. Moreover, high TRAP1 expression was associated with increased risk of disease recurrence (Amoroso et al. 2014).
Human genetic studies have identified a number of homozygous mutations in mitochondrial serine/threonine kinase PINK1 causing autosomal recessive, early onset of Parkinson’s disease. It has been demonstrated that PINK1 protects against oxidative stress-induced cell death by suppressing cytochrome c release from mitochondria and that PINK1 protective function depends on its kinase activity to phosphorylate TRAP1. The ability of PINK1 to promote TRAP1 phosphorylation and cell survival is impaired by Parkinson’s disease-linked PINK1 mutations (Amoroso et al. 2014). In turn, overexpression of human TRAP1 is able to mitigate PINK1 loss-of-function phenotypes in Drosophila, rescuing morphological, locomotory, and mitochondrial defects of PINK1-mutant flies (Amoroso et al. 2014). Symmetrically, loss of TRAP1 in flies results in increased sensitivity to stress, decreased mitochondrial function, and decreased dopamine brain levels; its upregulation in PINK1-mutant neurons rescues mitochondrial impairment (Amoroso et al. 2014).
CAKUT and VACTERL Association
Homozygosity mapping coupled with whole-exome sequencing identified recessive mutations in TRAP1 gene in individuals with CAKUT and VACTERL association, a well-defined variable phenotype encompassing vertebral defects, anorectal malformations, cardiac defects, tracheoesophageal fistula/atresia, renal malformations, and limb defects (Saisawat et al. 2014). The absence of TRAP1 homozygous or compound heterozygous deleterious (recessive) variants in 800 patients with nephronophthisis testifies for the causal relationship between TRAP1 mutations and the onset of the disease.
Transformation of mild glomerulonephritis into end-stage disease coincides with shutdown of renal DNaseI expression in mouse models, damaging glomerular basement membranes. Translating the observations obtained in mice to human lupus nephritis, it has been observed that DNaseI shutdown coincides with TRAP1 overexpression, with a still unclear but structured molecular mechanism that could probably involve transcriptional interference, due to overlapping of the two transcripts on chromosome 16. Indeed, an inspection of the DNaseI gene organization reveals an overlap of 59 nucleotides in the annotated transcript with transcripts from the convergently transcribed TRAP1 gene in their 3′-untranslated regions. Given that transcription proceeds well beyond the 3′ end of the mature transcript, the overlap between the transcripts of the two genes is substantial, and it is unlikely that they are transcribed simultaneously. This peculiar gene organization, found in only a few percentage of human transcription units, provides to TRAP1 a new role in the progression of this disease (Fismen et al. 2013).
TRAP1 as an Antioxidant and Antiapoptotic Protein
TRAP1 binds the immunophilin CypD, a component of the mitochondrial permeability transition pore, and prevents its phosphorylation in the presence of apoptotic stimuli, thus avoiding permeability transition and initiation of apoptosis (Kang et al. 2007).
A functional interaction with the mitochondrial isoform of the calcium-binding/antiapoptotic protein sorcin was also characterized in colorectal cancer cells as a cytoprotective complex (Amoroso et al. 2014).
TRAP1 interference in different cancer cell lines (e.g., colorectal and breast cancers), as well as the use of dominant-negative mutants of TRAP1, sensitizes cells to multiple cell death inducers, ranging from inhibitors of protein synthesis and degradation to endoplasmic reticulum stress inducers and chemotherapeutics agents.
Cells expressing high levels of TRAP1 show increased levels of the scavenging tripeptide GSH and are more resistant to oxidative stress, also showing cross-resistance to chemotherapeutics.
As a downstream effector of PINK1, TRAP1 prevents oxidative stress-induced apoptosis in neurons, and the dysregulation of this mitochondrial pathway seems involved in Parkinson’s disease pathogenesis.
It may be reasonable also that the causal role of TRAP1 mutations in CAKUT could be ascribed to reactive oxygen species-mediated apoptosis, upregulation of a stress response gene signature, and autophagy, with subsequent perturbation of the normal kidney development.
TRAP1 as a Modulator of Cell Metabolism
In recent years, many steps have been taken to understand the regulatory role of TRAP1 in the cell, but its function and the molecular mechanisms involved remain controversial. TRAP1 has been found to interact with complex II and IV of the electron transport chain in mitochondria. The interaction with the complex II, also known as succinate dehydrogenase (SDH), leads to inhibition of its enzymatic activity, whereas activity of complex IV is not affected, nor expression levels of complex II and mitochondrial mass are altered. An inverse correlation between TRAP1 and SDH activity is also found in colorectal cancer tissue specimens. As a result, TRAP1 yields a reduced oxygen consumption rate (i.e., reduced mitochondrial respiration) in different cell lines, this inducing a metabolic shift toward glycolysis and a “Warburg phenotype,” and decreased fatty acid oxidation (Rasola et al. 2014). Through these mechanisms, TRAP1 also reduces cellular reactive oxygen species and lipid peroxidation, in keeping with the stress protective roles described in tumoral settings. Along with the direct interaction with the respiratory complexes, it has also been found an interaction between TRAP1 and the tyrosine-protein kinase c-Src, which is known to stimulate complex IV activity and enhance oxidative phosphorylation, suggesting that the impact of TRAP1 on mitochondrial respiration could be mediated by c-Src. However, much remains to be done to clarify the molecular mechanisms responsible for the regulation of cellular metabolism by TRAP1. Indeed, TRAP1 has also been described as a factor stabilizing SDH and thus ensuring its activity under metabolic stress conditions, namely, glucose deprivation (Rasola et al. 2014). Accordingly, knockout mice for TRAP1 displayed global upregulation of oxidative phosphorylation and glycolysis transcriptomes, deregulated mitochondrial respiration, oxidative stress, impaired cell proliferation, and a switch to glycolytic metabolism in vivo. This was associated to reduce incidence of age-associated pathologies, including obesity, inflammatory tissue degeneration, dysplasia, and spontaneous tumor formation (Lisanti et al. 2014).
TRAP1 as a Chaperone Linked to Protein Synthesis
Contributes to the regulation of energetic metabolism through direct binding to succinate dehydrogenase, the complex II of the mitochondrial respiratory chain
Is part of a pro-survival signaling pathway aimed at protecting mitochondria against the toxic effects of oxidants and anticancer drugs
Controls protein homeostasis through a direct involvement in the regulation of protein synthesis and co-translational protein degradation, binding either component of the translational machinery or of the proteasome
However, whether TRAP1 roles are oncogenic or not a matter of debate, TRAP1 levels are elevated in several malignancies (i.e., colon, breast, prostate, lung) and correlate with drug resistance, whereas it is downregulated in specific tumors with predominant oxidative metabolism (i.e., ovarian, bladder, and renal). TRAP1 role in metabolic reprogramming is also controversial: its capacity to enhance or suppress oxidative phosphorylation seems to be context-dependent. The challenge ahead is to recapitulate the puzzling functions of TRAP1 as a paradigm for the complex regulatory network allowing survival advantage of cancer cells. This could help to unravel the mechanisms used by cancer cells to face the metabolic stress and to identify new strategies to undertake, in order to disrupt networks of integrated control in cancer cells. In this regard, several HSP90 ATPase inhibitors have been developed and some of them entered in clinical trials. However, very few data are available on the selectivity profiles of different HSP90 isoform inhibitors. The redundancy in HSP90 chaperone functions usually represents a limitation for the understanding of TRAP1-specific contribution to cellular regulations. Recently, the crystal structure of full-length TRAP1 has been presented, opening new scenarios on the possibility to design novel, specific TRAP1 inhibitors. Despite the high level of sequence homology, crystallographic and biochemical studies have shown that, when bound to nucleotides, HSP90α, HSP90β, GRP94, and TRAP1 adopt distinctly different conformations and hydrolyze ATP with different rates. Thus, the overall structure and conformational flexibility of the paralogs have a central role in shaping their ATP-binding sites and can be exploited to achieve selectivity among inhibitors. A second issue, in this scenario, is whether compartmentalized TRAP1 inhibition may provide a clinical advantage in reducing toxicity and improving anticancer activity. Thus, a future goal is to generate and compare a novel class of mitochondria- and non-mitochondria-directed selective TRAP1 inhibitors and to evaluate their cytotoxic/cytostatic activity in vitro and in vivo.
However, TRAP1 is not only related to cancer. TRAP1 works downstream of PINK1, and the disruption of this network causes mitochondrial dysfunction and neurodegeneration in Parkinson’s disease. Moreover, TRAP1 gene mutations have been identified as disease-causing in rare pathologies that affect kidney and urinary tract. From the overall scenario, it is possible to hypothesize a role for TRAP1 in normal cellular and tissue development. Data coming from knockout animals (mouse and fruit fly) demonstrate that TRAP1 is dispensable for the organismal development, but its deregulation can be disease-causing. It is well known that heat shock proteins in general are related to development and that heat shock protein genes are phase- and tissue-specific during embryogenesis. It would not be surprising to find that most of the diseases in which TRAP1 is involved are yet to be discovered. One of the most challenging future perspectives is the understanding of the regulation of TRAP1 expression and activity in space and time, as well as its impact in health and disease. This will provide novel tools to improve the welfare of the population to a wide audience of molecular oncologists and clinical pathologists.
- Agliarulo I, Matassa DS, Amoroso MR, Maddalena F, Sisinni L, Sepe L, Ferrari MC, Arzeni D, Avolio R, Paolella G, Landriscina M, Esposito F. TRAP1 controls cell migration of cancer cells in metabolic stress conditions: correlations with AKT/p70S6K pathways. Biochim Biophys Acta. 2015;1853:2570–9.PubMedCrossRefGoogle Scholar
- Amoroso MR, Matassa DS, Agliarulo I, Avolio R, Lu H, Sisinni L, Lettini G, Gabra H, Landriscina M, Esposito F. TRAP1 downregulation in human ovarian cancer enhances invasion and epithelial–mesenchymal transition. Cell Death Dis. 2016; 7:e2522. doi: 10.1038/cddis.2016.400.
- Condelli V, Piscazzi A, Sisinni L, Matassa DS, Maddalena F, Lettini G, Simeon V, Palladino G, Amoroso MR, Trino S, Esposito F, Landriscina M. TRAP1 is involved in BRAF regulation and downstream attenuation of ERK phosphorylation and cell cycle progression: a novel target for BRAF-mutated colorectal tumors. Cancer Res. 2014;74(22):6693–704.PubMedCrossRefGoogle Scholar
- Condelli V, Maddalena F, Sisinni L, Lettini G, Matassa DS, Piscazzi A, Palladino G, Amoroso MR, Esposito F, Landriscina M. Targeting TRAP1 as a downstream effector of BRAF cytoprotective pathway: a novel strategy for human BRAF-driven colorectal carcinoma. Oncotarget. 2015;6(26):22298–309.PubMedPubMedCentralCrossRefGoogle Scholar
- Fismen S, Thiyagarajan D, Seredkina N, Nielsen H, Jacobsen S, Elung-Jensen T, Kamper AL, Johansen SD, Mortensen ES, Rekvig OP. Impact of the tumor necrosis factor receptor-associated protein 1 (Trap 1) on renal DNaseI shutdown and on progression of murine and human lupus nephritis. Am J Pathol. 2013;182(3):688–700.PubMedCrossRefGoogle Scholar
- Lettini G, Sisinni L, Condelli V, Matassa DS, Simeon V, Maddalena F, Gemei M, Lopes E, Vita G, Del Vecchio L, Esposito F, Landriscina M. TRAP1 regulates stemness through Wnt/β-catenin pathway in human colorectal carcinoma. Cell Death Differ. 2016;23(11):1792–803.PubMedPubMedCentralCrossRefGoogle Scholar
- Matassa DS, Agliarulo I, Amoroso MR, Maddalena F, Sepe L, Ferrari MC, Sagar V, D’Amico S, Loreni F, Paolella G, Landriscina M, Esposito F. TRAP1-dependent regulation of p70S6K is involved in the attenuation of protein synthesis and cell migration: Relevance in human colorectal tumors. Mol Oncol. 2014;8(8):1482–94.PubMedPubMedCentralCrossRefGoogle Scholar
- Matassa DS, Amoroso MR, Lu H, Avolio R, Arzeni D, Procaccini C, Faicchia D, Maddalena F, Simeon V, Agliarulo I, Zanini E, Mazzoccoli C, Recchi C, Stronach E, Marone G, Gabra H, Matarese G, Landriscina M, Esposito F. Oxidative metabolism drives inflammation-induced platinum resistance in human ovarian cancer. Cell Death Differ. 2016;23:1542–54.PubMedPubMedCentralCrossRefGoogle Scholar
- Saisawat P, Kohl S, Hilger AC, Hwang DY, Yung Gee H, Dworschak GC, Tasic V, Pennimpede T, Natarajan S, Sperry E, Matassa DS, Stajić N, Bogdanovic R, de Blaauw I, Marcelis CL, Wijers CH, Bartels E, Schmiedeke E, Schmidt D, Märzheuser S, Grasshoff-Derr S, Holland-Cunz S, Ludwig M, Nöthen MM, Draaken M, Brosens E, Heij H, Tibboel D, Herrmann BG, Solomon BD, de Klein A, van Rooij IA, Esposito F, Reutter HM, Hildebrandt F. Whole-exome resequencing reveals recessive mutations in TRAP1 in individuals with CAKUT and VACTERL association. Kidney Int. 2014;85(6):1310–7.PubMedCrossRefGoogle Scholar