- Open Access
Nuclear heterogeneous nuclear ribonucleoprotein D is associated with poor prognosis and interactome analysis reveals its novel binding partners in oral cancer
- Manish Kumar1,
- Ajay Matta2,
- Olena Masui3,
- Gunjan Srivastava2,
- Jatinder Kaur1,
- Alok Thakar4,
- Nootan Kumar Shukla5,
- Ajoy RoyChoudhury6,
- Meherchand Sharma7,
- Paul G. Walfish2, 8, 9, 10, 11,
- K. W. Michael Siu3, 12,
- Shyam Singh Chauhan†1Email author and
- Ranju Ralhan†2, 8, 9, 10Email author
© Kumar et al. 2015
Received: 15 January 2015
Accepted: 13 August 2015
Published: 30 August 2015
Post-transcriptional regulation by heterogeneous ribonucleoproteins (hnRNPs) is an important regulatory paradigm in cancer development. Our proteomic analysis revealed hnRNPD overexpression in oral dysplasia as compared with normal mucosa; its role in oral carcinogenesis remains unknown. Here in we determined the hnRNPD associated protein networks and its clinical significance in oral squamous cell carcinoma (OSCC).
Immunoprecipitation (IP) followed by tandem mass spectrometry was used to identify the binding partners of hnRNPD in oral cancer cell lines. Ingenuity pathway analysis (IPA) was carried out to unravel the protein interaction networks associated with hnRNPD and key interactions were confirmed by co-IP-western blotting. hnRNPD expression was analyzed in 183 OSCCs, 44 oral dysplasia and 106 normal tissues using immunohistochemistry (IHC) and correlated with clinico-pathological parameters and follow up data over a period of 91 months. Kaplan–Meier survival and Cox-multivariate-regression analyses were used to evaluate the prognostic significance of hnRNPD in OSCC.
We identified 345 binding partners of hnRNPD in oral cancer cells. IPA unraveled novel protein–protein interaction networks associated with hnRNPD and suggested its involvement in multiple cellular processes: DNA repair, replication, chromatin remodeling, cellular proliferation, RNA splicing and stability, thereby directing the fate of oral cancer cells. Protein–protein interactions of hnRNPD with 14-3-3ζ, hnRNPK and S100A9 were confirmed using co-IP-western blotting. IHC analysis showed significant overexpression of nuclear hnRNPD in oral dysplasia [p = 0.001, Odds ratio (OR) = 5.1, 95 % CI = 2.1–11.1) and OSCCs (p = 0.001, OR = 8.1, 95 % CI = 4.5–14.4) in comparison with normal mucosa. OSCC patients showing nuclear hnRNPD overexpression had significantly reduced recurrence free survival [p = 0.026, Hazard ratio = 1.95, 95 % CI = 1.0–3.5] by Kaplan–Meier survival and Cox-multivariate-regression analyses and has potential to define a high-risk subgroup among OSCC patients with nodal negative disease.
Our findings suggest novel functions of hnRNPD in cellular proliferation and survival, besides RNA splicing and stability in oral cancer. Association of nuclear hnRNPD with poor prognosis in OSCC patients taken together with its associated protein networks in oral cancer warrant future studies designed to explore its potential as a plausible novel target for molecular therapeutics.
Post-transcriptional regulation of mRNA stability and translation by RNA binding proteins (RBPs) is a key determinant of gene expression [1–3]. These RNA–protein interactions dictate the ultimate fate of the transcripts and are emerging as an important regulatory paradigm in cancer development [3, 4]. The mRNA decay kinetics is largely controlled by presence of specific cis-acting sequence and/or structural determinants within each transcript [5, 6]. About 16 % of all human protein coding genes are encoded by mRNAs that contain an adenylate-uridylate (AU)-rich element [ARE] motif within their 3′UTR [1, 2, 5, 6]. AU-rich RNA-binding factor (AUF1)/heterogeneous nuclear ribonucleoprotein D (hnRNPD) is an ARE-binding protein which regulates the mRNA stability of many genes involved in cell cycle, proliferation, survival, senescence and stress response [1, 2, 3, 4, 5, 7, 8, 9, 10, 11]. This protein harbors two RNA-binding domains arranged in tandem and a glycine-rich region in the C-terminus (2x RBD-Gly) and regulates the cellular half-life of many mRNAs by directly interacting with AREs in their 3′untranslated region [12–15]. Overexpression of hnRNPD in vivo resulted in deregulation of mRNAs including c-myc, c-jun, c-fos, and tumor necrosis factor-α (TNF-α) which promote tumorigenesis suggesting an oncogenic role of hnRNPD [1, 2, 3, 16, 17]. Increased hnRNPD expression also reduced the cell cycle checkpoint regulators p21 and p16Ink4a, a critical mediator of senescence [10, 18, 19]. Nuclear hnRNPD has been shown to activate the transcription promoter for telomerase catalytic subunit Tert, and links maintenance of telomere length and normal aging to attenuation of inflammatory cytokine expression and inhibition of cellular senescence .
Head and neck squamous cell carcinoma (HNSCC) ranks as the sixth leading cause of cancer related deaths worldwide . HNSCCs often show heterogeneous pathologic and clinical features and diverse outcome [22, 23]. HNSCC is among the most morbid human malignancies and the quality of life in survivors is poor. Moreover, HNSCC patients often have recurrence of the tumor at the same site, or develop second primary tumors, frequently attributed to field cancerization . Oral squamous cell carcinomas (OSCCs) comprise a large proportion of HNSCCs. The lack of clinically proven biomarkers limits therapeutic decisions to be solely based on tumor site and staging. However, tumors with similar clinical features can differ in disease outcome . A better understanding of the molecular pathogenesis of OSCC is urgently needed for rigorous disease management. The development of OSCCs is often preceded by clinically distinct oral lesions such as leukoplakia or erythroplakia with histological evidence of squamous cell hyperplasia or dysplasia; on an average about one percent of these lesions transform to cancer annually. The oral lesions with histologically proven dysplasia are called Oral premalignant or potentially malignant lesions (OPLs). Identification of OPLs at high risk of progression to cancer is a high priority to enable early intervention, prior to development of frank malignancy for more effective disease management and improve the quality of life in survivors . We reported overexpression of hnRNPD in human oral premalignant lesions by proteomic analysis . In this study, interactome analysis was undertaken to gain an insight into hnRNPD associated protein–protein networks, by identifying its binding partners in oral cancer cells using immunoprecipitation followed by liquid chromatography—tandem mass spectrometry (LC–MS/MS). Bioinformatic analysis based cellular networks and pathways were identified and protein–protein interactions were confirmed using oral cancer cells. Further, we also determined the significance of hnRNPD overexpression in clinical specimens of oral dysplasia and cancer and correlated with disease outcome.
Oral squamous cell carcinoma (OSCC) cell line, SCC4 was obtained from American Type Culture Collection (ATCC), HSC2 (JCRB0622) from Health Science Research Resources Bank, Japan (HSRRB); Tu167 and MDA1986 were a kind gift from MD Anderson Cancer Centre (Houston, Texas). All cell lines were characterized using short tandem repeat polymorphism (STR) analysis. OSCC cells were grown in monolayer cultures in Dulbecco’s modified eagle medium (DMEM) (Sigma Aldrich, St. Louis, MO) supplemented with 10 % fetal bovine serum (FBS) (Sigma), 1 mM l-glutamine, 100 μg/ml streptomycin and 100 U/ml penicillin in a humidified incubator (5 % carbon-dioxide, 95 % air) at 37 °C as described previously [26, 27].
Patients, tissue specimens, clinicopathological data and follow-up
The Institutional Human Ethics Committee of All India Institute of Medical Sciences (AIIMS), New Delhi, India, approved this study prior to its commencement (NO.IESC/T-261/03.06.2011). Tissue specimens were obtained from patients with oral dysplasia (n = 44) as revealed by H&E staining from Department of Otorhinolaryngology, All India Institute of Medical Sciences (AIIMS) and from 183 OSCC patients undergoing curative cancer surgery during the period 2002–2008, after obtaining patients’ consent, while 106 non-malignant oral tissues with histological evidence of normal epithelium constituted the normal group. Patient demographic, clinical, and pathological data were recorded in a pre-designed performa .
Of the 183 OSCCs, 144 cases could be followed-up in the head-and-neck cancer follow-up clinic at regular time intervals up to a maximum period of 91 months as of May, 2013. The patients were revisited clinically on a regular basis and the time to recurrence was recorded. Of these 144 OSCC patients, loco-regional relapse was observed in 60 cases (41.7 %), while 14 patients died (9.7 %) as determined from follow-up reports. If a patient died, the survival time was censored at the time of death; the medical history, clinical examination, and radiological evaluation were used to determine whether the death had resulted from recurrent cancer (relapsing patients) or from any other causes. Recurrence-free survivors (RFS) were defined as patients free from clinical and radiological evidence of local or regional recurrence or death at the time of the last follow-up.
Real time-PCR analysis of hnRNPD mRNA levels in OSCCs and normal oral mucosa
For this study, hnRNPD mRNA level was determined in 12 paired tumor and normal tissue samples. Total RNA was extracted from a small portion of the biopsies OSCCs and normal oral tissues with Trizol reagent (Invitrogen, CA) according to the manufacturer’s protocol. The quality of the isolated RNA was tested by its optical density (260/280 ratio is 2.0). The expression of hnRNPD was quantified by real-time PCR (RT-qPCR). Total RNA (1 μg) was reverse-transcribed using Reverse transcriptase (Thermo Scientific, Waltham, MA, USA) using oligo-dT primers according to the manufacturer’s instructions. Real-time PCR reactions were performed and quantified by Maxima SYBR Green (Thermo Scientific, Waltham, MA, USA) using CFX96 Touch™ Real-Time PCR Detection System (BioRad, Hercules, CA, USA) using the ribosomal 18S gene as an internal control for normalisation. All assays were performed in triplicate in a 20 μl two-step reaction. The specificity of the amplified PCR products was assessed by melting curve analysis and agarose gel electrophoresis of a small aliquot of the reaction followed by staining with ethidium bromide. The efficiency of the qPCR reaction was measured in separate assays using cDNA obtained from total RNA of SCC4, HSC2, TU167 and MDA1986 oral cancer cell lines. The primer sequences are shown: hnRNPD–Sense: GCCTTTCTCCAGATACACCTGAAG; Antisense: CTTATTGGTCTTGTTGTCCA TGGG and 18S ribosomal–Sense: GTAACCCGTTGAACCCCATT, Antisense: CCA TCCAATCGGTAGTAGCG.
Whole-cell lysates were prepared from OSCC cells (SCC4/HSC2/Tu167/MDA1986), oral normal tissues (n = 4), dysplasia (n = 2), and OSCC (n = 8) by homogenization in RIPA lysis buffer [26, 27]. Equal amounts of proteins (60 μg/lane) were resolved and electro-transferred onto polyvinylidene-difluoride (PVDF) membrane. After blocking blots were incubated with rabbit polyclonal hnRNPD antibody at 4 °C overnight. β-actin served as a control for equal protein loading in each lane. Membranes were incubated with their respective HRP-conjugated secondary antibody (DAKO Cytomation, Glostrup, Denmark), diluted at an appropriate dilution in 1 % BSA, for 2 h at room temperature. Protein bands were detected by the enhanced chemiluminescence method (ECL, Pierce, IL) on XO-MAT film.
Oral cancer cells (SCC4/MDA1986) were lyzed in IP-lysis buffer as described . Lysates were pre-cleared with Protein A-Sepharose (GE Healthcare Biosciences, Sweden), and immunecomplexes were obtained by incubation with polyclonal hnRNPD antibody and pulled down by incubating with Protein A-Sepharose. In negative controls, only Protein A Sepharose beads were added to eliminate proteins that bind non-specifically to the beads. Immunecomplexes were resolved on 10 % SDS-PAGE, stained with gel code blue and analyzed by reverse phase (RP)-liquid chromatography mass spectrometry (LC–MS/MS) as described [28, 29].
Reverse phase (RP)-liquid chromatography mass spectrometry (LC–MS/MS)
The proteins bands were excised from gels and digested with trypsin as described [28, 29]. The digested peptides from each band were analyzed in duplicates using a Nanobore LC system (LC Packings, Amsterdam, Netherlands) and a QSTAR Pulsar mass spectrometer (Applied Biosystems/MDS SCIEX, Foster City, CA) in positive ion mode, externally calibrated with bovine serum albumin tryptic peptides [28, 29]. MS data were acquired in information-dependent acquisition (IDA) mode using Analyst QS 1.1 software (Applied Biosystems/MDS SCIEX). The LC–MS/MS was performed using a 1 s TOF–MS survey scan from 400 to 1500 Da, followed by four, 2 s product-ion scans, from 80 to 2000 Da, of the five most-abundant peaks. The collision energy (CE) was automatically controlled by the IDA CE parameter script. Switching criteria were set for ions with m/z ≥ 400 and <1500, charge states of +2 to +4, and abundances of ≥10 counts. Former target ions were excluded for 30 s, and ions within a 100-ppm window were ignored. To minimize redundancy in subsequent iterations, precursor ion exclusion (PIE) list was added to LC–MS/MS method as described earlier [28, 29].
Identification of binding partners
LC–MS/MS data of each sample was used to identify proteins by searching a concatenated Swissprot/Panther database of 66082 distinct human protein entries (version June 2, 2010). The database was searched using Proteinpilot software, version 2.0.1 (AB SCIEX, Foster City), and the Paragon algorithm . Protein identification was performed at a confidence threshold of 95 % (Proteinpilot Unused score ≥1.3) with methyl methanethiosulfonate (MMTS) selected as cysteine modification, and with the search option ‘emphasis on biological modifications’ checked. Peptide and protein summaries were generated. Only proteins identified with local false discovery rate (FDR) equal to, or less than, 5 % were considered for further analysis [28, 29]. Redundant proteins and peptides, proteins identified in reverse sequence and in negative controls (i.e. beads only) were removed from the list of identified proteins.
Confocal laser scan microscopy (CLSM)
For CLSM, 5 × 104 OSCC cells (SCC4/MDA1986) were plated on cover slips and grown for 24 h fixed in acetone: methanol mixture (1:1) at −20 °C for 20 min. . Cells were permeabilized with PBS-0.1 % Tween 20, non-specific binding blocked with 5 % BSA for 1 h; cells were incubated with rabbit polyclonal hnRNPD (ab50692)/mouse monoclonal hnRNPK (ab23644, Abcam, CA) antibody at 4 °C overnight. Expression of proteins was determined using fluorescein isothiocyanate (FITC)/TRITC-labeled secondary antibodies (DAKO Cytomation, Denmark) .
Immunohistochemistry of hnRNPD, hnRNPK and 14-3-3ζ in oral tissues and scoring
Paraffin-embedded tissue sections were deparaffinized, antigen was retrieved, endogenous peroxidase activity was quenched with hydrogen peroxide (0.3 % v/v) and non-specific binding blocked with 1 % bovine serum albumin (BSA). The slides were incubated with either rabbit polyclonal anti-hnRNPD antibody (1 μg/ml, ab50692, Abcam, CA) or mouse monoclonal anti-hnRNPK antibody (ab23644) or rabbit polyclonal 14-3-3ζ antibody (sc-1019) for 16 h at 4 °C. The primary antibody was detected using the Dako Envision kit (Dako CYTOMATION, Glostrup, Denmark) with diaminobenzidine as the chromogen and counterstained with hematoxylin [26, 27]. The sections were evaluated by light microscopy and scored using a semi-quantitative scoring system for both staining intensity (nuclear/cytoplasmic) and percentage positivity as described earlier [26, 27]. The tissue sections were scored based on the % of immunostained cells as: 0–10 % = 0; >10–30 % = 1; >30–50 % = 2; >50–70 % = 3 and >70–100 % = 4. Sections were also scored semi-quantitatively on the basis of staining intensity as negative = 0; mild = 1; moderate = 2; intense = 3. Finally, a total score was obtained by adding the score of percentage positivity and intensity giving a score range from 0 to 7. IHC score thus obtained for different proteins were subjected to statistical analysis.
The immunohistochemical data were subjected to statistical analyses using the SPSS 20.0 software (Chicago, IL, USA). Sensitivity and specificity was calculated using receiver operating characteristic (ROC) analyses. Based on sensitivity and specificity values a cut-off ≥4 was defined as positive criterion for hnRNPD (nuclear/cytoplasmic). The relationships between hnRNPD and clinicopathological parameters were tested using Chi Square and Fischer’s exact test. Two-sided p-values were calculated and p < 0.05 was considered to be significant. Similarly, positive predictive value (PPV) and negative predictive value (NPV) were calculated for hnRNPD overexpression in oral lesions and OSCCs in comparison with normal oral mucosa. The correlation of hnRNPD staining with patient survival was evaluated using life tables constructed from survival data with Kaplan–Meier plots and Cox regression multivariate analysis. In order to confirm the association among hnRNPD, hnRNPK and 14-3-3ζ overexpression in clinical specimens of OSCCs, we performed Kappa analysis to determine the agreement of association between these proteins using their IHC scores. One of the most important features of the Kappa statistical analysis is its ability to measure the degree of agreement or reliability of agreement [26–28].
Expression of hnRNPD in oral cancer cells and tissues
Western blot analysis showed a single intense band of 37 kDa in all the oral cancer cell lines tested (SCC4, HSC2, Tu167 and MDA167, Fig. 1b), dysplasia (d1, d2) and OSCCs (t1–t8) demonstrating the presence of only p37/AUF1 isoform of hnRNPD. Faint or no expression of hnRNPD was observed in normal oral tissues (n1–n4), while an intense band was observed in OSCCs, thus confirming hnRNPD protein overexpression in OSCCs (Fig. 1b).
Identification of binding partners of hnRNPD in OSCC cells
To gain an insight into the role of hnRNPD in OSCCs, we identified its binding partners in oral cancer cell lines (SCC4 and MDA1986). Immunoprecipitates obtained from SCC4 and MDA1986 cells using hnRNPD specific antibody were separated on 10 % SDS-PAGE, stained with gel-code blue, 35 protein bands were excised from the immunoprecipated sample and from the mock treated sample, digested with trypsin and subjected to RP-LC–MS/MS analysis for identification of proteins (Additional file 1: Figure S1A, Additional file 2: Figure S1B). Our novel approach using multiple iterations and development of precursor ion exclusion (PIE) list for protein identification revealed 345 binding partners of hnRNPD in oral cancer cell lines (Additional file 3: Table S1). Our approach revealed interactions of hnRNPD with 17 members of hnRNP family including hnRNPA2/B1, hnRNPK, hnRNPU, hnRNPG suggesting that hnRNPD forms heterodimers with its family members (Additional file 3: Table S1). hnRNPD also showed interactions with proteins involved in short-interfering-RNA (RNAi)-mediated gene silencing (EIF2C1, EIF2C2, EIF2C3), DNA repair (XRCC5, XRCC6), chromatin remodeling (SMARCC1, SMARCC2, SMARCB1; histone family of proteins including H1, H2A, H2B and H4), tumor protein 63 (TP63), transcription factors (zinc finger domain proteins, ZC3HAV1, ZCCHC8), cell signaling proteins (IGFBP, G3BP1, GNB2L1, NCL), nuclear-shuttling proteins (14-3-3ζ), S100A9 (calcium binding protein) and several other proteins involved in RNA splicing, stability and decay (ribosomal proteins 28S, 40S and 60S; ATP dependent RNA helicases, mRNA cap guanine-N7 methyltransferase, RNMT, RAE1) supporting its function in translation (Additional file 3: Table S1).
Network analysis of hnRNPD protein interactions
Verification of interactions between hnRNPD and 14-3-3ζ, hnRNPK and S100A9
Using similar approach, we verified the interaction of hnRNPD with hnRNPK and S100A9 in oral cancer cells. Western blotting showed presence of hnRNPK and S100A9 in hnRNPD immunoprecipitates and these findings were also confirmed using reverse—IP (Fig. 3B, i). Confocal laser scan microscopy (CLSM) analysis confirmed cytoplasmic and nuclear co-localization of hnRNPD with hnRNPK in OSCC cell lines, SCC4 and MDA1986 (Fig. 3B, ii).
Immunohistochemical analysis of hnRNPD expression in oral normal mucosa, dysplasia and OSCCs
Analysis of hnRNPD protein expression and correlation with clinicopathological parameters
(95 % CI)
Age (median 45 years)
(range 15–85 years)
T1 + T2
T3 + T4
I + II
III + IV
Evaluation of nuclear hnRNPD overexpression as a prognostic marker for OSCC
Evaluation of nuclear hnRNPD as a prognostic marker of OSCCs
Evaluation of hnRNPD as a diagnostic marker
Dysplasia vs. normal
OSCCs vs. normal
hnRNPD overexpression correlates with 14-3-3ζ and hnRNPK expression in OSCCs tissues
As shown above, both hnRNPK and 14-3-3ζ were identified in as interaction partners of hnRNPD in oral cancer cells and tissues. On the basis of IHC analysis, the immunoreactivity score ≥4 for both hnRNPD and hnRNPK, and a score ≥5 for 14-3-3ζ were considered as overexpression and were used in Kappa analysis. In order to confirm such association among clinical specimens of OSCCs, we performed Kappa analysis to determine the agreement of association between the hnRNPD, hnRNPK and 14-3-3ζ expressions in OSCCs using their IHC scores. One of the most important features of the Kappa statistics is that it is a measurement of the degree of agreement or reliability of agreement. Among OSCCs, 61 % agreement with a Kappa score (κ = 0.236) was observed between nuclear hnRNPD and hnRNPK (p = 0.0003) (Additional file 3: Table S3). Similar agreement with a Kappa score (κ = 0.248) was also observed between nuclear hnRNPD and 14-3-3ζ. This further strengthened our results of IP-LC–MS/MS. However, no significant agreement was observed between cytoplasmic expression of hnRNPD with hnRNPK or 14-3-3ζ in OSCCs.
The ability of proteins to form complexes by physically binding to each other and alterations in sub-cellular distribution lead to perturbations in the cell circuitry underlying cancer development. In this study, we determined hnRNPD associated protein networks to get an insight in molecular pathogenesis of oral cancer. Our network analysis revealed involvement of hnRNPD in multiple cellular pathways involved in progression and metastasis of oral cancer. We identified novel binding partners of hnRNPD suggesting its involvement in DNA repair, chromatin remodeling, RNAi mediated gene silencing and several other cell signaling pathways involved in cellular proliferation and apoptosis, besides its role in RNA processing and turnover. Several reports have demonstrated involvement of different RNA-binding proteins (RBPs) in determining the cellular fate of mRNA transcripts in terms of stability and rate of translation [11, 22]. In this respect, we observed other members of the hnRNP family including hnRNPA2/B1, hnRNPK, hnRNPU, hnRNPG forming heterodimers with hnRNPD in oral cancer cells. In this support, interaction of hnRNPK, hnRNPC, hnRNPL and hnRNPA2/B1 with hnRNPD in cervical and lung cancer cells has been reported earlier [31–33]. In addition, our results also suggested interaction of hnRNPD with other RNA binding proteins (RBPs) such as ELAVL1, RALY, EWSR1 and FUS in oral cancer. The fate of RNAs regulated by these protein interactions among hnRNPD and other RNA binding proteins in oral cancer is currently under investigation. However, a recent study reported RNA—dependent direct physical interaction between ELAVL1 and hnRNPD in the nucleus influences the expression of cyclin D1 and p16, both of which are important for oral cancer development [34, 35]. Besides RBPs, microRNAs are also important contributors to the post-transcriptional control of gene expression . miRNAs act preferentially by binding to 3′-UTR region of target mRNA and are also involved in ARE-mediated mRNA instability. Precursor miRNAs are processed to mature miRNAs by multiprotein complexes including Drosha and Dicer and then incorporated into RNA induced silencing complex (RISC) . We identified RNA polymerase II, eIF4 and argonaute proteins (EIF2C1, EIF2C2 and EIF2C3) that are important components of RISC in our interactome analysis of hnRNPD in oral cancer. hnRNPD associates with endogenous DICER1 mRNA and destabilizes it; however knockdown of hnRNPD using siRNA increased half-life of DICER1 mRNA and elevated its expression, while overexpression of hnRNPD lowered DICER1 mRNA and protein levels .
Interestingly, we also identified an important nuclear-cytoplasmic shuttling protein, 14-3-3ζ, a member of 14-3-3 family of proteins, as a binding partner of hnRNPD and verified their interaction in oral cancer cells (SCC4/MDA1986). Moreover, both hnRNPD and 14-3-3ζ co-localization was observed in cytoplasm of OSCC cells, suggesting 14-3-3ζ retains cytoplasmic hnRNPD, similar to 14-3-3σ. 14-3-3 family of proteins is known for their overlapping functions in orchestrating their target proteins in cytoplasm. Our results also showed presence of 14-3-3 binding motif in hnRNPD polypeptide and verified the interaction of 14-3-3ζ with hnRNPD may be dependent on Ser83 phosphorylation (present in this motif), unlike its interaction with 14-3-3σ, as reported earlier [39, 40]. Previous reports have shown binding of 14-3-3σ to hnRNPD masks its nuclear localization signal, retaining it in cytoplasm and enhances the rapid turnover of its target protein expression [39, 40]. Further, hnRNPD demonstrated significant association with 14-3-3ζ expression in clinical specimens of OSCCs analyzed in this study. In thyroid cancer, cytoplasmic hnRNPD interacts with mRNAs encoding cyclins (A1, B1, D1 and E1) and cyclin-dependent kinase inhibitors . Stimulation of melanoma cells and monocytes by lipopolysaccharide resulted in cytoplasmic translocation of hnRNPD from nucleus and reduced levels of IL-6, IL-10 and TNF-α following activation of MKP-1 (MAPK phosphatase-1) [42–47]. This scenario might explain how nuclear-cytoplasmic translocation of hnRNPD can influence the cytoplasmic fate of mRNA.
Together, our network analysis suggested an important role of hnRNPD overexpression and its associated networks including protein interactions (direct/indirect) and regulation among these pathways is likely to play important role in determining the clinical outcome of OSCC patients. Supporting this hypothesis, our clinical findings demonstrated increased nuclear hnRNPD expression in clinical specimens of oral dysplasia and OSCCs as compared to normal oral mucosa. Notably, OSCC patients showing increased expression of nuclear hnRNPD had significantly reduced recurrence free survival. Nuclear hnRNPD overexpression in OSCCs and its emergence as a predictor of recurrence free survival in multivariate analysis in comparison with clinical and pathological parameters is an important novel finding in oral cancer, even though previous studies reported its overexpression in other malignancies including thyroid, melanoma, breast, cervix and murine lung tumors [4, 7, 16, 18, 48, 49, 35]. hnRNPD nuclear expression emerged as a significant predictor for recurrence with median RFS of 22 months as compared to patients who were node negative and had lower nuclear hnRNPD (median RFS = 69.0 months, p = 0.028). However, nuclear hnRNPD did not show any significant association with recurrence among OSCC patients with positive nodes at time of surgery. Thus, hnRNPD expression is likely to have the potential to define a high-risk subgroup among OSCC patients with nodal negative disease and might address the urgent need for more effective risk stratification strategies to improve patient care for these patients.
Our interactome analysis of hnRNPD protein provided an insight into its novel functions in oral cancer cells. Further, nuclear hnRNPD emerged as a prognostic marker for evaluating the risk of recurrence in OSCC patients. Based on these results, we suggest nuclear hnRNPD as a potential target for molecular therapeutics for oral cancer in future.
RR and SSC conceptualized the study. RR, SSC and AM contributed to the study design and to the manuscript. MK, AM, and OM conducted the experimental work. MK performed the chart reviews for clinical data, follow-up and data collection and established the clinical database. AT, NKS and ARC provided the clinical samples, clinical parameters and the follow-up data. MK and MCS performed the histopathology reporting of all the patients’ tissues analyzed. MK, AM and JK did the statistical analysis and had access to the raw data. OM performed the proteomic analysis under the guidance of KWMS. GS and MK performed the confocal microscopy. AM, SSC and RR interpreted the data. RR, PGW and SSC provided the infrastructural support for this study. The manuscript was drafted by MK and AM, edited by SSC and RR, and submitted for comments to all the authors. All authors read and approved the final manuscript.
The study was supported by International Science and Technology Partnerships (ISTP), Canada—Department of Biotechnology (DBT), India grant to RR and SSC. RR gratefully acknowledges the financial support from Canadian Institutes of Health Research (CIHR) for CIHR Chair in Advanced Cancer Diagnostics and PGW acknowledges the financial support from Alex and Simona Shnaider Chair in Thyroid Cancer. MK is the recipient of Senior Research Fellowship (SRF), Council of Scientific and Industrial Research (CSIR), New Delhi, India.
Compliance with ethical guidelines
Competing interests The authors declare that they have no competing interests.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- White EJ, Brewer G, Wilson GM. Post-transcriptional control of gene expression by AUF1: mechanisms, physiological targets, and regulation. Biochim Biophys Acta. 2013;1829:680–8.PubMed CentralView ArticlePubMedGoogle Scholar
- Gratacos FM, Brewer G. The role of AUF1 in regulated mRNA decay. Wiley Interdiscip Rev RNA. 2010;1:457–73.PubMed CentralView ArticlePubMedGoogle Scholar
- Benjamin D, Moroni C. mRNA stability and cancer: an emerging link? Expert Opin Biol Ther. 2007;7:1515–29.View ArticlePubMedGoogle Scholar
- Chen M, Zhang J, Manley JL. Turning on a fuel switch of cancer: hnRNP proteins regulate alternative splicing of pyruvate kinase mRNA. Cancer Res. 2010;70:8977–80.PubMed CentralView ArticlePubMedGoogle Scholar
- Barker A, Epis MR, Porter CJ, Hopkins BR, Wilce MC, et al. Sequence requirements for RNA binding by HuR and AUF1. J Biochem. 2012;151:423–37.View ArticlePubMedGoogle Scholar
- Gruber AR, Fallmann J, Kratochvill F, Kovarik P, Hofacker IL. AREsite: a database for the comprehensive investigation of AU-rich elements. Nucleic Acids Res. 2011;39:66–9.View ArticleGoogle Scholar
- Blaxall BC, Dwyer-Nield LD, Bauer AK, Bohlmeyer TJ, Malkinson AM, et al. Differential expression and localization of the mRNA binding proteins, AU-rich element mRNA binding protein (AUF1) and Hu antigen R (HuR), in neoplastic lung tissue. Mol Carcinog. 2000;28:76–83.View ArticlePubMedGoogle Scholar
- Braakhuis BJ, Tabor MP, Kummer JA, Leemans CR, Brakenhoff RH. A genetic explanation of Slaughter’s concept of field cancerization: evidence and clinical implications. Cancer Res. 2003;63:1727–30.PubMedGoogle Scholar
- Lapucci A, Donnini M, Papucci L, Witort E, Tempestini A, et al. AUF1 Is a bcl-2 A + U-rich element-binding protein involved in bcl-2 mRNA destabilization during apoptosis. J Biol Chem. 2002;277:16139–46.View ArticlePubMedGoogle Scholar
- Wang W, Martindale JL, Yang X, Chrest FJ, Gorospe M. Increased stability of the p16 mRNA with replicative senescence. EMBO Rep. 2005;6:158–64.PubMed CentralView ArticlePubMedGoogle Scholar
- Moore AE, Chenette DM, Larkin LC, Schneider RJ. Physiological networks and disease functions of RNA-binding protein AUF1. Wiley Interdiscip Rev RNA. 2014;5:549–64.View ArticlePubMedGoogle Scholar
- Wagner BJ, DeMaria CT, Sun Y, Wilson GM, Brewer G. Structure and genomic organization of the human AUF1 gene: alternative pre-mRNA splicing generates four protein isoforms. Genomics. 1998;48:195–202.View ArticlePubMedGoogle Scholar
- DeMaria CT, Sun Y, Long L, Wagner BJ, Brewer G. Structural determinants in AUF1 required for high affinity binding to A + U-rich elements. J Biol Chem. 1997;272:27635–43.View ArticlePubMedGoogle Scholar
- Zhang W, Wagner BJ, Ehrenman K, Schaefer AW, DeMaria CT, et al. Purification, characterization, and cDNA cloning of an AU-rich element RNA-binding protein, AUF1. Mol Cell Biol. 1993;13:7652–65.PubMed CentralPubMedGoogle Scholar
- Zucconi BE, Wilson GM. Modulation of neoplastic gene regulatory pathways by the RNA-binding factor AUF1. Front Biosci. 2011;16:2307–25.View ArticleGoogle Scholar
- Gouble A, Grazide S, Meggetto F, Mercier P, Delsol G, et al. A new player in oncogenesis: AUF1/hnRNPD overexpression leads to tumorigenesis in transgenic mice. Cancer Res. 2002;62:1489–95.PubMedGoogle Scholar
- Lu M, Pan C, Zhang L, Ding C, Chen F, et al. ING4 inhibits the translation of proto-oncogene MYC by interacting with AUF1. FEBS Lett. 2013;587:1597–604.View ArticlePubMedGoogle Scholar
- Al-Ansari MM, Hendrayani SF, Shehata AI, Aboussekhra A. p16(INK4A) represses the paracrine tumor-promoting effects of breast stromal fibroblasts. Oncogene. 2013;32:2356–64.PubMed CentralView ArticlePubMedGoogle Scholar
- Al-Khalaf HH, Colak D, Al-Saif M, Al-Bakheet A, Hendrayani SF, et al. p16(INK4a) positively regulates cyclin D1 and E2F1 through negative control of AUF1. PLoS One. 2011;6:e21111.PubMed CentralView ArticlePubMedGoogle Scholar
- Pont AR, Sadri N, Hsiao SJ, Smith S, Schneider RJ. mRNA decay factor AUF1 maintains normal aging, telomere maintenance, and suppression of senescence by activation of telomerase transcription. Mol Cell. 2012;47:5–15.PubMed CentralView ArticlePubMedGoogle Scholar
- Siegel R, DeSantis C, Virgo K, Stein K, Mariotto A, et al. Cancer treatment and survivorship statistics. CA Cancer J Clin. 2012;62:220–41.View ArticlePubMedGoogle Scholar
- Lucs AV, Saltman B, Chung CH, Steinberg BM, Schwartz DL. Opportunities and challenges facing biomarker development for personalized head and neck cancer treatment. Head Neck. 2013;35:294–306.PubMed CentralView ArticlePubMedGoogle Scholar
- Leemans CR, Braakhuis BJ, Brakenhoff RH. The molecular biology of head and neck cancer. Nat Rev Cancer. 2012;11:9–22.View ArticleGoogle Scholar
- Jemal A. Global burden of cancer: opportunities for prevention. Lancet. 2012;380:1797–9.View ArticlePubMedGoogle Scholar
- Ralhan R, Desouza LV, Matta A, Chandra Tripathi S, Ghanny S, et al. iTRAQ-multidimensional liquid chromatography and tandem mass spectrometry-based identification of potential biomarkers of oral epithelial dysplasia and novel networks between inflammation and premalignancy. J Proteome Res. 2009;8:300–9.View ArticlePubMedGoogle Scholar
- Matta A, DeSouza LV, Shukla NK, Gupta SD, Ralhan R, et al. Prognostic significance of head-and-neck cancer biomarkers previously discovered and identified using iTRAQ-labeling and multidimensional liquid chromatography-tandem mass spectrometry. J Proteome Res. 2008;7:2078–87.View ArticlePubMedGoogle Scholar
- Matta A, Tripathi SC, DeSouza LV, Grigull J, Kaur J, et al. Heterogeneous ribonucleoprotein K is a marker of oral leukoplakia and correlates with poor prognosis of squamous cell carcinoma. Int J Cancer. 2009;125:1398–406.View ArticlePubMedGoogle Scholar
- Masui O, White NM, DeSouza LV, Krakovska O, Matta A, et al. Quantitative proteomic analysis in metastatic renal cell carcinoma reveals a unique set of proteins with potential prognostic significance. Mol Cell Proteomics. 2013;12:132–44.PubMed CentralView ArticlePubMedGoogle Scholar
- Voisin SN, Krakovska O, Matta A, DeSouza LV, Romaschin AD, et al. Identification of novel molecular targets for endometrial cancer using a drill-down LC-MS/MS approach with iTRAQ. PLoS One. 2011;6:e16352.PubMed CentralView ArticlePubMedGoogle Scholar
- Shilov IV, Seymour SL, Patel AA, Loboda A, Tang WH, et al. The Paragon Algorithm, a next generation search engine that uses sequence temperature values and feature probabilities to identify peptides from tandem mass spectra. Mol Cell Proteomics. 2007;6:1638–55.View ArticlePubMedGoogle Scholar
- Mili S, Shu HJ, Zhao Y, Pinol-Roma S. Distinct RNP complexes of shuttling hnRNP proteins with pre-mRNA and mRNA: candidate intermediates in formation and export of mRNA. Mol Cell Biol. 2001;21:7307–19.PubMed CentralView ArticlePubMedGoogle Scholar
- Park HG, Yoon JY, Choi M. Heterogeneous nuclear ribonucleoprotein D/AUF1 interacts with heterogeneous nuclear ribonucleoprotein L. J Biosci. 2007;32:1263–72.View ArticlePubMedGoogle Scholar
- Li X, Johansson C, Glahder J, Mossberg AK, Schwartz S. Suppression of HPV-16 late L1 5′-splice site SD3632 by binding of hnRNP D proteins and hnRNP A2/B1 to upstream AUAGUA RNA motifs. Nucleic Acids Res. 2013;41:10488–508.PubMed CentralView ArticlePubMedGoogle Scholar
- Pascale A, Govoni S. The complex world of post-transcriptional mechanisms: is their deregulation a common link for diseases? Focus on ELAV-like RNA-binding proteins. Cell Mol Life Sci. 2012;69:501–17.View ArticlePubMedGoogle Scholar
- Sinha P, Bahadur S, Thakar A, Matta A, Macha M. Significance of promoter hypermethylation of p16 gene for margin assessment in carcinoma tongue. Head Neck. 2009;31:1423–30.View ArticlePubMedGoogle Scholar
- Wu X, Chesoni S, Rondeau G, Tempesta C, Patel R, et al. Combinatorial mRNA binding by AUF1 and Argonaute 2 controls decay of selected target mRNAs. Nucleic Acids Res. 2013;41:2644–58.PubMed CentralView ArticlePubMedGoogle Scholar
- Ying SY, Chang DC, Lin S. The MicroRNA. Methods Mol Biol. 2013;936:1–19.View ArticlePubMedGoogle Scholar
- Abdelmohsen K, Tominaga-Yamanaka K, Srikantan S, Yoon JH, Kang MJ, et al. RNA-binding protein AUF1 represses Dicer expression. Nucleic Acids Res. 2012;40:11531–44.PubMed CentralPubMedGoogle Scholar
- He C, Schneider R. 14-3-3 sigma is a p37 AUF1-binding protein that facilitates AUF1 transport and AU-rich mRNA decay. EMBO J. 2006;25:3823–31.PubMed CentralView ArticlePubMedGoogle Scholar
- Suzuki M, Iijima M, Nishimura A, Tomozoe Y, Kamei D, et al. Two separate regions essential for nuclear import of the hnRNP D nucleocytoplasmic shuttling sequence. FEBS J. 2005;272:3975–87.View ArticlePubMedGoogle Scholar
- Trojanowicz B, Brodauf L, Sekulla C, Lorenz K, Finke R. The role of AUF1 in thyroid carcinoma progression. Endocr Relat Cancer. 2009;16:857–71.View ArticlePubMedGoogle Scholar
- Langer CJ. Exploring biomarkers in head and neck cancer. Cancer. 2012;118:3882–92.View ArticlePubMedGoogle Scholar
- Sarkar S, Sinsimer KS, Foster RL, Brewer G, Pestka S. AUF1 isoform-specific regulation of anti-inflammatory IL10 expression in monocytes. J Interferon Cytokine Res. 2011;28:679–91.View ArticleGoogle Scholar
- Paschoud S, Dogar AM, Kuntz C, Grisoni-Neupert B, Richman L, et al. Destabilization of interleukin-6 mRNA requires a putative RNA stem-loop structure, an AU-rich element, and the RNA-binding protein AUF1. Mol Cell Biol. 2006;26:8228–41.PubMed CentralView ArticlePubMedGoogle Scholar
- Palanisamy V, Park NJ, Wang J, Wong DT. AUF1 and HuR proteins stabilize interleukin-8 mRNA in human saliva. J Dent Res. 2008;87:772–6.PubMed CentralView ArticlePubMedGoogle Scholar
- Brewer G, Saccani S, Sarkar S, Lewis A, Pestka S. Increased interleukin-10 mRNA stability in melanoma cells is associated with decreased levels of A + U-rich element binding factor AUF1. J Interferon Cytokine Res. 2003;23:553–64.View ArticlePubMedGoogle Scholar
- Grosset C, Chen CY, Xu N, Sonenberg N, Jacquemin-Sablon H, et al. A mechanism for translationally coupled mRNA turnover: interaction between the poly(A) tail and a c-fos RNA coding determinant via a protein complex. Cell. 2000;103:29–40.View ArticlePubMedGoogle Scholar
- Abdelmohsen K, Kuwano Y, Kim HH, Gorospe M. Posttranscriptional gene regulation by RNA-binding proteins during oxidative stress: implications for cellular senescence. Biol Chem. 2008;389:243–55.View ArticlePubMedGoogle Scholar
- Kaur J, Matta A, Kak I, Srivastava G, Assi J, et al. S100A7 overexpression is a predictive marker for high risk of malignant transformation in oral dysplasia. Int J Cancer. 2013;134:1379–88.View ArticlePubMedGoogle Scholar