Tumor Biol. DOI 10.1007/s13277-014-2909-6

RESEARCH ARTICLE

Overexpression of HOXC11 in clear cell renal cell carcinoma induces cellular proliferation and is associated with poor prognosis

Yu-Jun Liu & Yu Zhu & Hai-Xia Yuan & Jian-Ping Zhang & Jian-Ming Guo & Zong-Ming Lin

Received: 10 October 2014 /Accepted: 27 November 2014 # International Society of Oncology and BioMarkers (ISOBM) 2014

Abstract Novel evidence has confirmed the involvement of increasing tumor cell proliferation and imply that HOXC11 dysregulated expression of HOX in cancer. HOX genes may be an important determinant of RCC patient prognosis. are a family of 39 transcription factors, divided in four clusters (HOXA to HOXD), that during normal development regulate Keywords Renal cell carcinoma . Homeobox genes . cell proliferation and specific cell fate. The aim of this study HOXC11 . Proliferation . Prognosis was to investigate whether genes of the HOXC cluster might play a role in renal cancer. The expression of HOXC11 was detected through polymerase chain reaction and immunohis- Introduction tochemical staining, and we demonstrated that HOXC11 was significantly higher in renal cell carcinoma (RCC) compared Renal cell carcinoma (hereafter termed RCC) is a com- to normal kidney tissue. We further demonstrated that mon malignant neoplasm in the urinary system and the HOXC11 overexpression in HK-2 epithelial cell line sixth leading cause of cancer deaths in Western coun- promoted proliferation, whereas downregulation of HOXC11 tries. Each year, around 200,000 patients are diagnosed endogenous levels in human RCC cells (Caki-2 cells) de- with this malignancy resulting in approximately 100,000 creased proliferation. In RCC, expression of HOXC11 and deaths, and its incidence is increasing steadily in recent Ki67, a marker of proliferation, correlates strongly with each years [1, 2]. The majority of oncogenic and tumor sup- other (rs=0.47, p<0.003). High immunohistochemical ex- pressor signaling pathways involved with kidney cancer pression of HOXC11 was correlated with T stage (p=0.06), converge on sets of transcription factors that ultimately Nstage(p=0.07), disease stage (p=0.08), and Ki67 expres- control resulting in tumor formation and sion (p=0.07), and patients with tumors showing high number metastatic progression. The activity of these transcription of HOXC11-positive cells had shorter overall survival (p= factors is modulated by multiple mechanisms spanning 0.08) and shorter progression-free survival after treatment (p= from transcriptional regulation, deregulation of the splic- 0.08) compared with patients with tumors exhibiting low ing, maturation, export and location of messenger RNAs amount of HOXC11-positive cells. Our data suggest that (mRNAs), synthesis, and post-translational mod- HOXC11 may contribute to RCC carcinogenesis by ifications. The recent involvement of the epigenetic mechanisms in the generation and the evolution of can- cer has produced a great deal of interest [3, 4]. Yu-Jun Liu and Yu Zhu contributed equally to this work. HOX genes belong to the superfamily of homeobox gene that encodes for transcription factors with important roles in < : : < * : < * : < Y. J. Liu Y. Zh u J. P. Zhang ( ) J. M. Guo ( ) Z. M. Lin the development [5]. Homeoproteins control diverse cellular Department of Urology, Zhongshan Hospital, Fudan University, No. 180 Fenglin Road, Shanghai 200032, People’s Republic of China processes by regulating the expression of many downstream e-mail: [email protected] target genes. Thus, it is typical for an individual homeoprotein e-mail: [email protected] to confer pleiotropic effects on cell behavior, including alter- < ations in proliferation, survival, migration, and invasion H. X. Yuan – Department of Ultrasound, Zhongshan Hospital, Fudan University, [6 10]. In , the HOX genes are organized into four Shanghai 200032, China chromosomal clusters or loci (HOXA at 7p15.3, HOXB at Tumor Biol.

17q21.3, HOXC at 12q13.3, and HOXD at 2q31) each con- estimated that the tissue used for quantitative protein analyses taining from 9 to 11 genes [10]. Genes of the HOX family are was dominated by carcinoma cells (>50 %). highly expressed in leukemias, melanomas, and in breast, ovarian, cervical, esophageal, and prostate cancers [11–13]. Patients follow-up In addition, aberrant expression of HOX genes was observed in urogenital cancers including renal cancer [14]. However, To determine the clinicopathological correlation of the their relationship with the neoplastic phenotype remains HOXC11 expression, immunohistochemistry was performed unclear. in primary RCC of 36 patients diagnosed and treated at the In neuronal cells, forced expression of HOXC6 and Department of Urology, Zhongshan Hospital, Fudan Univer- HOXC11 has been shown to induce S100 beta expression sity, Shanghai. The RCC patients, 22 of which were males, [15]. Recently in breast cancer, a role for HOXC11 in con- ranged in age between 44 and 85 years, with a mean of 61± junction with the transcriptional activator steroid 12.9 years. Twenty-nine patients (80.6 %) had history of coactivator 1 (SRC-1) was reported in the regulation of tobacco smoking. The tumors were histologically classified S100 beta [16]. HOXC11 stimulates the proliferation and and staged according to the 1997 revised TNM system and survival of progenitor cells during normal development; how- graded according to Fuhrman’s nuclear grading system as ever, when it is expressed out of context, HOXC11 can follows: stage I (4 patients), stage II (8 patients), stage III aberrantly promote proliferation, contributing to tumorigene- (10 patients), and stage IV (14 patients) [18]. All patients were sis [17]. Although HOXC11 expression has been observed in staged as M0 at the time of diagnosis. Information regarding RCC, its role in human renal cancer has not been elucidated patient characteristics was based on patient records and regis- [14]. In this study, we analyzed the expression of HOXC11 tries. After treatment, patients were followed up bi-monthly, genes in normal kidney tissue and RCC, and investigated the and disease recurrence was histologically confirmed. The effect of HOXC11 on stimulation of proliferation and survival study was approved by the local institutional review board. of RCC cancer cells. Additionally, we have analyzed the Informed consent was obtained from all patients enrolled in prognostic significance of HOXC11 immunohistochemical the study. expression for RCC patients. Cell culture

Materials and methods The human Caki-2 cell line and HK-2 cell line were obtained from American Type Culture Collection (ATCC; Manassas, Patients and specimens VA, USA) and cultured as recommended in 1:1 mixture of Dulbecco’smodifiedEagle’s medium (DMEM) and Ham’s Thirty-six consecutive patients with RCC during February F12 medium (DMEM/F12; Invitrogen, Carlsbad, CA, USA) 2005 to July 2006 were included, all undergoing surgical supplemented with 10 % fetal bovine serum (FBS), 400 ng/ml resection for histologically verified clear cell renal cell carci- hydrocortisone (Sigma-Aldrich, St. Louis, MO, USA), noma, and they received care and regular follow-up at the 100 μg/ml penicillin, and 100 μg/ml streptomycin at 37 °C

Department of Urology, Zhongshan Hospital, Fudan Univer- in a humidified atmosphere of 5 % CO2. sity (Shanghai, China) for 5 years. Pretreatment examinations included a chest X-ray, urological ultrasound, and CT scan of Overexpression and knockdown of HOXC11 the abdomen. For detection of HOXC11 by reverse transcrip- tase polymerase chain reaction (RT-PCR), we analyzed ten Caki-2 cells were transfected with control vector pDEST47 fresh samples from normal kidneys obtained from ten healthy (400 ng) or pDEST47 HOXC11 (400 ng). Protein levels were kidney transplant donor, since the kidneys were routinely determined 24 h post-transfection by Western blot. biopsied before transplantation. And ten pairs with fresh sam- Predesigned and validated small interfering RNA (siRNA) ples each pair are from the same RCC patients. Immediately directed against HOXC11 (QIAGEN Ltd, Crawley, West Sus- after the nephrectomy operation, the removed samples were sex, UK) was used in the knockdown studies. Protein and brought to the department of clinical pathology, and the tissue mRNA levels were assessed 24 h post-transfection. from the tumor was sampled by a pathologist. All fresh samples were divided into two parts: Samples intended for Duplex RT-PCR later immunohistochemistry (IHC) was fixed in formalin and embedded in paraffin. The other samples were snap frozen in Total RNA from fresh tissues and cell lines was isolated with liquid nitrogen. The samples were frozen within 30 min of TRIzol reagent according to manufacturer’sprotocol surgical removal. Based on a microscopic examination of H & (Invitrogen). Following DNase I treatment in order to elimi- E-stained FFPE tumor sections, it was semiquantitatively nate genomic DNA contamination, 3 μg of total RNA per Tumor Biol. sample was used to generate complementary DNA (cDNA) incorporation in proliferating cells was estimated using using a superscript enzyme (Superscript II RT enzyme, an immunohistochemical analysis kit (GE Healthcare). Invitrogen). The resulting cDNAs were subsequently ampli- The BrdU-labeling index, expressed as the percentage of fied, analyzed, and quantified as previously described [19]. cells labeled with BrdU, was determined by counting Primer sequences, PCR conditions, and the amplified lengths 1500 cells in three independent reactions using the have been described [19]. Glyceraldehyde-3-phosphate dehy- Kontron 400 image analysis system (Zeiss, Geman). drogenase (GAPDH) was used as a reference gene. Linear amplification range for each gene was determined by prepar- Ki67 index ing six similar reactions but amplified by different numbers of cycles (26, 29, 32, 35, 38, and 41 cycles). For all genes, the Cells were cultured in culture chamber slides at 37 °C in chosen cycle was in the exponential phase of the PCR humidified air containing 5 % CO2 for 24 h. After cellular amplification. synchronism and cell cycle induction as described above, thecellswerefixedin70%ethanolfor1handwashed Western blot with PBS. Cells were then treated with 1 % BSA diluted in PBS for 1 h, incubated with monoclonal antibodies Cells were washed with cold phosphate-buffered saline against Ki67, and followed by the ABC method (PBS) and lysed in RIPA buffer (50 mM Tris-HCl pH (StrepABC Complex/HRP, Dako). Reactions were devel-

7.4,150mMNaCl,1mMEDTA,1%NP-40,1% oped with 0.6 mg/ml DAB containing 0.01 % H2O2.The deoxycholic acid, 0.5 % sodium dodecyl sulfate, 1 mM Ki67 index was calculated using an image analysis system phenylmethylsulfonyl fluoride, 1 mM N-ethylmaleimide, by counting labeled nuclei of 1500 cells in three indepen- 1 mM dithiothreitol, 10 μg/ml soybean trypsin inhibitor, dent reactions and expressed as percentage of Ki67- 1 μg/ml leupeptin, and 1 μg/ml aprotinin). After centri- positive cells. fugation, protein concentrations were measured using a proteinassayaccordingtomanufacturer’s instructions Apoptosis analysis (Bio-Rad Protein Assay; Bio-Rad, Hercules, CA). Eighty micrograms of total protein per sample was Apoptosis index was determined by annexin V-fluorescein resolved in a 10 % sodium dodecyl sulfate polyacryl- isothiocyanate (FITC) labeling. Briefly, cells were harvested, amide gel electrophoresis (SDS-PAGE) under reducing washed with PBS, and resuspended in the binding buffer conditions and transferred to nitrocellulose membranes. (10 mM HEPES pH 7.4, 150 mM NaCl, 5 mM KCl, 1 mM

Themembraneswereblockedfor2hwith10%nonfat MgCl2, and 1.8 mM CaCl2) containing annexin V-FITC at dry milk in PBS containing 0.1 % Tween-20, rinsed in 1:500. After 20 min of incubation in the dark at room temper- the same buffer, and incubated for 1 h with anti- ature, cells were also stained with propidium iodide (PI; HOXC11 antibodies (1:200, sc-788; Santa Cruz Biotech- Sigma-Aldrich). Apoptosis was analyzed on a FACSCalibur nology) diluted 1:200 or with anti-β-actin antibodies flow cytometer equipped with an argon laser (Becton- (A1978; Sigma-Aldrich China, Shanghai, China) diluted Dickinson, San Jose, CA, USA) and quantified as the number 1:7500 in 5 % milk in PBS. After washing, the mem- of annexin V-FITC-positive and PI-negative cells divided by branes were developed using an Enhanced Chemilumi- the total number of cells. A minimum of 10,000 events was nescence Western blot kit (GE Healthcare, Vienna, analyzed in each sample. Austria). Statistical analysis Bromodeoxyuridine-labeling (BrdU) index Kruskal-Wallis multiple comparison test was used to an- Cells were plated in 8-well chamber slides at a density alyze HOXC11 mRNA and protein levels between of 30,000 cells per well in 500 μl of medium contain- groups. Student’s t test (bi-caudal) was used to compare ing 10 % FBS. After 16 h, the cells were washed with the proliferation of transfectant cell lines and of Caki-2 PBS and cultured in serum-free medium for an addi- under siRNA treatment. Spearman rank correlation test tional 24 h. Following serum starvation, the medium (rs) was assessed to verify the association between immu- was replaced by 10 % FBS medium. Proliferation rates nohistochemical expression of HOXC11 and Ki67. Cor- were determined 24 h after incubation by measuring relations between immunohistochemical expression of BrdU incorporation into DNA. Briefly, BrdU antigen HOXC11 and clinicopathological parameters of the tu- was added to the cultures and kept for 1 h at 37 °C mors were performed by Fisher’s exact test and continu- in5%CO2. After incubation, cells were washed in ous variables by nonparametric Mann-Whitney U test. PBS and fixed in 70 % ethanol for 1 h. BrdU P<0.05 was considered to be statistically significant. Tumor Biol.

Results HOXC11 overexpression promotes a proliferative phenotype in RCC cells Expression of HOXC11 genes in normal kidney tissue and RCC To determine whether HOXC11 overexpression contributes to the tumorigenic phenotype and, more specifically, to prolifer- As the previous literature reported, most members of HOXC ation and apoptosis, we next overexpressed HOXC11 in the were expressed at very low levels in the adult kidney, among HK-2 cell line, which does not show detectable endogenous which the HOXC11 that is located at the chromosomal area levels of this protein. Stable HOXC11 and control 12q13-15 was highly expressed [14]. Considering the inten- transfectants were generated and examined for HOXC11 pro- sity levels of gene expression, as represented by the densito- tein levels. Three stable HOXC11-overexpressing clones metric ratio of the optical density of target transcript/GAPDH (HK-2-HOXC11), and three control clones (HK-2-Control) bands, we identified that the gene HOXC11 is significantly were chosen for further analyses (Fig. 3a). more expressed in RCC samples compared to normal kidney HOXC11-overexpressing cells showed a statistically tissue samples from healthy donor kidneys (p<0.001), and significant increase in proliferation, as assessed by mea- tumor-adjacent tissue from patients received RCC operation suring BrdU incorporation into DNA and Ki67 expression (p<0.05) (Fig. 1). (Fig. 3b, c). Consistently, the analysis of RCC samples To confirm HOXC11 higher expression in RCC samples depicted in Fig. 1 for Ki67 immunohistochemical expres- compared to normal kidneys, we performed immunohisto- sion demonstrated a statistically significant correlation chemical analysis. Immunoreactivity for HOXC11 was ob- between percentage of HOXC11-positive cells and Ki67- served as a nuclear stain restricted to the basal layer in the positive cells (rs =0.80, p<0.005). In order to verify normal kidneys, whereas a broad positivity with variable whether the endogenous function of HOXC11 is to pro- distribution and intensity was found in the RCC patients mote proliferation, it was knocked down in epithelial- (Fig. 2a, b). As expected, RCC showed HOXC11-labeling derived HK-2 cells (HK-2 cells). Strand-specific RNA index significantly higher than the mean of normal kidney oligonucleotides against HOXC11 were transfected into group (p<0.001; Fig. 2c). The percentage of HOXC11- HK-2 cells. When HOXC11 knockdown was fulfilled, a positive cells in normal kidney group ranged between 0 and rapid downregulation of HOXC11 mRNA and protein was 16 %, with a median of 5.5 % (mean 7.7±6.79 %), whereas observed (Fig. 4a, b). The decrease in HOXC11 levels RCC group ranged between 12 and 64 %, with a median of resulted in a concomitant decrease in proliferation. Cell 32.4 % (mean 32.4±15.67 %). proliferation assays (BrdU incorporation and Ki67

Fig. 1 HOXC11 is overexpressed in RCCs. Total RNA from fresh Densitometric analysis of the HOXC11 bands demonstrated a significant samples was isolated and cDNA synthesized by RT-PCR. a Duplex RT- higher expression in RCCs compared to the normal kidney, independent PCR analysis of HOXC11 on the normal kidney from patients without of the source (p<0.001 when compared to normal from patients without contact with the main risk factors to renal cancer. b HOXC11 analysis on recognized RCC risk factors, and p<0.05 when compared to normal from the pairs of RCC and adjacent normal kidney from the same patient. c patients with RCC) Tumor Biol.

Fig. 2 Immunohistochemical detection of HOXC11 in the normal suprabasal layers, whereas in the RCCs, HOXC11-positive cells were kidney and RCC. Representative samples of the normal kidney (a)and broadly observed in the tumor (original magnification ×200). c RCC (b) of this study. In the normal kidney, HOXC11 expression was Percentage of HOXC11-positive cells was significantly higher in RCCs clearly limited to the nucleus of the epithelial cells located in the basal and than in normal kidneys (p<0.001)

Fig. 3 We overexpressed HOXC11 in the HK-2 cell line, which does not Control) were chosen for further analyses. b, c HOXC11-overexpressing show detectable endogenous levels of this protein, to determine whether cells showed a statistically significant increase in proliferation, as HOXC11 overexpression contributes to the tumorigenic phenotype and, assessed by measuring BrdU incorporation into DNA and Ki67 more specifically, to proliferation and apoptosis. a Three stable HOXC11- expression. overexpressing clones (HK-2-HOXC11) and three control clones (HK-2- Tumor Biol.

Fig. 4 HOXC11 was knocked down in epithelial-derived HK-2 cells decrease in HOXC11 levels resulted in a concomitant decrease in (HK-2 cells) in order to verify whether its endogenous function is to proliferation. c, d Cell proliferation assays (BrdU incorporation and promote proliferation. a, b A rapid downregulation of HOXC11 mRNA Ki67 expression) showed a statistically significant decrease in and protein was observed when HOXC11-knockdown was fulfilled. The proliferation when HOXC11 was downregulated expression) showed a statistically significant decrease in in the tumor was correlated with T stage (p=0.06), N proliferation when HOXC11 was downregulated stage (p=0.07), and disease stage (p=0.08). Local re- (Fig. 4c, d). The number of annexin V-FITC-positive cells currence was observed in only three patients of our was very low in both HOXC11-overexpressing cells and group, but all demonstrated tumors classified as high HOXC11 downregulated cells, and no significant differ- expression of HOXC11. Expression of HOXC11 was ences from controls were observed. not associated with age, gender, tumor cellular differen- tiation, or tumor recurrence. We found that tumors that HOXC11 expression correlates with patients’ overexpressed HOXC11 showed statistically significant clinicopathological features and prognosis increase in the percentage of Ki67-positive cells com- pared with low-expressing tumors (Table 1). Further- To investigate whether HOXC11 expression is associat- more, there was a significant correlation between per- ed with clinicopathological features of RCC and with centage of HOXC11- and Ki67-positive cells in the the proliferative potential of the tumor, we performed studied tumors (rs=0.47, p=0.003). Patients whose tu- immunohistochemistry on 36 human RCC patients to mors contained high number of HOXC11-positive cells examine HOXC11 expression concurrent with Ki67 pro- had shorter overall survival than patients with low num- liferative marker. Parallel slides were analyzed for berofKi67(p=0.07; Fig. 5a). Similarly, patients with HOXC11 and Ki67 expression. The clinicopathological tumors showing high HOXC11 expression had shorter correlations with the expression of HOXC11 are listed progression-free survival after treatment (p =0.08; in Table 1. Elevated amount of HOXC11-positive cells Fig. 5b). Taken together, these data demonstrate that Tumor Biol.

Table 1 Relationship between HOXC11 immunohistochemical Discussion expression and clinic pathological variables of the RCCs

Characteristic HOXC11-positive cells (%) p value Homeobox genes are a superfamily of originally identified transcription factors mostly involved with the determination <40 (%) >40 (%) of the developmental identity of animal body plan [8]. The network organization has evolutionarily evolved Age 0.32 through subsequent replication and transposition events from ≤60 (y) 11 (61.11) 8 (44.44) a unique ancestral (proto-Hox) gene [20]. During mammalian >60 (y) 7 (38.89) 10 (55.56) development, HOX gene expression controls the identity of Gender 0.78 various regions along the body axis according to the rules of Male 12 (42.86) 10 (55.56) spatiotemporal colinearity, with 3′ HOX genes (retinoic acid Female 6 (57.14) 8 (44.44) responsive) expressed early in development and controlling T Stage 0.06 anterior regions, followed by progressively more 5′ genes T1 & T2 12 (63.16) 8 (47.06) (FGF responsive) expressed later and controlling more poste- T3 & T4 7 (36.84) 9 (52.94) rior regions [21]. Since the first description of altered expres- N Stage 0.07 sion in leukemias, dysregulation of HOX gene expression is N0 10 (52.63) 6 (35.29) commonly observed in human cancers [22]. Typically, those N+ 9 (47.37) 11 (64.71) genes that are overexpressed in cancers show normal expres- Stage 0.08 sion patterns that are restricted to the undifferentiated or Stage I + II 8 (44.44) 7 (38.89) proliferative cells, whereas those that are lost in cancer are Stage III + IV 10 (55.56) 11 (61.11) expressed in fully differentiated cells [23]. Moreover, the Fuhrman grade 0.10 classical mechanism of sequential activation of HOX genes 1 & 2 5 (35.71) 7 (31.82) is lost in cancers, and specific functions are played by single 3 & 4 9 (64.29) 15 (68.18) genes [8]. Although these powerful developmental regulators Recurrence 0.08 are known to affect numerous important processes in embryo- Yes 13 (76.47) 14 (73.68) genesis and tumorigenesis, a causal role for HOX genes in No 4 (23.53) 5 (26.32) tumor initiation and/or progression remains controversial [24]. Ki67-positive cells 0.03 To identify differentially expressed genes, this study ana- ≤40 % 11 (64.70) 7 (36.84) lyzed HOX genes belonging to cluster C in pairs of histolog- >40 % 6 (35.29) 12 (63.16) ically normal kidney and renal cancer and in samples of the normal kidney obtained from healthy donor. In normal kid- neys, paralogous group 11 HOX genes in triple-mutant mice overexpression of HOXC11 in RCC is associated with show bilateral kidney agenesis because of alterations in the several clinicopathological features, including elevated epithelial-mesenchymal interaction, consequently preventing proliferative potential and worsened survival. ureteric bud outgrowth [25]. According to published data,

Fig. 5 HOXC11 expression is associated with poor prognosis. Curves of exhibiting low (<40 % of positive cells) and high (>40 % of positive cells) a overall survival and b progression-free survival for the patients of this HOXC11 levels. Patients with high number of HOXC11-positive tumor study. The median value of HOXC11-positive cells was employed to cells had shorter global survival period and progression-free survival after divide tumors into two groups, below and above the respective medians, treatment than patients with low number of HOXC11-positive tumor cells Tumor Biol.

HOXA11, the effector of which is α8 integrin, is again active six genes (HOXC4, HOXC5, HOXC6, and HOXC11), con- at the 15th week of development; HOXC11 is expressed from stitutively silent in the normal bladder, became active in the the 18th week and HOXD11 resumes being active from the majority of the bladder cancer. Our study also revealed that 23rd week until birth and is always active in normal adult modulation of HOXC11 levels had no dramatic effect on cell human kidneys [26]. However, among these gene networks, death. Accordingly, Walsh et al. demonstrated no direct effect we detected the expression of HOXC11 and found it signifi- of HOXC11 overexpression on cell death in breast cancers; it cantly higher in RCC compared with the normal kidney. In is SRC-1 that targets a series of genes important in apoptosis addition to the mRNA quantification, we showed that and cellular differentiation [29]. Furthermore, through bioin- HOXC11 protein expression is limited to the normal epitheli- formatic analysis and liquid chromatography/mass spectrom- um, while that cancers show strong and broad immunoreac- etry, they identified AP1 and jumonji domain con- tivity in the neoplastic cells. As expected, the number of taining 2C (JMD2C/KDM4C) as members of the SRC-1 HOXC11-positive cells was significantly higher in RCC than interactome responsible for transcriptional repression [29]. that of controls. Noteworthy, we have further demonstrated that abun- HOXC11 plays a critical role in expanding progenitor cells dant presence of HOXC11-positive cells leads to a more during normal development via its stimulation of proliferation aggressive behavior of the RCC. Tumors with high and upregulation of HOXC11 activity is observed in several percent of HOXC11-positive cells tended to be in ad- tumors, suggesting that when expressed out of context, it may vanced clinical stage of the disease (higher T stage, contribute to tumorigenesis by reinstating properties normally positive lymph node metastasis, and late disease stage), conveyed on developing cells [27]. Since HOXC11 is impor- and survival analysis showed that patients whose tumors tant for the expansion of progenitor cell populations during had high number of HOXC11-positive cells had a short- embryonic development, an increase in proliferation coupled ened overall survival and progression-free survival. Poor with a decrease in apoptosis would have dire consequences in prognosis associated with high levels of HOXC11 was an adult epithelium. Literature reported that HOXC11 con- independent of other factors, suggesting that HOXC11 tributes to tumor cell proliferation in the breast cancer, in expression may be an important independent prognostic which it is closely related with resident cell dedifferentiation indicator of poor survival for RCC patients. In addition, [28]. Overexpression of HOXC11 served as a transcription a most recent study showed that a large amount of factor partner for steroid receptor coactivator 1 (SRC-1), noncoding RNAs (ncRNAs) has been identified inside which can activate genes to mediate increased tumor plasticity four human HOX loci [30]. Between them, a 2.2-kb and adaptation to the therapeutic environment [29]. Recently, ncRNA residing in the HOXC locus on Walsh et al. reported that the differentiation marker CD24 and 12q13-15 between HOXC11 and HOXC12, termed the apoptotic protein PAWR were the HOXC11/SRC-1 sup- HOTAIR, is able to control in-cis transcription of adja- pression targets [29]. Overexpression of HOXC11 inhibited cent genes and repress in-trans transcription across tumor onset while promoted tumor growth and metastasis. 40 kb of the 5′-lumbosacral HOXD locus (from Taken together, these data strongly support an oncogenic role HOXD9 to HOXD13). The upregulation of HOXC11 for HOXC11. that we detected in RCCs may be connected with (i) Overexpression of HOXC11 was observed in kidney can- the involvement of HOTAIR in kidney cancers and (ii) cer [14]. Paralogous group 11 HOX genes (HOXA11, the downregulation of posterior HOXD genes in RCCs. HOXC11, and HOXD11), the ablation of which determines Further investigation to address the role of HOXC11- kidney agenesis, seem to be also involved in kidney carcino- HOTAIR interaction in the etiology of RCC carcinogen- genesis [14]. In the kidney, HOXC11 expression was detected esis is required. in normal and neoplastic samples, with levels significantly In conclusion, the evidence presented in this report that higher in RCC than in the normal kidney [14]. The results HOXC11 is differentially expressed between normal kidney presented here clearly demonstrate that HOXC11 overexpres- tissues and RCC, induces proliferative activity, and its expres- sion accelerated cell cycle progression, and a decrease in sion correlates with features related to poor prognosis of RCC HOXC11 levels slowed the progression of cells through the patients strongly support a role for this gene in RCC carcino- cell cycle. In addition, HOXC11 levels correlated with the genesis. Further pivotal clinical trials with the purpose of proliferative potential of the oral tumor cells, as revealed targeting HOXC11 to achieve therapeutic effect in RCC pa- in vivo by significant correlation between HOXC11 and tients are expected. Ki67 levels. As a transcriptional factor with upstream activa- tors, downstream unknown targets, and functional partners, Acknowledgments Special thanks to the faculty of Department of various functions of HOXC11 can impinge on normal cell Pathology, Zhongshan Hospital, Fudan University. proliferation. It is already known that a dramatic variation is detected in the expression of the HOXC locus, as four of the Conflicts of interest None. Tumor Biol.

References 16. McIlroy M, McCartan D, Early S, Gaora OP, Pennington S, Hill AD, et al. Interaction of developmental HOXC11 with steroid receptor coactivator SRC-1 mediates re- 1. Miyamoto H, Miller JS, Fajardo DA, Lee TK, Netto GJ, Epstein JI, sistance to endocrine therapy in breast cancer. Cancer Res. et al. Non-invasive papillary urothelial neoplasms: the 2004 WHO/ 2010;70(4):1585–94. ISUP classification system. Pathol Int. 2010;60:1–8. 17. Akbas GE, Taylor HS. HOXC and HOXD gene expression in human 2. Montironi R, Santinelli A, Pomante R, Mazzucchelli R, Colanzi P, endometrium: lack of redundancy with HOXA paralogs. Biol Filho AL, et al. Morphometric index of adult renal cell carcinoma. Reprod. 2004;70(1):39–45. Comparison with the Fuhrman grading system. Virchows Arch. 18. Ficarra V, Galfano A, Mancini M, Martignoni G, Artibani W. TNM 2000;437(1):82–9. staging system for renal-cell carcinoma: current status and future 3. Robinson CM, Ohh M. The multifaceted von Hippel-Lindau tumour perspectives. Lancet Oncol. 2007;8(6):554–8. suppressor protein. FEBS Lett. 2014;588(16):2704–11. 19. Bitu CC, Destro MF, Carrera M, da Silva SD, Graner E, Kowalski LP, 4. Keefe SM, Nathanson KL, Rathmell WK. The molecular biology of et al. HOXA1 is overexpressed in oral squamous cell carcinomas and renal cell carcinoma. Semin Oncol. 2013;40(4):421–8. its expression is correlated with poor prognosis. BMC Cancer. 5. Philippidou P, Dasen JS. Hox genes: choreographers in neural devel- 2012;12:146. opment, architects of circuit organization. Neuron. 2013;80(1):12–34. 20. Gehring WJ, Kloter U, Suga H. Evolution of the Hox gene 6. Del Bene F, Wittbrodt J. Cell cycle control by homeobox genes in complexfromanevolutionarygroundstate.CurrTopDev development and disease. Semin Cell Dev Biol. 2005;16(3):449–60. Biol. 2009;88:35–61. 7. Chen H, Sukumar S. Role of homeobox genes in normal mammary 21. Mallo M, Vinagre T, Carapuço M. The road to the vertebral formula. gland development and breast tumorigenesis. J Mammary Gland Biol Int J DevBiol. 2009;53(8–10):1469–81. Neoplasia. 2003;8(2):159–75. 22. Kongsuwan K, Webb E, Housiaux P, Adams JM. Expression of 8. Samuel S, Naora H. Homeobox gene expression in cancer: insights multiple homeobox genes within diverse mammalian haemopoietic from developmental regulation and deregulation. Eur J Cancer. lineages. EMBO J. 1988;7(7):2131–8. 2005;41(16):2428–37. 23. Christensen KL, Patrick AN, McCoy EL, Ford HL. The six family of 9. Abate-Shen C. Deregulated homeobox gene expression in cancer: homeobox genes in development and cancer. Adv Cancer Res. cause or consequence. Nat Rev Cancer. 2002;2(10):777–85. 2008;101:93–126. 10. Chen H, Sukumar S. HOX genes: emerging stars in cancer. Cancer 24. Jin K, Sukumar S. BRCA1: linking HOX to breast cancer suppres- Biol Ther. 2003;2(10):524–5. sion. Breast Cancer Res. 2010;12(4):306. 11. Maeda K, Hamada J, Takahashi Y,Tada M, Yamamoto Y,Sugihara T, 25. Tripathi P, Guo Q, Wang Y, Coussens M, Liapis H, Jain S, et al. et al. Altered expressions of HOX genes in human cutaneous malig- Midline signaling regulates kidney positioning but not nephrogenesis nant melanoma. Int J Cancer. 2005;114(3):436–41. through Shh. Dev Biol. 2010;340(2):518–27. 12. Alharbi RA, Pettengell R, Pandha HS, Morgan R. The role of HOX 26. Saxèn L. Organogenesis of the kidney. Cambridge: Cambridge Univ. genes in normal hematopoiesis and acute leukemia. Leukemia. Press; 1987. p. 1–173. 2013;27(5):1000–8. 27. Wu MY, Eldin KW, Beaudet AL. Identification of chromatin remod- 13. Javed S, Langley SE. Importance of HOX genes in normal prostate eling genes Arid4a and Arid4b as leukemia suppressor genes. J Natl gland formation, prostate cancer development and its early detection. Cancer Inst. 2008;100(17):1247–59. BJU Int. 2014;113(4):535–40. 28. Hur H, Lee JY, Yun HJ, Park BW, Kim MH. Analysis of HOX gene 14. Cantile M, Franco R, Schiavo G, Procino A, Cindolo L, Botti G, et al. expression patterns in human breast cancer. Mol Biotechnol. The HOX genes network in uro-genital cancers: mechanisms and 2014;56(1):64–71. potential therapeutic implications. Curr Med Chem. 2011;18(32): 29. Walsh CA, Bolger JC, Byrne C, Cocchiglia S, Hao Y, Fagan A, et al. 4872–84. Global gene repression by the steroid receptor coactivator SRC-1 15. Zhang X, Hamada J, Nishimoto A, Takahashi Y, Murai T, Tada M, promotes oncogenesis. Cancer Res. 2014;74(9):2533–44. et al. HOXC6 and HOXC11 increase transcription of S100beta gene 30. Woo CJ, Kharchenko PV, Daheron L, Park PJ, Kingston RE. A in BrdU-induced in vitro differentiation of GOTO neuroblastoma region of the human HOXD cluster that confers polycomb-group cells into Schwannian cells. J Cell Mol Med. 2007;11(2):299–306. responsiveness. Cell. 2010;140(1):99–110.