Study on Hoxbs of Clear Cell Renal Cell Carcinoma and Detection of New Molecular Target

Hindawi Journal of Oncology Volume 2021, Article ID 5541423, 15 pages https://doi.org/10.1155/2021/5541423

Research Article Study on HOXBs of Clear Cell Renal Cell Carcinoma and Detection of New Molecular Target

Guangzhen Wu ,1 Xiaowei Li,1 Yuanxin Liu,1 Quanlin Li ,1 Yingkun Xu ,2 and Qifei Wang 1

1Department of Urology, e First Aﬃliated Hospital of Dalian Medical University, Dalian 116011, China 2Department of Urology, Shandong Provincial Hospital, Cheeloo College of Medicine, Shandong University, Jinan 250021, China

Correspondence should be addressed to Quanlin Li; [email protected], Yingkun Xu; [email protected], and Qifei Wang; [email protected]

Received 14 February 2021; Revised 4 April 2021; Accepted 12 June 2021; Published 8 July 2021

Academic Editor: Varun S. Nair

Copyright © 2021 Guangzhen Wu et al. 6is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Our study examined the transcriptional and survival data of HOXBs in patients with clear cell renal cell carcinoma (ccRCC) from the ONCOMINE database, Human Protein Atlas, and STRING website. We discovered that the expression levels of HOXB3/5/6/ 8/9 were significantly lower in ccRCC than in normal nephritic tissues. In ccRCC, patients with a high expression of HOXB2/5/6/ 7/8/9 mRNA have a higher overall survival (OS) than patients with low expression. Further analysis by the GSCALite website revealed that the methylation of HOXB3/5/6/8 in ccRCC was significantly negatively correlated to gene expression, while HOXB5/ 9 was positively correlated to the CCT036477 drug target. As DNA abnormal methylation is one of the mechanisms of tumorigenesis, we hypothesized that HOXB5/6/8/9 are potential therapeutic targets for patients with ccRCC. We analyzed the function of enrichment data of HOXBs in patients with ccRCC from the Kyoto Encyclopedia of Genes and Genomes pathway enrichment and the PANTHER pathway. 6e results of the analysis show that the function of HOXBs might be associated with the Wnt pathway and that HOXB5/6/8/9 was coexpressed with multiple Wnt pathway classical genes and proteins, such as MYC, CTNNB, Cyclin D1 (CCND1), and tumor protein P53 (TP53), which further confirms that HOXBs inhibit the growth of renal carcinoma cells through the Wnt signaling pathway. In conclusion, our analysis of the family of HOXBs and their molecular mechanism may provide a theoretical basis for further research.

1. Introduction increases clinical drug resistance [2, 3]. 6erefore, new indicators for clinical diagnosis and prognosis and thera- Morbidity due to renal carcinoma is the third most peutic targets for renal cell carcinoma are urgently needed. common among urinary tumors [1] and increases annually. On the etiology and pathogenesis of tumors, Reya and It is higher in males than in females [2]. 6e onset of renal Clevers [5] concluded that tumors may result from ab- carcinoma is unremarkable, with no specific symptoms and normal organogenesis as DNA methylation could regulate signs in the early stage, thus making an early diagnosis the opening and closing of gene expression, which is as- difficult. Moreover, metastases occur in approximately sociated with many diseases and physiological processes 20–30% of patients at the time of diagnosis. Conventional such as tumorigenesis and embryonic development. More treatment methods for metastatic renal carcinoma have interestingly, the protein encoded by HOXB acts as a reportedly poor outcomes [3, 4]. Recently, a number of new nuclear transcription factor in vertebrate embryonic de- methods have been introduced, such as targeted therapy velopment, which codes for limb formation, regulates the and immunotherapy. Despite their effectiveness, these still genitourinary system, and specifies species diversity in the have limitations. 6e administration of drugs gradually anterior and posterior axes of fetal bodies [6]. 2 Journal of Oncology

At present, it is well established that the composition of 2. Materials and Methods the HOXB cluster is highly conserved among all vertebrates, and the HOXB cluster is composed of B1–B9 and B13. 2.1. ONCOMINE. ONCOMINE (https://www.oncomine. HOXB genes all contain homeoboxes, with 70%–80% ho- org/resource/main.html) is by far the largest cancer data- mology between genes. 6ey have the ability to bind DNA base, with the complete cancer mutation spectrum, gene and thus participate in the regulation of gene expression. expression database, and related clinical information, which However, HOXB has different effects on each tumor species. is conducive for discovering new biomarkers and thera- For example, in lung cancer, HOXB9, as a classical Wnt/TCF peutic targets. 6e ONCOMINE online database is used to signaling pathway target, can significantly improve the in- analyze the mRNA expression levels of HOXBs in different vasion and metastatic potential of tumor cells and directly tumor tissues. mediate the brain metastasis of lung adenocarcinoma [7]. HOXB2 can promote the metastasis of NSCLC by regulating 2.2. GEPIA Dataset. GEPIA (http://gepia.cancer-pku.cn/) is the metastasis-related genes and serve as an index when a web page that integrates cancer data from 6e Cancer judging the prognosis of stage I lung cancer. In gastroin- Genome Atlas (TCGA) and normal tissue data from the testinal cancer, there is a stage differential HOXB13 ex- Genotype-Tissue Expression (GTEx) [16]. In this database, pression in esophageal cancer. 6e expression of HOXB13 is data can be collected more simply and quickly. We analyzed increased in the submucosa or muscular layer, until the the differential expression of HOXBs in tumors, normal cancer tissue infiltrates into the adventitia [8]. HOXB9 has a tissues, and different tumor stages. P < 0.05 and |log2FC| > 1 low expression trend in gastric carcinoma. 6e more inferior were considered statistically significant. expression, the lesser the differentiated degree of cancer tissue, the earlier the lymph node metastasis, and the shorter the OS of patients. 6erefore, the low expression of HOXB9 2.3. e Human Protein Atlas. 6e Human Protein Atlas may be an independent risk factor for poor prognosis of (http://www.proteinatlas.org/) provides the distribution of gastric carcinoma [9]. HOXB3/8/9 significantly induce co- human protein in human tissues and cells, using immu- lorectal carcinoma of the ascending colon. In breast cancer, nohistochemical technology to describe the distribution and HOXB7/9 could transform human mammary epithelial cells expression of each protein in the body [17–19]. 6rough this MCF10A and induce epithelial-mesenchymal transition database, we analyzed the OS of HOXBs in ccRCC and (EMT). calculated the hazard ratio and 95% confidence interval. Moreover, the high expression of HOXB9 reduces the P < 0.05 was considered as statistically significant. disease-free survival rate of patients with breast cancer, which may be useful as an independent prognostic factor [10, 11]. In ovarian cancer, HOXB7 can enhance cell pro- 2.4. GSCALite Website. GSCALite (http://bioinfo.life.hust. liferation by promoting ovarian cancer cells to secrete more edu.cn/web/GSCALite/) is a cancer genome analysis plat- bFGF. HOXB7/13 can improve the invasion and metastatic form [20]. 6e report provided by GSCALite includes gene potential of tumor cells. HOXB13 can also resist tamoxifen- differential expression, overall survival, single nucleotide induced apoptosis of cancer cells [12] and can inhibit the variation, copy number variation, methylation, pathway ac- growth of prostate cancer cells through the Wnt signal tivity, miRNA regulation, normal tissue expression, and drug pathway and androgen receptor signal pathway [13]. In sensitivity. We used this online tool to explore the relationship other tumors, the functional diversity of HOXB also shows between HOXBs and cancer pathways and finally to inves- that HOXB7 could promote the progression and metastasis tigate their sensitivity to multiple anticancer drugs. P < 0.05 of pancreatic cancer [14]; metastasis and poor prognosis of was regarded as statistically significant. bladder cancer are closely associated with the ectopic expression of HOXB13 in the cytoplasm [15]. 2.5. STRING Online Database. 6e STRING online database In this study, we analyzed the expression and prognosis (https//string-db.org/) was used to analyze the relationship of HOXBs, family member molecules of ccRCCs, using data between HOXBs and screen genes that interact with HOXB collected from the ONCOMINE database, Human Protein [21, 22]. 6rough DAVID (https//david.ncifcrf.gov/home.jsp), Atlas, STRING, and GSCALite. We found that expression we analyzed the GO, the KEGG pathway, and the PANTHER levels of HOXB3/5/6/8/9 were significantly lower in ccRCC pathway. P < 0.05 was considered for statistical significance. than in normal nephritic tissues. Furthermore, the expression level of the mRNA factor of HOXB2/5/6/7/8/9 in ccRCC was significantly associated with patient prognosis, 2.6. cBioPortal Analysis. 6e cBioPortal for Cancer Geno- comparatively, while that of HOXB3/4 was not. We used the mics (http://www.cbioportal.org/) provides molecular pro- STRING website to associate 40 genes with their closely filing and clinical prognostic relevance for cancer genome related HOXBs. After that, we used the GSCALite to analyze datasets. Using the cBioPortal online tool, we downloaded the GO and the KEGG pathway analysis on 40 genes. In the ccRCC with 537 samples (TCGA, Provisional) and used the end, we discovered that HOXBs participated in the em- CLUSTIVIS online tool for visualization. P < 0.05 was bryonic development of living organisms. considered statistically significant. Journal of Oncology 3

3. Results number of heterozygotes. CNV amplifications and deletions occur in a large number of heterozygotes in KIRP and KICH, 3.1. Transcriptional Levels of HOXB in Patients with RCC. respectively (Figures 5(b) and 5(d)). We also found that the 6e ONCOMINE database is a classic oncogene chip da- relevance between CNV and mRNA expression of HOXBs in tabase that integrates part of the data of the TCGA and GEO, RCC was different. In KICH, HOXB3/4 and HOXB1 CNV aids in the screening of valuable target molecules or pre- were negatively correlated to mRNA expression, while dicting phenotypes, and provides various analysis tools. 6is HOXB5/7/9 CNV was positively correlated to the same database allows visual display of analysis of cancer and (Figure 5(c)). DNA methylation aberrance is the primary normal tissue differential expressions and coexpression transcriptional regulation mechanism of tumorigenesis analysis and can be helpful in the analysis of drug sensitivity, [25, 26]. mutation, or methylation-induced changes in gene ex- In our study, the methylation of HOXB in KIRC and pression. In this study, we analyzed the mRNA levels of KIRP was upregulated compared to that of normal renal HOXBs in tumors and normal tissue and found that the tissue. Also, in KIRC, KIRP, and KICH, the methylation of mRNA levels of HOXB1/6/8/9 are significantly down- HOXB3/5/6/8 was significantly negatively correlated to the regulated in renal cell carcinoma, while those of HOXB3/4/7 expression (Figures 5(e) and 5(g)). We found that the are upregulated in nearly all the tumors, including renal cell expressions of HOXBs were different in three types of renal carcinoma. In addition, HOXB2/7 have an obvious high cancers, but all showed an anticancer effect. Using the expression of mRNA in a variety of tumors (Figure 1). GSCALite to analyze the cancer potential pathways of HOXBs, we found that HOXB9 promotes DNA damage response, HOXB9/13 promote cell apoptosis, HOXB6/7 3.2. Relationship between the mRNA Level of HOXBs and the suppress PI3K/AKTclassic cancer pathway, HOXB4/5/6/7/ Clinicopathological Characteristics of Patients with RCC. 9 suppress receptor tyrosine kinase (RTK) pathway, We used the GEPIA dataset to analyze the 10 genes of HOXB HOXB3/5 promote hormone androgen receptor (AR), in patients with kidney cancer. In comparison with normal HOXB3/4/5/9 suppress hormone estrogen receptor (ER), renal tissues, the expressions of HOXB3/5/6/8/9 in KRC, HOXB2/6 suppress cell cycle, HOXB3/6/7/9 suppress HOXB5/8 in KICH, and HOXB6 in KIRP were significantly hormone ER, and HOXB7 suppresses RTK pathway. In- downregulated (Figures 2(a)–2(k)). Because the morbidity of terestingly, HOXB2/4/6 are also inducers of EMT ccRCC is as high as 75% in all types of nephritic carcinoma (Figures 5(f) and 5(i)). In our analysis of the relationship [23] and HOXB shows a significant difference in ccRCC, we between HOXBs and a variety of traditional tumor drug decided to primarily investigate ccRCC and analyzed the targets, we found that HOXB5/7/9/13 were positively relationship between the expression of HOXBs and tumor correlated to CCT036477 drug targets; HOXB7/8/9/13 stage in ccRCC. Significant differences were observed in were closely correlated to tretinoin, KW-2449, doxorubi- HOXB5, HOXB6, HOXB7, and HOXB8 groups but not in cin, and SB-225002; and HOXB9 was positively correlated HOXB1, HOXB2, HOXB3, HOXB4, and HOXB9 groups to PX-12 drug activity. CCT036477 is a classical Wnt signal (Figures 3(a)–3(j)). inhibitor that inhibits the growth of various cancer cells and the development of embryos and the expression of Wnt 3.3. Difference of HOXBs Protein Levels between ccRCC and target genes (Figure 5(h)). Normal Kidney Tissues and eir Relationship with Prognosis. We used diverse databases to comprehensively analyze To investigate different protein levels of HOXBs in patients the expression of HOXBs and their relationship with OS. We with ccRCC and those with normal renal tissues, we used the discovered that the mRNA levels of HOXB1/6/8/9 were Human Protein Atlas to verify the differential expression of significantly downregulated in renal cell carcinoma, while HOXBs. We found that patients with a high mRNA level of HOXB3/4/7 were upregulated in almost all tumors, in- HOXB2/5/6/7/8/9 have a higher OS, but it was not signif- cluding RCC. Also, the high mRNA levels of HOXB2/5/6/7/ icantly different from that of HOXB3/4 (Figures 4(a)–4(i)). 8/9 were predicted to have a higher OS in patients, and there was no significant statistical difference in HOXB3/4. 6erefore, we considered that HOXB 5/6/8/9 could be used 3.4. Interactions, Copy Number Variation, Methylation, as a potential target for ccRCC. Cancer Potential Pathways, and Drug Sensitivity of Patients with ccRCC. Using the STRING online database to analyze the relationship between HOXBs, we found a close rela- 3.5. Coexpression of HOXBs in Patients with ccRCC. We tionship between HOXB1-9 and the other genes and none analyzed the HOXBs coexpression relation for ccRCC with with HOXB13 (Figure 5(a)). 6e expression of HOXs in 537 samples (TCGA, Provisional) through the cBioPortal tumor samples is significantly different from that of the online tool; and we used CLUSTIVIS online tool for visu- normal renal tissues [24], and there may be variations in alization. We found a clear positive correlation between DNA fragments. Using the GSCALite to analyze the copy HOXB1-9, among which HOXB5 and HOXB6 had the number variation (CNV), cancer potential pathways, and strongest correlation (Figure 6(a)). 6e following showed drug sensitivity, we found significant differences in the significant positive correlation: HOXB2 and HOXB4; expression of CNV among KIRC, KIRP, and KICH. In HOXB3 and HOXB4; HOXB5 and HOXB6; and HOXB6 KIRC, CNV amplification and deletion occur in a small and HOXB8 (Figure 6(b)). 4 Journal of Oncology

Cancer Cancer Cancer Cancer Cancer Cancer Cancer Cancer Cancer Cancer vs. vs. vs. vs. vs. vs. vs. vs. vs. vs. normal normal normal normal normal normal normal normal normal normal Analysis type by cancer

HOXB1 HOXB2 HOXB3 HOXB4 HOXB5 HOXB6 HOXB7 HOXB8 HOXB9 HOXB13

Bladder cancer 2 Brain and CNS cancer 22 1 2 Breast cancer 1 2 1 1 1 1 Cervical cancer 1 1 2 Colorectal cancer 2 14 6 1 5 Esophageal cancer 3 1 6 1 Gastric cancer 1 2 1 6 21 Head and neck cancer 1 1 2 Kidney cancer 1 4 5 1 1 41 Leukemia 221 1 1 Liver cancer Lung cancer 4 1 Lymphoma 11 Melanoma 1 Myeloma Other cancers 2 1 1 1 Ovarian cancer 1 Pancreatic cancer 3 2 33 Prostate cancer 1 1 Sarcoma 1 1 1 Significant unique analyses 2 14 3 12 1 2 1 7 11 7 28 2 5 1 12 1 5 6 Total unique analyses 389 424 387 269 429 417 432 339 359 408

1 5510 10 1

% Figure 1: Comparison of the mRNA level of HOXBs in diﬀerent tumor tissues by the ONCOMINE database. Among them, red represents high expression and blue represents low expression. KIRC KIRP KICH KIRC KIRP KICH KIRC KIRP KICH KIRC KIRP KICH KIRC KIRP KICH KIRC KIRP KICH KIRC KIRP KICH KIRC KIRP KICH KIRC KIRP KICH KIRC KIRP KICH 1 160 240 30 140 240 180 120 120 28 11 1 11

0.9 140 120 24 200 25 200 150 100 100 0.8 120 0.7 100 20 160 20 160 120 80 80 100 0.6 80 16

0.5 80 120 15 120 90 60 60

60 12 0.4 60 80 10 80 60 40 40 0.3 40 8 40 Transcripts per million (TPM) Transcripts per million (TPM) Transcripts per million (TPM) Transcripts per million (TPM) Transcripts per million (TPM) Transcripts per million (TPM) Transcripts per million (TPM) Transcripts per million (TPM) Transcripts per million (TPM) 0.2 Transcripts per million (TPM) 40 5 40 30 20 20 20 4 20 0.1

0 0 0 0 0 0 0 0 0 0 T (n = 65) T (n = 66) T (n = 66) T (n = 66) T (n = 66) T (n = 66) T (n = 66) T (n = 65) T (n = 66) T (n = 66) N (n = 60) N (n = 53) N (n = 60) N (n = 53) N (n = 60) N (n = 53) N (n = 60) N (n = 53) N (n = 60) N (n = 53) N (n = 60) N (n = 53) N (n = 60) N (n = 53) N (n = 60) N (n = 53) N (n = 60) N (n = 53) N (n = 60) N (n = 53) T (n = 523) T (n = 285) T (n = 523) T (n = 286) T (n = 523) T (n = 286) T (n = 523) T (n = 286) T (n = 523) T (n = 286) T (n = 523) T (n = 286) T (n = 523) T (n = 286) T (n = 523) T (n = 286) T (n = 523) T (n = 286) T (n = 522) T (n = 285) N (n = 100) N (n = 100) N (n = 100) N (n = 100) N (n = 100) N (n = 100) N (n = 100) N (n = 100) N (n = 100) N (n = 100) HOXB1 HOXB2 HOXB3 HOXB4 HOXB5 HOXB6 HOXB7 HOXB8 HOXB9 HOXB13 (a) Figure 2: Continued. Journal of Oncology 5

HOXB1 HOXB2 HOXB3 ∗ ∗ 1.5 8

6 6 1.0 (TPM + 1) (TPM + 1) (TPM + 1) 2 2 2 4 4

0.5 2 2 Expression − log Expression − log Expression − log 0.0 0 0

KIRC KIRP KICH KIRC KIRP KICH KIRC KIRP KICH (num (T) = 523; (num (T) = 286; (num (T) = 66; (num (T) = 523; (num (T) = 286; (num (T) = 66; (num (T) = 523; (num (T) = 286; (num (T) = 66; num (N) = 100) num (N) = 60) num (N) = 53) num (N) = 100) num (N) = 60) num (N) = 53) num (N) = 100) num (N) = 60) num (N) = 53) (b) (c) (d) HOXB4 HOXB5 HOXB6 ∗ ∗ ∗∗ 5 8

4 6 6 (TPM + 1) (TPM + 1) (TPM + 1) 3 2 2 2 4 4 2

2 2 1 Expression − log Expression − log Expression − log 0 0 0

2 2 2 1 Expression − log Expression − log Expression − log

0 0 0

6 (TPM + 1) 2 4

2 Expression − log 0

KIRC KIRP KICH (num(T) = 523; (num(T) = 286; (num(T) = 66; num(N) = 100) num(N) = 60) num(N) = 53) (k)

Figure 2: Expression of HOXBs in renal carcinoma (GEPIA). (a) 6e differential expression of HOXBs between normal kidney and KIRC, KIRP, and KICH using the GEPIA database. (b–k) Scatterplot of the differential expression of HOXBs between normal kidney and KIRC (kidney renal clear cell carcinoma), KIRP (kidney renal papillary cell carcinoma), and KICH (kidney chromophobe) using the GEPIA database (histogram). GEPIA, expression profile analysis.

F value = 1.77 F value = 2.29 F value = 2.6 1.5 Pr (>F) = 0.151 Pr (>F) = 0.0766 6 Pr (>F) = 0.0513 6 5 1.0 4 4 3 0.5 2 2 1 0.0 0 Stage I Stage II Stage III Stage IV Stage I Stage II Stage III Stage IV Stage I Stage II Stage III Stage IV (a) (b) (c) Figure 3: Continued. 6 Journal of Oncology

4 F value = 2.16 F value = 7.24 F value = 4.28 F F Pr (>F) = 0.0915 5 Pr (> ) = 8.46e − 05 Pr (> ) = 0.0052 6 3 4

4 2 3 2 1 2 1

0 0 0 Stage I Stage II Stage III Stage IV Stage I Stage II Stage III Stage IV Stage I Stage II Stage III Stage IV (d) (e) (f)

F value = 3.96 8 F value = 7.63 F value = 2.61 Pr (>F) = 0.00812 Pr (>F) = 4.89e − 05 Pr (>F) = 0.0504 6 6 6

4 4 4

2 2 2

0 0

Stage I Stage II Stage III Stage IV Stage I Stage II Stage III Stage IV Stage I Stage II Stage III Stage IV (g) (h) (i) 8 F value = 10.1 Pr (>F) = 1.57e − 06 6

0 Stage I Stage II Stage III Stage IV (j)

Figure 3: 6e diﬀerential expression of HOXBs in diﬀerent stages of renal cell carcinoma using GEPIA database (a–j).

1.0 1.0 Survival analysis Survival analysis 0.9 0.9 Current cut offi: 9.05 Current cut offi: 1.54 0.8 Best expression cut offi: 0.8 Best expression cut offi: Median expressioni: 9.229.05 Median expressioni: 2.651.54 0.7 Median follow up timei: 2.99 0.7 Median follow up timei: 2.99 P scorei 0.0030 P scorei 0.32 0.6 0.6 5-year survival highi 75% 5-year survival highi 67% i i 0.5 5-year survival low 63% 0.5 5-year survival low 73%

0.4 0.4 Dead (n = 226) Dead (n = 226) Survival probability Survival probability 0.3 Alive (n = 651) 0.3 Alive (n = 651) 0.2 Density dead 0.2 Density dead Density alive Density alive 0.1 Density dead under cut off 0.1 Density dead under cut off HOXB2 Density dead over cut off HOXB3 Density dead over cut off 0.0 0.0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 012345678910111213141516 Time (years) Time (years)

Low expression (n = 427) Low expression (n = 228) High expression (n = 450) High expression (n = 649) (a) (b)

1.0 1.0 Survival analysis Survival analysis 0.9 0.9 Current cut offi: 2.69 Current cut offi: 2.64 0.8 Best expression cut offi: 0.8 Best expression cut offi: Median expressioni: 1.922.69 Median expressioni: 1.812.64 0.7 Median follow up timei: 2.99 0.7 Median follow up timei: 2.99 P scorei 0.34 P scorei 0.0076 0.6 0.6 5-year survival highi 71% 5-year survival highi 73% i i 0.5 5-year survival low 68% 0.5 5-year survival low 67%

0.4 0.4 Dead (n = 226) Dead (n = 226) Survival probability Survival probability 0.3 Alive (n = 651) 0.3 Alive (n = 651) 0.2 Density dead 0.2 Density dead Density alive Density alive 0.1 HOXB4 Density dead under cut off 0.1 HOXB5 Density dead under cut off Density dead over cut off Density dead over cut off 0.0 0.0 0 1 2 3 4 5 6 7 8 9 10111213141516 012345678910111213141516 Time (years) Time (years)

Low expression (n = 659) Low expression (n = 562) High expression (n = 218) High expression (n = 315) (c) (d) Figure 4: Continued. Journal of Oncology 7

1.0 1.0 Survival analysis Survival analysis 0.9 0.9 Current cut offi: 7.32 Current cut offi: 20.7 0.8 Best expression cut offi: 0.8 Best expression cut offi: Median expressioni: 3.427.32 Median expressioni: 13.1420.7 0.7 Median follow up timei: 2.99 0.7 Median follow up timei: 2.99 P scorei 0.0025 P scorei 0.012 0.6 0.6 5-year survival highi 77% 5-year survival highi 74% i i 0.5 5-year survival low 66% 0.5 5-year survival low 67%

0.4 0.4 Dead (n = 226) Dead (n = 226) Survival probability Survival probability 0.3 Alive (n = 651) 0.3 Alive (n = 651) 0.2 Density dead 0.2 Density dead Density alive Density alive 0.1 HOXB6 Density dead under cut off 0.1 HOXB7 Density dead under cut off Density dead over cut off Density dead over cut off 0.0 0.0 012345678910111213141516 012345678910111213141516 Time (years) Time (years)

Low expression (n = 687) Low expression (n = 684) High expression (n = 190) High expression (n = 193) (e) (f)

1.0 1.0 Survival analysis Survival analysis 0.9 0.9 Current cut offi: 3.48 Current cut offi: 7.46 0.8 Best expression cut offi: 0.8 Best expression cut offi: Median expressioni: 3.873.48 Median expressioni: 1.687.46 0.7 Median follow up timei: 2.99 0.7 Median follow up timei: 2.99 P scorei 1.3e-7 P scorei 0.0023 0.6 0.6 5-year survival highi 77% 5-year survival highi 76% i i 0.5 5-year survival low 60% 0.5 5-year survival low 66%

0.4 0.4 Dead (n = 226) Dead (n = 226) Survival probability Survival probability 0.3 Alive (n = 651) 0.3 Alive (n = 651) 0.2 Density dead 0.2 Density dead Density alive Density alive 0.1 Density dead under cut off 0.1 Density dead under cut off HOXB8 Density dead over cut off HOXB9 Density dead over cut off 0.0 0.0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 012345678910111213141516 Time (years) Time (years)

Low expression (n = 409) Low expression (n = 625) High expression (n = 468) High expression (n = 252) (g) (h)

1.0 Survival analysis 0.9 Current cut offi: 0.03 0.8 Best expression cut offi: Median expressioni: 0.010.03 0.7 Median follow up timei: 2.99 P scorei 4.7e-11 0.6 5-year survival highi 56% i 0.5 5-year survival low 76%

0.4 Dead (n = 226) Survival probability 0.3 Alive (n = 651) 0.2 Density dead Density alive 0.1 Density dead under cut off HOXB13 Density dead over cut off 0.0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Time (years)

Low expression (n = 536) High expression (n = 341) (i)

Figure 4: Using Human Protein Atlas to investigate the relationship between the mRNA of HOXBs and the prognosis (a–i). 6e expression of HOXB1 was too low to be analyzed.

3.6. Predicting the Function and Pathways ofHOXBs and eir 0043565 (sequence-specific DNA binding), GO:0030528 Adjacent Gene Alteration in Patients with ccRCC. (transcription regulatory activity), GO:0003700 (tran- 6rough the STRING website, we linked 40 genes that are scription factor activity), and GO:0003677 (DNA biding) closely associated with HOXBs. Studying the genes’ inter- were also significantly controlled by HOXBs. Moreover, the action, we constructed a network of 10 HOXBs and 40 genes PANTHER pathway analysis was visualized using the R with their frequently altered neighbor genes (Figure 7(a)) language. We discovered that the overall genes were closely and then analyzed the enrichment of GO and KEGG and the associated with TP53 and MEIS1, but, interestingly, PANTHER pathway through DAVID websites; the data HOXBs were also closely associated with c-JUN, PBX1, and were visualized through the R language. CCNP1 (Figures 7(b)–7(d)). 6rough GO analysis, we predicted, identified, and 6e results analyzed by KEGG show that three pathways validated the function of the target gene from bio- were associated with the change of HOXB function in informatics and expression system. We found that the ccRCC: Wnt signaling, colorectal cancer, and RNA degra- change of HOXB in ccRCC has a significant regulatory dation (Figure 7(e)). In these pathways, the Wnt signaling effect on GO:0003002 (regionalization) (anterior/poste- pathway is mainly responsible for the occurrence and de- rior) (pattern formation) and GO:0048706 (embryonic velopment of ccRCC (Figure 8). Simultaneously, it is also skeletal system development), which is beneficial to prove confirmed that HOXBs are closely associated with cytokine that HOXBs are associated with the development of bio- transcription (Figures 7(b)–7(d)). 6is finding is consistent logical embryos. At the same time, we found that GO: with Giampaolo A’s findings [27]. 8 Journal of Oncology

Pearson correlation between CNV and mRNA RSEM. KIRP KIRC KICH CTRP HOXB3 HOXB1 YK 4−279 HOXB1 HOXB7 WZ8040 WP1130 Vorinostat Vincristine HOXB2 VER−155008 HOXB7 HOXB2 UNC0638 Tubastatin A Triazolothiadiazine HOXB5 Tretinoin TPCA−1 HOXB13 Tozasertib HOXB2 Topotecan HOXB9 Tivantinib Tipifarnib−P2 HOXB4 Tipifarnib−P1 HOXB3 HOXB13 TG−101348 Symbol Teniposide Temsirolimus HOXB6 Tacedinaline HOXB5 HOXB4 Sunitinib Staurosporine HOXB4 SR−II−138A SNX−2112 HOXB3 SNS−032 SKI−II HOXB13 Sirolimus HOXB5 SID 26681509 HOXB1 Serdemetan SB−743921 SB−225002 Ruxolitinib HOXB8 HOXB6 Rigosertib HOXB9 QW−BI−011 PX−12 PRIMA−1−Met KIRP KIRC

KICH PRIMA−1 HOXB7 PL−DI Piperlongumine Cancer types PIK−93 Pifithrin−mu PI−103 Phloretin HOXB8 Pearson correlation PHA−793887 (a) PF−573228 PF−3758309 PF−184 Pevonedistat HOXB9 −0.4 −0.2 0.0 0.2 Pazopanib Parbendazole Panobinostat Paclitaxel −Log10 (FDR) Ouabain Hete amp OSI−027 1.5 3.0 Omacetaxine mepesuccinate Olaparib Homo amp Obatoclax 2.0 3.5 NVP−BSK805 NVP−231 Hete del 2.5 NSC95397 Heterozygous amplification Heterozygous deletion NSC632839 NSC48300 Homo del NSC23766 HOXB9 NSC19630 Neuronal differentiation inducer III None Necrosulfonamide (c) Narciclasine Nakiterpiosin HOXB8 N9−isopropylolomoucine Myriocin MST−312 Momelotinib HOXB7 (b) ML311 ML239 ML210 ML031 HOXB6 MK−2206 MK−1775 Mitomycin MGCD−265 HOXB5 Methylstat Merck60 Mdivi−1 Interaction map of gene and pathway. Manumycin A LY−2183240

Symbol HOXB4 LRRK2−IN−1 Linifanib Leptomycin B HOXB13 DNA Damage Response KX2−391 HOXB3 KW−2449 KU−60019 KU−55933 KU 0060648 HOXB2 KPT185 KIRC JQ−1 ISOX HOXB13 Isoliquiritigenin Isoevodiamine HOXB13 Indisulam I−BET151 GW−843682X GW−405833 HOXB1 GSK525762A GSK461364 HOXB9 Apoptosis Gemcitabine GDC−0941 Fulvestrant KICH KIRC KIRP KICH KIRC KIRP FQI−2 FQI−1 Foretinib Cancer type Fluorouracil HOXB9 Fingolimod FGIN−1−27 Etoposide Hete CNV% SCNA Type Epigallocatechin−3−monogallate Elocalcitol 5 40 Deletion ELCPK Doxorubicin Docetaxel 10 60 Amplification Dinaciclib Decitabine 20 100 HOXB7 Hormone AR Darinaparsin Cytarabine hydrochloride Cyanoquinoline 11 Curcumin Cucurbitacin I CR−1−31B HOXB7 Compound 7d−cis (d) DNA Damage Response Clofarabine Ciclopirox CHM−1 Chlorambucil KICH Ch−55 Methylation difference between tumor and normal samples. Cerulenin Ceranib−2 CD−437 CCT036477 HOXB6 RTK BRD−K84807411 HOXB5 BRD−K80183349 Apoptosis BRD−K75293299 BRD−K70511574 BRD−K66532283 HOXB6 HOXB6 BRD−K66453893 BRD−K61166597 BRD−K55116708 BRD−K49290616 HOXB3 BRD−K35604418 BRD−K34222889 BRD−K33514849 HOXB8 Hormone AR BRD−K30748066 BRD−K29313308 BRD−K29086754 BRD−K26531177 HOXB13 BRD−K01737880 HOXB5 PI3K/AKT BRD−A94377914 BRD6340 HOXB2 BMS−345541

Symbol BMS−270394 HOXB5 BIX−01294 BI−2536 RTK Bardoxolone methyl HOXB4 Barasertib B02 AZD7762 AZD1480 HOXB7 Axitinib Avrainvillamide KIRP AT7867 HOXB1 AT13387 Apicidin HOXB4 Hormone ER Alvocidib PI3K/AKT Alisertib HOXB9 HOXB4 Abiraterone KIRP KIRC Cancer types HOXB1 HOXB8 HOXB4 HOXB2 HOXB3 HOXB9 HOXB5 HOXB7 HOXB6

Hormone ER HOXB13 Methylation diff (T − N) HOXB3 Cell Cycle Spearman correlation 0.0 0.1 0.2 HOXB3 0.0 0.1 0.2 0.3 −Log10 (FDR) Cell Cycle 10 40 Regulation −log10 (FDR) 20 50 Activate 5 15 30 Inhibit 10 20 (e) (f) (h)

Spearman correlation coefficient of methylation and gene expression.

HOXB13 HOXB9 19 −3 7 −12 0 −21 10 −9

HOXB5 HOXB7 77−12 −18 HOXB7 HOXB6 41−21 61−9 341−9 −18 0−9 HOXB8

HOXB6 HOXB4 16 −3 Symbol Symbol HOXB4 HOXB3 7 −12 HOXB3 HOXB2 41−15 90 HOXB9

HOXB2 HOXB13 16 −3

KIRP KIRC KICH Cancer types RTK_I EMT_I RTK_A Spearman correlation coefficient EMT_A Cell cycle_I Cell cycle_A Hormone ER_I Hormone AR_I Hormone ER_A −0.6 −0.5 −0.4 −0.3 −0.2 Hormone AR_A Pathway (A: activate; I: inhibit) −Log10 (P value) 10 Percent 20 −20 −10 0 10 (g) (i) Figure 5: (a) Using the STRING website to analyze the interaction between HOXBs. (b–d) Using the GSCALite website to analyze the CNV of HOXBs. (e, g) Analyzing the DNA methylation of HOXBs. (f, i) Analyzing the potential pathway of HOXBs. (h) Analyzing the potential drug target of HOXBs. Journal of Oncology 9

1 HOXB6 vs. HOXB5 1.00 0.08 0.02 0.09 0.26 0.23 0.07 0.16 0.17 −0.05 HOXB1

10 0.08 1.00 0.57 0.72 0.34 0.48 0.39 0.45 0.08 −0.07 HOXB2 Spearman: 0.77 0.8 (p = 1.57e – 88) 9 0.02 0.57 1.00 0.70 0.65 0.62 0.35 0.46 0.28 −0.02 HOXB3 Pearson: 0.80 (p = 1.68e – 101) 8 0.09 0.72 0.70 1.00 0.45 0.59 0.42 0.49 0.21 −0.02 HOXB4 0.6

7 0.26 0.34 0.65 0.45 1.00 0.77 0.47 0.57 0.44 0.06 HOXB5

6 0.23 0.48 0.62 0.59 0.77 1.00 0.62 0.71 0.49 0.03 HOXB6 0.4

5 0.07 0.39 0.35 0.42 0.47 0.62 1.00 0.65 0.50 0.21 HOXB7

4 0.2 0.16 0.45 0.46 0.49 0.57 0.71 0.65 1.00 0.56 0.07 HOXB8 3 0.17 0.08 0.28 0.21 0.44 0.49 0.50 0.56 1.00 0.34 HOXB9 0 2 mRNA expression (RNA seq V2 RSEM): HOXB5 (log2) −0.08 −0.00 −0.09 0.04 −0.09 −0.15 −0.10 −0.10 −0.06 1.00 HOXB13 3 4 5 6 7 8 9 10 11 12 mRNA expression (RNA seq V2 RSEM): HOXB6 (log2) HOXB1 HOXB2 HOXB3 HOXB4 HOXB5 HOXB6 HOXB7 HOXB8 HOXB9 y = 0.89x + –0.83 HOXB13 R2 = 0.64 HOXB6 mutated Neither mutated

(a) (b) Figure 6: 6e coexpression of HOXBs was analyzed using the cBioPortal online tool and visualized using an online tool (a, b).

BP P Pathway enrichment –log10 ( value) GO:0051252~regulation of RNA metabolic process GO:0048706~embryonic skeletal system development 30.0 GO:0048704~embryonic skeletal system morphogenesis GO:0048598~embryonic morphogenesis GO:0048568~embryonic organ development 27.5 GO:0048562~embryonic organ morphogenesis GO:0045449~regulation of transcription GO:0043009~chordate embryonic development 25.0 GO:0009952~anterior/posterior pattern formation Pathway name GO:0009792~embryonic development ending in birth or egg hatching GO:0007389~pattern speciﬁcation process 22.5 GO:0006355~regulation of transcription, DNA-dependent GO:0006350~transcription GO:0003002~regionalization 20.0 GO:0001501~skeletal system development 1.6900e – 20 2.6195e – 20 3.5490e – 20 7.6050e – 21 –1.6900e – 21 P value Count 15 30 20 35 25 (a) (b) CC MF P P Pathway enrichment –log10 ( value) Pathway enrichment –log10 ( value)

12.5 GO:0070013~intracellular organelle lumen GO:0043565~sequence-speciﬁc DNA binding 30 GO:0044451~nucleoplasm part GO:0030528~transcription regulator activity 10.0 GO:0043233~organelle lumen GO:0016564~transcription repressor activity 20 GO:0031981~nuclear lumen 7.5 GO:0008134~transcription factor binding GO:0031974~membrane-enclosed lumen Pathway name Pathway name GO:0003702~RNA polymerase II transcription factor activity 5.0 GO:0005829~cytosol 10 GO:0003700~transcription factor activity GO:0005667~transcription factor complex 2.5 GO:0005654~nucleoplasm GO:0003677~DNA binding

0.00 0.02 0.04 0.06 0.08 0.00 0.02 0.04 P value P value Count Count 4 8 10 6 10 20 30 (c) (d) Figure 7: Continued. 10 Journal of Oncology

KEGG and PANTHER P Pathway enrichment –log10 ( value)

PANTHER P00057: Wnt signaling pathway

KEGG hsa05210: colorectal cancer

Pathway name KEGG hsa04310: Wnt signaling pathway

KEGG hsa03018: RNA degradation

0.0000 0.0025 0.0050 0.0075 0.0100 0.0125 P value Count 5.0 10.0 7.5 12.5 (e)

Figure 7: (a) Constructing a network of HOXB factors and their adjacent gene alteration through the STRING website. (b–d) Using the Gene Ontology analysis by DAVID tools, we predicted the functions of HOXBs and genes associated with HOXBs alterations and then visualized them through R language. (e) 6rough the analysis of KEGG and PANTHER using DAVID tools, predict the functions of HOXBs and genes associated with HOXB alterations and visualize them using the R language.

3.7. Potential Mechanisms and Signaling Pathways of HOXBs. While HOXB5/9 were positively correlated with the It is found that CCND1, CCNDBP1, MYCBP2, MYC, CCT036477 drug target, CCT036477 was the classical Wnt CTNNB1, β-Catenin, and TP53 are the classical proteins and signal inhibitor; and DNA abnormal methylation is one of genes of the Wnt pathway [28, 29]. We analyzed protein- the mechanisms of tumorigenesis. 6erefore, our study protein coexpression on the HOXBs and Wnt pathway. implied that HOXB5/6/8/9 are potential therapeutic targets Using the cBioPortal database, we discovered that HOXB5 is for patients with ccRCC. We examined the function of negatively associated with CCNDBP1 and MYCBP2; enrichment data of HOXBs in patients with ccRCC from HOXB6 is negatively correlated with CCNDBP1 and KEGG enrichment and the PANTHER pathway. We also MYCBP2; HOXB8 is negatively correlated with CCNDBP1, found that the function of HOXBs might be associated with MYCBP2, and MYC; HOXB9 is negatively correlated with the Wnt pathway. In coexpression analysis, we discovered CCNDBP1, MYCBP2, and CTNNB1 and positively corre- that HOXB5/6/8/9 were coexpressed with multiple Wnt lated with TP53 (Figures 9(a)–9(d)). In our study, HOXB5/ pathway classical genes and proteins, such as MYC, CTNNB, 6/8/9 can be used as a potential treatment target for ccRCC. CCND1, and TP53, which further confirms that HOXBs inhibit the growth of renal carcinoma cells through the Wnt 4. Discussion signaling pathway. In conclusion, our systematic analysis of all members of Although we have partly confirmed the effect of HOXBs in HOXBs and their underlying molecular mechanisms may the development and prognosis of various cancers [30, 31], provide a basis for future research. Wnt/β-Catenin signaling there has not been an in-depth study on HOXBs in ccRCC. pathway activated EMT [36] to promote the binding of Studies have shown that bioinformatics analysis is a method β-Catenin to the nucleus and T cell-specific transcription that can help us quickly find biomarkers in the development factor/lymphoenhancer (TCF/LEF) to regulate the expres- of diseases [32–35]. Our study deeply investigates the ex- sion of E-cadherin, the characteristic factor of EMT, and pression and prognosis of multiple HOXBs in ccRCC. We then affect the occurrence of EMT, which plays a blocking hope that our research will help discover the effect of role in the process of tumor invasion and migration. In other HOXBs, improving the treatment design and accuracy of studies, it was found that HOXB5 protein could bind with its prediction for ccRCC. cofactor PBX1 to induce apoptosis, thus delaying tumor In this study, we examined the transcriptional and progression [37]. Our research found that HOXB5 inhibited prognosis data of HOXBs in patients with ccRCC from the the key genes CCND1 and MYC in the Wnt/β-Catenin ONCOMINE database, Human Protein Atlas, and STRING signaling pathway, inhibited the tumor’s occurrence, and website. We discovered that the expression levels of HOXB3/ inhibited the classical cancer pathway RTK pathway to 5/6/8/9 were significantly lower in ccRCC than in normal inhibit cancer and improve OS. However, the expression of nephritic tissues. Furthermore, the expression level of HOXB5 in different stages of RCC is significantly different, mRNA factor of HOXB2/5/6/7/8/9 in ccRCC was related which suggests that HOXB5 plays a positive role in the early considerably to patient prognosis. A high level of the mRNA stage of RCC. We discovered that reducing the expression of factor had a better OS in patients, comparatively, while HOXB5 in the late stage of ccRCC is more conducive to HOXB3/4 did not. We studied the GSCALite website’s reducing the metastasis of RCC. 6is may be a direction for analysis that the methylation of HOXB3/5/6/8 in ccRCC was further research. We discovered that HOXB5 and HOXB6 significantly negatively correlated with gene expression. are most closely related in HOXBs, similar to the regulatory Journal of Oncology 11

WNT SIGNALING PATHWAY

Adherens junction MAPK signaling Alzheimer’s Rspo Lgr4-6 pathway disease +p TAK1 NLK Canonical pathway Rnf43 PS-1 Znrf3 PKA CBY1 c-myc Notum FRP Idax +p c-jun Cer-1 Duplin CBP Dally Scaffold fra-1 ε Porc CKI GBP Pontin52 +p cycD Cell Frizzled Dvl GSK-3β β-Catenin TCF/LEF Wnt WISPl cycle LRP5/6 Axin APC CKIα SMAD4 DNA CK2 δ SMAD3 PPAR Wif-1 PEDF Axam ICAT Xsox17 Groucho BAMBI Uterine SOST CtBP δ-Catenin Dkk Phosphorylation p53 signaling +p independence pathway β-TrCP TGF-β signaling p53 pathway Siah-1 Skp1 Ubiquitin medeated PAR-1 Nkd SlP Cul1 proteolysis APC Skp1 Rbx1 TBL1 +u Phosphorylation Proteolysis independence

Planar cell polarity (PCP) pathway Focal adhesion Stbm Prickle Daam1 RhoA ROCK2 Cytoskeletal change Wnt11 Frizzled Dv1 INVS Rac JNK Gene transcription Knypek ROR1/2 MAPK signaling pathway RYK

Wnt/Ca2+pathway CaMKII 2+ Ca -p Wnt5 Frizzled PLC CaN NFAT G protein? DNA RYK PKC

Figure 8: 6e Wnt pathway. 6e red mark indicates enrichment through the analysis of the KEGG pathway of the functions of HOXBs and genes associated with HOXBs alterations on the DAVID website.

HOXB5 HOXB6 HOXB5 vs. CCNDBP1 HOXB5 vs. MYCBP2 HOXB6 vs. CCNDBP1 HOXB6 vs. MYCBP2 14 14 13 Spearman: –0.25 Spearman: –0.25 13 Spearman: –0.14 Spearman: –0.25 p e p e (p = 4.96e – 8) ( = 8.66 – 8) 13 ( = 9.66 – 8) (p = 3.384e – 3) 13 Pearsan: –0.26 12 Pearsan: –0.22 Pearsan: –0.28 12 Pearsan: –0.06 p e p e p e ( = 2.525 – 6) 12 ( = 2.05 – 9) (p = 0.240) 12 ( = 3.90 – 8)

11 11 11 11

10 10 10 10 MYCBP2 (log2) MYCBP2 (log2) CCNDBP1 (log2) CCNDBP1 (log2) 9 9 9 9 8 8 8 8 mRNA expression (RNA seq V2 RSEM): mRNA expression (RNA seq V2 RSEM): mRNA expression (RNA seq V2 RSEM): mRNA expression (RNA seq V2 RSEM): 7 7

2345678910 2345678910 3456789101112 3456789101112 mRNA expression (RNA seq V2 RSEM): mRNA expression (RNA seq V2 RSEM): mRNA expression (RNA seq V2 RSEM): mRNA expression (RNA seq V2 RSEM): HOXB5 (log2) HOXB5 (log2) HOXB6 (log2) HOXB6 (log2)

y = –0.09x + 10.67 y = –0.18x + 11.38 y = –0.03x + 10.37 y = –0.18x + 11.67 R2 = 0.05 R2 = 0.08 R2 = 0 R2 = 0.07 Neither mutated MYCBP2 mutated HOXB6 mutated HOXB6 mutated Neither mutated Neither mutated MYCBP2 mutated Neither mutated (a) (b) Figure 9: Continued. 12 Journal of Oncology

HOXB8 HOXB8 vs. CCNDBP1 HOXB8 vs. MYC HOXB8 vs. MYCBP2 14 13 Spearman: –0.19 Spearman: –0.12 14 (p = 6.611e – 5) (p = 8.691e – 3) 13 13 12 Pearsan: –0.17 Pearsan: –0.09 (p = 2.290e – 4) (p = 0.0606) 12 12 11 11 11 10

10 9 10

MYC (log2) 8 MYCBP2 (log2)

CCNDBP1 (log2) 9 9 7 Spearman: –0.12 p 6 ( = 0.0122) 8 Pearsan: –0.18 8 5 mRNA expression (RNA seq V2 RSEM): mRNA expression (RNA seq V2 RSEM): (p = 1.233e – 4) mRNA expression (RNA seq V2 RSEM): 7 4 024681012 024681012 024681012 mRNA expression (RNA seq V2 RSEM): mRNA expression (RNA seq V2 RSEM): mRNA expression (RNA seq V2 RSEM): HOXB8 (log2) HOXB8 (log2) HOXB8 (log2)

y = –0.03x + 10.34 y = –0.11x + 11.54 y = –0.08x + 10.85 R2 = 0.01 R2 = 0.03 R2 = 0.03 Neither mutated Neither mutated MYCBP2 mutated Neither mutated (c) HOXB9 HOXB9 vs. TP53 HOXB9 vs. CCND1 HOXB9 vs. CTNNB1 HOXB9 vs. MYCBP2 12 14.5 14 Spearman: 0.24 17 Spearman: –0.13 Spearman: –0.13 11.5 p e p e ( = 3.62 – 7) 14 (p = 5.465e – 3) ( = 6.862 – 3) 16 13 11 Pearsan: 0.10 Pearsan: –0.13 Pearsan: –0.14 (p = 0.0315) 13.5 (p = 5.776e – 3) (p = 3.694e – 3) 10.5 15 12 13 10 14 11 9.5 12.5 13 10

TP53 (log2) 9 CCND1 (log2) CTNNB1 (log2)

12 MYCBP2 (log2) 12 9 8.5 Spearman: –0.35 (p = 3.33e – 14) 11.5 8 11 8 Pearsan: –0.37 mRNA expression (RNA seq V2 RSEM): p e mRNA expression (RNA seq V2 RSEM): 11 mRNA expression (RNA seq V2 RSEM): 7.5 mRNA expression (RNA seq V2 RSEM): 10 ( = 1.21 – 15) 7

–8 0–2–4–6 24681012 0–2–4–6–8 2 4 6 8 10 12 –8 0–2–4–6 2 4 6 8 10 12 –8 –2–4–6 0 24681012 mRNA expression (RNA seq V2 RSEM): mRNA expression (RNA seq V2 RSEM): mRNA expression (RNA seq V2 RSEM): mRNA expression (RNA seq V2 RSEM): HOXB9 (log2) HOXB9 (log2) HOXB9 (log2) HOXB9 (log2)

y = 0.02x + 9.97 y = –0.19x + 14.96 y = –0.03x + 12.85 y = –0.05x + 10.62 R2 = 0.01 R2 = 0.13 R2 = 0.02 R2 = 0.02 TP53 mutated Neither mutated CTNNB1 mutated MYCBP2 mutated Neither mutated Neither mutated Neither mutated (d)

Figure 9: 6rough the cBioPortal website, we investigated the coexpression of the Wnt pathway classical protein with genes that have significance in expression and survival (a–d). mechanism of HOXB5. HOXB6 is negatively associated with and their expression proteins in the Wnt signaling pathway. MYC and CCND1. MYC and CCND1 are the critical genes We speculated that HOXB9 might block the Wnt signaling in the Wnt signaling pathway and inhibit the RTK pathway, pathway and inhibit tumor growth through MYC and thus inhibiting independent downstream signaling pathways CTNNB1. 6rough the integration and transmission of such as Ras/ERK or P13-K/Akt signaling pathway and thus information inside and outside the cell and then through the inhibiting cell cycle and proliferation [38]. key genes of regulatory points, the cell decides whether to Interestingly, HOXB6 inhibited Wnt/β-Catenin signal continue to divide, differentiate, undergo apoptosis, or enter pathway activation and enhanced EMT. It can be concluded the G0 stage to repair damaged DNA. 6ere are many that HOXB6 can inhibit cell proliferation but cannot control regulatory points of the cell cycle in the process from G1 to tumor invasion and migration. In this study, we found that S. We found that HOXB9 was negatively correlated with the expression of HOXB8 in normal kidney tissue was CCND1. It may inhibit the cell cycle from the G1 phase to significantly different from that in ccRCC. 6e low level of the S phase, inhibit transcription, regulate cell mitosis, and HOXB8 reduced the OS of ccRCC; however, we found that inhibit cell proliferation and immortalization by inhibiting HOXB8 may also control tumor invasion by inhibiting the CCND1 coding protein Cyclin D1 binding cyclin-de- CCND1 and MYC in the Wnt signaling pathway. However, pendent kinase CDK4 [39]. At the same time, HOXB9 there is no strong evidence. negatively regulates RTK, thus inhibiting the proliferation of We found that the expression of HOXB9 in renal car- cancer cells. In addition to inhibiting cancer cell prolifer- cinoma patients was lower than that in normal people. ation, HOXB9 can activate TP53, regulate expression, Besides, based on the Human Protein Atlas analysis, we promote cell apoptosis, inhibit growth, block the cell cycle, discovered that the higher the expression of HOXB9 mRNA, promote cell differentiation and DNA repair, maintain cell the higher the OS. At the same time, HOXB9 had a sig- genome stability, and inhibit tumor angiogenesis [40]. Some nificant negative correlation with MYC, CTNNB1, CCND1, studies have shown that androgen receptors can promote the Journal of Oncology 13 metastasis of renal cell carcinoma. At the same time, HOXB9 Data Availability can reduce the androgen receptor and promote the expression of cycHIAT1 by regulating the expression of the 6e data used to support the findings of this study are host (HIAT1) at the transcription level. As a result, regu- available from the corresponding author upon request. lating the expression of miR-195-5p/29A-3P/29C-3P can affect the migration and invasion of renal cell carcinoma Disclosure [41]. In this study, we found that the inhibition of HOXB9 may be associated with the above pathways, and the above Guangzhen Wu and Xiaowei Li are the co-first authors. mechanisms will be verified in our future experiments. 6ere is no significant difference in expression of Conflicts of Interest HOXB1/2/3/4/7 between normal kidney tissue and renal clear cell carcinoma in the current study. 6erefore, we need 6e authors declare that they have no conflicts of interest. to increase the sample size further or carry out experiments to verify the above results. Our research shows that the level Authors’ Contributions of HOXB13 mRNA in ccRCC and renal tissue is almost no expression. However, HOXB13 is associated with various Qifei Wang, Yingkun Xu, and Quanlin Li designed the renal carcinoma targets involved in the Wnt pathway but research methods and analyzed the data. Guangzhen Wu combined with the introduction of HOXB13 through the and Xiaowei Li participated in data collection. Guangzhen Wnt signal pathway and the androgen receptor signal Wu, Xiaowei Li, and Yuanxin Liu drafted and revised the pathway to inhibit the growth of prostate cancer cells. We manuscript. All authors approved the version to be released consider that although HOXB13 participates in the Wnt and agreed to be responsible for all aspects of the work. pathway, it may be different from the receptor of renal Guangzhen Wu and Xiaowei Li contributed equally to this carcinoma. At the same time, HOXB13 is not significantly study. associated with other genes in the HOXB genome, so we speculate that HOXB13 may not be associated with renal Acknowledgments cells’ progress. At present, there are few studies on HOXBs in kidney cancer. We systematically analyze all members of 6e authors acknowledge the Cancer Genome Atlas (TCGA) HOXBs and their potential mechanisms. Our study may be for providing publicly available data. 6is study was sup- used as a foundation for future research. ported by the Natural Science Foundation of Liaoning (no. 201801460), the Scientific Research Fund of Liaoning Pro- 5. Conclusions vincial Education Department (no. LZ2020071), and the Liaoning Province Doctoral Research Startup Fund Program In summary, we discovered that HOXB5/6/8/9 are potential (no. 209). targets of precision therapy for patients with ccRCC. 6e role of HOXB5/6/8/9 in ccRCC may be associated with the References Wnt pathway. However, our conclusion needs to be verified by further experiments, as there are few studies on HOXBs [1] K. Gupta, J. D. Miller, J. Z. Li, M. W. Russell, and in kidney cancer at present. We have conducted a com- C. Charbonneau, “Epidemiologic and socioeconomic burden prehensive analysis of all members of HOXBs and studied of metastatic renal cell carcinoma (mRCC): a literature re- their molecular mechanism, thereby laying a foundation for view,” Cancer Treatment Reviews, vol. 34, no. 3, pp. 193–205, future studies. 2008. [2] R. J. Motzer, J. Bacik, B. A. Murphy, P. Russo, and M. Mazumdar, “Interferon-alfa as a comparative treatment Abbreviations for clinical trials of new therapies against advanced renal cell carcinoma,” Journal of Clinical Oncology, vol. 20, no. 1, HOXB: Homeobox B pp. 289–296, 2002. ccRCC: Clear cell renal cell carcinoma [3] R. J. Motzer, C. H. Barrios, T. M. Kim et al., “Phase II ran- OS: Overall survival domized trial comparing sequential first-line everolimus and MYC: MYC protooncogene second-line sunitinib versus first-line sunitinib and second- CTNNB: Catenin beta line everolimus in patients with metastatic renal cell carci- CCND1: Cyclin D1 noma,” Journal of Clinical Oncology, vol. 32, no. 25, TP53: Tumor protein P53 pp. 2765–2772, 2014. GO: Gene Ontology [4] T. K. Choueiri, D. J. Figueroa, A. P. Fay et al., “Correlation of KEGG: Kyoto Encyclopedia of Genes and Genomes PD-L1 tumor expression and treatment outcomes in patients with renal cell carcinoma receiving sunitinib or pazopanib: TCGA: 6e Cancer Genome Atlas results from COMPARZ, a randomized controlled trial,” CNV: Copy number variations Clinical Cancer Research, vol. 21, no. 5, pp. 1071–1077, 2015. KIRC: Kidney renal clear cell carcinoma [5] T. Reya and H. Clevers, “Wnt signalling in stem cells and KIRP: Kidney renal papillary cell carcinoma cancer,” Nature, vol. 434, no. 7035, pp. 843–850, 2005. KICH: Kidney chromophobe [6] F. R. Goodman, C. Bacchelli, A. F. Brady et al., “Novel HIAT1: Hippocampus abundant transcript 1. HOXA13 mutations and the phenotypic spectrum of hand- 14 Journal of Oncology

foot-genital syndrome,” e American Journal of Human associated with the progression of adult T cell leukemia Genetics, vol. 67, no. 1, pp. 197–202, 2000. (ATL),” Japanese Journal of Cancer Research, vol. 91, no. 11, [7] D. X. Nguyen, A. C. Chiang, X. H.-F. Zhang et al., “WNT/TCF pp. 1103–1110, 2000. signaling through LEF1 and HOXB9 mediates lung adeno- [24] C. Cillo, P. Barba, G. Freschi, G. Bucciarelli, M. C. Magli, and carcinoma metastasis,” Cell, vol. 138, no. 1, pp. 51–62, 2009. E. Boncinelli, “HOX gene expression in normal and neoplastic [8] O. Takahashi, J.-i. Hamada, M. Abe et al., “Dysregulated human kidney,” International Journal of Cancer, vol. 51, no. 6, expression of HOX and ParaHOX genes in human esophageal pp. 892–897, 1992. squamous cell carcinoma,” Oncology Reports, vol. 17, no. 4, [25] C. H. Barrios, D. Hernandez-Barajas, M. P. Brown et al., pp. 753–760, 2007. “Phase II trial of continuous once-daily dosing of sunitinib as [9] S. Sha, Y. Gu, B. Xu et al., “Decreased expression of HOXB9 is first-line treatment in patients with metastatic renal cell related to poor overall survival in patients with gastric car- carcinoma,” Cancer, vol. 118, no. 5, pp. 1252–1259, 2012. cinoma,” Digestive and Liver Disease, vol. 45, no. 5, [26] J. L. Lee, M. K. Kim, I. Park et al., “RandomizEd phase II trial pp. 422–429, 2013. of sunitinib four weeks on and two weeks off versus two weeks [10] N. Chiba, V. Comaills, B. Shiotani et al., “Homeobox B9 on and one week off in metastatic clear-cell type REnal cell induces epithelial-to-mesenchymal transition-associated carcinoma: RESTORE trial,” Annals of Oncology, vol. 26, radioresistance by accelerating DNA damage responses,” no. 11, pp. 2300–2305, 2015. Proceedings of the National Academy of Sciences, vol. 109, [27] A. Giampaolo, P. Sterpetti, D. Bulgarini et al., “Key functional no. 8, pp. 2760–2765, 2012. role and lineage-specific expression of selected HOXB genes [11] H. Seki, T. Hayashida, H. Jinno et al., “HOXB9 expression in purified hematopoietic progenitor differentiation,” Blood, promoting tumor cell proliferation and angiogenesis is as- vol. 84, no. 11, pp. 3637–3647, 1994. sociated with clinical outcomes in breast cancer patients,” [28] R. Nusse and H. Clevers, “Wnt/β-catenin signaling, disease, Annals of Surgical Oncology, vol. 19, no. 6, pp. 1831–1840, and emerging therapeutic modalities,” Cell, vol. 169, no. 6, 2012. pp. 985–999, 2017. [12] J. Miao, Z. Wang, H. Provencher et al., “HOXB13 promotes [29] P. Polakis, “Wnt signaling in cancer,” Cold Spring Harbor ovarian cancer progression,” Proceedings of the National Perspectives in Biology, vol. 4, no. 5, 2012. Academy of Sciences, vol. 104, no. 43, pp. 17093–17098, 2007. [30] C. Li, J. Cui, L. Zou, L. Zhu, and W. Wei, “Bioinformatics [13] S. D. Kim, R.-Y. Park, Y.-R. Kim et al., “HOXB13 is co-lo- analysis of the expression of HOXC13 and its role in the calized with androgen receptor to suppress androgen-stim- prognosis of breast cancer,” Oncology Letters, vol. 19, no. 1, ulated prostate-specific antigen expression,” Anatomy & Cell pp. 899–907, 2020. Biology, vol. 43, no. 4, pp. 284–293, 2010. [31] S. Bhatlekar, J. Z. Fields, and B. M. Boman, “HOX genes and [14] A. Nguyen Kovochich, M. Arensman, A. R. Lay et al., their role in the development of human cancers,” Journal of “HOXB7 promotes invasion and predicts survival in pan- Molecular Medicine, vol. 92, no. 8, pp. 811–823, 2014. creatic adenocarcinoma,” Cancer, vol. 119, no. 3, pp. 529–539, [32] G. Wu, Y. Xu, L. Li et al., “Tripartite-motif family genes 2013. associated with cancer stem cells affect tumor progression and [15] L. Marra, M. Cantile, G. Scognamiglio et al., “Deregulation of can assist in the clinical prognosis of kidney renal clear cell HOX B13 expression in urinary bladder cancer progression,” carcinoma,” International Journal of Medical Sciences, vol. 17, Current Medicinal Chemistry, vol. 20, no. 6, pp. 833–839, 2013. no. 18, pp. 2905–2916, 2020. [33] Y. Zhang, Y. Yao, X. Qi et al., “Identification of a new [16] Z. Tang, C. Li, B. Kang, G. Gao, C. Li, and Z. Zhang, “GEPIA: a web server for cancer and normal gene expression profiling prognostic risk signature of clear cell renal cell carcinoma and interactive analyses,” Nucleic Acids Research, vol. 45, based on N(6)-methyladenosine RNA methylation regula- no. W1, pp. W98–w102, 2017. tors,” Journal of Immunology Research, vol. 2021, Article ID [17] M. Uhlen,´ L. Fagerberg, B. M. Hallstrom¨ et al., “Tissue-based 6617841, 23 pages, 2021. map of the human proteome,” Science, vol. 347, no. 6220, [34] G. Wu, Y. Xu, N. Ruan et al., “Genetic alteration and clinical Article ID 1260419, 2015. significance of SUMOylation regulators in multiple cancer [18] P. J. 6ul, L. Akesson,˚ M. Wiking et al., “A subcellular map of types,” Journal of Cancer, vol. 11, no. 23, pp. 6823–6833, 2020. the human proteome,” Science, vol. 356 6340 pages, 2017. [35] P. Wu, Y. Xu, J. Li et al., “Comparison of the fatty acid [19] M. Uhlen, C. Zhang, S. Lee et al., “A pathology atlas of the metabolism pathway in pan-renal cell carcinoma: evidence human cancer transcriptome,” Science, vol. 357 6352 pages, from bioinformatics,” Analytical Cellular Pathology, vol. 2021, 2017. Article ID 8842105, 25 pages, 2021. [20] C.-J. Liu, F.-F. Hu, M.-X. Xia, L. Han, Q. Zhang, and [36] R. Kalluri and R. A. Weinberg, “6e basics of epithelial- A.-Y. Guo, “GSCALite: a web server for gene set cancer mesenchymal transition,” Journal of Clinical Investigation, analysis,” Bioinformatics, vol. 34, no. 21, pp. 3771-3772, 2018. vol. 119, no. 6, pp. 1420–1428, 2009. [21] D. Szklarczyk, A. L. Gable, D. Lyon et al., “STRING v11: [37] R. Morgan, L. Plowright, K. J. Harrington, A. Michael, and protein-protein association networks with increased coverage, H. S. Pandha, “Targeting HOX and PBX transcription factors supporting functional discovery in genome-wide experi- in ovarian cancer,” BMC Cancer, vol. 10, no. 1, p. 89, 2010. mental datasets,” Nucleic Acids Research, vol. 47, no. D1, [38] J.-L. Schwachtgen, J. E. Remacle, N. Janel et al., “Oct-1 is pp. D607–d613, 2019. involved in the transcriptional repression of the von wille- [22] D. Szklarczyk, J. H. Morris, H. Cook et al., “6e STRING brand factor gene promoter,” Blood, vol. 92, no. 4, database in 2017: quality-controlled protein-protein associ- pp. 1247–1258, 1998. ation networks, made broadly accessible,” Nucleic Acids Re- [39] C. V. Patel, R. Sharangpani, S. Bandyopadhyay, and search, vol. 45, no. D1, pp. D362–d368, 2017. P. E. DiCorleto, “Endothelial cells express a novel, tumor [23] T. Kohno, R. Moriuchi, S. Katamine, Y. Yamada, necrosis factor-α-regulated variant of HOXA9,” Journal of M. Tomonaga, and T. Matsuyama, “Identification of genes Biological Chemistry, vol. 274, no. 3, pp. 1415–1422, 1999. Journal of Oncology 15

[40] D. Qu, M. Jiang, D. Huang et al., “Synergistic eﬀects of the enhancements to mitochondrial ROS, p53 activation and apoptosis generated by aspartame and potassium sorbate in HepG2 cells,” Molecules, vol. 24, no. 3, 2019. [41] K. Wang, Y. Sun, W. Tao, X. Fei, and C. Chang, “Androgen receptor (AR) promotes clear cell renal cell carcinoma (ccRCC) migration and invasion via altering the circHIAT1/ miR-195-5p/29a-3p/29c-3p/CDC42 signals,” Cancer Letters, vol. 394, pp. 1–12, 2017.