Comprehensive analysis of clinical significance and potential molecular mechanism of PAX gene family in lung squamous cell carcinoma

doi:10.21203/rs.3.rs-2993563/v1

Download PDF

Research Article

Comprehensive analysis of clinical significance and potential molecular mechanism of PAX gene family in lung squamous cell carcinoma

https://doi.org/10.21203/rs.3.rs-2993563/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Purpose The paired box(PAX) gene family encodes a group of putative transcription factors and associates with pathogenesis and development of various cancers. However, the role of PAX gene family in lung squamous cell carcinoma (LUSC) remains largely unknown. This study aimed to explore its clinical significance and potential molecular mechanism in LUSC.

Methods LUSC data was downloaded from TCGA and GEO database to analyze the correlation between PAX gene family and LUSC prognostic. The possible mechanism of PAX gene family acting on LUSC was explored via biological information annotation. The co-expression and interaction network analysis method were performed among PAX gene family and relative genes.

Results Significant differential expression of PAX gene family was found between LUSC and normal tissues. Prominently, PAX9 showed the strongest diagnostic ability for LUSC. According to Cox model, PAX9 might be an independent prognostic factor of LUSC. GO and KEGG analysis suggested PAX gene family was related to transcription, DNA binding, etc. PAX9 co-expressed with the genes included KRT13, TMPRSS4, etc. and might act on LUSC through pathways such as MYC pathway, EIF pathway, etc.

Conclusion PAX9 is expected as a potential target for the diagnosis and prognostic of LUSC.

PAX

LUSC

Diagnosis

Prognosis

Bioinformatics

Lung cancer (LC), with the highest morbidity and mortality among all malignancies, is one of the most common causes of death worldwide today(Bray et al. 2018). There were more than 2.2 million new cases and 1.8 million deaths of LC worldwide in 2020 (Sung et al. 2021). The five-year overall survival for LC varied from 5% to 18% based on the cancer region and stage (Nasim et al. 2019; Torre et al. 2016). LC includes small cell lung cancer (SCLC) and non-small cell lung cancer (NSCLC). Particularly, NSCLC accounts for 85% of LC (Duma et al. 2019).

LUSC represents a major subtype of NSCLC, shows a noted for the high heterogeneity (Mazieres et al. 2020). Clinically, it is considered that the development of LUSC is closely related to smoking behavior (Takeuchi et al. 2012). At present, about 400000 people worldwide are affected by LUSC every year, but there are not effective LUSC targeted therapys (Duma et al. 2019). LUSC and lung adenocarcinoma each has a group of driving genes. EGFR, ALK, KRAS, BRAF and ROS1 are the common driving genes in lung adenocarcinoma, most of which are currently available or potential targets for targeted therapy( Li K et al. 2017; Rossi A et al. 2014; Sumi T et al. 2018).It’s still necessary and valuable to explore the clinical significance and potential molecular mechanism in LUSC.

The paired-box (PAX) gene family is an evolutionarily conserved gene family, which encodes a series of transcription factors related to growth and development regulation(Grimley et al. 2018). PAX gene family is also involved in multiple intracellular signal transduction pathways and plays key role in embryonic development, tissue and organ formation, and differentiation (Lang et al. 2007; Grimley et al. 2018).

Abnormalition of the function and structure of the PAX gene family has been found in a variety of tumors. But its role in LUSC was not fully understood. Chen et al discovered that miRNA-744-5p could impact the proliferation and invasion of NSCLC cell lines by targeting PAX2 (Chen et al. 2019). PAX2 was found to be differentially expressed between NSCLC tissues and pneumonia tissues. Tissues along with lymphatic metastasis possess showed significantly higher positive PAX2 and PAX5 expression (Ren et al. 2015). It is found that PAX9 high expressed in human oral mucosal squamous cell carcinoma and salivary gland tumor (Lee et al. 2008), and low expressed in esophageal squamous cell carcinoma and melanoma (Tan et al. 2017; Xiong et al. 2018). PAX9 was relative with various cell signaling pathways such as Wnt, PI3K/AKT, and BMP (Gerber et al. 2002; Hsu et al. 2009). So we aim to study the role of PAX gene family in the occurrence and development of LUSC to provide reference for the diagnostic and therapeutic targets research.

Data collection

Transcriptome data and clinicopathological information from TCGA database (https://portal.gdc.cancer.gov/) and microarray data from the GEO database (https://www.ncbi.nlm.nih.gov/GEO/) were downloaded for analysis. Transcriptome data of PAX gene family in LUSC and adjacent normal lung tissues were retrieved. The prognostic signature was further validated through the GEO database.

Differential expression analysis for PAX gene family

The expression analysis of PAX gene family was conducted based on TCGA database using the limma package of R software. The transcriptome character of PAX gene family was further verify by GEO dataset analysis. The Student's t test was applied to inspect the statistical significance(P<0.05).

Diagnostic efficiency assessment

ROC curve was carried out to assesse the diagnostic efficiency of PAX gene family with AUC (Aera Under the Curve) serving as the criterion. AUC>0.8 and P<0.05 was considered to possess good diagnostic efficacy.

Correlation with clinicopathological factors

We explored the correlation between clinicopathological factors and PAX menbers. Chi-squared test was used to compare the constituent ratio of age, gender, smoking, residual resection, TNM and pathologic stage based on TCGA and GEO databases. P<0.05 was considered as statistically significant.

Survival analysis

K-M method with log-rank test and Cox proportional hazards model with univariate/multivariate survival analysis were used to study the prognostic significance of PAX gene family in LUSC patient based on clinicopathologic information in TCGA and GEO databases. P<0.05 was considered statistically significant difference in prognosis.

Nomogram model

To assess the 1-year, 3-year, and 5-year Overall Survival (OS) for LUSC patients in TCGA database, nomogram model integrated multiple predictive indicators, and validated their predictive prognosis ability by plotting calibration curves with rms package in R. PAX members and clinical characteristics were inspected in multivariate Cox proportional hazards model to construct the prognostic model. K-M method was used to compare the prognosis ability of high and low risk groups. Survival curve and ROC curve was build in R studio to further evaluate the prediction effect.

GO and KEGG pathway analysis

To explore biological process, cellular component and molecular function related with PAX gene family, function enrichment analysis by GO and KEGG pathway analysis was performed on Database for Annotation, Visualization, and Integrated Discovery (DAVID) (https://david.ncifcrf.gov/, 2021 Update). Homo sapiens was selected to be biological background. P<0.05 considered as statistically significant.

Genome-wide co-expression analysis

Pearson correlation coefficient between all genes and PAX gene family was calculated in genome-wide co-expression analysis. Genes with Pearson correlation coefficient> 0.40 and P< 0.05 were considered to be coexpressed genes. The gene-gene interaction were derived by GeneMANIA analysis (http://genemania.org/).

Gene set enrichment analysis (GSEA)

LUSC patients were separated into high and low expression groups for GSEA through TCGA dataset. Curated gene sets was obtained from MsigDB database (https://www.GSEa-msigdb.org/GSEa/downloads.jsp). Each analysis was repeated 1000 times with the False Discovery Rate (FDR) correction. FDR<0.25 and P<0.05 meanwhile was considered as statistically significant.

Statistical analysis

The significance of the expression difference between LUSC tissues and normal lung tissue was assessed with Student’s t-test. ROC curve was used to evaluated the diagnostic efficiency. The Chi-squared test was employed to estimate the correlativity between clinicopathological characteristics and expression. K-M analysis log-rank test and Cox proportional hazards regression model were respectively used for univariate and multiple-factor survival analysis. P< 0.05 was considered statistically significant in all analysis, and FDR<0.25 necessary for GSEA. Statistical analysis was performed on SPSS 22.0 (SPSS, Inc.).

Differential expression analysis of PAX gene family

Transcriptome data of 496 LUSC and 49 adjacent normal tissues were acquired from TCGA database. Since 8 patients missed prognostic data, a total of 488 cases were included. And 2 GEO datasets (GSE21933 and GSE33479) of LUSC and adjacent normal tissues were downloaded subsequently.

Compared with normal lung tissue, PAX gene family, except PAX4/5, shown significantly differential expression in LUSC in TCGA dataset(Fig.1). As shown in Fig.2, PAX2/6/9 were observed with differential expression between LUSC and normal lung tissue in GSE21933 dataset. According to Fig.3, PAX1/2/9 expressed significantly different between LUSC and normal lung tissue in GSE33478. Comprehensive evaluation of these results preliminarily indicated that PAX2/9 emerged to a consistent transcriptome tendency among TCGA, GSE21933 and GSE33479 datasets.

Diagnostic efficiency of PAX gene family in LUSC

The outcomes of ROC curve exhibited in Fig. 4, PAX3 (AUC=0.818, P<0.001) and PAX9 (AUC=0.900, P<0.001) performed higher diagnostic efficiency in TCGA dataset. As for GSE21933 dataset, PAX6 (AUC=0.940, P<0.001) and PAX9 (AUC = 0.960, P < 0.001) indicated superior predictive diagnostic accuracy (Fig.5). We further evaluated the diagnostic significance of PAX gene family in LUSC via GSE33479 dataset, only PAX9 (AUC = 0.833, P < 0.001) shown strong diagnostic ability. All datasets confirmed PAX9 a good concordance with the diagnostic prediction.

Correlation with clinicopathological factors

To explore correlation between PAX9 expression and clinicopathological factors, GSE73403 with 69 LUSC cases were also obtained from GEO database. We found no significant clinicopathological factors among age, gender, smoking degree, residual resection, TNM and pathologic stage in TCGA dataset (Table 1). And only high TNM stage shown significance between PAX9 expression in LUSC based on GSE73403 dataset (Table 2).

Table 1. correlation between PAX9 expression and clinicopathologic feature in TCGA

Variables	Number of cases	PAX9 expression		X²	P value
Variables	Number of cases	low	high	X²	P value
Age(years)
<70	268	128	140
≥70	215	114	101	1.321	0.250
Missing	5
Gender
Female	126	63	63
Male	362	181	181	0.000	1.000
Missing	0
Smoking degree
low	231	118	113
high	245	118	127	0.405	0.524
Missing	12
residual resection
R0	390	203	187
R1+R2	37	15	22	1.792	0.181
Missing	61
TNM stage
Ⅰ+Ⅱ	327	167	160
Ⅲ+Ⅳ	88	42	46	0.310	0.578
Missing	73
Pathologic stage
1+2	394	201	193
3+4	90	41	49	0.874	0.350
Missing	4

Table 2. correlation between PAX9 expression and clinicopathologic feature in GSE73403 dataset.

Variables	Number of cases	PAX9 expression		X²	P value
Variables	Number of cases	low	high	X²	P value
Age
<60	35	20	15
≥60	34	14	20	1.759	0.185
Gender
Femal	4	1	3
Male	65	33	32	1.001	0.317
Pathologic stage
Low	46	25	21
High	23	9	14	1.421	0.233
Smoking
No	39	19	20
Yes	30	15	15	0.011	0.916
TNM stage
I/II	43	14	29
III/IV	26	20	6	11.457	0.001

Survival analysis of PAX gene family in LUSC

In Cox regression with 488 LUSC patients whose basic characteristics were variables, level of smoking, residual resection, TNM stage(Ⅲ+Ⅳ) and pathologic stage(Ⅲ+Ⅳ) were observed to be associated with overall survival (OS) in TCGA dataset (Table 3). These clinicopathological factors were then included into Cox proportional hazards model to adjust for the correlation between PAX gene expression and prognosis. According to Fig.7 and Table 4, PAX9 expression was associated with the OS of LUSC (P = 0.007 adjusted to P = 0.002), and PAX5 was associated with OS of LUSC only after adjusted in the Cox proportional hazards model (P=0.157 adjusted to P=0.023), while PAX4 shown no significance after adjustion (P=0.017 adjusted to P=0.080). In addition, combined effect survival analysis of PAX5 and PAX9 expression showed statistically significant difference between high-low risk group (P = 0.007, Fig. 7J adjusted to P = 0.001, Table 5).

Table 3. Basic characteristics of 488 LUSC patients

Variables	Patients	Overall survival
Variables	(n=488）	No. of events	MST (days)	HR (95% CI)	P
Age(years)
<70	268	107	1655	1
≥70	215	103	1645	1.168(0.847-1.610)	0.350
Missing	5
Gender
Female	126	49	1856	1
Male	362	161	1640	1.138(0.867-1.493)	0.343
Missing	0
Level of smoking
low	231	107	1874	1
high	245	100	1189	1.369(1.040-1.802)	0.024
Missing	12
Residual resection
R0	390	169	1695	1
R1+R2	37	16	1075	1.879(1.118-3.160)	0.016
Missing	61
TNM stage
Ⅰ+Ⅱ	327	135	1933	1
Ⅲ+Ⅳ	88	47	951	1.669(1.196-2.330)	0.002
Missing	73
Pathologic stage
Ⅰ+Ⅱ	394	159	1841	1
Ⅲ+Ⅳ	90	49	965	1.559(1.131-2.149)	0.006
Missing	4

Table 4. Prognostic significant of PAX genes expression in LUSC of TCGA database.

Gene expression	Patients (n=488)	Overall survival
Gene expression	Patients (n=488)	NO. of event	MST (days)	Crude HR (95% CI)	Crude P	Adjusted HR (95% CI)	Adjusted P §
PAX1
Low	244	104	1655
High	244	106	1679	0.977(0.744-1.282)	0.867	1.003(0.733-1.373)	0.985
PAX2
Low	244	108	1485
High	244	102	1713	0.932(0.710-1.223)	0.611	1.057(0.733-1.445)	0.730
PAX3
Low	244	112	1485
High	244	98	1841	0.886(0.675-1.163)	0.383	0.869(0.656-1.223)	0.488
PAX4
Low	244	122	1154
High	244	88	1975	0.716(0.544-0.942)	0.017	0.755(0.551-1.034)	0.080
PAX5
Low	244	101	1984
High	244	109	1470	1.217(0.927-1.598)	0.157	1.444(1.052-1.981)	0.023
PAX6
Low	244	102	1470
High	244	108	1713	0.965(0.736-1.267)	0.799	1.056(0.771-1.446)	0.735
PAX7
Low	244	97	1915
High	244	113	1426	1.197(0.911-1.571)	0.196	1.229(0.900-1.680)	0.195
PAX8
Low	244	109	1841
High	244	101	1640	1.147(0.874-1.506)	0.322	1.107(0.809-1.514)	0.525
PAX9
Low	244	119	1315
High	244	91	2160	0.686(0.521-0.902)	0.007	0.606(0.440-0.835)	0.002

Table 5. Combined effect analysis of PAX5 and PAX9 expression in HCC patients RFS.

Group	PAX5	PAX9	Patients	NO. of event	MST (months)	Crude HR (95% CI)	Crude P	Adjusted HR (95% CI)	Adjusted P §
A	Low	High	137	54	2803	1		1
B	Low	Low	107	47	1426	1.239(0.835-1.840)	0.288	1.405(0.905-2.182)	0.13
C	High	High	107	37	1841	0.990(0.650-1.506)	0.962	1.139(0.717-1.809)	0.581
D	High	Low	137	72	1656	1.631(1.144-2.326)	0.007	1.985(1.340-2.940)	0.001

Further validation was conducted using GSE73403 dataset (PAX2 data missed), high pathologic stage and level of smoking were found to be associated with OS of LUSC (Table 6). These clinicopathological factors were then included in Cox proportional hazards model to adjust for the relationship between PAX9 expression and prognosis. PAX9 expression was associated with OS of LUSC(P=0.011, Fig 8I adjusted to P=0.001, Table 7).

Table 6. Basic characteristics of 69 LUSC patients in GSE73403.

Variables	Patients	overall survival
Variables	(n=69）	No. of events	MST (years)	HR (95% CI)	P
Age
<60	35	14	NA	1
>=60	34	14	3.830	0.902(0.429-1.895)	0.783
Gender
Femal	4	0	NA	1
Male	65	28	NA	22.017(0.010-42.839)	0.225
Pathologic stage
Low	46	15	5.080	1
High	23	13	2.420	2.386(1.114-5.111)	0.020
Smoking
No	39	11	NA	1
Yes	30	17	2.420	2.463(1.152-5.265)	0.015
TNM stage
Low	43	17	3.830	1
High	26	11	NA	1.179(0.550-2.528)	0.675

Table 7. Prognostic significant of PAX genes expression in LUSC of GSE73403.

Gene expression	Patients (n=69)	Overall survival
		NO. of event	MST (years)	Crude HR (95% CI)	Crude P	Adjusted HR (95% CI)	Adjusted P §
		NO. of event	MST (years)	Crude HR (95% CI)	Crude P	Adjusted HR (95% CI)	Adjusted P §
PAX1
Low	34	15	5.080	1.000
High	35	13	3.500	0.970(0.461-2.040)	0.935	1.068(0.490-2.328	0.868
PAX3
Low	34	14	3.830	1.000
High	35	14	3.500	1.481(0.698-3.141)	0.299	2.081(0.918-40715)	0.079
PAX4
Low	34	17	5.080	1.000
High	35	11	3.830	0.980(0.453-2.121)	0.959	0.812(0.354-1.863)	0.623
PAX5
Low	34	17	3.170	1.000
High	35	11	NA	0.603(0.280-1.299)	0.188	0.485(0.213-1.104)	0.085
PAX6
Low	34	15	5.080	1.000
High	35	13	3.830	1.216(0.576-2.567)	0.605	1.270(0.599-2.690)	0.533
PAX7
Low	34	17	3.830	1.000
High	35	11	5.080	0.976(0.454-2.097)	0.949	0.916(0.421-1.993)	0.825
PAX8
Low	34	14	3.830	1.000
High	35	14	3.500	1.416(0.673-2.977)	0.352	1.486(0.697-3.168)	0.305
PAX9
Low	34	19	2.580	1.000
High	35	9	NA	0.370(0.166-0.824)	0.011	0.240(0.102-0.565)	0.001

Nomogram

Nomograms were plotted to further evaluate the predictive significance of each variables, including PAX5 and PAX9 expression, smoking degree, residual resection, pathologic stage and TNM stage in TCGA dataset (Fig. 9A). The accuracy of nomograms was evaluated using calibration curve (Fig. 9B-D). Based on different risk scores, patients were divided into high and low risk groups using “R” package (Fig. 9E-F). The heat map of PAX5/9 expression in TCGA LUSC dataset shown in Fig. 9G. According to Fig. 9G, time-dependent survival curve shown significant prognostic ability (P<0.001). The outcomes of ROC curve of 1-, 2-, 3-, 4- and 5-year survival in prognostic model exhibited in Fig. 9I.

GO and KEGG analysis of PAX gene family

We attempted to explore the role of PAX gene family played in LUSC. The details about GO and KEGG pathways enriched for PAX gene family were displayed in Table S1. The results indicated PAX gene family might play a crucial role in transcription, sequence-specific DNA binding and so on. Corresponding relationship between biological functions and PAX gene family was visualized by chord diagram (Figure 10A). The corresponding log (P value) of each biological function was shown in the bubble diagram (Figure 10B).

Genome‐wide and co‐expression analysis

The purpose of genome‐wide co‐expression analysis was conducted to further sought the capabilities of PAX9. We screened out the genes associated with PAX9 expression used bioinformatics tools. The expression profile of these genes in conjunction with PAX9 were shown at the heat map (Figure 11A), including KRT13, TMPRSS4, CYP2S1, SOX2, P2RY1, RGMA, GBP6, PITX1, FOXA1, FBXO34, etc. Genes are positively(red) or negatively(blue) correlated with the expression of PAX9. Meanwhile, the absolute value of all the correlation coefficients is greater than 0.40. GO and KEGG pathways enriched for PAX9 and co‐expressed genes was exhibited in chord diagram and bubble diagram (Figure 11B, C), including transcription factor activity, sequence specific DNA binding, positive regulation of transcription from RNA polymerase II promoter, etc. The details list shown in Table S2. Gene-gene interaction among PAX9 and related genes was displayed in network diagram (Figure 11D).

GSEA

Comprehensively analysis of the whole genome characteristics was carried out using the GSEA computing platform. To evaluate the PAX9 related biological processes, we divided patients into high and low groups according to the expression of PAX9 in LUSC tissues. Results indicated that PAX9 was mainly associated with MYC pathway, EIF pathway, MCM pathway, p53 regulation pathway, Wnt canonical pathway, IGF1-mTOR pathway, mTOR pathway, PITX2 pathway, SMAD2/3pathway, ERK pathway, E2F pathway, FOXO pathway, DNA repair, base excision repair, DNA replication, mismatch repair and so on. The representative GSEA results were listed in the supplementary material (Table S3).

LUSC is a subtype of non-small cell carcinoma that accounts for about 40% of all pathological types of lung cancer (Mazieres et al. 2020; Coudray N et al. 2018). Although recent successes of immune checkpoint blockade-based combination therapies have provided significant insights into implementing combination strategies in LUSC, definitive predictive biomarkers remain to be elucidated (Yuan H et al. 2021). PAX gene family is an evolutionarily conserved gene family, which encodes a series of transcription factors related to growth and development regulation (Grimley et al. 2018). This study aimed to explore the correlation between PAX gene family and LUSC based on bioinformatics.

High-throughput microarray platforms are a promising and useful tool for detecting oncogene alterations to uncover disease biomarkers (Malo N et al. 2006; Fernandes TG et al.2009). Comprehensively analyzing molecular characteristics including the gene expression signature may help to better understand the role of genes/events in disease progression, and thus facilitate developing effective therapeutic and management strategies to improve patient survival. By integrating multiple microarray data sets, we can get even more convincing results. In this investigation, differential expression analysis revealed PAX2 and PAX9 had consistent significant transcriptome tendency in LUSC among TCGA, GSE21933 and GSE33479 datasets. PAX9 was confirmed with good diagnostic efficiency among all PAX members based on these databases. Besides, high TNM stage was significantly correlated with PAX9 expression in GSE73403 dataset. This suggested that PAX9 might be a potential diagnostic biomarker of LUSC.

Accumulating evidence showed that abnormally expressed genes have the potential prognosis efficiency in LUSC. Cox proportional hazards model shown PAX9 was significantly associated with OS of LUSC in TCGA and GSE73403 datasets. Based on prognostic related clinicopathological features, prognostic nomogram was established subsequently. According to the nomogram, we observed that PAX9 produced considerable impact on prognosis of LUSC combine with clinicopathological covariates such as smoking, residual resection and clinical stage. Results validated PAX9 might be a good prognostic biomarker in LUSC.

PAX9 locates at the chromosome 14 (14q13.3) (Yin W et al. 2015; Neubuser A et al. 1995), which characterized by the DNA binding domain and acts as a transcription factor in the tumorigenesis and progression. Previous studies on PAX9 mainly focused on embryonic development (Kist et al. 2014; Jaskoll T et al. 2002; Sivakamasundari et al. 2017), especially bone and tooth formation (Sweat et al. 2020; Arcuri C et al. 2011; Brook AH et al. 2009; Kula K, et al. 2008; Safari et al. 2020). Study showed that PAX9 mutations played a key role in cleft palate (Jia et al. 2017), oral squamous carcinoma cell (Lee et al. 2008), esophageal squamous cell carcinoma (Tan et al. 2017) and Head-Neck Squamous Cell Carcinoma (Bai et al. 2019). Besides, PAX9 deficiency was found to be associated with excessive proliferation and delayed differentiation of esophageal squamous epithelial cells in mice (Xiong et al. 2018).

In order to explore the mechanism of PAX gene family, we conducted GO and KEGG enrichment analysis on PAX members. The results showed that PAX gene family was related to transcription, DNA binding and other functions. Genes co-expressed with PAX9 included KRT13, TMPRSS4, CYP2S1, etc. Previous studies have shown that overexpression and knockout of KRT13 have a significant effect on the biological behavior of breast cancer cells (Yin et al. 2022). High levels of TMPRSS4 are associated with poor prognosis of squamous cell carcinoma (Larzabal et al. 2011), and TMPRSS4 gene knockout can improve the sensitivity of lung squamous cell carcinoma cells to chemotherapy (Exposito et al. 2019). PAX gene family might play a crucial role in transcription factor activity, sequence specific DNA binding, response to growth factors, and the enrichment signal pathway is mainly metabolic signal pathway.

GSEA results shown that PAX9 related biological processes include MYC pathway, EIF pathway, MCM pathway, p53 regulation pathway, so on. The upregulation of MYC promotes the proliferation of NSCLC (Hua et al. 2019). MCMs can lead to poor prognosis in patients with squamous cell carcinoma (Liu et al. 2017). The regulation of MDM2-p53 pathway can promote the proliferation of NSCLC cells (Sun et al. 2017). Up regulation of p53-miR pathway and inhibition of the growth of pc-629-3j cells (Pan et al. 2021). Thus, PAX9 can play a role through a variety of signal pathways acting on LUSC.

Based on the data from large-scale gene expression profiling and two step strategy including discovery and validation, this study provides solid evidence about the value of the gene expression signature in prognostic stratification of LUSC. Nevertheless, the insufficient sample size of GEO datasets caused limitation in this study. More comprehensive analyses are warranted to accurately predict diagnostic and prognosis effect in LUSC.

Taken together, this study preliminarily explained PAX9 was a diagnostic and prognostic biomarker in LUSC. The findings may improve the understanding of LUCS prognostication and facilitate risk stratified management.

LUSC	Lung squamous cell carcinoma
PAX	Paired box
TCGA	The Cancer Genome Atlas
GEO	Gene Expression Omnibus
ROC	Receiver Operator Characteristic
K-M	Kaplan-Meier
TNM	Tumor/ regional lymph node/ metastasis
GO	Gene Ontology
KEGG	Kyoto Encyclopedia of Genes and Genomes
KRT13	keratin 13
TMPRSS4	transmembrane serine protease 4
CYP2S1	cytochrome P450 family 2 subfamily S member 1
SOX2	SRY-box transcription factor 2
P2RY1	purinergic receptor P2Y1
RGMA	repulsive guidance molecule BMP co-receptor a
GBP6	guanylate binding protein family member 6
PITX1	paired like homeodomain 1
FOXA1	forkhead box A1
FBXO34	F-box protein 34
EIF	eukaryotic translation initiation factor-like protein
IGF	insulin like growth factor
mTOR	mechanistic target of rapamycin kinase
LC	Lung cancer
SCLC	small cell lung cancer
NSCLC	non-small cell lung cancer
LUAD	lung adenocarcinoma
EGFR	epidermal growth factor receptor
ALK	ALK receptor tyrosine kinase
KRAS	KRAS proto-oncogene, GTPase
BRAF	B-Raf proto-oncogene, serine/threonine kinase
ROS1	ROS proto-oncogene 1
PI3K	phosphatidylinositol 3-kinase
AKT	AKT serine/threonine kinase 1
AUC	Aera Under the Curve
OS	Overall Survival
DAVID	the Database for Annotation,Visualization and Integrated Discovery
GSEA	Gene set enrichment analysis
FDR	False Discovery Rate
PITX2	paired like homeodomain 2
SMAD2/3	Mothers against dpp 2/3
ERK	extracellular regulated MAP kinase
E2F	transcription factor E2F
FOXO	forkhead box, sub-group O
SNX31	sorting nexin 31
C12ORF54	chromosome 12 open reading frame 54
SLC25A21	solute carrier family 25 member 21
FUT6	fucosyltransferase 6
CYP2C18	cytochrome P450 family 2 subfamily C member 18
PART1	prostate androgen-regulated transcript 1
FOXP2	forkhead box P2
ZNF812P	zinc finger protein 812, pseudogene
GRHL3	grainyhead like transcription factor 3
MCM5	minichromosome maintenance complex component 5
MDM2	MDM2 proto-oncogene
EMT	Epithelial-mesenchymal transition

Acknowledgement The authors thank the contributors of GSE21933, GSE33479, GSE73403 and TCGA database for sharing the data of LUSC on open access.

Funding This work was supported by the 2020 Guangxi Health and Medical Self-raised Fund Project "Expression and Biological Mechanism of the Four-transmembrane Protein Tspan29 in Hepatocellular Carcinoma" (Project No. Z20201311, Project Duration: January 2022 to December 2023), the 2021 Wuming District Key R&D Project "Analysis of the Correlation between JOSD1 and Prognosis of Liver Cancer and Molecular Mechanism Research" (Project No. 20210124, Project Duration: July 2022 to June 2025), and the 2022 Open Project of Guangxi Key Laboratory of Medical Genetics and Genomics Research (Project No. GXKMGG202202, Project Duration: January 2023 to December 2024).

Author contributions WZ conceptualizes and designs most projects, and writes original drafts. HC design project, download and analyze data, and write original manuscript. JC organizes the charts and writes the original manuscript. Supervision and verification by XZ. YT and XC review and revise the manuscript. LW and YZ guide the project, review and revise the manuscript. All authors have read and approved the final manuscript.

Availability of data and materials The dataset used and/or analyzed in the current research field can be downloaded from the TCGA and GEO databases, or obtained from the corresponding author upon reasonable request.

Conflict of interest The authors declare no competing interests.

Ethics approval and consent to participate Not available.

Bray F, Ferlay J, Soerjomataram I, et al (2018) Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA: a cancer journal for clinicians. 2018, 68(6):394–424. https://doi.org/10.3322/caac.21492
Sung H, Ferlay J, Siegel RL, et al (2021) Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin. 2021, 71(3):209–249. https://doi.org/10.3322/caac.21660
Nasim F, Sabath BF, Eapen GA (2019) Lung Cancer. Med Clin North Am. 2019, 103(3):463–473. https://doi.org/10.1016/j.mcna.2018.12.006
Torre LA, Siegel RL, Jemal A (2016) Lung Cancer Statistics. Adv Exp Med Biol. 2016, 893:1–19. https://doi.org/10.1007/978-3-319-24223-1_1
Duma N, Santana-Davila R, Molina JR (2019) Non-Small Cell Lung Cancer: Epidemiology, Screening, Diagnosis, and Treatment. Mayo Clin Proc. 2019,94(8):1623–1640. https://doi.org/10.1016/j.mayocp.2019.01.013
Mazieres J, Cropet C, Montané L, et al (2020) Vemurafenib in non-small-cell lung cancer patients with BRAF and BRAF mutations. Ann Oncol. 2020,31(2):289–294. https://doi.org/10.1016/j.annonc.2019.10.022
Takeuchi K, Soda M, Togashi Y, et al (2012) RET, ROS1 and ALK fusions in lung cancer. Nature medicine. 2012, 18(3):378–381. https://doi.org/10.1038/nm.2658
Li K, Yang M, Liang N, Li S (2017) Determining EGFR-TKI sensitivity of G719X and other uncommon EGFR mutations in non-small cell lung cancer: Perplexity and solution (Review). Oncology reports. 2017;37(3):1347–1358. https://doi.org/10.3892/or.2017.5409
Rossi A, Maione P, Sacco PC, et al (2014) ALK inhibitors and advanced non-small cell lung cancer (review). Int J Oncol. 2014;45(2):499–508. https://doi.org/10.3892/ijo.2014.2475
Sumi T, Hirai S, Yamaguchi M, et al (2018) Survivin knockdown induces senescence in TTF–1-expressing, KRAS-mutant lung adenocarcinomas. Int J Oncol. 2018;53(1):33–46. https://doi.org/10.3892/ijo.2018.4365
Lang D, Powell SK, Plummer RS, et al (2007) PAX genes: roles in development, pathophysiology, and cancer. Biochem Pharmacol. 2007,73(1):1–14. https://doi.org/10.1016/j.bcp.2006.06.024
Grimley E, Dressler GR (2018) Are PAX proteins potential therapeutic targets in kidney disease and cancer? Kidney Int. 2018, 94(2):259–267. https://doi.org/10.1016/j.kint.2018.01.025
Chen S, Shi F, Zhang W, et al (2019) miR-744-5p Inhibits Non-Small Cell Lung Cancer Proliferation and Invasion by Directly Targeting PAX2. Technol Cancer Res Treat. 2019,18:1533033819876913. https://doi.org/10.1177/1533033819876913
Ren Y, Hou J, Xu A, et al (2015) Diagnostic utility of PAX2 and PAX5 in distinguishing non-small cell lung cancer from small cell lung cancer. Int J Clin Exp Pathol. 2015, 8(11):14709–14716.
Lee J C, Sharma M, Lee Y-H, et al (2008) Pax9 mediated cell survival in oral squamous carcinoma cell enhanced by c-myb. Cell Biochem Funct. 2008, 26(8):892–899. https://doi.org/10.1002/cbf.1522
Xiong Z, Ren S, Chen H, et al (2018) PAX9 regulates squamous cell differentiation and carcinogenesis in the oro-oesophageal epithelium. J Pathol. 2018, 244(2):164–175. https://doi.org/10.1002/path.4998
Tan B, Wang J, Song Q, et al (2017) Prognostic value of PAX9 in patients with esophageal squamous cell carcinoma and its prediction value to radiation sensitivity. Mol Med Rep. 2017, 16(1):806–816. https://doi.org/10.3892/mmr.2017.6626
Gerber J-K, Richter T, Kremmer E, et al (2002) Progressive loss of PAX9 expression correlates with increasing malignancy of dysplastic and cancerous epithelium of the human oesophagus. J Pathol. 2002, 197(3):293–297. https://doi.org/10.1002/path.1115
Hsu DS, Acharya CR, Balakumaran BS, et al (2009) Characterizing the developmental pathways TTF-1, NKX2-8, and PAX9 in lung cancer. Proc Natl Acad Sci USA. 2009, 106(13):5312–5317. https://doi.org/10.1073/pnas.0900827106
Coudray N, Ocampo PS, Sakellaropoulos T, et al (2018) Classification and mutation prediction from non-small cell lung cancer histopathology images using deep learning. Nat Med. 2018;24(10):1559–1567. https://doi.org/10.1038/s41591-018-0177-5
Yuan H, Liu J, Zhang J (2021) The Current Landscape of Immune Checkpoint Blockade in Metastatic Lung Squamous Cell Carcinoma. Molecules. 2021;26(5):1392. https://doi.org/10.3390/molecules26051392
Malo N, Hanley JA, Cerquozzi S, Pelletier J, Nadon R (2006) Statistical practice in high-throughput screening data analysis. Nat Biotechnol. 2006;24(2):167–175. https://doi.org/10.1038/nbt1186
Fernandes TG, Diogo MM, Clark DS, Dordick JS, Cabral JM (2009) High-throughput cellular microarray platforms: applications in drug discovery, toxicology and stem cell research. Trends Biotechnol. 2009;27(6):342–349. https://doi.org/10.1016/j.tibtech.2009.02.009
Yin W, Bian Z (2015) The Gene Network Underlying Hypodontia. J Dent Res. 2015;94(7):878–885.
Neubuser A, Koseki H, Balling R (1995) Characterization and developmental expression of Pax9, a paired-box-containing gene related to Pax1. Dev Biol. 1995;170(2):701–716. https://doi.org/10.1177/0022034515583999
Kist R, Watson M, Crosier M, et al (2014) The formation of endoderm-derived taste sensory organs requires a Pax9-dependent expansion of embryonic taste bud progenitor cells. PLoS Genet. 2014, 10(10):e1004709. https://doi.org/10.1371/journal.pgen.1004709
Jaskoll T, Zhou YM, Chai Y, et al (2002) Embryonic submandibular gland morphogenesis: stage-specific protein localization of FGFs, BMPs, Pax6 and Pax9 in normal mice and abnormal SMG phenotypes in FgfR2-IIIc(+/Delta), BMP7(-/-) and Pax6(-/-) mice. Cells Tissues Organs. 2002;170(2–3):83–98. https://doi.org/10.1159/000046183
Sivakamasundari V, Kraus P, Sun W, et al (2017) A developmental transcriptomic analysis of Pax1 and Pax9 in embryonic intervertebral disc development. Biol Open. 2017, 6(2):187–199. https://doi.org/10.1242/bio.023218
Sweat YY, Sweat M, Mansaray M, et al (2020) Six2 regulates Pax9 expression, palatogenesis and craniofacial bone formation. Dev Biol. 2020, 458(2):246–256. https://doi.org/10.1016/j.ydbio.2019.11.010
Arcuri C, Zito I, Santini F, Muzzi F, Panetta V, Squitti R (2011) Understanding the implications of the PAX9 gene in tooth development. Eur J Paediatr Dent. 2011;12(4):245–248.
Brook AH, Elcock C, Aggarwal M, et al (2009) Tooth dimensions in hypodontia with a known PAX9 mutation. Arch Oral Biol. 2009;54 Suppl 1:S57-62. https://doi.org/10.1016/j.archoralbio.2008.05.017
Kula K, Trimmell J, Lu Y, Briscoe P, Feng JQ (2008) Tooth agenesis in a family and homozygous PAX9 mutation in exon 3: a case report. World J Orthod. 2008;9(4):e55-61.
Safari S, Ebadifar A, Najmabadi H, et al (2020) Screening PAX9, MSX1 and WNT10A Mutations in 4 Iranian Families with Non-Syndromic Tooth Agenesis. Avicenna J Med Biotechnol.2020,12 (4): 236–240.
Jia S, Zhou J, Wee Y, et al (2017) Anti-EDAR Agonist Antibody Therapy Resolves Palate Defects in Pax9(-/-) Mice. J Dent Res. 2017, 96(11):1282–1289. https://doi.org/10.1177/0022034517726073
Lee JC, Sharma M, Lee YH, et al (2008) Pax9 mediated cell survival in oral squamous carcinoma cell enhanced by c-myb. Cell Biochem Funct. 2008, 26(8):892–899. https://doi.org/10.1002/cbf.1522
Bai G, Song J, Yuan Y, et al (2019) Systematic analysis of differentially methylated expressed genes and site-speci fi c methylation as potential prognostic markers in head and neck cancer. J Cell Physiol. 2019, 234(12):22687–22702. https://doi.org/10.1002/jcp.28835
Yin L, Li Q, Mrdenovic S, et al (2022) KRT13 promotes stemness and drives metastasis in breast cancer through a plakoglobin/c-Myc signaling pathway.Breast Cancer Res. 2022, 24(1):7. https://doi.org/10.1186/s13058-022-01502-6
Larzabal L, Nguewa PA, Pio R, et al (2011) Overexpression of TMPRSS4 in non-small cell lung cancer is associated with poor prognosis in patients with squamous histology. Br J Cancer. 2011, 105(10): 1608-14. https://doi.org/10.1038/bjc.2011.432
Hua Q, Jin M, Mi B, Xu F, Li T, Zhao L, Liu J, Huang G (2019) LINC01123, a c-Myc-activated long non-coding RNA, promotes proliferation and aerobic glycolysis of non-small cell lung cancer through miR-199a-5p/c-Myc axis. J Hematol Oncol. 2019 Sep 5;12(1):91. doi: 10.1186/s13045-019-0773-y. https://doi.org/10.1186/s13045-019-0773-y
Liu YZ, Wang BS, Jiang YY, et al (2017) MCMs expression in lung cancer: implication of prognostic significance.J Cancer. 2017,8(18):3641–3647. https://doi.org/10.7150/jca.20777
Sun Y, Cao L, Sheng X, et al (2017) WDR79 promotes the proliferation of non-small cell lung cancer cells via USP7-mediated regulation of the Mdm2-p53 pathway.Cell Death Dis. 2017,8(4):e2743. https://doi.org/10.1038/cddis.2017.162
Pan J, Li M, Yu F, et al (2021) Up-Regulation of p53/miR-628-3p Pathway, a Novel Mechanism of Shikonin on Inhibiting Proliferation and Inducing Apoptosis of A549 and PC-9 Non-Small Cell Lung Cancer Cell Lines.Front Pharmacol. 2021, 12:766165. https://doi.org/10.3389/fphar.2021.766165

The Supplementary Materials are not available with this version

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Comprehensive analysis of clinical significance and potential molecular mechanism of PAX gene family in lung squamous cell carcinoma

Status:

Version 1

Abstract

Figures

Introduction

Materials and Methods

Result

Discussion

Conclusion

Abbreviations

Declarations

References

Supplementary Materials

Additional Declarations

Status:

Version 1