Identication of a FLG-wide-type bladder cancer subtype with poor prognosis and prediction use in immune checkpoint

Bladder cancer (BLCA) is one of most common urinary tract malignant tumor and immunotherapy have generated a great deal of interest in BLCA. Immune checkpoint blockade (ICB) therapy has signicantly progressed the treatment of BLCA. Multiple studies have suggested that specic genetic mutations may serve as immune biomarkers for ICB therapy. In this study, we aimed to investigate the role of mutations genes and subtypes in prognosis and immune checkpoint prediction in BLCA. Mutation information and expression proles were acquired from The Cancer Genome Atlas (TCGA) database. Integrated bioinformatics analysis was carried out to explore the mutation genes of BLCA. Functional enrichment analysis Gene Ontology (GO) and Gene set enrichment analysis (GSEA) was conducted. The inltrating immune cells and the prediction of ICB between different subtypes group were explored using immuCellAI algorithm.


Introduction
Bladder cancer (BLCA) is one of most common urinary tract malignant tumor and accounts for nearly 170,000 deaths annually in the world [1,2]. It is the ninth most incident neoplasm in China and the 10th most common malignant tumor worldwide [3,4]. BLCA has become a serious public health problem because of its high incidence, high risk of recurrence, treatment frequently failure (e.g. intravesical bacillus Calmette-Guerin or platinum-based chemotherapy) [5,6]. However, the treatment for BLCA has remained conservative and the curative effect has not made a breakthrough [7]. Due to the large number of recognizable antigens, BLCA might be sensitive to immunotherapy [8]. With the rapid development of immunotherapy, the immune checkpoint blockade (ICB) have become a novel treatment strategy for BLCA [9]. Immune checkpoint blockade refer to inhibitory drugs developed for immune checkpoints, which can rejuvenate immune cells and kill tumor cells again [10]. Therefore, the prediction of immune checkpoints is of clinical signi cance.
The Filaggrin (FLG) gene can encode a related protein that accumulates in the intermediate laments of mammalian epidermal keratin [11,12]. Previous studies have shown variability in the frequency of FLG variants [13]. The mutation of FLG is associated with a variety of skin diseases [14], and it is mutated in a variety of cancers such as head and neck cancer, prostate cancer, urinary cancer and bronchus and lung cancer [15]. It is suggested that FLG gene mutation is one of the risk factors for cancer, the associations between FLG loss-of-function mutations and cancer were in subgroup analyses [16].
Besides, studies have shown that BLCA is a highly mutated tumor type [17]. However, the association between the mutation of FLG and BLCA need further dissection.
In present study, we aimed to investigate the prognosis and immune checkpoint prediction role of gene subtype in BLCA. Mutation information and expression pro les were acquired from The Cancer Genome Atlas (TCGA) database. Integrated bioinformatics analysis was carried out to explore the mutation genes of BLCA and FLG were identi ed. Following the study on its subtype and functional enrichment analysis, their relationships with prognosis and immune in ltration were also evaluated. Finally, we ensure the important role of FLG subtype in ICB. This research provides a new basis and ideas for guiding the clinical use of BLCA immunotherapy.

Method And Materials
Data source collection and processing Transcriptome data and clinical information of BLCA patients were downloaded from the TCGA data portal (http://cancergenome.nih.gov/). The "maftools" R package was used to analyze the Mutation Annotation Format (MAF) le and visualize the somatic mutation data [18].

Identi cation of gene expression-based subtypes
Unsupervised clustering was performed using the R package "ConsensusClusterPlus" for class discovery based on the comparison of gene expression pro le [19,20]. 80% item resampling, 50 resamplings and a maximum evaluated K of 10 were selected for clustering. The cumulative distribution function (CDF) and consensus heatmap were used to assess the optimal K [21].

Differential expression analysis
The differently expressed mRNAs presented as heatmap and volcano plot were screened using GDCRNATools package, with the criterion of |log 2(fold change [FC]| > 1 and FDR < 0.05).

Functional enrichment analysis
To assess the function of differential genes between subtypes in BLCA, Gene Ontology (GO) annotation analyses were performed by using the cluster pro ler package of R software [22]. P-value < 0.05 was set as the cut-off criterion. Gene set enrichment analysis (GSEA) was conducted to examine critical pathways represented under different conditions [23]. The ridgeline plot was presented by using clusterPro ler.

Immune in ltration analysis and Immunotherapy prediction
The in ltrating immune cells and the prediction of ICB between different subtypes group were explored using immuCellAI algorithm [24].

Overview of mutation information in BLCA patients
We download BLCA gene mutation data from the TCGA database and analyzed the somatic mutation information. Figure 1 showed the overall mutation status of BLCA. As shown in Figure 1A, missense mutation is the most common mutation types with a frequency far higher than that of others. Single nucleotide polymorphism (SNP) represented the largest fraction in the variant type than insertion or deletion ( Figure 1B). The most common single nucleotide variant (SNV) was C > T, followed by C > G and C > A ( Figure 1C). The median of per variants samples was 148 ( Figure 1D) and variants classi cation summary was shown in Figure 1E. Figure  Prognostic FLG wild-type splited into two subtypes Next, we conducted a prognostic analysis of the TOP 20 mutant genes and found that FLG gene mutations are strongly related to the prognosis of patients ( Figure 2A). The survival time of FLG wild-type patients is signi cantly shorter than that of FLG mutant patients. Patients (BLCA) who harbored FLG mutations showed a better prognosis. Figure 2B suggested the mutation information of FLG gene in BLCA, the main mutation forms are: missense mutation and nonsense mutation. Based on the above results, we have made it clear that the research object in BLCA was FLG wild-type samples. We combined the BLCA transcriptome data in the TCGA database to explore whether FLG wild-type patients with high mortality rates have different types. We used the ConsensusClusterPlus software to divide the patient samples into two categories (Sub1 and Sub2) ( Figure 2C), and further analyzed the prognosis of the two types of patients. The results showed that the survival time of Sub2 patients was signi cantly less than that of Sub1 patients ( Figure 2D). At the same time, it could be seen that the mortality rate of Sub2 was as high as 90% ( Figure 2E). The death rate of Sub1 was 46%, while the death rate of Sub2 was 90%. In this analysis, we found that there are two different types of patients in FLG wild-type patients. The prognosis of Sub2 patients is worse than that of Sub1 patients. Sub2 of FLG -wild type in BLCA is usually associated with poorer outcome.

The 2 subtypes of FLG wild-type differed in biological functions
We used transcriptome data to explore the biological signi cance of two different subtypes of patients. The differently expressed genes (DEGs) between the Sub1 and Sub2 were analyzed (|log2FC| > 1 and FDR < 0.05). A total of 778 up-regulated and 972 down-regulated DEGs were identi ed ( Figure 3A-B). The results of GO functional enrichment analysis of DEGs showed that related pathways such as in ammation, chemokine production, and T cell activation were signi cantly enriched ( Figure 3C). The biological process (BP) pathway was further analyzed by the GSEA method, and the results also showed that T cells plays a key role in different types of tumor microenvironments. In addition, the results suggests that the pathways related to immunity and in ammation are signi cantly down-regulated in Sub2, and the immune response involved in T cell activation is suppressed ( Figure 3D).

Association of 2 subtypes of FLG wild-type with in ltrated immune Cells
In ltrating immune cells in tumor are a hallmark of immune surveillance and a necessary part of complex microenvironment regulating tumor progression [25]. It is known that T cells play a key role in mediating the tumor immune microenvironment, and different subtypes of T cells have different divisions of labor [26]. Therefore, the ImmuCellAI software was carried out to analyze the T cell subtypes of FLG wildtype patients with different types. As seen in Figure 4, the results showed that the CD4 naive, central memory T (Tcm), NKT were signi cantly down-regulated in Sub2, while Th1 cells were signi cantly upregulated in Sub2. CD4+ T cells can directly eliminate tumor cells through cytolysis or indirectly regulate the tumor immune microenvironment to target tumor cells [27]. NKT cells are a type of T cell subgroup with speci c NK cell markers in immune cells. After activation, they can directly act as anti-tumor effector cells to play a killing effect, and they can also activate other immune effector cells, such as NK cells, to indirectly achieve anti-tumor effect. NKT cells play an important role in anti-tumor immunity, acquired immune response and immune regulation[28].
Sub2 of FLG wild-type patients respond worse to ICB ICB is bene cial to restore T lymphocyte activity and break through the physical barrier of tumor immune microenvironment to promote T cell homing. Therefore, it can activate anti-tumor immunity and improving the effect of immunotherapy. Based on the results of the current study, enhanced tumor immunogenicity predicts improved clinical response to ICB. Finally, we used immuCellAI to perform immune check-up prediction on FLG wild-type patients of different types. As shown in Figure 5, the reaction rate of Sub1 was 33.7%, while the reaction rate of Sub2 was 10%. The results showed that Sub2 patients had a worse response to ICB. The results indicated that the detection of Sub2 FLG wild-type might be used in the clinical immunotherapy.

Discussion
Tumors are related to the accumulation of somatic mutations in cells as a result of carcinogens [29].
Studies have shown that BLCA is a highly mutated tumor type [17]. For example, Zhang et al. found that BLCA patients with higher tumor mutational burden (TMB) had more survival bene ts [17]. The mutations of FGFR3, HRAS, KRAS, NRAS and PIK3CA in BLCA can present a companion diagnostic to de ne patients for targeted therapies [30,31]. Recognizing the genetic mutations that may allow early genetic screenings for BLCA and new therapies targeting cells with these mutations [32]. The FLG gene encode a related protein that accumulates in the intermediate laments of mammalian epidermal keratin and shown variability in the frequency variants [11][12][13]. FLG gene mutation is one of the risk factors for cancer, the associations between FLG loss-of-function mutations and cancer were in subgroup analyses [16]. In this study, it's the rst time to found the survival time of FLG wild-type patients was signi cantly less than that of FLG mutant patients based on the BLCA gene mutation data in the TCGA database. Furthermore, we found that there were two different types of patients in FLG wild-type patients, sub1 and sub2. Though the sample size of sub2 is 10, the prognosis of sub2 patients was worse than that of sub1 patients, the mortality rate of Sub2 is as high as 90%. Hence, we paid attention to the biological function and pathways of sub2 in the following research.
As one of the most common urinary system cancers worldwide, BLCA ranges from unaggressive and noninvasive tumors to aggressive and invasive tumors with high disease-speci c mortality[6]. However, traditional treatments did not make signi cantly improvement for its the 5-year survival rate [33]. Recently, newer immunotherapy have generated a great deal of interest in BLCA [34]. In this study, the results of functional enrichment analysis of differential genes showed that related pathways such as in ammation, chemokine production, and T cell activation were signi cantly enriched. GSEA analysis of the BP pathway suggests that the pathways related to immunity and in ammation are signi cantly downregulated in Sub2, and the immune response involved in T cell activation was suppressed. Analysis of the T cell subtypes of FLG wild-type patients with different types showed that Sub2 of FLG wild-type was correlated with in ltrating levels of immune cells (CD4 naive, Tcm and NKT cell). The results above indicated that The Sub2 was related to the immune microenvironment of BLCA. The investigation of the relationship and biological mechanisms between FLG subtypes and immune microenvironment will help to better understand the role of immunotherapy in BLCA treatment.
Immune checkpoint inhibitors refer to inhibitory drugs developed for immune checkpoints, which can rejuvenate immune cells and kill tumor cells again [35]. The application of ICB, such as anti-programmed cell death (ligand) 1( anti-PD-(L)1)[36] and cytotoxic T lymphocyte-associated antigen 4 (anti-CTLA-4) [37], is becoming emerging as a novel treatment strategy for BLCA [38]. Therefore, the prediction of immune checkpoints is of great clinical signi cance for BLCA. Studies have shown a correlation between speci c genetic mutations and the e cacy of immunotherapy [39,40]. It was reported that TMB, microsatellite instability, mismatch repair gene de ciency, T-cell in amed and IFN-γ gene expression pro les (GEPs), and DNA damage response (DDR) and antigen presentation defects may serve as potential biomarkers for immune checkpoint of immunotherapy [41]. Zhang et al. identi ed NTRK3 as a potential prognostic biomarker associated with tumor mutation burden and immune in ltration in BLCA [17]. Lin et al. found the effect of NCOR1 mutations on immune microenvironment and e cacy of ICB in patient with BLCA [42]. In this study, we predicted the immune checkpoints of FLG wild-type patients of different types, and the results showed that Sub2 patients respond is worse to immune checkpoint. The results suggested that FLG mutilation may prove to be a novel detection and contribute to the development of immunotherapy for BLCA.

Conclusion
In conclusion, this study systematically explored the relationship between FLG wild-type and prognosis as well as immune in ltration in BLCA. The prognostic FLG wild-type spitted into two subtypes and our data revealed that Sub2 of FLG wild-type was closely correlated with survival and immune response with further exploration. Accordingly, Sub2 of FLG wild-type may prove as a novel diagnosis method and contributed to the development of immunotherapy for BLCA. Further experiments in vivo and in vitro to investigate the underlying mechanism and clinical validation is needed.

Declarations
Ethics approval and consent to participate Not applicable.

Consent for publication
Not applicable.

Availability of data and material
The original contributions presented in the study are included in the article, further inquiries can be directed to the corresponding author/s.

Competing interests
The authors declare that they have no con ict of interest.

Funding
Not applicable.

Author contributions
Liang Chen, and Xiaobo Huang designed the research. Liulin Xiong, Weinan Chen, and Lizhe An performed the experiments. Huanrui Wang, Yang Hong and Huina Wang visualized the results. Liang Chen and Liulin Xiong wrote manuscript. Xiaobo Huang provided experimental resource. All authors reviewed and approved the nal manuscript.    Association of 2 subtypes of FLG wild-type with in ltrated immune Cells Predicting immune checkpoints for 2 subtypes of FLG wild-type.