Abstract

Background

   Evidence has been presented that the tumor protein D52 (TPD52) family
   plays a critical role in tumor development and progression. As a member
   of the TPD52 family, the changes in TPD52L2 gene status are
   instrumental in kinds of cancer development. However, its effects on
   patient prognosis and immune infiltration in Head and Neck Squamous
   Carcinoma (HNSCC) are still poorly understood.

Methods

   The Tumor Genome Atlas (TCGA), Gene Expression Omnibus (GEO), and
   c-BioPortal database was used to explore the expression pattern,
   prognostic value, and variation of gene status in HNSCC. The
   LinkedOmics database was used to obtain the co-expression genes of
   TPD52L2 and identify the diagnostic value of TPD52L2 in HNSCC. The
   correlations between TPD52L2 expression and six main types of immune
   cell infiltrations and immune signatures were explored using Tumor
   Immune Estimation Resource (TIMER). The correlation between TPD52L2
   expression and immune checkpoint genes (ICGs) was analyzed by TCGA
   database. Immunohistochemistry (IHC) was performed to validate the
   expression of three ICGs (PDL1, PDL2, EGFR) and TPD52L2 using 5 paired
   HNSCC and normal head and neck tissues. Polymerase Chain Reaction (PCR)
   and Western Blot (WB) of HNSCC and normal head and neck cell lines were
   performed to verify the high level of TPD52L2 mRNA and protein
   expression. protein expression of TPD52L2 in pan-cancer was also
   validated using UALCAN.

Results

   TPD52L2 was overexpressed in tumor tissues, and it predicted worse
   survival status in HNSCC. ROC analysis suggested that TPD52L2 had a
   diagnostic value. Multivariate Cox analysis identified TPD52L2 as an
   independent negative prognostic marker of overall survival. Functional
   network analysis suggested that TPD52L2 was associated with
   immune-related signaling pathways, cell migration pathways, and
   cancer-related pathways. High expression of TPD52L2 was associated with
   a more mutant frequency of TP53. Notably, we found that the expression
   of TPD52L2 was closely negatively correlated with the infiltration
   levels of 15 types of immune cells and positively correlated with
   several immune markers. PCR, WB experiments, and UALCAN database
   verified the high level of TPD52L2 mRNA and protein expression.

Conclusion

   TPD52L2 is upregulated in HNSCC, which is an independent factor for
   adverse prognosis prediction. It probably plays a role in the negative
   regulation of immune cell infiltration. TPD52L2 might be a promising
   prognostic biomarker and therapeutic target in HNSCC.

Supplementary Information

   The online version contains supplementary material available at
   10.1186/s12903-024-04977-1.

   Keywords: Head and neck squamous carcinoma, TPD52L2, Diagnosis,
   Prognosis, Immune cells, Bioinformatics

Introduction

   As the sixth most common cancer worldwide, HNSCC affects approximately
   900,000 patients annually. The incidence of this disease is rising,
   while therapeutic options have only marginally improved [[34]1, [35]2].
   Approximately 95% of head and neck cancers are HNSCC, which develops in
   the mouth, hypopharynx, throat, or oropharynx [[36]3]. Current
   therapies of HNSCC mainly include surgery, radiotherapy, chemotherapy,
   and multidisciplinary comprehensive sequence therapy. As reported,
   sixty percent of patients are already in advanced stages (stage III or
   IV) at the time of diagnosis [[37]4, [38]5]. The survival rate among
   advanced HNSCC patients is only 34.9% and the effectively
   multidisciplinary treatment for HNSCC is still limited [[39]6, [40]7].
   Thus, there is an urgent need for novel biomarkers and new therapeutic
   options for HNSCC.

   The tumor protein D52-like family is a group of coiled proteins. This
   family consists of TPD52, TPD52L1 (TPD53), TPD52L2 (TPD54), and TPD52L3
   (TPD55). The TPD52 family has been reported to play an important role
   in the proliferation and metastasis of various cancer cells [[41]8,
   [42]9], TPD52 was shown to be overexpressed in breast cancer and
   prostate cancer and identified as a candidate oncogene [[43]10,
   [44]11]. In oral squamous cell carcinoma (OSCC), TPD52 was elevated
   under hypoxic conditions in a HIF-independent manner and promoted cell
   proliferation and survival [[45]12]. These data suggested that
   inhibition of TPD52 could contribute to cancer treatment.

   TPD52L2, a member of the TPD52-like protein family, is reported to be
   involved in a new class of intracellular transport vesicles:
   intracellular nanovesicles [[46]13]. And proteomic analysis revealed
   that TPD52L2 was one of the most abundant proteins in HeLa cells,
   ranking 180th out of 8,804 proteins, with an estimated 3:3 × 106 copies
   per cell [[47]12]. TPD52L2 has been proved to be highly expressed and
   linked to poor prognosis in several types of cancers, including lung
   adenocarcinoma [[48]14], pancreatic adenocarcinoma [[49]15, [50]16],
   and liver cancer [[51]17]. Recent studies indicate TPD52L2 to be a
   potential prognostic biomarker of survival for many cancers, but to
   date, no systematic reviews have been conducted on the role of TPD52L2
   in HNSCC.

   In the current study, we performed a bioinformatic analysis by
   extracting and analysing TCGA, GEO, TIMER, LinkedOmics, c-BioPortal and
   UALCAN datasets to explore the potential oncogenic mechanism of TPD52L2
   in HNSCC, as well as by using Gene ontology (GO) and Kyoto Encyclopedia
   of Genes and Genomes (KEGG), pathway enrichment to investigate the
   potential functions of TPD52L2. Furthermore, we performed an IHC
   experiment to verify the correlation between TPD52L2 and tumor immune
   cell infiltration/immune checkpoint, whereas expression of TPD52L2 in
   HNSCC were validated by PCR and WB experiments. These findings offer
   fresh perspectives on TPD52L2’s role as well as brand-new targets for
   HNSCC diagnosis and prognosis.

Materials and methods

Data source and processing

   Pan-cancer TCGA dataset (PANCAN, N = 10,535, G = 60,499, N stands for
   the number of cases, G stands for genes) was downloaded from UCSC
   cancer genome browser ([52]https://xenabrowser.net/) The expression of
   TPD52L2 values were log[2](1 + x) transformed for downstream analysis.
   After excluding cancer species less than 3 samples, the final cancer
   species for the analysis was 26.

   The data of the TPD52L2 expression and the associated clinical outcome
   data in HNSCC patients were downloaded from TCGA
   ([53]https://portal.gdc.cancer.gov/), including 502 tumor samples and
   44 normal samples, and then the data was converted to log[2]
   Transcripts Per Million (TPM).

   [54]GSE39400 [[55]18] dataset downloaded from GEO database
   ([56]https://www.ncbi.nlm.nih.gov/) was chosen as a validation set.
   T-test was used to compare the distribution in the expression levels of
   TPD52L2 between tumor and normal samples.

   The Copy Number Variation (CNV) datasets of HNSCC were obtained from
   TCGA and analyzed using the R package “maftools” [[57]19]. In addition,
   unpaired Wilcoxon’s ranksum and signed rank tests in R were performed
   to determine the significance of the difference between sample groups.

Analysis of the clinicopathological features and diagnostic value

   Genomic data and clinicopathological features from a total of 502 tumor
   samples of HNSCC published were downloaded from TCGA for inclusion in
   this study. The patients were divided into two groups according to high
   and low TPD52L2 expression level, separated by the median value. To
   evaluate the relationship between the expression of TPD52L2 and
   clinicopathological features in HNSCC samples, Pearson’s chi-square
   test and Fisher’s Exact Test were performed. The diagnostic performance
   of TPD52L2 to differentiate between normal tissue and tumor was
   assessed by receiver operating characteristics (ROC) curve analysis.

Prognostic analysis

   The dataset ([58]GSE65858 [[59]20], which was derived from the GEO
   database including 270 HNSCC tumor samples and corresponding gene
   expression and survival data) was used to validate the OS curves and
   Cox regression analysis in HNSCC. The samples in GEO database were
   stratified into two groups of high- and low-expression as described
   above. Then, the Kaplan–Meier analyses with the log-rank test was used
   for survival analysis in pan-cancer. Univariate and multivariate Cox
   regression analyses were conducted to explore the influence of clinical
   variables and TPD52L2 on the survival of patients with HNSCC using R
   packages “survminer” and “survival”. The prognostic variables p < 0.1
   in the univariate Cox regression analysis entered the multivariate Cox
   regression model.

Identification and enrichment analysis of co‑expression genes

   Co-expression analysis with Pearson correlation test was performed to
   identify co-expressed genes of TPD52L2 in the data of HNSCC from TCGA
   via the database LinkedOmics
   ([60]http://www.linkedomics.org/login.php). The top 120 co-expression
   genes were selected to perform GO and KEGG pathway enrichment analysis
   using the R package “clusterProfiler” (version 3.14.3) to explore the
   TPD52L2-related molecular mechanisms.

Gene mutation and CNV status analysis

   TCGA-pan-cancer gene mutation and CNV data of TPD52L2 were retrieved
   from cBioPortal database ([61]https://www.cbioportal.org/). Gene
   expression data of TPD52L2 were extracted from the Pan-cancer TCGA
   dataset (PANCAN, N = 10,535, G = 60,499). GISTIC copy number data was
   downloaded from TCGA dataset. We calculated gene expression of TPD52L2
   in tumor samples with different CNV status with the R statistical
   software (version 3.6.4), then compared differences between groups by
   unpaired Wilcoxon’s signed-rank test. The Kruskal–Wallis nonparametric
   statistical tests were performed using “kruskal.test” function from the
   stats package in R.

Gene mutation frequency analysis

   We further graphed mutational landscape of TPD52L2 using ComplexHeatmap
   [[62]21] in R from TCGA database. The chi-square test was used to
   assess the difference in the frequency of mutations in each group of
   samples.

Tumor immunity estimation resource database analysis

   TIMER ([63]https://cistrome.shinyapps.io/timer/) was utilized to
   analyze the comprehensive correlation between TPD52L2 and
   tumor-infiltrating immune cells signatures. The correlation between the
   CNV of TPD52L2 and the abundance of six types of tumors infiltrating
   immune cells (TIICs) including B cells, CD4 + T cells, CD8 + T cells,
   neutrophils, macrophages, and dendritic cells were explored.

Gene expression and immune cell infiltration analysis

   Calculation of immune infiltration scores were based on the gene
   expression data previously identified. The level of infiltration of 24
   immune cells in the tumors was evaluated based on gene expression using
   the R packages IOBR (version 0.99.9,
   [64]https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8283787/).

   The relative enrichment score of these immune cells in HNSCC was
   assessed by single-sample GSEA, which was accomplished using the R
   package GSVA. The correlation between the expression of TPD52L2 and
   these immune cells was investigated using the Spearman’s correlation
   analysis, and the differences in the level of immune infiltration
   between the high and low TPD52L2 expression groups were evaluated using
   the Wilcoxon ranksum test. The Estimation of Stromal and Immune cells
   in Malignant Tumor tissues using Expression (ESTIMATE) algorithm was
   used to calculate and compare stromal scores, immune scores, and
   ESTIMATE scores between low and high TPD52L2 expression groups of HNSCC
   by Wilcoxon rank-sum test.

Immune checkpoint gene analysis

   Patients with high expression of immune checkpoint inhibitors (ICIs)
   will benefit more from ICI treatment. We analyzed the correlation
   between the expression of TPD52L2 and more than 50 ICGs using the TCGA
   database.

   Five paired HNSCC and normal tissue specimens from Cancer Hospital of
   Shantou University Medical College were used (The immunohistochemistry
   involved patients who had undergone surgery and had not received
   chemoradiotherapy). The Ethical Committee of the Hospital approved the
   study protocol. IHC analysis used UltraSensitive SP IHC Kit-9710 (MXB,
   China), according to the manual instructions. The primary antibody used
   in IHC included: CD274 (PDL1) Monoclonal Antibody (14–5983-82,
   eBioscience invitrogen), Recombinant Anti-PDL2 antibody (ab205921,
   abcam), Recombinant Anti-EGFR antibody (ab52894, abcam) and
   Anti-TPD52L2/D54 Antibody (A07219, bosterbio). The staining intensity
   was scored as (0 = negative, 1 = weak, 2 = moderate, 3 = strong)
   according to the expression level and the staining area was scored as
   (0 = 0%, 1 = 1–25%, 2 = 26–50%, 3 = 51–100%) based on the percentage of
   positive cells. The immunostaining score (IS) was calculated by adding
   the score of staining intensity and area [[65]22].

RT-PCR and WB experiments

   The mRNA expression was analyzed by reverse transcription-polymerase
   chain reaction (RT-PCR). Primers for TPD52L2 transcripts amplification
   were forward 5’-CTCCATCCGCCACTCAAT-3’, reverse5’-
   CACCCACAACCTTAGACTTTAT-3’. Total RNA was extracted from five HNSCC
   different cell lines (cells (FD-LSC-1, Fadu, cal27, HN-6, SCC-25) vs.
   normal head and neck cells (NOK, FM)) using TRIzol® Reagent, and the
   above cells were all purchased from Guangzhou Yeshan Biotechnology Co.,
   Ltd. After identification purity and concentration, RNA was reverse
   transcribed to cDNA using M-MLV Reverse Transcriptase (Promega,
   U.S.A.). Real-time PCR amplification was performed using an SYBR Green
   Realtime PCR Master Mix (Toyobo, Japan). The mRNA expression level was
   normalized to the expression level of the housekeeping gene GAPDH.
   Relative gene expression levels were calculated using 2^−△△ Ct. Each
   experiment was performed in triplicate.

   Proteins were also extracted from the above cell lines. From sample
   preparation to detection, the reagents for WB were in Western Blotting
   Application Solutions Kit. After sample preparation, equal amounts of
   sample protein and internal control protein were loaded onto two
   SDS-PAGE gels (12% and 10% concentrations, respectively). Under the
   premise of maintaining the same conditions as much as possible,
   electrophoresis and transfer of the samples to a nitrocellulose
   membrane were performed. The membrane was blocked for 2 h at room
   temperature and incubated overnight at 4℃. Primary antibodies used are
   as followed: anti-TPD52L2 (1:1000, A07219, Bosterbio), HRP labeled
   GAPDH antibody (1:10,000, KC-5G5, Aksomics), Rabbit Anti-Mouse IgG
   (H + L) (1:4000, 6170–05, Southern Biotech), Goat Anti-Rabbit IgG
   (H + L chain specific) (1:5000, 4050–05, Southern Biotech), peroxidase-
   Rabbit Anti- Goat IgG (1:3000, BA1060, Bosterbio), peroxidase- Rabbit
   Anti- Rat IgG(1:2000, BA1058, Bosterbio).Proteins were detected using
   Medical X-ray film (Kodak). Each experiment was performed in
   triplicate.

Protein expression analysis in UALCAN

   The protein of TPD52L2 expression between normal and tumor tissues in
   pan-caner was analyzed using CPTAC data from the UALCAN website
   ([66]http://ualcan.path.uab.edu). Differences in protein level were
   compared using the Student’s t-test, p < 0.05 was considered
   statistically significant.

Results

Elevated expression of TPD52L2 in HNSCC

   Analysis of the expression of TPD52L2 in pan-cancer showed higher
   expression in 18 cancer subtypes compared to their normal counterpart
   tissue (Fig. [67]1A). Especially, TPD52L2 was over-expressed in HNSCC,
   which was concluded from paired and unpaired analyses in TCGA database
   (p < 0.001) (Fig. [68]1B, C). In [69]GSE39400 cohort, higher expression
   of TPD52L2 in HNSCC patients in peripheral blood cells (PBC) was
   validated (p < 0.001) (Fig. [70]1D). As shown in Table [71]1, TPD52L2
   expression was observed to have a strong association with T stage
   (p = 0.019). However, TPD52L2 expression level showed no significant
   correlation with other clinicopathologic features. ROC analysis showed
   a good diagnostic accuracy in discriminating normal from cancer tissues
   based on TPD52L2 mRNA expression level, with an area under the curve
   (AUC) of 0.891 [95% confidence interval (CI): 0.856–0.926], a
   sensitivity of 95.5%, and a specificity of 71.7% (Fig. [72]1E).

Fig. 1.

   [73]Fig. 1
   [74]Open in a new tab

   Expression and diagonostic value of TPD52L2. A Pan-cancer expression of
   TPD52L2. Analyses of TPD52L2 expression in (B) unpaired and (C) paired
   samples in TCGA cohort. D TPD52L2 expression was significantly
   increased in HNSCC tissues compared to normal tissues from the GEO
   datasets([75]GSE39400). E ROC curve analysis of TPD52L2 expression in
   HNSCC

Table 1.

   Clinicopathological features of HNSCC patients from TCGA dataset
   Clinicopatogical features Low expression of TPD52L2 High expression of
   TPD52L2 P value
   n 251 251
   Gender, n (%) 0.545
   Female 70 (13.9%) 64 (12.7%)
   Male 181 (36.1%) 187 (37.3%)
   Age, n (%) 0.624
    <  = 60 125 (25%) 120 (24%)
    > 60 125 (25%) 131 (26.1%)
   T stage, n (%) 0.019
   T1 21 (4.3%) 12 (2.5%)
   T2 80 (16.4%) 64 (13.1%)
   T3 69 (14.2%) 62 (12.7%)
   T4 74 (15.2%) 105 (21.6%)
   N stage, n (%) 0.195
   N0 124 (25.8%) 115 (24%)
   N1 32 (6.7%) 48 (10%)
   N2 83 (17.3%) 71 (14.8%)
   N3 3 (0.6%) 4 (0.8%)
   M stage, n (%) 1.000
   M0 236 (49.5%) 236 (49.5%)
   M1 3 (0.6%) 2 (0.4%)
   Smoker, n (%) 0.422
   No 52 (10.6%) 59 (12%)
   Yes 195 (39.6%) 186 (37.8%)
   Alcohol history, n (%) 0.866
   No 81 (16.5%) 77 (15.7%)
   Yes 168 (34.2%) 165 (33.6%)
   [76]Open in a new tab

Prognostic significance of TPD52L2

   Survival analysis demonstrated that TPD52L2 over-expression predicted
   worse overall survival (OS) of patients with BLCA, MESO, LGG, LUADLUSC,
   LIHC, BRCA, HNSCC (Fig. [77]2A, C-H). A similar result in HNSCC was
   confirmed in the GEO validation dataset ([78]GSE65858) (Fig. [79]2B).

Fig. 2.

   [80]Fig. 2
   [81]Open in a new tab

   TPD52L2 is associated with survival outcome. A Overall survival (OS) of
   TPD52L2 in TCGA HNSCC cohort. OS of TPD52L2 analysis from Kaplan–Meier
   survival curves on [82]GSE65858 (B), in TCGA BLCA (C), MESO (D), LGG
   (E), LUADLUSC (F), LIHC (G), BRCA (H) cohort

   As shown in the forest plot (Fig. [83]3A, B), univariate analysis of
   TCGA showed age, gender, N stage, M stage and TPD52L2 were significant
   prognostic factors for OS in HNSCC patients, and multivariate analysis
   showed that TPD52L2 (HR = 1.439, p = 0.01) had an independent
   prognostic value. these results were consistent with that of the
   [84]GSE65858 dataset (Fig. [85]3C, D).

Fig. 3.

   [86]Fig. 3
   [87]Open in a new tab

   Univariate (A) and multivariate (B) cox regression analyses of TPD52L2
   associated with overall survival in HNSCC patients from TCGA.
   Univariate (C) and multivariate (D) regression analysis of TPD52L2 from
   GEO database

Tpd52L2 co-expression gene analysis

   As shown in Fig. [88]4A, 4653 genes (dark red dots) showed significant
   positive correlation with TPD52L2, while 9518 genes (dark green dots)
   had significant negative correlation. The top fifty significant genes
   positively or negatively associated with TPD52L2 are shown by heat map
   in Fig. [89]4B, C. TPD52L2 expression had a strong positive
   relationship with the expression levels of UBE2V1 (r = 0.749), PSMA7
   (r = 0.689), and ARFGAP1 (r = 0.655). It also had a strong negative
   association with MYLIP (r = -0.621), BCL2 (r = -0.621), and REPIN1
   (r = -0.617).

Fig. 4.

   [90]Fig. 4
   [91]Open in a new tab

   Co-expression genes of TPD52L2 in HNSCC analyzed by LinkedOmics
   database. A All genes highly correlated with TPD52L2 as identified by
   Pearson correlation test in the HNSCC cohort. B, C Heat maps showing
   the top 50 genes positively and negatively correlated with TPD52L2 in
   HNSCC. Red indicates positively correlated genes, and blue indicates
   negatively correlated genes. D – G Significantly enriched GO
   annotations and KEGG pathways of TPD52L2 in HNSCC cohort

   The outcomes of the GO analysis showed that TPD52L2 co-expressed genes
   were involved mainly in cell junction organization, negative regulation
   of calcineurin-NFAT signaling cascade, focal adhesion, cell-substrate
   adherent junction, structural constituent of cytoskeleton
   (Fig. [92]4D-F). The KEGG pathway analysis showed enrichment in
   Phagosome, Gap junction (Fig. [93]4G).

Gene alteration of TPD52L2

   The genomic alteration frequency of TPD52L2 was higher than 1% in HNSCC
   patients with “amplification” as the primary alterations (Fig. [94]5A).
   In evaluating the CNV status of TPD52L2 in pan-cancer, we found that
   TPD52L2 expression was positively correlated with CNA in HNSCC
   (p < 0.001) (Fig. [95]5B).

Fig. 5.

   [96]Fig. 5
   [97]Open in a new tab

   Gene alteration of TPD52L2. A Mutation and CNA status of TPD52L2 in
   TCGA-pan-cancer using the cBioPortal database. B Correlation between
   TPD52L2 mRNA and CNA in TCGA-pan-cancer

Tpd52L2 genetic alteration status analysis

   As the prognosis of patients with high expression of TPD52L2 is worse
   than that of the patients with low expression, we speculated whether
   the expression of TPD52L2 is related to the mutations of some common
   cancer-promoting genes. A total of 518 samples were tested for
   mutations, of which 422(81.5%) were plotted. We further analyzed the
   gene mutations (such as TP53, PIK3CA, and FAM135B) in high and low
   TPD52L2 expression groups (Fig. [98]6A). These results demonstrated
   that the expression of TPD52L2 was potentially related to the mutation
   of TP53, PIK3CA, and FAM135B genes.

Fig. 6.

   [99]Fig. 6
   [100]Open in a new tab

   Gene mutant frequency (top 20) in high- and low-TPD52L2 expression
   groups. A total of 518 samples were tested for mutations, of which
   422(81.5%) were plotted. Gene (such as TP53, PIK3CA, and FAM135B)
   mutations in high and low TPD52L2 expression groups were analyzed using
   ComplexHeatmap in R from TCGA database

Correlation between TPD52L2 and immune infiltration

   An immune infiltration score for 22 immune cells was assigned for 39
   cancers based on their gene expression signatures which were downloaded
   from TCGA pan-cancer dataset. Data analysis using the “psych” package
   showed that TPD52L2 expression was significantly correlated with immune
   cell infiltrate in 38 cancers, including HNSCC (Fig. [101]7A).

Fig. 7.

   [102]Fig. 7
   [103]Open in a new tab

   Correlation Between TPD52L2 and Immune Infiltration. A Based on TCGA
   pan-cancer dataset, an immune infiltration score of 39 cancers for 22
   immune cells was analyzed. B TPD52L2 CNV status affected the
   infiltrating levels of B cells, CD8 + T cells, CD4 + T cells,
   Macrophages, neutrophils and dendritic cells. C The distribution of 24
   subtypes of immune cells in low- and high- TPD52L2 expression groups.
   Box diagram showing the stromal scores (D), immune scores (E), and
   ESTIMATE scores (F) in the low- and high- TPD52L2 groups analyzed by
   the ESTIMATE algorithm

   The TIMER analysis suggested that the CNV of TPD52L2 was significantly
   related to the infiltration levels of B cells, CD8 + cells,
   CD4 + cells, macrophages, neutrophils and dendritic cells
   (Fig. [104]7B). Additionally, the differences of 24 types of TIICs
   between the high-TPD52L2 expression and low- groups were compared. The
   results indicated that the high-expression group had fewer activated
   DCs (aDC) (p < 0.001), B cells (p < 0.001), CD8 T cells (p < 0.001).
   Other TIICs showed no statistically significant intergroup differences
   (Fig. [105]7C).

   The stromal scores (Fig. [106]7D), immune scores (Fig. [107]7E) and
   ESTIMATE scores (Fig. [108]7F) of high-TPD52L2 and low- group were
   performed by R package “ESTIMATE”. The ESTIMATE algorithm suggested
   that the high-TPD52L2 group had lower immune scores (p < 0.001) and
   ESTIMATE scores than the low- group.

   Correspondingly, we evaluated the correlations between immune responses
   and the expression of TPD52L2 by Spearman's correlation coefficient.
   There were 14 types of immune cells negatively correlated with TPD52L2
   expression, including aDCs (r =  − 0.177, p < 0.001), B cells
   (r =  − 0.361, p < 0.001), CD8^+ T cells (r =  − 0.270, p < 0.001),
   cytotoxic cells (r =  − 0.322, p < 0.001), DCs (r =  − 0.141,
   p = 0.002), Mast cells (r =  − 0.177, p < 0.001), NK CD56 bright cells
   (r =  − 0.188, p < 0.001), NK CD56 dim cells (r =  − 0.163, p < 0.001),
   pDC (r =  − 0.241, p < 0.001), T cells (r =  − 0.332, p < 0.001), T
   helper cells (r =  − 0.119, p = 0.008), TFH (r =  − 0.274, p < 0.001),
   Th17 cells (r =  − 0.194, p < 0.001), and Treg (r =  − 0.223,
   p < 0.001) (Fig. [109]7E).

Correlation between the expression of TPD52L2 and icgs in HNSCC

   In this study, the correlation between the expression of TPD52L2 and
   that of the over 40 common immune checkpoint genes was analyzed using
   spearman correlation coefficient. In HNSCC, TPD52L2 expression was
   positively correlated with 4 immune checkpoint markers, including
   PDCD1LG2, EGFR, CD44, CD274 (Fig. [110]8A).

Fig. 8.

   [111]Fig. 8
   [112]Open in a new tab

   Correlation between TPD52L2 expression and ICGs in HNSCC. A Results
   based on TCGA database. B Immunohistochemistry on the tissue
   microarrays displayed CD274, EGFR, PD-L2, TPD52L2 protein levels in
   normal tissues (n = 5) and HNSCC tissues (n = 5) (40*10 times), and
   representative images scored by Image-Pro Plus 6.0 software

   In addition, protein levels of TPD52L2 and three ICGs (PDL1, PDL2,
   EGFR) were further explored in our own clinical samples (n = 5 paired)
   based on IHC assay. In our samples, the protein level of TPD52L2 and
   three ICGs PDL1, PDL2, EGFR) was also significantly up-regulated in
   HNSCC samples compared with the paired normal tissues (p < 0.05)
   (Fig. [113]8B).

   These results demonstrated that TPD52L2 gene may play a crucial role in
   tumor immunity and may have impactful implications for the use of ICIs
   against HNSCC.

Cell line experiment verifies the overexpression of TPD52L2 in HNSCC

   Seven cell lines (five HNSCC cell lines (FD-LSC-1, Fadu, cal27, HN-6,
   SCC-25)) vs. two normal control cell lines (NOK, FM) were used to
   explore the mRNA and protein expression by qRT-PCR and WB. As shown in
   Fig. [114]9, the expression level of TPD52L2 in the specimens of HNSCC
   cell lines was significantly higher than that in normal cell lines. The
   WB assay result showed that the protein levels of these HNSCC cells
   were also significantly higher than in normal control cells
   (p < 0.001). The original blots are presented in Supplementary Fig. 1A,
   B.

Fig. 9.

   [115]Fig. 9
   [116]Open in a new tab

   The expression of TPD52L2 in vitro. A PCR analysis of the TPD52L2 mRNA
   levels in HNSCC cells and normal head and neck cells. B Western
   blotting analysis of the expression of TPD52L2 in HNSCC cells and
   normal head and neck cells. GAPDH was used as a loading control. The
   quantitative analysis result is shown in the lower panel. C TPD52L2
   protein expression in pan-cancer cohorts in TCGA by UALCAN

   To validate the TPD52L2 protein levels of normal and tumor tissues, the
   UALCAN website (http:// ualcan.path.uab.edu) was utilized. Based on
   CPTAC data, the protein expression of TPD52L2 was upregulated in HNSCC
   compared with normal controls (p < 0.001).

Discussion

   The tumor protein D52 (TPD52) family includes TPD52, -53, -54 (also
   known as TPD52L2) and -55 [[117]23], and they have been demonstrated to
   associated with the tumorigenesis and progression in several cancers.
   To exemplify, the overexpression of TPD52 in nonmalignant 3T3
   fibroblasts induced an increase in the growth speed and enabled the
   cells to exist in an anchorage-independent manner in vitro as well as
   to undergo metastatic growth in vivo [[118]24], and overexpression of
   TPD52 led to increased tumor growth in cancer cells [[119]25]. TPD53
   regulates the cell cycle and is highly upregulated at the G2-M phase
   transition [[120]26].

   In this study, we used data from TCGA database and UALCAN website to
   explore the expression of TPD52L2 and its clinical significance in
   HNSCC. We found that mRNA level and protein level of TPD52L2 (TPD54) in
   HNSCC tissues was higher than that in normal tissues. Next, qRT-PCR and
   WB verified the previous bioinformatics analysis results. ROC analysis
   also showed the overexpression of TPD52L2 can be used to distinguish
   between HNSCC and normal tissues. In addition, we found that the
   patients with high expression of TPD52L2 showed a much worse prognosis.
   Our cox regression analysis demonstrated that the expression level of
   TPD52L2 and age, N stage and M stage were independent factors for the
   prognosis of HNSCC patients.

   In pan-cancer analysis, we determined the high expression of TPD52L2 in
   pan-cancer and the KM analysis showed that TPD52L2 was significantly
   correlated with OS prognostic indicators of BLCA, MESO, LGG, LUADLUSC,
   LIHC and BRCA. Recent research has also shown that TPD52L2(TPD54) was
   upregulated in a variety of cancer tissues, and its expression level
   was associated with tumor progression and poor prognosis in prostate
   cancer [[121]16], basal breast cancer [[122]27] and lung adenocarcinoma
   [[123]14]. Pan et al. [[124]17] found that depletion of TPD52L2 could
   remarkably inhibit proliferative and colony-forming ability of liver
   cancer cells (SMMC-7721).

   To better understand the potential mechanism and function of TPD52L2 in
   the development of HNSCC, we conducted GO and KEGG enrichment analysis
   of TPD52L2 co-expression genes. Interestingly, the results showed that
   cell senescence and immune-related pathways were significantly
   enriched, e.g., calcineurin-NFAT (nuclear factor of activated T cells)
   signal cascade negative regulation pathway and calcineurin-mediated
   signal transduction negative regulation pathway. Previous studies have
   shown that inhibition of calcineurin-NFAT can inhibit cancer cell
   senescence, downregulate p53 expression, and promote the occurrence of
   skin squamous cell carcinoma [[125]28]. Cell aging is of great
   significance to the life activities of human body, including the
   regulation of embryonic development, tissue repair and tumor
   inhibition. And the loss of ‘Cell Aging’ characteristics of cells is
   related to the occurrence and progression of tumors. Our results
   suggest that TPD52L2 may affect the aging process of cells by
   regulating the negative cascade of calcineurin-NFAT signal cascade and
   play an important role in the occurrence of cancer.

   At the same time, our gene enrichment results also found that cell
   migration-related pathways such as focal adhesion and ER to Golgi
   vesicle-mediated transport were significantly enriched. Larocque
   [[126]13] first found that TPD52L2 is involved in a variety of membrane
   transport pathways: anterograde transport, recycling and Golgi
   integrity. Another study [[127]29] found that the only known marker of
   intracellular nano-vesicles (INVs, diameter 30 nm) is TPD52L2, and
   TPD52L2 binds directly to the cell membrane and binds to INVs through
   its C-terminal conserved positively charged motif, which researchers
   point out may be related to cancer cell migration and invasion.
   Therefore, TPD52L2 is likely to use its own INVs to transport goods to
   help achieve tumor cell migration. It is well known that tumor cell
   migration plays an important role in the development of cancer, and
   focal adhesion kinase (FAK) [[128]30] has been proved to be involved
   not only in tumor proliferation, but also in many processes such as
   tumor cell migration and angiogenesis, which eventually leads to
   further metastasis and deterioration of tumor. In general, our
   enrichment analysis results suggest that TPD52L2 may affect the
   progression of HNSCC and the prognosis of patients by inhibiting
   immunity and promoting tumor cell migration.

   Apart from that, our study revealed that the mutation frequency of TP53
   in 442 HNSCC samples was 77.4% and it was significantly higher in
   high-TPD52L2 group than the low- TPD52L2 group. Mutations in the TP53
   gene can lead to loss of tumor suppressor activity of its encoded
   protein p53, which is seen as a driver of a variety of tumorigenesis.
   TP53 mutations often occur in cancers and are associated with poor
   prognosis in many cancers [[129]31].

   Growing evidence showed that tumor microenvironment (TME) is important
   for tumor pathogenesis, where tumor cells usually establish tumor
   microenvironment to enrich kinds of properties [[130]32, [131]33].
   Immune-related cells and stromal cells are the main components of TME.
   We found that overexpression of TPD52L2 and activation of dendritic
   cells (p < 0.001), B cells (p < 0.001), CD8^+ T cells (p < 0.001) were
   negatively correlated. There were also differences in immune score and
   ESTIMATE score between TPD52L2 high expression group and low expression
   group, and the two scores in high expression group were significantly
   lower than those in TPD52L2 low expression group (p < 0.001). These
   results suggest that the overexpression of TPD52L2 may affect the
   progression and prognosis of HNSCC by regulating the level of immune
   cell infiltration and inhibiting immune anti-tumor function.

   As a new treatment method, immunotherapy is based on a variety of ways
   to activate the immune system in order to achieve the purpose of cancer
   treatment, which has dramatically improved the outlook of cancer
   treatment [[132]34]. Now many literatures have reported the clinical
   application with the immune checkpoint inhibitor in patients with
   locally advanced, HPV-unrelated head and neck cancer [[133]35].
   Therefore, we explored the correlation between TPD52L2 expression and
   immune checkpoint molecules by TCGA database and IHC. We observed that
   the expression of TPD52L2 was positively correlated with four immune
   checkpoint molecules, including PDCD1LG2, EGFR, CD44 and CD274. The
   results show that TPD52L2 may play a key role in tumor immunity and may
   play an important role in the treatment of HNSCC with ICIs.

   However, it is important to acknowledge the limitations of our study.
   Firstly, our research primarily relied on bioinformatic analyses and
   experiments conducted on cell lines, lacking in vivo and in vitro
   functional experiment verification. A previous study reported that
   their in vitro experiments suggested that TPD52L2 has negative effects
   on cell proliferation and migration in OSCC cells, while this result
   was not confirmed by in vivo experiments [[134]36]. Although our study
   confirmed the elevated expression of TPD52L2 mRNA and protein in a
   variety of head and neck squamous cell carcinomas (including OSCC), no
   further in vitro and in vivo experiments were conducted. It is worth
   noting that in vivo experiments can be influenced by specific cytokines
   that interact with TPD52L2, thereby potentially contributing to the
   variability in cellular backgrounds. Subsequently, we intend to conduct
   in vitro functional experiments and develop a TPD52L2-deletion mouse
   model to substantiate the role of TPD52L2 in regulating tumor
   proliferation and metastasis. This verification is crucial for
   advancing our results towards potential clinical applications.

Conclusion

   Together, our study confirmed that TPD52L2 is elevated in HNSCC and it
   could act as a reliable independent prognostic factor. In addition,
   TPD52L2 was associated with TIICs and immune checkpoints. Our results
   suggest that TPD52L2 may serve as a prognostic biomarker and it may
   provide a new therapeutic option for patients with HNSCC.

Supplementary Information

   [135]Supplementary Material 1.^ (159.9KB, pdf)

Acknowledgements