3.1 Analysis of PIK3CD expression across tumour types

To investigate the potential functions of PIK3CD in carcinogenesis, we first examined its expression in 24 different kinds of human tumors. As shown in Fig. 1, PIK3CD was significantly elevated in 11 tumor types, including ESCA, HNSC, KIRC, KIRP, LAML, PAAD, SKCM, STAD, CHOL, STES, and TGCT, and significantly downregulated in 11 tumor types, including BLCA, BRCA, COAD, KICH, LIHC, LUAD, LUSC, OV, PRAD, THCA and UCEC. However, no significant difference in PIK3CD was seen in GBMLGG or WT.

Fig. 1

Expression analysis for PIK3CD in multiple tumors. Combining data from the TCGA and GTEx databases, we examined the levels of PIK3CD mRNA expression in 24 different types of human tumors and normal tissues. *p value < 0.05; **p value < 0.01; ***p value < 0.001

3.2 PIK3CD prognostic values in human tumors.

The survival analysis for PIK3CD was then performed in BLCA,BRCA, COAD, KICH,LIHC,LUAD,LUSC,OV,PRAD,THCA,ESCA,HNSC,KIRC,KIRP,LAML, PAAD, SKCM,STAD,CHOL, and TGCT. High levels of PIK3CD in LUAD and BRCA were associated with a good prognosis for OS (Fig. 2A–D). Just elevated levels of PIK3CD in PRAD and LUSC patients suggested a poor prognosis for RFS, but higher expression of PIK3CD in CHOL and LIHC patients showed a better prognosis. There was no statistical relevance of PIK3CD in predicting patient outcome in other kinds of cancer (Fig. 2E–H). In patients with BRCA, PIK3CD could be used as a predictive biomarker.

Fig. 2

GEPIA assessed the overall survival (OS/RFS) for PIK3CD in diverse human cancers (A–H). The OS/RFS plot of PIK3CD in BRCA (A), COAD (B), LUAD (C), STAD (D), CHOL (E), LUSC (F), PRAD (G) and LIHC (H)

3.3 Prediction and evaluation of PIK3CD's upstream miRNAs

MiRNAs are widely known to be responsible for gene expression regulation. To determine whether PIK3CD was controlled by miRNAs, we predicted upstream miRNAs that could potentially bind to PIK3CD and eventually discovered 6 miRNAs. Based on the competing endogenous RNA (ceRNA) hypothesis, miRNA regulates target gene expression and there should be an adverse correlation between miRNA and PIK3CD.As a result, the expression correlation analysis was carried out. In BRCA, PIK3CD was significantly negatively linked with hsa-miR-30b-5p(|R|= 0.169) and favorably correlated with hsa-miR-30e-5p and hsa-miR-7-5p, as shown in Table 1. PIK3CD and hsa-miR-7-5p have no statistically significant expression correlations. Finally, hsa-miR-30b-5p expression and prognostic significance in BRCA were determined. As shown in Fig. 3A and B, hsa-miR-30b-5p was significantly downregulated in BRCA patients, and its overexpression was associated with a better prognosis. All of these findings point to miRNA-30b-5p as the most likely regulating miRNA of PIK3CD in BRCA.

Table 1 The starBase database examined the expression connection between predicted miRNAs and PIK3CD in BRCAFig. 3

miRNA-30b-5p was markedly downregulated (A) in BRCA and miRNA-30b level was positively linked to patients’ prognosis (B)

3.4 Prediction and analysis of upstream circRNAs of hsa-miR-30b-5p

Circular RNAs (circRNAs) have been linked to the progression of cancer [4]. We used bioinformatics to analyze and identify the circRNAs that might probably mediate PIK3CD expression in BRCA.The upstream circRNAs of hsa-miR-30b-5p were then predicted utilizing the starBase database. We investigated the top ten circRNAs. The levels of expression of these circRNAs were subsequently shown to be negatively related to the levels of hsa-miR-30b-5p, which were found to be negatively related to the expression of PIK3CD in BRCA. All of these circRNAs’ expression was linked to PIK3CD (Tables 2 and 3).

Table 2 Correlation analysis between CircRNA and miR-30b-5p in BRCA determined by starBase databaseTable 3 Correlation analysis between circRNAs and PIK3CD in BRCA determined by starBase database

The starBase database was then used to analyze the amounts of expression of these circRNAs in BRCA. Only CD99 was not significantly regulated in BRCA as compared to normal controls, as seen in Fig. 4. The prognostic values of the ten circRNAs in BRCA were then assessed. BRCA patients with higher levels of CFH, GLI2, RAB32, LAMB1, MGAT2, CD99 and ITGA8 expression had a better prognosis,while those of COL6A3 and PRRX1 had a worse one, as illustrated in Fig. 5.

Fig. 4

Expression analysis for upstream circRNAs of miR-30b-5p in BRCA.The expression of CD99 (A), CFH (B), COL6A3 (C), GLI2 (D), ITGA5 (E), ITGA8 (F), LAMB1 (G), MGAT2 (H), PRRX1 (I) and RAB32 (J) in “TCGA BRCA” compared with “TCGA normal” or “TCGA and GTEx normal” data

Fig. 5

Survival analysis for upstream circRNAs of miR-30b-5p in BRCA. In BRCA the OS analysis for CD99 (A), CFH (B), COL6A3 (C), GLI2 (D), ITGA8 (E), LAMB1 (F), MGAT2 (G), PRRX1 (H) and RAB32 (I) and ITGA5 (J) in BRCA

Taking expression, survival, and correlation studies into account, CHF, GLI2, RAB32, LAMB1, MGAT2, ITGA8, COL6A3 and PRRX1 could be the eight most likely upstream circRNAs of the miR-30b-5p/PIK3CD axis in BRCA.

3.5 PIK3CD correlates with EMT, migration and invasion in BRCA

The link between PIK3CD and EMT variables was then investigated. In the majority of the 5 cancer types, PIK3CD expression was linked favorably with the expression of mesenchymal markers such as VIM (Vimentin), TWIST1, SNAI1 (SNAIL), SNAI2 (SLUG), and CDH2 (Fig. 6A). Furthermore, PIK3CD was found to be strongly inversely related to the epithelial marker CDH1 (E-Cadherin). The correlation heatmaps revealed a strong relationship between PIK3CD and the EMT variables. These findings showed that PIK3CD may be involved in EMT.

Fig. 6

PIK3CD correlates with EMT in BRCA. A Correlation of PIK3CD expression with expression of EMT biomarkers across 5 cancer types. B The levels of PIK3CD mRNA in BRCA cell lines were detected by using qRT-PCR. a MDA-MB-231 cells. b MCF-7 cells. C Upregulation of PIK3CD increased MDA-MB-231 BRCA cell migration and invasion. Transwell assay of cell migration ability following PIK3CD cDNA treatment. Transwell assessment of cell invasion capabilities following PIK3CD cDNA transfection (a and b). (c) Western blot examination of EMT-related marker expression alterations following PIK3CD cDNA transfection. The statistical analysis was displayed. D Upregulation of PIK3CD enhanced MCF-7 BRCA cell motility and invasion. Transwell experiment of cell migration ability following PIK3CD cDNA treatment. Transwell test of cell invasion capabilities following PIK3CD cDNA transfection (a and b). (c) Western blot study of EMT-related marker expression following PIK3CD cDNA transfection. The statistical analysis was displayed

We studied the impact of PIK3CD on cell migration, invasion, and EMT of BRCA cells after verifying its expression and prognostic value for BRCA. MCF-7 BRCA cell lines were transfected with PIK3CD cDNA. To evaluate the effect of PIK3CD on cell migration and invasion, Transwell Assay was utilized. BRCA cells that received PIK3CD cDNA displayed higher migration and invasion abilities (Fig. 6B–D). After PIK3CD cDNA transfection, the expression of EMT-related markers was evaluated using Western blot. E-cadherin expression was downregulated in the PIK3CD cDNA group, while VIM and SNAI1 expression were elevated.

These findings demonstrated that PIK3CD cDNA increased cell EMT, migration, and invasion, indicating that PIK3CD plays an active role in BRCA cell activities.

3.6 PIK3CD correlates with influx of immunological cells in BRCA

The TIMER database was used to demonstrate the link between PIK3CD expression and immune cell infiltration levels in BRCA cells. The amount of immune cell infiltration significantly changed with different PIK3CD copy counts in BRCA, as illustrated in Fig. 7A. Somatic copy number alterations (SCNA) of PIK3CD were closely correlated with the abundance of immune infiltration. Consequently, the relationship between PIK3CD expression and immune cell infiltration was assessed. According to Fig. 7B, the expression of PIK3CD was highly positively linked with every immunological cell type that was investigated, including B cells,CD4 + T cells, CD8 + T cells,neutrophils, macrophages and dendritic cells in BRCA.

Fig. 7

The association between immune cell infiltration and PIK3CD levels in BRCA. A The level of invasion of various immune cells in BRCA with varying copy levels of PIK3CD from the TIMER database. B The level of PIK3CD expression in BRCA is correlated with B cell (b), CD8 + T cell (c), CD4 + T cell (d), macrophage (e), neutrophil (f), or dendriticcell (g) infiltration. C Correlation of PIK3CD with immune cells 5 cancer types.The CIBERSORT technique was used in this work to estimate the fraction of 22 immune cells

Clinical treatment sensitivity and disease diagnosis are significantly impacted by the microenvironment, which includes immune cells, extracellular matrix, inflammatory chemicals, and other growth factors. Using the CIBERSORT technique, the fraction of 22 immune cells was estimated using a histogram in this study. The results showed that the tumor group had considerably larger proportions of most immune cells than the control groups, with the exception of some immune cells such as CD56 dim natural killer cells in BRCA,LUAD, and STAD (Fig. 7C).

3.7 Correlation of PIK3CD expression and immune cell indicators in BRCA

The GEPIA database was used to examine the expression connection of PIK3CD with immune cell indicators in BRCA, able to better investigate the function of PIK3CD in tumor immunity. According to the findings, PIK3CD expression was strongly positively linked with immunity-associated gene marker expression, such as biomarkers of B cells including CD19 and CD79A, and biomarkers of dendritic cells in BRCA (Table 4, *p value < 0.05; **p value < 0.01; ***p value < 0.001). These findings revealed that increased PIK3CD levels were related to increase immunological infiltration as well as immune exhaustion, suggesting that PIK3CD could be a potential immunotherapeutic target for BRCA treatment. These results help to some extent establish the favorable relationship between PIK3CD and immune cell infiltration.

Table 4 Correlation analysis between PIK3CD and biomarkers of immune cells in BRCA determined by GEPIA database3.8 Link between PIK3CD and immunological checkpoints in BRCA

Immune checkpoints like PD1/PD-L1 and CTLA-4 are essential for tumor immunity escape. Given the putative oncogenic involvement of PIK3CD in BRCA, the association of PIK3CD with PD1, PD-L1,or CTLA-4 was investigated.As demonstrated in Fig. 8A, PIK3CD expression was strongly positively linked with PD1, PD-L1 and CTLA-4 in BRCA,which was modified by purity. We also found that PD1, PD-L1, CTLA-4 and PIK3CD in BRCA had a highly significant relationship, similar to what we found in the analysis of the TIMER data.Our findings show that PIK3CD-mediated tumorigenesis of BRCA may entail tumor immune evasion.

Fig. 8

Relationship between PIK3CD and immune checkpoints/promoter methylation level of PIK3CD in BRCA. A Correlation of PIK3CD expression in BRCA with PD-1, PD-L1, and CTLA-4 expression. (a) TIMER-adjusted Spearman association of PIK3CD with PD-1 expression in BRCA. (b) The GEPIA database determined the expression connection of PIK3CD with PD1 in BRCA. (c) PIK3CD-PD-L1 TIMER-adjusted Spearman correlation in BRCA. (d) The GEPIA database determined the expression connection of PIK3CD with PD-L1 in BRCA. (e) TIMER Spearman association of PIK3CD with CTLA-4 expression in BRCA corrected for purity. (f) The GEPIA database determined the expression connection of PIK3CD with CTLA-4 in BRCA. B The correlations between PIK3CD expression, methylation level, and Tumor-infiltrating lymphocyte (TIL) quantity. (a) Relationships between 28 different types of tumor-infiltrating lymphocytes (TILs) and PIK3CD expression in several tumors. (b) Correlations between 28 different types of TILs and PIK3CD methylation in several tumors. (c) Correlations between the promoter methylation level of PIK3CD and the clinicopathological traits of BRCA patients. UALCAN database was used to examine the relative promoter methylation level of PIK3CD based on major subclasses

3.9 Relationships between PIK3CD expression, methylation level and the amount of lymphocytes that infiltrate tumors

We discovered that PIK3CD expression had a positive relationship with the abundance of 28 different types of tumor-infiltrating lymphocytes in the TISIDB database (Fig. 8B, Table 5). We investigated the association between PIK3CD methylation and tumor-infiltrating lymphocyte abundance and discovered that it was strongly inversely linked with tumor-infiltrating lymphocyte abundance (Fig. 8B, Table 5). The results indicated that PIK3CD level was favorably related to the high number of lymphocytes infiltrating tumors, but PIK3CD methylation was negatively associated, and it might be used as one of the immunotherapeutic targets for BRCA therapy.

Table 5 Associations between PIK3CD expression/methylation level and TIL abundance

We also discovered links between PIK3CD methylation and clinicopathological characteristics in BRCA patients. In TNBC and HER2-Positive breast cancer tissues, the promoter methylation level of PIK3CD was lower than in Luminal (Fig. 8B). Thus, reduced PIK3CD methylation could be an underlying indication reflecting clinical BRCA features.

3.10 Identification of the m1A/m5C/m6A regulators relevant to PIK3CD in BRCA

The pan-cancer correlations between PIK3CD and the expression levels of 40 kinds of the m1A/m5C/m6A regulators in 5 cancer types were shown in Fig. 9A. There was a strong association (P < 0.05) between the expression levels of the 41 different m1A/m5C/m6A regulators and PIK3CD in 5 different cancer types. Furthermore,in BRCA and LUAD, significant co-expression of PIK3CD with most regulator genes was detected.In LIHC, expression of PIK3CD was negatively correlated with some molecules such as ZC3H13, YTHDF3 and TRMT61A, though for some of them not to a significant degree. It is also interesting that expression of PIK3CD was negatively correlated with TRMT61B, FMR1, FTO and METTL14 in LIHC,but the levels of PIK3CD were positively associated with them in BRCA (Fig. 9A).

Fig. 9

The m1A/m5C/m6A regulators relevant to PIK3CD in BRCA. A Relations between the levels of PIK3CD and the expressions of 41 kinds of the m1A/m5C/m6A regulators in various cancers. B The m1A/m5C/m6A regulator expression variations between tumor and normal samples. The vioplot shows the top 10 genes (|R|) that are significantly differentially expressed in breast tumors and normal tissues. C The overall survival (OS) analysis for the top 10 m1A/m5C/m6A regulators significantly differentially expressed (|R|) in both breast tumor and normal samples determined by Kaplan–Meier Plotter

In BRCA patients, DNMT1, NOP2, DNMT3A, FMR1, RBM15 and YTHDF2 were highly elevated, while WTAP, TET2, ALKBH3 and TRDMT1 were dramatically downregulated (Fig. 9B). For OS, high expression of ALKBH3, DNMT1, DNMT3A, FMR1, NOP2, TET2, acted as high-risk factors in BRCA and had unfavorable prognosis, but RBM15, TRDMT1 and WTAP indicated better prognosis (Fig. 9C).