SLC4A4 promotes prostate cancer progression in vivo and in vitro via AKT-mediated signalling pathway

Tumour tissues and adjacent paired non-tumour tissues were gathered from patients who were diagnosed with PCa and underwent surgical excision in Renmin Hospital of Wuhan University (Wuhan, China) between June 2018 and January 2020. PCa and normal prostate tissues were gathered from 74 patients; the age range of the patients was between 26 and 87(mean age, 65 years).

Immunohistochemistry (IHC) was used to detect the expression of SLC4A4 in these tissues. Paraffin-embedded sections were dewaxed, the antigen was retrieved, and then the paraffin sections were incubated using primary antibody against SLC4A4 (cat. No. bs-21660R; Bioss) (1:200) and secondary antibody goat anti-rabbit (cat. No. A0208; Beyotime) (1:400). After staining, ten fields (× 100 magnification) were chosen to be captured and analyzed with the optical microscope (Olympus, Japan) for each section. The SLC4A4 staining intensity was scored on a range from 0 (negative), 1 (weak), 2 (positive + +) and 3 (positive +  + +). Median IHC score was used to distinguish the high/low expression of SLC4A4.

Human PCa cell lines DU 145, LNCaP and PC-3 were purchased from the Cell Bank of Chinese Academy of Sciences and cultured in RPMI-1640 medium at 37 °C in a humidified incubator containing 5% CO₂. The cell culture media were supplemented with 10% fetal bovine serum (FBS) and 1% sodium penicillin G/streptomycin sulfate (P/S). The normal prostate epithelial cell line RWPE-1 was purchased from the Cell Bank of Chinese Academy of Sciences and cultivated in keratinocyte serum-free medium (K-SFM) containing 0.05 mg/mL bovine pituitary extract (BPE), 5 ng/mL epidermal growth factor (EGF) and 1% P/S.

The modified BR-shRNA plasmid and pMD2.G and pSPAX2 helper plasmids were transfected three times with Lipofectamine 2000 into HEK-293 T cells to obtain lentiviruses. Next, the lentiviral particles were gathered, filtered, and preserved. The knockdown efficiency of SLC4A4 was evaluated by RT-qPCR and Western blotting.

PCa cells were respectively transfected with 1 × 10⁷ TU/ml of the lentivirus containing shRNA interfering with SLC4A4 (shSLC4A4) or shRNA for negative control (shCtrl), and the cells were then incubated at 37 °C for three days. The expression of green fluorescent protein (GFP, carried by the lentiviral vector) was observed by fluorescent microscope (EMD Millipore), and the ratio of fluorescent cells to total cells (viewed in white light) was used to evaluate the transfection efficiency.

PCa cells were gathered and centrifuged. Total RNA was extracted with the TRIzol reagent. Complementary DNAs (cDNAs) was synthesized with the PrimeScript™ RT reagent Kit (Takara). The qPCR reaction system was prepared using a real-time quantitative PCR instrument according to product specification. The reaction system was composed of the following reagents: TB Green® Premix Ex Taq™ II, Forward and Reverse primers (Sangon Biotech), reverse transcription products and RNase-free H₂O. The thermal cycling conditions were: pre-denaturation at 95 °C for 30 s, denaturation at 95 °C for 15 s, annealing at 60 °C for 10 s for a total of 42 cycles, and 72 °C for 5 min (extension). GAPDH was used as an internal reference. The relative expression levels of genes were calculated as the method of the 2^−ΔΔCt [17]. The sequences of the main primers are as follows: GAPDH forward, 5'-TGACTTCAACAGCGACACCCA-3' and reverse, 5'-CACCCTGTTGCTGTAGCCAAA-3'; SLC4A4 forward, 5'-AAGCTCTTTCGGCAATTCTCTTC-3' and reverse, 5'-GAAACTCTCCAACACGCCCTG-3'.

PCa cells were washed twice with ice-cold PBS, lysed with cell lysis buffer (Beyotime) containing protease inhibitors, and incubated on ice for 15 min. The supernatant was harvested by centrifugation, and the protein content was measured with the BCA Protein Assay Kit (cat. No. 23225; HyClone-Pierce). Equal amounts of proteins were separated and transferred onto polyvinylidene difluoride (PVDF) membranes (0.45 μM). Subsequently, the membranes were blocked with TBS + 0.1% Tween-20 (TBST) buffer containing 5% skim milk and incubated with various primary antibodies. Following being washed with TBST, the membranes were blotted with HRP-conjugated secondary antibodies. The target bands were visualized with immobilon Western Chemiluminescent HRP Substrote (cat. No. RPN2232; Millipore). GAPDH was used as an internal reference.

The antibodies used in this experimental study were BAX (cat. No. 50599–2-Ig; Proteintech), BCL-2 (cat. No. sc-7382; santa cruz), CDK4 (cat. No. 11026–1-AP; Proteintech), FAS (cat. No. 13098–1-AP; Proteintech), AKT (cat. No. 10176–2-AP; Proteintech), p-AKT (cat. No. 66444–1-Ig; Proteintech), GAPDH (cat. No. 60004–1-lg; Proteintech), horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG (cat. No. A0208; Beyotime), goat anti-mouse (cat. No. A0216; Beyotime), donkey anti-goat (cat. No. A0181; Beyotime).

Following infection with shRNA lentivirus, PCa cells were trypsinized, resuspended, counted and then inoculated in 96-well plates. Each group underwent a minimum of 3 duplicate wells. From the second day, the plates were tested by Celigo Imaging Cytometry System once a day for 5 consecutive days. The number of green fluorescent cells in each scan plate was precisely calculated by adjusting the input parameters of the analysis setup; the data were then plotted statistically, and the cell proliferation curve for 5 days was drawn.

CCK-8 (Dojindo, Shanghai) assay was used to determine the number of living cells. The cell suspension was inoculated with 2 × 10³ cells per well into a 96-well plate and pre-incubated for 24 h after transfection. Treated or un-treated cells were cultured as appropriate. 10 µl of CCK-8 solution was added to each well, and the cells were incubated for 2 h at 37 °C. The absorbance at 450 nm was recorded with a TECAN infinite M200 Multimode microplate reader (Tecan, Mechelen, Belgium). All detections were conducted in triplicate.

Apoptosis was detected by the flow cytometry analysis. The infected PCa cells were cultured until the cell density reached 85%. The cells were trypsinized and resuspended, centrifuged for 5 min. The supernatant was discarded, and cell precipitates were washed with D‑Hank's solution (pH = 7.2–7.4) precooled at 4 °C. The cells were washed with 1 × binding buffer, centrifuged, and resuspended. The cell suspensions (1 × 10⁵–1 × 10⁶ cells) were stained by 10 µl Annexin V-APC and protected from light for 15 min [18]. Subsequently, 400–800 µl of 1 × binding buffer was added depending on the amount of cells. At last, the cells were tested by Guava easyCyte HT flow cytometer and analysed with FlowJo VX10.

Cell cycle distribution was analysed with the flow cytometry. When 6 cm dish cells in each experimental group grew to about 80% coverage, the cells were fixed at least 1 h after the washing. Afterwards, cells were washed and resuspended in PBS containing PI and RNase A. Finally, the cells were tested by flow cytometer, and the percentage of the cells in G0/G1, S and G2/M phases were visualized by a ModFit. All detections were performed in triplicate.

The aim of this assay was to assess the migration ability of cells after transfection. When the cells were dense in the microscopic field, three standardized wounds per well were scratched with the tip of a sterile pipette. Then, the cells were cultured in serum-free medium. Wound sizes were photographed with a phase-contrast microscope at 0 and 24 h, respectively. The five randomly selected fields were adopted to calculate the rate of wound healing using ImageJ software.

Diluted Matrigel (Corning, USA) was added to transwell upper chambers 12 h prior to the experiment and placed at 37 °C for solidification. The PCa cells (1.0 × 10⁵ cells per chamber) were inoculated in the upper chambers by resuspension with basal serum-free medium, and the lower chambers was added with medium containing 20% FBS to attract cells to penetrate the membrane. Following incubation for 24 h, those cells on the outside surface of the chambers, having invaded through the membrane, were fixed using 4% paraformaldehyde (500 µl per chamber) for 20 min and stained using 0.1% crystal violet for 20 min. Cells on the upper surface and remaining Matrigel were wiped off by cotton swabs. At last, the numbers of stained cells in five different views per chamber were counted.

To study in vivo tumor growth, four-week-old BALB/c nude mice, weighing 18–20 g, were purchased from Beijing HFK Bioscience Co. Ltd. All 20 mice were placed in SPF housing conditions.

After a week of acclimatization in Animal Experiment Center of Renmin Hospital of Wuhan University, 20 mice were randomly divided into the control group (shCtrl) and the test group (shSLC4A4) (n = 10 mice/group). Since LNCaP was an androgen-dependent cell line, the tumorigenic rate of subcutaneous inoculation in nude mice alone was very poor [19]. DU 145, on the other hand, was an androgen-independent line with low differentiation and better tumorigenic effect [20, 21]. The shRNA lentivirus-infected DU 145 cells were digested, suspended and injected subcutaneously into right forelimb axilla of each mouse (serum-free medium containing 4 × 10⁶ cells). Those 20 mice were reared for 31 days, during which length and width of the tumors in mice were measured five times using the Vernier caliper. On day 31, the mice were injected intraperitoneally with D-Luciferin, anesthetized with an intraperitoneal injection of 0.7% pentobarbital sodium and placed under the animal multispectral living imaging system for imaging. Next, the mice were sacrificed with cervical dislocation, and tumors autopsied from the mice were weighed and photographed. Tumor volume in mm³ (V) was calculated based on the formula: V = π/6 × L × W × W, where L represents length and W represents width of the tumor.

Paraffin sections of tumor tissues taken from mice were dewaxed, rehydrated in a decreasing ethanol gradient, and incubated with the anti-Ki-67 antibody (cat. No. ab16667; Abcam). After washing with PBS, the paraffin sections were incubated with secondary antibody goat anti-rabbit (cat. No. ab97080; Abcam), counterstained with hematoxylin, and Ki‑67 expression was observed under an optical microscope. Ten fields of each section were captured for analysis, and this experiment was repeated three times.

To explore the potential downstream signal pathways and functional targets of SLC4A4 in PCa, Human Phospho-Kinase Array Kit (cat. no. ARY003C; Bio-Techne China Co., Ltd.) was employed. PCa cells transfected with shCtrl or shSLC4A4 were lysed. Meanwhile, 8 nitrocellulose membranes (4 Part A, 4 Part B, each containing 39 different capture antibodies printed in duplicate) were closed with 2 ml of Array Buffer 1 (block buffer) for 1 h. Then, the samples were piped to wells and incubated overnight. After washing, 1 × biotinylated antibody cocktail was added to each well and incubated. Afterwards, diluted Streptavidin-HRP was pipetted into each well and incubated. Followed by the washing of membranes, any excess Wash Buffer was blotted off, and 500 µl of Chemi Reagent Mix (equal vol. of Chemi Reagent 1 and 2) was added to each membrane. In the end, the signal density was measured with the chemiluminescence imaging system and analysed with the ImageJ. This experiment was performed in duplicate.

SPSS 23.0 and Graphpad Prism 8 were used for data analysis. The quantitative data were presented as the mean ± standard deviation (SD). Chi-squared tests were performed to compare the differences in SLC4A4 expression among PCa patients. Spearman rank correlation analysis was used to analyse the correlation between SLC4A4 expression and clinicopathological features. The histograms of SLC4A4-related signal molecules in carcinoma cell were plotted by SignaLink 2.0 analysis. P-value < 0.05 was considered a statistically significant difference.

For the purpose of determining the effect of SLC4A4 on the development of PCa, the expression of SLC4A4 in clinical PCa and normal prostate tissues was examined by IHC. As demonstrated in Fig. 1A, the results of 192 pathological sections confirmed the cytoplasmic localization of SLC4A4 and indicated that the expression of SLC4A4 in PCa tissues was distinctly higher than that in normal prostate tissues (P < 0.001; Table 1), which can be used for follow-up statistical analysis of clinicopathological data.

The relationship between the levels of SLC4A4 expression and the clinicopathological characteristics in PCa patients was assessed by statistical analysis. The high/low SLC4A4 groups were divided depending on median of IHC scores of all tissue specimens. The results demonstrated that the levels of SLC4A4 expression were remarkably different between patients at different clinical stages, T Infiltrate and lymphatic metastasis (P < 0.05; Table 2). According to Spearman rank correlation analysis, the clinical stage, T Infiltrate and the risk of lymphatic metastasis were positively correlated with high SLC4A4 expression (Table 3). These results suggested that the expression of SLC4A4 was increasing with the deepening of tumour malignancy.

The results of qRT-PCR verified that SLC4A4 mRNA expression was significantly higher in the PCa cell lines than in the normal prostate epithelial cell line RWPE-1 (Fig. 1B). For investigating the roles of SLC4A4 in PCa, SLC4A4-targeting shRNA was cloned into GFP-carrying lentiviral vector. Afterwards, shSLC4A4 or shCtrl lentivirus was transfected into human PCa cell lines. Where the detailed plots of DU 145 and LNCaP are presented in Fig. 1C. The fluorescent signal inside cells, which have been infected with shCtrl or shSLC4A4 for 72 h, observed by the microscope, reveal a > 80% transfection efficiency in both cell lines. The knockdown efficiency of SLC4A4 was examined using RT-qPCR on mRNA level. The results demonstrated that, compared with shCtrl group, the knockdown efficiency of SLC4A4 in shSLC4A4 group was 59.24% (P < 0.001) in DU 145 cells; the knockdown efficiency of SLC4A4 in shSLC4A4 group was 82.79% in LNCaP cells (P < 0.01; Fig. 1D). Furthermore, results of western blotting also indicated that the expression of SLC4A4 protein was distinctly down-regulated after infection in comparison with shCtrl cells (Fig. 1E). These results implied that the SLC4A4-knockdown cell models were successfully constructed.

To observe the effect of SLC4A4 on cell proliferation, the growth curves of PCa cells in 5 days were depicted by Celigo Imaging Cytometry System. This is presented in Fig. 2A, B, cell proliferation is obviously suppressed in shSLC4A4 group in comparison with shCtrl group. In both DU 145 and LNCaP cells, the cells of shSLC4A4 group exhibited a slower proliferation rate (P < 0.001, fold change = − 4 and − 6.5, respectively). Flow cytometry analysis was used for detecting the effect of SLC4A4 knockdown on PCa cells apoptosis. Compared with the shCtrl group, the percentage of apoptotic cells in shSLC4A4 group was increased by 2.9-fold in DU 145 cells and 9.4-fold in LNCaP cells (P < 0.001; Fig. 3A, B), implying that SLC4A4 knockdown facilitated apoptosis among PCa cells. Furthermore, it was found that, after SLC4A4 knockdown, more cells were stalled in G2 phase and fewer in G1 and S phases than those transfected with shCtrl (Fig. 3C, D). Collectively, knockdown of SLC4A4 inhibited PCa development in vitro.

Scratch test detected that, after being transfected with the corresponding lentivirus, in contrast to shCtrl group, the migration rate of cells in shSLC4A4 group (24 h) was reduced by 16% (P < 0.01) in DU 145 cells and 68% (P < 0.001) in LNCaP cells (Fig. 4A), respectively. Simultaneously, the outcomes of transwell assay suggested that, the invasion rate of cells in shSLC4A4 group was reduced by 89% (P < 0.001) in DU 145 cells and 93.6% (P < 0.001) in LNCaP cells (Fig. 4B).

In order to investigate the potential of shSLC4A4 as a therapeutic target for PCa, the nude mouse xenograft model was established with DU 145 cells. The transfection efficiency of infecting into DU 145 cells with shSLC4A4 or shCtrl lentivirus was identified to be > 80%, and these cells were then injected subcutaneously into the 20 mice. After intraperitoneal injection of D-fluorescein into the 20 mice, the bioluminescence intensities (µW/cm²) were calculated by in vivo imaging. In contrast to shCtrl group, the bioluminescence intensity of shSLC4A4 group was lowered by 93% (P < 0.001; Fig. 5A, B). In addition, compared to shCtrl group, the tumors from mice of shSLC4A4 group were smaller in diameter at all five measured stages, and were lower in weight (P < 0.05; Fig. 5C–E). The above results indicated that the tumor growth was slower in shSLC4A4 group (P < 0.05). Besides, Ki-67 staining showed that the proliferative potential of PCa cells was obviously restrained in shSLC4A4 group in contrast to shCtrl group (Fig. 5F). These results support that SLC4A4 knockdown can inhibit tumour progression in vivo. Altogether, the findings imply that shSLC4A4, which targets explicitly SLC4A4, may have a potent prohibitive effect on prostate tumorigenesis in vivo.

The potential downstream signalling pathways and functional targets mediating the effects induced by SLC4A4 knockdown on PCa were explored using Human Phospho-Kinase Array Kit. Among them, the expression levels of proteins Akt1/2/3 (T308), CREB (S133), Chk-2 (T68), c-jun (S63), GSK-3α/β (S21/S9), GSK-3β (S9), p53 (S46), JNK1/2/3 (T183/Y185, T221/Y223), p38α (T180/Y182), PDGF Rβ (Y751), PLC-γ1 (Y783), Src (Y419), PYK2 (Y402), Yes (Y426), STAT1 (Y701) and STAT3 (S727) were assayed to be a significant down-regulation upon SLC4A4 knockdown compared with the shCtrl-infected cells (P < 0.05; Fig. 6A–C), suggesting the potential functional targets for SLC4A4-related regulation of PCa and guiding future directions of our research.

The results of western blotting indicated that after SLC4A4 interfered, compared with shCtrl group, the expressions of proteins CDK4, BCL-2 were down-regulated. In contrast, the expressions of proteins BAX, FAS were up-regulated (Fig. 6D). These protein bands have further verified that the knockdown of SLC4A4 could induce PCa apoptosis and boost the BAX/BCL-2 ratio, exhibiting a potent effect of SLC4A4 on restraining apoptosis and cell cycle. In other words, the results indicate that SLC4A4 promotes the progression of PCa.

Activation of PI3K/AKT can lead to diverse biological activities, like immunity, inflammation, cell proliferation, apoptosis, and tumorigenesis [22‐24]. In the present study, we detected whether the AKT activator SC79 could reverse the influence of SLC4A4 knockdown on PCa cells. Compared with shCtrl group, the level of SLC4A4 protein was down-regulated, AKT had no significant change, p-AKT expression was down-regulated, BAX expression was up-regulated and BCL-2 expression was down-regulated in shSLC4A4 group (Fig. 7A). Compared to the shSLC4A4 group without SC79 treatment, the treatment with SC79 of the shSLC4A4 + SC79 group produced the opposite effects on the expression of the proteins (p-AKT, BAX, BCL-2). The p-AKT level was clearly enhanced upon SC79 treatment (Fig. 7A). These results demonstrated that SC79 partially reversed the inhibitory action of SLC4A4 knockdown on PCa cells. In order to research the roles of AKT pathway in SLC4A4-mediated cell viability, apoptosis and mobility, rescue experiments were conducted as well. In the existence of SC79, compared with shSLC4A4 group, the cell proliferation was clearly elevated and the apoptosis rate was obviously reduced in shSLC4A4 + SC79 group (Fig. 7B–D). Simultaneously, the migration and invasion abilities of the cells upon the SC79 treatment were meaningfully increased (Fig. 8A, B). Altogether, these findings confirmed that SLC4A4 could promote PCa progression through regulating the AKT pathway.