MCT4 Upregulates PD-L1 Expression and Denes Novel Composite Therapeutic Targets for Triple-Negative Breast Cancer

Immune checkpoint blocking therapy targeting the PD-1/PD-L1 axis has shown promising availability for triple-negative breast cancer (TNBC). Nevertheless, in some cases, targeting eciency is low and ecient gene interaction networks need to be sought, which inspired the exploration that MCT4 and PD-L1 co-expression network analysis and potential regulatory mechanism research. In the paper, bioinformatics, Western blot, qRT-PCR, ow cytometry, biochemical analysis, multiple immunohistochemistry, CRISPR/Cas9 and lentiviral transduction for MCT4 knockout (sgMCT4/231 KO) or overexpression (pEGFP-N1-MCT4/231) were adopted. Analysis of database basis showed MCT4 (SLC16A3) and PD-L1 (CD274) were functionally correlated and highly expressed in TNBC cells, further MCT4 and PD-L1 were co-expressed (more than 50% PD-L1 + MCT4 + cells) in tissue section of TNBC patients. The expression of PD-L1 in TNBC cell lines MDA-MB-231, MDA-MB-468 and BT-549 was sensitive to lactate concentration, and lowering MCT4 expression could downregulate PD-L1 expression through affecting the lactate concentration. These data suggests that MCT4 is positively associated with PD-L1 and the co-targeted therapy for TNBC may be a promising clinical treatment strategy. The expression of PD-L1 was synchronously altered following altered MCT4 expression. WB was used to identify the expression of MCT4 in cells (a). We detected lactate concentration in medium supernatant of cells utilizing ELISA (b) and representative images and measurement of cells migration using the wound closure assay(e). WB (c) and qRT-PCR (d) were used to conrm the positive correlation of PD-L1 and MCT4. The results are presented as the mean ± S.E.M. of at least three independent experiments. Statistical analysis: Three lines and were treated with lactate ranging in 0, 5, 10, 15, 20 mM concentrations after glucose starvation for 5h for measure of PD-L1 expression in protein level (a-c) and mRNA level (d). The results are presented as the mean ± S.E.M. of at least three independent experiments. of 7ACC1 to MCT4. with 7ACC1 (40μM) and lactate (10mM and 15mM) for 4h, and WB (d) was used to detect the expression of PD-L1 and MCT4. The results are presented as the mean ± S.E.M. of at least three independent experiments. Statistical analysis: one-way ANOVA: * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. (e) Schematic of MCT4-mediated expression of PD-L1 through lactate in TNBC cell.

Immune checkpoint blockade (ICB) has now revolutionized the treatment of cancer, notably melanoma, lymphoma, renal cell carcinoma, non-small cell lung cancer, and TNBC [4]. Immune checkpoint Programmed cell death-1 (PD-1) is mainly expressed on T cells [5], and PD-L1, a ligand of PD-1, is expressed on antigen-presenting cells as well as cancer cells, both of them mediate tumor immune escape [4,6]. A study of 650 BC cases presented the positive expression rate of PD-L1 was 23.4% with reducing overall survival [7]. Similarly, a recent meta-analysis of PD-L1 in BC showed PD-L1 expression was a marker of poor prognosis [8]. One of the mechanisms of PD-L1 upregulation is the ampli cation of PD-L1 gene (CD274) at 9p24.1, which has been con rmed in TNBC cancer cells [9]. PD-L1 is highly expressed in melanoma and bladder cancer and its target antibody Nivolumab has been approved by the FDA for use in the treatment. Recent studies have shown that Nivolumab is also available in TNBC patients [10][11][12]. Lately researchers demonstrated the e cacy of PD-L1 therapy for TNBC by detecting low levels of serum PD-L1 (sPD-L1) in the patients' blood [13]. Some researchers proposed that blocking both PD-L1 and cytotoxic T-lymphocyte-associated antigen-4 (CTLA-4) were required for TNBC [14].
Monocarboxylate transporter protein 4 (MCT4) primarily mediates the transmembrane transport of proton-coupled monocarboxylic acid, such as lactate, and promotes cancer progression [15,16]. TNBC cells was found to exhibit a higher glycolysis rate due to high glucose uptake rate, overexpression of glycolytic enzymes, high oxygen consumption rate (OCR) and high extracellular acidi cation rate (ECAR) [17][18][19]. To excrete glycolytic wastes, MCT4 is assigned to actively express in TNBC cells [19,20]. MCT4 is therefore considered a carcinogenesis factor, even some pharmacologists study its structure to block it effectively [21].
It is now recognized that excess lactate and acidic tumor microenvironment (TME) can stimulate tumor angiogenesis and increase metastasis and immune escape of cancer cells [22][23][24][25][26][27], suggesting there is a connection between glycolytic metabolic molecules and the immune related molecules in cancer cells.
The expression of PD-L1 has been reported to be sensitive to additional lactate [28]. Based on previous studies, we intend to study a close connection between MCT4 and PD-L1, and research the role of lactate for TNBC cells, which may provide a combination mode of blocking MCT4/PD-L1 therapy for TNBC.

Materials And Methods
Gene enrichment analysis Gene enrichment analysis of CD274 (PD-L1) and SLC16A3 (MCT4) was performed using the Metascape online database (http://www.metascape.org/) based on the dataset retrieved from the String database Wound closure assay and lactate concentration measurement Cells were seeded into 6-well plate (1*10 6 cells per well) and incubated for overnight at 37℃. Different processing monolayer cells were scraped using 10 ml pipette tips for 0 h and 24 h, and image analyses of the wound closure area were conducted. The supernatant of cell culture was detected utilizing Human Lactate Elisa Kit (J&L, Shanghai, China) for detecting lactate concentration according to the manual.

Quantitative real time PCR (qRT-PCR) analysis
Total RNA was extracted using lysis reagent Trizol kit (Vazyme, Nanjing, China), and cDNA was obtained by Hifair® 1st Strand cDNA Synthesis Kit (Yeasen, Shanghai, China). qRT-PCR was carried out using UltraSYBR Mixture (Cwbio, Beijing, China), and the data obtained was analysed by the ΔΔCt method. The primers used for qRT-PCR was listed in Table 1.

Immuno uorescence
Cells grown in round glass slides were xed by 75% alcohol and then were incubated with anti-MCT4 (G-9) (Santa Cruz, 1:100) and anti-PDL1 (Proteintech, 1:100) primary antibody overnight following goat serum blocking, and then incubated with HRP-conjugated uorescent secondary antibodies for 1h and DAPI for 10min at room temperature, and images were photographed by confocal microscope (Olympus, FV1000, Japan).

Flow cytometry analysis
The cells were stained by primary antibodies PDL1 (Proteintech, 1:100) and MCT4 (G-7) (Santa Cruz, 1:100) for 30 min and uorescent secondary antibody for 30min respectively. The stained cells were analysed by a ow cytometer, and data obtained were presented with FlowJo software.
Hematoxylin-Eosin (HE) staining and multiple immunohistochemical staining (mIHC) HE staining and PD-L1, MCT4 and EpCAM staining were performed on formalin-xed, para n-embedded human TNBC tumor tissue sections provided by Tianjin cancer hospital with the permission of Nankai University medical ethics committee. We trusted WiSee Biotechnology Company to perform HE and mIHC staining and assay (protocal seen as supplemental material 1), which provided simultaneous detection and quantitation of PD-L1, MCT4 and EpCAM, and we provided MCT4 (G-7) (Santa Cruz) and MCT4 (G-9) (Santa Cruz) antibody.

Statistical analysis
Statistical analyses were performed with analysis of variance (one-way ANOVA) and Student's t test using SPSS 13.0 Statistical Software (SPSS Inc., Chicago, IL, USA) and are presented as mean±s.d. from triplicated independent experiments. A signi cant difference was considered when the P-value was <0.05.

Results
Bioinformatics analysis revealed MCT4 and PD-L1 were functionally linked and were highly expressed in TNBC To progressively determine the correlation of MCT4 and PD-L1 in TNBC, we adopted String online database to retrieve genes closely related to PD-L1 and MCT4 respectively, and used the Metascape online database to perform a comprehensive analysis of these two gene sets. Fig. 1a showed a dense functional overlap between the two gene sets, and the heat map of the enrichment analysis (Fig. 1b) exhibited that the two gene sets were jointly enriched in the physiological process of carbon metabolism in tumors. We then merged the two gene sets into a total set and used Metascape to generate the results again, and found that the most predominant functional enrichment of the total gene set was in PD-1 signaling (Fig. 1c). All these results indicated that PD-L1 and MCT4 were functionally related, and provided a strong theoretical basis for our subsequent experiments. We further analysed the expression level of MCT4 and PD-L1 of BC subtypes according to the French-American-British (FAB) classi cation based on UALCN-TCGA (The Cancer Genome Atlas) online database, and found that MCT4 and PD-L1 were speci cally expressed on TNBC at mRNA level ( Fig. 1d-e).
mIHC of tissue samples of TNBC patients con rmed MCT4 and PD-L1 had an actively co-expression Further investigation about the pathological correlation of MCT4 and PD-L1 were performed in tissues of TNBC patients on the basis of molecular biological analysis. TNBC tissues with nests of glandular epithelial cells with pink cytoplasm and distinct cell borders were presented by HE staining (Fig. 2a). mIHC staining (supplemental material 1) presented the distribution and spatial location of MCT4 and PD-L1 proteins. Two group images showed that compared with the stroma (top row), MCT4 and PD-L1 were both strongly expressed in the cancer tissue (bottom row) (Fig. 2b). Additionally, the statistics (supplemental material 2) presented that the ratio of MCT4 (G-7) positive cells and PD-L1 positive cells to total cells were 54.56% and 53.28%, respectively, and the ratio of MCT4 (G-9) positive cells and PD-L1 positive cells to total cells were 55.84% and 55.02%, respectively (Fig. 2c). A nely-ratio of MCT4 + /PDL1 + /EpCAM + cells to EpCAM + cells accounted for 62.50% or 62.35%, respectively (Fig. 2d-e). The results con rmed that MCT4 and PD-L1 were co-expressed in TNBC tissues, which also offered advisory evidence for TNBC ICB therapy.
To clarify the effect of MCT4 on PD-L1 expression, MDA-MB-231 that arti cially knocked down or overexpressed MCT4 were established To prove whether MCT4 in uenced the expression of PD-L1, we altered the expression of MCT4 in TNBC cells. We rstly detected the expression of MCT4 and PD-L1 in three TNBC cell lines by WB and cellular immuno uorescence (Fig. 3a-b). Then we used CRISPR/Cas9 technology (Fig. 3c) to speci cally silence the expression of MCT4 in MDA-MB-231 cells (Ctr/231), and screening a stable MCT4-silenced cell line sgMCT4/231(KO) with a silencing e ciency of 98.5%±0.6% (Fig. 3d). Infectious lentivirus particles for MCT4 overexpression were constructed through transfecting 293T cells, and MDA-MB-231 cells were infected with the viral supernatant to obtain a stable MCT4 overexpression cell line pEGFP-N1-MCT4/231(wt), with an e ciency of 7.5 ± 1.5 times (Fig. 3e). Next, ow cytometry analysis displayed that the expression trend of PD-L1 was consistent with that of MCT4 in three cell lines representing Ctr/231, sgMCT4/231(KO) and pEGFP-N1-MCT4/231(wt) (Fig. 3f), which preliminarily suggested that the change of MCT4 positively affected the expression of PD-L1.

Altered MCT4 expression positively affected lactate e ux and PD-L1 expression
Given that the MCT4 KO or MCT4 overexpression are stably expressed by sgMCT4/231(KO) or pEGFP-N1-MCT4/231(wt) cells respectively and play pivotal roles in affecting lactate e ux, we analysed the expression of MCT4 by WB as shown in Fig. 4a, and detected the corresponding lactate concentration of supernatant of three cell lines and observed a signi cantly synchronous change in lactate concentration ( Fig. 4b) as well as cell migration (Fig. 4e) with MCT4 expression. Simultaneously, we also found PD-L1 protein expression showed a trend consistent with MCT4 as shown by WB result (Fig. 4c), and so did mRNA expression (Fig. 4d), which suggested altered MCT4 expression could positively affected lactate e ux and PD-L1 expression.

PD-L1 expression was sensitive to lactate in three different TNBC cell lines
To verify the susceptibility of PD-L1 to lactate stimuli in TNBC cells, we selected ve different lactate concentrations to stimulate MDA-MB-231 cells, and investigated the expression of membrane protein and total protein of PD-L1 by WB (Fig. 5a). We found that the expression of PD-L1 in MDA-MB-231 cells was sensitive to lactate, and then MDA-MB-468 and BT-549 were tested (Fig. 5b-c). The results showed that the expression of PD-L1 in protein levels, as did the mRNA expression of PD-L1 (Fig. 5d), increased signi cantly and showed a dependence on lactate concentrations. It suggested TNBC cell lines were sensitive to different concentrations of lactate, which could up-regulate the expression of PD-L1.
MCT4 upregualted PD-L1 expression via lactate was further credited in MDA-MB-231 cells To further con rm that MCT4 promoted PD-L1 expression via lactate in TNBC cells, we chose the MCT4 inhibitor 7-aminocarboxycoumarins (7ACC1) [29] to speci cally interfere with MCT4 expression and detected whether the lactate e uxes and PD-L1 expression was affected in representative MDA-MB-231 cells. We selected the appropriate concentrations of 7ACC1 [16] according to our previous study. Cells were treated with 7ACC1 (20μM, 30μM, 40μM, and 0.1mM) for 4 h and 8 h, and the lactate concentration decreased by Elisa analysis (Fig 6a-b). Meanwhile, WB result con rmed that MCT4 expression was effectively inhibited (Fig 6c). We nally determined the optimal concentration and reaction time as 40 µM and 4 hours. Then we designed experimental groups with the addition of lactate with or without the addition of 7ACC1, and we found that the trend of increased PD-L1 protein expression following lactate stimulation disappeared after MCT4 was inhibited by 7ACC1 (Fig. 6d). Although other factors could not be excluded, the positive regulating effect of MCT4 on PD-L1 could be explained. Therefore, it implied that treating TNBC patients with blocking PD-L1 immunotherapy combined with blocking MCT4 might reduce recurrence and metastasis risk of TNBC, which was probably achieved by inhibition of either lactate or tumor glycolysis (Fig. 6e).

Discussion
ICB is becoming standard treatments in several cancer types [30]. In 2019, the US Food and Drug Administration (FDA) approved the use of the blocking antibody Atezolizumab against PD-L1 combined with nab-paclitaxel for TNBC [31]. However, PD-L1 is expressed in 20% of TNBC [6], so ICB has limitations obviously. The response rate of single-agent ICB remains low and of limited durability in unselected patient populations with metastatic TNBC [32]. It is urgent to nd more signi cant molecules as cotargets of PD-L1 to enhance the therapeutic range and e cacy of TNBC.
TNBC cell widely exhibits a metabolic phenotype with a high glycolysis [19], which negatively affects immunotherapy, for example, elevated serum lactate dehydrogenase levels in TNBC patients are associated with bad ICB therapy [33][34][35]. Therefore, targeting hyperglycolysis combined with ICB may improve the therapeutic e cacy of TNBC. Monocarboxylic acid transporter MCT4, responsible for lactate metabolism, is key molecules in targeting hyperglycolysis. In this study, we used the online databases UALCAN to predict the impact of MCT4 expression in BC. We found that MCT4 expression was higher in invasive BC compared to normal breast tissue (n = 1211, p < 1E-12), and MCT4 was highly expressed in TNBC (n = 833, p < 0.05).
It was reported that the combination of ICB with the lactate reduction was bene cial to improve the treatment response of TNBC. A preclinical study showed that in tumors with high glycolysis rates, Vdomain immunoglobulin inhibitory effect on T cell activation (VISTA) was more prominent in acidic conditions. Blocking VISTA with monoclonal antibodies reversed immunosuppression, resulting in increased T-cell in ltration, inhibition of the expression of checkpoint receptors (PD-1, LAG2, and Tim-3) on T-cells, and subsequently increased antitumor activity in MC38 colorectal cancer mice [36]. These hints targeting MCT4 to reduce lactate levels is important for the treatment of ICB.
It was reported that MCT4 and PD-L1 were consistently highly expressed in BC tissues [37], and lactate could stimulate the expression of PD-L1 in cancer cells [28]. MCT4 was a transporter of lactate to increase the extracellular lactate concentration, therefore we speculated that in TNBC cells, under the dominant regulation of MCT4, lactate might stimulate the expression of PD-L1. Analysis of database basis showed a signi cant functional correlation between MCT4-centered gene sets and PD-L1-centered gene sets. Together, these two genes enriched the biological functions of carbon metabolism and glycolysis and gluconeogenesis in cancer centers. Then we veri ed the expression of PD-L1 in overexpressed or knockdown MCT4 cell lines through a series of methods, and found that the extracellular lactate level was positively correlated with the expression level of PD-L1, and the changes in the expression of MCT4 and PD-L1 were always synchronous. This is consistent with our expectation that lactate upregulates the level of PD-L1, while MCT4 upregulates lactate, which shapes a positive feedback regulation.
More importantly, our study might contribute to the accurate differentiation of TNBC molecule subtype.  39]. Lehmann also developed TNBCtype online parting tools, but clinical application was still limited, more evidence was needed to re ne TNBC molecule subtype for better therapeutic effect. We uncovered PD-L1 had the highest expression in TNBC IM, and more than 50% PD-L1 + MCT4 + cells in tissue section of TNBC patients might have a major impact on TNBC treatment. According to the 2018 ASCO/CAP guidelines, HER2 is classi ed as positive (3 + ) with more than 10% expression in BC by IHC analysis, and could be treated with targeted drugs [40]. Therefore, the co-expression of PD-L1 and MCT4 in TNBC IM subtype was found to be up to 50%, which is expected to become a new pathological staging or immunotherapy strategy, and is of great signi cance for the precision treatment of TNBC. Tables   Table 1 Quantitative real time PCR primer     and PD-L1 cells to total cells was higher than 50%. MCT4 and PD-L1 proteins were highly expressed in EpCAM positive tissue areas (d). MCT4 and PD-L1 co-expressed cells accounted for more than 60%

Declarations
among the EpCAM positive cells in the tissue samples of TNBC patients (e). Flow cytometry analysis (f) was utilized to con rm the positive correlation of PD-L1 and MCT4. The results are presented as the mean ± S.E.M. of at least three independent experiments. Statistical analysis: one-way ANOVA and T test: * p < 0.05, *** p < 0.001.

Figure 4
Page 17/19 The expression of PD-L1 was synchronously altered following altered MCT4 expression. WB was used to identify the expression of MCT4 in cells (a). We detected lactate concentration in medium supernatant of cells utilizing ELISA (b) and representative images and measurement of cells migration using the wound closure assay(e). WB (c) and qRT-PCR (d) were used to con rm the positive correlation of PD-L1 and MCT4. The results are presented as the mean ± S.E.M. of at least three independent experiments.

Figure 5
The expression level of PD-L1 was upregulated with the addition of lactate. Three TNBC cell lines MDA-MB-231 and MDA-MB-468 and BT-549 were treated with lactate ranging in 0, 5, 10, 15, 20 mM concentrations after glucose starvation for 5h for measure of PD-L1 expression in protein level (a-c) and mRNA level (d). The results are presented as the mean ± S.E.M. of at least three independent experiments. Statistical analysis: one-way ANOVA: * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. Figure 6