Advertisement

Inhibition of PAI-1 Blocks PD-L1 Endocytosis and Improves the Response of Melanoma Cells to Immune Checkpoint Blockade

Open ArchivePublished:May 14, 2021DOI:https://doi.org/10.1016/j.jid.2021.03.030
      Immune checkpoint molecules, especially PD-1 and its ligand PD-L1, act as a major mechanism of cancer immune evasion. Although anti–PD-1/PD-L1 monotherapy increases therapeutic efficacy in melanoma treatment, only a subset of patients exhibits long-term tumor remission, and the underlying mechanism of resistance to PD-1/PD-L1 inhibitors remains unclear. In this study, we demonstrated that cell surface retention of PD-L1 is inversely correlated with PAI-1 expression in vitro, in vivo, and in clinical specimens. Moreover, extracellular PAI-1 induced the internalization of surface-expressed PD-L1 by triggering clathrin-mediated endocytosis. The endocytosed PD-L1 was transported to lysosomes for degradation by endolysosomal systems, resulting in the reduction of surface PD-L1. Notably, inhibition of PAI-1 by pharmacological inhibitor with tiplaxtinin led to elevated PD-L1 expression on the plasma membrane, both in vitro and in vivo. Strikingly, targeting PAI-1 by tiplaxtinin treatment synergizes with anti–PD-L1 immune checkpoint blockade therapy in a syngeneic murine model of melanoma. Our findings demonstrate a role for PAI-1 activity in immune checkpoint modulation by promoting surface PD-L1 for lysosomal degradation and provides an insight into the combination of PAI-1 inhibition and anti–PD-L1 immunotherapy as a promising therapeutic regimen for melanoma treatment.

      Abbreviations:

      CAV (caveolin), CLA (clathrin), siPai-1 (Pai-1 small interfering RNA), siRNA (small interfering RNA), TPX (tiplaxtinin)

      Introduction

      Blockade of immune checkpoints enhances therapeutic antitumor immunity. One of the most studied immune checkpoints is the PD-L1/PD-1 pathway (
      • Guo L.
      • Wei R.
      • Lin Y.
      • Kwok H.F.
      Clinical and recent patents applications of PD-1/PD-L1 targeting immunotherapy in cancer treatment-current progress, strategy, and future perspective.
      ;
      • Zhang J.Y.
      • Yan Y.Y.
      • Li J.J.
      • Adhikari R.
      • Fu L.W.
      PD-1/PD-L1 based combinational cancer therapy: icing on the cake.
      ). PD-L1 is a transmembrane glycoprotein that is expressed on the surface of tumor cells and tumor-infiltrating immune cells. The binding of PD-L1 to its receptor, PD-1, causes T-cell inhibition and downregulation of T-cell activity, leading to inactivation of the antitumor immune response (
      • Patrinely Jr., J.R.
      • Dewan A.K.
      • Johnson D.B.
      The role of anti-PD-1/PD-L1 in the treatment of skin cancer.
      ). Although anti–PD-L1 monotherapy has yielded promising clinical effects in treating melanoma, more than half of patients either exhibit or develop disease progression after an initial response (
      • Nowicki T.S.
      • Hu-Lieskovan S.
      • Ribas A.
      Mechanisms of resistance to PD-1 and PD-L1 blockade.
      ;
      • Wang D.Y.
      • Eroglu Z.
      • Ozgun A.
      • Leger P.D.
      • Zhao S.
      • Ye F.
      • et al.
      Clinical features of acquired resistance to anti-PD-1 therapy in advanced melanoma.
      ). Hence, the development of a strategy in combination with anti–PD-L1 therapy is required for the improvement of therapeutic efficacy in melanoma treatment.
      Emerging evidence suggests that improved prognosis and clinical outcomes in patients with melanoma treated with anti–PD-L1 therapy are associated with PD-L1 positivity (
      • Larkin J.
      • Chiarion-Sileni V.
      • Gonzalez R.
      • Grob J.J.
      • Cowey C.L.
      • Lao C.D.
      • et al.
      Combined nivolumab and ipilimumab or monotherapy in untreated melanoma [published correction appears in N Engl J Med 2018;379:2185].
      ;
      • Patel S.P.
      • Kurzrock R.
      PD-L1 expression as a predictive biomarker in cancer immunotherapy.
      ). Recently, endocytosis of surface receptors, such as EGFR and PD-L1, has been proven to regulate antitumor response to mAb therapy in human tumors (
      • Chew H.Y.
      • De Lima P.O.
      • Gonzalez Cruz J.L.
      • Banushi B.
      • Echejoh G.
      • Hu L.
      • et al.
      Endocytosis inhibition in humans to improve responses to ADCC-mediating antibodies.
      ;
      • Joseph S.R.
      • Gaffney D.
      • Barry R.
      • Hu L.
      • Banushi B.
      • Wells J.W.
      • et al.
      An ex vivo human tumor assay shows distinct patterns of EGFR trafficking in squamous cell carcinoma correlating to therapeutic outcomes.
      ;
      • Mellman I.
      • Yarden Y.
      Endocytosis and cancer.
      ). Inhibition of endocytosis by clinically available drugs enhances the availability of surface target proteins largely and improves antibody-dependent cellular cytotoxicity in cancer treatment (
      • Chew H.Y.
      • De Lima P.O.
      • Gonzalez Cruz J.L.
      • Banushi B.
      • Echejoh G.
      • Hu L.
      • et al.
      Endocytosis inhibition in humans to improve responses to ADCC-mediating antibodies.
      ), thereby suggesting that retention of surface PD-L1 potentiates immunotherapeutic effects with anti–PD-L1 therapy.
      PAI-1 is a member of the serpin superfamily and serves as a potent inhibitor of tPA and uPA (
      • Li S.
      • Wei X.
      • He J.
      • Tian X.
      • Yuan S.
      • Sun L.
      Plasminogen activator inhibitor-1 in cancer research.
      ). Binding of PAI-1 to the uPA/uPAR complex results in the recruitment of LRP1 and triggers the endocytosis process, leading to internalization of the PAI-1–uPA–uPAR–LRP1 quaternary complex. The endocytosed PAI-1 and uPA then undergo lysosomal degradation, whereas uPAR and LRP1 are transported back to the plasma membrane by recycling endosomes (
      • Pasupuleti N.
      • Grodzki A.C.
      • Gorin F.
      Mis-trafficking of endosomal urokinase proteins triggers drug-induced glioma nonapoptotic cell death.
      ). PAI-1 also executes a broad spectrum of protumorigenic functions, such as promoting tumor vasculature, enhancing tumor growth and motility, and protecting tumor cell apoptosis (
      • Bajou K.
      • Peng H.
      • Laug W.E.
      • Maillard C.
      • Noel A.
      • Foidart J.M.
      • et al.
      Plasminogen activator inhibitor-1 protects endothelial cells from FasL-mediated apoptosis.
      ,
      • Bajou K.
      • Masson V.
      • Gerard R.D.
      • Schmitt P.M.
      • Albert V.
      • Praus M.
      • et al.
      The plasminogen activator inhibitor PAI-1 controls in vivo tumor vascularization by interaction with proteases, not vitronectin. Implications for antiangiogenic strategies.
      ;
      • Fang H.
      • Placencio V.R.
      • DeClerck Y.A.
      Protumorigenic activity of plasminogen activator inhibitor-1 through an antiapoptotic function.
      ;
      • Isogai C.
      • Laug W.E.
      • Shimada H.
      • Declerck P.J.
      • Stins M.F.
      • Durden D.L.
      • et al.
      Plasminogen activator inhibitor-1 promotes angiogenesis by stimulating endothelial cell migration toward fibronectin.
      ). However, little is known about the PAI-1–induced vesicle endocytic trafficking involved in the therapeutic efficacy of cancer treatment.
      In this study, we investigated the regulation of cell surface–expressed PD-L1 trafficking by PAI-1 in melanoma cells and the effect of PAI-1 inhibition by a small-molecule inhibitor on the efficacy of PD-L1 blockade therapy (anti–PD-L1 antibody) in a melanoma mouse model. Our findings unravel a previously unreported mechanism of regulation of PD-L1 levels at the plasma membrane by PAI-1, by promoting PD-L1 endocytosis and leading to anti–PD-L1 immunotherapy insensitivity. Inhibition of PAI-1–mediated PD-L1 endocytosis may be valuable in improving the response to immune checkpoint blockade.

      Results

      Lack of PAI-1 enriches plasma membrane PD-L1 level in melanoma cells

      To explore the effects of PAI-1 on surface PD-L1 expression in melanoma cells, B16-F10 cells were transfected with small interfering RNA (siRNA)-mediated Pai-1 silencing (Pai-1 siRNA [siPai-1] #1 or #2. The reduced extracellular PAI-1 expression in siPai-1–expressing cells was validated by ELISA (Figure 1a). Flow cytometry analysis showed that, compared with B16-F10 cells expressing nontargeted scramble siRNA, those with Pai-1 knockdown exhibited increased surface expression of PD-L1 (97.7% vs. 2.83% and 87.2% vs. 2.83%; Figure 1a). The mean fluorescence intensity of PD-L1 in siPai-1–expressing groups showed more than a six-fold increase (Figure 1a). Manipulation of PAI-1 expression by either knockdown or overexpression approaches did not interfere with the cellular level of PD-L1 at both transcriptional and translational steps (Supplementary Figure S1), suggesting the enhancement of plasma membrane translocation of PD-L1 resulting from PAI-1 silencing. To further clarify the relationship between extracellular PAI-1 levels and surface PD-L1 abundance, PAI-1 expression was titrated in murine B16-F10 cells by delivering increasing doses of siPai-1 (#1 and #2). In siPai-1 #1–expressing cells, a reduction in 20% of secreted PAI-1 resulted in the enhancement of surface PD-L1. The mean fluorescence intensity of PD-L1 increased by nine-fold when soluble PAI-1 was downregulated by 60% (R2 = 0.84, Supplementary Figure S2a). A similar correlation between extracellular PAI-1 and surface PD-L1 expression was observed in the siPai-1 #2–expressing cells (R2 = 0.88, Supplementary Figure S2b). These results suggest that a decrease in extracellular PAI-1 is correlated with increased surface levels of PD-L1 in cells containing siPai-1. Similar results were observed in human melanoma cell lines A375 and SH4 after short hairpin RNA–mediated SERPINE1 knockdown because a 50% reduction in secreted PAI-1 protein correlated with a three-fold increase in surface exposure of PD-L1 (Supplementary Figure S3). The cellular localization of PD-L1 was detected by immunofluorescent staining to further confirm the surface retention of PD-L1 after Pai-1 silencing. Significant induction of plasma membrane PD-L1 was visualized in siPai-1–expressing cells compared with the faint plasma membrane and cytoplasmic distribution of cellular PD-L1 observed in cells expressing scramble siRNA (Figure 1b). We further examined the effect of PAI-1 on the surface distribution of PD-L1 in vivo. C57BL/6 mice were injected subcutaneously with syngeneic B16-F10 cells expressing either scramble siRNA or siPai-1 (#1). Fluorescent immunohistochemistry analysis by confocal microscopy showed that allografts derived from Pai-1–deficient cells exhibited enhanced PD-L1 staining on the cell surface compared with those with scramble control (Figure 1c). Immunohistochemistry analysis on 96 paraffin-embedded sections from patients with melanoma was performed to examine the correlation between PAI-1 and PD-L1 by assigning protein levels into low (–) and high expression (+) (Figure 1d). Low expression of PAI-1 protein was detected in 44 specimens (45.8%), whereas low PD-L1 expression was observed in 37 specimens (38.5%) (Figure 1d). Strikingly, we found a marked inverse correlation between PAI-1 and PD-L1 (59.3%, P < 0.001; Figure 1e), supporting the findings from in vitro and animal studies. Notably, patients with low PAI-1 expression and high PD-L1 levels were associated with the early stage (stage I and II), whereas the PAI-1/PD-L1 (+/–) expression pattern was associated with the late stage (stage III and IV) (P = 0.001; Supplementary Table S1). Collectively, these results indicate an increase in the membrane retention pattern of PD-L1 in the absence of PAI-1 expression both in vitro and in vivo and in the majority of clinical samples.
      Figure thumbnail gr1
      Figure 1Pai-1 silencing promotes levels of plasma membrane PD-L1 in vitro, in vivo, and in clinical samples. (a) B16-F10 cells were transfected with scr or siPai-1 #1 or #2 for 48 hours. Supernatants were collected for detecting extracellular PAI-1 expression using ELISA (upper right). The ratio of PAI-1 levels was relative to that of the scr. The percentage of cell surface PD-L1 expression was determined by flow cytometry assay (left). The ratio of PD-L1 MFI was relative to scr and shown here as mean ± SD of three independent experiments (lower right). P-values were calculated by two-tailed Student’s t-test. ∗∗P < 0.01 ∗∗∗P < 0.001. (b) Cells treated as in (a) were stained with anti–PD-L1 antibody and processed for confocal immunofluorescence microscopy. (c) C57BL/6 mice were inoculated with B16-F10 scr or Pai-1 (#1) silenced cells subcutaneously. Tumor samples were collected and sectioned on day 16 after inoculation, followed by fluorescence immunohistochemical staining with anti–PD-L1 antibody. Images were acquired by confocal microscopy. Bar = 20 μm. (d) Immunohistochemical analysis of PAI-1 and PD-L1 expression in representative tumor specimens from patients with melanoma. Bar = 50 μm. (e) Concordance analysis of PAI-1 and PD-L1 (High level, +; low level, –) in patients with melanoma. The percentage of a concordant group (+/+ and –/–; 40.6%) and discordant group (+/– and –/+; 59.3%) is indicated in the plot. MFI, mean fluorescence intensity; scr, scramble control small interfering RNA; siPai-1, Pai-1 small interfering RNA; siRNA, small interfering RNA.

      Exogeneous PAI-1 reduces surface PD-L1 level

      It has been shown that PAI-1 can trigger its receptor-mediated internalization for endocytosis (
      • Jaiswal R.K.
      • Varshney A.K.
      • Yadava P.K.
      Diversity and functional evolution of the plasminogen activator system.
      ), whereas cell surface retention of PD-L1 can be maintained by endocytosis inhibition (
      • Chew H.Y.
      • De Lima P.O.
      • Gonzalez Cruz J.L.
      • Banushi B.
      • Echejoh G.
      • Hu L.
      • et al.
      Endocytosis inhibition in humans to improve responses to ADCC-mediating antibodies.
      ). We therefore hypothesized that extracellular PAI-1 induces PD-L1 endocytosis to reduce PD-L1 expression on the plasma membrane. To this end, exogenous recombinant PAI-1 was used to treat siPai-1–expressing cells because abundant surface PD-L1 was expressed in these cells. In fact, unlike in scramble-expressing cells, recombinant PAI-1 administration to cells with Pai-1 knockdown reduced the plasma membrane level of PD-L1 in a dose-dependent manner (Figure 2a). Furthermore, we treated siPai-1–expressing cells with biotin-labeled recombinant PAI-1 protein to trace the transport of PD-L1 invagination. As shown in Figure 2b, compared with treatment with biotin-labeled BSA, a significant overlapping of signal between biotinylated PAI-1 and PD-L1 was observed (Figure 2b), thereby suggesting a key role of extracellular PAI-1 in mediating the expression of plasma membrane PD-L1. We therefore attempted to manipulate the extracellular PAI-1 level by using chemical compounds to inhibit intracellular vesicle trafficking pathways. Brefeldin A (for inhibition of transport from the endoplasmic reticulum to the Golgi complex), nocodazole (for blockade of microtubule polymerization), and cytochalasin D (for impairment of actin filament) were used to test whether low secretion of PAI-1 was correlated with high expression of surface PD-L1. Compared with the control group, cells treated with brefeldin A dramatically reduced PAI-1 release, whereas stimulation through nocodazole or cytochalasin D did not change the levels of PAI-1 in supernatants (Figure 2c). Notably, a significant increase in expression of PD-L1 on the plasma membrane was also observed in cells treated with brefeldin A but not in the nocodazole- or cytochalasin D–treated groups (Figure 2d). Tiplaxtinin (TPX), an effective small-molecule inhibitor of PAI-1, was used to further evaluate the role of exogenous PAI-1 in the enhancement of plasma membrane–bound PD-L1 because TPX has been previously reported to block soluble PAI-1 bioactivity (
      • Hennan J.K.
      • Morgan G.A.
      • Swillo R.E.
      • Antrilli T.M.
      • Mugford C.
      • Vlasuk G.P.
      • et al.
      Effect of tiplaxtinin (PAI-039), an orally bioavailable PAI-1 antagonist, in a rat model of thrombosis.
      ). Figure 2e shows that TPX treatment significantly upregulates PD-L1 expression on the plasma membrane in a dose-dependent manner in B16-F10 cells. Accordingly, treatment of human melanoma cell lines A375, SK-MEL-24, and RPMI-7951 with TPX also revealed enrichment of cell surface PD-L1 (Supplementary Figure S4), suggesting that inhibition of exogenous PAI-1 activity can increase plasma membrane translocation of PD-L1.
      Figure thumbnail gr2
      Figure 2Exogeneous PAI-1 restrained surface PD-L1 expression. (a) B16-F10 cells were transfected with scr or siPai-1 #1 for 48 hours, followed by treatment with increasing doses of rPAI-1 for a further 4 hours or left untreated. Cell surface PD-L1 expression was analyzed by flow cytometry and the ratio of PD-L1 MFI was compared with scr and shown as mean ± SD of three independent experiments. (b) Confocal microscopy images showing the localization of PD-L1 and biotinylated PAI-1 protein in scr- or siPai-1–expressing B16-F10 cells. Biotin-labeled proteins (BSA and PAI-1) were used at 10 μg/ml for 4 hours and visualized by Alexa Fluor 594–conjugated streptavidin. Green: PD-L1 staining; red: biotin positive staining; blue: nucleus staining. Bar = 20 μm. (c) B16-F10 cells were treated with BFA, nocodazole, or cytD for 24 hours. Supernatants were collected and the levels of soluble PAI-1 were determined by ELISA assay. (d) PD-L1 expression at the cell surface was analyzed by flow cytometry following treatment with the compounds in (c). (e) TPX was used to treat B16-F10 cells at the indicated doses for 24 hours. Surface PD-L1 expression was analyzed by flow cytometry. ∗∗P < 0.01, ∗∗∗P < 0.001. n = 3. BFA, brefeldin A; ctrl, control; cytD, cytochalasin D; MFI, mean fluorescence intensity; rPAI-1, recombinant active PAI-1; scr, scramble control small interfering RNA; siPai-1, Pai-1 small interfering RNA; siRNA, small interfering RNA; TPX, tiplaxtinin.

      PAI-1–induced clathrin-dependent endocytosis internalizes cell surface PD-L1

      We speculated that PAI-1 promotes PD-L1 endocytosis through the receptor-mediated pathway. To test this hypothesis, we treated B16-F10 cells with the clinically approved antiemetic/antipsychotic drug prochlorperazine, which has been reported to block endocytosis through dynamin inhibition (
      • Daniel J.A.
      • Chau N.
      • Abdel-Hamid M.K.
      • Hu L.
      • von Kleist L.
      • Whiting A.
      • et al.
      Phenothiazine-derived antipsychotic drugs inhibit dynamin and clathrin-mediated endocytosis.
      ). Dose-dependent induction of surface PD-L1 was observed on treatment with prochlorperazine. Flow cytometry assay detected a more than three-fold increase in induction with a 10 μM dose of prochlorperazine (Figure 3a). In addition, confocal microscopic immunofluorescence images showed that treatment of B16-F10 cells with prochlorperazine localized PD-L1 and PAI-1 to the plasma membrane when compared with the untreated control (Figure 3b), thereby suggesting that the membrane colocalization of PD-L1 and PAI-1 can be induced by endocytosis inhibition.
      Figure thumbnail gr3
      Figure 3Inhibition of endocytosis enhances the level of surface PD-L1. (a) Surface PD-L1 expression of B16-F10 cells was analyzed by flow cytometry after treatment with PCZ at 0, 5, or 10 μM for 24 hours. The ratio of PD-L1 MFI was relative to the control group and has been shown here as mean ± SD of three independent experiments. P-values were determined by two-tailed Student’s t-test. ∗∗∗P < 0.001. (b) B16-F10 cells were treated with PCZ at a dose of 10 μM for 24 hours. Representative images of confocal microscopy showed colocalization of PAI-1 (red) and PD-L1 (green) at the plasma membrane. Bar = 20 μm. Ctrl, control; MFI, mean fluorescence intensity; PCZ, prochlorperazine.
      Clathrin (CLA)- and caveolin (CAV)-mediated endocytosis has been reported to control the internalization of most cell surface receptors and their ligands (
      • Manzanares D.
      • Ceña V.
      Endocytosis: the nanoparticle and submicron nanocompounds gateway into the cell.
      ). Hence, we examined whether CLA or CAV mediated PAI-1–dependent PD-L1 endocytosis. The effective silencing of CLA gene, Cla, and CAV gene, Cav, was validated by western blotting (Supplementary Figure S5a and b). B16-F10 cells with Cla silencing, but not knock down of Cav, expressed increased PD-L1 on the plasma membrane (Figure 4a). siRNA-mediated knock down of Lrp1 (Supplementary Figure S5c), the endocytic receptor for PAI-1 (
      • Cao C.
      • Lawrence D.A.
      • Li Y.
      • Von Arnim C.A.
      • Herz J.
      • Su E.J.
      • et al.
      Endocytic receptor LRP together with tPA and PAI-1 coordinates Mac-1-dependent macrophage migration.
      ), also enhanced membrane localization of PD-L1 (Figure 4a). All the silencing of endocytosis-related proteins used did not interfere with cellular PD-L1 expression (Supplementary Figure S5d). In addition, we used Pitstop 2 to block clathrin-mediated endocytosis and showed that inhibition of clathrin-coated pits formation significantly increased the level of surface PD-L1 compared with B16-F10 cells treated with inactive Pitstop 2 (Figure 4b). These results suggest that endocytosis of PD-L1 is mediated by clathrin-coated vesicle when PAI-1 triggers endocytic membrane trafficking by interacting with its endocytic receptor LRP1.
      Figure thumbnail gr4
      Figure 4Clathrin-mediated endocytosis contributes to the internalization of surface PD-L1. (a) Expression of clathrin gene, Cla; Lrp-1; or caveolin gene, Cav was silenced by specific siRNA-mediated knockdown in B16-F10 cells for 48 hours. Surface PD-L1 expression was analyzed by staining with an anti–PD-L1 antibody. The ratio of PD-L1 MFI to scr is shown in the histogram. (b) Cells were treated with either active or inactive Pitstop 2 at a dose of 10 μM for 24 hours, and PD-L1 expression was determined by flow cytometry. P-values were determined by two-tailed Student’s t-test. ∗P < 0.05, ∗∗P < 0.01. n = 3. (c) B16-F10 cells treated in (a) were subjected to confocal microscopic analyses by using anti–PD-L1 (green) and anti-LAMP1 (red) as the detection antibodies. Bar = 20 μm. MFI, mean fluorescence intensity; ns, not significant; scr, scramble control small interfering RNA; siCav, caveolin small interfering RNA; siCla, clathrin small interfering RNA; siLrp1, LRP1 interfering RNA; siRNA, small interfering RNA.
      Next, we tested whether PD-L1 internalization undergoes lysosome-mediated degradation after PAI-1–induced clathrin-coated pits formation. We investigated the intracellular colocalization of PD-L1 and lysosomes after inhibition of endocytosis routes by silencing of Cla, Lrp1, or Cav. Lysosomes were visualized by the detection of LAMP1 signals. Confocal immunofluorescence images showed that PD-L1 was distributed mainly in the cytoplasm and faintly expressed in the plasma membrane of nonsilenced cells. Colocalization signals of PD-L1 and LAMP1 were observed in the cytosolic compartment in scramble control and Cav knockdown cells, whereas significant plasma membrane localization of PD-L1 was detected in cells lacking clathrin or LRP1, which did not localize to lysosomal regions (Figure 4c). Together, these results suggest that clathrin and/or LRP1 play a key role in the endocytic trafficking of surface PD-L1 to the lysosomal regions.

      PAI-1–induced surface PD-L1 endocytosis is delivered by the endolysosomal system

      It has been reported that the tertiary urokinase plasminogen activator complex, composed of uPA, uPAR, and PAI-1, is internalized by clathrin-mediated endocytosis through interaction with LRP1. The resulting quaternary complex is then transported to early endosomes, and subsequently to late endosomes, where uPA–PAI-1 separates from uPAR. The uPA–PAI-1 complex–containing vesicles then fuse to lysosomes for PAI-1 degradation (
      • Conese M.
      • Nykjaer A.
      • Petersen C.M.
      • Cremona O.
      • Pardi R.
      • Andreasen P.A.
      • et al.
      Alpha-2 macroglobulin receptor/Ldl receptor-related protein(Lrp)-dependent internalization of the urokinase receptor.
      ;
      • Olson D.
      • Pöllänen J.
      • Høyer-Hansen G.
      • Rønne E.
      • Sakaguchi K.
      • Wun T.C.
      • et al.
      Internalization of the urokinase-plasminogen activator inhibitor type-1 complex is mediated by the urokinase receptor.
      ). Therefore, we investigated whether PAI-1–dependent PD-L1 internalization is transported through the endosomal membrane to the lysosomal pathway. Colocalization of PD-L1 with early and late endosomes, as determined by the markers EEA1 and Rab7, respectively, was examined by confocal immunofluorescence microscopy. As shown in Figure 5a, a portion of PD-L1 and EEA1 localized to the cytosolic compartment in scramble RNA–expressing groups. In contrast, after PAI-1 silencing, a marked increase of surface PD-L1 and no intracellular colocalization of PD-L1 and EEA1 were observed (Figure 5a). Similarly, cells with PAI-1 deficiency failed to identify colocalization of PD-L1 and Rab7 (Figure 5b). Moreover, decreased PD-L1 transport to lysosomes was detected after PAI-1 knockdown, as reduced colocalization signals of PD-L1 and LAMP1 were observed (Figure 5c). Similar colocalization of PD-L1 and adaptors of the endolysosomal system was found in human melanoma cells after SERPINE1 silencing (Supplementary Figure S6). We also treated cells with the lysosomal inhibitors chloroquine or bafilomycin A1 for 0, 0.5, and 2 hours and examined PD-L1 levels by western blotting. As expected, PD-L1 was accumulated in chloroquine- or bafilomycin A1–treated cells (Figure 5d). Taken together, these results suggest that PAI-1 stimulates the trafficking of PD-L1–containing endosomal vesicles to lysosomes for degradation.
      Figure thumbnail gr5
      Figure 5Lack of PAI-1 disrupts localization of PD-L1 in the endolysosome system. B16-F10 cells were transfected with scr or siPai-1 #1 for 48 h. The cellular localization of PD-L1 and (a) EEA1, (b) Rab7, and (c) LAMP1 were imaged and analyzed by confocal immunofluorescence microscopy. (d) Cells were either left untreated or exposed to 100 μM CQ (left) or 100 nM BafA1 (right) for 0, 0.5, or 2 h. The PD-L1 expression level was analyzed by western blots using a specific antibody against PD-L1. BafA1, bafilomycin A1; CQ, chloroquine; h, hour; scr, scramble control small interfering RNA; siPai-1, Pai-1 small interfering RNA.

      Targeting PAI-1 improves the antitumor activity of anti–PD-L1 therapy in a syngeneic mouse model of melanoma

      Next, we utilized a murine model with B16-F10–derived allografts to test whether TPX could improve the antitumor effects of the anti–PD-L1 antibody. C57BL/6 mice were inoculated with syngeneic B16-F10 cells. Therapeutic efficacy of TPX alone, or anti–PD-L1 antibody monotherapy, in tumor-bearing mice was rather modest, because only slight inhibition of primary tumor growth was observed compared with the control group (P = 0.048 for TPX administration; P = 0.04 for anti–PD-L1 treatment; Figure 6a). Of note, mice receiving the combination therapy of TPX and anti–PD-L1 antibody showed marked improvement in suppression of tumor growth compared with those treated either with TPX alone (P = 0.007) or anti–PD-L1 antibody alone (P = 0.018) (Figure 6a). These therapies caused neither notable side effects nor weight loss in the mice (Figure 6b). Moreover, mice treated with TPX alone and in combination with anti–PD-L1 antibody exhibited increased cell surface PD-L1 localization in the tumor bed, as detected by fluorescent immunohistochemical assay. In contrast, mice receiving administration of anti–PD-L1 antibody displayed similar PD-L1 expression to the control group in tumor regions (Figure 6c). Furthermore, tumor samples from untreated control or anti–PD-L1 monotherapy groups revealed a portion of PD-L1 in the lysosomes. Remarkably, staining of PD-L1 on the tumor cell surface was increased in mice receiving TPX alone or in combination with anti–PD-L1 antibody but did not target the lysosomal compartments (Figure 6d). Together, these findings suggest that inhibition of PAI-1–dependent PD-L1 endocytosis for blocking its lysosomal degradation enhances the tumor-suppressive activity of the anti–PD-L1 blockade.
      Figure thumbnail gr6
      Figure 6Inhibition of PAI-1 improves the therapeutic efficacy of anti–PD-L1 in tumor-bearing mice. (a) C57BL/6 mice were inoculated with B16-F10 cells subcutaneously. Mice were divided into four groups with PBS in combination with control antibody (IgG), TPX and IgG, anti–PD-L1 alone, or TPX and anti–PD-L1 on day 7 after inoculation. PBS and TPX were orally administered every 2 days. Anti–PD-L1 antibody was injected intraperitoneally twice a week. Tumor growth was monitored at the indicated time points. P-values were determined by two-tailed Student’s t-test. ∗P < 0.05, ∗∗P < 0.01. (b) Mice body weight in each group was measured every 2 days. n = 10 mice per group. (c) Tumor samples were collected on day 16 after inoculation, and the sections were stained with anti–PD-L1 antibodies followed by immunohistochemical analysis. (d) Tumor sections harvested from (c) were analyzed by fluorescence immunohistochemical analysis using anti–PD-L1 (green) and anti-LAMP1 (red) antibodies. Bar = 20 μm; insets, bar = 7.5 μm. ns, not significant; TPX, tiplaxtinin.

      Discussion

      The underlying mechanisms responsible for the regulation of cancer progression by PAI-1 are implemented not only through its stimulatory effect on growth and migration but also its role in anti-apoptotic regulation and proangiogenic activities (
      • Kubala M.H.
      • DeClerck Y.A.
      The plasminogen activator inhibitor-1 paradox in cancer: a mechanistic understanding.
      ). However, whether PAI-1 is involved in tumor immune surveillance remains unclear. In this study, we report that PAI-1 downregulates plasma membrane expression of PD-L1, a key factor involved in the immune evasion of tumor cells by conducting internalization of surface PD-L1 to lysosomal degradation. We observed that deficiency in PAI-1 activity by genetic knockdown or pharmacological inhibitor resulted in accumulation of PD-L1 on the plasma membrane in cell-based and animal studies. Elevated surface PD-L1 expression resulted from PAI-1 blocking, thereby improving the therapeutic efficacy of anti–PD-L1 blockade.
      It has been demonstrated that the therapeutic efficacy of immune checkpoint blockade–based immunotherapy relies on the expression of PD-L1 in melanoma cells (
      • Daud A.I.
      • Wolchok J.D.
      • Robert C.
      • Hwu W.J.
      • Weber J.S.
      • Ribas A.
      • et al.
      Programmed death-ligand 1 expression and response to the anti-programmed death 1 antibody pembrolizumab in melanoma.
      ;
      • Dong H.
      • Strome S.E.
      • Salomao D.R.
      • Tamura H.
      • Hirano F.
      • Flies D.B.
      • et al.
      Tumor-associated B7-H1 promotes T-cell apoptosis: a potential mechanism of immune evasion [published correction appears in Nat Med 2002;8:1039].
      ;
      • Taube J.M.
      • Anders R.A.
      • Young G.D.
      • Xu H.
      • Sharma R.
      • McMiller T.L.
      • et al.
      Colocalization of inflammatory response with B7-h1 expression in human melanocytic lesions supports an adaptive resistance mechanism of immune escape.
      ). Emerging evidence has also shed light on the significance of PD-L1 transport in antitumor immune response (
      • Deng S.
      • Zhou X.
      • Xu J.
      Checkpoints under traffic control: from and to organelles.
      ;
      • Eikawa S.
      • Nishida M.
      • Mizukami S.
      • Yamazaki C.
      • Nakayama E.
      • Udono H.
      Immune-mediated antitumor effect by type 2 diabetes drug, metformin.
      ;
      • Pereira F.V.
      • Melo A.C.L.
      • Low J.S.
      • de Castro Í.A.
      • Braga T.T.
      • Almeida D.C.
      • et al.
      Metformin exerts antitumor activity via induction of multiple death pathways in tumor cells and activation of a protective immune response.
      ). The widely expressed protein CMTM6 is essential for PD-L1 trafficking from the plasma membrane to recycling endosomes, a step that enhances PD-L1 expression at the cell surface and leads to the suppression of tumor-specific T-cell activity (
      • Burr M.L.
      • Sparbier C.E.
      • Chan Y.C.
      • Williamson J.C.
      • Woods K.
      • Beavis P.A.
      • et al.
      CMTM6 maintains the expression of PD-L1 and regulates anti-tumour immunity.
      ). Contrastingly, the expression of HIP1R protein acts as a negative regulator of surface PD-L1 expression by targeting endocytosed PD-L1 to the lysosomes for degradation through the lysosome-sorting sequence and the PD-L1–binding portion of HIP1R (
      • Wang H.
      • Yao H.
      • Li C.
      • Shi H.
      • Lan J.
      • Li Z.
      • et al.
      HIP1R targets PD-L1 to lysosomal degradation to alter T cell-mediated cytotoxicity.
      ). These findings suggest a tight regulation of cell surface PD-L1 by intracellular vesicle trafficking. Furthermore, PD-L1 is also reported to undergo CLA- but not CAV-mediated endocytosis (
      • Li C.W.
      • Lim S.O.
      • Chung E.M.
      • Kim Y.S.
      • Park A.H.
      • Yao J.
      • et al.
      Eradication of triple-negative breast cancer cells by targeting glycosylated PD-L1.
      ). In line with these observations, our results also showed that PAI-1 promotes PD-L1 internalization by inducing clathrin-coated endocytosis through endocytic guiding receptor protein LRP1 and targets PD-L1 to lysosomal degradation. Thus, our findings suggest a specific role for PAI-1–dependent clathrin-coated endosomes in the trafficking of PD-L1 to the lysosomes.
      Mounting evidence over the years has demonstrated that elevated PD-L1 expression at the plasma membrane contributes to the suppression of antitumor immunity in the tumor microenvironment. Although our results indicate that inhibition of PAI-1 activity results in the accumulation of PD-L1 at the cell surface both in vitro and in vivo, treatment with TPX alone still showed therapeutic efficacy in the mouse model of melanoma (Figure 6). The tumor growth inhibition through TPX treatment alone may be attributed to the dampening of the proangiogenesis and tumor growth stimulatory functions of PAI-1, which has been proposed as a protumoral factor in various cancer types (
      • Li S.
      • Wei X.
      • He J.
      • Tian X.
      • Yuan S.
      • Sun L.
      Plasminogen activator inhibitor-1 in cancer research.
      ). In melanoma studies, lack of PAI-1 activity in Pai-1 knockout mice has been known to suppress angiogenesis and restrain tumor growth, whereas mice overexpressing PAI-1 have increased tumor vasculature (
      • McMahon G.A.
      • Petitclerc E.
      • Stefansson S.
      • Smith E.
      • Wong M.K.
      • Westrick R.J.
      • et al.
      Plasminogen activator inhibitor-1 regulates tumor growth and angiogenesis.
      ). Administration of PAI-1 inhibitor SK-216 orally in a melanoma murine model reportedly blocks tumor angiogenesis and exerts an antitumor effect (
      • Masuda T.
      • Hattori N.
      • Senoo T.
      • Akita S.
      • Ishikawa N.
      • Fujitaka K.
      • et al.
      SK-216, an inhibitor of plasminogen activator inhibitor-1, limits tumor progression and angiogenesis.
      ). Evidence from preclinical murine models of cancer also supports the therapeutic efficacy of small-molecule PAI-1 inhibitors, at least to a limited extent (
      • Placencio V.R.
      • DeClerck Y.A.
      Plasminogen activator inhibitor-1 in cancer: rationale and insight for future therapeutic testing.
      ). Additionally, PAI-1 can also induce a type of neovascularization, termed amoeboid angiogenesis, by supporting the activation of the RhoA/Rho–associated protein kinase pathway, which is independent of protease activity (
      • Cartier-Michaud A.
      • Malo M.
      • Charrière-Bertrand C.
      • Gadea G.
      • Anguille C.
      • Supiramaniam A.
      • et al.
      Matrix-bound PAI-1 supports cell blebbing via RhoA/ROCK1 signaling.
      ;
      • Chillà A.
      • Margheri F.
      • Biagioni A.
      • Del Rosso M.
      • Fibbi G.
      • Laurenzana A.
      Mature and progenitor endothelial cells perform angiogenesis also under protease inhibition: the amoeboid angiogenesis.
      ). These results indicate the multifaceted role of PAI-1 in the regulation of tumor angiogenesis. In addition, active PAI-1 also promotes immunosuppression in the tumor microenvironment by modulation of macrophage phenotype. PAI-1 stimulation can enhance TGFβ expression in macrophages (
      • Zhu C.
      • Shen H.
      • Zhu L.
      • Zhao F.
      • Shu Y.
      Plasminogen activator inhibitor 1 promotes immunosuppression in human non-small cell lung cancers by enhancing TGF-β1 expression in macrophage.
      ). Recently, PAI-1 has been demonstrated to polarize macrophages toward the M2 phenotype to facilitate tumor progression, whereas PAI-1 deficiency is associated with decreased protumoral M2 macrophages in the tumor bed (
      • Kubala M.H.
      • Punj V.
      • Placencio-Hickok V.R.
      • Fang H.
      • Fernandez G.E.
      • Sposto R.
      • et al.
      Plasminogen activator inhibitor-1 promotes the recruitment and polarization of macrophages in cancer.
      ). Thus, an effective therapeutic strategy for targeting PAI-1 in cancer needs to be developed. Therefore, the combinatorial treatment of a PAI-1 inhibitor and anti–PD-L1 immunotherapy, as shown in this study, could achieve synergistic tumor inhibition.
      In summary, we have elucidated that PAI-1 downregulates the plasma membrane level of PD-L1 by conducting internalization of surface PD-L1 to lysosomal degradation. We have also observed that inducing deficiency in PAI-1 activity, either through genetic knockdown or pharmacological inhibition, resulted in the accumulation of PD-L1 on the plasma membrane in both cell-based and animal studies. Elevated surface PD-L1 expression resulted from PAI-1 blocking, thereby enhancing the therapeutic efficacy of anti–PD-L1 blockade. In addition, our results demonstrate a previously unreported regulation of PD-L1 exposure by inhibition of PAI-1–induced endocytosis. We thus provide an insight on the molecular basis of PAI-1–involved PD-L1 trafficking that rationalizes the combination of PAI-1 inhibition and immunotherapy as a promising strategy for overcoming resistance to anti–PD-L1 therapy in treating melanoma.

      Materials and Methods

      Cell culture, RNA interference, and transduction

      B16-F10 murine melanoma cell line was purchased and characterized at the Bioresource Collection and Research Center (Hsinchu, Taiwan). Human melanoma cell lines A375, SH4, SK-MEL-24, and RPMI-7951 were obtained from ATCC. The cell lines were cultured in DMEM (Gibco, Waltham, MA) supplemented with 10% fetal bovine serum (Gibco) and 1% penicillin/streptomycin (Gibco) at 37 °C in a humidifier with 5% carbon dioxide. For gene silencing, B16-F10 cells were transfected with the siRNA for knock down of murine Pai-1 at 20 nM or a dose indicated elsewhere (catalog number: 4390771, identification: s233904 [#1], identification: s233903 [#2]; Ambion, Thermo Fisher Scientific, Waltham, MA), Cla, Lrp1, or Cav (SMARTpool, Dharmacon, Lafayette, CO) by using RNAiMAX reagents (Thermo Fisher Scientific) for 48 hours, whereas A375 and SH4 cells were transduced by lentiviral-mediated short hairpin RNA, obtained from National RNAi Core Facility (Institute of Molecular Biology, Academia Sinica, Taipei, Taiwan), for targeting Pai-1 for 48 hours. Overexpression of PAI-1 in a plasmid containing FLAG-tagged murine Pai-1 cDNA (OriGene, Rockville, MA) was transfected for 24 hours using Lipofectamine 3000 (Thermo Fisher Scientific) according to the manufacturer’s instructions. After transfection/transduction, cell viability was analyzed by dye exclusion assay using trypan blue and was found to be >98%. For detecting secreted PAI-1 in the supernatants of Pai-1–silenced cells, media from siPai-1–transfected or short hairpin RNA–targeting cells were collected 48 hours after transfection/transduction and subjected to ELISA.

      Mice studies

      All experiments involving the use and care of mice were in compliance with the guidelines of the institutional committee. C57BL/6 mice were injected with B16-F10 tumor cells harboring scramble or siPai-1 expression (4 × 105) subcutaneously. Tumor volumes were determined using the equation: V = (a × b2)/2, where a and b represent the tumor length and width, respectively. The mice were killed after 16 days and tumor tissues were collected, fixed, and prepared for fluorescence immunohistochemistry staining. For therapeutic treatment, the mice were injected with B16-F10 cells subcutaneously, and at 7 days after inoculation, tumor-bearing mice were administrated 10 mg/kg TPX (Axon Medchem, Reston, VA) through oral gavage every 2 days or combined with an intraperitoneal injection of 200 μg anti–PD-L1 antibody or control IgG (Bio X Cell, Lebanon, NH) twice a week. Tumor growth was monitored every 2 days, and mice body weight was also recorded.

      Clinical specimens and statistical analysis

      We acquired 96 paraffin-embedded clinical melanoma sections from the tissue bank of Kaohsiung Chang Gung Memorial Hospital (Kaohsiung City, Taiwan) between 2008 and 2019 after receiving permission from the Institutional Review Board. All clinical specimens used were obtained per the guidelines of the Declaration of Helsinki. The distribution of age, sex, cancer staging, and with/without therapy is summarized in Supplementary Table S1. Among the 96 patients, 56 (58.3%) were older than 65 years, and 63 (65.6%) were men. A total of 26 (27.1%) were classified to be in advanced stage (stage III and stage IV, based on the 8th edition of the American Joint Committee on Cancer staging system), whereas 66 (68.8%) were classified to be in the early stages (stages I and II). More than 90% of the tumor biopsies used for immunohistochemical analysis were obtained from fresh cases (93.8%, Supplementary Table S1). Chi-square test was used for analysis of the correlation between PAI-1 and PD-L1 expression in melanoma tissues. Pearson chi-square test was used to compare the correlation of PAI-1/PD-L1 expression patterns and tumor stages. A two-tailed Student’s t-test was used for calculation of the three independent cell studies and animal studies unless indicated otherwise. Data were represented as mean ± SD. The level of statistical significance was taken as P-values: ∗P < 0.05, ∗∗P < 0.01, and ∗∗∗P < 0.001.
      A detailed description of materials and methods including Supplementary Table S2, which listed the primers used in this study, are described in the Supplementary Materials and Methods.

      Data availability statement

      The data supporting the findings of this study are available in this article and the supplementary material.

      Conflict of Interest

      The authors declare no conflict of interest.

      Acknowledgments

      This work was supported by the Taiwan Ministry of Science and Technology (Taipei, Taiwan; MOST 109-2320-B-182A-016 to HTT) and Chang Gung Medical Foundation (Taipei, Taiwan; grant number CMRPG8J1241-2 [HTT]). We thank Chang Gung Medical Foundation Kaohsiung Chang Gung Memorial Hospital Biobank and Tissue Bank Core Lab (CLRPG8L0081) for excellent technical support. We also thank Yu-Shin Tu for her technical support.

      Author Contributions

      Data Curation: HTT; Formal Analysis: WYC, JLY, HTT, YJT; Funding Acquisition: HTT; Investigation: WYC, JLY, HTT, YJT; Methodology: HTT; Project Administration: HTT; Writing – Original Draft Preparation: HTT, YJT, CHL; Writing - Review and Editing: CHL

      Supplementary Materials and Methods

      Cell treatments

      B16-F10 cells were transfected with scramble control small interfering RNA or small interfering RNA targeting Pai-1 at 20 nM for 48 hours followed by treatment with a recombinant active fraction of murine PAI-1 (Molecular Innovations, Southfield, MI) at increasing doses of 0–20 μg/ml for another 4 hours. Biotin-labeled BSA (Sigma-Aldrich, St. Louis, MO) and PAI-1 protein (Molecular Innovations) were used to treat B16-F10 cells at 10 μg/ml for 4 hours. Inhibitors for blocking vesicle trafficking, such as brefeldin A (BioLegend, San Diego, CA), nocodazole (Sigma-Aldrich), cytochalasin D (Sigma-Aldrich), chloroquine (Sigma-Aldrich), and bafilomycin A1 (Sigma-Aldrich), were used to treat B16-F10 cells at 5 μg/ml (brefeldin A), 2.5 μg/ml (cytochalasin D), 10 μM (nocodazole), 20 nM (bafilomycin A1), and 50 μM (chloroquine), whereas tiplaxtinin (Axon Medchem, Reston, VA) was used at 28–56 μM for 24 hours. Pitstop 2 and inactive Pitstop 2 were purchased from Abcam (Cambridge, United Kingdom) and used at 10 μM for 24 hours. Prochlorperazine was obtained from Sigma-Aldrich and used at the indicated doses.

      ELISA

      Supernatants from cells with in vitro treatments were analyzed for expression levels of murine PAI-1/human PAI-1 by using DuoSet ELISA Assay (R&D Systems, Minneapolis, MI) according to the manufacturer’s protocol.

      Flow cytometry analysis

      Cells were harvested by trypsin digestion, washed with PBS buffer, and further washed twice with FACS buffer (PBS, 2% v/w fetal bovine serum). Cells were labeled with fluorescent dye–conjugated anti–PD-L1 antibody or isotype-matched control antibody (BioLegend) in FACS buffer for 30 minutes at 4 °C and analyzed by flow cytometry for characterization of PD-L1–positive cells. The mean fluorescence intensity of PD-L1 relative to isotype control was calculated as the mean fluorescence of PD-L1–positive cells by using FlowJo software.

      RNA extraction and quantitative real-time reverse transcriptase–PCR assay

      Total RNA was extracted using Trizol reagent (Invitrogen, Foster City, CA), and the purification of total RNA was carried out using Direct-zol RNA Miniprep Kit (Zymo Research, Irvine, CA) according to the manufacturer’s instructions. cDNA was generated using the High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific, Waltham, MA). qPCR reactions were performed to analyze the mRNA expression of target genes using SYBR Green Master Mix (Invitrogen). Results were normalized to the housekeeping gene GADPH. The primers used for quantitative real-time reverse transcriptase–PCR analysis are listed in Supplementary Table S2.

      Western blotting

      Cell lysates were extracted and quantified by Bradford assay (Bio-Rad, Hercules, CA). Equal proteins were loaded onto a 10% SDS gel for PAGE and then transferred onto a polyvinylidene fluoride membrane. The membrane was blocked with a blocking buffer (3% BSA in PBS) for 1 hour. Primary antibodies were incubated with the membrane overnight at 4 °C. After washing with 0.05% Triton-X100 in PBS, the membrane was incubated with horseradish peroxidase–conjugated secondary antibodies for 1 hour at 25 °C, and the signals were detected using the Western Lightning (PerkinElmer, Waltham, MA). The antibodies used for western blotting are listed as follows: anti-GADPH (1:2,000, Merck Millipore, Burlington, MA), anti–PD-L1 (1:1,000, Proteintech, Rosemont, IL), anti-caveolin (1:1,000, Cell Signaling Technology, Danvers, MA), anti-clathrin (1:1,000, Abcam), and anti-LRP1 (1:1,000, Abcam).

      Immunofluorescence assay and immunohistochemistry staining

      For immunofluorescence staining, the cells were grown on chamber slides, and the slides were fixed and permeabilized. After blocking, the slides were incubated with primary antibodies against PD-L1 (1:200, Proteintech), PAI-1 (1:200, Novus Biologicals), EEA1 (1:100, Cell Signaling Technology), Rab7 (1:100, Cell Signaling Technology), or LAMP1 (1:100, Santa Cruz Biotechnology, Dallas, TX). The stained slides were incubated with fluorescent dye–conjugated secondary antibodies along with DAPI staining as the counterstain and analyzed by confocal microscopy. For immunohistochemistry staining, after the mice were killed, the allografts were removed and tumor samples were sectioned after 4% paraformaldehyde fixation. The tumor sections were deparaffinized, rehydrated, and antigen retrieved, then permeabilized with 0.5% Triton X-100, after which the sections were incubated in blocking solution followed by staining with primary antibody against PD-L1 (1:500, Proteintech) at 4 ° overnight. Diaminobenzidine tetrahydrochloride solution was used to detect peroxidase activity and counterstained by hematoxylin staining. Images acquirement were performed by the Vectra Polaris instrument and quantified and analyzed by the Phenochart software (PerkinElmer).
      Figure thumbnail fx1
      Supplementary Figure S1PD-L1 expression in cells with manipulation of PAI-1 expression. B16-F10 cells were transfected with either (a) FLAG-tagged PAI-1 cDNA for 24 hours or (b) siPai-1 #1 and #2 as well as scr at 20 nM for 48 hours. The RNA expression levels of Pai-1 and Pd-l1 were determined by QRT-PCR assays. The fold change of genes was calculated by using Gapdh as the internal control. P-values were determined by two-tailed Student’s t-test. ∗∗P < 0.01, ∗∗∗P < 0.001. (c) Expression of PAI-1 in the supernatants of transfected cells from (b) was determined using ELISA. (d) The cells from (b) were lysed to extract proteins for western blotting analysis using anti–PD-L1 as the detection antibody. Data are shown as mean ± SD of three independent experiments. ctrl, control; ns, not significant; QRT-PCR, quantitative real-time reverse transcriptase–PCR; scr, scramble control small interfering RNA; siPai-1, Pai-1 small interfering RNA.
      Figure thumbnail fx2
      Supplementary Figure S2Decrease in extracellular PAI-1 correlates with increase in surface PD-L1. B16-F10 cells were transfected with scr, (a) siPai-1 #1 or (b) #2 with the indicated doses for 48 hours. The level of PAI-1 in supernatants was determined by ELISA. The fold change of PAI-1 levels is relative to that of scr. The percentage of PD-L1 expression on the plasma membrane was measured by flow cytometry and analyzed by FlowJo software (upper panel). The fold change of PD-L1 MFI is relative to that of scr. Correlation between soluble PAI-1 and surface PD-L1 was analyzed by linear regression (lower panel). MFI, mean fluorescence intensity; scr, scramble control small interfering RNA; siPai-1, Pai-1 small interfering RNA.
      Figure thumbnail fx3
      Supplementary Figure S3Cell surface expression of PD-L1 in PAI-1–deficient human melanoma cells. (a) A375 and (b) SH4 cells were transduced by lentivirus harboring shRNA targeting SERPINE1 for 48 hours. Extracellular PAI-1 level was determined by ELISA. The fold change of soluble PAI-1 is relative to that of scr. Flow cytometric analysis of PD-L1 expression at the plasma membrane of (c) A375 and (d) SH4 cells is shown. The fold change of PD-L1 MFI is relative to that of scr (right panel). Data are presented as mean ± SD of three independent experiments. P-values were determined by two-tailed Student’s t-test. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001. MFI, mean fluorescence intensity; scr, scramble control small interfering RNA; shPAI-1, PAI-1 short hairpin RNA; shRNA, short hairpin RNA.
      Figure thumbnail fx4
      Supplementary Figure S4Inhibition of PAI-1 enhances cell surface PD-L1 expression in human melanoma cells. (a) A375, (b) SK-MEL-24, and (c) RPMI-7951 cells were treated with TPX for 24 hours at the indicated doses. PD-L1 level on the cell surface was determined by flow cytometry. P-values were determined by two-tailed Student’s t-test. ∗P < 0.05, ∗∗P < 0.01. n = 3. ctrl, control; TPX, tiplaxtinin.
      Figure thumbnail fx5
      Supplementary Figure S5PD-L1 expression in cells with knock down of proteins responsible for endocytosis. B16-F10 cells were transfected with scr or (a) siCav, (b) siCla, or (c) siLrp1 for 48 hours. The protein expression of whole-cell lysates was detected by western blotting using the respective antibodies. (d) Cell lysates with knock down of Pai-1; clathrin gene, Cla; Lrp1; or caveolin gene, Cav were used to perform western blotting for detection of PD-L1 expression. scr, scramble control small interfering RNA; siCav, caveolin small interfering RNA; siCla, clathrin small interfering RNA; siLrp1, Lrp1 small interfering RNA.
      Figure thumbnail fx6
      Supplementary Figure S6PAI-1 colocalizes with PD-L1 in the endolysosomal system in human melanoma cells. (a–c) RPMI-7951 cells transduced with lentiviral-mediated scr or SERPINE1 knockdown for 48 hours were conducted to stain PD-L1, (a) EEA1, (b) Rab7, or (c) LAMP1 using specific antibodies. Images were acquired and analyzed by confocal microscopy. Green: PD-L1 staining; red: EEA1, Rab7, and LAMP1 staining; blue: nucleus staining. Bar = 20 μm. scr, scramble control small interfering RNA; shPAI-1, PAI-1 short hairpin RNA; shRNA, short hairpin RNA.
      Supplementary Table S1PAI-1/D-L1 Expression Pattern with Clinical Characteristics of the Patients with Melanoma Included in This Study
      CharacteristicsPatient DistributionPAI-1/PD-L1
      N = 96 (100%)
      The total number of samples in the stage category is less than the overall number analyzed because clinical data were not available for these samples.
      +/–

      n = 32 (33.3%)
      The total number of samples in the stage category is less than the overall number analyzed because clinical data were not available for these samples.
      +/–

      n = 25 (26%)
      Age, n (%)
       <65 years old40 (41.7)
       ≥65 years old56 (58.3)
      Gender, n (%)
       Female33 (34.4)
       Male63 (65.6)
      Stage
      Clinical stage was classified according to the American Joint Committee on Cancer tumor-node-metastasis classification, 8th edition.
      , n (%)
       Early (I + II)66 (68.8)26 (39.4%)12 (19.7%) 0.001
       Late (III + IV)26 (27.1)3 (11.5%)13 (50.0%)
      Fresh case
      Patients receiving chemotherapy, IL-2-based immunotherapy, or targeted therapy were included in the No category of the Fresh case.
      , n (%)
       Yes90 (93.8)
       No6 (6.3)
      The data were analyzed using Pearson’s Chi-square test, and P-value is shown as superscript.
      a The total number of samples in the stage category is less than the overall number analyzed because clinical data were not available for these samples.
      b Clinical stage was classified according to the American Joint Committee on Cancer tumor-node-metastasis classification, 8th edition.
      c Patients receiving chemotherapy, IL-2-based immunotherapy, or targeted therapy were included in the No category of the Fresh case.
      Supplementary Table S2Primers Used in This Study
      GenePrimerSequences (5′–3′)
      Cd274ForwardGCTCCAAAGGACTTGTACGTG
      ReverseTGATCTGAAGGGCAGCATTTC
      Pai-1ForwardGACACCCTCAGCATGTTCATC
      ReverseAGGGTTGCACTAAACATGTCAG
      GapdhForwardTTGTCTCCTGCGACTTCA
      ReverseCACCACCCTGTTGCTGTA

      References

        • Bajou K.
        • Masson V.
        • Gerard R.D.
        • Schmitt P.M.
        • Albert V.
        • Praus M.
        • et al.
        The plasminogen activator inhibitor PAI-1 controls in vivo tumor vascularization by interaction with proteases, not vitronectin. Implications for antiangiogenic strategies.
        J Cell Biol. 2001; 152: 777-784
        • Bajou K.
        • Peng H.
        • Laug W.E.
        • Maillard C.
        • Noel A.
        • Foidart J.M.
        • et al.
        Plasminogen activator inhibitor-1 protects endothelial cells from FasL-mediated apoptosis.
        Cancer Cell. 2008; 14: 324-334
        • Burr M.L.
        • Sparbier C.E.
        • Chan Y.C.
        • Williamson J.C.
        • Woods K.
        • Beavis P.A.
        • et al.
        CMTM6 maintains the expression of PD-L1 and regulates anti-tumour immunity.
        Nature. 2017; 549: 101-105
        • Cao C.
        • Lawrence D.A.
        • Li Y.
        • Von Arnim C.A.
        • Herz J.
        • Su E.J.
        • et al.
        Endocytic receptor LRP together with tPA and PAI-1 coordinates Mac-1-dependent macrophage migration.
        EMBO J. 2006; 25: 1860-1870
        • Cartier-Michaud A.
        • Malo M.
        • Charrière-Bertrand C.
        • Gadea G.
        • Anguille C.
        • Supiramaniam A.
        • et al.
        Matrix-bound PAI-1 supports cell blebbing via RhoA/ROCK1 signaling.
        PloS One. 2012; 7: e32204
        • Chew H.Y.
        • De Lima P.O.
        • Gonzalez Cruz J.L.
        • Banushi B.
        • Echejoh G.
        • Hu L.
        • et al.
        Endocytosis inhibition in humans to improve responses to ADCC-mediating antibodies.
        Cell. 2020; 180: 895-914.e27
        • Chillà A.
        • Margheri F.
        • Biagioni A.
        • Del Rosso M.
        • Fibbi G.
        • Laurenzana A.
        Mature and progenitor endothelial cells perform angiogenesis also under protease inhibition: the amoeboid angiogenesis.
        J Exp Clin Cancer Res. 2018; 37: 74
        • Conese M.
        • Nykjaer A.
        • Petersen C.M.
        • Cremona O.
        • Pardi R.
        • Andreasen P.A.
        • et al.
        Alpha-2 macroglobulin receptor/Ldl receptor-related protein(Lrp)-dependent internalization of the urokinase receptor.
        J Cell Biol. 1995; 131: 1609-1622
        • Daniel J.A.
        • Chau N.
        • Abdel-Hamid M.K.
        • Hu L.
        • von Kleist L.
        • Whiting A.
        • et al.
        Phenothiazine-derived antipsychotic drugs inhibit dynamin and clathrin-mediated endocytosis.
        Traffic. 2015; 16: 635-654
        • Daud A.I.
        • Wolchok J.D.
        • Robert C.
        • Hwu W.J.
        • Weber J.S.
        • Ribas A.
        • et al.
        Programmed death-ligand 1 expression and response to the anti-programmed death 1 antibody pembrolizumab in melanoma.
        J Clin Oncol. 2016; 34: 4102-4109
        • Deng S.
        • Zhou X.
        • Xu J.
        Checkpoints under traffic control: from and to organelles.
        Adv Exp Med Biol. 2020; 1248: 431-453
        • Dong H.
        • Strome S.E.
        • Salomao D.R.
        • Tamura H.
        • Hirano F.
        • Flies D.B.
        • et al.
        Tumor-associated B7-H1 promotes T-cell apoptosis: a potential mechanism of immune evasion [published correction appears in Nat Med 2002;8:1039].
        Nat Med. 2002; 8: 793-800
        • Eikawa S.
        • Nishida M.
        • Mizukami S.
        • Yamazaki C.
        • Nakayama E.
        • Udono H.
        Immune-mediated antitumor effect by type 2 diabetes drug, metformin.
        Proc Natl Acad Sci USA. 2015; 112: 1809-1814
        • Fang H.
        • Placencio V.R.
        • DeClerck Y.A.
        Protumorigenic activity of plasminogen activator inhibitor-1 through an antiapoptotic function.
        J Natl Cancer Inst. 2012; 104: 1470-1484
        • Guo L.
        • Wei R.
        • Lin Y.
        • Kwok H.F.
        Clinical and recent patents applications of PD-1/PD-L1 targeting immunotherapy in cancer treatment-current progress, strategy, and future perspective.
        Front Immunol. 2020; 11: 1508
        • Hennan J.K.
        • Morgan G.A.
        • Swillo R.E.
        • Antrilli T.M.
        • Mugford C.
        • Vlasuk G.P.
        • et al.
        Effect of tiplaxtinin (PAI-039), an orally bioavailable PAI-1 antagonist, in a rat model of thrombosis.
        J Thromb Haemost. 2008; 6: 1558-1564
        • Isogai C.
        • Laug W.E.
        • Shimada H.
        • Declerck P.J.
        • Stins M.F.
        • Durden D.L.
        • et al.
        Plasminogen activator inhibitor-1 promotes angiogenesis by stimulating endothelial cell migration toward fibronectin.
        Cancer Res. 2001; 61: 5587-5594
        • Jaiswal R.K.
        • Varshney A.K.
        • Yadava P.K.
        Diversity and functional evolution of the plasminogen activator system.
        Biomed Pharmacother. 2018; 98: 886-898
        • Joseph S.R.
        • Gaffney D.
        • Barry R.
        • Hu L.
        • Banushi B.
        • Wells J.W.
        • et al.
        An ex vivo human tumor assay shows distinct patterns of EGFR trafficking in squamous cell carcinoma correlating to therapeutic outcomes.
        J Invest Dermatol. 2019; 139: 213-223
        • Kubala M.H.
        • DeClerck Y.A.
        The plasminogen activator inhibitor-1 paradox in cancer: a mechanistic understanding.
        Cancer Metastasis Rev. 2019; 38: 483-492
        • Kubala M.H.
        • Punj V.
        • Placencio-Hickok V.R.
        • Fang H.
        • Fernandez G.E.
        • Sposto R.
        • et al.
        Plasminogen activator inhibitor-1 promotes the recruitment and polarization of macrophages in cancer.
        Cell Rep. 2018; 25: 2177-2191.e7
        • Larkin J.
        • Chiarion-Sileni V.
        • Gonzalez R.
        • Grob J.J.
        • Cowey C.L.
        • Lao C.D.
        • et al.
        Combined nivolumab and ipilimumab or monotherapy in untreated melanoma [published correction appears in N Engl J Med 2018;379:2185].
        N Engl J Med. 2015; 373: 23-34
        • Li C.W.
        • Lim S.O.
        • Chung E.M.
        • Kim Y.S.
        • Park A.H.
        • Yao J.
        • et al.
        Eradication of triple-negative breast cancer cells by targeting glycosylated PD-L1.
        Cancer Cell. 2018; 33: 187-201.e10
        • Li S.
        • Wei X.
        • He J.
        • Tian X.
        • Yuan S.
        • Sun L.
        Plasminogen activator inhibitor-1 in cancer research.
        Biomed Pharmacother. 2018; 105: 83-94
        • Manzanares D.
        • Ceña V.
        Endocytosis: the nanoparticle and submicron nanocompounds gateway into the cell.
        Pharmaceutics. 2020; 12: 371
        • Masuda T.
        • Hattori N.
        • Senoo T.
        • Akita S.
        • Ishikawa N.
        • Fujitaka K.
        • et al.
        SK-216, an inhibitor of plasminogen activator inhibitor-1, limits tumor progression and angiogenesis.
        Mol Cancer Ther. 2013; 12: 2378-2388
        • McMahon G.A.
        • Petitclerc E.
        • Stefansson S.
        • Smith E.
        • Wong M.K.
        • Westrick R.J.
        • et al.
        Plasminogen activator inhibitor-1 regulates tumor growth and angiogenesis.
        J Biol Chem. 2001; 276: 33964-33968
        • Mellman I.
        • Yarden Y.
        Endocytosis and cancer.
        Cold Spring Harb Perspect Biol. 2013; 5: a016949
        • Nowicki T.S.
        • Hu-Lieskovan S.
        • Ribas A.
        Mechanisms of resistance to PD-1 and PD-L1 blockade.
        Cancer J. 2018; 24: 47-53
        • Olson D.
        • Pöllänen J.
        • Høyer-Hansen G.
        • Rønne E.
        • Sakaguchi K.
        • Wun T.C.
        • et al.
        Internalization of the urokinase-plasminogen activator inhibitor type-1 complex is mediated by the urokinase receptor.
        J Biol Chem. 1992; 267: 9129-9133
        • Pasupuleti N.
        • Grodzki A.C.
        • Gorin F.
        Mis-trafficking of endosomal urokinase proteins triggers drug-induced glioma nonapoptotic cell death.
        Mol Pharmacol. 2015; 87: 683-696
        • Patel S.P.
        • Kurzrock R.
        PD-L1 expression as a predictive biomarker in cancer immunotherapy.
        Mol Cancer Ther. 2015; 14: 847-856
        • Patrinely Jr., J.R.
        • Dewan A.K.
        • Johnson D.B.
        The role of anti-PD-1/PD-L1 in the treatment of skin cancer.
        BioDrugs. 2020; 34: 495-503
        • Pereira F.V.
        • Melo A.C.L.
        • Low J.S.
        • de Castro Í.A.
        • Braga T.T.
        • Almeida D.C.
        • et al.
        Metformin exerts antitumor activity via induction of multiple death pathways in tumor cells and activation of a protective immune response.
        Oncotarget. 2018; 9: 25808-25825
        • Placencio V.R.
        • DeClerck Y.A.
        Plasminogen activator inhibitor-1 in cancer: rationale and insight for future therapeutic testing.
        Cancer Res. 2015; 75: 2969-2974
        • Taube J.M.
        • Anders R.A.
        • Young G.D.
        • Xu H.
        • Sharma R.
        • McMiller T.L.
        • et al.
        Colocalization of inflammatory response with B7-h1 expression in human melanocytic lesions supports an adaptive resistance mechanism of immune escape.
        Sci Transl Med. 2012; 4: 127ra37
        • Wang D.Y.
        • Eroglu Z.
        • Ozgun A.
        • Leger P.D.
        • Zhao S.
        • Ye F.
        • et al.
        Clinical features of acquired resistance to anti-PD-1 therapy in advanced melanoma.
        Cancer Immunol Res. 2017; 5: 357-362
        • Wang H.
        • Yao H.
        • Li C.
        • Shi H.
        • Lan J.
        • Li Z.
        • et al.
        HIP1R targets PD-L1 to lysosomal degradation to alter T cell-mediated cytotoxicity.
        Nat Chem Biol. 2019; 15: 42-50
        • Zhang J.Y.
        • Yan Y.Y.
        • Li J.J.
        • Adhikari R.
        • Fu L.W.
        PD-1/PD-L1 based combinational cancer therapy: icing on the cake.
        Front Pharmacol. 2020; 11: 722
        • Zhu C.
        • Shen H.
        • Zhu L.
        • Zhao F.
        • Shu Y.
        Plasminogen activator inhibitor 1 promotes immunosuppression in human non-small cell lung cancers by enhancing TGF-β1 expression in macrophage.
        Cell Physiol Biochem. 2017; 44: 2201-2211

      Linked Article

      • A Possible Role for PAI-1 Blockade in Melanoma Immunotherapy
        Journal of Investigative DermatologyVol. 141Issue 11
        • Preview
          In their new article in the Journal of Investigative Dermatology, Tseng et al. (2021) confirm that the sensitivity of melanoma cells to anti‒PD-L1 checkpoint inhibitor therapy is correlated with high PD-L1 surface expression. By blocking PD-L1 membrane clearing, controlled by LRP1 and PAI-1, the expression of high-cell-surface levels of PD-L1 was maintained.
        • Full-Text
        • PDF
        Open Archive