Research ArticleCell Biology

A defect in KCa3.1 channel activity limits the ability of CD8+ T cells from cancer patients to infiltrate an adenosine-rich microenvironment

See allHide authors and affiliations

Science Signaling  24 Apr 2018:
Vol. 11, Issue 527, eaaq1616
DOI: 10.1126/scisignal.aaq1616

Reduced K+ channel activity curbs T cell migration

T cell accumulation in solid tumors is limited by multiple factors found within the tumor microenvironment, including the nucleoside adenosine. Chimote et al. analyzed the migration of CD8+ T cells in a 3D chemotaxis assay and found that adenosine inhibited the chemotaxis of T cells from cancer patients more than T cells from healthy donors. The increased sensitivity of patient CD8+ T cells to adenosine correlated with reduced KCa3.1 potassium (K+) channel activity, but not adenosine receptor expression or signaling. Treatment with a KCa3.1 channel agonist restored patient CD8+ T cell migration in the presence of adenosine, suggesting that K+ channel activators may help augment T cell infiltration of adenosine-rich solid tumors.

Abstract

The limited ability of cytotoxic T cells to infiltrate solid tumors hampers immune surveillance and the efficacy of immunotherapies in cancer. Adenosine accumulates in solid tumors and inhibits tumor-specific T cells. Adenosine inhibits T cell motility through the A2A receptor (A2AR) and suppression of KCa3.1 channels. We conducted three-dimensional chemotaxis experiments to elucidate the effect of adenosine on the migration of peripheral blood CD8+ T cells from head and neck squamous cell carcinoma (HNSCC) patients. The chemotaxis of HNSCC CD8+ T cells was reduced in the presence of adenosine, and the effect was greater on HNSCC CD8+ T cells than on healthy donor (HD) CD8+ T cells. This response correlated with the inability of CD8+ T cells to infiltrate tumors. The effect of adenosine was mimicked by an A2AR agonist and prevented by an A2AR antagonist. We found no differences in A2AR expression, 3′,5′-cyclic adenosine monophosphate abundance, or protein kinase A type 1 activity between HNSCC and HD CD8+ T cells. We instead detected a decrease in KCa3.1 channel activity, but not expression, in HNSCC CD8+ T cells. Activation of KCa3.1 channels by 1-EBIO restored the ability of HNSCC CD8+ T cells to chemotax in the presence of adenosine. Our data highlight the mechanism underlying the increased sensitivity of HNSCC CD8+ T cells to adenosine and the potential therapeutic benefit of KCa3.1 channel activators, which could increase infiltration of these T cells into tumors.

INTRODUCTION

The immune system plays an important role in cancer. In many solid malignancies, including head and neck squamous cell carcinoma (HNSCC), an increased infiltration of cytotoxic CD8+ T cells into the tumor mass is often associated with good prognosis and response to therapy (13). This knowledge is indeed at the foundation of immune therapies that increase the number and functionality of cytotoxic tumor-infiltrating lymphocytes (TILs). Adoptive T cell (ATC) transfer, chimeric antigen receptor (CAR) T cells, and checkpoint inhibitors have shown promising results in many forms of cancer. Although these therapies are very effective in increasing the functional capabilities of T cells, the modified T cells still maintain a limited ability to infiltrate the tumor mass and resist the immunosuppressive tumor microenvironment (TME) (47). The inability of tumor-specific T cells to traffic to a solid tumor represents a great challenge for effective immunotherapy. The unique features of the TME contribute to the reduced infiltration and functionality of TILs (8). Thus, understanding how the TME limits T cell infiltration is necessary for improving immune surveillance in cancer and developing effective immunotherapies.

The purine nucleoside adenosine accumulates in the TME and has been associated with tumor progression, enhanced metastatic potential, and poor prognosis (911). In vivo studies provide conclusive evidence of the importance of adenosine in cancer (1215). Abrogation of the adenosine signaling pathway, either through knockdown of the adenosine A2A receptor (A2AR), a G protein (heterotrimeric guanine nucleotide–binding protein)–coupled receptor (GPCR) expressed in immune cells, or by A2AR antagonists, reduces tumor burden in tumor-bearing mice, increases survival, and increases the efficacy of immunotherapies (5, 6, 9, 1618). Furthermore, knockdown of CD73, an enzyme necessary for adenosine production, completely restores the efficacy of ATC therapies and leads to long-term tumor-free survival of tumor-bearing mice (19, 20). Adenosine is thus emerging as an important checkpoint inhibitor of the antitumor T cell response (21). In addition, we have shown that adenosine limits cytokine release and motility in human peripheral blood T lymphocytes through calcium-activated KCa3.1 potassium (K+) channels (22).

Ion channels regulate multiple functions of T lymphocytes including cytokine, granzyme B production, and motility (2326). Two K+ channels, the voltage-dependent Kv1.3 and the Ca2+-activated KCa3.1, regulate the electrochemical driving force for Ca2+ influx that is necessary for nuclear factor of activated T cells nuclear translocation, gene expression, and effector functions (26). These two channels also mediate the response to two key immune suppressive elements of the TME: hypoxia (Kv1.3) and adenosine (KCa3.1) (22, 2729). Defects in Kv1.3 channels have been reported in TILs and are associated with their diminished cytotoxicity (30). The importance of K+ channels of T lymphocytes in cancer was confirmed in mice where overexpression of the Kv1.3 channel increased interferon-γ (IFN-γ) production, reduced tumor burden, and increased survival (31, 32). We have shown that in human T lymphocytes, KCa3.1 channels reside at the uropod of polarized mobile T cells and mediate the inhibitory effect of adenosine (22, 24). Adenosine, through A2AR, stimulates 3′,5′-cyclic adenosine monophosphate (cAMP) production and protein kinase A type 1 (PKAI) activation, inhibits KCa3.1 channels, and suppresses T cell motility (22). We speculated that this mechanism could have important implications in the ability of effector T cells to infiltrate the tumor mass. Furthermore, it may be particularly important in HNSCC, where effector T cells are more sensitive to adenosine than are their healthy counterparts; that is, adenosine inhibits proliferation and cytokine release more in HNSCC effector T cells than in healthy donor (HD) cells (33). This enhanced sensitivity has been attributed to a reduction in adenosine deaminase activity and increased A2AR signaling. To date, the chemotactic abilities of HNSCC CD8+ T cells and their response to adenosine have not been studied. Here, we investigated the effect of adenosine on the chemotaxis of circulating CD8+ T cells of HNSCC patients and the mechanisms that mediate their heightened response to adenosine. We provide evidence of a role for KCa3.1 channels in the adenosine-mediated suppression of the chemotaxis of HNSCC CD8+ T cells and suggest that KCa3.1 channels may have therapeutic potential to increase the ability of HNSCC CD8+ T cells to infiltrate an adenosine-rich TME.

RESULTS

The chemotaxis of circulating CD8+ T cells of HNSCC patients is impaired by adenosine

The TME is characterized by rapidly dividing tumor cells in a fibrous matrix with a variable degree of immune cell infiltrate. Adenosine accumulates in the hypoxic TME, and the increased adenosine concentration contributes to the inhibition of the antitumor immune response by cytotoxic CD8+ T cells (10, 34). Experiments were performed to assess whether adenosine had any effect on the chemotaxis of HNSCC CD8+ T cells (for the demographic and clinicopathologic features of the patients, see Table 1 and table S1, respectively). We studied the effect of adenosine on the chemotaxis of activated HNSCC CD8+ T cells in a tumor-like experimental setting and compared it to that of CD8+ T cells from HDs. To mimic the TME, we used the μ-Slide Chemotaxis chamber, which enables the generation of a stable, three-dimensional (3D) collagenous matrix in which CD8+ T cells migrate in response to a chemokine gradient (35, 36). All chemotaxis experiments were conducted with CXCL12 (37). Also, all experiments were conducted on CD8+ T cells activated in vitro with anti-CD3 and anti-CD28 antibodies for 3 to 4 days, unless otherwise specified. In the absence of CXCL12, cells exhibited random migration within the collagen matrix, whereas in the presence of a CXCL12 gradient, the cells migrated toward the highest CXCL12 concentration (Fig. 1, A and B, fig. S1, and table S2). We observed no significant differences in the baseline chemotactic response of HD and HNSCC CD8+ T cells as indicated by similar values for Y-COM (y coordinate of the center of mass, that is, the averaged position the cells achieved at the end of the experiment; red triangles in Fig. 1 and fig. S1), FMIy (y coordinate of the forward migration index), velocity, accumulated distance, directness, and Euclidean distance (table S2). Hereafter, Y-COM is used to quantify a chemotactic response. To evaluate what effect adenosine had on the chemotaxis of HD and HNSCC CD8+ T cells, we measured CXCL12-driven chemotaxis in the presence or absence of a concomitant adenosine gradient. Adenosine inhibited chemotaxis and this effect was significantly more pronounced in HNSCC CD8+ T cells than in HD T cells (Fig. 1, A and B). Adenosine decreased Y-COM in cells from five of seven HDs (Fig. 1C). In contrast, HNSCC CD8+ T cells displayed greater sensitivity to adenosine than did their healthy counterparts (Fig. 1D). Activated T cells from HDs exhibited 26% overall reduction in the Y-COM values, whereas adenosine inhibited the Y-COM values of HNSCC donor cells by 80% (Fig. 1E). Overall, HNSCC cells lost their chemotactic ability in the presence of adenosine, as indicated by loss of significant differences between Y-COM and X-COM values (the COM y and x coordinates), and between FMIy and FMIx (FMI in the x and y directions; Fig. 1B and Table 2). Note that adenosine did not induce any significant changes in other parameters that define T cell migration, such as cell velocity, directness, or accumulated distance in either HD or HNSCC CD8+ T cells (Table 2). Overall, these data show that adenosine inhibits the chemotaxis of HNSCC CD8+ T cells. To test whether adenosine accumulation might explain the inability of CD8+ T cells to infiltrate the adenosine-rich TME, we performed immunohistochemical staining of HNSCC tumors for CD8 and CD73 (ecto-5′-nucleotidase), an enzyme responsible for adenosine production that is used as a marker of adenosine abundance in solid tumors (19). We found a mixed degree of CD8+ infiltration and CD73 expression in HNSCC tumors (Fig. 2, A and B, and table S1). However, in the small cohort of patients with high intratumoral CD73 expression (n = 9), there was a negative correlation between the effect of adenosine on chemotaxis in vitro and CD8+ T cell tumor infiltration, that is, the patients whose CD8+ T cells chemotaxis was most inhibited by adenosine were also those that had the lowest CD8+ T cell infiltration into the tumor (Fig. 2C and table S1).

Table 1 Demographics of HNSCC patients enrolled in the study.

Patients matching the inclusion criteria (n = 39) were enrolled in the study. ECOG (Eastern Cooperative Oncology Group) performance status describes how the disease affects the daily living abilities of the patient. For evaluating smoking status, pack years are calculated by multiplying the number of packs of cigarettes smoked per day by the number of years the person has smoked. Tumor stage from T1 to T4 refers to the size and extent of the tumors. The involvement of regional lymph nodes is referred by N1 to N3 depending on the number and location of the lymph nodes involved. N0 denotes absence of cancer in the regional lymph nodes.

View this table:
Fig. 1 HNSCC CD8+ T cells exhibit reduced chemotaxis in the presence of adenosine.

(A and B) Trajectories of CD8+ T cells migrating along either a CXCL12 gradient (green triangles) or a combination gradient of CXCL12 with adenosine (ADO, blue triangles) in a representative HD (A) and HNSCC patient (B). Trajectories of at least 15 to 20 cells are shown for each condition, and the starting point of each cell trajectory is artificially set to the same origin. The red triangles represent Y-COM. (C and D) Y-COM values for cells migrating along either a CXCL12 gradient or a combination gradient of CXCL12 with adenosine in HDs (n = 7 donors) (C) and HNSCC patients (n = 16 patients) (D). (E) Percentage inhibition in the Y-COM values in the presence of CXCL12 and adenosine [values shown in (C) and (D)] in HD (n = 7 donors) and HNSCC (n = 16 patients). Horizontal red line represents mean values for each group. Data in (C) and (D) were analyzed by paired Student’s t test and in (E) by Mann-Whitney rank sum test.

Table 2 Effect of adenosine on the chemotaxis of HD and HNSCC CD8+ T cells.

Activated CD8+ T cells from HD and HNSCC patients were exposed to a gradient of either CXCL12 or CXCL12 and adenosine (ADO), and the indicated values were measured. Results are presented as means ± SEM for all measured values. Y-COM, center of mass along the y axis, along the chemokine gradient; X-COM, center of mass along the x axis, perpendicular to the chemokine gradient; FMIy, forward migration index in the direction of the y axis (represents the efficiency of forward migration toward the chemokine gradient); FMIx, forward migration index in the direction of the x axis (represents the efficiency of migration perpendicular to the chemokine gradient respectively).

View this table:
Fig. 2 Tumor infiltration is dependent on the sensitivity of circulating CD8+ T cells to adenosine.

(A) Immunohistochemistry of CD8 (top) and CD73 (bottom) expression (brown signal) in representative HNSCC tumor tissues showing low and high infiltration by CD8+ T cells and low and high CD73 expression (table S1). Scale bars, 100 μm. (B) Bar graph showing the number of CD8+ T cells (cells/mm2) within the tumor region in 16 HNSCC tumors. Please note that donor HNC-52 has a mean CD8+ T cell infiltration value of 5 cells/mm2. The broken red line represents the median value for the 16 HNSCC patients. The tumors with CD8+ T cell infiltration above the median value were considered to be “well infiltrated” (referred to as high in table S1), whereas the tumors with CD8+ T cell infiltration below the median value were considered to be “poorly infiltrated” (referred as low in table S1). The bars represent means ± SEM. (C) Correlation between CD8+ T cell infiltration and percentage reduction of the Y-COM values in the presence of CXCL12 and adenosine (values shown in Fig. 1E) in nine HNSCC patients that were scored as CD73High (see table S1). Correlation was measured by Spearman rank-order correlation test (P = 0.0301; correlation coefficient, ρ = −0.700).

The A2AR mediates the suppressive effect of adenosine on the chemotaxis of HNSCC CD8+ T cells

We conducted experiments to define whether the A2AR mediates the inhibitory effect of adenosine on HNSCC CD8+ T cells. The selective A2AR agonist CGS21680 suppressed the chemotaxis of HNSCC CD8+ T cells in a way comparable to adenosine (Fig. 3, A and B) (22, 38). We observed that the Y-COM of HNSCC CD8+ T cells was reduced by CGS21680 (Fig. 3A). In four of these six patients, we simultaneously tested the effect of adenosine, and the Y-COM was not significantly different from that in the presence of CGS21680 (Fig. 3A). Overall, the degree of Y-COM reduction was similar in the presence of CGS21680 and adenosine (Fig. 3B). The involvement of the A2AR was further confirmed with SCH58261, a selective A2AR competitive antagonist (22). The inhibitory effect of adenosine on HNSCC CD8+ T cells was abrogated when the cells were pretreated with SCH58261 (Fig. 3C). The cells pretreated with SCH58261 still migrated toward CXCL12 even in the presence of adenosine, as shown by the increased Y-COM values when compared to cells not treated with SCH58621 (Fig. 3C). The adenosine-induced reduction in the Y-COM value of HNSCC CD8+ T cell chemotaxis was blocked by SCH58261 (Fig. 3D). Overall, we showed that the adenosine-mediated inhibition of chemotaxis of HNSCC CD8+ T cells occurs through the A2AR. These data suggest that blockade of the A2AR could increase the ability of circulating CD8+ T cells of HNSCC patients to migrate toward a chemokine.

Fig. 3 A2AR mediates the suppressive effect of adenosine on the chemotaxis of HNSCC CD8+ T cells.

(A) Y-COM values for HNSCC CD8+ T cells migrating along either a CXCL12 gradient (n = 6 patients), a combination gradient of CXCL12 with CGS21680 (n = 6 patients), or CXCL12 with adenosine (n = 4 patients). (B) Percentage inhibition in the Y-COM values for each individual experiment shown in (A) after incubation with CGS21680 or adenosine. Horizontal red lines represent mean values for each group. (C) Y-COM values for HNSCC CD8+ T cells pretreated with or without 1 μM SCH58261 migrating toward CXCL12 in the presence of adenosine. Untreated CD8+ T cells in a CXCL12 gradient were used as controls (n = 5 patients). (D) Percentage inhibition in the Y-COM values by adenosine for each of the donors shown in (C) with or without SCH58261 pretreatment. Horizontal red line represents mean values for each group. Data in (A) and (C) were analyzed by one-way repeated measures analysis of variance (ANOVA) [P = 0.010 for (A) and P = 0.001 for (C)]; data in (B) and (D) were analyzed by Student’s t test.

A2AR expression and cAMP-PKA signaling are not altered in HNSCC CD8+ T cells

The A2AR signals through the activation of adenylate cyclase and, in turn, induces an increase in cAMP, PKAI activation, and inhibition of KCa3.1 channels (11, 22). Experiments were performed to assess whether alterations in components of the adenosine signaling pathway mediate the inhibitory effect of adenosine on the chemotaxis of HNSCC CD8+ T cells. Increased A2AR abundance, increased cAMP-PKAI signaling, or decreased KCa3.1 activity in HNSCC CD8+ T cells could explain their enhanced sensitivity to adenosine. We measured ADORA2A expression (Fig. 4A) as well as A2AR protein abundance (Fig. 4, B to D) in CD8+ T cells in HDs and HNSCC patients and found no significant differences in either mRNA or protein expression. We further investigated whether there were any differences in the signaling pathway downstream of the A2AR. We did not observe any significant difference in the intracellular cAMP concentrations in activated HD and HNSCC CD8+ T cells (Fig. 4E). Similarly, we could not detect statistically significant differences in PKA activity after A2AR stimulation between HD and HNSCC CD8+ T cells (Fig. 4F). Overall, our results show that decreased migration of HNSCC CD8+ T cells in the presence of adenosine was not due to any differences in A2AR abundance or proximal signaling.

Fig. 4 A2AR expression and A2AR signaling are not altered in HNSCC CD8+ T cells.

(A) ADORA2A expression in activated HD and HNSCC CD8+ T cells was quantified by reverse transcription quantitative polymerase chain reaction (RT-qPCR). Data are the fold change in ADORA2A expression relative to GAPDH expression. The data were normalized to the mean ADORA2A expression in HD. Data are means ± SEM for from four HD and five HNSCC patients. (B) Representative flow cytometry histograms showing A2AR expression in resting and activated CD8+ T cells from HD and HNSCC. (C and D) Mean fluorescence intensity (MFI) of A2AR measured in resting (C) and activated (D) CD8+ T cells from HD (n = 6 donors) and HNSCC patients (n = 7 patients). (E) cAMP concentration in CD8+ T cells from HD (n = 7 donors) and HNSCC patients (n = 7 patients). (F) Relative PKA activity in CD8+ T cells from HD (n = 3 donors) and HNSCC patients (n = 4 patients). Horizontal red line represents mean values for each group. Data in (C), (D), and (F) were analyzed by Mann-Whitney rank sum test; data in (A) and (E) were analyzed by Student’s t test.

KCa3.1 channel activity is reduced in HNSCC CD8+ T cells

KCa3.1 channels are downstream of PKA in the adenosine signaling pathway, and we have previously shown that adenosine reduces T cell motility by inhibiting KCa3.1 channels (22). Hence, we conducted experiments to measure whether alterations in KCa3.1 channels in HNSCC CD8+ T cells could explain the adenosine-dependent inhibition of chemotaxis. KCa3.1 channels are present at the uropod of migrating T cells and mediate their migration (24). KCa3.1 channels are also involved in the chemotaxis of HD and HNSCC CD8+ T cells because the KCa3.1-specific blocker TRAM-34 inhibited their migration (fig. S2) (39). We thus evaluated the expression and function of KCa3.1 channels in resting and activated HNSCC CD8+ T cells and compared them to those of HD CD8+ T cells (Fig. 5). KCa3.1 channel expression in T lymphocytes increases with activation (40). Patch-clamp experiments showed that the KCa3.1 activity (defined by the whole-cell KCa3.1 conductance normalized to the cell membrane capacitance) was significantly lower in activated HNSCC CD8+ T cells as compared to their HD counterparts (Fig. 5, A and B, and table S3). No differences were observed in resting HD or HNSCC CD8+ T cells (table S3). There were no changes in the Kv1.3 activity (reported here as current density: peak current/capacitance) in either resting or activated HD or HNSCC CD8+ T cells (Fig. 5C and table S3). The data were normalized for the capacitance because there was a significant difference in cell capacitance between activated, but not resting, HD and HNSCC CD8+ T cells (table S3). Because activation increased KCa3.1 channel expression and capacitance, which is a measure of the cell size, these data raise the possibility that HNSCC CD8+ T cells are less activated than HD CD8+ T cells. We thus measured KCa3.1 expression by flow cytometry in resting and activated CD8+ T cells in HD and HNSCC patients and observed that upon activation, KCa3.1 channel abundance was similarly increased in both HD and HNSCC CD8+ T cells (Fig. 5, D and E). In both HD and HNSCC CD8+ T cells, there were comparable increases in the MFI of KCa3.1 after activation (Fig. 5F). We also measured CD69 abundance as a marker of activation and observed no statistically significant difference between HNSCC and HD CD8+ T cells (fig. S3). This suggests that the decreased KCa3.1 channel activity in HNSCC CD8+ T cells is not likely due to a decrease in channel surface expression or defective T cell activation. Overall, these data show that there is reduced KCa3.1 activity, but not expression, in HNSCC CD8+ T cells.

Fig. 5 KCa3.1 channel activity is reduced in HNSCC CD8+ T cells.

(A) Representative KCa3.1 currents in CD8+ T cells recorded in whole-cell voltage clamp configuration from an HD and HNSCC patient. Currents were normalized for the maximum current at +40 mV to ease comparison of the KCa3.1 conductance at hyperpolarizing voltages. (B) KCa3.1 conductance (normalized to cell capacitance, G/C) measured in activated CD8+ T cells from HD (n = 30 cells, six donors) and HNSCC patients (n = 21 cells, four patients). (C) Kv1.3 channel current density measured in activated CD8+ T cells from HD (n = 25 cells, five donors) and HNSCC patients (n = 21 cells, four patients). For (B) and (C), the data are normalized to values measured in activated CD8+ T cells from HD, and the bars represent mean ± SEM. (D) Representative flow cytometry histograms showing KCa3.1 expression in resting and activated CD8+ T cells from HD and HNSCC patients. (E) MFI of KCa3.1 measured in resting and activated CD8+ T cells from HD (n = 6 donors) and HNSCC patients (n = 7 patients). (F) KCa3.1 MFI in activated CD8+ T cells from HD (n = 6 donors) and HNSCC (n = 7 patients). Horizontal red line represents mean values for each group. Data in (B), (C), and (F) were analyzed by Mann-Whitney rank sum test; data in (E) were analyzed by paired Student’s t test.

Activation of KCa3.1 restores the chemotaxis of HNSCC CD8+ T cells in the presence of adenosine

Experiments were conducted to determine whether activation of KCa3.1 channels by 1-ethyl-2-benzimidazolinone (1-EBIO), a selective positive modulator of KCa3.1 channels, abolished the inhibitory effect of adenosine on the chemotaxis of HNSCC CD8+ T cells (41). We evaluated whether increasing the activity of KCa3.1 channels with 1-EBIO would enable the HNSCC CD8+ T cells to migrate toward a chemokine even in the presence of adenosine. 1-EBIO increased the activity of KCa3.1 channels in activated HNSCC CD8+ T cells to a level comparable to that of the baseline conductance in HD CD8+ T cells (without 1-EBIO) (Fig. 6A). Consistent with our earlier findings (Fig. 5, A and B), KCa3.1 conductance in HNSCC CD8+ T cells in the absence of 1-EBIO was also significantly lower than that in HD CD8+ T cells. The Y-COM values of HNSCC CD8+ T cells that underwent chemotaxis toward CXCL12 were significantly reduced in the presence of adenosine (Fig. 6B), similar to our earlier finding (Fig. 1). However, when HNSCC CD8+ T cells from the same individuals were preincubated with 1-EBIO, the Y-COM values in the presence of adenosine were almost threefold greater than those of cells in the presence of adenosine without 1-EBIO. Thus, the inhibitory effect of adenosine on the chemotaxis of HNSCC CD8+ T cells was blocked by 1-EBIO (Fig. 6C). A similar effect was caused by NS309, a more potent activator of KCa3.1 channels (42). Preincubation of CD8+ T cells from two HNSCC patients with NS309 reversed the inhibitory effect of adenosine on chemotaxis to CXCL12 (fig. S4). These findings suggest that enhancing KCa3.1 function in HNSCC CD8+ T cells restores their ability to chemotax in the presence of adenosine.

Fig. 6 Activation of KCa3.1 channels restores the chemotaxis of HNSCC CD8+ T cells in the presence of adenosine.

(A) KCa3.1 channel conductance in the presence or absence of 100 μM 1-EBIO was measured in activated CD8+ T cells from HD (n = 17 cells, four donors) and HNSCC patients (n = 24 cells, five patients). The data were normalized to untreated (no 1-EBIO) activated cells from HD. The data are means ± SEM. (B) Y-COM values calculated for HNSCC CD8+ T cells migrating along either a CXCL12 gradient or a combination gradient of CXCL12 with adenosine with or without preincubation with 20 μM 1-EBIO (n = 5 patients). (C) Percentage inhibition in the Y-COM values (B) of the cells pretreated with 1-EBIO. Horizontal red line represents mean values for each group. Data in (A) were analyzed by two-way ANOVA, whereas data in (B) were analyzed with one-way repeated measures ANOVA (P = 0.009) and (C) with paired Student’s t test.

DISCUSSION

The unequivocal prognostic and therapeutic significance of CD8+ T cell infiltration in most solid tumors and the known immunosuppressive properties of adenosine have driven the studies reported herein. The data that we present suggest that the chemotaxis of circulating HNSCC CD8+ T cells may be compromised in an adenosine-rich microenvironment because these cells, contrary to their healthy counterparts, have reduced KCa3.1 channel activity. Furthermore, these data highlight the therapeutic potential of KCa3.1 activators to increase penetration of CD8+ T cells into tumors by abrogating the inhibitory effect of adenosine on the chemotaxis of CD8+ T cells.

The failure of immune surveillance in cancer has been attributed to the lack of effective major histocompatibility complex presentation by cancer cells and the immunosuppressive properties of the TME (8). New immunotherapies have been designed to increase the functionality of T cells and their ability to resist the TME (6, 43, 44). A prerequisite for cytotoxic functionality is direct contact with tumor cells. To this effect, a high intratumoral CD8+/Treg (regulatory T cell) ratio is associated with good prognosis and response to therapy in multiple solid malignancies, including HNSCC (45). Thus, infiltration of CD8+ T cells is a limiting step in the efficacy of immune surveillance and immunotherapies in cancer. The TME is rich in the immunosuppressant adenosine (10). We showed that adenosine inhibited CD8+ T cell chemotaxis, and this effect was enhanced in cells from HNSCC patients. This is consistent with immunohistochemical data of various solid tumors showing an inverse correlation between CD73 in the tumor and the infiltration of CD8+ TILs (4650). The advantage of the studies that we have performed here over the correlative studies in tissue samples is that we have used a collagen-rich 3D microenvironment where we have full control over the experimental conditions used, whereas in vivo, the TME is a complex mixture of metabolic and waste products and tumor cells. This 3D system enables us to exclusively study the effect of adenosine on chemotaxis. We found that adenosine inhibited chemotaxis of HNSCC CD8+ T cells. The chemotaxis experiments reported herein showed that HNSCC CD8+ T cells lost directionality in the presence of adenosine but maintained their velocity and overall distance traveled. The effect of adenosine in 3D chemotaxis experiments does not fully recapitulate what we have previously observed in a 2D migration assay on intercellular adhesion molecule–1 surfaces, where adenosine inhibited integrin-mediated random migration of HD CD3+ T cells by reducing their velocity (22). The different experimental conditions and signaling pathways triggered by the different migratory stimuli may explain these discrepancies.

We found that adenosine had a more substantial inhibitory effect on the migration of HNSCC CD8+ T cells than HD CD8+ T cells. The degree of chemotaxis inhibition by adenosine in HNSCC CD8+ T cells negatively correlated with the number of CD8+ TILs in CD73-positive tumors underscoring the importance of this effect on immune surveillance in cancer. This also raises the possibility that the chemotactic sensitivity to adenosine of circulating CD8+ T cells could be used as a minimally invasive biomarker of CD8+ TIL infiltration and, possibly, prognosis. Larger population studies are necessary to confirm this possibility.

The increased sensitivity to adenosine that we observed in HNSCC CD8+ T cells is in agreement with the heightened suppression of cytokine production and proliferation of HNSCC CD8+ T cells by adenosine reported previously by Mandapathil et al. (33). They attributed this increased sensitivity to the reduced ability of HNSCC effector T cells (Teff) to degrade and internalize adenosine as well as an amplified A2AR signaling as compared to HDs’ Teff (33). We observed a comparable effect of adenosine and the A2AR agonist CGS21680 on the chemotaxis of HNSCC CD8+ T cells, suggesting that differences in the adenosine signaling pathway and not adenosine degradation are responsible for the heightened effect of adenosine on chemotaxis. It is well established that A2AR mediates the effect of adenosine in human T cells (5, 16, 17, 51, 52). We have previously shown that adenosine inhibits the motility of human CD3+ T cells and cytokine release through A2AR stimulation, activation of adenylyl cyclase and PKAI, and ultimately inhibition of KCa3.1 channels (22). Therefore, it is anticipated that an increase in A2AR, cAMP, and PKAI in HNSCC CD8+ T cells could amplify the response to adenosine. Similarly, a reduction in functional KCa3.1 channels could reduce the functionality of the cells once they are exposed to adenosine, but it could also contribute to the reduced ability of HNSCC T cells to produce interleukin-2 and IFN-γ as compared to HDs’ cells, which was previously reported (33). Our data indicated no significant differences in A2AR expression, cAMP abundance, or A2AR-stimulated PKAI activation in HNSCC T cells. We instead detected a decrease in KCa3.1 activity. This was not due to reduced KCa3.1 expression. Activated HD and HNSCC CD8+ T cells had similar amounts of KCa3.1 membrane proteins as determined by flow cytometry. Despite the low KCa3.1 whole-cell conductance, we observed no difference in baseline chemotaxis (in the absence of adenosine) between HNSCC and HD CD8+ T cells. It appears that only a fraction of KCa3.1 channel activity may be sufficient to guarantee chemotaxis. Further single-channel electrophysiological studies as well as evaluation of the downstream signaling pathways are necessary to understand the mechanism by which KCa3.1 channels regulate chemotaxis in T cells.

The findings reported in this article indicate that restoring KCa3.1 activity in HNSCC CD8+ T cells by 1-EBIO and NS309 abrogated the inhibitory effect of adenosine. KCa3.1 are Ca2+-sensitive K+ channels that open upon an increase in intracellular Ca2+ concentration. 1-EBIO is a positive modulator of KCa3.1 channels, which increases their Ca2+ sensitivity (41). Eil et al. (31) showed that pharmacological activation of KCa3.1 channels by positive modulators such as 1-EBIO improved CD8+ T cell function in vitro and may potentially be of therapeutic use in cancer. The data that we presented suggest that KCa3.1 activators could have therapeutic benefit by restoring the chemotactic capacity of CD8+ T cells in an adenosine-rich TME, which would favor tumor penetration. The effect of KCa3.1 activation was comparable to that obtained by A2AR blockade, which has already shown to be of clear benefit in combination with anti-programmed death protein 1 (anti-PD1) antibody or CAR T cells in preclinical models of solid malignancies (5, 6, 17). A therapy based on KCa3.1 activation would have advantages over A2AR inhibition. KCa3.1 channels are downstream of other GPCRs. Prostaglandin E2, which is also present in the TME, inhibits KCa3.1 channels in mast cells (53). Thus, activation of KCa3.1 channels could be a more effective approach for improving immune surveillance and the response to immune therapies in cancer because it could simultaneously counteract multiple immune suppressive components of the TME.

MATERIALS AND METHODS

Human subjects

Studies were conducted on peripheral blood obtained from 39 deidentified HNSCC patients in the age range of 34 to 77. The eligibility criteria for patient inclusion in the study were a positive diagnosis for HNSCC confirmed by tissue biopsy and no administration of radiation or chemotherapy before the time of drawing the blood (see Table 1 for patient demographics and table S1 for clinical information). The data on the study subjects were collected and managed using Research Electronic Data Capture (REDCap) tools hosted at the University of Cincinnati. Peripheral blood was also drawn from 20 age-matched (±5 years) HDs (8 female and 12 male) in the age range of 24 to 67 years. Discarded blood units from Hoxworth Blood Center (University of Cincinnati) were used for preliminary chemotaxis validation experiments as well as for the ADORA2A RT-qPCR and the 1-EBIO electrophysiological experiments. Informed consent was obtained from all HNSCC patients and HDs. The study and informed consent forms were approved by the University of Cincinnati Institutional Review Board (IRB no. 2014-4755).

Reagents and chemicals

Human serum, l-glutamine, adenosine, SCH58261, 1-EBIO, and sodium hydroxide were purchased from Sigma-Aldrich. Hepes, RPMI 1640, fetal bovine serum, penicillin, streptomycin, and phosphate-buffered saline (PBS) were obtained from Gibco. Rat tail collagen I was obtained from Corning Inc. CGS21680 hydrochloride and NS309 were purchased from Tocris Bioscience, whereas CXCL12 was obtained from R&D Systems. TRAM-34 was a gift from H. Wulff (Department of Pharmacology, University of California Davis). Stock solutions of SCH58261, CGS21680, TRAM-34, 1-EBIO, and NS309 were prepared in dimethyl sulfoxide and used at 0.1% dilution. Stock solution of CXCL12 was prepared in sterile PBS containing 0.1% bovine serum albumin.

Cell isolation and activation

Peripheral blood mononuclear cells (PBMCs) were isolated from whole blood by Ficoll-Paque density gradient centrifugation (GE Healthcare Bio-Sciences), as described previously (54). CD8+ T cells were subsequently isolated from PBMCs by negative selection using the EasySep Human CD8+ T Cell Enrichment Kit (STEMCELL Technologies Inc.) according to the manufacturer’s instructions. The CD8+ T cells were maintained in RPMI 1640 medium supplemented with 10% human serum, penicillin (200 U/ml), streptomycin (200 μg/ml), 1 mM l-glutamine, and 10 mM Hepes (54). Cells were activated for 72 to 96 hours in a cell culture dish coated with mouse anti-human CD3 antibody (10 μg/ml) (BioLegend) and mouse anti-human CD28 antibody (10 μg/ml) (BioLegend).

Reverse transcription quantitative polymerase chain reaction

Total RNA was isolated from activated HD and HNSCC CD8+ T cells using the E.Z.N.A. total RNA isolation Kit (Omega Bio-tek) as per the manufacturer’s instructions. Six hundred and fifty nanograms of RNA was used to synthesize complementary DNA (cDNA) using the Maxima First Strand cDNA Synthesis Kit for RT-qPCR (Thermo Fisher Scientific) as per the manufacturer’s instructions. Predesigned primers for RT-qPCR were obtained using TaqMan Gene Expression Assays (Applied Biosystems, Thermo Fisher Scientific) to detect the expression of ADORA2A (assay ID: Hs001169123_m1) and GAPDH (assay ID: HS03929097_g1). The RT-qPCR was set up in a 96-well plate by adding 30 ng of cDNA, 1× TaqMan Gene Expression Master Mix (Applied Biosystems), and 1 μl of TaqMan Gene Expression Assay primers. All samples were run in quadruplicate. GAPDH was used as an internal control. RT-qPCR was cycled in Applied Biosystems StepOne Real-Time PCR System (Applied Biosystems). CT values were measured using StepOne software version 2.1 (Applied Biosystems). CT values for ADORA2A were normalized against measured CT values for GAPDH, and the ΔΔCT values were calculated as described previously (55). Relative quantity values, representing the fold change in ADORA2A gene expression in HNSCC CD8+ T cells compared to HD CD8+ T cells, were calculated as the 2−ΔΔCT values.

Flow cytometry

CD8+ T cells (~1 × 106 cells per condition) were fixed with 4% paraformaldehyde (Affymetrix) and stained with ATTO 488–conjugated mouse anti-human KCa3.1 antibody (clone 6C1, Alomone Labs). Cells were then permeabilized with 0.1% Triton X-100 (Sigma-Aldrich) and stained with Alexa Fluor 405–conjugated mouse anti-human A2AR antibody (clone 2D1, Novus Biologicals). To test T cell activation, CD8+ T cells were stained live with Alexa Flour 488–conjugated anti-CD69 antibody (Clone FN50, BioLegend) and fixed with 1% paraformaldehyde. Data were collected on an LSR II flow cytometer (BD Biosciences) and analyzed with FlowJo software (FlowJo LLC).

Chemotaxis

Three-dimensional chemotaxis was performed using the μ-Slide Chemotaxis assay (ibidi GmbH) according to the manufacturer’s instructions. Briefly, ~1 × 106 activated CD8+ T cells were incorporated in a type I rat tail collagen gel (Corning) and added to the central observation chamber of the μ-Slide Chemotaxis. In all experiments, CXCL12 (8 μg/ml) was added to the migration medium in the reservoir to the left of the observation chamber, while migration medium without chemokine was added to the reservoir to the right of the observation chamber, thus generating a chemokine gradient to measure the baseline chemotaxis. To assess the effects of adenosine or CGS21680 on chemotaxis, we established simultaneous adenosine (or CGS21680) and chemokine gradients in a separate chamber by injecting adenosine (10 μM) or CGS21680 (10 μM) and CXCL12 (8 μg/ml) into the left reservoir. The slide was mounted on the stage of an inverted Zeiss LSM 710 microscope (Carl Zeiss Microscopy GmbH) equipped with a 37°C incubator. Cell migration was recorded by time-lapse video microscopy, with bright-field images acquired every 3 s for up to 3 hours. Cell tracking on the time-lapse images was performed using the “Manual Tracking plugin” on ImageJ software (National Institutes of Health), and the data were analyzed using the Chemotaxis and Migration Tool (ibidi GmbH). On average, 15 to 20 cells were tracked per condition. The following chemotactic parameters were derived: (i) COM (the average position along the relevant axis that the cells reached by the end of the experiment), (ii) Euclidean distance (the linear distance between the starting point and ending point of a cell), (iii) accumulated distance (the total distance traveled by the cell during the course of the entire microscopy recording), (iv) FMI (the ratio between the net distance traveled along the relevant axis and the accumulated distance), (v) directness (the ratio between the Euclidean distance and the accumulated distance, denotes the tendency of the cells to migrate along a straight line), and (vi) velocity (36). We defined a positive chemotaxis effect if the cells migrated along the chemokine gradient (y axis) that is, if the Y-COM was significantly greater than X-COM and FMIy was significantly greater than FMIx.

cAMP determination

Activated CD8+ T cells were lysed using 0.1 M HCl at a final concentration of 1 × 106 cells/ml. Intracellular cAMP concentrations were measured in T cell lysates using the acetylated procedure of the Direct cAMP ELISA Kit (Enzo Life Sciences) according to the manufacturer’s instructions.

PKA kinase activity assay

Activated CD8+ T cells were stimulated with 1 μM CGS21680 for 30 min. Cell lysates were prepared as previously described, and total protein content was measured using BCA Protein Assay (55). The cell lysates were diluted to equal protein concentrations using kinase activity assay buffer. The PKA kinase activity was measured using a PKA kinase activity kit (Enzo Life Sciences) as described by the manufacturer, with a reaction time of 90 min at 30°C and a development time of 60 min. The relative kinase activity was determined by the absorbance of the sample at 450 nm divided by the amount (micrograms) of crude protein.

Immunohistochemistry

Slides prepared from formalin-fixed paraffin-embedded (FFPE) tumor biopsy specimens from 16 HNSCC cases were deparaffinized and stained with a monoclonal rabbit anti-human CD8 antibody (clone SP57, Ventana Medical Systems) in a Ventana BenchMark ULTRA automated IHC slide staining system (Ventana Medical Systems). For CD73 staining, FFPE sections were stained with a mouse monoclonal anti-human CD73 antibody (clone 1D7, Abcam). The ultraView Universal DAB Detection Kit (Ventana Medical Systems) containing a horseradish peroxide multimer and 3,3′-diaminobenzidine tetrahydrochloride (DAB) chromogen was used for indirect detection of the primary antibody. The slides were counterstained with hematoxylin, and images were obtained at ×10 magnification on an Olympus BX53 light microscope (Olympus Corporation) or on a Leica DMi8 inverted microscope with Leica Application Suite X software (Leica Microsystems Inc.). Tumor regions in the stained slides were identified by a pathologist, and at least 4 to 10 fields per slide were imaged. CD8+ T cell infiltration in the tumor area was digitally quantitated by drawing a region of interest (ROI) around the tumor region and counting the number of cells (brown signal) within the ROI using NIS-Elements Viewer software (Nikon Instruments Inc.). Data were expressed as cells counted per square millimeter, and the median value of CD8+ infiltration in all of the measured ROIs was determined. Tumors with CD8+ infiltration values (cells/mm2) above and below the median were considered as “high” and “low” infiltrated, respectively. Slides stained for CD73 were visually assessed for CD73 staining (brown signal) in the tumor and stromal regions by a pathologist and characterized as having high or low CD73 staining in these regions. To eliminate a subjective bias in reporting the CD8+ infiltration as well as CD73 staining, microscopy and the image analysis were performed blinded.

Electrophysiology

KCa3.1 and Kv1.3 currents in CD8+ T cells were measured in whole-cell voltage-clamp configuration using an AxoPatch 200B Amplifier (Molecular Devices). The external solution contained either 160 mM NaCl, 4.5 mM KCl, 2.0 mM CaCl2, 1.0 mM MgCl2, and 10 mM Hepes (pH 7.4) (Fig. 5, A to C), or 145 mM Na-aspartate, 5 mM KCl, 2.5 mM CaCl2, 1.0 mM MgCl2, 5.5 mM glucose, and 10 mM Hepes (pH 7.4) (Fig. 6A), and the pipette solution contained 145 mM K-aspartate, 10 mM K2EGTA, 8.5 mM CaCl2, 2 mM MgCl2, and 10 mM Hepes (pH 7.2), 290 to 310 mosmol (1 μM free Ca2+ concentration). Currents were induced by 200-ms ramp depolarization from −120 to +40 mV from a holding potential of −70 mV every 10 s. The macroscopic conductance of KCa3.1 channels (GKCa3.1) was calculated as a ratio of the linear fraction of macroscopic current slope to the slope of ramp voltage stimulus after subtraction of the leak current G(pS)={Islope(pAms)}{Vslope(Vms)}. The slope conductance was measured between −100 and −80 mV to avoid contamination by the Kv1.3 current. Kv1.3 currents were determined from the same ramp protocol at +40 mV after subtraction of the KCa3.1 current extrapolated by linear regression.

Statistical analysis

Statistical analyses were performed using Student’s t test (paired or unpaired), Mann-Whitney rank sum test, Wilcoxon signed-rank test (in experiments where samples failed normality), and ANOVA as indicated. Post hoc testing on ANOVA was done by multiple pairwise comparison procedures using the Holm-Sidak method. Statistical analysis was performed using SigmaPlot 13.0 (Systat Software Inc.). Outliers were determined by Grubb’s test (GraphPad Software). P value of less than or equal to 0.05 was defined as statistically significant. The correlation between CD8+ T cell infiltration and inhibition of Y-COM was analyzed by Spearman’s rank-order correlation.

SUPPLEMENTARY MATERIALS

www.sciencesignaling.org/cgi/content/full/11/527/eaaq1616/DC1

Fig. S1. HD and HNSCC CD8+ T cells chemotax toward CXCL12.

Fig. S2. KCa3.1 channel blockade inhibits the chemotaxis of HD and HNSCC CD8+ T cells.

Fig. S3. Activation of CD8+ T cells from HD and HNSCC patients.

Fig. S4. Activation of KCa3.1 channels by NS309 restores the chemotaxis of HNSCC CD8+ T cells in the presence of adenosine.

Table S1. Clinicopathologic characteristics of individual HNSCC patients.

Table S2. CD8+ T cells from HDs and HNSCC patients chemotax toward CXCL12 similarly.

Table S3. Electrophysiological parameters of resting and activated CD8+ T cells isolated from HD and HNSCC patients.

REFERENCES AND NOTES

Acknowledgments: We would like to acknowledge clinical coordinators from the University of Cincinnati Cancer Institute’s clinical trials office (UCCI CTO) for their assistance in collection of patient samples. We also thank H. Duncan (Division of Nephrology and Hypertension, Department of Internal Medicine, University of Cincinnati) for assistance with IRB regulatory affairs. We would like to thank R. Jandarov (Department of Biostatistics and Bioinformatics, University of Cincinnati) for valuable advice on statistical analysis. We would also like to thank the Center for Clinical and Translational Science and Training (CCTST), University of Cincinnati, for providing REDCap for effective patient data management. We are grateful to H. Wulff (Department of Pharmacology, University of California Davis) for providing TRAM-34. Confocal microscopy images were acquired at the Live Microscopy Core, Department of Pharmacology and Systems Physiology, University of Cincinnati. Flow cytometry experiments were performed at Shriners Hospital for Children Flow Cytometry Core, Cincinnati, OH. Funding: This work was funded by grant support from the NIH (grant R01CA95286), a Pilot grant from the UCCI, and a Just-In-Time Core Grant from CCTST, University of Cincinnati to L.C. T.W.-D. was supported by a Clinical and Translational Science Award (CTSA)–awarded KL2 Mentored grant, a Hematology Oncology Translational Science Award (HOTSA)/Hematology Oncology Pilot Grant Award (HOPGA) from the Division of Hematology Oncology at the University of Cincinnati and a grant from CCTST (1UL1TR001425-01). A.B. was cofinanced by Campus Hungary Program (B2/1SZ/12351) and National Excellence Program (TÁMOP 4.2.4. A/2-11-1-2012-0001 B). P.H. was supported in part by the Bolyai János Fellowship (Hungarian Academy of Sciences). Author contributions: Conception and design: A.A.C., A.B., and L.C. Development of methodology: A.A.C., L.C., A.B., M.J.A., and H.S.N. Acquisition of data: A.A.C., A.B., M.J.A., H.S.N., P.H., and J.Q. Provision and management of patients: T.W.-D. Provision of patient data: T.W.-D. and J.Q. Analysis and interpretation of data: A.A.C., A.B., M.J.A., H.S.N., P.H., J.Q., T.W.-D., and L.C. Writing, review, and/or revision of the manuscript: A.A.C. and L.C. Study supervision: L.C. All authors discussed the results and commented on the manuscript. Competing interests: The authors declare that they have no competing interests. Data and materials availability: All data needed to evaluate the conclusions in the paper are present in the paper or the Supplementary Materials.
View Abstract

Navigate This Article