Senescence-associated secretory proteins induced in lung adenocarcinoma by extended treatment with dexamethasone enhance migration and activation of lymphocytes

There is a need to improve response rates of immunotherapies in lung adenocarcinoma (AC). Extended (7–14 days) treatment of high glucocorticoid receptor (GR) expressing lung AC cells with dexamethasone (Dex) induces an irreversible senescence phenotype through chronic induction of p27. As the senescence-associated secretory phenotype (SASP) may have either tumor supporting or antitumor immunomodulatory effects, it was interest to examine the effects of Dex-induced senescence of lung AC cells on immune cells. Dex-induced senescence resulted in sustained production of CCL2, CCL4, CXCL1 and CXCL2, both in vitro and in vivo. After Dex withdrawal, secretion of these chemokines by the senescent cells attracted peripheral blood monocytes, T-cells, and NK cells. Following treatment with Dex-induced SASP protein(s), the peripheral blood lymphocytes exhibited higher cell count and tumor cytolytic activity along with enhanced Ki67 and perforin expression in T and NK cells. This cytolytic activity was partially attributed to NKG2D, which was upregulated in NK cells by SASP while its ligand MICA/B was upregulated in the senescent cells. Enhanced infiltrations of T and NK cells were observed in human lung AC xenografts in humanized NSG mice, following treatment with Dex. The findings substantiate the idea that induction of irreversible senescence in high-GR expressing subpopulations of lung AC tumors using Dex pretreatment enhances tumor immune infiltration and may subsequently improve the clinical outcome of current immunotherapies.


Introduction
Lung cancer is the leading cause of cancer-related mortality in the United States, accounting for approximately 24% of cancer deaths [1]. Lung adenocarcinoma (AC) accounts for half of all lung cancer cases with a 2-years survival rate of 10%. Current first-line treatment decisions for advanced lung AC are based on the presence of targetable genetic aberrations, such as sensitizing mutations of EGFR, BRAF, and MET, or translocations of ALK and ROS1 genes. However, these oncogenic drivers are present in only up to 20% of patients. Most patients are treated with the programmed death-1 (PD-1) immune checkpoint inhibitor pembrolizumab, either alone or in combination with platinumpemetrexed chemotherapy [2,3]. The advent of immune checkpoint inhibitors has brought a paradigm shift in the management of advanced lung AC with more than a twofold improvement in median overall survival in a subset of patients. However, development of resistance to these agents is inevitable and disease progression poses a major challenge. There is a paucity of effective therapies for patients who have progressed on immune checkpoint inhibitors and 1 3 platinum-doublet therapies. Therefore, new treatments are urgently needed.
Broadly, a major cause of cancer immunotherapy failure is inadequate infiltration of effector T cells and cytotoxic NK cells into the tumor tissue. However, it is well-established that drug-induced senescence in cancer models in vivo is often followed by secretion of chemokines as part of the senescence-associated secretory phenotype (SASP) [4,5]. Inducing release of these chemoattractants may offer a mechanism to recruit immune cells to the tumor, contributing to tumor clearance. We have previously reported that in lung AC cells and patient lung AC tumors expressing relatively high levels of the glucocorticoid receptor type α (GR), the synthetic steroid dexamethasone (Dex) induces reversible G1 arrest within 24 h [6,7]. More recently, we have reported that in these cells, extended treatment with Dex progressively induces irreversible cell cycle blockade and a senescence phenotype through chronic activation of the p27 Kip1 gene and accumulation of the p27 protein [8]. As Dex-induced senescence could potentially be developed as a means of improving the response of lung AC to immunotherapy, we sought to identify cytokines and chemokines of the SASP induced by extended Dex treatment of lung AC and to explore their potential role in modulating lymphocytes.

In vitro induction of senescence and collection of conditioned media (CM) containing SASP
To induce senescence and prepare conditioned media, H1299 and H1299GR Clone 4 cells were grown in phenolfree RPMI 1640 medium supplemented with 10% FBS, 100 units/ml penicillin, 100 μg/ml streptomycin and 2 mM glutamine. Two types of conditioned media were generated. In one case, cells were cultured for 7 days in the presence of Dex (100 nM) and the media harvested (7-days-Dex-CM). Alternatively, the cells were initially cultured for 7 days in the presence of Dex (100 nM); the culture media was then removed, the cells were washed 2X with PBS, and the cells were replenished with fresh culture media and incubated for an additional 48 h in the absence of Dex prior to harvesting the media (7-days-Dex + 2-day CM). Control conditioned media were generated by culturing cells in the absence of Dex (vehicle CM). In all cases, the conditioned media was collected, centrifuged, and filtered through 0.2 µM filter.

Isolation and measurement of mRNA
Total RNA was isolated from cells or tissues using the Purelink RNA Mini Kit (ThermoFisher Scientific, Grand Island, NY). Xenograft tissue lysates were prepared by suspending 30 mg of tissue in 500 µl of the lysis buffer and homogenized using the PRO200 tissue homogenizer (catalog no. 01-01200, BioGen) for 15 s on ice. Homogenized solution was centrifuged at high, 15,000×g, and supernatant was used for RNA extraction. Reverse-transcription PCRs were performed using high-capacity complementary DNA archive kit (Life Technologies, Inc., Carlsbad, CA). cDNA was measured by quantitative real-time RT-PCR using TaqMan probes and the StepOne Plus real-time PCR system (Life Technologies). The TaqMan probes were species-specific to human. All RNA measurements were performed in biological triplicates, and all Ct values were normalized in each case to GAPDH (intra-sample mRNA) values within the same tumor.

Isogenic tumor xenografts models
The RNA collected from the isogenic tumor xenografts for this study was extracted from tumors harvested in a previous published study from our group [8]. The tumors were collected and stored at − 80 °C. RNA extraction from tumors is described above.

Human LC xenograft model in humanized NSG mice
All animal procedures were conducted in accordance with the U.S. Public Health Service Policy on Use of Laboratory Animals and with approval by Wayne State University Institutional Animal Care and Use Committee. NSG mice ((NOD SCID Gamma; Strain NOD.Cg-Prkdc scid Il2rg tm1Wjl / SzJ) were purchased from Jackson Laboratory, H1299-GRα tumor was first passaged in SCID mice (Jackson Lab) and then transplanted into the flanks of NSG mice, as previously described [8]. Tumor growth was monitored by our animal care facility and measured with a vernier caliper two to three times per week. Tumor volume was calculated with the formula XY 2 /2 where X is the long axis and Y is the short axis of the tumor. When the tumor was approximately 600 cc 3 in volume, one group of mice started receiving Dex treatment, as described in our previous study [8]. On the same day, all mice were humanized with donor's PBMC (15 × 10 6 cells in 250 µl PBS, slowly injected i.v.). At indicated timepoints after the final Dex treatment, mice were euthanized and blood, spleen and tumor were harvested. The tumor was enzymatically digested, as previously described [9], and analyzed by flow cytometry.

Migration assay
Lymphocyte migration in response to chemokine cues was assessed via a standard transwell migration assay with some modifications. In brief, plasma and PBMC were isolated from the peripheral blood following standard percoll-gradient centrifugation. Monocytes (CD14 + ), T cells (CD3 + ) and NK (CD56 + ) cells were positively isolated from the PBMC using magnetic separation kit (StemCell) (98% purity per the manufacturer's protocol) as previously described [9,10]. Conditioned media from lung AC cultures were placed in the lower chamber of a 24-well plate containing 5 uM cell inserts layered with basement membrane matrix Geltrex (Gibco). PBMCs or isolated T cells, NK cells or monocytes were seeded in the upper chamber. Following 16 h of incubation, cells that had migrated into the lower chamber were imaged and enumerated using the ImageJ software. The migrated cells were also counted manually under an Olympus microscope.

Immune profiling
Multi-parametric flow cytometry was performed as described elsewhere [9]. Briefly, PBMC were isolated from the peripheral blood following standard percoll-gradient centrifugation as described above. PBMC were analyzed via a 10-color flow cytometry to assess: (1) alterations in the absolute number and relative frequency of T (CD3) cells and cytotoxic (CD56lowCD16high) and regulatory (CD56high-CD16low) NK cells; (2) Proliferation (Ki67) and activation (NKG2D, Perforin) markers; as well as (3) PD1 status of T and NK cell subsets. Flow cytometry data were analyzed, using FCS Express 7 software (DeNovo). Following antibodies were used (Table 1).

NK/T cell expansion and cytotoxicity assay
Monocyte-depleted peripheral blood lymphocytes (PBL) were cultured in complete RPMI in the presence or absence of low-dose IL-2 (50 U/ml) and IL-15 (10 ng/ml) for 3 days. PBL were then counted and co-cultured with the lung AC cells at an effector to target ratio of 12:1 for 5 h. LDH release from the tumor (target) cells was analyzed as a measure of cytotoxicity using CytoTox 96® Non-Radioactive Cytotoxicity Assay Kit (Promega).

Statistical analysis
All the experiments were performed three times with PBMC obtained from 3 different donors. Statistical significance was determined using one-way ANOVA followed by post hoc t-test to compare pairs of samples.

Dex irreversibly enhances secretion of select chemokines in lung AC cells with high GR expression
As a previously established model for studies of Dexinduced senescence in lung AC [7], we used an H1299 lung AC cell line (H1299-GR-Clone4) overexpressing recombinant GR and the parental low GR expressing H1299 cell line as an isogenic low GR control (Fig. 1A). Conditioned media were recovered from these cells, either after 7 days of exposure to Dex (100 nM) (7-days-Dex CM) or after the 7-days Dex exposure followed by a 2-day Dex washout period by culturing the cells for an additional 2-days in Dex-free media (7-days-Dex + 2-days CM). Proteins known to comprise the SASP complement were assayed in the conditioned media using a Bioplex assay ( Table 2). Several chemokines were induced by Dex treatment. Importantly, the 7-days-Dex + 2-days CM from H1299-GR-Clone4 cells showed elevated levels of the chemokines CCL2, CCL4, CXCL1 and CXCL2 compared to parental H1299 cells. CXCL8 appeared to show a higher level in H1299-GR-Clone4 cells at the end of the 7-days Dex treatment, but this difference in expression compared with the vehicle control group was not sustained after the Dex washout. There was also a concomitant reduction in PDGF level in media from H1299-GR-Clone4 cells at the end of the 7-days Dex treatment, compared with the parental cells, but PDGF expression partially recovered after the Dex washout (Table 2A).
Although the parental H1299 also showed some induction of chemokines CCL2 and CXCL2 following 7-days Dex treatment, the increases compared to the vehicle controls were not sustained following Dex washout (Table 2B).
To assess whether Dex induces the above chemokines in vivo, we used relevant tumors harvested in a previous published study [8]. Here, mice harboring tumor xenografts of H1299-GR-Clone4 cells or the control parental H1299 cell xenografts were implanted with slow-release Dex pellets or control placebo pellets for 2 weeks. To exclude chemokines expressed in the host mouse stromal cells, the mRNA levels for human species-specific CCL2, CCL4, CXCL1, CXCL2 and CXCL8 were measured in all the tumors. Compared to parental H1299, Dex-treated H1299-GR-Clone4 tumors expressed higher levels of mRNAs for human CCL2, CCL4, CXCL1, and CXCL8 (Fig. 1B, C). Thus, induction of mRNAs for select chemokines by Dex in vivo generally recapitulates induction of the chemokines in vitro.

Conditioned media (CM) from Dex pre-treated high GR expressing lung AC cells induces migration of immune cells
We examined the effect of 48-h CM from H1299-GR-Clone4 cells pre-treated with 100 nM Dex for 7 days (7-days-Dex + 2-days CM). The CM significantly (by ~ threefold) enhanced migration of human peripheral blood mononuclear cells (PBMC), compared to similarly prepared CM from the vehicle treated cells (vehicle CM) ( Fig. 2A, B). Similarly, the 7-days-Dex + 2-days CM significantly (> 2.5-fold) enhanced migration of isolated monocytes, NK cells and T cells, compared to the vehicle CM ( Fig. 2A, B). The results suggest that induction of SASP proteins induced by Dex in Fig. 1 GR expression in isogenic lung AC models and related expression of chemokine mRNAs during Dex-induced senescence. A RNA was extracted from isogenic lung adenocarcinoma cell lines H1299 and H1299-GR-Clone 4 and the levels of GR mRNA were measured by real time RT-PCR using TaqMan probes. B, C Mice harboring tumor xenografts of H1299-GR-Clone4 cells, (B) or the control parental H1299 cell xenografts (C) were implanted with slow-release Dex pellets or control placebo pellets for 2 weeks. RNA was extracted from the tumors. The mRNA levels for CCL2, CCL4, CXCL1, CXCL2 and CXCL8 were measured in all the tumors by real time RT-PCR using TaqMan probes. Statistical significance was determined using one-way ANOVA followed by post hoc t-test to compare pairs of samples. Data are mean ± SD of three different experiments. *p < 0.05 ▸ high GR expressing lung AC cells is associated with secretion of chemokines that attract various immune cells.

CM from Dex pre-treated high GR expressing lung AC cells induces expansion and cytolytic activity of peripheral blood lymphocytes
Monocyte-depleted peripheral blood lymphocytes (PBL) mostly contain T cells and NK cells, which are the major immune cells responsible for cytolytic clearance of senescent tumor cells. PBLs were cultured for 3 days in the presence of 7-days-Dex + 2-days CM or corresponding vehicle CM derived from H1299-GR-Clone4 cells either in the absence or in the presence of low dose IL-2 (50 U/ml) plus IL-15 (10 ng/ml). Treatment with 7-days-Dex + 2-days CM yielded a significantly higher number of PBLs compared to vehicle CM, both in the absence and in the presence of the exogenous interleukins (Fig. 3A), indicating an expansion of PBLs by cytokines induced by Dex treatment.
To assess the tumor-cytolytic potential of the SASPstimulated PBLs above, we performed a cytotoxicity assay using H1299-GR-Clone4 cells as the target cells. Accordingly, PBLs were cultured as above for 3 days in the 7-days-Dex + 2-days CM or corresponding vehicle control media either in the absence or in the presence of low dose IL-2 (50 U/ml) plus IL-15 (10 ng/ml). PBLs were then co-cultured with the target H1299-GR-Clone4 cells at an effector to target ratio of 12:1, and cytotoxicity was assessed by LDH release. As shown in Fig. 3B, the tumor-cytolytic activities of PBLs-treated with 48-h conditioned media from Dex pre-treated lung AC cells (7days Dex + 2days CM) were higher than the corresponding controls either with or without IL-2 and IL-15. Notably, the cytolytic effect was greater in the presence of IL-2 and IL-15 (CM + IL-2/IL-15) ( Fig. 3B and Supplementary   Fig. S1). The cytolytic activities of PBLs cultured with CM from vehicle-treated lung AC cells were not appreciably different from the corresponding controls that were not treated with any conditioned media ( Supplementary  Fig. S1).
Multi-color flow cytometry revealed higher frequencies of Ki67 and perforin positive cells in the PBLs treated with the combination of 7-days-Dex + 2-days CM and IL-2 + IL-15 (CM + IL-2/IL-15) in both T cell (CD3 gated) and NK cell (CD56 gated) fractions (Fig. 3C, middle and lower panels, respectively). Collectively, the above results indicate that SASP proteins from Dex pre-treated lung AC cells can induce proliferation (enhanced cell count and Ki67 staining) and activation (cytotoxicity and perforin staining) of both T cells and NK cells.

Neutralizing antibodies inhibit the ability of conditioned media from Dex pre-treated lung AC cells to induce PBL migration without inhibiting cytolytic activity
Migration of PBLs induced by 7-days-Dex + 2-days CM from H1299-GR-Clone4 cells was significantly inhibited by neutralizing antibodies against four of the Dex-induced chemokines at concentrations corresponding to two times their respective IC 50 values reported by the manufacturer (R&D Systems). Specifically, neutralizing antibodies to CCL2, CCL4 and a pan-GRO antibody (cross-reactive with both CXCL1 and CXCL2) were all effective inhibitors of the SASP-induced PBL migration, both individually and in combination (Fig. 4A).
We also investigated whether the SASP-induced cytolytic activity (shown in Fig. 3C) could be inhibited by neutralizing antibodies to the chemokines. We observed that the PBLs cultured with 7-days-Dex + 2-days CM plus IL-2 and

Tumor-cytolytic activity of NK cells induced by SASP from Dex-treated lung AC cells is partially mediated by NKG2D
NK cells have a major role in the clearance of senescent cancer cells via NKG2D-mediated cytotoxicity [4,5]. Therefore, we examined the NK cell specific phenotype and cytolytic activation markers in PBLs following stimulation with 7-days-Dex + 2-days CM from H1299-GR-Clone4 cells plus IL-2 and IL-15. We observed a significantly enhanced proportion of CD56 + NK cells in the stimulated PBLs (total 27.5% CD56 + ), compared with PBLs treated with the control vehicle CM (total 18.9% CD56 + ) (Fig. 5A). The relative frequency of NK cells expressing activation markers NKG2D and CD16 were also significantly enhanced in the stimulated group (79.1%) compared to the control (70.2%) (Fig. 5A). Next, we analyzed the expression of the NKG2D ligand MICA/B and MHC class I molecules in H1299-GR-Clone4 cells following treatment with 100 nM Dex for 72 h. We opted to treat the cells for 72 h instead of 7 days, because the induction of senescent phenotype in lung AC cells can be observed as early as 24 h following Dex treatment [6,7]. We observed a distinct enhancement of MICA/B expression in the Dex-treated group, compared to the control group (Fig. 5B), suggesting that Dex treatment makes GR-high lung AC cells susceptible to lysis by activated T/NK cells via interaction with NKG2D. However, the cytotoxicity of PBLs cultured with 7-days-Dex + 2-days CM was only partially inhibited by an NKG2D neutralizing antibody (Fig. 5C), indicating that the cytolytic activity of the SASP-stimulated PBLs is partially mediated by NKG2D.

Enhanced T and NK cell infiltration into human lung cancer xenografts in humanized NSG mice following treatment with Dex
In a small pilot study to establish a protocol to humanize NSG mice by intravenous (i.v.) injection of 15 × 10 6 donor derived PBMCs, we detected T cells and NK cells in the peripheral blood of mice 2 weeks after PBMC injection ( Supplementary Fig. S2). We observed that treatment with Dex (4 mg/kg/day for 7 days) caused a depletion of 40-50% lymphocytes in patients with lung AC (data not shown). Dex given at a similar dose and duration depleted 40-45% PBMC in humanized NSG mice (Supplementary Fig. S2). We then transplanted H1299-GR tumor into the flanks of NSG mice and randomly divided them into two groups, namely placebo and Dex. Due to rapid growth of the tumor, especially, in the vehicle control arm, mice were euthanized after 5 days of treatment with Dex. Even with only 5-day treatment, we observed considerably slower tumor growth in Dex group than in the control group ( Supplementary Fig.  S3). Flow cytometric analysis of the tumor tissues revealed significantly higher infiltration of T cells and NK cells in the tumor xenografts of Dex-treated mice than that in the control group (Fig. 6). While NK cells were virtually undetectable in the spleen of both groups, there was a moderate decline in the relative frequency of T cells in the spleen of Dex-treated mice, compared to that in the control group.
Analysis of sera samples revealed elevated levels of CCL2 and CXCL8 in Dex-treated NSG mice bearing H1299-GR xenografts, compared to the control-treatment group (Supplementary Figure S), suggesting that senescence induction could begin as early as 5 days after Dex treatment. It is worth mentioning here that our preliminary clinical study in patients with lung cancer has also shown enhanced serum levels of chemokines CCL2, CCL4, CXCL1, and CXCL2 as early as 7 days after treatment with Dex (data not included but provided for review). Overall, the results indicate that despite considerable decline in the peripheral T cell population, induction of tumor senescence by Dex enhances tumor infiltration of T and NK cells via elaboration of chemokines. Fig. 4 Effect of neutralizing anti-chemokine antibodies on PBL migration and tumor-cytotoxicity induced by Dex-induced SASP. A PBL migration assay was performed, as described for Fig. 2, using the 7d Dex + 2d CM as chemo-attractant. The conditioned media contained neutralizing antibodies to CCL2 (2.5 ug/ml), CCL4 (2.5 ug/ml) and Pan-GRO (2.5 ug/ml) (R&D Systems) individually or in combination or antibody isotype controls, as indicated. Cells that migrated to the lower chamber were enumerated manually and using ImageJ software. Statistical significance was determined using oneway ANOVA followed by post hoc t-test to compare pairs of samples. Data are mean ± SD of three different experiments. *p < 0.001; **p = 0.04; § p = 0.047; ‡ p = 0.024. B PBL treated as described for panel A were used to perform tumor cytotoxicity assay as described for Fig. 3. Data are mean ± SD of three different experiments. *p < 0.001 Fig. 5 Role of NKG2D in SASP-induced tumor-cytolytic activity of NK cells. A PBL were cultured with 7-days-Dex + 2-days CM for 72 h. NK cell phenotype (CD56) and activation markers (CD16 and NKG2D) were analyzed by flow cytometry. B H1299-GR-Clone 4 cells were cultured with Dex (100 nM) for 72 h. Expression of NKG2D ligand MIC-A/B was analyzed by flow cytometry. C PBL were cultured with 7-days-Dex + 2-days CM plus IL-2/IL-15 in absence or presence of neutralizing antibody to NKG2D or isotype control and the tumor cytotoxicity assay was performed as described for Fig. 3. Statistical significance was determined using one-way ANOVA followed by post hoc t-test to compare pairs of samples are representative from one experiment out of 3 performed with similar results (A and B) or mean ± SD of three different experiments (C). *p < 0.001; **p < 0.05

Discussion
The findings in this study further elucidate the molecular phenotypic characteristics of senescence induced by extended Dex treatment in GR expressing lung AC cells. As the previously established isogenic lung AC model cells used in this study only differ with respect to GR expression levels and as the effects of Dex are known to be virtually exclusively mediated by GR, the senescence characteristics observed in this study must be mediated by GR [7,8]. Moreover, components of the conditioned media that continue to be detected at elevated levels after withdrawal of Dex may be linked to the irreversible senescence phenotype previously characterized by us [8]. Accordingly, among a large panel of chemokines and cytokines tested, Dex-induced senescence resulted in sustained overexpression of the chemokines CCL2, CCL4, CXCL1 and CXCL2. An initial Dex-induced increase in CXCL8 as well as a reduction in PDGF appeared to be at least partially reversible. While enhanced tumor RNA and serum levels of select chemokines were observed in the xenograft models, Dex-treated patients with lung AC also showed enhanced serum levels of these chemokines in our preliminary clinical studies. The four senescence-associated chemokines that are irreversibly induced by Dex have a common receptor (CXCR4), but they can also occasionally utilize other receptors. CCL2 and CCL4 attract and stimulate monocytes/macrophages and lymphocytes, including T cells and NK cells while CXCL1 and CXCL2 are classic chemo-attractants for neutrophils [11]. Additionally, resting CD56 dim CD16 bright NK cells have been known to express CXCR2, the receptor for CXCL1 as well as CXCL2. Ligation of CXCR2 is known to enhance migration as well as the cytolytic activity of NK cells (reviewed in [12]. NK cells have also been reported to express both CCR2 and CCR5 and to migrate to various sites including the brain in response to their respective chemokines (reviewed in [11]. Following Dex-induced senescence in lung AC cells, secretory factors in the conditioned media strongly attracted T-cells, NK cells and monocytes in vitro. Humanized NSG mice studies revealed enhanced tumor infiltration by T cells and NK cells following treatment with Dex. In vitro antibody inhibition studies showed that the Dex-induced SASP chemokines CCL2, CCL4, CXCL1 and CXCL2 mediated this mobilization of immune cells. In parallel, SASP protein(s) from the senescent cells expanded the PBL count while also inducing tumor cell cytolytic activity of PBLs and enhancing the expression of proliferation and activation markers Ki-67 and perforin, respectively, in both T cell and NK cell fractions of the PBL. However, antibodies against the above chemokines failed to inhibit the cytolytic activity of PBLs induced by Dex pre-treatment. Antibody inhibition experiments showed that the cytolytic activity could be attributed partially to the NKG2D ligand MICA/B, which was upregulated in the senescent cells. Senescent cancer cells modulate the immune system in various ways depending on the cancer cell type, mode of induction of senescence and the tumor microenvironment, in a manner that may either support tumor growth or suppress it [13,14]. Depending on the nature of the SASP components, they can be tumor promoting or tumor suppressive. For example, if the SASP components elicit cytokines such as IL-6 and GROα they can be tumor promoting [15]. However, if the SASP components elicit chemokines such as CCL2 and CXCL2, these can attract and activate an immune-dependent clearance, termed 'senescence surveillance' by NK cell and CD4 + T-cell mediated immune response [4,15]. Resultant NK and T cells can be cytolytic to surrounding tumor cells. It has been shown that drug-induced senescence in cancer models result in SASP chemokines that recruit immune cells which contribute to tumor clearance [4,5,15]. In this study, we have identified specific members of the broader SASP protein complement that persist following Dex-induced senescence of high GR expressing lung AC cells which attract and promote the anti-tumor activity of PBLs including monocytes, T-cells, and NK cells.
Several studies have, in fact, shown that the effect of glucocorticoids on lymphocytes is contextual and is determined by the cytokine milieu. While treatment with Dex in the presence of IL-7 can enhance the functions of T cells [16], the presence of IL-12 or IL-15, in fact, can enhance the frequency and cytolytic activity of NK cells [17][18][19]. We observed greater enhancement of tumor-cytolytic activity of PBLs by conditioned media obtained from the senescent lung AC cells in the presence of basal levels IL-2 (50 U/ml) and IL-15 (10 ng/ml).
Dex-induced tumor cell senescence offers a potential treatment strategy for a cohort of lung AC patients harboring high GR expressing tumor lesions, either as a single agent or in combination with standard of care chemotherapy. However, lung AC tumors tend to be heterogeneous with respect to GR expression levels among tumor cell subpopulations [6][7][8]. Therefore, a more effective treatment approach may be to combine Dex-induced senescence with an immunotherapy, as a senescent subpopulation of tumor cells could attract an immune response against the entire tumor. A recent retrospective meta-analysis of randomized clinical trials involving 7155 NSCLC patients have shown that the administration of Dex in combination with immunotherapy (checkpoint inhibitors) and chemotherapy resulted in a longer progression-free survival and overall survival compared to all other treatment groups [20]. An important consideration in utilizing Dex to enhance response to immunotherapies is that glucocorticoids inhibit T cell mediated immune responses by preventing the maturation of immune synapses via reduction of F-actin content in stimulated T cells [21], thereby reducing T cell frequency. NK cells are relatively more resistant to glucocorticoids compared to T cells, requiring a higher dose (> 1 ug/ml) or more than 3 weeks of treatment [22] for the NK cell number or activity to diminish. The suppressive effect of Dex on T cells however, is reversible [23,24]. As the senescence in lung AC induced by Dex persists after Dex withdrawal, including its immunomodulatory SASP, it may be envisioned that immunotherapies would need to be administered after a duration of Dex withdrawal to allow for recovery of T cells when enhanced response to the therapies may be expected.