Tumor Secretome to Adoptive Cellular Immunotherapy: Reduce Me Before I Make You My Partner.

Today, it is widely recognized that chronic inflammation is a driver of cancer ( 1 ), being estimated that 15-20% of cancers are inflammation-related ( 2 ). This association has been observed in different contexts, such as persistent Helicobacter pylori infection or autoimmune diseases like inflammatory bowel disease that increase the risk of developing gastric cancer ( 3 ) or colorectal cancer ( 4 ), respectively. Numerous studies have found associations of inflammatory markers with a higher risk of developing cancer. For instance, 15% of patients with cardiovascular disease, after a median follow-up of 8.3 years, developed different types of cancer whose incidence was associated with high C-reactive protein (CRP) levels ( 5 ). In addition, IL6 levels are also associated with an increased risk of developing different types of cancer ( 6 ). Moreover, IL1β inhibition reduced CRP and IL6 levels and the incidence of developing lung cancer in patients with atherosclerosis who had a myocardial infarction ( 7 ). Both immune and tumor cells promote this pro-inflammatory microenvironment. Expressly, tumor cells release a secretome that displays an altered composition compared to the normal tissue from which they are derived ( 8 ). This secretome contains cytokines, chemokines, hormones, metabolites, and growth factors involved in cell-cell communication, angiogenesis, hypoxia, metastasis, extracellular matrix remodeling, and drug resistance ( 8 , 9 ), where tumor cells employ it as a mechanism of immune evasion ( 10 – 12 ). On the other side, the different subsets of immune cells will also release immunosuppressive and inflammatory factors that will shape the tumor microenvironment (TME), promoting or inhibiting cancer progression ( 13 ). The anti-tumor activity of immune cells infiltrating tumors led to the development of adoptive cellular immunotherapy administering natural killer (NK) cells, T cells, or genetically modified chimeric antigen receptor (CAR)-T cells in cancer patients ( 14 – 17 ). Clinical results administering different immune cells have been reviewed by others ( Table 1 ). However, despite promising results in these studies for some malignancies ( 26 ), immune cells do not persist long for other malignancies, and patients end up relapsing ( 27 ). Once immune cells achieve the tumor, they will have to face tumor cells and their secretome that may polarize their anti-tumor activity to a pro-tumoral one, increasing angiogenesis and enhancing tumor growth ( 28 ). Moreover, after chemotherapy treatment, tumor cells can reach a senescent state, known as therapy-induced senescence (TIS), that shapes the tumor secretome to a variety of pro-inflammatory and angiogenic proteins known as “senescence-associated secretory phenotype” (SASP). The SASP may enhance the immune response at initial stages and contribute to a favorable environment for tumor growth at late stages ( 29 ). For example, senescent fibroblasts, much more than pre-senescent fibroblasts, secrete VEGF that causes premalignant and malignant epithelial cells to form tumors, suggesting that although cellular senescence suppresses tumorigenesis early in life, it may also promote cancer ( 30 ). Reviews indicating clinical results with different types of immune cells administered in immunotherapy studies in cancer patients. NK, natural killer; CAR, chimeric antigen receptor; TILs, tumor infiltrating lymphocytes. Here, we review how the tumor secretome can shape the immune response achieving a state when immune cells no longer recognize tumor cells and instead, they secrete proteins that breed the TME. We will specifically focus on the impact on T cells, CAR-T cells, and NK cells, which are currently used in adoptive cellular immunotherapy ( 14 – 17 , 31 ), and macrophages due to their relevant role in removing dying/senescent tumor cells after cancer treatment ( 32 ). The impact of these molecules is summarized in Table 2 . Moreover, we will review the effect of the tumor secretome in the immune response when tumor cells become senescent due to chemotherapy treatments. Impact of secreted factors in the tumor microenvironment (TME) over the different immune cell populations and description of receptors acting as eat me or don’t eat me signals for phagocytic activity of macrophages. PB-NK, peripheral blood NK cells; dNK, decidual NK cells; T-reg, regulatory T cell; APCs, antigen presenting cells; IFN-γ, interferon-γ; TGFβ, transforming growth factorβ; FGL1, fibrinogen-like 1; GAL-9, galectin-9; IL, interleukin; HLA, human leukocyte antigen; miR, microRNA; OSM, oncostatin-M; VEGF, vascular endothelial growth factor; MIF, macrophages migration inhibitory factor; ST2, suppression of tumorigenicity 2.

All authors listed have made a substantial, direct, and intellectual contribution to the work, and approved it for publication.

Finally, we call the reader’s attention to the type of cell death activated in tumor cells after the attack of immune cells in adoptive cellular immunotherapy. Inflammatory forms of cell death include pyroptosis, which activates the NLRP3 inflammasome, leading to IL1β production ( 237 ). As previously mentioned, IL1 signaling controls the SASP production ( 204 ). Of interest, CAR-T cells and NK cells used in adoptive cellular immunotherapy activate pyroptosis when they encounter the tumor cell ( 145 , 238 ). These events suggest that the consequences of this IL1β release should be considered. Expressly, inflammasome activation and pyroptosis execution represent a double edge-sword in cancer immunotherapy, as on one side, pyroptosis executes cell death. On the other side, pyroptosis and IL1β production activate multiple signaling pathways and inflammatory mediators that promote tumor growth and metastasis in cancer models ( 239 , 240 ), triggering TAMs to boost tumor angiogenesis ( 241 ). Moreover, the role of pyroptosis is highly relevant to attracting other immune cells through IL1β and IL18 secretion. These events are observed in microbial infections, where pyroptosis attract immune cells to kill the previously trapped pathogen and remove the infected cell ( 208 , 242 ). In adoptive cellular immunotherapy, removing dead tumor cells after being killed by immune cells is required, suggesting an advantage of pyroptosis in this context.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Studies have demonstrated that chemotherapy treatment can lead to acquired resistance and the emergence of more aggressive tumor cells. In this regard, the tumor secretome is shaped by chemotherapy treatment that will impact the immune response and increase tumor aggressiveness. For instance, in breast cancer, IL6 release after treatment converts differentiated tumor cells to cancer stem cells through the IL6-JAK1-STAT3 pathway ( 151 ). In non-small cell lung cancer, cisplatin induces IL6 secretion that increases tumor progression and resistance to treatment through up-regulation of anti-apoptotic proteins and DNA repair associated genes ( 152 ). Paclitaxel enhances IRE1 RNase activity that leads to the production of IL6, IL8, CXCL1, GM-CSF, and TGFβ2 in breast cancer cells contributing to the expansion of tumor-initiating cells ( 153 ). Doxycycline treatment in squamous cell carcinoma leads to TGFβ secretion that activates the TGF-β/SMAD pathway increasing tumorigenic potential ( 154 ). Treatment with kinase inhibitors causes secretion of positive mediators of the AKT pathway, including IGF1, EGF, ANGPTL7, and PDGFD, accelerating the expansion and dissemination of drug-resistant clones ( 155 ). Docetaxel induces secretion of extracellular vesicle-encapsulated miRNAs, including miR-9-5p, miR-195-5p, and miR-203a-3p, which down-regulate the transcription factor ONECUT2, leading to up-regulation of stemness-associated genes, that stimulate cancer stem-like cells and resistance to therapy in breast cancer ( 156 ). In addition, chemotherapy and radiotherapy treatments trigger a premature state of senescence in tumor cells termed “therapy-induced senescence” (TIS) that will shape the tumor secretome ( 29 , 157 ). TIS can reactivate the cell cycle and bring on cancer daughter cells that survive therapy more transformed than the original population ( 158 , 159 ). This secretome is unique because it is induced by senescence, being termed senescence-associated secretory phenotype (SASP). SASP includes various cytokines, chemokines, growth factors, and matrix metalloproteinases, such as IL1α, IL1β, IL6, IL8, CXCL1, CCL2, VEGF, and CXCR2 ( 29 , 160 , 161 ), that interfere with the paracrine activity of senescent cells. Of interest, SASP released by tumor cells after TIS induces transmission of senescence to non-senescent neighboring cells ( 162 , 163 ). The SASP can foster an immunosuppressive environment favoring metastasis ( 160 ), and on the other side, attracts immune cells including macrophages, neutrophils, and NK cells to remove senescent cells, a process known as “senescence surveillance” ( 164 – 167 ). Moreover, cancer is associated with aging. A physiological consequence of aging is the development of immunosenescence due to a functional degradation of the thymus, resulting in decreased functional naïve CD4 and CD8 T cells and a peripheral oligo-clonal expansion of memory T cells. These events provide a contracted T cell antigen receptor (TCR)-repertoire diversity with secretion of SASP ( 29 , 168 ). Immunosenescence associated with aging also occurs due to exposure to virus infections or chronic inflammation ( 169 ); and additional factors such as nutrition, sex, genetics, previous diseases, or tumors ( 170 , 171 ). Therefore, the immune cells of elderly cancer patients will probably be already senescent; and moreover, SASP secretion by senescent tumor cells after chemotherapy will accelerate this immunosenescence process ( 171 ). Here, we will mention some SASP factors released by tumor cells that impact the anti-tumor immune response. Studies have shown a significant accumulation of senescent T cells in certain types of cancer patients ( 172 ), and that tumor SASP induces T cell senescence leading to suppression of responses of naïve/effector T cells ( 173 ), suggesting that this might be a strategy used by malignant cells to evade immune surveillance. Transformed senescent T cells are in cell cycle arrest and develop significant phenotypic alterations, such as down-regulation or loss of CD27 and CD28.Through SASP factors including pro-inflammatory cytokines or inhibitory molecules like IL10 or TGFβ, senescent T cells will amplify the immunosenescence process. Moreover, the development of exhaustion with high expression of immune checkpoints, such as TIM-3 and other co-inhibitory receptors as CD57 or KLRG-1, will promote replicative senescence of T cells ( 174 ). TGFβ1 and TGFβ3 are early SASP factors that regulate thymic T cell homeostasis, inhibit cytotoxic T cell proliferation, and promote T-reg generation ( 175 ). Tumor senescent cells up-regulate NOTCH1 and drive a TGFβ-rich secretome that suppresses the release of a pro-inflammatory SASP and contributes to the transmission of senescence through cell-cell interaction via NOTCH-JAG1 pathway. Of interest, NOTCH1 inhibition recovers the secretion of pro-inflammatory cytokines, promoting lymphocyte recruitment and senescence surveillance ( 176 ). Senescent cells, after genotoxic stress, secrete IL6 and IL8 that promote epithelial-mesenchymal transition, increasing tumor cells’ invasiveness. Moreover, IL6 recruits myeloid cells that inhibit T cell responses ( 177 ). MAPK signaling is a relevant pathway that controls T cell senescence ( 178 ) through activation of p53, p21, and p16 ( 179 ). Recent research demonstrated that tumor-derived T-reg cells exhibit an accelerated glucose uptake, competing with effector T cells for glucose through TLR8 signaling, leading to MAPK activation, which induces T cell senescence ( 180 ). Another study showed that T-reg cells, through p38, ERK1/2 signaling, p16, p21, and p53 induce senescence in responder naïve and effector T cells. This event is reverted by the block of TLR8 signaling and or by specific ERK1/2 and p38 inhibition ( 181 ). Moreover, the p53 isoforms Δ133p53 and p53β regulate proliferation and senescence in human T lymphocytes. Thus, decreased Δ133p53 and increased p53β expression in healthy individuals and lung cancer patients associated with age-dependent accumulation of senescent CD8 T cells ( 182 ). The hypoxic TME leads to the accumulation of adenosine and tumor-derived cAMP. This cAMP is a SASP factor that induces T cell senescence in naïve/effector T cells. Of interest, activation of TLR8 signaling in tumor cells reverses this event resulting in enhanced anti-tumor immunity ( 183 ). Moreover, the accumulation of adenosine in the TME also inhibits the anti-tumor activity of T cells through the adenosine receptor A2AR, which in healthy conditions regulates immune cells protecting from inflammatory damage ( 184 ). Whereas the immunosenescence process has been widely studied in T cells, there is a lack of information related to CAR-T cell senescence. It could be exciting to delve into the mechanisms of senescence of this type of cells to find pathways to inhibit senescence without impacting their anti-tumor activity. Specifically, CAR-T cells undergo a significant in vitro expansion ( 185 ) to obtain enough CAR-T cells to treat the patients. This expansion might impact the development of senescence due to continuous in vitro proliferation. Moreover, the transfer of senescence from tumor cells in the TME mediated by cell-cell contact or through factors present in the SASP will impact CAR-T cell activity. CAR-T cells can be engineered to avoid these events. Thus, recently, CAR-T cells have been used as senolytic agents in lung adenocarcinoma to remove chemically induced senescent cells by targeting the urokinase-type plasminogen activator receptor ( 186 ). A tempting option that could be tested is to reverse early-stage senescent CAR-T cells by blocking critical mediators of this process, such as proteins involved in the DDR, p38, p53, p21, or ATM ( 187 ). However, these changes could also decrease T cell functionality by impacting other relevant functions. For instance, p38 is involved in the induction of senescence and IFNγ and TNFα secretion ( 188 ), and its inhibition have diminished these cytokines in different inflammation or virus infection models ( 189 , 190 ). Moreover, blockage of DDR and p53 involves a risk of DNA damage on T cells that might induce malignancy ( 191 ). NK cells have an essential role in the senescence surveillance of tumor cells. Senescence surveillance is initiated by the SASP that activates immune cells to clear senescent cells preventing tumor initiation ( 167 ), where both macrophages and NK cells have an important task ( 32 , 192 , 193 ). Proteins present in the SASP, such as CCL2, attract PB-NK cells to remove senescent cells through NKG2D ( 194 ). Of interest, this role of PB-NK cells removing senescent cells is also observed by decidual uterine NK cells to control embryo implantation. Specifically, dNK cells after being activated by IL15, present in the SASP, target and clear decidual cells that became senescent in an IL8 dependent manner. This mechanism of NK cells is mediated through granule exocytosis and involvement of NKG2D ( 195 ). SASP secretion by senescent tumor cells up-regulates HLA-E, the ligand of the inhibitory NKG2A NK receptor ( 196 ), and cleave NKG2D ligands inhibiting NK cell activity ( 197 ). Soluble Thrombospondin-1 (TSP1), released in the SASP, is involved in Ras-induced senescence ( 198 ). Moreover, TSP1 released by tumor cells binds CD47 on NK cells inhibiting its activity ( 199 ). CD47 is described as a relevant modulator of NK cell function in virus infection ( 200 ). Of interest, after TIS, binding of soluble TSP1 to CD47 causes emergence of tumor-resistant cells and metastasis in triple-negative breast cancer ( 201 ), and inhibits anti-melanoma NK cell activity with reduced granzyme B and IFNγ production ( 202 ) ( Figure 3 ). IL1β is another crucial molecule present in the SASP with a relevant pro-tumor activity ( 203 ). In detail, IL1 signaling controls the SASP production ( 204 ), and transmission of IL1β to neighboring cells induces cell senescence ( 205 , 206 ). A dual role for IL1β is observed in NK cell activity. For example, IL1β is required by CD56 bright NK cells to produce IFNγ ( 207 ) to activate pyroptosis, necessary for the anti-microbial ( 208 ) and anti-tumor ( 145 ) activity of NK cells. In addition, IL1β released by M1 macrophages increases NK cell cytotoxicity up-regulating NKp44 and NKG2D and triggering IFNγ production by NK cells. Of interest, these IL1β-primed NK cells can reverse M2 macrophage polarization ( 209 ). On the other side, a negative impact of IL1β has been described over NK activity. Thus, tumor-derived IL1β induces accumulation of MDSCs that impair NK cell development and functions ( 210 ). Moreover, a higher secretion of IL1β in endometrial cancer patients compared to healthy tissues correlates with infiltrating CD56 bright NK cells in the tumor with exhausted phenotype, indicated by TIGIT and TIM3 expression ( 211 ). IL6 and IL8, present in the SASP, favor the acquisition of migration/invasion and stem-like features, increasing tumor aggressiveness in breast cancer cells ( 212 , 213 ). Moreover, IL6 also inhibits NK cytotoxic activity by down-regulating perforin and Granzyme B ( 214 ). In esophageal squamous cell carcinoma, tumor cells activate the STAT3 pathway on NK cells through IL6 and IL8, leading to down-regulation of NKp30 and NKG2D on NK cells and tumor progression ( 215 ) ( Figure 3 ). In addition, increased levels of IL6 in the peritoneal fluid of endometriosis patients reduced the cytolytic activity of NK cells with down-regulation of granzyme B and perforin ( 216 ). IL8 activates and recruits immune cells ( 217 ) but also has tumor-promoting functions ( 218 ). IL8 is produced by CD56 bright NK cells ( 219 ), and stimulation with IL18 and IL12 induces higher IL8 production by NK cells ( 220 ). PGE2 secretion, present in the tumor secretome, inhibits NK cell activity ( 10 , 52 , 53 ). Moreover, PGE2 is also present in the SASP at early tumorigenesis stages, secreted by COX-2, a critical regulator of the SASP, and promotes senescence surveillance ( 221 ) ( Figure 3 ). Senescent cells show high ROS levels and lactate production that induce and maintain cell senescence ( 222 , 223 ). ROS can present contradictory effects on the activity of NK cells. Specifically, lactate production by metastatic colorectal cancer cells induces mitochondrial stress, increased ROS, and apoptosis in NK cells ( 224 ). On the other side, ROS is required for the anti-tumor activity of NK cells ( 225 ). Moreover, TIS up-regulates NKG2D ligands (MICA, MICB, and PVR) in an oxidant-dependent manner, resulting in enhanced NK cell activity against myeloma cells ( 226 ). This up-regulation of NKG2D ligands upon oxidative stress was also observed in colon carcinoma cells, leading to improved NK cell killing ( 227 ). However, in established tumors, ROS down-regulates NKp46 and NKG2D on mature CD56 dim NK cells inducing suppression of NK activity against melanoma ( 228 ) and acute myeloid leukemia cells ( 229 ). Of interest, we previously observed that cord blood-derived NK cells reduce ROS levels in multiple myeloma cells ( 230 ). This negative role of ROS in tumors has led to antioxidant treatments in cellular immunotherapy studies. For instance, as previously mentioned, in solid tumors, CAR-T cells modified to express the enzyme catalase presented an anti-oxidant capacity to protect bystander T cells and NK cells ( 116 ). All these studies suggest the beneficial and detrimental role of the SASP at early and late stages of tumorigenesis, respectively. As high levels of SASP inhibit NK cell activity, a strategy to treat advanced cancer patients with cellular immunotherapy, could be to administer senescence inhibitors to decrease the number of senescent cells. Once reduced levels of SASP are achieved, immune cells could be administered, that would be attracted to remove the remaining senescent tumor cells. Macrophages are attracted and stimulated by SASP factors including MCP-1, MIP-1α, and GM-CSF to remove senescent cells ( 231 ). Macrophages are also affected by age-related immunosenescence and the consequences of inflammaging, a chronic inflammation occurring with aging, leading to macrophage dysfunction. Increased levels of A20, a suppressor of the NFκB and MAPK signaling, mediated this dysfunction, leading to poor NFκB and MAPK activation following TLR stimulation ( 232 ). There is a disparity in the impact of TIS and the SASP in macrophage polarization and their phagocytic activity. Thus, in a model of skin aging, macrophage activity is inhibited when there are a high number of senescent cells ( 233 ). Specifically, through TNFα release, macrophages induce apoptosis in senescent fibroblasts, leading to the expression of phosphatidylserine on their surface. Phosphatidylserine is recognized by the STAB1 receptor on macrophages to promote their phagocytosis. However, SASP factors, including IL1α and GM-CSF, down-regulate STAB1 and TNFα expression, avoiding the killing and phagocytosis of macrophages, with no impact observed in the macrophage polarization ( 233 ). In a model of thyroid cancer, monocytes exposed to conditioned media from senescent thyrocytes and thyroid tumor cells, undergo M2-like polarization displaying tumor-promoting. These events were related to the production of PGE2 ( 234 ). In liver fibrosis and cirrhosis, hepatic stellate cells made senescent by carbon tetrachloride treatment produce cytokines that recruit M1 macrophages, promoting a tumor-suppressive environment. However, in the absence of p53, a promoter of senescence, the released secretome induces M2 polarization, enhancing premalignant cells’ proliferation ( 235 ). In a model of pancreatic cancer with oncogene-induced senescence, the SASP factor CXCL1 activates CXCR2 that leads to recruitment of M1 macrophages, inhibiting carcinogenesis. However, oncogene-induced senescence and SASP are bypassed at late stages, and M2 macrophages are recruited to enhance the proliferation of the transformed pancreatic cancer cells ( 236 ).

To conclude, adoptive cellular immunotherapy has emerged as a promising treatment to treat cancer patients in the last years. However, results still need to be improved in a variety of malignancies. Immune cells present a high capacity of plasticity when they receive stimuli from secreted molecules in the TME. Thus, if immune cells do not remove tumor cells, tumor secretome could modify their killer activity to an angiogenic or immunosuppressive one. A highly relevant aspect that needs to be considered to avoid these events is an efficient removal by macrophages of dying/dead tumor cells after the attack of immune cells, such as NK cells or CAR-T cells. Of interest, NK cells present additional functions to their classic killer activity that might help in this tumor cell surveillance. Inflammatory forms of cell death activated by in vitro expanded immune cells might also impact these processes. In summary, to achieve complete and permanent responses in cancer patients treated with adoptive cellular immunotherapy, all these aspects together need to be considered and count on the activity of the whole immune response and not just one immune cell population.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

This research was funded by Fondos Feder with a grant of the Institute of Health Carlos III, grant number PI20/00991.