Anlotinib Combined with Anti‐PD1 Potentiates Anti‐Tumor Immunity via Immunogenic Cell Death and Macrophage Reprogramming

The combination therapy of targeted drugs and immune checkpoint inhibitors has shown prominent success. In addition to blocking mutated oncogene downstream signaling, the immunological mechanism(s) underlying the anti‐tumor effect of targeted‐immuno‐therapy is not clear. In this study, anlotinib, a novel pan‐targeted tyrosine kinase receptor inhibitor (pTKI), is combined with anti‐PD1 (αPD1) as a therapeutic regimen applying to an immunocompetent mouse tumor model. Anlotinib induces immunogenic cell death (ICD), elicits anti‐tumor inflammation and infiltration, and activation of DCs and CD8+ T cells, which are enhanced by αPD1. Furthermore, anlotinib reduces KC/MCP‐1 secretion by attenuating educational effect that cancer cells imposed on tumor‐associated macrophages (TAMs) and prevents their M2 polarization by inhibiting AKT/mTORC1 and Pparδ pathways. Importantly, anlotinib plus αPD1 prolongs median progression‐free survival time compared with standard chemotherapy plus pembrolizumab as the 1st line treatment in non‐small cell lung cancer (NSCLC) patients. Thus, anlotinib treatment elicits both innate and adaptive anti‐tumor immune responses while αPD1 enhances its potency. This study provides strong evidence that combination of targeted therapy and immunotherapy is a promising regimen for treating NSCLC.

In this study, we show that anlotinib induces tumor cell ICD and thus promotes the activation and phagocytosis of DC, stimulates macrophage to release TNF- and HMGB1.Anlotinib and its combination therapy increase DC and T cell infiltration and activation, and inhibit macrophage M2 polarization by inhibit-ing AKT/mTORC1 and peroxisome proliferator-activated receptor  (Ppar)/suppressor of cytokine signaling 1 (Socs1) pathway while retains mitochondrial oxidative phosphorylation (OX-PHOS) capacity.We also demonstrate that tumor cell endows macrophage with a tumor-promoting property, which could be largely prevented by anlotinib treatment.Thus, anlotinib combined with PD1 potentiates the anti-tumor effect by eliciting both innate and adaptive immune responses and reshaping tumor-immune microenvironment, suggesting that the combination therapy may improve the efficacy for NSCLC.

𝜶PD1 Enhances the Tumor Suppressive Effect of Anlotinib
To investigate the anti-tumor effect of anlotinib, we inoculated 1 × 10 6 LLC s.c. in C57BL/6 mice, which is a classic syngeneic tumor model, and gave a two-week on/one-week off treatment (Figure 1A).We found that about 40% mice in control group and PD1 group survived on day 21, while anlotinib treatment increased the survival rate to 75% and all mice in the combination group survived (Figure 1B).Anlotinib and the combination groups had limited body weight loss (Figure S1A, Supporting Information), suggesting that the treatments had low toxicity.
Compared with control and PD1 monotherapy groups, anlotinib significantly inhibited tumor growth.When combined with PD1, anlotinib further reduced tumor size (Figure 1C,D) and splenomegaly (Figure S1B, Supporting Information), suggesting a synergistic effect of the two drugs.
Next, we compared histological evidence of proliferation, necrosis, and inflammation in tumors.H&E staining showed that anlotinib induced tumor cell death and shattered tumor mass into a loose state (Figure 1E,F).Tumor section i combination group exhibited extensive necrosis regions (Figure 1E, black arrows) with more inflammatory cell infiltration (Figure 1E, red taper) than did anlotinib group.Ki67 staining showed a marked tumor cell proliferation in control and PD1 groups while opposite in anlotinib group.The combination group had the least cell proliferation (Figure 1F).
Thus, anlotinib suppressed tumor growth and proliferation, improved survival rate of tumor-bearing mice and induced intratumoral inflammation, and this anti-tumor efficacy could be enhanced by PD1, suggesting a synergistic effect for the targetedimmunotherapy combination.

Anlotinib Elicits Enhanced Innate and Adaptive Immune Responses in Tumors
We next asked whether anlotinib altered tumor microenvironment and the immune cells in tumors were analyzed using FACS (Figure S2A, Supporting Information).Compared with control and PD1 groups, anlotinib and combination groups had increased proportions of CD4 + and CD8 + T cell and the combination group had more NKT cells than anlotinib group (Figure 2A).Importantly, anlotinib combination group had higher proportions of IFN- or TNF- producing CD8 + T cell (Figure 2B,C, Figure S2B, Supporting Information).Consistently, anlotinib and combination groups had more CD4 + and CD8 + T cells and natural killer (T) cells in their spleens (Figure S2C,D, Supporting Information).CD107a, a marker as degranulation of effector cell activation was upregulated on CD8 + T cells (Figure S2E,F, Supporting Information).Thus, anlotinib treatment elicited robust T cell response in tumors.
The increased expression of PD1 on T cell represents an active state, but persistent high expression of PD1 may cause T cell exhaustion. [23]In fact, the expression of PD1 could be intrinsic or therapy-induced in a dynamical way within tumor microenvironment, and therefore the effector T cell-mediated anti-tumor immunity would be attenuated as treatment progresses. [24]Thus, it is reasonable to combine ICI therapy as a common backbone with conventional monotherapy under proper situations. [25,26]e found that PD1 expressed at low levels on CD4 + and CD8 + T cells in control and PD1 groups, whereas its expression was boosted after anlotinib treatment and even higher in combination group (Figure 2D-G).These results suggested that anlotinib treatment resulted in T cell activation, which was partially reinforced by PD1.The upregulation of PD1 after anlotinib treatment justified anlotinib and PD1 combination therapy.
To investigate whether anlotinib directly promoted T cell activation in vitro, naïve CD4 + and CD8 + T cells were isolated from WT mouse spleen and treated with anti-CD3/CD28 antibodies.LLC, anlotinib or PD1 was added overnight.Compared to T cells alone, the addition of LLC impaired TNF- secretion (Figure S2G,H, Supporting Information) and PD1 expression (Figure S2I,J, Supporting Information), suggesting impairment of T cell activation.Anlotinib treatment didn't affect CD4 + or CD8 + T cell activation in the presence of LLC.However, under the same condition, anlotinib combined with PD1 induced more IFN- and TNF- secretion from CD4 + and CD8 + T cells (Figure S2G,H, Supporting Information), suggesting that anlotinib didn't regulate T cell response directly, but it weakened the suppressive effect of tumor cell to T cell when PD1 was present.

Anlotinib Inhibits M2-Like Macrophage Polarization and Induces Inflammation in Combination with 𝜶PD1
As the adaptive immune responses were instructed by innate immunity, we next studied the innate immune responses in tumors.The percentage of CD45 + immune cells was increased after anlotinib treatment, as opposed to the CD45 − non-immune cells (mainly tumor cells) that had increased PD-L1 expression (Figure 2H).NK cells, monocytes, and antigen-presenting cells (APCs) expanded in the combination group (Figure 2I,J).Anlotinib reduced the total proportion of tumor-infiltrating macrophage and upregulated PD1 (Figure 2K).The proportion of CD11b + F4/80 + CD206 + M2-like macrophage was dramatically decreased by anlotinib, while that of CD11b + F4/80 + MHC II + M1-like macrophage was increased by combination treatment (Figure 2L).The ratio of M1/M2 rose upon anlotinib treatment, which was even higher in combination group (Figure 2M).These results suggested that anlotinib treatment inhibited M2 polarization in tumors.
As for pro-inflammatory response, combination treatment promoted IFN-, TNF-, IL-6 release (Figure 2N) and induced pro-inflammatory cytokine gene expression in tumors (Figure 2P).Cxcl10, a chemoattractant that promotes lymphocyte recruitment, was also upregulated in mRNA level (Figure 2P).MCP-1 and KC were decreased in serum after anlotinib or its combination treatment (Figure 2O).The expression of M2 marker Arg-1 and tumor angiogenesis marker Vegf- was significantly reduced by anlotinib (Figure 2Q).Taken together, anlotinib and its combination of PD1 promoted pro-inflammatory response and reprogrammed macrophage polarization towards M1 state.
Considering the enhanced immune responses and multiple death patterns, we asked whether anlotinib could induce ICD and thus promoted DAMP expression or release. [27,28]Anlotinib induced cell surface exposure of calreticulin (calre) (Figure 3E-G) and ATP release (Figure 3H), hallmarks of ICD, in a dosedependent manner.To further confirm the ICD effect, we vaccinated mice with anlotinib-treated LLC (AL) and found the incidence of oncogenesis in AL-immunized mice was much lower than control groups (Figure 3I).As DCs can be activated by recognizing ICD, we generated bone marrow-derived dendritic cells (BMDCs) and co-cultured them with GFP-conjugated LLC that was pre-treated with anlotinib.We found that anlotinib treatment increased the uptake of tumor cell by BMDCs (Figure 3J).The expression of CD11c and co-stimulatory molecules were upregulated on the surface of BMDCs (Figure 3K,L) and IL-1 release was elevated (Figure 3M).When AL was used to stimulate macrophage, TNF- and HMGB1 were secreted in a dosedependent manner (Figure 3N), but IL-6 secretion was not affected (Figure 3O).These results demonstrated that anlotinib induced ICD and enhanced the phagocytic activity and activation of DCs.Anlotinib-induced tumor cell death also promoted TNF- and HMGB1 secretion from macrophage.

Anlotinib Induces ICD and DC Activation In Vivo
Having found the property of ICD induced by anlotinib in vitro, we further verified these findings in vivo.In mouse tumors (Figure 1A), the expression of calre on CD45 -cells (Figure 3P) and proportion of CD11b + MHCII + DCs among CD45 + cells (Figure 3Q) increased in anlotinib group, which was even higher in anlotinib and PD1 combination group.The expression of CD80, CD86 and MHC II on DCs and IL-1 in tumor tissue were upregulated in anlotinib and the combination groups (Figure 3R-U).Hence, anlotinib induced ICD and promoted DC activation in vivo, which could be enhanced by PD1.

Interaction between Macrophage and Tumor Cell Boosts MCP-1 and KC Production, Which Could be Attenuated by Anlotinib
In tumor microenvironment, tumor cell convert macrophage from "guarders" towards "turncoats" to contribute to tumor progression.Previous study reported that MCP-1 and KC in blood recruited CD11b + Gr1 + monocytes and promoted tumor growth and metastasis. [29,30]By treating BMDM with anlotinib, the expression of genes related to adhesion, communication, and chemokine receptors was downregulated (Figure 4A).Anlotinib increased PD-L1 expression on BMDM and PD-L1 level was further elevated upon IFN- treatment in both BMDM and LLC (Figure 4B,C), which indicated that anlotinib might regulate cell state and function before reaching toxic effect.Given all this, we used 4-8 μm anlotinib as a proper concentration range, which preserved drug activity without causing cell death (Figures 3A,  4B,C).To explore the interaction between macrophage and tumor cell, primary peritoneal macrophages (PM) were co-cultured with LLC.LLC stimulated PM to secret MCP-1 and KC (Figure 4D-G, Figure S4A, Supporting Information) and LLC secreted much more MCP-1 in the co-culture system (Figure 4D,F).When PM and LLC were separated by transwell, the amount of MCP-1 and KC decreased by half and that was almost inhibited by anlotinib (Figure 4D,E).As toll-like receptor 4 (TLR4) is important for KC production in macrophage by LPS stimulation, [31] we co-cultured LLC with Tlr4 −/− PM and the amount of MCP-1 from Tlr4 −/− macrophage was comparable to Tlr4 +/+ ones (Figure 4D), suggesting that production of MCP-1 was independent of TLR4.Culture medium from LLC stimulated Tlr4 −/− macrophage to secret KC (Figure 4G), but not much MCP-1 (Figure 4F).Production of MCP-1 and KC from LLC was largely elevated no matter which conditioned medium was added (Figure 4F,G).These results strongly demonstrated that direct or indirect interaction between tumor cell and macrophage promoted MCP-1 and KC secretion, which could be attenuated by anlotinib.The production of MCP-1 was independent of TLR4.
Using neutralizing antibody, we found that KC neutralization promoted MCP-1 secretion (Figure 4H) in LLC or macrophage-LLC co-culture system.However, KC secretion was not affected by MCP-1 neutralization (Figure 4I), thus illustrating KC played a negative regulatory role on MCP-1 production in the upstream signal.

Anlotinib Impairs M2 Polarization through AKT/mTORC1 and Ppar𝜹/Socs1 without Altering M1 Polarization and Metabolism
Animal model showed anlotinib plus PD1 promoted M1 polarization and anlotinib inhibited M2 marker expression in vivo (Figure 2L,M,P,Q), but the mechanism is not clear.RNA-Seq results showed that, after anlotinib treatment, the expression of M2-related genes was downregulated, but the expression of M1-related genes did not change (Figure 5A).KEGG annotation analysis suggested that anlotinib widely engaged in immune response and tumor-related signaling pathways (Figure 5B).We found that anlotinib didn't induce IL-6, TNF- and MIP-2 production in macrophage (Figure S5A, Supporting Information), and that the expression of Nos2 and IL-6 was not affected when LPS/IFN- existed (Figure 5C).Anlotinib didn't in-MFI) value of PD1 expression gating on CD4 + or CD8 + T cells.Representative images of immunofluorescence sections of tumor for CD4 (green), CD8 (red), PD1 (yellow), and nuclei (blue).Representative histograms represent quantification of CD4 and PD1 double positive or CD8 and PD1 double positive cells in the merge images.Data were shown as mean ± SEM, each with n = 5 samples/group.Original bar, 50 μm.H-K) Innate immune patterns and PD1/PD-L1 relative expression.L,M) TAM polarization was analyzed by FACS.N) Cytokine in tumor homogenates was normalized by total protein and O) concentration of MCP-1 and KC in serum.P,Q) RT-PCR analysis of gene expression within tumor.Data were combined from two independent experiments and unpaired, two-tailed Mann-Whitney U test was used unless otherwise stated.Throughout, *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.5D), suggesting that anlotinib did not alter pro-inflammatory response in macrophage directly.
40] We found that the expression of Ppar was inhibited by anlotinib (Figure 5E, Figure S5B, Supporting Information).Socs1 and Arg-1 were largely inhibited by anlotinib and GSK3787 (Ppar antagonist) (Figure 5E,F).Then, anlotinib restrained GW501516 (Ppar agonist)-induced activation of M2 polarization (Figure 5G).There was no significant change in Klf4 expression (Figure 5E,G).As Klf4-mediated M2 polarization is functioned by suppressing the after treating LLC with anlotinib.I) LLC treated with anlotinib, freeze-thawed (F-H) three times or only medium were inoculated s.c.into the lower right flank of C57 mice.Mice were rechallenged with live LLC two weeks later.Percent of oncogenesis was recorded and Log-rank (Mantel-Cox) test was used, n = 14-16.J-M) BMDCs were stimulated with anlotinib-treated LLC-GFP and the percent of CD11c + BMDCs of total cells was shown (K) and LLCphagocyted BMDCs was gating GPF + CD11c + subset on total CD11c + BMDCs (J).Co-stimulatory molecules (L) expressed on CD11c + BMDCs and IL-1 (M) were detected.N,O) BMDMs were stimulated with anlotinib-treated LLC for 16 h.Pro-inflammatory cytokines were determined by ELISA.Data were combined from 2-3 experiments.P,Q) As in animal model, abundance of calre in CD45 − tumor cell and percentage of DCs were shown. [71]Expression of co-stimulatory molecules on DCs and IL-1 in tumor homogenates were quantitated (U).Data were combined from two independent experiments and unpaired, two-tailed Mann-Whitney U test was used unless otherwise stated.Throughout, *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.Next, we accessed glycolysis and mitochondrial OXPHOS of anlotinib-treated macrophage and tumor cell.When treating BMDMs with LPS/IFN-, the glycolysis, glycolytic capacity and glycolytic reserve didn't show significant alteration upon addition of anlotinib (Figure 5I).However, anlotinib elevated the maximal and spare respiration capacity of mitochondria, which were even higher than IL-4-treated BMDMs (Figure 5J), suggesting a higher efficiency of mitochondrial respiration against external stimulus.These results demonstrated that anlotinib differentially affected macrophage metabolism in the context of distinct polarization, mainly increasing OXPHOS without impairing glycolysis (Figure S5C, Supporting Information).As for tumor cell, anlotinib inhibited ECAR and OCR (Supplementary Figure S5D, 5E), demonstrating the energy metabolism of tumor cell was more sensitive to anlotinib.In animal model, anlotinib and its combination treatment inhibited the phosphorylation of AKT, S6k and S6, as well as Arg-1 and Socs1 expression in protein level (Figure 5K) within tumors.The decreased expression of Ppar, Socs1, and Arg-1 in mRNA level was consistent with in vitro results (Figures 5L, 2Q).The expression of Klf4 didn't show significant difference among different groups (Figure 5L).These data suggested that anlotinib inhibited M2-tendency signaling pathway in vivo.

Anlotinib-Induced Anti-Tumor Immunity Mainly Depends on CD8 + T Cells, While TAMs Preferentially Promote Tumor Progression
Finally, we asked which kind(s) of immune cells exerted more efficient tumor-suppression efficacy under anlotinib combination therapy.By depleting CD4 + T cells, CD8 + T cells and macrophages (Figure 6A, Figure S6A,B, Supporting Information), we observed that CD4 + T cell depletion almost resembled anlotinib combination treatment (blue curve, as a treatment control) (Figure 6B), whereas depletion of CD8 + T cells or both CD8 + and CD4 + T cells exacerbated tumor progression indistinguishably (Figure 6B).Tumor volume further shrunk after macrophage depletion, suggesting macrophage aggravated tumor progression in general (Figure 6B).MCP-1 and KC in serum didn't decrease when only CD4 + T cells were depleted, while both of them elevated as macrophages or CD8 + T cells were depleted (Figure 6C,D).TNF- was also decreased at protein or mRNA levels after depletion, instead of IL-6 (Figure 6E,F).IFN- was reduced when macrophages or CD8 + T cells were depleted (Figure 6F).The upregulation of PD1 and PD-L1 in tumor was dependent on macrophages and CD8 + T cells, in line with IFN- levels (Figure 6G), suggesting that the expression of immune checkpoint molecule fluctuated as immune patterns changed and perhaps it impacted on the long-term efficacy of immunotherapy.
Macrophage depletion reduced intratumoral MCP-1 and KC (Figure S6C, Supporting Information), further confirming that deficiency of interaction between tumor cells and macrophages reduced tumor-related chemokine/inflammation.The decreased TNF-, IFN-, and PD1/PD-L1 may be partially attributed to the impaired activation of T cells for lack of presentation from M1like macrophages. [41,42]

Anlotinib and Anlotinib Plus 𝜶PD1 Suppress Tumor Progression of NSCLC Patients
In clinical practice, anlotinib plus PD1 and PD1 monotherapy both reduces tumor size from basal level to best response (BR) level (Figure 7A,B,E), but anlotinib plus PD1 exhibited better efficacy (Figure 7C) as the 1 st line therapeutic strategy in NSCLC patients.Anlotinib monotherapy significantly reduced tumor size as the 3 rd line therapy (Figure 7D,E).However, the efficacy of anlotinib monotherapy in NSCLC as the 1 st line therapy remains unknown, since anlotinib has not been approved for monotherapy in the 1 st line treatment yet.Data from our center showed that the mPFS was 6 months for anlotinib as the 3 rd line therapy and 16.6 months for anlotinib plus PD1 as the 1 st line therapy (Figure 7F,G).In comparison, a retrospective study revealed that the mPFS was 3.8 months for docetaxel as the 3 rd line treatment, [43] and clinical trials showed that mPFS of pemetrexed/paclitaxel combined pembrolizumab in NSCLC as the standard 1 st line treatment was 8.8 months and 6.4 months respectively. [44,45]Although data from these studies were biased due to different clinical backgrounds, the prolonged mPFS suggests a better efficacy of anlotinib and its combination of PD1.Both CXCL1 and CCL2 (MCP-1 and KC homologs in human) levels in patient sera decreased after combined therapy (Figure 7H,I).However, PD1 monotherapy didn't decrease CXCL1, and even elevated CCL2 level in patient sera (Figure 7J,K).The increased CCL2 in blood might be a rational explanation why PD1 worked less efficient than combining with anlotinib or as monotherapy.These results demonstrated that anlotinib suppressed advanced tumor progression in posterior line treatment, and its combination therapy worked more efficiently as the 1 st line treatment in NSCLC patients.

Discussion
This research identifies a heretofore unappreciated role of anlotinib in driving multiple tumor-suppression mechanisms, whereby combination with PD1 augments anti-tumor immunity.Distinguishing from previous studies mainly focused on tumor itself, we herein try to explore how anlotinib affects innate of M2-related genes of BMDMs treated with IL-4, anlotinib, GSK3787, GW501516, or combinations were as indicated for 16 h, and Immunoblotting of BMDM lysates was performed.I,J) BMDMs were treated by LPS/IFN-, IL-4, anlotinib, or specified combinations as indicated for 16 h.OCRs and ECARs were accessed using a Seahorse extracellular flux analyzer.Data were 5-6 wells from a representative of two independent experiments.K,L) Tumor tissue homogenate was subjected to western blot analysis.Results were quantitated by Image J and plotted with bar graph.Data were representative of two independent experiments with n = 5-7 per group.Data were representative of at least three independent experiments and presented as mean ± SEM, and unpaired, two-tailed Mann-Whitney U test was used unless otherwise stated.Throughout, *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.and immune responses and reprograms macrophage.These results demonstrate that anlotinib, especially combined with PD1, sharply inhibits tumor growth by inducing ICD and protective inflammation, thus conferring a prerequisite for T cell infiltration and activation.We also demonstrate that anlotinib weakens interaction between tumor cell and macrophage to decrease MCP-1/KC production and inhibits M2-polarization while efficient mitochondrial respiration capacity is reserved.Collectively, these findings identify the potently comprehensive anti-tumor mechanisms underlying anlotinib and its combination with PD1 (Figure 8), which evidences it a therapeutically exploitable modality for clinical application.
Anlotinib plus sintilimab as the 1 st line therapy in phase 1b trial (NCT03628521) shows objective response rate of 72.7% and mPFS 15 months, [46] and anlotinib combined with TQ-B2450 (PD-L1 mAb) exhibits inspiring therapeutic effects in biliary tract cancer, [47] ovarian cancer, [48] and drive-gene negative NSCLC patients as the 2 nd line treatment (NCT03910127).All these clinical trials verify that anlotinib plus ICI has promising anti-tumor efficacy and manageable toxicity.In our research, we confirm that anlotinib plus PD1 works efficiently compared to single drug.However, it cannot be denied that some differences between anlotinib and combination group are tiny or even no significance.We think it may attribute to the unique character-istic of anlotinib that possesses a much higher tissue affinity and longer terminal half-lives (t 1/2 ) [11,49,50] distinguished from other TKIs, such as imatinib and sorafenib.Surprisingly, the level of tissue exposure of anlotinib after oral administration is as much as 197 times higher than the associated systemic exposure level, and is 13 times higher in tissue than in blood of tumor-bearing mice. [50]Anlotinib itself would markedly inhibit tumor growth reaching a relatively high concentration, whereas PD1, acting as an "immune brake controller," usually exerts effector function after a certain amount of T cell infiltration induced by chemotherapy or targeted therapy, [9,10] thus indicating PD1 may work in a relatively slow and long-term manner compared to anlotinib in the combined therapy.Although the anti-tumor effect from anlotinib dominates in the combination therapy in some extent, the further shrinkage of tumor size in combination therapy still makes sense in clinical practice.
Based on the enhanced immune responses that transform "cold" tumor into a more "hot" one that is sensitive to ICIs, [51] we consider ICD as a probable consequence caused by anlotinib.[54] From a general point of view, an ICD-inducer should possess at least two properties that a) induce high immunostimulatory DAMPs release or surface exposure by stressed or injured tumor cells without any adjuvant and b) elicit a protective immune response against tumorigenesis as a vaccine in immunocompetent mice.All these features validate ICD-inducer drugs a long-lasting and potentiated anti-tumor effect in prolonging OS and PFS of patients. [55,56]In our settings, both in vivo and in vitro data manifest that anlotinib induces ICD, thus facilitating DC maturation and activation, phagocytosis and IL-1 secretion to promote T cell activation. [57,58]particularly plus PD1 as a combined therapy.
Accumulating evidence suggests that macrophage plays a nonnegligible role in regulating tumor progression and constantly reshapes tumor-immune soil. [59]M1-like M is commonly considered to exert host defense and anti-tumor effect by producing pro-inflammatory cytokines and reactive oxygen/nitrogen species.M2-like M produces anti-inflammatory and immune suppressive cytokines, thus not only attenuating devastating immunity against parasites and overwhelmed inflammation, but also promoting matrix remodeling, angiogenesis, and immunosuppression to facilitate tumor progression and metastasis. [60,61]nhibitors targeting colony-stimulating factor-1 receptor (CSF-1R) [62] and mAbs neutralizing CC-chemokine ligand 2 [63] are shown to restrain tumor growth and metastasis. [64]The Bruton tyrosine kinase inhibitor ibrutinib reprograms macrophage toward an M1-like phenotype and thereby stimulates CD8 + T cells in a macrophage-dependent manner. [65]Evidence above strongly demonstrates that targeting TAM or transferring macrophage from M2-like to M1-like phenotype could activate anti-tumor immune response and suppress tumor progression.
The shifts in mitochondrial metabolism are critical for macrophage polarization.Generally, M1-like M extremely depends on glycolysis and exhibits an impaired OXPHOS transforming arginine into nitric oxide via iNOS activity. [66,67]Whereas M2-like M prefers utilizing OXPHOS to support energy metabolism, which, compared to glycolysis, works slower but produces a larger amount of ATP [68,69] and metabolizes arginine by enhanced arginase-1. [70]Anlotinib does not alter LPS/IFN--induced increase of glycolysis and glycolytic reserve, which is identical to the fact that glycolysis fuels pro-inflammatory response.As to OXPHOS, anlotinib improves maximal respiration and spare respiratory capacity in M2-like M, even if it inhibits M2 polarization.It implies a unique feature of anlotinib, before achieving the significant cytotoxic effect, that reprograms macrophage metabolism toward an improved respiratory efficiency in anti-tumor status, which needs further investigation.
Up to now, anlotinib monotherapy is not approved for the 1 st line clinical therapy, so it is not possible to compare the efficacy between anlotinib and PD1 in the initial treatment.Thus, we take the mPFS of patients who used the standard 3 rd line chemotherapy as a control, and anlotinib exhibits mPFS 2.2 months longer than docetaxel in the 3 rd line treatment.The mPFS is 16.6 months in anlotinib plus PD1 group, which is much longer than that of chemotherapy plus pembrolizumab as the 1 st line treatment.However, the main limitation of our study is that the patient sample size in the 1 st line anlotinib combination group is relatively small and four of total twentyseven follow-up patients have PFS of more than 30 months with a favorable prognosis.In conclusion, anlotinib and its combination of PD1 prove to be potent therapeutic modality for NSCLC management.

Conclusions
This study elaborates that anlotinib, identified as an ICD inducer, suppresses tumor progression and activates anti-tumor innate and adaptive immunity via immunogenic cell death which promotes activation of DC and confers TNF- secretion from macrophage.Anlotinib weakens the boosted MCP-1 and KC production induced by tumor-macrophage interaction, and restrains macrophage M2 polarization through inhibiting AKT/mTORC1 and ppar without impairing mitochondria function.Anlotinib combined with PD1 restrains tumor progression and prolongs mPFS of NSCLC patients as the 1 st line treatment.

Experimental Section
Mice: Six-to ten-week-old C57BL/6J male mice were purchased from SPF (Beijing) Biotechnology Co., Ltd.B6.B10ScN-Tlr4lps-del/JthJ (Tlr4 −/− ) mice were obtained from Fudan University, Shanghai Medical College (Mingfang Lu).All mice were housed in a specific pathogen-free facility and all animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC).All protocols followed the Guide for the Care and Use of Laboratory Animals.
Weight-matched male C57 mice were injected s.c. with 1 × 10 6 LLC into lower right flank.Anlotinib (AL-3818), PD1 (clone# RMP1-14, BioXCell), or combination treatment was commenced when the average tumor diameter reached 2-3 mm.Anlotinib dissolved in normal saline was administrated 2 mg kg −1 i.g.consecutively 5 days a week with last 2 days withdrawal in terms of a standard 2-week on/1-week off therapeutic strategy for a total 3-week regimen.PD1 was injected i.p. 200 μg mouse −1 every 3 days till the end of experiment.Mice were sacrificed by inner canthus vein bleeding and their spleens, tumor mass and serum were collected for further analysis.For the immunization study, 1 × 10 6 LLC that was either freeze-thawed three times or treated with 50 μm anlotinib 16 h were injected s.c.into the lower right flank of C57 mice.Two weeks later, the immunized mice were re-challenged with 1 × 10 6 live LLC and tumor progression was monitored.Tumor volume was calculated using the standard formula: ½ × length × width 2 , where the length was the longer dimension.Anlotinib was developed and generated by Chia-tai Tianqing Pharmaceutical Co., Ltd, China.
Single Cell Preparation and Flowcytometry: The isolated tumor mass was weighted and dissociated by gentleMACS (Miltenyi Biotec) and digestion buffer containing RPMI, 1 mg mL −1 collagenase IV (Sigma), 10 U mL −1 DNase I (Sigma), and 0.5 mg mL −1 hyaluronidase (Sigma) at 37 °C for 30 min.After lysing erythrocytes, single cell suspension was filtered by 70 μm strainer and incubated with FcRIII/II-blocking antibody (antimouse CD16/32, BioLegend) to prevent non-specific binding before cell viability staining (Zombie Dye, BioLegend).Then, incubated with FACS antibodies on ice for 30 min, the samples were analyzed by BD FACSCanto II or LSRFortessa and the data were processed using Flow Jo software (TreeStar, Inc).Apoptosis induced by anlotinib was detected by Annexin V FITC and PI detection kit (catalog# 556 547, BD).All operating procedures were performed on ice and samples were detected within 1 h after staining.
In Vitro BMDM Experiments: Bone marrow cells were harvested from femurs and cultured in cRPMI supplemented with 20 ng mL −1 M-CSF for 7 days with refreshing culture medium every 2 days to generate BMDMs.BMDMs were stimulated with 4 μm anlotinib, 10 ng mL −1 LPS and 10 ng mL −1 IFN- or 10 ng mL −1 LPS and 10 ng mL PM and LLC Co-Culture Experiment: Briefly, PM was obtained from Tlr4 +/+ or Tlr4 −/− mice by peritoneal lavage using cold PBS containing 5 mm EDTA.Total cells from lavage liquid were collected and re-suspended in cRPMI.After adherence for 5 h at 37 °C and washing non-sticking cells away, Tlr4 +/+ or Tlr4 −/− PMs and prepared LLC were cultured with fresh cRPMI for another 8 h and collected supernatant as the conditioned medium to stimulate naïve LLC or PMs for 16 h.PMs co-cultured with LLC that pre-treated with or without 8 μm anlotinib overnight in the presence of an 8 μm pore size transwell for 16 h.As for chemokine neutralization, 5 μg mL −1 anti-mouse MCP-1 (clone# 2H5, BioLegend) or 2.5 μg mL −1 anti-mouse KC (clone# 48 415, R&D), and the same amount of hamster IgG isotype (clone# HTK888, BioLegend) or rat IgG2a isotype (clone# RTK2758, BioLegend) was added into conditioned medium or co-culture system to deal with indicated cell types for 16 h.
ATP was quantified by an Enhanced ATP Assay Kit (S0027, Beyotime).PM was obtained and prepared as outlined above and then was stimulated by LLC which was pre-treated with gradient anlotinib of indicated concentration overnight for 16 h.The supernatant was collected for measuring IL-6, TNF- (BD), and HMGB1 (Genie) by ELISA.
T Cell and Macrophage Depletion Experiment: To deplete CD4 + and CD8 + T cells, anti-mouse CD4 (clone# YTS119, BioXCell) or CD8 (clone# YTS169.4,BioXCell) neutralizing mAb was intraperitoneally injected 200 μg mouse −1 once a week starting from the day before LLC inoculation.As for macrophage depletion, clodronate liposomes (F70101C-N, FormuMax) was intraperitoneally injected 50 μL mouse −1 every 3 days.The efficiency of depletion of three cell types was determined by flow cytometry.At the endpoint of experiment, peritoneal lavage fluid or blood were collected to verify the depletion efficiency by detecting CD11b + F4/80 + macrophage or CD4 + /CD8 + T cells using FACS.

Figure 1 .
Figure 1.PD1 potentiates anlotinib-caused suppression of tumor progression.A) Schematic diagram of animal model.B) Survival fraction after drug administration was monitored.Log-rank (Mantel-Cox) test was used.C) Tumor volume was calculated against ½ × length × width 2 , where the length was the longer dimension and tumor growth curve was plotted (n = 9-13).Two-way ANOVA test was used.D) Tumor volume and weight were shown as mean ± SEM.Unpaired, two-tailed Mann-Whitney U test was used.E,F) H&E and Ki67 staining sections of tumor were shown.Black arrows indicated necrosis of tumor tissue and red tapers indicated inflammation.Blue dots indicated nucleus and brown dots indicated Ki67 expression.Original bar, 200 μm; Insert bar, 50 μm.Data were combined from two independent experiments.Throughout, *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Figure 2 .
Figure 2. Anlotinib elicits innate and adaptive immune responses in tumors.A-C) Percent of CD4 + T, CD8 + T, and NKT cells in total live cells and percent of activated CD4 + and CD8 + T cells in total CD4 + and CD8 + T cells were analyzed by FACS.D) Geometric mean fluorescence intensity (Geo

Figure 3 .
Figure 3. Anlotinib induces ICD and activates DCs.A) IC50 of anlotinib in MLE-12, LLC, LA795, and KLN205 was indicated by vertical dashed lines.B) RNA-Seq analysis was performed.Data relating to cell death and growth with >2.0-fold was shown, n = 3. C) Cell lines were treated with anlotinib for indicated timepoint before apoptosis detected by Annexin V and PI.D) LLC was treated with diverse death inhibitors 1hr before adding anlotinib for total 16 h.Data were combined from three experiments.E-H) Abundance of calreticulin on cell surface and ATP released in medium were determined

Figure 4 .
Figure 4. Anlotinib attenuates the secretion of KC and MCP-1.A) BMDMs were treated with anlotinib for 16 h before RNA-Seq analysis was performed.Data relating to cell adhesion or communication >2.0-fold difference was shown, n = 3. B,C) BMDMs and LLC were treated by anlotinib with or without IFN- for 16 h.PD-L1 expressing on cell surface was detected by FACS.D,E) Anlotinib-treated LLC co-cultured with Tlr4 +/+ or Tlr4 −/− PMs, which was separated by the 8 μm pore size transwell.MCP-1 and KC were detected by ELISA.F,G) Culture medium collected from LLC, PM, or co-culture system was diluted 1:20 as the conditioned medium to replenish different cell plate and continued culturing for another 16 h.MCP-1 and KC were detected.H,I) Five μg mL −1 MCP-1 neutralizing mAb, or hamster IgG isotype, and 2.5 μg mL −1 KC neutralizing mAb, or rat IgG2a isotype, were added into culture system along with conditioned medium.MCP-1 and KC were measured.Data were presented as mean ± SEM and unpaired, two-tailed Mann-Whitney U test was used unless otherwise stated.Throughout, *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Figure 5 .
Figure 5. Anlotinib inhibits M2 polarization and regulates mitochondria OXPHOS.A,B) RNA-Seq analysis relating to M1-and M2-polarized gene expression was shown, n = 3.The top ten Immune response-related signaling pathways with greatest changes in KEGG annotation enrichment were listed.C,D) BMDMs were treated with LPS/IFN-, anlotinib, or both.Il6, Nos2, and Immunoblotting of lysates of BMDMs were examined.E-H) Relative expression

Figure 6 .
Figure 6.Depletion of immune cell makes disparate impacts on tumor progression in context of anlotinib.A) Schematic diagram for depletion of immune cells and drug administration.B) Tumor growth curve was calculated.Two-way ANOVA test was used.C,D) Concentration of MCP-1 and KC in serum was detected by ELISA.E-I) Pro-inflammatory cytokines, chemokines and immune checkpoint molecules of tumor tissue in protein and mRNA levels were measured.Data were combined from two independent experiments and presented as mean ± SEM.Unpaired, two-tailed Mann-Whitney U test was used unless otherwise stated.Throughout, *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Figure 7 .
Figure 7. Anlotinib and combined therapy suppress tumor progression in NSCLC patients.A,B,E) Tumor size of basal and BR levels of patients who received anlotinib plus PD1 (n = 26) or PD1 monotherapy (n = 20) as the 1 st line treatment.Wilcoxon matched-pairs signed rank test was used.C) Therapeutic efficacy of anlotinib plus PD1 and PD1 monotherapy.Mann-Whitney U test was used.D) Tumor size of basal and BR levels of patients who received anlotinib monotherapy (n = 45) as the 3 rd line treatment.E) CT results of patients who received anlotinib plus PD1 or PD1 monotherapy as the 1 st line treatment, or received anlotinib as the 3 rd line treatment.F,G) PFS of patients received anlotinib (n = 47) and anlotinib plus PD1 (n = 27) treatment.H,I) CCL2 and CXCL1 in serum of patients who received anlotinib plus PD1 (n = 20) or PD1 (n = 18) as the 1 st line treatment were measured by ELISA.Throughout, *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Figure 8 .
Figure 8.The enhanced anti-tumor immunity underlying anlotinib and its combination therapy.Anlotinib induces cell death and ICD which strengthens anti-tumor immunity in three aspects: first, released DAMPs stimulate macrophages transforming from M2 phenotype towards M1; second, promote DC maturation and activation, and third, collaboratively activate T cells to produce IFN- and TNF-.Monocytes recruited by MCP-1 to tumor site, turning into mono-macs, suffer from "domestication" acquiring tumor-promoting properties within tumor microenvironment.Anlotinib attenuates the "adverse education" of tumor cell imposed on macrophage, which results in decrease of MCP-1 and KC in blood stream.As an anti-angiogenesis drug, anlotinib not only causes multi-type tumor death and suppresses tumor metabolism, but also exhibits immune-modulating function that targets Ppar and AKT/mTORC1 pathways to inhibit M2 polarization.The high expression of PD1 of activated T cells might be blocked by PD1 and thus interdicts its interaction with PD-L1.The comprehensive effects of anlotinib, especially anlotinib combined with PD1, demonstrate outstanding anti-tumor efficacy that a) increases tumor cell death, b) activates innate and adaptive immune cells, c) optimizes immune response pattern, d) converts a relative "cold" immune-soil into a "hot" one.