Endogenous CCL21-Ser deficiency reduces B16–F10 melanoma growth by enhanced antitumor immunity

The chemokine CCL21 regulates immune and cancer cell migration through its receptor CCR7. The Ccl21a gene encodes the isoform CCL21-Ser, predominantly expressed in the thymic medulla and the secondary lymphoid tissues. This study examined the roles of CCL21-Ser in the antitumor immune response in Ccl21a-knockout (KO) mice. The Ccl21a-KO mice showed significantly decreased growth of B16–F10 and YUMM1.7 melanomas and increased growth of MC38 colon cancer, despite no significant difference in LLC lung cancer and EO771 breast cancer. The B16–F10 tumor in Ccl21a-KO mice showed melanoma-specific activated CD8+ T cell and NK cell infiltration and higher Treg counts than wild-type mice. B16–F10 tumors in Ccl21a-KO mice showed a reduction in the positive correlation between the ratio of regulatory T cells (Tregs) to activated CD8+ T cells and tumor weight. In Ccl21a-KO tumor, the intratumoral Tregs showed lower co-inhibitory receptors TIM-3 and TIGIT. Taken together, these results suggest that endogenous CCL21-Ser supports melanoma growth in vivo by maintaining Treg function and suppressing antitumor immunity by CD8+ T cells.


Introduction
The homeostatic chemokine CCL21 has two isoforms: CCL21-Ser, encoded by the Ccl21a gene, with a serine at the 65th position, and CCL21-Leu, encoded by multiple genes, including Ccl21b, with leucine at that position [1,2].CCL21-Ser is expressed in the stromal cells of secondary lymphoid tissues and thymic epithelial cells [2][3][4], whereas CCL21-Leu is expressed in the lymphatic endothelial cells of peripheral tissues [1,4].Both isoforms transmit intracellular signals by binding to the chemokine receptor CCR7, expressed on thymocytes, naïve lymphocytes, central memory T cells, regulatory T cells, and mature dendritic cells [5].Knockout (KO) studies in mice have demonstrated the critical roles of CCL21 in immune cell migration and tissue localization.Ccl21a-KO mice develop autoimmune disease-like symptoms due to an incomplete negative selection of autoreactive T cells during lymphocyte differentiation in the thymus [3].The Ccr7-KO mice showed attenuated thymocyte migration (from the cortex to the medulla), reduced T cell and dendritic cell localization in the secondary lymphoid tissues [6][7][8].The dysregulation of immune response in those gene-KO mice indicates that CCR7 signaling is essential for proper immune cell regulation in vivo.Growing evidence suggests that the interaction between CCL21 and CCR7 in the tumor microenvironment is associated with poor prognosis in various types of human cancer.CCL21 produced by lymphatic endothelial cells generates a concentration gradient that induces migration of cancer cells to the lymph nodes (LNs) in a CCR7-dependent manner [9,10].Consistent with these findings, clinical studies in breast cancer, melanoma, gastric cancer, esophageal squamous epithelial cells carcinoma, and chronic lymphocytic leukemia, have demonstrated that CCR7 expression is correlated with cancer cell metastasis to LNs [11][12][13].Thus, CCL21 expressed by cancer cells and stromal cells in the tumor microenvironment confers metastatic capacity to cancer cells to the regional LNs.
The ability of CCL21 to stimulate T cell and DC migration suggests that CCL21 plays a key role in an effective immune response in the tumor microenvironment.A significant number of studies have demonstrated that systemic or local CCL21 expression promotes anti-tumor response by recruiting lymphocytes and DCs into tumor tissue [14].The expression of CCL21 in cancer cells including melanoma, liver cancer, and prostate cancer cells, induces immune responses and inhibits tumor growth [15][16][17].CCL21-Ser in melanoma and lung carcinoma cells suppressed tumor growth by developing blood vessels similar to the high endothelial venules in LNs, leading to an enhanced antitumor immunity by mobilizing naïve lymphocytes to the tumor tissue [18].In addition, CCL21 promotes antitumor immune response by inducing tumor antigen presentation and suppressing the production of immunosuppressive cytokines in human breast cancer cells [19].Intratumoral administration of DCs expressing CCL21 in mice induced anti-tumor immunity and markedly suppressed the growth of subcutaneous lung cancer, suggesting the benefit of CCL21 expression in DC-based cancer immunotherapy [20].In contrast, in another study, CCL21-Ser produced by murine melanoma cells induced peritumoral LN-like structures and recruited immunosuppressive cells, enhancing the melanoma growth [21].Thus, although the impact of the CCL21/CCR7 signaling on anti-tumor immunity remains controversial, the fact that CCL21 expression can induce an immune response provides a rationale for the use of CCL21 in cancer immunotherapy.For optimal therapeutic use of CCL21, the contribution of host-derived CCL21-Ser to tumor growth and its role in vivo must be fully understood.Therefore, this study assessed the role of endogenous CCL21-Ser in tumor growth and immune cell infiltration in tumor tissues using Ccl21a-KO mice.

Cells
The mouse melanoma B16-F10 and YUMM1.7 were obtained from the American Type Culture Collection.The mouse Lewis lung carcinoma (LLC) and colon adenocarcinoma MC38 cell lines were kindly provided by Drs.K. Moriwaki of Osaka Medical Collage and CM. Lee of Osaka University, respectively.Breast cancer EO771 cell line was obtained from CH3 BioSystems (Amherst, NY, USA).YUMM1.7 were cultured in Dulbecco's Modified Eagle medium/Ham's F-12 (FUJIFILM Wako Pure Chemicals, Osaka, Japan) supplemented with 10% (v/v) fetal calf serum.Other cell lines were cultured in Dulbecco's Modified Eagle medium (FUJIFILM Wako Pure Chemicals, Osaka, Japan) supplemented with 10% (v/v) fetal calf serum.

Animals
C57BL/6 N and B6 Albino (B6N-Tyr c-Brd /BrdCrCrl) were purchased from Nihon SLC (Hamamatsu, Japan) and Charles River Laboratories Japan, Inc. (Yokohama, Japan), respectively.The original Ccl21a-KO mice, CDB1030K http://www.clst.riken.jp/arg/mutant%20mice%20list.html [3], were backcrossed three times onto the C57BL/6 or B6 Albino backgrounds before continuing with sibling mating.The mice were euthanized by a lethal dose of isoflurane via inhalation.For in vivo tumor formation, B16-F10, LLC and MC38 cells (1 × 10 6 per mouse) were subcutaneously injected into 6-12-week-old mice of either sex and YUMM1.7 cells (1 × 10 5 per mouse) were injected into 6-week-old female mice with 27-G needles.EO771 cells (5 × 10 5 per mouse) were injected into the fifth right mammary fat pad of female mice.The volume of mammary tumor was determined by caliper measurements and calculated using formula V = width 2 × length × 0.5.All experiments were conducted in accordance with the approved guidelines from Kindai University.The experimental protocols for use of laboratory animals were approved by the Ethics Review Committee of Kindai University (KASE-2021-001).

Immunohistochemistry
Fresh-frozen tumor specimens dissected two weeks after cell injection were subjected to immunohistochemical staining.Frozen  sections of 10-μm thickness were fixed with ice-cold acetone, incubated with 50% Immunoblock (DS Pharma) in PBS for 30 min, and were treated with APC-conjugated anti-B220, Alexa Fluor 594-conjugated anti-CD3, and rabbit anti-mouse CCL21 antibodies listed on Table 1 overnight at 4 • C. CCL21 was detected by a tyramide signal amplification system (Thermo Fisher Scientific Inc.).Tissues were treated with horseradish peroxidase (HRP)-conjugated donkey anti-goat IgG for 30 min at room temperature, followed by Alexa Fluor 488 tyramide reagent at 1:800 dilution (B40953, Thermo Fisher Scientific Inc.) for 10 min at room temperature.The tissues were stained with 2 μg/ml Hoechst 33342 (Invitrogen) for 10 min, mounted with Fluoromount-G (SouthernBiotech), and observed with a confocal laser scanning microscope (FV1200-D, Olympus).

Flow cytometry
The tumor tissues were enzymatically treated at 37 • C for 1.5 h with RPMI 1640 medium containing 0.1% bovine serum albumin, 1 mg/ml collagenase type І (9001-12-1, Wako) and 2 μg/ml deoxyribonuclease I (9003-98-9, Sigma).They were subsequently treated with 0.83% NH 4 Cl and 0.17 M Tris-Cl (pH 7.65) for erythrocyte hemolysis.LN cells were treated with HB-197 cell culture supernatant containing an anti-CD16/32 (Fc-γ receptor) antibody.The cell suspension was treated with antibodies and fluorescent reagents listed on Table 1.The cell number was counted with Flow-count Fluorospheres (Beckman Coulter).The data analysis and interpretation were performed on a BD LSRFortessa system and the FlowJo software (BD Biosciences).

In vivo depletion of CD8 + T cells using mAb treatment
Six-week-old Ccl21a-KO mice were injected intraperitoneally with 100 μg of GoInVivo purified anti-mouse CD8 mAb (clone 53-6.7,BioLegend) or allogeneic isotype immunoglobulin (control Ig) three days prior to B16-F10 injection.Antibody injection was repeated on the day of B16-F10 cell injection and on day 7 post injection.Spleens and tumors were harvested from each mouse on day 14 and the CD8 population was examined by flow cytometry.

Analysis of the effector activity of in vitro cultured splenocytes
A detailed description of the protocol is presented previously [22].Briefly, splenocytes of tumor-bearing or control mice were cultured with or without MMC-treated B16-F10 cells in culture media containing 100 U/ml recombinant mouse IL-2 (BioLegend).To assess effector activity of CD8 + T cells by target cell viability, CD8 + T cells co-cultured with or without B16-F10 cells were added to B16-F10 or LLC cells expressing firefly luciferase.Bioluminescence was measured by real-time luminometer (Kronos Dio, Atto).

Statistics
Data were evaluated by F-test for equal variances, then analyzed by Student's t-test for equal variances, and by Welch's t-test or Mann-Whitney U test for unequal variances using StatPlus software version v7 (AnalystSoft Inc.).

Lack of host-derived CCL21-Ser reduces B16-F10 melanoma growth
The B16-F10-derived melanoma in the Ccl21a-KO mice had significantly lower tumor weight than that of the wild-type (WT) mice (Fig. 1A).In vivo bioluminescence imaging showed B16-F10 growth in WT mice until day 11 (Fig. 1B), whereas the luminescence in Ccl21a-KO mice was already lower than that in WT by day 4, suggesting that tumor growth was suppressed relatively at the early phase in the Ccl21a-KO mice.Similarly, the growth of YUMM1.7 melanoma was significantly lower in Ccl21a-KO mice, whereas not significantly different in LLC lung cancer and EO771 breast cancer, and significantly increased in MC38 colon cancer (Fig. 1C-F).Therefore, the reduction in tumor growth in Ccl21a-KO mice is likely to be a melanoma-selective phenomenon.The expression of Ccl21a and CCR7 are undetectable in B16-F10 cells cultured in vitro and those cells in tumor tissues (Supplementary Fig. 1A and B).Immunohistochemistry detected no ectopic lymphoid structures with B and T cell aggregates in tumors of WT and Ccl21a-KO mice (Supplementary Fig. 1C).Moreover, CCL21-Ser expression was undetectable in tumor tissues, suggesting the contribution of hostderived CCL21-Ser and CCR7 on nontumor cells to B16-F10 tumor growth.

Ccl21a-KO melanomas are frequently infiltrated with activated CD8 + T cells
To decipher the mechanisms underlying decreased B16-F10 tumorigenicity in Ccl21a-KO mice, we hypothesized that a stronger antitumor immunity was induced in the KO mice.A flow cytometric analysis of CD3 + , CD4 + , and CD8 + T cell populations within the intratumoral cells (Fig. 2A) revealed that the proportion of tumor infiltrating T cells, especially CD8 + T cells, was significantly higher in the KO mice than in WT (Fig. 2B and C).Additionally, a significant increase in activated T cells and a decrease in naïve T cells was observed in the KO mice (Fig. 2D).The depletion of CD8 + T cells by anti-CD8 mAb enhanced tumor growth in Ccl21a-KO tumors, indicating that the decrease in B16-F10 growth in Ccl21a-KO mice is CD8 + T cell-dependent (Fig. 3A).The frequency of CD8 + T cells expressing tumor-reactive T cell markers, PD-1, and Ki-67, was comparable between Ccl21a-KO and WT mice; however, the frequency of H-2Kb TRP-2 tetramer-specific IFN-γ + CD8 + T cells was elevated in Ccl21a-KO tumors, suggesting that anti-melanoma immunity was enhanced in the KO mice (Fig. 3B).The frequency of NK cells was also significantly increased in Ccl21a-KO tumor compared with WT, suggesting a possible contribution of NK cells to the decreased B16-F10 growth in Ccl21a-KO mice (Supplementary figure 2).In contrast, the frequencies of macrophage and DC infiltration were comparable between Ccl21a-KO and WT mice.We hypothesized that the KO mice have an elevated tumor-specific immune response in draining LNs (dLNs).We observed no significant difference in CD4 + or CD8 + T cell counts before and after tumor formation in WT dLNs, whereas they were significantly increased in Ccl21a-KO dLNs (Fig. 4A).Although the CD4/CD8 ratio of tumor-draining and non-dLNs was comparable (Fig. 4B), the frequencies of naïve T cells were decreased and those of effector memory T cells were increased in Ccl21a-KO dLNs (Fig. 4C).
We assessed the effector function of CD8 + T cells in tumor-bearing Ccl21a-KO and WT mice.To this end, we cultured CD8 + T cells derived from tumor-bearing mice and monitored the time course of target B16-F10 cell viability using a bioluminescence-based method [22].The tumor-bearing splenocytes from Ccl21a-KO mice showed a higher frequency of activated CD8 cells and a higher impact on B16-F10 cell viability than WT (Fig. 5A and B, respectively).They did not exert a bystander effect on LLC cells (Fig. 5C), suggesting that more B16-F10-specific effector T cells were generated in Ccl21a-KO mice.

Ccl21a-KO melanomas are infiltrated with Tregs with decreased levels of co-inhibitory receptors
Regulatory T cells (Tregs) regulate the differentiation of naïve CD8 + T cells into memory T cells and suppress melanoma growth [23,24].Although the number of CD4 + CD25 + CD127 lo cells in Ccl21a-KO dLNs was significantly lower than that of WT dLNs without tumors (Supplementary Fig. 3A, Fig. 6A), the number increased more significantly in Ccl21a-KO dLNs (approximately 6-fold) compared with the WT dLNs (approximately 1.7-fold) as the tumor grew.The majority of CD4 + CD25 + CD127 lo cells in the tumors were Tregs, as approximately 80% of them express FoxP3 (Supplementary Fig. 3B).Although the proportion of CD4 + T cells was lower in Ccl21a-KO than WT tumors, the Treg frequency of CD4 + T cells was higher in Ccl21a-KO (Fig. 6B), suggesting that Treg accumulation was more efficiently induced in Ccl21a-KO tumor.
In contrast to the higher Treg frequency, the Ccl21a-KO tumor showed an apparent reduction in the positive correlation between the ratio of Tregs to activated CD8 + T cells and tumor weight (Fig. 6C).Therefore, we hypothesized that the immune suppressive   function is impaired in Ccl21a-KO mice, regardless of their efficient recruitment to the tumor tissues.To understand the functional abnormalities in Ccl21a-KO intratumoral Tregs, we examined the expression of co-inhibitory receptors TIM-3 and TIGIT, which are associated with Treg function in the tumor microenvironment [25,26].The frequencies of TIM-3 + and TIGIT + Tregs were reduced in Ccl21a-KO compared with control mice (Fig. 6D), consistent with the idea that Ccl21a-KO Tregs are functionally impaired.The frequency of myeloid-derived suppressor cells (MDSCs), immune cells from the myeloid lineage with immunosuppressive phenotypes, was not significantly different between Ccl21a-KO and wild-type tumor (Supplementary Fig. 2B), suggesting that the MDSCs do not significantly contribute to the decreased tumor growth in Ccl21a-KO mice.

Discussion
Our results shed light on the function of host-derived CCL21 in tumor growth in vivo.We have demonstrated for the first time that B16-F10, a poorly immunogenic cell line widely used in the studies of tumor-induced T cell anergy [24], efficiently elicited immune response in the absence of host-derived CCL21-Ser.B16-F10 tumors formed in Ccl21a-KO mice showed higher T cell ratios and increased the numbers of activated CD4 + , CD8 + T cells, and NK cells.In addition, the activated CD8 + T cells recognizing TRP-2 were significantly increased in Ccl21a-KO tumors than in WT tumors, suggesting that host-derived CCL21-Ser has a suppressive effect on anti-melanoma immunity.
The results using a mouse melanoma model are consistent with cohort studies showing that melanoma patients with high CCL21 expression have shorter overall survival (the Human Protein Atlas, https://www.proteinatlas.org/ENSG00000137077-CCL21/pathology/melanoma).Among the five cell lines used, tumor growth of B16-F10 and YUMM1.7 was decreased in Ccl21a-KO mice, whereas that of LLC and EO771 was unchanged, and that of MC38 was increased, suggesting that the effect of CCL21-Ser deficiency is selective for melanoma.We speculate that differences in immunogenicity between B16-F10 and MC38 may have caused opposite effects in these cell lines.B16-F10 and MC38 are a poorly immunogenic and an immunogenic cell line, respectively [27].In MC38 tumors, CD3 + and Treg frequencies were unchanged in CCL21a-KO mice, whereas the CD8/CD4 ratio was increased (Supplementary figure 4A).Notably, the frequency of macrophages was significantly reduced in CCL21a-KO mice compared with WT mice (Supplementary figure 2B and 4B).A pervious paper has demonstrated that M1 macrophages, well-characterized activated macrophages with antitumorigenic properties, inhibit MC38 tumor growth [28] and they express CCR7 and migrate toward a CCL21 gradient [29].Since CCL21 was detectable in MC38 tumors by immunohistochemical analysis (Supplementary figure 4C), it is possible that the tumor growth of MC38 in CCL21a-KO mice is due to decreased infiltration of M1 macrophages by CCL21-Ser deficiency, resulting in a reduced anti-tumor immune response.Thus, differences in tumor microenvironment and immunogenicity may account for the different effects of Ccl21a deficiency.Our study does not clarify whether CCL21-Ser supports melanoma growth only in certain genetic backgrounds.To understand this point, it is necessary to analyze allogeneic transplants of various cancer cells in animals with different genetic backgrounds.
Given that activated CD8 + T cells and NK cells were more efficiently infiltrated in Ccl21a-KO melanoma compared with WT counterpart, these cytotoxic immune cell population primarily contribute to suppression of melanoma growth.CD8 + T and NK cells are major effector cells against various type of tumors and their anti-tumor effects are increased by Treg depletion [30][31][32].In our study, however, the increased activated CD8 + T cells was not associated with a numerical decrease in Tregs in Ccl21a-KO mice.The ratio of Treg to activated CD8 + T cells did not positively correlate with tumor weight, we therefore speculated that Tregs in Ccl21a-KO tumors have reduced immunosuppressive activity.In line with this idea, we examined the expression of TIM-3 and TIGIT, co-inhibitory molecules which are crucial for Treg function [26,33], and exhibited that TIM-3 + and TIGIT + Treg frequencies were reduced in Ccl21a-KO tumors.In Ccl21a-KO tumors, both TIM-3 + and TIGIT + Tregs were reduced by less than 10%, however, since the two molecules have been reported to act synergistically [18], the simultaneous reduction of TIM-3 and TIGIT may have resulted in a significant decrease in Treg suppressive activity.Unfortunately, it is technically impossible to directly analyze the intratumoral Treg activity, because the number of intratumoral Tregs is approximately 1~5 × 10 3 (0.1% of total KO tumor sample), which is far less than the number required for in vitro suppression assay.Further experiments with high sensitivity are needed to confirm intratumoral Treg dysfunction in Ccl21a-KO tumors.
The Ccr7-KO mice showed increased infiltration of inflamed tissues with inactive Tregs, suggesting that the CCL21-Ser/CCR7 axis is critical for functional Treg development in vivo [34].Our findings are consistent with this idea and indicate that endogenous CCL21-Ser expression is essential for the development of functional Tregs in antitumor immunity.Since the predominant Treg subset in human and mouse tumors is the thymus-derived naturally occurring Treg [35], a lack of CCL21-Ser expression in the Ccl21a-KO thymus may cause intrinsic dysfunction of naturally occurring Tregs and accelerated CD8 + T-mediated tumor rejection.However, we speculate that this is unlikely, as our preliminary data do not demonstrate dysfunction of Ccl21a-KO splenic Tregs on effector CD4 + T cell proliferation in vitro.Alternatively, the ineffective suppressive function of Ccl21a-KO Tregs is possibly by their defective LN positioning, as demonstrated in Ccr7-KO Tregs [34].
In mouse melanoma models, multiple studies have demonstrated that depletion of Tregs by anti-CD25 antibody enhances antitumor immune responses [23,36].However, a clinical trial of transient Treg depletion in melanoma patients with anti-CD25 antibodies did not improve the efficacy of the DC vaccine due to a concomitant depletion of effector T cells [37].Although further studies are needed on the molecular mechanisms leading to Treg dysfunction in Ccl21a-KO mice with B16-F10, following our idea that the reduced Treg recruitment in LNs is responsible for tumor regression in Ccl21a-KO mice, inactivating Treg function by blocking Treg entry into tumor-dLNs could be a new approach to effective immunotherapy.

Fig. 1 .
Fig. 1.The effect of Ccl21a deficiency on melanoma growth in vivo.(A) Representative images of B16-F10-derived tumors formed in the wildtype (WT) or Ccl21a-KO mice (left).Scale bar: 10 mm.B16-F10-derived tumors formed in WT (n = 55) or KO (n = 45) were weighed two weeks after injection (right).Each dot represents a sample of the individual mouse, and green lines show the median.Mann-Whitney's U test was used to compare differences between the two groups.***, P < 0.001, NS: Not Significant.(B) B16-F10 cells stably expressing click beetle luciferase were subcutaneously injected into WT or Ccl21a-KO on the B6 Albino background (n = 6).An IVIS Lumina system at the indicated time points measured luminescence counts.The average radiance of IVIS-imaged mice (p/s/cm 2 /sr) was shown.Welch's t-test was used to compare the differences between the two groups.*, P < 0.05, **, P < 0.005 (C) The effect of Ccl21a expression on YUMM1.7,(D) LLC carcinoma, and (E) MC38 colon carcinoma growth in WT and Ccl21a-KO mice.Tumors derived from YUMM1.7 (WT; n = 5, KO; n = 5) and MC38 (WT; n = 6, KO; n = 6) were weighed three-weeks after injection.Tumors derived from LLC (WT; n = 9, KO; n = 9) were weighed two-weeks after injection.(F) EO771 tumor volume was measured using a vernier caliper (WT; n = 5, KO; n = 5).Mann-Whitney's U test was used as the significance test.NS: Not Significant.

Fig. 2 .
Fig. 2. Analysis of tumor-infiltrated T cell subsets in wild-type and Ccl21a-deficient mice.The B16-F10-derived tumor generated in WT and Ccl21a-KO mice were harvested two weeks after injection.The tumor-infiltrated T cell subset was analyzed by flow cytometry.(A) The lymphocyte subset was distinguished by the forward and side scatter profiles, and the proportion of CD8 + or CD4 + cells in the CD3 + cell population was analyzed.(B) The percentage (n = 9 in WT and n = 8 in KO) of CD3 + cells in the tumor.Each dot represents an individual sample.The data were analyzed by Mann-Whitney U test (**, P < 0.01).Red lines represent the median.(C) The percentage of CD8 + or CD4 + cells in the CD3 + cell fraction (n = 6).The data are shown as mean ± SD analyzed by Student's t-test (*, P < 0.05).(D) The percentage of naïve (CD44 − CD62L + ) or activated (CD44 + CD62L − ) T cells in the CD4 + (WT: n = 11, KO: n = 16) and the CD8 + cell subsets (WT: n = 9, KO: n = 8) were analyzed.The data are shown as mean ± SD analyzed by Student's t-test (*, P < 0.05, NS: Not Significant).R. Fujie et al.

Fig. 3 .
Fig. 3.The effect of intratumoral CD8 þ T cells on melanoma growth.(A) The effect of CD8 + T cell depletion on tumor weight in the Ccl21a-KO mice.Anti-CD8 antibody or control immunoglobulin (control Ig) was administered to the Ccl21a-KO mice (n = 3) three days before, on days 0 and 7 of transplantation.The percentage of CD8 + cells and tumor weight in each mouse two weeks after transplantation was measured.(B) The expression of tumor-reactive T cell marker molecules (PD-1, Ki-67, IFN-γ, TRP2 tetramer) in the intratumoral CD8 + T cells.Results are representative of two biological repeats.

Fig. 4 .
Fig. 4. Changes in T cell subsets in the tumor-draining LNs of Ccl21a-KO mice.The number and percentage of T cells in the tumor-draining inguinal LNs of tumor-bearing or control mice were analyzed by flow cytometry.(A) The number of CD3 + , CD4 + , and CD8 + cells in the inguinal LNs of control or tumor-bearing mice (n = 9).(B) The percentage of CD4 + or CD8 + T cells in the CD3 + cell subset.(C) Naïve (CD62L high CD44 lo ), effector (CD62L lo CD44 mid ), and effector memory (CD62L lo CD44 high , Tem) T cell percentage of CD4 + and CD8 + T cells in WT and Ccl21a-KO mice of tumorbearing mice (n = 10).The data are shown as mean ± SD analyzed by Student's t-test (*, P < 0.05, NS: Not Significant).

Table 1
List of reagent and resource used in this study.