Follistatin-like 1 deficiency impairs T cell development to promote lung metastasis of triple negative breast cancer

Our study aims to detect the underlying mechanism of the suppressive effect of Follistatin-like 1 (FSTL1) on lung metastasis of triple negative breast cancer (TNBC). We found that FSTL1 had no effect on the proliferation and metastasis of 4T1 cells in vitro, while in the tumor-bearing Fstl1 heterozygous (Fstl1+/-) mice, the number of anti-tumor T lymphocytes in the lung was significantly reduced with the increase in lung metastasis. Impaired development of T cells can cause dysfunction of adaptive immune system, which promotes cancer metastasis. Therefore the effect of FSTL1 on T cell development was further investigated. Lower population of T cells in periphery and decreased proliferation of CD4- CD8- double negative (DN) thymocytes and impairment development of T cells were found in Fstl1+/- mice. Furthermore, high expression of FSTL1 in medullary thymus epithelial (mTEC) cells and decreased mRNA expression of inducible costimulator on activated T-cell ligand (Icosl) in mTECsh Fstl1 were detected. Combining other studies that the generation of ICOSL by mTEC cells promotes CD4+ single positive (SP) thymocytes to produce IL-2, which promotes T cell development. Our results indicate FSTL1 deficiency in mTEC cells impairs T cell development to promote the lung metastasis of TNBC.


INTRODUCTION
Globally, breast cancer is the most common malignant tumor and a leading cause of cancer deaths among women [1,2]. Triple negative breast cancer (TNBC), a highly aggressive type of breast cancer, is characterized by a high propensity for metastasis. Patients with metastatic TNBC typically exhibit drug resistance and poor outcomes of targeted therapy [3,4]. The host immune system, which can destroy nascent malignant cells through immunosurveillance, plays a critical role in breast cancer progression and metastasis [5]. Anti-tumor immunotherapy has been considered as a promising avenue for the treatment of TNBC.
Leukocyte subpopulations at distant sites play a role in the growth metastatic tumors [6]. Both CD4 + and CD8 + T lymphocytes act as the cellular effectors of adaptive immunity [7]. These pivotal anti-tumor immune cells exhibit a strong relationship with favorable outcomes and improved metastasis-free survival in patients with breast cancer [8]. Cytotoxic CD8 + T cells directly attack tumor cells through tumor specific adaptive immune response [9]. CD4 + T cells differentiate into two distinct subsets of T helper (Th) cells, Th1 and Th2 cells, which produce different cytokines and perform opposite function. Th1 cells produce interferon gamma (IFN-γ) and tumor necrosis factor alpha (TNF-α), which promote cellular immunity to initiate effective antitumor response [10]. Th2 cells secrete interleukin (IL)-4 and IL-10, which recruit immunosuppressive cells to negatively regulate the Th1 anti-tumor response, and support tumor growth and metastasis [11]. The shift of the Th1/Th2 balance towards Th2 response has been AGING observed in different types of cancers including breast cancer [12,13]. Furthermore, breast cancer patients with Th1 dominant response show higher overall survival rates and lower tendency for metastatic invasion [14]. To achieve successful immunotherapeutic intervention, it is necessary to restore tumor specific Th1 response and reduce Th2-skewed immune response [15]. Therefore, targeting Th1 cell and CD8 + T cell responses are promising strategies for breast cancer and lung metastasis treatment.
Follistatin-like 1 (FSTL1) is a secreted glycoprotein belonging to the secreted protein acidic and rich in cysteine (SPARC) family [16]. FSTL1 was first identified as a TGF-β1-inducible protein, which plays roles in many physiological and pathological processes [17]. FSTL1 exhibits opposite effects on the progressions of different tumors. In nasopharyngeal carcinoma, FSTL1 not only improved the antigen presentation ability of dendritic cells, but also activated macrophages to enhance the function of T lymphocytes and attenuate immune evasion [18][19][20]. On the contrary, FSTL1 promoted the generation of immunosuppressive mesenchymal stromal/stem cells to induce immune tolerance, which indirectly promoted progression of melanoma lung metastasis [21,22].
In this study, we explored immune microenvironment in the lungs of Fstl1 +/tumor-bearing mice. FSTL1 deficiency reduced anti-tumor T cells in lung affected by metastasis. This is the first study to demonstrate impaired T cell development in Fstl1 +/mice, which can promote lung metastasis of TNBC.

FSTL1 deficiency in the microenvironment indirectly facilitated metastatic growth in the lungs
In our previous study, FSTL1 deficiency was found to accelerate the growth of lung metastasis of TNBC [23]. Here, we sought to further clarify the reason for the significant increase in lung metastasis of tumor-bearing Fstl1 +/mice. Fourteen days after orthotropic implantation of 4T1 cells, lung tissues from wild-type (WT) and Fstl1 heterozygous (Fstl1 +/-) female BALB/c mice were collected and stained with HE to observe lung metastasis. Consistent with previous results, a significant increase of lung metastatic lesions was found in Fstl1 +/mice ( Figure  1A). As shown in Figure 1B, the metastatic lung tissues of Fstl1 +/mice showed higher expression of cyclindependent kinases 2 (CDK2) and phosphorylation of CDK2 (p-CDK2) ( Figure 1B), which indicated high proliferation and malignant phenotypes [24]. These data indicated that FSTL1 deficiency promoted the growth of metastatic lesions in the lungs.
The expression of FSTL1 was hardly detected in 4T1 cells, while it was enriched in the lungs of WT mice ( Figure 1C). This suggested that the presence of FSTL1 in the lung microenvironment may affect the biological behavior of 4T1 cells. To validate the effect of FSTL1 on the proliferation of 4T1 cells, 4T1 cells were treated with recombinant mouse FSTL1 (rmFSTL1) protein, and the cell viability was evaluated by CCK-8 assay. However, the cell viability did not change after treatment with different doses of rmFSTL1 ( Figure 1D). The protein level of CDK2 was measured to assess proliferation of 4T1 cells after rmFSTL1 treatment. However, no significant difference was observed ( Figure 1E). The expressions of Caspase-3 and cleaved Caspase-3 showed no significant difference between the control groups and the rmFSTL1 treated groups ( Figure 1E), which indicated no effect of rmFSTL1 on the apoptosis of 4T1 cells.
Malignant behavior of tumors is also manifested by their ability for invasion and migration. Therefore, we detected invasive and migratory markers by qRT-PCR. Neither the epithelial markers E-cadherin and Zo-1 nor the mesenchymal marker Vimentin showed any significant change. The expressions of Mmp-9 and Tgf-β (makers of invasiveness in advanced cancers [25,26]) were also not affected by rmFSTL1 treatment ( Figure 1F). These data indicate that rmFSTL1 has no effect on epithelial to mesenchymal transition (EMT) of 4T1 cells. Collectively, these results demonstrate that FSTL1 deficiency in the microenvironment indirectly facilitated the metastatic growth of breast cancer cells in lungs.

Fstl1 +/mice displayed decreased anti-tumor immune cells in metastatic lung
Tumor-infiltrating lymphocytes (TILs) and lung microenvironment infiltrated T lymphocytes were stained by IHC in the lung slices of WT and Fstl1 +/mice after orthotropic implantation of 4T1 cells. Fstl1 +/mice showed significant decrease in TILs and lung microenvironment infiltrated T lymphocytes (including CD4 + and CD8 + T cells) (Figure 2A, 2B). Flow cytometry also revealed a decreased in the proportions of CD4 + and CD8 + T cells in lungs of Fstl1 +/mice ( Figure 2C), these findings implied that FSTL1 deficiency reduced the infiltration of anti-tumor T cells in metastatic lung lesions of TNBC.
Tumor-reactive Th1 cells and tumor-promoting Th2 cells are two subtypes of CD4 + T helper cells. Shift from Th1 to Th2 response indicates dominant immunosuppression response in the tumor microenvironment. Reduced proportion of infiltrated Th1 (IFN-γ + CD4 + ) cells was observed in Fstl1 +/mouse lungs, whereas the proportion of Th2 (IL-4 + CD4 + ) cells was not changed ( Figure 2D). These findings indicated the immunosuppression dominant AGING response in metastatic lung tissue of Fstl1 +/mice. Collectively, these data demonstrate that FSTL1 deficiency in the microenvironment not only diminishes CD8 + T cells, but also inhibits the activity of Th1 cells in lung metastasis of breast cancer.

Fstl1 +/mice had decreased T cells in periphery
An important reason for the decreased anti-tumor T cells infiltration at the metastatic site is the impairment of T cell development. We observed a significant  The expression levels of CDK2, Caspase-3 and cleaved Caspase-3 in 4T1 cells treated with 500 ng/mL rmFSTL1 for 48h or 72h. Densitometric measurement of band intensity normalized to that of β-actin. (F) The mRNA levels of genes related to EMT in rmFSTL1 treated groups and control groups, which was normalized to that of β-actin. Data are presented as mean ± SD. Each dot in the graphs represents an individual mouse. Data in the line chart represent three sets of independent experiments. n.s., not significant; *p < 0.05, **p < 0.01. AGING decreased in the percentages of CD4 + and CD8 + T cells in the lungs and peripheral blood of Fstl1 +/tumor free mice ( Figure 3A, 3B). Since peripheral T cells reflect the function of thymus [27,28], our findings suggest that FSTL1 may play a role in T cell development and generation in thymus. Quantification of the proportions of CD4 + and CD8 + T cells within the gated live cells in the WT and Fstl1 +/mouse peripheral blood (n=5, WT; n=6, Fstl1 +/-). Data presented as mean ± SD. Each dot in the graphs represents an individual mouse. *p < 0.05, *** p < 0.001.

Fstl1 +/mice exhibited impaired T cell development
Thymus provides a unique microenvironment for T cell development by allowing proliferation, differentiation and selection of T cell precursors. T cell precursors enter the thymus and increase their number before they develop into double positive (DP) thymocytes. Subsequently, the DP thymocytes undergo positive selection and differentiate into CD4 + single positive (SP) or CD8 + SP thymocytes, which then reach the medulla to their negative selection [29,30].
Since Fstl1 -/mice died of breath failure after birth [16], the thymuses from WT and Fstl1 -/mice were isolated at embryonic (E) day 18.5 (E18.5) to investigate the effect of FSTL1 on thymic development. As shown in Figure  4A, Fstl1 -/mice had a significant reduction in thymus size and overall cell number compared to those in WT mice. Obvious reduction in the size and weight of thymuses was also found in 8-week-old Fstl1 +/mice compared to that in WT mice ( Figure 4B, 4D, 4E). Although percentages of double negative (DN), DP, SP thymocytes of Fstl1 +/mice were normal ( Figure. 4F), there was a decrease in the numbers of thymocytes and the cell numbers of all thymocyte subpopulations in Fstl1 +/mice ( Figure 4C, 4G). Furthermore, the mRNA level and protein level of proliferation marker Ki67 in thymuses of Fstl1 +/mice was significantly lower compared with that in WT mice ( Figure 4H, 4I). Since DN thymocytes are the most proliferative subpopulation in thymus, these results indicate reduction in the proliferation of DN thymocytes and impairment of T cell development in FSTL1 deficiency mice.

FSTL1 in mTECs supported T cell development
Recent studies have demonstrated that FSTL1 expression is restricted to non-hematopoietic cell lines, especially the mesenchymal lineage cells [31]. We sought to ascertain the cell types in which FSTL1 plays a vital role during thymic organogenesis. Medullary thymus epithelial (mTEC) cells are the major stromal cells in the thymic medulla, these cells have large, palestaining nuclei and enriched cytoplasm. FSTL1 was found mainly expressed in the medulla of thymus in WT mice by IHC staining ( Figure 5A). The expression of Fstl1 was measured in isolated DP, CD4 + SP, CD8 + SP thymocytes and mTEC cells using qRT-PCR. As shown in Figure 5B, mTEC cells were the major cellular source of FSTL1 in the thymus, whereas the mRNA level of Fstl1 in thymocytes were almost undetectable.
The mRNA level of Il-2 in thymuses of Fstl1 +/mice was decreased compared to that in WT mice ( Figure  5C). Studies have shown that IL-2 is secreted by CD4 + SP thymocytes [32], and production of inducible costimulator on activated T-cell ligand (ICOSL) by mTEC cells induces CD4 + SP thymocytes to secrete high levels of IL-2 in coculture [33]. In our research, knockdown of Fstl1 in mTEC cells significantly reduced the Icosl mRNA level compared to the mTEC sh con group in vitro ( Figure 5E). This suggests that FSTL1 in mTEC cells sustains the expression of Icosl, and further, ICOSL induced IL-2 production from CD4 + SP thymocytes ( Figure 5F).

DISCUSSION
Aberrant expression of FSTL1 has been demonstrated in tumor cell lines and clinical tumor biopsy specimens, which implies that FSTL1 may play different roles in different types of cancers [17,34,35]. In our previous study, the expression of FSTL1 in 231-BR cells, the brain metastatic cell line of MDA-MB-231 cells, was significantly higher than that in its parental cell line, while its proliferation ability was significantly lower than that of MDA-MB-231 cells [36]. In other studies, FSTL1 has been demonstrated to induce apoptosis and inhibit invasion and metastasis in endometrial carcinoma and ovarian carcinoma [37]. Anti-FSTL1 therapy was more effective in tumor models established by 3LL cells and colon 26 cells with high expression of FSTL1 than in tumor models that barely expressed FSTL1 [22]. It implied that endogenous FSTL1 may regulate the proliferation and metastasis of tumor cells. Interestingly, FSTL1 was hardly detected in 4T1 cells, while high expression level of FSTL1 was observed in lung. It indicated that the proliferation and metastasis of 4T1 cells may be influenced by FSTL1 from the lung microenvironment. However, rmFSTL1 did not significantly affect the proliferation and EMT of 4T1 cells in vitro.
We also did not observe significant difference in the growth of the primary 4T1 murine mammary tumor in WT and Fstl1 +/mice, whereas more metastatic nodules were detected in the lungs of Fstl1 +/− mice [23]. The vast difference of FSTL1 function in vivo and in vitro drew our attention to the tumor microenvironment. TILs are the most widely studied population of tumorinfiltrating immune cells and have been reported to be associated with good prognosis in breast cancer [8,38]. Some studies demonstrated that FSTL1 as a critical effector molecule in cancer progression via affecting host immunity [19,20,22]. In our research, the Fstl1 +/mice had the same volume of primary tumor as the WT mice; however, IHC staining showed that the number of infiltrated CD4 + T cells in the primary tumor was significantly decreased in Fstl1 +/mice compared to that in WT mice. The number of infiltrated CD8 + T cells in primary tumor of Fstl1 +/− mice showed a slight AGING AGING decreasing tendency (Supplementary Figure 1). Bidwell BN et al indicated that the growth of 4T1 primary tumor couldn't be restrained by the immune system [39]. Mammary tumor-infiltrating T cells differentiated into different subsets of effector cells, which impacted pulmonary metastasis, however, it might not affect the growth of the primary tumor [40]. Since the volume of the primary tumor was same in the WT and Fstl1 +/mice, the decreased number of infiltrated CD4 + T cells in the primary tumor may promote the escape of cancer cells and accelerate the metastasis of 4T1 cells.
We also found that FSTL1 deficiency significantly reduced the infiltration of CD4 + and CD8 + T cells in the metastatic nodules and the lung microenvironment. Furthermore, Th1 cells also reduced in the lung of Fstl1 +/tumor-bearing mice. Therefore, given the obvious increases in lung metastases in Fstl1 +/mice, we conclude that decreased Th1 cells and CD8 + T cells promote metastasis to lungs.
A previous study demonstrated that very low expression of FSTL1 was detected in T cells, therefore, the deficiency of FSTL1 may have little effect on T cell activation and differentiation [41]. Impaired T cell development may be an important reason for the decrease in anti-tumor T cells at metastatic site. Thymus, as the central immune organ and the place of T cell development, generates and maintains an important arm of host adaptive immune response [42]. Fstl1 +/tumor free mice exhibited significantly reduced proportions of T lymphocyte subpopulations in the periphery blood and lungs, which indirectly reflected the dysfunction of thymus. Reduction in thymus size and number of thymocytes in Fstl1 +/mice not only exhibited thymus dysplasia, but also could lead to the dysfunction of immune system. Furthermore, the Fstl1 +/tumor-bearing mice also displayed a decrease in thymus size and thymocyte number compared to that in WT mice (Supplementary Figure 2).
Since relatively small numbers of T cell progenitors migrate into the thymus per day, the DN thymocytes need to rapidly expand their numbers to maintain the pool size of precursor cells to initiate the process of T cell differentiation [43]. Therefore, reduced cell numbers of DN, DP and SP thymocytes and decreased expression of Ki67 in thymus tissues of Fstl1 +/mice suggested reduced proliferation of DN thymocytes and impaired T cell development in Fstl1 +/mice.
Developing thymocytes receive a wide array of signals from the thymic microenvironment for proliferation, differentiation and selection. IL-2 is a soluble cytokine that has been shown to promote the proliferation of DN thymocytes and to weakly induce the differentiation of DP thymocytes to CD8 + SP thymocytes [44][45][46]. In previous studies, Il-2 -/mice were shown to exhibit altered thymic architecture and significantly reduced numbers of DN, DP thymocyte [47]. These findings suggest that IL-2 promotes the proliferation and differentiation of T cells, and that the decreased expression of IL-2 may be the reason for impaired T cell development in thymuses of Fstl1 +/mice.
Establishment and maintenance of thymus requires TEC cells to support the shape of thymus and to nurse T cells [48], therefore, TEC dysfunction can affect the thymic microenvironment. Our studies indicated that mTEC cells were the major cellular source of FSTL1 in the thymus, and knockdown of Fstl1 in mTEC cells inhibited the mRNA levels of Icosl. ICOSL binds to its receptor on CD4 + T cells, which induces the production of IL-2 [33]. The underlying mechanism of FSTL1 in promoting the expression of ICOSL in mTEC cells will be explored in future work.
We demonstrated that deficiency of FSTL1 in mTEC cells decreased the expression of Icosl, which inhibited the crosstalk between mTEC cells and CD4 + SP thymocytes. Then the reduction in IL-2 production from CD4 + SP thymocytes decreased the proliferation of DN thymocytes, which in turn might block the differentiation into CD4 + SP ( Figure 5F). The feedback suppressive loop may be critical to impair the development of T cells. Therefore, FSTL1 deficiency in mTEC cells decreased the proliferation of DN thymocytes and production of T cells, which could result in immunodeficiency and impairment of adaptive immunity. Once breast cancer cells have metastasized to the lung, there were not enough effector T cells to proliferate, differentiate and migrate to the metastatic site. The increased metastatic growth of 4T1 cells in lung inhibited the activity of Th1 and CD8 + T cells in lungs of FSTL1 deficiency mice, and in turn, fewer anti-tumour T lymphocytes accelerated the progression of lung metastasis.

Mice
All animal experiments in this study were approved by Administration Regulations on Laboratory Animals of Beijing Municipality. Fstl1 +/mice were generated by intercrossing E II a-Cre;Fstl1 flox/+ mice, and were purchased from Model Animal Research Center of Nanjing University. All mice were backcrossed to the BALB/c background for more than 10 generations. The mice used in the study were 8-week-old. All animals were housed and maintained under pathogen-free conditions. All mice were AGING provided with adequate water and food under controlled environmental conditions.

Cell lines
4T1 cell line was a gift from Ning's Lab, College of Life Sciences, Nankai University. 4T1 cells were cultured in RPMI-1640 (Gibco) containing 10% fetal bovine calf serum (Gibco) at 37° C and 5% CO2. mTEC cell line was obtained from Dr. Shao of Jiangsu University as a gift. mTEC cells were cultured in DMEM (Gibco) containing 10% FBS (Gibco) at 37° C and 5% CO2.

Animal models
The 4T1 cell suspension was adjusted to a concentration of 10 7 cells/mL, and 10 6 cells were injected into the mammary fat pad of female BALB/c WT (as control group) and Fstl1 +/mice. After the mice were euthanized at day 14, the lungs were collected. Since 4T1 colonies were invisible on the lung surface, histological analysis was carried out to observe the size of tumors in lung tissues.

Cell counting kit-8 (CCK-8) assay
The 4T1 cell suspension was adjusted to a concentration of 10 5 cells/mL in complete RPMI-1640 100μL cell suspension was seeded into each well of 96-well plate to adhere overnight (day 1) or continue to culture for another 24 h (day2) and 48 h (day3). At the indicated time, 10μL CCK-8 solution (Dojindo) was added to each well and incubated for another 1 h. The absorbance at 450 nm wavelength of each well was detected. Data pertaining to cell viability were normalized to day 1.

Quantitative RT-PCR (qRT-PCR)
The thymus tissues and 4T1 cells were collected, treated with TRIzol (Thermo Fisher Scientific). After extraction of total RNA, cDNA was synthesized with FastKing-RT SuperMix (TIANGEN, China). qRT-PCR was performed on a Bio-Rad CFX Connect RT-PCR detection system. The primer sequences used are presented as Supplementary Material (Supplementary  Table 1).

Immunohistochemistry (IHC) staining
Lung tissues and thymus tissues were fixed in 4% paraformaldehyde for 24 hours, dehydrated in 70% ethyl alcohol, and embedded in paraffin. For hematoxylin and eosin (H&E) staining, lung tissues were stained with H&E. The tumor sizes in lung tissues were measured using Image J software.
For IHC staining, 4 μm thymus tissue serial slices were incubated with antibody against FSTL1 (20182-1-AP, Proteintech). 5 μm lung tissue serial slices were incubated with antibodies against CD4 (25229s, Cell Signaling Technology), CD8 (98941s, Cell Signaling Technology). The expression of CD4 and CD8 were quantified by counting the number of positive cells in each image (×40). The average number of positive cells from five random fields was used for statistical analysis.

Flow cytometry analysis
Lung and thymus tissues were cut into 2-mm pieces, then mechanically disintegrated and passed through cell strainers (BD Bioscience) to obtain the single cell suspensions. Red blood cells from peripheral blood were directly lysed to obtain single cell suspensions. First, Fixable Viability Stain 450 (562247, BD Bioscience) was performed for live-dead marker in the absence of other antibodies. The cells were incubated with anti-CD16/32 (101301, Biolegend) for blocking Fc receptor. Dividing cells were stained with anti-CD3-PE (Biolegend), anti-CD4-FITC (Biolegend) and anti-CD8a-APC (Biolegend) for 30 min on ice. To determine intracellular cytokine levels, cells were further fixed, permeabilized, and stained with anti-IFNγ-AlexaFluor 700 (BD Bioscience) and anti-IL-4-APC (BD Bioscience). Data were acquired with LSR Fortessa flow cytometer (BD Bioscience), and analyzed with the Tree Star Flowjo software.

Statistical analysis
All data were analyzed using Prism6 (Graph Pad). Results are presented as mean ± standard deviation (SD). Unpaired two-tailed student's t-tests was used to assess between-group differences. One-way ANOVA was used for multi-group comparisons. P-values less The numbers of DN, DP, CD4 + SP and CD8 + SP thymocytes (n=8, WT; n=9, Fstl1 +/-). Data are presented as mean ± SD. Each dot in the graphs represents an individual mouse. n.s., not significant; *p < 0.05, **p < 0.01.