Plac1 Is a Key Regulator of the Inflammatory Response and Immune Tolerance In Mammary Tumorigenesis

Plac1 is an X-linked trophoblast gene expressed at high levels in the placenta, but not in adult somatic tissues other than the testis. Plac1 however is re-expressed in several solid tumors and in most human cancer cell lines. To explore the role of Plac1 in cancer progression, Plac1 was reduced by RNA interference in EO771 mammary carcinoma cells. EO771 “knockdown” (KD) resulted in 50% reduction in proliferation in vitro and impaired tumor growth in syngeneic mice; however, tumor growth in SCID mice was equivalent to tumor cells expressing a non-silencing control RNA, suggesting that Plac1 regulated adaptive immunity. Gene expression profiling of Plac1 KD cells indicated reduction in several inflammatory and immune factors, including Cxcl1, Ccl5, Ly6a/Sca-1, Ly6c and Lif. Treatment of mice engrafted with wild-type EO771 cells with a Cxcr2 antagonist impaired tumor growth, reduced myeloid-derived suppressor cells and regulatory T cells, while increasing macrophages, dendritic cells, NK cells and the penetration of CD8+ T cells into the tumor bed. Cxcl1 KD phenocopied the effects of Plac1 KD on tumor growth, and overexpression of Cxcl1 partially rescued Plac1 KD cells. These results reveal that Plac1 modulates a tolerogenic tumor microenvironment in part by modulating the chemokine axis.


Reduction of Plac1 inhibits EO771 cell growth and tumor formation.
To characterize the functional role of Plac1, several mouse mammary tumor cell lines were screened by qRT-PCR for Plac1 RNA expression; among these, EO771 cells expressed the highest level, which was substantial in comparison to mouse placenta (Fig. 1a). EO771 cells were then transduced with recombinant lentiviruses expressing shRNAs targeting four regions of Plac1 mRNA (Fig. 1b). shRNA490 produced >98% reduction of Plac1 expression, and EO771 cells transduced with this shRNA (EO771/shPlac1) were used for further studies. EO771/shPlac1 cells grew in monolayer culture at 50% of the rate of control cells expressing a non-silencing RNA (Fig. 1c). Gene expression profiling revealed that Plac1 markedly suppressed several chemokine genes, including Cxcl1, Ccl7, Ccl2, Ccl5 and Cxcl10, as well as immune-related factors Lif, Ly6a/Sca-1, Ly6c and CD274 (Table 1, Fig. 1d, Supplementary Table 2). Changes in the expression of several of these genes were confirmed by qRT-PCR and most were consistent with the array profile (Fig. 1e).

Plac1 reduction blocks tumor growth in syngeneic mice, but not in SCID mice.
To determine the influence of Plac1 on tumor development in a host, EO771/shPlac1 cells were implanted in syngeneic C57BL/6 mice, and growth monitored by caliper measurement (Fig. 2a). EO771/shPlac1 isografts grew transiently in syngeneic mice, but when transplanted into SCID mice, their growth was similar to EO771/Scr control cells (Fig. 2b). Although all control isografts growing in syngeneic mice after 21 days expressed Plac1, there was little residual staining in mammary tissue from EO771/shPlac1-engrafted mice (Fig. 2c). Although Plac1 reduced growth in vitro, the lack of sustained growth of EO771/shPlac1 cells in syngeneic mice suggested that interactions with the tumor microenvironment may have contributed to this effect.
Cxcr2 antagonist inhibits tumor growth. Since Cxcl1 exhibited the greatest change in expression among all chemokine genes after Plac1 downregulation (Table 1), mice were treated with a Cxcr2 antagonist to determine if it would affect tumor growth to a similar extent. Animals engrafted with EO771 cells were treated three times EO771 cells were transduced with lentiviruses expressing crambled RNA (Scr) or four Plac1 shRNAs designated sh81, sh187, sh300 and sh490; sh490 inhibited RNA expression >98%, and these cells were designated EO771/shPlac1. (c) EO771/Scr and EO771/shPlac1 cells were grown as monolayers, and the number of viable cells were quantified by sulforhodamine B staining. Shown is the mean ± S.D. of triplicate analysis of three samples. The growth of EO771/shPlac1 cells differed significantly (P < 0.001) from EO771/Scr cells by the two-sided Student's t test.
SCientifiC RepoRts | (2018) 8:5717 | DOI:10.1038/s41598-018-24022-w per week with vehicle or 2 or 20 mg/kg SB225002 beginning 10 days after transplantation (Fig. 3a). Whereas, the lower dose partially inhibited tumor growth, 20 mg/kg SB225002 completely blocked growth after a lag of two weeks. Tumor stasis at 17 days following cell inoculation was associated with reduction in expression of immune and chemokine genes, many of which were downregulated in EO771/shPlac1 cells (Fig. 3b). Cell sorting of tumor immune infiltrates indicated that SB225002 treatment reduced myeloid-derived suppressor cells (MDSC) and Treg cells, and increased CD8 + /CD4 + T cells, NK cells, macrophages and dendritic cells (Fig. 3c,d). Especially noteworthy was the greater infiltration of CD8 + T cells into the tumor bed following SB25002 treatment (Fig. 3e), which was accompanied by increased macrophage and Treg cell infiltration, reduction of Plac1 and increased apoptosis (Fig. 3f). Since we did not determine tissue levels of SB25002, we determined its cytotoxicity in EO771 cell culture ( Supplementary Fig. 2). SB25002 at concentration less than 100 nM were not cytotoxic, but produced cytotoxicity at concentrations exceeding 1000 nM. Since we have not carried out pharmacokinetics, the contribution of SB25002 cytotoxicity to its antitumor effect cannot be ascertained.

Cxcl1 reduction inhibits tumor growth and immune cell-related transcription.
To evaluate the role of Cxcl1/Cxcr2 signaling in tumor growth, EO771 cells were transduced with scrambled shRNA (Scr) or four Cxcl1 shRNAs (Fig. 4a). EO771 cells expressing sh174 (shCxcl1) exhibited >95% reduction of Cxcl1 RNA expression. After 48 hr in monolayer culture, EO771/shCxcl1 cells grew at approximately 30% of the rate of control cells (Scr) (Fig. 4b). Comparison of the gene expression profile of EO771/shCxcl1 cells with EO771/Scr cells revealed a small subset of genes with ≥3-fold changes, including Ly6a, IL23a, C3, Cxcl1 and CD68 (Table 2,  Supplementary Table 3). Transplantation of EO771/shCxcl1 cells into syngeneic mice resulted in impaired tumor growth in comparison to control cells (Fig. 4c), a result that was similar to EO771/shPlac1 cells (Fig. 2a). Changes in immune-related gene expression (   Growth of EO771 cells in syngeneic mice following treatment with a Cxcr 2 antagonist. (a) Syngeneic 57BL/6 mice were inoculated in the mammary gland with 1 × 10 6 at five weeks of age, and injected i.p. daily with vehicle (blue) or 2 mg/kg (red) or 20 mg/kg (green) SB225002 beginning 11 days after cell inoculation. SB225002 completely suppressed tumor growth after 14 days. Differences between vehicle-and 2 mg/kg SB225002-treated mice were not significantly different (P = 0.145); differences between vehicle-and 20 mg/ kg SB225002-treated mice were significantly different (P = 0.005) by the unpaired two-tailed Student's t test.
Shown is the mean ± SD, N = 5 per group. (b) Immune gene expression in tumors 17 days after treatment with 20 mg/kg SB225002. Shown is the relative expression in control and SB225002-treated mice in comparison to their changes in EO771/shPlac1 cells (Table 1) did not change significantly (Fig. 4d). Measurement of Cxcl1 in tumors or mammary tissue by IHC after 21days indicated the presence of Cxcl1 in EO771/Scr tumors, but not in EO771/shCxcl1 engrafted mammary tissue (Fig. 4e). Comparison of gene expression in EO771/shPlac1 vs. EO771/shCxcl1 cells indicated reduced expression in five genes in common, viz. CD68, Cxcl1, Ly6a, Plau and Rgs16 (Table 3), although CD68 was not changed significantly in EO771/shCxcl1 cells as measured by qRT-PCR (Fig. 4d).

Cxcl1 partially rescues Plac1 reduction in EO771 cells.
To determine the contribution of Cxcl1 to the effects of Plac1 downregulation on tumor growth, EOI771/sh490 cells were transduced with a retrovirus expressing Cxcl1-mCherry and EO771/Scr and EO771/sh490 were transduced with mCherry alone (Fig. 5). After selection in G418, a significant percentage of cells co-expressed GFP and mCherry (Fig. 5a) and Cxcl1 mRNA (Fig. 5b). The growth of EO771/sh490 cells in vitro was slower rate than control cells as shown in Fig. 1c, but cells expressing Cxcl1 largely rescued this effect (Fig. 5c). Isografts of these cell lines in syngeneic mice confirmed the  poor growth of EO771/sh490 cells, and further showed that Cxcl1 could partially rescue their poor tumorigenicity (Fig. 5d).
As an added proof of the function of Plac1 in tumorigenesis, MC cells, a cell line with low Plac1 expression (Fig. 1a), were transfected with Plac1 ( Supplementary Fig. 3, Supplementary Table 5). MC/Plac1 cells grew at a higher rate than control cells and exhibited less apoptosis, and upregulated a gene expression profile that included several of the chemokines as well as CD274 that were downregulated in EO771/shPlac1 cells.

Discussion
The present study establishes the first link between the trophoblast gene, Plac1, and adaptive immunity, through its ability to modulate chemokine expression and other immune cell regulators. In retrospect, this is not too surprising since the placenta may be regarded as a foreign allograft protected from host vs. graft rejection 23 in part by the presence of Treg cells in the uterine decidua 24 . However, the link between Plac1, chemokine signaling and immune tolerance in tumors is a novel and relevant finding since the latter processes are hallmarks of most, if not all, solid tumors 25,26 . The relevance of Plac1 to mammary tumorigenesis was first noted in MMTV-PPARd mice, where Plac1 was markedly upregulated at the onset and throughout tumor development 14 . This finding implicated  Table 3. Gene expression common to EO771/shPlac1 and EO771/shCxcl1 cells. Shown is the ratio between EO771/shPlac1 or EO771/shCxcl1 cells to EO771/Scr control cells for genes with ≥3.0-fold changes in expression and a raw score ≥300. nuclear receptor signaling in the transcriptional regulation of Plac1, as noted previously for its activation at alternate promoter regions in the Plac1 locus by LXR and RXRA 27 . However, from a mechanistic perspective, the downstream intracellular components interacting with Plac1have not been determined. Plac1 is predominantly extracellular with an N-terminal signal peptide, a small transmembrane domain and an extracellular ZP3 domain 28 , which promotes protein-protein interactions. By analogy, cytokine receptors lacking an intracellular signaling domain partner with co-receptors, adapter molecules and cytosolic protein tyrosine kinases to effect signaling 29 , and such a mechanism may also pertain to Plac1. In the present study, the panel of immune-related genes down-regulated by Plac1 'knockdown' (Table 1) suggest that one of its functions is to modulate chemokine effector pathways associated with immune evasion, such as antigen presentation, angiogenesis and myeloid cell, T cell and fibroblast activation (see scheme in Supplementary  Fig. 4). One dominant downstream effector pathway was the Cxcl1/Cxcr2 axis, as shown by inhibition of tumor growth by the Cxcr2 antagonist SB25002 (Fig. 3a), and the inability of EO771/shCxcl1 cells to sustain tumor proliferation in syngeneic mice (Fig. 4c). Inhibition of Cxcr2 by SB25002 was associated with reduction of Treg cells and MDSC, and an increasing the percentage of CD8 + T cells, macrophages, NK cells and dendritic cells, which was consistent with previous studies demonstrating the ability of SB225002 to suppress MDSC infiltration in breast tumor xenografts 30 and prostate tumors 31 , as well as metastasis via S100A8/A9 30 . Thus, our data suggest that one role of Plac1 may be to maintain the production of inflammatory and immunoregulatory chemokines to effect changes in the stromal microenvironment conducive to immune tolerance and poor outcome 30,32,33 , which would explain the poor tumorigenicity of EO771/shPlac1 cells in syngeneic mice, but not in SCID mice. Although, co-expression of Plac1 and Cxcl1 in breast cancer tissue has not been reported, Cxcl1 expression in breast cancer biopsies was found to be elevated in metastases, and inversely related to ERα expression and relapse-free survival 34 .
Despite focusing on the immunological aspects of Plac1 function, it was apparent that it affected several signaling pathways in EO771 cells (Supplementary Table 2). Changes in gene expression common to both EO771/ shPlac1 and EO771/shCxcl1 cells, apart from Cxcl1, included reduced expression of Plau, Ly6a, CD68 and Rgs16 (Table 3). Plau is a well-known marker of metastasis 35 , and its role in invasion has been noted in trophoblast migration 36 . Ly6a/Sca-1 is a mouse stem cell and tumor-initiating cell biomarker 37,38 that down-regulates several tumor suppressor pathways, including PPARγ, TGF-β and PTEN 39,40 . CD68 is a scavenger receptor involved in phagocytosis, particularly in M2 polarized macrophages 41 , and has been implicated in immunotolerance by tumor-associated macrophages 42 . Rgs16 is a G-protein-coupled receptor associated with vascular smooth muscle cell proliferation and angiogenesis, which play prominent roles in oncogenesis 43 . Thus, the functions of Plac1 in placental development 44 appear to phenocopy its functions in tumorigenesis, which supports the onco-placental nature of cancer 3 . From a therapeutic perspective, our data not only suggest that Plac1 may be a potential drug target, but that chemokine receptor antagonists developed for chronic inflammatory disorders, including COPD and psoriasis 45,46 , may be useful adjuvants when used in combination with other therapies to enhance the efficacy of cancer treatment 47,48 .

Methods
Cell lines. EO771 cells were originally isolated from a spontaneous mammary carcinoma in C57BL/6 mice 49 , and were provided by Dr. Louis M. Weiner, Georgetown University. EO771 cells tested negative against the IMPACT II panel of infectious agents (IDEXX). Plac1 and Cxcl1 expression were reduced with the piLenti-sRNA-GFP lentiviral vector targeting the sequence 5′-CCACCTTATGTCTACAATCAAAAGAGCAT-3′ of Plac11 mRNA (cat #i034429, ABM, Vancouver, Canada) or the sequence 5′-CTGCACCCAAACCGAAGTCATAGCCACAC-3′ of Cxcl1 mRNA (cat # i042697, ABM); a scrambled sequence was used as a control. Lentivirus expression of Cxcl1 utilized Lenti ORF clone Cxcl1 (MR220966L1) from Origene and lenti vector CMV-mCxcl1-IRES-mCherry (VB170524-1062) from VectorBuilder or the vector lacking Cxcl1 as a control. HEK293T cells were co-transfected at 50% confluence with the lentiviral shRNA plasmid, psPAX2 packaging plasmid and the VSV-G/pMD2 envelope plasmid at a ratio of 2:1:0.1 using Fugene 6 (Promega). After 18 hr, medium was replaced with fresh growth medium, and after 24-48 hr, the virus-containing supernatant was collected, filtered through a 0.45 um filter, mixed with fresh cell culture medium at a ratio of 7:1, and added to EO771 cells with 8 µg/ml polybrene. A lentivirus expressing a scrambled non-silencing control shRNA (shRNAmir, ABM) served as a negative control. Cells were selected for stable integration of the virus by incubation with 7.5μg/ml puromycin (Sigma-Aldrich Corp.) for 10 days. The efficiency of integration was monitored by GFP expression by the lentivirus. Cell lines 34T, 105T, 437T, MC and NeuT were described previously 50,51 . Animals. EO771 cells at an inoculum of 1 × 10 6 cells/0.1 ml were injected into the no. 4 mammary gland of C57BL/6 or SCID mice (Taconic), and tumor growth was monitored daily. Cxcr2 antagonist SB225002 (Sigma-Aldrich) was dissolved in a diluent containing 8% DMSO, 10% PEG-400 and 1.75% Tween-20 in water at a concentration of 0.4 or 4.0 mg/ml, and administered i.p. Monday through Friday at a dose of 2 or 20 mg/kg, respectively 52,53 . Other mouse mammary tumor cells lines tested for Plac1 expression were 34T, 105T, 437T and MC 40,51 . Animal studies were conducted under protocols approved by the Georgetown University Animal Care and Use Committee (protocol 2016-1143) in accordance with NIH guidelines for the ethical treatment of animals.
Cell growth and cytotoxicity assays. EO771 or MC cells were grown in 96-well plates at an initial density of 5,000 cells per well. The cytotoxicity of SB225002 was determined in EO771 cells by dissolving the drug in DMSO and diluting in medium to a final DMSO concentration of 0.001%. Cell density was determined after incubation for 24, 48 and 72 hr by sulforhodamine B staining and measuring optical density at 570 nm 54 . Cytotoxicity assay data are shown in Supplementary Fig. 3.
SCientifiC RepoRts | (2018) 8:5717 | DOI:10.1038/s41598-018-24022-w Histopathology and immunohistochemistry (IHC). Mammary tumors were excised, and formalin-fixed, paraffin-embedded sections were prepared for H&E staining and IHC by the Tissue and Histopathology Shared Resource, LCCC. Antigen retrieval was carried out by incubation of tissue sections in 10 mM sodium citrate buffer (pH 6.0) for 20 min at a sub-boiling temperature in an electric steamer as previously described 14,51,55 . Endogenous peroxidase activity was quenched with 3% hydrogen peroxide for 10 min, and incubated for 30 min with blocking solution (10% goat serum in Tris-buffered saline), followed by incubation overnight at 4 °C with the appropriate primary antibody diluted in blocking solution. Biotin-conjugated secondary antibodies were diluted in TBS containing 0.1% Tween-20 and incubated for 30 min at room temperature using the ABC Vectastain (Vector Laboratories) detection system and diaminobenzidine (Pierce), and slides were counterstained with Harris-modified hematoxylin (Thermo-Fisher, Inc.), dehydrated and mounted in Permount (Thermo-Fisher, Inc.). Apoptosis was determined with the SignalStain Apoptosis IHC Detection kit for cleaved caspase-3 (Cell Signaling Technology). Antibodies and their dilutions for IHC and FACS are listed in Supplementary Table 1.

Fluorescence-Activated Cell Sorting (FACS).
Tumor immune infiltrates were obtained by excising tumors, mincing them into small pieces and digestion with collagenase D (Roche) at a ratio of 15 ml collagenase solution per 2 g of tissue for 1 hr at 37 °C with shaking. The cell suspension was filtered through a 70 μm cell strainer (Falcon), washed, erythrocytes lysed and 1 × 10 6 cells were analyzed by FACS. Cells were stained using the Live/ Dead Fixable Dead Cell Stain Kit (Invitrogen) and excluded from analysis, and non-specific binding was blocked with Fc antibody CD16/32 (Biolegend). Cells were first sorted for CD45 (macrophages, MDSC, Treg cells, NK cells and dendritic cells), CD45 + CD3 + (T cells) or CD45 + CD4 + (Treg cells). Cells were further sorted for: macrophages: F4/80 + /MHCII + , MDSC: CD11b + /Gr-1 + , dendritic cells: CD11c + /MHCII + , T cells: CD4 + /CD8 + , NK cells: CD45 + /NK1. Gene microarray analysis. Microarray analysis was carried out as previously described 39,51,[55][56][57] . Briefly, tissue was snap-frozen in liquid nitrogen, pulverized in a mortar and pestle and RNA was extracted using an RNeasy Mini Kit (Qiagen) according to the manufacturer's protocol. RNA purity was assessed by the integrity of 18S and 28S rRNA using an Agilent microfluidic chip. Array analysis was carried out with cRNA prepared from equal amounts of RNA (1 μg) pooled from three replicates of cells per group. Biotin-labeled cRNA was fragmented at 94°C for 35 min and hybridized overnight to an Affymetrix mouse 430A 2.0 GeneChip ® , and scanned with an Agilent Gene Array scanner. Grid alignment and raw data generation used the Affymetrix GeneChip ® Operating software 1.1. A noise value (Q) based on the variance of low-intensity probe cells was used to calculate a minimum threshold for each GeneChip. Samples were averaged and data refined by eliminating genes with signal intensities <300 in both comparison groups, and heat maps were generated from ≥3-fold changes in gene expression normalized to control tissue using unsupervised hierarchical cluster analysis as previously described 58 . Gene expression data for EO771 control, EO881/shPlac1 and EO771/shCxcl1 cells are included in Supplementary  Tables 2 and 3, respectively. Gene expression data for mice treated with vehicle or 20 mg/kg SB25001 are included in Supplementary Table 4. Gene interaction analysis utilized Ariadne Pathway Studio version 9.1 ( Supplementary  Fig. 4). Data sets were deposited in the GEO public database under accession no. GSE78202. Quantitative real-time polymerase chain reaction (qRT-PCR). Total RNA was extracted as described above, and RNA (1 µg) from each of 3 samples per group was reverse transcribed using the Omniscript RT kit (Qiagen) as previously described 39,51,[55][56][57] . PCR was performed in triplicate using an ABI-Prism 7700 (Applied Biosystems, Foster City, CA) with SYBRGreen I detection (Qiagen) according to the manufacturer's protocol. Amplification using the appropriate primers (Supplementary Table 5) was confirmed by ethidium bromide staining of the PCR products on an agarose gel. The expression of each target gene was normalized to GAPDH and is presented as the ratio of the target gene to GADPH expression calculated using the formula, 2 −ΔCt , where ΔCt = Ct Target − Ct18s 39 . Statistical analysis. Statistical significance of means ± S.D. were evaluated using the two-tailed Student's t test at a significance of P < 0.05. Differences in tumor growth in vivo were determined by the unpaired two-tailed Student's t test at a significance of P < 0.05.