Golgi Phosphoprotein 2 Is a Novel Regulator of IL-12 Production and Macrophage Polarization

A growing number of studies have revealed B cell and T cell interactions. T cells help to induce the terminal differentiation of B cells to Ig class-switched plasma cells (1, 2). Reciprocally, B cells can induce direct tolerance of Ag-specific CD8 T cells (3), induce T cell anergy via TGF-β production (4), and also influence Th1/Th2 differentiation via the production of regulatory factors (5). Similarly, B cells can exert a regulatory function in in vivo models of T cell immunity, including tumor rejection (6), experimental autoimmune encephalitis (7), and rheumatoid arthritis (8). In mice, a relatively rare spleen B cell subset with IL-10–dependent negative T cell–regulating function has recently been identified and named as B10 cells (9, 10), which indirectly modulate T cell–mediated autoimmunity by inhibiting the Ag-presenting capacity of dendritic cells (DCs) (11). Furthermore, it was reported that homeostatic regulation of T cell trafficking across the endothelium is controlled by a B cell–secreted peptide in autoimmune and chronic inflammatory disease (12). However, how B cells regulate T cell–mediated antitumor immunity at the molecular level has not been clearly established.

IL-12, a heterodimer composed of p35 (encoded by Il12a) and p40 (encoded by Il12b), is primarily produced by APCs, including macrophages/monocytes and DCs, in response to stimulation by certain pathogens. It principally activates NK cells and induces naive CD4 T cells to differentiate into IFN-γ–producing Th1 cells, and it is a crucial link between innate and adapt immunity (13, 14). IL-12 together with other members of this unique heterodimeric cytokine family IL-23 (composed of p19 and IL-12p40), IL-27 (p28 and EBI3), and IL-35 (EBI3 and IL-12p35) have been shown to be crucial playmakers of autoimmune and inflammatory responses, disorders, and cancers (15). However, there are still many unresolved questions to be addressed concerning the sometimes enigmatic IL-12.

In an effort to explore how B cells regulate T cell–mediated antitumor immunity, we adopted a proteomics-based approach and identified a secreted activity by activated malignant B cells. This novel activity strongly suppressed T cell activation indirectly by selective inhibition of the production of IL-12 in DCs. Further biochemical analysis revealed this IL-12 inhibitor as Golgi phosphoprotein 2 (GOLPH2). During the past years, a large number of independent studies have identified high levels of GOLPH2 in the sera of patients with liver diseases, particularly hepatocellular carcinoma (HCC) (16–18). The most profound elevation of serum levels of GOLPH2 was observed in patients who had developed HCC on the background of hepatitis C virus (19) and hepatitis B virus infections (20). The level of serum GOLPH2 is also significantly elevated in lung cancer and prostate cancer (21, 22). In light of the seemingly broad role of GOLPH2, we focused our study on its IL-12–inhibiting activity in professional APCs (DCs and macrophages) in inflammatory settings.

GOLPH2 knockout (KO) mice were established by Shanghai Research Center of Model Organisms with TALEN Technology (Shanghai, China). Mice were genotyped by PCR analysis of tail biopsy DNA using primers as follows: P1, 5′-GATCCAGTCTAGCCACAGCTT-3′; P2, 5′-AACTTGACAGGATCCAGTCTAGCTGGAG-3′; P3, 5′-GACCCTGGAACACAACTCCC-3′. P1 and P3 and P2 and P3 were separately used for wild-type (WT) and KO mice genotyping. Mice were bred and maintained in a specific pathogen-free facility and were treated in accordance with protocols approved by the Institutional Animal Care and Use Committee of Shanghai Jiao Tong University. The mouse macrophage cell line RAW264.7 was purchased from the Cell Center of the Chinese Academy of Sciences (Shanghai, China).

LPS (0127:B8 or 011:B4) were purchased from Sigma-Aldrich (St. Louis, MO). All recombinant proteins used in the paper containing human and murine GM-CSF, IL-4, TNF-α, IL-13, and IFN-γ were from PeproTech (Rocky Hill, NJ). Human CD4 beads, CD14 beads, and CD19 beads and murine CD4 beads, CD8 beads, and CD11c beads were purchased from Miltenyi Biotec. The human IL-12p35 and IL-12p40 promoter–luciferase constructs were described as previously (23). GOLPH2 expression plasmid was generated by inserting the open reading frame into the pTriEx-3 Neo or PLVX-EF1α vector.

Bone marrow–derived macrophages (BMDMs) were generated as previously described. Briefly, bone marrows isolated from femurs and tibias (day 0) were cultured in DMEM supplemented with 10% FBS and 20% L929 conditioned medium with fresh medium replenished on day 4. On day 7, cells were collected and seeded for different experiments. BMDCs were generated by culturing isolated bone marrow cells for 7 d using RPMI 1640 medium containing 10% FBS (Life Technologies/Thermo Fisher Scientific) supplemented with murine GM-CSF (20 ng/ml) and IL-4 (20 ng/ml). Primary mouse peritoneal macrophages were obtained from peritoneal exudates. Generally, peritoneal exudates isolated from mice were washed with PBS twice and cultured in DMEM for 3 h. The nonadherent cells were removed by washing with warm PBS. The adherent cells were analyzed by an Accuri C6 flow cytometer with the mouse macrophage marker F4/80 and >90% of them were F4/80⁺ macrophages.

IL-12p40, p70, IL-10, and TNF-α levels of cell supernatants or serum were analyzed by ELISA kits from BD Pharmingen (San Diego, CA), according to the manufacturer’s instructions. Total RNA was isolated using the RNeasy kit (Qiagen) and subjected to cDNA using a Transcriptor first-strand cDNA synthesis kit (Roche). Real-time PCRs were performed in triplicate, using a sequence detection system (CFX; Bio-Rad Laboratories). The primers used in the real-time PCR (RT-PCR) were as follows: Golph2 (forward, 5′-AACTCCAGACGCGCATTGTAG-3′; reverse, 5′-GCCTGTGGTGATGTTATTCACT-3′), Il-12p35 (forward, 5′-CCC-TTGCCCTCCTAAACCAC-3′; reverse, 5′-TAGTAGCCAGGCAACTCTCG-3′), Il-12p40 (forward, 5′-GGAAGCACGGCAGCAGAATAAAT-3′; reverse, 5′-AACTTGAGGGAGAAGTAGGAATGG-3′), Tnfα (forward, 5′-CAAAGTCAAATCCTACCAAAGTGACC-3′; reverse, 5′-TGCTACTCCGAGCGTCAAAGACC-3′), Arg1 (forward, 5′-CCAGAAGAATGGAAGAGTCAGTGT-3′; reverse, 5′-GCAGATATGCAGGGAGTCACC-3′), Fizz1 (forward, 5′-GGTCCCAGTGCATATGGATGAGACCATAGA-3′; reverse, 5′-CACCTCTTCACTCGAGGGACAGTTGGCAGC-3′), Ym-1 (forward, 5′-TCACAGGTCTGGCAATTCTTCTG-3′; reverse, 5′-TTTGTCCTTAGGAGGGCTTCCTCG-3′), Ifit1 (forward, 5′-CTGAGATGTCACTTCACATGGAA-3′; reverse: 5′-GTGCATCCCCAATGGGTTCT-3′), Ifit2 (forward, 5′-AGTACAACGAGTAAGGAGTCACT-3′; reverse: 5′-AGGCCAGTATGTTGCACATGG-3′). Ifit3 (forward, 5′-AGTGAGGTCAACCGGGAATCT-3′; reverse, 5′-TCTAGGTGCTTTATGTAGGCCA-3′), Ido (forward, 5′-GGCTAGAAATCTGCCTGTGC-3′; reverse, 5′-AGAGCTCGCAGTAGGGAACA-3′), Cxcl10 (forward, 5′-CCAAGTGCTGCCGTCATTTTC-3′; reverse, 5′-GGCTCGCAGGGATGATTTCAA-3′), IFNβ (forward, 5′-CAGCTCCAAGAAAGGACGAAC-3′; reverse, 5′-GGCAGTGTAACTCTTCTGCAT-3′), Irf1 (forward, 5′-GCAAAACCAAGAGGAAGCTG-3′; reverse, 5′-GCTGCCACTCAGACTGTTCA-3′), Hprt (forward, 5′-TTATGGACAGGACTGAAAGAC-3′; reverse, 5′-GCTTTAATGTAATCCAGCAGGT-3′), Gapdh (forward, 5′-GTCAA CGGATTTGGTCGTATT-3′; reverse, 5′-GATCTCGCTCCTGGAAGATGG-3′).

Transfections of RAW264.7 cells were performed as previously described (24). Generally, RAW264.7 cells were transiently transfected with the Il-12p35 or Il-12p40 promoter–luciferase reporter construct together with expression vectors for GOLPH2 and then stimulated by murine IFN-γ (10 ng/ml) for 16 h, followed by LPS (1 μg/ml) for the amount time indicated. Controls were transfected with the empty vector. β-Galactosidase activity was measured for normalization.

Cell lysates were lysed with RIPA buffer supplemented with protease inhibitors. The samples were used for Western blot with Abs against GOLPH2 (ab109628, 1:1000; Abcam) and β-actin (sc-47778, 1:1000; Santa Cruz Biotechnology). Signal intensity was measured with the Odyssey fluorescence detection system (LI-COR Biosciences, Lincoln, NE).

In brief, nuclear extracts were incubated with Ab against GC-binding protein (GC-BP) (25) and agarose conjugate overnight, and the immunoprecipitated complex was collected after wash, followed by SDS-PAGE separation and immunoblot with the Ab against p-Tyr (PY99) (sc-7020, 1:500; Santa Cruz Biotechnology).

The chromatin immunoprecipitation (ChIP) procedure was performed by using an assay kit following the manufacturer’s instructions (Merck Millipore), with 1 × 10⁷ BMDMs per condition. The input DNA was diluted 100 times before PCR. The input and precipitant DNAs were amplified via PCR with the primer pair for mouse IL-12 p35 promoter. The following primers were used: forward, 5′-CGATCGACGCACTTGTCCTT-3′; reverse, 5′-AGCGTGATTGACA CATGCTG-3′.

Cell supernatants from resting and LPS-stimulated RAMOS cells cultured in the absence of FBS for 24 h and that of 2E2 were boiled for 30 min, followed by trypsin treatment (50 ng/ml) for 30 min, fractionation through an SDS-PAGE gel, and silver staining. The arrow-pointed bands from both resting and LPS-stimulated RAMOS cells were excised and analyzed by mass spectrometry at the Proteomics Resource Center of the Rockefeller University.

Total RNA isolated from BMDCs of GOLPH2 KO and WT mice were used for library construction using Illumina TruSeq strand-specific mRNA sample preparation system following the manufacturer’s protocols. High-throughput sequencing was performed on Illumina HiSeq 2000 with >39 million paired-end reads per sample. The reads were aligned to mm10 assembly (National Center for Biotechnology Information Assembly database) using TopHat (v2.0.9) (26). The quality of the raw sequencing reads were checked using FastQC (v0.10.1) and RSeQC (27). The count of gene reads was calculated by HTSeq (v0.6.0) (28). The differentially expressed genes (DEGs) were analyzed by DEGseq (v1.24.0) (29) in R software with a cutoff of p ≤ 0.001, false discovery rate (FDR) ≤ 0.1%, and fold change ≥ 1.5 that selected for further pathway analyses.

Functional analyses of DEGs were performed using DAVID for gene ontology (30) analysis and Ingenuity Pathway Analysis (IPA) (31) for ingenuity canonical pathway enrichment. The enriched terms of biology process in gene ontology analysis were selected using a cut-off of p < 0.05 and FDR < 25% and are shown by bar graphs. Enriched terms with the association to ingenuity canonical pathway (p < 0.05, FDR < 0.05, and z-score ≥ 1.5) are shown by a bar graph.

The LPS-induced shock model was established by i.p. injection of LPS (011:B4, 25 mg/kg) into 8-wk-old GOLPH2 KO and WT female mice. The mouse survival was recorded every 12 h.

A two-tailed Student t test was used for data analysis. A p value < 0.05 was considered significant. The survival data were analyzed by Kaplan–Meier curves and a log-rank (Mantel–Cox) test.

DCs from Ag-immunized, B-deficient mice produce higher levels of IL-12p70 than do DCs from WT animals, resulting in deviation toward Th1 (32), implying that B cells play a role in inhibiting the production of IL-12. To understand the molecular mechanisms whereby B cells exert their influence on DCs, we examined the effects of peripheral B cells from chronic lymphocytic leukemia (CLL) patients on cytokine production by myeloid DCs. As shown in Fig. 1A, LPS-stimulated DCs produced robust amounts of IL-12 and IL-10. Addition of B cells from CLL patients resulted in a strong inhibition of IL-12p40 and p70, but not TNF-α, IL-6, and TGF-β (data not shown). Moreover, use of a Transwell separating the B cells from DCs indicated that this IL-12–inhibiting activity was secreted by the B cells as a soluble factor. We thus designated it secreted IL-12 inhibitor (sIL12-I). To further understand the regulatory mechanism of sIL12-I at the transcriptional level, we used a well-established IL12 promoter–reporter gene system (33, 34). The human IL12p35 promoter linked to a firefly luciferase reporter was transiently transfected into RAW264.7 cells, to which culture supernatants from various types of B cells were added. Fig. 1B shows that malignant B cell lines tested contained the sIL12-I–like activity following LPS activation. Some of them, such as 2E2, were able to produce it constitutively. The sIL12-I–like activity, however, was not found in the supernatant of Jurkat cells, a T cell leukemia cell line. Primary murine splenic B lymphocytes, similar to its human counterpart, acquired highly potent sIL12-I activity to suppress IL12p35 transcription only when activated by LPS (Fig. 1C). Moreover, sIL12-I was able to suppress both IL12p35 and p40 transcriptional activities, albeit to different degrees, that is, it inhibited p35 more strongly than p40 (Fig. 1D).

One of the major roles of IL-12 is to induce IFN-γ production by activated T cells (35). We next assessed whether sIL12-I could influence the ability of T cells to produce IFN-γ in response to IL-12. As shown in Fig. 1E, resting murine splenic T cells secreted a low level of IFN-γ (column 2), which was boosted when LPS-stimulated DC supernatant (b) was added (column 7), indicating that DCs produced a soluble factor that enhanced T cell production of IFN-γ. When sIL12-I–treated, LPS-stimulated DC supernatant (d) was added to resting T cells (column 9), IFN-γ secretion was reduced to that of DC-produced IFN-γ (column 4). Adding LPS-stimulated DC supernatant (b) to Con A–activated T cells caused a strong boost of IFN-γ secretion (column 12), far more than the combined production of LPS-stimulated DCs (column 2) and Con A–stimulated T cells (column 10), indicating the effect of additional T cell–stimulating factors produced by DCs. This factor was not present when DCs were either not stimulated (column 11) or LPS-stimulated but also treated with sIL12-I (column 14). Importantly, adding the sIL12-I–treated LPS-stimulated DC supernatant to Con A–stimulated T cells only reduced the level of IFN-γ production to the combined level of LPS-stimulated DCs and Con A–stimulated T cells (column 2 plus column 10), indicating that sIL12-I does not affect T cells directly. Rather, it works via affecting the ability of DCs to produce the T cell stimulatory activity. This conclusion is supported by the level of IFN-γ production (column 13) where unstimulated but sIL12-I–treated DC supernatant did not inhibit the baseline IFN-γ production by Con A–stimulated T cells (column 10). Collectively, these data suggest that sIL12-I could suppress IL12 gene transcription and IL-12–induced IFN-γ production by activated T cells. Subsequently, by a combination of biochemical and proteomic approaches aided by mass spectrometric analysis, we identified a major component of sIL12-I from LPS-stimulated RAMOS and unstimulated 2E2 as GOLPH2 (Fig. 1F).

GOLPH2 is a type II Golgi transmembrane protein with multiple functions. Endosomal trafficking of the normally membrane-bound GOLPH2 leads to its secretion into the blood, making it a potential biomarker for several cancers, including HCC (36), prostate cancer (37), and others. GOLPH2 has also been studied in virus infections (38, 39). However, how GOLPH2 functions and the mechanisms of its regulation in normal and disease contexts were still largely unknown.

To determine whether GOLPH2 had the ability to inhibit IL-12 production, we treated BMDMs with rGOLPH2 plus IFN-γ and LPS and analyzed cytokine mRNA levels by RT-PCR. As shown in Fig. 2A, rGOLPH2 inhibited IFN-γ/LPS–stimulated mRNA levels of IL-12p35 and p40, but not those of TNFα or IL-10. Furthermore, in RAW264.7 cells transfected with the IL-12p35 promoter linked to a firefly luciferase reporter gene, rGOLPH2 inhibited IL-12p35 transcription in a dose-dependent manner (Fig. 2B). Similarly, results from RAW264.7 cells transfected with IL-12p35 or p40 promoter plus a GOLPH2-expressing vector or control also confirmed this conclusion (Fig. 2C). Critically, the use of a GOLPH2-blocking Ab unequivocally demonstrated the essentiality of the extracellular GOLPH2 in the inhibition of IL-12p35 transcription, as the Ab largely reversed the inhibition by sIL12-I in 2E2 supernatant (Fig. 2D). The specificity of the anti-GOLPH2 Ab was confirmed by Western blot (Supplemental Fig. 1B). After we observed the ability of GOLPH2 to inhibit IL-12 expression, we next tried to determine how GOLPH2 regulates IL-12p35 transcription by a strategy of sequentially deleting specific segments and elements in the IL-12p35 promoter, followed by transient transfection of these constructs into RAW264.7 cells. Unexpectedly, this approach revealed that GOLPH2’s IL12p35 transcription-inhibiting effect did not depend on the presence of the critical NF-κB element in the IL12p35 promoter (data not shown). Surprisingly, a previously identified promoter element named apoptotic cell (AC) response element (ACRE), that is, the sequence 5′-TGCCGCG-3′ located between +13 and +19 upstream of the IL-12p35 promoter’s transcription initiation site, appeared to mediate GOLPH2’s inhibition because its deletion of mutation resulted in a total loss of the inhibitory effect (Fig. 2E). As we previously reported, during phagocytosis of ACs by macrophages, a novel signaling pathway is activated via the externalized phosphatidylserine on the surface of ACs, resulting in tyrosine phosphorylation and activation of a novel nuclear zinc finger protein in the phagocyte, GC-BP, which binds directly to the IL-12p35 promoter at the ACRE, blocking the transcription (25). As shown in the Fig. 2F, rGOLPH2, similar to ACs, potently activated GC-BP by the hallmark tyrosine phosphorylation. To further validate GC-BP’s role of GOLPH2 in regulating IL12 transcription, we performed the ChIP assay in BMDMs. As shown in Fig. 2G, GC-BP binding activity was enhanced by rGOLPH2 treatment. Taken together, these results demonstrate that GOLPH2 inhibits IL12 transcription and this activity is mediated through the ACRE in the IL-12p35 promoter involving activation of GC-BP’s binding activity.

To further explore GOLPH2’s physical and pathological roles, we generated GOLPH2 KO mice via the TALEN Technology by targeting exon 4 of the gene (Fig. 3A, 3B). RT-PCR analysis confirmed the successful generation of homozygous null mice (Fig. 3C, 3D). GOLPH2 KO mice appeared to be normal in birth, body weight, and general behaviors, and they did not show obvious abnormalities in immune cell development (Fig. 3E). We concluded that GOLPH2 is not essential for steady-state physiological activities of the mouse. An extensive analysis of GOLPH2 mRNA levels in WT mice revealed that its expression at steady-state is most in tissues of the epithelial origin such as intestines, stomach, and colon (Supplemental Fig. 1A).

To further assess the role of GOLPH2 in the regulation of IL-12 production in primary cells and in mice, we analyzed cytokine expression in IFN-γ/LPS–stimulated BMDMs from WT and GOLPH2 KO mice by RT-PCR and ELISA. The results show that GOLPH2 KO BMDMs had higher levels of IL-12p35 and p40, but not IL-10 or TNF-α (Fig. 4A, 4B). Meanwhile, we also analyzed the expression of some well-known IFN-dependent genes. Cxcl10 and Ifnβ mRNA levels were downregulated in GOLPH2 KO BMDMs, but not Irf1 (Fig. 4C). To assess the role of GOLPH2 in vivo, we injected LPS into WT and GOLPH2 KO mice and found that serum IL-12 levels in KO mice but not those of IL-10 or TNF-α were elevated compared with the WT controls (Fig. 4D), consistent with the in vitro data. The ability of GOLPH2 to inhibit IL-12 expression indicates that GOLPH2 may be an important inflammation modulator. To test this notion, we compared the survival rates of adult GOLPH2 KO mice and WT controls in the LPS-induced septic shock model. Fig. 4E shows that GOLPH2 KO mice were significantly more susceptible to sepsis-induced death.

Accumulating evidence indicates that macrophage polarization plays a vital role in physiological and pathological processes. Macrophage subsets exist across an M1 and M2 spectrum in which M1 cells are defined as “classically activated” macrophages, arising in inflammatory settings primarily dominated by TLR and IFN signaling, whereas M2 cells are defined as “alternatively activated” macrophages, arising in anti-inflammatory settings dominated by a Th2 response (40). Once we identified GOLPH2 as a regulator of IL12 gene transcription in vitro and in vivo, we wondered whether GOLPH2 was also involved in macrophage polarization. To address this question, we first analyzed the expression of GOLPH2 in BMDMs, peritoneal macrophages (PMs), and RAW264.7 cells with M1 and M2 stimuli. As shown in Fig. 5A, the prototypical M1 stimulus LPS decreased Golph2 mRNA levels in BMDMs and RAW cells, whereas the M2 stimuli IL-4 and IL-13 increased Golph2 expression in all three types of macrophages. To further investigate the role of GOLPH2 in macrophage polarization, we performed both loss- and gain-of-function studies. For the loss-of-function model, we stimulated BMDMs and PMs derived from WT and GOLPH2 KO mice with IL-4 and observed decreased mRNA expression of the hallmarks of M2 macrophages, that is, arginase 1 (Arg1), inflammatory zone 1 (Fizz1), and Ym1 in GOLPH2-deficient cells (Fig. 5B, 5C). Conversely, overexpression of GOLPH2 in BMDMs enhanced IL-4–mediated induction of Arg1 and Ym1 mRNAs (Fig. 5D). Collectively, these data indicate that GOLPH2 plays a significant role in macrophage polarization.

To further explore the immune functions of GOLPH2 in DCs in a comprehensive way, we performed global gene expression analysis by RNA sequencing (RNA-seq) in activated IFN-γ/LPS–treated BMDCs from WT and GOLPH2 KO mice. The number of DEGs is shown in Supplemental Fig. 2A and the reliability of RNA-seq was corroborated by selective RT-PCR analyses of some DEGs containing IFN-stimulating genes such as Ifit2, Ifit3, and Ido1 (Supplemental Fig. 2B). As shown in the volcano plot in Fig. 6A, there were 160 DEGs identified in stimulated GOLPH2 KO cells, including 102 downregulated and 58 upregulated genes. IPA analysis of these DEGs (Fig. 6B) revealed that GOLPH2 may participate in mediating hepatic fibrosis/hepatic stellate cell activation, pattern recognition receptors in host defense against bacteria and viruses, DC maturation, and other important biological processes. Further in-depth analyses of these data combined with experimental validation will provide greater insights into GOLPH2’s role in regulating inflammation, macrophage polarization, and DC function.

In the present study, we identified GOLPH2 as a novel regulator of IL12 gene transcription. GOLPH2 deficiency increased IL-12 production in activated macrophages and aggravated LPS-induced sepsis in vivo. Mechanistically, extracellular GOLPH2 modulates IL-12 regulation through inducing GC-BP binding to the IL-12p35 promoter, suppressing the latter’s transcription. Furthermore, endogenous GOLPH2 also plays a significant role in macrophage polarization.

GOLPH2, a type II transmembrane protein located on the surface of the Golgi apparatus, has drawn considerable attention since its original discovery (41). These studies have shown that elevated expressions of GOLPH2 were strongly associated with hepatitis B virus and hepatitis C virus infections and several cancers such as HCC, prostate cancer, and renal cancer (36, 37, 42). Especially, several researches proved that GOLPH2 could be a valuable serum biomarker of HCC due to its better specificity and sensitivity than the most commonly used marker AFP (43). Recently, Ye et al. (44) showed that GOLPH2 promotes growth and metastasis of HCC cells via activating EGFR/RTK signaling. Another study (45) suggested that a cytokine QTL at the NAA35-GOLM1 locus markedly modulated IL-6 production in response to multiple pathogens and was associated with susceptibility to candidemia. These studies indicate that GOLPH2 may have multiple functions in cancers and inflammation. In our study, we found that the secreted GOLPH2 by malignant B cells inhibited IL-12 production from DCs and decreased IFN-γ production by activated T cells indirectly via modulating the activity of DCs. Our data in the present study are consistent with the previous report that GOLPH2 overexpression inhibited IL-12 p35 transcription and IFN-γ synthesis in human gastric cancer cells (46). Besides, we showed that GOLPH2’s ability to suppress IL-12 production is independent of TGF-β, IL-10, TNF-α, and PGE₂, all well-known inhibitors of IL-12 synthesis (data not shown). Further mechanistic studies showed that GOLPH2 suppressed IL-12 production at least partly through ACRE, the proximal element in the IL-12p35 promoter that was first described by our group (25) and known to involve the activation of GC-BP, a novel nuclear zinc finger protein blocking IL-12 p35 transcription. However, because GOLPH2 is not a transcription factor, further details of this transcriptional regulation remain to be elucidated. We analyzed Golph2 mRNA levels in various mouse organs/tissues and immune cells, including DCs and macrophages. The result shows that GOLPH2’s expression at steady-state is most abundant in tissues of the epithelial origin such as colon, stomach, lung, and prostate and very little in DCs and macrophages (Supplemental Fig. 1). We also examined RAW264.7 cells (mouse macrophages) and did not find detectable levels of secreted GOLPH2 protein (data not shown). Additionally, supernatant from Jurkat (a T cell line) does not suppress IL-12p35 transcription at all in stark contrast with that of B cells (Fig. 1B). Thus, it is unlikely that DC- and macrophage- or T cell–derived extracellular GOLPH2 plays a significant autocrine or paracrine role in IL-12 regulation.

Wright et al. (47) reported that mice expressing a C-terminally truncated GOLPH2 transgene had a significantly reduced survival compared with WT controls, particularly in females, and they developed varying degrees of renal and hepatic diseases. However, the GOLPH2-null mice we generated displayed no obvious abnormalities. Lack of normal GOLPH2 function or aberrant function of a truncated GOLPH2 may underline the mortality of these mice. Nevertheless, the precise reasons to account for this inconsistency need to be further explored.

GOLPH2 KO mice were hypersensitive to LPS shock, and GOLPH2-deficient macrophages produced more M1-type inflammatory mediator IL-12 after LPS activation in vitro, implying that GOLPH2 may be a repressor of M1 inflammatory response. Conversely, reduced expression levels of M2 genes in GOLPH2-deficent macrophages suggest that GOLPH2 is a positive regulator of M2 polarization. These results indicate that GOLPH2 plays a significant role in macrophage polarization. However, how GOLPH2 regulates this process is still unclear.

DCs comprise a heterogeneous population of professional APCs and are one of the major immune cells that produce IL-12. In this study, we performed RNA-seq of WT and GOLPH2 KO BMDCs to investigate the immune functions of GOLPH2 comprehensively and systemically. IPA analysis of DEGs in LPS-activated BMDCs suggests that the functions of GOLPH2 pertain to those of macrophages, fibroblasts, and endothelial cells in rheumatoid arthritis, hepatic fibrosis, and cancer signaling in resting states, whereas it participates in hepatic fibrosis and recognition of bacteria and viruses in activated states. The networks revealed in the RNA-seq analysis provide further insights into the mechanisms underlying GOLPH2-associated immunological processes and associated disorders. For example, we found that IFN-inducing antivirus proteins such as Ifit2, Ifit3, Mx2, and ApoL9b were downregulated in activated GOLPH2 KO BMDCs, which indicates that GOLPH2 may play a positive role in host resistance to viral infections. Abnormal expression of collagen proteins could lead to collagen diseases, which are a heterogeneous group of systemic inflammatory diseases of autoimmune origin that affect a wide range of organs and systems, including systemic lupus erythematosus and rheumatoid arthritis (48, 49). We found that col1a2, col3a1, col5a2, and col8a1 were upregulated in resting GOLPH2 KO cells, suggesting the potential involvement of GOLPH2 in the pathogenesis of these autoimmune diseases (Supplemental Fig. 3).

In conclusion, we report the biochemical identification of GOLPH2 from B cell lymphoma culture supernatant and show that the secreted protein could inhibit IL-12 production by DCs and IL-12–induced IFN-γ production by activated T cells. GOLPH2’s IL-12–inhibiting activity is partly mediated through the ACRE involving GC-BP phosphorylation. Furthermore, GOLPH2 is a regulator of macrophage polarization toward the M2 type. The comprehensive analysis of gene expression profiles of activated WT and GOLPH2-deficient DCs by RNA-seq reveals mechanistic insights into the way GOLPH2 may mediate a number of important physiological and pathological conditions. Our study of extracellular GOLPH2 helps advance the scientific understanding of the biological roles of this novel and intriguing molecule with great potential as a diagnostic and prognostic marker as well as a therapeutic target in several cancers, particularly of viral origins.

The authors have no financial conflicts of interest.