Mtor- and SGK-Mediated Connexin 43 Expression Participates in Lipopolysaccharide-Stimulated Macrophage Migration Through the Inos/Src/FAK Axis

The Journal of Immunology mTOR- and SGK-Mediated Connexin 43 Expression Participates in Lipopolysaccharide-Stimulated Macrophage Migration through the iNOS/Src/FAK Axis

Chen Shen,* Jin Hong Chen,* Youngyi Lee,† Md. Mehedi Hassan,‡ Su Jin Kim,* Eun Young Choi,x Seong-Tshool Hong,‡ Byung-Hyun Park,† and Ji Hyun Park*

Connexin 43 (Cx43) deﬁciency was found to increase mortality in a mouse model of bacterial peritonitis, and Cx43 is upregulated in macrophages by LPS treatment. In this study, we characterized a novel signaling pathway for LPS-induced Cx43 expression in RAW264.7 cells and thioglycolate-elicited peritoneal macrophages (TGEMs). LPS alone or LPS-containing conditioned medium (CM) upregulated Cx43. Overexpression or silencing of Cx43 led to the enhancement or inhibition, respectively, of CM-induced TGEM migration. This response involved the inducible NO synthase (iNOS)/focal adhesion kinase (FAK)/Src pathways. Moreover, CM-induced migration was compromised in TGEMs from Cx43+/2 mice compared with TGEMs from Cx43+/+ littermates. Cx43 was upregulated by a serum/glucocorticoid-regulated kinase 1 (SGK) activator and downregulated, along with inhibition of CM- induced TGEM migration, by knockdown of the SGK gene or blockade of the SGK pathway. LPS-induced SGK activation was abrogated by Torin2, whereas LPS-induced Cx43 was downregulated by both Torin2 and rapamycin. Analysis of the effects of FK506 and methylprednisolone, common immunosuppressive agents following organ transplantation, suggested a link between these immunosuppressive drugs and impaired macrophage migration via the Cx43/iNOS/Src/FAK pathway. In a model of Escherichia coli infectious peritonitis, GSK650349-, an SGK inhibitor, or Torin2-treated mice showed less accumulation of F4/80+CD11b+ macrophages in the peritoneal cavity, with a delay in the elimination of bacteria. Furthermore, following pretreatment with Gap19, a selective Cx43 hemichannel blocker, the survival of model mice was signiﬁcantly reduced. Taken together, our study suggested that Cx43 in macrophages was associated with macrophage migration, an important immune process in host defense to infection. The Journal of Immunology, 2018, 201: 2986–2997.

acrophages, which are present in almost all tissues, for TLR4 (2). Upon binding TLR4, LPS triggers common constitute an essential component of the innate im- downstream signaling pathways, most notably PI3K/protein M mune system and form the first line of defense against kinase B (Akt) and NF-kB, leading to upregulation of proin- pathogens. Under physiological conditions, macrophage recruit- flammatory mediators, such as cytokines and NO (3). ment to sites of infection is an important immune process in host Cellular Src is the prototype of Src family kinases (SFKs) of defenses (1). highly conserved proteins, including Blk, Fgr, Fyn, Hck, Lck, Lyn, LPS, a major constituent of the Gram-negative bacterial endo- and Yes (4). In resting macrophages, Src is barely detectable; toxin, is also an important pathogen-associated molecular pattern however, LPS treatment leads to upregulation of Src, resulting in activation of focal adhesion kinase (FAK) and cell motility (5). *Division of Endocrinology and Metabolism, Department of Internal Medicine, FAK, a Src substrate, plays a critical role in macrophage adhesion Chonbuk National University Medical School, Research Institute of Clinical Medi- and motility (6). Accumulating evidence has indicated that LPS- cine of Chonbuk National University–Biomedical Research Institute of Chonbuk elicited macrophage motility requires the participation of Src and National University Hospital, Jeonju 54907, Republic of Korea; †Department of Biochemistry, Chonbuk National University Medical School, Jeonju 54907, Republic FAK and is dependent on inducible NO synthase (iNOS) (7). of Korea; ‡Department of Biomedical Sciences and Institute for Medical Science, Hence, targeting the iNOS/Src/FAK axis may directly reflect Chonbuk National University Medical School, Jeonju 54907, Republic of Korea; and xDepartment of Biomedical Sciences, University of Ulsan College of Medicine, Seoul LPS-triggered macrophage motility. 05505, Republic of Korea Gap junctions (GJs) are intercellular channels that allow ORCIDs: 0000-0001-7026-3461 (C.S.); 0000-0002-0417-6236 (J.H.C.); 0000-0003- communication between contacting cells by mediating recipro- 3768-4285 (B.-H.P.). cal exchange of ions and small molecules (8). Connexin 43 Received for publication July 3, 2017. Accepted for publication September 12, 2018. (Cx43), the most ubiquitous GJ protein subunit, interacts with This work was supported by research funds from the Research Institute of Clinical many proteins and has channel-independent functions (9). Medicine of Chonbuk National University–Biomedical Research Institute of Chonbuk Moreover, Cx43 is expressed by almost all immune cells and can National University Hospital. be upregulated when the immune cells become exposed to in- Address correspondence and reprint requests to Dr. Ji Hyun Park, Division of Endo- crinology and Metabolism, Department of Internal Medicine, Chonbuk National flammatory factors (10). However, the role of Cx43 in signaling- University Medical School, 20 Geonji-Ro, Deokjin-Gu, Jeonju 54907, Republic of mediated immune regulation is poorly defined. Importantly, Korea. E-mail address: [email protected] several recent studies have reported the correlation between Abbreviations used in this article: bFGF, basic fibroblast growth factor; CM, con- Cx43 and invasion or migration in cancer and embryonic stem ditioned medium; Cx43, connexin 43; FAK, focal adhesion kinase; GJ, gap junction; iNOS, inducible NO synthase; MCP-1, monocyte chemotactic protein-1; cells (11, 12). mTOR, mammalian target of rapamycin; mTORC1, mTOR complex 1; SGK, Therefore, in this study, we investigated the impact of Cx43 serum/glucocorticoid-regulated kinase 1; si, small interfering; si-Scr, scrambled on LPS-induced macrophage migration to provide a theoreti- siRNA; TGEM, thioglycolate-elicited peritoneal macrophage. cal basis for Cx43 gene therapy in the treatment of infection or Copyright Ó 2018 by The American Association of Immunologists, Inc. 0022-1767/18/$37.50 immunosuppression. www.jimmunol.org/cgi/doi/10.4049/jimmunol.1700954 The Journal of Immunology 2987

FIGURE 1. Effect of CM on Cx43 expression in RAW264.7 cells and TGEMs. RAW264.7 cells (A and B)andTGEMs(C and D)weretreatedwith CM for 6 and 24 h. Protein levels of Cx43 from treated cell lysates were analyzed by Western blotting and normalized against GAPDH (n =4).Cx43 mRNA expression was measured by real-time PCR and normalized to 36B4 expression (n =5).(E) The concentration of TNF-a,IL-1b, IL-6, and MCP-1 from CM was analyzed by ELISA kit. (F) TGEMs were treated with 10 ng/ml TNF-a,IL-1b, IL-6, and MCP-1 for 24 h. Protein levels of Cx43 and p-serum/ glucocorticoid-regulated kinase 1 (SGK) from treated cell lysates were analyzed by Western blotting and normalized against GAPDH and SGK, respectively. Values are presented as means 6 SEMs of three or four (indicated with n =4)orﬁve(indicatedwithn = 5) independent experiments. *p , 0.05, **p , 0.01 versus vehicle by one-way ANOVA (A–D) or Student t test.

Materials and Methods methylprednisolone, and GSK650394 were obtained from Cayman a Animals Chemical (Ann Arbor, MI). Recombinant mouse TNF- monocyte chemotactic protein-1 (MCP-1), IL-1b, and IL-6 were purchased from Seven-week-old male C57BL/6 mice were purchased from Samtako ProSpec (Ness-Ziona, Israel). Gap19 was obtained from Apexbio Tech- Bio (Osan, Korea). Wild-type mice and Cx43-knockout mice were nology (Houston, TX). obtained from mating of mice heterozygous for the mutant allele (Gja1+/Gja1–), which were purchased from The Jackson Laboratory Cell culture (Bar Harbor, ME). Because homozygous Cx43-knockout mice Murine RAW 264.7 macrophages were purchased from American Type (Cx432/2) are embryonic, lethal, heterozygous mice (Cx43+/2)and +/+ Culture Collection (Manassas, VA). Cells were thawed and grown in age-matched wild-type littermates (Cx43 ) older than 8 wk were phenol-red DMEM (Lonza, Walkersville, MD). All cell cultures were used for experiments. The animals were housed under pathogen-free supplemented with 10% FBS (GenDEPOT, Barker, TX) and kept at 37˚C conditions and fed a standard laboratory normal chow diet ad libi- in a humidiﬁed 5% CO incubator. For treatment with FBS (Fig. 6A), tum. There were no signiﬁcant differences in average body weight 2 +/+ +/2 cells were starved for 2 h and treated with indicated concentration of between Cx43 littermates and Cx43 mice (26.2 versus 25.8 g, the serum. respectively; p = 0.36). All experimental procedures were approved by the Institutional Animal Care and Use Committee (approval no. Isolation of primary macrophages CBNU-2016-0080). Thioglycolate-elicited peritoneal macrophages (TGEMs) were obtained Materials from 8-wk-old C57BL/6 mice 3 d after i.p. injection of Brewer thio- glycollate (3 ml of 3% p/v; BD Life Sciences, Franklin Lakes, NJ). Routine chemicals, laboratory ware, and reagents were obtained from Peritoneal exudates were washed with PBS and plated in 10% FBS- Invitrogen Life Technologies (Carlsbad, CA) or Sigma-Aldrich (St. Louis, containing RPMI 1640 medium (Life Technologies, Grand Island, NY) MO). Anti–p-serum/glucocorticoid-regulated kinase 1 (SGK), anti-SGK, for 2 h. Exudates were then washed with medium to remove nonadherent and anti-iNOS Abs were purchased from Santa Cruz Biotechnology cells. The resulting macrophage monolayer was 98% pure and ready for (Dallas, TX). Anti-Cx43, anti–p-mammalian target of rapamycin experimentation. (mTOR), anti-mTOR, anti–p-AKT, anti-AKT, anti-Src, anti–p-FAK, and anti-FAK were obtained from Cell Signaling Technology (Danvers, MA), In vitro Transwell migration assay anti-HSP90 Ab was from Enzo Life Sciences (Farmingdale, NY), and rabbit anti-GAPDH Ab was from Bioworld (Dublin, OH). 1400W TGEMs were treated with 10 ng/ml LPS overnight, and the conditioned and FK506 were obtained from Enzo Life Sciences. Rapamycin, medium (CM) was used as a chemoattractant. Migration assays were 2988 mTOR/SGK REGULATE MACROPHAGE MIGRATION THROUGH Cx43

FIGURE 2. Effect of LPS on Cx43 expression in RAW264.7 cells. RAW264.7 cells were treated with LPS at the indicated concentrations for 8 h (A)or with 10 ng/ml LPS for the indicated times (B). Cx43 mRNA expression was measured by real-time PCR and normalized to 36B4 expression. RAW264.7 cells were treated with LPS at the indicated concentrations for 8 h (C) or with 10 ng/ml LPS for the indicated times (D). Levels of Cx43, p-AKT, p-mTOR, and p-SGK from treated cell lysates were analyzed by Western blotting and normalized to GAPDH, AKT, mTOR, and SGK expression, respectively. Values are presented as means 6 SEMs of three independent experiments. *p , 0.05, **p , 0.01 versus vehicle by one-way ANOVA. performed using Transwell migration assay chambers (BD Life Sciences) was performed in a 7900HT Fast Real-Time PCR System (Applied Bio- by adding CM to the lower chamber. systems, Carlsbad, CA), using SYBR Green Realtime PCR Master Mix (TOYOBO, Osaka, Japan). To confirm the use of equal amounts of RNA in Measurement of NO each reaction, all samples were checked in parallel for 36B4 RNA expression. Primers were as follows: 59-GGCGTGAGGAAAGTACCAAA-39 Biologically produced NO is rapidly oxidized to nitrite and nitrate in 9 9 aqueous solution. Thus, the production of NO was assessed by mea- (forward) and 5 -AGCTTTGAGCAGGAATCCAA-3 (reverse) for Cx43; 59-TCACTGTGCCAGCTCAGAAC-39 (forward) and 59-AATTTCAAT- suring the nitrite concentrations of cell culture supernatants via a 9 colorimetric assay. Briefly, 100-ml aliquots of the culture supernatant GGTGCCTCTGG-3 (reverse) for 36B4 RNA. were incubated at room temperature for 5 min with 100 ml modified Griess reagent with a 1:1 mixture of 1% sulfanilamide in 30% acetic Western blotting acid and 0.1% N-(1-naphthyl)ethylenediamine dihydrochloride in 60% Whole cell lysates were prepared in RIPA lysis buffer (ELPIS-Biotech, acetic acid, and the absorbance was measured at 540 nm. The NO Taejeon, Republic of Korea). The protein concentrations were deter- concentration was determined using a linear standard curve of serial mined using bicinchoninic acid solution (Sigma-Aldrich). The proteins dilutions of sodium nitrite in a working medium. were electrotransferred to polyvinylidene difluoride membranes (Merck RNA isolation and real-time RT-PCR Millipore, Darmstadt, Germany). The membranes were incubated with HRP-conjugated anti-rabbit or anti-mouse IgG Abs (Enzo Life Sciences). Total RNA was extracted using TRIzol (Favorgen Biotech, Ping Tung, Signals were visualized with Chemiluminescent HRP Substrate (EMD Taiwan) and treated with RNase-free DNase (Invitrogen, Carlsbad, CA). Millipore, Billerica, MA). The band images were obtained by Fusion FX7 cDNA was generated using a QuantiTect Reverse Transcription kit (Peqlab, Torcy, France). All signals were analyzed by Image J (National (Invitrogen). Quantitative real-time RT-PCR of mouse Cx43 and 36B4 RNA Institutes of Health, Bethesda, MD). The Journal of Immunology 2989

FIGURE 3. Effect of LPS on Cx43 expression in TGEMs. TGEMs were treated with LPS at the indicated concentrations for 8 h (A) or with 10 ng/ml LPS for the indicated times (B). Cx43 mRNA expression was measured by real-time PCR and normalized to 36B4 expression. TGEMs were treated with LPS at the indicated concentrations for 8 h (C) or with 10 ng/ml LPS for the indicated times (D). Levels of Cx43, p-AKT, p-mTOR, and p-SGK from treated cell lysates were analyzed by Western blotting and normalized to GAPDH, AKT, mTOR, and SGK expression, respectively. Values are presented as means 6 SEMs of three independent experiments. *p , 0.05, **p , 0.01 versus vehicle by one-way ANOVA.

Transfection with Cx43 and SGK small interfering RNA Mouse peritonitis model TGEMs were transfected with Cx43 small interfering (si) RNA, SGK Eight-week-old C57BL/6 male mice were divided randomly and pretreated siRNA (si-Cx43 and si-SGK, respectively), or scrambled siRNA (si-Scr; with an i.p. injection of 300 ml DMSO/PBS (10%), 4.5 mg/kg GSK650394, Santa Cruz Biotechnology) using Lipofectamine RNAiMAX Reagent 5.1 mg/kg Torin2, or 1.45 mg/kg Gap19. Eight hours later, mice were 8 (Invitrogen) according to the manufacturer’s protocol. Six hours post- challenged with 1 3 10 CFU Escherichia coli through i.p. injection. A transfection, cells were cultured in complete RPMI 1640 medium for an common strain of E. coli (JBE8-00101) was obtained from the Gut additional 48 h before further use. Microbiota Bank (www.gutmicrobiotabank.com). After 12 h, 4 ml PBS was injected into the peritoneal cavity and, after a slight massage, collected Preparation of recombinant adenovirus as the peritoneal lavage ﬂuid. E. coli CFU was calculated from the Luria– Bertani agar plates after overnight culture with 10-fold dilutions. The As described previously (13), adenovirus infection of TGEMs was per- percentage of macrophages in the lavage was quantiﬁed by FACS (BD formed using a multiplicity of infection of 200. Twelve hours postinfec- Biosciences, San Jose, CA) with allophycocyanin-conjugated CD11b and tion, cells were cultured in complete RPMI 1640 medium for an additional FITC-conjugated F4/80 Ab (Invitrogen, San Diego, CA). The PBS and 48 h before further analysis. Gap19 group was pretreated as described above and injected with 4 3 108 2990 mTOR/SGK REGULATE MACROPHAGE MIGRATION THROUGH Cx43

FIGURE 4. Cx43 participates in CM-induced macrophage migration. (A) Representative microphotographs of the migration assay after staining cells with crystal violet (original magniﬁcation 3100). TGEMs that were infected with Ad-lacZ or Ad-Cx43 (multiplicity of infection = 200) for 48 h were transferred to Transwell inserts and incubated for an additional 6 h for migration in the presence of RPMI 1640 or CM in the lower well. Cells that migrated to the bottom of the inserts were counted and normalized to Ad-lacZ vehicle. (B) Protein levels of Cx43, GAPDH, and HSP90 from adenovirus-infected TGEM lysates were analyzed by Western blotting. (C) Representative microphotographs of migration assays after staining cells with crystal violet (original magniﬁcation 3100). TGEMs that were transfected with 20 nM si-Scr or si-Cx43 for 48 h were transferred to Transwell inserts and incubated for an additional 6 h for migration in the presence of RPMI 1640 or CM in the lower well. Cells that migrated to the bottom of the inserts were counted and normalized to si-Scr vehicle (n = 4). (D) Protein levels of Cx43, GAPDH, and HSP90 from siRNA-transfected TGEM lysates were analyzed by Western blotting. (E) Representative microphotographs of migration assays after staining cells with crystal violet (Figure legend continues) The Journal of Immunology 2991

FIGURE 5. iNOS participated in CM-induced macrophage migration. (A) TGEMs were pretreated (or left untreated) with 100 mM 1400W for 1 h and then stimulated (or left unstimulated) with CM for 24 h. Protein levels from the treated cell lysates were analyzed by Western blotting. Src and iNOS were normalized to GAPDH expression, and p-FAK was normalized to FAK expression. (B) NO production was measured in the culture supernatant of TGEMs that were treated as described in (A). (C) Representative microphotographs of the migration assay after staining cells with crystal violet (original mag- niﬁcation 3100). TGEMs that were pretreated without or with 100 mM 1400 W for 1 h were transferred to Transwell inserts and incubated for an additional 6 h for migration in the presence of RPMI 1640 or CM in the lower well. Cells that migrated to the bottom of the inserts were counted and normalized to the vehicle. Values are presented as means 6 SEMs of three independent experiments. *p , 0.05, **p , 0.01 versus CM-stimulated vehicle by Student t test.

CFU E. coli after 8 h. The mice were checked every 4 h to monitor their treatment with CM for 24 h resulted in a significant increase in survival over the next 48 h. Cx43 mRNA and protein levels. Similar effects were observed in ELISA TGEMs (Fig. 1C, 1D). Besides, the ELISA results showed an increase of TNF-a, IL-1b, IL-6, and MCP-1 in the CM (Fig. 1E), For ELISA, TGEMs were treated with or without 10 ng/ml LPS overnight, and and a 24 h-treatment of these cytokines and chemokines upregu- the supernatant was collected. The concentration of TNF-a,IL-1b, IL-6, and MCP-1 in the supernatant was measured using ELISA kits according to the lated Cx43 expression with increased p-SGK in TGEMs, except manufacturer’s instructions (eBioscience, San Diego, CA). for IL-6 (Fig. 1F). Data analysis Effect of LPS on Cx43 expression in RAW264.7 cells Experimental values were expressed as a percentage of the control. Data are LPS is a well-known cell-wall component of most Gram-negative presented as mean 6 SEMs (n = 3–6). Statistical significance was evalu- bacteria that also acts as a potent initiator of inflammation (13). ated with the Student t test or one-way ANOVA, followed by Tukey test. Treatment with LPS resulted in a significant increase in Cx43 For survival curve analysis, the log-rank test was performed. Differences with p values ,0.05 were considered significant. mRNA in a concentration- and time-dependent manner (Fig. 2A, 2B). Consistent with the results of Cx43 mRNA expression, the Results results of Western blotting showed that Cx43 protein expression was increased by LPS (Fig. 2C, 2D). Moreover, the phosphory- Effect of CM on Cx43 expression in RAW264.7 macrophages lation of AKT, mTOR, and SGK was also increased. and TGEMs RAW264.7 macrophages and TGEMs were treated with 10 ng/ml Effect of LPS on Cx43 expression in TGEMs LPS overnight, and the CM was used to investigate the effects on To further confirm whether LPS treatment increased Cx43 ex- Cx43 expression. As shown in Fig. 1A and 1B, in RAW264.7 cells, pression in primary macrophages, we isolated TGEMs from

(original magniﬁcation 3100). TGEMs that were isolated from Cx43+/+ and Cx43+/2 mice were transferred to Transwell inserts. Migrated cells were counted and normalized to the Cx43+/+ vehicle (n = 4). (F) TGEMs that were transfected with si-Scr/si-Cx43 or (G) isolated from Cx43+/+ and Cx43+/2 mice were stimulated without or with CM for 24 h. NO production was measured in the culture supernatant. (H) Protein levels of cells treated as described in (F) were analyzed by Western blotting. (I) Protein levels of cells treated as described in (G) were analyzed by Western blotting. Cx43, Src, and iNOS were normalized to GAPDH expression, and p-FAK was normalized to FAK expression. Values are presented as means 6 SEMs of three or four (indicated with n = 4) independent experiments. *p , 0.05, **p , 0.01 versus CM-stimulated Ad-lacZ or si-Scr or Cx43+/2 mice by one-way ANOVA (H) or Student t test. 2992 mTOR/SGK REGULATE MACROPHAGE MIGRATION THROUGH Cx43

FIGURE 6. Effects of the SGK pathway on Cx43 expression and CM-induced macrophage migration. RAW264.7 cells and TGEMs were treated with FBS before being starved for 2 h (A)orbFGF(B) at indicated concentrations for 24 h. Levels of Cx43, p-SGK, and SGK from treated cell lysates were analyzed by Western blotting and normalized to GAPDH expression. (C) TGEMs were transfected with 20 nM si-Scr/si-SGK for 48 h and stimulated with 10 ng/ml LPS for 24 h. Levels of Cx43, p-SGK, and SGK from treated cell lysates were analyzed by Western blotting and normalized to GAPDH expression (n =5).(D) Representative microphotographs of the migration assay after staining cells with crystal violet (original magniﬁcation 3100). TGEMs that were transfected with 20 nM si-Scr/SGK siRNA for 48 h were transferred to the Transwell insert and incubated for an additional 6 h for migration in the presence of RPMI 1640 or CM in the lower well. Cells that migrated to the bottom of the inserts were counted and normalized against si-Scr vehicle. (E) TGEMs that were pretreated without or with GSK650349 at the indicated concentrations (Figure legend continues) The Journal of Immunology 2993

C57BL/6 mice and repeated the previous experiments. As shown in Regarding total SGK expressions, whereas LPS induced statistically Fig. 3, both mRNA (Fig. 3A, 3B) and protein (Fig. 3C, 3D) ex- significant expression of total SGK protein (p = 0.04) in Fig. 6C, the pression of Cx43 were increased by LPS in a concentration- and same was not observed in Figs. 2C, 2D, 3C, 3D, 6E, or 7A–C. Our time-dependent manner. These data suggested the potential role of explanation for this inconsistency is that the stressful cellular en- Cx43 in LPS-mediated immune responses in macrophages. vironment due to siRNA transfection for 48 h might result in dif- ferential response in the expression of SGK to LPS in TGEMs in Cx43 participated in CM-induced macrophage migration Fig. 6C (16). To explore the role of Cx43 in LPS-containing CM-induced im- Effects of mTOR inhibitors and methylprednisolone on Cx43 mune responses, we measured whether Cx43 expression affected expression in macrophages macrophage migration. TGEMs were transfected with Cx43 adenovirus for 48 h and layered onto a Transwell insert for migration in Considering the involvement of the mTOR pathway in Cx43 ex- the presence of CM in the bottom well. The results showed that pression by LPS or CM in macrophages, TGEMs were treated overexpression of Cx43 resulted in a significant enhancement in with the mTOR inhibitors rapamycin and Torin2, with or without CM-induced TGEM migration (Fig. 4A). Moreover, inhibition of LPS (Fig. 7A). Torin2, an ATP-competitive mTOR inhibitor that migration was found in si-Cx43–transfected TGEMs and TGEMs potently targets mTOR complex 1 (mTORC1) and mTORC2 (17), that were isolated from Cx43+/2 mice (Fig. 4C, 4E). The over- significantly reduced LPS-induced Cx43 expression with de- expression and knockdown of Cx43 was confirmed by Western creased SGK phosphorylation. Additionally, treatment with rapa- blotting (Fig. 4B, 4D). Additionally, Cx43 deficiency led to a mycin, an mTORC1 inhibitor (18), also reduced Cx43 expression significant reduction in NO production in CM-stimulated TGEMs without significantly altering p-SGK levels. Moreover, immuno- (Fig. 4F, 4G). To elucidate the molecular mechanisms of the in- suppressive drugs, such as FK506 (tacrolimus) and methylpred- hibitory action of Cx43 deficiency against macrophage migration, nisolone, significantly suppressed LPS-induced Cx43 and we further observed that the increases in Src and p-FAK/FAK by phospho-SGK levels in TGEMs. CM were prevented in either si-Cx43–transfected TGEMs or +/2 Effect of mTOR/SGK pathway and Cx43 hemichannel on TGEMs that were isolated from Cx43 mice (Fig. 4H, 4I). macrophage recruitment and host defense against invading iNOS participated in CM-induced macrophage migration bacteria in peritonitis mouse model Because the iNOS pathway plays a role in macrophage migration To further confirm the potential role of mTOR/SGK pathway in (7) and was found to be regulated by Cx43 expression under our host defense against invading bacteria, we set up a mouse peri- experimental conditions, we hypothesized that the iNOS pathway tonitis model by i.p. injection of E. coli to monitor macrophage may be involved in Cx43-mediated CM-induced TGEM migra- migration in vivo and track the clearance of bacteria. As shown in + + tion. As shown in Fig. 5A and 5B, 1400W, an iNOS inhibitor, Fig. 8A, the accumulation of F4/80 CD11b macrophages from suppressed the p-FAK/FAK, Src, and iNOS, resulting in reduced the peritoneal cavity was significantly less in GSK650349- and NO production in CM-induced TGEMs. Moreover, pretreatment Torin2-treated mice than in the control group. Additionally, the with 1400W resulted in a significant reduction in CM-induced total quantity of bacteria in the peritoneal cavity was increased TGEM migration (Fig. 5C). in GSK650349/Torin2-injected mice (Fig. 8B). In contrast, decreased macrophage percentage and enhanced residual bacteria Effect of the SGK pathway on Cx43 expression and from the abdominal cavity were also observed in the mice pre- CM-induced macrophage migration treated with Gap19, a selective Cx43 hemichannel blocker (19) To investigate the upstream molecules involved in LPS-induced (Fig. 8C, 8D). To further confirm the key role of Cx43 hemichannel Cx43 expression, we examined the effect of the SGK activator in antibacterial immune response, we infected the mice with E. coli FBS (14) and basic fibroblast growth factor (bFGF) (15) on Cx43 to monitor mouse survival. As shown in Figs. 8E, the survival of expression in RAW264.7 cells and TGEMs. As shown in Fig. 6A Gap19-pretreated mice was obviously decreased. These data fur- and 6B, the treatment of FBS and bFGF increased the levels of ther confirmed the protective role of Cx43 in bacterial infection. phospho- and total SGK and upregulated Cx43 protein expression. In contrast, knockdown of the SGK gene by SGK siRNA de- Discussion creased the level of Cx43 in LPS-stimulated TGEMs (Fig. 6C). In this study, we found that the expression of Cx43 was upregulated Additionally, upregulation of Cx43 by LPS was abrogated by by LPS-containing CM or LPS alone in macrophages and played an pretreatment with GSK650394, an SGK inhibitor, in TGEMs important role in CM-promoted macrophage migration. Moreover, (Fig. 6E). As shown in Fig. 6C, SGK siRNA suppressed Cx43 we observed that the iNOS/Src/FAK axis was related to Cx43- expression in resting and LPS-stimulated cells, although the re- mediated macrophage migration. We also found that the mTOR/ sponse was more prominent under LPS-stimulated conditions. SGK signaling pathway was involved in the regulation of Cx43 Further studies are needed to confirm this finding. expression, Cx43-mediated CM-induced TGEM migration, and Furthermore, silencing of SGK by siRNA or blockade of SGK F4/80+CD11b+ macrophage accumulation in the peritoneal cavity activity by GSK650394 inhibited the CM-induced migration of in an E. coli infectious peritonitis model; these effects may be TGEMs (Fig. 6D, 6F). These results suggested that the SGK associated with increased residual bacteria in the abdominal cavity pathway regulated Cx43 expression in CM-stimulated TGEMs. and reduced survival in our acute peritonitis mouse model.

for 1 h were stimulated with 10 ng/ml LPS for 24 h. Levels of Cx43 and p-SGK from treated cell lysates were analyzed by Western blotting and normalized to GAPDH and SGK expression, respectively. (F) Representative microphotographs of the migration assay after staining cells with crystal violet (original magniﬁcation 3100). TGEMs that were pretreated without or with 100 nM GSK650349 for 1 h were transferred to Transwell inserts and incubated for an additional 6 h for migration in the presence of RPMI 1640 or CM in the lower well. Cells that migrated to the bottom of the inserts were counted and normalized to the vehicle. Values are presented as means 6 SEMs of three or ﬁve (indicated with n = 5) independent experiments. *p , 0.05, **p , 0.01 versus vehicle (A and B), si-Scr (C), or CM-stimulated vehicle (D–F) by one-way ANOVA (A, B, and E) or Student t test. 2994 mTOR/SGK REGULATE MACROPHAGE MIGRATION THROUGH Cx43

FIGURE 7. Effects of mTOR inhibitors and methylprednisolone/FK506 on Cx43 expression in macrophages. (A), TGEMs that were pretreated without or with 100 nM rapamycin/Torin2 for 1 h were stimulated with 10 ng/ml LPS for 24 h. Levels of Cx43, p-mTOR, and p-SGK from treated cell lysates were analyzed by Western blotting and normalized to GAPDH, mTOR, and SGK expression, respectively. (B), TGEMs that were pretreated without or with 100 nM and 1 mM methylprednisolone or (C)1mM FK506 for 1 h were stimulated with 10 ng/ml LPS for 24 h. Levels of Cx43, p-mTOR, and p-SGK from treated cell lysates were analyzed by Western blotting and normalized to GAPDH, mTOR, and SGK expression, respectively. Values are presented as means 6 SEMs of three independent experiments. *p , 0.05, **p , 0.01 versus LPS-stimulated vehicle by one-way ANOVA (B) or Student t test.

Recently, Qin et al. (20) found that Cx43 is the most abundant those in classic intercellular communication in the innate immune connexin in RAW 264.7 cells and that its expression is increased system, remain unknown. by LPS, consistent with our findings. Additionally, growing evi- In the current study, we found that Cx43 expression affected dence has indicated that connexins are involved in inflammatory iNOS expression and NO production in CM-induced macrophages. signaling, such as adhesion molecule expression and leukocyte Previous studies also support a link between iNOS-derived NO and transmigration (21, 22). Moreover, Anand et al. (23) found a novel Cx43 downregulation in heart failure (25), and Li and colleagues role for Cx43 in the regulation of phagocytosis. In contrast, (26) established a reciprocal regulation loop between iNOS and findings published by Glass et al. (24) suggested that modulation GJs in bladder smooth muscle cells. Taken together, these findings of Cx43 levels in cultured macrophages does not have a significant suggested that Cx43 had regulatory effects on iNOS, although the impact on phagocytosis. However, the roles of Cx43, other than specific signaling mechanisms are still unclear. The Journal of Immunology 2995

FIGURE 8. Effects of mTOR/SGK pathway and Cx43 hemichannel on macrophage recruitment and host defense against invading bacteria in peritonitis mouse model. Mice received an i.p. injection of DMSO/PBS (10%, 300 ml, n = 3), GSK650349 (4.5 mg/kg, n =3), andTorin2(5.1mg/kg,n =3)andwere infected with E. coli (1 3 108 CFU) for 12 h. (A) The percentage of macrophage (F4/80+CD11b+) from peritoneal fluid was analyzed by FACS. (B) Peritoneal fluid was diluted 10-fold in PBS and cultured in solid Luria–Bertani (LB) medium for 12 h to count CFU. Mice received an i.p. injection of DMSO/PBS (10%, 300 ml, n = 3) and Gap19 (1.45 mg/kg, n =3)and were infected with E. coli (1 3 108 CFU) for 12 h. (C) The percentage of macrophage (F4/80+CD11b+) from peritoneal fluid was analyzed by FACS. (D) Perito- neal fluid was diluted 10-fold in PBS and cultured in solid LB medium for 12 h to count CFU. For (A)–(D), values are presented as means 6 SEMs (n =3). *p , 0.05, **p , 0.01 versus vehicle group by Student t test. (E) Mice received an i.p. injection of DMSO/PBS (10%, 300 ml, n = 10), and Gap19 (1.45 mg/kg, n = 10) before infection with 4 3 108 CFU E. coli. Then, the mice were checked every 4 h over the next 48 h. *p , 0.05 by log-rank test.

Increasing evidence has demonstrated that dysregulated mac- direct relationship between Cx43 and SGK, AKT/mTOR-mediated rophage migration is involved in tissue damage and chronic in- Cx43 regulation was reported by us and others (13, 31), and SGK is flammation due to misdirected immune responses (27). However, it known to act downstream of mTORC2 (32, 33). In addition, SGK was still unclear how Cx43 mediates CM-induced macrophage has been found to play an essential regulatory role in adaptive migration. In our study, we found that Cx43-regulated iNOS immunity (34, 35) and to be phosphorylated in response to LPS expression may partly explain this mechanism. Previously, re- stimulation in human monocytes (36). Consistent with these find- searchers have demonstrated that LPS increases macrophage ings, CM-induced macrophage migration by silencing or blockade migration through the activity of Src and FAK (28–30) and that of the SGK pathway was also found in our study. However, the the iNOS/NO system is essential for LPS-mediated Src/FAK in- involvement and functional role of SGK in innate immune reduction (7, 30). Consistent with this, we found that inhibition of sponses remain to be elucidated. iNOS expression was accompanied by decreased NO production In this study, we found that both rapamycin, a sensitive inhibitor and Src/FAK expression through silencing of Cx43 in CM- of mTORC1, and Torin2, which potently targets mTORC1 and induced macrophages; this effect was also observed following mTORC2, abolished the LPS-dependent increase in Cx43 ex- pretreatment with the iNOS inhibitor 1400W, indicating that the pression, suggesting that mTORC1 and mTORC2 may both iNOS/Src/FAK axis may be involved in Cx43-mediated CM- play roles in this process. However, rapamycin induced SGK- induced macrophage migration. Fig. 9 represents a simple model independent Cx43 reduction. Further studies are needed to eluci- illustrating the role of Cx43 in CM-induced macrophage migration. date these mechanisms. In our investigation of upstream molecules involved in LPS- FK506 (tacrolimus) and methylprednisolone are immunosup- induced Cx43 overexpression in macrophages, we found that pressive drugs that are commonly administered following solid organ Cx43 was upregulated by FBS or bFGF or SGK activators (14, 15) transplantation; these drugs significantly decreased LPS-induced and downregulated by knockdown of the SGK gene or blockade of Cx43 expression and SGK activation in TGEMs. FK506 has long the SGK pathway. Although no other studies have revealed the been the cornerstone of the immunosuppressive standard regimen 2996 mTOR/SGK REGULATE MACROPHAGE MIGRATION THROUGH Cx43

present study and Gap19 inhibits Cx43 hemichannel opening, to elucidate the channel-independent role of Cx43 signaling in host defense against invading bacteria, further experiment is needed. Regarding mTOR, because macrophages can use autophagy to defend against infection and mTOR signaling modulates autophagy, further studies are required to determine the extent to which autophagy influences the peptide–MHC class II repertoire and to evaluate how this may be affected by mTOR inhibition (38, 39). Although the objectives of this study were fulfilled, there were some limitations. First, in contrast to our results, there are reports of higher levels of iNOS and NO induced by IL-1b/TNF-a in Cx43- deficient fibroblasts (26) and increased iNOS expression in the mitochondria of Cx43-deficient mice hearts (40). The reasons for this discrepancy are presently unclear; however, it could be at- tributed to the different cell types and experimental settings used for the investigations. More detailed analysis to account for the difference may be needed in the future. In addition, autocrine and paracrine effect of NO on macrophage function were also not tested. Second, our results showed both mTORC1 and mTORC2 are involved in regulating Cx43 expression. However, more detailed investigations on mTORC1-mediated Cx43 expression might be required. Third, this study only focused on the role of Cx43 in macrophage migration. Alternative activation status of macrophages are reported to affect the migratory properties of macrophages (41), but whether the classical (Th1) or alternative (Th2) activation is associated with Cx43 or SGK activation has not yet been tested. In summary, our findings revealed a previously unrecognized signaling pathway and function of CM- or LPS-induced Cx43 FIGURE 9. Schematic for the working model of the role of Cx43 in expression in macrophages, confirming the roles of Cx43 in the host macrophage migration. response to microbial infection.

Disclosures in solid organ transplantation. FK506 blocks the action of lym- The authors have no financial conflicts of interest. phocytes that can cause the body to reject the transplanted organ (37). However, both FK506 and methylprednisolone are associated with a range of side effects, including infection. Although further confir- References matory experiments are needed, we propose that Cx43 regulation 1. Gordon, S. 2003. Alternative activation of macrophages. Nat. Rev. Immunol. 3: may be involved in mediating the risk of serious infection in pa- 23–35. tients treated with these drugs. 2. Dalpke, A., and K. Heeg. 2002. Signal integration following Toll-like receptor triggering. Crit. Rev. Immunol. 22: 217–250. In vivo, we observed the inhibited macrophage recruitment and 3. Guha, M., and N. Mackman. 2002. The phosphatidylinositol 3-kinase-Akt bacteria clearance in E. coli–infected peritonitis mouse model by pathway limits lipopolysaccharide activation of signaling pathways and ex- inhibition of mTOR/SGK signaling or blocking the Cx43 hem- pression of inflammatory mediators in human monocytic cells. J. Biol. Chem. 277: 32124–32132. ichannel, suggesting the role of mTOR/SGK/Cx43 signaling in 4. Miguelez, C., T. Morera-Herreras, M. Torrecilla, J. A. Ruiz-Ortega, and macrophage migration and immune responses. Although the L. Ugedo. 2014. Interaction between the 5-HT system and the basal ganglia: functional implication and therapeutic perspective in Parkinson’s disease. Front. population of myeloid cells from the peritoneal cavity was more Neural Circuits 8: 21. significantly reduced in GSK650349- and Torin2-treated mice 5. Leu, T. H., S. Charoenfuprasert, C. K. Yen, C. W. Fan, and M. C. Maa. 2006. than in Gap19-treated mice, the number of recovered E. coli was Lipopolysaccharide-induced c-Src expression plays a role in nitric oxide and TNFalpha secretion in macrophages. Mol. Immunol. 43: 308–316. reversed. These experimental data suggested that Cx43 bestowed 6. Owen, K. A., F. J. Pixley, K. S. Thomas, M. Vicente-Manzanares, B. J. Ray, specificity to macrophage migration. Specific inhibition of Cx43 A. F. Horwitz, J. T. Parsons, H. E. Beggs, E. R. Stanley, and A. H. Bouton. 2007. by Gap19 seemed to result in specific inhibition of macrophage Regulation of lamellipodial persistence, adhesion turnover, and motility in macrophages by focal adhesion kinase. J. Cell Biol. 179: 1275–1287. migration without affecting the movement of other myeloid 7. Maa, M. C., M. Y. Chang, Y. J. Chen, C. H. Lin, C. J. Yu, Y. L. Yang, J. Li, cells. In contrast, GSK650349 and Torin2 inhibited the upstream P. R. Chen, C. H. Tang, H. Y. Lei, and T. H. Leu. 2008. Requirement of inducible nitric-oxide synthase in lipopolysaccharide-mediated Src induction and macro- components of the mTOR/SGK/Cx43 signaling pathway, thereby phage migration. J. Biol. Chem. 283: 31408–31416. suppressing the movement of a broad range of myeloid cells. 8. Saez, J. C., V. M. Berthoud, M. C. Branes, A. D. Martinez, and E. C. Beyer. Additionally, inhibition of macrophage recruitment by blocking 2003. Plasma membrane channels formed by connexins: their regulation and functions. Physiol. Rev. 83: 1359–1400. the hemichannel may partly contribute to weakened bacteria 9. Leithe, E., M. Mesnil, and T. Aasen. 2017. The connexin 43 C-terminus: a tail of clearance because Cx43-mediated UDP release may be involved many tales. Biochim. Biophys. Acta 1860: 48–64. in mediating the effects of Gap19 treatment (20). Furthermore, 10. Eugenıń, E. A., M. C. Bran˜es, J. W. Berman, and J. C. Saéz. 2003. TNF-alpha plus IFN-gamma induce connexin43 expression and formation of gap junctions mice survival in the acute peritonitis mouse model was de- between human monocytes/macrophages that enhance physiological responses. pressed by Gap19, which further confirmed the pivotal role J. Immunol. 170: 1320–1328. 11. Kim, M. O., Y. J. Lee, and H. J. Han. 2010. Involvement of Cx43 phosphory- of Cx43 hemichannel in the fight against bacterial infection. lation in 59-N-ethylcarboxamide-induced migration and proliferation of mouse However, as mTOR/SGK is upstream of Cx43 according to our embryonic stem cells. J. Cell. Physiol. 224: 187–194. The Journal of Immunology 2997

12. Zang, J. P., and R. Wei. 2015. Effects of Cx43 gene modification on the pro- 26. Li, K., J. Yao, L. Shi, N. Sawada, Y. Chi, Q. Yan, H. Matsue, M. Kitamura, and liferation and migration of the human lung squamous carcinoma cell line NCI- M. Takeda. 2011. Reciprocal regulation between proinflammatory cytokine- H226. Genet. Mol. Res. 14: 13110–13119. induced inducible NO synthase (iNOS) and connexin43 in bladder smooth 13. Shen, C., M. R. Kim, J. M. Noh, S. J. Kim, S. O. Ka, J. H. Kim, B. H. Park, and muscle cells. J. Biol. Chem. 286: 41552–41562. J. H. Park. 2016. Glucocorticoid suppresses Connexin 43 expression by inhib- 27. Zicha, D., W. E. Allen, P. M. Brickell, C. Kinnon, G. A. Dunn, G. E. Jones, and iting the Akt/mTOR signaling pathway in osteoblasts. [Published erratum ap- A. J. Thrasher. 1998. Chemotaxis of macrophages is abolished in the Wiskott- pears in 2016 Calcif. Tissue Int. 99: 98.] Calcif. Tissue Int. 99: 88–97. Aldrich syndrome. Br. J. Haematol. 101: 659–665. 14. Webster, M. K., L. Goya, Y. Ge, A. C. Maiyar, and G. L. Firestone. 1993. 28. Parsons, J. T., K. H. Martin, J. K. Slack, J. M. Taylor, and S. A. Weed. 2000. Characterization of sgk, a novel member of the serine/threonine protein kinase Focal adhesion kinase: a regulator of focal adhesion dynamics and cell move- gene family which is transcriptionally induced by glucocorticoids and serum. ment. Oncogene 19: 5606–5613. Mol. Cell. Biol. 13: 2031–2040. 29. Shimojo, N., R. Hashizume, K. Kanayama, M. Hara, Y. Suzuki, T. Nishioka, 15. Mizuno, H., and E. Nishida. 2001. The ERK MAP kinase pathway mediates M. Hiroe, T. Yoshida, and K. Imanaka-Yoshida. 2015. Tenascin-C may accel- induction of SGK (serum- and glucocorticoid-inducible kinase) by growth fac- erate cardiac fibrosis by activating macrophages via the integrin aVb3/nuclear tors. Genes Cells 6: 261–268. factor-kB/interleukin-6 axis. Hypertension 66: 757–766. 16. Leong, M. L., A. C. Maiyar, B. Kim, B. A. O’Keeffe, and G. L. Firestone. 2003. 30. Maa, M. C., M. Y. Chang, J. Li, Y. Y. Li, M. Y. Hsieh, C. J. Yang, Y. J. Chen, Expression of the serum- and glucocorticoid-inducible protein kinase, Sgk, is a Y. Li, H. C. Chen, W. E. Cheng, et al. 2011. The iNOS/Src/FAK axis is critical in cell survival response to multiple types of environmental stress stimuli in Toll-like receptor-mediated cell motility in macrophages. Biochim. Biophys. mammary epithelial cells. J. Biol. Chem. 278: 5871–5882. Acta 1813: 136–147. 17. Liu, Q., J. Wang, S. A. Kang, C. C. Thoreen, W. Hur, T. Ahmed, D. M. Sabatini, 31. Liu, L., X. Hu, G. Y. Cai, Y. Lv, L. Zhuo, J. J. Gao, S. Y. Cui, Z. Feng, B. Fu, and and N. S. Gray. 2011. Discovery of 9-(6-aminopyridin-3-yl)-1-(3-(trifluoromethyl) X. M. Chen. 2012. High glucose-induced hypertrophy of mesangial cells is re- phenyl)benzo[h][1,6]naphthyridin-2(1H)-one (Torin2) as a potent, selective, and versed by connexin43 overexpression via PTEN/Akt/mTOR signaling. Nephrol. orally available mammalian target of rapamycin (mTOR) inhibitor for treatment of Dial. Transplant. 27: 90–100. cancer. J. Med. Chem. 54: 1473–1480. 32. Alessi, D. R., L. R. Pearce, and J. M. Garcıá-Martıńez. 2009. New insights into 18. Giubellino, A., P. Bullova, S. No¨lting, H. Turkova, J. F. Powers, Q. Liu, mTOR signaling: mTORC2 and beyond. Sci. Signal. 2: pe27. S. Guichard, A. S. Tischler, A. B. Grossman, and K. Pacak. 2013. Combined 33. Jhanwar-Uniyal, M., J. L. Gillick, J. Neil, M. Tobias, Z. E. Thwing, and inhibition of mTORC1 and mTORC2 signaling pathways is a promising thera- R. Murali. 2015. Distinct signaling mechanisms of mTORC1 and mTORC2 in peutic option in inhibiting pheochromocytoma tumor growth: in vitro and in vivo glioblastoma multiforme: a tale of two complexes. Adv. Biol. Regul. 57: 64–74. studies in female athymic nude mice. Endocrinology 154: 646–655. 34. Heikamp, E. B., C. H. Patel, S. Collins, A. Waickman, M. H. Oh, I. H. Sun, 19. Wang, N., E. De Vuyst, R. Ponsaerts, K. Boengler, N. Palacios-Prado, P. Illei, A. Sharma, A. Naray-Fejes-Toth, G. Fejes-Toth, et al. 2014. The AGC J. Wauman, C. P. Lai, M. D. Bock, E. Decrock, M. Bol, et al. 2013. Selective kinase SGK1 regulates TH1 and TH2 differentiation downstream of the inhibition of Cx43 hemichannels by Gap19 and its impact on myocardial mTORC2 complex. Nat. Immunol. 15: 457–464. ischemia/reperfusion injury. Basic Res. Cardiol. 108: 309. 35. Wu, C., N. Yosef, T. Thalhamer, C. Zhu, S. Xiao, Y. Kishi, A. Regev, and 20.Qin,J.,G.Zhang,X.Zhang,B.Tan,Z.Lv,M.Liu,H.Ren,M.Qian,and V. K. Kuchroo. 2013. Induction of pathogenic TH17 cells by inducible salt- B. Du. 2016. TLR-activated gap junction channels protect mice against sensing kinase SGK1. Nature 496: 513–517. bacterial infection through extracellular UDP release. J. Immunol. 196: 36. Zhou, H., S. Gao, X. Duan, S. Liang, D. A. Scott, R. J. Lamont, and H. Wang. 1790–1798. 2015. Inhibition of serum- and glucocorticoid-inducible kinase 1 enhances TLR- 21. O’Donnell, III J. J., A. A. Birukova, E. C. Beyer, and K. G. Birukov. 2014. Gap mediated inflammation and promotes endotoxin-driven organ failure. FASEB J. junction protein connexin43 exacerbates lung vascular permeability. PLoS One 29: 3737–3749. 9: e100931. 37. Kelly, P. A., G. J. Burckart, and R. Venkataramanan. 1995. Tacrolimus: a new 22. Tsuchida, S., Y. Arai, T. Kishida, K. A. Takahashi, K. Honjo, R. Terauchi, immunosuppressive agent. Am. J. Health Syst. Pharm. 52: 1521–1535. H. Inoue, R. Oda, O. Mazda, and T. Kubo. 2013. Silencing the expression of 38. Martinez, J., K. Verbist, R. Wang, and D. R. Green. 2013. The relationship connexin 43 decreases inflammation and joint destruction in experimental ar- between metabolism and the autophagy machinery during the innate immune thritis. J. Orthop. Res. 31: 525–530. response. Cell Metab. 17: 895–900. 23. Anand, R. J., S. Dai, S. C. Gribar, W. Richardson, J. W. Kohler, R. A. Hoffman, 39. Thomson, A. W., H. R. Turnquist, and G. Raimondi. 2009. Immunoregulatory M. F. Branca, J. Li, X. H. Shi, C. P. Sodhi, and D. J. Hackam. 2008. A role for functions of mTOR inhibition. Nat. Rev. Immunol. 9: 324–337. connexin43 in macrophage phagocytosis and host survival after bacterial peri- 40. Kirca, M., P. Kleinbongard, D. Soetkamp, J. Heger, C. Csonka, P. Ferdinandy, toneal infection. J. Immunol. 181: 8534–8543. and R. Schulz. 2015. Interaction between connexin 43 and nitric oxide synthase 24. Glass, A. M., B. J. Wolf, K. M. Schneider, M. F. Princiotta, and S. M. Taffet. in mice heart mitochondria. J. Cell. Mol. Med. 19: 815–825. 2013. Connexin43 is dispensable for phagocytosis. J. Immunol. 190: 4830–4835. 41. Vogel, D. Y., P. D. Heijnen, M. Breur, H. E. de Vries, A. T. Tool, S. Amor, and 25. Jackson, P. E., Q. P. Feng, and D. L. Jones. 2008. Nitric oxide depresses con- C. D. Dijkstra. 2014. Macrophages migrate in an activation-dependent manner to nexin 43 after myocardial infarction in mice. Acta Physiol. (Oxf.) 194: 23–33. chemokines involved in neuroinflammation. J. Neuroinflammation 11: 23.