The ability of macrophages to respond to chemoattractants and inflammatory signals is important for their migration to sites of inflammation and immune activity and for host responses to infection. Macrophages differentiated from the bone marrow (BM) of UV-irradiated mice, even after activation with LPS, migrated inefficiently toward CSF-1 and CCL2. When BM cells were harvested from UV-irradiated mice and transplanted into naive mice, the recipient mice (UV-chimeric) had reduced accumulation of elicited monocytes/macrophages in the peritoneal cavity in response to inflammatory thioglycollate or alum. Macrophages differentiating from the BM of UV-chimeric mice also had an inherent reduced ability to migrate toward chemoattractants in vitro, even after LPS activation. Microarray analysis identified reduced reticulon-1 mRNA expressed in macrophages differentiated from the BM of UV-chimeric mice. By using an anti-reticulon-1 Ab, a role for reticulon-1 in macrophage migration toward both CSF-1 and CCL2 was confirmed. Reticulon-1 subcellular localization to the periphery after exposure to CSF-1 for 2.5 min was shown by immunofluorescence microscopy. The proposal that reduced reticulon-1 is responsible for the poor inherent ability of macrophages to respond to chemokine gradients was supported by Western blotting. In summary, skin exposure to erythemal UV radiation can modulate macrophage progenitors in the BM such that their differentiated progeny respond inefficiently to signals to accumulate at sites of inflammation and immunity.
In both humans and experimental animals, exposure of the skin to erythemal or chronic low-dose UV radiation (UVR) causes a systemic immunosuppression such that responses to experimental Ags applied to nonirradiated sites are reduced (1, 2). UVR-induced systemic immunosuppression has been implicated in the positive latitude gradients (higher latitude, less solar UVR, more disease) reported for several chronic immune diseases such as multiple sclerosis and type 1 diabetes (3, 4), as well as growth of tumors at nonirradiated sites (1, 5). Many pathways in skin are activated by UV irradiation, resulting in the transit of cells and mediators to the draining lymph nodes. Published reviews have extensively covered the numerous molecular and cellular changes in UV-irradiated skin that have been implicated in downstream suppression of both local and systemic immunity (1, 5). UVR-induced suppression of a response to a contact sensitizer applied to a distant skin site can still be detected 1 mo after UVR exposure (6). The long-lasting nature of UVR-induced immunosuppression suggests bone marrow (BM) stem and progenitor cell involvement, as does the season-of-birth effect reported for many autoimmune diseases (6, 7). In light of the suppressed immune responses in UV-irradiated mice, analyses have concentrated on the development and activation of tolerogenic dendritic cells (DCs) and regulatory T and B cells (1, 5, 8–11), with little emphasis on the effect of UV irradiation of skin on macrophages, a cell type with diverse functions in tissue homeostasis and inflammation (12, 13). Previous studies have noted that there is a large influx of macrophages into UV-irradiated human and mouse skin (14, 15). It is proposed that the macrophages respond to UVB-induced ligands via their chemokine receptor CCR2, and in neonatal skin, the macrophages accumulating in response to the CCR2 ligand produce IFN-γ (16) but are less efficient at presenting skin Ags associated with UV damage (17). Two decades ago, it was reported that UV irradiation of skin reduced the phagocytosis and intracellular killing of mycobacteria by macrophages isolated from the spleen and peritoneal cavity, and their ability to produce NO (18). Thus, there is some evidence that macrophage function is altered by UV-irradiation of skin.
The ability of macrophages to respond to chemoattractants and inflammatory signals is important for their attraction to inflammatory sites and localization to sites of immune activity. Most tissue macrophages have their origin prenatally in yolk sac and fetal liver, but monocyte-derived macrophages that respond during inflammation in postnatal life differentiate from the BM (19). This links with studies in our laboratory of chimeric mice; to create them, BM cells are transplanted into BM-ablated naive, recipient mice. The chimeric model allows the study of changes to BM cells without the ongoing influence of UVR-induced changes to cells in the periphery of donor mice. In the recipient mice, after ∼12 wk, the progeny of the transferred BM cells have populated all tissues. We have previously shown that, in both UV-irradiated mice and UV-chimeric mice, DCs differentiating from the BM are poorly immunogenic. Further investigations suggested that DCs differentiating from the BM of UV-chimeric mice had a reduced ability upon activation to migrate to draining lymph nodes, resulting in dampened immune responses (6, 20).
Because macrophages and DCs differentiate from a common hematopoietic precursor in the BM (21), we used a similar experimental model to examine the effect of UVR on macrophages differentiating from the BM of both UV-irradiated mice and UV-chimeric mice. A reduced chemotactic response to CSF-1 and CCL-2 was measured. Furthermore, the poor mobility was detected even if the macrophages were activated with LPS. Monocytes/macrophages differentiated in vivo from the BM of UV-chimeric mice were also inefficient in responding to an inflammatory signal; significantly fewer accumulated in the peritoneal cavity following injection of thioglycollate or alum. In these studies, the basis for the functional changes in the macrophages was sought by microarray, flow cytometry, Western blotting, immunofluorescence microscopy, and use of a blocking Ab to reticulon-1 in assays of macrophage migration.
Materials and Methods
Mice and ethics statement
All experiments were performed with the approval of the Telethon Kids Institute Animal Ethics Committee (approvals 278, 296), and standard operating procedures for euthanasia and anesthesia defined according to guidelines of the National Health and Medical Research Council of Australia. Female C57BL/6J (CD45.2 alloantigen) and B6.SJL-Ptprca (CD45.1 alloantigen) mice were obtained from the Animal Resources Centre (Murdoch, WA).
UV irradiation of mice
A bank of TL40W/12RS lamps (Philips, Amsterdam, the Netherlands) emitting broadband UVR with 65% UVB (280–320 nm) and peak emission at 313 nm was used as previously described (22). UVR (8 kJ/m2) was delivered; this is equivalent to three to four minimal erythemal doses and is measured by skin edema.
Culture of BM cells for differentiation of macrophages
BM-derived macrophages were isolated as previously described with some modifications (23). In brief, BM cells were collected from tibias and femurs of mice and cultured in RPMI 1640 (HyClone; GE Health Care Life Sciences, Logan, UT) supplemented with 10% FCS (HyClone) (RPMI 10) and 0.6 ng/ml (50 IU/ml) CSF-1 (kind gift of Dr. E.R. Stanley) for 24 h. Nonadherent cells were then collected and cultured in RPMI 10 supplemented with 12 ng/ml CSF-1 (1000 IU/ml) for 3 d in 10-cm Petri dishes to induce differentiation (some cells were frozen at day 2). To induce further spreading and adhesion as they continue to differentiate into mature macrophages, at day 4 the immature adherent macrophages were lifted using 2 mM EDTA in PBS, reseeded at 105 cells/ml in 10 ml RPMI 10 supplemented with 120 ng/ml CSF-1 (10,000 IU/ml), and cultured unless stated otherwise for a further 5 d prior to use (total of 9 d in culture) (24, 25).
Prior to staining of cultured BM cells, or cells isolated from the peritoneal cavity, they were incubated for 5 min with rat anti-mouse CD16/CD32 Fc receptor Ab to prevent nonspecific binding. Cells were then stained for 30 min with fluorescently labeled rat anti-mouse Abs: FITC anti-F4/80, PE or allophycocyanin.Cy7 anti-CD11b, PE anti-CD11c, allophycocyanin anti-Gr1, allophycocyanin anti-CCR2 (CD192), or allophycocyanin anti–CSF-1R (CD115) (BD Biosciences, San Jose, CA). For surface staining of reticulon-1 on BM-differentiated macrophages, a mouse monoclonal IgG (MON162) to reticulon-1A (ab9274) (Abcam, Cambridge, MA) was used. After washing, these cells were incubated with an AF647 goat anti-mouse IgG (Ab1505115; Abcam). For intracellular staining of reticulon-1, BM-differentiated macrophages were incubated in fixation/permeabilization buffer and fixed for 30 min at room temperature. The cells were then washed, and Ab staining was performed as above in permeabilization buffer (eBioscience, San Diego, CA). Data were collected on a BD LSR-II flow cytometer (BD Biosciences), and flow cytometric analyses were performed using FlowJo software (version 10.0; Tree Star, Ashland, OR).
Macrophage chemotaxis assay to CSF-1 and CCL2
Replicate cultures of BM-differentiated macrophages (2.5 × 105) were seeded into transwell inserts with 8-μm pores (BD Biosciences) in 200 μl of RPMI 10 (CSF-1– or CCL2-free), and inserts were placed into a 24-well companion plate with RPMI 10 containing 120 ng/ml CSF-1 or 20 ng/ml CCL2, respectively. Where indicated, BM-macrophages were resuspended in RPMI 10 containing the anti-mouse reticulon-1A IgG (5 μg/ml) or, as a control, an anti-mouse IgG (5 μg/ml) prior to seeding into transwell inserts. The cells were allowed to migrate for 5 h before fixation in 4% paraformaldehyde. The cells were then stained for 5 min with NucBlue fixed cell stain (Life Technologies, VIC, Australia), as per manufacturer’s instructions, and imaged using a Nikon C2 Plus confocal microscope (Nikon, Tokyo, Japan). The number of migrated cells was counted for 10 representative fields per sample at 20× magnification, and results were normalized to the number of migrated control cells. Where indicated, BM-macrophages were stimulated for 24 h with 1 μg/ml LPS from Escherichia coli 0111:B4 (Sigma-Aldrich, St. Louis, MO).
Generating UV-chimeric mice
Donor BM cells were isolated from congenic B6.SJL-Ptprca (CD45.1 alloantigen) mice that had been administered 8 kJ/m2 UVR to their shaved backs 3 d previously (6). BM cells from tibias and femurs were disaggregated and passed through sterile cotton wool to remove bone debris. RBCs were lysed with ammonium chloride. The recipient mice were 8-wk-old C57BL/6J mice (CD45.2 alloantigen) that were gamma-irradiated (2 × 550 rad) using a 137Cs source (Gammacell 3000 Elan; MDS Nordion, Ottawa, ON, Canada) prior to injection of 2 × 106 donor BM cells. Control chimeric mice were gamma-irradiated and injected with 2 × 106 BM cells from naive congenic B6.SJL-Ptprca mice (shaved but not UV-irradiated). The engraftment of cells in the chimeric mice was tracked for 12 wk in the BM, spleen, and lymph nodes (6, 20).
Assay of monocyte/macrophage migration into the peritoneal cavity
The migration capabilities of monocytes/macrophages into the peritoneal cavity of control- and UV-chimeric mice (12–15 wk since injection of donor cells for their establishment) were examined by i.p. injection of 1 ml thioglycollate (3.8% Medium Brewer Modified; BD Biosciences), 0.2 ml alum (aluminum hydroxide; 2 mg/mouse) (SERVA Electrophoresis, Heidelberg, Germany) or 1 ml 0.9% saline. After 3 d, the peritoneal cavity was washed out with saline. The harvested cells were counted and objectively identified (26) by staining with fluorescently labeled Abs directed against F4/80, CD11b, CD11c, and Gr1.
Measurement of cytokine and chemokine production
Cytokines in culture supernatants were measured using Bio-Plex Pro Mouse Cytokine 23-plex panel (Bio-Rad Laboratories, Hercules, CA) as per the manufacturer’s instructions. Before analysis, samples were diluted 1:2, 1:5, or 1:25 using Bio-Plex mouse serum diluent (Bio-Rad Laboratories) as recommended by the manufacturer. Prepared plates were run on a Luminex 100 instrument (Luminex, Austin, TX). The cytokines analyzed included IL-1α, IL-1β, IL-2, IL-3, IL-4, IL-5, IL-6, IL-9, IL-10, IL-12 p40, IL-12 p70, IL-13, IL-17, Eotaxin (CCL11), G-CSF, GM-CSF, IFN-γ, KC, MCP-1, MIP-1α, MIP-1β, RANTES, and TNF-α.
Assays of macrophage lactate production and responses to glycolytic stress
Concentrations of secreted l-lactate (subsequently referred to as lactate) in culture supernatants were quantified using the Glycolysis Cell-Based Assay Kit (catalog no. 600450; Cayman Chemical, Ann Arbor, MI). In brief, BM cells after 8 d in culture were lifted using 2 mM EDTA in PBS and then recultured, in triplicate, in flat-bottom 96-well plates at a density of 0.8 × 106/ml in 120 μl RPMI 1640 medium containing 0.5% FCS and 120 ng/ml CSF-1. Samples were cultured with or without 1 μg/ml LPS, and supernatants were collected after 24 h. Supernatants were immediately centrifuged (1500 rpm, 5 min) to ensure that they were cell-free and frozen immediately at −80°C.
Bioenergetics was measured using a Seahorse XFe96 Extracellular Flux Analyzer (Seahorse Bioscience, North Billerica, MA) and the Seahorse Glycolytic Stress Test as previously reported (27). The extracellular acidification rate (ECAR) along with the seeding density and reagents were optimized according to manufacturer’s specification. In short, 1.6 × 105 cells/well were cultured in RPMI 10 supplemented with 120 ng/ml CSF-1 for 24 h in poly-d-lysine–coated (Sigma-Aldrich) culture microplates (Seahorse Bioscience). After 24 h, the medium was replaced with serum-free, DMEM Base medium (1 mM sodium pyruvate, 2 mM glutamine, and 0 mM glucose). The following were injected (at final concentration): 25 mM glucose, 1 μM oligomycin, and 100 mM 2-deoxyglucose. Three measurements were taken after each drug addition (6 min apart); the last measurement prior to subsequent injection was used for the statistical calculations. ECARs were normalized according to protein levels in each well using radioimmunoprecipitation buffer and the bicinchoninic acid protein assay (Thermo Fisher). Data were analyzed as published previously (28), with ECAR initial measurements standardized to 100% and all subsequent changes calculated as a percentage change relative to that 100%.
Microarray of BM-macrophages after differentiation for 9 d
Total RNA was extracted from macrophages differentiated for 9 d from BM using TRIzol (Life Technologies) followed by an RNeasy MinElute kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. At least 500 ng at 50 ng/ml (evaluated by the absorbance 260/280 ratio using a NanoDrop ND 1000 spectrophotometer; NanoDrop Technologies, Wilmington, DE) was sent to the Ramaciotti Centre for Genomics, University of New South Wales, Sydney, NSW, Australia, for analysis using Affymetrix GeneChip HT MG-430 PM Array. The Affy package from Bioconductor 3.2 was used to read the Affymetrix CEL files into R 3.3.1 (https://www.r-project.org/) and to preprocess the arrays using robust multichip average methodology. Quality control of the arrays was performed using the arrayQualityMetrics package (29). After initial exploration of all of the array data by PCA (total n = 32), we found the data primarily separated by LPS exposure, and thus we conducted the differential gene analysis of UV versus control in the LPS (n = 16) and non-LPS (n = 16) environments separately. The limma package (30) was used to conduct differential expression analysis, with gene targets with an unadjusted p ≤ 0.05 considered of significance for further analyses. The raw CEL files and processed data are available from NCBI Gene Expression Omnibus under accession number GSE98840 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE98840).
Detection of reticulon-1, paxillin, and motility-related proteins by Western blotting
Macrophages were grown to subconfluence (8-d culture) before CSF-1 starvation for 16 h to upregulate CSF-1R expression and then incubated with or without 120 ng/ml CSF-1 at 37°C for 2.5 min (24, 25). After incubation, the cells were washed twice with PBS (4°C) and lysed by scraping in lysis buffer (50 mM Tris-HCl, 150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 5 μM ZnCl2, 500 μM Na3VO4, 1 mM benzamidine, 10 μg/ml leupeptin, and 10 μg/ml aprotinin [pH 7.2]) for SDS-PAGE analysis. Polyvinylidene difluoride membranes were incubated with HRP substrates (EMS-Millipore), and the chemiluminescent signal was detected with an ImageQuant LAS 4000 biomolecular imager (GE Healthcare, Piscataway, NJ). The Abs used included anti–reticulon-1A as described earlier, anti-phosphotyrosine (4G10; EMS-Millipore, Billerica, MA), anti–β-actin (Sigma-Aldrich), anti-pY397 FAK (BD Transduction Laboratories, Lexington, KY), anti-pY402 Pyk2 (BD Transduction Laboratories), anti-p85 PI3K (Millipore, Billerica, MA), anti-pS473 Akt (Cell Signaling Technology, Danvers, MA), anti-p44/42 Erk (Cell Signaling Technology), anti-paxillin (BD Transduction Laboratories), and anti-phosphoY118paxillin (Invitrogen, Camarillo, CA). CSF-1R Ab (clone C19) was a kind gift from Dr. E.R. Stanley. HRP-conjugated secondary Abs were obtained from Cell Signaling Technology.
Detection of reticulon-1 subcellular localization by immunofluorescence microscopy
BM-differentiated macrophages (9 d in culture) were seeded onto fibronectin-coated coverslips (BD Biosciences) and grown for 48 h before CSF-1 starvation for a further 16 h, then incubated with or without 120 ng/ml CSF-1 at 37°C for 2.5 min. After stimulation, coverslips were sequentially fixed in 4% paraformaldehyde and permeabilized in 0.25% Triton X-100 before staining for F-actin (Alexa568 phalloidin, 1:100) as previously described (31). Anti–reticulon-1A Ab was used at 1:50 dilution; the secondary Ab was an Alexa 488–conjugated goat anti-mouse F(ab′)2 fragment Ab from Life Technologies (Invitrogen). Staining with DAPI (part of the mounting step) was performed using Prolong Diamond +DAPI (Invitrogen). Mounted coverslips were examined under an Olympus 1X-81 inverted microscope with images recorded using a cooled FluoView II CCD digital monochrome camera.
Yield and phenotype of cells differentiating from the BM of UV-irradiated mice
BM cells were harvested from nonirradiated mice or mice 3 d after administration of 8 kJ/m2 UVR to their shaved backs, and cultured with increasing concentrations of CSF-1. There were no significant differences in the number of BM cells harvested per mouse, or the rate of growth and differentiation of macrophages from BM progenitors (Table I). There was also no difference in cell yields from the BM of control- and UV-chimeric mice (12 wk after their establishment), and from these cells cultured with increasing concentrations of CSF-1 (Table I). After 9 d in culture, the cells were >98% F4/80+CD11b+. The expression of CD11c by harvested cells was low relative to that of BM-differentiated DCs (6).
Migration of BM-macrophages toward CSF-1 and CCL2
Although CSF-1 is necessary for macrophage growth and differentiation, it is also a potent chemokine that stimulates macrophage migration (24). The chemotaxis of fully differentiated (day 9) macrophages was investigated. In three independent experiments, macrophages from the BM of UV-irradiated mice migrated significantly less efficiently than those differentiated from the BM of nonirradiated mice (Fig. 1A, 1B). This was not due to differences in expression of the CSF-1R (CD115), which was expressed at similar levels by macrophages differentiated from the BM of nonirradiated and UV-irradiated mice (Fig. 1C). In three experiments, aliquots of macrophages were activated with LPS (1 μg/ml) for 24 h before assessing their migration. Expression of the CSF-1R was reduced by LPS exposure (Fig. 1C) and correlated with a decrease in the number of migrated cells (Fig. 1A, 1B). Thus, the inefficient migration by macrophages from the BM of UV-irradiated mice remained even after LPS stimulation. Migration toward CCL2 was also reduced for macrophages differentiated from the BM of UV-irradiated mice (± LPS activation) (Fig. 1D). CCL2 was originally called MCP-1 because of its ability to recruit monocytes to sites of inflammation associated with tissue injury or infection (32). CCL2 has been implicated in the pathogenesis of diseases such as rheumatoid arthritis and atherosclerosis, which are characterized by monocytic infiltrates. There were no significant differences in CCR2 (CD192) expression by macrophages differentiated from the BM of control or UV-irradiated mice (Fig. 1E). Upon activation by LPS, CCR2 expression was not changed (Fig. 1E), and the cells differentiated from the BM of UV-irradiated mice continued to have reduced mobility toward CCL2 (Fig. 1D).
Migration of BM-macrophages from chimeric mice toward CSF-1 and CCL2
To further analyze the longevity of the reduced migratory capacity of macrophages differentiating from the BM of UV-irradiated mice, we used BM cells from UV-irradiated mice to establish UV-chimeric mice. Control-chimeric mice were created with BM cells from nonirradiated mice. After 12 wk and ∼90% engraftment of donor CD45.1 cells into the BM, spleen, and peripheral lymph nodes of recipient mice (Table II), macrophages were differentiated as described earlier from the BM of the chimeric mice, and their migratory properties were analyzed. As shown in Fig. 2A, 2B, and 2D, macrophages differentiated from the BM of UV-chimeric mice migrated less efficiently toward CSF-1 and CCL2. Activation by LPS exposure significantly reduced CSF-1 receptor expression (Fig. 2C) but did not override or negate the migration differences measured for macrophages differentiated from BM of control- and UV-chimeric mice. In contrast, there were no changes in CCR2 (CD192) expression on the macrophages with LPS exposure, and macrophages differentiated from BM of control- and UV-chimeric mice expressed similar levels of CCR2 (data not shown).
Responses to inflammatory stimuli injected into the peritoneal cavity of chimeric mice
To investigate the migratory capacity of monocytes/macrophages differentiated in vivo, but still with an origin from the donor BM of UV- or nonirradiated mice, we injected inflammatory mediators into the peritoneal cavity of chimeric mice. i.p. injection of inflammatory stimuli such as thioglycollate and LPS induces the recruitment of monocytes to the peritoneal cavity that differentiate into macrophages. After LPS exposure, peritoneal macrophages are ∼90% similar to BM-differentiated macrophages at the gene expression level (33). Fully engrafted chimeric mice (12 wk since BM cell transfer) were injected with 1 ml of 0.9% saline, 1 ml of thioglycollate (3.8% Medium Brewer Modified), or 2 mg of alum in a volume of 0.2 ml. After 3 d, cells that had collected in the peritoneal cavity were measured (Fig. 3). In response to thioglycollate injection, a significantly greater number of cells were harvested from the control-chimeric mice than from the UV-chimeric mice (Fig. 3A). There was a trend for a similar finding in the peritoneal cavity of UV-chimeric mice injected with alum. The phenotyping of peritoneal cells harvested after saline and thioglycollate injection is shown in Fig. 3B and 3C, respectively. The number of macrophages in the peritoneal cavity of control-chimeric mice was greater than that measured in UV-chimeric mice after injection of thioglycollate and alum; these were principally defined as elicited macrophages (Fig. 3D, 3E). The responses to thioglycollate and alum differed; other than large numbers of macrophages, the former stimulated an influx of eosinophils, and the latter neutrophils (Fig. 3F, 3G). In summary, monocytes/macrophages developing in vivo from the BM of UV-chimeric mice were poor to accumulate in the peritoneal cavity in response to different inflammatory agents; this result supports the findings of the in vitro migration assays.
Cytokine production by macrophages differentiating from the BM of UV-chimeric mice
The possibility was investigated that autocrine production of cytokines or chemokines was controlling the differences in migration of the BM-differentiated macrophages from the control- and UV-chimeric mice. Following 8 d in culture, BM cells from six control-chimeric and six UV-chimeric were incubated for 24 h, with or without LPS (1 μg/ml), and 23 different cytokines and chemokines were analyzed in the culture supernatants (Fig. 4A and 4B for two potential macrophage chemokines). No differences in cytokine or chemokine levels were detected in the absence of LPS exposure. Generally, cytokine and chemokine levels were increased upon LPS exposure, and statistically significant differences were recorded for some cytokines in supernatants from cells differentiated from the BM of control- and UV-chimeric mice (up to a 10% change for some cytokines). However, because significant differences in macrophage migratory potential were detected for both LPS-stimulated and non-LPS-exposed macrophages, it was unlikely that autocrine cytokine production controlled differences in macrophage migration capabilities.
Lactate production by macrophages differentiated from the BM of UV-irradiated and UV-chimeric mice
Because the cytoskeletal rearrangements necessary for migration are one of the most energy-consuming cellular processes (34–36), metabolic responses were analyzed. Levels of lactate, an end-product of glycolysis, were assessed upon incubation of the differentiated macrophages for 24 h in medium containing 0.5% FCS and 120 ng/ml CSF-1. LPS significantly stimulated lactate production and consequently glycolysis in macrophages over 24 h, but there were no differences in the levels produced by macrophages from the BM of nonirradiated and UV-irradiated mice, and control- and UV-chimeric mice, respectively (Fig. 4C, 4D). Macrophages differentiated from the BM of chimeric mice were also investigated in a glycolytic stress test using real-time metabolic flux analysis under CSF-1–free conditions (Seahorse Flux Analyzer). LPS exposure increased the glycolytic response, but there were no differences detected between macrophages from the BM on control- and UV-chimeric mice (±LPS), confirming the absence of metabolic adaptions (Fig. 4E).
Microarray of genes expressed by macrophages differentiated from the BM of UV-chimeric mice
To further probe the mechanism underlying differences in migration both in vitro and in vivo by macrophages differentiating from the BM of control- and UV-chimeric mice, we examined mRNA from fully differentiated BM-macrophages (after 9 d in culture) by microarray. Because exposure to LPS does not cancel differences in migration ability, cells that had been exposed to LPS, or not, for 24 h (between days 8 and 9 of differentiation) were analyzed by microarray; this approach allowed identification of genes responsible for variations in migration against a background of different activation states. Microarray of RNA from cells under different activation states also provided an inbuilt control for replication of results. RNA was extracted from eight paired populations of BM-differentiated macrophages, each exposed for 24 h to LPS or control medium. The paired cell preparations had differentiated from the BM of eight pairs of control- and UV-chimeric mice, each pair euthanized on separate days between 12 and 15 wk after establishing the chimeric mice. When the array data were examined for subsets of genes that have a nominal (unadjusted) p ≤ 0.05 and were similarly regulated in cells both exposed and not exposed to LPS, there were 15 genes that were reduced in macrophages differentiated from the BM of UV-chimeric mice compared with those from the BM of control-chimeric mice, and 25 genes that were increased in expression (Supplemental Table I).
When the genes with an unadjusted p ≤ 0.05 were further filtered for a log2 fold change ≥0.4 (considered a biologically significant change), we found one gene, for reticulon-1, that was differentially expressed in the same direction in both the LPS-exposed and not exposed macrophages from the BM of UV-chimeric mice compared with those from control-chimeric mice. In macrophages differentiated from the BM of UV-chimeric mice, reticulon-1 mRNA levels were reduced by 27 and 36% for (−LPS) and (+LPS) cells, respectively.
Expression of reticulon-1 protein in macrophages differentiated from the BM of control- and UV-chimeric mice
Reticulon-1 expression was first examined by flow cytometry using an Ab to reticulon-1A previously recorded for staining macrophages (http://www.ProteinAtlas.org/ENSG00000139970-Rtn1/tissue). As shown in Fig. 5A and 5B, >90% BM-differentiated macrophages after 9 d in culture expressed reticulon-1 on their surface, as well as intracellularly. Differences in expression were not detected between cells from the BM of control- and UV-chimeric mice. For examination by Western blotting, macrophages differentiated for 8 d from the BM of four control- and four UV-chimeric mice were starved of CSF-1 for 16 h and then pulsed for 2.5 min with 120 ng/ml CSF-1, or control medium, before lysis and analysis. Again, the total levels of reticulon-1 expression were significant, and there were no detected differences (as a measure of actin) in lysates from macrophages differentiated from the BM of control- or UV-chimeric mice (Fig. 5C, Supplemental Fig. 1), confirming the earlier flow cytometry data. Unexpectedly, reticulon-1 levels were significantly reduced in all lysates of BM-macrophages after the 2.5-min CSF-1 stimulation (Fig. 5C, Supplemental Fig. 1). The reduction in the reticulon-1 band for lysates of BM-macrophages from UV-chimeric mice was significantly greater (49.8%) than that measured for BM-macrophages from control-chimeric mice (30.8%). These results suggest that reticulon-1 is involved in responses to CSF-1 because it is either activated and downregulated or becomes inaccessible to the Ab through protein complex formation following activation of CSF-1R signaling pathways. Importantly, reticulon-1 levels are reduced to a significantly greater extent in BM-macrophages from UV-chimeric mice.
Lysates were also examined for expression of adhesion- and migration-associated proteins (Supplemental Table II), particularly as macrophage spreading, adhesion, and migration through the extracellular matrix requires dynamic remodeling of the actin cytoskeleton associated with integrin clustering in podosomes (for enhanced migration) and focal adhesions (for greater adhesion and reduced chemotaxis) (24, 37, 38). When normalized to the expression of actin, there were no apparent changes in the levels of phosphorylated motility-related proteins and signaling molecules in macrophages differentiated from the BM of UV-chimeric mice compared with those from control-chimeric mice. However, macrophages differentiated from the BM of UV-chimeric mice expressed significantly higher levels of paxillin (with proportionally similar amounts of expression of phosphorylated paxillin) after CSF-1 stimulation (Fig. 5D, Supplemental Fig. 1) and suggested that CSF-1R signaling in these macrophages may stimulate enhanced adhesion and, in turn, reduce their capacity to migrate toward a chemoattractant.
Reticulon-1 translocates to the plasma membrane upon CSF-1 stimulation
Reticulon-3 was recently identified as a critical regulator of nonclathrin-mediated endocytosis of the epidermal growth factor (EGF) receptor (EGFR) (39). EGF stimulated contact between the endoplasmic reticulum and the plasma membrane in a reticulon-3–dependent manner. We determined whether CSF-1 induced plasma membrane reticulon-1 localization by starving macrophages differentiated from the BM of control-chimeric mice of CSF-1 for 16 h and then pulsing them for 2.5 min with 120 ng/ml CSF-1 or control medium. As shown in Fig. 6, reticulon-1 localized to the F-actin–rich cell cortex following CSF-1 stimulation, suggesting that CSF-1 triggers contact between the endoplasmic reticulum and the plasma membrane.
Reticulon-1 is involved in macrophage migration toward CSF-1
Because reticulon-1 is expressed on the surface by most BM-differentiated macrophages in the presence of CSF-1, its role in macrophage migration was investigated. In three independent experiments examining the migration of macrophages toward CSF-1, incubation with an Ab to reticulon-1 significantly reduced the ability of BM-differentiated macrophages to migrate toward CSF-1 (Fig. 7A). Incubation with the Ab to reticulon-1 also significantly reduced the ability of BM-differentiated macrophages prepared from four independent control-chimeric mice to migrate toward CCL2 (Fig. 7B). This result suggests that reticulon-1 is required for macrophage migration toward both CSF-1 and CCL2, and if reticulon-1 levels are reduced, macrophage migration toward chemoattractants is impaired.
Macrophages fully differentiated from the BM of UV-irradiated and UV-chimeric mice, compared with those from nonirradiated and control-chimeric mice, migrate less efficiently in vitro to a gradient of chemoattractants. The different behaviors could not be explained by altered expression of the receptors for those chemoattractants (CSF-1, CCL2). Reduced migration was also seen in the UV-chimeric mice after injection of thioglycollate or alum; in both scenarios, the cells accumulating would have differentiated from monocytes and, in turn, their precursors in the BM (40). The UV-chimeric mice were generated with BM from UV-irradiated mice and allowed analysis of the longevity of the effects of UVR on hematopoietic precursor cells in the BM. The hypothesis proposed was that in response to UVR, macrophage progenitor cells in BM were altered epigenetically such that their progeny that differentiate in response to CSF-1 have sustained reduced migration in response to multiple chemoattractants. We sought mRNA differences in these BM-differentiated macrophages by microarray; reticulon-1 transcripts were significantly reduced in macrophages from UV-chimeric mice, both in a basal state and after activation by LPS for 24 h. Expression of reticulon-1 on the surface of BM-differentiated macrophages was demonstrated, and by use of a blocking Ab, it was confirmed that reticulon-1 was functionally important in macrophage migration toward both CSF-1 and CCL2. When reticulon-1 was examined by Western blotting in BM-macrophages after a CSF-1 pulse, levels were reduced more significantly in cells from UV-chimeric mice than from control-chimeric mice. The same CSF-1 pulse also induced translocation of reticulon-1 to the plasma membrane.
The reticulons are a group of membrane-associated proteins with much experimentation supporting their involvement in shaping the tubular endoplasmic reticulum network (41). There are four members in mammals with varied N-terminal domains that help confer specific functions. They are expressed in the endoplasmic reticulum, Golgi, and plasma membranes, and can be involved in secretory pathways, membrane trafficking, apoptosis, and with relevance to our study, cell migration. Reticulon-1 is an interacting partner of the GTPase activation protein, TBC1D20 (42). The small GTPase activated by TBC1D20 is Rab1, a molecule that can regulate integrin β1 recycling to lipid rafts to promote cell migration (43). In our microarrays, a significant reduction in TBC1D20 transcripts was measured in non-LPS-exposed macrophages differentiated from the BM of UV-chimeric mice (p = 0.03). The involvement of another reticulon, reticulon-4, in macrophage migration has been previously reported; macrophages from mice unable to express reticulon-4 had reduced spreading and chemotaxis (44). In our analyses, reticulon-4 transcripts were not differentially expressed in BM-macrophages from control- and UV-chimeric mice. A new role for reticulon-3 located at endoplasmic reticulum–plasma membrane contact sites has recently been described as a critical regulator of EGFR nonclathrin endocytosis (39). At higher concentrations of EGF, ∼40% of EGFRs use nonclathrin endocytosis to traffic to the lysosome for degradation, with reticulon-3 acting as a tethering factor between the endoplasmic reticulum–plasma membrane (39).
This study has shown for the first time, to our knowledge, that reticulon-1 is expressed both on the surface and intracellularly of BM-differentiated macrophages. The macrophages required reticulon-1 to migrate sufficiently toward the chemoattractants, CSF-1 and CCL2, supporting reticulon-1 as an important membrane protein. The Western blotting and immunofluorescence experiments also confirmed involvement of reticulon-1 in responses to CSF-1. Just as reticulon-3 can regulate the type of endocytosis induced by EGFRs (39), reticulon-1 may determine responses of macrophages to CSF-1. We hypothesize that reticulon-1 will also localize to the cell membrane after a short exposure to CCL2. Migratory responses toward thioglycollate and alum in vivo were reduced when monocytes/macrophages were differentiated from the BM of UV-chimeric mice. Thus, we propose that reticulon-1 may be involved more broadly in monocyte and macrophage responses to chemoattractants and inflammatory agents, and that surface expression of reticulon-1 is reduced in macrophages differentiated in vivo from the BM of UV-chimeric mice. The mechanisms by which reticulon-1 and migratory responses may be linked are the subject of further experimentation. The outcomes of Western blotting may represent proteolytic cleavage or decreased expression of reticulon-1 after 2.5 min of CSF-1 exposure. However, the immunofluorescence staining suggested that reticulon-1 moves within the cell. The Western blotting experiments also detected increased levels of paxillin, with proportionally similar amounts of expression of phosphorylated paxillin, in lysates of macrophages from the BM of UV-chimeric mice. This suggests paxillin involvement in responses to CSF-1 by the less migratory BM-macrophages from UV-chimeric mice, possibly through increased formation of adhesion structures (45). Links between reticulon-1 and paxillin in macrophages from the BM of UV-chimeric mice after a short exposure to CSF-1, or CCL2, warrant further examination.
Activation of the BM-differentiated macrophages by LPS did not help dissect the basis of different migratory responses. In fact, differences in response to CSF-1 and CCL2 were similar regardless of the biochemical pathways activated by LPS. The possibility of other functional differences in macrophages differentiated from the BM of control- and UV-chimeric mice were examined; there were relatively small differences in cytokine and chemokine production (with or without LPS), which suggested that their autocrine activity was not at play. Differences in flux through activation pathways were accounted for in the completed microarrays that examined inherent differences in the macrophages when not stimulated and after LPS exposure. The accumulation of cells in the peritoneal cavity was also examined, and confirmed that the outcomes we were measuring were not an artifact of the in vitro chemotaxis assay with a semipermeable membrane. Instead, monocytes, macrophages, and other cells were being attracted to migrate into, and accumulate in, the peritoneal cavity.
Like BM-macrophages differentiated in vitro, monocyte-derived peritoneal macrophages in UV-chimeric mice inefficiently migrated toward the chemoattractants thioglycollate and alum. We had proposed that differences in metabolic programming may influence different migratory responses by macrophages. However, there were no differences in lactate production or metabolic flux through glycolytic or mitochondrial processes, the latter performed in the absence of CSF-1. Furthermore, LPS could stimulate significant glycolytic activity, stimulated lactate production, and control metabolic pathways in macrophages, but was unable to influence differences in migration capability between macrophages from nonirradiated and UV-irradiated mice and control- and UV-chimeric mice. In addition, BM-derived and peritoneal monocyte-derived macrophages have different metabolic phenotypes that include divergent mitochondrial responses to LPS activation (40), yet both types of macrophages have poor migratory capabilities if they have differentiated from the BM of UV-chimeric mice.
Macrophages, like all hematopoietic cells in the periphery of control- and UV-chimeric mice, did not appear to differ numerically. This suggests similar replacement of radiation-sensitive macrophage populations following BM transplantation. In our previous publications (6, 20), detailed patterns of engraftment also did not suggest time-dependent differences in engraftment. We are uncertain whether responses differ to physiological versus inflammatory chemoattractants (e.g., thioglycollate). We have also recorded reduced migration capacity by fully differentiated macrophages; it is unknown whether less differentiated macrophages from the BM of UV-irradiated and UV-chimeric mice also have reduced migration abilities.
In summary, macrophages differentiated from the BM of UV-irradiated and UV-chimeric mice, compared with those from nonirradiated or control-chimeric mice, have an inherent reduced ability to migrate toward chemoattractants and inflammatory agents. Furthermore, the effect of UV exposure is long lasting with a sustained reduction in macrophage migration toward chemoattractants for 12–15 wk in the UV-chimeric mice. Microarray analysis of nonstimulated and LPS-activated macrophages and Western blotting support involvement of reticulon-1, a confirmed surface membrane protein, in determining the differences in macrophage migration to CSF-1 and CCL2, and we propose thioglycollate and alum. The studies using immunofluorescence microscopy suggested that reticulon-1 moves around the cell and can translocate to the cell membrane in response to chemoattractants. Thus, future experiments should concentrate on the subcellular localization of reticulon-1, rather than measure levels of reticulon-1 in lysates from macrophages differentiated from the BM of both nonirradiated and control-chimeric mice, in direct comparison with those from UV-irradiated or UV-chimeric mice, respectively. Our current data (from gene array analysis, migration response of macrophages incubated with the anti–reticulon-1 Ab, and immunofluorescence microscopy) suggest that there is less reticulon-1 localized to the cell membrane of macrophages differentiated from the BM of UV-chimeric mice, although confirmatory experiments are required. From this study, we report that macrophages in UV-irradiated mice are indeed slow to collect at a site of inflammation, and thus pathways stimulated by UV irradiation of skin may help regulate macrophage-associated outcomes during infections and inflammatory challenges. Whether these outcomes contribute to UVR-induced immunosuppression in humans is a matter for further study. More broadly, reticulon-1, and pathways involved in determining its function, may become a new target for biochemical manipulation to regulate monocyte and macrophage migration capabilities.
We thank Deborah Strickland for assistance creating the chimeric mice and gratefully acknowledge Bright Blue for donation of the Nikon C2 confocal microscope.
This work was supported by the Cancer Council Western Australia (to P.H.H. and F.J.P.) and the National Health and Medical Research Council, Australia (Grants 572660 and 1067209 to P.H.H.).
The raw CEL files and processed data in this article have been submitted to the National Center for Biotechnology Information Gene Expression Omnibus (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE98840) under accession number GSE98840.
The online version of this article contains supplemental material.
The authors have no financial conflicts of interest.