Galanin is a potent modulator of cytokine and chemokine expression in human macrophages

The regulatory peptide galanin is broadly distributed in the central- and peripheral nervous systems as well as in non-neuronal tissues, where it exerts its diverse physiological functions via three G-protein-coupled receptors (GAL1-3-R). Regulatory peptides are important mediators of the cross-communication between the nervous- and immune systems and have emerged as a focus of new therapeutics for a variety of inflammatory diseases. Studies on inflammatory animal models and immune cells revealed both pro- and anti-inflammatory functions of galanin. Here, we probed specific immune-related functions of the galanin system and found galanin and GAL1-R and GAL2-R mRNA to be expressed in a range of human immune cells. In particular, macrophages displayed differentiation- and polarization-dependent expression of galanin and its receptors. Exposure to exogenous galanin affected the cytokine/chemokine expression profile of macrophages differently, depending on their differentiation and polarization, and mainly modulated the expression of chemokines (CCL2, CCL3, CCL5 and CXCL8) and anti-inflammatory cytokines (TGF-β, IL-10 and IL-1Ra), especially in type-1 macrophages. Cytokine/chemokine expression levels in interferon-gamma- and lipopolysaccharide-polarized macrophages were upregulated whereas in unpolarized macrophages they were downregulated upon galanin treatment for 20 hours. This study illuminates the regulation of important cytokines/chemokines in macrophages by galanin, depending on specific cell activation.


Results expression and regulation of the galanin system in immune cells.
Recently, we reported that peripheral monocytes express moderate levels of galanin, which increased upon GM-CSF-induced differentiation of monocytes to macrophages (M0-GM-Mϕ) 36 , whereas we detected no galanin expression in PMNs and NK cells 36,37 . In the present study, we extended this analysis to human B cells and CD4 + and CD8 + T cells, and all were found to express galanin at levels similar to monocytes (Table 1). Interestingly, all immune cells showed low mRNA levels of spexin, whereas no GALP expression was observed ( Table 1).
The galanin system comprises three receptors, GAL 1-3 -R, which exhibit different binding affinities to galanin ligands and are linked to diverse inflammatory processes. Therefore, we screened diverse immune cells for their galanin receptor expression. We previously showed that PMNs, NK cells and monocytes express GAL 2 -R but not GAL 1 -R or GAL 3 -R 36,37 . In the present study, we observed that human B cells and CD4 + /CD8 + T cells also express GAL 2 -R but not GAL 1 -R or GAL 3 -R (Table 1). Furthermore, M0-GM-Mϕ and M0-M-Mϕ showed consistent GAL 1 -R and GAL 2 -R expression and sporadically low levels of GAL 3 -R (Table 1, Fig. 2). A more detailed analysis of macrophage subtypes revealed that GAL 2 -R expression is significantly lower in polarized macrophages (M1-GM-Mϕ, M2a-M-Mϕ and M2c-M-Mϕ) compared to M0-GM-Mϕ (Fig. 2). We observed no significant difference of GAL 1 -R expression in macrophage subtypes (Fig. 2). GAL 3 -R was only sporadically expressed in four of 10 donors at very low levels (Fig. 2).
To evaluate the in vivo relevance of GAL 1 -R and GAL 2 -R expression on macrophages we determined galanin receptor protein expression in human xanthelasma deposits by immunohistochemistry. A xanthelasma, also referred as xanthoma, is a cluster of foam cells in the connective tissue of the skin. The foam cells are formed from macrophages accumulating lipids by phagocytosis 45 . We found membrane-associated GAL 1 -R (Fig. 4a) as well as GAL 2 -R ( Fig. 4b) staining on macrophages in the xanthelasma deposits. In agreement with immunofluorescence microscopy, not all but some macrophages were positive for receptor expression. Interestingly, the more membrane-associated staining of GAL 1 -R in the xantelasma macrophages is similar to the staining observed in the monocyte derived M0-GM-Mϕ. Whereas the vesicular staining of GAL 2 -R in the xantelasma macrophages is similar to the staining observed in the monocyte derived M0-M-Mϕ.
Galanin modulates cytokine/chemokine expression in macrophages. As galanin is secreted by macrophages and because monocytes/macrophages express galanin receptors, we investigated the effect of galanin on the cytokine/chemokine profile of macrophages during differentiation and polarization. To examine the impact of galanin on macrophage differentiation and subsequent polarization, we differentiated monocytes into macrophages with GM-CSF or macrophage colony-stimulating factor (M-CSF) in the presence or absence of 10 nM galanin. Afterwards, the differentiation factors were removed and the macrophages were either left untreated, treated with 10 nM galanin (for 20 hours) or polarized with appropriate inducers (IFNγ + LPS or IL-4 or IL-10) plus 10 nM galanin (or no galanin) for 20 hours. The expression data of cytokines/chemokines were analyzed in terms of: 1) differentiation of monocytes in the presence or absence of galanin, and 2) polarization of macrophages in the presence or absence of galanin (Fig. 5). Fold changes of cytokine/chemokine mRNA expression levels after exposure to galanin during differentiation and polarization compared to control samples without galanin treatment are summarized in an expression heat map (Fig. 6).
Macrophages differentiated for 6 days with GM-CSF in the presence of galanin (M0-GM + GAL-Mϕ) and cultured for further 20 hours without any additional treatment (unpolarized) showed similar levels of expression of a range of cytokines and chemokines as unpolarized macrophages differentiated without galanin (M0-GM-Mϕ) (Figs 5, 6, Supplementary Table S1). Only IL-18 mRNA levels differed significantly, showing a 1.2-fold reduction in the macrophages exposed to galanin during differentiation (p = 0.0106) (  Galanin modulates chemokine secretion by macrophages. As IL-10, TGF-β, CCL2, CCL3 and CXCL8 mRNA expression in GM-CSF-differentiated macrophages was highly affected by galanin treatment, we www.nature.com/scientificreports www.nature.com/scientificreports/ analyzed if galanin is also able to stimulate the secretion of these cytokines and chemokines into cell culture supernatants. TGF-β levels were under the detection limit (data not shown). We did not observe a difference in secreted IL-10 and CCL2 levels in M0-GM + GAL-Mϕ upon galanin treatment for 20 hours (Fig. 7a,b). Interestingly, galanin stimulation resulted in increased CCL3 and CXCL8 levels in the supernatant of M0-GM + GAL-Mϕ (Fig. 7c,d), although mRNA levels were reduced after 20 hours.

Discussion
Receptors of diverse regulatory peptides such as CRH, Calcitonin gene-related peptide (CGRP), SP, α-Melanocyte-stimulating hormone (α-MSH), VIP, NPY and others were reported to be expressed by various immune cells 46 . Furthermore, nerve endings and different immune cells in primary and secondary lymph nodes release regulatory peptides 46,47 . These findings imply an essential role of regulatory peptides in controlling www.nature.com/scientificreports www.nature.com/scientificreports/ immune cell functions. Thus, diverse effects of regulatory peptides on immune cell migration, proliferation, maturation and cytokine/chemokine regulation have been described for PMNs, dendritic cells, monocytes, macrophages, NK cells and T cells [46][47][48] . Expression of the galanin peptide system parallels that of other regulatory www.nature.com/scientificreports www.nature.com/scientificreports/ peptides, in that leukocytes can serve as both a source and acceptor of galanin receptor ligands. Because macrophages secrete full-length galanin and express galanin receptors, an autocrine signaling mechanism can be assumed. Accordingly, immunocytes positive for EP1 (a marker for monocytes/macrophages) were shown to produce increased amounts of galanin during inflammation in rats 24 . In addition, galanin might also act as a paracrine factor, as various immune cells, including important macrophage-interacting cells like CD4 + T cells, express GAL 2 -R. GAL 2 -R signaling leads to activation of PKC, and several isoforms of PKC were reported to modulate T cell activation, polarization and survival, and also affect B cell and NK cell functions 18 . GAL 2 -R is expressed by diverse immune cells; nevertheless, subtype expression analysis of monocytes, B cells and T cells needs to be performed in future studies. As galanin affected the expression profile of macrophages in a subtype-specific manner and activation of PKC modulates T cell activation and polarization, we assume that a specific activation and/or T cell polarization could be facilitated or inhibited by galanin via GAL 2 -R. Similarly, a galanin-GAL 2 -R interaction  Fig. S1). Values are colored when a main factor or an interaction of the main factors (differentiation or polarization) was significant by two-way ANOVA. www.nature.com/scientificreports www.nature.com/scientificreports/ on B cells could lead to a specific cell activation. The expression of specific galanin receptors and the combination of them on different types of immune cells might enable galanin to exert its effects in an immune cell type-and activation/tissue type-specific manner. However, expression of galanin, GAL 1 -R and GAL 3 -R (but not GAL 2 -R) and involvement of the galanin peptide in proliferation and apoptosis of immature rat thymocytes have already been reported 49 . An observed difference in receptor expression between immature rat thymocytes and human T cells isolated from peripheral blood 49 may be due to species-specific differences or be related to the different developmental stages of the immune cells.
Macrophages expressed GAL 1 -R and GAL 2 -R, albeit only a part of M0-GM-Mϕ showed membrane-associated GAL 1 -R and even less cells GAL 2 -R protein expression. The only partially membrane-associated GAL 1 -R and GAL 2 -R positive cells can be explained by a receptor internalization. It has been demonstrated that especially GAL 2 -R can be internalized upon ligand binding 4,50,51 . As macrophages already secrete high amounts of galanin into the supernatant galanin-receptor interaction could be followed by receptor internalization. In addition, GAL 1 -R and GAL 2 -R positive macrophages could have different characteristics as receptor negative cells, which has to be clarified in future studies. M0-M-Mϕ did not reveal a clear GAL 1 -R positive protein expression; however, the cells exhibited an intracellular GAL 2 -R staining, which was also observed in GAL 2 -R overexpressing neuroblastoma cells 44 and PC12 cells 51 . The lack of membrane-associated galanin receptors could explain the moderate response of M0-M-Mϕ upon exogenous administered galanin. In vivo relevance of galanin receptor expression on macrophages was demonstrated by immunohistochemical staining of GAL 1 -R and GAL 2 -R on macrophages in human xanthelasma of the skin. These cholesterol deposits consist of foam cell-forming macrophages and it was reported that lipid oxidation inside the foam cells induced M-CSF synthesis 45 .
We cannot exclude that only a small population of macrophages were stimulated by galanin. However, the stimulation of this subgroup might have resulted in the secretion of cytokines and chemokines which then further stimulated neighboring cells.
A co-expression of both galanin receptors could result in a more flexible signaling cascade, as these receptors were reported to assemble into both homo-and heterodimers 52 . Interestingly, macrophage subtypes differed in the proportion of GAL 1 -R and GAL 2 -R expression and thus could have distinct signaling upon galanin exposure. GAL 1 -R and GAL 3 -R activation leads to inhibition of the second messenger cAMP and inactivation of PKA 4 , which can influence macrophage polarization [19][20][21][22] .
The importance of galanin ligand-receptor-dependent communication within inflammatory processes was previously demonstrated in diverse inflammatory animal models, including IBD, arthritis, dermatitis, psoriasis and pancreatitis [28][29][30][31][32][33][34] . Interestingly, the findings reveal a complexity in how the galanin system affects immune responses, in that galanin and its receptors were reported to exhibit both pro-and anti-inflammatory properties [28][29][30][31][32][33][34] . These biphasic properties might depend on the type and duration of inflammation and consequently the types of immune cells involved. The microenvironment or cytokine/chemokine milieu might also play a role. However, the impact of galanin on immune cell functions, especially on specific subsets, has received little attention up to now.
To our knowledge, the present study provides the first insight into how galanin affects diverse macrophage subtypes differently. Galanin was downregulated in macrophages by IFNγ (an important effector cytokine of T h 1 responses) and LPS (M1-GM-Mϕ), whereas IL-4 (M2a-M-Mϕ), a main effector of T h 2 responses, had no effect on galanin production by macrophages. Furthermore, these polarizations revealed different relative involvements of GAL 1 -R and GAL 2 -R in type-1 (M1-GM-Mϕ) and type-2 (M2a-M-Mϕ and M2c-M-Mϕ) immune responses. This observation needs to be studied in more detail in vivo when considering galanin receptors as targets for therapeutic strategies in inflammatory diseases. The M2c subtype, in turn, showed similar levels of GAL 1 -R/GAL 2 -R and downregulation of galanin secretion, indicating fine-tuning of the galanin system during an immune challenge, especially as an immune response is not a static process. Interestingly, unpolarized type-2 macrophages (M0-M-Mϕ) where unaffected by galanin stimulation though they expressed both GAL 1 -R and GAL 2 -R. Thus, an extra stimulus might be needed to change galanin functions, similar to what we observed in NK cells and PMNs 36,37 . Furthermore, secretion of chemokines CCL3 and CXCL8 was increased upon galanin treatment of unpolarized type-1 macrophages (M0-GM-Mϕ), although gene expression after 20 hours was significantly downregulated. This indicates an initial boost of CCL3 and CXCL8 expression and downregulation at a later phase. Polarization of macrophages with IFNγ and LPS in the presence of galanin (M1-GM + GAL-Mϕ) increased the gene expression of chemokines, including CCL3 and CXCL8, compared to the control polarized without galanin, which leads us to speculate that macrophages require a specific immune challenge to maintain their pro-inflammatory properties. The finding of increased CXCL8 production upon galanin stimulation was already reported in human keratinocytes 53 . CXCL8 mainly targets granulocytes like neutrophils or basophils via CXCR1 and CXCR2. CCL3 interact with CCR1 or CCR5 expressing cells including monocytes/macrophages, dendritic cells, granulocytes, NK cells but also subpopulations of T cells and B cells. Following, the augmented secretion of the chemokines CCL3 and CXCL8 by macrophages upon the activation of galanin receptor signaling they potentially influence leucocyte migration or perhaps other chemokine related functions 54 . Previously we observed biphasic properties of galanin affecting PMNs and NK cells 36,37 . Galanin was able to sensitize and desensitize murine PMNs toward CXCL1 depending on the galanin concentration 37 .
Thus, the pro-or anti-inflammatory properties of the galanin system depend on many factors, which have to be considered in future in vitro and in vivo studies. Alteration of chemokine expression and secretion by galanin indicates a role of galanin in immune cell migration, which was also reported for other regulatory peptides, such as the ability of SP to promote monocyte migration upon induction of chemokine expression 55 . Furthermore, similar to galanin, diverse regulatory peptides like SP, NPY or VIP can alter cytokine expression and secretion by immune cells 46 . SP was shown to upregulate TNFα secretion by mast cells 56 and IL-1 and IL-6 secretion by monocytes 57 , and VIP was reported to exert an anti-inflammatory effect, inhibiting IL-12 secretion and IL-10 induction by LPS-induced macrophages 46 . NPY was shown to induce cell migration and secretion of (2019) 9:7237 | https://doi.org/10.1038/s41598-019-43704-7 www.nature.com/scientificreports www.nature.com/scientificreports/ IL-6 and IL-10 by monocyte-derived immature dendritic cells. Moreover, NPY promoted T h 2 cell polarization 58 . Interestingly, CGRP was shown to modulate macrophage polarization by counter-regulating the expression of IL-1β and expression/activity of NACHT, LRR and PYD domain-containing protein 3 (NALP3), a marker for classic LPS-activated type-1 macrophages. Furthermore, CGRP and IL-4 co-stimulated the expression of IL-10, resistin-like beta (Fizz1) and mannose receptor C-type 1 (CD206) as well as the expression/activation of arginase 1, a marker for an alternative mode of type-2 macrophage activation, compared to IL-4 stimulation alone 59 . Similarly, SP was reported to support type-2 macrophage polarization 60 .
We conclude that the galanin system plays an important role in diverse immunoregulatory mechanisms, spanning from the early immune response to the later phase of inflammation, as we showed that galanin and its receptors are expressed and regulated in cells of both the innate and adaptive immune systems. Furthermore, galanin might affect immune cell migration, as chemokine expression and secretion by macrophages were clearly altered upon galanin treatment. In addition, galanin could modulate cell polarization, because it effected a more potent change in the cytokine/chemokine profile of type-1 macrophages than type-2 macrophages. Our study establishes that galanin is a potent modulator of cytokine and chemokine expression and thus a player in fine-tuning the macrophage-driven immune response. Immune cell isolation and differentiation. All immune cells were isolated from surplus buffy coat material from healthy donors. Buffy coats were generated and preselected as recently described 36 . Peripheral blood mononuclear cells (PBMCs) were isolated by Ficoll-Paque 1.077 g/l (GE Healthcare, Chicago, IL, USA) gradient centrifugation. Human PMNs and NK cells were isolated as described recently 36,37 . CD3 + T cells were positively selected from PBMCs using an anti-CD3-FITC and anti-FITC MultiSort Kit (Miltenyi Biotec, Bergisch Gladbach, Germany) following the manufacturer's instructions. After removal of MultiSort MicroBeads, CD4 + and CD8 + T cells were separated using CD4 MicroBeads (Miltenyi Biotec). B cells were positively selected from PBMCs using CD19 MicroBreads (Miltenyi Biotec). CD14 + monocytes were isolated by using the Human Pan Monocytes Isolation Kit (Miltenyi Biotec).
Determination of in vivo galanin receptor protein expression by immunohistochemistry. Formalinfixed paraffin-embedded (FFPE) tissue blocks of a skin biopsies of a patient with a xantelasma were analyzed for GAL 1 -R and GAL 2 -R by immunohistochemistry. All reagents used in immunohistochemistry were obtained from Leica (Wetzlar, Germany). FFPE tissues were sectioned to 5 μm, mounted and dried at 60 °C for 1 h. After deparaffinization and rehydration, heat-induced epitope retrieval was performed with BOND Epitope Retrieval solution for 20 min at 100 °C. Primary antibodies against human GAL 1 -R (GTX108207, 1:200) and GAL 2 -R (customized: S4510-1, 1:200 were diluted in BOND Primary Antibody Diluent. Slides were incubated with the primary antibody for 15 min at RT. After washing with Wash Solution, the Novolink Max polymer (second antibody) was applied for 30 min at RT. Another washing step was followed by visualization with Mixed Red Refine for 10 min at RT followed by another 5 min at RT. Slides were washed in deionized water and counterstained in Mayer's hemalum solution (VWR, Radnor, PA, USA) for 10 min at RT. Slides were washed in deionized water, wash solution and 96% ethanol and after dehydration with 2-propanol, the slides were incubated in xylene and mounted. Digital micrographs were taken with an Olympus BX43 (Olympus Corporation, Tokyo, Japan).

Determination of galanin peptide secretion by macrophage subtypes using an in-house eLIsA kit.
Complete mini protease inhibitor cocktail (Roche, Basel Switzerland) was added to cell culture supernatants before storage at −80 °C. Cell supernatants were analyzed for IL-10, TGF-β, CCL2, CCL3 and CXCL8 concentrations using specific ELISA kits (Thermo Fischer Scientific). For detection of secreted full-length galanin in supernatants of macrophages, an in-house galanin sandwich ELISA was used as described recently 43 . statistical analysis. Statistical analyses were performed by using GraphPad Prism 7 (GraphPad Software, Inc., San Diego, CA, USA). Significances of galanin secretion and expression were calculated for GM-CSF-differentiated macrophages with a paired t-test and M-CSF-differentiated macrophages with a matched one-way ANOVA followed by Tukey's multiple comparison test. Correlation between galanin secretion and mRNA levels was computed using Pearson correlation coefficients. Statistical analysis of galanin receptor expression in macrophage subtypes was performed by using a two-way ANOVA (Experimental design: repeated measures by both factors; receptor and cell subtype) followed by a Tukey multi comparison test. Cytokine and chemokine expression data sets were analyzed by using a two-way ANOVA (Experimental design: repeated measures by both factors; differentiation and polarization) followed by a Sidak multi comparison test when the interaction of main factors was significant. Data sets generated from ELISA were tested for Gaussian distribution using the D' Agostino-Pearson omnibus normality test and further analyzed for significance with a paired t test (if the values passed the normality test) or a Wilcoxon matched-pairs signed rank test (if the values did not pass the normality test).

Data Availability
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.