The Chemokine Decoy Receptor M3 Blocks CC Chemokine Ligand 2 and CXC Chemokine Ligand 13 Function In Vivo

Mice

RIPCXCL13 mice were generated by microinjection of the RIPCXCL13 transgene into B6D2F₁ mouse eggs (The Jackson Laboratory) and transferred into oviducts of ICR foster mothers (Charles River Laboratories), according to published procedures (15). The RIPCXCL13 transgene was constructed as follows: a 3.45 kb, BstXI fragment of CXCL13 genomic DNA was isolated from a bacterial artificial chromosome clone and subcloned into the EcoRV/SalI sites of the rat insulin promoter (RIP)-TNF-α-pBS plasmid (16). TNF-α was released from RIP-TNF-α-pBS by EcoRV/SalI digestion, and replaced with the CXCL13 genomic DNA.

RIPCCL2 mice were generated using a transgene described in Grewal et al. (17) using conventional techniques (15). Five independent lines were established. RIPM3 transgenic mice were described previously (14).

Identification of the transgenic mice was accomplished by PCR amplification of mouse tail DNA. Specifically, the primers used for detection of the RIPCXCL13 transgene were: 5′-AAGACTCCAGGGATTTGAGG-3′ and 5′-GAGGCAGCTGGCCAGGAG-3′. For the RIPCCL2 transgene, the primers were: 5′-AAAGCTAGTGGGGATTCAGTAACCCCCAGC-3′ and 5′-GGCCCAGAAGCATGACAGGGACCTGC-3′. For the RIPM3 transgene, the primers were: 5′-ACCTGGTACGAGTTTTGGTGG-3′ and 5′-ATGGAGTGAGCAGAGACATCT-3′. The endogenous low density lipoprotein gene, used as an internal control, was amplified with the following primers: 5′-CAGCTCTACATCACCTGCCA-3′ and 5′-CGCAGTGCTCCTCATCTGACTTGT-3′. PCR conditions were as follows: 94°C, 30 s; 60°C, 30 s; 72°C, 60 s. All mice were housed under specific-pathogen-free conditions in individually ventilated cages at the Mount Sinai School of Medicine Animal Facility. All experiments were performed following institutional guidelines.

Histology

Tissues for light microscopic examination were fixed by immersion in 10% phosphate-buffered formalin and then processed for paraffin sections. Routinely, 5-μm sections were cut and stained with H&E. For immunostaining, fresh frozen sections were first fixed with ice-cold acetone for 20 min, dried, and stored at −20°C. Slides were incubated for 1 h at room temperature with purified primary Abs followed by incubation with the appropriate labeled secondary Abs for 30 min. Primary Abs used were anti-CD45 (no. 550539), CD3 (no. 550275), CD11c (no. 550283), B220 (no. 550286) and CD11b (no. 553308) from BD Biosciences/BD Pharmingen; anti-F4/80 (no. MCAP497) from Serotec; anti-CCL2 (AF479) and anti-CXCL13 (AF470) from R&D Systems; and guinea pig polyclonal anti-insulin (no. A0564) from DakoCytomation. Secondary Abs used were Alexa Fluor 488 and 594 goat anti-rat IgG (nos. A-11006 and A-11007), Alexa Fluor 488 goat anti-rabbit IgG (no. A-11034), and Alexa Fluor 594 donkey anti-goat IgG (no. A-11058) from Molecular Probes; and Cy3 goat anti-Armenian hamster (no. 127-165-160) and FITC anti-guinea pig (no. 706-095-148) from Jackson ImmunoResearch Laboratories.

To determine the degree of islet infiltration we analyzed histological sections of the pancreas of different animals. Sections were stained with H&E or with Abs against CD45 and insulin. Between 40 and 100 islets were examined for each mouse. Insulitis was scored as follows: no lesions; small or peri-insular leukocytic aggregates, usually periductal infiltrates; medium or moderate insulitis with mononuclear cells infiltrating <50% of the islet architecture; and large or severe insulitis with >50% of the islet tissue infiltrated by mononuclear cells. Data are presented as mean insulitis score ± SD for the indicated experimental group.

Isolation of pancreatic islets of Langerhans

Islets of Langerhans were isolated as previously described (18). Briefly, the common bile duct was clamped distal to the pancreatic duct junction at its hepatic insertion. The proximal common bile duct was then cannulated using a 27-gauge needle, and the pancreas was infused by retrograde injection of 2 ml of ice-cold collagenase solution (1.0 mg/ml; Sigma-Aldrich) in HBSS (Invitrogen Life Technologies). Pancreatic tissue was recovered and subjected to a 14-min digestion at 37°C. Subsequently, ice-cold HBSS was added and the suspension was vortexed at full speed for 10 s. Islets were hand-picked under a dissection microscope. Islets were used immediately after isolation to obtain RNA. In other experiments, islets were cultured for 24 h in RPMI 1640 supplemented with 10% FBS, 20 mM HEPES, 1% penicillin-streptomycin, 1% l-glutamine, and 2.5 ml of 7.5% NaHCO3 (all from Invitrogen Life Technologies) at 37°C, 5% CO₂. Supernatants and serum from transgenic mice were stored at −20°C until assayed. CCL2 and CXCL13 concentrations were measured by ELISA following the manufacturer’s guidelines (R&D Systems).

FACS analysis

Islets were isolated from different lines of transgenic mice as described above. Isolated islets were passed through a 40-μm diameter mesh to obtain a single-cell suspension. Islet cell suspensions were centrifuged at 1400 rpm for 5 min at 4°C and resuspended in FACS staining buffer (PBS containing 2% FCS and 0.01% sodium azide). Cells were incubated for 20 min at 4°C with 5 μg/ml Fc block (BD Pharmingen) and then stained with directly conjugated primary mAbs (BD Pharmingen). Samples were analyzed in a FACSCanto instrument (BD Biosciences). Data were analyzed using the FlowJo software (Tree Star).

Quantitative real-time PCR

Total RNA was extracted from islets using the RNeasy Mini Kit (Qiagen) according to the manufacturer’s instructions. Reverse transcription was performed from 2 μg of RNA. Quantitative real-time PCR was conducted in duplicates from 25 ng of cDNA and with 0.4 μM of each primer in a 30-μl final reaction volume of 1× SYBR Green PCR Master Mix (Applied Biosystems) using the ABI PRISM 7700 instrument. PCR cycling conditions were as follows: 1 cycle of 50°C for 2 min, 1 cycle of 95°C for 10 min, 40 cycles of 95°C for 15 s, and 1 cycle of 60°C for 1 min. Relative expression levels were calculated as 2^{cycle threshold (Ct) ubiquitin − Ct gene} (for details see ABI PRISM 7700 User Bulletin no. 2; Applied Biosystems) using ubiquitin RNA as endogenous control. The primers used were as follows: M3, forward GAGGGCCAGTCTCAAAATGTG and reverse GCTGACCCCAACCAGGTTAG; CCL2, forward GCTGGAGCATCCACGTGTT and reverse ATCTTGCTGGTGAATGAGTAGCA; CXCL13, forward CATAGATCGGATTCAAGTTACGCC and reverse TCTTGGTCCAGATCACAACTTCA; preproinsulin, forward GGGAGCAGGTG ACCTTCAGA and reverse TGATCTACAATGCCACGCTTCT; and ubiquitin, forward TGGCTATTAATTATTCGGTCTGCAT and reverse GCAAGTGGCTAGAGTGCAGAGTAA.

Statistical analysis

Data are expressed as means ± SD. Paired, unpaired Student’s t test and one-way ANOVA were used to determine statistical significance. Differences were considered significant when p < 0.05.

Expression of CCL2 by pancreatic β cells promotes migration of mononuclear cells into the islets of Langerhans

Eleven transgenic founders were generated by microinjection of the RIPCCL2 transgene (17) into fertilized mouse eggs. Five transgenic lines were established from these founders. All transgenic mice developed normally and were fertile. To examine whether CCL2 was produced by the transgenic islets, we measured CCL2 secreted in the supernatants of cultured pancreatic islets from transgenic mice and controls. As shown in Fig. 1⇓A, islets from control mice secreted small amounts of CCL2 (0.03 ± 0.006 ng/ml/islet). In contrast, RIPCCL2 transgenic islets secreted 20- (line 251) to 700-fold (line 1) more CCL2 than control islets (Fig. 1⇓A). We next asked whether the expression of CCL2 promoted recruitment of leukocytes into the islets of Langerhans. To this end, we analyzed leukocyte infiltration in the pancreas of RIPCCL2 and wild-type animals by immunostaining using an anti-CD45 Ab (Fig. 1⇓B). Few CD45⁺ cells were found in the islets of wild-type mice. In marked contrast, large number of CD45⁺ cells were found in pancreatic islets of RIPCCL2 mice. The most pronounced infiltrates were observed in animals of lines 1 and 254, which expressed the highest levels of CCL2 (Fig. 1⇓B).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Expression of CCL2 by β cells promotes migration of mononuclear cells to the islets. A, Isolated islets from mice of different RIPCCL2 lines and nontransgenic littermates were cultured overnight and supernatants were analyzed for CCL2 production by ELISA. Results represent three separate experiments (one-way ANOVA; ∗, p < 0.0001). B, Representative pictures of CD45 staining (green) in islets from pancreata of control and different RIPCCL2 lines (p = lines). C, H&E-stained paraffin section of wild-type pancreas and RIPCCL2 pancreas from lines 251 and 254. Representative picture of F4/80 (green) and CD11c (red) staining in islets from pancreata of control and RIPCCL2 mice from lines 251 and 254. Scale bars, 25 μm.

To study in detail the effect of CCL2 expression in islets of Langerhans, we expanded lines 251 and 254. Analysis of H&E-stained sections of transgenic pancreas showed that pancreata of animals from line 251 had a mild insulitis, with mononuclear cells predominantly accumulating in the periphery of the islets (Fig. 1⇑C, upper panel). Pancreata from mice in line 254 showed a severe insulitis, with mononuclear cells accumulating in the periphery and within the islets (Fig. 1⇑C, upper panel). The frequency and the size of these infiltrates did not change during the life of the animals (data not shown). Next, we phenotyped the leukocyte infiltrates using Abs against common surface markers. In agreement with Grewal et al. (17), most of the CD45⁺ cells in RIPCCL2 islets were monocytes or macrophages (Fig. 1⇑C, middle panel). Next in abundance in the infiltrates were CD11c-positive cells (dendritic cells (DCs); Fig. 1⇑C, bottom panel). There were also a few scattered T and B cells (data not shown). We observed a direct correlation between the levels of CCL2 expressed and the number of cells in the infiltrates. Islets from mice of line 254 showed more monocytes and DCs than islets from mice of line 251 and control islets (Fig. 1⇑, B and C and not shown). These results indicated that CCL2 promoted recruitment of monocytes and DCs into the islets and that this effect was dependent on the amount of CCL2 produced.

Expression of M3 in β cells blocks migration of monocytes and DC induced by CCL2 expression

To test the hypothesis that M3 expression blocks CCL2 function in vivo, we crossed RIPCCL2 mice from both lines with mice expressing M3 in the islets (RIPM3 mice) (14). Double transgenic mice are referred to as RIPM3/CCL2 mice.

We examined histological sections of pancreata of control, single transgenic mice (RIPCCL2) and double transgenic mice (RIPM3/CCL2) of lines 251 and 254. Semiquantitative analysis of infiltrated islets was performed in pancreata from adult mice (6–12 wk old). No infiltrates were found in islets of control or RIPM3 transgenic mice regardless of age (n = 8 each; data not shown). As expected, mononuclear infiltrates of varying sizes were found in the pancreatic islets of RIPCCL2 transgenic mice. 65 ± 10% of the islets of RIPCCL2 mice from line 251 (n = 10) had infiltrates (Fig. 2⇓A), mostly arranged around the islets (Fig. 2⇓B). In striking contrast, none of islets from RIPM3/CCL2 mice of the same line (n = 15) had infiltrates (Fig. 2⇓, A and C) resembling islets of control mice (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Expression of M3 in β cells blocks migration of mononuclear cells induced by CCL2. A, Semiquantitative analysis of islet infiltrates in pancreas from RIPCCL2 and RIPM3/CCL2 mice from line 251. B and C, Representative immunostaining for CD45 (red) and insulin (green) in pancreata of RIPCCL2 (B) and RIPM3/CCL2 (C) from line 251. D, Semiquantitative analysis of islet infiltrates in pancreas from RIPCCL2 and RIPM3/CCL2 mice from line 254. E and F, Representative immunostaining for CD45 (red) and insulin (green) in pancreata of RIPCCL2 (E) and RIPM3/CCL2 (F) from line 254. Scale bars, 25 μm.

To determine whether M3 could block leukocyte recruitment induced by higher amounts of CCL2, we analyzed pancreata from mice of line 254, which produced 15-fold more CCL2 than pancreata from mice of line 251. Semiquantitative analysis of infiltration in islets of Langerhans showed that 97 ± 2% of islets of RIPCCL2 mice and 96 ± 2% of the islets of RIPM3/CCL2 mice were infiltrated (n = 9 each; Fig. 2⇑D). In both cases, these infiltrates contained large clusters of CD45⁺ cells around the islets. But, in contrast to RIPCCL2 254 islets (Fig. 2⇑E), CD45⁺ cells were rarely found in the center of the RIPM3/CCL2 254 islets (Fig. 2⇑F).

Taken together, these results indicate that expression of M3 in pancreatic islets blocked the accumulation of mononuclear cells induced by the ectopic expression of CCL2, and that this effect was more pronounced in the line expressing lower levels of CCL2 in the pancreas.

Coexpression of M3 in β cells does not affect CCL2 production

A possible explanation for the results described above would be that expression of M3 negatively affected the expression of CCL2, thus accounting for reduced leukocyte recruitment. To test whether expression of M3 affected the expression of CCL2 and vice versa, we analyzed CCL2, M3, and preproinsulin mRNA expression in isolated islets from control, single transgenic mice, and RIPM3/CCL2 mice of both lines. All mice showed comparable levels of preproinsulin mRNA (Fig. 3⇓A), suggesting that the presence of the transgenes did not affect global islet function. No CCL2 mRNA expression was detected in freshly isolated islets from control and RIPM3 mice. As expected, mice from RIPCCL2 lines 251 and 254 expressed CCL2 mRNA at very different levels, but transcription of CCL2 was not affected by the introduction of the M3 transgene. There were also no differences between the levels of M3 mRNA detected in single or double transgenic islets (Fig. 3⇓A). These results indicate that the transcription of M3 and CCL2 genes is not affected in islets of double transgenic mice.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Coexpression of M3 in β cells does not affect CCL2 production. A, Relative expression of M3, CCL2, and preproinsulin mRNA in pancreatic islets of control mice, RIPM3, RIPCCL2, and RIPM3/CCL2 mice from both lines (normalized to ubiquitin; n = 2 experiments). B–D, Representative picture of immunostaining for insulin (blue, B), M3 (green, C), and CCL2 (red, D) in the pancreata of RIPM3/CCL2 transgenic mice. E, Western blot analysis of the serum from control, RIPM3, and RIPM3/CCL2 mice of both lines (n = 6) using M3 Ab. The results are representative of two separate experiments. F, CCL2 levels in serum from mice of RIPM3/CCL2 lines and control littermates (n = 12 mice in each group; error bars, SD). Scale bars, 25 μm.

We next examined whether M3 and CCL2 proteins were produced by double transgenic islets. Immunostaining of pancreata from animals from RIPM3/CCL2 line 251 showed that CCL2 and M3 proteins were produced by the islets of Langerhans (data not shown). Within the islets, M3 (Fig. 3⇑C) and CCL2 (Fig. 3⇑D) were expressed by insulin-producing β cells (Fig. 3⇑B). Both M3 and CCL2 proteins were expressed by the same cells (data not shown).

Increased production of M3 or reduced production of CCL2 protein could explain the decrease in leukocyte recruitment in double transgenic mice. To test whether coexpression of M3 and CCL2 affected their secretion, we measured their levels in serum. M3 was detected at equivalent amounts in the serum of RIPM3 or RIPM3/CCL2 mice, but not in the serum of control mice (Fig. 3⇑E). Similarly, CCL2 was detected in the serum of single- or double-positive mice, and at very low levels in the serum of controls (Fig. 3⇑F). Of note, the levels CCL2 were higher in the serum of double transgenic than in the single transgenic mice (Fig. 3⇑F).

Altogether, these results indicate that coexpression of M3 and CCL2 in the islets does not cause changes in the transcription, nor changes in protein production that could explain the reduced recruitment of leukocytes to the islets.

M3 blocks organized accumulation of lymphocytes promoted by expression of CXCL13 within the islets of Langerhans

Expression of CXCL13 in pancreatic islets promotes recruitment of B and T cells (19). To generate transgenic mice expressing CXCL13 in the pancreas, we injected a transgene encoding a genomic segment of CXCL13 under the control of the RIP. A total of six founders were generated from these injections and two transgenic lines were derived from them. CXCL13 expression was detected by immunostaining in the pancreatic islets of both transgenic lines (Fig. 4⇓A), but not in islets of control mice (data not shown). CXCL13 protein colocalized with insulin, indicating specific expression by β cells (Fig. 4⇓B). To confirm that CXCL13 could induce lymphocyte recruitment in vivo, we examined H&E-stained paraffin sections of transgenic pancreas at different ages. In agreement with the findings reported by Luther et al. (19), mononuclear infiltrates of varying sizes were found in the pancreatic islets of transgenic mice regardless of age. One transgenic line was expanded for the analysis shown in this study. In this RIPCXL13 line, ∼30 ± 14% of the islets were infiltrated by mononuclear cells (10 ± 9% of the islets had large infiltrates, 8 ± 5% medium, and 12 ± 7% small infiltrates) (Fig. 4⇓C).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Expression of M3 in β cells blocks development of lymphoid aggregates induced by CXCL13. A and B, Representative picture of CXCL13 (red, A) and insulin (green, B) immunostaining in the pancreata of RIPCXCL13 mice. C, Semiquantitative analysis of islet infiltrates in pancreas from RIPCXCL13 (n = 23) and RIPM3/CXCL13 (n = 26) mice (Student’s t test, ∗, p < 0.05 and ∗∗∗, p < 0.0001). D and E, Immunostaining for CD45 (red) and insulin (green) in pancreata of RIPCXCL13 (D) and RIPM3/CXCL13 mice (E). F, Flow cytometric analysis of isolated islets from control RIPM3, RIPCXCL13, and RIPM3/CXCL13 transgenic mice. Total numbers of CD45, CD19, and CD3 cells are shown (error bars, SD; Student’s t test, ∗, p < 0.05 and ∗∗, p < 0.001; four mice in each group). G, Relative expression of M3, CXCL13, and preproinsulin mRNA in pancreatic islets of control mice, RIPM3, RIPCXCL13, and RIPM3/CXCL13 transgenic mice (normalized to ubiquitin, n = 2 experiments). H, CXCL13 levels in serum from RIPCXCL13, RIPM3/CXCL13, RIPM3, and control littermates (n = 10 mice were analyzed in each line; error bars, SD). Scale bars, 25 μm.

To further characterize the cellular infiltrates, we performed immunohistochemical staining using Abs against several leukocyte cell surface makers. We observed that the majority of the infiltrating cells were B220⁺ B lymphocytes and CD3⁺ T cells. The infiltrates appeared to be organized into separate compartments with T cells at the center of the infiltrates. CD11c⁺ cells, and very small numbers of F4/80⁺ cells were also found among the cellular infiltrates (data not shown).

To evaluate the effect of M3 expression on the migration of cells induced by CXCL13 in vivo, we crossed RIPCXCL13 mice with RIPM3 mice to obtain RIPM3/CXCL13 mice that simultaneously expressed M3 and CXCL13 in the islets. Next, we evaluated H&E-stained paraffin sections of pancreata from control, single transgenic mice, and RIPM3/CXCL13 mice. No infiltrates were found in islets of control or RIPM3 transgenic mice regardless of age (n = 16 in each group; data not shown). As described above, mononuclear infiltrates of varying sizes were found in ∼30% of the pancreatic islets of RIPCXCL13 mice (n = 23). In marked contrast, mononuclear infiltrates were found in <10% of the RIPM3/CXCL13 islets, and the size of the infiltrates was significantly reduced (n = 26; Fig. 4⇑C). Furthermore, there were also changes in the distribution of the infiltrates. Mononuclear cells occupied the center of islets in RIPCXCL13 animals (Fig. 4⇑D), whereas they accumulated predominately near ducts or in the periphery of the islets of the RIPM3/CXCL13 animals (Fig. 4⇑E).

Flow cytometric analysis confirmed the histological findings. Control mice had on average 3 × 10³ CD45⁺ cells in the pancreatic islets, whereas the number of CD45⁺ cells from islets of RIPCXCL13 mice was increased 15-fold (45 × 10³ cells). Coexpression of M3 with CXCL13 reduced the number of CD45⁺ cells by 85% (6.5 × 10³ cells) (n = 4 mice in each group; Fig. 4⇑F). Despite the notable decrease in the number of leukocytes, the infiltrates in RIPM3/CXCL13 mice were still dominated by B cells (∼3 × 10³ cells, 7-fold fewer B cells than were found in RIPCXCL13 mice). CD3⁺ T cells represented a very small fraction of the infiltrates (0.7 × 10³ cells; Fig. 4⇑F).

To exclude the possibility that expression of both transgenes in islets affected the production of CXCL13 or M3, we measured the levels of preproinsulin, M3, and CXCL13 mRNA in the islets of control, RIPM3, RIPCXCL13, and RIPM3/CXCL13 mice by quantitative real-time PCR. As shown in Fig. 4⇑G, all mice showed equivalent levels of preproinsulin mRNA. CXCL13 was not expressed in isolated islets from control and RIPM3 mice. The levels of M3 and CXCL13 mRNA observed in the double transgenic animals were comparable to those detected in the single transgenic animals, indicating that transcription of both transgenes was not affected by their coexpression. To test whether CXCL13 and M3 proteins were released from the islets we measured CXCL13 and M3 levels in the serum of control, RIPM3, RIPCXCL13, and RIPM3/CXCL13 mice. CXCL13 levels were higher in RIPCXCL13 and RIPM3/CXCL13 mice than in controls or RM3 mice. The levels of CXCL13 in RIPCXCL13 and RIPM3/CXCL13 mice were not statistically different (Fig. 4⇑H). M3 was also detected in the serum of transgenic mice (data not shown), indicating that expression of M3 did not reduce the production of CXCL13 by the islets or vice versa.

These results indicate that expression of CXCL13 in islets of Langerhans is sufficient to promote recruitment and topological segregation of B and T cells, and that these activities can be significantly blunted by coexpression of M3 by the islets.