Published online before print February 19, 2008
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||

* Department of Immunology, Erasmus MC, Rotterdam, The Netherlands; and
Division of Gastroenterology and Hepatopancreatology, Erasme Hospital, Brussels, Belgium
1Correspondence: Dept. of Immunology, Erasmus MC, Dr. Molewaterplein 50, 3015 GE Rotterdam, The Netherlands. E-mail: m.wildenberg{at}erasmusmc.nl
|
|
|---|
Key Words: chemokine receptor Sjögrens syndrome autoimmunity regulation
|
|
|---|
DC are potent APC capable of stimulating naïve and memory T cells and boosting B cell responses, thus playing a key role in the initiation of the immune response [3 , 4 ]. Furthermore, DC are crucial to the development of tolerance, preventing harmful immune responses against harmless antigens [5 , 6 ]. To fully exert their sentinel function, DC migrate from the blood into tissues and further to the draining lymph nodes. This migration is a complex, multi-step process, tightly regulated by numerous molecules including chemokines and their receptors [7 , 8 ].
The chemokine family contains a large number of proteins, characterized by their small size and conserved cysteine residues [9 ]. Most chemokines bind and signal through several receptors, and one receptor usually has multiple ligands [9 , 10 ], creating a highly complex network of interactions. Chemokine receptors are G-protein-coupled 7-transmembrane receptors, which have differential expression patterns on various leukocyte subsets [11 ]. On DC, the expression of chemokine receptors is linked to their state of maturation. For example, immature DC express CXCR4 and CCR1, -2, and -5, most of which are down-regulated after exposure to an activating stimulus [12 , 13 ]. Besides cell trafficking, chemokine receptors have been described to regulate several other processes in DC, including endocytosis, dendrite formation, cell survival, and maturation of DC [14 15 16 17 18 ].
Given the importance of DC in the immune response, the expression of chemokine receptors in the DC accumulating in the NOD salivary glands was determined. Most strikingly, a complete lack of CCR5 expression was observed in these cells. Earlier, it has been shown that mice deficient in CCR5 display excessive immune responses, for example, in an influenza model [19 ]. Therefore, we hypothesized that the lack of CCR5 results in a proinflammatory environment. The results indeed show an immunoregulatory role for CCR5 in DC, both by scavenging of inflammatory ligands and by regulating the secretion of IL-12.
|
|
|---|
Human materials
Heparinized blood and serum samples were obtained from female pSjS patients (mean age 52.7±16.4; mean disease duration 6.6±7.8 years) at Erasmus MC (Rotterdam, The Netherlands). Inclusion criteria consisted of a positive diagnosis, according to the criteria of the American-European consensus group [20
], and the absence of other rheumatic or autoimmune disorders. Control blood samples were obtained from female volunteers (mean age 45.8±6.1) not suffering from autoimmune diseases or taking immunosuppressive medication. The Medical Ethical Committee of Erasmus MC approved this study, and all subjects gave informed consent.
PBMC were isolated from peripheral blood using a Ficoll gradient (Amersham, Little Chalfont, UK).
Isolation of CD11c+ cells from tissue
Organs were removed and cleared of adipose tissue and lymph nodes. Tissue was digested by Liberase (Roche, Woerden, The Netherlands) treatment at 37°C and washed twice in PBS containing 10% FCS. Subsequently, cells were cleared from debris using a Ficoll (Amersham Biosciences, Uppsala, Sweden) gradient. Cells appearing at the interface were incubated with mouse CD11c microbeads (Miltenyi, Bergisch Gladbach, Germany), according to the manufacturers protocol and isolated using the AutoMACS (Miltenyi) system.
Flow cytometry
Cells were resuspended in PBS containing 1% BSA and incubated with primary antibodies, which were biotinylated MHC class II (C57BL/6 clone ER-TR3; NOD clone 10.2.16), biotinylated CD11c, CD8
-FITC, CD80-PE, CD86-FITC, CD40-PE, CXCR4-PE, and CCR5-PE (all BD Biosciences, San Jose, CA, USA). For human material, CCR5-PE-Cy5 (BD Biosciences) was used. Afterwards, cells were washed twice and when appropriate, incubated with streptavidin-allophycocyanin (BD Biosciences). For intracellular CCR5 staining, surface CCR5 was blocked by incubation with a tenfold excess of unlabeled anti-CCR5 antibody (BD Biosciences). Next, cells were fixed using 2% paraformaldehyde and permeabilized using 0.5% saponin. Cells were then incubated with labeled CCR5 antibody diluted in 0.5% saponin, washed, and resuspended in 1% BSA. Cell suspensions were analyzed using a FACSCalibur (Becton Dickenson, San Diego, CA, USA) and WinMDI software (Microsoft Co., Redmont, WA, USA). Expression was calculated as mean fluorescence intensity (MFI) specific staining – MFI isotype control. For calculation of relative expression levels, expression of C57Bl/6 was set to 1, and NOD levels were calculated accordingly.
Cytokine determination
Levels of IL-12p40 in supernatant were determined using an IL-12p40 Duoset kit (R&D Systems Inc., Minneapolis, MN, USA), and serum levels of IL-12 were determined using a human cytokine 25-plex AB bead kit (Biosource, Camarillo, CA, USA), according to the manufacturers suggestions. Induction of IL-12 was calculated as secretion after LPS stimulation divided by secretion without stimulus.
Bone marrow-derived DC (BMDC) culture
BM was isolated from femur and cultured in RPMI-1640 medium (BioWhittaker, Verviers, Belgium), supplemented with 10% FCS (BioWhittaker), 50 µM β-ME (Sigma Aldrich, Deisenhofen, Germany), and 20 ng/mL recombinant mouse GM-CSF (Biosource) for 9 days. To enrich immature DC, cells were incubated with anti-CD86 antibody (clone GL-1), washed, labeled with goat anti-rat microbeads (Miltenyi), and separated using the AutoMACS (Miltenyi) system. For stimulation experiments, BMDC were incubated with Met-RANTES (10 µg/mL, R&D Systems Inc.), where appropriate, and stimulated using LPS (5 ng/mL, Sigma Aldrich) for 16 h, after which supernatants were collected.
RNA isolation and RT-PCR
RNA was isolated from submandibular glands (SMG) using the RNeasy Mini kit (Qiagen, Valencia, CA, USA), according to the manufacturers suggestions. cDNA was generated using Superscript II (Invitrogen, Carlsbad, CA, USA) and random hexamer primers (Amersham). RT-PCR analysis was carried out using predesigned primer/probe sets (Applied Biosystems, Foster City, CA, USA), according to the manufacturers suggestions. For calculation of relative expression, all samples were normalized against expression of the household gene Abl.
Statistics
Data were analyzed using Students t-test, except for relative data (C57BL/6 set to 1), where the Mann Whitney U-test was used. Error bars represent SD. Correlations were calculated using Spearmans Rank test. All statistical analyses were performed using SPSS software (SPSS Inc., Chicago, IL, USA) and were considered significant if P < 0.05.
|
|
|---|
![]() View larger version (19K): [in a new window] |
Figure 1. Chemokine receptor and activation markers on SMG CD11c+ cells, which were isolated from SMG by Ficoll gradient followed by magnetic bead isolation. Cells were then stained for surface markers and analyzed by flow cytometry. (A) Gray lines represent NOD mice; black lines represent C57BL/6 mice; one representative sample of each strain is shown. (B) Average expression in C57BL/6 mice was set to 1, and expression levels of NOD mice were calculated accordingly (n=4–5); error bars represent SD; *, P < 0.05. (C) CD11c+ cells were isolated from SMG of mice at 5 or 15 weeks of age. Expression of CCR5 was calculated as MFI specific staining – MFI control.
|
Decreased CCR5 expression in NOD DC derived from other sources
To determine if the lack of CCR5 expression is tissue-specific or a general phenomenon in NOD mice, CD11c+ cells from other in vitro and in vivo sources were studied. CD11c+/autofluorescencelow DC derived from the lungs of NOD mice showed a clearly decreased expression of CCR5 when compared with control mice (Fig. 2A
). Similarly, DC cultured from BM (BMDC) in the presence of GM-CSF showed a significantly decreased expression of CCR5 (Fig. 2B)
.
![]() View larger version (25K): [in a new window] |
Figure 2. CCR5 expression on lung and BM CD11c+ cells, which were isolated from lungs of C57BL/6 and NOD mice (A) or generated in vitro from BM (B) and analyzed by flow cytometry. Results of gated autofluorescentlow cells (A) or gated CD11c+ cells (B) are shown; black lines represent specific staining; gray lines represent isotype control. One representative image for each strain is shown. Bars represent average expression; error bars represent SD. (C, D) CD11c+ cells were isolated from salivary glands (C) or cultured from BM (D) and analyzed for intracellular expression of CCR5 by flow cytometry. Average expression in C57BL/6 mice was set to 1, and expression levels of NOD mice were calculated accordingly (n=3); error bars represent SD; *, P < 0.05.
|
Lack of CCR5 receptor expression may lead to increased free ligand
The lack of CCR5 expression on the surface of CD11c+ cells in the SMG of NOD mice may influence biologically available levels of CCR5 ligands CCL3, CCL4, and CCL5. When measured in total salivary gland, mRNA levels of these chemokines tended to be somewhat higher in NOD than in control mice (Fig. 3A
). Furthermore, in control mice, these chemokines can be scavenged from the environment by surface CCR5 on tissue CD11c+ cells, thus decreasing the level of free, active ligand. This results in a balanced ratio between CCR5 ligands and CCR5 itself. As a result of the lack of surface CCR5 expression, this mechanism is not available in NOD mice, which would be indicated by an increased CCR5 ligand:CCR5 ratio. Indeed, ratios of CCL3, CCL4, and CCL5 relative to the level of CCR5 were increased significantly in the NOD salivary gland (Fig. 3B)
.
![]() View larger version (31K): [in a new window] |
Figure 3. Ratio of CCR5 ligands:CCR5 is increased in NOD SMG. Total mRNA was isolated from SMG, and expression levels of CCL3, CCL4, CCL5, and CCR5 were determined by RT-PCR (n=5). Error bars represent SD; *, P < 0.05.
|
![]() View larger version (10K): [in a new window] |
Figure 4. IL-12 secretion in the absence of CCR5 signaling. (A) BMDC were generated in the presence of GM-CSF from wild-type (WT) and CCR5–/– animals. Cells were then stimulated using LPS (5 ng/mL), and IL-12p40 was measured in the supernatant by ELISA (n=6). Error bars represent SD. (B) BMDC were generated in the presence of GM-CSF from wild-type animals, treated with met-RANTES (met), and stimulated using LPS (5 ng/mL). IL-12p40 levels in the supernatants were measured by ELISA (n=3). Bars represent mean values; error bars represent SD.
|
![]() View larger version (16K): [in a new window] |
Figure 5. Decreased CCR5 expression in pSjS patients. (A) PBMC were isolated from pSjS patients and healthy controls and analyzed by flow cytometry. CCR5 expression was determined as MFI-specific staining – MFI isotype control. Each dot represents one individual patient; *, P < 0.01 (B) Serum levels of IL-12p40 were determined by the AB bead kit and correlated to the monocyte expression of CCR5, as determined by flow cytometry. Each dot represents one individual patient; *, P < 0.05.
|
|
|
|---|
CCR5 was initially thought to be involved in directing immature DC to tissue sites [13 ]. Surprisingly, in the steady-state, CCR5–/– mice develop normally and show cell numbers and lymphoid architecture similar to wild-type animals [25 ]. Furthermore, several reports show normal macrophage recruitment in CCR5–/– mice during inflammation [19 , 25 26 27 ]. These findings may be explained in light of the high redundancy in the chemokine system. In CCR5-deficient cells, migration is probably regulated by other chemokine molecules, such as CCR1 and CCR2. RNA levels of these receptors were comparable in NOD and control SMG (data not shown), which may explain the presence of DC in NOD SMG, despite the lack of CCR5.
In accordance with previous studies in CCR5–/– mice [19 , 26 , 28 ], mRNA of CCR5 ligands tended to be increased in NOD SMG. In addition, the lack of expression of CCR5 in DC isolated from NOD SMG leads to a lack of scavenging by this receptor, thus contributing to increased amounts of free, biologically available ligands. Such scavenging by CCR5 has been reported to play a role in the control as well as the resolution of immune responses [29 , 30 ]. The increased availability of free CCL3, -4, and -5 may lead to an enhanced influx of other leukocytes, such as T cell and macrophages, as is seen in the salivary glands of NOD mice. One could argue that the increased levels of CCR5 ligand are causative rather than the result of decreased expression of CCR5 as a result of receptor internalization. However, as DC cultured from BM of NOD mice in vitro also show strong defects in expression of CCR5, this appears to be unlikely. Furthermore, extended exposure of wild-type DC to CCR5 ligands does not lead to decreased expression of the receptor, with the exception of CCL3 at very high concentrations (data not shown). However, even is this case, expression of CCR5 remained at clearly detectable levels, unlike NOD DC in vivo.
In a mouse model for experimental autoimmune encephalitis, oral tolerance could be induced in wild-type, but not in CCR5–/– mice [22 ]. Furthermore, in the CCR5–/– mice, increased levels of IL-12 were observed in the GALT. The current study shows that DC derived from CCR5–/– or DC treated with a CCR5 antagonist show increased secretion of IL-12p40 compared with wild-type BMDC. Although levels were low, results showed a similar trend for IL-12p70 (data not shown). Earlier, increased secretion of IL-12 in response to various stimuli was reported in NOD DC and NOD macrophages, which also show decreased expression of CCR5, and this was indeed the result of autocrine signaling [31 , 32 ]. Interestingly, in pSjS patients, decreased levels of CCR5 were observed in peripheral monocytes, and this correlated negatively with the level of IL-12 in the serum, supporting the hypothesis that CCR5 plays a role in the regulation of IL-12 in DC.
IL-12 secreted by DC is a well-known inducer of Th1 cells and plays an important role in the development of various immune responses [33 ]. In the NOD mouse model, addition of IL-12 leads to acceleration of the development of diabetes, whereas inhibition of IL-12 using an antagonist has a protective effect [34 , 35 ]. Unfortunately, the salivary glands were not studied in these experiments, but increased expression of IL-12 is found in the salivary glands of pSjS patients [36 ]. Furthermore, transgenic mice overexpressing IL-12 develop infiltrations in their salivary glands similar to those seen in NOD mice, indicating an important role for this cytokine in the pathogenesis of pSjS [37 ].
In a recent report, CCR5–/– mice showed an increased antibody production in the setting of a cardiac transplant [38 ]. These antibodies were highly functional, as CCR5–/– mice showed acute humoral rejection, which could be transferred by serum [39 ]. Similar increases in levels of antibody were observed earlier in an infection model in CCR5–/– animals [25 ]. In both studies, numbers of IL-4-producing T cells were increased, suggesting a skewing of the immune system toward antibody production in the absence of CCR5. Increased levels of autoantibodies are a well-known phenomenon in human pSjS patients as well as NOD mice [40 41 42 ]. The lack of CCR5 on DC may play a role in this by increasing the numbers of IL-4-producing T cells, thus stimulating the generation of autoantibodies. Indeed, in NOD mice, IL-4 is of crucial importance to the development of pSjS-like symptoms, as NOD.IL-4–/– mice do not develop salivary dysfunction [43 ]. Interestingly, in the patient cohort described in this study, CCR5 expression was lower in patients who were positive for the well-known autoantibody SS-B, although this did not reach statistical significance (P=0.067; data not shown).
The hypothesis that the lack of CCR5 plays a role in the increased expression of IL-12 as well as IL-4 appears somewhat at odds with the original Th1/Th2 paradigm. However, it has now been clearly shown that the two responses are not as mutually exclusive as previously thought, and in pSjS patients, increased numbers of IFN-
- and IL-4-producing T cells are found [44
45
46
]. Furthermore, the NOD mouse, which is regarded as a strongly Th1-biased animal, displays a more aggressive form of the stereotypical Th2 condition asthma [47
].
In humans, the role of CCR5 polymorphisms [48 , 49 ] has been studied in various autoimmune diseases. In diabetes and systemic lupus erythematosus, no association was found between the frequency of CCR5 polymorphisms and disease incidence [50 51 52 ], but more complications and interestingly, higher levels of autoantibodies were found in patients with polymorphic CCR5 alleles [52 53 54 ]. Only one earlier study focused on CCR5 in pSjS and concluded that a CCR5 mutation may somewhat protect against pSjS, contradicting our results [55 ]. However, in that study, only one polymorphism was included, leaving the possibility that other polymorphisms influenced the results. In the current study, CCR5 was determined at the protein level, thus resulting in actual expression levels rather than genetics.
The function of CCR5 in the migration of various cell types has been studied extensively. This study shows that in addition, CCR5 is a regulatory molecule and that its expression on immature DC is likely to contribute to the tolerogenic phenotype of this subset. In the salivary gland of the NOD mouse model for pSjS, the lack of CCR5 on DC contributes to a more proinflammatory environment.
Received November 27, 2007; revised January 14, 2008; accepted January 25, 2008.
|
|
|---|
B activation in nonobese diabetic mouse dendritic cells results in enhanced APC function J. Immunol. 168,188-196
and IL-10 define a unique cytokine profile in macrophages from young nonobese diabetic mice Diabetes 49,1106-1115[Abstract]
-fodrin antibodies in Sjogren syndrome and lupus erythematosus Arch. Dermatol. 135,535-539
-fodrin immune response with Th1-cytokine profile in the NOD mouse model of Sjogrens syndrome Eur. J. Immunol. 28,3336-3345[CrossRef][Medline]
32 with type 1 diabetes and latent autoimmune diabetes in adults Hum. Immunol. 64,629-632[Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||