The adipokine CTRP-3 (C1q/TNF-related protein-3) belongs to the C1q/TNF-related protein family which antagonizes the effects of lipopolysaccharide (LPS). The aim was to investigate the antiinflammatory and antifibrotic role of CTRP-3 in Crohn's disease (CD).
Mesenteric adipose tissue (MAT) of patients with CD or colonic cancer (CC) was resected. Human primary colonic lamina propria fibroblasts (CLPF) were isolated from controls and CD patients. Concentrations of chemokines and cytokines in the supernatants were measured by enzyme-linked immunosorbent assay (ELISA). Expression of connective tissue growth factor (CTGF), collagen I, and collagen III was analyzed by real-time polymerase chain reaction (PCR). Recombinant CTRP-3 expressed in insect cells was used for stimulation experiments.
CTRP-3 is synthesized and secreted by MAT resected from patients with CD, ulcerative colitis (UC), CC, and sigma diverticulitis as well as by murine and human mature adipocytes. CTRP-3 had no effect on the basal secretion of MCSF, MIF, or RANTES in MAT of CD and control patients. LPS-stimulation (10 ng/mL) significantly increased IL-8 release in CLPF of CD patients and, to a lesser extent, in cells of controls and of fibrotic CD tissue. CTRP-3 significantly and dose-dependently reduced LPS-induced IL-8 secretion in CLPF within 8 hours after LPS exposure, whereas LPS-induced IL-6 and TNF release was not affected. CTRP-3 inhibited TGF-β production and the expression of CTGF and collagen I in CLPF, whereas collagen III expression remained unchanged.
CTRP-3 exerts potent antiinflammatory and antifibrotic effects in CLPF by antagonizing the LPS pathway and by targeting the TGF-β–CTGF–collagen I pathway.
Mesenteric adipose tissue hypertrophy, also named “creeping fat” (inflamed adipose tissue enveloping the circumference of the gut), represents a hallmark of Crohn's disease1,–3 (CD) and is present from the onset of the disease. The associated connective and adipose tissue changes are characterized by mesenteric fat hypertrophy, fibrofatty proliferation, macrophage infiltration, regional lymphadenopathy, tissue fibrosis, perivascular and transmural inflammation, and intimal/medial thickening of small vessels.1,–4 Moreover, creeping fat correlates with ulceration, stricture formation, transmural inflammation, wall thickness, and internal bowel diameter.5 There are arguments both for a causative role and a secondary role of mesenteric adipose tissue inflammation in CD. Transmural inflammation together with chronic intestinal and lymphoid tissue cytokine release might cause adipose tissue hypertrophy with a role of creeping fat in exerting a putative barrier function.3,4 However, the presence of creeping fat at the onset of the disease, the axial polarity of inflammation and ulcers along the mesenteric border, and the active release of a broad variety of proinflammatory mediators such as interleukin (IL)-6, tumor necrosis factor (TNF), monocyte chemotactic protein 1 (MCP-1), macrophage colony-stimulating factor (M-CSF), and adipokines such as adiponectin, leptin, and resistin from the adipose tissue might argue for a more active role of mesenteric adipose tissue in the pathogenesis of CD.1,–3 Our recent work demonstrated6,–8 that ex vivo cultures of total adipose tissue resected from creeping fat in CD patients release significant quantities of IL-10, regulated on activation, normal T cell expressed and secreted (RANTES), vascular epithelial growth factor (VEGF), migration inhibitory factor (MIF), M-CSF, MCP-1, IL-6, adiponectin, leptin, and resistin. A specific overexpression in creeping fat of CD patients was shown for TNF,4 leptin,9 M-CSF,10 and adiponectin.11
During the last 5 years, our group12,–18 and others19,–22 characterized a novel secretory protein of the adipose tissue with antiinflammatory properties, CTRP-3 (also named CORS-26, cartonectin, cartducin). CTRP-3 (C1q/TNF-related protein-3) was identified as a member of the C1q/TNF-related protein (CTRP) superfamily. The protein consists of an N-terminal collagenous region and a C-terminal globular head domain like adiponectin. CTRP-3 is not only expressed in cartilage, where it was originally described, but also in murine and human adipose tissue, murine preadipocyte cell lines,13,14 and in monocytes.16 It circulates in human sera,18 exerts potent antiinflammatory effects,16 and upregulates the secretion of adiponectin and resistin in adipocytes.15 Recently, CTRP-3 was characterized as a novel endogenous lipopolysaccharide (LPS)-antagonist of the adipose tissue inhibiting Toll-like receptor ligand- and fatty acid-induced proinflammatory effects.23 Since fibroblasts are important cellular components of both mesenteric adipose tissue and the lamina propria in the colon, the interplay of these two cell types might be of pathophysiological interest regarding transmural inflammation, adipose tissue hypertrophy, and stricture formation.
Based on these observations, it was the aim of the present study: 1) to prove the hypothesis that short-term cultures of resected creeping fat samples release significant quantities of CTRP-3 into the supernatants; 2) to investigate whether total adipose tissue explants of creeping fat samples show a distinct profile of cytokine and chemokine release in the absence and presence of LPS, recombinant human CTRP-3, and palmitate, a major proinflammatory dietary fatty acid; and 3) to investigate whether recombinant human CTRP-3 is able to regulate inflammatory and fibrotic activity in colon lamina propria fibroblasts (CLPF) isolated from CD patients.
Materials and Methods
Ex Vivo Short-term Culture of Human Total Adipose Tissue Explants
Surgical specimens of intraabdominal visceral adipose tissue were obtained during laparotomy of four patients with CD (two males, two females, mean age 31.3 ± 6.6 years) and six patients with colonic cancer (three males, three females, mean age 65.2 ± 13.3 years) requiring bowel resection. The total adipose tissue specimens were obtained from the adjacent mesenteric adipose tissue. In patients with CD, the creeping fat contiguous to the involved intestine was resected. Importantly, fat lobules were prepared from different sites of the resected bowel fragment. Each patient gave written informed consent and the study was approved by the local ethical committee. Since a high interindividual variation of adipocytokine secretion was expected and since ex vivo experiments with short-term culture of tissue specimens are characterized by a high variation of adipocytokine secretion, n = 3 tissue samples were incubated of the same patient in one well and n = 4 wells were incubated per patient and per experimental group. Based on this procedure, n = 12 fat tissue lobules were incubated per patient.
The fat tissue samples were placed in 1× PBS (phosphate-buffered saline) containing 0.1% BSA (bovine serum albumin). Fat tissue lobules (≈0.5 cm3) were prepared under sterile conditions by microdissection. Vessels and adjacent soft tissue were carefully removed. After preparation, the fat tissue samples were washed twice with 1× PBS / 0.1% BSA. The tissue samples were incubated in 1 mL medium for 24 hours at 37°C in a 95% O2 and 5% CO2 incubator. The medium consisted of DMEM:F-12/glutamate (Dulbeccos Modified Eagle's Medium Ham's F12 (1:1 mix) (BioWhittaker, Cambrex, Belgium) supplemented with 100 U/mL penicillin and 100 μg/mL streptomycin (GIBCO BRL, Berlin, Germany). Supernatants were collected after 24 hours and stored at −20°C. The wet weight of the remaining fat tissue was determined to calculate adipokine secretion as pg or ng/g fat/24 h. As outlined above, four wells with three fat lobules were used for each experimental group: control group (vehicle without stimulation), stimulation with LPS (1 ng/mL), stimulation with CTRP-3 (10 ng/mL), stimulation with palmitic acid (100 μM).
Isolation and Culture of Human CLPF
Primary CLPF cultures were obtained from the intestinal mucosa of control patients and CD patients with active CD (CD-CLPF) and from fibrotic (fibrosis-CLPF) areas of the mucosa. Primary human CLPF cultures were prepared of endoscopic biopsies or surgical specimens taken from the inflamed or fibrotic mucosa of patients with CD. All biopsy or surgical specimens were clinically evaluated by an experienced examiner and histologically evaluated by an experienced pathologist. Control-CLPF of three patients (mean age 60.3 ± 12.0 years, two males, one female), CD-CLPF of three patients (mean age 36.7 ± 18.2 years, two males, one female), and fibrosis-CLPF of three patients (mean age 23.7 ± 6.1 years, three females) were examined. This study was approved by the Ethics Committee of the University of Regensburg and performed according to the Declaration of Helsinki.
Primary human CLPF were isolated and cultured as previously described.24 Isolated cells were washed with Quantum 333 medium for fibroblasts (Quantum; PAA, Cölbe, Germany) and cultured in 25 cm2 culture flasks (Costar, Bodenheim, Germany) in the presence of penicillin (100 U/mL), streptomycin (100 μg/mL), ciprofloxacin (8 μg/mL), gentamycin (50 μg/mL), and amphotericin B (1 μg/mL). Nonadherent cells were removed by subsequent changes of medium. The remaining cells were used between passages 3 and 8.
Measurement of Adipokine, Cytokine, and Chemokine Concentrations by Enzyme-linked Immunosorbent Assay (ELISA)
Supernatant adipokine concentrations were measured by ELISA: MIF (Human MIF Quantikine, R&D Systems, Wiesbaden-Nordenstadt, Germany), M-CSF (Human M-CSF Quantikine, R&D Systems, Wiesbaden-Nordenstadt, Germany). RANTES was measured in supernatants by ELISA using the human RANTES/CCL5 ELISA system (R&D Systems Europe, Abingdon, UK). This assay system is specific for human RANTES/CCL5 and has no crossreactivity or interference with human IL-8, MCP-1, GROα, MIP-1α, or MIP-1β. For data normalization, total protein concentration was measured. For the detection of potential cytotoxic effects, LDH concentration was measured in the supernatants (cytotoxicity detection kit, Roche, Mannheim, Germany). Each sample was measured in duplicate by ELISA. Values were normalized to total protein content of each well and are expressed as means ± standard error of the mean (SEM).
Recombinant Expression of CTRP-3
Recombinant CTRP-3 protein expression was performed in High Five insect cells (Invitrogen, Karlsruhe, Germany) using the BacPAK Baculovirus Expression System (BD Biosciences, Palo Alto, CA) as published previously by our group.16 Supernatants were collected 3 days after viral infection of these cells for purification of CTRP-3 using the BD Talon Purification kit (BD Biosciences). Integrity and purity of the protein was analyzed by immunoblot and silver staining of sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). In contrast to E. coli-based expression systems, the recombinant expression in insect cells usually maintains glycosylation and phosphorylation. Based on this, we could demonstrate previously that our expression system generates trimeric CTRP-3 that was used for stimulation experiments.16,18
Western Blot Analysis
Cultured cells or tissue homogenates were washed twice with PBS, harvested gently with a cell scraper, centrifuged, and resuspended in 100 μL PBS. Equal amounts of total protein were submitted to gel electrophoresis after a 1:1 dilution with Laemmli loading buffer. SDS-PAGE was performed following standard procedures. Proteins were transferred to a Fluotrans Transfer Membrane (Pall, Portsmouth, UK). Transfer was confirmed by Ponceau Red staining of the membrane and Coomassie Blue staining of the gel after the electroblot. For detection of CTRP-3, antibodies were raised (Pineda-abservice, Berlin, Germany) in chicken against a synthetic peptide of the human sequence (NH2-CYSYEMKGKSDTSSNA-COOH) from the noncollagenous region of the protein (amino acids 193–207) and used for western blot analysis as demonstrated previously by our group.16,18 These antibodies recognize the corresponding protein of human origin as a trimeric, high molecular weight (HMW) form. The CTRP-3 antibody was used with a 1:200 dilution, the secondary peroxidase-coupled antichicken antibodies of rabbit origin (Sigma-Aldrich, Steinheim, Germany) with a 1:5000 dilution in 5% nonfat dry milk/PBS suspension. Detection of the immune complexes was carried out with the enhanced chemiluminescence western blot detection system (Amersham ECL Western Blotting System, GE healthcare, Freiburg, Germany).
Stimulation of Human CLPF and Analysis of Cytokine Production
For stimulation experiments, 1.5 × 105 cells were seeded into 6-well culture dishes and cultured for 24 hours at 37°C. Medium was removed, the cells were washed twice with PBS and stimulation medium (Dulbecco's Modified Eagle's medium [DMEM; PAA]) containing penicillin (100 U/mL), streptomycin (100 μg/mL), ciprofloxacin (8 μg/mL), gentamycin (50 μg/mL), and amphotericin B (1 μg/mL) was added. CLPF were incubated in duplicates for 8–48 hours with LPS (10 ng/mL; Sigma-Aldrich), and/or CTRP-3 (10–500 ng/mL) at 37°C. Supernatants were collected and analyzed for IL-6 (BD Biosciences, San Diego, CA), IL-8 (antibodies from Thermo Scientific, Rockford, IL), and transforming growth factor beta (TGF-β) (R&D Systems) by ELISA.
RNA Isolation and cDNA Synthesis
For the isolation of total RNA, stimulated and unstimulated primary human CLPF cultures were washed twice with PBS. CLPF were scraped off, centrifuged for 5 minutes, and resuspended in lysis buffer using the RNeasy kit (Qiagen, Hilden, Germany). Total RNA was prepared according to the manufacturer's protocol and stored at −80°C. The isolated RNA was reverse transcribed using the Affinity Script Multi Temperature cDNA synthesis Kit (Agilent Technologies, Santa Clara, CA).
Quantitative mRNA Analysis by Real-time Polymerase Chain Reaction (PCR)
Stimulation experiments were performed in duplicate or quadruplicate. Connective tissue growth factor (CTGF), collagen I, and collagen III mRNA were quantified by real-time PCR as previously described.25 RT-PCR from each sample was performed in triplicate. Amplicons were defined by gene-specific oligonucleotide primers (Table 1). Sequences of primers and fluorigenic probes were designed using Primer Express 1.5 software (PE Applied Biosystems, Foster City, CA) and synthesized by MWG Biotech (Ebersberg, Germany). The primers for glyceraldehyde-3-phosphate-dehydrogenase (GAPDH) were synthesized by PE Applied Biosystems. Released reporter dye fluorescence during 40 cycles of amplification was monitored using Sequence Detector software (SDS v. 2.0, PE Applied Biosystems). Reporter dye fluorescence versus PCR cycles was plotted. A threshold was set in the exponential phase of the fluorescence curves. The threshold cycle numbers (Ct) were calculated. Ct values of GAPDH were subtracted from those of target cDNAs: ΔCt = Ct (target cDNA) − Ct (GAPDH). The mean value of ΔCt values was calculated. The values of cDNA from stimulated CLPF were subtracted from those of untreated cDNA of the same CLPF: ΔΔCt = ΔCt (stimulated) − ΔCt (unstimulated). The relative start quantity of cDNA was calculated in consideration of the exponential amplification: x = 2−ΔΔCt.
Statistical Analysis
Statistical analysis was performed using the Mann–Whitney rank sum test. Measured values are expressed as median ± 5th/95th percentile. Statistically significant differences were accepted when P < 0.05.
Results
Detection of CTRP-3 in Supernatants of Human Visceral Adipose Tissue
As shown by western blot analysis (Fig. 1), CTRP-3 protein was detected in the supernatants of visceral adipose tissue resected of patients with CC (Fig. 1, line 1), CD (Fig. 1, line 2), sigma diverticulitis (Fig. 1, line 3), and UC (Fig. 1, line 4). CTRP-3 was not detectable in the supernatants of murine 3T3-L1 preadipocytes (Fig. 1, line 5), whereas it was found in the supernatants of mature 3T3-L1 adipocytes (Fig. 1, line 6) and purified mature human adipocytes (Fig. 1, line 7). The polyclonal antibody recognizes the trimeric high molecular weight (HMW) form of CTRP-3 at 90 kDa as well as a low molecular weight form (LMW). The recombinant protein was used as a positive control (Fig. 1, line 8). In supernatants, LMW consisting of probably monomeric and dimeric forms (52 kDa) were detectable.
Western blot analysis of CTRP-3 in supernatants of visceral adipose tissue and adipocytic cells. Supernatants of cells or tissues were immunoblotted using a polyclonal antibody against CTRP-3 recognizing multiple isoforms of monomeric, dimeric, and higher molecular weight forms (HMF) of CTRP-3. The polyclonal antibody recognizes the trimeric HMF of CTRP-3 at 90 kDa and the recombinant protein was used as a positive control (line 8). In supernatants, lower molecular weight forms (LMW) consisting of probably monomeric and dimeric (52 kDa) forms are detectable. Supernatants were used from visceral adipose tissue resected of patients with colonic cancer (line 1), CD (line 2), sigma diverticulitis (line 3), and UC (line 4). Supernatants of murine 3T3-L1 preadipocytes (line 5), murine mature 3T3-L1 adipocytes (line 6), and isolated human mature adipocytes (line 7) were also used.
Effect of LPS, CTRP-3, and Palmitic Acid on Adipose Tissue Derived Chemokines
The effects of LPS, CTRP-3, and palmitic acid on the concentration of MCSF, MIF, and RANTES in the supernatants of visceral adipose tissue resected of patients with CC and CD are summarized in Figure 2A,B. Basal unstimulated concentrations (line 1) of MCSF, MIF, and RANTES were not significantly different in the supernatants of visceral adipose tissue of patients with CC (Fig. 2A) and CD (Fig. 2B). In CC patients (Fig. 2A), neither LPS (line 2) nor palmitic acid (line 4) affected basal concentrations of MCSF, MIF, and RANTES. Stimulation by CTRP-3 (line 3) was unable to modulate the release of MCSF, MIF, and RANTES.
Visceral adipose tissue supernatant chemokine concentrations. Short-term ex vivo cultures of visceral adipose tissue (n = 3 fat lobules per well and n = 4 wells per patient) were used for stimulation experiments. Adipose tissue was resected of patients with colonic cancer (A) and CD patients (B). Chemokine concentrations were measured after 24 hours by ELISA. Line 1 = control/vehicle, line 2 = stimulation with LPS (1 ng/mL), line 3 = stimulation with CTRP-3 (10 ng/mL), line 4 = stimulation with palmitic acid (100 μM). *Colonic cancer versus CD.
Similarly, in patients with CD (Fig. 2B), neither LPS (line 2) nor palmitic acid (line 4) affected basal concentrations of MCSF, MIF, and RANTES. Stimulation by CTRP-3 (line 3) was unable to modulate the release of MCSF, MIF and RANTES.
However, when CD patients (Fig. 2B) were compared to those with CC (Fig. 2A), tissue of CD patients produced significantly (P = 0.04) lower quantities of MCSF (1025 ± 110 pg/mL versus 1565 ± 211 pg/mL) in the supernatants than colonic cancer patients after stimulation with CTRP-3.
Moreover, creeping fat supernatant concentrations of MIF were significantly (P < 0.01) higher (290 ± 37 pg/mL versus 136 ± 11 pg/mL) in CD patients when compared to adipose tissue of patients with CC after stimulation with CTRP-3.
Effect of LPS, CTRP-3, and Palmitic Acid on Adipose Tissue Derived Cytokines
The effects of LPS, CTRP-3, and palmitic acid on the concentration of IL-6 and TNF in the supernatants of visceral adipose tissue are summarized in Figure 3A,B. Basal unstimulated concentrations of IL-6 (line 1) were not significantly different in CC (Fig. 3A) and CD patients (Fig. 3B). LPS (line 2), palmitic acid (line 4), and CTRP-3 did not significantly influence the release of IL-6 or TNF from adipose tissue explants of patients with CC (Fig. 3A) or CD (Fig. 3B).
Visceral adipose tissue supernatant cytokine concentrations. Short-term ex vivo cultures of visceral adipose tissue (n = 3 fat lobules per well and n = 4 wells per patient) were used for stimulation experiments. Adipose tissue was resected of patients with colonic cancer (A) and CD patients (B). Cytokine concentrations were measured after 24 hours by ELISA. Line 1 = control/vehicle, line 2 = stimulation with LPS (1 ng/mL), line 3 = stimulation with CTRP-3 (10 ng/mL), line 4 = stimulation with palmitic acid (100 μM). *Colonic cancer versus CD (control), **Colonic cancer versus CD (LPS stimulation), ***Colonic cancer versus CD (stimulation with CTRP-3); §Colonic cancer versus CD (palmitic acid stimulation).
However, supernatant TNF concentrations were significantly lower in supernatants of adipose tissue of CD patients when compared to CC patients, regardless of the stimuli (Fig. 3A,B, all lines).
CTRP-3 Suppresses Proinflammatory Immune Responses of Primary Human CLPF
Since CTRP-3 is produced by mesenteric adipose tissue, it is conceivable that it targets cells within the colonic lamina propria such as CLPF. These stromal cells play an important role in wound healing processes and participate in inflammatory reactions. Therefore, the in vitro effect of CTRP-3 on primary human CLPF of control (without inflammatory bowel disease [IBD]) and CD patients was evaluated. Basal unstimulated IL-8 concentrations in the supernatants were 207 ± 111 pg/mL for control-CLPF, 23 ± 10 pg/mL for CD-CLPF, and 397 ± 88 pg/mL for fibrosis CD-CLPF (data not shown). When CLPF were stimulated with LPS, cells of control tissue as well as of CD and fibrosis-CD tissue displayed a profound production of the proinflammatory cytokine IL-8, with CLPF of CD tissue showing the strongest IL-8 secretion, while IL-8 responses of control-CLPF and fibrosis CD-CLPF were moderate (Fig. 4A). Coincubation with increasing concentrations of CTRP-3 resulted in a dose-dependent reduction of LPS-induced IL-8 responses in control CLPF (Fig. 4B), CD-CLPF (Fig. 4C), and fibrosis CD-CLPF (Fig. 4D). A dose of 100 ng/mL CTRP-3 was sufficient to significantly suppress LPS-induced IL-8 production in control CLPF, whereas at least 200 ng/mL CTRP-3 were necessary to achieve a similar effect in CD-CLPF. A significant reduction of IL-8 production in fibrosis CD-CLPF was only seen at a concentration of 500 ng/mL (Fig. 4D). These effects of CTRP-3 on LPS-induced IL-8 responses of CLPF were detected after 8 hours of stimulation. At later timepoints, the antiinflammatory activity of CTRP-3 was less distinctive (data not shown). Interestingly, although CTRP-3 incubation alone had no effect on IL-8 concentrations in the supernatants of control and CD-CLPF cultures (data not shown), basal IL-8 secretion of fibrosis CD-CLPF was significantly reduced by CTRP-3, an effect which was observed at doses of 200 ng/mL (Fig. 4E).
Antiinflammatory effect of CTRP-3 on primary human CLPF. CLPF isolated of control, CD, and fibrosis CD tissue were stimulated with LPS (10 ng/mL) in the presence of increasing concentrations of CTRP-3 for 8 hours (n = 2–3 independent experiments and n = 2 wells per patient). (A) Comparison of LPS-induced IL-8 responses in control-, CD-, and fibrosis CD-CLPF. Effect of CTRP-3 on LPS-induced IL-8 production of control-CLPF (B), CD-CLPF (C), and fibrosis CD-CLPF (D). (E) Effect of CTRP-3 on basal IL-8 secretion of CLPF of fibrosis CD tissue. B–D: P-values compared to LPS; E: P-values compared to untreated controls.
The observed antiinflammatory effect of CTRP-3 on CLPF of control, CD, and fibrosis CD tissue was specific for LPS-induced IL-8 production, since neither IL-6 levels nor TNF-induced cytokine responses were modulated in the presence of this adipokine (data not shown).
Taken together, these data indicate that CTRP-3 exerts antiinflammatory effects on primary human CLPF by suppression of IL-8 production.
CTRP-3 Inhibits Fibrotic Activity of CLPF from Strictured CD Tissue
As the antiinflammatory effect of CTRP-3 was even seen on basal IL-8 production of fibrosis CD-CLPF, we aimed to investigate whether this adipokine has the capacity to modulate fibrotic processes. In a first approach, TGF-β production in CTRP-3-treated fibrosis CD-CLPF was measured. Average basal TGF-β concentrations in unstimulated CLPF were 277 ± 290 pg/mL. Similar to IL-8 levels, TGF-β concentrations in the supernatants were significantly diminished in the presence of CTRP-3 (Fig. 5A). TGF-β is a known inducer of CTGF, which then triggers the expression of extracellular matrix genes. Therefore, to confirm the antifibrotic activity of CTRP-3, expression of the fibrotic factors CTGF, collagen I, and collagen III in CTRP-3-treated fibrosis CD-CLPF was determined. Stimulation with CTRP-3 resulted in a significant reduction of both CTGF and collagen I expression (Fig. 5B,C). In contrast, collagen III expression remained unchanged after stimulation with CTRP-3 (Fig. 5D). These findings indicate that CTRP-3 exerts potent antifibrotic effects by targeting the TGF-β–CTGF–collagen I pathway.
Fibrosis-suppressing effect of CTRP-3 on CLPF from fibrosis CD tissue. CLPF isolated of fibrosis CD tissue were left untreated or stimulated with CTRP-3 for 24 hours. (A) Effect of CTRP-3 on TGF-β secretion. Effect of CTRP-3 on the expression of CTGF (B), collagen I (C), and collagen III (D). P-values compared to untreated controls.
Discussion
The present data demonstrate that the novel adipokine CTRP-3 is produced by visceral adipose tissue resected of patients suffering from inflammatory (diverticulitis, UC, CD) and noninflammatory intestinal diseases (colonic cancer). Whereas adipose tissue in CD and in colonic cancer does not respond to CTRP-3, our results provide evidence that CTRP-3 not only exerts antiinflammatory effects on immune cells such as monocytes,16,23 but also on mesenchymal cells such as CLPF. The crosstalk between “creeping fat” and CLPF in CD is of high interest since hypertrophy and inflammatory transformation of mesenteric adipose tissue can be regarded either as a consequence of transmural inflammation or as a pathophysiological dysregulation causing CLPF activation and subsequent fibrosis.1,3
CLPF represents one of the essential components in wound healing that takes place after an injury of the intestinal mucosa.26 These cells facilitate wound healing via tissue contraction and deposition of extracellular matrix (ECM) deposition. Moreover, CLPF have been shown to express genes encoding for Toll-like receptors as well as NOD1 and NOD2. This ability enables CLPF to respond to bacterial products and to participate in immune reactions.27 In our study, IL-8 release in response to the bacterial cell wall component LPS is significantly and dose-dependently suppressed by CTRP-3. The most pronounced effect on LPS-induced IL-8 was observed for CLPF from non-IBD control tissue, where a dose of 100 ng/mL was sufficient to significantly suppress IL-8 responses. Although CD-CLPF displayed a low basal IL-8 production, higher concentrations of CTRP-3 were necessary to block LPS-induced IL-8. Fibrosis CD-CLPF showed the highest levels of basal IL-8 and in these cells LPS-induced IL-8 secretion could only be suppressed using 500 ng/mL CTRP-3. The variation of cytokine concentrations concerning spontaneous cytokine production may be due to the fact that primary CLPF cultures were used. These cells were isolated in the context of different inflammatory microenvironments and of patients with different medical therapies and disease courses. Nevertheless, these data are in accord with a study in monocytes where doses of 500 ng/mL and higher inhibited LPS-induced proinflammatory cytokine production.16 In contrast, neither LPS-induced IL-6 secretion nor TNF-induced cytokine responses were modulated in the presence of CTRP-3. This indicates that the adipokine CTRP-3 specifically blocks LPS-triggered IL-8 in CLPF.
Although CLPF are essential for intestinal tissue homeostasis and repair after injury, their prolonged presence and activation increases ECM production and might cause tissue fibrosis.24,–26 As a consequence, about one-third of CD patients require surgery due to intestinal obstructions and strictures.26 Nevertheless, the molecular mechanisms of intestinal wound healing are only partially understood.26,28,29 Maintenance of connective tissue function is regulated by an interplay of the growth factors TGF-β and CTGF,30 which stimulate proliferation of fibroblasts and the production of connective tissue components such as collagen and fibronectin.31
This study demonstrates that CTRP-3 is capable to modulate fibrotic processes of CLPF from strictured intestinal tissues. By targeting TGF-β production as well as expression of CTGF and collagen I genes, CTRP-3 reduces the fibrotic activity of fibrosis CD-CLPF. Based on this, CTRP-3 can be regarded as a novel molecular mediator derived from intestinal/mesenteric adipose tissue which inhibit CPLF activation and profibrotic mechanisms. Thus, CTRP-3 is an interesting and potential drug target in the pathophysiology of CD and its related complications such as fibrosis and stricture. Further studies have to clarify whether CTRP-3 production is altered in creeping fat compared to adipose tissue of non-IBD patients. Several studies aimed to characterize the specific cytokine and adipokine secretion profile of mesenteric adipose tissue.6,–8 As reported previously, MIF release was higher from CD adipose tissue when compared to colonic cancer patients.6,–8 Whereas MCSF release was reported previously to be elevated in CD adipose tissue,6 MCSF concentrations were lower in the present study. Since induction of RANTES in CD adipose tissue was shown to be a steroid-induced effect,8 we could not detect a specific upregulation of RANTES in CD patients. However, results from total adipose tissue cultures of different patients have to be interpreted with caution since there is a high grade of inter- and intraindividual variation.
As a next step, it has to be investigated whether exogenous treatment with CTRP-3 might be a therapeutic option for both inflammation and fibrosis in murine models of intestinal inflammation.
Acknowledgements
We thank K. Neumeier, M. Stieber-Gunckel, and N. Dunger for technical assistance.
References
Author notes
Reprints: Department of Internal Medicine I, University of Regensburg, D-93042 Regensburg, Germany e-mail: claudia.hofmann@klinik.uni-regensburg.de
Supported by a DFG grant to Andreas Schäffler and Florian Obermeier (SCHA 789/5-1).
