SEARCH

SEARCH BY CITATION

Keywords:

  • HLA-G;
  • IL-10;
  • Crohn's disease;
  • ulcerative colitis

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Background: HLA-G antigens are nonclassical major histocompatibility complex (MHC) class I molecules characterized by tolerogenic and antiinflammatory properties. Recently, a different expression of HLA-G antigens has been observed between intestinal biopsies of ulcerative colitis (UC) and Crohn's disease (CD) patients. These data suggested a functional role for HLA-G molecules in the diseases and proposed the HLA-G modulation as a marker for the diagnosis of UC and CD. The soluble HLA-G antigens (sHLA-G) are circulating molecules mainly produced by activated peripheral blood CD14+ monocytes.

Methods: We tested, by specific enzyme-linked immunosorbent assay (ELISA), the sHLA-G molecule levels in the supernatants of unstimulated and bacterial lipopolysaccharide (LPS)-stimulated cultures of peripheral blood mononuclear cells (PBMC) from 30 healthy subjects, 10 CD, and 18 UC patients. The data were not influenced by treatment or disease activity.

Results: The results confirmed a different sHLA-G expression between the diseases, with a spontaneous secretion of sHLA-G in CD patients but not in UC and healthy subjects. Moreover, a lack of sHLA-G antigens has been reported in UC patient cultures after LPS activation but not in healthy subjects and CD patients. The defective sHLA-G production was related to an impaired IL-10 secretion in UC but not in CD.

Conclusions: Overall, these results confirm the presence of a different biological characteristic between CD and UC patients and suggest sHLA-G production by PBMC as a noninvasive diagnostic tool in the early phases of the diseases.

(Inflamm Bowel Dis 2007)

Several genetic and immunoregulatory factors concur in the development of inflammatory bowel diseases (IBDs). In this context, the defects in T-cell regulation have been suggested as involved in the relapse and maintenance of a chronic inflammation. Several data are consistent with the persistence of a Th1/Th2 disequilibrium in Crohn's disease (CD) and ulcerative colitis (UC)1 that seems to be at the basis of the autoimmune and inflammatory condition.

Recently, a protective role for HLA-G antigens has been proposed in inflammatory diseases.2 HLA-G is defined as a nonclassical Ib gene that differs from classical HLA class I for its limited polymorphism, a restricted tissue distribution, and an alternative translation of spliced mRNA transcripts that encode membrane-bound (HLA-G1, G2, G3, G4) and soluble isoforms (HLA-G5, G6, G7).3–5 Following the identification in cytotrophoblasts, HLA-G products have been observed in physiological conditions in thymus and in the peripheral blood monocyte population activated by IL-10 or interferons.6–8 The expression of HLA-G antigens has been further observed in some solid tumors,9–12 virally infected cells,13, 14 transplanted organs,15, 16 and cutaneous inflammatory diseases,17, 18 while soluble molecules have been detected in the cerebral spinal fluid of patients with multiple sclerosis.19 Several studies have confirmed that HLA-G antigens, mainly in their HLA-G1 membrane-bound and HLA-G5 soluble isoforms, can mediate the inhibition of natural killer (NK) cells and CD8+ T-lymphocyte cytotoxic activity and CD4+ T-cell alloproliferation by the interaction with inhibitory receptors and apoptosis induction.20–22 Furthermore, HLA-G can suppress antigen-presenting cell functions and HLA-G treatment of human monocyte-derived dendritic cells inhibits allogeneic T-cell proliferation.23, 24 Overall, these findings confirmed its immunosuppressive functions against innate and adaptative cellular responses. Therefore, the pathobiological involvement of HLA-G in IBD could be associated with the immunoinhibitory function of this molecule. It could be involved in the regulation of T-cytotoxic and NK cell activity, which have a pivotal role in IBD. It is known that CD NK cell function and cell number are significantly below normal levels in both the remissive and active stages.25, 26 This could be affected by HLA-G molecules that could reduce the NK activity by inhibitory receptor interaction.20 As these cells have a particular importance in the regulation of a variety of gut infectious and inflammatory conditions,27 their apoptosis could have a negative effect on disease control.

Recently, Torres et al28 observed a significantly different HLA-G expression in intestinal biopsies from CD and UC patients. In detail, the results have shown the modulation of HLA-G molecules in UC but not in CD biopsies. Overall, the data suggested a functional role for HLA-G molecules in IBD diseases and proposed the different HLA-G expression as a potential diagnostic marker to distinguish CD and UC. However, this analysis of HLA-G molecules on intestinal biopsies is invasive and this convinced us to develop a noninvasive method to discriminate UC and CD.

Taking into consideration the recently reported impaired Th1/Th2 cytokine balance in IBD patients1 and the role of Interleukin-10 (IL-10) as 1 of the main up-modulators of HLA-G production,22 we analyzed IL-10 and sHLA-G production in IBD patients.

As no differences have been reported in IL-10 serum levels between UC and CD patients29 and the possible effects of therapy on HLA-G serum and plasma levels,30 we developed an ex vivo system to evaluate a possible different state of activation in UC and CD peripheral blood mononuclear cells (PBMCs). We compared with a specific enzyme-linked immunosorbent assay (ELISA) the IL-10 and sHLA-G levels in the culture supernatants of unstimulated and LPS-stimulated PBMC of 30 healthy subjects, 17 patients affected by CD, and 20 with UC. Western blot analysis was performed to confirm the ELISA data.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Patients

Blood samples were obtained from 27 subjects with different disease activities attending the Gastroenterology and Digestive Endoscopy Unit of the Ferrara Sant'Anna Hospital. The patients had not been treated with biological immunosuppressive drugs.

Of the 27 subjects, 17 had CD and 20 had UC. The diagnosis was based on standard clinical, radiological, and histopathological features according to the criteria of Lockhart-Mummery and Morson.31 At the time when blood was taken for the assay the disease activity had been assessed clinically. The disease activity in CD was documented using the Crohn's Disease Activity Index (CDAI) and for UC using the Colitis Activity Index (CAI) according to Rachmilewitz32 and the criteria of Truelove and Witts.33 On the basis of this assessment the disease in each case was classified as active or inactive. Patient data are summarized in Table 1.

Table 1. Characteristics of Patients with Crohn's Disease and Ulcerative Colitis.
 Crohn's DiseaseUlcerative Colitis
  • *

    Assessment of disease activity using Crohn's disease activity index (CDAI)32 in Crohn's disease and colitis activity index (CAI) in ulcerative colitis.33

No. of patients1720
Sex (F/M)9/811/9
Age (mean range)54 (33–83)51 (27–75)
Disease activity*  
 Active1013
 Inactive77
Localization  
 Rectocolitis 8
 Pancolitis 7
 Rectosigmoid 5
 Ileocolic12 
 Ileal5 
Medication  
 Steroids2321
 5-ASA3633
 Antibiotics88
 No treatment01

The 30 healthy control subjects were matched for sex (13 males and 17 females) and age 50.5 ± 9.7 year (range, 30–62). Informed consent for blood drawing was obtained from each subject.

PBMC Cultures

PBMC obtained by Ficoll centrifugation (Cederlane, Hornby, Ontario, Canada) was resuspended in Iscove's medium (Biochrom, Berlin, Germany) + 10% fetal calf serum at a concentration of 1 × 106/mL and cultured for 48 hours. Activated cultures were obtained by 10 ng/mL of LPS (Calbiochem, La Jolla, CA).34

To restore sHLA-G secretion in UC, the LPS-activated cultures were supplemented with 20 ng/mL of exogenous recombinant IL-1034 (PeproTech, Rocky Hill, NJ). CD14+ cells viability and percentages were evaluated by cytofluorimetry analysis (FACS Vantage, Becton Dickinson, San Jose, CA), using propidium-iodide staining and an antihuman CD14-FITC-conjugated antibody (Cymbus Biotechnology, Hants, UK).

sHLA-G Levels

sHLA-G1, obtained from the proteolytic cleavage of the membrane-bound HLA-G1 and HLA-G5, generated by mRNA alternative splicing, was assayed in culture supernatants in triplicate as reported in Essen Workshop on sHLA-G quantification35 using as capture antibody the MoAb MEM-G9 (Exbio, Praha, Czech Republic), which recognizes HLA-G molecule, in β2-microglobulin-associated form, at a concentration of 20 μg/mL. As detecting antibody an anti-β2 microglobulin MoAb HRP-conjugated (Dako, Glostrup, Denmark) was used. HeLa cell wildtype culture supernatants were used as negative control, transfected HeLa-G5 cell (kindly provided by Prof. E. Weiss, Institut fur Anthropologie and Genetik, LMU, Munchen, Germany) as positive control. They were cultured in CD hybridoma AGt medium (GIBCO, Auckland, NZ) with 1% FCS and antibiotics added. Culture supernatants were collected at cell confluence and concentrated by a lyophilization procedure. Following depletion of albumin with the Albumin Depletion Kit (Enchant Life Science kit, Pall, East Hills, NY), the purification of the sHLA-G proteins was carried out as previously reported.23 The sHLA-G molecules obtained were used as standard at different dilutions.

The intraassay coefficient of variation (CV) was 1.4% and the interassay CV was 4.0%. The limit of sensitivity was 1.0 ng/mL. The detection limit was calculated with repeated measurements of a blank and a standard deviation (SD) was calculated, with the value of the lower limit of detection being 3.29 SD.36 In this case there is only a 5% chance of classifying a result in the wrong population and the lower limit of detection sample determinations are above this midway concentration with a probability of 95%.

IL-10 Concentrations

IL-10 concentrations were determined in triplicate for undiluted samples using the commercially available Human IL-10 BioSource Immunoassay Kit (Human IL-10 US; BioSource, Camarillo, CA) with a detection limit of 0.2 pg/mL.

Western Blotting

The presence of sHLA-G molecules in PBMC culture supernatants was analyzed by Western blot in unstimulated and LPS-stimulated PBMC cultures from UC, CD, and healthy subjects as previously reported.37 Briefly, concentrated and albumin-depleted (Enchant Life Science kit, Pall) PBMC culture supernatants were loaded on 8% SDS-polyacrilamide gel, electrophoresed at 80V for 2 hours, and blotted onto PVDF membrane (Immobilon-P Millipore, Billerica, MA) by electrotransfer at 100V for 45 minutes in 25 mM Tris Buffer, 190 mM glycine, 2% SDS, and 20% (V/V) methanol. Blocking was carried out with 5% nonfat dry milk, Tris 100 mM pH 7.5, NaCl 150 mM overnight at 4°C. After 2 washes, the membrane was incubated with MEM-G1 (10 μg/mL) for 3 hours at room temperature with gentle shaking. The sHLA-G molecules were detected using Protein-G HRP (BioRad, Hercules, CA) at a dilution of 1:5000 in 10 mM Tris pH 8.0, 150 mM NaCl, and 0.1% Tween 20. Reactions were developed by chemiluminescence with SuperSignal enhanced chemiluminescence kit (Super Signal West Pico system; Pierce, Rockford, IL) and exposed to Hyperfilm ECL (Amersham Bioscience, Piscataway, NY). The culture supernatants of transfected HeLa-G5 cells (kindly provided by Prof. E. Weiss), obtained as previously reported, were used as positive control and HeLa wildtype as negative control. The molecular weights were determined with the BenchMark (Invitrogen, La Jolla, CA) prestained protein ladder (range, 10–200 kD).

Statistics

Statistical evaluations were performed by Fisher's exact test and the comparisons between variables by Kruskal–Wallis 1-way analysis of variance and Mann–Whitney U-test with the Stat View software package (SAS Institute, Cary, NC). All P-values < 0.05 were considered significant.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

The data obtained by cytofluorimetry analysis showed similar percentages of CD14+ cells in the 3 groups (healthy subjects: 7.9 ± 3.2% (mean ± SD); CD: 5.9 ± 3.2%; UC: 8.3 ± 4.6%). Specific ELISA assays for sHLA-G and IL-10 molecules confirmed that CD and UC are characterized by fundamental differences in sHLA-G1/HLA-G5 and IL-10 production. The frequencies of unstimulated and LPS-stimulated PBMC cultures positive for IL-10 were similar in the 3 groups (P = ns, Fisher's exact test). No differences were observed in IL-10 levels in unstimulated PBMC cultures (P = ns, Kruskal–Wallis test) (Fig. 1a), whereas a difference was found in the IL-10 concentrations between the 3 groups in LPS-stimulated cultures (P < 0.0001, Kruskal–Wallis test) (Fig. 1b). The cytokine levels were reduced in UC patients (2.6–79.2 pg/mL; 43.0 pg/mL; range; median) when compared to healthy subjects (35.6–425.3 pg/mL; 141.9 pg/mL) (P < 0.0001, Mann–Whitney test) and CD patients (16.0–325.6 pg/mL; 113.3 pg/mL) (P = 0.0008, Mann–Whitney test).

thumbnail image

Figure 1. IL-10 levels in unstimulated (a) and LPS-stimulated (b) PBMC culture supernatants from healthy subjects (HS), Crohn's disease (CD), and ulcerative colitis (UC) patients. Median values are shown. Unstimulated: P = NS, Kruskal–Wallis test. LPS-stimulated: UC versus CD: P = 0.0008. UC versus HS: P < 0.0001.

Download figure to PowerPoint

The 3 groups significantly differ for sHLA-G concentrations in unstimulated (P < 0.0001, Kruskal–Wallis test) and LPS-stimulated PBMC cultures. In CD patients a spontaneous production of sHLA-G antigens (0–20 ng/mL; 4.0 ng/mL) was observed in unstimulated cultures (16/17), while no detectable sHLA-G1/HLA-G5 was found in the culture supernatants from UC patients (0/18) (P = 2.8 × 10−4, Fisher's exact test) (P < 0.0001, Mann–Whitney test) and healthy controls (0/30) (P = 3.6 × 10−6, Fisher's exact test) (P < 0.0001, Mann–Whitney test) (Fig. 2a). Following LPS stimulation similar levels of sHLA-G1/HLA-G5 molecules were reported in the culture supernatants of healthy subjects (1.4–20.8 ng/mL; 3.4 ng/mL) and CD patients (0–21.4 ng/mL; 3.2 ng/mL) (P < 0.0001, Mann–Whitney test) (Fig. 2b), while PBMC cultures of UC patients showed no detectable sHLA-G (0/18) (HS/UC P = 4.8 × 10−5; CD/UC P = 1.3 × 10−4, Fisher's exact test) (P < 0.0001, Mann–Whitney test). The addition of exogenous recombinant IL-10 to PBMC cultures from UC patients restored a normal sHLA-G production (2.4–11.1 ng/mL; 3.0 ng/mL) (P = ns, Kruskal–Wallis test) (Fig. 2b).

thumbnail image

Figure 2. HLA-G5/sHLA-G1 (sHLA-G) levels in unstimulated (a) and LPS-stimulated (b) PBMC culture supernatants from healthy subjects (HS), Crohn's disease (CD), and ulcerative colitis (UC) patients. Median values are shown. Unstimulated: CD versus UC: P < 0.0001. CD versus HS: P < 0.0001. LPS-stimulated: UC versus CD: P < 0.0001. UC versus HS: P < 0.0001.

Download figure to PowerPoint

The different expression of sHLA-G molecules in unstimulated and LPS-stimulated PBMC-cultures was confirmed by Western blot analysis (Fig. 3). These results were not influenced by therapeutic treatment, disease localization, or activity (P = ns, Kruskal–Wallis test).

thumbnail image

Figure 3. Western blot analysis of healthy subjects (HS), Crohn's disease (CD), and ulcerative colitis (UC) unstimulated and LPS-stimulated PBMC culture supernatants probed with MEM-G1 MoAb. The molecular weights were determined with the BenchMark (Invitrogen) prestained protein ladder (range, 10–200 kD). Transfected HeLa-G5 (+) cell culture supernatants were used as positive control and HeLa wildtype cell culture supernatants (−) as negative control. sHLA-G was revealed at 37 kD.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Our study has documented a clear difference in the production of sHLA-G molecules in IBD patients, confirming the presence of a different etiology and immune response mechanisms in UC and CD. In particular, LPS-stimulated PBMCs from UC patients produce lower sHLA-G levels, associated with lower IL-10 production. The restored capacity to produce normal sHLA-G levels following the addition of exogenous rIL-10 has confirmed that the reduced amounts of IL-10 molecules could be responsible for the defective production of sHLA-G antigens by activated PBMCs. As a consequence, the low production of these tolerogenic molecules could be involved in the development and persistence of inflammatory conditions of UC, where there is an atypical Th2 immune response (IL-4, TGFβ) with a relative deficiency of IL-10.38 This immunological imbalance in UC could be at the basis of the impaired sHLA-G production.

On the other hand, the detectable amounts of sHLA-G molecules observed in the supernatants of unstimulated PBMC of CD patients could be ascribed to the elevated IL-10 levels detected in the culture supernatants (Fig. 1). In these patients LPS activation induced an increase in IL-10 production and sHLA-G levels comparable to healthy controls. The persistence of inflammatory conditions in CD patients, in spite of the normal production of sHLA-G antigens, could suggest a chronic inflammatory condition that cannot be efficiently counteracted by the production of HLA-G antiinflammatory molecules. It is known that the colon from a patient with CD already secretes higher levels of IL-10 compared with control colons.39 These data confirm our results on PBMCs and suggest that sHLA-G may not have a beneficial effect in CD.

The discrepancy between the deficit in sHLA-G production by UC PBMC and the reported expression of these tolerogenic molecules in intestinal biopsies28 could be explained by the different specific microenvironment.2 Several reports have proposed that the microenvironment is fundamental for HLA-G modulation and a recent study has reported the loss of HLA-G expression after in vitro stabilization of several HLA-G-positive tumor biopsies.40 Moreover, it has been described that CD intestinal mucosa expresses increased IFN-γ and IL-10 compared to controls, whereas UC patients have IFN-α and IL-10 levels comparable to controls.41 These data seem to confirm our results on culture IL-10 levels. We can suggest that CD has a more important modification in cytokine balance than UC, where the mucosa IL-10 production is less affected by the disease. This situation is clearly demonstrated by sHLA-G levels, too, which follow IL-10 behavior.

The therapeutic treatment, disease localization, and disease activity have not influenced the sHLA-G production that seems to be a peculiar behavior of the specific disease. These results strengthen the power of this test, which could be used without limitations.

It is known that a specific IBD diagnosis is characterized by some difficulties at the early stages of the diseases in those patients with an inflammatory disease limited to the colon. The endoscopic and radiologic results supported by histological features and laboratory tests permit distinguishing between UC and CD, but with many difficulties. While it is clear that our data need to be confirmed in a larger number of patients, the sHLA-G production by “in vitro” LPS-activated PBMC could be suggested as a simple and noninvasive diagnostic aid in IBD to discriminate between UC and CD.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES