SEARCH

SEARCH BY CITATION

Keywords:

  • Crohn's disease;
  • inflammatory bowel disease;
  • autoantibody;
  • ubiquitination factor E4A;
  • enteroendocrine cells

Abstract

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

Background: Identification of a disease- and organ-specific autoantigen can potentially contribute to understanding molecular mechanisms involved in Crohn's disease (CD) and lead to development of clinically useful markers. The aim was to identify potential intestinal autoantigens specific to patients with CD and evaluate their diagnostic value.

Methods: A cDNA expression library from normal terminal ileum was created and screened with pooled sera from 3 randomly selected patients with CD. For evaluation of the diagnostic value of antibody screening, serum samples obtained from 39 patients with CD, 28 patients with ulcerative colitis (UC), and 60 healthy controls were examined for IgG against the strongest clone.

Results: We identified an intestinal cDNA clone encoding ubiquitination factor E4A (UBE4A), a U-box-type ubiquitin-protein ligase. The prevalence of anti-UBE4A IgG in patients with CD was significantly higher than that in patients with UC or healthy controls (46.2% versus respectively 7.1%, P = 0.0006; 3.3%, P < 0.0001). Anti-UBE4A-positive patients with CD were more likely to require surgery (P = 0.0013). The level of anti-UBE4A IgG was correlated with disease activity (r = 0.777, P < 0.0001) and associated with stricturing or penetrating disease (P = 0.0028). Immunohistochemical studies showed upregulation of UBE4A in enteroendocrine cells of inflamed ileal mucosa with CD.

Conclusions: Anti-UBE4A antibodies are potentially useful markers for detection and prediction of clinical activity and outcome in patients with CD.

(Inflamm Bowel Dis 2007)

Crohn's disease (CD) and ulcerative colitis (UC) are the 2 major forms of inflammatory bowel disease (IBD). CD is an inflammatory disease of the gut with a chronic relapsing and remitting course. CD is characterized by transmural inflammation of the gastrointestinal tract, most often affecting the ileocecal region, but which may affect any part of the digestive tract from the mouth to the anus.1 The etiology of CD is unknown, but current hypotheses suggest that an inappropriate mucosal immune response to the gut lumen microflora leading to an excessive Th1-mediated chronic inflammation may be the critical component of CD.2 Genetic factors such as NOD2/CARD15 mutations may have an important role in determining susceptibility to CD.3, 4 A large number of genetically modified mice develop spontaneous intestinal inflammation.5 Many of these animal models develop colitis under conventional conditions but not under germ-free conditions, emphasizing a role for microbial flora in the pathogenesis of chronic mucosal inflammation.6

The hypothesis that CD might be an autoimmune reaction has been considered for several decades, but it seems unlikely that CD is a simple autoimmune disease directly caused by autoantibodies or self-reactive lymphocytes.7 However, involvement of the autoimmune process in the pathogenesis of CD cannot be excluded.8 It has been suggested that a number of autoimmune diseases are caused, or triggered, by microbes.9, 10 Extraintestinal manifestations in CD such as inflammation of skin, eye, and joint support the hypothesis that CD is a systemic disease.11 Several autoantibodies have been described for patients with CD such as anti-lymphocytes,12 anti-cytoskeletal proteins,13 and pancreatic autoantibodies.14

Although there is no conclusive evidence for a direct pathogenic role for any autoantibody in CD, detection of autoantibodies may have clinical and research potential. Identification of CD-related autoantigens may contribute to understanding the underlying dysregulation of the intestinal immune system, crossreactivity with enteric microbial antigens, and immunological alterations related to genetic susceptibility.15 Organ-specific autoantigens may shed light on the novel mechanisms involved in the sustained intestinal inflammation. Only 2 organ-specific autoantigens have been reported in IBD: epithelial cell-associated components (ECAC)16 and human tropomyosin isoform 5 (hTM5).17 ECAC-specific reactivity was present in both CD and UC, whereas autoantibodies against hTM5 were found in UC, but not CD patients.

The aim of the present study was to identify specific autoantigens potentially contributing to the pathogenesis in CD and which may serve as useful serological markers for the clinical diagnosis of CD. By screening a cDNA library from normal terminal ileum with sera from patients with CD, we identified 1 strongly immunoreactive cDNA clone that encodes the C-terminal subunit of ubiquitination factor E4A (UBE4A), a U-box-type ubiquitin-protein ligase. To evaluate anti-UBE4A serum immunoreactivity, we measured specific IgG levels in patients with CD and compared them with those in patients with UC and in healthy controls. We then evaluated the predictive value of anti-UBE4A IgG level for clinical and phenotypic characteristics of CD. Finally, we investigated the expression of UBE4A protein in normal and inflamed ileal mucosa by immunohistochemistry.

MATERIALS AND METHODS

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

Serum Samples

Following informed consent, serum samples were obtained from 39 patients with CD, 28 patients with UC, and 60 healthy controls. The diagnoses of CD and UC were based on standard clinical, radiological, endoscopic, and pathological criteria.

Clinical information of CD patients was collected by clinical investigators blinded to patients' anti-UBE4A levels. Disease activity was assessed using the Crohn's Disease Activity Index (CDAI).18 The majority of patients with CD had taken oral 5-aminosalicylic acid derivatives and received elemental diet therapy. Seven patients with CD had taken steroids at some point during their disease. None of cases had been treated with immunomodulatory agents (e.g., azathioprine, 6-mercaptopurine, or antitumor necrosis factor-alpha monoclonal antibodies).

Patients with CD were classed according to the Vienna classification.19 This classification includes 3 variables: age at diagnosis (under 40 years [A1], equal to or above 40 years [A2]), disease location (terminal ileum [L1], colon [L2], ileocolon [L3], upper gastrointestinal [L4]), and disease behavior (nonstricturing nonpenetrating [B1], stricturing [B2], penetrating [B3]). All serum samples were stored at −70°C until analysis.

Tissue Samples

Ileal tissue samples were obtained from 8 control patients who underwent right colectomy for colonic cancer and from 6 patients with CD undergoing surgery for presence of fistulae or stenosis with clinical signs of obstruction. All samples were obtained with informed consent and in accordance with the Declaration of Helsinki.

Immunoscreening of the Ileal cDNA Expression Library

Poly(A)+ RNA from normal terminal ileum was extracted with the PolyATract system (Promega, Madison, WI) according to the manufacturer's protocol. A λgt11-cDNA library was constructed with SuperScript choice system (Invitrogen, Carlsbad, CA).

The cDNA expression library in Escherichia coli Y1090 was plated on LB agar and incubated for 7 hours at 42°C, then overlaid with 20 mM isopropyl β–D-thiogalactopyranoside (IPTG) (Nacalai tesque, Kyoto, Japan) -impregnated nitrocellulose membranes (Schleicher and Schuell, Dassel, Germany) for 7 hours at 37°C. After blocking with 5% (w/v) skim milk / Tris-buffered-saline (TBS), membranes were incubated with sera, which were diluted at 1:100 with 5% skim milk/TBS and preabsorbed with lysates of E. coli Y1090, for 15 hours at 4°C. The membranes were incubated in horseradish peroxidase (HRP)-conjugated goat antihuman IgG (ICN Pharmaceuticals, Aurora, OH) diluted at 1:1000 with 5% skim milk/TBS for 1 hour at 22°C, and with 4-chloro-1-naphthol (Nacalai tesque) to detect antibody-reactive phage plaques. Membranes were washed 3 times with TBS containing 0.05% Tween 20 (TBS-T) after each incubation. Positive clones were isolated by the second and third screenings, which were carried out in the same manner as the initial screening.

Sequence Analysis of cDNA Clones

cDNA inserts in immunoreactive phage clones were amplified by polymerase chain reaction (PCR) using λgt11 forward and reverse primers. cDNA inserts were sequenced using an ABI Prism 310 (PE Applied Biosystems, Foster, CA). Sequence alignments were performed using the software and searching programs provided by the National Center of Biotechnology Information (NCBI, http://www.ncbi.nlm.nih.gov).

Preparation of Recombinant C-terminal UBE4A Protein

The construct used for making the GST-C-terminal UBE4A fusion protein was made by ligating a PCR product corresponding to nucleotide position 2953–3348 of UBE4A into the vector pGEX-6P-1 (Amersham Biosciences, Piscataway, NJ). Following sequencing to verify the construct, the vector was transformed into BL21 strain of E. coli and the protein was expressed by addition of 0.1 mM IPTG. The GST-C-terminal UBE4A fusion protein was purified with B-PER GST fusion protein purification kit (Pierce Biotechnology, Rockford, IL) according to the manufacturer's protocol. Recombinant C-terminal UBE4A protein was produced by cleavage of GST-C-terminal UBE4A with PreScission protease (Amersham Biosciences). The purity of the recombinant protein was estimated by 12.5% SDS-PAGE.

Enzyme-linked Immunosorbent Assay (ELISA) for Anti-UBE4A IgG

Microtiter plates (NUNC, Roskilde, Denmark) were coated with recombinant C-terminal UBE4A protein in phosphate-buffered saline (PBS, pH 7.4) for 15 hours at 4°C. After blocking with SuperBlock blocking buffer (Pierce Biotechnology), the plates were incubated for 1 hour at 22°C with 150 μL serum samples diluted at 1:100 with PBS containing 0.05% Tween 20 (PBS-T). The plates were washed in PBS-T and 150 μL HRP-conjugated goat antihuman IgG (ICN Pharmaceuticals) diluted at 1:2000 with PBS-T were added to each well followed by incubation for 1 hour at 22°C. The plates were washed with PBS-T and 100 μL 3,3′,5,5′-tetramethylbenzidine (TMB) substrate (1-Step Slow TMB-ELISA; Pierce) was added to each well. After 30 minutes the reaction was stopped by adding 100 μL of 1 M H2SO4, and optical density (OD) at 450 nm was determined using a Titerteck Multiskan Plus MKII reader (ICN Pharmaceuticals). Data are presented as the mean OD corrected for background (wells without coated antigen). An OD value more than 2 SD above the mean of the healthy control values (0.150) were considered the cutoff level for positive reactions.

Anti-Saccharomyces cervisiae Antibodies (ASCA) ELISA

ASCA IgG + IgA ELISA kits (Genesis Diagnostics, Cambridgeshire, UK) were used to measure ASCA IgG + IgA in sera of patients with CD according to the manufacturer's instruction. A positive ASCA was defined as either a positive IgG, IgA, or both.

Statistical Analysis

Comparisons of discrete variables were analyzed by the χ2 test or, when appropriate, by the Fisher's exact test. Mann–Whitney unpaired test and Kruskal–Wallis tests were used to compare the quantitative variables. Pearson product-moment correlation coefficients were calculated to quantitate the linear association among the variables. All statistical analyses were performed with StatView 5.0 (SAS Institute, Cary, NC). Values are expressed as mean ± SD. P-values less than 0.05 were considered significant.

Immunohistochemistry

After deparaffinization the sections were treated with 0.3% hydrogen peroxide in TBS for 30 minutes to block endogenous peroxidase activity and rinsed in distilled water. Antigen retrieval was performed by microwave boiling for 5 minutes in a 0.01 M citrate buffer pH 6.0 (Target retrieval solution, S1700, DakoCytomation, Kyoto, Japan). Nonspecific binding was blocked by incubating sections in a 0.25% casein solution (Protein Block Serum-Free, X0909, DakoCytomation) for 5 minutes at room temperature, followed by blowing off the casein solution. The sections were subsequently incubated with polyclonal rabbit antihuman UBE4A C-term antibody (1:100, Abgent, San Diego, CA) diluted in antibody diluent (S2022, DakoCytomation) for 1 hour at 22°C and then rinsed 3 times with TBS-T. The reaction product was visualized by treating with HRP- and a secondary antibody-labeled polymer reagent (ChemMate Envision, K5027, DakoCytomation), rinsing, and reacting with Vector NovaRED Substrate Kit (Vector Laboratories, Burlingame, CA). Nuclear counterstaining was carried out by incubating sections in a hematoxylin solution for 1 minute. For chromogranin A staining, polyclonal rabbit antihuman chromogranin A (N1535, DakoCytomation) was used and the reaction was visualized by Metal Enhanced DAB Substrate Kit (Pierce Biotechnology). A section without primary antibody served as control.

RESULTS

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

Identification of UBE4A as a Candidate Autoantigen in CD

To identify autoantibodies present in the sera of patients with CD, a cDNA expression library was constructed from normal terminal ileum, an area frequently affected in patients with CD. A total of 5 × 105 phage plaques were screened with pooled sera from 3 randomly selected patients with CD. Positive phage clones were then screened with pooled sera from 10 randomly selected healthy controls (negative screening). Identities of phage clones were determined by sequencing. One clone that exhibited a strong reaction with sera from patients with CD was homologous with UBE4A (Access. No. NM_004788), which covers C-terminal 132 amino acids (12.3%) of whole UBE4A consisting of 1073 amino acids and contains a U-box domain (Fig. 1A). Reactivity of sera from patients with CD and healthy controls for C-terminal UBE4A is shown in Figure 1B.

thumbnail image

Figure 1. Amino acid sequence and immune reactivity of C-terminal UBE4A. (A) The amino acid sequence predicted from the nucleotide sequence of the cloned cDNA insertion is 132 residues long. The sequence has 100% identity with the C-terminal subunit of UBE4A. The underlined segment indicates the U-box domain. (B) Seroreactivity of patients with CD against phage clones encoding C-terminal UBE4A. λgt11 phages without insert were equally mixed and served as negative controls. Reactivity of sera from patients with CD (left) and healthy controls (right) for C-terminal UBE4A is shown.

Download figure to PowerPoint

ELISA for Anti-UBE4A IgG

To determine if a humoral immune response directed against UBE4A is associated with CD, serum samples obtained from patients with CD, patients with UC, and healthy controls were tested for their reactivity to a recombinant UBE4A protein. Because of difficulty in production of a recombinant full-length UBE4A, we used a recombinant C-terminal UBE4A to detect anti-UBE4A antibodies in patients' sera. In preliminary experiments, λgt11 phages encoding whole UBE4A exhibited equivalent seroreactivity with sera from patients with CD that were reactive with phages encoding C-terminal UBE4A (data not shown); thus considered, the C-terminal UBE4A contains the immunodominant epitopes.

Recombinant C-terminal UBE4A protein was produced by cleavage of GST-C-terminal UBE4A with site-specific protease to exclude the effect of reactivity against GST. The molecular size predicted from the amino acid sequence of 15.2 kDa and the purity of the recombinant protein was confirmed by SDS-PAGE (data not shown).

Anti-UBE4A IgG was measured using ELISA. The mean OD value of anti-UBE4A IgG in patients with CD was significantly higher than that in patients with UC or healthy controls (0.170 ± 0.119 versus, respectively, 0.089 ± 0.042, P = 0.0006; 0.086 ± 0.032, P < 0.0001, Mann–Whitney unpaired test) (Fig. 2A). Anti-UBE4A IgG was positive in 46.2% (18/39) of CD patients, in 7.1% (2/28) of UC patients, and in 3.3% (2/60) of healthy controls. The prevalence of anti-UBE4A IgG in patients with CD was also significantly higher than that in patients with UC (P = 0.0006) or that in healthy controls (P < 0.0001).

thumbnail image

Figure 2. (A) Quantitation of anti-UBE4A IgG measured by ELISA in serum from patients with CD, patients with UC, and healthy controls. The mean OD value for each group are shown as horizontal lines. The cutoff OD value for positivity of anti-UBE4A IgG was defined as an OD value more than 2 SD above the mean of the healthy control values. The broken line indicates a cutoff level (0.150). (B) Positive correlation of serum anti-UBE4A IgG levels with CDAI in patients with CD (r = 0.777, R2 = 0.603, P < 0.0001).

Download figure to PowerPoint

Clinical Characteristics of CD and Anti-UBE4A IgG

Patients with CD were divided into anti-UBE4A IgG-positive and anti-UBE4A IgG-negative groups. The 2 groups were compared with regard to gender, age, age at onset, duration of disease, family history, smoking, CDAI, perianal disease, ASCA, use of steroids, and history of surgery (Table 1).

Table 1. Clinical Characteristics of CD Patients and Anti-UBE4A IgG
CharacteristicsAnti-UBE4A IgGP-value
Negative (n = 21)Positive (n = 18)
  1. UBE4A, ubiquitination factor E4A; CD, Crohn's disease; CDAI, Crohn's Disease Activity Index; ASCA, Anti-Saccharomyces cervisiae antibodies; N.S., not significant.

Gender: male gender (%)17 (81.0)9 (50.0)N.S.
Age: mean years ± SD28.0 ± 9.932.9 ± 10.1N.S.
Age at onset: mean years ± SD21.2 ± 7.020.9 ± 7.8N.S.
Duration of disease: mean years ± SD7.8 ± 6.411.5 ± 5.8N.S.
Family history: n (%)1 (4.8)3 (16.7)N.S.
Smoker: n (%)6 (28.6)4 (21.1)N.S.
Disease activity (CDAI): mean ± SD137.5 ± 37.7219.6 ± 76.2P = 0.0002
Perianal disease: n (%)15 (71.4)11 (61.1)N.S.
ASCA positive: n (%)14 (66.7)15 (78.9)N.S.
Steroids use: n (%)4 (19.0)3 (16.7)N.S.
Previous surgery: n (%)4 (19.0)13 (68.4)P = 0.0013

CDAI values of anti-UBE4A IgG-positive patients were significantly higher than that of anti-UBE4A-negative patients (219.6 ± 76.2 versus 137.5 ± 37.7, P = 0.0002). Anti-UBE4A IgG-positive patients with CD had a higher incidence of surgery (P = 0.0013), but had no association with gender, age, age at onset, duration of disease, family history, smoking, perianal disease, ASCA status, or use of steroids.

The relationship between the level of anti-UBE4A IgG and CDAI in patients with CD is shown in Figure 2B. A significant correlation was found between serum anti-UBE4A IgG levels and CDAI scores (r = 0.777, R2 = 0.603, P < 0.0001).

Phenotype of CD and Anti-UBE4A IgG

Table 2 shows the relationship between the levels of anti-UBE4A IgG and phenotype of CD according to the Vienna classification. Because all patients with CD recruited into this study were diagnosed below age 40 years, the A1 category was subdivided into juvenile onset (0–16) and early adult onset (17–39). Comparison between these 2 groups was not significant. No significant association was found between disease location and anti-UBE4A IgG level. Comparison of anti-UBE4A IgG with disease behavior showed that patients with stricturing (B2) or penetrating (B3) disease had significantly higher anti-UBE4A IgG levels than patients with nonstricturing nonpenetrating (B1) disease (P = 0.0028).

Table 2. Vienna Classification of CD Patients and Corresponding Anti-UBE4A IgG
Vienna classificationAnti-UBE4A IgG (OD)P-valueAnti-UBE4A IgG
mean ± SDPositive/Total (%)
  • Statistical comparisons of the quantitative anti-UBE4A IgG levels between patient groups were performed using the

  • a

    Mann-Whitney unpaired test or the

  • b

    Kruskal-Wallis test.

Age at diagnosis    
 Below 40 years (A1)    
  Juvenile onset (0–16)0.227 ± 0.166N.S.a5/9(55.6)
  Early adult onset (17–39)0.166 ± 0.097 13/30(43.3)
 Over 40 years (A2) 0/0
Location    
 Terminal ileum (L1)0.186 ± 0.133N.S.b10/19(52.6)
 Colon (L2)0.084 ± 0.020 0/3(0)
 Ileocolon (L3)0.167 ± 0.108 8/17(47.1)
 Upper GI (L4) 0/0
Behavior    
 Nonstricturing    
 Nonpenetrating (B1)0.070 ± 0.018P = 0.0028b0/6(0)
 Stricturing (B2)0.182 ± 0.129 11/20(55.0)
 Penetrating (B3)0.197 ± 0.111 7/13(53.8)

Localization of UBE4A in Ileal Mucosa

UBE4A expression in ileum of normal and CD patients was examined by immunohistochemistry. Weak cytoplasmic expression of UBE4A, mostly above the nucleus, was seen in enterocytes and goblet cells of ileal mucosa from normal controls (Fig. 3A). In inflamed ileal mucosa of patients with CD, cytoplasmic expression of enterocytes and goblet cells was similar to that of controls, but strong UBE4A expression was detected in the cells that exhibited a characteristic morphology of enteroendocrine cells (Fig. 3B). Chromogranin A staining confirmed the localization of UBE4A to enteroendocrine cells (Fig. 3C).

thumbnail image

Figure 3. Immunohistochemical localization of UBE4A in ileal mucosa from controls and patients with CD (original magnification ×800). (A) Normal ileal mucosa; enterocytes and goblet cells show weak UBE4A immunostaining above nuclei. (B) Ileal specimens of patients with CD; there is a similar UBE4A expression in enterocytes and goblet cells, but intense immunoreactivity is seen in enteroendocrine cells (arrows). (C) Immunohistochemistry with chromogranin A in a serial section of that seen in (B). The cells indicated by arrows in (B) are positive in chromogranin A staining (arrowheads).

Download figure to PowerPoint

DISCUSSION

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

In recent years, several antibody markers such as antibodies to microbial antigens (ASCA, antibodies against E. coli outer membrane porin C [anti-OmpC], against a CD-related protein from Pseudomonas fluorescens [anti-I2], and against bacterial flagellin [anti-CBir1]) and pancreatic autoantibodies have been found to be useful for the diagnosis and differentiation of CD. ASCA has been shown to be found in about two-thirds of patients with CD.20, 21 Anti-OmpC, anti-I2, and anti-CBir1 were detected in approximately half of CD patients.22–24 Pancreatic autoantibodies which are present in about one-third of patients with CD have a high degree of specificity for CD,25 but most of the consistently detectable antibodies associated with CD do not correlate with disease activity.15, 23, 24 Anti-UBE4A levels are significantly correlated with CDAI, and also associated with requirement of surgery and stricturing or penetrating behavior; therefore, anti-UBE4A may be helpful in predicting disease severity. It should be noted that anti-UBE4A level was determined on a single sample for each patient in this study. A longitudinal study of anti-UBE4A levels over time in relationship to clinical activity or behavior changes needs to confirm the relationship. In addition, although the prevalence is relatively low, anti-UBE4A is highly disease-specific. The combined search for anti-UBE4A and ASCA may help increase the diagnostic precision and detection percentage for CD, because anti-UBE4A was not associated with ASCA status.

UBE4A is a mammalian homolog of yeast Ufd2, a U-box-type ubiquitin-protein ligase (E3). The ubiquitination-mediated protein degradation has been shown to play a significant role in multiple cellular processes such as cell cycle regulation, cellular signaling in response to stress and to extracellular signals, morphogenesis, secretory pathway, DNA repair, and organelle biogenesis.26, 27 When a cellular protein degrades, a ubiquitin-activating enzyme (E1) forms a thiolester bond with ubiquitin and then transfers the activated ubiquitin to the target protein via a ubiquitin-conjugating enzyme (E2) and an E3. The polyubiquitylated target proteins are recognized by the 26S proteasome and degraded. Ubiquitin on the target are removed by deubiquitylating enzymes and recycled. The U-box proteins were originally described as an E4 ubiquitination factor,28 but mammalian U-box proteins are thought to constitute a third family of E3 enzymes, in addition to the HECT and RING-finger families.29 UBE4A also might play diverse roles in collaboration with molecular chaperone systems. Mouse UFD2b, a homolog of human UBE4A, binds to DnaJc7,30 which appears to interact with Hsp90.31 DnaJc7 possesses 2 functional domains related to chaperone function: a TPR domain, which interacts with the COOH-terminal region of Hsp90 or Hsp70 (EEVD motif),32 and a J domain, which mediates the interaction with Hsc70.33

UBE4A is reported to be expressed in the skeletal muscle, kidney, and liver at high levels, but the expression levels of UBE4A in small and large intestine are low.34 Immunohistochemical analysis in the present study revealed that upregulation of UBE4A in enteroendocrine cells was seen in the inflamed ileum of CD but not in the normal ileum. Generation of antibodies to UBE4A in patients with CD might be the result of upregulation of UBE4A in inflamed intestinal mucosa.

The various peptide hormones produced by enteroendocrine cells control important physiological functions including growth and repair of the gut epithelium and motility of the gut wall.35 An increased number of enteroendocrine cells has been demonstrated in CD36 and TNBS-induced ileitis.37 Since secretory products from enteroendocrine cells could contribute to mucosal repair and protection from further epithelial injury in inflamed mucosa of CD, upregulation of UBE4A in these cells may be associated with altered secretion of gut hormones. However, the physiological role of UBE4A in production of gut hormones requires further study.

Another possibility of the generation of anti-UBE4A antibodies is the result of crossreactivity with enteric microbial antigens. Mounting evidence exists that serum antibody responses to self-antigens in IBD are likely to be triggered by crossreactivity with luminal antigens.38, 39 Epitope-specific crossreactivity between microbial antigens and self-antigens has been shown in some animal models to initiate autoimmunity.40 Identification of an immunodominant epitope of C-terminal UBE4A might provide information on the mimicking microbial antigen.

In conclusion, UBE4A was identified as a novel autoantigen in CD. The presence of autoantibodies against UBE4A in sera was significantly specific to CD and had a significant association with disease activity, behavior, and requirement of surgery. Our results suggest that anti-UBE4A antibodies are potentially useful markers to diagnose and predict disease severity in patients with CD. Whether UBE4A plays a role in the etiopathogenesis of CD remains to be determined.

Acknowledgements

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

The authors thank Drs. Mark W. Musch, Eugene B. Chang, and Laura E. Harrell of the University of Chicago for helpful suggestions and critical reading of the article.

REFERENCES

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