Anti-α-enolase Antibodies in Serum from Pediatric Patients Affected by Inflammatory Diseases: Diagnostic and Pathogenetic Insights

Human glycolytic enzyme α-enolase was associated with human diseases and with inflammation. An ELISA test was developed to measure anti-α-enolase AAE IgG and AAE IgA in the serum from patients affected by inflammatory diseases with the purpose to evaluate it as a novel diagnostic marker. 80 healthy blood donors and 194 paediatric patients affected by Juvenile idiopathic arthritis (JIA), celiac disease (CD), Crohn's Disease (CrD), hereditary periodic fever (HPF), and PFAPA syndrome were included in the study. HPF patients showed high levels of AAE antibodies, whereas JIA, CD, and CrD presented only partial results. Benign fevers such as PFAPA were almost negative for AAE Abs. These findings suggested that the genetic dysfunction of inflammasome associated with HPF could lead to the formation of AAE Abs that could be used for an early and easy diagnosis.

Recently it was shown that during the inflammatory process, α-Enolase, together with other glycolytic enzymes, might be a substrate of caspase-1 in the newly described caspase-1 digestosome [10].Caspase-1 leads to the activation of IL-1β and NF-kB, through the inflammasome, and also to the inflammatory cell death called pyroptosis.Pyroptosis, eliminating macrophages, probably contributes to the end of inflammation.Because glycolysis is essential for macrophage survival and activation, the cleavage of α-Enolase and the glycolysis substrates, which results in reduction of cellular glycolysis, seems to be an essential step toward cell death [10].
We hypothesized that high amount of α-Enolase-derived peptides and/or their inefficient removal from the milieu could be immunogenic and that high amount of antibodies anti-α-enolase could be an epiphenomenon in inflammatory processes that are characterized by an important caspase-1 activation.The evaluation of antibodies anti-α-enolase (AAE Abs) could provide a novel tool for better definition of differential diagnosis in patients with inflammatory disorders.

Materials and Methods
184 patients were recruited from the Paediatric Division of the Institute of Maternal and Child Health IRCCS "Burlo Garofolo" (Trieste, Italy) and 10 (5 FMF and 5 CINCA) from the II Paediatric Division of the Institute of Child Health IRCCS "G.Gaslini" (Genoa, Italy).All the sera were collected at the moment of the diagnosis, before starting any kind of treatment.
JIA was diagnosed according to the criteria of the International League of Associations for Rheumatology [11].The CD diagnosis was based on the evaluation of clinical features, ELISA anti-transglutaminase antibody (tTG) quantification (Eu-tTG kit, Eurospital, Trieste, Italy), and the presence of the HLA DQ2/DQ8 heterodimer (Eu-Gen Risk, Eurospital) and confirmed by intestinal biopsy following the indications recommended by the European Society for Paediatric Gastroenterology, Hepatology and Nutrition [12].CrD was diagnosed according to clinical, endoscopic, and histologic criteria, as suggested by international guidelines [13].HPFs were diagnosed according to clinical parameters specifically related to each kind of disease [14] and to the molecular analysis of respective genes (CINCA syndrome: CIAS1/NALP3; FMF: MEFV; MKD MVK; TRAPS: TNFRS1A).PFAPA syndrome was diagnosed by the meaning of clinical signs [15] and after the exclusion of any HPF by the meaning of the analysis of the HPF genes.
Viral or pathogen infections as well as cancer have been previously excluded.
Moreover, 80 healthy Italian subjects (35 males/45 females; mean age 7.1 years ± 4.8 SD) matched for age and ethnicity with the patients, and not related to the patients group, were included as controls (HC).One patient affected by vasculitis was also included as positive control [16].
The study protocol was approved by independent ethics committees at each study site.Patients or their parents provided written informed consent.

ELISA for Anti-α-enolase
IgG and IgA Antibodies.Two ELISA tests were developed using recombinant α-Enolase Microtiter plates (96-well Nunc Maxisorp) were coated with purified unlabelled recombinant α-Enolase at 9.5 μg/mL in PBS overnight at 4 • C. Plates were then washed with 0.05% Tween-20 in PBS (wash buffer) and saturated with 0.1% Tween-20 in PBS for 20 • C at room temperature (RT).
The ELISA was performed by diluting healthy controls and patients serum 1 : 100 in PBS for IgG test and 1 : 200 for IgA test.The serum from children affected by vasculitis was used as positive control as previously reported [16].All incubations were performed at RT, and all volumes were 100 μl per well.The plates were incubated for 1 hour and washed three times with wash buffer.
The phosphatise-conjugated anti-human IgA (diluted 1 : 2000; Sigma Aldrich) or anti-human IgG (diluted 1 : 2000; Sigma Aldrich) was then added to the wells, incubated for 1 hour, and plates were washed three times.The assay was developed by the addition of 1 g/L p-nitrophenyl phosphate (Sigma Aldrich) to the plate.The data was acquired by measuring absorbance at 405 nm and analysed using GraphPad Prism version 5.0.Results were expressed as arbitrary unit (AU)/mL.

Data Analysis.
Statistical analysis was carried out using the GraphPad Prism software version 5.0.For the comparison of groups, independent Student t-test and ANOVA with Scheffe's post hoc test were used.P values of less than 0.05 were regarded as significant.

Results
Demographic characteristics (mean age, male/female ratio) were compared between healthy controls (HCs) and each group of patients and data were reported in Table 1.
80 sera from healthy blood donors included as controls were tested for the presence of AAE IgA and IgG antibodies.The mean and standard deviation (SD) was 15.44 UA/mL ± 5.21 for AAE IgA and 16.79 UA/mL ± 5.81 for AAE IgG.The cut-off values calculated for each Ig class (mean + 2SD) were 25.86 and 28.41, respectively.
The results for the vasculitis patient serum were 29.92 AU/mL for AAE IgA and 54.08 AU/mL for AAE IgG.
A larger number of serum samples from patients with chronic inflammatory diseases were procured to test the utility of the ELISA in detecting AAE antibodies.
Among the patient groups significant differences were reported for PFAPA versus all the other groups as AAE IgA and IgG, JIA versus CD (IgG: P = 0.006), CD versus CrD (IgG: P = 0.003), and CrD versus HPF (IgG: P = 0.025) (See Supplementary Data S1 in supplementary material available online at doi: 10.1155/2011/870214).
Overall, the samples represent an average of 41% positive AAE IgA and 64% AAE IgG above the cut-offs established for the assay.Despite the high specificity of the AAE Abs test (100% for IgA and IgG), the sensitivity varied within the cohorts ranging from 33 to 65% for AAE IgA and from 65 to 87% for AAE IgG (Table 2).These results show that the AAE ELISA can distinguish differences between serum from healthy donors and patients with an important inflammatory disease.
When patients affected by JIA were stratified for joint involvement and severity of the disease (polyarticular, pauciarticular, and systemic JIA) a significant difference was observed for AAE IgA titre in polyarticular JIA versus HC (P = 2.0exp−4) and poliarticular JIA versus systemic JIA (P = 0.004); for AAE IgG all the subgroups were significantly different compared to HC (Table 3).
Among the group of hereditary periodic fever, 67% of CINCA patients (6/9) were AAE IgA+ and all the 9 CINCA patients (100%) were AAE IgG+.80% (4/5) of FMF were both AAE IgA+ IgG+, and 20% (1/5) were double negative for the test.60% (3/5) of MKD were AAE IgA+, 60% (3/5) were AAE IgG+, and 40% (2/5) were AAE IgA+ IgG+.The results of 1 MKD patient, to note the one with the most severe clinical presentation, were fully negative.The Table 2: AAE IgA and AAE IgG in healthy controls and patients.Levels and percentage of positivity of anti-α-enolase IgA and IgG in healthy controls (HCs) and patients with juvenile idiopathic arthritis (JIA), celiac disease (CD), Crohn's disease (ChD), hereditary periodic fevers (HPF), and PFAPA syndrome (PFAPA) were reported as mean and standard deviation (SD).AAE levels are expressed as arbitrary units per mL (AU/mL) and AAE positivity as number and percentage of sera with titre above the cut-off.Patients' AAE levels are compared to healthy control by t-test and P values are reported.TRAPS patient serum resulted in being negative for the AAE detection.Also in this cohort the ELISA results did not report any significant difference within the 4 HPFs (Table 3).

Discussion
This study showed that autoantibodies directed against human α-Enolase are present in different inflammatory conditions in paediatric patients, such as juvenile idiopathic arthritis, celiac disease, Crohn's disease, and hereditary periodic fevers, but not in benign fever PFAPA.Initially described as diagnostic marker for rheumatoid arthritis (RA), AAE antibodies seem to occur in several inflammatory disorders [17], lacking the proposed role of specific diagnostic test for RA.
As previously reported for RA in adults, AAE antibodies were present in JIA.In particular, AAE IgG was found in 87% of JIA sera (Table 2) in respect to 25% and 6% of positivity in RA sera demonstrated with immune-blotting technique by other authors [18,19].The rate of positive sera among our paediatric group was particularly high in the systemic presentation of the disease (AAE IgG 100%), whereas AAE IgA levels were more heterogeneous (Table 3).Although JIA does not present mucosa involvement, these results could be explicated taking in account a hypothesis of polyclonal activation in this inflammatory disease with autoimmune etiopathology, as described by other authors [17].Moreover the presence of AAE Abs did not correlate with a bad prognosis for patients with JIA as suggested by Saulot in early rheumatoid arthritis [19], even if we could only observe patients in a short follow-up period (6-12 months).AAE antibodies were previously described, with a proteomic approach, in adults affected by celiac disease [5] and by inflammatory bowel disease [20].In celiac subjects we observed a low titre of AAE IgA (33%) and a quite elevated one of AAE IgG (65%) (Table 2).This suggests that the activation of autoreactive B cells against α-Enolase could be not mucosa specific, but perhaps a systemic event.
74% of Crohn's patients presented AAE IgG compared to data reported by Vermeulen et al. (AAE IgG 50%) [20], highlighting the massive inflammatory state of these subjects.
These results confirm previously reported findings about the presence of AAE Abs in the serum of patients affected by important inflammatory conditions, although the variable sensitivity of the test does not sustain our initial enthusiasm.
For the first time to our knowledge AAE antibodies were described in subjects affected by recurrent fevers.Our results clearly showed a difference between benign fever PFAPA, which was almost negative for AAE Abs, and hereditary periodic fevers (Table 2).Moreover, among the HPFs, the ones characterized by a genetic defect in inflammasomesuch as CINCA syndrome and Familial Mediterranean fever-are more frequently positive for the test (Table 3).
Hereditary periodic fevers represent a big diagnostic challenge due to the wide and often unspecific clinical spectrum that accounts for possible difficulties in differential diagnosis with chronic inflammatory disorders and also with the benign PFAPA syndrome.However an early diagnosis is very important in these conditions to start the appropriate treatment and to prevent unwanted long-term complications in HPF or, on the other hand, to rapidly exclude a severe disease in PFAPA and avoid inappropriate and often expensive investigations.For this reason, despite the limited size of our cohorts, we believe that AAE antibodies titre could be a novel useful inflammatory marker for these rare syndromes.
Finally, our findings are interesting also considering their ethiopathologic implications.The quite different levels of anti-α-enolase antibodies that we found in our patients seem to support our hypothesis that the immunogenicity of α-Enolase could be a characteristic event in caspase-1-related diseases.We underlined that the highest AAE Abs levels were observed in CINCA syndrome, Familial Mediterranean fever, and the inflammasome-related syndromes.HIDS patients presented high AAE Abs, and this disease was recently associated with the autoactivation of caspase-1, too [21].
Moreover, one of the most known susceptibility genes for Crohn's disease is NOD2.NOD2 belongs to the NOD-Like Receptor (NLR) superfamily comprising also CIAS1/NALP3 (mutated in CINCA syndrome).NOD2 induces NF-kB, similarly to inflammasome, and it was supposed to activate caspase-1 interacting with NALP1 [22], so it is possible that the 2 signalling cascades (inflammasome and NODsome), could converge and affect the same molecular targets, such as glycolytic enzymes destruction before cell death.A dysfunction in inflammasome was also hypothesized for early rheumatoid arthritis [23] and celiac disease [24] but Table 3: Serum level and percentage of AAE IgA+ and AAE IgG+ in JIA stratified for clinical presentation and in HPF classified for genetic defect.Levels of anti-α-enolase IgA and IgG in patients with pauciarticular, polyarticular, or systemic juvenile idiopathic arthritis (JIA) CINCA syndrome, FMF, MKD, and TRAPS hereditary periodic fevers (HPF) were reported as mean and standard deviation (SD).Results are expressed as arbitrary units per mL (AU/mL).Patients' AAE levels are compared by t-test ( * = P < 0.05).Percentage of AAE IgA and IgG positivity in each disease class is also reported.

Table 1 :
Patients and controls demographic data.Mean age and males/female ratio were reported for healthy controls (HC) and in patients with juvenile idiopathic arthritis (JIA), celiac disease (CD), Crohn's disease (ChD), hereditary periodic fevers (HPFs), and PFAPA syndrome (PFAPA).Age and males/females ratio from patients were compared with HC using an independent t-test and a Fisher exact test, respectively.
(a gift from Dr. Edomi; University of Trieste) as the capture reagent to detect AAE IgA or AAE IgG serum antibodies.Phosphatase-conjugated anti-human IgA or anti-human IgG (Sigma Aldrich) were used as the detection antibody.