Xiaoli Wu1, Zhangyuan Liao2, Jing Ye2, Huiqing Dong2, Chaodong Wang1, Piu Chan1

1Department of Neurology and Key Laboratory Ministry of Education on Neurodegenerative Diseases, Xuanwu Hospital of Capital Medical University, Beijing 100053, China
2Department of Neurology, Xuanwu Hospital of Capital Medical University, Beijing 100053, China

Abstract
A total of 66 samples (from 27 cases with neuromyelitis optica, 26 cases with multiple sclerosis, and 13 cases with optic neuritis) were tested for aquaporin-4 antibody by a cell-based immunofluores-cence assay and an enzyme-linked immunosorbent assay. The sensitivities and specificities of the two assays were similar. We further analyzed an additional 68 patients and 93 healthy controls using the enzyme-linked immunosorbent assay. A Kappa test showed good consistency between the two methods in terms of detection of anti-aquaporin-4 antibody in the sera of neuromyelitis optica patients. No significant correlations were identified with onset age or disease duration, suggesting that aquaporin-4 antibody is a good marker for neuromyelitis optica. The enzyme-linked immuno-sorbent assay can be used for quantifying aquaporin-4 antibody concentrations and may be useful to dynamically monitor changes in the levels of aquaporin-4 antibody during disease duration.
Key Words: neuromyelitis optica; cell-based immunofluorescence assay; anti-aquaporin 4 antibody; enzyme-linked immunosorbent assay; long and extended spinal cord lesions; neural regeneration
 

INTRODUCTION
  
Neuromyelitis optica (NMO) (Devic’s syndrome) occurs more frequently in the oriental population[1-4], and is clinically distinct from multiple sclerosis (MS) with higher female preponderance, poorer prognosis, rarer brain MRI abnormalities, more longitudinally extensive (≥ three vertebral segments) spinal cord lesions, and an absence of oligoclonal IgG bands[5].
NMO-IgG was identified as a novel serological marker for NMO with a high sensitivity (58-76%) and specificity (94-100%) and has now been incorporated into the revised diagnostic criteria for NMO[6-7]. Subsequently, the water channel protein aquaporin-4 (AQP4), a component of the dystroglycan protein complex that is densely expressed in astrocytic foot process at the blood-brain barrier, was found to be the antigen for NMO IgG[8-9]. Brain lesions in NMO patients have been reported to be localized to the periventricular and periqueductal areas[1], and loss of AQP4 was observed in the active perivascular lesions of NMO patients[10]. These findings strongly suggest that anti-AQP4 antibody (AQP4-Ab) plays a crucial role in the pathogenesis of NMO, and that measurement of AQP4-Ab levels could be of considerable value in distinguishing NMO from MS. However, it is still controversial whether NMO is an independent inflammatory disease of the central nervous system.
The role of AQP4-Ab in NMO has been investigated in several studies with small samples, providing limited information on its relationship with clinical features and its usefulness in differential diagnosis. Moreover, AQP4-Ab has not been extensively studied in Chinese mainland.
The cell-based immunofluorescence assay (CBA) has been the best method for testing for AQP4-Ab, but it has not been extensively used in the clinic because it is time consuming and costly. A new enzyme-linked immunosorbent assay (ELISA) has the potential to be automated, which would make it possible to screen a large number of samples at a relatively low cost and within a short time. The present study tested for both
a semi-quantitative CBA and the quantitative ELISA to investigate the consistency of these two methods, and retrospectively analyzed clinical characteristics related to AQP4-Ab in NMO patients.

RESULTS

Clinical characteristics of subjects
One-hundred and thirty four patients, including 34 NMO patients, 12 atypical NMO patients, 63 MS patients, 12 acute transverse myelitis patients, 13 optic neuritis patients, and 93 healthy controls, were studied. Their clinical and image features are shown in Table 1. Compared with MS patients, NMO patients had higher female to male ratios and Expanded Disability Status Scale (EDSS) scores, and were more likely to have optic nerve involvement, fixed attack-related weakness in one or more limbs, rare brain MRI abnormalities (brain lesions not fulfilling the Barkhof brain lesions[11]) and transverse myelitis. There were no significant differences in onset age (P = 0.192) between NMO and MS patients. In addition, NMO patients had more frequent longitudinally extensive spinal cord lesions (LESCLs) (100% vs. 12.6%, χ2 = 68.55; P = 0.000), and normal brain MRI scans at the time of onset.
Sensitivity and specificity of the CBA and ELISA assays
There were initially 66 samples (27 cases with NMO, 26 cases with MS, 13 cases with optic neuritis) tested for AQP4-Ab by CBA. Twenty-three of the NMO cases (85.2%) tested positive (titer ranging from 1: 4 to 1:
1 024), while five of the optic neuritis cases (38.5%) tested positive (titer ranging from 1: 20 to 1:320). All 26 MS cases were negative, giving 100% specificity for differential diagnosis of NMO from MS. When tested by ELISA, 17 of the NMO cases (63%), three of the optic neuritis cases (23.1%) and two of the MS cases (7.6%) tested positive with a concentration ranging from 7.1-160 U/mL (mean 85.6 ± 64.6 U/mL), giving a specificity of 95.8% for differential diagnosis of NMO from MS. There was no difference in seropositivity rates of AQP4-Ab detection between the ELISA and CBA  (χ2 = 3.471, P = 0.062). There was good consistency in the detection of AQP4-Ab in NMO patients (Kappa = 0.681; P = 0.00), and a significant correlation (r = 0.689, P = 0.000) was found between the results of the two methods (supplementary Figure 1 online).
AQP4 antibody seropositivity among different types of multiple sclerosis patients
We further analyzed an additional 68 patients and 93 health controls by ELISA raising the total number of samples to 227. Elevated serum AQP4-Ab was found in 76.5% of NMO patients, 66.7% of atypical NMO patients, 25% of acute transverse myelitis patients, 23.1% of optic neuritis patients, and 4% of MS patients, with concentrations ranging from 6.2-160 U/mL (mean 92.4 ± 62.6 U/mL), none of the healthy controls were positive, giving 97.4% specificity for diagnosis of NMO from healthy controls and MS patients (Table 1).
Associations between serum AQP4 antibody concentrations and clinical features
Among NMO cases (including the atypical NMO cases), only 33% (2/6) of the males were positive for AQP4-Ab, while 80% (32/40) of the females were positive (P = 0.033). No significant correlations were determined between AQP4-Ab concentration and onset age (r = 0.049; P = 0.784) or disease duration (r = 0.316; P = 0.069). Eight of the 34 typical NMO patients (23.5%) were negative for AQP4-Ab by ELISA. High EDSS scores were more frequent in NMO patients positive for AQP4-Ab (n = 26) than in those who were negative for AQP4-Ab (n = 8) (t = 2.28, P = 0.029). Multivariate linear regression revealed a positive correlation between serum AQP4-Ab concentration and EDSS score (r = 0.269; P = 0.003; Figure 1). There was also a positive correlation between EDSS scores and the number of lesioned segments of the spinal cord on MRI (r = 0.359, P = 0.014). However, no correlation was found between the concentration of AQP4-Ab and the number of lesioned segments of spinal cord (P = 0.686).


DISCUSSION

Studies to date in Westerners, Japanese, Korean and Hong Kong residents have mainly focused on the detection of AQP4-Ab and the distinction between NMO and MS, while studies of the seropositivity rate and clinical data of NMO patients from Chinese mainland are rare (supplement Table 1 online)[1-4, 12-20]. Our results confirmed that both CBA and ELISA facilitate distinction of NMO from MS by detection of AQP4-Ab, which mostly exists in NMO patients (elevated in 76.5% of NMO cases by ELISA) but not MS patients (elevated in 6.3% of MS patients by ELISA), and showed that NMO patients positive for serum AQP4-Ab are clinically different from MS patients: women were more susceptible, and had more severe clinical manifestations and normal brain imaging, suggesting that AQP4-Ab is a good marker for NMO.
In this study, AQP4-Ab concentration was significantly positively associated with EDSS, and higher EDSS scores were more frequent in NMO patients positive for AQP4-Ab than in those who were negative. There was no correlation between antibody positivity and either onset age or disease duration among NMO patients. These results are consistent with those of previous reports[2, 13] and demonstrate that NMO patients positive for serum AQP4-Ab have more severe clinical manifestations than those who are negative for the antibody, suggesting that AQP4-Ab is not only a good marker for NMO, but also that it is associated with disease severity[12].
Several studies have reported that AQP4-Ab is more likely to appear in patients with LESCLs[2, 21-22], especially during the initial attack in one-third of these patients[13, 23]. Although all NMO patients had LESCLs, and only 12.6% of MS patients had LESCLs, there was no direct relationship between AQP4-Ab positivity and LESCLs. A positive correlation between EDSS scores and LESCLs was reported by Matsuoka et al [24]. Eight of 12 (66.7%) atypical NMO patients were positive for AQP4-Ab and met the 2006 NMO criteria[7]. The other four atypical NMO patients, who were negative for AQP4-Ab, could be defined as having MS with durations of 4, 18, 24 and 48 months. From our experience of the definition of NMO, without knowing AQP4-Ab status, patients with optic neuritis and myelitis with LESCLs were easily diagnosed from clinical practice, while those without LESCLs should be monitored for clinical progress in follow-ups, and serum AQP4-Ab positivity should be determined to facilitate differentiation of NMO from MS.
The AQP4-Ab titers may be correlated with disease activity in NMO attacks, but the role of the AQP4-Ab in lesion formation and pathophysiology of NMO is not yet known. Studies in vitro have demonstrated that human AQP4-Ab significantly exacerbates inflammation and the demyelization processes of experimental autoimmune encephalomyelitis[25-26]. A growing amount of evidence suggests that NMO-IgG is pathophysiological effector in NMO through its binding to the astrocytic transmembrane protein AQP4 and initiating classical complement activation[27]. Traditionally, optic nerve and spine are particularly vulnerable to antibody-mediated injury due to the inherent weakness of the blood-brain barrier[28-29], and human serum AQP4-Ab is pathogenic only when reaching the central nervous system at sites of brain inflammation[30]. It seemed that AQP4-Ab, inflammation of central nervous system, astrocytic transmembrane protein (AQP4) and complement are necessary conditions for the pathogenesis of NMO, but how can the low level of AQP4-Ab positivity in MS patients be explained? Familial NMO is indistinguishable from sporadic NMO[31], and the different frequencies of seronegative and seropositive NMO in the pediatric population appear to be dependent on ethnicity[32]. Taken together, these data suggest the existence of complex genetic susceptibility in NMO. Therefore, the presence of AQP4-Ab could not directly lead to NMO; other conditions, such as specific genotypes and differently activated immune cell populations, are required to modify disease development.
The ELISA in our study had similar sensitivity (76.5% vs. 85.2%) for detecting AQP4-Ab to the CBA. However, ELISA has some advantages in clinical use with relatively low cost and time savings. Four NMO cases had serum that was negative by both methods; clinical data showed that these serum samples were obtained within 1–3 months after high-dose intravenous methyl-prednisolone therapy. The antibody titer in these patients might have been below the cutoff level for detection after high-dose intravenous methyl- prednisolone treatment[2]. A larger-scale study is needed to compare the AQP4-Ab status before and after high-dose intravenous methyl-prednisolone treatment, and to address the reason why typical or atypical NMO patients might be truly negative for AQP4-Ab. Four MS cases were positive for AQP4-Ab, with concentrations of 7.520, 17.974, 150.429, and 17.101 U/mL. All of these patients had extensive brain lesions with a duration ≥ 3 years and multiple patchy or LESCL spinal lesions. Previous studies have reported that AQP4-Ab titers were higher in young adult patients (<45 years old) with extensive or large cerebral lesions on MRI than in older patients (≥ 45 years old) with rare brain lesions, with a nearly 3-fold increase in AQP4-Ab titers before relapses[2, 16]. The present study showed that the presence of AQP4-Ab might have prognostic significance, indicating a more severe disease course in relapse. From our experience, CBA may be a more specific assay for making a differential diagnosis between NMO and MS, while quantitative ELISA assay may be more useful for monitoring disease severity and more easy to use in the clinic.
In conclusion, both ELISA and CBA showed good consistency in detecting serum AQP4-Ab in NMO patients. However, ELISA can also be used to quantify AQP4-Ab concentrations and may be useful to dynamically monitor changes in the level of AQP4-Ab during disease duration. The current study demonstrates that AQP4-Ab is a reliable marker for NMO and may predict disease severity.

SUBJECTS AND METHODS

Design
Molecular biology and clinical prospective analysis.
Time and setting
The study was performed at the Xuanwu Hospital of Capital Medical University, Beijing, China from May 2004 to February 2010.
Subjects
One-hundred and thirty four patients were recruited from the clinical center for MS of the Xuanwu Hospital from May 2004 to February 2010, including 68 subjects from the baseline of Phase II clinical trial of interferon 1β between May 2005 and May 2007. Patients were diagnosed according to the following criteria. Clinically definite MS was diagnosed according to the revised McDonald criteria (2005)[33]. NMO was diagnosed if fulfilling all items of the 2006 NMO criteria [7], except for NMO-IgG status. Patients with optic neuritis and myelitis (spanning less than three vertebral segments) who fulfilled any other items of the 2006 NMO criteria, except for NMO-IgG status, were still recorded as having atypical NMO. Acute transverse myelitis was diagnosed according to the recommendations of the transverse myelitis consortium working group[32]. LESCLs were defined as lesions extending over more than three vertebral segments on spinal MRI [10]. Optic neuritis was diagnosed in patients that experienced only a single attack and showed no brain lesions on MRI. The MRI imaging results were reviewed by the authors who were blinded to the outcomes of serological assays, and all subjects with tumors, infarction (including lacunar infarction) and white matter degeneration under an axial T1-weighted spin-echo sequence MRI, were excluded. Clinical severity was measured using the EDSS[34].
All 93 controls were recruited from the Beijing community and excluded if they had any neurological diseases. All participants provided written informed consent. All blood samples for analysis were collected during relapse before receiving interferon 1β or corticosteroids, or immunosuppressants, and stored at -80°C until use. The study was performed in accordance with the World Medical Association Declaration of Helsinki.
Methods
CBA for anti-AQP4 antibody
The CBA was performed using human embryonic kidney 293 cells (HEK-293) transfected with AQP4 as reported previously[35]. Samples were tested with the examiner blinded to the clinical category. The cutoff of antibody titer was 1: 4, and antibody-positive sera were titrated in serial two-fold dilutions to ascertain the maximum dilution for positive staining.
ELISA for anti-AQP4 antibody 
The RSR AQP4 autoantibody ELISA kit was used for the quantitative determination of AQP4-Ab in human serum as described in its instruction booklet (RSR Ltd., Avenue Park Pentwyn Cardiff CF23 8HE, UK; website: www.rsrltd.com). In brief, samples and calibrators were added to 96 wells coated with AQP4, to which biotinylated AQP4 was then added. After incubation at room temperature for 2 hours, a second incubation step involving addition of streptavidin-peroxidase was performed. After washing, addition of 3, 3’, 5, 5’-tetramethlybenzidine substrate would result in formation of a blue color, which was stopped by the addition of the stop solution after which the color changed to yellow. The absorbance of the yellow reaction mixture was read at 450 nm using an ELISA plate reader (Thermo Scientific Multiskan MK3, Thermo Fisher Company; MA). A calibration curve was established by plotting calibrator concentration on the X-axis against the absorbance of the calibrators on the Y-axis. Then, the concentrations in patients’ sera could be read off the calibration curve. A concentration of AQP4-Ab < 5.0 U/mL (arbitrary RSR units) was noted as negative and ≥ 5.0 U/mL was noted as positive. Samples were tested with the examiners blinded to the clinical category.
Statistical analysis
The consistency of the two methods was analyzed by Kappa test and the relationship between the results was analyzed by Spearman correlation analysis. Clinical, serological and radiological characteristics were analyzed using Fisher’s test, the Mann-Whitney U test, and multiple linear regression analysis with the duration of disease as s covariate. Statistical analyses were performed using SPSS 11.5 software (SPSS, Chicago, IL, USA). A P value < 0.05 was considered statistically significant.

Author contributions: Xiaoli Wu and Piu Chan participated in study concept and design. Xiaoli Wu and Zhangyuan Liao were in charge of acquisition of data. Xiaoli Wu was responsible for drafting of the manuscript. Chaodong Wang and Piu Chan participated in critical revision of the manuscript for important intellectual content. Piu Chan obtained funding. Xiaoli Wu, Zhangyuan Liao, Jing Ye and Huiqing Dong participated in administrative, technical, and material support. Xiaoli Wu, Chaodong Wang and Piu Chan supervised the study.
Conflicts of interest: None declared.
Funding: This work was supported by grants from the Ministry of Sciences and Technology of China, No. 2006AA02A408, 2008ZX09312-014.
Ethical approval: This study was approved by the Ethics Committee of Xuanwu Hospital, China.
Acknowledgments: We are very grateful to Professor Shengli Xu from Xuanwu Hospital of Capital Medical University, China for technical help.
Supplementary information: Supplementary data associated with this article can be found in the online version by visiting www.nrronline.org and entering Vol. 6, No. 32, 2011 after selecting the “NRR Current Issue” button on the page.

REFERENCES

[1] Nakashima I, Fujihara K, Miyazawa I, et al. Clinical and MRI features of Japanese patients with multiple sclerosis positive for NMO-IgG. J Neurol Neurosurg Psychiatry. 2006;77(9):1073-1075.
[2] Takahashi T, Fujihara K, Nakashima I, et al. Anti-aquaporin-4 antibody is involved in the pathogenesis of NMO: a study on antibody titre. Brain. 2007;130(Pt 5):1235-1243.
[3] Chan KH, Ramsden DB, Yu YL, et al. Neuromyelitis optica-IgG in idiopathic inflammatory demyelinating disorders amongst Hong Kong Chinese. Eur J Neurol. 2009;16(3):310-316.
[4] Chan KH, Kwan JS, Ho PW, et al. Aquaporin-4 autoantibodies in neuromyelitis optica spectrum disorders: comparison between tissue-based and cell-based indirect immunofluorescence assays. J Neuroinflammation. 2010;7:50.
[5] Wingerchuk DM, Hogancamp WF, O'Brien PC, et al. The clinical course of neuromyelitis optica (Devic's syndrome). Neurology. 1999;53(5):1107-1114.
[6] Lennon VA, Wingerchuk DM, Kryzer TJ, et al. A serum autoantibody marker of neuromyelitis optica: distinction from multiple sclerosis. Lancet. 2004;364(9451):2106-2112.
[7] Wingerchuk DM, Lennon VA, Pittock SJ, et al. Revised diagnostic criteria for neuromyelitis optica. Neurology. 2006;66(10): 1485-1489.
[8] Lennon VA, Kryzer TJ, Pittock SJ, et al. IgG marker of optic-spinal multiple sclerosis binds to the aquaporin-4 water channel. J Exp Med. 2005;202(4):473-477.
[9] Vizuete ML, Venero JL, Vargas C, et al. Differential upregulation of aquaporin-4 mRNA expression in reactive astrocytes after brain injury: potential role in brain edema. Neurobiol Dis. 1999;6(4): 245-258.
[10] Misu T, Fujihara K, Nakamura M, et al. Loss of aquaporin-4 in active perivascular lesions in neuromyelitis optica: a case report. Tohoku J Exp Med. 2006;209(3):269-275.
[11] Barkhof F, Filippi M, Miller DH, et al. Comparison of MRI criteria at first presentation to predict conversion to clinically definite multiple sclerosis. Brain. 1997;120 ( Pt 11):2059-2069.
[12] Akman-Demir G, Tüzün E, Waters P, et al. Prognostic implications of aquaporin-4 antibody status in neuromyelitis optica patients. J Neurol. 2011;258(3):464-470.
[13] Takahashi T, Fujihara K, Nakashima I, et al. Establishment of a new sensitive assay for anti-human aquaporin-4 antibody in neuromyelitis optica. Tohoku J Exp Med. 2006;210(4):307-313.

[14] Kim SH, Kim SM, Vincent A, et al. Clinical characteristics, prognosis, and seropositivity to the anti-aquaporin-4 antibody in Korean patients with longitudinally extensive transverse myelitis. J Neurol. 2010;257(6):920-925.
[15] Marignier R, De Sèze J, Vukusic S, et al. NMO-IgG and Devic's neuromyelitis optica: a French experience. Mult Scler. 2008;14(4): 440-445.
[16] Jarius S, Frederikson J, Waters P, et al. Frequency and prognostic impact of antibodies to aquaporin-4 in patients with optic neuritis. J Neurol Sci. 2010;298(1-2):158-162.
[17] Waters P, Jarius S, Littleton E, et al. Aquaporin-4 antibodies in neuromyelitis optica and longitudinally extensive transverse myelitis. Arch Neurol. 2008;65(7):913-919.
[18] Fazio R, Malosio ML, Lampasona V, et al. Antiacquaporin 4 antibodies detection by different techniques in neuromyelitis optica patients. Mult Scler. 2009;15(10):1153-1163.
[19] Wu JS, Matsushita T, Carroll WM, et al. Low sensitivity of anti-aquaporin-4 antibody in multiple sclerosis, longitudinally extensive spinal cord lesions and neuromyelitis optica in Australians. Neurology Asia. 2007;12:149-150.
[20] Adoni T, Lino AM, Marchiori PE, et al. Seroprevalence of NMO-IgG antibody in Brazilian patients with neuromyelitis optica. Arq Neuropsiquiatr. 2008;66(2B):295-297.
[21] Lu Z, Qiu W, Zou Y, et al. Characteristic linear lesions and longitudinally extensive spinal cord lesions in Chinese patients with neuromyelitis optica. J Neurol Sci. 2010;293(1-2):92-96.
[22] Tanaka K, Tani T, Tanaka M, et al. Anti-aquaporin 4 antibody in selected Japanese multiple sclerosis patients with long spinal cord lesions. Mult Scler. 2007;13(7):850-855.
[23] Weinshenker BG, Wingerchuk DM, Vukusic S, et al. Neuromyelitis optica IgG predicts relapse after longitudinally extensive transverse myelitis. Ann Neurol. 2006;59(3):566-569.
[24] Matsuoka T, Matsushita T, Kawano Y, et al. Heterogeneity of aquaporin-4 autoimmunity and spinal cord lesions in multiple sclerosis in Japanese. Brain. 2007;130(Pt 5):1206-1223.
[25] Bradl M, Misu T, Takahashi T, et al. Neuromyelitis optica: pathogenicity of patient immunoglobulin in vivo. Ann Neurol. 2009; 66(5):630-643.
[26] Kinoshita M, Nakatsuji Y, Kimura T, et al. Neuromyelitis optica: Passive transfer to rats by human immunoglobulin. Biochem Biophys Res Commun. 2009;386(4):623-627.
[27] Hinson SR, Pittock SJ, Lucchinetti CF, et al. Pathogenic potential of IgG binding to water channel extracellular domain in neuromyelitis optica. Neurology. 2007;69(24):2221-2231.
[28] Lucchinetti CF, Mandler RN, McGavern D, et al. A role for humoral mechanisms in the pathogenesis of Devic's neuromyelitis optica. Brain. 2002;125(Pt 7):1450-1461.
[29] Guy J, Rao NA. Acute and chronic experimental optic neuritis. Alteration in the blood-optic nerve barrier. Arch Ophthalmol. 1984; 102(3):450-454.
[30] Matiello M, Kim HJ, Kim W, et al. Familial neuromyelitis optica. Neurology. 2010;75(4):310-315.
[31] Huppke P, Blüthner M, Bauer O, et al. Neuromyelitis optica and NMO-IgG in European pediatric patients. Neurology. 2010;75(19): 1740-1744.
[32] Transverse Myelitis Consortium Working Group. Proposed diagnostic criteria and nosology of acute transverse myelitis. Neurology. 2002;59(4):499-505.
[33] Polman CH, Reingold SC, Edan G, et al. Diagnostic criteria for multiple sclerosis: 2005 revisions to the "McDonald Criteria". Ann Neurol. 2005;58(6):840-846.
[34] Kurtzke JF. Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology. 1983;33(11):1444-1452.
[35] Liao ZY, Ye J, Guan YQ, et al. Construction and clinical application of lentivirus-AQP4 expressing vector. Zhonghua Yi Xue Za Zhi. 2010;90(3):208-212.
 (Edited by Cheng XD, Gu T/Qiu Y/Song LP)