- Email: [email protected]
Elevated plasma high-mobility group box 1 protein is a potential marker for neuromyelitis optica
Neuroscience 226 (2012) 510–516
ELEVATED PLASMA HIGH-MOBILITY GROUP BOX 1 PROTEIN IS A POTENTIAL MARKER FOR NEUROMYELITIS OPTICA K.-C. WANG, a,b,c C.-P. TSAI, b,c C.-L. LEE, c,d S.-Y. CHEN, e L.-T. CHIN a,f* AND S.-J. CHEN g,h*
included a limited sample size, we attempted to determine an optimized cutoff point for HMGB1 (P2 ng/ml), which provided 89.7% sensitivity and 95.0% specificity for the diagnosis of NMO. These results indicate that plasma HMGB1 level might serve as a surrogate marker for NMO disease activity and aid in the differentiation of NMO from MS at the early disease stage. Ó 2012 IBRO. Published by Elsevier Ltd. All rights reserved.
a
Graduate Institute of Medical Sciences, National Defense Medical Center, Taipei, Taiwan
b
Department of Neurology, Cheng Hsin General Hospital, Taipei, Taiwan
c The Neurological Institute, Taipei Veterans General Hospital, Taipei, Taiwan d
Department of Neurology, I-Lan Hospital, National Yang-Ming University School of Medicine, I-Lan, Taiwan
Key words: cytokine, ELISA, high-mobility group box 1 protein, multiple sclerosis, neuromyelitis optica.
e
Section of Hyperbaric Oxygen Medicine, Cardinal Tien Hospital, Taipei, Taiwan
f
Department of Microbiology, Immunology and Biopharmaceuticals, National Chiayi University, Chiayi, Taiwan
INTRODUCTION
g
Department and Graduate Institute of Microbiology and Immunology, National Defense Medical Center, Taipei, Taiwan
Neuromyelitis optica (NMO), previously described as Devic disease, involves repeated clinical symptoms of optic neuritis and/or myelitis, resulting in loss of vision, decreased strength and coordination, sensory impairment, and paraplegia or even tetraplegia (Hazin et al., 2009). It is a rare disease in Taiwan and other countries, with an incidence of <1 in 100,000 persons (Lai and Tseng, 2009). Many patients with NMO are initially diagnosed with multiple sclerosis (MS), owing to similar initial neurologic presentations (Wingerchuk et al., 1999; de Seze et al., 2003; Cree, 2008). However, the clinical course of NMO is usually more severe than that of classical MS (Cornelio et al., 2009). Within 5 years of onset, 50% of patients with NMO either lose functional vision in at least one eye or are unable to walk unassisted (Wingerchuk et al., 1999; Cabre et al., 2009). Clinically, interferon (IFN) treatment for MS may exacerbate the severity of NMO (Shimizu et al., 2010). The likelihood that NMO may be mistaken for MS underscores the importance of early detection. The diagnosis of NMO has previously depended on the presence of various peripheral autoantibodies such as nuclear autoantibodies (O’Riordan et al., 1996) or non-organ-specific autoantibodies (Hummers et al., 2004); however, these are found to be associated with, but not specific for, the disease. Aquaporin 4 (AQP4) antibody (AQP4 Ab/NMO-IgG), an antibody against a water channel protein, has recently been described as a diagnostic marker for NMO (Lennon et al., 2004). However, its sensitivity varies between studies, being particularly low for Asian patients (e.g., 40% of patients with NMO show positivity for AQP4 Ab in Taiwan compared to 60–70% in Western countries) (Matsuoka et al., 2007; Wang et al., 2011). Therefore, AQP4 is not an optimal marker for NMO. The potential use of proinflammatory cytokines as markers for NMO is
h
Department of Pediatrics, Tri-Service General Hospital, National Defense Medical Center, Graduate Institute of Microbiology and Immunology, Taipei, Taiwan
Abstract—High-mobility group box 1 protein (HMGB1) has cytokine activities and mediates systemic inflammation as well as immune responses. The aim of this study was to determine if plasma HMGB1 level can be used as a marker for neuromyelitis optica (NMO) and to differentiate NMO from multiple sclerosis (MS). We measured plasma levels of HMGB1, tumor necrosis factor-a (TNF-a), interferon-c (IFN-c), and interleukin 17 (IL-17) in 29 patients with NMO and 20 patients with MS at enrollment and at 2 years follow-up (at the time of definitive diagnosis) by enzyme-linked immunosorbent assay. Plasma HMGB1 level was significantly greater in the NMO group compared to the MS group (P < 0.001). Plasma levels of TNF-a, IFN-c, and IL-17 were significantly greater in the NMO group compared to the MS group, and HMGB1 level was positively correlated with TNF-a, IFN-c, and IL-17 levels. Univariate logistic regression analysis showed a significant association of HMGB1 level, and IFN-c level with NMO diagnosis. Although this study
*Correspondence to: L.-T. Chin, Department of Microbiology, Immunology and Biopharmaceuticals, National Chiayi University, No. 300 Syuefu Road, Chiayi City 60004, Taiwan. Tel: +886-5-2717830; fax: +886-5-2717831. S.-J. Chen, Graduate Institute of Microbiology and Immunology, National Defense Medical Center, 114 No. 161, Sec. 6, Minquan E. Road, Neihu District, Taipei City 114, Taiwan. Tel: +886-2-87927025; fax: +886-2-87927923. E-mail addresses: [email protected] (L.-T. Chin), chensjou@ yahoo.com.tw, [email protected] (S.-J. Chen). Abbreviations: 95% CI, 95% confidence intervals; AQP4, aquaporin 4; AUC, area under curve; EDSS, Expanded Disability Status Scale; ELISA, enzyme-linked immunosorbent assay; HMGB1, high-mobility group box 1 protein; IFN-c, interferon-c; IL-17, interleukin 17; MS, multiple sclerosis; NMO, neuromyelitis optica; OR, odds ratios; ROC, receiver operating characteristic; TNF-a, tumor necrosis factor-a.
0306-4522/12 $36.00 Ó 2012 IBRO. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.neuroscience.2012.08.041 510
K.-C. Wang et al. / Neuroscience 226 (2012) 510–516
also not optimal, given their association with many diseases. Thus, there are presently no available biomarkers that can detect NMO with high sensitivity and specificity. High-mobility group box 1 protein (HMGB1) was originally described as an architectural chromatin protein (Bustin, 1999) and was later recognized as a cytokine that is released mainly by monocytes/macrophages to mediate endotoxin lethality (Wang et al., 1999), propagate inflammation, prolong macrophage activation, and act as a mesoangioblast chemoattractant. Because HMGB1 can leak passively from necrotic cells and is actively secreted by stimulated macrophages, it has become evident that extracellular HMGB1 can function as a cytokine, activating inflammatory cells to produce proinflammatory cytokines (i.e., interleukin 17 [IL-17], tumor necrosis factor-a [TNF-a], IL-1b, and IL-8) (Andersson et al., 2000; Shi et al., 2012); upregulating adhesion molecules such as intercellular adhesion molecules and vascular adhesion molecules, leading to the recruitment of macrophages and monocytes and promoting cell migration (Ribeiro et al., 2007); and activating dendritic cells to enhance antigen presentation (Leslie et al., 2004). HMGB1 has been shown to be involved in several autoimmune diseases such as rheumatoid arthritis (Wittemann et al., 1990), type 1 diabetes (Nejentsev et al., 2004), and systemic lupus erythematosus (Uesugi et al., 1998). HMGB1 has also been found in active lesions of MS (Andersson et al., 2008). Because of the severity of NMO, and the importance for it to be treated differently from MS, early detection, differentiation from MS, monitoring of progression or remission at multiple stages, and prevention of disabling attacks are highly desirable. Although HMGB1 has been investigated in various inflammatory conditions (e.g., diabetic retinopathy, pneumonia, ischemia, kidney disease, sepsis, liver injury), and HMGB1 and its corresponding receptors have been shown to be increased in active lesions of MS and in experimental autoimmune encephalomyelitis (Andersson et al., 2008), a role of HMGB1 in the pathogenesis of NMO has not been described. The aim of the present study was to determine if plasma HMGB1 level can be used as a marker for NMO and to differentiate early stage NMO from MS. To this end, we used enzyme-linked immunosorbent assay (ELISA) to perform seroscreening for HMGB1 and related cytokines in patients with NMO and those with MS.
EXPERIMENTAL PROCEDURES
511
criteria were as follows: presence of optic neuritis and acute long spinal cord myelitis (more than three segments). Exclusion criteria were as follows: recurrent myelitis or recurrent optic disease activity. Patients fulfilling the criteria for MS, as established by Poser et al. (1983), were also enrolled. We used the Poser criteria over other criteria (e.g., the McDonald criteria [McDonald et al., 2001; Polman et al., 2011]) because NMO and MS are difficult to distinguish at the early stages, particularly in Asian patients, who generally have less brain involvement than patients in Western countries (Nakamura et al., 2009). Patients with associated rheumatologic disorders, including Sjo¨gren syndrome and systemic lupus erythematosus, were excluded at initial enrollment. The Institutional Review Board of Cheng Hsin General Hospital approved this study (IRB #[240]100-02). All patients signed informed consent forms and were subjected to a thorough neurologic examination, routine laboratory tests, and clinical evaluation at regular intervals in the clinic.
Disease monitoring Patient clinical characteristics, including sex, age, age of symptom onset, and annual relapse rate, were recorded and analyzed together with the Kurtzke Expanded Disability Status Scale (EDSS) score (Kurtzke, 1983). According to the clinical presentation, patients with NMO were further categorized into an active group (n = 16), with more than one relapse per year, and an inactive group (n = 13), with less than one relapse per year, according to retrospective medical record review. All of the patients with NMO were treated with the immunosuppressant azathioprine once diagnosis was made. All of the patients with MS were treated with interferon. High-dose steroids were administered for acute relapses for both NMO and MS.
Detection of AQP4 Ab/NMO-IgG Blood samples were drawn >3 months after any previous episode (i.e., during disease remission), before the 2-year follow-up. Sera were harvested and stored at 20 °C until use. Sera were assayed for anti-AQP4 Ab/NMO-IgG by immunofluorescence using Green Fluorescent Protein (GFP)AQP4 fusion protein-transfected human embryonic kidney (HEK)-293T cells, as described by Matsuoka and associates (Matsuoka et al., 2007).
ELISA for HMGB1 and other cytokines Two-step sandwich ELISA was performed with an HMGB1 ELISA Kit (Shino-Test Corp., Tokyo, Japan), a Human TNF-a ELISA MAX kit, a Human IFN-c ELISA MAX kit, and a Human IL-17 ELISA MAX kit (all from BioLegend, San Diego, CA, USA), according to the manufacturers’ instructions. Given that sample handling and factors such as hemolysis can affect ELISA results, all samples were handled by a single technician experienced with HMGB1 ELISA, and all samples from both patient groups were handled at the same time and in the same manner.
Study design and subjects Statistical analysis This was a prospective study in which subjects were enrolled and assessed at baseline and at a 2-year follow-up. Consecutive patients visiting our clinic during the period July 2008 to July 2009 who had experienced at least two attacks of CNS demyelination were included. Diagnosis of NMO or MS was made after a continuous 2-year clinical follow-up. Patients who fulfilled the clinical diagnostic criteria for NMO, as published by Wingerchuk and associates (2006), were enrolled. Inclusion
Continuous data were not normally distributed and are therefore presented as median and interquartile range, with the exception of age, which was normally distributed and is presented as mean ± standard deviation (SD). Categorical variables are presented as count and percentage. Differences in continuous variables between the NMO and MS groups were tested by independent two-sample t test and nonparametric Mann–
512
K.-C. Wang et al. / Neuroscience 226 (2012) 510–516
Whitney test for normally and nonnormally distributed variables, respectively. Associations between categorical variables were assessed by Fisher exact test. Univariate logistic regression analysis was performed to obtain odds ratios (ORs) and 95% confidence intervals (95% CIs) to show the strength of association between baseline characteristics and NMO. The association of plasma HMGB1 level and NMO diagnosis was assessed by the area under the receiver operating characteristic (ROC) curve (AUC), with 95% CI. Although the sample size was relatively small, we attempted to identify an optimal HMGB1 cutoff point; the highest accuracy for NMO diagnosis was determined by the Youden index (maximum of sensitivity + sensitivity-1). The Pearson correlation coefficient (r) was used to show the correlation between HMGB1 versus cytokine levels. The level of significance was set at an alpha level of 0.05. Statistical analyses were performed with SPSS software version 15.0 (SPSS Inc., Chicago, IL, USA).
RESULTS
compared to the MS group at follow-up (2.5 vs. 1.0; P < 0.001), similar to at baseline. HMGB1 level in NMO and MS The plasma level of HMGB1 in the NMO group was significantly greater than that in the MS group (median, 3.99 ng/ml vs. 1.33 ng/ml; P < 0.001) (Table 2). In addition, the active NMO group showed a significantly greater plasma HMGB1 level compared to that in the inactive NMO group (median, 4.72 ng/ml vs. 2.94 ng/ml, P = 0.015) (Fig. 1). We also conducted an analysis of HMGB1 level in AQP4-positive versus AQP4-negative patients with NMO (Table 4). Whereas the plasma HMGB1 level was greater in AQP4-positive patients compared to AQP4negative patients (median: 4.33 vs. 3.72), the difference did not reach statistical significance.
Patient characteristics During the period July 2008 to July 2009, a total of 60 consecutive patients who had experienced at least two attacks of CNS demyelination were selected. Diagnosis of NMO or MS was made after a continuous 2-year clinical follow-up. A total of 29 patients with NMO and 20 patients with MS were included in the analysis (11 patients were excluded, 5 with recurrent optic neuritis, and 6 with recurrent myelitis; none had lesions as determined by brain imaging). Demographic and clinical characteristics at baseline and at 2 years are listed in Table 1. Age, sex, disease duration, and EDSS score were comparable between the two disease groups. Significantly greater value for AQP4 Ab positivity (34.5% vs. 0%; P = 0.003) were observed in the NMO group compared to the MS group. At the 2-year follow-up, the median annual relapse rate had increased significantly from 1.0 to 2.5 in the NMO group and from 0.6 to 1.0 in the MS group (Table 1). This rate was significantly greater in the NMO group
Correlation of inflammatory cytokines with plasma HMGB1 level Plasma levels of TNF-a and IFN-c were significantly greater in the NMO group compared to the MS group (TNF-a: median, 4.91 pg/ml vs. 0 pg/ml, P = 0.021; IFN-c: 10.61 pg/ml vs. 0 pg/ml, P = 0.001) (Table 2), and the plasma HMGB1 level was positively correlated with TNF-a and IFN-c levels (r = 0.517, P = 0.040 and r = 0.659, P = 0.005, respectively) (Table 3). In addition, the plasma IL-17 level was significantly greater in the NMO group compared to the MS group (median, 25.4 pg/ml vs. 0 pg/ml; P = 0.007) (Table 2), and the HMGB1 level was positively correlated with the IL-17 level (r = 0.723, P = 0.028) (Table 3). Diagnosis of NMO Results of univariate logistic regression analysis showed significant association of the following baseline
Table 1. Demographic and clinical characteristics for the NMO and MS groups in this study
Baseline characteristic Age (years)a Sexc Male Female Disease duration (years)b EDSSb AQP4 Abc Positive Negative At 2 years Total number of relapses during disease durationb Annual relapse rateb EDSSb
NMO (n = 29)
MS (n = 20)
P value
41.5 ± 9.7
41.7 ± 12.6
0.946
1 (3.4) 28 (96.6) 5.0 (2.6, 9.0) 3.0 (2.5, 4.5)
4 (20.0) 16 (80.0) 4.0 (3.0, 6.0) 2.8 (1.0, 4.0)
0.144
10 (34.5) 19 (65.5)
0 (0) 20 (100.0)
0.003*
5.0 (3.0, 8.0) 2.5 (1.5, 4.0) 3.0 (2.5, 5.5)
2.0 (2.0, 3.0) 1.0 (1.0, 1.5) 3.0 (0.8, 3.8)
AQP4 Ab, serum AQP4 antibody; EDSS, Expanded Disability Status Scale score; MS, multiple sclerosis; NMO, neuromyelitis optica. a Data are presented as mean ± standard deviation. b Data are presented as median (interquartile range). c Data are presented as count and percentage. * Statistically significant difference between groups. Statistically significant difference between baseline and follow-up.
0.875 0.139
<0.001* <0.001* 0.098
513
K.-C. Wang et al. / Neuroscience 226 (2012) 510–516 Table 2. Comparison of baseline serum HMGB1 and cytokine levels in the NMO and MS groups
HMGB1 (ng/ml) TNF-a (pg/ml) IFN-c (pg/ml) IL-17 (pg/ml)
NMO (n = 29)
MS (n = 20)
P value
3.99 (2.80, 4.97) 4.91 (0, 7.29) 10.61 (3.31, 12.43) 25.38 (15.33, 74.39)
1.33 (1.12, 1.52) 0 (0, 0) 0 (0, 2.57) 0 (0, 0)
<0.001* 0.021* 0.001* 0.007*
Data are presented as median (interquartile range). HMGB1, high-mobility group box 1 protein; IFN-c, interferon-c; IL-17, interleukin 17; MS, multiple sclerosis; NMO, neuromyelitis optica; TNF-a, tumor necrosis factor-a. * Statistically significant difference between groups. Most values for TNF-a and IL-17 were 0 in the MS group, with only three patients positive for TNF-a (2.93, 3.81, 35.74 pg/ml) or IL-17 (3.73, 29.55, 58.09 pg/ml).
Table 4. Plasma HMGB1 level according to AQP4 antibody status in patients with NMO AQP4 Ab-positive AQP4 Ab-negative P (n = 10) (n = 19) value HMGB1 (ng/ml) 4.33 (2.71, 5.48)
3.72 (2.80, 4.80)
0.484
Data are presented as median (interquartile range). AQP4 Ab, serum AQP4 antibody; HMGB1, high-mobility group box 1 protein; NMO, neuromyelitis optica.
Table 5. Univariate logistic regression results identifying baseline factors associated with NMO
Fig. 1. Plasma HMGB1 levels in MS, inactive NMO, and active NMO groups. ⁄Significant difference compared to MS group; significant difference compared to inactive group. MS, multiple sclerosis; HMGB1, high-mobility group box 1 protein; MS, multiple sclerosis; NMO, neuromyelitis optica.
Table 3. Pearson correlation coefficients (r) for baseline plasma HMGB1 level versus cytokine levels TNF-a
IFN-c
IL-17
MS
r P value
0.284 0.325
0.167 0.568
0.216 0.457
NMO
r P value
0.517 0.040*
0.659 0.005*
0.723 0.028*
HMGB1, high-mobility group box 1 protein; IFN-c, interferon-c; IL-17, interleukin 17; MS, multiple sclerosis; NMO, neuromyelitis optica; TNF-a, tumor necrosis factor-a. * Statistically significant.
Variable
OR (95% CI)
P value
Age (years) Sex (female vs. male) Disease duration (years) EDSS HMGB1 (ng/ml) TNF-a (pg/ml) IFN-c (pg/ml) IL-17 (pg/ml)
0.998 (0.95, 1.05) 7.00 (0.72, 68.15) 1.03 (0.91, 1.15) 1.35 (0.97, 1.88) 18.37 (3.06, 110.17) 1.06 (0.95, 1.17) 1.31 (1.07, 1.59) 1.04 (1.00, 1.09)
0.944 0.094 0.666 0.076 0.001* 0.314 0.008* 0.064
The OR for AQP4 Ab (positive vs. negative) was not applicable, owing to a zero count. AQP4 Ab, serum AQP4 antibody; EDSS, Expanded Disability Status Scale score; HMGB1, high-mobility group box 1 protein; IFN-c, interferon-c; IL-17, interleukin 17; NMO, neuromyelitis optica; OR, odds ratio; TNF-a, tumor necrosis factor-a. * Statistically significant.
sensitivity of 89.7% and a specificity of 95.0% (Fig. 2). A cutoff point for HMGB1 of P1.3 ng/ml resulted in a sensitivity of 100% but a lower specificity of 45.0%.
DISCUSSION characteristics with NMO diagnosis (among all study patients, including both the NMO and MS groups): plasma HMGB1 level, and IFN-c level (Table 5). Those with greater plasma HMGB1 level were more likely to have NMO, and the likelihood increased with every unit increase in HMGB1 level (OR = 18.37, P = 0.001). The sample size of 49 was too small to conduct multivariate logistic regression analysis to identify independent factors. An OR for AQP4 Ab (positive vs. negative) was not applicable, owing to a lack of AQP4 Ab-positive patients in the MS group. For diagnosis of NMO according to plasma HMGB1 level, the AUC for the ROC was 0.955, which showed good accuracy. An optimized cutoff point for HMGB1 of P2 ng/ml was determined by the Youden index, which provided a
We performed ELISA for plasma HMGB1 and related cytokines in patients with demyelinating disorders at the initial stages and correlated these levels with the later diagnosis of NMO or MS. Our results showed that plasma HMGB1 level could differentiate NMO from MS in the early stages. A substantial amount of plasma HMGB1 was detected in patients with NMO, and the level correlated well with the levels of disability and recurrence, as assessed by EDSS score and annual relapse rate. In addition, patients with NMO showed an elevated plasma HMGB1 level compared to patients with MS. Furthermore, plasma HMGB1 level was significantly greater in AQP4-positive patients compared to AQP4-negative patients, consistent with NMO status. The rate of AQP4 Ab positivity among patient with NMO
514
K.-C. Wang et al. / Neuroscience 226 (2012) 510–516
Fig. 2. ROC curve and sensitivity of plasma HMGB1 level in the diagnosis of NMO. AUC = 0.955; 95% CI = 0.903, 1.0. AUC, area under the curve; HMGB1, high-mobility group box 1 protein; MS, multiple sclerosis; NMO, neuromyelitis optica; ROC, receiver operating characteristic.
(10/29 patients. 34.5%) was consistent with the relative low seroprevalence of AQP4 Ab in Asian patients with NMO (Matsuoka et al., 2007; Wang et al., 2011). Whereas increased plasma levels of TNF-a, IFN-c, and IL-17 might suggest specificity for NMO, these cytokines are affected by many inflammatory diseases and are associated with the severity of inflammation. Plasma HMGB1 level was a more specific marker for NMO in the present study, given that it was significantly increased in patients who subsequently received a diagnosis of NMO versus MS. This finding provides clinicians with a tool, in addition to severity of clinical attacks and AQP4 Ab positivity, for the diagnosis of NMO at the early stage. When encountering a patient presenting with optical neuritis, myelitis, or both, assessment of plasma HMGB1 level might facilitate an earlier diagnosis of NMO versus MS. The resulting appropriate and different treatment strategies might also minimize damage and prevent subsequent attacks. The pleiotropic cytokine IFN-c, secreted by activated T lymphocytes and natural killer cells, is believed to play a role in immune-mediated demyelinating disorders such as MS (Panitch et al., 1987; Petereit et al., 2002). In addition, IFN-c inhibits central nervous system remyelination (Lin et al., 2006). Our present results indicate a potential association between IFN-c and HMGB1 in patients with NMO. IFN-c has been reported to induce HMGB1 secretion (Strachan et al., 2008), and increased HMGB1 plasma level correlated with active inflammation in our patients. HMGB1 has been reported to activate human neutrophils to produce proinflammatory mediators such as TNF-a, IL-1b, and IL-8 (Li et al., 2004). HMGB1 itself also acts as an immunostimulatory signal to induce maturation of
dendritic cells and the secretion of proinflammatory cytokines including TNF-a, IL-1a, IL-17, IL-8, and IL-12 (Gupta et al., 2004; Marioni et al., 2010). Thus, HMGB1 could play a role in immune-related central nervous system demyelinating diseases. It is well established that HMGB1 is secreted by activated macrophages (Wang et al., 1999; Bonaldi et al., 2003), mature dendritic cells, and natural killer cells (Semino et al., 2005) in response to injury, infection, or other inflammatory stimuli. In the case of NMO, in which a severe inflammatory response is often associated clinically with transverse myelitis and optic neuritis, a role of HMGB1 is possible, given that factors inducing its release, such as extensive necrosis, demyelination, and perivascular macrophage infiltration are common clinical features. There is evidence that autocrine HMGB1 signaling is necessary for the upregulation of B7 costimulation and for IL-12 production of human dendritic cells, which are subsequently required for efficient antigen presentation (Dumitriu et al., 2005). It has also been reported that HMGB1 exerts an inhibitory effect on regulatory T cells in autoimmune disease (Zhang et al., 2008) may induce clonal expansion, survival, and functional polarization of naive T cells (Han et al., 2008). In vivo, HMGB1 has been shown to enhance primary responses to soluble antigens (Kimura et al., 2010). This provides a potential rationale for why patients with NMO are prone to the expression of autoantibodies such as AQP4 Ab/NMOIgG (Lennon et al., 2004), serum nuclear and anti-SS-A/ SS-B autoantibodies (O’Riordan et al., 1996), and non-organ-specific autoantibodies (Hummers et al., 2004). Potential limitations of the present study include the relatively small sample size and the fact that it was conducted at a single institution. The cutoff point determined in the present study was calculated because clinicians generally need a reference point to differentiate between an initial diagnosis of NMO or MS; however, the limited sample size suggests caution in relying on this cutoff point for all cases. It should be mentioned that both the NMO and MS groups showed an increased relapse rate at follow-up (with IFN treatment). In Taiwan, patients receiving IFN often visit the clinic less and refuse continued IFN treatment because they are experiencing fewer attacks. However, some patients experience more attacks and therefore visit the clinic more often. These patients, showing a relatively greater relapse rate and more severe disease, were more likely to be recruited for the present study.
CONCLUSION In conclusion, differentiating NMO from MS, especially at the initial stage, is a real challenge in the clinic. In the past, NMO (not MS) was inferred when abnormalities were detected only in the optic nerve and spinal cord. While not definitive, our present results suggest that HMGB1 might be a diagnostic marker for NMO at the early stage. Additional research is needed to determine whether HMGB1 might also be a mediator of or therapeutic target for NMO. We hope to elucidate
K.-C. Wang et al. / Neuroscience 226 (2012) 510–516
differences in HMGB1 between patients with NMO and those with systemic lupus erythematosus, Sjo¨gren syndrome, or other inflammatory neurologic diseases in future studies.
CONTRIBUTORS Kai-Chen Wang: study design; experimental studies; data analysis; manuscript preparation; manuscript editing. Ching-Piao Tsai: clinical studies; manuscript review. Chao-Lin Lee: clinical studies; experimental studies; data acquisition. Shao-Yuan Chen: literature research; statistical analysis. Li-Te Chin: study concepts; definition of intellectual content. Shyi-Jou Chen: guarantor of integrity of the entire study; manuscript review. All authors have materially participated in the research and manuscript preparation and have approved the final version of the manuscript. Acknowledgments—The authors thank Miss Ruei-Sia Wang for technical assistance. Editorial assistance was provided by Lauren P. Baker, PhD, ELS, in association with MedCom Asia Inc. This work was supported partly by grant from NSC99-2314-B016-002-MY3 to SJ Chen.
REFERENCES Andersson A, Covacu R, Sunnemark D, Danilov AI, Dal Bianco A, Khademi M, Wallstro¨m E, Lobell A, Brundin L, Lassmann H, Harris RA (2008) Pivotal advance: HMGB1 expression in active lesions of human and experimental multiple sclerosis. J Leukoc Biol 84:1248–1255. Andersson U, Wang H, Palmblad K, Aveberger AC, Bloom O, Erlandsson-Harris J, Janson A, Kokkola R, Zhang M, Yang H, Tracey KJ (2000) High mobility group 1 protein (HMG-1) stimulates proinflammatory cytokine synthesis in human monocytes. J Exp Med 192:565–570. Bonaldi T, Talamo F, Scaffidi P, Ferrera D, Porto A, Bachi A, Rubartelli A, Agresti A, Bianchi ME (2003) Monocytic cells hyperacetylate chromatin protein HMGB1 to redirect it towards secretion. EMBO J 22:5551–5560. Bustin M (1999) Regulation of DNA-dependent activities by the functional motifs of the high-mobility-group chromosomal proteins. Mol Cell Biol 19:5237–5246. Cabre P, Gonza´lez-Quevedo A, Bonnan M, Saiz A, Olindo S, Graus F, Smadja D, Merle H, Thomas L, Cabrera-Gomez JA (2009) Relapsing neuromyelitis optica: long term history and clinical predictors of death. J Neurol Neurosurg Psychiatry 80:1162–1164. Cornelio DB, Braga RP, Rosa MW, Ayub AC (2009) Devic’s neuromyelitis optica and pregnancy: distinction from multiple sclerosis is essential. Arch Gynecol Obstet 280:475–477. Cree B (2008) Neuromyelitis optica: diagnosis, pathogenesis, and treatment. Curr Neurol Neurosci Rep 8:427–433. de Seze J, Lebrun C, Stojkovic T, Ferriby D, Chatel M, Vermersch P (2003) Is Devic’s neuromyelitis optica a separate disease? A comparative study with multiple sclerosis. Mult Scler 9:521–525. Dumitriu IE, Baruah P, Valentinis B, Voll RE, Herrmann M, Nawroth PP, Arnold B, Bianchi ME, Manfredi AA, Rovere-Querini P (2005) Release of high mobility group box 1 by dendritic cells controls T cell activation via the receptor for advanced glycation end products. J Immunol 174:7506–7515.
515
Gupta R, Sheikh A, Strachan DP, Anderson HR (2004) Burden of allergic disease in the UK: secondary analyses of national databases. Clin Exp Allergy 34:520–526. Han J, Zhong J, Wei W, Wang Y, Huang Y, Yang P, Purohit S, Dong Z, Wang MH, She JX, Gong F, Stern DM, Wang CY (2008) Extracellular high-mobility group box 1 acts as an innate immune mediator to enhance autoimmune progression and diabetes onset in NOD mice. Diabetes 57:2118–2127. Hazin R, Khan F, Bhatti MT (2009) Neuromyelitis optica: current concepts and prospects for future management. Curr Opin Ophthalmol 20:434–439. Hummers LK, Krishnan C, Casciola-Rosen L, Rosen A, Morris S, Mahoney JA, Kerr DA, Wigley FM (2004) Recurrent transverse myelitis associates with anti-Ro (SSA) autoantibodies. Neurology 62:147–149. Kimura R, Shiibashi R, Suzuki M, Hayashizaki Y, Chiba J (2010) Enhancement of antibody response by high mobility group box protein-1-based DNA immunization. J Immunol Methods 361:21–30. Kurtzke JF (1983) Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology 33:1444–1452. Lai CH, Tseng HF (2009) Population-based epidemiological study of neurological diseases in Taiwan: I. Creutzfeldt–Jakob disease and multiple sclerosis. Neuroepidemiology 33:247–253. Lennon VA, Wingerchuk DM, Kryzer TJ, Pittock SJ, Lucchinetti CF, Fujihara K, Nakashima I, Weinshenker BG (2004) A serum autoantibody marker of neuromyelitis optica: distinction from multiple sclerosis. Lancet 364:2106–2112. Leslie SJ, Johnston N, Strachan FE, Bagnali A, Gray GA, Newby DE, Denvir MA, Webb DJ (2004) Endothelin-1[1–31] is not elevated in men with chronic heart failure. J Cardiovasc Pharmacol 44(Suppl. 1):S96–S99. Li B, Kawatkar SP, George S, Strachan H, Woods RJ, Siriwardena A, Moremen KW, Boons GJ (2004) Inhibition of Golgi mannosidase II with mannostatin A analogues: synthesis, biological evaluation, and structure–activity relationship studies. Chembiochem 5:1220–1227. Lin W, Kemper A, Dupree JL, Harding HP, Ron D, Popko B (2006) Interferon-gamma inhibits central nervous system remyelination through a process modulated by endoplasmic reticulum stress. Brain 129(Pt. 5):1306–1318. Marioni RE, Deary IJ, Strachan MW, Lowe GD, Rumley A, Murray GD, Price JF (2010) Blood rheology and cognition in the Edinburgh Type 2 Diabetes Study. Age Ageing 39:354–359. Matsuoka T, Matsushita T, Kawano Y, Osoegawa M, Ochi H, Ishizu T, Minohara M, Kikuchi H, Mihara F, Ohyagi Y, Kira J (2007) Heterogeneity of aquaporin-4 autoimmunity and spinal cord lesions in multiple sclerosis in Japanese. Brain 130:1206–1223. McDonald WI, Compston A, Edan G, Goodkin D, Hartung HP, Lublin FD, McFarland HF, Paty DW, Polman CH, Reingold SC, Sandberg-Wollheim M, Sibley W, Thompson A, van den Noort S, Weinshenker BY, Wolinsky JS (2001) Recommended diagnostic criteria for multiple sclerosis: guidelines from the international panel on the diagnosis of multiple sclerosis. Ann Neurol 50:121–127. Nakamura M, Houzen H, Niino M, Tanaka K, Sasaki H (2009) Relationship between Barkhof criteria and the clinical features of multiple sclerosis in northern Japan. Mult Scler 15: 1450–1458. Nejentsev S, Cooper JD, Godfrey L, Howson JM, Rance H, Nutland S, Walker NM, Guja C, Ionescu-Tirgovisßte C, Savage DA, Undlien DE, Rønningen KS, Tuomilehto-Wolf E, Tuomilehto J, Gillespie KM, Ring SM, Strachan DP, Widmer B, Dunger D, Todd JA (2004) Analysis of the vitamin D receptor gene sequence variants in type 1 diabetes. Diabetes 53:2709–2712. O’Riordan JI, Gallagher HL, Thompson AJ, Howard RS, Kingsley DP, Thompson EJ, McDonald WI, Miller DH (1996) Clinical, CSF, and MRI findings in Devic’s neuromyelitis optica. J Neurol Neurosurg Psychiatry 60:382–387.
516
K.-C. Wang et al. / Neuroscience 226 (2012) 510–516
Panitch HS, Hirsch RL, Schindler J, Johnson KP (1987) Treatment of multiple sclerosis with gamma interferon: exacerbations associated with activation of the immune system. Neurology 37:1097–1102. Petereit HF, Nolden S, Schoppe S, Bamborschke S, Pukrop R, Heiss WD (2002) Low interferon gamma producers are better treatment responders: a two-year follow-up of interferon beta-treated multiple sclerosis patients. Mult Scler 8:492–494. Polman CH, Reingold SC, Banwell B, Clanet M, Cohen JA, Filippi M, Fujihara K, Havrdova E, Hutchinson M, Kappos L, Lublin FD, Montalban X, O’Connor P, Sandberg-Wollheim M, Thompson AJ, Waubant E, Weinshenker B, Wolinsky JS (2011) Diagnostic criteria for multiple sclerosis: 2010 revisions to the McDonald criteria. Ann Neurol 69:292–302. Poser CM, Paty DW, Scheinberg L, McDonald WI, Davis FA, Ebers GC, Johnson KP, Sibley WA, Silberberg DH, Tourtellotte WW (1983) New diagnostic criteria for multiple sclerosis: guidelines for research protocols. Ann Neurol 13:227–231. Ribeiro MA, Cabral HO, Costa PF (2007) Modulatory effect of NO on sodium currents in a neuroblastoma cell line: aspects of cell specificity. Neurosci Res 58:361–370. Semino C, Angelini G, Poggi A, Rubartelli A (2005) NK/iDC interaction results in IL-18 secretion by DCs at the synaptic cleft followed by NK cell activation and release of the DC maturation factor HMGB1. Blood 106:609–616. Shi Y, Sandoghchian Shotorbani S, Su Z, Liu Y, Tong J, Zheng D, Chen J, Liu Y, Xu Y, Jiao Z, Wang S, Lu L, Huang X, Xu H (2012) Enhanced HMGB1 expression may contribute to Th17 cells activation in rheumatoid arthritis. Clin Dev Immunol 2012:295081. Shimizu J, Hatanaka Y, Hasegawa M, Iwata A, Sugimoto I, Date H, Goto J, Shimizu T, Takatsu M, Sakurai Y, Nakase H, Uesaka Y, Hashida H, Hashimoto K, Komiya T, Tsuji S (2010) IFNb-1b may
severely exacerbate Japanese optic-spinal MS in neuromyelitis optica spectrum. Neurology 75:1423–1427. Strachan MW, Reynolds RM, Frier BM, Mitchell RJ, Price JF (2008) The relationship between type 2 diabetes and dementia. Br Med Bull 88:131–146. Uesugi H, Ozaki S, Sobajima J, Osakada F, Shirakawa H, Yoshida M, Nakao K (1998) Prevalence and characterization of novel pANCA, antibodies to the high mobility group non-histone chromosomal proteins HMG1 and HMG2, in systemic rheumatic diseases. J Rheumatol 25:703–709. Wang H, Bloom O, Zhang M, Vishnubhakat JM, Ombrellino M, Che J, Frazier A, Yang H, Ivanova S, Borovitkova L, Manogue KR, Faist E, Abraham E, Andersson J, Andersson U, Molina PE, Abumrad NN, Sama A, Tracey KJ (1999) HMG-1 as a late mediator of endotoxin lethality in mice. Science 285:248–251. Wang KC, Tsai CP, Lee CL, Chen SY, Chen SJ (2011) The prevalence of long spinal cord lesions and anti-aquaporin 4 antibodies in neuromyelitis optica patients in Taiwan. Eur Neurol 65:99–104. Wingerchuk DM, Hogancamp WF, O’Brien PC, Weinshenker BG (1999) The clinical course of neuromyelitis optica (Devic’s syndrome). Neurology 53:1107–1114. Wingerchuk DM, Lennon VA, Pittock SJ, Lucchinetti CF, Weinshenker BG (2006) Revised diagnostic criteria for neuromyelitis optica. Neurology 66:1485–1489. Wittemann B, Neuer G, Michels H, Truckenbrodt H, Bautz FA (1990) Autoantibodies to nonhistone chromosomal proteins HMG-1 and HMG-2 in sera of patients with juvenile rheumatoid arthritis. Arthritis Rheum 33:1378–1383. Zhang Y, Yao YM, Yu Y, Wu Y, Sheng ZY (2008) Influence of high mobility group box-1 protein on the correlation between regulatory T cells and CD4+ CD25 T cells of spleen in mice. Chin J Surg 46:217–220 (in Chinese).
(Accepted 17 August 2012)
ELEVATED PLASMA HIGH-MOBILITY GROUP BOX 1 PROTEIN IS A POTENTIAL MARKER FOR NEUROMYELITIS OPTICA K.-C. WANG, a,b,c C.-P. TSAI, b,c C.-L. LEE, c,d S.-Y. CHEN, e L.-T. CHIN a,f* AND S.-J. CHEN g,h*
included a limited sample size, we attempted to determine an optimized cutoff point for HMGB1 (P2 ng/ml), which provided 89.7% sensitivity and 95.0% specificity for the diagnosis of NMO. These results indicate that plasma HMGB1 level might serve as a surrogate marker for NMO disease activity and aid in the differentiation of NMO from MS at the early disease stage. Ó 2012 IBRO. Published by Elsevier Ltd. All rights reserved.
a
Graduate Institute of Medical Sciences, National Defense Medical Center, Taipei, Taiwan
b
Department of Neurology, Cheng Hsin General Hospital, Taipei, Taiwan
c The Neurological Institute, Taipei Veterans General Hospital, Taipei, Taiwan d
Department of Neurology, I-Lan Hospital, National Yang-Ming University School of Medicine, I-Lan, Taiwan
Key words: cytokine, ELISA, high-mobility group box 1 protein, multiple sclerosis, neuromyelitis optica.
e
Section of Hyperbaric Oxygen Medicine, Cardinal Tien Hospital, Taipei, Taiwan
f
Department of Microbiology, Immunology and Biopharmaceuticals, National Chiayi University, Chiayi, Taiwan
INTRODUCTION
g
Department and Graduate Institute of Microbiology and Immunology, National Defense Medical Center, Taipei, Taiwan
Neuromyelitis optica (NMO), previously described as Devic disease, involves repeated clinical symptoms of optic neuritis and/or myelitis, resulting in loss of vision, decreased strength and coordination, sensory impairment, and paraplegia or even tetraplegia (Hazin et al., 2009). It is a rare disease in Taiwan and other countries, with an incidence of <1 in 100,000 persons (Lai and Tseng, 2009). Many patients with NMO are initially diagnosed with multiple sclerosis (MS), owing to similar initial neurologic presentations (Wingerchuk et al., 1999; de Seze et al., 2003; Cree, 2008). However, the clinical course of NMO is usually more severe than that of classical MS (Cornelio et al., 2009). Within 5 years of onset, 50% of patients with NMO either lose functional vision in at least one eye or are unable to walk unassisted (Wingerchuk et al., 1999; Cabre et al., 2009). Clinically, interferon (IFN) treatment for MS may exacerbate the severity of NMO (Shimizu et al., 2010). The likelihood that NMO may be mistaken for MS underscores the importance of early detection. The diagnosis of NMO has previously depended on the presence of various peripheral autoantibodies such as nuclear autoantibodies (O’Riordan et al., 1996) or non-organ-specific autoantibodies (Hummers et al., 2004); however, these are found to be associated with, but not specific for, the disease. Aquaporin 4 (AQP4) antibody (AQP4 Ab/NMO-IgG), an antibody against a water channel protein, has recently been described as a diagnostic marker for NMO (Lennon et al., 2004). However, its sensitivity varies between studies, being particularly low for Asian patients (e.g., 40% of patients with NMO show positivity for AQP4 Ab in Taiwan compared to 60–70% in Western countries) (Matsuoka et al., 2007; Wang et al., 2011). Therefore, AQP4 is not an optimal marker for NMO. The potential use of proinflammatory cytokines as markers for NMO is
h
Department of Pediatrics, Tri-Service General Hospital, National Defense Medical Center, Graduate Institute of Microbiology and Immunology, Taipei, Taiwan
Abstract—High-mobility group box 1 protein (HMGB1) has cytokine activities and mediates systemic inflammation as well as immune responses. The aim of this study was to determine if plasma HMGB1 level can be used as a marker for neuromyelitis optica (NMO) and to differentiate NMO from multiple sclerosis (MS). We measured plasma levels of HMGB1, tumor necrosis factor-a (TNF-a), interferon-c (IFN-c), and interleukin 17 (IL-17) in 29 patients with NMO and 20 patients with MS at enrollment and at 2 years follow-up (at the time of definitive diagnosis) by enzyme-linked immunosorbent assay. Plasma HMGB1 level was significantly greater in the NMO group compared to the MS group (P < 0.001). Plasma levels of TNF-a, IFN-c, and IL-17 were significantly greater in the NMO group compared to the MS group, and HMGB1 level was positively correlated with TNF-a, IFN-c, and IL-17 levels. Univariate logistic regression analysis showed a significant association of HMGB1 level, and IFN-c level with NMO diagnosis. Although this study
*Correspondence to: L.-T. Chin, Department of Microbiology, Immunology and Biopharmaceuticals, National Chiayi University, No. 300 Syuefu Road, Chiayi City 60004, Taiwan. Tel: +886-5-2717830; fax: +886-5-2717831. S.-J. Chen, Graduate Institute of Microbiology and Immunology, National Defense Medical Center, 114 No. 161, Sec. 6, Minquan E. Road, Neihu District, Taipei City 114, Taiwan. Tel: +886-2-87927025; fax: +886-2-87927923. E-mail addresses: [email protected] (L.-T. Chin), chensjou@ yahoo.com.tw, [email protected] (S.-J. Chen). Abbreviations: 95% CI, 95% confidence intervals; AQP4, aquaporin 4; AUC, area under curve; EDSS, Expanded Disability Status Scale; ELISA, enzyme-linked immunosorbent assay; HMGB1, high-mobility group box 1 protein; IFN-c, interferon-c; IL-17, interleukin 17; MS, multiple sclerosis; NMO, neuromyelitis optica; OR, odds ratios; ROC, receiver operating characteristic; TNF-a, tumor necrosis factor-a.
0306-4522/12 $36.00 Ó 2012 IBRO. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.neuroscience.2012.08.041 510
K.-C. Wang et al. / Neuroscience 226 (2012) 510–516
also not optimal, given their association with many diseases. Thus, there are presently no available biomarkers that can detect NMO with high sensitivity and specificity. High-mobility group box 1 protein (HMGB1) was originally described as an architectural chromatin protein (Bustin, 1999) and was later recognized as a cytokine that is released mainly by monocytes/macrophages to mediate endotoxin lethality (Wang et al., 1999), propagate inflammation, prolong macrophage activation, and act as a mesoangioblast chemoattractant. Because HMGB1 can leak passively from necrotic cells and is actively secreted by stimulated macrophages, it has become evident that extracellular HMGB1 can function as a cytokine, activating inflammatory cells to produce proinflammatory cytokines (i.e., interleukin 17 [IL-17], tumor necrosis factor-a [TNF-a], IL-1b, and IL-8) (Andersson et al., 2000; Shi et al., 2012); upregulating adhesion molecules such as intercellular adhesion molecules and vascular adhesion molecules, leading to the recruitment of macrophages and monocytes and promoting cell migration (Ribeiro et al., 2007); and activating dendritic cells to enhance antigen presentation (Leslie et al., 2004). HMGB1 has been shown to be involved in several autoimmune diseases such as rheumatoid arthritis (Wittemann et al., 1990), type 1 diabetes (Nejentsev et al., 2004), and systemic lupus erythematosus (Uesugi et al., 1998). HMGB1 has also been found in active lesions of MS (Andersson et al., 2008). Because of the severity of NMO, and the importance for it to be treated differently from MS, early detection, differentiation from MS, monitoring of progression or remission at multiple stages, and prevention of disabling attacks are highly desirable. Although HMGB1 has been investigated in various inflammatory conditions (e.g., diabetic retinopathy, pneumonia, ischemia, kidney disease, sepsis, liver injury), and HMGB1 and its corresponding receptors have been shown to be increased in active lesions of MS and in experimental autoimmune encephalomyelitis (Andersson et al., 2008), a role of HMGB1 in the pathogenesis of NMO has not been described. The aim of the present study was to determine if plasma HMGB1 level can be used as a marker for NMO and to differentiate early stage NMO from MS. To this end, we used enzyme-linked immunosorbent assay (ELISA) to perform seroscreening for HMGB1 and related cytokines in patients with NMO and those with MS.
EXPERIMENTAL PROCEDURES
511
criteria were as follows: presence of optic neuritis and acute long spinal cord myelitis (more than three segments). Exclusion criteria were as follows: recurrent myelitis or recurrent optic disease activity. Patients fulfilling the criteria for MS, as established by Poser et al. (1983), were also enrolled. We used the Poser criteria over other criteria (e.g., the McDonald criteria [McDonald et al., 2001; Polman et al., 2011]) because NMO and MS are difficult to distinguish at the early stages, particularly in Asian patients, who generally have less brain involvement than patients in Western countries (Nakamura et al., 2009). Patients with associated rheumatologic disorders, including Sjo¨gren syndrome and systemic lupus erythematosus, were excluded at initial enrollment. The Institutional Review Board of Cheng Hsin General Hospital approved this study (IRB #[240]100-02). All patients signed informed consent forms and were subjected to a thorough neurologic examination, routine laboratory tests, and clinical evaluation at regular intervals in the clinic.
Disease monitoring Patient clinical characteristics, including sex, age, age of symptom onset, and annual relapse rate, were recorded and analyzed together with the Kurtzke Expanded Disability Status Scale (EDSS) score (Kurtzke, 1983). According to the clinical presentation, patients with NMO were further categorized into an active group (n = 16), with more than one relapse per year, and an inactive group (n = 13), with less than one relapse per year, according to retrospective medical record review. All of the patients with NMO were treated with the immunosuppressant azathioprine once diagnosis was made. All of the patients with MS were treated with interferon. High-dose steroids were administered for acute relapses for both NMO and MS.
Detection of AQP4 Ab/NMO-IgG Blood samples were drawn >3 months after any previous episode (i.e., during disease remission), before the 2-year follow-up. Sera were harvested and stored at 20 °C until use. Sera were assayed for anti-AQP4 Ab/NMO-IgG by immunofluorescence using Green Fluorescent Protein (GFP)AQP4 fusion protein-transfected human embryonic kidney (HEK)-293T cells, as described by Matsuoka and associates (Matsuoka et al., 2007).
ELISA for HMGB1 and other cytokines Two-step sandwich ELISA was performed with an HMGB1 ELISA Kit (Shino-Test Corp., Tokyo, Japan), a Human TNF-a ELISA MAX kit, a Human IFN-c ELISA MAX kit, and a Human IL-17 ELISA MAX kit (all from BioLegend, San Diego, CA, USA), according to the manufacturers’ instructions. Given that sample handling and factors such as hemolysis can affect ELISA results, all samples were handled by a single technician experienced with HMGB1 ELISA, and all samples from both patient groups were handled at the same time and in the same manner.
Study design and subjects Statistical analysis This was a prospective study in which subjects were enrolled and assessed at baseline and at a 2-year follow-up. Consecutive patients visiting our clinic during the period July 2008 to July 2009 who had experienced at least two attacks of CNS demyelination were included. Diagnosis of NMO or MS was made after a continuous 2-year clinical follow-up. Patients who fulfilled the clinical diagnostic criteria for NMO, as published by Wingerchuk and associates (2006), were enrolled. Inclusion
Continuous data were not normally distributed and are therefore presented as median and interquartile range, with the exception of age, which was normally distributed and is presented as mean ± standard deviation (SD). Categorical variables are presented as count and percentage. Differences in continuous variables between the NMO and MS groups were tested by independent two-sample t test and nonparametric Mann–
512
K.-C. Wang et al. / Neuroscience 226 (2012) 510–516
Whitney test for normally and nonnormally distributed variables, respectively. Associations between categorical variables were assessed by Fisher exact test. Univariate logistic regression analysis was performed to obtain odds ratios (ORs) and 95% confidence intervals (95% CIs) to show the strength of association between baseline characteristics and NMO. The association of plasma HMGB1 level and NMO diagnosis was assessed by the area under the receiver operating characteristic (ROC) curve (AUC), with 95% CI. Although the sample size was relatively small, we attempted to identify an optimal HMGB1 cutoff point; the highest accuracy for NMO diagnosis was determined by the Youden index (maximum of sensitivity + sensitivity-1). The Pearson correlation coefficient (r) was used to show the correlation between HMGB1 versus cytokine levels. The level of significance was set at an alpha level of 0.05. Statistical analyses were performed with SPSS software version 15.0 (SPSS Inc., Chicago, IL, USA).
RESULTS
compared to the MS group at follow-up (2.5 vs. 1.0; P < 0.001), similar to at baseline. HMGB1 level in NMO and MS The plasma level of HMGB1 in the NMO group was significantly greater than that in the MS group (median, 3.99 ng/ml vs. 1.33 ng/ml; P < 0.001) (Table 2). In addition, the active NMO group showed a significantly greater plasma HMGB1 level compared to that in the inactive NMO group (median, 4.72 ng/ml vs. 2.94 ng/ml, P = 0.015) (Fig. 1). We also conducted an analysis of HMGB1 level in AQP4-positive versus AQP4-negative patients with NMO (Table 4). Whereas the plasma HMGB1 level was greater in AQP4-positive patients compared to AQP4negative patients (median: 4.33 vs. 3.72), the difference did not reach statistical significance.
Patient characteristics During the period July 2008 to July 2009, a total of 60 consecutive patients who had experienced at least two attacks of CNS demyelination were selected. Diagnosis of NMO or MS was made after a continuous 2-year clinical follow-up. A total of 29 patients with NMO and 20 patients with MS were included in the analysis (11 patients were excluded, 5 with recurrent optic neuritis, and 6 with recurrent myelitis; none had lesions as determined by brain imaging). Demographic and clinical characteristics at baseline and at 2 years are listed in Table 1. Age, sex, disease duration, and EDSS score were comparable between the two disease groups. Significantly greater value for AQP4 Ab positivity (34.5% vs. 0%; P = 0.003) were observed in the NMO group compared to the MS group. At the 2-year follow-up, the median annual relapse rate had increased significantly from 1.0 to 2.5 in the NMO group and from 0.6 to 1.0 in the MS group (Table 1). This rate was significantly greater in the NMO group
Correlation of inflammatory cytokines with plasma HMGB1 level Plasma levels of TNF-a and IFN-c were significantly greater in the NMO group compared to the MS group (TNF-a: median, 4.91 pg/ml vs. 0 pg/ml, P = 0.021; IFN-c: 10.61 pg/ml vs. 0 pg/ml, P = 0.001) (Table 2), and the plasma HMGB1 level was positively correlated with TNF-a and IFN-c levels (r = 0.517, P = 0.040 and r = 0.659, P = 0.005, respectively) (Table 3). In addition, the plasma IL-17 level was significantly greater in the NMO group compared to the MS group (median, 25.4 pg/ml vs. 0 pg/ml; P = 0.007) (Table 2), and the HMGB1 level was positively correlated with the IL-17 level (r = 0.723, P = 0.028) (Table 3). Diagnosis of NMO Results of univariate logistic regression analysis showed significant association of the following baseline
Table 1. Demographic and clinical characteristics for the NMO and MS groups in this study
Baseline characteristic Age (years)a Sexc Male Female Disease duration (years)b EDSSb AQP4 Abc Positive Negative At 2 years Total number of relapses during disease durationb Annual relapse rateb EDSSb
NMO (n = 29)
MS (n = 20)
P value
41.5 ± 9.7
41.7 ± 12.6
0.946
1 (3.4) 28 (96.6) 5.0 (2.6, 9.0) 3.0 (2.5, 4.5)
4 (20.0) 16 (80.0) 4.0 (3.0, 6.0) 2.8 (1.0, 4.0)
0.144
10 (34.5) 19 (65.5)
0 (0) 20 (100.0)
0.003*
5.0 (3.0, 8.0) 2.5 (1.5, 4.0) 3.0 (2.5, 5.5)
2.0 (2.0, 3.0) 1.0 (1.0, 1.5) 3.0 (0.8, 3.8)
AQP4 Ab, serum AQP4 antibody; EDSS, Expanded Disability Status Scale score; MS, multiple sclerosis; NMO, neuromyelitis optica. a Data are presented as mean ± standard deviation. b Data are presented as median (interquartile range). c Data are presented as count and percentage. * Statistically significant difference between groups. Statistically significant difference between baseline and follow-up.
0.875 0.139
<0.001* <0.001* 0.098
513
K.-C. Wang et al. / Neuroscience 226 (2012) 510–516 Table 2. Comparison of baseline serum HMGB1 and cytokine levels in the NMO and MS groups
HMGB1 (ng/ml) TNF-a (pg/ml) IFN-c (pg/ml) IL-17 (pg/ml)
NMO (n = 29)
MS (n = 20)
P value
3.99 (2.80, 4.97) 4.91 (0, 7.29) 10.61 (3.31, 12.43) 25.38 (15.33, 74.39)
1.33 (1.12, 1.52) 0 (0, 0) 0 (0, 2.57) 0 (0, 0)
<0.001* 0.021* 0.001* 0.007*
Data are presented as median (interquartile range). HMGB1, high-mobility group box 1 protein; IFN-c, interferon-c; IL-17, interleukin 17; MS, multiple sclerosis; NMO, neuromyelitis optica; TNF-a, tumor necrosis factor-a. * Statistically significant difference between groups. Most values for TNF-a and IL-17 were 0 in the MS group, with only three patients positive for TNF-a (2.93, 3.81, 35.74 pg/ml) or IL-17 (3.73, 29.55, 58.09 pg/ml).
Table 4. Plasma HMGB1 level according to AQP4 antibody status in patients with NMO AQP4 Ab-positive AQP4 Ab-negative P (n = 10) (n = 19) value HMGB1 (ng/ml) 4.33 (2.71, 5.48)
3.72 (2.80, 4.80)
0.484
Data are presented as median (interquartile range). AQP4 Ab, serum AQP4 antibody; HMGB1, high-mobility group box 1 protein; NMO, neuromyelitis optica.
Table 5. Univariate logistic regression results identifying baseline factors associated with NMO
Fig. 1. Plasma HMGB1 levels in MS, inactive NMO, and active NMO groups. ⁄Significant difference compared to MS group; significant difference compared to inactive group. MS, multiple sclerosis; HMGB1, high-mobility group box 1 protein; MS, multiple sclerosis; NMO, neuromyelitis optica.
Table 3. Pearson correlation coefficients (r) for baseline plasma HMGB1 level versus cytokine levels TNF-a
IFN-c
IL-17
MS
r P value
0.284 0.325
0.167 0.568
0.216 0.457
NMO
r P value
0.517 0.040*
0.659 0.005*
0.723 0.028*
HMGB1, high-mobility group box 1 protein; IFN-c, interferon-c; IL-17, interleukin 17; MS, multiple sclerosis; NMO, neuromyelitis optica; TNF-a, tumor necrosis factor-a. * Statistically significant.
Variable
OR (95% CI)
P value
Age (years) Sex (female vs. male) Disease duration (years) EDSS HMGB1 (ng/ml) TNF-a (pg/ml) IFN-c (pg/ml) IL-17 (pg/ml)
0.998 (0.95, 1.05) 7.00 (0.72, 68.15) 1.03 (0.91, 1.15) 1.35 (0.97, 1.88) 18.37 (3.06, 110.17) 1.06 (0.95, 1.17) 1.31 (1.07, 1.59) 1.04 (1.00, 1.09)
0.944 0.094 0.666 0.076 0.001* 0.314 0.008* 0.064
The OR for AQP4 Ab (positive vs. negative) was not applicable, owing to a zero count. AQP4 Ab, serum AQP4 antibody; EDSS, Expanded Disability Status Scale score; HMGB1, high-mobility group box 1 protein; IFN-c, interferon-c; IL-17, interleukin 17; NMO, neuromyelitis optica; OR, odds ratio; TNF-a, tumor necrosis factor-a. * Statistically significant.
sensitivity of 89.7% and a specificity of 95.0% (Fig. 2). A cutoff point for HMGB1 of P1.3 ng/ml resulted in a sensitivity of 100% but a lower specificity of 45.0%.
DISCUSSION characteristics with NMO diagnosis (among all study patients, including both the NMO and MS groups): plasma HMGB1 level, and IFN-c level (Table 5). Those with greater plasma HMGB1 level were more likely to have NMO, and the likelihood increased with every unit increase in HMGB1 level (OR = 18.37, P = 0.001). The sample size of 49 was too small to conduct multivariate logistic regression analysis to identify independent factors. An OR for AQP4 Ab (positive vs. negative) was not applicable, owing to a lack of AQP4 Ab-positive patients in the MS group. For diagnosis of NMO according to plasma HMGB1 level, the AUC for the ROC was 0.955, which showed good accuracy. An optimized cutoff point for HMGB1 of P2 ng/ml was determined by the Youden index, which provided a
We performed ELISA for plasma HMGB1 and related cytokines in patients with demyelinating disorders at the initial stages and correlated these levels with the later diagnosis of NMO or MS. Our results showed that plasma HMGB1 level could differentiate NMO from MS in the early stages. A substantial amount of plasma HMGB1 was detected in patients with NMO, and the level correlated well with the levels of disability and recurrence, as assessed by EDSS score and annual relapse rate. In addition, patients with NMO showed an elevated plasma HMGB1 level compared to patients with MS. Furthermore, plasma HMGB1 level was significantly greater in AQP4-positive patients compared to AQP4-negative patients, consistent with NMO status. The rate of AQP4 Ab positivity among patient with NMO
514
K.-C. Wang et al. / Neuroscience 226 (2012) 510–516
Fig. 2. ROC curve and sensitivity of plasma HMGB1 level in the diagnosis of NMO. AUC = 0.955; 95% CI = 0.903, 1.0. AUC, area under the curve; HMGB1, high-mobility group box 1 protein; MS, multiple sclerosis; NMO, neuromyelitis optica; ROC, receiver operating characteristic.
(10/29 patients. 34.5%) was consistent with the relative low seroprevalence of AQP4 Ab in Asian patients with NMO (Matsuoka et al., 2007; Wang et al., 2011). Whereas increased plasma levels of TNF-a, IFN-c, and IL-17 might suggest specificity for NMO, these cytokines are affected by many inflammatory diseases and are associated with the severity of inflammation. Plasma HMGB1 level was a more specific marker for NMO in the present study, given that it was significantly increased in patients who subsequently received a diagnosis of NMO versus MS. This finding provides clinicians with a tool, in addition to severity of clinical attacks and AQP4 Ab positivity, for the diagnosis of NMO at the early stage. When encountering a patient presenting with optical neuritis, myelitis, or both, assessment of plasma HMGB1 level might facilitate an earlier diagnosis of NMO versus MS. The resulting appropriate and different treatment strategies might also minimize damage and prevent subsequent attacks. The pleiotropic cytokine IFN-c, secreted by activated T lymphocytes and natural killer cells, is believed to play a role in immune-mediated demyelinating disorders such as MS (Panitch et al., 1987; Petereit et al., 2002). In addition, IFN-c inhibits central nervous system remyelination (Lin et al., 2006). Our present results indicate a potential association between IFN-c and HMGB1 in patients with NMO. IFN-c has been reported to induce HMGB1 secretion (Strachan et al., 2008), and increased HMGB1 plasma level correlated with active inflammation in our patients. HMGB1 has been reported to activate human neutrophils to produce proinflammatory mediators such as TNF-a, IL-1b, and IL-8 (Li et al., 2004). HMGB1 itself also acts as an immunostimulatory signal to induce maturation of
dendritic cells and the secretion of proinflammatory cytokines including TNF-a, IL-1a, IL-17, IL-8, and IL-12 (Gupta et al., 2004; Marioni et al., 2010). Thus, HMGB1 could play a role in immune-related central nervous system demyelinating diseases. It is well established that HMGB1 is secreted by activated macrophages (Wang et al., 1999; Bonaldi et al., 2003), mature dendritic cells, and natural killer cells (Semino et al., 2005) in response to injury, infection, or other inflammatory stimuli. In the case of NMO, in which a severe inflammatory response is often associated clinically with transverse myelitis and optic neuritis, a role of HMGB1 is possible, given that factors inducing its release, such as extensive necrosis, demyelination, and perivascular macrophage infiltration are common clinical features. There is evidence that autocrine HMGB1 signaling is necessary for the upregulation of B7 costimulation and for IL-12 production of human dendritic cells, which are subsequently required for efficient antigen presentation (Dumitriu et al., 2005). It has also been reported that HMGB1 exerts an inhibitory effect on regulatory T cells in autoimmune disease (Zhang et al., 2008) may induce clonal expansion, survival, and functional polarization of naive T cells (Han et al., 2008). In vivo, HMGB1 has been shown to enhance primary responses to soluble antigens (Kimura et al., 2010). This provides a potential rationale for why patients with NMO are prone to the expression of autoantibodies such as AQP4 Ab/NMOIgG (Lennon et al., 2004), serum nuclear and anti-SS-A/ SS-B autoantibodies (O’Riordan et al., 1996), and non-organ-specific autoantibodies (Hummers et al., 2004). Potential limitations of the present study include the relatively small sample size and the fact that it was conducted at a single institution. The cutoff point determined in the present study was calculated because clinicians generally need a reference point to differentiate between an initial diagnosis of NMO or MS; however, the limited sample size suggests caution in relying on this cutoff point for all cases. It should be mentioned that both the NMO and MS groups showed an increased relapse rate at follow-up (with IFN treatment). In Taiwan, patients receiving IFN often visit the clinic less and refuse continued IFN treatment because they are experiencing fewer attacks. However, some patients experience more attacks and therefore visit the clinic more often. These patients, showing a relatively greater relapse rate and more severe disease, were more likely to be recruited for the present study.
CONCLUSION In conclusion, differentiating NMO from MS, especially at the initial stage, is a real challenge in the clinic. In the past, NMO (not MS) was inferred when abnormalities were detected only in the optic nerve and spinal cord. While not definitive, our present results suggest that HMGB1 might be a diagnostic marker for NMO at the early stage. Additional research is needed to determine whether HMGB1 might also be a mediator of or therapeutic target for NMO. We hope to elucidate
K.-C. Wang et al. / Neuroscience 226 (2012) 510–516
differences in HMGB1 between patients with NMO and those with systemic lupus erythematosus, Sjo¨gren syndrome, or other inflammatory neurologic diseases in future studies.
CONTRIBUTORS Kai-Chen Wang: study design; experimental studies; data analysis; manuscript preparation; manuscript editing. Ching-Piao Tsai: clinical studies; manuscript review. Chao-Lin Lee: clinical studies; experimental studies; data acquisition. Shao-Yuan Chen: literature research; statistical analysis. Li-Te Chin: study concepts; definition of intellectual content. Shyi-Jou Chen: guarantor of integrity of the entire study; manuscript review. All authors have materially participated in the research and manuscript preparation and have approved the final version of the manuscript. Acknowledgments—The authors thank Miss Ruei-Sia Wang for technical assistance. Editorial assistance was provided by Lauren P. Baker, PhD, ELS, in association with MedCom Asia Inc. This work was supported partly by grant from NSC99-2314-B016-002-MY3 to SJ Chen.
REFERENCES Andersson A, Covacu R, Sunnemark D, Danilov AI, Dal Bianco A, Khademi M, Wallstro¨m E, Lobell A, Brundin L, Lassmann H, Harris RA (2008) Pivotal advance: HMGB1 expression in active lesions of human and experimental multiple sclerosis. J Leukoc Biol 84:1248–1255. Andersson U, Wang H, Palmblad K, Aveberger AC, Bloom O, Erlandsson-Harris J, Janson A, Kokkola R, Zhang M, Yang H, Tracey KJ (2000) High mobility group 1 protein (HMG-1) stimulates proinflammatory cytokine synthesis in human monocytes. J Exp Med 192:565–570. Bonaldi T, Talamo F, Scaffidi P, Ferrera D, Porto A, Bachi A, Rubartelli A, Agresti A, Bianchi ME (2003) Monocytic cells hyperacetylate chromatin protein HMGB1 to redirect it towards secretion. EMBO J 22:5551–5560. Bustin M (1999) Regulation of DNA-dependent activities by the functional motifs of the high-mobility-group chromosomal proteins. Mol Cell Biol 19:5237–5246. Cabre P, Gonza´lez-Quevedo A, Bonnan M, Saiz A, Olindo S, Graus F, Smadja D, Merle H, Thomas L, Cabrera-Gomez JA (2009) Relapsing neuromyelitis optica: long term history and clinical predictors of death. J Neurol Neurosurg Psychiatry 80:1162–1164. Cornelio DB, Braga RP, Rosa MW, Ayub AC (2009) Devic’s neuromyelitis optica and pregnancy: distinction from multiple sclerosis is essential. Arch Gynecol Obstet 280:475–477. Cree B (2008) Neuromyelitis optica: diagnosis, pathogenesis, and treatment. Curr Neurol Neurosci Rep 8:427–433. de Seze J, Lebrun C, Stojkovic T, Ferriby D, Chatel M, Vermersch P (2003) Is Devic’s neuromyelitis optica a separate disease? A comparative study with multiple sclerosis. Mult Scler 9:521–525. Dumitriu IE, Baruah P, Valentinis B, Voll RE, Herrmann M, Nawroth PP, Arnold B, Bianchi ME, Manfredi AA, Rovere-Querini P (2005) Release of high mobility group box 1 by dendritic cells controls T cell activation via the receptor for advanced glycation end products. J Immunol 174:7506–7515.
515
Gupta R, Sheikh A, Strachan DP, Anderson HR (2004) Burden of allergic disease in the UK: secondary analyses of national databases. Clin Exp Allergy 34:520–526. Han J, Zhong J, Wei W, Wang Y, Huang Y, Yang P, Purohit S, Dong Z, Wang MH, She JX, Gong F, Stern DM, Wang CY (2008) Extracellular high-mobility group box 1 acts as an innate immune mediator to enhance autoimmune progression and diabetes onset in NOD mice. Diabetes 57:2118–2127. Hazin R, Khan F, Bhatti MT (2009) Neuromyelitis optica: current concepts and prospects for future management. Curr Opin Ophthalmol 20:434–439. Hummers LK, Krishnan C, Casciola-Rosen L, Rosen A, Morris S, Mahoney JA, Kerr DA, Wigley FM (2004) Recurrent transverse myelitis associates with anti-Ro (SSA) autoantibodies. Neurology 62:147–149. Kimura R, Shiibashi R, Suzuki M, Hayashizaki Y, Chiba J (2010) Enhancement of antibody response by high mobility group box protein-1-based DNA immunization. J Immunol Methods 361:21–30. Kurtzke JF (1983) Rating neurologic impairment in multiple sclerosis: an expanded disability status scale (EDSS). Neurology 33:1444–1452. Lai CH, Tseng HF (2009) Population-based epidemiological study of neurological diseases in Taiwan: I. Creutzfeldt–Jakob disease and multiple sclerosis. Neuroepidemiology 33:247–253. Lennon VA, Wingerchuk DM, Kryzer TJ, Pittock SJ, Lucchinetti CF, Fujihara K, Nakashima I, Weinshenker BG (2004) A serum autoantibody marker of neuromyelitis optica: distinction from multiple sclerosis. Lancet 364:2106–2112. Leslie SJ, Johnston N, Strachan FE, Bagnali A, Gray GA, Newby DE, Denvir MA, Webb DJ (2004) Endothelin-1[1–31] is not elevated in men with chronic heart failure. J Cardiovasc Pharmacol 44(Suppl. 1):S96–S99. Li B, Kawatkar SP, George S, Strachan H, Woods RJ, Siriwardena A, Moremen KW, Boons GJ (2004) Inhibition of Golgi mannosidase II with mannostatin A analogues: synthesis, biological evaluation, and structure–activity relationship studies. Chembiochem 5:1220–1227. Lin W, Kemper A, Dupree JL, Harding HP, Ron D, Popko B (2006) Interferon-gamma inhibits central nervous system remyelination through a process modulated by endoplasmic reticulum stress. Brain 129(Pt. 5):1306–1318. Marioni RE, Deary IJ, Strachan MW, Lowe GD, Rumley A, Murray GD, Price JF (2010) Blood rheology and cognition in the Edinburgh Type 2 Diabetes Study. Age Ageing 39:354–359. Matsuoka T, Matsushita T, Kawano Y, Osoegawa M, Ochi H, Ishizu T, Minohara M, Kikuchi H, Mihara F, Ohyagi Y, Kira J (2007) Heterogeneity of aquaporin-4 autoimmunity and spinal cord lesions in multiple sclerosis in Japanese. Brain 130:1206–1223. McDonald WI, Compston A, Edan G, Goodkin D, Hartung HP, Lublin FD, McFarland HF, Paty DW, Polman CH, Reingold SC, Sandberg-Wollheim M, Sibley W, Thompson A, van den Noort S, Weinshenker BY, Wolinsky JS (2001) Recommended diagnostic criteria for multiple sclerosis: guidelines from the international panel on the diagnosis of multiple sclerosis. Ann Neurol 50:121–127. Nakamura M, Houzen H, Niino M, Tanaka K, Sasaki H (2009) Relationship between Barkhof criteria and the clinical features of multiple sclerosis in northern Japan. Mult Scler 15: 1450–1458. Nejentsev S, Cooper JD, Godfrey L, Howson JM, Rance H, Nutland S, Walker NM, Guja C, Ionescu-Tirgovisßte C, Savage DA, Undlien DE, Rønningen KS, Tuomilehto-Wolf E, Tuomilehto J, Gillespie KM, Ring SM, Strachan DP, Widmer B, Dunger D, Todd JA (2004) Analysis of the vitamin D receptor gene sequence variants in type 1 diabetes. Diabetes 53:2709–2712. O’Riordan JI, Gallagher HL, Thompson AJ, Howard RS, Kingsley DP, Thompson EJ, McDonald WI, Miller DH (1996) Clinical, CSF, and MRI findings in Devic’s neuromyelitis optica. J Neurol Neurosurg Psychiatry 60:382–387.
516
K.-C. Wang et al. / Neuroscience 226 (2012) 510–516
Panitch HS, Hirsch RL, Schindler J, Johnson KP (1987) Treatment of multiple sclerosis with gamma interferon: exacerbations associated with activation of the immune system. Neurology 37:1097–1102. Petereit HF, Nolden S, Schoppe S, Bamborschke S, Pukrop R, Heiss WD (2002) Low interferon gamma producers are better treatment responders: a two-year follow-up of interferon beta-treated multiple sclerosis patients. Mult Scler 8:492–494. Polman CH, Reingold SC, Banwell B, Clanet M, Cohen JA, Filippi M, Fujihara K, Havrdova E, Hutchinson M, Kappos L, Lublin FD, Montalban X, O’Connor P, Sandberg-Wollheim M, Thompson AJ, Waubant E, Weinshenker B, Wolinsky JS (2011) Diagnostic criteria for multiple sclerosis: 2010 revisions to the McDonald criteria. Ann Neurol 69:292–302. Poser CM, Paty DW, Scheinberg L, McDonald WI, Davis FA, Ebers GC, Johnson KP, Sibley WA, Silberberg DH, Tourtellotte WW (1983) New diagnostic criteria for multiple sclerosis: guidelines for research protocols. Ann Neurol 13:227–231. Ribeiro MA, Cabral HO, Costa PF (2007) Modulatory effect of NO on sodium currents in a neuroblastoma cell line: aspects of cell specificity. Neurosci Res 58:361–370. Semino C, Angelini G, Poggi A, Rubartelli A (2005) NK/iDC interaction results in IL-18 secretion by DCs at the synaptic cleft followed by NK cell activation and release of the DC maturation factor HMGB1. Blood 106:609–616. Shi Y, Sandoghchian Shotorbani S, Su Z, Liu Y, Tong J, Zheng D, Chen J, Liu Y, Xu Y, Jiao Z, Wang S, Lu L, Huang X, Xu H (2012) Enhanced HMGB1 expression may contribute to Th17 cells activation in rheumatoid arthritis. Clin Dev Immunol 2012:295081. Shimizu J, Hatanaka Y, Hasegawa M, Iwata A, Sugimoto I, Date H, Goto J, Shimizu T, Takatsu M, Sakurai Y, Nakase H, Uesaka Y, Hashida H, Hashimoto K, Komiya T, Tsuji S (2010) IFNb-1b may
severely exacerbate Japanese optic-spinal MS in neuromyelitis optica spectrum. Neurology 75:1423–1427. Strachan MW, Reynolds RM, Frier BM, Mitchell RJ, Price JF (2008) The relationship between type 2 diabetes and dementia. Br Med Bull 88:131–146. Uesugi H, Ozaki S, Sobajima J, Osakada F, Shirakawa H, Yoshida M, Nakao K (1998) Prevalence and characterization of novel pANCA, antibodies to the high mobility group non-histone chromosomal proteins HMG1 and HMG2, in systemic rheumatic diseases. J Rheumatol 25:703–709. Wang H, Bloom O, Zhang M, Vishnubhakat JM, Ombrellino M, Che J, Frazier A, Yang H, Ivanova S, Borovitkova L, Manogue KR, Faist E, Abraham E, Andersson J, Andersson U, Molina PE, Abumrad NN, Sama A, Tracey KJ (1999) HMG-1 as a late mediator of endotoxin lethality in mice. Science 285:248–251. Wang KC, Tsai CP, Lee CL, Chen SY, Chen SJ (2011) The prevalence of long spinal cord lesions and anti-aquaporin 4 antibodies in neuromyelitis optica patients in Taiwan. Eur Neurol 65:99–104. Wingerchuk DM, Hogancamp WF, O’Brien PC, Weinshenker BG (1999) The clinical course of neuromyelitis optica (Devic’s syndrome). Neurology 53:1107–1114. Wingerchuk DM, Lennon VA, Pittock SJ, Lucchinetti CF, Weinshenker BG (2006) Revised diagnostic criteria for neuromyelitis optica. Neurology 66:1485–1489. Wittemann B, Neuer G, Michels H, Truckenbrodt H, Bautz FA (1990) Autoantibodies to nonhistone chromosomal proteins HMG-1 and HMG-2 in sera of patients with juvenile rheumatoid arthritis. Arthritis Rheum 33:1378–1383. Zhang Y, Yao YM, Yu Y, Wu Y, Sheng ZY (2008) Influence of high mobility group box-1 protein on the correlation between regulatory T cells and CD4+ CD25 T cells of spleen in mice. Chin J Surg 46:217–220 (in Chinese).
(Accepted 17 August 2012)