This article has Open Peer Review reports available.
Passively transferred human NMO-IgG exacerbates demyelination in mouse experimental autoimmune encephalomyelitis
© Saini et al.; licensee BioMed Central Ltd. 2013
Received: 21 January 2013
Accepted: 1 August 2013
Published: 8 August 2013
Neuromyelitis optica (NMO) is a devastating inflammatory disorder of the optic nerves and spinal cord characterized by frequently recurring exacerbations of humoral inflammation. NMO is associated with the highly specific NMO-IgG biomarker, an antibody that binds the aquaporin-4 water channel. Aquaporin-4 is present on glial endfeet in the central nervous system (CNS). In humans, the NMO-IgG portends more frequent exacerbations and a worse long-term clinical outcome.
We tested the longer-term outcome of mice with EAE injected with NMO-IgG and followed them for 60 days. Clinical exams and pathology of the spinal cord and optic nerves were compared to mice that received control human IgG.
Passively transferred human NMO-IgG leads to more severe neurology disability over two months after onset of disease. Clinical worsening is associated with an increased concentration of large demyelinating lesions primarily to subpial AQP4-rich regions of the spinal cord.
NMO-IgG is pathogenic in the context of EAE in mice.
Neuromyelitis optica (NMO) is a devastating neuroinflammatory disorder that preferentially targets the optic nerves, brainstem and spinal cord . Also known as Devic’s disease, NMO is associated with the highly specific NMO-IgG antibody found in up to 63% of patients . The target of the NMO-IgG is the aquaporin-4 (AQP4) water channel expressed in multiple tissues in the body. AQP4 is the major aquaporin found in the CNS and is highly localized to the endfeet of astrocytes, especially along the pia limitans and on the abluminal surface of blood vessels in the brain .
The NMO-IgG is hypothesized to be pathogenic; binding of the antibody to its glial target triggers a humoral inflammatory cascade involving IgG, IgM, complement deposition and recruitment of neutrophils and eosinophils . This model of disease is supported by two reports of passively transferred NMO-IgG in which the NMO-IgG exacerbates behavioral signs of rat experimental autoimmune encephalomyelitis (EAE) and induces a pathology similar to human NMO: areas of acute inflammation with granulocytes, a dramatic loss of aquaporin-4 staining and complement deposition [5, 6]. While EAE induced by myelin basic protein in complete Freund’s adjuvant (CFA) in Lewis rats generally leads to a complete neurologic recovery , EAE induced by myelin oligodendrocyte glycoprotein peptide 35–55 (MOG35-55) in C57Bl6 mice causes demyelination and axon loss in the spinal cord with limited behavioral recovery, the latter of which may better represent a more suitable animal model system for severe human neuromyelitis optica disease.
One of the hallmarks of the NMO-IgG seropositive testing in humans with NMO disease is the prognostic implication of more frequent recurrences and worse neurologic outcome with increased disability compared to individuals with NMO whose serum does not react with AQP4 . In our study, we tested the consequences of NMO-IgG in mouse EAE passively transferred at disease onset on long-term outcome and found that the NMO-IgG results in a worse neurological outcome which is maintained as late as two months after immunization. Pathological evaluation revealed larger, primarily subpial demyelinated lesions in the spinal cord and optic nerves of EAE mice receiving passively transferred NMO-IgG.
Adult female C57/BL6 mice between 6 – 8 weeks of age were purchased from The Jackson Laboratory and housed in a 12-hour artificial light–dark cycle and had ad libitum access to food and water. The Johns Hopkins Institutional Animal Care and Use Committee approved all experimental procedures.
Human IgG fractions were purified from the plasma of patients undergoing plasma exchange using a resin based purification method (Melon Gel IgG Purification kit, Thermo Scientific) two days prior to injection. The purified IgG was concentrated by spin column centrifugation (Amicon Ultra, 100kD MW cut off) and the final protein concentration was adjusted to 25 mg/ml for 100 μl intra-peritoneal injection. All NMO patients tested seropositive for the NMO-IgG by the Mayo clinical NMO-IgG assay and the NMO plasma samples from 3 patients were pooled prior to purification. Human control IgG fraction (control-IgG) was obtained from a non-NMO patient undergoing plasma exchange for ABO incompatibility. All samples were obtained through a protocol approved by the Johns Hopkins Institutional Review Board (NA_ 00003551) in a de-identified manner with informed consent to use the samples for research.
EAE induction and scoring
Experimental autoimmune encephalomyelitis (EAE) was induced in 3 groups of 10 mice. Each animal received a single subcutaneous injection of one hundred microliters of an emulsified solution of 1:1, 5 mg/ml Myelin Oligodendrocyte Glycoprotein (MOG) peptide 35–55 in phosphate buffered saline (PBS) and incomplete Freund’s adjuvant containing 12.5 mg/ml heat-killed Mycobacterium tuberculosis; each animal therefore received 250 μg MOG35-55 and 625 μg M. tuberculosis (day 0). Pertussis Toxin (300 ng) was administered intraperitoneally on days 0 and 2. Animals were weighed daily and scored on a standardized 5-point EAE disability scale by a blinded examiner . A series of 4 intraperitoneal injections of human IgG purified from either pooled NMO plasma or control human plasma were administered on days 13, 14, 18, and 19 for a total of 10 mg/animal , with the initial pair of injections corresponding to onset of EAE disease and the second pair of injections provided during the initial remission of EAE. Vehicle controls received an equal volume of PBS.
Tissue processing and histology
Animals were anesthetized with isofluorane and perfused via cardiac puncture first with PBS and then with freshly prepared 4% formaldehyde solution. The optic nerves and spinal cords were harvested, fixed overnight, cryopreserved in 30% sucrose and frozen for sectioning. After embedding tissue in in O.C.T. Compound (Tissue-Tek® ), ten to thirty micron slices sections were mounted on Superfrost Plus Microscope Slides (Fisher brand). The first cohort of animals was sacrificed 20 minutes after the last intraperitoneal injection of human IgG for the purpose of tracking human antibody entrance into the mouse central nervous system. The second cohort of animals was sacrificed and their tissue was prepared in a similar fashion on day 62 post-EAE induction.
Eriochrome cyanine staining for myelin
Eriochrome cyanine was used to identify demyelinating lesions in the sectioned tissue. Slides with frozen sections were thawed at room temperature. Eriochrome cyanine solution was prepared by dissolving eriochrome cyanine in 450 ml 0.5% H2SO4(0.2%) and 10% FeCl3 added to a final concentration of 0.4%. The sectioned tissue was hydrated by serial washes in 100% ethanol, 95% ethanol, 70% ethanol and distilled water for 10 minutes each and then immersed for 15 minutes in eriochrome cyanine solution. After staining, differentiation was carried out in freshly made 0.1% NH4OH for 20–30 seconds and halted by thorough washing in distilled water. Slides were mounted as described in the preceding section.
Details of the primary antibody reagents used for immunohistochemical experiments in these studies
Anti-Aquaporin 4, C-terminus
5% BSA-PBS + 1:100 NGS + 0.1% Triton X-100
Purified Rat anti-mouse CD45
All cells of hematopoietic origin, except erythrocytes
5% BSA-PBS + 1:100 NGS
Purified Rat anti-mouse CD5
Thymocytes, T lymphocytes, thymic NK-T cells, subset of B lymphocytes
5% BSA-PBS + 1:100 NGS + 0.1% Triton X-100
Purified Rat anti-mouse CD45R (B220)
5% BSA-PBS + 1:100 NGS + 0.1% Triton X-100
Purified Rat anti-mouse Ly-6G and Ly-6C
5% BSA-PBS + 1:100 NGS + 0.1% Triton X-100
Purified Goat anti-human
5% BSA-PBS + 1:100 NGS
Six to eight areas of eriochrome-stained spinal cord were photographed at high resolution (4080x3072) using a 4x objective and background corrected to correct for uneven illumination. Image analysis was performed in a blinded fashion. Using ImageProPlus5 software, two measurements were acquired from each image: (1) a total measure of the white matter area (including eriochrome-stained regions and demyelinated areas) and (2) a measure of each demyelinated lesion’s area, outlined as separate areas of interest (AOIs). During analysis, each lesion was noted to be largely bordered by the pia mater or not; while most EAE lesions contact the pia, this analysis was to indicate lesions whose major axis contacted the pia. For total demyelinated area, lesioned areas were summed. For percent demyelinated area, this measure was divided by the total white matter area. “Large pial lesions” were classified as those comprised of ≥10,000μm2. For percent large pial lesions, the total area of these large lesions was divided by the total demyelinated area for that animal to demonstrate the prevalence of these large subpial confluent lesions in these mice. For percent pial lesions, the area of all lesions blindly classified as primarily pial was summed and divided by the total demyelinated area for that animal.
For comparison of EAE scores, a non-parametric (Mann–Whitney U-test) was performed. For comparison of demyelinated areas, Student t-tests were performed. Data was analyzed and depicted using GraphPad 5.0 software. Results were deemed statistically significant for p<0.05.
Our mouse model of NMO resembles human NMO in several ways. First, the neurological phenotype of mouse EAE worsens in the presence of NMO-IgG and remains persistently worse for the duration of our 2-month study. This correlates with large subpial demyelinated lesions in the spinal cord. Second, the optic nerves and spinal cord are the predominant sites of pathology, similar to human NMO disease. Thirdly, pathologic evaluation of EAE mice exposed to NMO-IgG showed evidence of infiltration by granulocytes in acute lesions.
In the context of ameliorating NMO severity, one may conclude from this study that NMO-IgG may be harmful when present in acute inflammatory lesions of the CNS and should be targeted for removal. Data on patients is mixed with regards to the correlation of NMO-IgG titer with either the severity of an acute attack or the degree of recovery following treatment [10, 11]. However, there is good evidence that presence of the NMO-IgG is predictive of more frequent relapses and a worse neurologic outcome , the latter of which we observed in our mouse model.
There has been debate on the role of the NMO-IgG as a pathogenic antibody versus a biomarker of disease [12–14]. In our model, there is strong evidence in favor of the NMO-IgG as a contributor to the neurological pathology and behavioral outcome. As in rat models of NMO [5, 6], there is no effect of the NMO-IgG in the absence of EAE in this strain of mouse and there is no penetration of NMO-IgG into the CNS parenchyma in the absence of EAE. However, NMO-IgG appears to convert the pathology and disease course in mice from acute behavioral disease with small lesions within superficial white matter of the dorsal columns to confluent pial lesions in the dorsal and lateral spinal cord. Although the total demyelination is similar, large confluent lesions involving the dorsolateral corticospinal tracts may lead to greater disability in this mouse model. We also found limited evidence for perivascular granulocyte recruitment in the NMO-IgG treated animals but the role of these cells in altering the distribution of lesions and in causing persistent behavioral disease in our EAE mice needs further study. Our hypothesis is that NMO-IgG targets AQP4, which is preferentially expressed on the astrocytic foot processes of the pia-glia limitans. As a result of the antibody binding, a humoral-mediated inflammatory process ensues that includes granulocyte infiltration. It is temporary because the human NMO-IgG is cleared shortly after passive transfer. Interestingly, while behavioral disease persisted for 2 months, human IgG and granulocytes were not detectable in the CNS at the 2-month time point. This suggests that transient exposure to the NMO-IgG has long-lasting consequences. This might explain why NMO-IgG titer is not easily linked to disease severity: transient spikes in titer coinciding with blood–brain permeability in the spinal cord and optic nerve may cause tissue damage and disease, but the overlap of these two events may be rare.
In contrast to our mouse model, rats with EAE that received control-IgG did not manifest behavioral worsening defined as a higher EAE disability score [5, 6]. In addition, some patients’ NMO-IgG have more toxic potential than others . While this may be due to the presence of autoantibodies, which are not reactive to the animal model species being used, it is still unclear as to why we observed transient worsening of behavioral signs in EAE mice injected with the control-IgG . The patient we isolated our control-IgG from suffered from chronic kidney disease that required plasma exchange to remove ABO incompatible antibodies from circulation prior to kidney transplant. In addition to not being positive for AQP4-immunoreactivity, we found no evidence of other autoantibodies binding to antigens in the mouse CNS. The observed transient worsening in mice receiving the control-IgG did not significantly alter the distribution of lesions as did the NMO-IgG. Because this worsening was not observed in rat models of EAE, this effect of control-IgG could be unique to this patient’s IgG. The difference in rodent species could also influence the course of EAE and reactivity of autoantibodies.
Another difference between this mouse NMO model and the previous rat NMO models is that the behavioral worsening in our mouse model was not accompanied by AQP4 depletion. This could mean that higher EAE disability score with NMO-IgG does not require overt depletion of astrocytes or of AQP4 from lesions but does not rule out AQP4 involvement. Interestingly, even severe depletion of AQP4 and astrocytes from lesioned areas in rat animal models does not lead to myelin damage, suggesting that myelin can survive at least transient loss of AQP4. The long term consequences of short term NMO-IgG exposure in rat models has not yet been reported. Also, purified preparations of NMO-IgG have been difficult to obtain due to the polyclonal nature of the antibody and the uncertainty about which antigenic targets on AQP4 are important in the pathogenesis of disease . Further work in this area will allow better specificity of passive transfer studies in NMO.
In this experimental model of EAE in C57/BL6 mice in which passively transferred IgG from NMO patients is infused at onset of neurologic disease, we conclude that the anti-AQP4 antibody contributes to the pathogenicity by targeting subpial astrocytes leading to large, confluent demyelinated lesions in the spinal cord and optic nerve.
This study was funded by the Guthy Jackson Charitable Foundation. We would like to thank Dr. Carlos Pardo for his assistance with the neuropathology.
- Graber DJ, Levy M, Kerr D, Wade WF: Neuromyelitis optica pathogenesis and aquaporin 4. J Neuroinflammation. 2008, 5: 22-10.1186/1742-2094-5-22.View ArticlePubMedPubMed CentralGoogle Scholar
- McKeon A, Fryer JP, Apiwattanakul M, Lennon VA, Hinson SR, Kryzer TJ, Lucchinetti CF, Weinshenker BG, Wingerchuk DM, Shuster EA, Pittock SJ: Diagnosis of neuromyelitis spectrum disorders: comparative sensitivities and specificities of immunohistochemical and immunoprecipitation assays. Arch Neurol. 2009, 66: 1134-1138. 10.1001/archneurol.2009.178.PubMedGoogle Scholar
- Roemer SF, Parisi JE, Lennon VA, Benarroch EE, Lassmann H, Bruck W, Mandler RN, Weinshenker BG, Pittock SJ, Wingerchuk DM, Lucchinetti CF: Pattern-specific loss of aquaporin-4 immunoreactivity distinguishes neuromyelitis optica from multiple sclerosis. Brain. 2007, 130: 1194-1205. 10.1093/brain/awl371.View ArticlePubMedGoogle Scholar
- Lucchinetti CF, Mandler RN, McGavern D, Bruck W, Gleich G, Ransohoff RM, Trebst C, Weinshenker B, Wingerchuk D, Parisi JE, Lassmann H: A role for humoral mechanisms in the pathogenesis of Devic's neuromyelitis optica. Brain. 2002, 125: 1450-1461. 10.1093/brain/awf151.View ArticlePubMedGoogle Scholar
- Bradl M, Misu T, Takahashi T, Watanabe M, Mader S, Reindl M, Adzemovic M, Bauer J, Berger T, Fujihara K, Itoyama Y, Lassmann H: Neuromyelitis optica: pathogenicity of patient immunoglobulin in vivo. Ann Neurol. 2009, 66: 630-643. 10.1002/ana.21837.View ArticlePubMedGoogle Scholar
- Kinoshita M, Nakatsuji Y, Kimura T, Moriya M, Takata K, Okuno T, Kumanogoh A, Kajiyama K, Yoshikawa H, Sakoda S: Neuromyelitis optica: Passive transfer to rats by human immunoglobulin. Biochem Biophys Res Commun. 2009, 386: 623-627. 10.1016/j.bbrc.2009.06.085.View ArticlePubMedGoogle Scholar
- Swanborg RH: Experimental autoimmune encephalomyelitis in the rat: lessons in T-cell immunology and autoreactivity. Immunol Rev. 2001, 184: 129-135. 10.1034/j.1600-065x.2001.1840112.x.View ArticlePubMedGoogle Scholar
- Jarius S, Wildemann B: AQP4 antibodies in neuromyelitis optica: diagnostic and pathogenetic relevance. Nat Rev Neurol. 2010, 6: 383-392. 10.1038/nrneurol.2010.72.View ArticlePubMedGoogle Scholar
- Jones MV, Nguyen TT, Deboy CA, Griffin JW, Whartenby KA, Kerr DA, Calabresi PA: Behavioral and pathological outcomes in MOG 35–55 experimental autoimmune encephalomyelitis. J Neuroimmunol. 2008, 199: 83-93. 10.1016/j.jneuroim.2008.05.013.View ArticlePubMedGoogle Scholar
- Hinson SR, McKeon A, Fryer JP, Apiwattanakul M, Lennon VA, Pittock SJ: Prediction of neuromyelitis optica attack severity by quantitation of complement-mediated injury to aquaporin-4-expressing cells. Arch Neurol. 2009, 66: 1164-1167. 10.1001/archneurol.2009.188.View ArticlePubMedGoogle Scholar
- Takahashi T, Fujihara K, Nakashima I, Misu T, Miyazawa I, Nakamura M, Watanabe S, Shiga Y, Kanaoka C, Fujimori J, Sato S, Itoyama Y: Anti-aquaporin-4 antibody is involved in the pathogenesis of NMO: a study on antibody titre. Brain. 2007, 130: 1235-1243. 10.1093/brain/awm062.View ArticlePubMedGoogle Scholar
- Cayrol R, Saikali P, Vincent T: Effector functions of antiaquaporin-4 autoantibodies in neuromyelitis optica. Ann N Y Acad Sci. 2009, 1173: 478-486. 10.1111/j.1749-6632.2009.04871.x.View ArticlePubMedGoogle Scholar
- Frohman EM, Kerr D: Is neuromyelitis optica distinct from multiple sclerosis?: something for “lumpers” and “splitters”. Arch Neurol. 2007, 64: 903-905. 10.1001/archneur.64.6.903.View ArticlePubMedGoogle Scholar
- Weinstock-Guttman B, Miller C, Yeh E, Stosic M, Umhauer M, Batra N, Munschauer F, Zivadinov R, Ramanathan M: Neuromyelitis optica immunoglobulins as a marker of disease activity and response to therapy in patients with neuromyelitis optica. Mult Scler. 2008, 14 (8): 1061-1067. 10.1177/1352458508092811.View ArticlePubMedGoogle Scholar
- Kinoshita M, Nakatsuji Y, Kimura T, Moriya M, Takata K, Okuno T, Kumanogoh A, Kajiyama K, Yoshikawa H, Sakoda S: Anti-aquaporin-4 antibody induces astrocytic cytotoxicity in the absence of CNS antigen-specific T cells. Biochem Biophys Res Commun. 2010, 394: 205-210. 10.1016/j.bbrc.2010.02.157.View ArticlePubMedGoogle Scholar
- Lapointe BM, Herx LM, Gill V, Metz LM, Kubes P: IVIg therapy in brain inflammation: etiology-dependent differential effects on leucocyte recruitment. Brain. 2004, 127: 2649-2656. 10.1093/brain/awh297.View ArticlePubMedGoogle Scholar
- Tani T, Sakimura K, Tsujita M, Nakada T, Tanaka M, Nishizawa M, Tanaka K: Identification of binding sites for anti-aquaporin 4 antibodies in patients with neuromyelitis optica. J Neuroimmunol. 2009, 211: 110-113. 10.1016/j.jneuroim.2009.04.001.View ArticlePubMedGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2377/13/104/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.