Pathogenic implications of cerebrospinal fluid barrier pathology in neuromyelitis optica

Acta Neuropathol DOI 10.1007/s00401-017-1682-1 ORIGINAL PAPER Pathogenic implications of cerebrospinal fluid barrier pathology in neuromyelitis optica Yong Guo1 · Stephen D. Weigand2 · Bogdan F. Popescu3 · Vanda A. Lennon1,4,5,6 · Joseph E. Parisi1,4 · Sean J. Pittock1,6 · Natalie E. Parks1 · Stacey L. Clardy1  · Charles L. Howe1,5,6,7 · Claudia F. Lucchinetti1,6 Received: 6 August 2016 / Revised: 12 January 2017 / Accepted: 25 January 2017 © The Author(s) 2017. This article is published with open access at Springerlink.com Abstract Pathogenic autoantibodies associated with neuromyelitis optica (NMO) induce disease by targeting aquaporin-4 (AQP4) water channels enriched on astrocytic endfeet at blood–brain interfaces. AQP4 is also expressed at cerebrospinal fluid (CSF)–brain interfaces, such as the pial glia limitans and the ependyma and at the choroid plexus blood–CSF barrier. However, little is known regarding pathology at these sites in NMO. Therefore, we evaluated AQP4 expression, microglial reactivity, and complement deposition at pial and ependymal surfaces and in the fourth ventricle choroid plexus in 23 autopsy cases with clinically Electronic supplementary material The online version of this article (doi:10.1007/s00401-017-1682-1) contains supplementary material, which is available to authorized users. * Charles L. Howe * Claudia F. Lucchinetti 1 Department of Neurology, Mayo Clinic, 200 First Street SW, Rochester, MN 55905, USA 2 Department of Health Sciences Research, Mayo Clinic, Rochester, MN, USA 3 Department of Anatomy and Cell Biology, Cameco MS Neuroscience Research Center, University of Saskatchewan, Saskatoon, SK, Canada 4 Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, MN, USA 5 Department of Immunology, Mayo Clinic, Rochester, MN, USA 6 Center for Multiple Sclerosis and Autoimmune Neurology, Mayo Clinic, Rochester, MN, USA 7 Department of Neuroscience, Mayo Clinic, Rochester, MN, USA and/or pathologically confirmed NMO or NMO spectrum disorder. These findings were compared to five cases with multiple sclerosis, five cases of choroid plexus papilloma, and five control cases without central nervous system disease. In the NMO cases, AQP4 immunoreactivity was reduced relative to control levels in the pia (91%; 21/23), ependyma (56%; 9/16), and choroid plexus epithelium (100%; 12/12). AQP4 immunoreactivity was normal in MS cases in these regions. Compared to MS, NMO cases also showed a focal pattern of pial and ependymal complement deposition and more pronounced microglial reactivity. In addition, AQP4 loss, microglial reactivity, and complement deposition colocalized along the pia and ependyma only in NMO cases. Within the choroid plexus, AQP4 loss was coincident with C9neo immunoreactivity on epithelial cell membranes only in NMO cases. These observations demonstrate that NMO immunopathology extends beyond perivascular astrocytic foot processes to include the pia, ependyma, and choroid plexus, suggesting that NMO IgG-induced pathological alterations at CSF–brain and blood–CSF interfaces may contribute to the occurrence of ventriculitis, leptomeningitis, and hydrocephalus observed among NMO patients. Moreover, disruption of the blood–CSF barrier induced by binding of NMO IgG to AQP4 on the basolateral surface of choroid plexus epithelial cells may provide a unique portal for entry of the pathogenic antibody into the central nervous system. Keywords Choroid plexus · Leptomeninges · Astrocyte · Complement · Immunopathology · Hydrocephalus Introduction Neuromyelitis optica (NMO) is a disabling inflammatory disorder of the central nervous system (CNS) that is marked 13 Acta Neuropathol by expression of a pathogenic IgG autoantibody directed against the ectodomain of aquaporin-4 (AQP4), the major water channel in the CNS [34, 35]. AQP4 functions to couple bidirectional fast water transport to active ion flux across the plasma membrane, thereby controlling astrocyte homeostasis and CNS osmotic stability [76]. The enrichment of AQP4 on astrocytic endfeet at the blood–brain barrier is consistent with a crucial role in maintaining physiological water balance and with responding to pathological perturbations of this balance associated with ischemia or trauma [75]. Early NMO pathology is characterized by the presence of reactive astrocytes, intramyelinic edema, loss of AQP4 expression, variable perivascular deposition of IgG and complement components, and granulocytic leukocyte infiltration [40, 61]. Advanced lesions demonstrate more profound complement deposition, loss of myelin, and astrocyte destruction. Astrocytic responses in NMO range from sublytic gliosis to overt lysis, and these responses are frequently observed in regions without myelin loss, suggesting that NMO is a primary astrocytopathy associated with secondary demyelination [39]. Such a model for NMO pathogenesis is consistent with observations of water dyshomeostasis [62], lesion reversibility [41, 42, 80], and behavioral sequelae in NMO patients [59]. AQP4 is also expressed at the pial glia limitans, the ependyma, and the choroid plexus [53, 61]. Notably, differential cell–cell junction expression at these barrier sites [78] may provide a unique route for NMO IgG to enter the cerebrospinal fluid (CSF) and access AQP4-expressing targets in the brain parenchyma. Recently, loss of AQP4 expression was observed in cortical layer I in NMO tissue and was associated with cognitive impairment and a corresponding loss of neurons in cortical layer II [65]. This finding suggests that subpial AQP4 was targeted by NMO IgG and indicates that astrocytopathy outside of the typical periventricular lesion may have profound pathogenic and neurologic consequences. However, at present, little is known about the pathology of CSF–brain and blood–CSF barriers in NMO patients. Therefore, in this study, we analyzed astrocyte and microglia reactivity, AQP4 expression level, and complement deposition at these interfaces. Materials and methods Study design and series This study was approved by the Institutional Review Board of Mayo Clinic, Rochester, MN (IRB 2067-99). Inclusion criteria were (i) clinical and pathological diagnosis of NMO or NMOSD; (ii) sufficient archival tissue for pathological analysis; and (iii) no evidence of alternative diagnosis. Twenty-three autopsy cases met the inclusion criteria (315 total tissue blocks) (Supplemental Figure 1). Table 1 13 provides demographics for the patient cohort. As controls, we included five multiple sclerosis (MS) cases (79 blocks; 4 relapsing remitting MS, 1 secondary progressive MS) and five control cases without known CNS disease (45 blocks). As additional controls for choroid plexus disease, we included five hydrocephalus cases (five blocks) and five papilloma cases (five blocks). We analyzed supratentorial brain (including optic nerve), brain stem, cerebellum, and spinal cord for histopathological abnormalities in the (...truncated)