Lipocalin 2 is present in the EAE brain and is modulated by natalizumab

Abstract

Multiple sclerosis (MS) is a demyelinating disease that causes major neurological disability in young adults. A definitive diagnosis at the time of the first episode is still lacking, but since early treatment leads to better prognosis, the search for early biomarkers is needed. Here we characterized the transcriptome of the choroid plexus (CP), which is part of the blood-brain barriers (BBBs) and the major site of cerebrospinal fluid production, in the experimental autoimmune encephalomyelitis (EAE) mouse model of MS. In addition, cerebrospinal fluid samples from two cohorts of patients with MS and with optic neuritis (ON) were analyzed to confirm the clinical relevance of the findings. Genes encoding for adhesion molecules, chemokines and cytokines displayed the most altered expression, supporting the role of CP as a site of immune-brain interaction in MS. The gene encoding for lipocalin 2 was the most up-regulated; notably, the cerebrospinal fluid lipocalin 2 levels coincided with the active phases of the disease. Immunostaining revealed that neutrophils infiltrating the CP were the source of the increased lipocalin 2 expression in this structure. However, within the brain, lipocalin 2 was also detected in astrocytes, particularly in regions typically affected in patients with MS. The increase of lipocalin 2 in the cerebrospinal fluid and in astrocytes was reverted by natalizumab treatment. Most importantly, the results obtained in the murine model were translatable into humans since patients from two different cohorts presented increased cerebrospinal fluid lipocalin 2 levels. The findings support lipocalin 2 as a valuable molecule for the diagnostic/monitoring panel of MS.

Keywords: astrocytes; cerebrospinal fluid; experimental autoimmune encephalomyelitis; lipocalin 2; multiple sclerosis; natalizumab.

Figure 1

LCN2 is increased in EAE. (A) Immunization of SJL female mice was performed on days 0 and 7. Animals were sacrificed at the different phases of the disease and control animals at day 14. The CP transcriptome from the onset, remission, and relapse phases were compared with the CP transcriptome of non-induced animals. (B) Disease course of SJL mice immunized with PLP/CFA emulsion. (C) Augmented levels of CSF LCN2 were observed in the EAE active phases: onset and relapse. (D) Immunohistochemistry analysis showed that LCN2 staining is observed in the CP stroma cells but not in the CP epithelial cells (panels a and b). Staining for Ly6G identified neutrophils as the cells labeling for LCN2 in the CP stroma (panel c–e). LV, lateral ventricle. Data represent mean ± SE. Scale bars, 50 μm.

Figure 2

Transcriptome signature of the EAE CP. (A) Number of genes whose expression was found altered in the onset, remission, and relapse phases. Up-regulated genes are in red and down-regulated are in green. (B) Clustering of the genes whose expression was altered in the CP upon EAE induction. The number of genes altered in each pathway in the onset remission, and relapse phases are represented in red, green and brown, respectively. Microarray data are representative of one single experiment. For each time point at least 2 pools of CP were analyzed and each pool contained CP's from three animals.

Figure 3

qRT-PCR for CP Lcn2. Lcn2 expression was strongly induced during the onset phase of the disease returning to control levels in the remission and relapse phases. Data are representative of one independent experiment (at least 5 pools of CP, each containing CP's from three animals). Data represent mean ± SE.

Figure 4

Schematic representation of the LCN2 expression distribution in the EAE brains. Staining was mostly observed in the olfactory bulb, periventricular regions surrounding the lateral and 3rd ventricles, cerebellum and brain stem in the active phases of EAE. The cerebellum and the brain stem were the regions more highly stained in the relapse phase. No DAB-positive staining was observed in the brain parenchyma of control animals or in the remission phase of the disease. +++, high expression; + +, medium expression; +, low expression. Data are representative of two independent experiments each containing four animals/group.

Figure 5

LCN2 is produced by astrocytes. (A) Double immunofluorescence labeling for LCN2 (green) and of GFAP for astrocytes (red) indicated double-labeled cells in the onset and in the relapse phases (panels d–f and j–l), while no staining was observed in the remission phase or in controls (panels a–c and g–i). (B) Localization of the double positive staining was restricted to white matter and to the granular layer of the cerebellum; in the onset and relapse phases of the disease. None or few cells were found in the cerebellum molecular layer. Mo, molecular layer; Gr, granular layer; WM, white mater; n.d., not found. Data represent mean ± SE. Scale bars, 100 μm.

Figure 6

Natalizumab abrogates the expression of LCN2 and LCN2 is increased in the CSF of MS patients. (A) After EAE induction animals were injected intraperitoneally with natalizumab or IgG (control for the treatment) 1, 3, and 5 days after the appearance of clinical symptoms. Animals were sacrificed at the onset phase for the analysis of LCN2. (B) Treatment attenuated the disease clinical score. (C) CSF LCN2 levels were modulated and normalized by natalizumab treatment. (D and E) LCN2 astrocytic labeling was strongly reduced upon natalizumb treatment both in the white matter and in the granular layer of the cerebellum. (F) LCN2 was detected in 5 out of 20 control samples; 1 out of 9 ON patients and in 12 out of 20 patients with MS. CSF levels were significantly higher in MS patients when compared to the non-MS individuals. Data are representative of two independent cohorts from Portugal and Denmark. Data represent mean ± SE. CSF samples below the detection (0.5 ng/ml) were given the value of 0.5 ng/ml. Mo, molecular layer; Gr, granular layer; WM, white mater. Scale bars, 100 μm.