Neurotoxic autoantibodies mediate congenital cortical impairment of offspring in maternal lupus

Abstract

Systemic lupus erythematosus (SLE) is an autoimmune disease mediated by autoantibodies and preferentially affecting women of childbearing age. Because the offspring of mothers with SLE show a high frequency of learning disorders, we hypothesized that maternally transferred autoantibodies that bind DNA and the N-methyl-D-aspartate receptor (NMDAR) could have a pathogenic role during fetal brain development. Here we describe a maternal SLE mouse model wherein pregnant dams harbored DNA-specific, NMDAR-specific autoantibodies throughout gestation. High titers of these autoantibodies in maternal circulation led to histological abnormalities in fetal brain and subsequent cognitive impairments in adult offspring. These data support a paradigm in which in utero exposure to neurotoxic autoantibodies causes abnormal brain development with long-term consequences. This paradigm may apply to multiple congenital neuropsychiatric disorders.

Figure 1

MP brains exposed to maternal NMDAR-specific AAbs in utero display morphological abnormalities at E15. (a) Serology of BALB/c adult females (n = 5 per group, 3 cohorts per group) used in fetal and behavioral studies, expressed as optical density (OD, means ± s.d.). Mice were immunized with DWEYSVWLSN peptide (MP, 100 μg) or polylysine core (MC, 100 μg). MP dams were segregated into high (MPH) and low (MPL) AAb titer groups. (b) TUNEL assays, in coronal sections, show increased cell death in MP brains. Insets (top right) correspond to the boxed areas. (c) DAPI (left panels) detects cell nuclei and PH3 (middle panels) assesses mitotic activity. Most PH3+ cells are located near the ventricular zone (VZ, single arrows), corresponding to the expected labeling pattern. In MP brain, PH3+ cells are also detected in the marginal zone (MZ, double arrows in merged MP panel). (d) The cortical plate (CP) is visualized as a DAPI-dense cellular band lacking nestin (arrows in merged panels). CW, cortical width. (e) TUNEL+ cells and PH3+ cells per unit area (means ± s.d., ** P < 0.0001, Mann-Whitney U test). (f) CP thickness (at left), measured from merged sections, is significantly reduced in MP neocortex (P < 0.0001, Mann-Whitney U test). Ratio of CP to CW, at right, is significantly lower in MP neocortex (means ± s.e.m, ** P < 0.0001, Mann-Whitney U test). For each assay, 2-3 fetal brains per dam were analyzed; n = 3–4 dams by group. Scale bar = 250 μm (inset = 25 μm).

Figure 2. Morphological abnormalities in E15 brains exposed to murine anti-NMDAR AAbs in utero

The monoclonal AAb R4A or the control IgG2b was intravenously administered to pregnant dams (200 μg per dam), at E12. Embryos were harvested at E15 for analysis. Coronal sections of E15 fetal brains were stained with PH3, DAPI and nestin. (a) R4A shows reactivity to MP (peptide) and double-stranded DNA (dsDNA) by ELISA (mean ± s.d., ** P < 0.0001, Mann-Whitney U test). (b) Mitotic activity is assessed with PH3 expression. The VZ harbors most PH3+ cells (single arrows) in both groups, but PH3+ cells are found outside of the VZ in R4A-exposed neocortex (double arrows). (c) VZ and (SVZ) of R4A-exposed fetal brains show significantly more PH+ cells (mean ± s.d., ** P < 0.0001, Mann-Whitney U test). (d) Combined DAPI/nestin stained sections of R4A or IgG2b neocortex are used to measure the CP (arrows in the merged panels). Fetal brains exposed to R4A in utero show CP thinning. (e) CP thickness is significantly decreased in the neocortex exposed to R4A (P < 0.0001, Mann-Whitney U test). CP to CW ratio is significantly reduced in R4A-treated neocortex (mean ± s.e.m., ** P < 0.0001, Mann-Whitney U test), demonstrating that passive transfer of R4A from maternal circulation to the fetal neocortex disrupts proper development of the CP. For each assay, 2 fetal brains per dam were analyzed; n = 3 dams for each group. Scale bar = 250 μm.

Figure 3. Morphological abnormalities in E15 brains exposed to human lupus anti-NMDAR AAbs in utero

Monoclonal Abs (G11 and B1) were intravenously administered to pregnant dams (150 μg per dam), at E12. Embryos were harvested at E15. Coronal sections of fetal brains were stained with PH3, DAPI and nestin. (a) G11 and B1 were tested for peptide, dsDNA, NR2A and NR2B reactivity by ELISA. G11 shows robust reactivity with each antigen, while B1 shows negligible reactivity (mean ± s.d., ** P < 0.0001, Mann-Whitney U test). (b) PH3 staining shows that the VZ (single arrows) harbors most PH3+ cells. Similar to R4A-treated fetal cortex, PH3+ cells are found outside of the VZ in G11-exposed neocortex (double arrows). The graph displays the number of PH3+ cells per unit area (mean ± s.d., ** P < 0.0001, Mann-Whitney U test). (c) Coronal sections stained with DAPI and nestin were used to visualize the CP (arrows in merged sections). G11-exposed brains show CP thinning. (d) CP thickness is significantly decreased in G11-exposed neocortex when compared to B1 neocortex (P < 0.0001, Mann-Whitney U test). CP to CW ratio demonstrates that passive transfer of G11 from maternal circulation to fetal circulation causes disruption of the proper organization of the CP (mean ± s.e.m., ** P < 0.0001, Mann-Whitney U test). For each assay, 3 fetal brains per dam were analyzed; n = 2 dams for each group. Scale bar = 250 μm.

Figure 4. Congenital impairments in neonatal reflexes and cognitive tasks

(a) Offspring born to MPH (n = 19), MPL (n = 14) and MC (n = 15) dams were examined in the negative geotaxis reflex. MPH offspring show significant delay in reflex acquisition (mean ± s.e.m., ** P < 0.001, t test MPH vs. MC). (b) The extinction task quantifies the ability to erase fearful responses to stimuli previously associated with foot-shocks. Adult MC (n = 10) and MPL (n = 8) groups show negligible freezing after 20 trials, whereas MPH (n = 10) mice display significantly higher freezing (means ± s.e.m., ** P < 0.001, ANOVA with trials as repeated measure). (c) The novel object recognition task (left diagram) shows that MPH mice are impaired (mean discrimination ratio ± s.e.m., ** P = 0.001, Mann-Whitney U test), whereas MPL mice behave normally. (d) The topological task (left diagram) shows that MPH mice fail to recognize novel configurations (mean topological ratio ± s.e.m., * P = 0.0015, Mann-Whitney U test). (e) Coronal sections of parietal cortex (cresyl violet) of adult mice show disorganization and thinning of cortical layers in MPH mice. Scale bar = 250 μm. (f) Cumulative probability plots for neuronal size are abnormal for MPH animals in the sensorimotor (left) and parietal (right) regions (* P < 0.05, ** P < 0.001, Kolmogorov-Smirnov test). (g) Regional volumes (mean ± s.d.) of adult neocortex show significant decrease in cortical volume in MPH mice (P < 0.0001, Kruskal-Wallis test).