Blocking Placental Class G Immunoglobulin Transfer Prevents NMDA Receptor Antibody Effects in Newborn Mice

Human Serum Samples, IgG Purification, and Immunoabsorption

IgG was isolated by ammonium sulfate precipitation from serum of 7 patients with anti-NMDAR encephalitis and serum of 7 healthy blood donors. All patients with anti-NMDAR encephalitis were women (median age 20 years, range 16–26 years) with classical anti-NMDAR encephalitis; none of the patients had teratoma, and in all instances, serum samples were obtained before treatment. The presence of NMDAR antibodies was determined with rat brain tissue immunohistochemistry and a cell-based assay (CBA) with human embryonic kidney (HEK) 293T cells expressing GluN1 and GluN2B receptor subunits, as reported.¹ The absence of other brain-specific antibodies was established using previously reported experiments^13,22 in which patients' NMDAR antibodies were immunoabsorbed, resulting in abrogation of patients' IgG effects in cultured neurons.²³

Animals

Pregnant C57BL/6J mice (9 weeks old, 25–30 g, Charles River) were housed in cages of 5 until pregnancy day 16 (E16), at which time they were caged alone in previously reported room conditions.¹⁶ Representative offspring from every litter were separated by sex on postnatal day (PD) 21, and kept in the same caging conditions until the indicated time points (i.e., PD 21 or 43). Overall, 21 pregnant female mice, 22 fetuses, and 91 pups were used for behavioral, electrophysiologic, morphological, and synaptic brain studies.

Standard Protocol Approvals, Registrations, and Patient Consents

Written informed consent was obtained from all patients; the study was approved by the local institutional review board at Hospital Clínic de Barcelona (registration number HCB/2018/0192). All procedures were approved by the local ethical committee of the University of Barcelona following European (2010/63/UE) and Spanish (RD 53/2013) regulations about the use and care of experimental animals.

Infusion of Human IgG and FcRn-ab to Pregnant Mice

Pooled IgG (800 µg) from patients or controls was injected via tail vein to pregnant mice on days 14, 15, and 16 (E14, E15, and E16) of gestation (Figure 1), as previously reported.¹⁶ Mouse monoclonal FcRn-ab (3 mg/kg, GTX14550, GeneTex, Irvine, CA) or sterile saline solution was injected via tail vein on the same days of injection E14, E15, and E16, 6 hours before the human IgG administration to potentially block FcRn-mediated transplacental transfer of IgG. Three experimental groups were established: pregnant mice injected with controls' IgG, or injected with patients' IgG; or with patients' IgG and treated with FcRn-ab. These experiments were planned according to the window of time in which the FcRn is expressed in placental tissue, and the immature fetal blood-brain barrier (BBB) does not restrict the crossing of IgG (e.g., around gestational day 16, the BBB becomes significantly more restrictive to maternal antibody penetration into the fetal brain) as shown by us and others.^16,24 FcRn-ab administration schedule was based on the findings from a passive transfer rat model of autoimmune myasthenia gravis that used the same monoclonal antibody.¹⁹ The dosage was adapted from the 5 mg/kg intraperitoneally used in the FNIT mouse model²⁰ to a lower dose of 3 mg/kg since FcRn-ab administration route was changed to IV to avoid possible fetal damage.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Experimental Design

IV injection of FcRn-ab (or saline solution) was performed on gestational days E14, E15, and E16; 6 hours before the administration of controls' or patients' IgG on these days. Fetal brain samples were collected on E17 for immunohistochemical studies of NMDAR cluster density and cortical plate thickness. Electrophysiology to determine hippocampal long-term potentiation was performed on PD 21. Newborns were assessed daily for neurodevelopmental milestones of innate reflexes from birth until day 12 on alternate days (surface righting [PDs 1, 3, 5, 7, 9, and 11] and negative geotaxis [days 2, 4, 6, 8, 10, and 12]). After breastfeeding withdrawal (day 21), mice underwent a simplified battery of behavioral tests at age 1 month including novel object location task and locomotor activity measurement. Gestational period is marked in pink and postnatal period in green. FcRn-ab = FcRn antibody; NMDAR = NMDA receptor; PD = postnatal day.

Processing of Brain and Blood Samples

In subsets of mice, brain and blood from fetuses and blood from pregnant mice were collected on day 17 of gestation (E17, figure 1). Fetal brains were fixed with 4% paraformaldehyde for 1 hour and cryopreserved in 40% glucose for 48 hours. Then, brains were embedded in optimal cutting temperature compound and snapped frozen in isopentane chilled with liquid nitrogen. Blood samples were centrifuged to obtain serum.

Antibody Determination in Blood From Pregnant Mice

The presence of human NMDAR antibodies in blood from pregnant mice was demonstrated by CBA with fixed HEK293T cells expressing GluN1/2B subunits of the NMDAR, as reported.¹ Briefly, transfected HEK cells were blocked and coincubated with the collected serum sample (1:10) and with a commercial GluN1 monoclonal mouse antibody (MAB363, 1:20.000, Sigma-Aldrich, St. Louis, MO) at 4°C overnight. Then, cells were washed and immunolabeled with Alexa Fluor 488 goat anti-human IgG (A11013) and Alexa Fluor 594 goat anti-mouse IgG (A11005; both diluted 1:1000, from Thermo Fisher, Waltham, MA) as secondary antibodies for 1 hour at room temperature (RT). After washing, coverslips were mounted and scanned under a Zeiss LSM 710 confocal microscope (Carl Zeiss GmbH, Jena, Germany).

Determination of Human IgG, NMDAR Clusters, and Cortical Plate Thickness in Fetal Brains

To determine whether FcRn blockade prevented human NMDAR antibodies injected to pregnant mice from reaching the brain of fetuses, E17 brain tissue sections were blocked with 1% bovine serum albumin for 1 hour at RT and then incubated with the same Alexa Fluor 488 goat anti-human IgG as above diluted at 1:1000 for 2 hours at RT. The fluorescence intensity in the developing hippocampus was quantified using Imaris suite v.8.1 software (Oxford Instruments, Belfast, United Kingdom). Mean intensity of IgG immunostaining in control animals was defined as 100%.¹⁶

To determine whether treatment of pregnant mice with FcRn-ab abrogated the effects of patients' antibodies on the number of clusters of NMDAR and postsynaptic density protein 95 (PSD95) in fetuses, nonpermeabilized 5-μm-thick brain sections were blocked as above and serially incubated with a human CSF NMDAR antibody sample (1:20, used as primary antibody) for 2 hours at RT and the secondary Alexa Fluor 488 goat anti-human IgG (1:1000, A11013, Thermo Fisher) for 1 hour at RT. Tissue sections were then permeabilized with 0.3% Triton X-100 for 10 minutes at RT and serially incubated with rabbit polyclonal anti-PSD95 (1:250, ab18258 Abcam, Cambridge, United Kingdom) overnight at 4°C and the corresponding secondary Alexa Fluor 594 goat anti-rabbit IgG (1:1000, A-11012, Thermo Fisher) for 1 hour at RT. Slides were then mounted with ProLong Gold antifade reagent, containing 6-diamidino-2-phenylindole dihydrochloride (DAPI, P36935; Thermo Fisher) and results scanned with the Zeiss LSM710 confocal microscope (Carl Zeiss) with EC-Plan NEOFLUAR CS 100×/1.3 NA oil objective. Standardized z-stacks including 25 optical images were acquired from 5 different areas of the developing hippocampus. Images were then deconvolved using theoretical point spread functions of Huygens Professional 17.04 software (Scientific Volume Imaging, Hilversum, NL). The median density of clusters of NMDAR or PSD95 was obtained using a spot detection algorithm from Imaris 8.1 software (Oxford instruments, Belfast, United Kingdom) and the cluster density expressed as spots/μm.³ Three-dimensional colocalization of clusters (e.g., NMDAR and PSD95) was performed using a spot colocalization algorithm implemented in Imaris. Synaptic localization was defined as colocalization of NMDAR with PSD95 and expressed as colocalized spots/µm.³ For each experimental group, the median cluster density of total cell surface and synaptic NMDAR and PSD95 were normalized to that of brains of fetuses exposed to controls' IgG (100%). The effects of FcRn-ab treatment on the thickness of the cortical plate were examined in DAPI-stained E17 fetal sagittal brain sections, acquired by confocal microscopy, and quantified using Fiji ImageJ software (fiji.sc/Fiji) as previously reported.^12,16

Field Potential Recordings and Analysis of Hippocampal LTP and Paired-Pulse Facilitation

Acute sections of the hippocampus obtained on PDs 19–23 were used to determine long-term plasticity by the classical paradigm of theta-burst stimulation at the Schaffer collateral and field potential recording at the CA1 dendritic area. We recently reported this technique applied to the current model of placental transfer of NMDAR antibodies.^16,23

Neurobehavioral Assessment in Postnatal and Early Adult Stages

From birth to PD 12, newborn pups were assessed daily for achievement of innate reflexes on alternate days, that is, surface body righting on PDs 1, 3, 5, 7, 9, and 11; and negative geotaxis on PDs 2, 4, 6, 8, 10, and 12; from a simplified Fox battery for developmental milestones previously reported.²⁵

After breastfeeding withdrawal at PD 21, subsets of mice underwent a battery of behavioral tests consisting of novel object location (NOL) test (PD 35) and locomotor activity assessment (PD 36) around age 1 month. Selection and timing of these tasks were based on findings from our previously reported animal model of placental transfer of NMDAR antibodies.¹⁶ All experiments were performed by researchers blinded to animals' experimental conditions, as reported.^13,16

Statistical Analysis

Comparison of human IgG intensity in the brain of fetuses among the 3 experimental groups was performed with 1-way ANOVA with Tukey correction for multiple comparisons. Comparison of confocal densities of NMDAR and PSD95 clusters and cortical plate thickness among the 3 experimental groups was analyzed with the Welch ANOVA test and Dunnett T3 corrections for multiple comparisons of small n groups (lower than 50) comparing normally distributed nonhomoscedastic populations. Comparison of field excitatory postsynaptic potential (fEPSP) slope recordings in mice born to mothers from the indicated experimental groups was assessed by the Kruskal-Wallis test comparing ranks (as populations were not normally distributed according to D'Agostino-Pearson test) with Dunn corrections for multiple comparisons. For behavioral paradigms, longitudinal analyses were performed by generalized estimated equations (GEEs) using an AR(1) matrix to account for intraindividual correlations. All models include litter size, group, and interaction group by time as fixed factors. Pairwise comparison in interaction factor was used to analyze differences between group time by time, showed as estimated means and their 95% CIs. Mean values of fEPSP slope change recordings and of times for developmental reflexes were presented with standard error of the mean (±SEM). All experiments were assessed for outliers with Robust regression and Outlier removal (ROUT) method applying Q = 1%. In all statistical analyses, we used a 2-sided type I error of 5%. All tests were performed using GraphPad Prism (Version 8, San Diego, CA) or SPSS (Version 25, IBM Corp., Armonk, NY) for GEE models.

Data Availability

Data supporting these findings are available on reasonable request.