An integrative study of the microbiome gut-brain-axis and hippocampal inflammation in psychosis: Persistent effects from mode of birth

Abstract

The mechanism producing psychosis appears to include hippocampal inflammation, which could be associated with the microbiome-gut-brain-axis (MGBS). To test this hypothesis we are conducting a multidisciplinary study, herein described. The procedures are illustrated with testing of a single subject and group level information on the impact of C-section birth are presented.

Method: Study subjects undergo research diagnostic interviews and symptom assessments to be categorized into one of 3 study groups: psychosis, nonpsychotic affective disorder or healthy control. Hippocampal volume and metabolite concentrations are assessed using 3-dimensional, multi-voxel H1 Magnetic Resonance Imaging (MRSI) encompassing all gray matter in the entire hippocampal volume. Rich self-report information is obtained with the PROMIS interview, which was developed by the NIH Commons for research in chronic conditions. Early trauma is assessed and cognition is quantitated using the MATRICS. The method also includes the most comprehensive autonomic nervous system (ANS) battery used to date in psychiatric research. Stool and oral samples are obtained for microbiome assessments and cytokines and other substances are measured in blood samples.

Results: Group level preliminary data shows that C-section birth is associated with higher concentrations of GLX, a glutamate related hippocampal neurotransmitter in psychotic cases, worse symptoms in affective disorder cases and smaller hippocampal volume in controls.

Conclusion: Mode of birth appears to have persistent influences through adulthood. The methodology described for this study will define pathways through which the MGBA may influence the risk for psychiatric disorders.

Keywords: Autonomic nervous system; Inflammation; Magnetic resonance spectroscopy; Microbiome; Psychosis; Schizophrenia.

Fig. 1

Analysis of the gut and oral microbiome in the case example, with respect to all others in the study, highlighting the clear individual variability in microbiome composition which may play an important role in disease pathogenesis. (A) Increased diversity in the case report in the oral microbiome but not in gut microbiome. (B) Major differences in relative taxa abundances between the case report and the whole sample, again showing more pronounced differences in the oral microbiome as compared to the gut microbiome. (C) Principal coordinate analysis (PCoA) of beta diversity. The individual patient demonstrates a distinct and unique oral microbiome.

Fig. 2.

Top, left, a: T₁-weighted MP-RAGE MRI of an axial slice from the 38 year old female patient with schizophrenia, on the level of the bilateral hippocampi (outlined in yellow, and traced manually by the neuroradiologist according to Frizoni et al. (16,17), as described in the Methods IV) and superimposed with the size and position of the 9×6 cm² LR×AP ¹H-MRSI VOI (orange frame). Note the left hippocampus trace is not continuous, reflecting the subjects head-tilt which prevented capturing both structures on the same slice (see GM on b) and reflecting another advantage of 3D multi-slice ¹H-MRSI acquisition. Top, right and bottom, b-c: Gray, white matter and cerebrospinal fluid (GM, WM, CSF) masks generated by the SPM12 software from the T₁-wighted MP-RAGE (a), and used by our software to estimate their tissue-fractions within the bilateral hippocampi masks, e.g., outlined on a. The ¹H-MRSI VOI size and location is outlined on the MRI on each tissue-type mask (orange frame). Note the mostly GM nature of the medial temporal lobe (hippocampi) on b and small WM or CSF involvement.

Fig. 3.

Top, a, b, c: Sagittal, coronal and axial T₁-weighted MRI from the 38 year old female patient with schizophrenia, superimposed with the 9×6×2 cm³ (LR×AP×IS) VOI, 12×12 cm² axial CSI FOV and grid (solid and dashed lines on a-c and gray grid on c) 1×1 cm³ voxels and bilateral hippocampi masks (transparent yellow on c, from Fig. 1). Yellow horizontal arrows on a, b indicate the level of c and yellow vertical arrow on c, the level of a. Bottom, left, d: Real part of the 9×6 axial (LR×AP) ¹H spectra matrix from the VOI slice on c, all on a common frequency (ppm) and intensity scales. The spectra in the tallow frames are those from the highlighted voxels on c and represent the bilateral hippocampi on this slice. Bottom, right, e, the spectra in the right hippocampus in c and d, expanded for greater detail (black lines), superimposed with their spectral fits (thick gray lines). Note: (i) the spectral SNR; (ii) spectral resolution (8.1±3.0 Hz linewidth) in these 1.0 cm³ voxels; (iii) the fidelity of the spectral fit; and (iv) relatively elevated Cho in the expanded hippocampal spectra (marked with red arrows on e) as well as in the contralateral hippocampus on d.