Cholinergic Stress Signals Accompany MicroRNA-Associated Stereotypic Behavior and Glutamatergic Neuromodulation in the Prefrontal Cortex

Abstract

Stereotypic behavior (SB) is common in emotional stress-involved psychiatric disorders and is often attributed to glutamatergic impairments, but the underlying molecular mechanisms are unknown. Given the neuro-modulatory role of acetylcholine, we sought behavioral-transcriptomic links in SB using TgR transgenic mice with impaired cholinergic transmission due to over-expression of the stress-inducible soluble 'readthrough' acetylcholinesterase-R splice variant AChE-R. TgR mice showed impaired organization of behavior, performance errors in a serial maze test, escape-like locomotion, intensified reaction to pilocarpine and reduced rearing in unfamiliar situations. Small-RNA sequencing revealed 36 differentially expressed (DE) microRNAs in TgR mice hippocampi, 8 of which target more than 5 cholinergic transcripts. Moreover, compared to FVB/N mice, TgR prefrontal cortices displayed individually variable changes in over 400 DE mRNA transcripts, primarily acetylcholine and glutamate-related. Furthermore, TgR brains presented c-fos over-expression in motor behavior-regulating brain regions and immune-labeled AChE-R excess in the basal ganglia, limbic brain nuclei and the brain stem, indicating a link with the observed behavioral phenotypes. Our findings demonstrate association of stress-induced SB to previously unknown microRNA-mediated perturbations of cholinergic/glutamatergic networks and underscore new therapeutic strategies for correcting stereotypic behaviors.

Keywords: acetylcholinesterase; cholinergic; glutamatergic; hippocampus; microRNAs; motor control; prefrontal cortex; stereotypic behavior.

Figure 1

TgR mice show SB profiles. (A) Forced swim test. Shown are wall swim (number of swims around the circumference of the circular pool), line/crossing swim (number of episodes in which the mouse made a straight-line path instead of adhering to the wall), body axis circling (lateral circling in the water around the body axis) and asymmetry % (index of asymmetry in direction of swimming; 0 = perfect symmetry, 50% = complete asymmetry). Sample sizes: FVB/N (n = 10), TgR (n = 14). (B–E) Open field tests. Shown are locomotion time, episodes, rearing events and the index of asymmetry. (F,G) Circadian light: dark rhythm and body temperatures in TgR and FVB/N mice. Sample sizes: FVB/N (n = 8), TgR (n = 6). Asterisks represent significant ANOVA: * 1, * 2, * 3, in the first, second, and third quarters respectively.

Figure 2

Pilocarpine suppresses the motor hyperactivity but not the locomotor asymmetry component of SB. Shown are the effects on motor activity in the open field. Pilocarpine, a muscarinic agonist, at a dose of 25 mg/kg, suppressed motor activity without affecting motor asymmetry. * 1 marks a significant transgene effect. * 2 marks a significant drug effect.

Figure 3

(A) Simplified structure of the human ACHE gene with its synaptic (AChE-S) and “readthrough” (AChE-R) mRNA 3′ alternative splicing products. (B,C). Quantitative PCR shows selective increases of the stress-induced AChE-R variant in the cholinoceptive hippocampi of TgR mice, and more modest increases of both AChE-S and AChE-R in the cholinergic prefrontal cortex. Results were normalized to RPL19. (D–F). AChE-R overexpression in matrix cells positive for the common AChE domain, but not in patches. No hAChE-R labeling appeared in cholinergic nuclei with intense choline-acetyltransferase (ChAT) staining in the medial septum (MS) (G), diagonal band (DB) (H), and the prepontine tegmental nucleus (PPTG) (I). hAChE-R labeling co-appeared in cholinergic nuclei expressing ChAT, with moderate to weak staining, e.g., in the nucleus basalis magnocellularis (NBM) (J), striatum (ST) (K) and spinal cord (ventral horn) (L). Red = ChAT, Green = hAChE-R, Yellow–Orange = co-localization of ChAT and AChE-R.

Figure 4

AChE-R shows hierarchical distribution in brain regions controlling motor behavior. (A) In FVB/N mice, subsets of CA1 neurons displayed higher than background expression of mouse AChE-R mRNA. (B) Neuronal activation (c-fos) patterns show inversed association to AChE-R. In the CA1 region of TgR mice, similar neuronal subsets expressed higher than background levels of the human AChE-R (hAChE-R) protein. (C) Using similar procedures, TgR mice presented lower basal level of neuronal activation reflected as c-fos staining in the dorsolateral striatum (LST), medial striatum (MST), anterior cingulate cortex (AC), lateral septum (LS) but not in hippocampal sub-regions CA1-2, CA3, and dentate gyrus (DG).

Figure 5

A. DE miRNAs in the TgR Hippocampus. Enriched DE miRNAs out of global miRNAs in the TgR hippocampus (minimal average expression of 300 normalized counts). Eight miRNAs targeting at least 5 cholinergic genes each and which are also known to be expressed in humans were DE in the TgR hippocampus, out of a total of 36 DE miRNAs. B. Network analysis for down- and up-regulated DE cholinergic-targeting miRNAs in the TgR hippocampus. Analysis was done using miRwalk 3.0 using maximum binding p-value of 1 and filter to 3’UTR binding and using miRDB database [73].

Figure 6

Microarray-validated individual diversity in PFC’s cholinergic-related transcripts compared to the CP. Shown are variabilities between the CP and the PFC in FVB/N mice and the TgR strain. Higher individual variability can be seen in the expression profiles between the PFC and the CP, especially in the SB-prone TgR strain.

Figure 7

Glutamatergic transcripts are conspicuously changed in the TgR PFC. (A) Log10 Scatter plots demonstrating greater differences between microarray pools of 4 TgR mice PFCs (R1 R2) as compared with pools of 4 FVB/N mice PFCs (C1 C2), and yet greater differences between cross pools (R1 C2, R2 C1, R1 C1, R2 C2). Gray dots depict genes the expression level of which was unaltered between the two tests; black dots depict genes with log ratio higher than 1. (B) GOC pathways terms modified in TgR mice relative to parent strain mice. Terms for the biological process and molecular function are shown at the top and bottom, respectively. The lines indicate the hierarchical relationship between the terms. Values indicate the probability to observe the given number of changed transcripts within a term by chance. PFC = prefrontal cortex; GOC = gene ontology categories; pr = process; comm. = communication; act = activity.

Figure 8

qPCR-validated change in cholinergic signaling related genes. Most changes between FVB/N mice and the TgR strain occur in the hippocampus. Different expression profiles can be observed between the PFC and hippocampus.