Identification of novel mouse and rat CB1R isoforms and in silico modeling of human CB1R for peripheral cannabinoid therapeutics

Abstract

Targeting peripheral CB1R is desirable for the treatment of metabolic syndromes without adverse neuropsychiatric effects. We previously reported a human hCB1b isoform that is selectively enriched in pancreatic beta-cells and hepatocytes, providing a potential peripheral therapeutic hCB1R target. It is unknown whether there are peripherally enriched mouse and rat CB1R (mCB1 and rCB1, respectively) isoforms. In this study, we found no evidence of peripherally enriched rodent CB1 isoforms; however, some mCB1R isoforms are absent in peripheral tissues. We show that the mouse Cnr1 gene contains six exons that are transcribed from a single promoter. We found that mCB1A is a spliced variant of extended exon 1 and protein-coding exon 6; mCB1B is a novel spliced variant containing unspliced exon 1, intron 1, and exon 2, which is then spliced to exon 6; and mCB1C is a spliced variant including all 6 exons. Using RNAscope in situ hybridization, we show that the isoforms mCB1A and mCB1B are expressed at a cellular level and colocalized in GABAergic neurons in the hippocampus and cortex. RT-qPCR reveals that mCB1A and mCB1B are enriched in the brain, while mCB1B is not expressed in the pancreas or the liver. Rat rCB1R isoforms are differentially expressed in primary cultured neurons, astrocytes, and microglia. We also investigated modulation of Cnr1 expression by insulin in vivo and carried out in silico modeling of CB1R with JD5037, a peripherally restricted CB1R inverse agonist, using the published crystal structure of hCB1R. The results provide models for future CB1R peripheral targeting.

Keywords: alternative splicing; cannabinoids; cell- and tissue-based therapy; gene expression.

Fig. 1

a Mouse Cnr1 (4A5) and b rat Cnr1 gene (5q21) structure and transcript variants. The open boxes represent the exons, with the exon number marked inside the box, and horizontal lines represent introns. A fused exon–intron is shown by an open box with a middle line. The intra-exonal introns are shown by smaller shaded boxes, and arrows represent intra-exonal splicing sites. The alternatively spliced transcript variants are shown under the genes, and GenBank accession numbers are included in the parenthesis. The TaqMan probes were designed to hybridize in junctions of spliced exons represented by thin lines under exon–exon junctions. RNAscope ZZ pair probes are marked as black bars under exon 1, retaining intron 1 and the 3′UTR of the mouse Cnr1 gene

Fig. 2

a Hypothalamus (HYP), hippocampus (HIP), cortex (CTX), cerebellum (CER), skeletal muscle (SKM), and islet (ISL) expression of mCB1A, mCB1B, and mCB1a in wild-type and beta-cell CB1R-knockout mice (strain background: C57BL/6). One-way ANOVA (n = 4, the mean ± SEM, *P < 0.05). b Expression of mCB1A, mCB1B, and mCB1a in hypothalamus (HYP), hippocampus (HIP), cortex (CTX), and cerebellum (CER) of mice treated by vehicle or S961. c Expression of mCB1A, mCB2A and mCB2B isoform in liver of wild-type and β-CB1R^−/− knockout mice fed a standard (SD) or high-fat (HF) diet, and in liver of wild-type mice treated with vehicle or S961. Unpaired Student’s t-test between control and experimental groups (n = 4, the mean ± SEM, *P < 0.05). The y-axis is fold change of mCB1R isoforms in tissues of β-CB1-KO using the relevant wild-type mouse tissue as reference (a), vehicle vs S961 (b, c), and standard diet vs high-fat diet (c)

Fig. 3

RNAscope ISH colocalization (20×) of mCB1A and mCB1B isoform-specific probes with the common CB1R probe in mouse brain. mCB1A and mCB1B probes (green) hybridize with mCB1R (red) positive cells in hippocampus (a mCB1A; b mCB1B) and cortex (c mCB1A; d mCB1B). Yellow arrows indicate the colocalization of mCB1 isoforms with the common mCB1 probe

Fig. 4

RNAscope ISH colocalization (20×) of mCB1A and mCB1B isoform-specific probes (green) with the GABAergic Gad1 (red) probe in hippocampus (a mCB1A; b mCB1b) and cortex (c mCB1A; d mCB1B). Yellow arrows indicate the colocalization of mCB1 isoforms with Gad1. Corresponding hippocampus and cortex stereotaxic coordinates are shown by black arrows (e)

Fig. 5

rCB1A, rCB2A, and rCB2B cell type expression. a rCB1A, rCB2A, and rCB2B expression in rat astrocytes, microglia, and neurons; b astrocyte marker Gfap, microglia marker Iba-1, and neuron marker NeuN expression in rat astrocytes, microglia, and neurons (n = 3, mean ± SEM). The y-axis is fold change of rCB1A, rCB2A, rCB2B, NeuN, Gfap, and Iba-1 gene expression in different primary cell types using the corresponding astrocyte gene as reference

Fig. 6

Chemical structures of JD5037 (a) and 2-AG (b) and predicted interactions of hCB1R with JD5037 (c) and 2-AG (d). JD5037 and 2-AG are indicated by green, red, and blue sticks. The key residues in hCB1R that interact with JD5037 and 2-AG are indicated by light blue

Fig. 7

Alignment of human hCB1R, hCB1a, and hCB1b with mouse mCB1a and mCB1b N-terminal variants. Identical amino acids are marked by (*), conservative changes by (.), and the gaps by dashed lines. Transmembrane domain 1 (TM1) is underlined, and a potential N-linked glycosylation site is marked by italics (N)