RNA editing of the serotonin 2C receptor and expression of Galpha(q) protein: genetic mouse models do not support a role for regulation or compensation

Abstract

The serotonin 2C (5-HT(2C)) receptor undergoes RNA editing at five bases in a region of the pre-mRNA encoding the second intracellular loop, generating many unique 5-HT(2C) receptor isoforms. Mechanisms regulating in vivo expression of different edited 5-HT(2C) receptor isoforms are poorly understood, as are the adaptive consequences of variation in editing profiles. Recent findings suggest a putative relationship between expression levels of Galpha(q/11) protein and the degree of editing of 5-HT(2C) receptor transcripts. To elucidate the potential regulatory or adaptive role of Galpha(q/11) protein levels, we quantified editing of 5-HT(2C) receptor RNA transcripts in Galpha(q) null mice and protein levels of Galpha(q) and Galpha(11) in transgenic male mice solely expressing either the non-edited (INI) or the fully edited (VGV) isoforms of the 5-HT(2C) receptor. Pyrosequencing of RNA isolated from amygdaloid cortex in Galpha(q) null and wild-type mice revealed no significant differences in 5-HT(2C) receptor mRNA editing profiles. Cortical tissue from INI/y, VGV/y, and wild-type mice was assayed for expression of Galpha(q) and Galpha(11) subunits by Western blotting. No differences in signal density between wild-type and INI/y or VGV/y groups were found, indicating equivalent levels of Galpha(q) and Galpha(11) protein. Together, these data do not support a causal or compensatory relationship between 5-HT(2C) receptor RNA editing and G(q) protein levels.

Figure 1

Dot plot of the mean percentage editing in Gαq null (−/−) and Gαq WT (+/+) mice for individual edit sites (top), RNA isoforms (middle), and predicted protein isoforms (bottom). This figure conveys at a glance the overall similarity in editing profiles between genotypes. Means and standard deviations for each genotype are tabulated in Table 2–Table 4, along with statistical test results. The figure also reveals that only 6 of 22 observed RNA isoforms were above 1% in frequency. In other words, relatively few isoforms captured most of the observed editing events.

Figure 2

(A) Western blots detecting Gα₁₁ at ~45 kDa (top) and Gα_q at 45 kDa (bottom) and α-tubulin at 50 kDa (loading control) in cortical tissue of male Gα_q null mice. Notice that Gα_q protein was undetectable in null mice, yet Gα₁₁ protein expression was robust. (B) Western blots detecting Gα₁₁ (top) and Gα_q (bottom) and α-tubulin (loading control) in cortical tissue of WT mice and mice expressing only the 5-HT_2C-INI receptor isoform. (C) Western blots detecting Gα₁₁ (top) and Gα_q (bottom) and α-tubulin (loading control) in cortical tissue from WT mice and mice expressing only the 5-HT_2C-VGV receptor isoform.

Figure 3

(A) The mean optical density ratios of Gα₁₁ and Gα_q to α-tubulin from Western blots of VGV/y cortical tissue are compared to the mean optical density ratios from Western blots in cortical tissue from wild-type mice. (B) The mean optical density ratios of Gα₁₁ and Gα_q to α-tubulin from Western blots of INI/y cortical tissue are compared to the mean optical density ratios from Western blots in cortical tissue from wild-type mice. Unpaired t-tests revealed no significant differences in the levels of Gα₁₁ or Gα_q in cortex of VGV/y (p = 0.94 and p = 0.72, respectively) or INI/y (p = 0.77 and p = 0.56, respectively) mice compared to their respective wild-type litter mates.