Surprises from modulating GNAT1 and GNAT2 expression by CRISPR/Cas9-mediated gene targeting

Abstract

Purpose : Phototransduction is thought to depend on GNAT1 and GNAT2 in rod and cone, respectively. Here we seek to down-regulate or eliminate the expression of GNAT1 or GNAT2 to see if and how rod and cone phototransduction is compromised and whether elimination of both genes result in complete loss of image-forming vision in mice.

Methods : We used CRISPR/Cas9 mediated genome editing technique to modifiy the immediate upstream Nrl site in the Gnat1 locus and to delete the coding region of the Gnat2 gene. We also generated rabbit antibodies against GNAT1 or GNAT2 and performed biochemical, electroretinographical and behavioral characterizsations of all mutant mice including the spontaneous Gnat2^cpfl3/cpfl3 mice. Protein expression level is determined by immunohistochemistry and immunoblotting. Behavioral assessment of image-forming vision is done by scoring mouse optokinetic reflex responses at both photopic and scotopic conditions.

Results : We have obtained several mouse lines with various deletions in and around the Nrl site in the Gnat1 gene. Deletion of the central four base pairs of the Nrl site is sufficient to lower GNAT1 level to below detection by immunoblotting. Deletion of six base pairs immediately downstream of the Nrl1 site has no effect on GNAT1 expression level. Interestingly, deletion of the first 9 base pair of the Nrl site in the Tux22 mouse line results in downregulation but not complete elimination of GNAT1. Using the UUTA2bcm antibody against GNAT2, we showed that the spontaneous mutant mouse, Gnat2^cpfl3/cpfl3, is not a null but a hypomorph. This is confirmed by a small but readily detectable photopic ERG b-wave response in Gnat2^cpfl3/cpfl3 mice which also lack GNAT1. GNAT2 expression is not detectable in mice where the Gnat2 structure gene is deleted. However, we can still record a small photopic ERG b-wave response from them.

Conclusions : We have confirmed the requirement of NRL for GNAT1 expression and further refined the critical sequences within the Nrl site required for this regulation. Our results indicate that the use of Gnat2^cpfl3/cpfl3 mice as a strain where cone function is lost is inappropriate and should therefore be avoided in the future and all previous findings using this strain be revisited. Finally, the source of the small photopic ERG b-wave detected in the GNAT2 null mice needs to be identified should the mouse be used as a replacement for the Gnat2^cpfl3/cpfl3 strain.

This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.