Fig. 3

Cone-specific Cre-LoxP recombination activity in Arr3^T2ACreERT2 mouse retina. A Noninvasive in vivo fundus short-wavelength autofluorescence (SW-AF) showed a homogenous background in the C57BL/6J mouse, increased intensity, and a “mosaic” pattern of SW-AF in male hemizygous Arr3^T2ACreERT2Ai14D^+/- and female heterozygous Arr3^T2ACreERT2/+Ai14D^+/- mice. There was no difference in the intensity or mosaicism of SW-AF images when tamoxifen induction was delivered intragastrically at postnatal day (PD) 4–6 or intraperitoneally at PD21-24. B Immunohistochemistry (IHC) staining on retinal whole mounts and crysections from male hemizygous Arr3^T2ACreERT2Ai14D^+/- mice showed distinct tdTomato signal in cone cells (red) that did not co-label with anti-ARR3 antibody staining (green) (1st and 2nd rows; Additional file 1: Movie S1). When the same sections were labeled for another cone cell marker, glycogen phosphorylase (GlyPh), we found that 40% of GlyPh-positive cones also expressed tdTomato (bottom; Additional file 1: Movie S2), suggesting that the Cre-LoxP recombination efficiency was 40% and not every cone cell in hemizygous Arr3^T2ACreERT2Ai14D^+/-± mice expresses tdTomato. C–D IHC staining for the anti-ARR3 antibody on a retinal whole mount and cryosections from a female heterozygous Arr3^T2ACreERT2/+Ai14D^+/- mouse showed a mosaic pattern of green (ARR3-positive cone cells) and red (tdTomato expressing cone cells) in retinal whole mounts and cryosections (1st and 2nd rows; Additional file 1: Movies S3, S4). IHC staining for anti-GlyPh and PNA antibodies on retinal cryosections from a female heterozygous Arr3^T2ACreERT2/+Ai14D^+/- mouse showed around 50% of GlyPh-positive or PNA-positive cone cells express tdTomato (3^rd and 4^th rows; Additional file 1: Movie S5). PD postnatal day, Tam tamoxifen, OPL outer plexiform layer, ONL outer nuclear layer, IS inner segment, Ai14D: B6.Cg-Gt(ROSA)26Sor^{tm14(CAG−tdTomato)Hze}/J (JAX #007914). Scale bars, 200 μm in A, 20 μm in B and D, and 1000 μm in C