Cell type-specific translational profiling in the Xenopus laevis retina.
Translating Ribosome Affinity Purification (TRAP), a method recently developed to generate cell type-specific translational profiles, relies on creating transgenic lines of animals in which a tagged ribosomal protein is placed under regulatory control of a cell type-specific promoter. An antibody is then used to affinity purify the tagged ribosomes so that cell type-specific mRNAs can be isolated from whole tissue lysates. Here, cell type-specific transgenic lines were generated to enable TRAP studies for retinal ganglion cells and rod photoreceptors in the Xenopus laevis retina. Using real time quantitative PCR for assessing expression levels of cell type-specific mRNAs, the TRAP method was shown to selectively isolate mRNAs expressed in the targeted cell and was efficient at purifying mRNAs expressed at both high and low levels. Statistical measures used to distinguish cell type-specific RNAs from low level background and non-specific RNAs showed TRAP to be highly effective in Xenopus. TRAP can be used to purify mRNAs expressed in rod photoreceptors and retinal ganglion cells in X. laevis. The generated transgenic lines will enable numerous studies into the development, disease, and injury of the X. laevis retina.
PubMed ID: 23074098
PMC ID: PMC4536902
Article link: Dev Dyn.
Grant support: 1R01EY019960 NEI NIH HHS , R01 EY019960 NEI NIH HHS
Genes referenced: barhl2 crx fabp7 gnat1 notch1 pde6a rbpms rbpms2 rho rpe rpl10a
Article Images: [+] show captions
|Figure 1. X. laevis lines have specific expression of EGFP-Rpl10a in rods or RGCs. Representative images show GFP fluorescence of the EGFP-rpl10a transgene in retinal sections of stage 57 F1 progeny (green; left column) from lines expressing the EGFP-rpl10a cDNA under the control of zebrafish islet2b (isl2b), X. laevis rhodopsin (rho), or X. tropicalis fatty acid binding protein 7 (fabp7) upstream regulatory sequences. Immunostaining using cell type–specific antibodies (red; second column) specific for RGCs (Sncg), rods (Rho), and Müller cells (Fabp7) confirms cell type-specificity of the Tg(rho:EGFP-rpl10a) and Tg(isl2b:EGFP-rpl10a) lines but not Tg(fabp7:EGFP-rpl10a) line. DAPI shows the nuclear cell layers (blue; 3rd column) while a composite image of the GFP fluorescence, antibody, DAPI and brightfield identifies any overlap between endogenous GFP fluorescence and immunostaining (Merge + Phase; 4th column). Arrowheads point to a presumed displaced RGC. Arrows point to Müller cell processes. RPE, retinal pigment epithelium; ONL/INL, outer and inner nuclear layers; IPL, inner plexiform layer; GCL, ganglion cell layer. Scale bar = 20 μm.Download figure to PowerPoint|
|Figure 2. EGFP-Rpl10a protein localizes properly and can be used to selectively enrich for RNAs. A: Western analysis of whole eye lysates from transgenic tadpoles expressing cytoplasmic GFP (GFP-cyto) and EGFP-Rpl10a, both under control of the same islet2b regulatory sequence, and from wildtype control tadpoles (Wt) show transgene protein products of the expected size. B: Representative image shows a 2D projection of a series of optical sections (Z-stack) showing nucleolar and inner segment localization of the EGFP-Rpl10a protein in rods. C: Representative image shows a 2D projection of a Z-stack showing nucleolar and cytoplasmic localization of the transgenic EGFP-Rpl10a protein in RGCs. Arrowheads in B and C point to nucleoli. D: RNA recovery from unconjugated beads (−) or EGFP-coated beads (+) using lysates of Tg(isl2b:EGFP-Rpl10a) (L10a), Tg(isl2b:GFP) (GFP-cyto), and from non-transgenic wildtype (Wt) retinas, show significant RNA recovery only in Tg(isl2b:EGFP-Rpl10a) lysates in the presence of the EGFP antibody. Grey horizontal line shows the 0.5-ng limit of detection of the bioanalyzer. E: The residual unbound RNA fraction from the samples shown in D was comparable in all samples. Data in D and E represent the mean RNA levels from 12 retinas averaged for three replicates, and statistical significance between individual samples was established by pair-wise ANOVA comparisons (***P < 0.001). F: Representative superimposed electropherogram traces from TRAP EGFP-rpl10a transgenic samples affinity purified with EGFP conjugated beads (red trace) and unconjugated beads (black trace) show high abundance 18S and 28S ribosomal RNAs and lower abundance mRNAs only in EGFP-rpl10a samples extracted using EGFP- conjugated beads. Right insets show corresponding RNA gels for each of the traces. Scale bars = 2 μm.Download figure to PowerPoint|
|Figure 3. In situ hybridization shows expression patterns of a subset of the genes used to assess cell-type enrichment by TRAP. Representative images of retina sections from wildtype stage 57 tadpoles show mRNA expression patterns for genes selectively expressed in RGCs, rods, and Müller cells (rbpms, gnat1, rlpb1, respectively) and a gene expressed widely in multiple cell types (gapdh). A composite image of the in situs (shown in top row), DAPI and brightfield is used to identify cell layers (Merge DAPI + Phase; 2nd row). RPE, retinal pigment epithelium; ONL/INL, outer and inner nuclear layers; IPL, inner plexiform layer; GCL, ganglion cell layer. Scale bar = 20 μm.Download figure to PowerPoint|
|Figure 4. Lines expressing EGFP-rpl10a in rods and RGCs enable efficient enrichment of cell type–specific mRNAs. Graphs show absolute expression levels of mRNA for affinity-purified TRAPed mRNAs from transgenic lines (blue, Tg(rho:EGFP-Rpl10a); red, Tg(isl2b:EGFP-Rpl10a); black, Tg(fabp7:EGFP-Rpl10a)) and from the unbound fraction of wildtype control eyes (grey, Wildtype). To determine the gene expression copy number, a standard curve was generated for each of the genes tested by qPCR (Supp. Fig. S2). Due to large differences in absolute gene expression values, genes expressed at low levels require a different y-axis scale (see inset bar graphs for barhl2 [A], pde6a and crx [B], and notch1 [C]). Asterisks reflect the level of significance for individual pair-wise comparisons between the marked genotype and either one other genotype (bracket) or all other genotypes (Tukey test; **P < 0.005; ***P < 0.001).Download figure to PowerPoint|
|Figure 5. The IP/Total ratio and Specificity Index (SI) provide independent measures to identify genes enriched above background levels in different cell types. A,B: Data for each of three biological replicates are plotted as the ratio of the absolute expression level for the IP for Tg(isl2b:EGFP-Rpl10a) (A) and Tg(rho:EGFP-Rpl10a) (B) TRAPed mRNAs over the absolute expression level for RNAs purified from total unbound fraction from a reference wildtype sample (Total) for each gene. Standard error of the mean (error bars) was calculated based on the standard deviation for a ratio as described in the Experimental Procedures section. C: Scatter plot of the SI for all cell types plotted against individual genes shows a clustering of cell type–specific enrichment for RGCs (red highlighted cluster) and rod phototoreceptor cells (blue highlighted cluster). Genotypes on graph are as indicated isl2b = Tg(isl2b:EGFP-rpl10a); rho = Tg(rho:EGFP-rpl10a); fabp7 = Tg(fabp7:EGFP-rpl10a). P values are significant at an SI less than 4 (dashed line, P < 0.1).Download figure to PowerPoint|
|Figure 6. TRAP provides cell-type enrichment for genes of different absolute expression levels and reproducibility across different transgenic lines in the X. laevis retina. A: PCA of individual samples showing qPCR-based expression levels from TRAPed replicates from the same GFP-rpl10a frog lines as well as from different lines. B: PCA of the contribution of each of the 16 analyzed genes on the overall variance. The placement and length of the vectors show the contribution to the variance for each gene. The gene expression data have been standardized to a value of 1, demarked by the tan circle (see Experimental Procedures section).Download figure to PowerPoint|