Fig 1.
RpS12-dependent gene expression in Rp+/- wing discs.
A) 201 of 253 mRNAs altered in Rp+/- wing discs were rpS12-dependent. B) rpS12- and Xrp1-dependent Rp+/- wing disc genes were largely overlapping. Shown in parentheses are the numbers of genes with predicted Xrp1 binding motifs (See also S3 Table). C. Heat map of fold changes in mRNA levels between wing discs from wild type and from indicated genotypes. Upregulation in Rp+/- genotypes was overwhelmingly dependent on both rpS12 and Xrp1. Genes shown here include all those corresponding to the enriched GO terms mature ribosome assembly (R), sulfur compound metabolic process (S), glutathione metabolic process (G), telomere maintenance (T), and DNA repair (D). The individual fold changes are shown in the S1 Table and S2 Table. Panels D-F show GstD1-LacZ expression in wing discs. Panels G-I show Upd3.3-LacZ labeling of wing discs. Genotypes: wild type (D,G); M(3)i55 ubi-GFP FRT80B/+ (E,H); M(3)i55 ubi-GFP FRT80B/FRT82 Xrp1m2-73 (F,I).
Fig 2.
Growth and translation in Rp+/- wing discs.
All panels show wing pouch regions of third-instar wing imaginal discs. Genotypes are as indicated below each figure; the font colors correspond to the labeling of the corresponding genotype (the genotype corresponding to the most brightly labeled cells is shown in white on a black background). A,B). Xrp1 protein levels were lower in RpS3+/- rpS12D97/D97 clones than in RpS3+/- rpS12D97/+ cells. Small RpS3+/- rpS12+/+ clones were detected rarely (eg arrow) C) RpS3+/- rpS12+/+ clones are rarely detected in RpS3+/- rpS12D97/+ wing discs, unlike the reciprocal RpS3+/- rpS12D97/D97 clones. D) In contrast to panel C, in the Xrp1+/- background RpS3+/- rpS12+/+ clones were recovered similarly to the reciprocal RpS3+/- rpS12D97/D97 clones. E,F). Overall translation was reduced in RpS17+/- cells, as was described previously [16]. G,H). Overall translation was not affected in RpS17+/- cells in rpS12D97/D97 wing discs. I,J) Translation in RpS18+/- cells is reduced compared to wt cells, as was described previously [16]. K,L). Further data related to this figure is shown in the S1 Fig. Genotypes: A-C) y w hsF; rpS12D97 FRT80B/P{arm-LacZ} FRT80B M(3)95A armLacZ. D) y w hsF; rpS12D97 FRT80B/P{arm-LacZ} FRT80B Xrp1m2-73 M(3)95A. E,F) y w hsF; RpS174 P{ubi-GFP} FRT80B/FRT80B. G,H) y w hsF; rpS12D97 RpS174 P{ubi-GFP} FRT80B/rpS12D97 FRT80B. I,J) y w hsF; FRT42D P{arm-LacZ} M(2)56i/FRT42D. K,L) y w hsF; FRT42D P{arm-LacZ} M(2)56i/FRT42D; rpS12D97 FRT80B/rpS12D97 FRT80B.
Fig 3.
Contribution of RpS12 to rate of development.
All panels show the cumulative percentage of adult flies emerged according to time after egg laying in hours. A,B) A genomic transgene including the rpS12+ locus had little effect on wild type development but significantly retarded development of RpS3+/- flies. C,D) Heterozygosity for rpS12 had little effect on wild type development but modestly accelerated development of RpS18+/- flies. E,F) the genotype rpS12D97/EP3025 modestly delayed development of otherwise wild type flies. G,H) the genotype rpS12D97/D97 modestly delayed development of otherwise wild type flies and this delay was suppressed by a genomic transgene including the rpS12+ locus. I,J) An experiment in which rpS12D97/D97 and rpS12D97/+ genotypes had little effect on the development of RpS3+/- flies. K,L) An experiment in which rpS12D97/D97 and rpS12D97/+ genotypes suppressed the developmental delay of RpS3+/- flies, to a similar extent to RpS3+/- rpS12D97/D97 Xrp1+/- flies. In Fig 3K and 3L, the development of RpS3+/- rpS12D97/D97 flies and RpS3+/- rpS12D97/D97 Xrp1+/- flies are significantly faster than RpS3+/- flies (p<0.00001 in all cases, Mann Whitney procedure). The detailed genotypes used and numerical data corresponding to these graphs is tabulated in the S4 Table.
Fig 4.
Contributions of RpS12 to longevity and organ size.
A) Survival curve of female rpS12G97D flies compared to w11-18 (wild type) controls. Shown is the mean are 95% confidence limits from 3 replicates of 120 flies each. B) Survival curve of male rpS12G97D flies compared to w11-18 (wild type) controls. The difference is significant at p<0.0001 Log-rank (Mantel-Cox) test. See the S3 Fig for individual replicates and the S5 Table for raw data. C) Adult wing size was significantly smaller for the rpS12G97D females compared to w11-18 (p = 0.0165, two-tailed Welch’s t-test) and for rpS12G97D/- females compared to FRT80B (p = 0.000006, two-tailed Welch’s t-test). Wing size was also smaller for rpS12G97D/- males compared to FRT80B (p = 0.00014, two-tailed t-test) but not for rpS12G97D males compared to w11-18 (p = 0.126, two-tailed t-test). Raw data are tabulated in the S7 Table. D) Wings from rpS12G97D/- and FRT80B females overlaid to illustrate the difference. The rpS12G97D/- wing is smaller.
Fig 5.
Cross-genotype co-expression analysis.
A. Transcripts that differed significantly between control and rpS12D97/D97 wing discs showed little overlap with those affected in RpS3+/- and RpS17+/- wing discs. B) Transcripts dependent on RpS12 in both Rp+/- and Rp+/+ wing discs showed little overlap with those affected in both RpS3+/- and RpS17+/- wing discs. C) Transcripts that differed significantly between control and Xrp1+/- wing discs showed little overlap with those affected in RpS3+/- and RpS17+/- wing discs. D) Transcripts dependent on Xrp1 in both Rp+/- and Rp+/+ wing discs showed little overlap with those affected in both RpS3+/- and RpS17+/- wing discs. E) RpS12- and Xrp1-dependent transcripts showed little overlap. F) Hierarchical clustering of wing disc samples by their similarities (as Pearson correlation coefficients) in gene expression profiles distinguished samples by genotypes. The Xrp1+/- replicates grouped with wild type controls. G) Co-expressed and similarly-regulated genes (module #2) whose expression was altered in RpS3+/- and RpS17+/- wing discs in an RpS12-dependent, Xrp1-dependent manner. H) Co-expressed and similarly-regulated genes (module #8) whose expression was altered in RpS3+/- and RpS17+/- wing discs in an RpS12-independent, Xrp1-independent manner. I) Co-expressed and similarly-regulated genes (module #6) whose expression was altered in an RpS12-dependent manner regardless of RpS3 genotype. In G-I, the columns in the heatmaps are genes and the rows are samples (3 replicates each as indicated); color scale indicates relative expression across samples. J) Genes in the Module 6 that were downregulated in rpS12D97 genotypes and present in the GO term ‘secretion’. Their fold changes (from DESeq2) are shown here. Bold values were statistically significant (Padj < 0.05). Neither Module 6 genes that were upregulated in rpS12D97 genotypes, not Module 6 as a whole, was enriched for any particular GO terms.
Fig 6.
Ribosomal protein transcripts.
All panels show fold changes of Rp mRNA levels from mRNA-seq analysis of mutant wing imaginal discs in comparison to wild type controls. Loci are arranged from most reduced to most increased in each case. Transcripts of duplicated Rp genes that are expressed in at low levels because their paralog dominates in wing discs are not included. RpS9, rpS12, RACK1, RpS27A transcripts highlighted. A) Rp genes show an overall reduction in transcript levels in RpS3+/- and RpS17+/- wing discs, with RpS9, rpS12, Rack1, and RpS27A affected to a greater extent. Fold changes are the mean of RpS3+/- and RpS17+/- values except for RpS3 and RpS17 themselves; transcript levels of these genes is 50% reduced in their own mutants so only the value for the other mutant is shown here. B) Rp mRNAs in RpS3+/- Xrp1+/- wing discs resembled those in RpS3+/- wing discs. RpS3 mRNA levels are not included. C) RpS9 and Rack1 transcript levels remained reduced in rpS12D97/D97 RpS3+/- but rpS12 and RpS27A were restored to wild type or higher levels. RpS3 mRNA levels are not included. D) The only Rp whose transcript levels were affected in rpS12D97/D97 wing discs was rpS12 itself.
Fig 7.
Effects of Xrp1 and of tetracycline on rate of development.
All panels show the cumulative percentage of adult flies emerged according to time after egg laying in hours. A,B) An experiment in which rpS12D97/D97 and rpS12D97/+ genotypes did not accelerate the development of RpS3+/- flies, and in which RpS3+/- rpS12D97/D97 Xrp1+/- flies developed more slowly than RpS3+/- Xrp1+/- flies. C,D) An experiment in which rpS12D97/D97 and rpS12D97/+ genotypes did not accelerate the development of RpS17+/- flies, but in which RpS17+/- rpS12D97/D97 Xrp1+/- flies developed as rapidly as RpS17+/- Xrp1+/- flies. E,F) Tetracyclin feeding modestly retarded development of wild type flies. G,H) Tetracyclin feeding modestly retarded development of RpS3+/- and RpS17+/- flies. I,J) After tetracycline feeding, rpS12D97/D97 and rpS12D97/+ genotypes did not accelerate the development of RpS3+/- flies, but in which RpS3+/- rpS12D97/D97 Xrp1+/- flies developed as rapidly as RpS3+/- Xrp1+/- flies. K,L) After tetracycline feeding, rpS12D97/D97 and rpS12D97/+ genotypes did not accelerate the development of RpS3+/- flies, but in which RpS3+/- rpS12D97/D97 Xrp1+/- flies developed as rapidly as RpS3+/- Xrp1+/- flies. The genotypes used and numerical data corresponding to these graphs is tabulated in the S8 Table.
Fig 8.
Model for regulatory effects of Rp mutations.
When one of the many haploinsufficient Rp is limiting, an RpS12-dependent signal is activated that elevates Xrp1 expression, inhibiting imaginal disc cell translation and growth and making Rp+/- cells less competitive. Analyses of imaginal disc gene expression and genetic epistasis studies provide compelling evidence that RpS12 acts through Xrp1 in imaginal discs. Defects in imaginal disc growth delay overall organism development, which may depend on Xrp1-dependent signaling by Dilp8 [21]. Other data suggest that RpS12 also contributes positively to the rate of development, independently of Minute Rp mutations, and that Minute Rp mutations and Xrp1 may also slow organism development independently of RpS12.