Figure 1.
Possible outcomes of genome evolution.
As a genome evolves, the accumulated mutations can be neutral, having no impact on the molecular phenotype (that is, the functions encoded in the genome and the ways that these are regulated), or they can lead to adaptation via changes in heritable phenotype due to changes in the molecular phenotype. Developmental System Drift (DSD) describes a third possibility: while the overall phenotype of the organism remains identical, the underlying genetic networks underpinning this phenotype have changed. A key outcome of this is that some orthologous genes play different in vivo roles in phenotypically identical, related species.
Figure 2.
Outline of screening procedures.
A. Manual comparison of RNAi phenotypes in both species. RNAi phenotypes are screened by eye on 12 well plates. For each gene being examined, three replicates of the C. elegans RNAi and six replicates of the C. briggsae RNAi clone are screened; in each replicate, phenotypes are examined in the progeny of a single adult worm that has been exposed to dsRNA-expressing bacteria for 72 hrs. Each plate was scored by two people at two time-points (24 hrs and 48 hrs after removal of adult), as described in Kamath et al. In the example shown, knockdown of Y39G10AR.7 produces a sterile phenotype in C. elegans but not in C. briggsae. B. Automated analysis of RNAi phenotypes. RNAi was carried out in liquid cultures in 96-well plates as described in Methods. Each well was sampled with a commercially available worm sorter and we quantified the number of animals in each well, as well as the size (Time of Flight, TOF) and optical density (Extinction, Ext) of each animal. The scatter plots show Time of Flight (TOF, x-axis) and Extinction (EXT, y-axis) for the population in a representative well of either C. elegans or C. briggsae following RNAi targeting Y39G10AR.7. In addition, the corresponding brood size defect is shown (see Methods for calculation).
Figure 3.
Functional enrichment in genes with different in vivo functions in C. elegans and C. briggsae.
A. All 1333 genes analyzed were manually placed into the functional categories described in Kamath et al. [14]. The graph shows the proportion of genes that have different RNAi phenotypes in several different functional classes: all genes analyzed (‘All’), genes annotated to have roles in Protein Synthesis (Prot. Synth.), Transcription Factors (‘TF’), or genes of Unknown function (‘Unk’). Classes with significantly fewer genes with different RNAi phenotypes are shown in blue; those with statistically increased numbers are shown in red. Enrichments are significant with an FDR of 0.05 (Hypergeometric test). B. RNAi phenotypes differ most for more recently evolved genes. All 1333 genes analyzed were placed into five classes based on their evolutionary age as described in Methods. The most ancient genes could be dated back to the emergence of the Opisthokta lineage (‘Opis.’), then becoming progressively younger, we could date sets of genes back to the emergence of the Coelomata (‘Coel.’), Nematoda (Nem.), Chromadorea (‘Chro.’), and finally some genes had arisen so recently that they were only detectable in Caenorhabditis species (‘Caen.’). In each case, the graph shows the proportion of genes in each evolutionary class that had a different RNAi phenotype.
Figure 4.
Schematic illustrating transgenic rescue approach.
The transgenic rescue approach illustrated here shows the possible molecular events driving changes in gene function and how transgenic rescue with various hybrid rescue constructs would be interpreted to differentiate between these. In each case regions of constructs in red are derived from C. elegans while regions of construct in blue are derived from C. briggsae. Coding regions are shown as coloured boxes.
Figure 5.
in vivo expression of a subset of genes with different RNAi phenotypes.
C03D6.1, K04G7.1, C27F2.7, and sac-1 had strongly different RNAi phenotypes in C. elegans and C. briggsae. We generated transgenic C. elegans strains (N2) expressing GFP under control of the promoter of the C. elegans orthologue or C. briggsae strains (AF16) expressing mWormCherry under control of the orthologous C. briggsae promoter for each gene. In each case, four panels are shown: DIC image of N2 worms transgenic for the C. elegans promoter driving GFP, fluorescence image of N2 worms transgenic for the C. elegans promoter driving GFP, DIC image of AF16 worms transgenic for the C. briggsae promoter driving mWormCherry expression, fluorescence image of AF16 worms transgenic for the C. briggsae promoter driving mWormCherry expression. Images are confocal projections at 200× magnification, and scale bars represent 100 µm, except for C. elegans K04G7.1 which is at 400× magnification with a scale bar representing 50 µm. Images are representative of 3 independent lines. A. Expression difference for C03D6.1. Arrow heads indicate tail cells (white), intestine (red) and hypodermis (blue). B. Expression difference for K04G7.1. Arrow heads indicate head neurons (white) and body wall muscle (red). C. Expression difference for C27F2.7. Arrow heads indicate head neurons (white), hypodermis (blue), intestine (yellow), vulva (green) and tail neurons (red). D. Expression difference for sac-1. Shown are 3 confocal projections along the body of the animals 400× magnification. Scale bars represent 50 µm. Arrowheads indicate the intestine (blue), pharynx and pharyngeal neurons (white), spermatheca (yellow), and tail cells (red).
Figure 6.
Differences in sac-1 RNAi phenotype are due to differences in sac-1 promoter function.
We generated C. elegans lines transgenic either for the C. elegans sac-1 ORF under control of the C. elegans sac-1 promoter, the C. briggsae sac-1 ORF under control of the C. briggsae sac-1 promoter, or for the two hybrid constructs shown. In each case, we examined the ability of the transgenic array to rescue the developmental arrest phenotype of sac-1(ok1602) homozygous animals — the graph shows the percentage of animals that reached the adult stage that are homozygous for the sac-1(ok1602) allele, indicating rescue. Either the C. elegans sac-1 ORF or the C. briggsae sac-1 ORF under control of the C. elegans sac-1 promoter could partially rescue; no rescue was seen for the remaining constructs, indicating that the sac-1 promoter has diverged in the two species, while the sac-1 coding regions appear to be functionally interchangeable.
Figure 7.
bli-5 and bli-4 have an identical gene function despite showing different RNAi phenotypes.
A. RNAi phenotype of bli-5 in C. elegans (N2) and C. briggsae (JU1018). B. Quantification of the phenotype shown in panel A. C. Rescue of the bli-5(e518) phenotype by either C. elegans or C. briggsae bli-5 genes. We generated transgenic bli-5(e518) lines in which either C. elegans bli-5 coding region was expressed under the control of C. elegans bli-5 promoter (C. elegans rescue) or the C. briggsae bli-5 under the control of the C. briggsae bli-5 promoter (C. briggsae rescue). We examined adult animals and assessed the proportion with blistered cuticles; the results were combined across lines, with a minimum of 3 lines. Error bars represent the standard error on the binomial proportion. D. Similar data to panel C but instead showing rescue of the bli-4(e937) allele with analogous C. elegans and C. briggsae bli-4 constructs.