A genetic pathway encoding double-stranded RNA transporters and interactors regulates growth and plasticity in Caenorhabditis elegans

The environment and genes shape the development, physiology and behaviour of organisms. Many animal species can take-up double-stranded RNA (dsRNA) from the environment. Environmental dsRNA changes gene expression through RNA interference (RNAi). While environmental RNAi is used as a laboratory tool, e.g. in nematodes, planaria and insects, its biological role remains enigmatic. Here we characterise the environmental dsRNA receptor SID-2 to understand the biological function of dsRNA uptake in Caenorhabditis elegans. First we determine that SID-2 localises to the apical membrane and the trans-Golgi-network (TGN) in the intestine, implicating the TGN as a central cellular compartment for environmental dsRNA uptake. We demonstrate that SID-2 is irrelevant for nucleotide uptake from the environment as a nutritional (nitrogen) source. Instead RNA profiling and high-resolution live imaging revealed a new biological function for sid-2 in growth and phenotypic plasticity. Surprisingly, lack of the ability to uptake environmental RNA reduces plasticity of gene expression. Furthermore, using genetic analyses we show that the dsRNA pathway genes sid-2, sid-1 and rde-4 together regulate growth. This work suggest that environmental RNA affects morphology and plasticity through gene regulation.

previous reports [26], in addition it localised to numerous foci in the cytoplasm (Fig 1B). sid-2 does not function in dsRNA uptake for nutritional reasons 119 Because SID-2 localisation is conserved in C. elegans and C. briggsae, but mediates 120 artificial dsRNA uptake only C. elegans, we wondered if SID-2 conserved function is 121 enhancing the uptake of dsRNA for nutritional purposes. Since the nutrient rich 122 laboratory environments might mask such a function, we use a C. elegans strain with a 123 compromised PYRimidine biosynthesis pathway pyr-1 [67] to test if SID-2 takes up 124 dsRNA for nutritional reasons. Specifically, we asked if exogenous sources of 125 pyrimidines contribute nutritionally in the pyr-1 sensitised background, and if the 126 uptake of exogenous sources requires sid-2 (Fig 2A). In this experiment, C. elegans    Table), was slower in wild-type animals than in sid-2 mutants (Fig 3C), suggesting that 196 the present of SID-2 slows development perhaps do to dsRNA uptake.

197
From the same individuales the ex-utero development time and the time from 198 hatching until the first egg was laid were collected (S3 Table). Together with the three 199 estimated parameters from the logistic function (logistic max, logistic rate, logistic  indicating that wild-type animals are more variant that sid-2 mutant animals (S3 205   Table). To test if the phenotypic plasticity of sid-2 mutants was different from wild-type 206 animals, we performed a statistical analysis testing for equality of the variation of the 207 projection phenotypic data on the first four principle components containing more than 208 99% if the variance comparing sid-2 mutant and wild-type animals (Box's M tests p 209 <0.0016). This results suggest that SID-2 increases phenotypic plasticity. sid-2 slows growth rate and increases phenotypic plasticity A) MA plot visualising embryo transcriptome comparison of wild type (n = 5) and sid-2 mutants (total n = 9, sid-2(qt142) n = 4, sid-2(mj465) n = 5). Each red circle represents a statistically significant (DE) transcript (FDR <0.01). B) Embedded wild type and sid-2 mutant embryo transcriptome in principle component space of embryo development data (Boeck et al 2016). C) Growth curve visualising body length from hatching to egg laying adult of wild-type animals (n = 20) and sid-2(qt142) (n = 31), sid-2(mj465) (n = 33) mutants, line represent median and error-bar represent the 95% confidence interval of the median. D) Principle component analysis representing the combined phenotypic data along the first and second principle component of the analysed wild-type animals (n = 20) and sid-2(qt142) (n = 31), sid-2(mj465) (n = 33) mutants. Individual circles represent aggregated phenotypical data of a individual animal. The line represent the 33% and 66% contour line. Probability density estimate of the phenotypic data is plotted left and top of the PCA plot.
To understand at what stages of development SID-2 may affect growth and phenotypic 212 plasticity, western blot analysis was performed throughout development. The specificity 213 of the SID-2 antibody was tested by comparing wild-type adults and sid-2(qt142) null 214 mutant adults (S5A Fig). SID-2's predicted molecular weight is 33 kDa, however bands 215 were detected at ≈ 40 kDa and at ≈ 20 kDa that were absent in sid-2 mutants. At ≈ 216 40 kDa two bands were detected indicating that two isoforms of SID-2 exist, potentially 217 with different post-translational modifications (e.g. glycosylation, a modification placed 218 in the Golgi [70]). The smallest band at ≈ 20 kDA is potentially a degradation product 219 of full length SID-2. The band observed at ≈ 60 kDa was detected as well in sid-2 220 mutants, and therefore likely result from non-specific binding to the SID-2 antibody.  To identify a potential exogenous RNA responsible for the sid-2 phenotype, we 230 sequenced long and short RNA of wild-type and sid-2 L4 animals. We then compared 231 the amount of E. coli derived reads from wild-type and sid-2 mutant animals. No 232 significant difference in the total amount of bacteria-derived RNA was detected (S6B 233 Fig). Next, we asked if sid-2 is required for the uptake of a specific transcript. Using 234 differential expression analysis we were unable to identify E. coli transcripts that were 235 significantly lower in abundance in sid-2 mutants than in wild-type animals (S6C Fig). 236 Together these experiments indicate that, we were unable to identify a potential 237 causative environmental RNA, suggesting that body length could be regulated by  To further support the involvement for a dsRNA in the body length, we performed 267 the length measurement in C. briggsae wild-type animals and compared them to 268 C. briggsae sid-2 mutants animals. The body length did not differ between wild-type 269 and C. briggsae sid-2 mutant animals ( Fig 4B) showing that sid-2 does not regulate involved in dsRNA processing [26,27,29,32,74,75], we speculate that naturally 306 occurring dsRNA is transported and processed by these proteins to regulate growth.

307
Furthermore, due to the localisation of SID-2 in the apical intestine membrane we 308 suspect, that dsRNA from the environment is regulating growth (Fig. 5).  determining on how a more natural environment shapes C. elegans through RNA. We 333 expect that such RNA-based interactions will be of broad relevance in many species.     from Dharmacon using the concentrations indicated in (Table. 1) according to [82]. For 443 C. briggsae sid-2 CRISPR dpy-10 crRNA was replaced with plasmid pCFJ90  Imaging plates for developmental analysis were made using the standard NGM recipe, 507 but without peptone and cholesterol. Furthermore, agarose was substituted with 0.8% 508 Gelzan (Sigma G1910-250G) for a more transparent gel. Imaging plates were seeded 509 with with one µL of HB101 concentrate at optical density 20.  Table. SID-2 and AMA-1 protein sequences in Caenorhabditis.

621
S2 Table. sid-1 and sid-2 animals are resistant to rpb-2 RNAi by feeding. 622 RNAi feeding experiment on L4 larva fed with either control (L4440) or rpb-2 RNAi. 623 The presents (+) or absence (-) of F1 larvae was scored after 48  Sequencing data is available in the European Nucleotide Archive under the study 645 accession number PRJEB32813.