RNA molecules carry, in addition to their four canonical bases, a diverse range of chemically modified ribonucleosides. The methylation of carbon-5 of cytosines (m5C) in RNAs is a highly conserved and widely distributed feature in biological systems. Yet, the molecular functions of this modification, as well as the mechanisms through which this mark is interpreted are still poorly understood. In this thesis, I used Caenorhabditis elegans as a model organism for the study of genes and molecular functions associated to the m5C pathway. Using CRISPR-Cas9, I developed the first eukaryotic organism in which m5C is not detectable in RNA – the noNSUN strain. By exploiting this genetic tool, I uncoupled phenotypes associated to loss of function and loss of catalytic activity of m5C-methyltransferases, and showed that, while m5C is a non-essential modification, it is required for C. elegans normal development and fertility. In addition, I partially unravelled the synthetic pathway of hm5Cm, an oxidative derivative of m5C. By taking advantage of the noNSUN strain as a negative control in whole transcriptome bisulfite sequencing experiments, I determined the distribution of m5C sites throughout C. elegans’ transcriptome, and found that most methylated sites localise to the variable loop of tRNAs. In addition, I found four modified positions in rRNAs, and discovered an unprecedented role of NSUN-4 in mitochondrial tRNA methylation. I also investigated the presence of m5C sites in coding transcripts, and detected novel modified positions in small non-coding RNAs. I discovered that m5C has a relevant role in the response to heat stress, as the noNSUN strain shows temperature-sensitive fertility and developmental phenotypes. Upon temperature changes, phenotypical aggravation is accompanied by a global reduction in protein synthesis and by alterations in tRNA homeostasis. I found that the noNSUN strain shows transcriptional alterations when compared to the wild-type, and establishes a different transcriptional response to heat stress. Furthermore, I showed that translation is mostly affected by increased ribosome pausing at UUG codons upon loss of m5C. Finally, I attempted to identify proteins that bind to m5C and its derivatives. I produced preliminary evidence that ALY-1 and ALY-2 preferentially bind to m5C-modified RNA in vitro. In summary, this thesis presents a systematic characterisation of the m5C pathway in C. elegans. It provides detailed phenotypical and molecular analyses of the first metazoan mutant strain devoid of this modification in RNA, as well as the first list of m5C residues in C. elegans. The results presented here suggest a participation of m5C in the translational adaptation to heat stress.