We observed that heat shock of Caenorhabditis elegans leads to the formation of nuclear double-stranded RNA (dsRNA) foci, detectable with a dsRNA-specific monoclonal antibody. These foci significantly overlap with nuclear HSF-1 granules. To investigate the molecular mechanism(s) underlying dsRNA foci formation, we used RNA-seq to globally characterize total RNA and immunoprecipitated dsRNA from control and heat shocked worms. We find a subset of both sense and antisense transcripts enriched in the dsRNA pool by heat shock overlap with dsRNA transcripts enriched by deletion of tdp-1 , which encodes the C . elegans ortholog of TDP-43. Interestingly, transcripts involved in translation are over-represented in the dsRNAs induced by either heat shock or deletion of tdp-1 . Also enriched in the dsRNA transcripts are sequences downstream of annotated genes (DoGs), which we globally quantified with a new algorithm. To validate these observations, we used fluorescence in situ hybridization (FISH) to confirm both antisense and downstream of gene transcription for eif-3 . B , one of the affected loci we identified.

Numerous reports have suggested that infectious agents could play a role in neurodegenerative diseases, but specific etiological agents have not been convincingly demonstrated. To search for candidate agents in an unbiased fashion, we have developed a bioinformatic pipeline that identifies microbial sequences in mammalian RNA-seq data, including sequences with no significant nucleotide similarity hits in GenBank. Effectiveness of the pipeline was tested using publicly available RNA-seq data and in a reconstruction experiment using synthetic data. We then applied this pipeline to a novel RNA-seq dataset generated from a cohort of 120 samples from amyotrophic lateral sclerosis (ALS) patients and controls, and identified sequences corresponding to known bacteria and viruses, as well as novel virus-like sequences. The presence of these novel virus-like sequences, which were identified in subsets of both patients and controls, were confirmed by quantitative RT-PCR. We believe this pipeline will be a useful tool for the identification of potential etiological agents in the many RNA-seq data sets currently being generated.

Numerous reports have suggested that infectious agents could play a role in neurodegenerative diseases, but specific etiological agents have not been convincingly demonstrated. To search for candidate agents in an unbiased fashion, we have developed a bioinformatic pipeline that identifies microbial sequences in mammalian RNA-seq data, including sequences with no significant nucleotide similarity hits in GenBank. Effectiveness of the pipeline was tested using publicly available RNA-seq data. We then applied this pipeline to a novel RNA-seq dataset generated from a cohort of 120 samples from amyotrophic lateral sclerosis (ALS) patients and controls, and identified sequences corresponding to known bacteria and viruses, as well as novel virus-like sequences. The presence of these novel virus-like sequences, which were identified in subsets of both patients and controls, were confirmed by quantitative RT-PCR. We believe this pipeline will be a useful tool for the identification of potential etiological agents in the many RNA-seq data sets currently being generated.

43Abstract 44 We observed that heat shock of Caenorhabditis elegans leads to the formation of nuclear 45 double-stranded RNA (dsRNA) foci, detectable with a dsRNA-specific monoclonal antibody.46 These foci significantly overlap with nuclear HSF-1 granules. To investigate the molecular 47 mechanism(s) underlying dsRNA foci formation, we used RNA-seq to globally characterize total 48 RNA and immunoprecipitated dsRNA from control and heat shocked worms. We find antisense 49 transcripts are generally increased after heat shock, and a subset of both sense and antisense 50 transcripts enriched in the dsRNA pool by heat shock overlap with dsRNA transcripts enriched 51 by deletion of tdp-1, which encodes the C. elegans ortholog of TDP-43. Interestingly, transcripts 52 involved in translation are over-represented in the dsRNAs induced by either heat shock or 53 deletion of tdp-1. Also enriched in the dsRNA transcripts are sequences downstream of 54 annotated genes (DoGs), which we globally quantified with a new algorithm. To validate these 55 observations, we used fluorescence in situ hybridization (FISH) to confirm both antisense and 56 downstream of gene transcription for eif-3.B, one of the affected loci we identified. 57 58 Introduction 59 Cytoplasmic proteotoxic stress induced by temperatures outside of the optimal range for 60 cells or organisms triggers the heat shock response (HSR) [1]. The response to heat shock is 61 multi-faceted and regulation of both transcription and translation occurs. Transcriptional 62 responses include formation of stress granules, alternative splicing, and aberrant transcriptional 63 termination [2-5]. The HSR is a highly conserved transcriptional response and is driven largely 64 by the heat shock transcription factor HSF1 [6]. Under basal level conditions, HSF1 is a 65 monomer in the cytoplasm and nucleus. Upon stress, HSF1 undergoes homotrimerization and 66 binds to DNA heat shock elements (HSE) and initiates the transcription of heat shock protein 3 67 genes [7,8]. In addition, translation of non-heat shock mRNAs is reduced through pausing of 68 translation elongation as well as inhibition of translation initiation [9-11]. Regulation and 69 clearance of misfolded proteins by heat shock proteins has been implicated in neurodegenerative 70 diseases such as Huntington's disease (HD), Parkinson's disease (PD), Alzheimer's disease 71 (AD), and amyotrophic lateral sclerosis (ALS) [12]. 72 Aside from the canonical binding of HSF1 to HSE loci, heat shock can cause HSE-73 independent transcriptional changes [2]. In mammalian cells, HSF1 granules co-localize with 74 markers of active transcription where HSF1 binds at satellite II and III repeat regions [13]. In the 75 worm Caenorhabditis elegans, HSF-1 granules also show markers of active transcription but the 76 putative sites of HSF-1 stress granule binding are unknown [14]. 77 In addition to formation of HSF1 stress granules, heat shock can cause reduced efficiency 78 of transcription termination and the accumulation of normally untransc...