Ribosomal stalk proteins RPLP1 and RPLP2 promote biogenesis of flaviviral and cellular multi-pass transmembrane proteins

The ribosomal stalk proteins, RPLP1 and RPLP2 (RPLP1/2), which form the ancient ribosomal stalk, were discovered decades ago but their functions remain mysterious. We had previously shown that RPLP1/2 are exquisitely required for replication of dengue virus (DENV) and other mosquito-borne flaviviruses. Here, we show that RPLP1/2 function to relieve ribosome pausing within the DENV envelope coding sequence, leading to enhanced protein stability. We evaluated viral and cellular translation in RPLP1/2-depleted cells using ribosome profiling and found that ribosomes pause in the sequence coding for the N-terminus of the envelope protein, immediately downstream of sequences encoding two adjacent transmembrane domains (TMDs). We also find that RPLP1/2 depletion impacts a ribosome density for a small subset of cellular mRNAs. Importantly, the polarity of ribosomes on mRNAs encoding multiple TMDs was disproportionately affected by RPLP1/2 knockdown, implying a role for RPLP1/2 in multipass transmembrane protein biogenesis. These analyses of viral and host RNAs converge to implicate RPLP1/2 as functionally important for ribosomes to elongate through ORFs encoding multiple TMDs. We suggest that the effect of RPLP1/2 at TMD associated pauses is mediated by improving the efficiency of co-translational folding and subsequent protein stability.

Methods
Data collection:

A549 cells were plated at 1.5 × 106 cells per 10 cm dish. Three 10cm dishes were transfected with NSC siRNA whereas three other dishes were transfected with either siP1_1, siP2_1, or siP2_4 siRNAs as described above. 48 h later cells were infected with DENV-2 (NGC strain) at MOI of 10 in a total volume of 10 ml, rocked every 15 min for 1 h and the infection was allowed to proceed for an additional 1.5 h. Cells were then flash-frozen in liquid nitrogen without cycloheximide (CHX) pretreatment and a cold lysis buffer containing CHX was used to lyse the cells on ice. The RIBOseq strategy was adapted from Ingolia and colleagues with a few modifications. After nuclease digestion, samples were run on a 15–50% sucrose gradient and the 80S ribosome fractions were collected. As described by Reid and colleagues, fractions were extracted using Trizol LS (Thermo Fisher Scientific), and rRNAs were removed using the Ribo-Zero gold rRNA removal kit (Illumina, San Diego, CA, USA) according to the manufacturer's protocol. The remaining RNA was treated with PNK and then size selected by 15% denaturing PAGE. For adapter ligation and library building, we used the NEBNext Small RNA Library Prep Set (Illumina). Data can be accessed at the gene expression omnibus repository: GSE133111.

Data processing:

RNA-seq and footprinting reads were jointly mapped to the human transcriptome and DENV genome using the riboviz pipeline (22). Sequencing adapters were trimmed from reads using Cutadapt 1.10 using parameters ‐‐trim-n -O 1 ‐‐minimum-length 5. Trimmed reads that aligned to human/mouse rRNA were removed using HISAT2 version 2.1.0. The remaining reads were mapped to a set of 19,192 principal transcripts for each gene in the APPRIS database using HISAT2. Only reads that mapped uniquely were used for all downstream analyses. For genes with multiple principal transcripts, the first one on the list was chosen. Codes for selecting these transcripts were obtained from the riboviz package (https://github.com/shahpr/riboviz).