Transcriptome-wide measurement of translation by ribosome profiling

Abstract

Translation is one of the fundamental processes of life. It comprises the assembly of polypeptides whose amino acid sequence corresponds to the codon sequence of an mRNA's ORF. Translation is performed by the ribosome; therefore, in order to understand translation and its regulation we must be able to determine the numbers and locations of ribosomes on mRNAs in vivo. Furthermore, we must be able to examine their redistribution in different physiological contexts and in response to experimental manipulations. The ribosome profiling method provides us with an opportunity to learn these locations, by sequencing a cDNA library derived from the short fragments of mRNA covered by the ribosome. Since its original description, the ribosome profiling method has undergone continuing development; in this article we describe the method's current state. Important improvements include: the incorporation of sample barcodes to enable library multiplexing, the incorporation of unique molecular identifiers to enable to removal of duplicated sequences, and the replacement of a gel-purification step with the enzymatic degradation of unligated linker.

Keywords: High-throughput sequencing; RNA; RNA-sequencing; Ribosome; Ribosome profiling; Translation.

Figure 1. A schematic of the ribosome profiling protocol

The major steps of the updated ribosome profiling protocol, highlighting several of the most important developments. Each step has the same name and number as the corresponding part of this protocol. UMI – Unique Molecular Index.

Figure 2. Footprint fragment size-selection gel

An image of a typical gel prepared according to Section 5.3. Footprinting samples were prepared from cultures of wild-type S. cerevisiae strain BY4742, as described in this protocol. NI-800 is used to determine the lower position on the gel to cut in order to isolate large (~28 nt) footprints. It is still routinely run, but has been made somewhat redundant by the discovery of small footprints and the use of the NEB miRNA marker.

Figure 3. Specific degradation of pre-adenylated linker by combined treatment with Yeast 5′ deadenylase and RecJ

The presence of a reaction component is indicated by a filled circle, the absence of a component by a empty circle. Y5pDA – Yeast 5′ deadenylase. L – 10 bp ladder. The 100 nt band and selected other bands of the 10 bp ladder is indicated on the right-hand side of the gel. Ligated product is indicated. The effect of Yeast 5′ deadenylase and RecJ treatment on the pre-adenylated linker band is indicated by the box.

Figure 4. Hotter reverse transcription reduces untemplated extension without reducing yield

Reverse transcription of ligated footprints (left), linker only (middle), or RT primer alone (right) with Protoscript II at temperatures ranging from 46°C – 57°C. Reaction were resolved by polyacrylamide gel electrophoresis as described in Section 5.3. An example of non-templated addition to the reverse transcription (RT) primer is indicated. L – 10 bp ladder, the 100 nt band and selected other bands are indicated on the left-hand side of the gel.

Figure 5. cDNA size-selection gel

An image of a typical cDNA size-selection gel prepared according to Section 5.6. In order to capture the greatest diversity of sequences we routinely cut just above the no-insert control (lane 3) to somewhat above the positive control (NI-805, lane 5). The efficiency of the reverse transcription can be judged the amount of unutilized primer (NI-802) in the footprinting sample (lane 6). Selected bands from the 10 bp ladder are indicated on the left-hand side of the gel.

Figure 6. Library construction PCR gel

As outlined in Section 5.9, each sample is loaded across two lanes of the gel. Note the downwards smear in lanes 4 & 5 that is absent from lanes 2 & 3, below the pink box denoting the area of the gel that was excised for purification of the full-length library. This smear likely represents no-insert product and unelongated reverse transcription primer that has persisted through the cDNA size-selection gel. The size (nucleotides)of several bands from the 10 bp ladder are indicated on the left-hand side of the gel.

Figure 7. Library length distribution analysis using the Agilent 2200 TapeStation

An electropherogram image of the length distribution of the ribosome profiling library generated by the TapeStation analysis software. Note the single major peak at 168 bp, within the range of what we typically expect for a ribosome profiling library. The small proportion of larger sequences, and the absence of a discernable no-insert product peak are also fairly typical. Insert: The false gel image generated from the same sample.

Figure 8. Characteristics of ribosome footprints

All panels are based on the analysis of a single sample from the described library (barcode ATCGT), prepared from wild-type S. cerevisiae BY4742 grown to OD₆₀₀ 0.6 in YEPD. A The frequency distribution of fragment lengths between 16 – 36 nucleotides (nt). B A meta-gene plot of fragments of all lengths whose 5′ end aligns within 100 nt of the start codon. The height of each bar represents the number of fragment 5′ ends aligning to a given nucleotide position. The color of the bar represents the frame of that nucleotide relative to the A in the AUG start codon. The lower panel illustrates the portion 55 – 85 nt after the start codon with greater resolution. C & D are heatmaps illustrating the log10 of the number of fragment 5′ ends (after the addition of a pseudo count of 1) of each fragment length at each nucleotide around the start codon (C) and stop codon (D).