Normalized Ribo-Seq for Quantifying Absolute Global and Specific Changes in Translation

Abstract

Ribosome profiling (Ribo-Seq) is a highly sensitive method to quantify ribosome occupancies along individual mRNAs on a genome-wide scale. Hereby, ribosome-protected fragments (= footprints) are generated by nuclease digestion, isolated, and sequenced together with the corresponding randomly fragmented input samples, to determine ribosome densities (RD). For library preparation, equal amounts of total RNA are used. Subsequently, all transcript fragments are subjected to linker ligation, cDNA synthesis, and PCR amplification. Importantly, the number of reads obtained for every transcript in input and footprint samples during sequencing depends on sequencing depth and library size, as well as the relative abundance of the transcript in the sample. However, the information pertaining to the absolute amount of input and footprint sequences is lost during sample preparation, hence ruling out any conclusion whether translation is generally suppressed or activated in one condition over the other. Therefore, the RD fold-changes determined for individual genes do not reflect absolute regulation, but have to be interpreted as relative to bulk mRNA translation. Here, we modified the original ribosome profiling protocol that was first established by Ingolia et al. (2009), by adding small amounts of yeast lysate to the mammalian lysates of interest as a spike-in. This allows us to not only detect changes in the RD of specific transcripts relative to each other, but also to simultaneously measure global differences in RD (normalized ribosome density values) between samples. Graphic abstract: Global changes in translation efficiency can be detected with polysome profiling, where the proportion of polysomal ribosomes is interpreted as a proxy for ribosome density (RD) on bulk mRNA. Ribo-Seq measures changes in RD of specific mRNAs relative to bulk mRNA. The addition of a yeast-lysate, as a spike-in for normalization of read counts, allows for an absolute measurement of changes in RD.

Keywords: Normalized ribosome density; Regulation of translation; Ribo-seq; Ribosome profiling; Spike-in.

Figure 1.. Schematic overview of the protocol.

On the left side, the experimental procedure is illustrated, highlighting the addition of yeast lysate as spike-in, after preparation of human cell lysates. On the right side, the data analysis workflow is depicted, highlighting the sample-wise normalization to the sum of yeast ORF reads, to retrieve information about global changes in RD.

Figure 2.. Identification of monosomal fractions after sucrose density gradient centrifugation.

A. UV absorbance at 254 nm during gradient elution with and without RNase I digestion of RPEI lysates. The position of fraction borders is marked by sharp drops in the profile. B. As in A, for RAW264.7 cells.

Figure 3.. Periodicity and read distribution around the start and stop codon.

The distance between the 5’ end of reads and the start or stop codon is determined separately for different read lengths. The three frames are indicated in different colors (1: red, 2: purple, 3: blue).

Figure 4.. Normalized RD as scatter plot.

Human ORF read counts are normalized to the sum of yeast ORF read counts. Ribosome densities (RD) are calculated as the ratio of footprint counts to input counts (FP/IN). The mean of this ratio is determined from three independent biological replicates. Differentially regulated mRNAs are identified with the DESeq2 package. RDs with a p-value <0.05 after Benjamini-Hochberg multiple testing correction and a fold-change of >2 or <0.5 are highlighted. The total least squares regression line is depicted in red, and a line through the origin with slope =1 in black.