To detect intragenic drop-off at levels of Marimastat 20 or less, let alone the 1? frameshifting, previously measured by metabolic labelling, on the UCC_UUU_CGU shift site in the influenza A virus sequence context. Nonetheless, we inspected ASXL genes in ribosome profiling datasets on the GWIPS-viz genome browser [37]. Most datasets did not show noticeable drop-off at the end of the TF ORF for either ASXL1 or ASXL2, though a few data sets showed moderate to strong drop-off. To more clearly visualize ribosome drop-off, we removed intronic regions by remapping datasets to ASXL1 and ASXL2 mRNA transcripts. Two examples ?Jurkat cells from Gawron et al. [38] and MDA-MB-231 cells from Rubio et al. [39] ?are shown in Additional file 1: Figure S6. For ASXL1, the mean ribosome footprint density downstream of the TF ORF was 0.15 times the mean density upstream of the TF ORF in Jurkat cells, whereas in MDA-MB-231 cells the ratio was 1.01, indicating that 85 of ribosomes drop off theASXL1 mRNA within the TF region in Jurkat cells but not in MDA-MB-231 cells. For ASXL2 the ratios were less extreme and much closer to each other ?0.69 and 0.85 in Jurkat and MDA-MB-231 cells respectively ?making it harder to distinguish TF-specific drop-off from potential generic decreases in ribosome footprint density perhaps due to other causes. While the Jurkat ASXL1 ribosome drop-off ostensibly supports efficient PRF in certain cell types, we are suspicious that this particular result is an artefact of heterozygous somatic mutations. Jurkats are a pseudodiploid cell line, with polyploidy occurring in a moderate percentage of cells. Analysis of Gawron et al. RiboSeq reads (typically 30 nt) mapping PubMed ID:https://www.ncbi.nlm.nih.gov/pubmed/16474207 to the ASXL1 TF region revealed two indel mutations ?a GCCCG to GCCCCG insertion present in 29 of 74 reads and a AGGGGGGGGU to AGGGGGGGU deletion present in 5 of 5 reads. The former mimics a -1 frameshift leading to premature termination in the middle of the TF ORF. The latter mimics a +1 frameshift leading to termination at the TF stop codon. Interestingly, to our knowledge Jurkat cells have not been reported to have indels in ASXL1 (https://cansar.icr.ac.uk/cansar/cell-lines/JURKAT/), and analysis of genomic sequencing datasets (SRX2596625, SRX2596624; 150-nt reads) also did not reveal indels at these sites. Thus the Gavron et al. indels may be specific to their isolate of the Jurkat cell line. In summary, therefore, current ribosome profiling datasets neither support nor contradict the PRF hypothesis. However, given that the PRF efficiencies involved may be PubMed ID:https://www.ncbi.nlm.nih.gov/pubmed/8627573 of order a few percent, it is likely that ASXL PRF would not be detectable by this approach.Discussion Hitherto, most documented cases of PRF have been found to occur within mobile genetic elements and particularly in the genomes of RNA viruses [1]. Less attention has been paid to frameshifting as a gene expression mechanism in cellular organisms; however, there are a small number of notable instances of both prokaryotic and eukaryotic chromosomal genes whose expression is dependent upon PRF [1]. The efficiency of frameshifting in two such cases (prokaryotic release factor 2, and eukaryotic antizyme) is regulated via feedback loops. For example, the antizyme genes of yeast and metazoa consist of two partially overlapping reading frames (ORF1 and ORF2), and +1 PRF at the last codon of the former is required for the translation of full-length, biochemically active antizyme proteins [40]. Frameshifting at these sites is resp.