Introduction
In all forms of life, the ribosome is responsible for protein biosynthesis. In eukaryotes the 80S ribosome (70S in bacteria) is comprised of a large subunit (60S for eukaryotes and 50S for prokaryotes) and a small subunit (40S for eukaryotes and 30S for prokaryotes) (1). During ribosome profiling (RP), the removal of rRNA from samples prior to sequencing is important as rRNA contaminants reduce the fraction of mappable reads, restrict the sequencing depth and consequently increase the cost of sequencing (2,3)
rRNA contaminants make up to 90% of all RNA reads upon sequencing (4). This large abundance of rRNA is no surprise considering that after RNA digestion each ribosomal complex is composed of only 28-32bp (eukaryotes) or 15-40bp (prokaryotes) ribosomal protected fragments (RPFs) and several kb of rRNA(5). This contamination is exacerbated in ribosome profiling studies (sequencing of these RPFs ) when the nuclease used to generate RPFs creates widespread nicks in the rRNA (6). This complicates the detection of low abundant transcripts as it requires substantially higher depth of sequencing. To increase the sequence coverage of the transcripts of interest, contaminating rRNAs fragments need to be removed from the pool of all RNA fragments prior to sequencing (7).
Ribo-Zero, long considered to be the gold standard rRNA depletion – due to its ability to target fragmented sequences as well as in-tact sequences – was discontinued as a stand-alone product (8,9). As such, efforts to develop an alternative rRNA depletion technique compatible with ribosome profiling has intensified.
Physical Subtractive Hybridisation
A common approach for rRNA depletion is subtractive hybridisation (Figure 1A), which uses antisense DNA oligonucleotides that are complementary to target rRNA sense sequences (9). These oligonucleotide probes are often biotinylated (10) to allow capturing and removal by streptavidin-coated magnetic beads (11) after hybridisation to the complementary rRNA fragments. Subtractive hybridisation is the underlying mechanism of a number of commercial kits, including the now-discontinued Ribo-Zero. However, many of these kits are designed to target the entire rRNA sequence rather than specific high-abundance contaminants (12). A more reliable and affordable approach to subtractive hybridisation for RP purposes is to design custom oligonucleotide probes targeting the most abundant contaminating rRNA fragments (5). To help the design of species-specific hybridisation probes Alkan et al. (12) developed a software tool Ribo-ODDR for identifying the most abundant rRNA fragments and predicts the rRNA depletion efficiency of potential oligonucleotides. However, even when using custom-designed biotinylated oligonucleotides, streptavidin beads remain very expensive.
Enzymatic Subtractive Hybridisation
An alternative method for rRNA depletion involves RNase H digestion (Figure 1B) of the RNA in RNA:DNA hybrids (13). This strategy is similar to subtractive hybridisation in that that it is based on using deoxyoligonucleotides with sequences complementary to rRNA that hybridise to targeted rRNA fragments. However, instead of physical removal, RNAse H digests RNA strands in a RNA:DNA hybrid. DNase addition afterwards digests remaining DNA strands, leaving untargeted RPFs intact. RNase H has previously not been recommended for use in RP studies due off-target trimming and degradation of RPFs (2).
cDNA Normalisation with Duplex-specific nucleases
cDNA normalisation (Figure 1C) with duplex-specific nuclease (DSN) offers another option for rRNA depletion. The aim of this approach is to utilise second order kinetics so that relative transcript sequences can be equalised to some extent. DSN is an enzyme that degrades duplex DNA in preference to single-stranded DNA at high temperatures (14). Unlike some of the previous mentioned techniques, this type of rRNA depletion would be carried out after first-strand cDNA synthesis (it does not work effectively on amplified cDNA libraries (14)). Firstly, cDNA is denatured followed by reassociation, however, as the hybridisation rate is proportional to the square of its concentration (15) higher abundant transcripts will form double-stranded DNA more effectively than less abundant transcripts. Following this reassociation step, DSN digestion takes place, but as DSN degrades duplex DNA in preference to single-stranded DNA this allows degradation of those higher abundant transcripts (derived from rRNA) which have reannealed before the less abundant transcripts (derived from reads of interest such as mRNA), thus, normalising levels of low abundant transcripts within the overall cDNA pool.
Chung et al. (16) found that DSN was able to deplete rRNA to a level comparable to RiboZero treatment in RP eukaryotic samples, while Giannoukos et al. (8) was one of the first studies to show the use of this method in prokaryotes, however, it had poor rRNA depletion efficiency and sequence-specific GC biases in Rhodbacter sphaeroides. Some limitations apply to this strategy, for example it has been shown to be better suited to poly(A) enriched mRNA libraries (15) making it more appropriate for RNA-seq libraries as RPFs do not contain poly(A) tails. In addition, extensive optimisation of timing DSN digestion step as well as hybridisation buffer concentrations would need to take place on a sample by sample basis.
Targeted Amplification
An alternative method to deplete rRNA involves the use of specific, not-so-random (NSR) hexamer/heptamer primers (Figure 1D) (17,18) that bind to target rRNA during cDNA synthesis. This method is commercialised by the Ovation RNA-Seq kit from NuGen, a kit that is designed to target a collection of bacterial and archaeal rRNA sequences (8,19), but has not been shown to be suitable for RP purposes. NSR primers have been utilised for RP purposes in the literature (20). However, Song et al. (21) observed that when NSR primers were applied to Rhodopseudomonas palustris 3’ end bias was observed. In addition, this mechanism was also determined unsuitable for studying bacterial operon relationships (co-expression of genes and their organisation within an operon) where uniform coverage is required (17).
CRISPR-Cas9 Depletion
An emerging approach for rRNA depletion is the use of Clustered Regularly Interspaced Palindromic Repeats (CRISPR) and CRISPR associated (Cas) proteins (Figure 1E). CRISPR-Cas9 depletion of abundant sequences was first performed by Gu et al. (22) in a method termed DASH (Depletion of high abundance sequences) as a way to remove unwanted mitochondrial rRNA from HeLa cell line RNA and pathogen-derived RNA from human cerebrospinal fluid (CSF). This method utilises the precision of CRISPR Cas9 to target specific sequences acting much like a restriction enzyme. For target recognition, the 20nt spacer sequence in the Cas9 complex must form complimentary base pairing with 20nt protospacer (target rRNA sequence ) (23) however in order for Cas9 to bind DNA a protospacer adjacent motif (PAM) must be present directly next to the 20nt target site on the 3’ end (24,25) A bonus to this method is that it allows depletion at the cDNA level, reducing the input necessary as depletion can now be done on an amplified library and multiplexing becomes a possibility at this point. While in theory subtractive hybridisation is possible at the cDNA level, the expense of conducting this method of depletion on an amplified library is not feasible due to the volume of beads that would be required.
Montefiori et al. (26) used CRISPR Cas9 to deplete contaminating reads of mitochondrial origin in their ATAC-seq (Assay for Transposase-Accessible Chromatin sequencing) technique – a technique for identifying DNA sequences located in open chromatin. While Hardigan et al. (27) adapted the DASH method to deplete microRNA molecules (miRNAs) naming it MAD-DASH (Depletion of adapter dimer and overabundant miRNAs). MAD-DASH effectively reduced adapter dimer formation and more accurately quantified miRNAs from tissue-derived RNA. Prezza et al. (28) adapted DASH for prokaryotes using two model bacteria: Salmonella enterica and Bacteroides thetaiotaomicron. Not only did this study effectively remove 56-86% rRNA reads from their RNA-seq library, but they also did this with sub-nanogram input of RNA.
Since the publication of these methods, there has been one publication (29) and one pre-print article released utilising CRISPR-Cas9 for depletion of rRNA in RP libraries (30). Wilkins & Ule (29) developed what they termed RiboCutter and found that Cas9 treatment for rRNA depletion did not alter footprint distribution and did not perturb sub-codon level resolution of footprints while effectively depleting rRNA from RP libraries. Han et al. (28) used CRISPR Cas9 for depletion of rRNA in both disome and monosome libraries to which they had a successful increase in mRNA mapping reads. A limitation to this strategy is the requirement for sequences to contain a PAM site, but one way to overcome this would be to include a PAM in the 5’ or 3’ adapter as done by Hardigan et al.(27).
Figure 1: rRNA depletion techniques. A) rRNA depletion by Subtractive Hybridisation with biotinylated probes and streptavidin magnetic beads. B) rRNA depletion by RNaseH digestion of RNA in RNA:DNA heteroduplexes. C) rRNA depletion by DSN digestion of dsDNA. D) Selective amplification of desired mRNA sequences by addition of hexamer primers. E) CRISPR-Cas9 depletion of rRNA. Image Created with Biorender.com.
Conclusions
Because most of these methods deplete rRNA at the RNA level, high RNA input levels are required which makes them less suitable for low-input samples. Overall, the mentioned methods range in their effectiveness, with custom oligonucleotide beads being the most commonly used option for rRNA depletion of RP libraries in the literature (12). The biggest issue for most commercial kits is that they are designed to work on RNA-seq libraries, that is in-tact fragments that exist in equimolar ratios, but RPFs are biased by nuclease digestion and size selection (12). CRISPR-mediated depletion of rRNA offers a promising alternative in that it offers a much more affordable and time-efficient approach due to the possibility of multiplexing samples prior to rRNA depletion.
References
- Schuwirth BS, Borovinskaya MA, Hau CW, Zhang W, Vila-Sanjurjo A, Holton JM, et al. Structures of the bacterial ribosome at 3.5 Å resolution. Science (1979). 2005;310(5749):827–34.
- Zinshteyn B, Wangen JR, Hua B, Green R. Nuclease-mediated depletion biases in ribosome footprint profiling libraries. RNA. 2020;26(10):1481–8.
- Latif H, Szubin R, Tan J, Brunk E, Lechner A, Zengler K, et al. A streamlined ribosome profiling protocol for the characterization of microorganisms. Biotechniques. 2015;58(6):329–32.
- Morlan JD, Qu K, Sinicropi D v. Selective depletion of rRNA enables whole transcriptome profiling of archival fixed tissue. 2012;
- Ingolia NT, Brar GA, Rouskin S, McGeachy AM, Weissman JS. The ribosome profiling strategy for monitoring translation in vivo by deep sequencing of ribosome-protected mRNA fragments. Nat Protoc. 2012;7(8):1534–50.
- Ingolia NT, Ghaemmaghami S, Newman JRS, Weissman JS. Genome-wide analysis in vivo of translation with nucleotide resolution using ribosome profiling. Science (1979). 2009;324(5924):218–23.
- Kraus AJ, Brink BG, Siegel TN. Efficient and specific oligo-based depletion of rRNA. Sci Rep. 2019;9(1):1–8.
- Giannoukos G, Ciulla DM, Huang K, Haas BJ, Izard J, Levin JZ, et al. Efficient and robust RNA-seq process for cultured bacteria and complex community transcriptomes. Genome Biol. 2012;13(3):1–13.
- Petrova OE, Garcia-Alcalde F, Zampaloni C, Sauer K. Comparative evaluation of rRNA depletion procedures for the improved analysis of bacterial biofilm and mixed pathogen culture transcriptomes. Sci Rep. 2017;7(1):1–15.
- Engelhardt F, Tomasch J, Häussler S. Organism-specific depletion of highly abundant RNA species from bacterial total RNA. Access Microbiol. 2020;2(10).
- Pang X, Zhou D, Song Y, Pei D, Wang J, Guo Z, et al. Bacterial mRNA purification by magnetic capture‐hybridization method. Microbiol Immunol. 2004;48(2):91–6.
- Alkan F, Silva J, Pintó Barberà E, Faller WJ. Ribo-ODDR: oligo design pipeline for experiment-specific rRNA depletion in Ribo-seq. Bioinformatics. 2021;
- He S. Ribosomal RNA removal methods for microbial transcriptomics. In: Metagenomics for Microbiology. Elsevier; 2015. p. 39–53.
- Shagin DA, Rebrikov D v, Kozhemyako VB, Altshuler IM, Shcheglov AS, Zhulidov PA, et al. A novel method for SNP detection using a new duplex-specific nuclease from crab hepatopancreas. Genome Res. 2002;12(12):1935–42.
- Yi H, Cho YJ, Won S, Lee JE, Jin Yu H, Kim S, et al. Duplex-specific nuclease efficiently removes rRNA for prokaryotic RNA-seq. Nucleic Acids Res. 2011;39(20):e140–e140.
- Chung BY, Hardcastle TJ, Jones JD, Irigoyen N, Firth AE, Baulcombe DC, et al. The use of duplex-specific nuclease in ribosome profiling and a user-friendly software package for Ribo-seq data analysis. Rna. 2015;21(10):1731–45.
- Hirakawa H, Oda Y, Phattarasukol S, Armour CD, Castle JC, Raymond CK, et al. Activity of the Rhodopseudomonas palustris p-coumaroyl-homoserine lactone-responsive transcription factor RpaR. J Bacteriol. 2011;193(10):2598–607.
- Chugani S, Kim BS, Phattarasukol S, Brittnacher MJ, Choi SH, Harwood CS, et al. Strain-dependent diversity in the Pseudomonas aeruginosa quorum-sensing regulon. Proceedings of the National Academy of Sciences. 2012 Oct 9;109(41).
- Peano C, Pietrelli A, Consolandi C, Rossi E, Petiti L, Tagliabue L, et al. An efficient rRNA removal method for RNA sequencing in GC-rich bacteria. Microb Inform Exp. 2013;3(1):1–11.
- Shiber A, Döring K, Friedrich U, Klann K, Merker D, Zedan M, et al. Cotranslational assembly of protein complexes in eukaryotes revealed by ribosome profiling. Nature. 2018 Sep 29;561(7722):268–72.
- Song Y, Milon B, Ott S, Zhao X, Sadzewicz L, Shetty A, et al. A comparative analysis of library prep approaches for sequencing low input translatome samples. BMC Genomics. 2018 Dec 21;19(1).
- Gu H, Sun YH, Li XZ. Novel rRNA-depletion methods for total RNA sequencing and ribosome profiling developed for avian species. Poult Sci. 2021;101321.
- Jiang F, Doudna JA. CRISPR–Cas9 Structures and Mechanisms. Annu Rev Biophys. 2017 May 22;46(1).
- Francisco J M Mojica 1, César Díez-Villaseñor, Jesús García-Martínez, Elena Soria. Intervening Sequences of Regularly Spaced Prokaryotic Repeats Derive from Foreign Genetic Elements. J Mol Evol. 2005 Feb;60(2).
- Jinek M, Chylinski K, Fonfara I, Hauer M, Doudna JA, Charpentier E. A Programmable Dual-RNA-Guided DNA Endonuclease in Adaptive Bacterial Immunity. Science (1979). 2012 Aug 17;337(6096).
- Montefiori L, Hernandez L, Zhang Z, Gilad Y, Ober C, Crawford G, et al. Reducing mitochondrial reads in ATAC-seq using CRISPR/Cas9. Sci Rep. 2017;7(1):1–9.
- Hardigan AA, Roberts BS, Moore DE, Ramaker RC, Jones AL, Myers RM. CRISPR/Cas9-targeted removal of unwanted sequences from small-RNA sequencing libraries. Nucleic Acids Res. 2019;47(14):e84–e84.
- Prezza G, Heckel T, Dietrich S, Homberger C, Westermann AJ, Vogel J. Improved bacterial RNA-seq by Cas9-based depletion of ribosomal RNA reads. RNA. 2020;26(8):1069–78.
- Han P, Shichino Y, Schneider-Poetsch T, Mito M, Hashimoto S, Udagawa T, et al. Genome-wide Survey of Ribosome Collision. Cell Rep. 2020 May;31(5).
- Wilkins OG, Ule J. Ribocutter: Cas9-mediated rRNA depletion from multiplexed Ribo-seq libraries. bioRxiv. 2021;