Sample preparation for mass spectrometry, enrichment and purification strate-

To ease interpretation of MS and MS/MS spectra of cross-linked heteroconjugates, both proteins and RNA need to be hydrolyzed thoroughly prior to and/or following enrichment or isolation strategies^{[62, 63]}. Disassembly and denaturation of protein–RNA complexes can be achieved in 1 M urea or guanidine hydrochloride, larger amounts of other detergents like SDS should be avoided (see above)^[49]. As for most proteomic approaches, trypsin is favored as the enzyme for proteolysis as it cleaves after the basic residues lysine and arginine. This usually leads to peptide fragmentation series starting from the peptide C-terminus, aiding data interpretation. Chymotrypsin has also been applied successfully, especially in studies of snurportin 1 and U1 snRNA as well as reconstituted human [15.5K-61K-U4atac snRNA(-U6atac snRNA)][47, 64–66]. Use of different endoproteinases can lead to the identification of additional cross-linking sites within the same protein^[45]. Since increas-ing length of the cross-linked RNA oligonucleotide leads to suppression of the peptide fragment signals, RNA hydrolysis to single or a low number of nucleotides is desirable, especially for ESI-MS^[62]. This can also be achieved by complete hydrolysis of the oligonucleotide with HF^[67]. Due to the low cross-linking yield and the usually limited amounts of starting material, the purifi-cation or enrichment of cross-linked heteroconjugates is a crucial step in the sample preparation for mass spectrometric analysis. The high excess of noncross-linked peptides and oligonucleotides would otherwise hinder cross-link detection and identification. Through enrichment, sample complexity is greatly reduced, which is beneficial in maximizing the number of potential cross-links chosen for MS/MS fragmentation and for data analysis. In addition, signal suppression by noncross-linked RNA oligonucleotides and peptides is decreased.

Figure 1.10: Isolation of cross-linked heteroconjugates from noncross-linked peptides by size ex-clusion chromatography. After proteolysis under denaturing conditions, full-length RNA together with cross-linked peptides can be isolated from peptides by size ex-clusion. After RNA hydrolysis, noncross-linked oligonucleotides need to be removed through suitable methods.

Purification by reversed phase high performance liquid chromatography (RP-HPLC) or size exclu-sion (SE) chromatography as well as enrichment via immobilized metal-ion affinity chromatography (IMAC) or titanium dioxide material have been applied successfully.

Size exclusion chromatography can be applied if proteins and RNA differ considerably in size or for isolation of RNA with and without cross-linked peptides following proteolysis. Both approaches were combined in cross-linking studies of the prokaryotic ribosome[36, 46, 68]. In a first SE, ribosomal RNA (rRNA) with cross-linked proteins was separated from noncross-linked ribosomal proteins.

After proteolysis, a second SE step separated rRNA with cross-linked peptides from noncross-linked peptides. rRNA containing fractions were hydrolyzed by nucleases and cross-noncross-linked peptide–

RNA heteroconjugates were separated by RP-HPLC. Isolated cross-links were subjected to Edman sequencing to identify the sequence of the cross-linked peptide. In many cases, the cross-linked amino acid led to a gap in the sequence analysis and could thus be identified.

In contrast to ribosomes, the size difference between uridine-rich small nuclear RNAs (U snRNAs) and their associated proteins is not sufficient to permit their separation by size exclusion chro-matography. Therefore in studies of human small nuclear ribonucleoprotein particles (snRNPs), SE was only applied after proteolysis (as outlined in Figure 1.10). RNA containing fractions were subsequently hydrolyzed with nucleases and endoproteinases. The mixture was then separated by RP-HPLC. Monitoring absorption at both 220 nm (peptides) and 260 nm (RNA) allowed the detection of heteroconjugates (e.g. ^{[64, 66]}).

After size exclusion chromatography and hydrolysis of RNA-containing fractions, the mixture can also be directly subjected to on-line LC-ESI-MS/MS. This was demonstrated in a cross-linking study of the human U1 snRNP and the reconstituted [15.5K-61K-U4atac snRNA] complex. An

1.3 UV induced protein–RNA cross-linking 17 extensive washing step was included to remove the nonlinked RNA oligonucleotides, cross-linked heteroconjugates and residual peptides were retained on the trapping column^[63].

Figure 1.11: Enrichment of cross-linked heteroconjugates with C18 and titanium dioxide chro-matography. The protein–RNA complex is UV irradiated, for native RNA at 254 nm.

Next, the complex is hydrolyzed under denaturing conditions with RNases and en-doproteinases. Desalting with C18 material removes the majority of noncross-linked RNA oligonucleotides. Finally, titanium dioxide chromatography separates noncross-linked peptides from the cross-noncross-linked heteroconjugates. These are then subjected to LC-ESI-MS/MS analysis.

Figure originally published in ^[48].

Several enrichment protocols for phosphopeptides are based on the interaction of the phosphate groups with metal ions. These can be adapted to enrich peptide–RNA oligonucleotide heterocon-jugates via the phosphate groups in the RNA backbone. At first, enrichment protocols were based on immobilized metal-ion affinity chromatography (IMAC) with Fe(III) ions for cross-link to either RNA^{[62, 66]} or DNA^{[67, 69]}.

More recently, enrichment based on titanium dioxide (TiO₂) chromatography was applied. Protocols initially established for phosphopeptides use competitive binding with 2,5-dihydroxy benzoic acid (DHB) to TiO₂ to reduce co-enrichment of acidic peptides^[70]. This approach could be directly conveyed to cross-linked peptide–RNA oligonucleotide heteroconjugates^[71]. After UV irradiation and ethanol precipitation, the protein–RNA complexes are hydrolyzed by RNases and trypsin. The sample is then desalted and the cross-links are subsequently enriched over TiO₂ columns in the presence of DHB^[71](see Figure 1.11). In contrast to IMAC agarose beads, titanium dioxide can be integrated into an HPLC setup. A two dimensional LC approach, combining C18 and TiO₂columns, has been used for enrichment and subsequent automatic spotting for MALDI-MS analysis^[72]. Overall, the most frequently applied method in our laboratory is TiO₂enrichment with spin columns, as it requires less sample amounts compared to size exclusion or reversed phase isolation and is more selective than IMAC enrichment strategies. It also allows for the enrichment of several samples in parallel and is compatible with LC-ESI-MS/MS analysis, which is advantageous compared to the 2D LC setup with subsequent MALDI analysis in terms of sample processing time and quality of the obtained MS information.