Deuterated formaldehyde and sodium cyanoborohydride have been used in a process called reductive methylation of ubiquitinated isopeptides (RUbI) . This produces one deuterated dimethyl group at the a-amino group of N-terminus and the iso-N-terminus of multiply charged Ub-isopeptides enhancing their analysis using QTOF-based LC-MS/MS with low-energy collision-cell CID . The presence of two methyl groups replacing two hydrogen atoms on the amine at the N-terminus of a peptide imparts favorable fragmentation behavior of an N-terminal dimethylated peptide and N-terminal product-ion generation under low-energy collision-cell CID by (i) preventing the proton transfer that would usually occur from a non-dimethylated amine, from the ai ion to the yx ions in the ai-yx peptide fragmentation pathway, resulting in the generation of a highly abundant a1'-type ion [55, 56]; (ii) improved coverage of predominantly b/b' and also y/y' product ions due to the dimethyl modification reducing the a-amino group's basic character in the gas phase, enabling more random protonation of the peptide backbone, facilitating fragmentation by CID ; (iii) the electron-donating nature of the dimethyl groups stabilizes the a1 ion resulting in its enhancement abundance postfragmentation of the N-terminal Calpha-Camide bond of the peptide in a similar way to the a1-yx peptide fragmentation pathway ; and (iv) Harrison demonstrated that N-terminal acetylated peptides prevent sequence scrambling of peptides under low-energy collision-cell CID conditions and advantageously improved the presence of higher-order b-ions . This could suggest that the presence of the dimethyl group similarly provides steric hindrance at the N-terminus of a Ub-isopeptide, advantageously resulting in the presence of higher-order b-type product ions. The three advantages that the dimethyl groups impart on the fragmentation of peptides (and N-terminal product-ion formation) were also observed from both N-termini of Ub-isopeptides . The dimethyl groups at each N-termini of a Ub-isopeptide result in the addition of 32 Da at each N-termini, thereby increasing the mass of the Ub-isopep- tide by 64 Da. The CID MS/MS spectra of a dimethylated di-Ub-K48-isopeptide ion resulted in the generation of comprehensive b- and y-type product ions and the generation of (i) a characteristic abundant a1 ion from its N-terminus, which identifies the first N-terminal amino acid to be leucine/isoleucine at m/z 118.16; (ii) a characteristic abundant a1' ion from its iso-N-terminus, which identifies the first iso-N-terminal amino acid of the iso-chain to be G at m/z 62.09; and (iii) a b2' ion at m/z 147.11 ion from the iso-N-terminus, thus confirming the full identity of the iso-chain attached to the acceptor lysine of the Ub-isopeptide to be GG. The comprehensive a1 product ion, b- and y-type product ions generated from the Ub-isopeptide enabled manual structural elucidation of its backbone and the ai' and b2' ions enabled the manual spectro- metric structural elucidation of its GG iso-chain (Figure 6.5a). By comparison, the CID MS/MS spectrum of the equivalent dimethylated K48-peptide, that is, the peptide without the GG iso-chain covalently attached to the acceptor lysine (Figure 6.5b), only shows the presence of the a1 ion at m/z 118.16 from the N-terminus of the peptide but does not show the additional a1' and b2', which represent the GG iso-chain as expected.
The combined diagnostic utility of the a1' and b2' ions was assessed by extraction postacquisition from LC-MS/MS data representing the tryptic digest of an anti-Ub pulldown of HEK293-T cell lysate samples spiked with a tryptic digest of six-protein mix. Extracting the a1' and b2' ions results in a significant reduction in chromatographic data complexity toward enhancing the analysis of potential Ub-isopeptide candidates. The retention times of peak profiles from the two extracted ion chromatogram (XIC) traces that perfectly co-eluted were identified as potential Ub-isopeptides. The unequivocal identity of all these Ub-isopeptides was confirmed by manual interrogation of their CID MS/MS spectra. The observation of the a1' and b2' ions thus enabled
Figure 6.5 (a) The mass range between m/z 60 and 280 shows the presence of the diagnostic a1'-type product ion (m/z 62.09) and the b2'-type product ion (m/z 147.11) generated from the GG iso-chain. (b) These two diagnostic ions are shown to be absent in the linear version of the peptide without the GG iso-chain modification. Source: Chicooree, 2013 . Reproduced with permission from Springer.
comprehensive spectrometric structural elucidation of the GG iso-chain, in addition to the presence of a comprehensive series of b- and y-type product ions, enabling the structural elucidation of the backbone sequence of the Ub- isopeptide. This demonstrated the robust diagnostic utility and enhanced selectivity of the ai' and b2' ions toward the analysis of Ub-isopeptides in complex samples. From the Ub-isopeptides identified, the selectivity of these diagnostic a1' and b2' ions in the RUbI approach enabled the identification of a “false-negative” large Ub-isopeptide which carried 6+ charges and contained two miscleavages within its peptide sequence and its site-specific GG isochain, which was proved to be beyond the searching and identification capabilities of Mascot when these dimethylated groups were added as variable modifications. Furthermore, Mascot suggested the presence of a Ub-isopeptide that was not detected during the RUbI approach. However, when analyzing the CID MS/MS spectra of the Ub-isopeptide there were no diagnostic a1' and b2' ions to indicate the presence of a GG iso-chain attached to the acceptor lysine, thereby confirming the Ub-isopeptide identified by Mascot to be a falsepositive result. A minor caveat to the RUbI approach is that the a1' and b2' diagnostic ions would also be generated from tryptic linear peptides with a GG at their N-terminus; however, experimental and theoretical agreement is that this type of false-positive result would occur at a rate of approximately 0.5%. This demonstrates the important role the information-rich MS/MS spectrum has to play as it facilitates structural confirmation or rebuttal of the conclusions made based on extraction of the a1' and b2' ions . Furthermore, a proportion of this 0.5% would also contain linear peptides that may be derived from target proteins that have undergone N-terminal protein ubiquitination, which occurs significantly less frequently than lysine protein ubiquitination.
The RUbI approach has also found utility in the enhanced analysis of SUMO(2/3)-isopeptides atypically generated from tryptic digestion of simple di-SUMO(2/3)-ylated proteins that have been spiked into a mixture of six protein tryptic peptides . It was found that the presence of these dimethyl groups at the N-termini and iso-N-termini of these SUMO(2/3)-isopeptides also imparted favorable N-termini fragmentation behavior of the isopeptide and N-termini product-ion generation under low-energy collision-cell CID conditions. This resulted in the generation of (i) a set of modification-specific diagnostic a1', b2', and b3' ions from the iso-N-termini of a SUMO(2/3)- isopeptide with an atypically cleaved TGG iso-chain and (ii) a set of modification-specific diagnostic a1', b2', and b4' ions from the iso-N-termini of a SUMO(2/3)-isopeptide with an atypically cleaved QTGG iso-chain. For example, the CID MS/MS spectra of a dimethylated di-SUMO(2)-isopeptide ion (Figure 6.6) resulted in the generation of comprehensive b- and y-type product ions and the generation of (i) a characteristic abundant a1 ion from its N-terminus, which identifies the first N-terminal amino acid to be glutamic acid at m/z 134.10; (ii) a characteristic abundant a1' ion from its iso-N-terminus,
Figure 6.6 The low-energy collision-cell CID MS/MS spectrum of dimethyl-labeled SUMO(2)-isopeptide bearing a QTGG iso-chain derived from atypical tryptic digestion of a SUMO(2)ylated SUMO(2) protein, showing the presence of abundant diagnostic ions generated from the QTGG iso-chain; a,' (m/z 133.13) (shown in the region zoomed in between m/z 130 and 140), b2' (m/z 262.17) and b4' (m/z 376.22) ions annotated. Source: Chicooree, 2013 . Reproduced with permission from Springer.
which identifies the first iso-N-terminal amino acid of the iso-chain to be Q at m/z 133.13; (iii) a b2' ion at m/z 262.17 from the iso-N-terminus, which identifies the first two amino acids of the iso-chain to be QT; and (iv) a b4' ion at m/z 376.22 from the iso-N-terminus, thus confirming the full identity of the isochain attached to the acceptor lysine of the SUMO(2)-isopeptide to be QTGG. The comprehensive a1 product ion and b- and y-type product ions generated from the SUMO(2)-isopeptide enabled manual structural elucidation of its backbone and the a1' and b2' ions enabled the manual spectrometric structural elucidation of its QTGG or TGG iso-chain.
XlCs of the modification-specific diagnostic iso-N-termini ions result in a significant reduction in chromatographic data complexity. Perfect co-elution of the XIC traces enabled the identification of potential SUMO(2)-isopeptides from simple and semi-complex digestion mixtures. This demonstrated the robust diagnostic utility and enhanced selectivity of the ai' and b' ions toward the analysis of SUMO(2/3)-isopeptides in semi-complex digestion mixtures.