Supporting Information Lysine Propionylation to Boost Proteome Sequence Coverage and Enable a Silent SILAC Strategy for Relative Protein Quantification Christoph U. Schräder 1, Shaun Moore 1,2, Aaron A. Goodarzi 1,2 and David C. Schriemer 1,3 * 1 Department of Biochemistry and Molecular Biology, University of Calgary, Calgary, Alberta, Canada T2N 4N1 2 Robson DNA Science Centre, Arnie Charbonneau Cancer Institute, University of Calgary, Calgary, Alberta, Canada, T2N 4N1 3 Department of Chemistry, University of Calgary, Calgary, Alberta, Canada T2N 4N1 *Corresponding author: David C. Schriemer, Ph.D. Department of Biochemistry and Molecular Biology The University of Calgary, Room 300 Heritage Medical Research Building 3330 Hospital Drive NW Calgary, Alberta, Canada T2N 4N1 Email: dschriem@ucalgary.ca 1
Figure S1. Analysis of protein mix composed of BSA, α-amylase and myoglobin, untreated or propionylated followed by GluC digestion. (A) LC/MS chromatogram showing the TIC of eluting peptides for the untreated sample (black) and the propionylated sample (orange). (B) Observed sequence coverage for the three proteins under the two conditions as indicated, along with their molecular weight. 2
Figure S2. Illustration of propionylation efficiency. (A) XIC for peptide RALKAWSVARLSQKFPKAE showing high and specific labeling efficiency of Lys residues. (B) Radar plot representation of propionylation yield shown for five peptides as indicated. Complete propionylation is shown in red, under- and over-propionylation are shown in yellow and blue, respectively. Line width represents standard deviation. 3
Figure S3. Cleavage specificity of trypsin, LysargiNase and GluC, with and without propionylation, as visualized as IceLogo plots. Differences are displayed as the amino acid occurrence in a certain position normalized to the natural occurrence in H. sapiens. Number of unique cleavage sites that were taken into account for each condition are provided in the figure. 4
Figure S4. Characteristics of precursor and fragment ions for the different HeLa digests. (A) Charge-state distribution of precursor ions selected for MS2 fragmentation using HCD. (B) Total fragment ion abundance observed for the different digests. Only native immonium ions were taken into account, and no other reporter ions (e.g. m/z 84). Error bars represent the standard deviation of two biological replicates. Legends are provided in the figures. Derivatization did not affect general trends in fragment ion formation, where we observed mostly b ions after LysargiNase digestion (ArgN-like), y ions after tryptic (ArgC-like) and a mixture of both for peptides released by GluC. 5
Figure S5. HeLa cell lysate (with and without propionylation) digested with GluC. (A) Total ion chromatogram (TIC) of lysates. (B) Number of Lys residues as a function of peptide length plotted against corresponding retention time. Each peptide is shown as a dot and further visualized as box and violin plots per 10-minute interval. A value of 1 means that each residue in the corresponding peptide is Lys, whereas 0 means that the peptide does not contain any. All identified peptides (n) combined across both biological replicates were taken into account. 6
Figure S6. Gain in protein groups per replicate analysis using the same enzyme with and without propionylation. Plot shows additional protein groups per replicate as indicated compared to replicate 1 (untreated sample). Number of protein groups was here 2101 (GluC), 2590 (LysargiNase) and 3494 (Trypsin), respectively. 7
Figure S7. Comparison of identified protein groups for untreated HeLa lysate digested with LysargiNase, GluC or trypsin. Numbers of protein groups per condition, and in total (underlined), are provided. Only protein groups are displayed and compared that were identified in both biological replicates at a protein FDR of 1%. 8
Figure S8. Identified and in silico calculated peptide lengths/masses of enzymatically released peptides. (A) Box plot representation of identified peptide mass among the different conditions for two biological replicates of HeLa lysate digests. Detected and all MS2 triggered peptide masses are shown in grey and red, respectively. (B) In silico digestion of the human proteome using five digestion strategies as indicated. Note that ArgN/ArgC- and trypsin/lysarginase produced peptide lengths completely overlap. (C) Alternative visualization of the in silico digestion, trimmed of the peptides <600 Da. 9
Figure S9. Analysis of missed cleavage sites for all conditions tested. Total missed cleavage site for untreated LysargiNase and trypsin samples that can cleave Lys and Arg residues are further subdivided for missed Lys and Arg residues, respectively. Legend of missed cleavages is provided in the figure. 10
Figure S10. Chemical structures for Lys and ions commonly seen upon HCD fragmentation. (A) Representative peptide containing Lys. Curved lines represent b- and y-type cleavages. (B) Structure of the α-amino-ε-caprolactam ion at m/z 129, along with its monoisotopic mass, representing a cyclic rearrangement of Lys upon fragmentation as shown in (A). (C) Structure of the cyclic imine at m/z 84 along with its monoisotopic mass, formed from the Lys side chain upon further breakdown of Lys. Backbone 1-C carbonyl and amino groups (in blue) lost during fragmentation are indicated in blue. 11
Figure S11. Product ion spectra of the peptide KKLFYSTFATDDRKE (HCD fragmentation at 32% NE, Uniprot accession no. P30084, enoyl-coa hydratase, mitochondrial) released by GluC from untreated (top; m/z 924.9782, z = 2, RT = 16.52 min) and propionylated (bottom; m/z 1009.0152, z = 2, RT = 86.13 min) HeLa cell lysate in the m/z range 80 to 1800. Propionylated Lys residues are marked by a brown circle. Immonium ions and related ions are shown in green and the ion at m/z 129, representing protonated α-amino-ε-caprolactam, is shown in purple. Respective m/z values are further provided for clarity. The Kpr derived immonium ion (m/z 157) and reporter ion (Rep-Kpr) at m/z 140 are labeled accordingly. We did not observe the propionylated counterpart for α-amino-ε-caprolactam (theoretical m/z at 185). PEAKS peptide scores are shown in the upper or lower right corners. 12
Figure S12. In silico analysis of Lys residues per protein in the human proteome. 13
Figure S13. Number of Lys residues per identified peptide as a function of their relative abundance in propionylated HeLa cell lysate. The dotted red line shows the cut-off between no Lys residues present (to the left) and those peptides having at least one Lys (to the right). Their summed abundance is further given for each protease. Error bars represent the deviation of two biological replicates. 14
Figure S14. Product ion spectra of three peptides (HCD fragmentation at 32% NE), containing ten propionylated Lys residues each. The corresponding protease which was used is provided in the figure along with the respective PEAKS peptide scores. Peptide identities and additional characteristics, TOP: Uniprot accession no. Q14978, nucleolar and coiled-body phosphoprotein 1 (m/z 931.7900, z = 4, RT = 100.32 min, protease: GluC), MIDDLE: Uniprot accession no. Q14683, structural maintenance of chromosomes protein 1A (m/z 919.9213, z = 5, RT = 108.17 min, protease: LysargiNase), BOTTOM: Uniprot accession no. Q02878, 60S ribosomal protein L6 (m/z 1048.5974, z = 4, RT = 116.75 min, protease: Trypsin). Propionylated Lys residues are marked by a brown circle. Immonium ions and related ions are shown in green. The Kpr derived reporter ion (Rep-Kpr) at m/z 140 is labeled. 15
Figure S15. Incorporation efficiency of 1-13 C/6-13 C-Lys into human A549 adenocarcinoma cells (page 17). (A) Precursor ion spectra of peptide QEILAALEK (Uniprot accession no. P07602; 16
Prosaposin) metabolically labeled either with 1-13 C-Lys (top, solid line) or 6-13 C-Lys (bottom, dotted line) along with their respective retention time. Labeled Lys residue is underlined. The abundance of the monoisotopic peak corresponding to the unlabeled peptide is approx. 2% compared to the labeled peptide. Samples were digested with trypsin without propionylation. (B) Centroid HCD fragment ion spectra of peptide QEILAALEK, labeled either with 1-13 C-Lys (top; m/z 508.2971, z = 2, RT = 50.49) or 6-13 C-Lys (bottom; m/z 508.2970, z = 2, RT = 49.74). Spectra were acquired in the orbitrap analyzer at a resolution of 15,000. Reporter ions are shown in green, b-ions in blue and y-ions in red. Unlabeled peaks represent internal ions or those resulting from neutral loss. Both spectra show the same sequence-specific fragment ions, which have the same accurate mass and can only be differentiated in the low mass region by the corresponding reporter ions highlighted in bold at either m/z 84 (1-13 C-Lys) or m/z 85 (6-13 C- Lys), shown as a zoom scan to the right (displaying the range m/z 83.9 86.5 of the same fragment ion spectra). 17
Figure S16. Isobaric ions at m/z 140 and 141. (A) Analysis of all theoretical isobaric ions at m/z 140 (representing 1-13 C Rep-Kpr; left) and m/z 141 (representing 6-13 C Rep-Kpr; right) that could be released from peptides upon HCD fragmentation. No other post-translational modifications were taken into account. 1-13 C Rep-Kpr is shown in blue and 6-13 C Rep-Kpr is shown in red, both at a theoretical resolution of 15,000. Isobaric ions are shown in dotted lines together with their possible identities. The theoretical relative mass difference between the isobaric ions is shown. (B) Actual relative intensity of isobaric ions at m/z 140 and 141, sorted by 18
their relative abundance from GluC-digested propionylated HeLa cell lysate. Identities are provided. Figure S17. Overall identification metrics for A549 cells grown in 1-13 C/6-13 C-Lys SILAC media and mixed at either 1:1 or 5:1 ratio. Mixed lysates were propionylated and subsequently digested with trypsin. The number of triggered MS2 (MS/MS) scans is shown on the right y-axis and represented in grey bars. The number of PSMs, identified peptides and protein groups is indicated on the left y-axis. Results for two biological replicates are shown. 19
Figure S18. Hypothetical chemical structures of isotopes used to extend the 2-plex workflow to a 4-plex experiment. (A) Chemical structure of propionylated lysine along with atoms that will be isotopically tagged. Blue and green marked atoms will be lost as a neutral mass during reporter ion formation and red and orange marked atoms remain part of the reporter ion, thus encoding for m/z values between 140 and 143. Extension to a 4-plex experiment will involve the use of 18 O labeled propionic anhydride and 18 O labeled Lys. (B) Chemical structures of the four isotopically tagged reporter ions that will be generated together with their monoisotopic masses. 20