Selective reaction monitoring approach using structure-defined synthetic glycopeptides for validating glycopeptide biomarkers pre-determined by bottom-up glycoproteomics

Clusterin is a heavily glycosylated protein that is upregulated in various cancer and neurological diseases. The findings by the Hancock and Iliopoulos group that levels of the tryptic glycopeptide derived from plasma clusterin, 372Leu-Ala-Asn-Leu-Thr-Gln-Gly-Glu-Asp-Gln-Tyr-Tyr-Leu-Arg385 with a biantennary disialyl N-glycan (A2G2S2 or FA2G2S2) at Asn374 differed significantly prior to and after curative nephrectomy for clear cell renal cell carcinoma (RCC) patients motivated us to verify the feasibility of this glycopeptide as a novel biomarker of RCC. To determine the precise N-glycan structure attached to Asn374, whether A2G2S2 is composed of the Neu5Acα2,3Gal or/and the Neu5Acα2,6Gal moiety, we synthesized key glycopeptides having one of the two putative isomers. Selective reaction monitoring assay using synthetic glycopeptides as calibration standards allowed “top-down glycopeptidomics” for the absolute quantitation of targeted label-free glycopeptides in a range from 313.3 to 697.5 nM in the complex tryptic digests derived from serum samples of RCC patients and healthy controls. Our results provided evidence that the Asn374 residue of human clusterin is modified dominantly with the Neu5Acα2,6Gal structure and the levels of clusterin bearing an A2G2S2 with homo Neu5Acα2,6Gal terminals at Asn374 decrease significantly in RCC patients as compared with healthy controls. The present study elicits that a new strategy integrating the bottom-up glycoproteomics with top-down glycopeptidomics using structure-defined synthetic glycopeptides enables the confident identification and quantitation of the glycopeptide targets pre-determined by the existing methods for intact glycopeptide profiling.


Supplementary Information
Page(s) Experimental S3-12 Figure S1 NMR spectra of compound 1 S13 Figure S2 NMR spectra of compound 2 S14 Figure S3 NMR spectra of compound 3 S15 Figure S4 NMR spectra of compound 5 S16 Figure S5 HPLC profiles to determine the concentrations S17 Figure S6 XICs for the preparation of calibration curves S18-19

Figure S7
XICs in the SRM assay of human serum samples S20-21 Table S1 Chemical shifts of glycopeptides 1, 2, 3, and 5 S22-26 Table S2 Mass parameters used in this study S27 Table S3 Amino acid analysis of glycopeptide 3 S28 Table S4 Concentrations of glycopeptides 1 and 2 S28 Serum samples from newly diagnosed RCC patients (n=8) were collected before treatment at Hirosaki University S1 and healthy control samples (n=8) were collected at Showa University Hospital, respectively. After serum collection, all samples were immediately aliquoted, frozen, and stored at -80°C until pre-processing by our experimental workflow.
All commercially available solvents and reagents were used without further purification. Trityl-OH-ChemMatrix ® resin was purchased from Biotage Japan Ltd. N--Fluorenylmethoxycarbonyl-amino acid (Fmoc-amino acid) derivatives except for glycosylated amino acids were purchased from Merck Millipore. N--Fluorenylmethoxycarbonyl-N--(2-N-acetylamido-2-deoxy--D-glucopyranosyl)-Lasparagine [Fmoc-Asn(Ac3β-GlcNAc)-OH] was synthesized by the procedure reported previously S2  μmol, 4 eq.) in DMF and allowed to react with resin at 50°C for 30 min under microwave irradiation. After washing with DMF three times, unreacted amino groups were capped with acetyl group by treating with a solution (Ac2O/DIEA/DMF = 85/10/5). The resin was washed with DMF six times. The streamlined procedures such as Fmoc-removal, amino acid coupling, and acetyl capping as mentioned above were performed repeatedly.
After completing the total synthesis, cleavage cocktail (TFA: H2O: triisopropylsilane = 95:2.5:2.5) was added to the resin and the reaction mixture was shaken at room temperature for 2 h. The solution was filtered and TFA was removed by evaporation, and then 10 mL of t-butyl methyl ether was added. The filtrate in t-butyl methyl ether was centrifuged and supernatant was removed. The residual solid was dissolved in 50% acetonitrile/water and lyophilized. Next, the residue was dissolved in methanol and pH of the solution was adjusted to 12.5 with 1N NaOH aq. The solution was kept at room temperature for 1 h to remove acetyl groups from sugar residues and neutralized with 1 N AcOH aq., then concentrated. Crude product was purified by RP-HPLC to afford compound 3 (9.1 mg, 15% overall yield  Figure S5).  Figure S4).  Table S2.

Concentration of the synthetic clusterin glycopeptides.
Concentrations of the standard solution of clusterin glycopeptides 1 and 2 was determined by combining amino acid analysis and RP-HPLC analysis because of the limited amounts of the glycopeptides 1 and 2 to determine the correct concentrations from their weights. Firstly, amino acid analysis of the glycopeptide 3 was carried out and the result was used for the determination of the concentration to be 590 M (Table S3). Then, the standard solution was subjected to spike with the measurement of RP-HPLC profiles of glycopeptides 1 and 2. By using the RP-HPLC profiles of glycopeptides 1 and 2 in the presence of concentration-defined glycopeptide 3, the concentrations were estimated to be 168 M and 146 M, respectively ( Figure S7 and Table S4). These solutions were employed by dilution to adjust 10 M before the SRM experiments.         Serum samples from 8 RCC patients (entry 1-8) were collected before treatment at Hirosaki University and 8 healthy subjects (entry 9-16) were collected at Showa University Hospital, respectively.

S22
Immediately after serum collection, samples were aliquoted, frozen, and stored at -80°C until preprocessing by an experimental workflow.