Charge state deconvoluted ESI mass spectra of intact DBP isolated directly from the blood plasma of 3 cancer patients and 3 healthy individuals with matching DBP diploid genotype. The base peaks at approximately 51,200 Da represent unmodified DBP (which varies in exact mass according to DBP genotype), a peak at Δm + 162 Da, if present, represents nonenzymatically glycated DBP, a peak at Δm + 203.2, if present, represents DBP modified by a single GalNAc residue, a broadened peak at approximately Δm + 365.4 likely represents an O-linked disaccharide (DS) glycoform (Gal₁GalNAc₁),8 and a peak at Δm + 656.6 represents the O-linked trisaccharide glycoform (TS, NeuNAc₁Gal₁GalNAc₁).8,13 Displayed masses correspond to the observed m/z value of an “MH⁺” species. (A) Homozygous 1F/1F DBP from a breast cancer patient (calculated unmodified 1F DBP m/z = 51,189.2) (B) Homozygous 1F/1F DBP from a healthy individual (C) Homozygous 1S/1S DBP from a colorectal cancer patient (calculated unmodified 1S DBP m/z = 51203.2) (D) Homozygous 1S/1S DBP from a healthy individual (E) Homozygous Gc2/Gc2 DBP from a colorectal cancer patient (calculated unmodified Gc2 DBP m/z = 51216.3) (F) Homozygous Gc2/Gc2 DBP from a healthy individual. Though not shown directly, the mass resolution is sufficient to readily classify the diploid genotype of all heterozygous combinations of DBP allele products.

Relative percent of DBP modified with a GalNAc monosaccharide at T420 in cancer patients compared to age and gender matched healthy individuals. Data were tabulated only when a mass spectral peak representing the target species was observed (i.e., samples with a relative percentage of zero are not included). This molecular species was not observed on any of the 32 observed Gc2 allele products-even when found paired in heterozygous fashion with Gc1 protein carrying the residual GalNAc residue-hence the T420 positional assignment. Error bars represent standard deviation. *P < 0.02, Student's t-test.

Allele-specific relative abundance of DBP O-linked trisaccharide glycosylation in cancer patients compared to healthy individuals. n-values for each sub-group are as follows: Total Healthy Individuals = 29, Total Cancer Patients = 56; 1F/1F Healthy = 3, 1F/1F Cancer = 4; 1S/1S Healthy = 10, 1S/1S Cancer = 15, Gc2/Gc2 Healthy = 4, Gc2/Gc2 Cancer = 4; 1F/1S Healthy = 3, 1F/1S Cancer = 10, 1S/Gc2 Healthy = 5, 1S/Gc2 Cancer = 18; 1F/Gc2 Healthy = 3, Cancer = 3. A few rare, unclassified DBP genotypic variants were also observed. Error bars indicate sample standard deviation. There are no statistically significant differences between cancer patients and healthy controls for any diploid genotype. As depicted in the plot and shown in raw format in Figure 1, Gc2 allele products completely lack trisaccharide glycosylation; the presence of trisaccharide glycosylated-DBP in heterozygous individuals is unambiguously due to modification of 1F or 1S DBP.8,9