The identification of phosphoglycerate kinase-1 and histone H4 autoantibodies in pancreatic cancer patient serum using a natural protein microarray

Abstract

Protein microarrays have been used to explore whether a humoral response to pancreatic cancer-specific tumor antigens has utility as a biomarker of pancreatic cancer. To determine if such arrays can be used to identify novel autoantibodies in the sera from pancreatic cancer patients, proteins from a pancreatic adenocarcinoma cell line (MIAPACA) were resolved by 2-D liquid-based separations, and then arrayed on nitrocellulose slides. The slides were probed with serum from a set of patients diagnosed with pancreatic cancer and compared with age- and sex-matched normal subjects. To account for patient-to-patient variability, we used a rank-based non-parametric statistical testing approach in which proteins eliciting significant differences in the humoral response in cancer compared with control samples were identified. The prediction analysis for microarrays classification algorithm was used to explore the classification power of the proteins found to be differentially expressed in cancer and control sera. The generalization error of the classification analysis was estimated using leave-one-out cross-validation. A serum diagnosis of pancreatic cancer in this set was predicted with 86.7% accuracy, with a sensitivity and specificity of 93.3 and 80%, respectively. Candidate autoantibody biomarkers identified using this approach were studied for their classification power by performing a humoral response experiment on recombinant proteins using an independent sample set of 238 serum samples. Phosphoglycerate kinase-1 and histone H4 were noted to elicit a significant differential humoral response in cancer sera compared with age- and sex-matched sera from normal patients and patients with chronic pancreatitis and diabetes. This work demonstrates the use of natural protein arrays to study the humoral response as a means to search for the potential markers of cancer in serum.

Figure 1

Humoral response experimental overview. Proteins are first extracted from cell line and separated in two orthogonal dimensions. Separated fractions are spotted by non-contact means on nitrocellulose slides, which are then probed with serum from normal or cancer sera. Antibody–antigen response is detected using antihuman IgG conjugated to a fluorophore. Following statistical analysis, proteins of interest are identified by MS/MS.

Figure 2

2-D UV chromatogram of separated MIAPACA cell lysate. On the horizontal axis are fractions from CF starting from the lowest pH to the highest pH. On the vertical axis is increasing retention time or hydrophobicity of the separated protein.

Figure 3

Selected microarray shots of differential humoral response as well as selected tandem mass spectrum for sequence confirmation of (A) fibrillarin and (B) cathepsin D.

Figure 4

A grid of p-values from Wilcoxon rank-sum tests between cancer and normal sera per separated fraction in (A) foreground only and (B) background-subtracted data. The grid is arranged according to the 2-D fractionation of the whole cell lysate and colored according to the level of significance of the direction of the difference between cancer and normal sera where gray indicates no evidence of change. (C) z-score plot for proteins separated from pH fraction 5.1–4.9. On the vertical axis are all fractions by increasing retention time and on the horizontal axis are each of the serum samples with which samples were probed. Red and yellow blocks represent responses significantly higher than the mean of the normal sample (4<z<25 and 2<z<4, respectively), whereas blue and green blocks represent responses significantly lower than the mean of the normal sample (−25<z< −4 and −4<z< −2, respectively).

Figure 5

(A) This ROC curve shows error rates attributed to each of the 30 threshold values considered by PAM, corresponding to 30 subsets of the 98 proteins. The circle highlights the chosen threshold that uses only nine proteins for classification, which provides the lowest error using the smallest number of proteins. The AUC for this ROC curve was estimated to be 0.85. (B) Boxplots of the intensities per diagnosis for each of the nine proteins included in the classifier panel built using all 30 samples.

Figure 6

Heatmap showing median-centered responses of all serum samples to selected proteins of interest. The scale from dark to light represents lower response to higher response on a scale of −2 to 2. The arrows indicate the protein spots that formed the panel of nine potential markers with highest sensitivity and specificity.

Figure 7

Scatterplot illustrating the differential humoral response in recombinant human (A) histone H4 and (B) PGK-1 used for validating initial experimental results. Code: cancer, diamond; diabetes, square; pancreatitis, triangle; normal, x.