Dataset for: Tumour antigen-specific antibody signatures as diagnostic biomarkers for melanoma
Description
Circulating IgG and IgM Abs were profiled in serum samples from two independent melanoma cohorts, early-stage (n = 65) and advanced-stage (n = 77); as well as healthy controls (n = 86) and individuals with any prior cancer in remission (n = 37). A protein microarray immobilized with 263 cancer–testis (CT) antigens was used to detect serum Ab reactivity. Univariate and multivariate ROC analysis were applied to identify diagnostic Ab signatures.
Files
Steps to reproduce
Serum Ab profiling assay This study utilized the Sengenics CT antigen protein microarray (Sengenics Corporation, Singapore) containing 263 tumour antigens (CT and non-CT antigens) (Supplementary Table 1) to identify Abs to cognate tumour antigens in the serum of melanoma patients, healthy donors, and those with prior cancers. This standard protocol has been previously published. The printed array slides were first thawed at room temperature (RT) for 1 hour in the dark, then blocked on ice with freshly prepared blocking buffer containing 50 μM biotin, 50 mM KCl, 20% glycerol, 0.1% Triton X-100, and 25 mM HEPES (pH 7.5 in dH₂O), while agitated at 20–40 rpm using an orbital shaker. Following blocking, slides were washed three times with PBST (1× PBS, 1% Tween-20) for 5 minutes at RT, and dried by centrifugation at 500 × g for 5 minutes at RT. To prevent cross-contamination between arrays, each slide was tightly sealed using a 4-plex array gasket (Grace Bio-Labs, Oregon, USA). Diluted patient serum (1:100 dilution in PBST) was then applied and incubated on each array for 1 hour at RT with gentle agitation (20–40 rpm). After incubation, slides were washed three times with PBST for 5 minutes at RT. Bound cognate IgG Abs were detected using a Cy3-conjugated polyclonal anti-human IgG Ab (20 μg/mL; Invitrogen, Cat. no. A-21089) and IgM Abs using a Cy5-conjugated polyclonal anti-human IgM Ab (20 μg/mL; Invitrogen, Cat. no. A-21249) diluted in PBST. Arrays were wrapped in foil to protect from light and incubated for 1 hour at RT under gentle shaking (20–40 rpm). Slides were then washed three times with PBST for 5 minutes at RT. Slides were imaged using a GenePix® 4000B microarray scanner at fixed gain settings, and the resulting raw image files were extracted for downstream data processing and analysis (Fig. 1(b)). Statistical Analysis Background-corrected median net intensities were averaged across replicas, and then converted to Z-scores. Z-scores above 2 were deemed significant for subsequent analyses. The sensitivity and specificity of potential individual IgG and IgM Abs were separately evaluated using receiver operating characteristic (ROC) curve analyses, and univariate AUC values were obtained. CombiROC was then used to identify optimal antigen combinations. Multivariate ROC curves were generated and AUC values calculated to define diagnostic Ab signatures. Mann–Whitney U tests were also used to determine differences between melanoma patients and controls. These analyses were performed using the GraphPad Prism v6.0c software. A p-value of < 0.05 was defined as statistically significant. Hierarchical clustering using one minus Spearman ranking of candidate antigens was performed using the Morpheus software (https://software.broadinstitute.org/morpheus/index.html, Broad Institute, Cambridge, MA, USA).
Institutions
- Monash UniversityVictoria, Melbourne
- La Trobe UniversityVictoria, Melbourne