Genomics Day 2005: Genomics Day 2005: Public Health Genomics at CDC
Group 2: Biomarker Discovery
Protein profiling in cervical mucous samples for biomarker discovery
Rollin DD, Whistler T, Lee DR , Vernon SD , Gurbaxani B, Unger ER.
Division of Viral and Rickettsial Diseases, NCID, CDC
The purpose of this work was to evaluate surface enhanced laser desorption and ionization-time-of-flight (SELDI-TOF) mass spectrometry protein profiling of cervical mucous. Cervical mucus is an ideal choice for cervical cancer marker discovery because it can be collected non-invasively and is produced in the affected region and therefore likely to include proteins produced by the lesion and the host response to the lesion. For this pilot study we used samples collected as part of a larger study of women evaluated at colposcopy for possible cervical intraepithelial neoplasia (CIN). We included 30 patients with no dysplasia identified (CIN 0) and 30 patients with biopsy-proven CIN 3 matched for age, race and HPV status. (All women were high risk-HPV positive, 15 in each group were HPV 16 positive and 15 were other HR-HPV positive.) We chose to match on HPV because we wanted the markers to reflect disease not simply HPV infection. The samples were collected using Weck-Cel sponges and stored at –80°C. We used a mammalian protein extraction buffer followed by centrifugation to extract the proteins from the Weck-Cel and added a protease inhibitor cocktail. The protein concentration of each sample was determined by Bradford protein assay and SELDI-TOF profiling was performed on a fixed protein quantity, using two ProteinChip chemistry surfaces (strong anion exchanger, SAX-2 and metal-binding, IMAC3-Cu). Data processing included spectral calibration, baseline subtraction, and normalization to total ion current. Ciphergen's Biomarker Wizard was used to identify “clusters” or groups of peaks of similar mass, for the sample groups (either CIN or HPV status).
The IMAC chip provided the best resolution of multiple peaks over the 2,000 to 20,000 M/Z range with approximately 140 peaks identified. Protein profile complexity appeared to be adequate to determine statistically significant differences between groups. Several protein peaks were identified that discriminated between women with no CIN and women with CIN III. Three peaks in the 10,000 to 10,600 Da range show extremely good discrimination of CIN 3 (p<0.004). For these a search of the Swiss-Prot data base for a protein of 10,200 ± 300 Da returned 59 possibilities, some of which included HPV E7 (pI ±4.5) and Follicle-stimulating hormone alpha chain (pI 8.4), a glycosylated protein which is indicated by the broad peak profile.
Supported in part by the National Cancer Institute's Early Detection Research Network (EDRN), Interagency Agreement Y1-CN-0101-01.