Coskun Guzel

Biomarker research in biofluids is carried out on a large scale and performed, to a lower extent, on tissue level related to the origin of the disease. This thesis describes a method to identify biomarkers in affected tissue. As a proof of principle two diseases were investigated, namely preeclampsia and cervical cancer.

Preeclampsia is a pregnancy-specific syndrome associated with serious perinatal and maternal morbidity and mortality. Endovascular remodeling and invasion of the spiral arteries are impaired, which results in reduced and irregular placental perfusion. The maternal spiral artery remodeling leads to placental oxidative stress, resulting in endoplasmic reticulum-related impaired protein synthesis. We previously found two calcyclin (SNOOAS) peptides significantly higher expressed in preeclampsia measured by MALDI mass spectrometry. This observation was further confirmed through Multiple Reaction Monitoring (MRM), as described in Chapter 2. We extended these MRM measurements to Formalin-Fixed Paraffin-Embedded (FFPE) placental material. Laser capture microdissected cells were obtained from controls and preeclamptic FFPE placental material. The number of protein identifications obtained from equal areas of microdissected paired FFPE and frozen placenta tissue was determined by Orbitrap mass spectrometry and compared with each other. Two SNOOAS peptides were detected by MRM and quantified in the microdissected cells from preeclampsia patients and controls. SNOOAS levels were significantly higher in trophoblast cells of preeclampsia patients compared to controls. An overlap of 60% of identified proteins was observed between paired FFPE and frozen tissue.

In Chapter 3, we described through immunohistochemistry, that preeclamptic trophoblast cells stained heavily with antibodies specific for SNOOAS in contrast to the preterm controls. To extrapolate these findings, we measured SNOOAS and its binding partner heat shock protein 90 (HSP90) in serum samples described in Chapter 4. A Parallel Reaction Monitoring (PRM) assay was developed for quantitative measurements of SNOOAS and HSP90 in serum using stable isotope-labeled peptides. Sera from preeclamptic women throughout pregnancy from the first trimester to term were compared with sera of age-matched pregnant controls. To our knowledge, the two proteins SNOOAS and HSP90 were for the first time studied in all trimesters of pregnancy in patients with preeclampsia and normotensive controls. Both interacting proteins were notably changed in preeclamptic patients as compared to controls; however, without differences in the onset of preeclampsia. SNOOAS was already decreased before the onset of preeclampsia in the second trimester, and HSP90 was strongly increased in the third trimester. This suggests that these proteins may play a role in the pathogenesis of preeclampsia and ought to be investigated in large cohort studies to determine the clinical value. SNOOAS and HSP90 are both expressed in trophoblast cells, as confirmed by double immunofluorescence staining.

In Chapter 5, we described an HSP90a SRM and PRM assay that was compared with an HSP90a immunoassay. We developed 2D-LC-MS/MS-based SRM and PRM assays to measure HSP90a in serum and compared the results with a commercially available HSP90a immunoassay (ELISA). Serum samples were trypsin-digested and fractionated by SCX chromatography prior to SRM and PRM measurements. PRM data obtained by high-resolution mass spectrometry correlated better with ELISA measurements than SRM data measured on a triple quadrupole mass spectrometer. All three methods (SRM, PRM, ELISA) were able to quantify HSP90a in serum at the nanograms per ml level, but the use of PRM on a high-resolution mass spectrometer reduced variation. To rule out that the observed differences in SRM and PRM were due to different mass spectrometry systems, the SCX fractions were measured on the same high-resolution instrument (Orbitrap) in the ion trap mode (IT-PRM). These measurements showed that intense co-eluting signals were present in the SRM method, but these interfering peaks were mostly eliminated in the high-resolution PRM mode. Results showed that it is possible to measure nanograms per ml levels of HSP90a in a reproducible, selective, and sensitive way using PRM in serum. This opens the possibility to quantify low levels of multiple proteins in complex samples based on a fractionation strategy on tryptic peptides, followed by PRM.

In cervical cancer research (Chapter 6), we studied the landscape of proteomes found in cervical tissue through laser capture microdissection (LCM). Cervical cancer is characterized by a well-defined pre-malignant phase, cervical intraepithelial neoplasia (CIN). Identification of high-grade CIN lesions by population-based screening programs and their subsequent treatment has led to a significant reduction of the incidence and mortality of cervical cancer. Cytology-based testing of cervical smears is the most widely used cervical cancer screening method, but is not ideal, as the sensitivity for detection of CIN2 and higher (CIN2+) is 50%. Therefore, more sensitive and specific biomarkers for cervical cancer and its precancerous stages are needed.

By comparing early stage, late stage cervical squamous cell cancer tissue with epithelium and stromal cells from healthy subjects, 30 significant differentially expressed proteins were found after correction for multiple testing (Bonferroni correction). Among these 30 proteins, the MCM protein family was strongly up-regulated (up to 32-fold increase). The presence of four proteins (MCMP, CEACAMS, SNOOP, ICAMN) that were found highly significant in cervical cancer tissue using the LCM approach was confirmed in digests of whole tissue lysates by PRM measurements. We assume that the up-regulation of MCM proteins is critical for tumor progression of cervical cancer. We propose that MCMP can be implemented together with the hrHPV testing in the current cervical cancer screening program to reach higher initial specificity and to decrease unnecessary referrals for colposcopy.