Searching for biomarkers of acute rejection in renal transplant recipients : development and optimization of a urinary proteomic approach

Abstract

In Paper I a method for sample preparation of urinary proteins was developed and optimized. The main steps were desalting/enrichment by cut-off centrifugation (5 kDa), albumin depletion and tryptic digestion followed by 2D-LC-MS. Emphasize was put on maximizing protein recovery and improving downstream compatibility. A 2D-separation approach combining ZIC-HILIC and RP was also tested and gave a separation system with a high degree of orthogonality. Finally, the suitability of the method was assessed in a comprehensive proteomic experiment using urine from renal transplants. A high number of urinary proteins were identified and the variability of the whole method was in the range of 11 to 30 % (RSD). In Paper II enzymatic digestion using immobilized trypsin beads was investigated. Evaluation of different reactor formats and conditions like digestion temperature and reaction time were carried out to find the optimum setup. Larger proteins demanded longer digestion time and BSA was digested in 89 minutes at 37 °C. The optimized procedure was compared with digestion in-solution with respect to time consumption, sequence coverage and degree of unsuccessful cleaving. The final digestion set-up was carried out in urine samples yielding good signal intensities and reproducibility. In Paper III a multidimensional on-line system including Strong Anion Exchange Chromatography (SAX) separation of native proteins, reduction, alkylation, C4 separation and tryptic digestion of the alkylated proteins followed by MS detection was tested as an alternative to the off-line method developed. Proof of concept was shown and the efficiency of the reduction and alkylation was equivalent with established methods. On-line tryptic digestion was satisfactory for several proteins but needs further optimization to cover the full proteome. The system was evaluated using both model proteins and human urine sample and has shown potential as a tool to identify biomarkers offering short analysis time and minimum manual sample handling. In Paper IV proteolytic 18O-labeling of peptides was investigated and improved in order to optimize the labeling efficiency and accelerate the process. Optimization was carried out using BSA and cyt c as model proteins and the best efficiency was achieved at pH 6 yielding complete labeling during 2 hours at 37 °C with immobilized trypsin beads. An approach integrating tryptic digestion developed in Paper II with 18O-labeling, both using immobilized trypsin beads was also developed. This enabled tryptic digestion and 18O-labeling by 3.5 hours, without any sample transfer steps. The procedure was evaluated in urine, first by spiking it with model proteins and then by analyzing the true human urinary proteome after implementation in the workflow developed in Paper I. In Paper V the method developed in Paper I, II and IV was used to identify urinary proteins associated with acute rejection episodes in kidney transplanted patients. A large degree of regulation was found and 11 proteins were identified as up-regulated in the rejection group (n=6) compared with the control group (n=6) according to strict criteria. The up-regulated proteins could be grouped by biological function in 2 main groups; proteins involved in growth and proteins involved in immune response. The growth proteins were statistically significantly up-regulated (P=0.03) while the immune proteins only showed an overall trend towards up-regulation in the rejection group compared with the control group (P=0.13).