There were 351 non-redundant protein sequences in the plasma-only subproteome. Currently, there was no evidence that they were present in urine based on available proteomic data. Due to the profiling depth of the urine proteome is significantly exceeded that of the plasma proteome, this group of proteins is more reliable. These proteins are supposed to be difficult to pass through the kidney black box. There were only 4 proteins with experimental MWs, which were relatively small proteins. They were P11226 (26.3 kDa), P02745 (26.3 kDa), P20742 (51.6 kDa), and Q14289 (65.1 kDa). The theoretical MWs were in a range from 3 kDa to more than 400 kDa, while the theoretical PIs were in a range from 4 to 11.6. According to estimated concentration of the proteins in plasma provided by plasma proteome dataset, there were many moderate- or even high-abundant proteins that were present only in plasma but not in urine. Particularly, there were some high-abundant Ig proteins present only in plasma. This group included proteins that could not be filtered at the glomerular capillaries or filtered but reabsorbed completely back into the blood from the tubules.

There were 1,424 non-redundant protein sequences that existed in both plasma and urine. They may pass the kidney black box in various forms. There were only 15 proteins with experimental MW information available from both plasma and urine, which range from 11 to 133 kDa in plasma and 11 to 77 kDa in urine. Comparing their MWs in plasma and in urine, six proteins were within 20 % variation, suggesting that they may pass through the kidney in an intact form; four had MWs 20 % higher in plasma than in urine, and five had MWs 20 % lower in plasma than in urine. These differences reflect functions of the kidney. In addition to one protein whose theoretical MW was 3,816 kDa, the theoretical MWs of the proteins were in a range from 3 kDa to more than 600 kDa, while the theoretical PIs were in a range from 3.6 to 12.

For the kidney considered as a black box, the quantitative changes of different proteins from plasma proteome to urine proteome reflect the kidney protein handling function. Therefore, the ranking order of the plasma and urine proteins sorted by their concentrations in plasma was compared with that in urine. The estimated concentrations of the proteins in plasma were provided by plasma proteome dataset, which were used to generate their ranking order in plasma. Many urine proteins were collected from a large-scale dataset identified in our institution using MASCOT search engine (data not published). The MASCOT search engine has been incorporated the exponentially modified protein abundance index (emPAI), which offers approximate, label-free, relative quantitation of the proteins in a mixture based on protein coverage by peptide matches [2]. Each identified urine protein had an emPAI value, which can be used to approximately estimate the absolute protein contents in urine. The emPAI values of these plasma and urine proteins were extracted from the large-scale dataset. These proteins were sorted from most to least abundant. Proteins not identified in the large-scale dataset were at the end. The ranking order of these proteins approximately represents their abundance in human urine. After comparing their ranking order in plasma with that in urine, we found that the two ranking orders did not correspond well with each other. Many proteins have a significantly changed ranking order when they passed through the kidney. This suggested that the kidney performed different handling functions for the different proteins.