Traditionally, biochemical purification has been one of the most powerful methodologies for the molecular identification of endogenous enzymes and biologically active substances from complex sources. However, combining traditional biochemistry with proteomics has been shown to resolve the challenges associated with physico-chemical stability and solubilization — especially with membrane-bound proteins — faster and more efficiently.1
Researchers led by Kazuishi Kubota, Senior Director, Functional Genomics & Proteomics Research Group at Daiichi Sankyo RD Novare (Japan) have now applied the technique of proteomic correlation profiling to the identification of drug-metabolizing enzymes. The team used chromatography rather than centrifugation as a separation method, and kinase activity as a basis for comparison. It is the first study to use the technology for an enzyme class other than kinases, with a tissue extract as the purification source.2 In a previous study, the researchers successfully identified a kinase responsible for phosphorylation of peptide substrates in a single-step chromatography.1
The biological material is first fractionated by column chromatography, followed by calculation of each protein’s correlation coefficient between the enzyme activity and proteomic profile of the fractions. The researchers then select possible enzyme candidates among proteins with a high correlation value, using informatics tools for domain predictions. The technology obviates the need to observe the separated bands correlating with enzyme activity. The streamlined proteomic correlation profiling protocol entails fewer purification steps and very little starting material, as compared to traditional biochemistry.
Incorporating label-free, intensity-based protein quantitation as described in the MaxQuant software suite,3,4 the investigators illustrate their unique proteomic correlation profiling approach to identify the drug-metabolizing enzyme alkaline phosphatase, non-specific protein (ALPL). ALPL is a phosphatase of CS-0777 phosphate (CS-0777-P), a selective sphingosine-1-phosphate receptor 1 modulator used for treating multiple sclerosis. Identification of the CS-0777-P phosphatase in the kidney holds the key to fully understanding the pharmacokinetics and pharmacodynamics of CS-0777. The Daiichi Sankyo researchers say that the phosphorylated CS-0777-P has more than 300-fold selectivity for S1P1 receptors, compared to S1P3 and weaker effects on S1P5 and no activity on S1P2.5 In multiple sclerosis patients, single oral doses of CS-0777 caused dose-dependent decreases in circulating lymphocytes, especially the CD4+ T cells. Multiple sclerosis, the most common autoimmune disorder of the central nerve system, results from demyelination and scarring of the axons of the brain and spinal cord.
In their previous work, the Japanese researchers identified CS-0777 kinases, the activating enzymes of a prodrug, in human blood. And now, they report successful identification of the alkaline phosphatase, non-specific isozyme (ALPL), the inactivating enzyme of an active metabolite, as the major CS-0777-P phosphatase candidate in the human kidney. (The high kinase activity in blood is balanced by phosphatases.)
To identify the unknown phosphatase or phosphatases capable of dephosphorylating CS-0777-P in humans, the researchers used proteomic correlation profiling to identify proteins that co-purified with CS-0777-P dephosphorylation activity upon fractionation, after fractionating the CS-0777-P phosphatase activity from 1.1 g of human kidney.
The proteomic correlation profiling experiment assumes that protein quantity correlates with protein activity. The chromatographic samples were subjected to proteomic analysis to quantify all proteins, in parallel with a biological assay (CS-0777-P phosphatase activity). Strong correlation with a specific enzyme activity suggests that the protein was linked to the biological activity. The researchers spiked all fractions exhibiting target enzyme activity, following gel filtration chromatography and several earlier fractionation steps, with equal amounts of Bovine Serum Albumin (BSA) as the internal standard, and subject to trypsin digestion prior to liquid chromatography tandem mass spectrometry (LC-MS/MS) analysis. The total protein concentration was determined by modified Bradford protein assay (Coomassie Plus Protein Assay, Thermo Scientific), using BSA as a standard protein. The researchers used the LTQ-Orbitrap (Thermo Scientific) equipped with an Agilent 1100 LC system, which was modified to a 200-300 nL/min flow rate by an in-house flow splitter.
Proteins were identified and quantified by MaxQuant using an intensity-based, label-free algorithm. After normalization by BSA, the Pearson correlation coefficients were calculated, comparing CS-0777-P phosphatase activity levels with protein abundance profiles for each of the 266 identified proteins. The findings revealed that among the top 25 most correlated proteins, ALPL was the only candidate that had a possible phosphatase domain and a high correlation efficient (r= 0.9965), ranking second among all 266 proteins. The proteomic correlation profiling applied in the third-step fraction, anion exchange chromatography, revealed ALPL as the highest-ranked protein with a phosphatase domain; only one other protein had the domain.
The findings suggest that ALPL is the major CS-0777-P phosphatase candidate for the human kidney, based on the single broad active peak observed in all purification steps. The concentration-dependent inhibition of the CS-0777-P phosphatase activity by levamisole as an ALPL-specific inhibitor and additional immunodepletion studies also add to the evidence.
The active enzyme was identified from only about 1 g of human kidney with four purification steps, from only a 200-fold purification of human kidney extract, which clearly demonstrates the advantages of proteomic correlation profiling. However, the proteomic correlation profiling may still have a few limitations. Any post-translational modifications or multiple proteins underlying enzyme activity, and endogenous inhibitor proteins in the fractions, can interfere with the correlation. Multiple candidates can affect the ability to select candidate proteins based on protein domain information. Additional purification steps might then be required.
Nonetheless, the authors conclude with the belief that “this approach can be further extended well beyond kinases and phosphatases to a wide variety of enzyme activities, and that this study in particular establishes a foundation for proteomic correlation profiling to be used as a general method.”
1. Kubota, K., et al. (2009) “Sensitive multiplexed analysis of kinase activities and activity-based kinase identification,” Nature Biotechnology, 27 (pp. 933–940), doi: 10.1038/nbt.1566.
2. Sakurai, H., et al. (2013, May 14) “Identification of a metabolizing enzyme in human kidney by proteomic correlation profiling,” Molecular and Cellular Proteomics, doi: 10.1074/mcp.M112.023853.
3. Cox, J., et al. (2009) “A practical guide to the MaxQuant computational platform for SILAC-based quantitative proteomics,” Nat. Protoc. 4 (pp. 698–705), doi: 10.1038/nprot.2009.36.
4. Cox, J., et al. (2011) “Andromeda: a peptide search engine integrated into the MaxQuant environment,” Journal of Proteome Research, 10 (pp. 1794–1805), doi: 10.1021/pr101065j.
5. Nishi, T., et al. (2011) “Discovery of CS-0777: A Potent, Selective, and Orally Active S1P1 Agonist,”
ACS Medicinal Chemistry Letters, 2 (pp. 368–372), doi: 10.1021/ml100301k.
Post Author: Sridhar Nadamuni.