A Method for Assaying Deubiquitinating Enzymes

1 ,2 ). Ub is covalently ligated to a variety of intracellular proteins through an isopeptide linkage. Ubs by themselves or that have already been conjugated to proteins may also be ligated to additional Ub molecules to form branched poly-Ub chains. This ubiquitination has been implicated in the regulation of diverse cellular processes, such as selective protein breakdown, cell cycle regulation, and stress response (3 -5 ). In addition, protein ubiquitination in vivo is a dynamic, reversible process that is under control responding to external stimuli, such as heat shock and starvation (6 -8 ). Therefore, the enzymes that proteolytically remove Ub from Ub-protein conjugates should be of importance in maintaining the steady-state levels of free Ub for its diverse cellular functions.

Ubs are encoded by two distinct gene classes. One is a poly-Ub gene that encodes a poly-protein of tandemly repeated Ubs (9 ,10 ). The other encodes a fusion protein in which a single Ub is linked to a ribosomal protein consisting of 52 or 76-80 amino acids. The transient association of Ub with the ribosomal proteins has been suggested to promote their incorporation into ribosomes (11 ). Therefore, proteolysis at the peptide bonds between Ub and the extension proteins is required for generation of ribosomal proteins for ribosome biogenesis as well as of free Ubs.

Deubiquitinating enzymes (DUBs) are known to consist of a large protein family in eukaryotes. For example, the budding yeast has 17 genes for DUBs (4 ). Moreover, so far more than 60 full-length DUB sequences have been identified in eukaryotes (12 ). However, comprehensive searches for DUBs, particularly in mammalian cells, were hampered due to the lack of rapid and efficient methods for assaying the enzymes. We have recently reported that ¹²⁵ I-labeled Ub-PESTc serves as an excellent substrate for the sensitive and quantitative assay of various DUBs (13 ,14 ). Using this assay, we have also isolated a number of novel DUBs in chick skeletal muscle and yeast (13 , 15 -18 ). Here we describe the detailed protocol for assaying DUBs using ¹²⁵ I-labeled Ub-PESTc. We also compare the sensitivity of this method to that using a fluorogenic peptide substrate, carbobenzoxy-LRGG-7-amido-4-methylcoumarin (Cbz-LRGG-AMC), that has also been used as a substrate for DUBs (19 ).

20 ). The purified Ub-specific protease-6 (yUBP-6) in yeast and cUBP41, Ub C-terminal hydrolase-1 (cUCH-1), cUCH-6, and cUCH-8 in chick skeletal muscle were prepared as described (13 , 15 -18 ). Ub-PESTc was purified from an E. coli strain AR13 carrying pNMHUB-PESTc as described by Yoo et al. (21 ). Ub-aldehyde was prepared as described (13 ). Cbz-LRGG-AMC was kindly provided by Dr. K. Tanaka (Tokyo Metropolitan Institute of Medical Science, Japan).

Ub-PESTc was radiolabeled with Na¹²⁵ I using IODO-BEADS (Pierce) by following the procedure recommended by the manufacturer and Markwell (22 ). One IODO-BEAD was incubated in 0.2 ml of 0.1 M Tris-HCl (pH 7) in a microfuge tube for 5 min at room temperature. After incubation, 0.2 mg of the purified Ub-PESTc and 200 μCi of Na¹²⁵ I were added to the tube and further incubated for the next 15 min. The radioiodinated Ub-PESTc (i.e. , the liquid portion) was then removed and subjected to gel filtration on a Sephadex G-10 column equilibrated with the Tris buffer to remove free iodine. Upon the iodination procedure, approximately 85% of ¹²⁵ I was incorporated into Ub-PESTc.

Reaction mixtures (final 0.1 ml) contained proper amounts of the purified DUBs, 10-20 μg of ¹²⁵ I-labeled Ub-PESTc, 0.1 M Tris-HCl (pH 7.8), 1 mM EDTA, 1 mM dithiothreitol, and 5% (v/v) glycerol. After incubation of the mixtures for appropriate periods at 37 °C, the reaction was terminated by adding 50 μl of 40% (v/v) trichloroacetic acid and 50 μl of 1.2% (w/v) bovine serum albumin. The samples were vortexed and centrifuged for 10 min at 10,000 x g using a microfuge, and aliquots (0.1 ml) of the resulting supernatants were counted for radioactivity using a gamma-counter (13 ).

Reaction mixtures (final 0.1 ml) contained appropriate amounts of the purified DUBs, 0.2 mM Cbz-LRGG-AMC, 0.1 M Tris-HCl (pH 7.8), 1 mM EDTA, 1 mM dithiothreitol, and 5% (v/v) glycerol. After incubation of the mixtures for various periods at 37 °C, the reaction was terminated by adding 0.1 ml of 1% (w/v) SDS and 0.8 ml of H₂ O. Release of free AMC by the enzyme reaction was determined by measuring its fluorescence at 380 nm (excitation) and 440 nm (emission) (23 ). Proteins were quantified as described by Bradford (24 ).

Polyacrylamide gel electrophoresis in the presence of SDS and 2-mercaptoethanol was performed using Tris-Tricine buffer as described by Schägger and von Jagow (25 ). The discontinuous slab gels contained 4, 10 and 16% polyacrylamide to improve resolution of small proteins. The sample buffer contained 150 mM Tris-HCl (pH 6.8), 1.5% (w/v) SDS, 2% (v/v) 2-mercaptoethanol, 0.002% (w/v) bromophenol blue and 7% (v/v) glycerol. After electrophoresis, the gels were stained with Coomassie blue R-250 or subjected to autoradiography.

21 ) have constructed Ub-αNH-peptide extensions containing "PEST" sequences. Of these, Ub-PESTc contains a peptide extension of 18 amino acids, which carries a single tyrosine residue that can be radioiodinated and is short enough to be released as an acid-soluble product when incubated with DUBs. Moreover, they have reported that the recombinant Ub-PESTc protein can be easily purified by heating, such as at 85 °C, because fusion of the short peptide to Ub does not alter the heat resistance of Ub molecule. In addition, they have demonstrated that Ub-PESTc appears correctly processed to yield free Ub upon incubation with the chromatographic fractions of rabbit reticulocytes. Therefore, we chose Ub-PESTc for labeling with ¹²⁵ I and hence for using the labeled protein as a substrate for the assay of DUBs. Overall procedure for the enzyme assay is summarized in Fig. 1

By following the protocol, we determined the activity of YUH1 by measuring its ability to release radioactive PESTc into an acid-soluble form from ¹²⁵ I-labeled Ub-PESTc. The purified enzyme was incubated with the radioiodinated substrate for various periods in the absence and presence of Ub-aldehyde, which is known as a specific inhibitor of DUBs (26 ). Table 1 shows that the acid-soluble radioactivity increases in an incubation time-dependent fashion and this increase can be completely blocked by the treatment of Ub-aldehyde. Since YUH1 as well as other DUBs are known to specifically cleave the carboxyl side of the C-terminal Gly residue of Ub, the acid-soluble products should represent the PESTc peptide released from the Ub-peptide extension.

13 ). Thus, it is clear that the acid-soluble radioactivity represents PESTc released from ¹²⁵ I-labeled PESTc by the action of YUH1.

Surprisingly, however, the band intensity of Ub upon the autoradiography was far lower than the others, despite the fact that Ub itself has a single tyrosine residue at the 59th position that can also be labeled by ¹²⁵ I. Typically, chloramine T is used for radioiodination of Ub molecules. In our studies, we used IODO-BEADS, in which chloramine T is immobilized on non-porous polystyrene beads. Perhaps, the tyrosine residue in Ub is not accessible to electrophilic idodine species (i.e. , I⁺ ), which are produced by the immobilized chloramine T, due to structural barrier, unlike that in PESTc, that is fused to the flexible C-terminal region of Ub. In any event, we could obtain ¹²⁵ I-labeled Ub-PESTc, in which PESTc was almost exclusively radioiodinated. This fortuitous finding allowed us to determine rapidly the activity of YUH1 as well as of other DUBs and to quantify precisely cleavage products by simple measurement of the radioactivity released into the acid-soluble products.

Using the assay method, we have previously purified several novel DUBs from chick skeletal muscle and yeast, including cUBP41 (15), cUCH-1 (16), cUCH-6 (13), cUCH-8 (17), and yUBP-6 (18), using conventional chromatographic procedures. To compare the activities of these DUBs against ¹²⁵ I-Ub-PESTc, the same amount of each enzyme was incubated with the substrate for various periods. Fig. 3 shows that the hydrolytic rates differ markedly from each other. Among the DUBs, cUCH-6 hydrolyzed Ub-PESTc most rapidly. The K_m values for the enzymes were also different, ranging from 5 to 65 μM (e.g., 5.1 μM for cUCH-6 and 64.5 μM for yUBP-6.)

Ub ends with the amino acid sequence of -RLRGG⁷⁶ . Based on the C-terminal sequence, Stein and coworkers (19 ) have synthesized a variety of fluorogenic peptides by conjugating AMC to the C-termini of the peptides and used them as substrates for determination of the specificity of isopeptidase T, which exists in all mammalian cells and hydrolyzes the isopeptide linkages of poly-Ub chains (27 ). Of these, Cbz-LRGG-AMC was used in the present studies to compare its sensitivity to the purified DUBs to that of ¹²⁵ I-Ub-PESTc. Table 2 shows that all of the DUBs have at least three order higher specific activities against ¹²⁵ I-Ub-PESTc than those against Cbz-LRGG-AMC, indicating that the assay method using ¹²⁵ I-Ub-PESTc is much more sensitive for determining the activities of DUBs than that using the fluorogenic peptide substrate. In addition, we have recently found that the extracts of E. coli by themselves are capable of releasing AMC from the peptide substrate (data not shown), implying that E. coli contains a protease(s), which specifically cleaves off AMC from the peptide substrate. On the other hand, the same extracts could not hydrolyze ¹²⁵ I-Ub-PESTc at all. Thus, the protease(s) is likely to interfere with the assay for overproduced eukaryotic DUBs in E. coli cells, when Cbz-LRGG-AMC was used as a substrate. Furthermore, extracts prepared from most of eukaryotic cells also contain a protease(s) that rapidly cleaves Cbz-LRGG-AMC but not ¹²⁵ I-Ub-PESTc. Thus, the assay method using ¹²⁵ I-Ub-PESTc should be appropriate for identification of DUBs in eukaryotic cells and their clones expressed in E. coli . In fact, we were able to identify and partially purify at least 10 different DUBs from the extract of chick skeletal muscle (13 ) and to purify several novel chick DUBs cloned and expressed in E. coli (15 ,28 ).

29 ) have developed a new assay method for DUBs based on the substrate Ub C-terminal AMC (Ub-AMC). They showed that the rate constants (kc/K_m ) for the hydrolysis of Ub-AMC are 10⁴ - and 10⁷ -fold over those for the cleavage of Cbz-LRGG-AMC for isopeptidase T andUCH-L3 respectively. UCH--L3 is a 26 kDa Ub C-terminal hydrolase (30 ) isolated from rabbit reticulocytes. However, it has not yet been tested whether Ub-AMC is susceptible to any protease in bacterial or eukaryotic cells other than DUBs.