|Home | About | Journals | Submit | Contact Us | Français|
Small ubiquitin-related modifier (SUMO) is an ubiquitin-like protein that is covalently attached to a variety of target proteins. Unlike ubiquitination, sumoylation does not target proteins for proteolytic breakdown, but is instead involved in regulating a variety of different protein functional properties including protein-protein interactions and subcellular targeting, to name a few. Protein sumoylation has been particularly well characterized as a regulator of many nuclear processes as well as nuclear structure, making the characterization of this modification vital for understanding nuclear structure and function. Because sumoylation plays an important role in regulating so many important cellular processes, there has been intense interest in identifying new proteins that are targets of this modification and determining what role it plays in regulating their functions. This chapter presents methodologies for determining whether a particular protein is a substrate of sumoylation, and for identifying the lysine residue(s) where the modification occurs.
Small ubiquitin–related modifier (SUMO) was discovered as a modifier of mammalian proteins in 1997 (1–2). SUMO has since been demonstrated to be a modifier of many important proteins, giving this modification a vital role in modulating a large number of important cellular processes (3–5). SUMO proteins are very similar to ubiquitin structurally, but sumoylation does not promote degradation of proteins and instead regulates key functional properties of target proteins. These properties include subcellular localization, protein partnering, and transactivation functions of transcription factors, among others (3–5). Protein sumoylation plays a particularly vital role in regulating many important processes occurring in the nucleus, and although sumoylation can be found on proteins that exist in a number of cellular compartments, most of the sumoylation characterized to date occurs on nuclear proteins (4, 5). Indeed, proteins of the SUMO conjugation machinery have been found to be localized to nuclear pore complexes, in addition to other locations in the nucleus.
SUMO proteins are covalently attached to lysine residues of proteins, which are generally found within the consensus motifΨKXE where Ψ is a hydrophobic amino acid and X is any residue. Like ubiquitination, the covalent attachment of SUMO to other proteins involves a series of enzymatic steps (see Figure 1), but the proteins involved are distinct from those in the ubiquitin conjugation pathway. First, the SUMO proteins have to undergo proteolytic processing near their C-terminal end to form the mature proteins, a step which is performed by SUMO proteases (Ulp’s). These proteases are dual-functional, as they are also responsible for cleaving SUMO groups from substrate proteins by cleaving the isopeptide bonds by which they are joined (3–5). The mature processed SUMO protein is covalently attached via a thioester bond to the Uba2 subunit of the heterodimeric SUMO E1 activating enzyme in an ATP-dependent reaction (6–9). The SUMO moiety is transferred from the E1 to ubc9, the SUMO E2 enzyme, which then binds to the ΨKXE consensus sequence in target proteins and forms an isopeptide bond between the ε-amino group of the lysine within this sequence and the carboxyl group of the C-terminal glycine of the SUMO polypeptide (10–13). SUMO E3 proteins have been identified that enhance the efficiency of SUMO attachment by interacting with both ubc9 (the E2 enzyme) and the substrate, thereby acting as bridging factors (3–5). Vertebrate cells contain three SUMO paralogs. SUMO-2 and SUMO-3 are very similar to each other in sequence, and have approximately 50% sequence identity with SUMO-1, which is the best characterized of the three vertebrate SUMO proteins.
In this chapter we describe two different experimental approaches for determining whether a specific protein is sumoylated. One method employs immunoprecipitation of the protein of interest, either endogenous or transfected epitope-tagged protein, followed by Western blot with SUMO antibodies. The second method involves incubating the protein, either as a 35S-labeled in vitro translation product or purified recombinant protein, with a reconstituted in vitro sumoylation enzymatic reaction, followed by SDS-PAGE and autoradiography or Western blot, respectively, to look for the appearance of higher molecular weight bands indicative of sumoylation. By comparing wild type protein constructs with those containing non-sumoylatable arginine substitutions of candidate target lysine residues, these protocols can also allow identification of those lysine residues where SUMO attachment actually occurs in a given protein. This information then provides the critical reagents for testing the functional consequences of blocking sumoylation of that particular protein. To illustrate the types of data that can be obtained using these methodologies we present figures showing the results of immunoprecipitation and in vitro sumoylation analyses of a transcription factor we study called HSF2.
In this protocol, proteins to be tested for sumoylation are immunoprecipitated using lysis buffers designed to block the action of desumoylating enzymes (see Note 1) and then subjected to Western blot analysis using anti-SUMO antibodies to look for the appearance of a band with a size consistent with a sumoylated form of the protein. Although the theoretical molecular weight of the SUMO proteins is approximately 11 kDa, the size increase for each SUMO added on SDS-PAGE is typically in the range of 15–17 kDa. In the case of a protein with multiple sumoylation sites, or where SUMO chains form on a lysine target site (see Note 2), multiples of this size increase are expected, sometimes yielding very large shifts in mobility. Multiple bands representing different sumoylation states of the protein are also possible. This approach can be used to analyze endogenous proteins, or transfected proteins containing an epitope tag (FLAG, myc, etc.) (see Note 3). The transfection approach can also be used to determine the lysine residue(s) to which the SUMO group is attached, by comparing the sumoylation of wildtype constructs to ones in which candidate lysines have been changed to non-sumoylatable arginines. Two different lysis conditions are described, one using SDS to inhibit desumoylase enzymes and the other containing N-ethylmaleimide, a chemical inhibitor of these enzymes.
In this protocol the protein of interest is in vitro translated (typically with 35S-methionine incorporation) and then incubated in a reaction containing the SUMO E1 and E2 enzymes and SUMO-1, followed by SDS-PAGE and autoradiography to determine whether a lower mobility band appears that would be consistent with a sumoylated form of the target protein. Because of the high concentrations of SUMO E1 and E2 enzymes used in this in vitro sumoylation reaction, the need for SUMO E3 proteins is diminished and thus sumoylation can be detected without their addition. The expression and purification of the recombinant proteins required for the in vitro sumoylation assay is described in subheadings 3.2.1, and the protocol for performing assay itself is described in subheading 3.2.3.
The SUMO proteins are expressed and purified as fusion proteins with GST or 6xHis at the N-terminal end of the SUMO proteins, and these affinity tags do not need to be removed prior to using these SUMO proteins for the sumoylation reaction. The size difference of GST-SUMO vs. 6xHis-SUMO can also provide a useful control for the sumoylation reaction as it yields a predictable size difference between the sizes of the sumoylation products (e.g. see Figure 3). The SUMO E1 is a heterodimer of SAE1/SAE2, and is active when expressed in E. coli from a bicistronic contruct of GST-SAE2 and untagged SAE1; the two proteins complex and can be purified using glutathione-agarose affinity chromatography (14). The SUMO E2 can be expressed which appears to be more active, at least in our hands when the GST tag is removed by thrombin cleavage. The following is the general protocol for expressing and purifying these recombinant proteins needed for the in vitro sumoylation assay. Once purified these recombinant proteins can be stored at −80°C for extended periods of time.
The in vitro sumoylation assay uses 35S-methionine radiolabelled protein as the substrate. We generate radiolabelled protein using the TNT T7 Quick for PCR DNA kit (Promega, Madison WI) following the manufacturers instructions. Described below is the SUMO modification procedure to be utilized with radiolabelled translated proteins.
We are very grateful to Mike Matunis (Johns Hopkins), Ron Hay (University of St. Andrews), and Chris Lima (Sloan Kettering Institute) for providing constructs and reagents.
1SUMO-modified proteins are highly susceptible to SUMO proteases. The SDS in the lysis buffer described in this protocol inactivates the SUMO proteases allowing for easier detection of sumoylated proteins. However, a common complication with the SDS-lysis method is that the cell lysates tend to be very viscous and sticky due to genomic DNA in the lysate. This problem is remedied by brief sonication which shears the DNA and makes the samples easier to manipulate. SUMO proteases can also be inhibited by the addition isopeptidase inhibitor N-ethylmaleimide (20mM) to standard lysis buffers such as NP-40 lysis buffer (50 mM Tris-HCl (pH 8.0), 150 mM NaCl, 1% NP-40) if another lysis buffer besides the SDS-lysis is more desirable.
2The size of putative sumoylated forms of a protein on SDS-PAGE will depend on how many different SUMO attachment sites the protein has. In addition, SUMO-2 and SUMO-3 have been reported to form polymeric chains reminiscent of ubiquitin, which could result in large increases in size for a sumoylated protein on SDS-PAGE compared to the non-sumoylated form (14). Thus, it is possible to observe bands that are multiples of the approximate 15–17 kDa size of each SUMO unit, as well as multiple bands representing different sumoylation states.
3Investigating sumoylation may also be done using cells transfected with fusion-tagged plasmid constructs of the protein thought to be sumoylated with immunoprecipitation utilizing fusion tag antibodies which are available from commercial sources (e.g. GFP, FLAG, Myc, and 6xHis tags). In these types of experiments it is advisable to co-transfect the cells with a SUMO expression construct (often this is epitope-tagged) to ensure that sufficient SUMO protein is present in the transfected cells to allow for efficient sumoylation of the transfected target protein being tested.