(Received for publication, February 28, 1997, and in revised form, March 23, 1997)
From the Department of Pathology, Harvard Medical School, Boston, Massachusetts 02115
The cellular protein E6AP functions as an E3 ubiquitin protein ligase in the E6-dependent ubiquitination of p53. E6AP is a member of a family of functionally related E3 proteins that share a conserved carboxyl-terminal region called the Hect domain. Although several different E2 ubiquitin-conjugating enzymes have been shown to function with E6AP in the E6-dependent ubiquitination of p53 in vitro, the E2s that cooperate with E6AP in the ubiquitination of its normal substrates are presently unknown. Moreover, the basis of functional cooperativity between specific E2 and Hect E3 proteins has not yet been determined.
Here we report the cloning of a new human E2, designated UbcH8, that was identified in a two-hybrid screen through specific interaction with E6AP. We demonstrate that UbcH7, an E2 closely related to UbcH8, can also bind to E6AP. The region of E6AP involved in complex formation with UbcH8 and UbcH7 was mapped to its Hect domain. Furthermore, we show that UbcH5 and UbcH6, two highly homologous E2s that were deficient for interaction with E6AP, could associate efficiently with another Hect-E3 protein, RSP5. Finally, only the E6AP-interacting E2s could function in conjunction with E6AP in the ubiquitination of an E6 independent substrate of E6AP, whereas the noninteracting E2s could not. Taken together, these studies demonstrate for the first time complex formation between specific human E2s and the Hect domain family of E3 proteins and suggest that selective physical interaction between E2 and E3 enzymes forms the basis of specificity for functionally distinct E2:E3 combinations.
Ubiquitin-dependent proteolysis constitutes a major
pathway in the cell for selective protein degradation (1-3). The
covalent attachment of multiple ubiquitin molecules to lysine residues of a target protein serves to signal its recognition and rapid degradation by the 26 S proteasome. Ubiquitin conjugation can also
result in nonproteolytic modification of target proteins (4-7).
Ubiquitination of a protein substrate requires the concerted action of
three classes of enzymes; the ubiquitin activating enzyme E11 initially activates ubiquitin in an
ATP-dependent reaction through the formation of a thiol
ester bond between the carboxyl terminus of ubiquitin and the thiol
group of a specific cysteine residue of E1. Ubiquitin is then
transferred to a specific cysteine residue on one of several
ubiquitin-conjugating enzymes (Ubcs or E2s). E2 enzymes in turn may
transfer the ubiquitin either directly to a substrate or to the final
class of enzymes known as ubiquitin protein ligases (or E3s). The E3
enzymes catalyze the formation of an isopeptide bond between the
carboxyl terminus of ubiquitin and the -amino group of lysine
residues on a target protein (3, 8, 9). A substrate may be multiply
ubiquitinated through the attachment of additional ubiquitin molecules
to specific lysine residues (lysine 48 or 63) of ubiquitin itself,
although the processive nature of a multiubiquitination reaction is
presently unclear (3, 4, 7). In order for this process to be efficient, it is likely that the E1, E2, and E3 enzymes involved form multiprotein complexes to allow rapid thiol ester transfer of ubiquitin molecules (4, 7). Whereas multiubiquitination of some proteins leads to their
rapid degradation by the proteasome, in other instances ubiquitination
may serve as a modification resulting in functional regulation (5, 6).
At present it is unclear how specific proteins are recognized as
substrates for the ubiquitin system and what precise roles the E2 and
E3 enzymes play in the recognition as well as in the ubiquitination of
a substrate.
While only one functional E1 ubiquitin activating enzyme has been identified thus far, over 30 different E2 ubiquitin-conjugating enzymes have been isolated from various organisms (10). All E2s contain a conserved domain of approximately 14 kDa (~130 amino acids) and an active site cysteine residue that is required for thiol ester formation with ubiquitin. E2 enzymes that consist almost exclusively of the conserved Ubc domain (class I E2s) are unable to transfer ubiquitin to protein substrates in vitro, suggesting that this class of E2s may require E3 ubiquitin protein ligases for substrate recognition. A second group of E2 enzymes (class II) contain unique carboxyl-terminal extensions (e.g. cdc34, Rad6, Ubc6) (10) that may contribute to substrate specificity and cellular localization (10).
The E3 ubiquitin protein ligases may be the key enzymes that provide
substrate specificity for the ubiquitin conjugation system. Although
two E3 enzymes have been previously identified from rabbit reticulocytes (E3 and E3
) and one from yeast (UBR1), it was not
until the cloning and characterization of E6AP that the structural and
functional features of a new class of E3 enzymes was revealed (10-13).
E6AP was initially identified as a 100-kDa cellular protein that in
conjunction with the E6 oncoprotein of the human papillomavirus type 16 (HPV) constituted the E3 activity in the ubiquitination of p53 (8,
14-16). E6AP can also promote the ubiquitination of cellular proteins
in the absence of E6, indicating that E6AP can function as an E3
independent of E6 (8). Sequence analysis of E6AP revealed a region of
approximately 350 amino acids in the carboxyl terminus that was highly
conserved among a number of proteins from various organisms (17). This
region, termed the Hect domain (homologous to
E6AP carboxyl terminus), also contains a conserved cysteine residue that serves as the active site for thiol
ester formation with ubiquitin (17). In addition to E6AP, two other
Hect domain proteins, RSP5 and rat p100, have been shown to form thiol
ester complexes with ubiquitin (9, 17). A total of six genes encoding
Hect proteins have been identified in Saccharomyces cerevisiae and several in Drosophila melanogaster,
Caenorhabditis elegans, and mammals (17).
The question of substrate specificity is key to understanding how the ubiquitin system is regulated. Although the recognition of specific proteins as substrates appears to involve protein-protein interactions with specific E3s, the relative contribution of E2 enzymes in substrate recognition and ubiquitination has not yet been established. The existence of multiple E2s and Hect E3 proteins, however, suggests that specific combinations of these proteins are likely to function together in the ubiquitination of a substrate. At least 12 different genes encoding E2 enzymes have been isolated from S. cerevisiae and shown to be involved in a variety of cellular functions, including DNA repair, cell cycle control, protein translocation, stress response, and chromosomal organization (3, 10). Moreover, gene inactivation experiments in mice have indicated very specific roles for certain E2s, suggesting that, despite belonging to a large multigene family, their functions are not redundant. For example, proviral integration and inactivation of UbcM4 (mouse homologue of UbcH7) results in placental defects and embryonic lethality (18). In another recent study, it was demonstrated that inactivation of mHR6B (one of the mouse homologues of yeast Rad6/Ubc2) in mice causes male sterility associated with chromatin modification (19). This is an intriguing result in terms of functional specificity of E2 enzymes since another mouse homologue of Rad6, mHR6A, which shares over 90% sequence identity with mHR6B and is expressed in all the same organs and tissues as mHR6B, was intact in these mice (19). In addition, both human homologues of Rad6, hHR6A and hHR6B, were shown to complement Rad6 function in DNA repair but not in sporulation in yeast (20). These data strongly indicate that individual E2 enzymes, despite having closely related homologues, carry out very specific functions in the cell.
The critical question with regard to different E2s is how functional specificity for each enzyme is achieved. Although the expression profiles, tissue distribution and subcellular localization of individual E2s may contribute to functional specificity, an equally important aspect is the ability of E2s to cooperate with specific E3 proteins in substrate ubiquitination. However, the basis underlying the specificity of why certain E2 enzymes and not others function with particular E3s is not yet known.
To identify potential substrates and regulators of E6AP, we have carried out a yeast two-hybrid screen using E6AP as bait. In this study, we report the cloning of a new human E2 (designated UbcH8) that was isolated as an E6AP-interacting protein and demonstrate that UbcH8 can transfer ubiquitin to E6AP. We have extended this analysis to additional E2s and Hect domain E3s. Our results demonstrate that only a subset of structurally related E2s physically interact with E6AP and function in the ubiquitination of an E6AP substrate (22).2 Furthermore, we show that a different subfamily of structurally related E2s bind to the S. cerevisiae Hect protein RSP5 (17, 23). Taken together, these studies demonstrate specific complex formation between E2 enzymes and the Hect domain family of E3 proteins, and suggest that the ability of E2s to physically associate with specific Hect E3s constitutes the basis of specificity for functionally distinct E2:E3 combinations.
A modified version of the yeast
two-hybrid screen was carried out to identify E6AP-interacting proteins
(21, 24-26). The bait vector was constructed by inserting a
catalytically inactive mutant of E6AP (C833A) into pPC97, in-frame with
the Gal4 DNA binding domain (amino acids 1-147). The prey cDNA
library contained the Gal4 activation domain fused to cDNAs derived
from activated human T cells in pPC86 vector. The bait and prey
constructs were transformed into yeast strain MaV103 (MATa
ura3-52 leu2-3, 112 trp1-901 his3200 ade2-101 gal4
gal80
GAL1::HIS3 @ lys2 SPAL10::URA3) and
transformants plated in the absence of histidine. Since increased HIS3 expression resulting from an interacting clone will
render the yeast resistant to 3-aminotriazole (3-AT), selection of
interacting clones was performed in the presence of 25 mM
3-AT. pPC86-derived interaction positive cDNAs were rescued by
transformation of competent HB101 with total yeast DNA. The DNA
sequence of isolated clones was determined by dideoxynucleotide
sequencing using appropriate oligonucleotide primers. The pPC97 vector,
pPC86-derived human T cell cDNA library and the yeast host strain
MaV103 were kindly provided by Dr. Marc Vidal (Massachusetts General
Hospital, Charlestown, MA).
The ubch8, ubch7,
ubch5, and ubch9 (27) genes were amplified by
polymerase chain reaction (PCR) using appropriate oligonucleotide primers and inserted into pET23a vector (Novagen) for bacterial expression. For expression in MaV103 yeast strain, PCR-derived E2
cDNAs (UbcH8, UbcH7, UbcH6, UbcH5, and UbcH9) were cloned in-frame with the Gal4 activation domain in the pPC86 plasmid. E6AP, RSP5, and
rat p100 cDNAs harboring a substitution of the active site cysteine
residue with alanine were amplified by PCR and cloned into the pPC97
vector in-frame with the Gal4 DNA binding domain. E6AP deletion mutants
were generated by PCR using appropriate oliginucleotide primers and
inserted in-frame with the Gal4 DNA binding domain in pPC97 (see Fig.
3B). dC109, dC217, and dC325 delete 109, 217, and 325 residues, respectively, from the carboxyl terminus of E6AP. dHect is a
deletion of 350 amino acids from the carboxyl terminus corresponding to
the Hect domain. pPC97-Hect expresses the Hect domain of E6AP fused to
the Gal4 DNA binding domain and dNHect is a deletion of 150 residues
from the amino terminus of the Hect domain. The full-length HHR23A
cDNA was cloned into pGem-1 by PCR for in vitro
expression (Promega).
Protein Expression
E1, UbcH8, UbcH7, UbcH5, and UbcH9 were expressed in Escherichia coli BL21 using the pET expression system (Novagen) (23, 28, 29). The relative amounts of expression of various E2 proteins was determined by Coomassie Blue staining. Construction, purification, and radioactive labeling of GST-ubiquitin has been described previously (8). Preparation of E6AP from Hi5 cells infected with recombinant baculovirus expressing WT or mutant E6AP proteins has been described (8, 17, 23). In vitro expression of HHR23A was performed in TNT-coupled wheat germ extracts (WGE) in the presence of [35S]methionine as per manufacturer's instructions (Promega). The expression of fusion proteins in yeast strain MaV103 was confirmed by Western blotting with anti-Gal4 DNA binding domain and anti-Gal4 activation domain antibodies from Santa Cruz Biotechnology.
Thiol Ester AssayUbiquitin thiol ester formation on UbcH8 and E6AP was determined as described previously (9). Reaction mixtures contained approximately 5-10 ng of E1, 100 ng of UbcH8, 200 ng of WT-E6AP or E6AP (C833A), and 500 ng of 32P-labeled GST-ubiquitin in 20 mM Tris-HCl, pH 7.6, 50 mM NaCl, 4 mM ATP, 10 mM MgCl2, and 0.2 mM dithiothreitol for 3 min at 25 °C. Reactions were terminated by incubating the mixtures for 15 min at 30 °C in SDS-sample buffer (50 mM Tris-HCl, pH 6.8, 2% SDS, 10% glycerol, and 0.2% bromphenol blue) in the absence of reducing agents and resolved by SDS-PAGE. Radioactively labeled proteins were visualized by autoradiography.
Protein Interaction AssayThe yeast strain MaV103 was co-transformed with pPC97-derived Hect protein constructs and pPC86-derived E2 constructs using standard lithium acetate procedures and plated in the absence of leucine and tryptophan (21). Individual colonies were picked and equivalent number of cells, as determined by OD reading at 600 nm, were spotted on histidine drop-out plates containing 25 mM 3-AT. Interactions were scored based on growth in the presence of 3-AT after incubation at 30 °C for 3 days.
Ubiquitination AssayHHR23A protein was synthesized in vitro in the presence of [35S]methionine for 90 min at 30 °C using TNT-coupled wheat germ extracts (22) (Promega). 5-µl aliquots of in vitro translated HHR23A were incubated with 5-10 ng of E1, approximately 100 ng of various E2s, 200 ng of E6AP, and 4 µg of ubiquitin (Sigma) in 20 mM Tris-HCl, pH 7.6, 50 mM NaCl, 4 mM ATP, 10 mM MgCl2, and 0.2 mM dithiothreitol. Reactions were terminated after 2 h. at 30 °C by the addition of SDS-sample buffer. Samples were boiled for 5 min, resolved by SDS-PAGE, and visualized by autoradiography.
To identify cDNA clones that encode
E6AP-interacting proteins, a modified version of the yeast two-hybrid
screen was performed using E6AP as bait (21). A catalytically inactive
form of E6AP in which the active site cysteine residue has been
substituted with alanine (C833A) was used to avoid potential
degradation of interacting proteins. The yeast reporter strain MaV103
expressing E6AP (C833A) as a fusion protein with the Gal4 DNA-binding
domain was transformed with the prey cDNA library derived from
human T cells (21). A total of 12 independent interacting clones were isolated by plating the transformants on histidine drop-out plates containing 25 mM 3-AT. Here we describe the identification
of two independent clones consisting of identical cDNA inserts that were determined to code for a protein of 152 residues with a predicted molecular mass of approximately 17 kDa (Fig.
1A). Amino acid sequence comparison of the
isolated clone revealed strong homology with ubiquitin conjugating
enzymes from various organisms, indicating that the isolated clone
encodes a previously unidentified human E2 enzyme. In keeping with the
nomenclature, we have designated the new E2 UbcH8 (30). The amino acid
sequence alignment and percent homology of UbcH8 with other known human
E2s was obtained using the DNAsis program is shown in Fig.
1B. The primary sequence of UbcH8 indicates that it belongs
to class I type E2s as it consists primarily of the Ubc domain and does
not contain amino- or carboxyl-terminal extensions (10). A previously
described E2, UbcH7, that was shown to transfer ubiquitin to E6AP
in vitro, was found to have the highest degree of homology
to UbcH8 (45.7% identity, Fig. 1B) (30).
Ubiquitin Thiol Ester Formation and Transfer
To ascertain
whether UbcH8 can form thiol ester complexes with ubiquitin and
transfer the ubiquitin to the active site cysteine residue of E6AP, the
UbcH8 cDNA was inserted into pET23a vector and expressed in
E. coli (23, 28, 29). Bacterially expressed UbcH8 was
incubated in the presence of ATP with E1, wild type or catalytically
inactive E6AP, and radiolabeled ubiquitin. As shown in Fig.
2 (lanes 2 and 3), UbcH8 could
efficiently form thiol ester complexes with ubiquitin. Furthermore,
UbcH8 could transfer ubiquitin to wild type E6AP and not to the E6AP
mutant (C833A) containing a substitution of the active site cysteine residue with alanine (Fig. 2, lanes 2 and 3). The
complex of E6AP and ubiquitin could be disrupted by the addition of
dithiothreitol, indicating the labile nature of a reduction sensitive
thiol ester bond (data not shown). No thiol ester adduct was detected
on E6AP in the absence of UbcH8 (lane 1). It should be noted
that in the presence of mutant E6AP (C833A), the amount of ubiquitin
thiol ester complexes remaining on UbcH8 was greater than in the
presence of wild type E6AP indicating the specificity of ubiquitin
transfer from UbcH8 to E6AP. These results demonstrate that like other E2s that have been characterized thus far, UbcH8 can form thiol ester
complexes with ubiquitin. In addition, UbcH8 has the ability to
transfer ubiquitin to the active site cysteine residue of E6AP.
Interaction of E6AP with UbcH8 and UbcH7
The finding that UbcH8 can not only transfer ubiquitin to E6AP but also physically interact with E6AP, suggested that the binding of E2 enzymes to Hect E3 proteins might be a general property important for functional cooperation between E2 and E3 enzymes. We therefore examined the ability of other human E2s to associate with E6AP using the yeast two-hybrid assay. The yeast strain MaV103 was co-transformed with E6AP (C833A) and various E2s, and interactions were scored by growth on histidine-minus plates containing 25 mM 3-AT. Whereas both UbcH8 and the closely related UbcH7 (30) interacted efficiently with E6AP, UbcH5 and UbcH9 (23, 27) did not show significant levels of interaction (Fig. 3A). Thus E6AP can associate with only a subset of E2s, namely UbcH8 and UbcH7. In addition, as suggested by the high degree of homology between UbcH8 and UbcH7 (shown in Fig. 1B), these E2s may represent a structurally and functionally related subfamily.
To map the region of E6AP that is required for its interaction with UbcH8 and UbcH7, several deletion constructs of E6AP were generated and assayed for their ability to interact with the different E2s. As summarized in Fig. 3B, the Hect domain of E6AP was found to be required and sufficient for interaction with UbcH8 and UbcH7, while the amino-terminal sequences of E6AP were dispensable. Each of the deletions within the Hect region resulted in a loss of binding, further indicating the requirement of this region for interaction with the specific E2s (Fig. 3B). Finer deletions within the Hect domain will be required to determine whether the entire region is involved in the interaction with E2s or whether specific segments mediate the binding. Nonetheless, these results suggest that the conserved Hect domains of this family of proteins may enable them to associate with specific E2 enzymes and as such provide the specificity for functional cooperation in substrate ubiquitination.
Interaction of RSP5 with UbcH6 and UbcH5Since the conserved
Hect domain defines a family of proteins believed to function as E3s,
the Hect domains of two other proteins, RSP5 and the rat p100 protein
(17), were tested for their ability to interact with different E2s
using the two-hybrid assay. Whereas the E6AP Hect domain interacted
selectively with UbcH8 and UbcH7, the RSP5 Hect region was found to
interact with UbcH6 and UbcH5, and not with other E2s (Fig.
4). It should be noted that just as UbcH8 and UbcH7 are
closely related to each other (45.7% identity; Fig. 1B),
UbcH6 and UbcH5 share strong homology with each other (53.7% identity;
Fig. 1B). The rat p100 protein was unable to interact with
any of the five E2s that were tested in these experiments. On the basis
of these results, it seems that different Hect containing proteins interact with specific subsets of closely related E2s and that the selective association of E2 and E3 proteins may constitute the basis of specificity for distinct E2:E3 combinations.
Ubiquitination of an E6AP Substrate
Since UbcH8 and UbcH7
interact with E6AP, we next tested whether these E2s could function in
conjunction with E6AP in the ubiquitination of HHR23A, an E6
independent substrate of E6AP (22).2 The HHR23A protein was
synthesized in WGE in the presence of radiolabeled methionine and
incubated with E6AP either in the absence (Fig. 5, lane 2)
or in the presence of equivalent amounts of bacterially expressed E2s
as indicated. Due to the presence of endogenous E2 activity in WGE,
some ubiquitinated forms of HHR23A were observed even in the absence of
any exogenously added E2s (lane 2). However,
upon addition of bacterially expressed UbcH8 or UbcH7, the
ubiquitination of HHR23A was enhanced (lanes 3 and
5). In contrast, the addition of UbcH5 did not seem to
further facilitate the reaction over background levels (lane
4). Similar negative results were obtained using UbcH9 (data not
shown). Therefore, consistent with the interaction data, the two
E6AP-interacting E2s (UbcH8 and UbcH7) were also able to function with
E6AP in the ubiquitination of one of its substrates, whereas
noninteracting E2s were unable to do so. These results support the
hypothesis that physical interaction between specific E2 and E3 enzymes
may be important for functional cooperativity.
The ubiquitin system plays a major role in selective protein degradation and may also be an important pathway for protein modification (4-7). The specificity of substrate recognition by the ubiquitin system may be achieved by different E3 enzymes capable of interacting with specific substrates. However, specificity within the enzymatic components of the pathway, i.e. different E2:E3 combinations, may also influence substrate recognition and ubiquitination. As noted earlier, gene inactivation experiments in mice have indicated very specific roles for E2 enzymes (18, 19), although the way functional specificity for different E2s is achieved is not yet known. The existence of multiple E3 proteins and their likely role in substrate recognition strongly suggests that cooperation between specific E2 and E3 enzymes may play an important role in defining the specificity involved in substrate ubiquitination. Accordingly, ubiquitination of a substrate may depend upon the transfer of ubiquitin from E1 to a given E2, the ability of that E2 to cooperate with a specific E3, and finally the interaction between the E3 and the substrate.
To date, however, the only physical interaction demonstrated between E2 and E3 proteins involves the S. cerevisiae Rad6 (Ubc2) protein and UBR1, a non-Hect domain E3 involved in the N-end rule pathway for protein degradation (13, 31-33). UBR1 has not been shown to form thiol ester complexes with ubiquitin, suggesting that it may function more like an adaptor or docking protein by bringing in close proximity the relevant E2 and the substrate. The Hect domain family of E3 proteins, as established for E6AP, play a more direct role in substrate ubiquitination as they can form high energy thiol ester bonds with ubiquitin prior to the transfer and covalent attachment of ubiquitin to a substrate (9, 17). Therefore, the Hect domain family of E3s do not function simply as adaptors between E2s and substrate proteins but rather participate as intermediates in the cascade of ubiquitin transfers from E1 to E2s, from E2s to specific E3s, and finally to substrates (9, 17). However, the basis of specificity for ubiquitin thiol ester transfer from E2s to E3s has not been addressed, and prior to this study, physical association of specific E2 and Hect E3 proteins had not been demonstrated.
The studies presented here demonstrate for the first time complex formation between specific E2s and Hect E3 proteins in vivo, and suggest that the ability of E2s to interact selectively with Hect domain E3s forms the basis for functional cooperativity between E2 and E3 enzymes in substrate ubiquitination. UbcH8 and the closely related UbcH7 interacted specifically with E6AP. This interaction was mapped to the Hect domain of E6AP, suggesting that other Hect family E3 proteins may also have the ability to interact with specific E2s. Interestingly, two other E2s closely related to each other, UbcH6 and UbcH5, were found to interact selectively with RSP5, a Hect-domain protein from S. cerevisiae (17, 23, 30). None of the five E2s examined in this study interacted with the rat p100 protein. Finally, we demonstrate that the two E6AP-interacting E2s also functioned in conjunction with E6AP in the ubiquitination of HHR23A, an E6AP substrate (22).2 The noninteracting E2s were unable to enhance the ubiquitination of HHR23A above background levels observed in wheat germ extracts.
Prior to this report, the only assay used to define the specificity of ubiquitin thiol ester transfer from E2s to Hect E3s was based upon in vitro reactions in which relatively high amounts of partially purified components were used to detect the transfer of a radiolabeled GST-ubiquitin fusion protein to E6AP (9, 30). Using these in vitro thiol ester assays, E6AP was shown to accept ubiquitin from A. thaliana Ubc8, UbcH5, and UbcH7 (23, 30); and, RSP5 from UbcH5 (30). However, as shown in this study, UbcH5 fails to interact efficiently with E6AP in the yeast two-hybrid system whereas it can interact strongly with RSP5. The most probable explanation for a lack of efficient interaction between E6AP and UbcH5 may be that the two proteins do not have a high affinity for each other. Accordingly, it is possible that unlike the assay conditions for thiol ester transfer in vitro (9), the comparatively low levels of protein expression from centromeric plasmids in yeast cells (21) may preclude the detection of relatively low affinity interactions. Consistent with this possibility, UbcH5 was unable to enhance the ubiquitination of HHR23A under conditions where E6AP-interacting UbcH8 and UbcH7 did so efficiently.
In this regard, it is also important to note that UbcH5, which is very similar to Arabidopsis thaliana Ubc8, was cloned by reverse transcription-PCR using two degenerate primers corresponding to conserved sequences in A. thaliana Ubc8 and other similar E2s (23). As such, the isolation of UbcH5 was not based upon any functional criteria that link it directly to E6AP. On the other hand, UbcH7 (initially called E2-F1) was isolated as an E2 activity from reticulocyte lysates and shown to function with E6 and E6AP in the ubiquitination of p53 (34, 35). The cDNA for E2-F1 was subsequently cloned by Nuber et al. (30) based upon peptide sequences derived from direct protein sequencing (34) and designated as UbcH7 (30). UbcH7 was shown to be capable of transferring ubiquitin to E6AP in vitro (30). As described in this report, we isolated UbcH8 as a specific E6AP-interacting protein, and show that it also can transfer ubiquitin to E6AP. Sequence comparison revealed that UbcH8 was most closely related to UbcH7, and both E2s could interact specifically with E6AP. As noted above, UbcH5 and UbcH6 share a high degree of homology with each other. Both of these E2s failed to bind E6AP efficiently but interacted strongly with RSP5. Consistent with this observation, UbcH6 and to a lesser extent UbcH5, were also able to bind Pub-1, the Schizosaccharomyces pombe homologue of RSP5.3
Based on the interaction and functional data presented here, UbcH8 and UbcH7 represent a structurally related subfamily of E2s that may function physiologically with E6AP in the ubiquitination of its substrates, whereas UbcH6 and UbcH5 may belong to a subfamily that functions with the mammalian homologues of RSP5. Further studies are currently underway to extend the analysis of E2:E3 interactions to additional E2s and Hect E3s. As more substrates of the E3 enzymes become known, different E2:E3 combinations can be tested for their ability to function in the ubiquitination of specific substrates, thereby allowing the identification of E2:E3 pairs that cooperate in substrate ubiquitination.
We thank Dr. Marc Vidal, Massachusetts General Hospital, Charlestown, MA, for help and reagents for the yeast two-hybrid screen. We also thank Dr. Martin Scheffner for providing UbcH6 and UbcH7 cDNAs. We are grateful to Drs. Carl Maki, Lucienne Ronco, and Andrea Talis for helpful suggestions and comments on the manuscript.