Degradation signals in the lysine-asparagine sequence space
1999; Springer Nature; Volume: 18; Issue: 21 Linguagem: Inglês
10.1093/emboj/18.21.6017
ISSN1460-2075
Autores Tópico(s)Protein Degradation and Inhibitors
ResumoArticle1 November 1999free access Degradation signals in the lysine–asparagine sequence space Tetsuro Suzuki Tetsuro Suzuki Present address: Department of Virology II, National Institute of Infectious Diseases, 1-23-1 Toyama, Shinjuku-ku, Tokyo, 162-8640 Japan Search for more papers by this author Alexander Varshavsky Corresponding Author Alexander Varshavsky Division of Biology, California Institute of Technology, 1200 East California Boulevard, Pasadena, CA, 91125 USA Search for more papers by this author Tetsuro Suzuki Tetsuro Suzuki Present address: Department of Virology II, National Institute of Infectious Diseases, 1-23-1 Toyama, Shinjuku-ku, Tokyo, 162-8640 Japan Search for more papers by this author Alexander Varshavsky Corresponding Author Alexander Varshavsky Division of Biology, California Institute of Technology, 1200 East California Boulevard, Pasadena, CA, 91125 USA Search for more papers by this author Author Information Tetsuro Suzuki2 and Alexander Varshavsky 1 1Division of Biology, California Institute of Technology, 1200 East California Boulevard, Pasadena, CA, 91125 USA 2Present address: Department of Virology II, National Institute of Infectious Diseases, 1-23-1 Toyama, Shinjuku-ku, Tokyo, 162-8640 Japan *Corresponding author. E-mail: [email protected] The EMBO Journal (1999)18:6017-6026https://doi.org/10.1093/emboj/18.21.6017 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info The N-degrons, a set of degradation signals recognized by the N-end rule pathway, comprise a protein's destabilizing N-terminal residue and an internal lysine residue. We show that the strength of an N-degron can be markedly increased, without loss of specificity, through the addition of lysine residues. A nearly exhaustive screen was carried out for N-degrons in the lysine (K)–asparagine (N) sequence space of the 14-residue peptides containing either K or N (16 384 different sequences). Of these sequences, 68 were found to function as N-degrons, and three of them were at least as active and specific as any of the previously known N-degrons. All 68 K/N-based N-degrons lacked the lysine at position 2, and all three of the strongest N-degrons contained lysines at positions 3 and 15. The results support a model of the targeting mechanism in which the binding of the E3–E2 complex to the substrate's destabilizing N-terminal residue is followed by a stochastic search for a sterically suitable lysine residue. Our strategy of screening a small library that encompasses the entire sequence space of two amino acids should be of use in many settings, including studies of protein targeting and folding. Introduction Regulatory proteins are often short-lived in vivo, providing a way to generate their spatial gradients and to rapidly adjust their concentration or subunit composition through changes in the rate of their synthesis or degradation. Most of the damaged or otherwise abnormal proteins are metabolically unstable as well. Many other proteins, while long-lived as components of larger structures such as ribosomes and oligomeric proteins, are short-lived as free subunits (reviewed by Hochstrasser, 1996; Varshavsky, 1997; Hershko and Ciechanover, 1998; Scheffner et al., 1998; Koepp et al., 1999; Tyers and Willems, 1999). Features of proteins that confer metabolic instability are called degradation signals, or degrons (Laney and Hochstrasser, 1999). One class of degradation signals, called the N-degrons, comprises a protein's destabilizing N-terminal residue and an internal Lys residue (Bachmair et al., 1986; Varshavsky, 1996). A set of N-degrons containing different N-terminal residues that are destabilizing in a given cell defines a rule, termed the N-end rule, which relates the in vivo half-life of a protein to the identity of its N-terminal residue. The lysine determinant of an N-degron is the site of formation of a substrate-linked multi-ubiquitin chain (Bachmair and Varshavsky, 1989; Chau et al., 1989). The N-end rule pathway is thus one pathway of the ubiquitin (Ub) system. Ub is a 76-residue protein whose covalent conjugation to other proteins plays a role in a multitude of processes, including cell growth, division, differentiation, and responses to stress (Pickart, 1997; Varshavsky, 1997; Peters, 1998; Scheffner et al., 1998). In many of these settings, Ub acts through routes that involve the degradation of Ub–protein conjugates by the 26S proteasome, an ATP-dependent multisubunit protease (Coux et al., 1996; Hilt and Wolf, 1996; Baumeister et al., 1998; Rechsteiner, 1998). The N-end rule is organized hierarchically. In the yeast Saccharomyces cerevisiae, Asn and Gln are tertiary destabilizing N-terminal residues in that they function through their conversion, by the NTA1-encoded N-terminal amidase, into the secondary destabilizing N-terminal residues Asp and Glu (Baker and Varshavsky, 1995). The destabilizing activity of N-terminal Asp and Glu requires their conjugation, by the ATE1-encoded Arg-tRNA-protein transferase, to Arg, one of the primary destabilizing residues. The primary N-terminal residues are bound directly by the UBR1-encoded N-recognin, the E3 (recognition) component of the N-end rule pathway. In S.cerevisiae, N-recognin is a 225 kDa protein that binds to potential N-end rule substrates through their primary destabilizing N-terminal residues: Phe, Leu, Trp, Tyr, Ile, Arg, Lys and His (Varshavsky, 1996). Analogous components of the mammalian N-end rule pathway have been identified as well (Stewart et al., 1995; Grigoryev et al., 1996; Kwon et al., 1998, 1999). Studies with engineered N-end rule substrates indicated the bipartite organization of N-degrons and suggested a stochastic model of their targeting, in which specific lysines of an N-end rule substrate could be assigned different probabilities of being used as a ubiquitylation site (Bachmair and Varshavsky, 1989; Chau et al., 1989; Johnson et al., 1990; Hill et al., 1993; Varshavsky, 1996; Lévy et al., 1999). Most of the evidence for this model was produced with a set of N-degrons in which a destabilizing N-terminal residue X was linked to the ∼40-residue Escherichia coli Lac repressor-derived sequence termed eK [extension (e) bearing lysines (K)] (Figure 1A) (Bachmair and Varshavsky, 1989). The resulting X-eK sequence comprised a portable N-degron, which could confer short half-lives on test proteins such as E.coli β-galactosidase (βgal) or mouse dihydrofolate reductase (DHFR) (Varshavsky, 1996). At least one of two lysines (K) in eK, either K-15 or K-17, must be present for the N-degron to be active (Figure 1A) (Bachmair and Varshavsky, 1989; Johnson et al., 1990). Even though several other classes of N-degron, including the naturally occurring ones, have been described over the last decade (Townsend et al., 1988; deGroot et al., 1991; Dohmen et al., 1994; Sadis and Finley, 1995; Ghislain et al., 1996; Sijts et al., 1997; Tobery and Siliciano, 1999), the mechanistic understanding of these degradation signals remains confined largely to the eK-based N-degrons (Varshavsky, 1996; Lévy et al., 1999). Figure 1.Test proteins. (A) Fusions used in this work contained some of the following elements (see Materials and methods): DHFRha, a mouse dihydrofolate reductase moiety extended at the C-terminus by a sequence containing the hemagglutinin-derived ha epitope; the UbR48 moiety bearing the Lys→Arg alteration at position 48; a 40-residue E.coli Lac repressor-derived sequence, termed eK and shown in single-letter abbreviations for amino acids; a variable residue X (either Tyr, His or Met) between UbR48 and eK; the E.coli βgal moiety lacking the first 24 residues of wild-type βgal. The positions of single or multiple KRK insertions into eK are indicated. The endogenous KRK sequence of eK is underlined. The arrow indicates the site of in vivo cleavage by DUBs. (B) The fusion construct used for screening in the K/N sequence space. The 14 residues of eK immediately following the residue X were replaced by a set of 14-residue sequences that comprised a random permutation of Lys and Asn residues, followed by the sequence HGSGAWLLPVSLVRS, derived from residues 2–14 of the eK extension, followed by Arg-Ser. Download figure Download PowerPoint In the present work, we show that spiking an eK-based N-degron with additional Lys residues can markedly increase its activity. We also show, using a new approach of searching in the sequence space of lysine and asparagine, that simple-sequence N-degrons can be as strong and specific as any of the previously known N-degrons. These findings provide independent evidence for the model of a bipartite N-degron and stochastic targeting mechanism (Bachmair and Varshavsky, 1989). The strategy of exhaustive searching in the sequence space of two amino acids should be of use in many settings, including studies of protein folding and degradation. Results and discussion The Ub/protein/reference technique The assays below utilized the previously developed Ub/protein/reference (UPR) technique, which increases the accuracy of pulse–chase analysis by providing a 'built-in' reference protein (Lévy et al., 1996). This method employs a linear fusion in which Ub is located between a protein of interest and a reference protein moiety (Figure 1A). The fusion is co-translationally cleaved by Ub-specific de-ubiquitylating enzymes (DUBs) (Wilkinson and Hochstrasser, 1998) after the last residue of Ub, producing equimolar amounts of the protein of interest and the reference protein bearing the C-terminal Ub moiety. If both the reference protein and the protein of interest are immunoprecipitated in a pulse–chase assay, the relative amounts of the protein of interest can be normalized against the reference protein in the same sample. The UPR technique can thus compensate for the scatter of immunoprecipitation yields, sample volumes and other sources of sample-to-sample variation (Lévy et al., 1996, 1999). Two previously introduced terms, IDx, initial decay, i.e. the decay of a protein during the pulse of x min, and ty–z0.5, the protein's half-life averaged over the interval of y to z min of chase (Lévy et al., 1996), are used below to describe the decay curves of test proteins. The IDx term and the interval-specific term ty–z0.5 would be superfluous in the case of a strictly first-order decay, which is defined by a single half-life. However, the in vivo degradation of most proteins deviates from first-order kinetics. For example, the rate of degradation of short-lived proteins can be much higher during the pulse, in part because a newly labeled (either nascent or just-completed) polypeptide is conformationally immature and may, consequently, be targeted for degradation more efficiently than its mature counterpart. This enhanced early degradation, previously termed the 'zero-point' effect (Baker and Varshavsky, 1991), is described by the parameter IDx (Lévy et al., 1996). It was found that a large fraction of the zero-point effect results from the co-translational degradation of nascent (being synthesized) polypeptide chains, which never reach their mature size before their destruction by processive proteolysis (G.Turner and A.Varshavsky, unpublished data). The detection of a zero-point effect requires the comparison of a test protein's degradation between cells containing and lacking the relevant proteolytic pathway. Alternatively, the zero-point effect can be detected by comparing, through the UPR technique, the degradation of otherwise identical degron-containing and degron-lacking versions of a test protein (Lévy et al., 1996, 1999). Although the degradation of a protein during the pulse can be strikingly high (Lévy et al., 1996) (see also below), it is not detectable by a conventional, reference-lacking pulse–chase assay. Increasing the strength of N-degrons by spiking them with additional lysines The UPR constructs of the present work were DHFR-ha-UbR48-X-eK-βgal fusions. They contained the metabolically stable, ha-epitope-bearing DHFR-ha-UbR48 moiety as a reference protein, termed dha-Ub below. The dha-Ub-X-eK-βgal proteins were co-translationally cleaved in vivo, yielding the test protein X-eK-βgal and the reference dha-Ub (Figure 1A). To reduce the possibility that the C-terminal Ub moiety of dha-Ub could function as a ubiquitylation/degradation signal, the K-48 residue of Ub (a major site of isopeptide bonds in multi-Ub chains) was converted to Arg, which cannot be ubiquitylated, yielding UbR48 (Lévy et al., 1996). These and related fusions (Figure 1A) were expressed in S.cerevisiae from low copy plasmids and the copper-inducible PCUP1 promoter. X-eK-βgal is an extensively analyzed class of N-end rule substrates, which contain a variable N-terminal residue X (produced through the DUB-mediated cleavage of dha-Ub-X-eK-βgal at the Ub–X junction), a 40-residue N-terminal extension called eK (see Introduction), and a βgal moiety lacking the first 24 residues of wild-type E.coli βgal (Figure 1A). If K-15 and K-17, the only lysines of the eK extension (Figure 1A), are replaced by Arg residues, which cannot be ubiquitylated, the resulting X-eΔK-βgal is long-lived even if its N-terminal residue is destabilizing in the N-end rule (Bachmair et al., 1986; Johnson et al., 1990). The inactivity of N-degron in X-eΔK-βgal is caused by the absence of targetable lysines (Varshavsky, 1996). Specifically, the multiple lysines of the βgal moiety in X-eΔK-βgal (Chau et al., 1989) cannot serve as N-degron determinants, apparently because the most N-terminal Lys residue in X-eΔK-βgal, at position 239, is too far from the protein's N-terminus. One of our aims was to produce stronger N-degrons. We chose Tyr, a moderately destabilizing type 2 residue (Bachmair and Varshavsky, 1989; Varshavsky, 1996), as the N-terminal residue of an initial test protein (Figure 1A). More strongly destabilizing N-terminal residues, e.g. Leu or Arg, in the context of (expected) stronger N-degrons would have made the test proteins too short-lived for detection in a pulse–chase assay. Met was employed as a stabilizing N-terminal residue. The term ID5 below (see Materials and methods) conveys the extent of degradation of a protein during the 5 min pulse, in comparison with the degradation, during the same pulse, of a control (degron-lacking, i.e. Met-bearing) version of the same protein. To determine whether the degradation of Tyr-eK-βgal in S.cerevisiae could be enhanced through the addition of Lys residues while remaining dependent on the Ubr1p N-recognin, the sequence Lys-Arg-Lys (KRK), identical to the sequence at positions 15–17 of eK, was inserted at the indicated locations within eK (Figure 1A). The unmodified Tyr-esuper K-βgal had an ID5 of ∼48%, i.e. ∼48% of the labeled Tyr-eK-βgal was destroyed during the 5 min pulse, before time 0. The t0.50–10 (half-life between 0 and 10 min of chase) of Tyr-eK-βgal was ∼26 min (Figures 2A and 3A). The KRK sequence inserted at any of the indicated three positions within eK (Figure 1A) strongly destabilized the already short-lived Tyr-eK-βgal: for example, Tyr-K1eK-βgal (Figure 1A) had an ID5 of ∼75% and t0.50–10 of ∼5 min (Figures 2A and 3A). The increased degradation of Tyr-eK-βgal derivatives containing extra KRK remained completely Ubr1p-dependent: Tyr-eK-βgal, Tyr-K1eK-βgal, Tyr-K2eK-βgal and Tyr-K3eK-βgal were all long-lived proteins (t0.5 >10 h) in ubr1Δ cells (Figures 2A and 3A). In addition, Met-eK-βgal, Met-K1eK-βgal, Met-K2eK-βgal and Met-K3eK-βgal, the Met-bearing counterparts of the Tyr-based N-end rule substrates, were long-lived in either UBR1 or ubr1Δ cells (data not shown). Figure 2.Active N-degrons can be strongly enhanced by additional lysines. (A) Congenic UBR1 (wt) and ubr1Δ S.cerevisiae that expressed the UPR-based fusions Met-eK-βgal (Mek) (DHFR-ha-UbR48-Met-eK-βgal), Tyr-eK-βgal (Yek), Tyr-K1eK-βgal (YK1), Tyr-K2eK-βgal (YK2) and Tyr-K3eK-βgal (YK3) (see Figure 1A) were labeled with [35S]methionine/cysteine for 5 min at 30°C, followed by a chase for 0, 10 and 30 min, extraction, immunoprecipitation with anti-ha and anti-βgal antibodies, SDS–PAGE, and autoradiography (see Materials and methods). The bands of X-βgal (test protein) and DHFR-ha-UbR48 (reference protein) are indicated on the right. (B) As in (A), but with Met-eK-βgal (Mek), Tyr-eK-βgal (Yek), Tyr-K1eK-βgal (YK1), Tyr-KK1eK-βgal (YKK1) and Tyr-KKK1eK-βgal (YKKK1) (see Figure 1A). (C) As in (A), but with Met-eK-βgal (Mek), His-eK-βgal (Hek), His-K1eK-βgal (HK1), His-KK1eK-βgal (HKK1) and His-KKK1eK-βgal (HKKK1). (D) Metabolic stability of conformationally mature Tyr-eK-βgal and its KRK-spiked derivatives. JD54 (PGAL1-UBR1) cells expressing Tyr-eK-βgal (Yek), Tyr-K1eK-βgal (YK1) or Tyr-KK1eK-βgal (YKK1) were grown in SM-raffinose medium (no expression of Ubr1p), then labeled with [35S]methionine/cysteine for 10 min at 30°C. After a 20 min chase in SM-raffinose, galactose was added to 3% to induce Ubr1p expression, followed by a chase for 1, 3 and 6 h, and the analysis of immunoprecipitated test proteins. Lanes 1, 6 and 11, the end of 35S labeling (time 0). Lanes 2, 7 and 12, the end of 20 min chase in SM-raffinose. Lanes 3, 8 and 13, 1 h chase with galactose. Lanes 4, 9 and 14, 3 h chase. Lanes 5, 10 and 15, 6 h chase. Download figure Download PowerPoint Figure 3.Quantitation of degradation of the test proteins. (A) Pulse–chase patterns of Met-eK-βgal (MeK), Tyr-eK-βgal (YeK), Tyr-K1eK-βgal (YK1), Tyr-K2eK-βgal (YK2) and Tyr-K3eK-βgal (YK3) (see Figures 1A and 2A) were quantitated using the UPR technique and PhosphorImager (see Materials and methods). Time 0 refers to the end of the 5 min pulse; 100% refers to the relative amount of Met-eK-βgal, normalized against the reference protein dha-UbR48. (B) As in (A) but with Met-eK-βgal (MeK), Tyr-eK-βgal (YeK), Tyr-K1eK-βgal (YK1), Tyr-KK1eK-βgal (YKK1) and Tyr-KKK1eK-βgal (YKKK1) (see Figures 1A and 2B). (C) As in (A) but with Met-eK-βgal (MeK), His-eK-βgal (HeK), His-K1eK-βgal (HK1), His-KK1eK-βgal (HKK1) and His-KKK1eK-βgal (HKKK1) (see Figures 1A and 2C). (D) As in (A), but quantitation of the post-translational degradation of Tyr-eK-βgal (YeK), Tyr-K1eK-βgal (YK1) and Tyr-KK1eK-βgal (YKK1) in the PGAL1-UBR1 strain JD54 (see Figures 1A and 2A), following the induction of Ubr1p by galactose. Time 0 refers to the end of the 10 min labeling in raffinose (no Ubr1p). The cells were incubated for another 20 min in raffinose, followed by the addition of galactose to induce Ubr1p (see Materials and methods). (E) Relative enzymatic activities of βgal in UBR1 cells (filled bars) and ubr1Δ cells (striped bars) expressing one of the following test proteins: Met-eK-βgal (MeK), Tyr-eK-βgal (YeK), Tyr-K1eK-βgal (YK1), Tyr-KK1eK-βgal (YKK1), Tyr-KKK1ueK-βgal (YKKK1), His-eK-βgal (HeK), His-K1eK-βgal (HK1), His-KK1eK-βgal (HKK1) and His-KKK3eK-βgal (HKKK1) (see Materials and methods). The activities of βgal were normalized to the activity of Met-eK-βgal in each cell. Values shown are the means from duplicate measurements, which yielded results within 10% of the mean values. Download figure Download PowerPoint These results led us to examine the effects of adding more than one KRK sequence to eK (Figure 1A). The resulting Tyr-euper K-βgal derivatives, bearing either two (Tyr-KK1eK-βgal) or three (Tyr-KKK1eK-βgal) KRK sequences, in addition to the original KRK of eK, were extremely short-lived proteins, even though N-terminal Tyr is a weakly destabilizing residue (Varshavsky, 1996). For example, Tyr-KK1eK-βgal (Figure 1A) had t0.50–10 of ∼4 min (in comparison with ∼26 min in the case of Tyr-eK-βgal) and an ID5 of ∼94%. In other words, ∼94% of the labeled Tyr-KK1eK-βgal was destroyed during the 5 min pulse, before time 0 (Figures 2B and 3B). At the same time, all of these proteins were long-lived in ubr1Δ cells (Figures 2B and 3B). The N-terminal Tyr is bound by the type 2 site of N-recognin (Ubr1p) that recognizes substrates bearing bulky hydrophobic N-terminal residues (Varshavsky, 1996). We asked whether the above findings were also relevant to the type 1 (basic) destabilizing N-terminal residues Arg, Lys and His, which are bound by the type 1 site of Ubr1p. Counterparts of the Tyr-βgal fusions that bore N-terminal His (a weak type 1 destabilizing residue) were constructed (Figure 1A) and tested in pulse–chase assays. The His residue was chosen as a type 1 destabilizing residue in these tests for the same reason as Tyr in the preceding tests: a stronger destabilizing residue would have made the measurements impractical with extremely short-lived substrates. The results (Figures 2C and 3C) confirmed the generality and specificity of degradation enhancement by the additional KRK sequences. For example, His-K1eK-βgal had an ID5 of ∼95% (i.e. ∼95% of the labeled His-K1eK-βgal was destroyed during the 5 min pulse, before time 0), in comparison with the ID5 of ∼41% for His-eK-βgal; the corresponding t0.50–10 values were ∼5 min and ∼8 min for His-K1eK-βgal and His-eK-βgal, respectively (Figure 3C). Similar to the results with Tyr-bearing substrates, their His-bearing, multiple KRK-containing counterparts were long-lived in ubr1Δ cells (Figures 2C and 3C). The only exception was His-KKK1eK-βgal (Figure 1A), which contained three KRK sequences, in addition to the KRK of the original eK: in contrast to Tyr-KKK1eK-βgal, His-KKK1eK-βgal was stabilized strongly but incompletely in the ubr1Δ genetic background (Figure 3C). Thus, the His-KKK1eK extension, in contrast to the Tyr-KKK1eK extension (Figure 1A), appears to contain a Ubr1p-independent degron. Previous work (Madura et al., 1993; Kwon et al., 1999) has shown that the steady-state level of an X-βgal protein (determined by measuring the enzymatic activity of βgal in yeast extracts) is a sensitive measure of its metabolic stability. The results of this steady-state assay were in agreement with those derived from pulse–chase measurements: the level of Tyr-eK-βgal in UBR1 cells was 53% of the level of the long-lived Met-eK-βgal, whereas Tyr-K1eK-βgal was present at 4% of the Met-eK-βgal level, and the concentration of Tyr-KKK1eK-βgal was virtually indistinguishable from the assay's background (cells transformed with vector alone) (Figure 3E). Crucially, the levels of these extra KRK-bearing Tyr-eK-βgal fusions in ubr1Δ cells became similar to that of Met-eK-βgal (Figure 3E), in agreement with the pulse–chase data (Figures 2A, B and 3A, B). Although the addition of extra KRK sequences to eK yielded considerable decreases in the t0.50–10 of the corresponding N-end rule substrates, by far the major effect of multiple KRK sequences was on the decay curves' ID5 term, which conveys the extent of degradation of a protein during or shortly after its synthesis (Figures 2A–C and 3A–C). To examine this issue in a different way, Tyr-eK-βgal, Tyr-K1eK-βgal and Tyr-KK1eK-βgal (Figure 1A) were produced in the JD54 S.cerevisiae strain, which expressed Ubr1p from the galactose-inducible, dextrose-repressible PGAL1 promoter. JD54 cells expressing one of the test proteins were labeled in raffinose-containing SR medium (no Ubr1p), incubated for 20 min in the same medium and thereafter shifted to galactose, where Ubr1p was induced. Even though the N-end rule pathway became hyperactive in the presence of galactose (Madura and Varshavsky, 1994; Ghislain et al., 1996; data not shown), the pre-labeled substrates Tyr-eK-βgal and Tyr-K1eK-βgal were barely degraded after the induction of Ubr1p; Tyr-KK1eK-βgal was degraded only slightly (Figures 2D and 3D). These findings were consistent with the earlier evidence for a strong retardation of the post-translational degradation of Arg-eK-βgal under the same conditions (R.J.Dohmen and A.Varshavsky, unpublished data). Thus, in contrast to a newly formed, conformationally immature βgal-based test protein, a conformationally mature βgal tetramer is a poor substrate of the N-end rule pathway even in the presence of N-degron enhancements such as the additional KRK sequences. It is the βgal moiety of these test proteins (Figure 1A) that was responsible for the time-dependent decline in the rate of degradation, because the kinetics of in vivo degradation of eK-DHFR-based N-end rule substrates was much closer to first-order decay (Lévy et al., 1999; data not shown). Locating N-degrons in the lysine–asparagine sequence space The earlier work, which led to the bipartite model of N-degron (Bachmair and Varshavsky, 1989; Hill et al., 1993), and particularly the present findings about the effects of adding KRK sequences to an eK-based N-degron (Figures 2 and 3) suggested that a substrate's destabilizing N-terminal residue and a sterically suitable internal Lys residue (or residues) are the two necessary and sufficient components of an N-degron. However, since both the eK-based and other previously analyzed N-degrons are embedded in complex sequence contexts (deGroot et al., 1991; Dohmen et al., 1994; Varshavsky, 1996), we wished to address the bipartite-degron model by constructing an N-degron from much simpler sequence motifs. Should this prove feasible, we also wanted to explore constraints on the structure of N-degrons through a screen in a simpler sequence setting. If the sequence space could be reduced strongly enough, one advantage of such a screen would be its exhaustiveness. The AAA codon for lysine differs by just one third-letter substitution from the codon for asparagine (AAU), a polar uncharged residue. Thus, one could attempt a screen for N-degrons in the sequence space of two amino acids: Lys (K) and Asn (N). A double-stranded oligonucleotide that encoded random 14-residue K/N sequences (see Materials and methods) was used to replace the sequence encoding 14 residues of eK immediately following the residue X (Figure 1B). In the resulting test proteins, this latter sequence, HGSGAWLLPVSLVRS (plus the sequence RS), followed the quasi-random 14-residue K/N sequence (Figure 1A). The resulting K/N-based extensions either lacked the lysines or contained a variable number of them between residues 2 and 16. The K-17 of eK (the only other lysine in eK) was replaced by Arg. In these test fusions, dha-UbR48-Arg-(K/N)14-eΔ-βgals, Arg was used as a destabilizing N-terminal residue (Figure 1B). The number of different 14-residue sequences containing exclusively K or N is 214 = 16 384. The bulk of a library of this complexity could be encompassed with conventional screening methods. Testing of the pRKN14-based library by amplifying it in E.coli indicated that >90% of the plasmids contained an oligonucleotide insert. The pRKN14 library was introduced into S.cerevisiae JD54 (Ghislain et al., 1996), which expressed Ubr1p from the PGAL1 promoter, and screened for colonies that stained blue with XGal [high levels of Arg-(K/N)14-eΔ-βgal] on dextrose (SD) plates but stained white [low levels of Arg-(K/N)14-eΔ-βgal] on replica-plated galactose (SG) plates. Approximately 20 000 colonies were screened this way. A total of 68 isolates were identified in which the activity of βgal was significantly higher in the absence than in the presence of the N-end rule pathway. The corresponding Arg-(K/N)14-eΔ-βgal test proteins were expressed in congenic ubr1Δ and UBR1 strains, and the ratio of βgal activities was determined for each of the test proteins. The results are summarized in Figure 4, which shows the K/N sequences of the 30 most active N-degrons, and the ratios of the corresponding βgal activity in the ubr1Δ strain to that in the UBR1 strain (higher ratios indicate stronger N-degrons). Remarkably, the strongest K/N-based N-degron was found to be more active than the strongest eK-based N-degron (Figure 4). Black bars in Figure 4 denote βgal activity derived from constructs carrying K/N-based N-degrons with lysines present at positions 3 and 15; the strongest N-degrons were largely of this class (Figure 4). K-15 was present in the 15 strongest K/N-based N-degrons except one (clone 132), which had K at position 14, and was also, presumably in compensation for the absence of K-15, one of the most lysine-rich N-degrons in this set (Figure 4). Similarly, K-3 was present in the 15 strongest K/N-based N-degrons except three (clones 3, 77 and 138) (Figure 4). All of these exceptional clones bore K-15; in addition, one of them (clone 77) bore K-4 and K-5, as well as K-14 and K-15 (Figure 4). Figure 4.N-degrons in the K/N sequence space. The deduced sequences of the identified K/N N-degrons are shown in conjunction with the bar diagram of their relative activity, defined as the ratio of βgal activities in the ubr1Δ versus UBR1 cells expressing a given Arg-(K/N)14-eΔ-βgal test protein. The top bar (vertical stripes) indicates the relative activity of N-degron in the original Arg-eK-βgal (ReK). Thirty 14-residue K/N extensions with the highest Ubr1p-dependent destabilizing activity are listed, out of the total of 68 isolates that were metabolically unstable in the presence but not in the absence of Ubr1p. The total number of possible 14-residue K/N sequences is 16 384 (see the main text). The Lys and Asn residues are denoted as the letter K and a hyphen, respectively. Position 1 in each clone was occupied by Arg. The extensions in which lysines were present either at positions 3 and 15, or only at 15, or only at 3, are marked, respectively, by the filled, striped and open bars. Download figure Download PowerPoint A completely uniform feature of all 68 K/N-based N-degrons was the absence of K from position 2 (Figure 4; data not shown), consistent with the fact that all of the previously examined N-degrons (Varsha
Referência(s)