Syd, a SecY-interacting Protein, Excludes SecA from the SecYE Complex with an Altered SecY24 Subunit*

Syd is an Escherichia coli cytosolic protein that interacts with SecY. Overproduction of this protein causes a number of protein translocation-related phenotypes, including the strong toxicity against the secY24 mutant cells. Previously, this mutation was shown to impair the interaction between SecY and SecE, the two fundamental subunits of the membrane-embedded part of protein translocase. We have now studied in vitrothe mechanisms of the Syd-directed inhibition of protein translocation. Pro-OmpA translocation into inverted membrane vesicles (IMVs) prepared from the secY24 mutant cells as well as the accompanied translocation ATPase activity of SecA were rapidly inhibited by purified Syd protein. In the course of protein translocation, high affinity binding of preprotein-bearing SecA to the translocase on the IMV is followed by ATP-driven insertion of the 30-kDa SecA segment into the membrane. Our experiments using 125I-labeled SecA and the secY24 mutant IMV showed that Syd abolished both the high affinity SecA binding and the SecA insertion. Syd was even able to release the inserted form of SecA that had been stabilized by a nonhydrolyzable ATP analog. Syd affected markedly the proteolytic digestion pattern of the IMV-integrated SecY24 protein, suggesting that Syd exerts its inhibitory effect by interacting directly with the SecY24 protein. In accordance with this notion, a SecY24 variant with a second site mutation (secY249) resisted the Syd action bothin vivo and in vitro. Thus, Syd acts against the SecY24 form of translocase, in which SecY-SecE interaction has been compromised, to exclude the SecA motor protein from the SecYE channel complex.

Protein translocation across the E. coli plasma membrane is facilitated by multiple Sec factors (for reviews, see Refs. 1 and 2). A membrane-interacting ATPase, SecA, binds with high affinity to the membrane containing the functional SecY/SecE components (3). This is then followed by the ATP-dependent insertion of a SecA segment into the membrane (4). ATP hydrolysis-dependent release of the preprotein and deinsertion of SecA allow the next cycle of polypeptide movement (4). SecY, SecE, and SecG are the essential core components in the mem-brane, having 10, 3, and 2 transmembrane segments, respectively. They not only provide the high affinity binding site for SecA (3) but allow productive SecA insertion (4,5), and they are thought to provide a channel-like pathway for the movement of polypeptide chain. These membrane proteins form a trimeric complex (6 -8) in which SecY interacts independently with SecE and SecG (9). Proteoliposomes reconstituted only from SecY and SecE exhibit some basic translocation activities in conjunction with SecA (10). In addition, the inclusion of SecG markedly stimulates the translocation activities of the SecYE proteoliposomes (8,11).
In vivo, SecY requires SecE as the stabilizing partner (12,13). Thus, uncomplexed forms of SecY are degraded rapidly by the FtsH proteolytic system (14). We proposed previously that a dominant negative variant of SecY, SecY Ϫd 1, with a small internal deletion, sequesters SecE in an inactive complex (15). Intragenic suppressor mutations that abolish the dominant effect preferentially occur within or around cytoplasmic domain 4 (C4) of SecY (16). The secY24 amino acid substitution (Gly 240 3 Asp) in C4 (17), which also suppresses secY Ϫd 1 intragenically, abolishes the SecY-SecE affinity co-isolation (9,16). Thus, C4 is an essential cytoplasmic domain that is important for the interaction between SecY and SecE. The importance of the second cytoplasmic segment (C2) of SecE for SecY-SecE interaction was also shown (18).
The secY Ϫd 1 mutation allowed us to identify a new gene, syd, as a multicopy suppressor against this mutation (19). The syd gene product (Syd) is a hydrophilic protein of 181 amino acid residues, which is associated weakly with the membrane, presumably via its interaction with the SecY protein. When Syd is overproduced in SecY Ϫd 1-expressing cells, it preferentially stabilizes the chromosomally encoded wild-type SecY protein, which is now given a greater opportunity to associate with SecE. This is the proposed suppression mechanism. Another striking phenotype associated with overproduction of Syd is its strong toxicity against the secY24 mutant cells. Presumably because of the defect in SecY-SecE interaction, the SecY24 mutant protein is degraded at 42°C. This degradation is primarily responsible for the protein export defect (14). However, the Syd-mediated impairment of protein export and cell viability is not accompanied by degradation of the SecY protein (19). Thus, Syd seems to directly interfere with the functioning of the SecY24 mutant protein. These observations indicate that Syd provides a useful means to investigate into the nature of protein interactions involving SecY.
Through our reversion studies of the secY24 mutant with respect to its susceptibility to Syd overproduction, we identified three new alleles of secY (20). While two of them affected residue 240 (the residue altered by secY24) of SecY, the third mutation, termed secY249 (for Ala 249 3 Val) alleviated the Syd sensitivity of the secY24 alteration (when placed in cis configuration with secY24) without restoring the temperature resistance or the SecY-SecE interaction. Thus, the secY249 mutation may affect the interaction between SecY and Syd.
This work was focused on the ability of Syd to inhibit protein translocase when its SecY subunit is altered to the SecY24 mutant form. Our in vitro experiments show that Syd interferes with productive interaction between SecA and the membrane-integrated SecYE complex.

EXPERIMENTAL PROCEDURES
Escherichia coli Strains and Plasmids--Plasmid pST67 contained the syd gene placed under the lac promoter and a translation-initiating SD sequence (5Ј-AAGGAG-3Ј) and was used as a Syd-overproducing plasmid. For its construction, a polymerase chain reaction was programmed on pST30 (19) with primers 5Ј-GGGGTACCCCAAGGAGAACA-GCTATGGACGATTTGACCGC-3Ј and 5Ј-GCCCTAGGGCAATGTCTT-CTCCGAATT-3Ј to produce a 0.57-kilobase pair fragment, which was then treated with KpnI and BamHI and cloned into pUC18. The resulting plasmid was confirmed for the Syd-coding region by sequencing. E. coli strain AD202 was a derivative of MC4100 (21) carrying ompT::kan (22). EM147 was a derivative of MC4100 carrying syd::kan (19). EM166 was a derivative of MC4100 carrying syd::cat and constructed essentially as described previously (19), except that it carried the chloramphenicol resistance determinant (cat) (instead of kan previously used) in the middle of the syd gene. Mutants carrying the secY24 or secY24 -249 mutation were described previously (20). These mutations were transduced into the genetic background of strain TW155 (a ⌬uncB uncC derivative of AD202; Ref. 5) and named EM153 and EM152, respectively. For these constructions, zhd-33::Tn10 (17) was used as a selective marker, and transductants were examined for temperature-sensitivity and Syd sensitivity (20); an isogenic secY ϩ strain was named EM154. Their further derivatives that received syd::cat from EM166 were named EM168 (secY24 syd::cat), EM167 (secY24 -249 syd::cat), and EM169 (secY ϩ syd::cat). IMVs 1 from strains of a particular secY phenotype exhibited essentially indistinguishable activities whether they were from strains of the syd ϩ or the syd::cat background.
Purification of Syd-Syd-overproducing cells (EM147/pST67 or CU141 (23)/pST67) were grown at 37°C to a middle to late log phase in 1.5 liters of peptone medium (20) supplemented with 0.017 M KH 2 PO 4 , 0.072 M K 2 HPO 4 , and ampicillin (50 g/ml). Isopropyl ␤-D-thiogalactopyranoside (1 mM) was added to induce the lac promoter-directed overexpression of syd. Syd was purified essentially as described previously (19), desalted by a NAP-25 column (Amersham Pharmacia Biotech) equilibrated with 50 mM Tris-HCl (pH 7.5) and 10% glycerol and stored at Ϫ80°C. Protein concentration was determined by a Micro BCA protein assay kit (purchased from Pierce) using bovine serum albumin as a standard.
In Vitro Translocation Assay-IMVs were prepared as described previously (24) from cells of the specified strains that had been grown at 30°C. They were treated with 6 M urea for 1 h at 0°C (25) when specified. Pro-OmpA was synthesized in vitro in the presence of [ 35 S]methionine as described previously (26). The transcription-translation mixture was then mixed with chloramphenicol (100 g/ml), sodium succinate (5 mM), IMV (100 g of protein/ml), and SecA (40 g/ml). Various concentrations of Syd were added before SecA when specified. After incubation at 37°C, samples were placed on ice and treated with proteinase K (100 g/ml) at 0°C for 30 min. Reactions were stopped by the addition of 1 mM phenylmethylsulfonyl fluoride followed by trichloroacetic acid precipitation. Samples were separated by SDS-PAGE and visualized by autoradiography as described previously (26), and the bands were quantified by a Fuji BAS2000 phosphor imager.
SecA Insertion and Binding Assays-SecA was iodinated with 125 I and subjected to the insertion assay essentially as described previously (4,5), except that Syd was added when indicated. SecA binding assays were similarly performed according to previously published procedures (3,5).
Trypsin Digestion of IMV--Urea-treated IMVs were incubated in the presence or absence of Syd (100 g/ml) with various concentrations of trypsin at 0°C for 20 min. A purified preparation of pro-OmpA (26 g/ml) (5, 28) was added in some experiments, but it did not affect the digestion patterns. Reactions were stopped by the addition of 1 mM phenylmethylsulfonyl fluoride followed by trichloroacetic acid precipitation. The precipitates were washed with acetone, solubilized in SDS sample buffer at 37°C for 5 min, and separated by SDS-PAGE (19). SecY and SecE were then visualized by immunoblotting (19), using antiserum against the N-terminal 22 residues of SecY (19) and antiserum against residues 64 -81 of SecE, respectively. The latter antiserum, a custom product of Sawady Technology, was raised in a rabbit using the synthetic peptide conjugated with keyhole limpet hemocyanin. Kaleidoscope prestained standards (purchased from Bio-Rad) were used as molecular weight markers in electrophoresis.

Syd Inhibits Protein Translocation into IMVs with the Mutated SecY24 Subunit of Protein Translocase-Overproduction
of Syd in the secY24 mutant cells leads to the severest extent of inhibition of protein export known for E. coli sec mutants (19) and appeared to merit investigations in vitro. We prepared IMVs from wild-type and secY24 mutant cells grown at 30°C, the permissive temperature for the mutant. An in vitro translation product of pro-OmpA, labeled with [ 35 S]methionine, was subjected to posttranslational translocation into IMV. When IMVs from secY24 mutant cells were preincubated at 0°C with increasing concentrations of Syd, translocation of pro-OmpA was progressively inhibited (Fig. 1). In contrast, pro-OmpA translocation into the wild-type IMV was not appreciably affected by Syd of up to 100 g/ml (Fig. 1). Dose-dependent inhibition of translocation was also observed when a crude cytosol fraction from the Syd-overproducing cells was added to the reaction with the SecY24 IMV. Immunoblotting experiments showed that the concentrations of Syd protein required for a particular level of inhibition were similar between the purified sample and the crude cytosolic fraction, indicating that Syd had not been inactivated during the purification procedures (data not shown). Neither bovine serum albumin, up to 1 mg/ml, nor heat-treated Syd exerted any appreciable inhibitory effect (data not shown). Thus, this inhibition was due to the functional Syd protein and not to some nonspecific factors such as increased protein concentration.
When the SecY24 IMV was first incubated with excess Syd, sedimented, and subjected to the translocation assay, substantial levels of translocation of pro-OmpA were observed. Syd concentration in the second incubation mixture was about 15% of that before centrifugation. We suggest that Syd inhibits translocation while in an equilibrium of association and dissociation with IMV. In fact, preincubation was not necessary, since Syd added after the initiation of translocation reaction also prevented the subsequent increase of the translocated molecules (data not shown). None of the translocase subunits was proteolyzed during the pretreatment with Syd or after the incubation at 37°C with Syd, as confirmed by direct immunoblotting experiments (data not shown).
Syd Rapidly Uncouples SecA Functions from Protein Translocation into the SecY24 IMV-The ATP hydrolysis activity of SecA that is stimulated by a precursor protein and membrane vesicles with intact SecYEG complex is termed translocation ATPase (29). We found that Syd was a potent inhibitor of translocation ATPase when SecA was combined with the SecY24 IMV. Whereas Syd (100 g/ml) inhibited this activity almost completely (Fig. 2, solid triangles), the same concentration of Syd only negligibly affected the activity in combination with the wild-type IMV (Fig. 2, solid circles). The inhibitory effect of Syd was specific for ATP hydrolysis that was coupled with translocation, such that neither the membrane ATPase activity (in the absence of preprotein) nor the intrinsic ATPase activity (without any other macromolecules) was affected significantly (data not shown). The addition of Syd into the reaction mixture in which SecA translocation ATPase was already activated by pro-OmpA and SecY24 IMV resulted in almost instantaneous inhibition of the ongoing reaction (Fig. 3).
We examined the SecA insertion reaction (4) using 125 Ilabeled SecA and urea-washed wild-type and SecY24 IMVs. In the absence of added Syd, both wild-type IMV and SecY24 IMV supported the pro-OmpA-and ATP-dependent formation of a membrane-protected 30-kDa fragment (Fig. 4, lane 4), although the efficiency was slightly lower for the SecY24 IMV. A nonhydrolyzable ATP analog, AMP-PNP, supported apparently pro-OmpA-independent insertion of SecA (Fig. 4, lane 6) as reported previously (5,30). When Syd was included in these reactions, the SecY24 IMV did not support any detectable insertion of SecA, irrespective of whether ATP or AMP-PNP was used (Fig. 4, middle panel, lanes 1-3). In contrast, Syd inhibited SecA insertion into the wild-type IMV only slightly (Fig. 4,  upper panel, lanes 1-3). These results indicate that Syd uncouples the SecA functions from the SecY24-mediated protein translocation reaction.
Syd Interferes with High Affinity Binding of SecA to the SecY24 IMV-To define the target reaction of the Syd inhibition, we carried out SecA binding assays. Various concentrations of SecA and IMV were incubated at 0°C, and the membrane-bound SecA was isolated for the Scatchard analysis of the binding. SecA binding to the SecY24 IMV was found to be abolished almost completely by Syd. Thus, the high affinity phase of the binding was hardly observed in the presence of Syd (Fig. 5B, solid triangles). High affinity binding of SecA to the wild-type IMV was lowered by Syd to only a small extent (Fig.  5A, compare open and solid circles). These results suggest that the Syd action to the SecY24 IMV can essentially be ascribed to the elimination of the high affinity SecA binding.
We then examined whether Syd interferes with the SecA-SecY interaction passively, for instance by steric hindrance, or if it acts more actively, for instance to change the subunit arrangement of the SecY-SecE complex. For this purpose, we studied the effect of Syd on a state of SecA that had already been inserted and stabilized as such by AMP-PNP. 125 I-Labeled SecA was first allowed to insert into wild-type or SecY24 IMV in the presence of AMP-PNP and then subjected to the second incubation with either Syd or buffer alone. The second incubation in the presence of Syd resulted in the disappearance of the 30-kDa fragment inserted into the SecY24 IMV (Fig. 6, lower  panel, lane 1), whereas SecA inserted into the wild-type IMV persisted (Fig. 6, upper panel, lane 1). Incubation without Syd did not cause any change in the intensity of the inserted SecA  (compare Fig. 4, lane 6, and Fig. 6, lane 2). Thus, Syd is able to dissociate SecA that is otherwise locked in the inserted state by the lack of ATP hydrolysis. We also noted that the AMP-PNPstabilized SecA was partially released from the SecY24 IMV, but not from the wild-type IMV, when unlabeled excess SecA was added, suggesting that SecA insertion occurs somewhat weakly against the mutant IMV.

SecY24 Alteration of Subunit Arrangements of Translocase and Further Evidence for the Syd Interaction with SecY-The secY24 alteration in SecY weakens the interaction between
SecY and SecE such that the SecY24-SecE complex comes apart after solubilization with a nonionic detergent (9,26). We examined trypsin digestion patterns of SecY and SecE in the membrane-integrated state. Wild-type and SecY24 IMV were treated with increasing concentrations of trypsin, and SecY and SecE were detected by immunoblotting using antisera against the N-terminal sequence of SecY and the central cytoplasmic region of SecE, respectively. Treatment of the wildtype IMV with low concentrations of trypsin produced an Nterminal fragment of about 21 kDa (Fig. 7, lanes 17 and 18). Its apparent molecular size suggests that the site of this cleavage was within cytoplasmic region 4 (C4). The SecY24 IMV gave different trypsin digestion patterns in two respects. First, the overall trypsin sensitivity as assessed by the disappearance of the intact SecY molecules was significantly higher for the SecY24 IMV than the wild-type IMV (Fig. 7, compare lanes  6 -10 with lanes 16 -20). Second, the 21-kDa N-terminal fragment produced from the SecY24 IMV (Fig. 6, lane 7) was further shortened to a slightly faster migrating product when trypsin concentrations were increased (Fig. 7, lanes 8 and 9). Although the SecY24 amino acid alteration (Gly 240 to Asp) is also in C4 domain of SecY, it is unrelated to any amino acid residues directly relevant to the trypsin-mediated hydrolysis of peptide bonds. Thus, this sequence alteration may have caused a change in local conformation of SecY or in subunit arrangements of the SecYEG complex.
We found a striking difference in the trypsin susceptibility of SecE between the wild-type IMV and the SecY24 IMV. SecE in the former IMV resisted trypsin of the highest concentration examined (Fig. 7, lower panel, lanes 16 -20), although some decrease in the intensity of SecE, in parallel with that of the intact SecY, was observed at higher trypsin concentrations

FIG. 4. Effects of Syd on SecA insertion into IMVs. 125 I-Labeled
SecA (4 g) and urea-treated IMVs (60 g of protein) from secY ϩ (EM154; upper panel), secY24 (EM153; middle panel), and secY24 -249 (EM167; lower panel) cells were preincubated at 0°C in a total volume of 230 l, and IMV-SecA complex was isolated by centrifugation, resuspended, and divided into six portions, each of which was then subjected to a SecA insertion reaction at 37°C for 15 min in a total volume of 50 l containing the 125 I-SecA-bound IMV (100 g/ml), pro-3His-OmpAЈ (66 g/ml; lanes 1 and 4), either ATP (2 mM; lanes 1, 2, 4, and 5) or AMP-PNP (2 mM; lanes 3 and 6), and Syd (100 g/ml; lanes 1-3). The samples were treated with trypsin (1000 g/ml at 0°C for 15 min), and the 30-kDa fragment was visualized by SDS-PAGE and autoradiography. (Fig. 7, lanes 11-20, compare upper and lower panels). In contrast, SecE in the SecY24 IMV was digested almost completely by trypsin (Fig. 7, lower panel, lanes 9 and 10). SecE was more trypsin-sensitive than SecY24 itself (compare upper and lower panels for lane 5 of Fig. 7). These results lend further support to the notion that subunit arrangements of the SecYEG complex have been altered by the SecY24 amino acid alteration, resulting in significantly impaired SecY-SecE interaction.
We then repeated the trypsin digestion experiments in the presence of an excess of Syd protein. While the addition of Syd did not appreciably affect the trypsin digestion profiles of SecY or SecE in the wild-type IMV, it markedly changed the pattern of SecY cleavage in the SecY24 IMV. The 21-kDa N-terminal fragment was now shifted to an apparent molecular mass of 27 kDa (Fig. 7, upper panel, lanes 3 and 4). Thus, the original two cleavage sites (presumably in C4) were somehow protected by Syd. Thus, Syd interacts specifically with SecY24. However, it did not exert any protecting effect on SecE (Fig. 7, lower panel,  lanes 4 and 5).
A Second Site Mutation, secY249, Cancels the Inhibitory Effects of Syd-Previously, we isolated Syd-resistant variants from the secY24 mutant (20). The mutations converged into three alleles in addition to the true reversion. Two of the alleles change residue 240 affected by the secY24 mutation, restoring the SecY-SecE interaction (20). The other allele, secY249 (for an Ala 249 to Val change), is an intragenic suppressor mutation that suppresses only the sensitivity to overproduction of Syd, without any effect on the temperature-sensitive protein export defect of the secY24 mutation (20). IMVs prepared from the secY24 -249 double mutant supported pro-OmpA translocation with an efficiency similar to that of the SecY24 IMV. However, it was only slightly inhibited by Syd (Fig. 1). The double mutant IMV also supported the translocation ATPase activity of SecA even in the presence of Syd (data not shown). Insertion of the 30-kDa fragment into the double mutant IMV was not inhibited by Syd (Fig. 4, lower panel, lane 1). The SecY24 -249 IMV gave tryptic digestion patterns of SecY and SecE that were essentially identical to those of the SecY24 IMV. However, they were unaffected by the presence of excess Syd (data not shown). Thus, the effects of the secY249 mutation on the phenotypes of the secY24 mutation have been reproduced in vitro. These results suggest that the Syd effect is brought about through specific interaction between SecY and Syd, in which the residue 249 of SecY plays an essential role. DISCUSSION Syd, originally identified as a multicopy suppressor of the secY Ϫd 1 mutation (19), directly interacts with SecY. Overproduction of Syd stabilizes the wild type SecY protein when it alone is overproduced or when it fails to associate with SecE because of the presence of the excess amounts of SecY Ϫd 1, the dominant negative SecY variant. The association of Syd with the plasma membrane is saturable such that SecY overproduction increases the membrane-bound state of Syd. We demonstrated in this study that Syd markedly altered the trypsin digestion pattern of SecY24 in IMV, lending further support to the notion that Syd interacts with SecY.
Evidence indicates that the secY24 mutation impairs SecY-SecE interaction. The temperature sensitivity of this mutant is partially suppressible by overproduction of SecE, 2 and the mutation can act as an intragenic suppressor against the SecEsequestering effect of the secY Ϫd 1 mutation (16,20). After solubilization of the membrane with a nonionic detergent, a complex between SecY24 and SecE cannot be isolated (9,16). The temperature-sensitive protein export defect of the secY24 mutant is due to the FtsH-mediated degradation of the SecY24 protein at 42°C (14,16). These observations suggest that the impaired SecY-SecE interaction is the primary cause of the secY24 defect. In further support of this notion, we found that SecE in the SecY24 IMV was more accessible to trypsin digestion than that in the wild-type IMV. The temperature sensitivity of the secY24 mutant could be due to temperature-dependent exacerbation of the interaction defect, temperaturedependent activation of the proteolytic system, or both.
When Syd is overproduced in the secY24 mutant, protein export is rapidly blocked followed by the loss of viability even at 30°C (19). The severity of this block makes it the most useful system to accumulate chemical amounts of a precursor protein with uncleaved signal sequence within the cell (24). The rapidity and the lack of accompanied SecY proteolysis suggest that this inhibition is brought about by direct interference by Syd with the translocase functions.
We reproduced the inhibitory effect of Syd in the in vitro translocation reaction of pro-OmpA. The concentration of Syd required for 50% inhibition of translocation was 5-10 g/ml (or 250 -500 nM) in a reaction containing IMV of 100 g/ml proteins. Assuming that a cell contains 500 molecules of SecY (31) and that 10% of total proteins are located in the inner membrane, the above concentration of IMV corresponds to 250 ng/ml (or 5 nM) of SecY. Thus, numbers of the Syd molecules required for efficient inhibition far exceed those of SecY molecules. We have shown that Syd inhibits the high affinity binding of SecA to the IMV containing the SecY24 form of the translocase subunit. This could be due to a steric hindrance by the SecY-bound Syd with respect to SecA binding. There may be a competition between Syd and SecA for the interaction with SecYEG translocase. Indeed, increased concentrations of SecA acted to partially overcome the inhibition caused by a low concentration (10 g/ml) of Syd. Although quantitative discussion requires knowledge about the K d value of Syd-IMV binding, our IMV re-isolation experiments suggested that the dissociation rate of Syd is considerably high. Thus, real time binding/dissociation measurements will be required to determine the K d value.
While the simple steric hindrance discussed above is possible, we believe it is more likely that Syd causes more drastic structural changes to the SecYE complex. Syd was found to inhibit the SecA insertion reactions. We previously noted that the SecY205 mutant form of SecY (32) did not allow the preprotein-and ATP-dependent mode of SecA insertion, while it still allowed SecA insertion in the presence of AMP-PNP (5). Our present results show, in contrast, that Syd abolishes both of these modes of SecA insertion reactions. Thus, it is conceivable that the SecY205 defect lies in a specific process of the SecA insertion reaction cascade, whereas Syd impairs the system more generally or inhibits a very early event. Indeed, Syd inhibited the high affinity binding between SecA and the SecY24 IMV. Furthermore, it acted to dissociate the inserted SecA segment that had been stabilized by AMP-PNP. In these cases, the subunit arrangement of the SecYE complex may have been altered, leading to the elimination of both the high affinity SecA binding site and the structural element (or channel) required for maintaining the "inserted" state of SecA.
Previous studies suggested that the high affinity binding of SecA requires both SecY and SecE (3,33,34). Thus, it seems to be a logical explanation that the presence of a high concentration of Syd acts to dissociate the SecY-SecE complex that is already compromised by the SecY24 alteration. An interesting possibility here is that Syd may bind preferentially to an uncomplexed state of SecY. This is consistent with the known ability of Syd to stabilize overproduced SecY (19). Although the physiological role of Syd is elusive (19), some speculations may be useful. For instance, Syd may act as a reservoir for unassembled forms of SecY, thus preventing its deleterious effects on the cell (14). Alternatively, it may serve as a regulatory factor that negatively controls the translocase function by interfering with either SecY-SecE or SecY-SecA binding. At any rate, interaction between Syd and SecY seems specific in that it is disrupted by the secY249 alteration of residue 249 of SecY. Syd will be useful to probe the architecture of protein translocase from the functional point of view.