| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 278, Issue 49, 49316-49322, December 5, 2003
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
From the Department of Structural Biology, Stanford University School of Medicine, Stanford, California 94305
Received for publication, August 11, 2003 , and in revised form, September 22, 2003.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
barrel structures. Thus, SurA recognizes a peptide motif that is characteristic of integral outer membrane proteins. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
The surA sequence delineates four separate domains following the leader sequence: an amino-terminal domain of 
150 amino acids, followed by two peptidyl-prolyl isomerase domains of 100 residues each (PPIase1 P1 and P2) that are homologous to the PPIase parvulin (6), and finally, a carboxyl-terminal domain of 40 amino acids. Although SurA was initially designated a PPIase because of the presence of two parvulin-like domains (3), it has been shown that the P1 and P2 domains can be deleted without loss of in vivo function, whereas both the N and C domains are required (7). Additionally, it has been shown that in vitro, both SurA and an amino-terminal fragment thereof (residues 21-133) bind a peptide derived from somatostatin (sequence AGSKNFFWKTFTSS) that is devoid of proline residues, giving an example where the PPIase domains are not required in the SurA protein, and proline is not required in the target peptide for peptide binding (8).
The x-ray crystallographic structure of SurA reveals a protein with a core module constituted of the N, P1, and C domains, and a satellite P2 domain tethered to and 30 Å distant from, the core module (9). The core module has an extended crevice 50 Å in length that could readily accommodate polypeptides. In the SurA crystals, a segment of peptide from one molecule is bound in the crevice of a neighbor molecule. Models for the molecular chaperone activity of SurA envisage binding of porin polypeptide segments; the crevice of the core module is an attractive candidate for the putative peptide binding activity.
As one approach to characterizing a peptide binding activity of SurA, we have used phage display to select short peptides that bind with significant affinity (10, 11). We find that both full-length SurA and its core module have similar peptide binding specificities and affinities for peptides with a consensus sequence of the form aromatic-polar-aromatic-nonpolar-proline (Ar-polar-Ar-nonpolar-Pro). Ar-X-Ar tripeptide motifs (where X can be any residue) are found with high frequency in the sequences of outer membrane proteins, where they typically localize to the aromatic bands in the structures of the proteins. The data we present here suggest that SurA recognizes peptides that would be characteristic of outer membrane proteins consistent with its proposed role in facilitating the correct folding and assembly of these proteins.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
P2), residues 21-281 and 390-428) was created using the overlap-extension methodology (12). In the first stage of PCR amplification, the coding sequence for the amino-terminal fragment of the deletion construct, residues 21-281, was amplified from the expression plasmid for full-length SurA using primers A (5'-AATTAACATATGGCCCCCCAGGTAGTCG-3') (NdeI site in bold) and B (5'-ACGATCTTTCTGCGCAGCGTCCGAGATATTTTTGCTTTCGCC-3'). Similarly, the coding sequence for residues 390-428 was amplified using primer C (5'-GACGCTGCGCAGAAAGATCGTGCATACCGC-3') (complementary overlap between primers B and C is underlined) and primer D (5'-AACATTGGTACCCTTGGCAAAGC-3') (KpnI site in bold). The PCR fragments from the first stage were used in a second PCR amplification along with primers A and D to produce the coding sequence for the deletion construct, which was then cloned into the pTYB1 intein-fusion expression vector (IMPACT system; New England Biolabs) using the NdeI and KpnI restriction sites. The SurA(P2) protein was expressed as an intein fusion and purified following the protocol used previously for the SurA protein (9).
Phage Display—Three rounds of phage display selection were performed independently for both mature SurA and SurA(P2) protein using a heptapeptide-presenting phage library (library Ph.D.-7 from New England BioLabs) following essentially the manufacturer's protocol. Briefly, 1.5 ml of a 100 µg/ml protein solution in 100 mM NaHCO3, pH 8.6, was added to a 60-mm Petri dish. The target protein was allowed to coat the Petri dish surface overnight in a closed humidified container at 4 °C. Unbound protein was discarded, and the surface of the dish was blocked with blocking buffer (100 mM NaHCO3, 5 mg/ml bovine serum albumin, pH 8.6) for 1 h at 4 °C. After this period, the dish was washed six times with TBST buffer (50 mM Tris/HCl, 150 mM NaCl, 0.1% (v/v) Tween 20, pH 7.5). Immediately after the wash step, 2 x 1011 pfu (from the phage library) pre-diluted in 1 ml of TBST buffer were poured on the dish and rocked on a rotary shaker for 30 min at 25 °C. Unbound phage remaining in the solution was discarded, and the dish was washed 10 times with TBST buffer. To elute the bound phage, 1 ml of 100 µg/ml SurA or SurA(P2) was added and rocked on a rotary shaker for 60 min at 25 °C. Eluted phage was amplified in E. coli ER2738, and titer was established using standard phage biology protocols. A second and third round of selection were performed using the same protocol as the first round, except for two variations: (a) Tween 20 concentration in the TBST buffer was increased to 0.5% (v/v) to achieve more stringent selection, and (b) the phage elution period was shortened from 60 min to 30 min. After the third round of selection, aliquots of the phage solutions were plated on LB/IPTG/Xgal (IPTG, isopropyl--D-thiogalactopyranoside; Xgal, 5-bromo-4-chloro-3-indolyl -D-galactopyranoside) plates. Ten plaques were selected and purified; single-strand DNA was isolated from phage derived from each plaque and sequenced in the region encoding the heptamer peptide.
Affinity Evaluation of Selected Clones by ELISA—The wells of 96-well ELISA plates were coated overnight at 4 °C in a humidified closed container with 100 µl/well of 100 µg/ml SurA or SurA(P2) protein in 100 mM NaHCO3, pH 8.6. As a control to identify plastic-binding phage, protein was omitted from some wells in parallel. The following day, protein solutions were discarded and wells were blocked for 2 h at 4 °C with blocking buffer (100 mM NaHCO3, 5 mg/ml bovine serum albumin, pH 8.6). The ELISA plates were then washed six times with TBST buffer. Phage dilutions were made in a separate plate that had been treated with blocking buffer; the first well contained 1011 pfu/200 µl of TBST, and ten sequential 4-fold dilutions were made resulting in a range of 105-1011 pfu/200 µl. Diluted phage solutions were transferred into ELISA plates coated with the target protein and incubated 1 h at room temperature with agitation. Wells were washed 6 times with TBST, after which 200 µl of horseradish peroxidase-conjugated anti-M13 antibody (Amersham Biosciences catalog number 27-9411-01) diluted 1:5000 times in blocking buffer was pipetted into the ELISA plate wells. After 1 h of incubation at room temperature, the antibody solution was discarded and the plate was washed six times in TBST. Finally, 200 µl of freshly prepared horseradish peroxidase substrate (36 µl of 30% H2O2 added to 21 ml of ABTS stock (22 mg of 2,2'-azino-bis(3-ethylbenzthiazoline)-6-sulfonic acid in 100 ml of 50 mM sodium citrate, pH 4.0)) was added to the ELISA plate wells, and the reaction was allowed to proceed for 15 min at room temperature. Product of the horseradish peroxidase reaction was monitored as optical absorbance at 
= 410 nm with a microplate reader.
Isothermal Titration Calorimetry—A heptapeptide of sequence WEYIPNV was synthesized and purified by high pressure liquid chromatography by the Stanford Protein and Nucleic Acid facility. A peptide of sequence YYY was purchased from Sigma. Peptides of sequence DNRDGNVYYF and DNRDGNVYQF were a generous gift from Dr. Robert Sauer (Massachusetts Institute of Technology). Protein and peptide concentrations were determined by spectrophotometer in 6 M guanidine hydrochloride, 20 mM NaH2PO4, pH 6.5, using calculated molar extinction coefficients 
276 = 28,850 M-1 cm-1 for SurA, 23,450 for SurA(P2), 6,850 for WEYIPNV, 2900 for DNRDGNVYYF, 1450 for DNRDGNVYQF, and 4350 for YYY (13).
Solutions of SurA or SurA(P2) at 36-160 µM protein concentration in 150 mM NaCl, 20 mM NaH2PO4, pH in the range of 4.8-9.5, were placed in a 1.4-ml sample cell for a model VP-ITC MicroCalorimeter (MicroCal, Inc.). The peptide WEYIPNV was dissolved in the same buffer as the protein at 1.5-2 mM concentration and placed into a 250 µM injection syringe. The system was allowed to equilibrate to the desired experimental temperature in the range of 20-37 °C. In accordance with recommended procedures from the manufacturer, a single initial injection of 1-2 µl of peptide solution was made into the sample cell followed by 240 s of equilibration. This data point was not included in the analysis. The initial injection was followed by 24-36 injections of 6-10 µl of volume at a rate of 2 µl/s with equilibration intervals of 300-420 s between injections. The cell contents were stirred continuously during the experiment at 310 rpm. A control experiment to evaluate the heat contribution from peptide dilution was performed at 30 °C using identical injections of peptide solution into the buffer in the absence of protein. The heat contribution was found to be minimal (0.1% of the experimental signal) and was not corrected for during the analysis. The experimental data were fit with programs supplied by the manufacturer to a model with one binding site parameterized by the equation,
 | (Eq. 1) |
H is enthalpy change, and Ka is the association constant; all parameters are in volume (V0) to obtain apparent binding stoichiometry n, association constant Ka, and H values. From these parameters and the equations G = RT ln(Ka) and G = H - TS, we computed G and S.
ITC experiments with other peptides were attempted using similar buffer conditions (150 mM NaCl, 20 mM NaH2PO4, pH 7.3, T = 30 °C) and generally higher protein concentrations: for DNRDGNVYYF, 1.95 mM peptide and 282 µM SurA(P2); for DNRDGNVYQF, 1.5 mM peptide and 106 µM SurA(P2); for YYY, 3.86 mM peptide and 250 µM SurA.
Computational Sequence Analysis—For proteome analysis, the predicted open reading frames of the E. coli K12 genome were used.2 The annotations available at the E. coli Cell Envelope Protein Data Collection (ECCE; www.cf.ac.uk/biosi/staff/ehrmann/tools/ecce/ecce.htm) were used to identify the inner membrane, and periplasmic and outer membrane proteins; the remaining proteins were designated to be cytoplasmic. After correcting for duplication resulting from alternative names used for some proteins, the data set consisted of 74 outer membrane, 129 periplasmic, 593 inner membrane, and 3,482 cytoplasmic proteins.
The normalized occurrence of the Ar-X-Ar motif per hundred amino acids (where Ar represents Trp, Tyr, or Phe and X is any amino acid) was computed as (number of occurrences) x 100/(protein length) for each protein. For each protein set, the normalized rate of occurrence of the Ar-X-Ar motif was computed as the arithmetic average for all proteins in the set. The frequency of occurrence of individual amino acids (freq(X)) was computed as the total number of occurrences in all proteins of that set divided by the total number of amino acid residues in the set. The expected frequency of a random occurrence of the Ar-X-Ar motif within a polypeptide of 100-residue length was computed from the frequencies of occurrence of individual amino acids as 98 x (freq(Trp) + freq(Tyr) + freq(Phe))2 for each protein set (outer membrane, periplasmic, inner membrane, and cytoplasmic).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
2.8 x 109 transformants (New England Biolabs), which can be compared with 1.3 x 109 different sequences that are possible for a heptapeptide so that the initial library included essentially all possible 7-mer amino acid sequences.
The bipartite structure of the SurA protein (in which the second PPIase domain, P2, is a satellite of the core structure of the molecule) suggests a partitioning of activities between the core module (which includes domains N, P1, and C) and the P2 domain. Further, an extended crevice within the core module is suggestive of a peptide binding capability that may be related to the molecular chaperone activity of SurA (9). In this context, we have made a construct in which the peripheral PPIase domain is deleted (SurA(P2)) and have characterized its peptide binding activity in parallel with that of SurA. Although we do not report the structure of the SurA(P2) protein here, we have crystallized it, collected data to 3.5 Å on one crystal form, and determined a solution for the structure by molecular replacement, attesting to the structural integrity of the protein from which the P2 domain has been deleted.
Peptides Selected by Phage Display—Three cycles of phage display selection were carried out independently with both SurA and SurA(P2) proteins each as the affinity targets. After the third cycle, ten plaques from each selection were chosen randomly and purified, and the DNA sequence encoding the heptapeptide tag of each phage was determined. (One phage stock selected on SurA(P2) gave an ambiguous sequence; because a clear consensus emerged from among the other nine, it was not pursued further). The derived amino acid sequences of the displayed peptides are shown in Table I.
|
P2) were determined by ELISA using standard protocols. Phage binding was detected with a chromogenic horseradish peroxidase reaction from bound horseradish peroxidase-conjugated anti-M13 antibody. Serial dilutions of phage spanning a range of 106 plaque-forming units (pfu) were used. Control experiments (in which SurA and SurA(P2) proteins were omitted) were done to identify phage with an artifactually high affinity for the ELISA plate wells (plastic binders).
Results of the ELISA experiments are shown in Fig. 1 and included in Table I. For both SurA and SurA(P2), the majority of the peptides cluster in a similar affinity versus phage profile with a midpoint in the range of 108-109 pfu. Three peptides were identified as having 2-3 orders of magnitude lower relative affinity than the majority of peptides; they are denoted as "low" relative affinity in Table I. One peptide (of sequence WEYIPNV) showed significantly higher affinity than others for SurA in ELISA assays; for SurA(P2), another peptide sequence (WTYIPPV) also had a higher than typical affinity (Fig. 1). One peptide (with sequence FPCFFCY) was initially suspected to be a plastic binder because of its high content of aromatic residues and was found to have a significant affinity for the ELISA plate wells at high phage concentration. However, separate measurements of its affinity versus phage profile revealed that it also has a high affinity for SurA at a low phage concentration.
|
P2); as a case in point, one identical peptide sequence (WTYIPPV) emerged in both selections from a library of 1.3 x 109 independent sequences. It is also notable that the ELISA profiles are similar for both SurA and SurA(P2); this supports a suggestion that the core domain of SurA is largely, if not wholly, responsible for the peptide binding we are observing. Finally, a consensus pattern that emerges is Ar-polar-Arnonpolar-Pro. Among the modest number of individual phage plaques that were sequenced, this consensus occurs more frequently for SurA(P2) protein (7 of 9 plaques) than for SurA (6 of 10).
One deviation from this consensus that still has high affinity is the sequence FDFNRRI selected by SurA(P2), which lacks the hydrophobic and proline residues; however, it retains the Ar-polar-Ar motif. The question this raises is whether the first three residues might constitute a minimal recognition sequence for binding.
A second variant is the sequence FPCFFCY, which also has high affinity. It has four aromatic residues and an Ar-polar-Ar motif in the final positions rather than the initial positions, suggesting a general preference for aromatic rich peptides by SurA.
A third variant from this consensus is the peptide PLPISPR selected by SurA; this peptide has an unusually high proline content (3/7). It is unclear why this peptide would have an affinity similar to those fitting the consensus pattern; we cannot rule out the possibility that it is binding at a different site or in a different manner than the other peptides.
Isothermal Titration Calorimetry—To quantify the peptide binding affinity, we have synthesized the peptide displaying high affinity in ELISA (WEYIPNV) and have carried out ITC experiments with it. A typical calorimetric titration is shown in Fig. 2A. The data on heat enthalpy change versus molar ratio of peptide to protein have been fit with a theoretical curve that models a single set of identical peptide binding sites on the protein; a typical fit is shown in Fig. 2B. Because this model fit the data well, models with greater complexity were not utilized. Values of the parameter representing the number of peptide molecules bound per protein molecule ranged from 0.86 to 0.88 for SurA and from 1.04 to 1.41 for SurA(P2); we interpret these numbers to indicate a single peptide binding site per protein.
|
P2) at 30 °C, pH 7.3, gave values for the dissociation constant Kd of 3.95 ± 0.12, 4.15 ± 0.10, and 3.66 ± 0.22 µM. By contrast, independent measurements under identical conditions on the same batch of protein typically agreed to three significant figures in values for fitted parameters. Consequently, the dependence of binding parameters on temperature, pH, and salt concentration were all measured on the same batch of protein.
|
P2) over the range of 20-37 °C. The Kd values of the peptide for both proteins lie in the range of 1-14 µM over this temperature range. SurA(P2) consistently has a slightly (less than 2-fold) higher affinity than full-length SurA for the peptide. This is consistent with the ELISA data for relative affinities of peptide-tagged phage particles for the two proteins where the binding midpoint occurred at a slightly lower phage concentration for SurA(P2) than for SurA for each of the high affinity peptides (Fig. 1). Both the absolute values of the binding parameters and their dependence on temperature were similar for SurA and SurA(P2).
Because the SurA(P2) protein is easier to produce in large quantities and is more well behaved in solution than SurA, a more extensive characterization of the binding was carried out with SurA(P2). We see that the peptide affinity decreases with increasing temperature and increasing pH. It does not show a significant dependence on the ionic strength of the solvent; the peptide affinity in the presence of 1 M NaCl is equal to its affinity in the presence of 0.15 M NaCl. There is a slight decrease in H with increasing temperature, suggesting that a significant component of the peptide-SurA interaction is hydrophobic, which would be expected from the nature of the consensus sequence.
Recently it was shown that peptides bearing the carboxyl-terminal Ar-X-Ar motif, which is prevalent in outer membrane porins, activate the periplasmic unfolded protein response by binding the DegS protease. In this context, and because the consensus SurA binding sequence derived from phage display experiments bears an amino-terminal Ar-X-Ar motif, we effected ITC experiments on peptides used in the DegS study that have a carboxyl-terminal Ar-X-Ar motif, as well as on a short tripeptide of aromatic residues. ITC with the peptides DNRDGNVYYF and DNRDGNVYQF (Ar-X-Ar motif underlined) gave heat absorption traces that failed to reach saturation under the conditions used (100-280 µM SurA(P2)) indicating substantially weaker binding of these peptides compared with the peptide of sequence WEYIPNV. Accurate affinity constants were not determined for these peptides because the calorimetry experiments would have required excessively high protein concentrations to compensate for the low affinities. For comparison, ITC was also carried out with the tripeptide of sequence YYY and 250 µM full-length SurA protein. Again, the heat absorption trace failed to saturate indicating that Kd is significantly greater than 250 µM for this tripeptide.
Prevalence of the Ar-X-Ar Motif—A genome-wide analysis of predicted E. coli open reading frames reveals (a) aromatic residues have a higher frequency of occurrence in membrane proteins than in cytoplasmic and periplasmic proteins; in particular, tyrosine is prevalent in outer membrane proteins, whereas phenylalanine and tryptophan are frequent in inner membrane proteins (Table III); (b) the Ar-X-Ar motif has a higher than random occurrence in membrane proteins, occurring almost twice as frequently as in periplasmic and cytoplasmic proteins where its average frequency of appearance is essentially random (Table IV and Fig. 3). The presence of a carboxyl-terminal Ar-X-Ar motif, which has previously been recognized as occurring with high frequency in integral outer membrane proteins, is found in 28% of outer membrane proteins in the E. coli genome. In contrast, it occurs in only 0.5% of inner membrane proteins, it is not found in periplasmic proteins, and it occurs in 33, or 1%, of proteins designated to be cytoplasmic (although inspection of the sequences suggests that several of these proteins may actually be outer membrane proteins). Caveats to bear in mind when evaluating these statistics include: (a) the classification of proteins into the different classes is an ongoing process, and it is likely that some of the current annotations are erroneous; and (b) not all proteins assigned to the outer membrane class are integral membrane proteins. Nonetheless, the results are derived from statistics on a large population, and hence the general trends we observe are unlikely to be altered by variances introduced by evolving annotations.
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
P2). ITC with one representative heptapeptide, WEYIPNV, shows the affinity to be 1-14 µM under conditions relevant to E. coli physiology (T,20-37 °C); SurA(P2) has a slightly higher affinity for the peptide than full-length SurA protein. These data demonstrate that the peptides bind the core domain SurA(P2); the peripheral PPIase domain P2 has little apparent effect on the peptide selectivity. Generally consistent with this scheme are earlier observations showing that SurA and an amino-terminal fragment of SurA bind the peptide AGSKNFFWKTFTSS, which has an internal Ar-X-Ar motif (underlined), and which does not have proline (8).
Sequences of bacterial outer membrane protein have a high occurrence of tripeptides with an Ar-X-Ar pattern; structures of bacterial porins and other outer membrane proteins reveal a rationale for this. The known structures are antiparallel barrels that have aromatic bands encircling the exterior surface of the molecule at levels corresponding to the extremities of the hydrophobic interior of the lipid bilayer (14, 15). The motif Ar-X-Ar occurring within the aromatic band can place two aromatic side chains on the outer face of the sheet. For example, the E. coli OmpF protein sequence has eight Ar-X-Ar motifs, all of which occur in the aromatic bands of the structure (and in seven of which the amino acid X is polar, either Asp, Asn, Glu, or Gln, consistent with a more restrictive consensus motif suggested by the phage display results in which the second residue is polar). This suggests that SurA is selecting for a peptide motif that is prevalent in, and to a large extent diagnostic of, outer membrane protein sequences.
In early work, Struyve et al. (16) showed that a carboxyl-terminal phenylalanine is essential for the correct assembly of the E. coli PhoE porin in the outer membrane and further that a carboxyl-terminal Ar-X-Ar motif (where X can be any amino acid) is common among outer membrane proteins of Gram-negative bacteria. More recently, it has been shown that peptides mimicking outer membrane proteins through a carboxyl-terminal YQF or YYF sequence activate the periplasmic DegS protease in vitro through specific binding to its PDZ domain (17). By implication, exposed carboxyl-terminal YXF motifs of outer membrane proteins, by activating DegS in vivo, would initiate the proteolytic cascade that induces the periplasmic unfolded protein response in E. coli. The peptide affinities for the DegS PDZ domain are 0.6 and 15 µM for model peptides ending in YYF and YQF respectively; the latter value is similar to the affinity of the peptide WEYIPNV for SurA. However, somewhat surprisingly, the peptides that terminate in YYF and YQF used in the DegS study do not bind with similar affinity to SurA nor does the control tripeptide of sequence YYY; their affinities are at least an order of magnitude weaker than that of the peptide WEYIPNV. These data imply that SurA does not compete directly with DegS for binding of carboxyl-terminal Ar-X-Ar motifs in outer membrane proteins and that SurA is not a direct negative regulator of the periplasmic unfolded protein response.
The data also suggest that there are features in the heptameric peptides selected by phage display in addition to the Ar-X-Ar motif that contribute to affinity for SurA. The emergence of proline at position five of the consensus peptide sequence argues that the minimum length of the peptide bound by SurA is at least five residues and that binding is optimal with proline at that position. An attractive candidate for a proline binding site that is consistent with this observation is the P1 prolyl isomerase domain of the core module. The SurA(P2) core module has a deep extended channel, described previously (9), which is a candidate site for peptide binding (Fig. 4). However, neither inspection of the surface of the channel nor examination of the intermolecular crystal contacts formed by a segment of peptide that binds in the channel (9) suggests specific unambiguous docking interactions through which the residues of the WEYIPNV peptide may bind SurA; the structural details of the peptide binding remain an unsolved problem.
|
A remaining question is, what is the functional role of the satellite P2 domain? Its presence has minimal influence on the peptide binding specificity and affinity; the spatial separation of the P2 domain and core module are suggestive of two separate activities for SurA, only one of which (the peptide binding activity of the core module) is revealed in the phage display experiments. SurA is proposed to facilitate the folding of individual porin protomers prior to their assembly into stable trimers. In this context, one folding mechanism that can be suggested is one in which the core module sequesters segments of polypeptide, whereas the P2 domain catalyzes peptide backbone isomerization at sites remote from the sequestered site during the folding process.
|
FOOTNOTES |
|---|
To whom correspondence should be addressed. Tel.: 650-723-6589; Fax: 650-723-8464; E-mail: Dave.McKay{at}Stanford.edu.
1 The abbreviations used are: PPIase, peptidyl-prolyl isomerase; Ar, aromatic; ELISA, enzyme-linked immunosorbent assay; ITC, isothermal titration calorimetry; pfu, plaque-forming units.
2 GenBankTM release March 7, 2003, NC_000913
[GenBank]
.gbk available at ftp.ncbi.nih.gov.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S. S. Justice, S. R. Lauer, S. J. Hultgren, and D. A. Hunstad Maturation of Intracellular Escherichia coli Communities Requires SurA. Infect. Immun., August 1, 2006; 74(8): 4793 - 4800. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. S. Justice, D. A. Hunstad, J. R. Harper, A. R. Duguay, J. S. Pinkner, J. Bann, C. Frieden, T. J. Silhavy, and S. J. Hultgren Periplasmic Peptidyl Prolyl cis-trans Isomerases Are Not Essential for Viability, but SurA Is Required for Pilus Biogenesis in Escherichia coli J. Bacteriol., November 15, 2005; 187(22): 7680 - 7686. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. A. Hunstad, S. S. Justice, C. S. Hung, S. R. Lauer, and S. J. Hultgren Suppression of Bladder Epithelial Cytokine Responses by Uropathogenic Escherichia coli Infect. Immun., July 1, 2005; 73(7): 3999 - 4006. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Hennecke, J. Nolte, R. Volkmer-Engert, J. Schneider-Mergener, and S. Behrens The Periplasmic Chaperone SurA Exploits Two Features Characteristic of Integral Outer Membrane Proteins for Selective Substrate Recognition J. Biol. Chem., June 24, 2005; 280(25): 23540 - 23548. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Molecular and Cellular Proteomics |
| Journal of Lipid Research | ASBMB Today |
P2). Sequences of phage-attached peptides: 
, WTYIPPV; 
, WEYIPNV; 
, WSYIPVV; 
, FTYIPNS; 
, PLPISPR; 
, GSQSHHI.
carbons of the first and last residues of a seven-residue peptide in a