Electrostatic Repulsion between Unique Arginine Residues Is

Subscriber access provided by UNIV OF PITTSBURGH

Article

Electrostatic Repulsion between Unique Arginine Residues is Essential for the Efficient In Vitro Assembly of the Transmembrane Domain of a Trimeric Autotransporter Eriko Aoki, Daisuke Sato, Kazuo Fujiwara, and Masamichi Ikeguchi Biochemistry, Just Accepted Manuscript • DOI: 10.1021/acs.biochem.6b01130 • Publication Date (Web): 30 Mar 2017 Downloaded from http://pubs.acs.org on April 4, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Biochemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Electrostatic Repulsion between Unique Arginine Residues is Essential for the Efficient In Vitro Assembly of the Transmembrane Domain of a Trimeric Autotransporter

Eriko Aoki, Daisuke Sato, Kazuo Fujiwara and Masamichi Ikeguchi*

†

Department of Bioinformatics, Soka University, 1-236 Tangi-machi, Hachioji, Tokyo

192-8577, Japan * Corresponding Author: E-mail : [email protected] Phone : +81-426-91-9444

1 ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abbreviations and Textual Footnotes CD, circular dichroism; C8E4, n-octyltetraoxyethylene; Hia, Haemophilus influenzae adhesin; HMWA, high-molecular-weight aggregate; OMP, outer membrane protein; PAGE, polyacrylamide gel electrophoresis; PBS, phosphate-buffered saline; PDB, protein data bank; SDS, sodium dodecyl sulfate; TD, translocator domain; WT, wild type.


Page 2 of 31

Page 3 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

ABSTRACT: Haemophilus influenzae adhesin (Hia) belongs to the trimeric autotransporter family and mediates the adherence of these bacteria to the epithelial cells of host organisms. Hia contains a passenger and a transmembrane domain. The transmembrane domain forms a 12-stranded β-barrel in which four strands are provided from each subunit. The β-barrel has a pore that is traversed by three α-helices. This domain has a unique arginine cluster, in which the side chains of the three arginine residues located at position 1077 (Arg1077) protrude into the pore of the β-barrel. This arrangement seems to be unfavorable for assembly, because of repulsion between the positive charges. In this study, we investigated the in vitro assembly of the Hia transmembrane minimum domain (mHiaTD) and found that the dissociated mHiaTD reassembled in detergent solution. To investigate the role of Arg1077 on trimer assembly, we generated mutant proteins in which Arg1077 was replaced with methionine or lysine. The reassembly kinetics of the mutants was compared with that of the wild-type protein. The methionine mutant showed misassembly, whereas the lysine mutant showed reversible assembly, similar to that observed for the wild-type protein. These results show that electrostatic repulsion between the positive charges of Arg1077 is important to prevent the formation of misassembled oligomers by mHiaTD in vitro.


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Autotransporters are a large family of integral membrane proteins located in the outer membrane that are produced by Gram-negative bacteria.1 The autotransporter family includes classical (monomeric) and trimeric autotransporter subfamilies. All autotransporters are composed of a signal peptide for transport across the inner membrane, a passenger domain containing a virulence factor, and a transmembrane β-barrel domain that is required for passenger domain secretion. After proteins are synthesized in the cytoplasm, the signal sequences target the nascent chains to the Sec machinery located at the inner membrane and are cleaved after translocation to the periplasm. The translocation of classical autotransporters into, and their assembly at, the outer membrane have been studied extensively.2, 3 A self-transport model was originally proposed, in which the β-barrel domain was first inserted into the outer membrane, followed by the formation of a hairpin inside the pore of the β-barrel by the C-terminus of the passenger domain and the N-terminal segment of the passenger domain was subsequently threaded through the pore.4 Recently, it was accepted that the transmembrane domain is inserted into the outer membrane by the β-barrel assembly machinery (Bam) complex.2, 3 In trimeric autotransporters, the transmembrane domain is sufficient for trimerization and secretion.5 The translocation of the trimeric autotransporter adhesin A of Yersinia enterocolitica (YadA) require the BamA, which is a central component of the Bam complex.6 The C-terminal motif of YadA is recognized by BamA. The current model of the transport of this protein proposes that the C-terminal motif is recognized by the Bam complex before the N terminus is released from the Sec complex, and that the transmembrane domain is inserted into the outer membrane with the help of the Bam machinery.7 The translocation of the passenger domain, which is kept unfolded by periplasmic chaperones, may occur in a C- to 4 ACS Paragon Plus Environment

Page 4 of 31

Page 5 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

N-terminal direction.7 The manner in which the transmembrane domain is assembled into the outer membrane and the passenger domain is exported remains uncertain. Haemophilus influenzae adhesin (Hia) belongs to the trimeric autotransporter subfamily and mediates the adherence of these bacteria to the epithelial cells of host organisms. The passenger domain contains repeats of structurally similar domains, and three-dimensional structures have been reported for some of them.8, 9 The crystal structure of the Hia transmembrane domain (HiaTD) has been solved.10 It is also known that the minimal unit of HiaTD (mHiaTD) corresponds to residues 1022–1098.11 mHiaTD forms a 12-stranded β-barrel in which four strands are provided from each subunit (Figure 1A). The β-barrel has a pore that is traversed by three α-helices, one each provided by the three subunits.10 The domain has a unique arrangement of the arginine 1077 stemming from each subunit. Three Arg1077 side chains protrude into the pore of the β-barrel and are close to each other (Figure 1B). The arrangement seems to be unfavorable for trimer assembly because of repulsion between the positive charges of these residues. Although Meng et al.10 have suggested that the positive charges of Arg1077 neutralize the negative pole of the helix dipole (Figure 1C), the role of Arg1077 has not been examined experimentally. In the present study, we addressed the role of Arg1077 in the in vitro assembly of mHiaTD. Although β-barrel membrane proteins are thought to be translocated with the support of the Bam machinery in vivo, it is also known that many of them are able to fold spontaneously in detergent micelles or unilamellar vesicles.12 Therefore, the role of specific residues in the assembly and stability of β-barrel membrane proteins may be investigated using in vitro assembly studies.13


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

MATERIALS AND METHODS Materials. Synthetic DNAs encoding HiaTD (residues 992–1098) and its shorter fragment mHiaTD (residues 1022–1098) were obtained from Eurofin Genomics K.K. (Japan). Vectors pASK-IBA12 and pET-3c were purchased from IBA and Novagen, respectively. A Quikchange site-directed mutagenesis kit was obtained from Agilent 6 ACS Paragon Plus Environment

Page 6 of 31

Page 7 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Technologies. The rabbit anti-Strep-tag antibody was purchased from GeneTex (USA). Peroxidase-conjugated goat anti-rabbit IgG (H+L) was purchased from Jackson ImmunoResearch (USA). EzWestLumiOne and P plus membranes were purchased from ATTO (Japan). Elugent and n-octyltetraoxyethylene (C8E4) were purchased from Calbiochem and Bachem AG, respectively. Nimble Juice was purchased from GeneDireX (USA). Expression and Purification. The Strep-tagII sequence followed by a thrombin recognition sequence and the mHiaTD DNA sequence was inserted between the NdeI and BamHI sites of pET-3c to generate pET-mHiaTD. Substitutions of arginine 1077 to methionine or lysine were performed using the Quikchange site-directed mutagenesis kit to generate pET-mHiaTD_R1077M and pET-mHiaTD_R1077K, respectively.

Escherichia coli (E. coli) BL21(DE3) cells harboring each vector were grown at 37 °C, and protein expression was induced by the addition of 0.4 mM isopropyl

β-D-1-thiogalactopyranoside. After 4 h, cells were harvested by centrifugation, resuspended in phosphate-buffered saline (PBS), and lysed by sonication. The proteins were solubilized with Elugent and purified using a Strep-Tactin column (IBA), as described previously.10 The concentration of the purified protein was determined spectrophotometrically using an extinction coefficient of 13,940 M–1cm–1, which was calculated from the amino acid composition.14 Dissociation and Reassembly of mHiaTD. The solutions of the purified mHiaTD and mutants were diluted 5-fold with distilled water. The proteins were then precipitated by adding 4 volumes of acetone, and supernatants were removed after centrifugation. The pellets were washed with 80% acetone, dried, and treated with 70% formic acid for 30 min, to dissociate mHiaTD trimers. After evaporating the formic acid using a 7 ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

centrifugal concentrator, the sample was kept dried until use. The integrity of the protein was confirmed by mass spectrometry (Figure S1). The reassembly experiment was performed as follows: first, the dissociated mHiaTD was dissolved in 8 M urea containing a trace of HCl. An aliquot of this solution was withdrawn at various time points and used for the reassembly reaction. Reassemblies were initiated by diluting with reassembly buffer (20 mM Tris-HCl (pH 8.0), 0.6% (v/v) C8E4, and 100 mM or 500 mM NaCl), followed by incubation at 25 °C. A series of reassembly solutions from the same dissociated mHiaTD solution was applied to SDS–PAGE using the dissociated mHiaTD as a control. The monomer, the trimer, and other forms were separated via SDS–PAGE, stained with Nimble Juice, and detected using LAS-3000 (Fujifilm). Each band was quantified using ImageJ15 and expressed as a proportion to the total intensity of the bands observed in each lane. The kinetics data were analyzed using KaleidaGraph (Synergy Software). CD measurements. Far-UV CD measurements were performed on a Chirascan spectropolarimeter (Applied Photophysics). Unfolded proteins in 8 M urea were diluted 15-fold with 20 mM Tris-HCl (pH 8.0), 100 mM NaCl, and 0.6% (v/v) C8E4. After incubation for 24 h at 25 °C, spectra were measured at 25 °C using cuvettes with a 1 mm path length. Fluorescence measurements. Fluorescence spectrum measurements were performed on an F-2500 spectrometer (Hitachi). Unfolded proteins in 8 M urea were diluted 15-fold with 20 mM Tris-HCl (pH 8.0), 100 mM NaCl, and 0.6% (v/v) C8E4. The excitation wavelength was 280 nm. Fluorescence spectra were measured at 25 °C using cuvettes with a 3 mm path length. The bandwidth was 5 and 10 nm for excitation and emission, respectively. Anisotropy (r) was obtained using the following equation: 8 ACS Paragon Plus Environment

Page 8 of 31

Page 9 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

=

∥

∥

,

(1)

where ∥ and are fluorescence intensities at 340 nm, at which excitation and emission polarizers are in parallel and orthogonal arrangements, respectively. Western blot analyses of whole-cell lysates. The pASK-IBA12 plasmid was modified using a Quikchange site-directed mutagenesis kit so that the SacII site was replaced with an EcoRI site. Then, synthetic DNA encoding HiaTD was inserted between the EcoRI and BamHI sites, to give the expression vector pASK-HiaTD. Mutation was introduced as described above. Because pASK-IBA12 has a nucleotide sequence that encodes the OmpA signal sequence followed by a Strep-tag and a thrombin recognition sequence, the expressed protein is expected to have an N-terminal extension that includes a secretion signal, the Strep-tag, and the thrombin recognition sequence. E. coli BL21(DE3) cells harboring pASK-HiaTD or its variants were grown at 37 °C until the optical density at 600 nm (OD600) reached 0.6–0.8; subsequently, protein expression was induced using 0.2 mg/L anhydrotetracycline. After 2 h of culture, cells were harvested by centrifugation for 10 min at 5000 × g and resuspended in PBS. The cells were disrupted by sonication, and cell lysates were analyzed by SDS–PAGE. Proteins were transferred to P plus membranes. The membranes were then blocked with 5% (w/v) skimmed milk in Tris-buffered saline containing Tween-20 (TBS-T; 20 mM Tris-HCl (pH 7.4), 150 mM NaCl, 0.1% (w/v) Tween-20). Proteins were probed with the anti-Strep-tag antibody (1:100000 in 5% skimmed milk/TBS-T) and visualized using horseradish peroxidase (HRP)-conjugated secondary antibodies (1:10000 in 5% skimmed milk/TBS-T) and EzWestLumiOne. Signals were detected using LAS-3000.


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

RESULTS Reassembly of mHiaTD. It is known that mHiaTD forms a stable trimer when expressed in E. coli and solubilized from the membrane with detergents.10 The trimer is stable under SDS–PAGE conditions; therefore, its assembly can be easily monitored using this method. The trimeric mHiaTD was stable against 8 M urea (not shown). Conversely, the mHiaTD trimer is known to dissociate to the monomeric form after treatment with 95% formic acid.11 In this study, the purified mHiaTD was dissociated by 70% formic acid treatment after the detergent was removed by acetone precipitation. The dissociated mHiaTD dissolved in 8 M urea was detected as a band that corresponded to the molecular weight of the monomer (~10 kDa) on SDS–PAGE (Figure 2A, lane 1). The far-UV CD spectrum showed that the dissociated mHiaTD in 8 M urea existed as a randomly coiled polypeptide (Figure 2C). Reassembly was achieved by diluting the urea with the solution containing the detergent. The reassembled mHiaTD was detected as a band located at the same position as that of the purified mHiaTD trimer (Figure 2A, lane 8), and the CD spectrum of the reassembled mHiaTD agreed well with that of the purified mHiaTD trimer (Figure 2C). Therefore, the disassembled mHiaTD seems to be able to reassemble into the native trimer. To confirm the stability of the reassembled mHiaTD, we measured the CD spectrum of the reassembled mHiaTD in the presence of 2% SDS (Figure 2C). The spectrum of mHiaTD was not affected by the addition of SDS, which indicates that the reassembled mHiaTD showed resistance to SDS, as did the mHiaTD purified from E. coli.


Page 10 of 31

Page 11 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

The time course of trimer reassembly was assessed using SDS–PAGE (Figure 2A and 2B). The amount of monomer (~10 kDa band) was decreased as the reaction 11 ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 31

proceeded, and had almost disappeared at 24 h. In contrast, the trimer (~35 kDa band) increased as the reaction proceeded (Figure 2A and 2B). Although a negligible amount of aggregates with a higher molecular weight (including a molecular weight of ~70 kDa) was observed, the assembly reaction was cooperative and no intermediate was observed. Therefore, we assumed a simple model, in which three unfolded monomers (M) simultaneously assembled into the folded trimer:

3M → T,

(2)

where k is the rate constant. In this kinetic model, the fraction of the trimer was expressed as a function of time, t:

Fraction (%) = 1 −

( × 100,

! [#]% & '

(3)

where [M]0 is the initial concentration of the monomer and Y is the yield of the trimer, which is 1 for the fully reversible assembly. This equation fitted the data well, as shown in Figure 2B, and the rate constant estimated is shown in Table 1. The assembly rate should be accelerated at a higher protein concentration, because the assembly reaction is multimolecular. To confirm this, we examined the reassembly kinetics at a higher protein concentration (113 µg/mL). The results are shown in Figure 3. In contrast to the results shown in Figure 2, in which the protein concentration was 23 µg/mL, the accumulation of high-molecular-weight aggregates (HMWAs) (~70 kDa and others) observed in Figure 3A was remarkable, and their abundance reached 27% (Figure 3B). Although the trimer yield that was attained finally was decreased, the rate of trimer formation was indeed accelerated (Figure 3B vs Figure 2B). The rate constant estimated by curve fitting was approximately one third of the value obtained at 23 µg/mL (Table 1). If our model (Equation 2) is valid, the value obtained at 113 µg/mL should coincide 12 ACS Paragon Plus Environment

Page 13 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

with the value obtained at 23 µg/mL. The discrepancy observed between the two values may have been caused by the formation of HMWAs at the higher protein concentration. It is not clear at present whether the aggregation reaction proceeded in parallel with, or after the correct trimer formation. If it occurred in parallel, the monomeric molecule that is necessary for correct trimer formation may have been reduced, which results in a lower estimation of the rate constant. Conversely, if it occurred after trimer formation, the trimer concentration may have been reduced, thus also resulting in a lower estimation of the rate constant. Taking these issues into consideration, we think that equation 3 is valid if the yield is close to 100%.


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 31

Table 1. Rate constants and yield of reassembly reactions a Condition 1 Condition 2 Condition 3

WT

[Protein] (µg/mL)

23

113

23

[NaCl] (mM)

100

100

500

3.77 ± 0.39

1.15 ± 0.17

24.8 ± 2.8

100

67 ± 2

95 ± 2

7.29 ± 2.33

−

21.7 ± 4.4

81 ± 7

−

86 ± 3

Rate constant ×10–10 (1/M2h) Yield (%)

R1077K

Rate constant ×10–10 (1/M2h) Yield (%)

a

The error indicated in this table is the error estimated by the curve fitting procedure.

Note that actual errors are much larger because they involve the errors in the protein concentration estimations.

If the electrostatic repulsion between the three Arg1077 positive charges is a determinant of the assembly rate, the addition of salt should increase the assembly rate. Furthermore, the repulsion between the net charges of the three subunits may also have a significant effect on the assembly rate. Sato et al.16 addressed the role of the net charges of the subunits of ferritin on its assembly kinetics and showed that the subunit net charge affected the assembly rate, although the effect was remarkable only at an ionic strength lower than 0.1. They also suggested that the local electrostatic interactions are important, even at high ionic strength. To assess the effect of salt on the


Page 15 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

assembly kinetics of mHiaTD, we observed its assembly kinetics in the presence of 500 mM NaCl (Figure 4). The assembly rate observed at 500 mM NaCl was approximately six times faster than that detected at 100 mM NaCl (Table 1). Therefore, the electrostatic repulsion between local charges is an important factor in the determination of the assembly rate.

Assembly reaction monitored by fluorescence spectroscopy. To investigate the folding and assembly of mHiaTD, we monitored the change in the fluorescence spectrum and anisotropy during the assembly reaction. Figure 5A shows the change in the fluorescence spectrum during the assembly reaction at 23 µg/mL of mHiaTD. The 15 ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 31

maximum fluorescence intensity was observed at 342 nm at the earliest time point (15 min), and it shifted to the shorter wavelength as the assembly reaction proceeded. This spectrum change is typical of folding-induced sequestration of tryptophan side chains from the solvent.17 mHiaTD has two tryptophan residues, one of which is located near the C terminus (W1098); the other is included in the Strep-tag. mHiaTD also has two tyrosine residues, which may contribute to the fluorescence spectrum. Concomitantly with the spectrum shift, the fluorescence intensity increased (Figure 5A). The time course of the intensity at 340 nm is shown in Figure 5B. We also measured fluorescence anisotropy changes (Figure 5C). The anisotropy increased as the reaction proceeded, indicating that the Brownian motion of the tryptophan side chains was slowed down. Both fluorescence intensity and anisotropy changes were fitted to the following equation:

+(,) = ∆+ 1 −

! [#]% & .

( + +0 ,

(4)

where A0 is the initial fluorescence intensity or anisotropy and ∆A is the amplitude. The estimated rate constants (Table 2) are consistent with the rate constant of assembly within the experimental error (Table 1), suggesting that the folding and assembly of mHiaTD occur in a concerted manner.


Page 17 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Table 2. Rate constants of folding and assembly measured by fluorescence intensity and anisotropy (Condition 1)a Rate constant ×10–10 (1/M2h) WT

R1077K

a

Intensity

7.6 ± 3.2

Anisotropy

8.4 ± 2.4

Intensity

1.2 ± 0.6

Anisotropy

3.4 ± 2.1

The error indicated in this table is the error estimated by the curve fitting procedure.

Note that actual errors are much larger because they involve the errors in the protein concentration estimations.

Effect of mutation. To investigate the role of Arg1077, we replaced this residue with methionine (R1077M) or lysine (R1077K). The two amino acids were selected to determine the role of positive charges located on the side chain. The CD spectra of the two mutant proteins agreed well with the spectrum of the WT protein (Figure S2), indicating that none of these mutations perturb the secondary structure. The reassembly reaction was initiated as described for the WT protein and was monitored by SDS–PAGE, the fluorescence spectrum, and fluorescence anisotropy. The populations of the trimeric and monomeric forms were quantified by SDS–PAGE, and the time courses of the trimer and monomer were assessed as described above. The effect of the methionine substitution was drastic. The reaction exhibited at least two kinetic phases.


Page 18 of 31

Page 19 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

Under the standard conditions (23 µg/mL protein, 100 mM NaCl), the trimer of R1077M was formed up to 30% at the earliest observed time (15 min) (Figure 6B). This indicates that the trimer forms much faster than does the wild-type mHiaTD (Figure 6B vs Figure 2B). This was also shown by the fluorescence spectrum (Figure 7A). At the earliest time point (15 min), the maximum wavelength was 337 nm, which is identical to the value of obtained for the WT protein at 24 h, and it did not change further. This result indicates that the tryptophan side chains were already sequestered from the solvent within 15 min of the reaction. Thus, the fast phase was too fast to allow the determination of the rate constant. Therefore, the values of R1077M are not presented in Tables 1 and 2. Conversely, the amount of trimer formed at 24 h was much lower than that observed for the WT. This deficiency of R1077M regarding trimer formation was not attributable to the formation of HMWA, as observed for the WT protein at 113

µg/mL. In contrast to WT assembly at 113 µg/mL (Figure 3), negligible bands were observed in the higher-molecular-weight region of the gel (Figure 6A). A significant amount of monomeric R1077M remained after 24 h of reassembly (Figure 6B). Therefore, the trapped species is a monomer with a conformation that is not competent for assembly, or a misassembled oligomer that dissociates during SDS–PAGE. To characterize the kinetic trap, we reassembled R1077M at a higher protein concentration (Figure S3). Although the trimer was formed faster, the trimer fraction at 24 h was lower than that detected at 23 µg/mL (Figure 6B); that is, a larger amount of the trapped species was formed at a higher protein concentration, suggesting that the trapped species is formed via multimolecular reactions.


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

To characterize the trapped species further, we measured the CD spectrum of R1077M after reassembly for 24 h (Figure 6C). The CD spectrum of the reassembled 20 ACS Paragon Plus Environment

Page 20 of 31

Page 21 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

R1077M was more intense than that of the native R1077M, although the fractional population of the trimer was much lower than that of the WT protein. This result suggests that the trapped species forms an intermolecular β-sheet structure. The CD intensity of this spectrum was decreased by adding SDS (Figure 6C), although the spectrum of the reassembled WT protein did not change after this treatment (Figure 2C). These results suggest that a portion of the R1077M molecules forms a misassembled oligomer that is dissociated in the presence of SDS and is observed as monomers in SDS–PAGE. This incorrect structure is considered to include an intermolecular β-sheet, as assessed based on its CD spectrum.


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

R1077M also exhibited unexpected changes in fluorescence intensity (Figure 7C) and anisotropy (Figure 7D). Both fluorescence intensity and anisotropy decreased at first, and then increased. Because the fluorescence life time of tryptophan is relatively short (probably 1~10 ns), the anisotropy of tryptophan fluorescence is expected to be insensitive to the molecular mass change beyond 30kDa, of which correlation time is expected to be >20 ns.18 Although the anisotropy value of R1077M was not large compared with that of the WT, we could not exclude the possibility of formation of large aggregates. In contrast to the drastic effect of the R1077M mutation, the R1077K mutant exhibited an assembly kinetics that was similar to that of the WT protein (Figure 8 and S4). The observation that the fraction of the trimer formed was slightly lower than that of the WT protein was attributable to a slight increase in the number of HMWA. The rate constants estimated by curve fitting were similar to those of the WT species at two different NaCl concentrations (Table 1). The changes in fluorescence spectrum and anisotropy were also similar to those detected in the WT protein (Figure 7). The CD spectrum of the reassembled R1077K was the same as that observed before dissociation and did not change after the addition of 2% SDS (Figure 8C). These results indicate that R1077K can reassemble in a manner that is similar to that of the WT protein. In conclusion, the close proximity of positive charges at residue 1077 is important in preventing the formation of misassembled oligomers of mHiaTD.


Page 22 of 31

Page 23 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Expression level of mutant proteins. To address the effect of the R1077M mutation in vivo, we expressed HiaTD and its mutants, and whole-cell lysates were assayed by SDS–PAGE followed by immunoblotting. The results of these experiments are shown in Figure 9. Mature trimeric and monomeric HiaTD, in addition to precursor monomeric HiaTD, were observed in all cell lysates, and the level of expression and the trimer/monomer ratio were similar among all variants. Thus, the assembly deficiency of R1077M was not detected in vivo.

DISCUSSION Physical origin of misfolding and/or misassembly. In this study, we showed that mHiaTD reassembled in vitro and that the unique arginine residue located at position 1077 was important for its correct assembly. Unusual clustering of arginine side chains


Page 24 of 31

Page 25 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

was found previously in proteins,19 especially at the interface of oligomeric proteins.20 Among the arginine clusters found in the Protein Data Bank in which the Cζ–Cζ distance is less than 5 Å, the positive charges of more than 90% of clusters were neutralized by the negative charges of nearby carboxyl groups or bound ions.20 In the case of mHiaTD, however, there is no carboxyl group in the proximity of the Arg1077 guanidino group. As suggested by Meng et al.,10 the negative pole of the helix dipole may neutralize the positive charge of Arg1077. Water molecules are also suggested as a stabilizing factor and are involved in hydrogen bonds with 79% of the clustered arginine residues.20 Water molecules that interact with Arg1077 were found in the crystal structure of the Hia transmembrane domain (Figure 10). A crystal of the Hia transmembrane domain 1022–1098 (PDB ID: 2GR8) contained two trimers in the asymmetric unit.10 One of them (chains A, C, and D) had four water molecules that can interact with the Nη atoms of Arg1077 (HOH 87, 101, 146, 191). These water molecules seem to form a hydrogen bond network with the Nη atoms of Arg1077 and the backbone carbonyl oxygen atoms of Ala1034, Ser1035, and Leu1037 (Figure 10 and Table S1). In the other trimer (chains B, E, and F), however, only one water molecule (HOH 62) could be identified at the position corresponding to HOH 146 when chain B was fitted to chain A. In another crystal of the Hia transmembrane domain 992–1098 (PDB ID: 2GR7), there were also two trimers in the asymmetric unit. One of them (chains A, B, and C) had only one water molecule (HOH 161) at the position corresponding to HOH 87 of 2GR8 chains A, C, and D when the chain A of 2GR7 was fitted to the chain A of 2GR8. Water molecules seemed to be located at some consensus positions for the formation of hydrogen bond networks with Arg1077. A water molecule, HOH 232, was found at an exceptional position in the crystal structure of the other trimer of 2GR7 25 ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(chains D, E, and F), which corresponds to the geometric center of HOH 87, 101, 146 of 2GR8. In conclusion, the hydrogen bond network between water molecules, the protein backbone (Ala1034, Ser1035, and Leu1037), and the side chains of Arg1077 likely stabilizes the trimer, although it may change dynamically. A molecular dynamics simulation of HiaTD showed that the hydrogen bond formed between Gln1036 and Arg1077 or that formed between Gln1039 and Arg1077 is present in 40% or 45% of the trajectory frames, respectively, although they are not present in the X-ray structure.21

Was the incorrect assembly of the R1077M mutant caused by the loss of the stabilizing interaction described above? The mutational study showed that lysine could act in the place of arginine, which suggests the importance of the positive charge in this process. Although a crystal of R1077K has not been obtained, the three-dimensional structure of another member of the trimeric autotransporter family, YadA, has been solved by solid-state NMR22. YadA contains a lysine residue (Lys434) at the position corresponding to Arg1077 of mHiaTD. The side chains of Lys434 exhibited different 26 ACS Paragon Plus Environment

Page 26 of 31

Page 27 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

conformations in 20 NMR models. Some of them protrude into the pore of the β-barrel. The shortest Nζ–Nζ distance is 8.8 Å, which is longer than the Cζ–Cζ distance of Arg1077 (6.2 Å). This distance seems to be too long for the formation of a hydrogen bond network such as that of mHiaTD, although this is not a conclusive result, because water molecules cannot be identified in the NMR structure. However, it is unlikely that the loss of the hydrogen bond network is at the origin of misfolding and/or misassembly. We found 66 membrane anchor sequences for trimeric autotransporters in the daTAA server (https://toolkit.tuebingen.mpg.de/dataa).23 Among these 66 sequences, 27 and 8 possess lysine and arginine, respectively, at the site corresponding to Arg1077 of Hia. Positively charged amino acid residues are not conserved but are preferred for this site. Furthermore, we cannot exclude the possibility that positively charged residues are in close proximity at other positions, because the three-dimensional structure of the transmembrane domain of trimeric autotransporters has been determined only for Hia8, 10

and YadA.22 The fact that the assembly rate of R1077M was much faster than that of the WT

protein even at 500 mM NaCl suggests that electrostatic repulsion between the three Arg1077 residues retards the assembly reaction, because it has been shown that electrostatic repulsion between subunit net charges is suppressed by ions giving an ionic strength of 0.1.16 Electrostatic interactions are well known as being important factors for molecular assembly.24, 25 Furthermore, Zlotnick and coworkers have shown that weak protein–protein interactions appear to be a common theme in viral assembly, and that strong interactions tend to result in a kinetic trap.26, 27 The electrostatic repulsion that occurs between Arg1077 residues is essential because the interactions between the 27 ACS Paragon Plus Environment

Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

subunits of mHiaTD may be too strong for its correct assembly if repulsion between the three Arg1077 residues is absent. Implication for in vivo assembly. For the monomeric outer membrane protein (OMP), it has been suggested that BamA assists membrane insertion via the lateral opening of its own β-barrel, followed by the transient formation of a hybrid β-barrel with the guest OMP and release of the OMP.28 Does BamA work similarly for trimeric autotransporter? Does one BamA molecule form a hybrid β-barrel with three molecules of trimeric autotransporter? If this is the case, the electrostatic repulsion between the Arg1077 side chains should occur at the process released from the hybrid β-barrel in vivo. The elimination of this repulsion may accelerate the rate of the release from the hybrid β-barrel. Although it is not clear whether the assembly of R1077M is faster than that of WT in vivo, it is noteworthy that the 60 and 75 kDa bands were weaker in R1077M than in WT or R1077K (Figure 9). The 45, 60, and 75 kDa bands present in the whole-cell lysate may correspond to the barrels that consist of four, five, and six molecules, i.e., incorrectly assembled HiaTD. Similar HMWAs were also observed for WT in vitro, especially when the protein concentration was high (Figure 3). The fact that HMWAs did not accumulate significantly for R1077M, even at a high protein concentration (Figure 6), suggests that HMWAs are not stable for R1077M either in vivo or in vitro. Because we were not able to obtain any evidence indicating the presence of misfolded and/or misassembled forms of R1077M in vivo, the absence of the electrostatic repulsion between the Arg1077 side chains may not induce any misfolding and/or misassembly during the Bam-assisted membrane insertion (in vivo), although we cannot exclude the possibility that the misfolding and/or misassembly of R1077M occur also in vivo but incorrectly assembled proteins may be rapidly degraded 28 ACS Paragon Plus Environment

Page 28 of 31

Page 29 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

and undetectable.

ASSOCIATED CONTENT Supporting Information. Mass spectra of protein samples (Figures S1), CD spectra of purified proteins (Figure S2), SDS–PAGE image of R1077M reassembly at the protein concentration of113 µg/mL (Figure S3), reassembly kinetics of R1077K at 500 mM NaCl (Figure S4), and distances of possible hydrogen bonding atomic pairs relevant to Arg1077 (Table S1) are supplied as Supporting Information.

Notes The authors declare no competing financial interest.

REFERENCES [1] Grijpstra, J., Arenas, J., Rutten, L., and Tommassen, J. (2013) Autotransporter secretion: varying on a theme, Res. Microbiol. 164, 562-582. [2] Bernstein, H. D. (2015) Looks can be deceiving: recent insights into the mechanism of protein secretion by the autotransporter pathway, Mol. Microbiol. 97, 205-215. [3] Fan, E., Chauhan, N., Udatha, D. B., Leo, J. C., and Linke, D. (2016) Type V Secretion Systems in Bacteria, Microbiol. Spectr. 4, 1-21. [4] Pohlner, J., Halter, R., Beyreuther, K., and Meyer, T. F. (1987) Gene structure and extracellular secretion of Neisseria gonorrhoeae IgA protease, Nature 325, 458-462. [5] Mikula, K. M., Leo, J. C., Lyskowski, A., Kedracka-Krok, S., Pirog, A., and Goldman, A. (2012) The translocation domain in trimeric autotransporter adhesins is necessary and sufficient for trimerization and autotransportation, J. Bacteriol. 194, 827-838. [6] Lehr, U., Schutz, M., Oberhettinger, P., Ruiz-Perez, F., Donald, J. W., Palmer, T., Linke, D., Henderson, I. R., and Autenrieth, I. B. (2010) C-terminal amino acid residues of the trimeric autotransporter adhesin YadA of Yersinia enterocolitica are decisive for its recognition and assembly by BamA, Mol. Microbiol. 78, 932-946. [7] Leo, J. C., Grin, I., and Linke, D. (2012) Type V secretion: mechanism(s) of autotransport through the bacterial outer membrane, Philos. Trans. R. Soc. Lond. B Biol. Sci. 367, 1088-1101. [8] Meng, G., St Geme, J. W., 3rd, and Waksman, G. (2008) Repetitive architecture of the Haemophilus influenzae Hia trimeric autotransporter, J. Mol. Biol. 384, 824-836. [9] Laarmann, S., Cutter, D., Juehne, T., Barenkamp, S. J., and St Geme, J. W. (2002) The Haemophilus influenzae Hia autotransporter harbours two adhesive pockets that reside in the passenger domain and recognize the same host cell receptor, Mol.


Biochemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Microbiol. 46, 731-743. [10] Meng, G., Surana, N. K., St Geme, J. W., 3rd, and Waksman, G. (2006) Structure of the outer membrane translocator domain of the Haemophilus influenzae Hia trimeric autotransporter, EMBO J. 25, 2297-2304. [11] Surana, N. K., Cutter, D., Barenkamp, S. J., and St Geme, J. W., 3rd. (2004) The Haemophilus influenzae Hia autotransporter contains an unusually short trimeric translocator domain, J. Biol. Chem. 279, 14679-14685. [12] McMorran, L. M., Brockwell, D. J., and Radford, S. E. (2014) Mechanistic studies of the biogenesis and folding of outer membrane proteins in vitro and in vivo: what have we learned to date?, Arch. Biochem. Biophys. 564, 265-280. [13] Tamm, L. K., Hong, H., and Liang, B. (2004) Folding and assembly of beta-barrel membrane proteins, Biochim. Biophys. Acta 1666, 250-263. [14] Gill, S. C., and von Hippel, P. H. (1989) Calculation of protein extinction coefficients from amino acid sequence data, Anal. Biochem. 182, 319-326. [15] Schneider, C. A., Rasband, W. S., and Eliceiri, K. W. (2012) NIH Image to ImageJ: 25 years of image analysis, Nat. Methods 9, 671-675. [16] Sato, D., Takebe, S., Kurobe, A., Ohtomo, H., Fujiwara, K., and Ikeguchi, M. (2016) Electrostatic Repulsion during Ferritin Assembly and Its Screening by Ions, Biochemistry 55, 482-488. [17] Duy, C., and Fitter, J. (2006) How aggregation and conformational scrambling of unfolded states govern fluorescence emission spectra, Biophys. J. 90, 3704-3711. [18] Ortega, A., and Garcia de la Torre, J. (2005) Efficient, accurate calculation of rotational diffusion and NMR relaxation of globular proteins from atomic-level structures and approximate hydrodynamic calculations, J. Am. Chem. Soc. 127, 12764-12765. [19] Magalhaes, A., Maigret, B., Hoflack, J., Gomes, J. N., and Scheraga, H. A. (1994) Contribution of unusual arginine-arginine short-range interactions to stabilization and recognition in proteins, J. Protein Chem. 13, 195-215. [20] Neves, M. A., Yeager, M., and Abagyan, R. (2012) Unusual arginine formations in protein function and assembly: rings, strings, and stacks, J. Phys. Chem. B 116, 7006-7013. [21] Holdbrook, D. A., Piggot, T. J., Sansom, M. S., and Khalid, S. (2013) Stability and membrane interactions of an autotransport protein: MD simulations of the Hia translocator domain in a complex membrane environment, Biochim. Biophys. Acta 1828, 715-723. [22] Shahid, S. A., Bardiaux, B., Franks, W. T., Krabben, L., Habeck, M., van Rossum, B. J., and Linke, D. (2012) Membrane-protein structure determination by solid-state NMR spectroscopy of microcrystals, Nat. Methods 9, 1212-1217. [23] Szczesny, P., and Lupas, A. (2008) Domain annotation of trimeric autotransporter adhesins--daTAA, Bioinformatics 24, 1251-1256. [24] Parent, K. N., Doyle, S. M., Anderson, E., and Teschke, C. M. (2005) Electrostatic interactions govern both nucleation and elongation during phage P22 procapsid assembly, Virology 340, 33-45. [25] Klug, A. (1983) From Macromolecules to Biological Assemblies (Nobel Lecture), Angew. Chem. Int. Ed. Engl. 22, 565-582. [26] Zlotnick, A., Johnson, J. M., Wingfield, P. W., Stahl, S. J., and Endres, D. (1999) A theoretical model successfully identifies features of hepatitis B virus capsid assembly, Biochemistry 38, 14644-14652. [27] Zlotnick, A. (2003) Are weak protein-protein interactions the general rule in capsid assembly?, Virology 315, 269-274. [28] Noinaj, N., Kuszak, A. J., Balusek, C., Gumbart, J. C., and Buchanan, S. K. (2014) Lateral opening and exit pore formation are required for BamA function, Structure 22, 1055-1062.


Page 30 of 31

Page 31 of 31

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Biochemistry

For Table of Contents Use Only 89x35mm (300 x 300 DPI)

ACS Paragon Plus Environment

Electrostatic Repulsion between Unique Arginine Residues Is

Recommend Documents