biomolecules

Article Role of Lipopolysaccharide in Protecting OmpT from Autoproteolysis during In Vitro Refolding

1, 1, 2 2 Gaurav Sinsinbar † , Sushanth Gudlur †, Kevin J Metcalf , Milan Mrksich , Madhavan Nallani 1 and Bo Liedberg 1,*

1 Center for Biomimetic Sensor Science, School of Materials Science and Engineering, Nanyang Technological University, Singapore 637553, Singapore; [email protected] (G.S.); [email protected] (S.G.); [email protected] (M.N.) 2 Departments of Chemistry and Biomedical Engineering, Northwestern University, Evanston, IL 60208, USA; [email protected] (K.J.M.); [email protected] (M.M.) * Correspondence: [email protected] G.S. and S.G. contributed equally to this work. †  Received: 8 May 2020; Accepted: 14 June 2020; Published: 18 June 2020 

Abstract: Outer membrane (OmpT) is a 33.5 kDa aspartyl protease that cleaves at dibasic sites and is thought to function as a defense mechanism for E. coli against cationic antimicrobial peptides secreted by the host immune system. Despite carrying three dibasic sites in its own sequence, there is no report of OmpT autoproteolysis in vivo. However, recombinant OmpT expressed in vitro as inclusion bodies has been reported to undergo autoproteolysis during the refolding step, thus resulting in an inactive protease. In this study, we monitor and compare levels of in vitro autoproteolysis of folded and unfolded OmpT and examine the role of lipopolysaccharide (LPS) in autoproteolysis. SDS-PAGE data indicate that it is only the unfolded OmpT that undergoes autoproteolysis while the folded OmpT remains protected and resistant to autoproteolysis. This selective susceptibility to autoproteolysis is intriguing. Previous studies suggest that LPS, a co-factor necessary for OmpT activity, may play a protective role in preventing autoproteolysis. However, data presented here confirm that LPS plays no such protective role in the case of unfolded OmpT. Furthermore, OmpT mutants designed to prevent LPS from binding to its putative LPS-binding motif still exhibited excellent protease activity, suggesting that the putative LPS-binding motif is of less importance for OmpT’s activity than previously proposed.

Keywords: LPS; autoproteolysis; OmpT; heat modifiable ; omptin family; outer membrane protease; E. coli

1. Introduction Gram-negative bacteria have evolved several different strategies to evade or counter antimicrobial peptides (AMPs) secreted by the host immune system. One such strategy, adopted by the Enterobacteriacea family, involves recruiting omptin to cleave AMPs and making them ineffective. Omptins are a family of outer membrane proteases that share 40–80% amino acid sequence similarity and a highly conserved active site [1]. Examples include OmpT () Pla (), PgtE (), IcsP (), and CroP (Citrobacter rodentium)[1,2]. E. coli OmpT is the most notable omptin protease and is thought to play an essential role in the bacteria’s defense system as well as in their virulence [3–5]. It is overexpressed in uropathogenic E. coli (UPEC) associated with the majority of community-acquired urinary tract infections [6,7], making UPECs more resistant to AMPs than commensal E. coli strains [6]. Besides UPEC, OmpT expression is also upregulated in certain E. coli strains involved in foodborne infections [8,9], indicating an active role in the pathogenicity of E. coli strains. Among the two

Biomolecules 2020, 10, 922; doi:10.3390/biom10060922 www.mdpi.com/journal/biomolecules Biomolecules 2020, 10, 922x FOR PEER REVIEW 2 of 15 16 role in the pathogenicity of E. coli strains. Among the two genetically related gastrointestinal pathogens—enterohemorrhagicgenetically related gastrointestinal E. pathogens—enterohemorrhagic coli (EHEC) and enteropathogenicE. coli (EHEC) E. coli and(EPEC)—that enteropathogenic cause E.severe coli (EPEC)—that diarrheal diseases cause severein humans diarrheal [10], diseases OmpT in overexpression humans [10], OmpT in EHEC overexpression allows the in EHECactive degradationallows the active of AMPs degradation to promote of AMPs bacterial to promote survival bacterial [5,11–13]. survival [5,11–13]. OmpT isis a a 33.5 33.5 kDa kDa membrane membrane protein protein that adopts that adopts a β-barrel a β structure-barrel structure consisting consisting of 10 antiparallel of 10 antiparallelβ-strands spanning β-strands the spanning outer membrane the outer with membrane an active site with facing an theactive extracellular site facing environment the extracellular [1,3,14] (Schemeenvironment1A). Previous[3,14] (Scheme work 1A). reports Previous that OmpT work reports requires that lipopolysaccharide OmpT requires lipopolysaccharide (LPS) as a co-factor (LPS) for asits a proteolytic co-factor for function its proteolytic [15]. However, function the [15]. exact However, mechanism the exact of how mechanism LPS activates of how OmpT LPS isactivates poorly OmpTunderstood. is poorly It is understood. proposed that It is the proposed binding that of LPS the tobinding the putative of LPS binding to the putative motif in binding OmpT causesmotif in a OmpTslight conformational causes a slight changeconformational in the protease’s change activein the protease’s site which isactive associated site which with is an associated active OmpT with [15 an]. Anactive active OmpT OmpT [15]. preferentially An active OmpT cleaves preferentially host-generated cleaves antimicrobial host-generated peptides antimicrobial at dibasic sites peptides [15–17 at]. dibasicDespite sites carrying [15–17]. three Despit dibasice carrying sites in itsthree own dibasic sequence, sites there in its is own no report sequence, of OmpT there autoproteolysisis no report of OmpTin vivo .autoproteolysis The biogenesis ofin OmpTvivo. Thein vivo biogenesisis a complex of OmpT process in vivo starting is a withcomplex the cytosolic process productionstarting with of the proteincytosolic that production is translocated of the across protein the that inner is membrane translocated through across translocases the inner membrane in a partially through folded translocasesstate [18]. Upon in a enteringpartially thefolded periplasmic state [18]. space, Upon the entering protein the is fully periplasmic folded and space, inserted the protein into the is outerfully foldedmembrane and withinserted the into help the of periplasmicouter membrane chaperones with the and help the ofβ -barrelperiplasmic assembly chaperones machinery and (BAM)the β- barrelcomplex assembly [18]. Once machinery in the outer (BAM) membrane, complex [18]. OmpT Once becomes in the outer active, membrane, most likely OmpT when becomes it binds active, to the mostLPS that likely is foundwhen it only binds in to the the outer LPS leafletthat is offound Gram-negative only in the outer bacteria. leaflet An of inactive Gram-negative OmpT during bacteria. its Antransport inactive from OmpT the cytosolduring its to thetransport outer membranefrom the cytosol possibly to the serves outer as membrane a safety mechanism possibly serves evolved as toa preventsafety mechanism any accidental evolved proteolysis to prevent of cytosolic any accidental . proteolysis of cytosolic proteins.

Scheme 1.1. StructureStructure of of outer outer membrane membrane protease protease (OmpT) (OmpT) and itsand autoproteolyzed its autoproteolyzed fragments. fragments. (A) Ribbon (A) Ribbonrepresentation representation of OmpT of with OmpT the with five predicted the five pr lipopolysaccharideedicted lipopolysaccharide (LPS) binding (LPS) residues binding (blue) residues and (blue)the three and dibasic the three sites dibasic represented sites re inpresented licorice (green). in licorice Black (green). arrow Black points arrow at the points location at the of thelocation dibasic of thesite ondibasic loop site 4 (red) on thatloop is 4 prone (red) tothat autoproteolysis is prone to autoproteolysis yielding the two yielding autoproteolytic the two fragmentsautoproteolytic (B,C). fragmentsNote: Crystal (B,C structure). Note: Crystal of wild structure type OmpT of wild is not type available. OmpT Theis not Protein available. Data The Bank Protein (PDB) Data structure Bank

(1I78)(PDB) ofstructuremtOmpT (1I78)DB, an of OmpT mtOmpT mutantDB, an (mt) OmpT that hasmutant one of(mt) its dibasicthat has (DB) one sites of its mutated dibasic (K217-R218 (DB) sites mutated to(K217-R218 K216-G217-R218), mutated isto used K216-G217-R218), here to illustrate is theused different here to keyillustrate points, the e.g., different the locations key points, of the e.g.,autoproteolytic the locations and of predictedthe autoproteoly LPS bindingtic and residues. predicted LPS binding residues.

A better understandingunderstanding of of OmpT OmpT autoproteolysis autoproteolysis can can help help in designingin designing antibiotics antibiotics that that induce induce and andaccelerate accelerate such detrimentalsuch detrimental events eventsin vivo in, thereby vivo, thereby making making the bacteria the susceptiblebacteria susceptible to host generated to host AMPs.generated Similarly, AMPs. understandingSimilarly, understanding LPS’s role inLPS’s OmpT role activity in OmpT can activity help devise can help strategies devise to strategies interfere orto interfereblock LPS or binding block to LPS OmpT, binding thus making to OmpT, the protease thus ma inactiveking andthe incapableprotease ofinactive degrading and host-generated incapable of AMPsdegrading that host-generated attack the bacteria. AMPs Most that of attack what wethe knowbacteria. today Most about of what OmpT we is know based today on in about vitro studiesOmpT is based on in vitro studies of recombinant OmpT. In vitro expression of recombinant OmpT can be Biomolecules 2020, 10, 922 3 of 16 of recombinant OmpT. In vitro expression of recombinant OmpT can be achieved either in the E. coli outer membrane or overexpressed as cytoplasmic protein aggregates called inclusion bodies (IBs) that are easily isolated and subsequently solubilized, refolded, and purified. The latter method is favored since it offers higher yields and ease of purification; however, OmpT undergoes autoproteolysis during in vitro refolding and purification resulting in an inactive protease [19] (Scheme1B,C). Among its three dibasic sites (R37-K38,K217-R218, and K259-K260), only K217-R218 is prone to autoproteolysis [19]. Located on a flexible extracellular L4 loop on top of the beta-barrel structure of OmpT, this exposed site makes it more susceptible to autoproteolysis than the other two sites that are located on the body of the barrel (Scheme1) [ 1]. Previously, autoproteolysis of recombinant OmpT during in vitro refolding was overcome via two different strategies. In the first strategy adopted by Kramer et al. (2000) [19], the protease sequence was mutated in the vicinity of the autoproteolysis site, K217-R218 was mutated to K216-G217-R218 [19]. However, this mutant was reported to have a lower activity (~30% lower) in comparison with the wild type protease [19]. In addition, we recently reported that this minor mutation in OmpT resulted in an altered substrate preference [20]. A second strategy, adopted by Wu et al. (2012) [21] and described in detail in Wu’s PhD thesis, involved the addition of LPS to refolded OmpT immediately after cation exchange purification to avoid autoproteolysis [21,22]. In this case, the author suspected a protective role of LPS against OmpT autoproteolysis, but did not characterize it any further. In this study, we aim at gaining a deeper understanding of the in vitro conditions that promote autoproteolysis and the role of LPS in autoproteolysis. Using SDS-PAGE data, we demonstrate that a folded OmpT does not undergo autoproteolysis, but instead, it is the unfolded protein that undergoes proteolytic cleavage. Additionally, we show that LPS neither protects unfolded OmpT from autoproteolysis nor is its binding to the putative LPS-binding motif in OmpT essential for activity, as was previously thought, suggesting that there may be more than one allosteric LPS binding site on OmpT that is critical for its activity.

2. Materials and Methods

2.1. Materials The plasmid ompT_pET28a was a gift from Prof. Neil Kelleher (Northwestern University, Evanston IL, USA) while plasmid mut_ompT_pET28a was synthesized and sequenced by Bio Basic Inc (Singapore). E. coli Rosetta™ (DE3) cells with T7 RNA polymerase under the lac promoter were used for expressing wild type and mutant OmpTs as IBs and were purchased from Merck Millipore (Singapore). LB Broth was purchased from BD Difco (Franklin Lakes, NJ, USA). Kanamycin was purchased from Gibco by Life technologies (Now, ThermoFisher Scientific, Waltham, MA, USA). Isopropyl thio-β-D-galactopyranoside (IPTG) was purchased from Gold Biotechnology (St. Louis, MO, USA). N-Dodecyl-N, N-dimethyl-3-ammonio-1-propanesulfonate (DodMe2NPrSO3) and lipopolysaccharides (LPS) from E. coli 0111:B4 purified by phenol extraction, Corning® Black, flat bottom 384-well plates were purchased from Sigma Aldrich (Singapore). Urea was purchased from Affymetrix USB (Santa Clara, CA, USA). PD-10 Desalting Columns containing Sephadex G-25 resin were purchased from GE Healthcare Life Sciences (Singapore). HiPPR™ Detergent Removal Spin Column was purchased from ThermoFisher Scientific (Singapore). Synthetic peptide, Abz-ARRA-Tyr (NO2)-NH2, Ac-GARKVG-NH2, and Ac-GARIIG-NH2 were purchased as lyophilized powder and certified >95% purity from Biomatik (Cambridge, ON, Canada).

2.2. Recombinant OmpT Expression The ompT gene corresponds to a polypeptide of 317 amino acids whose first 20 amino acids act as a signal peptide responsible for transferring OmpT into the periplasm [23]. The plasmid ompT_pET28a had a construct where the mature OmpT sequence corresponding to 297 amino acids in addition to methionine and alanine at the N-terminal was cloned without the signal peptide sequence between NcoI Biomolecules 2020, 10, 922 4 of 16 and EcoRI into the pET28a vector to produce ompT_pET28a. The N-terminal methionine, incorporated for initiation of the translation, is eventually excised upon expression [24]. The mutant ompT gene (G216K and K217G), which was synthesized and subcloned into the Xbal and Xhol site into the pET28a vector to produce mut_ompT_pET28a, was purchased from Bio Basic Inc. The wild type and mutant OmpTs (298 amino acids) were expressed as IBs in E. coli Rosetta™ (DE3) cells using the method previously described by Kramer, R.A. et al. (2000) [19]. Briefly, Rosetta™ (DE3) cells with ompT_pET28a and mut_ompT_pET28a plasmids were grown separately overnight in 5 mL LB Broth with 50 µg/mL kanamycin at 37 ◦C in a shaker set to 220 RPM. This primary culture was used to inoculate 200 mL LB Broth secondary cultures that were later induced with IPTG (1 mM final concentration) when the OD600 reached ~0.8–1. The cells were harvested after 4 h and washed with 50 mM Tris-HCl, pH 7.5 before suspending them in 5 mL lysis buffer (50 mM Tris-HCl, 40 mM EDTA, pH 8.0). These cells were kept on ice for 20 min before adding a further 5 mL of pre-chilled lysis buffer, followed by a second incubation on ice for 20 min. After treatment with the lysis buffer, the cells were sonicated for 7 min (2 s on, 4 s off, and 50% amplitude), upon which, the solution turned viscous. This viscous solution was centrifuged at 4500 g for 30 min to separate the IBs as pellets which were then washed twice with × 30 mL wash buffer (10 mM Tris-HCl, 1 mM EDTA, pH 8.0). During the final wash, the solutions were aliquoted into separate tubes and centrifuged to obtain IBs as pellets.

2.3. Refolding of OmpT IBs The IBs as pellets were solubilized in a solubilizing buffer (20 mM Tris-HCl, 8 M urea, 50 mM glycine, pH 8.3) and kept on ice for 1 h before the addition of pre-chilled refolding buffer (20 mM Tris-HCl, 31 mM DodMe2NPrSO3, pH 7.5) such that the final concentration of protein and urea was between 0.8 and 1 mg/mL and 0.8 M, respectively. The samples were kept on ice for 3–4 h. For samples requiring 4.5 mg/mL LPS during the refolding step, LPS was added to the refolding solution before mixing with the IBs in the solubilizing buffer. The refolded OmpT samples were centrifuged at 10,000 g × for 10 min to pellet the protein aggregates. To remove the remaining urea from the samples, PD-10 columns were used. The columns were equilibrated with buffer (20 mM Tris, 10 mM DodMe2NPrSO3, pH 7.5). Refolded protein was loaded on the PD-10 column and fractions were collected. The fractions were checked for protein concentration and pooled together. Samples were aliquoted and stored at 20 C for further use. − ◦ 2.4. OmpT Protease Activity in the Presence of an Inhibitor Based on a method previously described by Kramer et al. (2000) [19], OmpT IBs refolded in refolding buffer with 31 mM DodMe2NPrSO3 (RC-31), refolding buffer with 10 mM DodMe2NPrSO3 (RC-10), and refolding buffer with no DodMe2NPrSO3 (RC-0) were assayed for their protease activity in the presence of increasing concentrations of an inhibitor. All fluorescence assays were performed in 384-well plates at a final reaction volume of 100 µL. Synthetic peptide substrate, Abz-ARRA-Tyr (NO2)-NH2, is reported to be a good substrate for OmpT and has been used for determining OmpT activity [19]. The peptide substrate was prepared as stock solutions of 0.5 mM in Milli-Q water. For the assay, the peptide stock solution was diluted with a mixture of Milli-Q water and 10 assay × buffer (200 mM Bis-tris, 20 mM DodMe2NPrSO3, 50 mM EDTA, pH 6.7) such that the final peptide concentration was 15 µM in 1 assay buffer (20 mM Bis-tris, 2 mM DodMe NPrSO , 5 mM EDTA, × 2 3 pH 6.7). The concentrations of all three OmpT samples were kept the same and confirmed using a NanoDropTM 2000c spectrophotometer (ThermoFischer Scientific, Singapore) and OmpT’s molar 1 1 absorption coefficient of 74,960 M− cm− [19]. A stock solution of 2 mg/mL of LPS from E. coli was prepared in water. All three samples of OmpT were diluted 10 times and incubated with 20 µg/mL LPS for ~15 min. An amount of 10 µL of OmpT-LPS mixture was mixed with 10 µL of ZnCl2 at different concentration (1000, 500, 250, 62.5, 15.6, 7.8, and 3.9 10 µM). An amount of 10 µL of this OmpT-LPS-ZnCl2 mixture was added to 90 µL of 15 µM peptide substrate in assay buffer. The enzymatic reaction was performed at 37 ◦C. The increase in fluorescence intensity upon OmpT addition was Biomolecules 2020, 10, 922 5 of 16 collected over time using a TECAN infinite® M200 PRO (Männedorf, Switzerland) fluorescence spectrophotometer fitted with a plate reader. Fluorescence intensity signals were collected every 10 s for 60 min with the excitation wavelength set to 325 nm and emission at 420 nm, and the data plotted in OriginPro 9.1 (OriginLab Corporation, Northampton, MA, USA). The initial velocity of the 1 enzymatic reaction (µMs− ) was calculated using the data points in the linear range of the kinetic data. The reaction velocity was plotted for different inhibitor (ZnCl2) concentrations for all the three OmpT samples.

2.5. OmpT Protease Activity Assay Based on a method previously described by Kramer et al. (2000) [19], wild type and mutant OmpT samples were assayed for their protease activity. All fluorescence assays were performed in 384-well plates at a final reaction volume of 100 µL. The synthetic peptide substrate Abz-ARRA-Tyr (NO2)-NH2 was prepared as stock solutions of 0.5 mM in Milli-Q water and diluted with a mixture of Milli-Q water and 10 assay buffer (200 mM Bis-tris, 20 mM DodMe NPrSO , 50 mM EDTA, pH 6.7) to a final × 2 3 peptide concentration of 50 µM in 1 assay buffer (20 mM Bis-tris, 2 mM DodMe NPrSO , 5 mM EDTA, × 2 3 pH 6.7) for the assay. The concentration of the wild type and mutant OmpT samples were confirmed by a BCA assay. A stock solution of 2 mg/mL of LPS from E. coli was prepared in water. All four OmpT samples were diluted 80 times in a buffer devoid of LPS. The diluted OmpT samples were incubated with 20 µg/mL LPS for ~15 min, to ensure OmpT activity. An amount of 10 µL of the OmpT-LPS mixture was added to 90 µL of 50 µM peptide substrate in assay buffer. The enzymatic reaction was performed at 37 ◦C. The increase in fluorescence intensity upon OmpT addition was collected over time using a TECAN infinite® M200 PRO (Männedorf, Switzerland) fluorescence spectrophotometer fitted with a plate reader. Fluorescence intensity signals were collected every 10 s for 60 min with the excitation wavelength set to 325 nm and emission at 420 nm, and the data plotted in OriginPro 9.1. Specific activity was calculated for each OmpT sample based on the fluorescence data.

2.6. Liquid Chromatography-Mass Spectrometry (LC-MS) and MALDI-TOF OmpT samples were prepared for LC-MS analysis of soluble (9 kDa peptide) and detergent-stabilized (33.5 and 24.5 kDa proteins) species. The frozen OmpT sample was thawed on ice. The 9 kDa peptide was observed by dilution of the thawed sample in water to a final concentration of ~10 µM for injection to the LC-MS system. The two higher molecular weight species were prepared for LC-MS analysis using methanol/chloroform/water precipitation to remove the detergent [25]. The sample was resuspended in 5% acetonitrile and 0.1% formic acid in water to a concentration of 30 µM and diluted with water to a final concentration of ~1 µM. LC-MS analysis was performed on an Agilent 1200 series HPLC connected to an Agilent 6210A time-of-flight (TOF) mass spectrometer. A 10 µL injection of each sample was captured on a C18 trap column (Waters) and eluted using a gradient from 5% to 95% acetonitrile and 0.1% formic acid in water with a flowrate of 0.25 mL/min. Data were analyzed with Agilent MassHunter Qualitative Analysis B.04.00. For the MALDI-TOF analysis, two synthetic peptides, Ac-GARKVG-NH2 and Ac-GARIIG-NH2, with and without a dibasic site, respectively, were employed to confirm the substrate specificity of the RC-31 (LPS) samples. A stock solution of 0.5 mM for each peptide was prepared and was stored at 20 C. Each peptide was diluted using a mixture of Milli-Q water and 10 assay buffer − ◦ × (200 mM Bis-tris, 20 mM DodMe2NPrSO3, 50 mM EDTA, pH 6.7) to a final peptide concentration of 10 µM in 1 assay buffer (20 mM Bis-tris, 2 mM DodMe NPrSO , 50 mM EDTA, pH 6.7), to which × 2 3 OmpT was added at 1 µg/mL. The reaction mixture was incubated at 37 ◦C for 1 h with shaking at 500 rpm. HiPPR™ Detergent Removal Spin Column was used, as directed in the product guide, to remove the detergent from the reaction mixture. The reaction samples for the two peptides and the matrix, α-Cyano-4-hydroxycinnamic acid, were mixed (1:1), and 1 µL of the mixture was spotted on a MALDI-TOF array plate. The matrix and reaction sample was allowed to dry before loading into the instrument. The samples were analyzed using a Shimadzu Biotech Axima Performance Mass Biomolecules 2020, 10, 922 6 of 16 spectrometer (Nakagyo-ku, Kyoto, Japan) that contained a nitrogen UV laser. The intensity vs. m/z spectrum was collected at a laser power of 100 mV in the reflection hi-resolution mode, and other settings were adjusted for the best signal-to-noise ratio.

3. Results and Discussion

3.1. Refolding and Heat-Modifiable Migration of OmpT Biomolecules 2020, 10, x FOR PEER REVIEW 6 of 15 A mutant OmpT (mtOmpTDB), with its dibasic site K217-R218 mutated to K216-G217-R218, was expressed in vitro as3.1. IBs. Refolding High and yields Heat-Modifiable and ease Migration of separation of OmpTfrom unwanted soluble proteins is an important advantage ofA expressing mutant OmpT the protein(mtOmpT asDB IBs), with in E. its coli dibasic[19]. However,site K217-R218 IBs mutated are partially to K216 folded-G217-R218 or, was misfolded insolubleexpressed protein aggregates in vitro as IBs. that High need yields to be and solubilized ease of separation and completely from unwanted unfolded soluble before proteins properly is an refolding them backimportant into their advantage native of state expressing [26]. Therefore, the proteinmt asOmpT IBs in E.DB coliIBs [19]. were However, solubilized IBs are andpartially unfolded folded in 8 M urea beforeor misfolded they were insoluble refolded. protein The aggregates unfolding that was need confirmed to be solubilized by analyzing and completely the sample unfolded on before an SDS gel properly refolding them back into their native state [26]. Therefore, mtOmpTDB IBs were solubilized where the protein migrated to a distinct 33.5 kDa band (Figure1, Lane 1). mtOmpT IBs solubilized and unfolded in 8 M urea before they were refolded. The unfolding was confirmed by analyzingDB the and unfoldedsample inon 8an M SDS urea gel werewhere subsequently the protein migrated refolded to a indistinct the presence 33.5 kDa band of 31 (Figure mM DodMe1, Lane 21).NPrSO 3 detergent.mtOmpT The criteriaDB IBs solubilized for selecting and unfolded DodMe in2NPrSO 8 M urea3 aswere the subsequently refolding detergent refolded in was the presence based on of previous31 studies bymM Kramer DodMe et2NPrSO al. (2000)3 detergent. [19] who The screenedcriteria for Triton selecting X-100, DodMe Tween2NPrSO 20,3 octylpolyoxyethyleneas the refolding detergent (OPOE), and n-hexa-decylphosphorylcholinewas based on previous studies as by alternatives Kramer et al to. (2000) DodMe [19]2NPrSO who screened3. They Triton did not X-100, observe Tween any 20, protease activity inoctylpolyoxyethylene these alternate detergents (OPOE), despite and these n-hexa-dec detergentsylphosphorylcholine appearing to help as properly alternatives refold theto OmpT. DodMe2NPrSO3. They did not observe any protease activity in these alternate detergents despite Unlike most proteins, OmpT and other Gram-negative outer membrane beta-barrel proteins these detergents appearing to help properly refold the OmpT. are resistantUnlike to denaturation most proteins, by OmpT SDS and alone other dueGram-neg to theative extensive outer membrane hydrogen beta-barrel bonding proteins between are the β-strandsresistant [27–30 to]. denaturation As a result, by OmpTs SDS alone treated due to with the extensive SDS alone hydrogen continue bonding to retainbetween a the compact β-strands globular shape allowing[27–30]. As itto a migrateresult, OmpTs further treated along with the SDS SDS-PAGE alone continue gel and to retain be mistaken a compact for globular a lower shape molecular weight proteinallowing of it to 27 migrate kDa (Figure further 1along, Lane the 2,SDS-PAGE lower MW gel and band). be mistaken This anomalous for a lower molecular migration weight is termed “gel-shifting”,protein as of the 27 migrationkDa (Figure does 1, Lane not 2, correlate lower MW with band). molecular This anomalous weight. Formigration a detailed is termed description “gel- and shifting”, as the migration does not correlate with molecular weight. For a detailed description and discussiondiscussion of the migration of the migration pattern pattern of folded of folded and and unfolded unfolded OmpT OmpT seesee Figure S1. S1. Therefore, Therefore, in order in order for OmpT tofor migrate OmpT atto itsmigrate expected at its molecularexpected molecular weight ofweight 33.5 kDa,of 33.5 a kDa, combination a combination of SDS of SDS and and heat heat treatment is requiredtreatment to unfold is required a folded to unfold OmpT a folded (Figure OmpT1, Lane (Figure 4). 1, Additionally,Lane 4). Additionally, it is important it is important to to note note that in previouslythat reported in previously work reported [19] on work the [19] heat-modifiable on the heat-modifiable migration migration of OmpT, of OmpT, unfolding unfolding was was achieved achieved within only 5 min of heat denaturation at 100 °C. However, we still observed a faint lower within only 5 min of heat denaturation at 100 ◦C. However, we still observed a faint lower molecular molecular weight band at 27 kDa (Figure 1, lane 3) even after 10 min of heating at 100 °C that weight band at 27 kDa (Figure1, lane 3) even after 10 min of heating at 100 C that disappeared only disappeared only upon prolonged heating at 100 °C for 45 min (Figure 1, lane ◦4). Nevertheless, this upon prolongedheat-modifiable heating migration at 100 serves◦C for as 45a reliable min (Figure indicator1, of lane OmpT 4). refolding Nevertheless, efficiency this since heat-modifiable it allows migrationfor serves the determination as a reliable of indicator the relative of amounts OmpT of refolding folded and effi unfoldedciency since protein it allowsin a given for OmpT the determination sample of the relative[19,31,32]. amounts of folded and unfolded protein in a given OmpT sample [19,31,32].

Figure 1. Gel-shifting of mtOmpTDB on SDS-PAGE: Lane 1—mtOmpTDB inclusion bodies (IBs), Lane Figure 1. Gel-shifting of mtOmpTDB on SDS-PAGE: Lane 1—mtOmpTDB inclusion bodies (IBs), M—protein molecular weight marker. mtOmpTDB after refolding in DodMe2NPrSO3 detergent buffer Lane M—protein molecular weight marker. OmpT after refolding in DodMe NPrSO detergent when heated at 100 °C for 0 min (Lane 2), 10mt min (LaneDB 3), and 45 min (Lane 4). UF—unfolded2 protein,3 buffer whenF—folded heated protein. at 100 ◦C for 0 min (Lane 2), 10 min (Lane 3), and 45 min (Lane 4). UF—unfolded protein, F—folded protein. 3.2. Autoproteolysis of OmpT In order to study OmpT autoproteolysis, wild type OmpT was expressed in vitro as IBs that were later solubilized and unfolded in 8 M urea. The unfolded proteins were subsequently refolded in Tris-HCl buffer containing 31, 10, or 0 mM DodMe2NPrSO3 detergent. For the sake of convenience, we refer to these three refolding conditions as RC-31, RC-10, and RC-0 denoting the high, intermediate, and zero detergent concentration, respectively. When OmpT IBs were refolded in the Biomolecules 2020, 10, 922 7 of 16

3.2. Autoproteolysis of OmpT In order to study OmpT autoproteolysis, wild type OmpT was expressed in vitro as IBs that were later solubilized and unfolded in 8 M urea. The unfolded proteins were subsequently refolded in Tris-HCl buffer containing 31, 10, or 0 mM DodMe2NPrSO3 detergent. For the sake of convenience, we refer to these three refolding conditions as RC-31, RC-10, and RC-0 denoting the high, intermediate, and zero detergent concentration, respectively. When OmpT IBs were refolded in the absence of the detergent (RC-0), the protein migrated as a single band at 33.5 kDa in an SDS-PAGE gel regardless of heat treatment, indicating thatBiomolecules all of the2020, OmpT 10, x FOR wasPEER unfoldedREVIEW (Figure2, non-heated and 45 min-heated samples,7 of 15 Lane RC-0). As previouslyabsence reported, of the detergent and confirmed (RC-0), the here, protein the migrated presence as a of single the detergentband at 33.5 duringkDa in an the SDS-PAGE refolding process aids in the propergel regardless refolding of heat of OmpTtreatment, from indicating the IBs that [33 all– 35of the]. At OmpT this was point, unfolded the apparent (Figure 2, refoldingnon-heatedefficiencies for RC-31 andand RC-10 45 min-heated equal ~69% samples, and Lane 32%, RC-0). respectively, As previously based reported, on the and relative confirmed intensities here, the of presence the two bands at of the detergent during the refolding process aids in the proper refolding of OmpT from the IBs [33– 33.5 and 27 kDa35]. At for this the point, non-heated the apparent samples refolding (Figure efficiencies2, non-heated for RC-31 samples, and RC-10 Lane equal RC-31 ~69% andand RC-10),32%, a clear indication thatrespectively, excess detergent based on the is relative essential intensities to achieve of the two higher bands refolding at 33.5 and efficiencies. 27 kDa for the However,non-heated this is not the completesamples story as (Figure will be2, discussednon-heated below.samples, The Lane specific RC-31 activityand RC-10), for thea clear RC-31, indication RC-10, that and excess RC-0 samples detergent is essential to achieve higher refolding efficiencies. However, this is not the complete story 1 1 measured usingas willa be previously discussed below. reported The specific FRET activity assay for [ 19the] RC-31, was calculated RC-10, and RC-0 to be samples13.01 measured nmol min − mg− , 1 1 1 5.64 nmol minusing− mga previously− , and reported close to FRET 0 pmol assay s− [19](Figure was calculated S3—Day to be 0, 13.01 data nmol not shownmin−1 mg for−1, 5.64 the nmol RC-0 sample), and these valuesmin−1 mg correlated−1, and close well to 0 pmol with s the−1 (Figure refolding S3—Day efficiency 0, data not (Figure shown for2, the 0 h RC-0 for sample), the non-heated and these samples). values correlated well with the refolding efficiency (Figure 2, 0 h for the non-heated samples).

Figure 2. Solubilized OmpT IBs reconstituted in detergent micelles were monitored for Figure 2. Solubilized OmpT IBs reconstituted in detergent micelles were monitored for autoproteolysis autoproteolysis over a time period of 72 h: Solubilized OmpT IBs were refolded in the presence of 31,

over a time10, period or 0 mM of DodMe 72 h:2NPrSO Solubilized3 detergent OmpT and denoted IBs were as RC-31, refolded RC-10, and in theRC-0, presence respectively. of Samples 31, 10, or 0 mM DodMe2NPrSOincubated3 detergent at 4 °C were and collected denoted after as 0,RC-31, 24, and 72 RC-10, h of refolding and RC-0, and analyzed respectively. using SDS-PAGE: Samples Left incubated at panel—samples heated for 45 min prior to gel loading, and right panel—samples loaded without any 4 ◦C were collected after 0, 24, and 72 h of refolding and analyzed using SDS-PAGE: Left panel—samples heating (non-heated samples) on SDS-PAGE. The thickness of the red and blue dashed lines reflects heated for 45increasing min prior and to decreasing gel loading, protein and band right intensities. panel—samples Note: We observed loaded at withoutleast one additional any heating higher (non-heated samples) onmolecular SDS-PAGE. weight The band thickness (slightly ofabove the the red 33.5 and kD bluea) in the dashed non-heated lines RC-31 reflects and increasing RC-10 samples and of decreasing protein band24 intensities.and 72 h which Note: were Wealso observed by atKramer least et one al. (2000) additional [19]. These higher are most molecular likely OmpT weight band (slightly abovecomplexed the 33.5 with kDa) LPS. in the non-heated RC-31 and RC-10 samples of 24 and 72 h which were also observed by Kramer et al. (2000) [19]. These are most likely OmpT complexed with LPS. BiomoleculesBiomolecules 20202020,, 1010,, x 922 FOR PEER REVIEW 88 of of 15 16

Despite the prolonged heat treatment for 45 min at 100 °C intended to unfold all OmpTs, both the RC-31Despite and the RC-10 prolonged samples heat exhibited treatment a faint for 45 band min at 10027 kDa◦Cintended (Figure 2, to bands unfold underlined all OmpTs, with both red the dashedRC-31 and line RC-10 for 0 h). samples This 27 exhibited kDa band a faint is not band due atto 27 folded kDa(Figure OmpT,2 ,but bands in fact underlined attributable with to red the dashed larger ofline the for two 0 h). autoproteolytic This 27 kDa band fragments is not due resulting to folded from OmpT, OmpT but in autoproteolysis fact attributable (Scheme to the larger 1B). ofIt theis unfortunate,two autoproteolytic but the fragmentslarger autopr resultingoteolytic from fragment OmpT autoproteolysisas well as the folded (Scheme intact1B). OmpT It is unfortunate, display an almostbut the identical larger autoproteolytic migration pattern fragment through as well the as gel the with folded an intactapparent OmpT molecular display weight an almost of ~27 identical kDa [19],migration thus making pattern data through interpretation the gel with difficult an apparent and sometimes molecular confusing. weight of The ~27 LC-MS kDa [19 analysis], thus making of the samplesdata interpretation analogous ditoffi RC-31cult and and sometimes RC-10 revealed confusing. peaks The for LC-MS the full-length analysis of theprotein samples at 33.5 analogous and two to peaksRC-31 at and 9.1 RC-10 and 24.4 revealed kDa for peaks the autoproteolyzed for the full-length fragments, protein at 33.5thus and confirming two peaks autoproteolysis at 9.1 and 24.4 kDaas well for the autoproteolyzed fragments, thus confirming autoproteolysis as well as the site of autoproteolysis as the site of autoproteolysis to be between K217 and R218 (Figure 3 A-B; Supplementary Figure S2 and Supplementaryto be between KTable217 and T1). R218 The(Figure RC-313A,B; andSupplementary RC-10 samples Figure exhibited S2 and Supplementarylow levels (~12%) Table S1of). autoproteolysisThe RC-31 and RC-10for the samples freshly refolded exhibited samples, low levels based (~12%) on the of autoproteolysisfaint bands observed for the at freshly 27 kDa refolded (Figure 2,samples, bands underlined based on the with faint red bands dashed observed line for at 270 h). kDa The (Figure band2, intensities bands underlined gradually with increased red dashed to 20% line andfor 0to h). 21–24% The band for intensitiesthe samples gradually analyzed increased after incubation to 20% and at 4 to °C 21–24% for 24 and for the72 h, samples respectively analyzed (Figure after 2,incubation bands underlined at 4 ◦C for with 24 andred 72dashed h, respectively line for 24 (Figure h, and2 72, bands h). During underlined the same with time red frame, dashed the line non- for heated24 h, and RC-31 72 h). and During RC-10 the samples same timeexhibited frame, a thegradual non-heated loss in RC-31intensity and from RC-10 31% samples to 6% and exhibited 68% to a 32%,gradual respectively, loss in intensity of the from33.5 kDa 31% band to 6% (Figure and 68% 2, tobands 32%, underlined respectively, with of the blue 33.5 dashed kDa band line (Figurefor 0, 24,2, andbands 72 underlinedh). Autoproteolysis with blue of dashedunfolded line OmpTs for 0, 24,is the and most 72 h). likely Autoproteolysis explanation for of unfoldedthis observation. OmpTs If is foldedthe most OmpTs likely explanationwere to undergo for this autoproteolysis, observation. If there folded would OmpTs be were a gradual to undergo decrease autoproteolysis, in protease activitythere would for the be RC-31 a gradual and decreaseRC-10 samples. in protease However, activity the for protease the RC-31 activity and RC-10 of the samples. above RC-31 However, and RC- the 10protease samples activity did not of change the above significantly RC-31 and over RC-10 time samples and retained did not ~80% change activity significantly even after over40 days time when and storedretained at ~80%4 °C, activitysuggesting even that after the 40 daysfolded when OmpT stored did atnot 4 ◦ undergoC, suggesting any extensive that the folded autoproteolysis OmpT did (Supplementarynot undergo any Figure extensive S3). autoproteolysis Taken together, (Supplementary the gradual increase Figure S3in). the Taken level together, of autoproteolytic the gradual fragment,increase in a theconcomitant level of autoproteolytic loss of the unfolded fragment, OmpT, a concomitant and a constant loss protease of the unfolded activity OmpT,of the folded and a OmpTconstant sample protease over activity time ofclearly the folded suggest OmpT that sample the overunfolded time clearlyportion suggest of OmpT that theis prone unfolded to autoproteolysis.portion of OmpT is prone to autoproteolysis. Alternatively,Alternatively, itit wouldwould bebe possible possible to to follow follow OmpT OmpT autoproteolysis autoproteolysis by by monitoring monitoring the the levels levels of theof thesmaller smaller (9.1 kDa)(9.1 kDa) autoproteolyzed autoproteolyzed fragment fragment using tricine-basedusing tricine-based gels. However, gels. However, this smaller this fragment smaller fragmentcarries a dibasiccarries sitea dibasic in its sequence site in its at Ksequence259-K260 whichat K259 is-K prone260 which to undergo is prone further to undergo autoproteolysis further autoproteolysisyielding two fragments yielding oftwo ~4.9 fragments and 4.2 of kDa. ~4.9 Following and 4.2 kDa. autoproteolysis Following autoproteolysis by monitoring by the monitoring changing thelevels changing of three levels low molecularof three low weight molecular bands weight with two bands of themwith belowtwo of 5them kDa below is relatively 5 kDa diisffi relativelycult and difficultis likely toand be is error likely prone to be when error it comesprone towhen gathering it comes quantitative to gathering data quantitative based on cumulative data based band on cumulativeintensities. Inband comparison, intensities. monitoring In comparison, the ~27 monitoring kDa band, the although ~27 kDa challenging, band, although is relatively challenging, less error is relativelyprone due less to itserror dependence prone due on to aits single dependence band intensity on a single and band more intensity importantly, and providesmore importantly, details of providesprotein refolding details of e ffiprotciency.ein refolding efficiency.

FigureFigure 3. LC-MSLC-MS characterization characterization of theof the OmpT OmpT samples: samples: (A) The (A larger) The autoproteolyzedlarger autoproteolyzed fragment fragment (24.4 kDa) (24.4andintact kDa) OmpT and intact (33.5 kDa)OmpT after (33.5 detergent kDa) removalafter detergent via methanol removal/chloroform via methanol/chloroform/water/water precipitation, and (B) precipitation,the smaller proteolyzed and (B) the fragment smaller (9.1proteolyzed kDa) from fragment the soluble (9.1 phase kDa) from (i.e., withoutthe soluble precipitation). phase (i.e., without precipitation). Biomolecules 2020, 10, 922 9 of 16 Biomolecules 2020, 10, x FOR PEER REVIEW 9 of 15

Interestingly, thethe RC-0RC-0 samplessamples displayeddisplayed littlelittle autoproteolysis despitedespite having the highest levels of unfolded OmpT among the three refolding conditions (Figure2 2,, Lanes Lanes RC-0). RC-0). ThisThis isis likelylikely duedue toto the lacklack of of su sufficientfficient amounts amounts of properlyof properly refolded refolded andactive and active OmpT OmpT necessary necessary to cleave to other cleave unfolded other OmpTs.unfolded This OmpTs. was confirmedThis was confirmed by conducting by conducti FRET-basedng FRET-based protease activity protease measurements activity measurements where RC-0 sampleswhere RC-0 exhibited samples poor exhibited activity whenpoor comparedactivity wh withen compared the RC-31 andwith RC-10 the RC-31 samples and (Figure RC-104 A,samples Blue). Furthermore,(Figure 4A, Blue). active Furthermore, site titration measurementsactive site titration using measurements ZnCl2 as the inhibitorusing ZnCl confirmed2 as the the inhibitor RC-31 samplesconfirmed to the contain RC-31 a samples higher level to contain of active a higher OmpT level as comparedof active OmpT with RC-10,as compared with the with lowest RC-10, values with obtainedthe lowest for values the RC-0 obtained samples for (Figure the RC-04B). Thesesamples activity (Figure observations 4B). These correlated activity observations well with the correlated refolding ewellfficiencies with the of refolding the three samplesefficiencies (Figure of the2, thr non-heatedee samples samples, (Figure 0 2, h). non-heated samples, 0 h).

Figure 4. FRETFRET cleavage cleavage and and inhibitor inhibitor assays assays for for OmpT OmpT in in three three refolding refolding buffers buffers containing containing different different concentrations of detergent, RC-31, RC-10, and RC-0. ( A) Change in fluorescencefluorescence emission intensity due to cleavage of FRET-peptide by OmpT. (B) ProteaseProtease activity measurements in the presence of increasing inhibitor (ZnCl ) concentration and plotted as specific activity vs. inhibitor concentration. increasing inhibitor (ZnCl2) concentration and plotted as specific activity vs. inhibitor concentration. Note: To avoidavoid the 0 valuevalue on thethe XX-axis-axis ofof thethe semi-logarithmicsemi-logarithmic plot in (B), the specificspecific activity corresponding to 0 µM ZnCl was plotted as 0.01 µM ZnCl . corresponding to 0 μM ZnCl22 was plotted as 0.01 μM ZnCl2. Although the initial refolding efficiency of ~69% and 32% for the RC-31 and RC-10 samples, respectively, Although the initial refolding efficiency of ~69% and 32% for the RC-31 and RC-10 samples, were based on the relative intensities of the 33.5 and 27 kDa bands for the non-heated samples in Figure2, respectively, were based on the relative intensities of the 33.5 and 27 kDa bands for the non-heated this did not take into account the contribution of the autoproteolytic fragment to the 27 kDa band. samples in Figure 2, this did not take into account the contribution of the autoproteolytic fragment to The refolding efficiency was therefore recalculated as follows: Refolding Eff (%) = [(% intensity of 27 kDa band the 27 kDa band. The refolding efficiency was therefore recalculated as follows: Refolding Eff (%) = for 0 h, non-heated samples) (% intensity of 27 kDa band for 0 h, heated samples)]. Basically, the contribution [(% intensity of 27 kDa band− for 0 h, non-heated samples) – (% intensity of 27 kDa band for 0 h, heated from the autoproteolyzed fragment is removed when calculating the refolding efficiency. For the RC-31 samples)]. Basically, the contribution from the autoproteolyzed fragment is removed when calculating and RC-10 samples, the refolding efficiency was now close to ~60% and 20%, respectively. However, for the the refolding efficiency. For the RC-31 and RC-10 samples, the refolding efficiency was now close to RC-31 samples, the refolding efficiency varied between 50–70% for different batches (data not shown) and ~60% and 20%, respectively. However, for the RC-31 samples, the refolding efficiency varied between was difficult to control using the protocol adopted from Kramer et al. (2000) [19]. Despite this difference 50–70% for different batches (data not shown) and was difficult to control using the protocol adopted in the protein refolding efficiency between the different batches, the gradual increase in the level of from Kramer et al. (2000) [19]. Despite this difference in the protein refolding efficiency between the autoproteolysis and a concomitant loss of unfolded OmpT clearly suggest that the unfolded OmpT different batches, the gradual increase in the level of autoproteolysis and a concomitant loss of underwent autoproteolysis (Supplementary Figure S2). unfolded OmpT clearly suggest that the unfolded OmpT underwent autoproteolysis (Supplementary 3.3.Figure Role S2). of LPS in Autoproteolysis of Unfolded OmpT

3.3. RoleLPS of has LPS previously in Autoproteolysis been shown of Unfolded to be OmpT essential for OmpT’s activity [15,19,36–39]. Binding of LPS to a putative LPS-binding motif in OmpT is thought to cause a slight conformational change LPS has previously been shown to be essential for OmpT’s activity [19,36–39]. Binding of LPS to in the protease’s active site which is associated with an active OmpT [15,38]. Although no LPS was a putative LPS-binding motif in OmpT is thought to cause a slight conformational change in the added during the refolding stage for RC-31, RC-10, and RC-0, autoproteolysis was still observed in protease’s active site which is associated with an active OmpT [15,38]. Although no LPS was added all three samples, revealing that active OmpT was present during the refolding process. This could during the refolding stage for RC-31, RC-10, and RC-0, autoproteolysis was still observed in all three be due to the unavoidable residual LPS from the membrane fraction during the IB isolation process. samples, revealing that active OmpT was present during the refolding process. This could be due to Attempts to quantify this residual LPS were unsuccessful, mostly due to interference from detergent the unavoidable residual LPS from the membrane fraction during the IB isolation process. Attempts molecules. While LPS appears to be critical for autoproteolysis, it is also important to consider that to quantify this residual LPS were unsuccessful, mostly due to interference from detergent molecules. not all of the OmpT is autoproteolyzed. Active OmpT, presumably with an LPS bound to it, seems to While LPS appears to be critical for autoproteolysis, it is also important to consider that not all of the OmpT is autoproteolyzed. Active OmpT, presumably with an LPS bound to it, seems to be resistant Biomolecules 2020, 10, 922 10 of 16 Biomolecules 2020, 10, x FOR PEER REVIEW 10 of 15 beto autoproteolysis, resistant to autoproteolysis, suggesting a suggesting protective arole protective for LPS.role Similarly, for LPS. Wu Similarly, et al. (2012) Wu [21] et al. suspected (2012) [21 a] suspectedprotective arole protective of LPS role against of LPS OmpT against autoproteo OmpT autoproteolysislysis and presented and presented evidence evidence to support to support this thishypothesis hypothesis by performing by performing experiments experiments where where LPS was LPS added was added to refolded to refolded OmpT OmpT immediately immediately after aftercation cation exchange exchange purificati purificationon [21,22]. [21, 22We]. Wetherefore therefore wondered wondered if LPS if LPS is isplaying playing a adual dual role, role, that of activating OmpT and of protectingprotecting itit fromfrom autoproteolysis.autoproteolysis. To exploreexplore thethe rolerole ofof LPSLPS inin OmpTOmpT autoproteolysis,autoproteolysis, OmpT IBs solubilized in 8 M urea were refolded in a bubufferffer containingcontaining 31 mMmM detergent (RC-31)(RC-31) along with excess LPS (4.5 mgmg/mL)/mL) or no LPS. For the the sake sake of of convenience, convenience, we we refer refer to tothese these two two refolding refolding conditions conditions as RC-31(LPS) as RC-31(LPS) and andRC- RC-3131 to differentiate to differentiate between between samples samples refolded refolded in inthe the presence presence of of excess excess or or no no externally externally added added LPS, LPS, respectively. TheThe RC-31(LPS) RC-31(LPS) and and RC-31 RC-31 samples samples collected collected at three at three different different intervals intervals over aover 72 h perioda 72 h andperiod run and on anrun SDS-PAGE on an SDS-PAGE exhibited exhibited an almost an identical almost identical band intensity band intensity profile (Figure profile5). (Figure The intensity 5). The ofintensity the 33.5 of kDa the 33.5 bands kDa corresponding bands corresponding to the unfolded to the unfolded protein decreasedprotein decreased over time over from time 35% from to 35% 17% andto 17% 39% and to 20%39% forto 20% RC-31(LPS) for RC-31(LPS) and RC-31, and respectively RC-31, respectively (Figure5 A,(Figure Lanes 5NH A, Lanes (0, 72 h)NH and (0, 572 B h) Lanes and NH5 B (0,Lanes 72 h)).NH During (0, 72 theh)). sameDuring time, the the same 27 kDa time, protein the 27 band kDa corresponding protein band tocorresponding the autoproteolytic to the fragmentautoproteolytic increased fragment over time increased for both over conditions time for fromboth conditions 0% to 35% andfrom 0% 0% to to 23% 35% for and RC-31(LPS) 0% to 23% and for RC-31,RC-31(LPS) respectively and RC-31, (Figure respectively5A, Lanes (Figure H (0, 72 5 A, h) andLanes 6 BH Lanes(0, 72 Hh) (0,and 72 6 h)).B Lanes These H almost (0, 72 h)). identical These autoproteolysisalmost identical profilesautoproteolysis for the RC-31(LPS) profiles for andthe RC-31(LPS) RC-31 samples and indicateRC-31 samples that LPS indicate did not that protect LPS did the unfoldednot protect OmpT the unfolded from autoproteolysis. OmpT from Theautoproteolysis. overall rate of The autoproteolysis overall rate of in autoproteolysis the RC-31(LPS) andin the RC-31 RC- samples31(LPS) and during RC-31 the samples observed during time the period observed were time calculated period were to be calculated 0.5%/h and to 0.3%be 0.5%/h/h, indicating and 0.3%/h, no significantindicating dinoff erencesignificant between difference the two samples,between suggestingthe two samples, that residual suggesting LPS is su ffithatcient residual to keep LPS OmpT is insufficient an active to state.keep Furthermore,OmpT in an active the overall state. refolding Furthermore, efficiency the overall of the RC-31(LPS)refolding efficiency and RC-31 of samplesthe RC- analyzed31(LPS) and immediately RC-31 samples after refoldinganalyzed wereimmediately calculated after to berefolding 65% and were 61%, calculated respectively, to be again 65% indicating and 61%, norespectively, significant again difference indicating in the refoldingno significant efficiency difference between in the tworefolding samples. efficiency In summary, between excess the LPStwo didsamples. not a ffInect summary, autoproteolysis excess of LPS the did unfolded not affect OmpT au nortoproteolysis did it affect of the the refolding unfolded e ffiOmpTciency nor of OmpTdid it IBs.affect Additionally, the refolding varying efficiency the of levels OmpT of LPSIBs. duringAdditionally, OmpT varying IB refolding the levels did not of significantlyLPS during OmpT alter the IB aboverefolding results did (datanot significantly not shown). alter the above results (data not shown).

Figure 5. OmpT refolding in the presence and absence of LPS. (A) OmpTOmpT IBs refoldedrefolded inin absenceabsence ofof LPS. ((BB)) OmpTOmpT IBs IBs refolded refolded in in the the presence presence of of 4.5 4.5 mg mg/mL/mL LPS. LPS. Samples Samples were were collected collected at 0, at 24, 0, and 24, 72and h.

Samples72 h. Samples were were either either heated heated for 45 for min 45 at min 100 at◦C 100 (H) °C or (H) not or heated not heated (NH) (NH) prior toprior gel to loading. gel loading.

MALDI-TOF experiments were undertaken to confirm that the substrate specificity of OmpT (RC-31(LPS) samples) for the dibasic sites was preserved (Figure 6), in agreement with the established substrate specificity reported in the literature [19,20]. When a short synthetic peptide (Seq: GARKVG, Biomolecules 2020, 10, 922 11 of 16

MALDI-TOF experiments were undertaken to confirm that the substrate specificity of OmpT Biomolecules(RC-31(LPS) 2020, samples)10, x FOR PEER for the REVIEW dibasic sites was preserved (Figure6), in agreement with the established11 of 15 substrate specificity reported in the literature [17,19,20]. When a short synthetic peptide (Seq: GARKVG, MWMW 627 627 Da; Da; Figure Figure 6A)6A) carrying carrying a dibasic sitesite inin itsits sequencesequence was was treated treated with with OmpT, OmpT, the the signal signal for for thethe full-length full-length peptide peptide was was no no longer longer visible visible in thethe MALDIMALDI spectra spectra (Figure (Figure6C), 6C), indicating indicating complete complete cleavagecleavage of of the the peptide. peptide. A control peptidepeptide (Seq: (Seq: GARIIG, GARIIG, MW MW 626 626 Da; Da; Figure Figure6B) lacking 6B) lacking a dibasic a dibasic site sitein in its its sequence sequence was was not cleavednot cleaved by OmpT, by OmpT, as seen as from seen the from peak the arising peak from arising the full-lengthfrom the full-length peptide peptide(Figure (Figure6D). 6D).

FigureFigure 6. 6.CharacterizationCharacterization of of substrate substrate specificity specificity of RC-31(LPS)RC-31(LPS) forfor the the dibasic dibasic sites: sites: MALDI-TOF MALDI-TOF spectraspectra of of a a short short syntheticsynthetic peptide peptide with with a dibasic a dibasic site insite the in absence the absence (A) and (A presence) and presence (C) of RC-31(LPS). (C) of RC- 31(LPS).MALDI-TOF MALDI-TOF spectra ofspectra a short of synthetic a short synthetic peptide lacking peptide a dibasiclacking site a dibasic in the absencesite in the (B) absence and presence (B) and presence(D) of RC-31(LPS). (D) of RC-31(LPS). Red arrows Red point arrows to the expected point to location the ex ofpected the m /locationz peaks for of thethe uncleaved m/z peaks peptide. for the uncleaved peptide. 3.4. Mutation of the Predicted LPS Biding Residues in OmpT 3.4. MutationFrom the of abovethe Predicted section, LPS it is evidentBiding Residues that LPS doesin OmpT not protect unfolded OmpT from autoproteolysis, but to understand whether LPS protects the folded OmpT from autoproteolysis, it is important to obtain From the above section, it is evident that LPS does not protect unfolded OmpT from folded OmpT without any LPS bound to it. This is easier said than done, since it is extremely difficult to autoproteolysis, but to understand whether LPS protects the folded OmpT from autoproteolysis, it is separate and purify OmpT IBs without residual LPS from the membrane fraction during the isolation important to obtain folded OmpT without any LPS bound to it. This is easier said than done, since it process. Chromatographic columns and kits designed for endotoxin removal were not considered is extremely difficult to separate and purify OmpT IBs without residual LPS from the membrane due to the harsh conditions the protein must endure during the process [40]. Moreover, they are not fractionknown during to completely the isolation remove LPS,process. especially Chromato thosegraphic that are boundcolumns to the and protein kits designed as co-factors. for The endotoxin other removalalternative were is not to express considered a mutant due OmpTto the harsh whose condit LPS bindingions the site protein is modified must suchendure that during it can nothe longer process [40].bind Moreover, LPS. According they are to not Vandeputte-Rutten known to completely et al. (2001) remove [1], LPS, one LPSespecially molecule those per that OmpT are is bound sufficient to the proteinfor OmpT as co-factors. activity. This The one other LPS alternative is thought is to to bind express to a putative a mutant LPS-binding OmpT whose motif LPS in OmpT binding whose site is modifiedlocation such was predicted that it can based no longer on the crystalbind LPS. structure According of FhuA to in Vandeputte-Rutte complex with LPSn [1 ].et FhuAal. (2001) is an [1], outer one LPSmembrane moleculeβ -barrelper OmpT protein is sufficient in E. coli that for sharesOmpT some activi structuralty. This one similarity LPS is with thought OmpT, to andbind has to aa known putative LPS-bindingLPS-binding motif motif in [41 OmpT]. In total, whose five location residues was of FhuA’s predicted LPS-binding based on motif the crystal are also structure present inof OmpTFhuA in complexat almost with the exactLPS [1]. same FhuA positions, is an and outer these membrane are: Arg138 β,-barrel Arg175, Lysprotein226, Glu in136 E., andcoli Tyrthat134 shareson OmpT some structural(Scheme 1similarity) [1]. with OmpT, and has a known LPS-binding motif [41]. In total, five residues of FhuA’sTwo LPS-binding recombinant motif mutant are also OmpTs present with in either OmpT the at three almost basic the residues exact same (mt3OmpT positions,LPS) or and all five these are:(mt5 ArgOmpT138, ArgLPS)175 of, Lys the226 predicted, Glu136, LPSand bindingTyr134 on residues OmpT modified(Scheme were1) [1]. expressed and purified (Table1). Two recombinant mutant OmpTs with either the three basic residues (mt3OmpTLPS) or all five (mt5OmpTLPS) of the predicted LPS binding residues modified were expressed and purified (Table 1). LPS should not bind to the modified putative LPS-binding motif in the two mutants and ought to remain inactive and exhibit no autoproteolysis.

BiomoleculesBiomolecules 2020,2020 10, ,x10 FOR, 922 PEER REVIEW 12 of12 16 of 15

Table 1. Specific activity of wild type OmpT and the various mutants used in this study. LPS should not bind to the modified putative LPS-binding motif in the two mutants and ought to Specific Activity remainOmpT inactive I.D. and exhibit no autoproteolysis. Residues Mutated * (pmol min−1 mg−1) OmpTTable 1. Specific activity of wild type None OmpT and the various mutants used 1.3 in this× 10 study.4 mtOmpTDB G216K and K217G 1.6 × 101 3 1 OmpT I.D. Residues Mutated * Specific Activity (pmol min− mg− ) 4 mt3OmpTOmpTLPS R None138E, R175E, K226E 1.3 104 1.5 × 10 × 4 mt5OmpTLPS Y134A, E136A, R138E, R175E, K226E 3 1.6 × 10 mtOmpT G K and K G 1.6 10 DB 216 217 × * For example, Y134A means the phenylalanine at position 134 was mutated to an alanine. OmpT R E, R E, K E 1.5 104 mt3 LPS 138 175 226 × OmpT Y A, E A, R E, R E, K E 1.6 104 Before wemt5 couldLPS discern134 the136 role of138 LPS175 in au226toproteolysis, a FRET-based× activity assay was * For example, Y134A means the phenylalanine at position 134 was mutated to an alanine. performed to confirm that the two mutants were inactive. Surprisingly, mt3OmpTLPS and mt5OmpTLPS exhibitedBefore low levels we could of protease discern activity the role that of LPS sign inificantly autoproteolysis, improved a FRET-basedwith the addition activity of assay LPS (Figure was 7, Tableperformed 1 and Figure to confirm S4) thatindicating the two that mutants the putative were inactive. LPS-binding Surprisingly, motifmt3 inOmpT OmpTLPS isand notmt5 criticalOmpTLPS for its activity.exhibited Nor can low we levels confirm of protease the role activity of LPS that bindin significantlyg to the improved putative with LPS-binding the addition motif of LPS in (Figure protecting7, activeTable OmpT1 and from Figure autoproteolysis. S4) indicating However, that the putative we cannot LPS-binding conclusively motif inrule OmpT out the is not possibility critical for that its LPS still bindsactivity. to NorOmpT can and we confirmmodifies the its role activity of LPS as binding well as to protects the putative it from LPS-binding autoproteolysis. motif in It protecting is therefore temptingactive to OmpT hypothesize from autoproteolysis. that there may However, be other we cannotLPS binding conclusively sites on rule OmpT out the that possibility are critical that LPS for its activitystill and binds capable to OmpT of andprotecting modifies it itsfrom activity autoprot as welleolysis. as protects In an it attempt from autoproteolysis. to independently It is therefore determine tempting to hypothesize that there may be other LPS binding sites on OmpT that are critical for its the number of LPS molecules that bind to OmpT, as well as to its mutants, we performed isothermal activity and capable of protecting it from autoproteolysis. In an attempt to independently determine titration calorimetry (ITC) experiments. Unfortunately, we were unable to get any clear results due the number of LPS molecules that bind to OmpT, as well as to its mutants, we performed isothermal to technicaltitration calorimetrydifficulties (ITC)in distinguishing experiments. Unfortunately, between LPS we binding were unable to OmpT to get anyand clearLPS resultsbinding due to to the detergenttechnical micelles difficulties in which in distinguishing the OmpT between was recons LPS bindingtituted to OmpT(data not and LPSshown). binding However, to the detergent previous studiesmicelles with inOprH, which an the outer OmpT membrane was reconstituted protein (data similar not shown). to OmpT, However, report previous that LPS studies binds with non- specificallyOprH, anto outerOprH membrane and it is possible protein similar that this to OmpT,is also reportthe case that with LPS OmpT binds non-specifically [42,43]. Their observation to OprH is in andline itwith is possible our hypothesis that this is and also prompts the case withfor further OmpT [in-depth42,43]. Their studies observation to identify is in the line LPS with binding our sites hypothesisthat are critical and prompts for OmpT’s for further activity in-depth and studiescapable to of identify protecting the LPS OmpT binding from sites autoproteolysis, that are critical if such forsites OmpT’s do exist. activity and capable of protecting OmpT from autoproteolysis, if such sites do exist.

FigureFigure 7. Comparison 7. Comparison of enzyme of enzyme activities activities of of the the fourOmpT OmpT variants variants (OmpT, (OmpT,mtOmpT mtOmpTDB, mt3DB,OmpT mt3OmpTLPS, LPS, and mt5andOmpTmt5OmpTLPS): LPS(A)): FRET-based (A) FRET-based activity activity measurements; measurements; ( (BB)) initial initial velocityvelocity measurements measurements used used to to calculatecalculate specific specific activity, activity, Table Table 1.1 .

4. Conclusions Based on the above findings, it is clear that only the unfolded OmpT undergoes autoproteolysis in vitro and LPS was unable to protect unfolded OmpT from undergoing autoproteolysis. A practical implication of this finding is that in vitro OmpT autoproteolysis can be reduced by improving protein refolding efficiencies such that there is less unfolded OmpT available for autoproteolysis. Folded active OmpT remains resistant to autoproteolysis since its protease activity did not significantly change over almost 40 days. However, we were unable to confirm if this protection from Biomolecules 2020, 10, 922 13 of 16

4. Conclusions Based on the above findings, it is clear that only the unfolded OmpT undergoes autoproteolysis in vitro and LPS was unable to protect unfolded OmpT from undergoing autoproteolysis. A practical implication of this finding is that in vitro OmpT autoproteolysis can be reduced by improving protein refolding efficiencies such that there is less unfolded OmpT available for autoproteolysis. Folded active OmpT remains resistant to autoproteolysis since its protease activity did not significantly change over almost 40 days. However, we were unable to confirm if this protection from autoproteolysis was due to LPS. Additionally, the possible incorporation of LPS into the detergent micelles may alter the normal interaction of LPS with OmpT, thus influencing the level of autoproteolysis and activity. In this regard, we are currently investigating OmpT reconstitution in supported lipid bilayers composed of different ratios of phospholipid and LPS, and whether these different membrane compositions affect OmpT (to be reported elsewhere). Together with additional structural data focusing on the OmpT–LPS interaction, we should be able to gain better insights into the role and mechanism of LPS in activating OmpT in vitro and in vivo. Moreover, the previously proposed [15] putative LPS-binding motif in OmpT appears not to be critical for its activity since mutant OmpTs, with the predicted LPS-binding residues mutated, exhibited excellent protease activity. Thus, it is possible that there might be multiple LPS binding sites present on OmpT or that LPS binding to OmpT is purely non-specific [42,43]. Finally, the data presented herein reiterate the need to exercise caution when interpreting the SDS-PAGE profiles of heat-modifiable proteins where reliable results are obtained only upon prolonged heating of the samples.

Supplementary Materials: The following are available online at http://www.mdpi.com/2218-273X/10/6/922/s1, Figure S1: Autoproteolysis of OmpT over time: OmpT samples were collected at 0, 24, and 72 h after refolding. The samples were run on SDS-PAGE: (A) Without heating the samples (B) Samples that were heated for 45 min before loading on SDS-PAGE. Figure S2: Mass spectrum of OmpT collected by LC-MS before deconvolution. Figure S3: Protease activity of RC-31 and RC-10 samples monitored over 40 days. OmpT IBs were refolded in presence of different concentrations of detergent and their kinetic activity was measured using the FRET peptide substrate (Abz-ARRA -Tyr (NO2)-NH2) over 40 days. Figure S4. OmpT protease activity in the presence and absence of externally added LPS. FRET based assay was used to measure protease activity of (A) wild type OmpT, (B) mtOmpTDB, (C) mt3OmpTLPS, (D) mt5OmpTLPS in the presence (red) and absence (blue) of LPS. Black lines indicate fluorescence intensity of control samples containing FRET peptide with LPS, but no OmpT added. Table S1: Molecular weight of OmpT and the fragments of OmpT. Author Contributions: M.N. and B.L. conceived the project and jointly supervised the entire project. K.J.M. performed the LC-MS experiments. G.S. performed the expression, purification, and other experimental work with inputs from S.G., G.S. and S.G. analyzed and interpreted the data and prepared the manuscript with final revisions from M.M., M.N., and B.L. All authors have read and agreed to the published version of the manuscript. Funding: This research work was funded by the Singapore Ministry of Education Academic Research Fund Tier 2 (MOE2018-T2-1-025) and the NTU-NU Institute for NanoMedicine located at the International Institute for Nanotechnology, Northwestern University, USA and the Nanyang Technological University, Singapore; Agmt10/20/14. KJM was supported by NIH/NCI training grant 5T32CA186897-02. This work made use of the IMSERC at Northwestern University, which has received support from the Soft and Hybrid Nanotechnology Experimental (SHyNE) Resource (NSF NNCI-1542205); the State of Illinois and International Institute for Nanotechnology (IIN). This research was carried out in collaboration with the National Resource for Translational and Developmental Proteomics under Grant P41 GM108569 from the National Institute of General Medical Sciences, National Institutes of Health. Conflicts of Interest: The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

References

1. Vandeputte-Rutten, L.; Kramer, R.A.; Kroon, J.; Dekker, N.; Egmond, M.R.; Gros, P. Crystal structure of the outer membrane protease OmpT from Escherichia coli suggests a novel catalytic site. EMBO J. 2001, 20, 5033–5039. [CrossRef][PubMed] 2. Hritonenko, V.; Stathopoulos, C. Omptin proteins: An expanding family of outer membrane proteases in Gram-negative . Mol. Membr. Biol. 2007, 24, 395–406. [CrossRef][PubMed] Biomolecules 2020, 10, 922 14 of 16

3. Stumpe, S.; Schmid, R.; Stephens, D.L.; Georgiou, G.; Bakker, E.P. Identification of OmpT as the protease that hydrolyzes the antimicrobial peptide protamine before it enters growing cells of Escherichia coli. J. Bacteriol. 1998, 180, 4002–4006. [CrossRef][PubMed] 4. Brannon, J.R.; Thomassin, J.-L.; Desloges, I.; Gruenheid, S.; Le Moual, H. Role of uropathogenic Escherichia coli OmpT in the resistance against human cathelicidin LL-37. FEMS Microbiol. Lett. 2013, 345, 64–71. [CrossRef] 5. Hui, C.-Y.; Guo, Y.; He, Q.-S.; Peng, L.; Wu, S.-C.; Cao, H.; Huang, S.-H. Escherichia coli outer membrane protease OmpT confers resistance to urinary cationic peptides. Microbiol. Immunol. 2010, 54, 452–459. [CrossRef] 6. Nielubowicz, G.R.; Mobley, H.L.T. Host-pathogen interactions in urinary tract infection. Nat. Rev. Urol. 2010, 7, 430–441. [CrossRef] 7. Flores-Mireles, A.L.; Walker, J.N.; Caparon, M.; Hultgren, S.J. Urinary tract infections: Epidemiology,mechanisms of infection and treatment options. Nat. Rev. Microbiol. 2015, 13, 269–284. [CrossRef] 8. Thomassin, J.-L.; Brannon, J.R.; Gibbs, B.F.; Gruenheid, S.; Le Moual, H. OmpT outer membrane proteases of enterohemorrhagic and enteropathogenic Escherichia coli contribute differently to the degradation of human LL-37. Infect. Immun. 2012, 80, 483–492. [CrossRef] 9. Montero, D.; Orellana, P.; Gutiérrez, D.; Araya, D.; Salazar, J.C.; Prado, V.; Oñate, Á.; Canto, F.D.; Vidal, R. Immunoproteomic Analysis To Identify Shiga Toxin-Producing Escherichia coli Outer Membrane Proteins Expressed during Human Infection. Infect Immun. 2014, 82, 4767–4777. [CrossRef] 10. Hartland, E.L.; Leong, J.M. Enteropathogenic and enterohemorrhagic E. coli: Ecology, pathogenesis, and evolution. Front. Cell. Infect. Microbiol. 2013, 3, 1–3. [CrossRef] 11. Thomassin, J.L.; Brannon, J.R.; Kaiser, J.; Gruenheid, S.; Le Moual, H. Enterohemorrhagic and enteropathogenic Escherichia coli evolved different strategies to resist antimicrobial peptides. Gut Microbes. 2012, 3, 556–561. [CrossRef][PubMed] 12. Urashima, A.; Sanou, A.; Yen, H.; Tobe, T. Enterohaemorrhagic Escherichia coli produces outer membrane vesicles as an active defence system against antimicrobial peptide LL-37. Cell. Microbiol. 2017, 19, e12758. [CrossRef][PubMed] 13. Premjani, V.; Tilley, D.; Gruenheid, S.; Le Moual, H.; Samis, J.A. Enterohemorrhagic Escherichia coli OmpT regulates outer membrane vesicle biogenesis. FEMS Microbiol. Lett. 2014, 355, 185–192. [CrossRef][PubMed] 14. Kramer, R.A.; Vandeputte-Rutten, L.; De Roon, G.J.; Gros, P.; Dekker, N.; Egmond, M.R. Identification of essential acidic residues of outer membrane protease OmpT supports a novel active site. FEBS Lett. 2001, 505, 426–430. [CrossRef] 15. Kramer, R.A.; Brandenburg, K.; Vandeputte-Rutten, L.; Werkhoven, M.; Gros, P.; Dekker, N.; Egmond, M.R. Lipopolysaccharide regions involved in the activation of Escherichia coli outer membrane protease OmpT. Eur. J. Biochem. 2002, 269, 1746–1752. [CrossRef] 16. Sugimura, K.; Nishihara, T. Purification, characterization, and primary structure of Escherichia coli protease VII with specificity for paired basic residues: Identity of protease VII and OmpT. J. Bacteriol. 1988, 170, 5625–5632. [CrossRef] 17. Dekker, N.; Cox, R.C.; Kramer, R.A.; Egmond, M.R. Substrate Specificity of the Integral Membrane Protease OmpT Determined by Spatially Addressed Peptide Libraries. Biochemistry. 2001, 40, 1694–1701. [CrossRef] 18. Rollauer, S.E.; Sooreshjani, M.A.; Noinaj, N.; Buchanan, S.K. Outer membrane protein biogenesis in Gram-negative bacteria. Philos. Trans. R. Soc. B Biol. Sci. 2015, 370, 20150023. [CrossRef] 19. Kramer, R.A.; Zandwijken, D.; Egmond, M.R.; Dekker, N. In vitro folding, purification and characterization of Escherichia coli outer membrane protease OmpT. Eur. J. Biochem. 2000, 267, 885–893. [CrossRef] 20. Wood, S.E.; Sinsinbar, G.; Gudlur, S.; Nallani, M.; Huang, C.-F.F.; Liedberg, B.; Mrksich, M. A Bottom-Up Proteomic Approach to Identify Substrate Specificity of Outer-Membrane Protease OmpT. Angew. Chemie Int. Ed. 2017, 56, 16531–16535. [CrossRef] 21. Wu, C.; Tran, J.C.; Zamdborg, L.; Durbin, K.R.; Li, M.; Ahlf, D.R.; Early, B.P.; Thomas, P.M.; Sweedler, J.V.; Kelleher, N.L. A protease for “middle-down” proteomics. Nat. Methods. 2012, 9, 822–824. [CrossRef] [PubMed] Biomolecules 2020, 10, 922 15 of 16

22. WU, C. Development and Application of Ompt Based Middle down Proteomics. Ph.D. Thesis, University of Illinois at Urbana-Champaign, Champaign, IL, USA, 2013. 23. Kudva, R.; Denks, K.; Kuhn, P.; Vogt, A.; Müller, M.; Koch, H.-G. Protein translocation across the inner membrane of Gram-negative bacteria: The Sec and Tat dependent protein transport pathways. Res. Microbiol. 2013, 164, 505–534. [CrossRef][PubMed] 24. Hirel, P.H.; Schmitter, M.J.; Dessen, P.; Fayat, G.; Blanquet, S. Extent of N-terminal methionine excision from Escherichia coli proteins is governed by the side-chain length of the penultimate amino acid. Proc. Natl. Acad. Sci. USA 1989, 86, 8247–8251. [CrossRef][PubMed] 25. Wessel, D.; Flügge, U.I. A method for the quantitative recovery of protein in dilute solution in the presence of detergents and lipids. Anal. Biochem. 1984, 138, 141–143. [CrossRef] 26. Carrió, M.M.; Villaverde, A. Localization of chaperones DnaK and GroEL in bacterial inclusion bodies. J. Bacteriol. 2005, 187, 3599–3601. [CrossRef] 27. Noinaj, N.; Kuszak, A.J.; Buchanan, S.K. Heat Modifiability of Outer Membrane Proteins from Gram-Negative Bacteria. Methods Mol. Biol. 2015, 1329, 51–56. 28. Koebnik, R. Structural and Functional Roles of the Surface-Exposed Loops of the β-Barrel Membrane Protein OmpA from Escherichia coli. J. Bacteriol. 1999, 181, 3688–3694. [CrossRef] 29. Sánchez, S.; Arenas, J.; Abel, A.; Criado, M.-T.; Ferreirós, C.M. Analysis of Outer Membrane Protein Complexes and Heat-Modifiable Proteins in Neisseria Strains Using Two-Dimensional Diagonal Electrophoresis. J. Proteome Res. 2005, 4, 91–95. [CrossRef] 30. Minetti, C.A.; Tai, J.Y.; Blake, M.S.; Pullen, J.K.; Liang, S.M.; Remeta, D.P. Structural and functional characterization of a recombinant PorB class 2 protein from Neisseria meningitidis. Conformational stability and porin activity. J. Biol. Chem. 1997, 272, 10710–10720. [CrossRef] 31. Visudtiphole, V.; Thomas, M.B.; Chalton, D.A.; Lakey, J.H. Refolding of Escherichia coli outer membrane protein F in detergent creates LPS-free trimers and asymmetric dimers. Biochem. J. 2005, 392, 375–381. [CrossRef] 32. Hong, H.; Tamm, L.K. Elastic coupling of integral membrane protein stability to lipid bilayer forces. Proc. Natl. Acad. Sci. USA 2004, 101, 4065–4070. [CrossRef][PubMed] 33. Tulumello, D.V.; Deber, C.M. Efficiency of detergents at maintaining membrane protein structures in their biologically relevant forms. Biochim. Biophys. Acta Biomembr. 2012, 1818, 1351–1358. [CrossRef][PubMed] 34. Anandan, A.; Vrielink, A. Detergents in membrane protein purification and crystallisation. Adv. Exp. Med. Biol. 2016, 922, 13–28. [PubMed] 35. Kleinschmidt, J.H.; Wiener, M.C.; Tamm, L.K. Outer membrane protein A of E. coli folds into detergent micelles, but not in the presence of monomeric detergent. Protein Sci. 1999, 8, 2065–2071. [CrossRef] [PubMed] 36. Brandenburg, K.; Garidel, P.; Schromm, A.B.; Andrae, J.; Kramer, A.; Egmond, M.; Wiese, A.; Andra, J.; Kramer, A.; Egmond, M.; et al. Investigation into the interaction of the bacterial protease OmpT with outer membrane lipids and biological activity of OmpT:lipopolysaccharide complexes. Eur. Biophys. J. 2005, 34, 28–41. [CrossRef] 37. Ebbensgaard, A.; Mordhorst, H.; Aarestrup, F.M.; Hansen, E.B. The Role of Outer Membrane Proteins and Lipopolysaccharides for the Sensitivity of Escherichia coli to Antimicrobial Peptides. Front. Microbiol. 2018, 9, 1–13. [CrossRef] 38. Eren, E.; Van Den Berg, B. Structural basis for activation of an integral membrane protease by lipopolysaccharide. J. Biol. Chem. 2012, 287, 23971–23976. [CrossRef] 39. Kukkonen, M.; Suomalainen, M.; Kyllönen, P.; Lähteenmäki, K.; Lång, H.; Virkola, R.; Helander, I.M.; Holst, O.; Korhonen, T.K.; Kyllonen, P.; et al. Lack of O-antigen is essential for plasminogen activation by Yersinia pestis and Salmonella enterica. Mol. Microbiol. 2004, 51, 215–225. [CrossRef] 40. Ongkudon, C.M.; Chew, J.H.; Liu, B.; Danquah, M.K. Chromatographic Removal of Endotoxins: A Bioprocess Engineer’s Perspective. ISRN Chromatogr. 2012, 2012, 1–9. [CrossRef] 41. Ferguson, A.D.; Welte, W.; Hofmann, E.; Lindner, B.; Holst, O.; Coulton, J.W.; Diederichs, K. A conserved structural motif for lipopolysaccharide recognition by procaryotic and eucaryotic proteins. Structure 2000, 8, 585–592. [CrossRef] Biomolecules 2020, 10, 922 16 of 16

42. Kucharska, I.; Liang, B.; Ursini, N.; Tamm, L.K. Molecular Interactions of Lipopolysaccharide with an Outer Membrane Protein from Pseudomonas aeruginosa Probed by Solution NMR. Biochemistry 2016, 55, 5061–5072. [CrossRef][PubMed] 43. Edrington, T.C.; Kintz, E.; Goldberg, J.B.; Tamm, L.K. Structural basis for the interaction of lipopolysaccharide with outer membrane protein H (OprH) from Pseudomonas aeruginosa. J. Biol. Chem. 2011, 286, 39211–39223. [CrossRef][PubMed]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).