The Binding of Antibiotics in Ompf Porin

Structure Article

The Binding of Antibiotics in OmpF Porin

Brigitte K. Ziervogel1 and Benoıˆt Roux1,* 1Department of Biochemistry and Molecular Biology, Gordon Center for Integrative Science, The University of Chicago, 929 East 57th Street, Chicago, IL 60637, USA *Correspondence: [email protected] http://dx.doi.org/10.1016/j.str.2012.10.014

SUMMARY OmpF and OmpC porins has been linked to antibiotic resistance in several reports, especially for E. coli and Salmonella typhimu- The structure of OmpF porin in complex with three rium (Nikaido, 2003). Additional studies have also shown that common antibiotics (zwitterionic ampicillin, anionic mutations of key OmpF residues can substantially alter suscep- ertapenem, and di-anionic carbenicillin) was deter- tibility to antibiotics (Bredin et al., 2002, 2003; Simonet et al., mined using X-ray crystallography. The three antibi- 2000). Since a decrease in membrane permeability results in otics are found to bind within the extracellular and reduced accumulation of cellular antibiotic levels, it often func- periplasmic pore vestibules, away from the narrow tions synergistically with the other resistance mechanisms by allowing time for drug inactivation or efflux pump expression OmpF constriction zone. Using the X-ray structures (Delcour, 2009). Therefore, a better understanding of how modi- as a starting point, nonequilibrium molecular fication of membrane permeability triggers bacterial resistance dynamics simulations with an applied membrane is necessary for the development of new antibiotic therapy voltage show that ionic current through the OmpF strategies. channel is blocked with bound ampicillin, but not Crystal structures for the E. coli OmpF and PhoE proteins with bound carbenicillin. The susceptibility of Es- (Cowan et al., 1992, 1995), and more recently for the E. coli cherichia coli expressing OmpF mutants to ampicillin OmpC (Basle´ et al., 2006), Klebsiella pneumoniae OmpK36 and carbenicillin was also experimentally character- (Dutzler et al., 1999), and Salmonella typhi OmpF (Balasubra- ized using microbiologic assays. These results maniam et al., 2012) proteins reveal large homotrimers of show that general diffusion by OmpF porins allows b-barrels consisting of a conserved antiparallel 16-strand struc- for transfer of molecules with varied charged states ture. The large number and configuration of b strands allows each monomer to form a separate central pore, which functions and give insights into the design of more efficient independently of the others in terms of solute permeation. antibiotics. A better understanding of this mecha- b-turn motifs connect the b strands on the periplasmic side nism will shed light on nature’s way of devising chan- of the channel, and a number of long loops protrude out on nels able to enhance the transport of molecules the extracellular side. One of these extracellular loops (L2) through membranes. contacts the adjacent porin monomer to stabilize the trimer structure, while a second long loop (L3) folds back into the middle of the barrel to form a ‘‘constriction zone,’’ which INTRODUCTION defines the size exclusion limit for permeation. The E. coli OmpF constriction zone is marked by a cluster of acidic resi- b-lactam antibiotics, the most widely used antibacterial agents, dues found on L3, which faces a cluster of basic residues on function by inactivating penicillin-binding proteins involved in the adjacent b-barrel wall. This gives rise to a strong transverse cell wall synthesis within the periplasmic space of gram-negative electric field that is thought to affect ion permeation and bacteria. The treatment of bacterial infections, however, is solute transfer (Im and Roux, 2002a, 2002b). A recent X-ray increasingly complicated by the ability of bacteria to develop structure of OmpF porin from KCl or RbCl-soaked crystals resistance to antimicrobial agents, now prevalent in a number has also provided experimental evidence for a specific ion of key gram-negative bacterial pathogens, including Escherichia pathway across the OmpF channel (Dhakshnamoorthy et al., coli, Salmonella, Klebsiella, Enterobacter, and Pseudomonas 2010). (Page` s et al., 2008). Bacterial resistance to b-lactam antibiotics The high-resolution crystal structure of OmpF porin has can be caused by three factors: modification by inactivating enabled detailed molecular dynamics (MD) simulation studies enzymes, expulsion by multidrug efflux pumps, and reduction of antibiotic permeation (Ceccarelli et al., 2004; Danelon of outer membrane (OM) permeability, the latter being the least et al., 2006; Hajjar et al., 2010a, 2010b, 2010c; Kumar et al., well understood. The OM, unique to gram-negative bacteria, 2010). Specifically, MD simulations showed that the zwitterionic acts as a selectivity barrier by providing an extra layer against ampicillin molecule remained stably bound when docked at the harmful compounds in the environment. General diffusion OmpF constriction zone. On the other hand, di-anionic carbeni- porins, expressed at high levels (>105 copies per cell) in the cillin moved toward the extracellular side of the OmpF constric- OM (Delcour, 2009), allow nonspecific transport of charged tion zone, reorienting parallel to the channel axis, where it and zwitterionic nutrient molecules. E. coli produce three major closely associated with the positively charged pore wall and general diffusion porins: OmpF, OmpC, and PhoE. The loss of blocked the channel lumen only partially (Danelon et al.,

OmpC into proteoliposomes. In these studies, zwitterionic compounds penetrated rapidly, flux of mono-anionic compounds was related to the hydrophobicity of the molecule, and di-anionic compounds penetrated at the slowest rates (Yoshi- mura and Nikaido, 1985). However, some 50%–70% of swelling may be due to the influx of counter ions for mono-anionic and di- anionic compounds, respectively (Yoshimura and Nikaido, 1985). For this reason, direct evidence for the interaction of antibiotics with OmpF porin has been sought from electrophysiology measurements of OmpF inserted in a lipid bilayer. In these single channel recording experiments, brief interruptions of ionic current were observed in the presence of ampicillin and amoxicillin (Danelon et al., 2006; Nestorovich et al., 2002), but no such current interruptions were observed in the presence of three anionic compounds: carbenicillin, azlocillin, and piperacil- lin (Danelon et al., 2006), reflecting either a lack of binding or reduced binding affinity within the pore. Therefore, the single channel recording experiments, together with the results from MD simulations of antibiotic permeation, suggest that zwitterionic antibiotics—but not their anionic analogs—make specific and favorable interactions near the OmpF constriction zone. Given that liposome swelling assays report fast diffusion rates for zwitterionic antibiotics, this counter-intuitively implies that binding at the constriction zone is necessary to facilitate efficient antibiotic transfer across OmpF porin. However, antibiotic binding within the pore can only be detected in single channel recordings if the flow of ionic current through the channel is Figure 1. Three b-Lactam Antibiotics Are Considered in the Present Investigation completely blocked or substantially reduced. It seems conceiv- The zwitterionic ampicillin molecule (top) is neutral but carries a large dipole able that some compounds may bind within the wide aqueous + moment owing to the positively charged amine group (NH3 ) and the nega- extracellular or periplasmic OmpF pores without actually altering À tively charged carboxylate group (COO ). The di-anionic carbenicillin (middle) the flow of ionic current. Ultimately, whether interactions in the + differs only by a substitution of ampicillin’s NH3 by a second negatively pore facilitate or hinder efficient antibiotic transfer across charged carboxylate moiety located at the segment linked to the aromatic À OmpF porin remains unclear. group (referred as a-COO while the carboxylate linked to the penem group will be referred as p-COOÀ). The anionic ertapenem (bottom) is longer than To address these issues, the OmpF channel was crystallized ampicillin and carbenecillin and harbors one positively and two negatively in complex with three common b-lactam antibiotics, ampicillin, charged groups. The segment of ertapenem that is observed in the X-ray carbenicillin, and ertapenem (Figure 1), allowing determination structure lies on the right of the dashed line. of their binding sites within the lumen of the pore for the first time, to our knowledge. The structures reveal a bound ampi- 2006). Computational techniques have also been used to cillin oriented perpendicular to the pore axis on the extracellular assess total antibiotic diffusion profiles across OmpF by using side of the constriction zone, while the bound carbenicillin and a metadynamics biasing algorithm to overcome the long time- ertapenem are oriented parallel to the pore axis near the peri- scale of antibiotic translocation (Ceccarelli et al., 2004; Danelon plasmic mouth or extracellular loops of the pore, respectively. et al., 2006; Hajjar et al., 2010a, 2010b, 2010c; Kumar et al., In addition, all-atom MD simulations based on the X-ray struc- 2010). Such calculations suggested that ampicillin visits an tures show that ionic current through the OmpF channel is energy minimum within the constriction region of the OmpF blocked with bound ampicillin, but not with bound carbenicillin. pore, and led to the proposal that interaction with residues in Lastly, the susceptibility of E. coli expressing mutants of the OmpF constriction zone would help facilitate antibiotic trans- OmpF porin to ampicillin and carbenicillin showed that dis- fer by partially compensating for the entropic and desolvation ruption of key drug-protein interactions resulted in a general costs of drug confinement in this narrow region (Danelon et al., increase in antibiotic susceptibility. These results provide 2006). However, due to the relatively short length of the MD evidence for the role of drug-protein interactions in bacterial trajectories, these simulation studies remain somewhat limited. permeation of antibiotics. Experimental data for antibiotic permeation through OM porins are scarce. Measurement of antibiotic flux in whole RESULTS cells reported rates on the order of 10–50 3 10À5 cm/s for the permeation of zwitterionic drugs across OmpF, but were Ampicillin Binds within the Extracellular Pore greatly reduced for anionic compounds (Nikaido et al., 1983). The structure of OmpF (Figure S1 available online) in complex Evidence for a direct role of general diffusion porins in mediating with ampicillin was solved at 2.0 A˚ resolution after subsequent the diffusion of b-lactams was further elucidated by liposome ampicillin soaking of Mg2+-free OmpF crystals (Figures 2A and swelling assays, involving reconstitution of purified OmpF or S2; Table 1). An Fobs À Fcalc difference map, with ligand

Figure 2. Crystal Structure of OmpF Porin in Complex with Three b-Lactam Antibiotic Molecules Three views are provided to show OmpF porin in complex with ampicilin (A), carbenicilin (B), and ertapenem (C). Left: Ribbon diagram depiction of a single OmpF monomer with the antibiotic (sticks) binding orientation shown. Middle: Zoomed-in view of OmpF residues (green or yellow sticks) within hydrogen-bonding distance (3.3 A˚ ) of the antibiotic molecule (orange sticks). Right: Space-ﬁlling view of the OmpF monomer with the bound antibiotic from the extracellular side (ampicillin and ertapenem) or from the periplasmic side (carbenicillin). Ampicillin is also coordinated by two water molecules shown as red spheres. The Fobs À Fcalc omit map at 2.0 s cutoff (purple mesh) shows the electron density for the antibiotic and the 2Fobs À Fcalc density map at 1.0 s cutoff (cyan mesh) shows the electron density for the modeled OmpF residues. L3 is colored yellow. See also Figures S1–S3.

Table 1. Intensity and Refinement Statistics: OmpF in Complex with Antibiotic OmpF/ OmpF/ OmpF/ OmpF Structure Ampicillin Carbenicillin Ertapenem Crystal Space group P3 P3 P3 Cell constants a, b, c (A˚ ) 116.6, 116.6, 116.7, 116.7, 116.8, 116.8, 53.3 53.9 52.8 a, b, g () 90, 90, 120 90, 90, 120 90, 90, 120 No. molecules in ASU 2 2 2 Data collection Wavelength 0.97918 0.97918 0.97918 Resolution (A˚ ) 50–1.98 50–2.3 50–1.87 (2.09–1.98) (2.42–2.3) (1.97–1.87) Measured reflections 245,574 224,254 471,080 Figure 3. Antibiotic Binding Sites Unique reflections 56,274 36,275 65,981 (A) The crystallographic positions of ertapenem (blue sticks), ampicillin (purple Redundancy 2.2 (2.1) 2.1 (2.0) 2.0 (1.9) sticks), and carbenicillin (orange sticks) are superimposed on a single OmpF monomer (ribbons with front b strands removed for clarity). I/s (I) 16.8 (2.0) 17.1 (2.0) 9.5 (2.1) (B) OmpF residues assessed in this study are shown and colored based on Completeness (%) 98.7 (97.9) 98.3 (98.5) 96.5 (96.9) those found in either the ampicillin (purple sticks) or carbenicillin (orange sticks) binding sites, or additional constriction zone residues (gray sticks). L3 is Rmerge 0.071 (0.424) 0.074 (0.413) 0.085 (0.339) colored yellow. Refinement Resolution (A˚ ) 1.98 2.3 1.87 R 0.202 0.219 0.188 G119 backbone oxygen atom, the side chains of D121 and Y32, and an additional water molecule. Finally, the phenyl ring R 0.239 0.272 0.227 free of ampicilin is positioned in a hydrophobic pocket near several Total atoms 5471 5405 5620 aromatic residues (Y22, Y32, F118), which could provide favor- Protein residues/ 666/2/263 666/2/215 666/2/424 able pi stacking interactions. Comparison with the apo structure ligand/water in ASU of OmpF indicates that the bound ampicillin does not induce ˚ 2 B factors (A ) a significant conformational change of the surrounding protein Protein 37.7 39.6 26.0 atoms; the total RMSD (including side chain atoms) for residues Antibiotic 58.7 76.1 44.2 that interact with ampicillin is only 0.13 A˚ compared with the apo Waters 42.9 38.6 38.1 structure (PDB 2ZFG). However, the B-factor values for ampi- ˚ 2 RMSD from ideal cillin (average of 59 A , compared with an average B-factor of ˚ 2 ˚ 2 Bond lengths (A˚ ) 0.025 0.023 0.024 38 A for all protein atoms and 33 A for interacting residue atoms), suggest that ampicillin experiences thermal fluctuations Bond angles () 2.12 2.06 2.16 within its binding site (Figure S3). This is consistent with the weak Values in parentheses are for the outermost shell. À equilibrium binding constant (1.9 M 1) of ampicillin to the OmpF pore (Nestorovich et al., 2002). omitted from the atomic model, showed a well-defined electron Carbenicillin Binds within the Periplasmic Pore density in the OmpF pore, which was absent from crystals grown The crystal structure of OmpF in complex with carbenicillin in the same condition without ampicillin. The ampicillin binding (2.3 A˚ ), a di-anionic molecule with a size and structure similar site was located 6A˚ above the OmpF constriction zone, within to ampicillin was also determined (Figure 2B; Table 1). The the larger extracellular pore vestibule. According to the electron OmpF-carbenicillin complex was also obtained by subsequent density, ampicillin is oriented perpendicular to the channel axis, soaking OmpF crystals in antibiotic solution. An Fobs À Fcalc in a configuration that allows a number of stabilizing interactions map, with carbenicillin absent in the atomic model, shows extra with OmpF residues on both sides of the pore. Specifically, electron density within the OmpF pore as well, but far from the ampicillin’s negatively charged COOÀ group is positioned in ampicillin binding site (Figure 3A). Carbenicillin is oriented between two arginine residues (R167, R168) and is also within parallel to the channel axis and closely associates with the hydrogen-bonding distance of the S125 side chain OH and b-barrel wall in the intracellular OmpF pore vestibule, about backbone nitrogen groups. Furthermore, a nearby ordered water 10 A˚ below the constriction zone. In its binding orientation, the molecule, found at this same position in the apo OmpF structure phenyl group of carbenicillin faces upward (toward the extra- (Protein Data Bank [PDB] 2ZFG), extends the hydrogen-bonding cellular side of the pore), and is positioned close to Y14. This network by forming a bridge between R168 and the ampicillin allows the nearby COOÀ group, located at the linkage segment COOÀ group. At the other end of the ampicillin molecule, terminating at the aromatic group (referred as a-COOÀ), to + the positively charged NH3 moiety is positioned between the make several contacts with basic residues (K16, R42, R82) just

A defined for the ligand at this position, it could not be used to model the entire ertapenem molecule. Ertapenem consists of Bound RMSD R167 Y22 0.9 Å two rigid planar structures (including phenyl pyrroline rings on Y32 one side and the penem group on the other) brought together by a flexible bond at a sulfur atom (indicated by a dashed line R168 F118 in Figure 1). Therefore, several rotameric states could exist for D121 Free this molecule. It has been previously shown that ertapenem S125 G119 samples different isomerization states around this sulfur atom in the active sites of b-lactamase enzymes (Kalp and Carey, 2008). For example, a recent crystal structure of ertapenem in 0.1 Å complex with a b-lactamase protein shows that two alternative B ertapenem conformations could be modeled in the X-ray density due to isomerization at the sulfur atom (PDB 3M6B and 3M6H) R42 RMSD K16 R82 (Tremblay et al., 2010). Furthermore, in one b-lactamase binding Bound Free 0.9 Å orientation, only half of ertapenem (the segment lying on the right E62 of the dashed line displayed in Figure 1) was modeled, which was the same segment of ertapenem that was modeled in complex Y14 with OmpF. It was, however, still possible to predict the most probable orientation for ertapenem based on the partial experimental electron density. In this conformation, the ertapenem K46 carboxylate group can hydrogen-bond with R168, an interaction 0.5 Å that was also identified for the ampicillin carboxylate moiety.

Figure 4. Antibiotic Dynamics in the OmpF Binding Site Molecular Dynamics Simulations of Bound Antibiotics Left: A superposition of 1 ns snapshots shows ampicillin (A) or carbenicillin (B) To assess the validity of the antibiotic binding sites for ampicilin conformations (purple sticks with the X-ray conformation shown in orange and carbenicilin, MD simulations were carried out starting from sticks) along the 10 ns MD trajectory. OmpF protein is depicted as ribbons with the X-ray structures. The binding site of ertapenem was not ˚ residues that form hydrogen bonds (within 3.3 A) with the antibiotic shown as simulated because the antibiotic is incomplete in the X-ray struc- sticks. L3 is colored yellow. Right: The average RMSD along the MD trajectory ture. To reduce the computational cost, a single OmpF monomer for the antibiotic in its OmpF binding site was calculated and compared with the RMSD for free antibiotic simulated in a water box (10 ns). A snapshot of the was considered. The monomer was embedded in a palmitoyl MD system is shown in Figure S4. oleoyl phosphatidyl choline (POPC) lipid membrane and fully solvated, including 0.15 M KCl (Figure S4). Despite fluctuations during the MD simulations, ampicillin did not leave its binding below the constriction zone. A hydrogen bond between a carbe- site as shown by a superposition of snapshots taken along the nicillin nitrogen moiety and E62 further stabilizes this binding trajectory (Figure 4A). In fact, the hydrogen bonds predicted orientation. On the other hand, only a single hydrogen bond, from the X-ray structure (ampicillin COOÀ and OmpF residues À + between the COO attached to the penem group (referred as S125, R167, and R168 and ampicillin NH3 and OmpF residues p-COOÀ) and K46, is observed in the periplasmic pore, allowing Y32 and D121) were maintained along the trajectory. Further- for considerable fluctuations of this half of carbenicillin. This is more, the molecule’s fluctuations were greatly reduced in the reflected in the B-factor values (Figure S3), which are 25% OmpF binding site when compared with free ampicillin simulated higher for atoms near the penem group (average B-factor of in a bulk water (Figure 4A). Carbenicillin also remained bound in 86 A˚ 2), compared with atoms on the phenyl side of the carbeni- its X-ray orientation during the MD trajectory (Figure 4B) and cillin molecule (average B-factor of 63 A˚ 2). Carbenicillin binding OmpF residues found to closely associate with carbenicillin in does not perturb the OmpF residues, as shown by an RMSD of the crystal structure (carbenicillin a-COOÀ and OmpF residues 0.26 A˚ for residues that can interact with carbenicillin, when K16, R42, and R82; carbenicillin nitrogen atom and OmpF compared with the corresponding residues from the apo struc- residue E62; and carbenicillin p-COOÀ and OmpF residue K46) ture (PDB 2ZFG). displayed hydrogen-bonding interactions along the MD trajectory. RMSD analysis also showed that while the phenyl side of Ertapenem Binds in the Extracellular Loop Region carbenicillin underwent little fluctuation, the penem side dis- The structure of OmpF in complex with ertapenem, a b-lactam played an increase in thermal motion, consistent with the large molecule belonging to the carbapenem family and harboring B-factor values also observed for these atoms in the crystal one positive and two negative charges, was determined at structure (Figures 4B and S3). 1.9 A˚ resolution after soaking OmpF crystals in ertapenem solution (Figure 2C; Table 1). The Fobs – Fcalc omit map, with erta- Antibiotic Effect on Ion Permeation and Selectivity penem unmodeled, showed clear density at the extracellular of OmpF Porin mouth of the OmpF pore, located 17 A˚ above the constriction MD simulations were used to examine the effect of a bound zone. The bound molecule is closely associated with the extra- antibiotic on ion permeation across the OmpF pore. Here, the cellular loops on one side of the OmpF pore. Although the KCl concentration was increased to 1 M and a membrane experimental electronic density from the omit map was well potential of ±150 mV was applied across the membrane along

À Table 2. Conductance and Ion Selectivity of OmpF Porin in 1 M ation event for the 150 mV simulation (Figure S5C). Most likely KCl because of long-range electrostatic effects, the binding of carbenicillin in the periplasmic OmpF vestibule even reduces the Vapp = +150 mV Vapp = À150 mV number of anions that entered the extracellular vestibule on the G (nS) I /I G (nS) I /I + K Cl - K Cl other side of the pore. Apo 0.43 1.00 0.58 1.46 Ampicillin 0.06 2.00 0 0 Antibiotic Diffusion Pathways across OmpF Carbenicillin 0.41 – 0.30 13.03 Analysis of the crystal structures suggests that interactions with

Vapp, applied voltage; G, conductance; IK/ICl, ion selectivity. See also specific residues may affect the binding and transfer rates of Figure S5. antibiotics through OmpF porin (Figure 3). Therefore, nonequilibrium steered molecular dynamics (SMD) simulations were carried out to further examine the ampicillin and carbenicillin the Z-axis to reproduce the conditions used in single channel transfer paths. Here, translocation of the molecule was driven recordings. In a 50 ns trajectory, ion permeation was observed, through the pore with an applied external force, thereby acceler- corresponding to ionic conductance values of 0.43 nS and ating this process from 100 ms to a few nanoseconds. The initial 0.58 nS for ±150 mV simulations, respectively (Table 2). These state was defined by the coordinates from the antibiotic-bound ion conductance values are reduced by 50% from experi- OmpF crystal structures to ensure that the experimentally mental measurements and previous results from Brownian located binding sites would be visited along the pathway. The dynamics simulations (Im and Roux, 2002a). Presumably, the antibiotic molecule was then dragged upward or downward reduced conductance observed in the present MD reflects an from the starting position, and is subsequently discussed as increase in the ion permeation energy barrier (on the order of a continuous permeation event from the extracellular side.

kBT) due to the loss of favorable interactions with the high Results of the SMD trajectories show the interaction profiles dielectric regions from the two OmpF aqueous pores that were within the OmpF pore for zwitterionic ampicillin and di-anionic not included in the present simulations. Nonetheless, a slight carbenicillin. This was defined by mapping protein residues cation selectivity was observed for the OmpF monomer (Fig- along the SMD trajectory within hydrogen-bonding distance ure S5A), which has also been reported for the trimer channel (3.3 A˚ ) of the translocating molecules. Along its transition using both experimental and computational approaches (Charrel pathway (Figures 5A, 5C, and S6A), ampicillin enters the extra- et al., 1996; Gill et al., 1998; Goldstein et al., 1983; Nikaido, 1998). cellular mouth of the OmpF channel near residues E29, G33, Alteration of ion permeation and selectivity for the OmpF pore T241, K243, and N246. It then continues to its extracellular was also investigated in the presence of antibiotic by all-atom binding site identified in the crystal structure, involving residues MD. When ampicillin was restrained in its X-ray conformation, Y32, D121, S125, R167, and R168, where ampicillin is oriented + ion current through OmpF was dramatically reduced. Total ion perpendicular to the channel axis with the NH3 moiety near conductance was either decreased over 7-fold (to 0.06 nS) the acidic L3 and the COOÀ group near the b-barrel wall lined or completely abolished for the ±150 mV simulations in the with basic residues. The interaction with R167 and R168 was presence of ampicillin (Figure S5B; Table 2). This finding is destabilized first, allowing the ampicillin COOÀ group to sample consistent with previous electrophysiology recordings for the area near the cluster of arginine residues (R42, R82, R132) in OmpF channels inserted in a bilayer lipid membrane, which the narrow OmpF constriction. Establishing these new contacts, + showed time-resolved current blockages across single channels while maintaining the interaction between the ampicillin NH3 in the presence of ampicillin (Danelon et al., 2006; Hajjar et al., group and D121, facilitates a reorientation of the ampicillin 2010a, 2010c; Nestorovich et al., 2002). Interestingly, the binding conformation from perpendicular to parallel relative to the pore of ampicillin in the extracellular vestibule even decreased the axis (Figures 5A and S6A). After crossing the constriction, ampi- number of ions entering from the periplasmic side of the pore cillin does not come within hydrogen-bonding distance of by 2-fold (Figure S5B), providing evidence that long-range other residues until it exits the periplasmic side of the pore electrostatic properties of porins are not solely controlled by near residues D149, E182, and Y182. the narrow constriction zone. On the other hand, previous Along the translocation pathway determined from the SMD electrophysiology data have shown that, in the presence of car- simulations (Figures 5B, 5D, and S6B), carbenicillin enters the benicillin and other anionic antibiotic molecules, ion permeation extracellular mouth near a similar region as ampicillin, close to across the OmpF channel is not disrupted, suggesting that residues T165, N246, T247, and S248, but is immediately di- anionic molecules do not cross the OmpF channel (Danelon verted to the b-barrel wall away from the acidic loop L3. In the et al., 2006). MD simulations of ion transfer across the OmpF extracellular pore, carbenicillin forms favorable interactions pore with carbenicillin bound in its X-ray site indicate that ions with S125, R167, and R168, contacts that were also identified flow freely in the presence of the antibiotic molecule. During for ampicillin. However, carbenicillin stays oriented parallel the 50 ns trajectory, the total ion conductance across the to the channel axis with its phenyl group pointed toward OmpF pore is only slightly decreased in the presence of carbeni- the extracellular mouth while it associates with the b-barrel cillin (0.41 and 0.30 ns for the ±150 mV simulations, respectively) wall and interacts with residues K80 and R132. Carbenicillin compared with conductance values calculated in the absence of transfers across the constriction zone by forming favorable inter- antibiotic (Table 2). However, carbenicillin binding induced actions with the arginine cluster and then settles into its high- a dramatic increase in cation selectivity, with zero ClÀ crossing affinity periplasmic binding site defined by interactions of the the pore in the + 150 mV simulation and only a single ClÀ perme- a-COOÀ moiety with K16, R42, and R82 and the p-COOÀ moiety

that disrupted the carbenicillin binding site in the periplasmic vestibule as shown by the disk diffusion test with no significant difference in ampicillin uptake for expression of Y14S, K16S, K46S, or E62Q OmpF proteins. Finally, mutations made to residues found in the narrow OmpF constriction zone showed varying levels of ampicillin susceptibility, with the R42S mutation, disrupt- ing the conserved arginine cluster, displaying the largest effect. Bacterial susceptibility to carbenicillin also followed the general trend of increased antibiotic uptake with expression of OmpF mutants that disrupt its binding interactions (Figure 6B). There was a significant increase in carbenicillin uptake for E. coli expressing the R42S mutant porin. Mutation of this residue, found in the middle of the three basic residues that hydrogen bond with the carbenicillin a-COOÀ group (K16, R42, R82), would severely destabilize this binding site. On the other hand, single mutations of either OmpF residues K16 or R82 display less of an effect. Furthermore, mutation of K46, found near the carbenicillin p-COOÀ group, showed little difference in carbenicillin uptake when compared with bacteria expressing wild-type (WT) OmpF. Carbenicillin uptake was also affected by mutation of OmpF residues found in the extracellular vestibule, near the high-affinity ampicillin binding site. For example, there was an 20% increase in carbenicillin susceptibility for bacteria expressing the D121N mutant OmpF porin as shown by the disk diffusion test. Finally, hydrophobic interactions did not affect bacterial uptake of carbenicillin because the Y14S OmpF mutation did not alter carbenicillin susceptibility.

DISCUSSION Figure 5. SMD Simulation of Antibiotic Transition Pathways across the OmpF Pore Crystallization of the OmpF channel in complex with b-lactam Conformations of ampicillin (A) or carbenicillin (B) as well as OmpF residues antibiotics allowed visualization of their specific binding sites that hydrogen-bond (within 3.3 A˚ ) with either ampicillin (C) or carbenicillin (D) within the lumen of the large aqueous pore and showed wide during the total antibiotic diffusion paths across the OmpF pore are shown as variation for their preferred binding locations. In contrast with sticks. See also Figure S6. broadly held views, these results, summarized in Figure 3A, show that the larger pore vestibular entryways of OmpF, rather with K46 in the periplasmic pore. Carbenicillin then continues than the narrow constriction zone, provide the lowest-energy along the b-barrel wall and exits the periplasmic mouth of the binding sites for these common antibiotic molecules. Specifi- channel near residues A1 and Q339. The large number of cally, ampicillin in its zwitterionic form binds to the extracellular contacts between carbenicillin and OmpF periplasmic residues pore vestibule, oriented perpendicular to the pore axis where it is in sharp contrast from the lack of interactions observed with simultaneously associates with both sides of the pore. The di- ampicillin in the periplasmic vestibule. anionic carbenicillin binding site is located in the periplasmic pore vestibule, where carbenicillin is oriented parallel to the Antibiotic Uptake of E. coli Bacteria Expressing OmpF pore axis and closely associates with the b-barrel wall. Finally, Mutant Porins harboring one positive and two negative charges, ertapenem is Susceptibility of E. coli expressing mutations of OmpF to ampi- positioned within the extracellular vestibule, but located at the cillin and carbenicillin was assayed using disk diffusion and mouth of the extracellular vestibule near the extracellular loops. agar dilution tests. Results of both tests showed a general It adopts an orientation parallel to the pore axis, and on the same increase in both ampicillin and carbenicillin susceptibility values face of the OmpF pore as carbenicillin. for E. coli bacteria expressing the mutant OmpF porins designed The ampicillin and carbenicillin binding sites identified by X-ray to disrupt antibiotic binding in the pore (Figure 6). In the case of crystallography were further validated by MD simulations. Such ampicillin (Figure 6A), the largest increase in antibiotic suscepti- simulations can offer an important test of the overall correctness bility was observed by mutation of OmpF residues (D121N, of the antibiotic binding configurations, since these were deter- + R167S, R168S) that hydrogen-bond with the charged NH3 and mined principally from the limited information provided by omit COOÀ ampicillin moieties. On the other hand, the Y22S mutant, maps and subsequently refined with B-factors higher than the involved only in hydrophobic interactions with the ampicillin surrounding protein atoms. It is thus encouraging to observe phenyl group, showed no effect on antibiotic susceptibility that, during the MD simulations, ampicillin and carbenicillin measured with either microbiologic test. Bacterial susceptibility form stable hydrogen bonds with OmpF residues and remain in to ampicillin was not affected by expression of OmpF mutants the specific configurations determined by X-ray crystallography.

Figure 6. Microbial Assay Measurements of Antibiotic Susceptibility The relative increase in measured bacterial growth inhibition zone diameters for expression of mutant OmpF porins (relative to WT) in the presence of either ampicillin (A) or carbenicillin (B) is shown by the disk diffusion assay (top). Bars are colored based on OmpF residues found in the extracellular ampicillin binding site (purple), periplasmic carbenicillin binding site (orange), constriction zone (gray), or both carbenicillin binding site and constriction zone (orange and gray). Error bars represent the deviation among three different measurements. Bottom: Minimal inhibitory concentrations (MICs) of ampicillin (A) or carbenicillin (B) are shown for bacteria expressing WT or mutant OmpF protein by the agar dilution assay. Bars are colored based on the WT residue characteristics: hydrophobic (green), hydrophilic (purple), acidic (red), and basic (blue). Results for bacteria expressing WT OmpF protein are colored gray. See also Figure S7 and Table S1.

Interaction profiles of ampicillin and carbenicillin along the entire in the constriction zone substantially modify antibiotic suscepti- OmpF channel length were further identified by carrying out SMD bility profiles. However, it is also possible that the impact of these simulations. While limited in scope, these SMD simulations are mutations was caused by alteration in pore size. For example, suggestive of the dynamical interactions during antibiotic diffu- a noticeable increase in cavity volume was observed for the sion through the pore. The results confirm that ampicillin adopts R132A and R132D mutants by protein modeling (Bredin et al., an orientation perpendicular to the pore axis while it is in the wide 2002). Therefore, it remains to be seen whether charged residues extracellular vestibule, most likely due to its large electric dipole. within the narrow constriction zone play the largest role in Similar interactions were observed in a previous metadynamics determining total solute diffusion rates. simulation study, which showed that the polar oxygens of ampi- A recent study of clinical strains of multidrug resistant E. coli cillin could form hydrogen bonds with extracellular residues identified sequence alterations of charged and polar residues R167 and R168 (Kumar et al., 2010). Furthermore, a reorientation within the pore of the related OmpC porin that altered antibiotic of ampicillin from perpendicular to parallel with respect to the susceptibility profiles without affecting overall pore size (Lou pore axis is observed for crossing the narrow constriction et al., 2011). To achieve a more complete assessment of the zone, as suggested previously (Ceccarelli et al., 2004; Danelon role of charged OmpF residues in antibiotic uptake, microbial et al., 2006; Hajjar et al., 2010a, 2010b, 2010c; Kumar et al., susceptibility assays were carried out for mutations of the resi- 2010). Finally, ampicillin does not interact with specific residues dues directly interacting with either ampicillin or carbenicillin in in the periplasmic vestibule. The complete path identified from the crystal structures. Results of microbial assays showed SMD for carbenicillin is very different. Carbenicillin maintains a general increase in both ampicillin and carbenicillin suscepti- interactions with the b-barrel wall along its entire translocation bility for E. coli expressing OmpF mutations that disrupt electro- path, including finding many favorable contacts in the periplas- static and hydrogen-bonding, but not hydrophobic, interactions mic vestibule. A previous MD simulation study also observed in the X-ray binding sites. Although alteration of residues in the similar interactions between carbenicillin and the b-barrel wall, extracellular OmpF vestibule affects bacterial infection by both allowing simultaneous interactions with the cluster of conserved ampicillin and carbenicillin, these assays show that only carbeni- arginine residues found in the constriction zone (R42, R82, R132) cillin susceptibility is affected by periplasmic pore mutations. and two additional arginine residues (R167, R168) in the extracel- Therefore, the rates of channel exit for zwitterionic and charged lular pore vestibule (Danelon et al., 2006). Therefore, while none molecules may be completely altered by a single charge sub- of the previous simulation studies correctly predicted the binding stitution, even in the large, aqueous OmpF periplasmic pore site of ampicillin or carbenicillin in the OmpF pore, many obser- vestibule. vations were broadly consistent with the present results. Ion permeation measurements through the OmpF channel in Previous antibiotic flux and microbial susceptibility assays for planar lipid bilayers has been an important technique to directly mutations of key charged OmpF constriction zone residues probe the interactions of antibiotics with the pore. These exper- showed an increase in bacterial susceptibility to several b-lac- iments revealed brief interruptions of ionic current in the pres- tams upon truncation of side chain residues (K16D, D113A, ence of ampicillin (Danelon et al., 2006; Nestorovich et al., D121A, R132A, R132D) (Bredin et al., 2002, 2003; Simonet 2002 ), suggesting that the zwitterionic molecule can act as et al., 2000; Vidal et al., 2005) and an increase in b-lactam resis- a pore blocker. Interestingly, no similar ion current blockage tance for substitution of a glycine residue for a protruding acidic was observed in the presence of carbenicillin or other anionic side chain on L3 (G119D and G119E) (Bredin et al., 2003; Simo- molecules. This led to the suggestion that di-anionic molecules net et al., 2000). This led to the conclusion that charge alterations do not pass through the OmpF pore and that efficient

Figure 7. Comparison of E. coli OmpF, E. coli OmpC, K. pneumoniae OmpK36, and S. typhi OmpF (A) Sequence alignment generated by the Dali server using E. coli OmpF (PDB 2OMF) as the query structure. Key constriction zone residues (acidic residues colored red and basic residues colored blue) and residues found within the ampicillin binding site (colored purple) are indicated. Lowercase letters represent residues without a match in the query structure. Numbering is based on the OmpF sequence. Ec, E. coli; Kp, K. pneumoniae; Ea, E. aerogenes; St, S. typhi. (B) Bound ampicillin, based on the OmpF-ampicillin crystal structure (top left), is superimposed with the structures of OmpC, OmpK36, and S. typhi OmpF and proposed hydrogen-bonding interactions (within 3.3 A˚ ) are indicated.

a diffusing antibiotic not only affects its own transfer pathway, but also that of the ions permeating through the channel. The functional advantage for a bacte- rium to have weak-afﬁnity antibiotic binding sites within the OmpF pore may be counter-intuitive. Previously proposed models suggest that an afﬁnity site close to the mouth of the channel might favor the translocation of antibiotics across OmpF (Ceccarelli et al., 2012). One possi- bility is that, at a concentration approach- ing the equilibrium dissociation constant

Kd of the binding site, the probability that the OmpF pore would be occupied and blocked by a bound molecule would increase such that the net flux would saturate to a finite maximum value inde- pendent of the surrounding antibiotic concentration. This imperfect defense barrier could provide an effective protec- translocation of antibiotics through the OmpF pore is associated tion until slower mechanisms, such as regulation of porin expres- with high affinity drug-protein interactions. However, the present sion or translation of catalytic enzymes or efflux pumps, are X-ray structures suggest a different interpretation of these activated. One additional series of experiments were carried results. Visual inspection of space-filling models of the OmpF out to test this conjecture. It confirms that bacterial resistance pore with the bound antibiotics (Figure 2, right) indicates that against high concentrations of ampicillin from co-expression of ion current should be blocked by ampicillin, but not by carbeni- the b-lactamase enzyme with OmpF is reduced 2-fold upon cillin or ertapenem. Furthermore, OmpF mutations that destabi- mutation of key residues either found within the constriction lize drug-protein interactions are shown here to increase uptake zone or identified from the X-ray structure (Table S1). of both zwitterionic ampicillin and di-anionic carbenicillin by It is then perhaps not surprising that mutation of single OmpF microbial assays. Results from non-equilibrium MD simulations residues can show dramatic differences in the infection pro- carried out in the presence of a membrane potential show that, perties for ampicillin and carbenicillin, two antibiotics that are indeed, ion current through the channel is blocked when ampi- widely prescribed in the clinical setting. This has also been cillin is bound in the configuration from the crystal structure. observed previously by a comparison of biophysical measure- On the other hand, ion current is not blocked when carbenicillin ments of diffusing ampicillin across three different bacterial por- is bound as in the X-ray structure, showing that single channel ins: E. coli OmpF, E. coli OmpC, and Enterobacter aerogenes recording was an unreliable approach for detecting the interac- Omp36 (Danelon et al., 2006; James et al., 2009; Nestorovich tion of this molecule with OmpF. Interestingly, although carbeni- et al., 2002). Interestingly, these proteins show high sequence cillin has only a small effect on the magnitude of ion current, the conservation within the narrow constriction zone, the previously bound di-anionic molecule yields a significant increase in cation assumed antibiotic binding site (Figure 7A), but there was no ion selectivity. Therefore, it appears that the molecular charge of conductance defect measured across OmpC and Omp36 in

the presence of ampicillin (James et al., 2009). Analysis of induce pKa shifts of key acidic OmpF residues and affect ion selectivity of the the OmpC crystal structure shows a slightly more constricted channel (Queralt-Martıń et al., 2011), a reduced concentration of MgCl2 pore (radius of 2.9 A˚ at the narrowest region, compared with (200 mM) was used, as previously reported (Reitz et al., 2009). Crystals of E. coli OmpF protein were obtained at 20C in a hanging drop (VDX plates, 3.2 A˚ for OmpF) (Basle´ et al., 2006), but examination of the Hampton Research) with the reservoir solution, 100 mM sodium cacodylate, K. pneumoniae OmpK36 crystal structure (Dutzler et al., 1999), 200 mM MgCl2, 50% (w/v) PEG 200, pH 6.5 mixed in a 1:1 ratio with protein which is 91% homologous to E. aerogenes Omp36, shows solution, 20 mM Tris-HCl, 0.1 M NaCl, 0.8% octyl-POE, pH 8.0; OmpF con- a slightly increased constriction zone (3.3 A˚ radius). Therefore, centration 5–8 mg/ml. Crystals were transferred to mother liquor solution it seems likely that residues outside the constriction zone are containing 1–2 M antibiotic and 25% glycerol for cryoprotection, but lacking 2+ responsible for providing the ampicillin binding site. Several resi- MgCl2 for two overnight soaks to ensure that Mg did not affect antibiotic binding (Figure S2). X-ray diffraction data were collected at beamlines NE- dues found within the ampicillin binding site for E. coli OmpF are CAT 24-ID at the Advanced Photon Source (Argonne National Lab). Process- altered in the X-ray structures of OmpC and OmpK36, as well as ing of diffraction data was carried out with iMosflm (Battye et al., 2011). The in a recent crystal structure of the OmpF homolog from S. typhi structures were solved by the molecular replacement method using the (Balasubramaniam et al., 2012)(Figure 7B). Specifically, binding program Phaser v2.1 (McCoy et al., 2007) implemented in the CCP4 suite interactions of the ampicillin carbonyl moiety would be destabi- (Winn et al., 2011). The structure of PDB 2ZFG was used as the starting lized in the OmpC, OmpK36, and S. typhi OmpF pores due to model and refined using REFMAC (Murshudov et al., 1997). Model building alteration of the proteins near the R167 and R168 residues. was carried out using the program COOT (Emsley and Cowtan, 2004). Since b-lactams are flexible molecules that can adopt a number of different con- Furthermore, the OmpC, OmpK36, and S. typhi OmpF proteins formations in solution, conformers that fit within the experimental density each lack the Y32 residue, which is found in the hydrophobic were generated by AFITT, a fully automatic ligand fitting process distributed pocket close to the ampicillin phenyl group and can also by OpenEye Scientific Software (Wlodek et al., 2006). The AFITT program + À hydrogen-bond with the ampicillin NH3 moiety. This not only was supplied with the OmpF coordinates and the Fobs Fcalc electron provides an explanation for the surprising loss of pore-blocking map (without antibiotic), and the coordinates for the antibiotic molecule by ampicillin for the OmpC and Omp36 proteins, but also gives were supplied separately. The AFITT program searched for all unmodeled density that was the correct size and shape for the ligand and the lowest further support that single point mutations within general diffu- energy solution was used for REFMAC refinement. sion porins can dramatically alter the uptake of antibiotics. It might therefore be technically possible to fine-tune the rate of Disk Diffusion and Agar Dilution Antibiotic Susceptibility Assays antibiotic uptake by gram-negative bacteria by controlling the E. coli BL21(DE3)Omp5 bacteria transformed with pRSF-1b plasmid encoding porin residues that an antibiotic molecule contacts during its the WT or mutant OmpF protein sequence were incubated in LB broth supple- translocation. mented with kanamycin (30 mg/ml), at 37 C, 250 rpm. OmpF expression was induced with 1 mM IPTG for 3 hr once an OD600 of 0.6 was reached. For the disk diffusion tests, cultures were spread on Mueller-Hinton (MH) agar plates EXPERIMENTAL PROCEDURES and either ampicillin (10 mg) or carbenicillin (100 mg) charged disks (Becton, Dickinson and Company) were placed on inoculated agar plates, supple- Bacterial Strains, Media, and Expression of OmpF Protein mented with 30 mg/ml kanamycin. Slow diffusion of antibiotic into the agar OmpF was purified from E. coli strain BL21(DE3)Omp5, a porin-deficient surrounding the disk created an antibiotic gradient, which decreased in mutant strain (Prilipov et al., 1998), which was kindly provided by Professor concentration outward from the disk. The diameter of the ‘‘inhibition zone’’ Ralf Koebnik. BL21(DE3)Omp5 cells were transformed with pRSF-1b plasmid around each disk was measured after overnight incubation at 37C. For the (Novagen) encoding the OmpF protein sequence engineered with a His6-tag agar dilution test, induced cultures were diluted 20-fold and spotted (5 ml) on and a tobacco etch virus (TEV) cleavage site between the OM targeting MH-agar plates containing 30 mg/ml kanamycin as well as 2-fold serial dilu- sequence and OmpF mature domain, leaving one glycine at the N-terminus tions of either ampicillin or carbenicillin (16, 8, 4, 2, 1, 0.5, 0.25, 0.125, and after TEV cleavage (Figure S1). This construct was engineered by the 0 mg/ml). In a separate agar dilution experiment, bacteria were co-transformed GenScript Corporation (GenScript USA) and is similar to that for PhoE (Van with pRSF-1b plasmid encoding OmpF and pGEM-3Z plasmid encoding Gelder et al., 1996). Mutants were constructed using Quikchange site-directed b -lactamase (Ampr) (Promega), incubated in LB broth supplemented with mutagenesis (Agilent). Bacteria were grown at 37 C in Luria-Bertani (LB) broth kanamycin (30 mg/ml) and ampicillin (100 mg/ml), and spotted as above on and supplemented with kanamycin (30 mg/ml). plates containing 30 mg/ml kanamycin and serial dilutions of ampicillin (160, 80, 40, 20, 10, 5, 2.5, 1.25, and 0.1 mg/ml). Plates were incubated overnight Inner Membrane Solubilization and OmpF Purification at 37C. MICs were defined as the concentration of antibiotic, which prevented Isolation of OM fractions was carried out as described previously (Taylor et al., bacterial growth (less than 10 colonies per spot). WT and mutant porin ex- 1998), which involves inner membrane solubilization by Triton X-100. OMs pression was compared by western blot quantification of E. coli lysates using were extracted overnight (room temp) with 3% n-octyl-b-D-glucopyranoside an anti-His primary antibody (QIAGEN) for OmpF and an antibody to DnaK (Anatrace) in 20 mM Tris-HCl, 1 mM EDTA, pH 8.0. Protein extractions were (Abcam) as a loading control after induction (Figure S7). centrifuged for 30 min at 125,000 3 g (room temp). OmpF protein was purified using Ni affinity chromatography (Ni-NTA agarose, QIAGEN) and eluted with All-Atom Molecular Dynamics Simulations 100 mM imidazole. To remove the N-terminal His6-tag, purified OmpF protein The simulation model, constructed by CHARMM-GUI (Jo et al., 2007, 2008), was incubated with TEV protease at a ratio of 20:1 (w/w) in 50 mM Tris-HCl, comprises one OmpF monomer, 103 POPC lipid molecules, and solvated 0.5 mM EDTA, 1 mM DTT, pH 8.0 and dialyzed overnight (room temp) into with 9,190 water molecules in 150 mM KCl (39 K+ and 12 ClÀ), for a total of 20 mM Tris-HCl, 0.1 M NaCl, 0.8% octyl-polyoxyethylene (OPOE), 0.5 mM 47,000 atoms. Tetragonal periodic boundary conditions were applied with PMSF, 1 mM sodium azide, pH 8.0. Final purification was conducted by gel- a distance of 70.29 A˚ in the XY-direction and 91.82 A˚ in the Z-direction. Elec- filtration chromatography (Superdex 200 10/300 GL, GE Healthcare) and trostatic interactions were calculated using the Particle-Mesh Ewald (PME) eluted with 20 mM Tris-HCl, 0.1 M NaCl, 0.8% OPOE, pH 8.0. Protein was method with a 75 3 75 3 100 grid. Protein residues E296, D312, and D127 concentrated in Amicon Ultra spin filters (Millipore) for crystallization. were protonated (Im and Roux, 2002a, 2002b; Varma et al., 2006). The simulations were performed at constant pressure (1 atm) and temperature (300 K) X-Ray Crystallography, Data Collection, and Refinement with a time step of 2 fs. Harmonic restraints (1 kcalmolÀ1A˚ À1) were applied Crystallization conditions with high Mg2+ were inspired by a previous study of to the Ca atoms at their crystallographic positions to prevent lateral movement OmpF (Yamashita et al., 2008). However, since divalent cations are known to of the protein. A single antibiotic molecule (ampicillin modeled in its

zwitterionic form, or carbenicillin in its anionic form) was positioned in the pore Ceccarelli, M., Danelon, C., Laio, A., and Parrinello, M. (2004). Microscopic using the coordinates extracted from crystallography. Restraints were initially mechanism of antibiotics translocation through a porin. Biophys. J. 87, 58–64. applied to antibiotic atoms and then released to relax the system. For SMD, Ceccarelli, M., Vargiu, A.V., and Ruggerone, P. (2012). A kinetic Monte Carlo ˚ 2 a dummy atom attached to the antibiotic center-of-mass (5 kcal/molA approach to investigate antibiotic translocation through bacterial porins. ˚ spring constant) was pulled at a constant velocity (0.00001 A/timestep) along J. Phys. Condens. Matter 24, 104012–104019. the ± Z-axis until the molecule escaped from either the extracellular or peri- Charrel, R.N., Page` s, J.-M., De Micco, P., and Malle´ a, M. (1996). Prevalence of plasmisic mouths of the pore. The KCl concentration was increased to 1 M outer membrane porin alteration in b-lactam-antibiotic-resistant Enterobacter (92 K+ and 83 ClÀ, respectively) for the ion permeation simulations to produce aerogenes. Antimicrob. Agents Chemother. 40, 2854–2858. a larger conductance. An electric field corresponding to a transmembrane potential of ± 150 mV was applied along the Z-direction (Roux, 2008), pressure Cowan, S.W., Schirmer, T., Rummel, G., Steiert, M., Ghosh, R., Pauptit, R.A., control was removed to avoid distortion of the system by the electric field, and Jansonius, J.N., and Rosenbusch, J.P. (1992). Crystal structures explain func- Langevin friction was removed for the ions. The all-atom CHARMM force field tional properties of two E. coli porins. Nature 358, 727–733. was used for protein (MacKerell et al., 1998) and lipids (Feller et al., 1997), and Cowan, S.W., Garavito, R.M., Jansonius, J.N., Jenkins, J.A., Karlsson, R., TIP3P (Jorgensen et al., 1983) for water. All the MD simulations were carried Ko¨ nig, N., Pai, E.F., Pauptit, R.A., Rizkallah, P.J., Rosenbusch, J.P., et al. out using the NAMD scalable molecular dynamics program (Phillips et al., (1995). The structure of OmpF porin in a tetragonal crystal form. Structure 3, 2005). CHARMM residue topology and parameter files for the antibiotic 1041–1050. molecules were constructed using Antechamber (Wang et al., 2006) with Danelon, C., Nestorovich, E.M., Winterhalter, M., Ceccarelli, M., and a restrained electrostatic potential (RESP) charge fitting procedure. Bezrukov, S.M. (2006). Interaction of zwitterionic penicillins with the OmpF channel facilitates their translocation. Biophys. J. 90, 1617–1627. ACCESSION NUMBERS Delcour, A.H. (2009). Outer membrane permeability and antibiotic resistance. Biochim. Biophys. Acta 1794, 808–816. The Protein Data Bank accession numbers for the atomic coordinates and Dhakshnamoorthy, B., Raychaudhury, S., Blachowicz, L., and Roux, B. (2010). structure factors for the reported crystal structures of OmpF in complex with Cation-selective pathway of OmpF porin revealed by anomalous X-ray diffrac- amipicillin, carbenicillin, and ertapenem are 4GCP, 4GCQ, and 4GCS, tion. J. Mol. Biol. 396, 293–300. respectively. Dutzler, R., Rummel, G., Albertı´, S., Herna´ ndez-Alle´ s, S., Phale, P.S., Rosenbusch, J.P., Benedı´, V.J., and Schirmer, T. (1999). Crystal structure SUPPLEMENTAL INFORMATION and functional characterization of OmpK36, the osmoporin of Klebsiella pneumoniae. Structure 7, 425–434. Supplemental Information includes seven figures and one table and can be found with this article online at http://dx.doi.org/10.1016/j.str.2012.10.014. Emsley, P., and Cowtan, K. (2004). Coot: model-building tools for molecular graphics. Acta Crystallogr. D Biol. Crystallogr. 60, 2126–2132. ACKNOWLEDGMENTS Feller, S.E., Yin, D., Pastor, R.W., and MacKerell, A.D., Jr. (1997). Molecular dynamics simulation of unsaturated lipid bilayers at low hydration: parameter- The help and advice from Balasundaresan Dhakshnamoorthy, Erin Adams, ization and comparison with diffraction studies. Biophys. J. 73, 2269–2279. Eduardo Perozo, and Anthony Kossiakoff, Suchismita Raychaudhury, Kay Gill, M.J., Simjee, S., Al-Hattawi, K., Robertson, B.D., Easmon, C.S.F., and Perry, Raj Rajashankar, John M. Tang, and Bob Eisenberg and the support Ison, C.A. (1998). Gonococcal resistance to b-lactams and tetracycline of Lydia Blachowicz are gratefully acknowledged. This work was supported involves mutation in loop 3 of the porin encoded at the penB locus. by the National Institutes of Health (NIH) through grant R01-GM062342 and Antimicrob. Agents Chemother. 42, 2799–2803. training grant GM007183 (to B.K.Z.). Diffraction data were collected at the Goldstein, F.W., Gutmann, L., Williamson, R., Collatz, E., and Acar, J.F. (1983). Advanced Photon Source (APS) of the Argonne National Laboratory through In vivo and in vitro emergence of simultaneous resistance to both b-lactam and the Northeastern Collaborative Access Team beamlines of the APS, which is aminoglycoside antibiotics in a strain of Serratia marcescens. Ann. Microbiol. supported by award RR-15301 from the National Center for Research (Paris) 134A, 329–337. Resources at the NIH and by the U.S. Department of Energy, Office of Basic Hajjar, E., Bessonov, A., Molitor, A., Kumar, A., Mahendran, K.R., Winterhalter, Energy Sciences through grant DE-AC02-06CH11357. M., Page` s, J.-M., Ruggerone, P., and Ceccarelli, M. (2010a). Toward screening for antibiotics with enhanced permeation properties through bacterial porins. Received: July 31, 2012 Biochemistry 49, 6928–6935. Revised: October 19, 2012 Accepted: October 26, 2012 Hajjar, E., Kumar, A., Ruggerone, P., and Ceccarelli, M. (2010b). Investigating Published online: November 29, 2012 reaction pathways in rare events simulations of antibiotics diffusion through protein channels. J. Mol. Model. 16, 1701–1708. REFERENCES Hajjar, E., Mahendran, K.R., Kumar, A., Bessonov, A., Petrescu, M., Weingart, H., Ruggerone, P., Winterhalter, M., and Ceccarelli, M. (2010c). Bridging time- Balasubramaniam, D., Arockiasamy, A., Kumar, P.D., Sharma, A., and scales and length scales: from macroscopic flux to the molecular mechanism Krishnaswamy, S. (2012). Asymmetric pore occupancy in crystal structure of of antibiotic diffusion through porins. Biophys. J. 98, 569–575. OmpF porin from Salmonella typhi. J. Struct. Biol. 178, 233–244. Im, W., and Roux, B. (2002a). Ion permeation and selectivity of OmpF porin: Basle´ , A., Rummel, G., Storici, P., Rosenbusch, J.P., and Schirmer, T. (2006). a theoretical study based on molecular dynamics, Brownian dynamics, and Crystal structure of osmoporin OmpC from E. coli at 2.0 A. J. Mol. Biol. 362, continuum electrodiffusion theory. J. Mol. Biol. 322, 851–869. 933–942. Im, W., and Roux, B. (2002b). Ions and counterions in a biological channel: Battye, T.G.G., Kontogiannis, L., Johnson, O., Powell, H.R., and Leslie, A.G. a molecular dynamics simulation of OmpF porin from Escherichia coli in an (2011). iMOSFLM: a new graphical interface for diffraction-image processing explicit membrane with 1 M KCl aqueous salt solution. J. Mol. Biol. 319, with MOSFLM. Acta Crystallogr. D Biol. Crystallogr. 67, 271–281. 1177–1197. Bredin, J., Saint, N., Malle´ a, M., De´ , E., Molle, G., Page` s, J.-M., and Simonet, James, C.E., Mahendran, K.R., Molitor, A., Bolla, J.-M., Bessonov, A.N., V. (2002). Alteration of pore properties of Escherichia coli OmpF induced by Winterhalter, M., and Page` s, J.M. (2009). How b-lactam antibiotics enter mutation of key residues in anti-loop 3 region. Biochem. J. 363, 521–528. bacteria: a dialogue with the porins. PLoS ONE 4, e5453. Bredin, J., Simonet, V., Iyer, R., Delcour, A.H., and Page` s, J.-M. (2003). Jo, S., Kim, T., and Im, W. (2007). Automated builder and database of protein/ Colicins, spermine and cephalosporins: a competitive interaction with the membrane complexes for molecular dynamics simulations. PLoS ONE 2, OmpF eyelet. Biochem. J. 376, 245–252. e880.

Jo, S., Kim, T., Iyer, V.G., and Im, W. (2008). CHARMM-GUI: a web-based scale production of mutant porins from E. coli. FEMS Microbiol. Lett. 163, graphical user interface for CHARMM. J. Comput. Chem. 29, 1859–1865. 65–72. Jorgensen, W.L., Chandrasekhar, J., Madura, J.D., Impey, R.W., and Klein, Queralt-Martıń, M., Garcıá-Gime´ nez, E., Mafe´ , S., and Alcaraz, A. (2011). M.L. (1983). Comparison of simple potential functions for simulating liquid Divalent cations reduce the pH sensitivity of OmpF channel inducing the water. J. Chem. Phys. 79, 926–935. pK(a) shift of key acidic residues. Phys. Chem. Chem. Phys. 13, 563–569. Kalp, M., and Carey, P.R. (2008). Carbapenems and SHV-1 b-lactamase form Reitz, S., Cebi, M., Reiss, P., Studnik, G., Linne, U., Koert, U., and Essen, L.-O. different acyl-enzyme populations in crystals and solution. Biochemistry 47, (2009). On the function and structure of synthetically modified porins. Angew. 11830–11837. Chem. Int. Ed. Engl. 48, 4853–4857. Kumar, A., Hajjar, E., Ruggerone, P., and Ceccarelli, M. (2010). Molecular Roux, B. (2008). The membrane potential and its representation by a constant simulations reveal the mechanism and the determinants for ampicillin translo- electric field in computer simulations. Biophys. J. 95, 4205–4216. cation through OmpF. J. Phys. Chem. B 114, 9608–9616. Simonet, V., Malle´ a, M., and Page` s, J.-M. (2000). Substitutions in the eyelet Lou, H., Chen, M., Black, S.S., Bushell, S.R., Ceccarelli, M., Mach, T., Beis, K., region disrupt cefepime diffusion through the Escherichia coli OmpF channel. Low, A.S., Bamford, V.A., Booth, I.R., et al. (2011). Altered antibiotic transport Antimicrob. Agents Chemother. 44, 311–315. in OmpC mutants isolated from a series of clinical strains of multi-drug resis- Taylor, R., Burgner, J.W., Clifton, J., and Cramer, W.A. (1998). Purification and tant E. coli. PLoS ONE 6, e25825. characterization of monomeric Escherichia coli vitamin B12 receptor with high MacKerell, A.D., Bashford, D., Bellott, M., Dunbrack, R.L., Evanseck, J.D., affinity for colicin E3. J. Biol. Chem. 273, 31113–31118. Field, M.J., Fischer, S., Gao, J., Guo, H., Ha, S., et al. (1998). All-atom empirical Tremblay, L.W., Fan, F., and Blanchard, J.S. (2010). Biochemical and struc- potential for molecular modeling and dynamics studies of proteins. J. Phys. tural characterization of Mycobacterium tuberculosis b-lactamase with the Chem. B 102, 3586–3616. carbapenems ertapenem and doripenem. Biochemistry 49, 3766–3773. McCoy, A.J., Grosse-Kunstleve, R.W., Adams, P.D., Winn, M.D., Storoni, L.C., Van Gelder, P., Steiert, M., El Khattabi, M., Rosenbusch, J.P., and and Read, R.J. (2007). Phaser crystallographic software. J. Appl. Cryst. 40, Tommassen, J. (1996). Structural and functional characterization of a His- 658–674. tagged PhoE pore protein of Escherichia coli. Biochem. Biophys. Res. Commun. 229, 869–875. Murshudov, G.N., Vagin, A.A., and Dodson, E.J. (1997). Refinement of macro- molecular structures by the maximum-likelihood method. Acta Crystallogr. D Varma, S., Chiu, S.-W., and Jakobsson, E. (2006). The influence of amino acid Biol. Crystallogr. 53, 240–255. protonation states on molecular dynamics simulations of the bacterial porin OmpF. Biophys. J. 90, 112–123. Nestorovich, E.M., Danelon, C., Winterhalter, M., and Bezrukov, S.M. (2002). ` b Designed to penetrate: time-resolved interaction of single antibiotic molecules Vidal, S., Bredin, J., Pages, J.-M., and Barbe, J. (2005). -lactam screening by with bacterial pores. Proc. Natl. Acad. Sci. USA 99, 9789–9794. specific residues of the OmpF eyelet. J. Med. Chem. 48, 1395–1400. Wang, J., Wang, W., Kollman, P.A., and Case, D.A. (2006). Automatic atom Nikaido, H., Rosenberg, E.Y., and Foulds, J. (1983). Porin channels in type and bond type perception in molecular mechanical calculations. J. Mol. Escherichia coli: studies with b-lactams in intact cells. J. Bacteriol. 153, Graph. Model. 25, 247–260. 232–240. Winn, M.D., Ballard, C.C., Cowtan, K.D., Dodson, E.J., Emsley, P., Evans, Nikaido, H. (1998). Antibiotic resistance caused by gram-negative multidrug P.R., Keegan, R.M., Krissinel, E.B., Leslie, A.G., McCoy, A., et al. (2011). efflux pumps. Clin. Infect. Dis. 27(Suppl 1 ), S32–S41. Overview of the CCP4 suite and current developments. Acta Crystallogr. D Nikaido, H. (2003). Molecular basis of bacterial outer membrane permeability Biol. Crystallogr. 67, 235–242. revisited. Microbiol. Mol. Biol. Rev. 67, 593–656. Wlodek, S., Skillman, A.G., and Nicholls, A. (2006). Automated ligand place- Page` s, J.-M., James, C.E., and Winterhalter, M. (2008). The porin and the ment and refinement with a combined force field and shape potential. Acta permeating antibiotic: a selective diffusion barrier in Gram-negative bacteria. Crystallogr. D Biol. Crystallogr. 62, 741–749. Nat. Rev. Microbiol. 6, 893–903. Yamashita, E., Zhalnina, M.V., Zakharov, S.D., Sharma, O., and Cramer, W.A. Phillips, J.C., Braun, R., Wang, W., Gumbart, J., Tajkhorshid, E., Villa, E., (2008). Crystal structures of the OmpF porin: function in a colicin translocon. Chipot, C., Skeel, R.D., Kale´ , L., and Schulten, K. (2005). Scalable molecular EMBO J. 27, 2171–2180. dynamics with NAMD. J. Comput. Chem. 26, 1781–1802. Yoshimura, F., and Nikaido, H. (1985). Diffusion of b-lactam antibiotics through Prilipov, A., Phale, P.S., Van Gelder, P., Rosenbusch, J.P., and Koebnik, R. the porin channels of Escherichia coli K-12. Antimicrob. Agents Chemother. (1998). Coupling site-directed mutagenesis with high-level expression: large 27, 84–92.