Infection control by antibody disruption of bacterial quorum sensing signaling

Design and Synthesis of AIP-4 Hapten

Based on the reported structural information of AIP-4 [26], we designed and synthesized the hapten AP4-5 to elicit an anti-AIP-4 antibody immune response in mice (Fig. 2). Our reasoning for the chemical switch from the native thiolactone to a lactone-containing hapten was based on a lactones greater aminolytic stability [27]. This strategy ensured that the hapten conjugates remained structurally intact during the immunization process and subsequent immune response; thus, avoiding the generation of degradation products with unknown chemical and biological properties as previously uncovered for other QS molecules by our laboratory [28]. Furthermore, this substitution was also intended to prevent a possible intramolecular thiol exchange between the conserved thiolactone and the pendant cysteine thiol. Therefore, Fmoc-Serine(Trt)-OH was incorporated at position 4 in place of the native cysteine residue.

The hapten 5 was conjugated to the carrier proteins keyhole limpet hemocyanin (KLH) and bovine serum albumin (BSA) via a bifunctional linker (see Supplemental Fig. 1). Balb/c mice were immunized with the KLH conjugate using standard protocols [19]. Overall, the immunizations resulted in moderate titers (1600 - 3200), and based on ELISA analysis, 20 monoclonal antibodies (mAbs) were prepared. The affinities of the AP4-mAbs were determined against all four natural AIPs using competition ELISA methodology (see Supplemental Table 1). One of the mAbs, namely AP4-24H11, possessed strong binding affinity (K_{d AIP-4} ≈ 90 nM) and high specificity to AIP-4 while displaying little cross reactivity for the other AIPs (K_{d AIP-1} ≈ 5 μM, K_{d AIP-2} = >25 μM, K_{d AIP-3} = >25 μM). The ability of AP4-24H11 to discriminate between AIP1 and AIP4 is noteworthy as these two oligopeptides differ only at position 5 with an aspartic acid residue in AIP-1, and a tyrosine moiety in AIP-4. Thus, AP4-24H11 was selected for further biological evaluation.

AP4-24H11 alters expression of the virulent factors in S. aureus

α-Hemolysin and protein A are two major virulence factors in S. aureus, and expression of these proteins is tightly regulated by S. aureus signaling networks including the AIP-based agr QS system. The agr QS system positively regulates expression of α-hemolysin while protein A production is down-regulated by QS signaling. In order to test our rationale that anti-AIP antibodies are able to interfere with QS signaling in S. aureus, we examined whether the anti-AIP-4 mAb AP4-24H11 could modulate the expression of α-hemolysin and protein A in agr group IV strains, RN4850 and NRS168. First, we observed that AP4-24H11 affects the expression and/or secretion of S. aureus exoproteins, some of which might also be regulated by the agr QS circuits (see Supplemental Fig. 2a). As seen in Figure 3a, mAb AP4-24H11 can successfully reduce the α-hemolysin expression in S. aureus, furthermore, no hemolytic activity was observed on blood agar plates with the AP4-24H11 treated supernatant (Supplemental Fig. 2b). In contrast, protein A expression was significantly increased by mAb AP4-24H11 in RN4850, which is also consistent with agr QS inhibition.

The only structural difference between AIP-1 and AIP-4 is position 5, and our data suggest that AP4-24H11 is able to bind to AIP-1 with moderate affinity (≈ 5 μM). Therefore, we investigated whether AP4-24H11 could affect QS signaling in an agr group I strain, namely Wood 46. Clearly, AP4-24H11 was not able to block α-hemolysin expression in Wood 46 as effectively as in RN4850; however, a notable decrease in α-hemolysin production in Wood 46 grown in the presence of AP4-24H11 was evident (Fig. 3a). These data suggest that it might be possible to generate cross-reactive mAbs that suppress S. aureus QS signaling of two or more different agr groups.

While it could be argued that the decrease in toxin production and overall protein secretion is caused by an antibody-mediated growth defect, it is important to note that no significant growth changes of S. aureus were observed over a 24-hour growth period in the presence of AP4-24H11 (Fig. 3b). In addition, no discernable growth effects were observed with mAb SP2-6E11, an unrelated isotype control (κγ_2a) for AP4-24H11.

One of the important bacterial virulent factors regulated by QS is biofilm formation. In S. aureus, biofilm formation is known to be negatively regulated by agr QS signaling [29], which is indeed one of the problems in controlling S. aureus virulence through agr QS inhibition [30]. Consistent with previous studies, AP4-24H11-mediated QS inhibition led to increased biofilm formation in RN4850 (Fig. 3c). Although the increase of biofilm formation poses a significant problem in chronic infection of S. aureus, it represent a lesser predicament in acute infections and thus, mAb AP4-24H11 might be still an effective way to control such S. aureus infections.

Real time PCR analysis

To further examine agr QS inhibition by AP4-24H11, we performed real time-polymerase chain reaction (RT-PCR) analysis to evaluate if the observed changes in virulent factor expression were indeed caused by interference with the agr QS system, i.e. whether the presence of AP4-24H11 affects the transcription of rnaIII, the immediate product of agr autoinduction and the main QS effector in S. aureus [31]. As expected, the rnaIII transcriptional level in RN4850 during stationary growth phase was reduced significantly (> 50 fold), by AP4-24H11; this finding advocates that the alteration of α-hemolysin and protein A expression is a direct result of the interference of AIP-4-mediated QS signaling by AP4-24H11 (Fig. 3d). Yet, the subtle changes in overall exoprotein expression (see Supplemental Fig. 2) might be misconstrued to mean that AP4-24H11 does not block the QS signaling efficiently, however, our RT-PCR analysis provides evidence that AP4-24H11 significantly inhibits AIP4-based QS in S. aureus RN4850.

To gain greater appreciation of the specificity and potential limitations of antibody-based QS interference in S. aureus, we investigated the transcriptional level of two additional virulence regulators, namely sarA (staphylococcal accessory regulator) and saeR (staphylococcal accessory protein effector) [6, 32, 33], which control the response to environmental stresses as well as virulence factor expression in S. aureus. Importantly, no significant changes (≤ 2-fold) were observed in either sarA or saeR transcription, supporting that AP4-24H11 only affects agr QS system (Fig. 3d).

The transcription of α-hemolysin and protein A was analyzed by RT-PCR. As stated, (vide supra), significant changes were seen in protein expression level. In terms of transcription, the hla and spa genes were suppressed and elevated respectively ≈ 3 to 5 fold, again confirming that rnaIII affects not only transcription but also translation of these proteins as reported previously [31]. Finally, exofoliatin A (eta) transcription was investigated, which is another agr QS regulated toxin exclusively produced by AIP-4 utilizing S. aureus strains [34]. Gratifyingly, our data indicated that AP4-24H11 also decreased eta transcription by ≈ 10 fold (Fig. 3d).

Inactivation of AP4-24H11 by the synthetic AIP-4

Balaban and co-workers have proposed that the RNAIII activating protein (RAP) regulates agr QS signaling in S. aureus in an AIP-independent manner [35]. In addition, the same group has reported that a linear peptide, termed RNAIII-inhibiting peptide (RIP), could blunt RAP-mediated QS signaling resulting in a decrease of RNAIII transcription [36, 37]. Although we have presented evidence that AP4-24H11 inhibited agr QS through binding to AIP-4 and sequestering it from the cell growing medium, there was still the possibility that AP4-24H11 might affect other signaling systems in S. aureus including RAP, which in turn could affect agr QS network. As such, we investigated whether external addition of AIP-4 could restore the agr QS signaling network in S. aureus RN4850 in the presence of AP4-24H11. Thus, we treated AP4-24H11 with an equimolar amount of synthetic AIP-4 before addition to the S. aureus growth medium; this then would assure saturation of the antibody binding sites with the AIP-4 peptide. As seen in Figure 3e, the addition of synthetic AIP-4 efficiently reduced the quorum quenching effect of AP4-24H11, and as a result fully restored expression of α-hemolysin in S. aureus RN4850. As expected, this finding provides additional confirmation that AP4-24H11 indeed sequesters AIP-4 in S. aureus growth medium and inhibits AIP-dependent QS signaling in S. aureus in a strictly AIP-4-dependent manner.

AP4-24H11 inhibits S. aureus-induced apoptosis in mammalian cells

Recent studies have shown that incubation of Jurkat T cells with supernatant of S. aureus culture results in induction of apoptosis [38]. We treated Jurkat cells with the supernatants of S. aureus (RN4850 and Wood 46) cultures grown in the presence or absence of AP4-24H11. After incubation for 4 hours with the supernatant, the cleavage of poly(ADP-ribose) polymerase (PARP), a biochemical marker indicative of apoptosis induction, was evaluated in Jurkat cell protein extracts. As shown in Figure 4, AP4-24H11 prevented RN4850 supernatant (1 %) to induce PARP cleavage in Jurkat cells, and also partially inhibited the effect of Wood 46 supernatant. In fact, previous mechanistic studies proposed that one of the major factors in S. aureus supernatant responsible for inducing apoptosis might be α-hemolysin; our findings, (Fig. 3a and Fig. 4), support this positive correlation between expression of α-hemolysin and S. aureus-induced apoptosis.

AP4-24H11 blocks S. aureus-induced dermal injury in mice

Next, we investigated the potential of mAb AP4-24H11 to mitigate S. aureus induced injury in vivo by employing a murine subcutaneous infection model. Freshly grown log phase S. aureus RN4850 were suspended in PBS containing Cytodex beads and where indicated AP4-24H11 or control IgG. Subcutaneous injections of bacterial suspension or vehicle control were made in the flank of SKH1 hairless mice followed by close monitoring over seven days. Doses administered were 10⁷ or 10⁸ bacteria (colony forming units; cfu) and 0.6 or 0.06 mg AP4-24H11 or control IgG. Mice receiving 10⁷ cfu developed minimal hyperemia/edema followed by limited induration over 7 days (see Supplemental Fig. 3a). However as early as six hours after injection, mice receiving 10⁸ cfu suspended in saline or control IgG showed early-stage hyperemia / redness at the injection site and extending 3-5 mm horizontally and 5-10 mm vertically in a diagonal pattern along the flank (Fig. 5a). Upon reexamination at 18 hours, the same areas surrounding the injection site were devitalized, and the skin was transformed to a brittle, reddish-brown scab. Over the 7-day observation period, the hardened scab began to detach from the surrounding relatively normal appearing skin, and small amounts of purulent exudate were observed at the normal/necrotic junction. In contrast, skin injury was abrogated in mice that received 10⁸ bacteria with 0.6 mg AP4-24H11 (Fig. 5c). As anticipated, the lower dose of AP4-24H11 (0.06 mg) was not protective (Fig. 5b), and control mice receiving 10⁸ cfu with 0.6 mg control IgG were not protected (see Supplemental Fig. 3b). Mice that received an injection of PBS/Cytodex alone or containing 0.6 mg AP4-24H11 remained normal over the observation period with the exception of occasional local induration (Fig. 5d). Excitingly, animals that had received the protective dose of 0.6 mg AP4-24H11 in combination with S. aureus RN4850 did not develop any significant lesions over the 7 day observation period.

Passive immunization with AP4-24H11 protected mice from S. aureus-induced fatality

To evaluate the effectiveness of a passive immunization approach using AP4-24H11 against a lethal challenge with S. aureus, SKH1 hairless mice received a 1 ml i.p. injection of AP4-24H11, control IgG or vehicle (DPBS) followed 2 hours later by 0.5 ml DPBS^- containing 3 × 10⁸ S. aureus RN4850. As shown in Figure 6, all of the mice receiving AP4-24H11 (6/6) survived through the 8-day observation period. In contrast, only one of the DPBS treated control mice (1/6) and none of the control IgG treated mice (0/6) survived longer than 24 hours. These data further validated our immunopharmcaothereutic approach for combating acute S. aureus infections.

Significance

We have sought to elucidate the consequences of quorum sensing signaling using highly specific anti-autoinducer antibodies elicited against a rationally-designed synthetic hapten. Our data indicate that the scavenging of an AIP by a monoclonal antibody is sufficient to suppress QS-controlled expression of virulence factors in S. aureus, which establishes this as the first report of AIP removal from bacterial culture without genetic manipulations of the organism. Antibodies now comprise over one third of the molecules undergoing clinical evaluation, mostly for the cancer immunotherapy [39]. However, antibody-based therapies for the treatment of bacterial infections using QS as a molecular target are still in their infancy [39-41]. Conventional QS interference strategies utilizing small molecule antagonists are based on a competition between the bacterial autoinducer and the inhibitor for the bacterial AI receptor protein. In contrast, quorum quenching antibodies engage in competition with the AI receptor for the autoinducer. In this context, an immunopharmacotherapeutic strategy is highly appealing as the antibody acts as a “decoy-receptor” while possessing well documented pharmacokinetic behavior and pharmacodynamic response.

In total, antibodies generated against bacterial autoinducers represent a new and valuable set of immunological tools for both the study of QS-controlled processes and potentially an alternative strategy engaging immunopharmacotherapy for the prevention or treatment of infections in which QS signaling contributes to bacterial pathogenesis.

Synthesis of the linear protected peptide (3)

All N-α-Fmoc protected amino acids, coupling reagents and the resins for peptide synthesis were purchased from EMD Biosciences, Inc. (San Diego, CA). All other chemicals were purchased from Sigma-Aldrich Corp. (St. Louis, MO). ESI-MS analyses were performed with API150EX (PESCIEX, Foster City, CA), and HITACHI L-7300 and SHIMADZU SCL-10A were used for analytical and preparative HPLC experiments, respectively.

The peptide was synthesized by Fmoc SPPS on 2-chlorotrityl resin preloaded with the Fmoc-Met 1. An Fmoc-Ser(Trt)-OH was incorporated at the position of lactonization. All other residues were chosen with side chain protecting groups stable to dilute TFA and labile in 95% TFA. A short flexible linker was incorporated penultimate to the N-terminus by coupling Fmoc-8-amino-3,6-dioxaoctanoic acid. The N-terminal residue was Boc-Cys(SEt)-OH for eventual use in conjugation to carrier proteins. Specific Conditions: Batch synthesis was carried out on 1 mmol of resin swollen in DMF for at least 1 h. A solution of the protected amino acid, DIC, and HOBt (4 eq each) in 5 mL DMF was prepared and allowed to sit for 5 min for pre-activation, followed by the addition of 0.5 mL symcollidine. The cocktail was added to the resin for coupling, which was generally complete in 1 h. The resin was then washed with DMF and subjected to Fmoc deprotection with 20 % (v/v) piperidine in DMF (2 × 7 min). The resin was then washed with DMF and the next coupling reaction was carried out. When synthesis was complete, the resin was washed with DMF, then CH₂Cl₂, and finally with ether before it was placed in the desiccator. Cleavage (and Trityl Deprotection): The resin was added to a cocktail of 4 % TFA, 4 % triisopropylsilane (TIS) and 0.5 % H₂O in chloroform, and shaken for 6 h. The mixture was filtered, allowing the filtrate to drip into cold ether to precipitate the peptide. The ether mixture was centrifuged and the supernatant was decanted. The peptide was then washed (2x) with ether by re-suspending the solid in ether, centrifugation, and decanting the supernatant. The resulting solid was placed in a desiccator. Purification: The fully protected peptide 3 was dissolved in methylene chloride and purified by normal phase silica gel chromatography eluted with 5 % methanol in methylene chloride.

Lactonization of (3)

The protected linear peptide 3 was dissolved in 1,2-dichloroethane (previously dried over anhydrous MgSO₄) to give a final concentration of no greater than 1.0 mM. The solution was stirred and heated to 80 °C and 3 eq each of EDC and 4-DMAP were added; another equivalent each of EDC and 4-DMAP were added at both 24 h and 48 h into the reaction. The reaction was monitored by HPLC. After 4 days, the reaction mixture was cooled to room temperature, washed with 2 × 200 mL of 0.2 M KHSO₄ (aq), dried over anhydrous Na₂SO₄, and evaporated to dryness. The cyclized peptide 4 was purified by prep-HPLC. Yields range from 30-60 % as determined by analytical HPLC integration.

Global Deprotection and Disulphide Deprotection of (4)

The solid, purified peptide was dissolved in TFA containing 2 % TIS and stirred for 1 h. The mixture was then evaporated to dryness. Water was added and the mixture was frozen and lyophilized. The lyophilized solid was then dissolved in H₂O with tris(2-carboxyethyl)phosphine hydrochloride (TCEP). The mixture was stirred for 1 h and injected directly into the prep-HPLC for purification yielding AP4 hapten 5. The collected pure fractions were pooled, frozen, and lyophilized. ESI-MS: m / z calcd for C₅₇H₈₀N₁₀O₁₇S₂ (M + H), 1241.5; found, 1242.2.

Conjugation of (5) to KLH/BSA

Analysis of exoprotein secretion in S. aureus

After overnight growth on an agar plate at 37 °C, a single colony of S. aureus (RN4850 or Wood 46) was inoculated into 3 mL CYGP medium and grown for overnight (18 h) [42]. The overnight cultured cells were diluted to OD₆₀₀ ≈ 0.03 in fresh CYGP medium, and distributed to 5 mL polystyrene cell-culturing tube, where each tube contained 0.5 mL of the diluted cells and the appropriate antibody (0.2 mg/mL). After growth for 20-24 h at 37 °C in a humid incubator without agitation, the samples were transferred to the microcentrifuge tubes (1.5 mL) and centrifuged at 13,000 rpm for 5 min. The supernatants were sterilized by filtration through a Millex®-GV filter unit (0.22 μm; Millipore, Ireland), and analyzed by SDS-PAGE (10 % Bis-Tris gel, Invitrogen, Carlsbad CA). To confirm α-hemolysin and protein A expression, Western blot analyses were performed using the HRP conjugated sheep polyclonal α-hemolysin antibody (abcam Inc., Cambridge MA) and anti-Protein A mouse monoclonal antibody (Sigma-Aldrich, St. Louis MO) and murine mAb SP2-6E11 (Park and Janda, unpublished data) was used as a control antibody. To test hemolytic activity, the S. aureus supernatants (75 μL × 3) were applied onto the sheep blood agar plate, and the plates were incubated at 37 °C for 18 hours and at room temperature for another 24 hours.

Static Biofim Analysis

The biofilm assay was conducted by following a literature procedure with a few modifications [43]. After S. aureus cells (200 μL) were grown in tryptic soy broth (TSB) medium containing 0.2 % glucose with or without the antibody (0.2 mg/mL) in the polystyrene 96-well plate for 20-24 h without agitation, the plate was washed by submersion in water and dried. A crystal violet solution (200 μL, aq. 0.1 %) was added to stain the biofilm, and then the plate was washed vigorously with water followed by adding acetic acid (250 μL, aq. 30 %) to solubilize the remaining crystal violet. Absorbance was measured at 570 nm with Spectramax 250 (Molecular Devices, Sunnyvale CA).

Real Time-PCR Analysis

Overnight cultured S. aureus RN4850 cells were diluted to OD₆₀₀ ≈ 0.03 in fresh CYGP medium (1 mL) containing the antibody and grown for 20-24 h (OD₆₀₀ ≈ 2) at 37 °C without shaking. RNA from the cells was isolated using RNeasy® Mini Kit (QIAGEN Inc., Valencia CA) according to the manufacturer’s instructions. Isolated RNA was further purified by treating with RNase-Free DNase (QIAGENE Inc.) for 30 min at room temperature. The first-strand DNA was synthesized using SuperScript^TM First-Strand Synthesis System for RT-PCR (Invitrogen) using ≈ 300 ng of purified RNA. RT-PCR experiments were performed with at least two independent samples, and each experiment was set up in duplicate using LightCycler® FastStart DNA Master^PLUS SYBR Green I (Roche Applied Science, Indianapolis, IN). Generic SYBR Green Protocol (Roche) was used for the PCR conditions, and relative quantification analyses were performed with LightCycler® 2.0 system (Roche Applied Science) using the housekeeping GyrA gene was a reference. The sequence information of the primers used in our experiments is listed in the Supplemental Data [44].

Dermal infection model in mice

All experiments on mice were performed in accordance with TSRI guidelines and regulations. SKH1 euthymic hairless mice, 6-8 weeks old were obtained from Charles River Laboratories and housed in the biocontainment vivarium for one week before use in experiments. Brain heart infusion agar was from BBL (#211065) and CYGP broth contained 1% casamino acids (Fisher BP1424) 1% yeast extract (EMD 1.03753) 0.59% sodium chloride, 0.5% dextrose and 60 mM β-glycerol phosphate disodium salt (Fluka 50020) as described by Novick [42]. Cytodex 1 beads (GE Healthcare 17-0448-01) were suspended (1 gram in 50 ml) in Dulbecco’s Phosphate Buffered Saline without calcium/magnesium (Gibco) overnight at 20°C. The supernatant was decanted and the beads washed three times by suspension in DPBS and 1G sedimentation followed by autoclaving (121°C, 15 psi, 15 minutes). Staphylococcus aureus RN4850 (AIP4) was grown from frozen stock (BHI + 20% glycerol) on brain heart infusion agar plates 35°C overnight. Three representative colonies were combined to inoculate 2 ml CYGP broth, and after overnight incubation without shaking, 0.25 ml of the culture was used to inoculate 5 ml of CYGP followed by incubation at 35°C, 200 rpm for 3 hours. The culture was centrifuged 1,300 × G at 4°C for 20 minutes, the supernatant poured off, and the bacterial pellet was suspended in 1 ml DPBS without calcium/magnesium. The SKH1 received 200 μl intradermal flank injections containing S. aureus (1 × 10⁷ or 1 × 10⁸ bacteria), 4 μl packed volume Cytodex beads, DPBS, anti-AIP4 antibody or control IgG (0.6 or 0.06 mg). Additional control animals received 200 μl intradermal injections containing Cytodex beads or beads plus antibody. After injections were made the mice were monitored at least three times each day over a period of 4-7 days. At the conclusion of the monitoring period the mice were euthanized and tissues harvested for bacteriologic and histologic analysis.

Passive immunization of mice with AP4-24H11

S. aureus RN4850 were stored at -80°C in 20% glycerol/BHI medium, thawed and grown on BHI-agar plates overnight, and three separate colonies sampled to inoculate 2 ml CYGP medium. The inoculum culture was maintained 1 hour at 35°C without shaking, followed by shaking at 200 rpm for 3 hours. Aliquots of the freshly grown inoculum culture were transferred to 5 ml CYGP medium in 50 ml conical polypropylene tubes (1/20 dilution) followed by shaking at 200 rpm, 35°C for 3 hours. The bacteria were pelleted by centrifugation at 3,000 rpm (1300 × G) for 10 minutes, 4°C. The bacterial pellets were resuspended in Dulbecco’s phosphate buffered saline without calcium or magnesium (DPBS^-), and enumerated using a Petroff-Hausser counting chamber. Final dilutions were made in DPBS^- so that 3 × 10⁸ bacteria were administered i.p. in 0.5 ml. To maintain viability bacteria were administered within two hours of harvest.

MAb AP4-24H11, isotype-matched control IgG (1 mg each) or DPBS was administered i.p. in DPBS to SKH1 mice (6-9 weeks old; 6 animals per treatment group) followed two hours later by 0.5 ml DPBS^- i.p. containing 3 × 10⁸ S. aureus. The mice were monitored several times on the day of injection and twice each day on subsequent days, observing ambulation, alertness, response to handling and skin temperature measured by infrared thermometry (Raytek MiniTemp MT4) using a 1 cm diameter infrasternal skin site. Animals showing surface temperature consistently below 30°C and also diminished response to handling and weakened righting reflex were considered moribund and were euthanized.