Cyclization increases bactericidal activity of arginine-rich cationic cell-penetrating peptide for Neisseria gonorrhoeae

ABSTRACT We previously reported that a linear cationic 12-amino acid cell-penetrating peptide (CPP) was bactericidal for Neisseria gonorrhoeae. In this study, our objectives were to determine the effect of cyclization of the linear CPP on its antibacterial activity for N. gonorrhoeae and cytotoxicity for human cells. We compared the bactericidal effect of 4-hour treatment with the linear CPP to that of CPPs cyclized by a thioether or a disulfide bond on human challenge and multi-drug resistant (MDR) strains of N. gonorrhoeae grown in cell culture media with 10% fetal bovine serum (FBS). The effect of lipooligosaccharide (LOS) sialylation on bactericidal activity was analyzed. We determined the ability of the CPPs to treat human cells infected in vitro with N. gonorrhoeae, to reduce the inflammatory response of human monocytic cells to gonococci, to kill strains of three commensal Neisseria species, and to inhibit gonococcal biofilms. The cyclized CPPs killed 100% of gonococci from all strains at 100 µM and >90% at 20 µM and were more potent than the linear form. The thioether-linked but not the disulfide-linked CPP was less cytotoxic for human cervical cells compared to the linear CPP. LOS sialylation had minimal effect on bactericidal activity. In treating infected human cells, the thioether-linked CPP at 20 µM killed >60% of extra- and intracellular bacteria and reduced TNF-α expression by THP-1 cells. The potency of the CPPs for the pathogenic and the commensal Neisseria was similar. The thioether-linked CPP partially eradicated gonococcal biofilms. Future studies will focus on determining efficacy in the female mouse model of gonorrhea. IMPORTANCE Neisseria gonorrhoeae remains a major cause of sexually transmitted infections with 82 million cases worldwide in 2020, and 710,151 confirmed cases in the US in 2021, up 25% from 2017. N. gonorrhoeae can infect multiple tissues including the urethra, cervix, rectum, pharynx, and conjunctiva. The most serious sequelae are suffered by infected women as gonococci ascend to the upper reproductive tract and cause pelvic inflammatory disease, chronic pelvic pain, and infertility in 10%–20% of women. Control of gonococcal infection is widely recognized as increasingly challenging due to the lack of any vaccine. N. gonorrhoeae has quickly developed resistance to all but one class of antibiotics and the emergence of multidrug-resistant strains could result in untreatable infections. As such, gonorrhea is classified by the Center for Disease Control (CDC) as an urgent public health threat. The research presented herein on new therapeutics for gonorrhea has identified a cyclic cell-penetrating peptide (CPP) as a potent molecule targeting N. gonorrhoeae.

in the US alone in 2021 (1,2).Resistance has developed to all classes of drugs used to treat gonorrhea making control of the disease more challenging, especially because there is no vaccine.Current management strategies reflect the limited number of antibiotics that are available for treatment and the relatively few that are in the process of development (3).
The third-generation cephalosporin, ceftriaxone, has been a mainstay of treatment since 1989.MICs for ceftriaxone have been quite stable in the US despite their rise internationally and the rapid increase in isolates with elevated MICs for azithromycin.The Sexually Transmitted Infections Treatment Guidelines issued by the Center for Disease Control (CDC) in 2021 recommend a single injection of 500 mg ceftriaxone intramuscu larly for uncomplicated gonorrhea at all anatomic sites with the addition of doxycycline 100 mg twice daily for 7 days if chlamydia coinfection is possible.There are few options for the treatment of individuals allergic to cephalosporin drugs (3).
Mucosal membranes of the cervix, uterus, and fallopian tubes in women and the urethra, mouth, throat, eyes, and rectum in both women and men can be infected by N. gonorrhoeae.Women have more asymptomatic or subclinical urogenital infec tions compared to men whereas rectal and pharyngeal infections are most commonly asymptomatic in both men and women (4,5).Untreated infections can lead to seri ous complications such as pelvic inflammatory disease, ectopic pregnancy, infertility in women, epididymitis, urethral stricture, reduced fertility in men, and blindness in neonates (6,7).
Several groups including our own have reported on the antibacterial activity of specific cell-penetrating peptides (CPPs) (8)(9)(10)(11).Bacterial membranes are composed of a relatively greater percentage of negatively charged molecules compared to the outer leaflet of mammalian cell membranes that mainly contain zwitterionic phosphatidylcho line and sphingomyelin, which enables targeting bacteria in vivo with cationic peptides such as CPPs that contain the basic amino acids lysine and arginine.Classically CPPs have been mainly explored as vehicles to facilitate translocation of cargos across mammalian cell membranes (12).
Endogenous cationic antimicrobial peptides (AMPs) such as LL-37, which are expressed ubiquitously in mammals, typically have activity against a broad range of pathogens.There are AMPs and CPPs that are both cationic and amphipathic and some AMPs can penetrate mammalian cells and some CPPs have antimicrobial activity (9).Thus, the differentiation of CPPs and AMPs based on their structure or bioactivity is not always clear.
AMPs have been characterized as having one or more types of simplistic interac tions with bacteria including detergent-like disruption of bacterial membranes in carpet models, formation of amphipathic pores in bacterial membranes in barrel stave models, and first aggregating on the bacterial surface followed by insertion into the bacterial membrane and inducing monolayer bending in barrel stave models (13).Studies also indicate that some AMPs kill bacteria primarily by binding to intracellular nucleic acids, proteins, or peptidoglycan after membrane penetration (14,15).
The lack of highly effective resistance to endogenous cationic AMPs, which are evolutionarily ancient, could be viewed as surprising when compared to the relatively rapid resistance that develops to most antibiotics.It has been proposed that bacterial resistance mechanisms co-evolved with AMPs as part of host-pathogen adaptation (16).The relatively non-specific nature and multiplicity of the activity of AMPs-membrane disruption, pore formation, and complexation with nucleic acids, peptidoglycan, and essential proteins-is thought to slow the development of resistance (17,18).Recent studies have examined the mechanisms of the limited bacterial resistance that does occur such as by the expression of the MtrCDE multi-drug resistant (MDR) exporter in N. gonorrhoeae and Neisseria meningitidis (16,(19)(20)(21)(22).
Analyses of cationic CPPs and AMPs have shown that the presence of arginine, which can form two hydrogen bonds and has greater charge distribution compared to lysine, enhances their seemingly contradictory capabilities of penetrating or disrupting non-polar membranes (12,(23)(24)(25)(26). Greater relative affinity of CPPs or AMPs for bacterial versus mammalian membranes has been identified as a determinant of antibacterial activity in vivo (27).However, if the affinity of the peptide for components of the bacterial outer membranes is too high, potency can be reduced by decreased translocation of the peptide into and/or through the lipid bilayer.The dichotomous effects of increased binding of AMPs to bacterial membranes or membrane components have been termed the Goldilocks dilemma (28).
We previously reported that a cationic 12-amino acid CPP with a linear (RXR) 4 structure, where X is 6-aminohexanoic acid (Ahx) and R is Arg, was bactericidal for N. gonorrhoeae (11).The presence of Ahx moieties is thought to reduce susceptibility to proteolysis and increase the flexibility and amphipathic nature of CPPs (29,30).Cyclizing antimicrobial peptides via stapling with covalent bridges between side chains or forming amide bonds at the termini can reduce susceptibility to proteolytic digestion and limit the formation of secondary structures, which could reduce the free energy required for binding the peptides to bacterial membranes, thus, enhancing bacterial cell wall penetration and intracellular activity (31)(32)(33).Increased selectivity for bacterial cells due to cyclization has also been demonstrated (33,34).Here we compare the bactericidal activity of the linear CPP to two forms that we cyclized to increase bactericidal efficacy and to reduce cytotoxicity for human cells.

Cyclic CPPs are more potent in bactericidal assays with human challenge and MDR strains of N. gonorrhoeae
We produced two cyclic forms of the linear (lCPP) peptide (Fig. 1A).One was achieved via a disulfide bond (dcCPP) that was formed after the addition and deprotection of Cand N-terminal cysteine residues (Fig. 1B, arrow).The second was made through chemical ligation of an N-terminal bromoacetyl group with a C-terminal cysteine to form a thioether (tcCPP) bond (Fig. 1C, arrow).The peptides were analyzed by electrospray liquid chromatography-mass spectrometry (LC-MS) with UV detection at 220 nm.The UV chromatograms revealed a single major peak for each peptide with overall purity >95% (data not shown).
Four-hour treatment with the lCPP, dcCPP, and tcCPP was bactericidal for human challenge (FA1090) and MDR gonococcal strains (F89 and H041) (Fig. 2A through C).The gonococci were significantly more sensitive to both cyclic CPP (cCPPs) compared to the lCPP at both 4 and 20 µM.There was a complete killing of the F89 and H041 isolates and >95 killing of FA1090 by the tcCPP at 20 µM.At 4 µM, the F89 bacteria were significantly more, and the H041 bacteria tended to be more sensitive to the tcCPP compared to the dcCPP.

tcCPP is not cytotoxic for human cells
To determine cytotoxicity for human cells, we analyzed lactate dehydrogenase (LDH) release from human cells, End1/E6E7 endocervical epithelial cells, ME-180 cervical epithelial cells, and activated THP-1 monocytes following 18 hours treatment (Fig. 3A  through C).Results showed that at 100 µM, which represents 10-fold more than the approximate minimum bactericidal concentration resulting in 50% sterilization (MBC 50 ) (Fig. 2), cell death was significantly increased only in THP-1 cells treated with the dcCPP.At 200 µM, the dcCPP was more toxic compared to the tcCPP for all three cell types.At 200 µM, the lCPP also was more cytotoxic compared to the tcCPP for the End1/E6E7 cells, but at this concentration, the lCPP was less cytotoxic for the THP-1 cells.We analyzed the cytotoxicity of the endogenous AMP LL-37 as a comparator.Treatment of the Endo1/ E6E7 or ME-180 cells with LL-37 at 20 µM (Fig. 3D) resulted in LDH levels that were relatively more than fivefold higher than those seen with the CPPs at 100 µM.

CPPs do not cause hemolysis
The in vitro assessment of the hemolysis of red blood cells is traditionally used as a preclinical assay of toxicity (35).The procedure we followed employed 0.1% Triton as the positive control and a 1-hour incubation period (36).The results (Fig. 3E) showed that there was no hemolysis from the lCPP, dcCPP, or tcCPP at either 100 or 200 µM.

tcCPP is more bactericidal for sialylated gonococci than the lCPP
We focused subsequent efforts on comparing the lCPP with the tcCPP, which had more potent bactericidal activity and less cytotoxicity than the dcCPP (Fig. 2 and 3).Although N. gonorrhoeae do not make CMP-Neu5Ac, the bacteria can utilize host CMP-Neu5Ac to sialylate the outer membrane lipooligosaccharide (LOS), which increases the anionic charge and confers greater resistance to complement-mediated killing by increasing binding to human factor H (37,38).We cultured the N. gonorrhoeae on gonococcal (GC) agar containing CMP-Neu5Ac to determine the relative bactericidal activity of the CPPs for bacteria with sialylated LOS.The results (Fig. 4A through C) showed that at both 4 and 20 µM the tcCPP was more potent than the lCPP for all three strains.
Our results showed that sialylation of the LOS had variable effects on the dosedependent bactericidal activity of the CPPs for the gonococci, reducing potency at some concentrations and increasing it at others.For example, at 4 µM the potency of the tcCPP increased after sialylation of the LOS on the FA1090 bacteria compared to the unsialylated FA1090 bacteria (Fig. 2A and 4A) but the effect of sialylation on potency was the reverse at 20 µM.The results demonstrated that the CPPs have bactericidal activity for sialylated N. gonorrhoeae, that the tcCPP was more potent than the lCPP regardless of sialylation state, and that there were variable CPP dose-dependent differences in the effect of sialylation on potency.

tcCPP can eradicate N. gonorrhoeae that infect human cells
Four-hour bactericidal assays were performed with the FA1090 strain in the presence of three types of human cells, End1/E6E7, ME-180, and activated THP-1 (Fig. 5).Gonococci can adhere to and invade human cells over this period (39,40), thus, 1% saponin was added terminally to enable surveying both extracellular and intracellular bacteria.The tcCPP was more potent than the lCPP at both 4 and 20 µM in the presence of all three cell types.At 100 µM, the tcCPP was able to completely eradicate the FA1090 N. gonorrhoeae infecting human End1/E6E7 and THP-1 cells, and 99% infecting ME-180 cells.The results demonstrated bactericidal activity ex vivo and provided additional evidence of the greater efficacy of the tcCPP compared to the lCPP.

Lower TNF-α levels expressed by THP-1 cells infected with gonococci treated with tcCPP versus lCPP
We previously showed that treatment with the lCPP reduced inflammatory signaling of infected THP-1 cells (11).To determine the relative effect of the tcCPP, we incubated activated THP-1 cells with each of the three bacterial strains at a multiplicity of infection (moi) of 1 in the presence of the lCPP and the tcCPP at 4, 20, and 100 µM (Fig. 6).The tcCPP treatment significantly decreased TNF-α expression relative to the lCPP at 20 µM in cells infected with FA1090 (Fig. 6A) and at 20 µM and 100 µM in cells infected with F89 (Fig. 6B).Relative to vehicle-only controls there were significant reductions in TNF-α expressed by monocytes infected by all three strains and treated with tcCPP at the highest concentration (100 µM).This was also true for monocytes infected with the FA1090 and F89 bacteria treated with tcCPP at 20 µM.There was little effect of 4 µM treatments likely due to the relative levels of TNF-α induction by LOS released from dead or dying bacteria (41).

CPPs exhibit bactericidal activity against three Neisseria commensal species
The rates of carriage of commensal Neisseria in the upper respiratory tract range from 3% to >45% of the population (42).We compared the bactericidal activity of the lCPP and tcCPP for strains of three commensal species, N. flavescens, N. lactamica, and N. subflava.As shown in Fig. 7, the tcCPP had significantly more potent bactericidal activity than the lCPP for all three species at 20 µM and for N. flavescens and N. lactamica also at lower concentrations.The sensitivity to the CPPs varied between the three species.The sensitivity of N. lactamica (Fig. 7B) to the tcCPP was most comparable to that of gonococcal strains FA1090 and H041 (Fig. 2A and C) whereas N. flavescens (Fig. 7A) had greater sensitivity, similar to the sensitivity of strain F89 (Fig. 2B).N. subflava (Fig. 7C) exhibited the least sensitivity to the tcCPP of the six commensal and gonococcal strains tested.

tcCPP partially eradicated N. gonorrhoeae biofilms
Gonococcal biofilms have been shown to occur in human cervical infections and in vitro on both primary urethral epithelial and cervical cells (43)(44)(45).Growth in biofilms may play a role in the shift of the gonococcus to anaerobic growth and in the asympto matic nature of N. gonorrhoeae infections (44,46).Biofilms are notorious for their ability to protect bacteria from antibiotics (47).We compared the capability of the linear and the tcCPP to eradicate gonococcal biofilms formed by plating bacteria in 96-well cell culture plates followed by incubation for 40 hours prior to treatment.The biofilm extent was assessed by staining with crystal violet followed by detection at 570 nm.The biofilm was not reduced by exposure to the lCPP at any of the concentrations tested.However, our results showed that the tcCPP reduced the biofilm by >35% at 100 µM and by >45% at 200 µM and was more efficacious than the lCPP at all three concentrations (Fig. 8).

DISCUSSION
Herein we have shown that cyclization of a linear (RXR) 4 CPP using a thioether bond increased the bactericidal potency for N. gonorrhoeae approximately 10-fold with a similar increase in sensitivity observed with bacteria having sialylated LOS (Fig. 2A  through C and 4).The potency of the dcCPP was intermediate to that of the lCPP and tcCPP.
For bactericidal assays, we added serum to the media, tested the potency of bacteria with sialylated LOS, and determined efficacy in the treatment of human cells infected with N. gonorrhoeae in vitro.In vitro testing with serum and host cells has been pro posed as essential in evaluating therapeutic potential because host-cell interactions such as binding to serum albumin and other serum proteins, and host cell membranes could reduce the antimicrobial activity of AMPs (48).Conversely, complement and other serum proteins can have antibacterial effects (49) that are, however, controlled by our comparisons to the vehicle-only samples.Thus, the potency of the tcCPP in the tested conditions supports the potential efficacy of the peptide in vivo.
The structural significance of cyclization to the increased potency is supported by the results with the dcCPP.Compared to thioether bonds, disulfide bonds can be more readily reduced entropically or through reduction by mammalian and bacterial NADPH and oxoreductases such as glutathione, which can occur both intra-and extracellularly (50)(51)(52).We postulate that the linearity of some of the dcCPP molecules could be the main determinant of the less potent bactericidal activity of this construct.
We recently showed that the FA1090 1-81-S2 human challenge strain expresses an oligosaccharide moiety on the LOS that can be mono-or disialylated (53).The presence of sialic acid on the bacterial LOS would increase its negative charge and, thus, should increase the affinity and localization of the CPPs on the surface of the bacteria.However, there is evidence from studies of 6-amino acid positively-charged cyclic AMPs that binding to the bacterial membrane was not simply equated with bactericidal activity because if the affinity of the AMPs for lipopolysaccharide (LPS) in the membrane was too high, bactericidal activity was reduced due to less translocation of AMPs into the bacterial membrane (28).A significant mechanism of bacterial resistance to AMPs is the reduction of the anionic charge on the bacterial cell membrane, for example, by alteration of the negative-charge state of LPS or LOS by dephosphorylation or by derivatization with galactosamine and phosphoethanolamine (PEA) (54)(55)(56)(57)(58)(59)(60)(61).The endogenous expression of PEA on lipid A of pathogenic Neisseria increases polymyxin resistance (60,62,63).Our data showed that all three CPPs were bactericidal despite sialylation of the LOS, or PEA substitution of lipid A as present on both N. lactamica and FA1090 N. gonorrhoeae LOS (53,64).
All three commensal Neisseria strains were sensitive to the tcCPP.N. flavescens, which lacks PEA on the diphosphoryl lipid A, was the most susceptible.Some commensal Neisseria has been reported to be highly resistant to antibiotic treatment and, being capable of horizontal gene transfer to N. gonorrhoeae, can serve as a reservoir of MDR-conferring genes (65)(66)(67).The potential of the commensal bacteria to harbor resistance genes recently prompted a call to expand gonococcal surveillance to include resistance profiling and whole genome sequence of commensal Neisseria isolates (68).Although they are a component of normal microbiota in the oropharynx, perhaps bactericidal activity against the commensal Neisseria could be advantageous in this era of increasing MDR gonorrhea.The bactericidal potency observed of the tcCPP for N. gonorrhoeae and commensal Neisseria spp.suggests that efficacy also may be observed for N. meningitidis, which we plan to explore in the future.This postulate is supported by the potency reported of LL-37 for N. meningitidis, which was similar to that which we reported for N. gonorrhoeae (11,69,70).This is even more significant because of the adaptation of a non-encapsulated meningococcal clade of a hyperinvasive lineage to the urogenital tract that could better enable horizontal gene transfer between these two pathogenic Neisseria species (71).
Treatment with the tcCPP at 100 and 200 µM (Fig. 8) produced dose-responsive partial eradication (~30% and 50%) of a preformed gonococcal biofilm.This is important because decreased bactericidal activity of antibiotics which can be as high as three to four orders of magnitude has been observed with biofilm compared to the planktonic forms of bacteria (72).The ability to survive in the biofilm form may play a role in two important characteristics of gonococcal infection, the increased MDR of N. gonorrhoeae and the persistent asymptomatic infections that occur (43,45).Numerous differences have been identified between bacteria in the planktonic and in the biofilm form, which gonococci and virtually all bacteria can form via the creation of a protective extracellular glycocalyx that promotes strong binding interactions with a surface enabling persistence with limited growth in a favorable environment that functionally mimics eukaryotic tissue (44,72).
The lack of cytotoxicity exhibited for any of the mammalian cell types by either the lCPP or the tcCPP at 100 µM supports the potential for clinical utility.This is in contrast to the more significant cytotoxicity exhibited by the endogenous human cathelicidin, LL-37.There have been clinical trials of several AMPs of synthetic or natural origin for bacterial infections within the last few years including LL-37, PL-5 (peceleganan), POL7080 (murepavadin), which is derived from protegrin-1, and DPK-060, which is derived from the human protein kininogen.Only two cationic AMPs, polymyxin B and colistin (polymyxin E), currently are used clinically.The latter is administered by intramuscular or intravascular injection as the inactive prodrug, colistin methanesulfo nate, that is converted to colistin in vivo.Systemic use of polymyxin B and colistin, which can give rise to dose-limiting toxicity, has re-emerged as salvage treatment for MDR Gram-negative bacterial infections (73)(74)(75).
Potential for clinical application is supported by the results from analysis of the stability of the lCPP conjugated to a phosphorodiamidate morpholino oligomer in human serum after 24 hours, which revealed significant levels of the full-length intact conjugate along with apparently similar levels of proteolytic products produced by cleavage of the R-R bonds in the lCPP (76).For successful treatment of gonococcal infections in humans, we expect that the tcCPP could be administered systemically by injection similarly to colistin, or by newer methods such as by incorporation of hyaluroni dase to enable subcutaneous auto injection of larger volumes (77).
In summary, the dcCPP and the tcCPP had enhanced potency in bactericidal assays with N. gonorrhoeae compared to the lCPP.The tcCPP compared to the dcCPP was more potent in bactericidal assays with strain F89 and was more potent than the lCPP for sialylated gonococci and for gonococci infecting cervical and monocytic cells in vitro.This mirrored the greater reduction in expression of TNF-α by monocytes incubated with bacteria treated with the tcCPP compared to lCPP.The selective toxicity of the tcCPP for bacteria compared to mammalian cells was improved relative to the lCPP.The tcCPP was bactericidal to gonococci with sialylated LOS and showed efficacy in eradicating the biofilm form of N. gonorrhoeae.Based on these promising results we plan to continue testing the antimicrobial efficacy of cyclized analogs of the lCPP and specifically the tcCPP based on its lack of toxicity for mammalian cells, the potential for slow development of resistance, and its bactericidal activity.

Bacterial strains
Origins of the FA1090 1-81-S2 strain of N. gonorrhoeae and the more recent MDR isolates (H041 and F89, WHO X and Y, respectively) were described previously (78).N. flavescens strain 4322, N. lactamica strain 1934, and N. subflava strain 52 were from the collections of the late Dr. Herman Schneider of the Walter Reed Army Institute of Research and of the Neisseria Repository formerly housed at UC Berkeley.

Cell-penetrating peptides
The CPPs and LL-37 were made by Anaspec (Fremont, CA, USA) using Fmoc solid-phase peptide synthesis.Analysis by reverse-phase (C18 column) LC-MS with electrospray ionization revealed all three peptides were >95% pure.A run time of 8.5 minutes was used for the LC with a 5%-60% gradient of buffer A (H 2 O) and buffer B (acetonitrile), both containing 0.1% trifluoroacetic acid.The lCPP has a 12 residue sequence H-R-X-R-(R-X-R) 2 -R-X-R-OH in which X is 6-aminohexanoic acid.The dcCPP was produced by the addition of a cysteine on each terminus and oxidation by incubation with iodine (I 2 ) and, thus, had a structure corresponding to cyclo where S is sulfur.The tcCPP was formed via a thioether bond and had a structure corresponding to cyclo[R-X-R-(R-X-R) 2 -R-C(S-)]-NH 2 , where S is sulfur, by reaction of an N-terminal bromoacetyl group with a C-terminal cysteine.

Bactericidal assays
Bacteria were grown overnight on Difco GC agar (Becton, Dickinson, Sparks, MD, USA) containing 1% BBL Isovitalex Enrichment (Becton, Dickinson), and then harvested and suspended in calcium-and magnesium-free phosphate-buffered saline (PBS).Absorb ance (Abs) at 595 nm was used to serially dilute bacterial suspensions with RPMI 1640 media (Gibco, ThermoFisher Scientific, Pittsburgh, PA, USA) containing 10% fetal bovine serum (FBS) in 96-well plates (CellBIND, Corning, Glendale, AZ, USA) in 190 µL volumes/well.Aliquots (10 µL) of CPPs or DMSO-containing vehicles in RPMI 1640 with 10% FBS were added and plates were incubated at 37°C in 5% CO 2 for 4 hours as in our previous study of the lCPP and N. gonorrhoeae (11).After mixing gently, 10 µL aliquots were plated on agar plates and colonies were counted after overnight incubation.To determine the effect of LOS sialylation, bacteria were grown overnight on agar contain ing 10 µg/mL CMP-NANA (Sigma-Aldrich, St. Louis, MO, USA).

Cytotoxicity of CPPs for human cells
We analyzed the levels of LDH released from human End1/E6E7 endocervical epithelial (79), ME-180 cervical epithelial (80), and THP-1 monocytic cells (81) (all from ATCC) after exposure to the CPPs or LL-37 using the Cyquant LDH Kit (Invitrogen Thermo Fisher).THP-1 cells were activated overnight as previously described (11).The THP-1 and ME-180 cells were cultured in Opti-MEM (Gibco) with 1% FBS.The End1/E6E7 cells were cultured in Keratinocyte-Serum Free Media containing supplements (ThermoFisher).Stock solutions of the CPPs or LL-37 were diluted in PBS and aliquots were added to wells (total 100 µL volumes/well).Negative control wells were treated with DMSO-containing vehicles only.Plates were gently mixed and incubated at 37°C and 5% CO 2 for 18 hours.Positive controls were treated with 10 µL of the Triton X-100 Lysis Buffer.Levels of LDH were quantified on an Epoch plate reader (Biotek Instruments, Winooski, Vermont, USA) with detection at 490-680 nm.Results were reported as a percentage of LDH released from cells treated with CPPs or LL-37 relative to LDH released from cells lysed by Triton X-100.

Hemolysis assay
Washed pooled sheep red blood cells (10% vol/vol, Rockland, Philadelphia, PA, USA) were diluted to 2% in PBS, then added to a 96-well plate (Corning CellBIND) along with 10 µL aliquots of dilute CPP stock solutions as previously described (36).Wells were treated with 0.1% Triton X-100 as a positive control.Plates were incubated at 37°C and 5% CO 2 for 1 hour, centrifuged, and then the supernatants (100 µL) were transferred to a new plate and analyzed on the Epoch microtiter plate reader.Results are expressed as % hemolysis according to the following: % hemolysis = [(405 nm Abs of RBCs in peptide solution − 405 nm Abs of RBCs in PBS)/(405 nm Abs of RBCs in 0.1% Triton X-100 -405 nm Abs of RBCs in PBS)] × 100.Values of <10% were defined as non-hemolytic (35).

CPP treatment of gonococci in the presence of human cells
Assays were performed in 96-well CellBIND plates with activated human THP-1, and adherent End1/E6E7 or ME-180 cells (~30,000 cells/well) in RPMI-1640 media with 10% FBS for the THP-1 and the ME-180 cells, or Keratinocyte Serum Free Media for the End1/ E6E7 cells to form a nearly confluent monolayer.Prior to infection, the THP-1 cells were activated overnight, treated with trypsin, plated, and allowed to adhere according to the protocol we previously used in testing the lCPP with N. gonorrhoeae (11).CPP stock solutions or DMSO-containing vehicles were added concurrently with 4,000 bacteria followed by incubation for 4 hours.Then 100 µL of 3% saponin was added to each well and the plates were incubated at 37°C for 15 minutes prior to plating on GC agar.

Effect of CPP on induction of TNF-α expression from gonococcal infection of THP-1 cells
The induction of TNF-α expression by THP-1 cells infected with N. gonorrhoeae at a moi of 1 was determined as previously described (78).Aliquots of CPP-containing solutions were added to the wells to achieve final concentrations of 4, 20, or 100 µM in total volumes of 200 µL/well.Plates were incubated at 37°C and 5% CO 2 for 18 hours, and then supernatants were transferred to a fresh 96-well plate, sealed, and frozen at −80°C until analysis.A kit (Human TNF-α Uncoated ELISA kit, Invitrogen, ThermoFisher) was used for detection and results were analyzed on the Epoch plate reader at 450 nm.

Assay for dispersion of abiotic N. gonorrhoeae biofilms
Assays of the eradication of N. gonorrhoeae in abiotic biofilms were performed as previously described (82)(83)(84).FA1090 was grown overnight on Difco GC agar supplemen ted with Isovitalex and then suspended in PBS containing 10% GC broth containing Isovitalex.Absorbance at 595 nm was used to dilute bacterial suspensions to 1.12 × 10 8 cells/mL and 100 µL/well was added to 96-well CellBIND plates to determine the effect of the CPPs on a biofilm formed on an abiotic surface.Plates were incubated at 37°C and 5% CO 2 for 40 hours, washed with GC broth, and then lCPP, tcCPP, or vehicle control was added.After incubation with the CPPs for 24 hours, plates were washed twice.Biofilms were dried, stained with 125 µL of 0.1% (wt/vol) crystal violet solution for 15 minutes, and then washed with water.The plates were dried in ambient conditions and then 125 µL of 30% glacial acetic acid was added to each well and the plates were shaken until the biofilm was dissolved.Plates were analyzed on the Epoch microtiter plate reader at 570 nm.

Statistical analyses
SigmaPlot version 12.5 (Systat Software, San Jose, CA) was used for statistical analysis.Multigroup comparisons were performed using a one-way analysis of variance with Newman-Keuls post hoc tests.Significance was defined as P < 0.05 for all comparisons.

FIG 1
FIG 1 Structures of the lCPP (A), which is ArgAhxArgArgAhxArgArgAhxArgArgAhxArg and Ahx is 6-aminohexanoic acid, the dcCPP (B), which is cyclo[Cys(S-)ArgAhxArgArgAhxArgArgAhxArgArgAhxArgCys(S-)] that contains a disulfide bond between the Cys at positions 1 and 14, and tcCPP (C), that corresponds to cyclo[ArgAhxArgArgAhxArgArgAhxArgArgAhxArg)Cys(S-)] and is cyclized by a thioether (sulfide) bond between the Cys at position 13 and an N-terminal acetate group.

FIG 5 FIG 6
FIG 5 Bactericidal activity of increasing concentrations (4, 20, and 100 µM) of lCPP and tcCPP for FA1090 gonococci infecting human End1/E6E7 (A), ME-180 (B), and THP-1 cells (C).The cultures were incubated with RPMI-1640 media with 10% FBS at 37°C for 4 hours.Prior to plating 10 µL aliquots of the cultures for colony counting, 100 µL of 3% saponin was added to each well and the plates were incubated at 37°C for 15 minutes to lyse the human cells.Bars represent the mean ± SEM from three experiments that were each performed in duplicate.**P < 0.01, ***P < 0.001.