Polybiguanides, particularly polyethylene hexamethylene biguanide, have activity against human immunodeficiency virus type 1

Biomedicine & Pharmacotherapy 59 (2005) 438–445 http://france.elsevier.com/direct/BIOPHA/

Dossier: HIV/AIDS: New approaches in chemotherapy and immunotherapy

Polybiguanides, particularly polyethylene hexamethylene biguanide, have activity against human immunodeficiency virus type 1 Fred C. Krebs a, Shendra R. Miller a, Mary Lee Ferguson a, Mohamed Labib b, Robert F. Rando b, Brian Wigdahl a,* a

Department of Microbiology and Immunology, and Center for Sexually Transmitted Disease, Center for Molecular Therapeutics, Institute for Molecular Medicine and Infectious Disease, Drexel University College of Medicine, Philadelphia, Pennsylvania 19129 USA b Novaflux Biosciences, Inc., 1 Wall Street, Princeton, NJ 08540 USA Received 26 July 2005 Available online 18 August 2005

Abstract Polyhexamethylene biguanide (PHMB) is a polybiguanide (PBG) oligomer with antimicrobial activity that is used extensively and safely as a disinfectant. The reported mechanism of PHMB antimicrobial activity, which involves interactions with cell membrane components, suggested that PHMB or other PBG-based compounds might also have antiviral or virucidal activity against the human immunodeficiency virus type 1 (HIV-1). PHMB had modest in vitro activity against both cell-free and cell-associated HIV-1, as well as the ability to interfere with viral binding and entry. However, PHMB was comparable in cytotoxicity to the spermicidal agent nonoxynol-9 (N-9), a compound that has been characterized in previous studies as generally cytotoxic and detrimental to cervicovaginal epithelial integrity. To identify structural variants of PHMB with greater anti-HIV-1 activity and/or less cytotoxicity, modified versions of PHMB incorporating length changes in the hydrocarbon linker units were synthesized and evaluated for in vitro cytotoxicity and inhibition of HIV-1 infectivity. These experiments demonstrated that the PHMB variant polyethylene hexamethylene biguanide (PEHMB) was just as active against HIV-1 as PHMB, yet was much less cytotoxic than either N-9 or PHMB, resulting in an in vitro therapeutic index (TI) approximately 114-fold greater than the TI of N-9. PEHMB, which has been identified in these studies as a promising microbicidal candidate in this family of compounds, will be the focus of further in vitro and in vivo evaluations of anti-HIV-1 activity, toxicity, and mechanisms of action. © 2005 Elsevier SAS. All rights reserved. Keywords: HIV-1; Polybiguanide; Microbicide

1. Introduction Increasing numbers of women worldwide are being affected by the acquired immune deficiency syndrome (AIDS) epidemic. Globally, just under half of all people infected with the human immunodeficiency virus type 1 (HIV-1) are women [12]. In Sub-Saharan Africa, infection rates of women and girls are particularly high; almost 57% of the total HIV-1infected population is female. Because more effective means to reduce or eliminate sexual transmission of HIV-1 to women are clearly necessary, increasing efforts have been concen* Corresponding author. Mailing address: Drexel University College of Medicine, Department of Microbiology and Immunology (G44), 2900 Queen Lane, Philadelphia, Pennsylvania 19129,Tel.: (215) 991-8352, fax: (215) 848-2271. E-mail address: [email protected] (B. Wigdahl). 0753-3322/$ - see front matter © 2005 Elsevier SAS. All rights reserved. doi:10.1016/j.biopha.2005.07.007

trated on the development of microbicidal compounds with activity against HIV-1, as well as other sexually transmitted disease (STD) pathogens. Compounds with the potential for use as topical microbicides effective against HIV-1 transmission should be (i) highly effective against HIV-1 transmission during sexual intercourse, (ii) safe for repetitive in vivo use, (iii) inexpensive to manufacture and formulate, (iv) stable during long-term storage under a wide range of environmental conditions, and (v) female-controlled as necessary [4]. Some of the initial efforts focused on identifying potential microbicidal products were concentrated on the evaluation of nonoxynol-9 (N-9), which was already in public use as a contraceptive agent. Although N-9 does have in vitro activity against HIV-1 [10,17], it also has considerable in vitro cytotoxicity [9,15,16] and has been shown in animal models of toxicity [3,25] and human clini-

F.C. Krebs et al. / Biomedicine & Pharmacotherapy 59 (2005) 438–445

cal trials [31,36] to cause considerable epithelial damage and inflammation when applied to the female reproductive tract. Furthermore, clinical studies have demonstrated that vaginal application of N-9-containing products appears to have little or no impact on the risk of HIV-1 transmission [18,27,36] and may, during highly repetitive use, actually increase the risk of infection [35]. In light of the apparent failure of N-9 to meet the requirements for a microbicidal agent, a multitude of microbicides with a range of activities and mechanisms of action are now under pre-clinical and clinical development [4,29,33]. Our efforts have focused on the development of polybiguanide (PBG)-based compounds as microbicidal agents (Fig. 1). PBGs are polycationic compounds comprised of biguanide repeat units separated by hydrocarbon chain spacers of variable length. PBGs are normally present as chloride or other salts at physiologic pH (Fig. 1 A). PBGs represent a unique class of compounds with a convergence of safety, broad-spectrum activity, and structural diversity. Chlorhexidine digluconate (CHG), a bis-biguanide (two biguanide groups), has been used as a safe, general vaginal disinfectant for over thirty years [30,32] and has more recently been evaluated as a microbicidal agent effective against chlamydial infection [26]. Biguanide-based drugs, which have also been used safely and successfully for decades, include Porguanil, which is used as an anti-malarial agent [24], and Metformin, which is used to treat type 2 diabetes [14]. The PBG compound polyhexamethylene biguanide (PHMB, Fig. 1B) is used as the broad-spectrum active ingredient in anti-bacterial contact lens solutions [13], as a treatment for Acanthamoeba keratitis [19], as a potential antiseptic mouthrinse [28], as a topical wound disinfectant [20,21], and as an environmental biocide [37]. PBGs such as PHMB have many characteristics that are consistent with the attributes of an ideal microbicide

Fig. 1. Polybiguanides (PGBs) are polymeric, polycationic molecules. (A) A prototypical polybiguanide is distinguished by cationic biguanide repeat units separated by hydrocarbon chain linkers of identical or dissimilar lengths. The molecule is accompanied by anions and can be capped on one or both ends with additional functional groups. (B) Polybiguanides used in these studies included polyhexamethylene biguanide (PHMB), polyethylene biguanide (PEB), polytetramethylene biguanide (PTMB), and polyethylene hexamethylene biguanide (PEHMB). PBG molecules used in these studies were accompanied in solution by chloride anions.

439

[4], including (i) ease of synthesis and preparation, (ii) the absence of odor, and (iii) chemical stability. The antibacterial activity of PHMB is attributed to interactions with cellular membrane components, specifically anionic phospholipids [2,11] and perhaps proteins and lipopolysaccharides. The ability to interact with membranes suggests that PHMB may also be effective against HIV-1, since infection of cells susceptible to HIV-1 is mediated by components in both the cellular membrane and viral envelope. Indeed, PHMB was demonstrated to be effective in vitro against herpes simplex virus type 1 (HSV-1) at concentrations as low as 0.01% [34]. The following studies, which were conducted to assess the in vitro impact of PHMB and several related compounds on HIV-1 infectivity and cell viability, explore the hypothesis that PBG-based compounds are active against HIV-1. These studies represent the first steps in the process of PBG microbicide discovery and development, and the identification of PBG-based compounds to be used as safe and effective microbicides.

2. Materials and Methods 2.1. Compound synthesis PHMB is commercially available (Avicia, Wilmington, DE) and consists of biguanide repeat units flanked by hexamethylene spacers (designated 6-6; Fig. 1B). PHMB can be synthesized using a two-step process. In the first step, hexamethylenediamine is reacted with sodium dicyandiamide. The product of this reaction, hexamethylene bis-cyanoguanidine, is subsequently reacted neat with hexamethylenediamine hydrochloride at 160-185 °C for 4h under constant stirring to form oligomeric molecules of PHMB. PHMB synthesized using this protocol contains an average of eight biguanide subunits per molecule [6,38]. PHMB produced by this method is pure (>99% as determined by nitrogen analysis) and readily soluble in water. Backbone modifications of PHMB (Fig. 1B) were synthesized at Novaflux Biosciences, Inc., by varying the hydrocarbon chain lengths within the bis-cyanoguanidine and/or diamine co-monomers used in the synthesis reactions. Polyethylene biguanide (PEB, 2-2), polytetramethylene biguanide (PTMB, 4-4), and polyethylene hexamethylene biguanide (PEHMB, 2-6) were synthesized by substituting monomers with the appropriate chain lengths into the synthesis strategy described above. The numbers that accompany each compound name indicate the number of methylene groups in the linkers that flank the biguanide moieties. 2.2. Tissue culture HeLa cells (ATCC designation CCL-2) were maintained in Dulbecco’s modified Eagle’s media (DMEM). P4-CCR5 cells (AIDS Reagent Program #3580) were cultured in DMEM supplemented with puromycin at a concentration of

440

F.C. Krebs et al. / Biomedicine & Pharmacotherapy 59 (2005) 438–445

1 µg/ml [5]. Sup-T1 human T lymphocytes (ATCC designation CRL-1942) were cultured in RPMI 1640. All three cell types were cultured in media supplemented with 10% fetal bovine serum (FBS), L-glutamine (0.3 mg/ml), antibiotics (penicillin, streptomycin, and kanamycin at 0.04 mg/ml each), and 0.05% sodium bicarbonate. Cells of the Vk2/E6E7 human vaginal keratinocyte cell line (kindly provided by Dr. Raina Fichorova, Harvard Medical School, Boston, MA) were cultured as described [8,15].

4 x 104 cells per well in 96-well culture plates 24 h before the assay. Compounds were diluted in media to achieve the final exposure concentrations. At the conclusion of the exposure period, cells were washed, supplied with new media, and assessed for viability using an MTT assay as previously described [17].

2.3. Inactivation of cell-free HIV-1 (CFI)

3.1. PHMB has activity against cell-free HIV-1

P4-CCR5 cells were plated at a density of 8 x 104 cells per well in 12-well culture plates 24 h before each assay. Concentrated, cell-free preparations (~108 TCID50/ml, Advanced Biotechnologies, Inc., Columbia, MD) of HIV-1 strains IIIB (X4 phenotype) or BaL (R5 phenotype) were mixed in a 1:1 ratio with dilutions of each compound (10 µL total volume) to achieve the final concentrations indicated for each experiment. After incubation for 10 min at 37 °C, the mixtures were diluted 1:100 in RPMI 1640 (with 10% FBS) and added (300 µL per well to triplicate wells) to P4-CCR5 cells. Following a 2-h adsorption period, new media (2 ml) was added to each well. After incubation for 48 h at 37 °C, whole cell lysates were prepared and assayed for b-galactosidase activity using the Galacto-Star b-Galactosidase Reporter Gene Assay System for Mammalian Cells (Applied Biosystems, Bedford, MA). 2.4. Inactivation of cell-associated HIV-1 (CAI) Sup-T1 cells were infected with HIV-1 IIIB (3 µL virus per 3 x 106 cells in 30 mL media) 72 h prior to the assay. HIV-1-infected cells were subsequently pelleted and resuspended in RPMI 1640 at a concentration of 1 x 106 cells per 95 µL and mixed with 5 µL of compound to achieve the desired final concentrations. After a 10-min incubation at 37 °C, the mixture was diluted 1:10 and added to P4CCR5 cells (300 µL per well in triplicate wells). Following a 2-h adsorption period, the cells were washed once with PBS and supplied with new media (2 ml). After a 48-h incubation, the cells were harvested and assayed as described above. 2.5. Inhibition of HIV-1 binding and entry (VBI) Cell-free HIV-1 strains IIIB or BaL were incubated with each compound in triplicate at the desired concentration (300 µL total volume) in the presence of P4-CCR5 cells for 2 h at 37 °C. Following the incubation period, the cells were washed once with PBS and supplied with new media (2 mL). After a 48-h incubation, the cells were harvested and assayed as described above. 2.6. In vitro cytotoxicity Assessments of in vitro cytotoxicity were performed using HeLa, P4-CCR5, or Vk2/E6E7 cells. Cells were seeded at

3. Results

Experiments were first conducted to assess the activity of the commercially available PBG PHMB against cell-free strains of HIV-1. In assays using strain IIIB (X4 phenotype), exposure to PHMB resulted in concentration-dependent inactivation of infectivity, achieving complete inactivation at 0.32% (w/v) and 50% inactivation (IC50; the concentration at which infectivity was reduced to 50% of the control value) at approximately 0.18% (Fig. 2 A). N-9, which was used as a comparative standard in these and subsequent assays, had greater activity than PHMB in this assay (0.02% IC50), resulting in complete inactivation at concentrations greater than 0.03%. Similarly, both PHMB and N-9 were evaluated for activity (Fig. 2B) against cell-free preparations of HIV1 strain BaL (R5 phenotype). In these experiments, the activities of N-9 (0.02% IC50) and PHMB (0.03% IC50) were more comparable, since PHMB was more active against HIV1 BaL than against HIV-1 IIIB, while N-9 had similar activity against both strains (presumably due to its non-specific, surfactant-based virucidal activity). 3.2. PHMB has activity against cell-associated HIV-1 comparable to that of N-9 Potential microbicidal compounds must also be assessed for inactivation of cell-associated forms of infectivity, since both cell-free and cell-associated forms of HIV-1 may mediate sexual transmission. In these assays, HIV-1 IIIB-infected cells of the human Sup-T1 T lymphocyte cell line were used as the source of cell-associated virus (Fig. 2C). Similar levels of inactivation of cell-associated infectivity were achieved by 10-min exposure to PHMB (0.01% IC50) or N-9 (0.02% IC50). 3.3. HIV-1 binding and entry is inhibited by PHMB Because assays of both cell-free and cell-associated viral inactivation include dilution of the virus and compounds after the inactivation period, the compound may not be present at sufficient concentrations to interfere with viral binding and entry events that take place at the surface of the cell used as an indicator of infection. To assess the ability of these compounds to inhibit HIV-1 binding and/or entry, assays were performed in which each compound was present during the 2-h incubation of virus with the target/indicator P4-

F.C. Krebs et al. / Biomedicine & Pharmacotherapy 59 (2005) 438–445

441

Fig. 2. PHMB has activity against HIV-1. N-9 and PHMB were assessed for cell-free inactivation (CFI) of HIV-1 strains (A) IIIB or (B) BaL, for inactivation of cell-associated infectivity (CAI) from HIV-1 IIIB-infected Sup-T1 T cells (C), and for (D) interference with viral binding and entry (VBI-IIIB). Each panel illustrates representative results from one of two independent assays in which each concentration was tested in duplicate. Assays were performed as described in Materials and Methods. [Y-axis: infectivity remaining (%)]

CCR5 cells (Fig. 2D). Dextran sulfate (Dextralip 50, Sigma), which has been shown to effectively inhibit HIV-1 infection by blocking viral adsorption to the host cell [1], was used as a comparative control and was highly active in these experiments (43% infectivity remaining at 0.0001%). PHMB and N-9 appeared to have similar capacities for interfering with viral binding and entry. However, qualitative observations of cell viability during these experiments indicated that exposure of the target cells to either N-9 or PHMB resulted in concentration-dependent cytotoxicity above 0.001%, suggesting that the observed anti-HIV-1 activities were, at least in part, due to reductions in target cell viability. To further explore this hypothesis, both compounds were assessed for in vitro cytotoxicity.

ity (Fig. 3B). After 24 h (Fig. 3C), cells were again slightly more sensitive to PHMB (0.001% CC50) than N-9 (0.003% CC50). Similar experiments were conducted using Vk2/E6E7 human vaginal epithelial cells, which are cells derived from human vaginal keratinocytes transformed with human papillomavirus 16/E6E7 [8]. After a 10-min exposure, cell viability was similarly reduced by N-9 and PHMB (Fig. 3D). Following 2-h (Fig. 3E) or 24-h exposures (Fig. 3F) to each compound, Vk2/E6E7 cell viability was reduced slightly more by PHMB compared to N-9.

3.4. PHMB is comparable in cytotoxicity to N-9

One of the inherent strengths of the PBG family of compounds is the flexibility of molecular design. A large number of related but distinct PBG-based compounds can be synthesized using one or more of the following strategies: (i) altering the backbone structure, which results in changes in molecular weight, backbone flexibility, hydrophobicity, and charge density; (ii) adding functional groups to one or both ends of the molecule (end caps) to provide additional activity; (iii) changing the molecular weight (i.e. the number of biguanide repeats); and (iv) changing the identity of the accompanying anion to alter the electrostatic and/or hydrophobic nature of the PBG-anion complex. Previous studies indicated that alkyl chains of varied lengths attached to single biguanide units caused notable differences in both cytotoxicity and activity against vaccinia and influenza viruses [7]. In

To estimate the relative cytotoxicity of PHMB and N-9, and to gauge the impact of cytotoxicity on the preceding assays, in vitro cytotoxicity was measured using cells of both human cervical (HeLa cells) and vaginal (Vk2/E6E7) origin. Assays using HeLa cells, which we have shown in previous experiments [17] to be similar to P4-CCR5 cells in sensitivity to microbicidal agents, were performed to measure cytotoxicity after exposure durations of 10 min, 2 h or 24 h. After a 10-min exposure (Fig. 3 A), HeLa cell viability was reduced to a greater extent by PHMB (0.006% CC50; the concentration at which cell viability was reduced to 50% of the control value) compared to N-9 (0.054% CC50). After a 2-h exposure, PHMB and N-9 caused similar reductions in cell viabil-

3.5. PBG backbone modifications alter both in vitro cytotoxicity and anti-HIV-1 activity

442

F.C. Krebs et al. / Biomedicine & Pharmacotherapy 59 (2005) 438–445

Fig. 3. Human cervicovaginal cell lines are similarly sensitive to PHMB and N-9 exposure in vitro. HeLa (human cervical) and VK2/E6E7 (human vaginal) cells were exposed to N-9 or PHMB for (A, D) 10 min, (B, E) 2 h, or (C, F) 24 h and subsequently assayed for viability as described in Materials and Methods. Representative results from one of two independent cytotoxicity (CTS) assays, in which each concentration was tested in duplicate, are depicted. [Y-axis: viability index]

this regard, several PBG compounds similar to PHMB were synthesized with hydrocarbon chain linkers that varied in length from two to six methylene groups (Fig. 1B). These

variant PBG compounds were compared in assays of both cytotoxicity and antiviral activity (Fig. 4). In 2-h cytotoxicity assays (Fig. 4 A), N-9 and PHMB (6-6 backbone configura-

Fig. 4. PEHMB is characterized by a combination of low in vitro cytotoxicity and considerable inhibition of HIV-1 binding and entry. Compounds used in these experiments included N-9, dextran sulfate (Dextralip 50, Sigma), PHMB (6-6 backbone configuration), PEB (2-2), PTMB (4-4), and PEHMB (2-6). (A) P4-CCR5 cells were exposed to each compound for 2 h and subsequently assayed for viability as described. (B) P4-CCR5 cells were incubated with cell-free HIV-1 strain IIIB in the presence of each compound for 2 h at 37 °C and subsequently assayed for infection as described in Materials and Methods. Representative results from one of two independent assays, in which each concentration was tested in duplicate, are depicted in each panel. [Y-axis: A viability index; B: infectivity remaining (%)]

F.C. Krebs et al. / Biomedicine & Pharmacotherapy 59 (2005) 438–445

tion) had similar effects on cell viability (0.007% and 0.009% CC50, respectively), while polytetramethylene biguanide (PTMB, 4-4 backbone) was less cytotoxic (0.055% CC50). Polyethylene hexamethylene biguanide (PEHMB, 2-6 backbone) had even less effect on cell viability (0.889% CC50) and was approximately 100-fold less cytotoxic than N-9 and PHMB. Polyethylene biguanide (PEB, 2-2 backbone) was the least cytotoxic of the PBG compounds examined (~2.2% CC50, estimated by extrapolation). In assays of viral binding and entry inhibition, there was a similar divergence of activity as a consequence of backbone modifications (Fig. 4B). The presence of N-9 (0.007% IC50), PHMB (0.001% IC50), or PTMB (0.003% IC50) resulted in similar levels of viral binding and entry inhibition, while PEB was considerably less effective (0.160% IC50). PEHMB also had considerable activity against HIV-1 (0.007% IC50) comparable to N-9. By comparison, dextran sulfate, a prototypical, polyanionic inhibitor of HIV-1 binding, completely eliminated infectivity at 0.001%. 3.6. PBG backbone modifications alter compound effıcacy The measure of microbicide efficacy is its ability to reduce or eliminate the risk of HIV-1 transmission compared to its effect on the viability and integrity of the vaginal and cervical epithelial tissues. Because there are many factors that can influence in vivo anti-HIV-1 activity and toxicity (i.e. changes in pH, the three-dimensional structure of the vaginal and cervical epithelium, the presence of cervical fluid, mucus, and semen), in vitro assays that do not account for these variables cannot provide an absolute measure of compound efficacy. However, they may provide indications of relative efficacy across assays and within families of compounds. In vitro therapeutic indices (TIs) calculated using the results illustrated in Fig. 4 clearly demonstrate the range of efficacies produced by the modifications made to the PBG backbone (Table 1). In these experiments, the TI of N-9 was approximately 1. In contrast, the TI of PHMB was 9.5, reflecting its greater ability to inhibit viral binding and entry relative to N-9. Using the extrapolated CC50 value, the TI of PEB was estimated to be approximately 14. PTMB, which had activity comparable to PHMB but less cytotoxicity, had a TI of 17.5. Finally, PEHMB, which was characterized by a combination of anti-HIV-1 activity and low cytotoxicity, had a TI of 125.6, suggesting that this modified variant of PHMB should be considered for further study as a promising candidate microbicide. Table 1 PEHMB has a much greater in vitro therapeutic index than N-9, PHMB or other PBG backbone variants Compound N-9 PHMB PEB PTMB PEHMB a

Backbone 6-6 2-2 4-4 2-6

CC50 (%) 0.007 0.009 ~2.2 0.055 0.889

IC50 (%) 0.007 0.001 0.160 0.003 0.007

TI Valuea 1.1 9.5 ~13.7 17.5 125.6

TIs were calculated by dividing the CC50 value (Fig. 4 A) by the IC50 concentration (Fig. 4B).

443

4. Discussion Using a series of in vitro efficacy and toxicity experiments, we have explored the anti-HIV-1 activity and cytotoxicity of a series of PBG-based compounds differing with respect to the lengths of their methylene linkages. These experiments represent the first steps in ongoing efforts to design and characterize PBG-based compounds with high activity against HIV-1 and minimal cytotoxicity. The results presented in this manuscript illustrate the potential of PBGbased compounds to form the basis of topical vaginal microbicide formulations. In addition, these results preliminarily suggest several mechanisms of action for PGB-based compounds and provide insights into strategies that can be used to further optimize their efficacy. While assays of cell-free and cell-associated HIV-1 inactivation (Fig. 2) indicated that both PHMB and N-9 had antiviral or virucidal activity, assays such as these may be affected by the inherent cytotoxicity of compounds under examination if those compounds are both cytotoxic and active against HIV-1 over similar concentration ranges. However, because both N-9 and PHMB were diluted (1:100 in the CFI assay, Figs. 2 A and B; 1:10 in the CAI assay, Fig. 2C) prior to the 2-h incubation with the target/indicator cells, the effects of compound cytotoxicity during the 2-h exposure (Fig. 3B) were restricted to concentrations at or above 0.1% and 0.01% in the CFI and CAI assays, respectively, which allowed for a clear interpretation of the efficacy data at or below these concentrations. PHMB anti-HIV-1 activity was still evident in both the CFI and CAI assays. PHMB was particularly active against HIV-1 BaL relative to its activity against HIV-1 IIIB, suggesting that PHMB may be acting specifically against cellfree virus rather than through a non-specific disruption of viral integrity. Specific activity against cell-free virus may be attributable to interactions between PHMB and components of the virion membrane. These interactions may cause decreases in infectivity by binding to lipids or proteins within the viral envelope, or by disrupting the integrity of the viral particle, perhaps involving the same mechanisms that result in cytotoxicity at higher concentrations. Because PHMB has been shown to interact with membrane-bound anionic phospholipids [11], strain-dependent activity of PHMB might be a consequence of differences in lipid composition (or possibly protein content) and membrane organization between the BaL and IIIB viral envelopes, which are acquired from the macrophages and T cells used to produce the respective strains. Alternatively, these results may be a consequence of differences in the amino acid sequences between the respective gp120 molcules. Modeling of the electrostatic potentials of the conserved coreceptor binding surfaces and the V3 loop of gp120 indicated that the primary amino acid sequences of two X4 HIV-1 strains (HXBc2 and MN) and a single X4R5 virus (89.6) render the V3 loops highly basic (highly cationic) [22]. In contrast, the V3 loop of an R5 virus (JRFL) is only weakly basic (less cationic) [22]. The lower cationic

444

F.C. Krebs et al. / Biomedicine & Pharmacotherapy 59 (2005) 438–445

surface charge of an R5 gp120 molcule may permit higher affinity interactions with cationic PBGs like PHMB, resulting in increased activity against cell-free R5 virus relative to activity against an X4 virus. This hypothesis is consistent with the observation that PHMB had increased activity against HIV-1 BaL (R5) relative to strain IIIB (X4) in the cell-free inactivation assays. Although analyses of activity in the VBI assays must also be tempered by the observation that PHMB and N-9 are both active against HIV-1 infection (Fig. 2D) and cytotoxic (Fig. 3B) over the same concentration ranges, our results indicated that PHMB had greater activity in assays of viral binding and entry inhibition (65% infectivity remaining at 0.001%) compared to assays of cell-free inactivation (78% infectivity remaining at 0.1% and ineffective below 0.0316%). For this reason, subsequent examinations of PBG compounds synthesized with length variations in the hydrocarbon chains flanking the PBG repeat units were conducted using assays of viral binding and entry inhibition (Fig. 4B). These experiments paralleled previous studies that demonstrated changes in cytotoxicity and antiviral activity associated with length variations in alkyl chains attached to monomeric biguanide molecules [7]. The results reported herein indicated that changes in linker length, which alter charge density, hydrophobicity, and polymer length, had a substantial impact on compound cytotoxicity and a smaller effect on activity against HIV-1. The balance between hydrophobicity, charge density, and chain flexibility struck by PEHMB resulted in a molecule with a considerably better therapeutic index compared to the starting compound PHMB. The greatly reduced cytotoxicity of PEHMB permitted the demonstration of anti-HIV1 activity (Fig. 4B) in a non-cytotoxic concentration range (Fig. 4 A) and increased the in vitro therapeutic index of PEHMB by 114-fold relative to the TI of N-9 in the same assays (Table 1). Additionally, results of mouse cervicovaginal toxicity studies using PEHMB [3] strongly suggest that the in vivo TIs of PBGs such as PEHMB are much higher than the in vitro values and also much higher than the in vivo TI of N-9. Experimental results to date demonstrated that PHMB (and PEHMB, data not shown) had greater activity in assays that examined inhibition of viral binding and entry compared to experiments that assessed inactivation of cell-associated and cell-free virus. Although studies concerning the mechanism of action of PBG-based compounds against HIV-1 have not yet been completed, these results, as well as studies of biguanide-based compounds as bacteriocides, suggest that PBGs may act through at least two mechanisms. As was discussed above, the results of the cell-free virus inactivation experiments imply that direct interactions between PHMB and the cell-free virions could be responsible for the observed decreases in infectivity. However, the greater activities of these compounds in assays of cell-associated inactivation and inhibition of viral binding and entry suggest that PBG-based compounds also act through interactions with membrane phospholipids or membrane-associated components (i.e. proteins

or glycoproteins) found on the cell, but not on the virus. Preliminary analyses of the effect of PBG compounds on HIV1 cell surface receptors and coreceptors have indicated that PEHMB treatment temporarily decreased the ability to detect specific epitopes of CXCR4 but not CD4 (F. C. Krebs and B. Wigdahl, unpublished observations). These results indicate that PBG-based compounds may interfere with viral binding or entry events by blocking gp120-coreceptor interactions, perturbing the spatial relationships between CD4 and CXCR4 or CCR5 necessary for viral binding and entry, altering the conformations of the coreceptors, or initiating the removal of the coreceptor from the cell surface. Such effects might be achieved by changes at the cell surface initiated by direct interactions between the PBG molecules and the cellular coreceptors CXCR4 or CCR5 or by indirect interactions with anionic phospholipids in the cell membrane. PBG compounds may also interfere with other membrane-proximal activities, as suggested by studies in which phagocytosis of latex beads or heat-killed yeast coated by PHMB was inhibited [23]. Finally, changes in cytotoxicity and anti-HIV1 activity that resulted from alterations in PBG structure support the hypothesis that specific interactions between PBG molecules and membrane-bound components of the cell membrane (and, perhaps, the virion envelope) form the basis for the activities of these molecules. More extensive studies are now in progress to more precisely define the anti-HIV1 mechanisms of PBG-based compounds. By taking advantage of the design flexibility of the PBG family of compounds, we have advanced our search for PBGbased microbicides from a compound with activity against HIV-1 (PHMB) combined with considerable cytotoxicity to a greatly improved compound (PEHMB) with much lower cytotoxicity in combination with considerable anti-HIV1 activity. PEHMB is an excellent example of the potential of PBG-based compounds and serves as a springboard for further investigations of PBG-based compounds with even greater microbicidal efficacy. This compound, which holds promise as a microbicidal candidate, is now being examined in more extensive in vitro and in vivo studies to better characterize its anti-HIV-1 activity, cytotoxicity, and mechanism of action.

Acknowledgements These studies were supported by Public Health Service grants P01 AI37829 (M. K. Howett, Program Director; B. Wigdahl, Project 3C Director) and U19 HD48958 (M. Labib, Principal Investigator and B. Wigdahl, Co-Principal Investigator; M. Labib, Project I Director; F. Krebs, Project II Director; B. Wigdahl, Project III Director). We thank Lori Schlipf and Nina Thakkar for their critical reviews of the manuscript. We wish to thank Dr. Richard F. Stockel for his help in synthesizing the PHMB variants.

F.C. Krebs et al. / Biomedicine & Pharmacotherapy 59 (2005) 438–445

References [1]

[2]

[3]

[4]

[5]

[6]

[7]

[8]

[9]

[10] [11]

[12]

[13]

[14] [15]

[16]

[17]

[18]

Baba M, Pauwels R, Balzarini J, Arnout J, Desmyter J, De Clercq E. Mechanism of inhibitory effect of dextran sulfate and heparin on replication of human immunodeficiency virus in vitro. Proc Natl Acad Sci USA 1988;85:6132–6. Broxton P, Woodcock PM, Heatley F, Gilbert P. Interaction of some polyhexamethylene biguanides and membrane phospholipids in Escherichia coli. J Appl Bacteriol 1984;57:115–24. Catalone BJ, Kish-Catalone TM, Budgeon LR, Neely EB, Ferguson M, Krebs FC, et al. Mouse model of cervicovaginal toxicity and inflammation for preclinical evaluation of topical vaginal microbicides. Antimicrob Agents Chemother 2004;48:1837–47. Catalone BJ, Krebs FC, Labib M, Wigdahl B. Broad-spectrum microbicides as a strategy for reducing sexually transmitted disease prevalence and HIV-1 transmission. Res Adv in Antimicrob Agents Chemother 2001;2:55–80. Charneau P, Mirambeau G, Roux P, Paulous S, Buc H, Clavel F. HIV-1 reverse transcription. A termination step at the center of the genome. J Mol Biol 1994;241:651–62. East GC, McIntyre JE, Shao J. Polybiguanides: Synthesis and characterization of polybiguanides containing hexamethylene groups. Polym 1997;38:3973–84. Fara GM, Lugaro G, Galli MG, Giannattasio G. Antiviral activity of selected biguanide derivatives. Pharmacol Res Commun 1974;6:117– 26. Fichorova RN, Rheinwald JG, Anderson DJ. Generation of papillomavirus-immortalized cell lines from normal human ectocervical, endocervical, and vaginal epithelium that maintain expression of tissue-specific differentiation proteins. Biol Reprod 1997;57:847– 55. Fichorova RN, Tucker LD, Anderson DJ. The molecular basis of nonoxynol-9-induced vaginal inflammation and its possible relevance to human immunodeficiency virus type 1 transmission. J Infect Dis 2001;184:418–28. Harrison C, Chantler E. The effect of nonoxynol-9 and chlorhexidine on HIV and sperm in vitro. Int J STD AIDS 1998;9:92–7. Ikeda T, Ledwith A, Bamford CH, Hann RA. Interaction of a polymeric biguanide biocide with phospholipid membranes. Biochim Biophys Acta 1984;769:57–66. Joint United Nations Programme on HIV/AIDS (UNAIDS) / World Health Organization (WHO). AIDS epidemic update - December 2004. Geneva: Switzerland; 2004. Keeven J, Wrobel S, Portoles M, DeCicco BT. Evaluating the preservative effectiveness of RGP lens care solutions. CLAO J 1995;21: 238–41. Kirpichnikov D, McFarlane SI, Sowers JR. Metformin: an update. Ann Intern Med 2002;137:25–33. Krebs FC, Miller SR, Catalone BJ, Fichorova R, Anderson D, Malamud D, et al. Comparative in vitro sensitivities of human immune cell lines, vaginal and cervical epithelial cell lines, and primary cells to candidate microbicides nonoxynol 9, C31G, and sodium dodecyl sulfate. Antimicrob Agents Chemother 2002;46:2292–8. Krebs FC, Miller SR, Catalone BJ, Welsh PA, Malamud D, Howett MK, et al. Sodium dodecyl sulfate and C31G as microbicidal alternatives to nonoxynol 9: comparative sensitivity of primary human vaginal keratinocytes. Antimicrob Agents Chemother 2000;44:1954– 60. Krebs FC, Miller SR, Malamud D, Howett MK, Wigdahl B. Inactivation of human immunodeficiency virus type 1 by nonoxynol-9, C31G, or an alkyl sulfate, sodium dodecyl sulfate. Antiviral Res 1999;43: 157–73. Kreiss J, Ngugi E, Holmes K, Ndinya-Achola J, Waiyaki P, Roberts PL, et al. Efficacy of nonoxynol 9 contraceptive sponge use in preventing heterosexual acquisition of HIV in Nairobi prostitutes. JAMA 1992;268:477–82.

445

[19] Larkin DF, Kilvington S, Dart JK. Treatment of Acanthamoeba keratitis with polyhexamethylene biguanide. Ophthalmology 1992;99:185– 91. [20] Lee WR, Tobias KM, Bemis DA, Rohrbach BW. In vitro efficacy of a polyhexamethylene biguanide-impregnated gauze dressing against bacteria found in veterinary patients. Vet Surg 2004;33:404–11. [21] Motta GJ, Milne CT, Corbett LQ. Impact of antimicrobial gauze on bacterial colonies in wounds that require packing. Ostomy Wound Manage 2004;50:48–62. [22] Moulard M, Lortat-Jacob H, Mondor I, Roca G, Wyatt R, Sodroski J, et al. Selective interactions of polyanions with basic surfaces on human immunodeficiency virus type 1 gp120. J Virol 2000;74:1948– 60. [23] Noble JA, Ahearn DG, Avery SV, Crow Jr. SA. Phagocytosis Affects Biguanide Sensitivity of Acanthamoeba spp. Antimicrob Agents Chemother 2002;46:2069–76. [24] Overbosch D, Schilthuis H, Bienzle U, Behrens RH, Kain KC, Clarke PD, et al. Atovaquone-proguanil versus mefloquine for malaria prophylaxis in nonimmune travelers: results from a randomized, double-blind study. Clin Infect Dis 2001;33:1015–21. [25] Patton DL, Kidder GG, Sweeney YC, Rabe LK, Hillier SL. Effects of multiple applications of benzalkonium chloride and nonoxynol 9 on the vaginal epithelium in the pigtailed macaque (Macaca nemestrina). Am J Obstet Gynecol 1999;180:1080–7. [26] Patton DL, Sweeney YT, McKay TL, DeMers SM, Clark AM, Rabe LK, et al. 0.25% chlorhexidine gluconate gel. A protective topical microbicide. Sex Transm Dis 1998;25:421–4. [27] Roddy RE, Zekeng L, Ryan KA, Tamoufe U, Weir SS, Wong EL. A controlled trial of nonoxynol 9 film to reduce male-to-female transmission of sexually transmitted diseases. N Engl J Med 1998;339: 504–10. [28] Rosin M, Welk A, Bernhardt O, Ruhnau M, Pitten FA, Kocher T, et al. Effect of a polyhexamethylene biguanide mouthrinse on bacterial counts and plaque. J Clin Periodontol 2001;28:1121–6. [29] Shattock RJ, Moore JP. Inhibiting sexual transmission of HIV-1 infection. Nat Rev Microbiol 2003;1:25–34. [30] Shubair M, Stanek R, White S, Larsen B. Effects of chlorhexidine gluconate douche on normal vaginal flora. Gynecol Obstet Invest 1992;34:229–33. [31] Stafford MK, Ward H, Flanagan A, Rosenstein IJ, Taylor-Robinson D, Smith JR, et al. Safety study of nonoxynol-9 as a vaginal microbicide: evidence of adverse effects. J Acquir Immune Defic Syndr Hum Retrovirol 1998;17:327–31. [32] Stray-Pedersen B, Bergan T, Hafstad A, Normann E, Grogaard J, Vangdal M. Vaginal disinfection with chlorhexidine during childbirth. Int J Antimicrob Agents 1999;12:245–51. [33] Turpin JA. Considerations and development of topical microbicides to inhibit the sexual transmission of HIV. Expert Opin Investig Drugs 2002;11:1077–97. [34] Valluri S, Fleming TP, Laycock KA, Tarle IS, Goldberg MA, GarciaFerrer FJ, et al. In vitro and in vivo effects of polyhexamethylene biguanide against herpes simplex virus infection. Cornea 1997;16: 556–9. [35] Van Damme L, Ramjee G, Alary M, Vuylsteke B, Chandeying V, Rees H, et al. Effectiveness of COL-1492, a nonoxynol-9 vaginal gel, on HIV-1 transmission in female sex workers: a randomised controlled trial. Lancet 2002;360:971–7. [36] Wilkinson D, Tholandi M, Ramjee G, Rutherford GW. Nonoxynol9 spermicide for prevention of vaginally acquired HIV and other sexually transmitted infections: systematic review and meta-analysis of randomised controlled trials including more than 5000 women. Lancet Infect Dis 2002;2:613–7. [37] Woodcock PM. Biguanides as industrial biocides. In: Payne KR, editor. Critical Reports on Applied Chemistry: Industrial Biocides. New York, NY: John Wiley & Sons; 1988. [38] Zhang Y, Jiang J, Chen Y. Synthesis and antimicrobial activity of polymeric guanidine and biguanidine salts. Polym 1999;40:6189–98.