Sublimable C5A Delivery Provides Sustained and Prolonged Anti-HIV Microbicidal Activities

Abstract

We have identified a short amphipathic helical peptide, called C5A, which exhibits potent microbicidal activities in vitro and which offers protection from vaginal HIV transmission in vivo in a humanized mouse model. However, there are many obstacles to overcome before C5A can be considered a true microbicidal candidate. First, it must be stabilized against enzymatic degradation in a continuously warm and moist environment. Second, it must be delivered in a controlled manner to achieve long-term and coitally independent efficacy. We demonstrate in this in vitro study that the combination of two matrices with different subliming properties ((hexamethylcyclotrisiloxane [HMCS] and cyclododecane [CDD]) containing 10% labile C5A yielded the best results in terms of controlled release and preserved anti-HIV activity of the peptide when pre-exposed to cell-free medium or cell culture at body temperature for up to 2 months.

INTRODUCTION

More than 33 million people are living with the human immunodeficiency virus (HIV), and approximately 2.5 million people are newly infected per year. Worldwide, nearly half of all individuals living with HIV are now women, who acquire the virus largely by heterosexual exposure (9, 11, 15, 17). Many women, because of limited economic options and gender inequality, cannot reliably negotiate sexual encounters, leaving them vulnerable to unwanted pregnancy and sexually transmitted infections, including HIV infections. In the absence of an effective vaccine, topical microbicide formulations, which are applied vaginally or rectally, represent an attractive solution to stop HIV transmission.

Importantly, a proof of concept for vaginal microbicides was obtained in 2010, when the phase IIb CAPRISA 004 clinical study showed a statistically significant 39% reduction in HIV incidence after 30 months of tenofovir gel use compared to placebo gel use in 900 HIV-negative, sexually active women between the ages of 18 and 40 in South Africa (1). More recently, the phase IIb VOICE (Vaginal and Oral Interventions to Control the Epidemic) clinical study has been investigating the use of both an oral pre-exposure prophylaxis (PrEP) and a vaginal microbicide gel to prevent the transmission of HIV (13). The trial, involving 5,029 women in Africa, was composed of five arms with approximately 1,000 women in each. The women were randomized to receive daily (i) tenofovir pills, (ii) tenofovir-emtricitabine pills, (iii) placebo pills, (iv) vaginal tenofovir gel, or (v) vaginal placebo gel. The final clinical data are expected in 2013. However, three arms of the study have been halted due to futility (14). The tenofovir-emtricitabine and placebo pills are the only arms of the study still ongoing. The VOICE results were unexpected and caused the field to re-evaluate the microbicide paradigm (5). Therefore, there is an urgent need to design new microbicide approaches using either antivirals with new mechanisms of action or new drug delivery systems.

We previously identified a short amphipathic helical peptide (SWLRDIWDWICEVLSDFK), called C5A (3, 6), which represents a novel class of microbicidal candidates. C5A neutralizes HIV in the nanomolar to micromolar range without cytotoxicity to human cells (3). C5A corresponds to a small (18-amino-acid [aa]) N-terminal alpha-helical region (aa 3 to 20) (Fig. 1) of the hepatitis C virus (HCV) nonstructural protein 5A (NS5A) (477 amino acids). The sequence of C5A encompasses the region responsible for the anchoring of NS5A to the endoplasmic reticulum (ER) membrane by an in-plane amphipathic alpha-helix (4, 16). In contrast to C5A (18 aa), full-length HCV NS5A (477 aa) does not inhibit HIV infection (3). We obtained several lines of evidence that C5A disrupts the HIV membrane but preserves the integrity of the plasma membrane of cells even when used at high doses (3). We showed that the amphipathic property of C5A is vital for its antiviral action (3, 6). Indeed, disturbing the amphipathicity of parental C5A (SWLRDIWDWICEVLSDFK) by displacing residues in the helical structure of C5A (SWRLDIWDWICESVLDFK) eliminates its anti-HIV activity (3). More recently, we demonstrated that unformulated C5A offers complete protection against vaginal HIV transmission in vivo in a humanized bone marrow-liver-thymus (BLT) mouse model (7). Furthermore, we showed that C5A efficiently inhibited both herpes simplex virus 1 (HSV-1) and HSV-2 infection in epithelial cells in vitro as well as in an ex vivo epidermal infection model (8). These data show that C5A is a multipurpose microbicide candidate which blocks both HIV-1 and HSV (3, 8), representing a prototype of a new generation of microbicidal agents that may have promise for HIV prevention. However, there are many obstacles to overcome before the short peptide C5A can be considered a true microbicidal candidate. It must be stabilized against enzymatic degradation in a continuously warm and moist environment such as the genital tract. C5A must also be delivered in a controlled manner to achieve the long-term and coitally independent efficacy necessary for prophylaxis under real-world conditions. To address these critical delivery issues, we utilized a new subliming solid delivery system.

Fig 1.

Amino acid sequence and three-dimensional structure of the short helical peptide C5A (amino acids 3 to 20) deduced from the experimental NMR structure of NS5A N-terminal membrane anchor in dodecylphosphocholine detergent used as a membrane mimetic (PDB entry: 1R7G) (17). Residues are color coded according to their physicochemical properties. Hydrophobic and polar residues are gray and yellow, respectively. Positively and negatively charged groups of basic (Lys and Arg) and acidic (Asp and Glu) residues are blue and red, respectively. The SH group of Cys is green. (Bottom) Ribbon-and-stick model of the C5A peptide. The image was generated from structure coordinates using VMD (http://www.ks.uiuc.edu/Research/vmd/).

Subliming matrices can provide a broad range of release rates and durations independent of the presence of drug and independent of the environment (other than temperature) in which the device sublimes (S. Gunaseelan et al., submitted for publication). Subliming matrices are both hydrophobic and chemically inert and, unlike essentially all other sustained release delivery systems, can stabilize antiviral agents. Subliming matrices stabilize incorporated drug by being continuously hydrophobic (i.e., they do not absorb water during drug release), since drug release is due to a surface erosion process. This has allowed labile drugs or proteins stored within (and releasing from) such matrices to be even more stable than when stored in the solid state in moist air (Gunaseelan et al., submitted). Therefore, the incorporation of a labile peptide such as C5A into subliming matrices may enhance its endurance and maintain its antiviral potency within the genital environment. In this study, we investigated whether the release of C5A, which occurs during sublimation of the matrix material as a result of surface erosion, provides sustained and prolonged anti-HIV microbicidal activities in vitro.

MATERIALS AND METHODS

C5A and subliming matrices.

Both parental amphipathic (SWLRDIWDWICEVLSDFK) and nonamphipathic (SWRLDIWDWICESVLDFK) C5A peptides were synthesized at The Scripps Research Institute, La Jolla, CA. The subliming matrices used in this study were hexamethylcyclotrisiloxane (HMCS) and cyclododecane (CDD). HMCS was purchased from Sigma-Aldrich (St. Louis, MO), and CDD was obtained from Kremer Gmbh (Aichstettin, Germany). Matrix materials were further purified by resublimation at atmospheric pressure.

C5A incorporation into subliming matrices.

C5A was micronized using a mortar and pestle. C5A powder was then homogenously dispersed into subliming HMCS or CDD powder at 0.1, 0.3, and 1.0% (wt/wt) loading through serial dilution in a mortar and pestle. Cylindrical solids (5 mm by 5 mm by 3 mm) of C5A-loaded subliming matrices, which were longitudinally enclosed in impermeable polyvinyl chloride (PVC) sheaths, were prepared by weighing C5A-loaded powders (to yield final pellet weights of ∼175 mg) into a steel mold and then compressing with a force of 545 lb for 1 s using the corresponding cylindrical die in a Carver hydraulic press (Carver Inc., Wabash, IN). Empty placebo HMCS and CDD subliming matrices were prepared in an identical manner.

Viruses.

293T cells were transfected with proviral plasmid (9 μg) and vesicular stomatitis virus G (VSV-G) envelope plasmid (1 μg) (a generous gift from C. Aiken). At day 2, VSV-G-pseudotyped viruses (5 μg of p24) were harvested and used to acutely infect activated human peripheral blood mononuclear cells (PBMC) (2 × 10⁶ cells). Two days postinfection, viruses were harvested, and p24 content was measured by enzyme-linked immunosorbent assay (ELISA) (Perkin Elmer Life Sciences). The following proviral constructs were used: wild-type pNL4.3 (X4); pNL4.3-BaL (R5), in which wild-type NL4.3 envelope was replaced with the R5 BaL envelope; pNL4.3-ΔEnv, which lacks gp160; and pNL4.3-GFP (X4) and pNL4.3-BaL-GFP (R5), which carry the green fluorescent protein (GFP) gene instead of the nef gene (3). The primary HIV-1 isolate JR-CSF was obtained through the NIH AIDS Research and Reference Reagent Program and amplified in PBMC (5 × 10⁶ cells).

Cells.

Blood-derived CD4⁺ T lymphocytes, macrophages, and dendritic cells (DC) were isolated as described previously (7; Gunaseelan et al., submitted). Briefly, human PBMC were purified from fresh blood by banding on Ficoll-Hypaque (Amersham Biosciences; 30 min, 800 × g, 25°C). Primary human CD4⁺ T cells were purified from PBMC by positive selection with anti-CD4 Dynabeads and subsequent release using Detachabead (Dynal Biotech). Cells were cultured in RPMI 1640 (Gibco) supplemented with 10% fetal calf serum (FCS) (HyClone), modified Eagle medium (MEM) amino acids, l-glutamine, MEM vitamins, sodium pyruvate (Gibco), and penicillin plus streptomycin (Cellgro) and were subsequently activated with the bacterial superantigen staphylococcal enterotoxin B (SEB; 100 ng/ml) and mitomycin C-killed PBMC from another donor (10:1 PBMC/CD4⁺ cell ratio). Three days following stimulation, cells were split 1:2 in medium containing interleukin 2 (IL-2) (NIH AIDS Research and Reference Reagent Program; 200 U/ml, final concentration). Cultures were then split 1:2 every 2 days in IL-2 medium and infected with HIV at 7 days poststimulation.

To generate primary human macrophages, monocytes were purified from human PBMC by negative selection (Dynal), activated and cultured at a concentration of 10⁶/ml in Dulbecco's modified Eagle medium (DMEM) supplemented with 10% FCS (HyClone), 0.4% MEM amino acids, 2 mM l-glutamine, 0.4% MEM vitamins, 1 mM sodium pyruvate (Gibco), and 100 U/ml penicillin, 100 mg/ml streptomycin, and 50 ng/ml recombinant human granulocyte-macrophage colony-stimulating factor (rhGM-CSF) (R&D systems), and maintained at 37°C in a humidified atmosphere supplemented with 5% CO₂. To obtain monocyte-derived macrophages, cells were allowed to adhere to plastic and cultured for 6 days to allow differentiation.

Immature DC were cultured from monocytes in the presence of IL-4 and GM-CSF (500 and 800 U/ml, respectively) At day 7, the phenotype of the cultured DC was confirmed by flow-cytometric analysis.

Primary genital epithelial cells (PGEC) of vaginal origin were provided by B. Kahn of the Department of Obstetrics and Gynecology at Scripps Clinic. By rotation of cotton swabs against the vaginal walls, several million cells were collected per individual. A group of reproductive-age, premenopausal Caucasian women 18 to 45 years of age without significant medical problems was enrolled under informed consent and with human subject research ethics committee review approval from the Scripps Clinic. Cells were immediately placed in sterile phosphate-buffered saline (PBS), held at 4°C for 2 h, and transported to the laboratory. After centrifugation (300 × g for 5 min), the cell pellet was digested in 1 mg/ml of collagenase-dispase (Roche Molecular Biochemicals) containing 1 mg/ml of DNase (Sigma) and 0.15 mg/ml of sodium p-tosyl-l-lysine chloromethyl ketone (Sigma) for 1 h at 37°C. The digest was spun down (1,000 × g for 20 min) and resuspended in 250 mg/ml of PBS-bovine serum albumin (PBS-BSA). After additional centrifugation, the pellet was resuspended in 5 mg/ml of PBS-BSA and loaded onto a 50% Percoll gradient. PGEC were then isolated from contaminating cells by fluorescence-activated cell sorting (FACS) as previously described (2) and propagated in collagen type I-coated T-25 flasks in Dulbecco's modified Eagle medium-F12 medium containing 10% FCS and epithelial cell growth supplement (100 μg/ml) (Sigma). PGEC were passaged fewer than three times prior to use in order to maintain their original features.

Cellular toxicity of subliming matrices.

PGEC (100,000) were plated in 6-well plates (BD Biosciences) together with placebo or C5A-loaded subliming matrices in 3 ml complete DMEM. After a week, CellQuanti-MTT reagent (Gentaur) was added, and cells were incubated for 4 h at 37°C. Solubilization solution was then added, and plates were shaken for 1 h at room temperature. The optical density at 570 nm (OD₅₇₀) was measured on a Molecular Devices SpectraMax 384 Plus reader. Saponin (0.01%) was used as positive control. No washes were performed, in order to maintain a continuous exposure of cells to C5A and subliming matrices.

Infection assays. (i) Matrix sublimation in the absence of target cells.

Placebo or C5A-loaded subliming matrices were placed into 6-well for an extended period of time (>60 days) in 3 ml of complete DMEM. At various time points, aliquots of supernatants were collected and immediately frozen at −80°C until use. Collected aliquots were then thawed and individually added to TZM reporter target cells (18) together with HIV (JR-CSF 200 pg of p24). Cells were washed after 3 h. Two days postinfection, TZM cells were lysed and β-galactosidase activity in cell lysate quantified. Data were expressed in relative light units.

(ii) Matrix sublimation in the presence of target cells.

Placebo- or C5A-loaded subliming matrices were placed together with activated PBMC or macrophages (500,000 cells) in 6-well plates for an extended period of time (2 weeks for PBMC and 4 weeks for macrophages) in 3 ml of complete RPMI. At various time points, cells were exposed to pNL4.3 (X4) or pNL4.3-BaL (R5) viruses, and infection was analyzed after 48 h by FACS. Ten thousand cells were analyzed per time point and GFP-positive cells were gated. Data (duplicates) are expressed as percent GFP-positive cells. In some experiments, clarified semen (GeneWay Biotech, Inc., San Diego, CA) and cervicovaginal fluids ((Lee Bio Solutions, St. Louis, MO) were added to cell cultures.

Transcytosis assay.

An HIV transcytosis assay using PGEC seeded onto polycarbonate membrane Transwells was conducted as described previously (2). Briefly, PGEC were seeded onto the upper face of collagen I- and fibronectin-coated, 12-mm-diameter, 3 μm-pore-size polycarbonate membrane Transwells at a density of 10⁵ cells/well and were cultured for 2 days until formation of tight junctions was achieved. Placebo- or C5A-loaded subliming matrices were then placed delicately on the surface of the PGEC monolayer (total volume of 2 ml). The inserts were fed every 2 days. The monolayer on the filter effectively divides the well into an apical compartment and a basolateral compartment. The integrity of each cell monolayer was measured using an epithelial volt-ohm meter (Millipore). To ensure the integrity of the PGEC barrier, we monitored the elevated transepithelial electrical resistance of each cell monolayer and measured the paracellular passage of the extracellular marker inulin by determination of the permeability coefficient as evaluated by a diffusion assay using ¹⁴C-carboxylated inulin (Sigma) (molecular weight, 5,000) in the upper chamber as we described previously (2). After the integrity of the monolayer on the Transwell filters was verified, PGEC were exposed to HIV (50 μl of JR-CSF added to the upper chamber) at different time points, and HIV transcytosis monitored by measuring the amounts of particulate capsid in the basal chamber by p24 ELISA. We eliminated free capsid from particulate capsid by microcentrifuging (14,000 rpm) medium collected from the basal chamber for 90 min at 4°C. Using a biotinylated version of C5A, we verified by ELISA (streptavidin-horseradish peroxidase detection) that the peptide did not cross the PGEC monolayer (2), indicating that the peptide cannot act on transcytosed viruses. After particulate capsid measurement by p24 ELISA, the infectivity of transcytosed viruses (1 ng of p24) was scored on TZM indicator cells as described above.

DC transmission assay.

Blood-derived immature DC were plated at 150,000 cells/well in 6-well plates (BD Biosciences) in DMEM supplemented with 10% FCS (HyClone), 0.4% MEM amino acids, 2 mM l-glutamine, 0.4% MEM vitamins, 1 mM sodium pyruvate (Gibco), 100 U/ml penicillin, 100 mg/ml streptomycin, and 50 ng/ml rhGM-CSF (R&D systems). Cells were incubated with NL4.3-GFP (X4), NL4.3-BaL-GFP (R5), or the single-round NL4.3ΔEnv-GFP pseudotyped with NL4.3 gp160 (R5 JR-CSF) (25 ng p24) for 2 h at 37°C. Cells were washed 3 times with warm medium, and CCR5⁺ Jurkat T cells (100,000 cells) were added. Cells were cultured in a flat-bottom 96-well plate, harvested after 3 days, and fixed in 4% paraformaldehyde-PBS, and GFP expression was measured by FACS. The percentage of infected Jurkat T cells was selectively quantified by gating T cells using an anti-CD3 antibody. Ten thousand cells were analyzed per time point.

RESULTS

Rate of controlled release of C5A from subliming matrices.

To determine whether the sublimable delivery of C5A could provide sustained and extended anti-HIV microbicidal activity, we first asked whether parental C5A (SWLRDIWDWICEVLSDFK) incorporated into two selected subliming matrices—hexamethylcyclotrisiloxane (HMCS) and cyclododecane (CDD)—retains its antiviral activity. Matrix erosion rate has been shown to be primarily a function of the sublimation enthalpy of each matrix material. The sublimation rate (K₀) of HMCS is 108.3, whereas that of CDD is 1.6 (Gunaseelan et al., submitted). Thus, HMCS sublimes in air or water ∼100-fold more rapidly than CDD. Although these matrices remain to be tested for safety in humans, they did not exert toxicities when applied to mucosal surfaces of ewes (Gunaseelan et al., submitted). Increasing amounts of C5A (0.1, 1, and 10% C5A) were incorporated into HMCS and CDD solids as described above. Empty placebo HMCS and CDD solids were used as controls. As an additional control, a nonamphipathic version of C5A (SWRLDIWDWICESVLDFK), which lacks antiviral activity (3, 6), was incorporated into HMCS and CDD subliming matrices (10% C5A). Solids were then immersed in complete medium in 6-well plates in triplicate. To determine whether the incorporation of C5A into subliming matrices would preserve its antiviral activities over time, amounts of unformulated C5A identical to those incorporated into the solids were added directly to cell-free medium (DMEM supplemented with 10% FCS, 0.4% MEM amino acids, 2 mM l-glutamine, 0.4% MEM vitamins, 1 mM sodium pyruvate, 100 U/ml penicillin, and 100 mg/ml streptomycin) at 37°C in 10% CO₂ incubators. Supernatant aliquots were collected at various time points over a period of 2 months. Each aliquot was then tested for antiviral activities using TZM reporter cells. Briefly, HIV (JR-CSF, 200 pg of p24) was added to TZM reporter target cells together with an aliquot (5 μl) of collected supernatants. Two days later, infection was scored by measuring amounts of β-galactosidase activity in cell lysates.

We found that in contrast to unformulated nonamphipathic C5A, unformulated parental C5A exhibits potent anti-HIV activities during the first 3 days of incubation in cell-free medium (Fig. 2A). This further confirms that the amphipathic structure of C5A is required for its antiviral activity (3, 6). We also found that empty placebo HMCS and CDD subliming matrices do not influence HIV infection (Fig. 2B and C), demonstrating that the subliming material has no intrinsic antiviral activity. Importantly, incorporation of C5A into the subliming matrices greatly prolonged the anti-HIV potency of C5A. Indeed, unformulated C5A lost its antiviral activity in a few days, whereas C5A in HMCS and CDD retained its antiviral activity for 2 and 8 weeks, respectively (Fig. 2B and C). C5A incorporated into HMCS rapidly established a antiviral state, in less than 2 days, but failed to maintain it after 2 weeks (Fig. 2B). In sharp contrast, C5A incorporated into CDD established an antiviral state late (only after 1 to 2 weeks) but maintained it over a period of more than 2 months. This is in accordance with the fact that HMCS is a fast-subliming material, whereas CDD is a slowly subliming material (Gunaseelan et al., submitted). Increasing the concentration of C5A in the subliming matrices both accelerated and extended the antiviral activities of subliming matrices (Fig. 2B and C). Adding two C5A-loaded HMCS or two C5A-loaded CDD subliming matrices per well did not significantly change the kinetics of the antiviral coverage (Fig. 2D). However, adding a C5A-loaded HMCS solid together with a C5A-loaded CDD solid to a single well established an ideal antiviral state from day 1 to 60. It is likely that the C5A-loaded HMCS solid is responsible for the rapid establishment of antiviral activity, whereas the C5A-loaded CDD solid is responsible for the sustained antiviral state (>2 months).

Fig 2.

(A) Unformulated peptides were added to cell-free complete medium in 24-well plates and placed in 10% CO₂ incubators. At the indicated days, supernatant aliquots were collected and stored at −80°C. After 60 days, collected aliquots were thawed and added to TZM reporter target cells together with HIV (JR-CSF; 200 pg of p24). The x axis shows the time, in days, that C5A-loaded matrices were kept in cell-free medium. Cells were washed twice after 3 h. Two days postinfection, TZM cells were lysed, and β-galactosidase activity in the cell lysate was quantified. Experiments were done in triplicate, and results are expressed in relative light units (RLU). (B and C) HMCS (B) or CDD (C) subliming matrices loaded with 0.1 to 10% C5A were immersed in cell-free complete medium in 24-well plates and placed in 10% CO₂ incubators. At the indicated time, supernatant aliquots (20 μl) were collected and analyzed for anti-HIV activities as described for panel A. All incubations were performed at 37°C; 20 μl of supernatant of cell-free medium was removed at each time point and replaced with fresh medium (20 μl). (D) The experiment was done as described for panels B and C, except that one HMCS solid was added together with one CDD solid to a single well. Error bars represent standard errors for duplicates. The results are representative of two independent experiments using blood from three donors.

Together, these data indicate that C5A incorporated into selected subliming matrices provides sustained and prolonged anti-HIV activities. They also suggest that the release of C5A from subliming matrices can be finely regulated by varying both the amounts of loaded peptide and the ratio of appropriate subliming matrix material.

C5A-loaded subliming matrices provide sustained and extended protection against HIV infection of primary target cells.

After demonstrating that subliming matrices release C5A with potent antiviral activities over an extended period of time (Fig. 2), we asked whether this delivery system is also applicable under more physiological conditions. Specifically, we used human activated PBMC and macrophages (Fig. 3)—the main HIV in vivo targets—rather than immortalized TZM cells (Fig. 2). Moreover, we examined whether C5A-loaded subliming matrices could provide antiviral coverage when placed directly in culture with primary cells in the presence of genital fluids, including seminal plasma and cervicovaginal fluids.

Fig 3.

Unformulated C5A or 10% C5A-loaded subliming matrices (one HMCS solid and one CDD solid) were added to wells containing blood-derived human PBMC or macrophages in the presence of a pool (n = 6) of human seminal plasma (A) or cervicovaginal fluids (B) (dilution, 1/10). At various time points, PBMC were exposed to NL4.3 X4-GFP and macrophages to NL4.3 R5-GFP viruses, and infection was analyzed after 48 h by FACS. All incubations were performed at 37°C; 200 μl of cell culture was removed at each time point and replaced with fresh medium (200 μl). The x axis shows the time, in days, that C5A-loaded matrices were kept in cell cultures. Error bars represent standard errors of triplicates. The results are representative of three independent experiments using blood from three donors.

To test this hypothesis, unformulated C5A or C5A-loaded subliming matrices (one HMCS solid and one CDD solid) were added to wells containing blood-derived human PBMC or macrophages. At various time points, an aliquot of cells was collected and incubated with NL4.3 X4-GFP or NL4.3 R5-GFP viruses and infection was analyzed after 48 h by FACS. In the absence of C5A (medium only), both PBMC and macrophages could be infected at any time point (Fig. 3). Specifically, NL4.3 X4-GFP infected PBMC from day 0 to 10 (Fig. 3, top), whereas NL4.3 R5-GFP infected macrophages from day 0 to 15 (Fig. 3, bottom). We also found that unformulated parental C5A provides protection, but only for about 2 days. Specifically, NL4.3 X4-GFP failed to infect PBMC from day 0 to 2 but infected them from day 4 to 10 (Fig. 3, top). Similarly, NL4.3 R5-GFP failed to infect macrophages from day 0 to 2 but infected them from day 4 to 15 (Fig. 3, bottom). It is likely that unformulated C5A peptide undergoes rapid degradation or inactivation in cell cultures, which would explain the limited antiviral activity. Remarkably, C5A-loaded subliming matrices provide sustained and extended protection (Fig. 3). Specifically, both NL4.3 X4-GFP and NL4.3 R5-GFP were unable to infect PBMC and macrophages from day 0 to 10 and 15, respectively (Fig. 3). We obtained similar results in cell cultures containing seminal plasma (Fig. 3A) and those containing cervicovaginal fluids (Fig. 3B). Together, these data suggest that incorporating C5A into subliming matrices greatly enhance C5A endurance in primary cell cultures containing genital fluids over an extended period of time.

Sustained and extended prevention of HIV transfer from DC to T cells by C5A-loaded subliming matrices.

The virological synapse transfer of infectious HIV particles from DC, which reside in the genital epithelial tissue, to T cells has been proposed to represent an important mode of HIV transmission, distinct from direct infection of T cells and macrophages (10, 12). We thus examined the capacity of C5A incorporated into subliming matrices to prevent DC-mediated transmission of HIV. We took advantage of a replication-defective virus (NL4.3ΔEnv-GFP) (3), which does not encode Env but which has been pseudotyped with HIV Env. Pseudotyped viruses can infect cells because they harbor Env. However, cells which have been infected by pseudotyped viruses cannot produce infectious viruses, because de novo viruses do not encode Env. Thus, the use of pseudotyped viruses allows us to determine the effect of C5A incorporated into subliming matrices on the transmission of infectious particles from DC to T cells independently of DC infection.

Unformulated C5A, empty subliming matrices, or C5A-loaded matrices (one HMCS solid and one CDD solid) were added to wells without cells. At various time points, wild-type NL4.3 X4-GFP and NL4.3-R5-GFP or pseudotyped NL4.3ΔEnv-GFP/gp160 JR-CSF R5 Env viruses (25 ng p24) were added to DC. Two hours later, at which time the entry of the virus into DC is completed, DC were washed to remove free virus and added to wells containing placebo or C5A-loaded subliming matrices. To measure DC-to-T-cell transmission, Jurkat T cells were added to wells for 3 days, and the percentage of T cells that were infected (gated with an anti-CD3 antibody) was analyzed by FACS. Only pseudotyped viruses which have been rapidly transferred from DC to T cells through the virological synapse (independently of DC infection) can infect T cells. Indeed, pseudotyped viruses that infect DC can no longer infect T cells, since they do not encode Env. Because DC were washed prior to the addition of T cells, the T-cell infection by the pseudotyped virus could arise only from pseudotyped particles that were transferred from DC to T cells. Importantly, C5A released from subliming matrices prevented the transfer of both wild-type and pseudotyped viruses from DC to T cells (Fig. 4). Remarkably, the C5A-mediated block was sustained over a period of 3 weeks. This suggests that the sublimable delivery of C5A provides continuous and extended prevention of HIV transfer from DC to T cells.

Fig 4.

At various time points, wild-type NL4.3 X4-GFP and NL4.3-R5-GFP viruses or pseudotyped NL4.3ΔEnv-GFP/gp160 JR-CSF R5 Env viruses (25 ng p24) were added to DC (100,000 cells). After 2 h, DC were washed and added to wells together with supernatants collected from cell-free medium containing unformulated C5A or empty or 10% C5A-loaded subliming matrices (one HMCS solid and one CDD solid). To measure DC-to-T-cell transmission, Jurkat T cells were added to wells for 3 days, and the percentage of infected T cells (gated with an anti-CD3 antibody) was analyzed by GFP expression. Error bars represent standard errors of duplicates. All incubations were performed at 37°C; 20 μl of cell-free medium was removed at each time point and replaced with fresh medium (20 μl). The x axis showing days represents the duration of C5A-loaded matrices kept in cell cultures. The results are representative of three independent experiments using blood from three donors.

C5A-loaded subliming matrices provide sustained and extended inhibition of HIV transcytosis through primary genital epithelial cells (PGEC).

We examined the antiviral effect of the sublimable C5A delivery system on HIV transcytosis through PGEC monolayers. To address this issue, we took advantage of an in vitro Transwell chamber assay which was developed in our laboratory and which mimics HIV transcytosis through PGEC (2). PGEC were seeded onto the upper face of membrane Transwells and cultured until formation of tight junctions was achieved. Placebo or C5A-loaded subliming matrices were then placed on the surface of PGEC monolayers, which divide the well into an apical compartment and a basolateral compartment. The integrity of each cell monolayer was carefully monitored. Both empty and C5A-loaded subliming matrices do not alter the integrity of the monolayer on the Transwell filters (data not shown). At various days, HIV (JR-CSF) was added to the upper chamber, which contains subliming matrices and PGEC monolayers. HIV transcytosis was monitored after 8 h at 37°C by measuring amounts of particulate capsid in the basal chamber by p24 ELISA. After particulate capsid measurement by p24 ELISA, infectivity of transcytosed viruses (1 ng of p24) was scored on TZM indicator cells as described above. We previously showed that HIV efficiently transmigrated through PGEC, in contrast to a virus lacking gp160 (2), demonstrating that the viral glycoprotein is required for transcytosis and that the PGEC layer does not allow nonspecific transcytosis of viruses. In contrast to empty subliming matrices, C5A-loaded subliming matrices totally prevented HIV transcytosis (Fig. 5A). Note that Fig. 4 shows the duration of PGEC exposure to C5A-loaded matrices. Importantly, the transcytosis block was maintained over a period of 3 weeks (Fig. 5A). As mentioned above, the incorporation of C5A into subliming matrices greatly enhances its endurance in culture, since unformulated C5A lost its antiviral effect in less than a week (Fig. 5A). Importantly, the p24 values in the basal chamber perfectly correlated with infectivity values obtained with TZM cells as targets (data not shown).

Fig 5.

(A) PGEC were seeded onto the upper face of Transwells and cultured until formation of tight junctions was achieved. Empty or 10% C5A-loaded subliming matrices were placed on the surface of PGEC. At various days, JR-CSF was added to the upper chamber, which contains subliming matrices and PGEC monolayers. HIV (JR-CSF) transcytosis was monitored after 8 h at 37°C by measuring amounts of particulate capsid in the basal chamber by p24 ELISA. Free capsid from particulate capsid was eliminated by microcentrifuging medium collected from the basal chamber for 90 min at 4°C. The x axis shows the duration, in days, of PGEC exposure to C5A-loaded matrices. Error bars represent standard errors of duplicates. Data are percent transcytosis, obtained by fixing at 100 the percentage of HIV transcytosis in the presence of medium only B. PGEC (100,000) were plated in 6-well plates together with 0.01% saponin, unformulated C5A, or empty or C5A-loaded subliming matrices in complete medium. CellQuanti-MTT reagent was added after a week, and cell viability was quantified by OD₅₇₀ reading. No washes were performed, in order to maintain a continuous exposure of cells to C5A and subliming matrices. Error bars represent standard errors of duplicates. The results are representative of two independent experiments using PGEC from two donors.

We also examined the possibility that C5A-loaded subliming matrices prevented HIV transcytosis due to their toxicity to PGEC. No washes were performed to maintain a continuous exposure of cells to C5A. As a control, cells were exposed to the detergent saponin. In contrast to 0.01% saponin, C5A released from subliming matrices was not toxic to vaginal epithelial cells (Fig. 5B). These data suggest that the solids are not toxic and that C5A delivered from sublimable matrices prevents HIV transcytosis without interfering with the integrity of the epithelial monolayers. Importantly, we obtained similar results with primary human cells other than PGEC, such as PBMC, macrophages, and DC (data not shown). Together, these data further indicate that matrices and C5A delivered from them do not have any toxicity for primary human cells even after extended exposure.

DISCUSSION

We have shown in this study that volatile water-insoluble organic subliming matrices represent an attractive sustained release delivery system for intravaginally administered HIV microbicides such as C5A. Specifically, we demonstrated that HMCS and CDD matrices in combination increased the biological availability of C5A after a prolonged period in cell-free medium or cell culture at body temperature. We showed that unformulated C5A loses its potency in a few days, whereas C5A incorporated into HMCS and CDD solids maintains its antiviral potency for several weeks. This suggests that the antiviral activities of C5A incorporated into HMCS and CDD subliming matrices are greatly preserved. This also suggests that the antiviral activities of unformulated C5A are rapidly lost in cell-free medium or cell culture. It is likely that unformulated C5A is more susceptible to degradation or inactivation in the milieu than C5A incorporated into the subliming matrices.

We demonstrated that the rate of C5A release from subliming matrices can be finely controlled. Specifically, by varying the amounts of peptide incorporated into the subliming matrices, we can accelerate or delay the time of establishment of the antiviral state. We also found that the degree of sublimation of the matrix dictates the release rate of C5A. Specifically, C5A-loaded HMCS solids, which sublime more rapidly than CDD solids, established an antiviral state more promptly than C5A-loaded CDD solids. In contrast, C5A-loaded CDD solids, which sublime more slowly than HMCS solids, provided a remarkably sustained and continued antiviral state (>2 months). Importantly, a combination of C5A-loaded HMCS (fast-subliming) and CDD (slowly subliming) matrices can provide both an early and a prolonged antiviral state. Altogether these in vitro data suggest that a rapid, continuous, and extended genital delivery of HIV microbicides can be reached by adjusting both the concentration of loaded microbicide and the choice and ratio of subliming matrix material.

We obtained several lines of evidence that C5A-loaded subliming matrices provide sustained and prolonged distinct microbicidal properties in vitro. First, C5A-loaded subliming matrices placed in culture with either human PBMC or macrophages prevented infection of these primary in vivo targets for an extended period of time even in the presence of genital fluids. Second, C5A-loaded subliming matrices precluded the transfer of infectious particles from human DC to T cells over a period of 3 weeks. Finally, C5A-loaded subliming matrices efficiently inhibited the passage of HIV through PGEC monolayers over a period of 3 weeks without altering the viability and the integrity of the monolayers. Together, these results suggest that C5A released from subliming matrices mediates a sustained and extended block of HIV transmission by interfering at least at three levels: macrophage and T-lymphocyte infection, DC-to-T-cell transfer, and genital epithelial cell transcytosis.

Sustained and controlled delivery of microbicidal candidates such as C5A through sublimation of matrices may represent a revolutionary unobtrusive genital delivery system. Controlled release rates and durations of C5A may represent a major step on the way to the establishment of a delivery system applicable to all microbicides. Further work is required to determine whether this novel in vitro delivery system is applicable in vivo. Specifically, C5A-loaded subliming matrices will be embedded in elastomer rings, which will be vaginally placed in mice or macaques and subsequently tested for safety and efficacy in HIV transmission models. The ultimate goal is to develop a unique therapeutic system which consists of highly potent HIV microbicidal candidates such as C5A incorporated into drug-stabilizing sustained-release subliming matrices to provide long-term and effective prophylaxis against sexual HIV transmission.