Safety of multiple administrations of spermicidal LL-37 antimicrobial peptide into the mouse female reproductive tract

Abstract

We have previously demonstrated spermicidal activity of LL-37 antimicrobial peptide on mouse/human sperm and its contraceptive effects in female mice. With its microbicidal action against Neisseria gonorrhoeae, LL-37 warrants development into a multipurpose prevention technology (MPT) agent for administering into the female reproductive tract (FRT). However, it is important to verify that multiple administrations of LL-37 do not lead to damage of FRT tissues and/or irreversible loss of fecundity. Herein, we transcervically injected LL-37 (36 µM–10× spermicidal dose) into female mice in estrus in three consecutive estrous cycles. A set of mice were sacrificed for histological assessment of the vagina/cervix/uterus 24 h after the last injection, while the second set were artificially inseminated with sperm from fertile males 1 week afterwards, and then monitored for pregnancy. Mice injected with PBS in parallel were regarded as negative controls, whereas those injected with vaginal contraceptive foam (VCF, available over the counter), containing 12.5% nonoxynol-9, served as positive controls for vaginal epithelium disruption. We demonstrated that the vagina/cervix/uterus remained normal in both LL-37-injected and PBS-injected mice, which also showed 100% resumption of fecundity. In contrast, VCF-injected mice showed histological abnormalities in the vagina/cervix/uterus and only 50% of them resumed fecundity. Similarly, LL-37 multiply administered intravaginally caused no damage to FRT tissues. While our results indicate the safety of multiple treatments of LL-37 in the mouse model, similar studies have to be conducted in non-human primates and then humans. Regardless, our study provides an experimental model for studying in vivo safety of other vaginal MPT/spermicide candidates.

Introduction

LL-37 is the only human cathelicidin antimicrobial peptide, with known microbicidal activities against more than 50 microbes including Neisseria gonorrhoeae, the cause of one of the most common sexually transmitted diseases globally (Tanphaichitr et al., 2016; Kiattiburut et al., 2018; Wang et al., 2019). LL-37 and its truncated peptides (GI-10 and GF-17) have been shown to exert spermicidal activity on both human and mouse sperm (Srakaew et al., 2014; Kiattiburut et al., 2018; Lee et al., 2022). The contraceptive effects of these peptides have also been demonstrated, when administered into the mouse female reproductive tract (FRT) (Srakaew et al., 2014; Kiattiburut et al., 2018; Lee et al., 2022). To date, LL-37 is the only human antimicrobial peptide that possesses spermicidal activity (Tanphaichitr et al., 2016).

These dual microbicidal and spermicidal activities of LL-37 give this peptide the potential to be developed into a multipurpose prevention technology (MPT) agent, which can empower women to protect themselves against unwanted pregnancies and sexually transmitted infections (Hynes et al., 2018; Hemmerling et al., 2020). Prior to its clinical development, it is important to demonstrate that the female reproductive physiology and tissues remain normal following multiple administrations of LL-37 into the reproductive tract. The setbacks seen with the attempt to develop nonoxynol-9 (N-9) to be an MPT agent must be avoided. The surfactant N-9 has been used for over 50 years as a spermicide (Raymond et al., 2004; Xu et al., 2022). With results from in vitro experiments and a few studies in animal models indicating microbicidal activity of N-9 against microbes that cause sexually transmitted infection, e.g. HIV/SIV/SHIV (Miller et al., 1992; Moench et al., 1993; Whaley et al., 1993; Patton et al., 1996; Weber et al., 2001; Doncel, 2006), ten clinical trials on N-9 as an anti-HIV in women were initiated (Wilkinson et al., 2002; Tanphaichitr et al., 2016). This was despite the lack of thorough studies on the safety of N-9 on the FRT/physiology. In fact, the last clinical trial was halted half way as women using N-9 vaginal gel had a 2-fold increase in the HIV infection rate, compared with those not using N-9 (Van Damme et al., 2002; Wilkinson et al., 2002). In this report, we used mice as an experimental model to evaluate whether multiple administrations of LL-37 into the FRT caused any histological damages to the vagina, cervix, and uterus and/or inflammation to the vagina. In addition, the reversibility of the contraceptive effects of LL-37 in mice that were pre-treated with the peptide in three consecutive estrous cycles was assessed. The results obtained from LL-37 treated mice were then compared with those from mice treated in parallel with N-9.

Materials and methods

Materials

LL-37 was chemically synthesized by CPC Inc (San Jose, CA, USA) with 85% purity as revealed by HPLC. VCF (Vaginal Contraceptive Foam, Apothecus Pharmaceutical, Oyster Bay, NY, USA) with 12.5% nonoxynol-9 as the active ingredient was purchased over the counter from a local drugstore. Nonoxynol-9 solution (USP 1467950) in a sealed ampule was available from Millipore Sigma (Oakville, ON, Canada).

Use and boarding of mice with ethical approval

CD-1 male (3–6 months old) and CF-1 female (8 weeks old) mice purchased from Charles River Laboratories (Senneville, QC, Canada) were boarded in a temperature-controlled (22.5°C) room with a light (14 h)/dark (10 h) cycle. Use and handling of mice were approved by the University of Ottawa Animal Care Committee (protocol 2567), following ARRIVE checklists and guidelines.

Transcervical injection and intravaginal injection of LL-37 into mice

Transcervical injection

Fifty microliters of LL-37 (36 µM), VCF, and PBS were separately injected into the uterine cavity of female mice on the estrus night (approximately 1 h before the expected earliest ovulation time, i.e. 5 h after the start of the dark cycle in the estrus phase). The transcervical injection was performed using a 23G needle cut to be 18 mm long with the end blunted, attached to a BD 1 ml Tuberculin syringe. The estrous cycle of these female mice was tracked daily through cytological assessment of the vaginal lavage (Byers et al., 2012) collected by intravaginal injection and rinsing of 50 µl of PBS 10–15 times in the vaginal lumen.

Intravaginal injection

Female mice were induced to be in diestrus by subcutaneous injection with 2.5 mg Depo-Provera (Depot medroxyprogesterone acetate (DMPA), Pfizer Canada, Kirkland, QC, Canada) (Day 1) (Galen et al., 2007). On Day 4, 20 µl of LL-37 (36 µM), VCF, and PBS were intravaginally injected (using the same type of syringe and needle as used for transcervical injection) into these mice. Vaginal lavages were then collected for cytokine/chemokine analyses (see below) on Day 5. The whole procedure was then repeated two more times starting with DMPA injection on Day 6 and Day 11.

The experimental scheme of transcervical and intravaginal injection of LL-37, VCF, and PBS is shown in Supplementary Fig. S1.

Histological assessment of the lower FRT tissues

The tissues (vagina, cervix, and uterus) were collected from mice sacrificed 24 h after multiple transcervical or intravaginal injections of LL-37 (36 μM), VCF, or PBS. The tissues were fixed in Bouin’s solution and then embedded in paraffin for tissue sectioning with 4 μm thickness, following standard methods. Tissue sections were stained with hematoxylin and eosin (Millipore Sigma) and digitally imaged on an Aperio CS2-Digital Pathology Scanner (Leica Biosystems, Buffalo Grove, IL, USA) or ZEISS Axio Scan.Z1 Slide Scanner (Carl Zeiss Canada Ltd, Toronto, ON, Canada). Images obtained from each slide were then analyzed by ImageScope software (Leica Biosystems) or ZEN Blue 3.2 imaging software (Carl Zeiss Canada). The normalcy and aberrancy of the histology of the vagina sections were assigned based on the previously described criteria (Westwood, 2008).

Assessment of the egg quality of females previously transcervically injected with PBS, LL-37, and VCF for three times

After the last of the three transcervical injections of PBS, LL-37, and VCF, three female mice in each injection group were boarded without any treatment for 1 week. They were then superovulated by intraperitoneal injection of pregnant mare serum gonadotropin (PMSG, 5 IU, Prospec, Ness-Ziona, Israel) and human chorionic gonadotropin (hCG, 5 IU, Millipore Sigma) (with a 48 h interval between the two injections). Oviductal cumulus–oocyte complexes were retrieved from these sacrificed females 14 h post-hCG injection. Cumulus free eggs were prepared by hyaluronidase treatment and assessed for their ability to be fertilized by sperm collected from fertile males, which were processed through Percoll gradient centrifugation and capacitated in Krebs–Ringer bicarbonate (KRB) medium supplemented with 0.3% bovine serum albumin (BSA). After 6 h of gamete co-incubation in KRB-0.3% BSA, eggs were microscopically scored for the presence of two pronuclei as evidence of fertilization in vitro. All procedures in these experiments were as described previously (Tantibhedhyangkul et al., 2002). As negative controls of all injections, in vitro fertilization was also assessed on oviductal eggs retrieved from superovulated females, which were not priorly injected with any compound. The experiment was performed twice. Data on percent fertilized eggs from six females were then analyzed.

Fertilized eggs were cultured in the same medium for another day, the time period which allowed normal fertilized eggs to develop into two-cell embryos. The number and morphology of two-cell embryos obtained were then microscopically recorded.

Fecundity assessment

Females previously transcervically injected with LL-37, VCF, or PBS in three consecutive estrous cycles were boarded without any treatment for 1 week. Their estrous cycle was then tracked and they were artificially inseminated through transcervical injection of sperm collected from fertile males during the estrus night following our previously described method (Srakaew et al., 2014). Pregnancies were monitored by abdominal enlargement and then pup delivery, and litter sizes were recorded. For mice that did not become pregnant, artificial insemination was repeated 1 month after the first insemination, and pregnancies and litter sizes were likewise documented. The experiment was performed twice with 4 mice in each treatment group in each experiment (i.e. the total of 8 mice in each treatment group).

Vaginal lavage collection and analyses of chemokines and cytokines

Eighteen hours after each intravaginal injection of LL-37, VCF, or PBS, 60 µl PBS containing Roche Protease Inhibitor Cocktail (Millipore Sigma, Catalog No. 05892970001) was injected into the vagina of each animal and pipetted up and down 10–15 times. The Protease Inhibitor Cocktail was prepared according to the manufacturer’s instruction. Briefly, the tablet provided was dissolved in 10 ml PBS and 60 µl of it was used directly in each vaginal injection. The vaginal lavage was then centrifuged (350 g, 5 min, room temperature) to pellet cells. The supernatant was then stored at −80°C for subsequent cytokine/chemokine measurement. Seven female mice were used for each treatment group.

The Milliplex Mouse Cytokine/Chemokine Magnetic Bead Panel was used to measure IL-6, MCP-1, TNF-α, IL-10, IL-1α, IL-1β, and MIP-2 (EMD Millipore, Billerica, MA, USA) in vaginal lavages of females previously injected with LL-37, VCF, and PBS. The protocol was followed as provided in the kit, with no additional dilution to the lavage samples. The assay was read on the Luminex Magpix running the Luminex xPonent for Magpix software (version 4.2, Toronto, ON, Canada). Analysis was completed by Milliplex Analyst (version 5.1.0.0., VigeneTech Inc, Carlisle, MA, USA) using 5 parameter curve fitting (log scale).

Statistical analyses

Significant differences of data among samples were analyzed by two-way ANOVA with Tukey’s or Dunnett’s multiple comparison.

Results

Multiple uterine exposure to LL-37 or VCF: duration of the estrous cycle, egg fertilizing ability, histology of the FRT tissues, and resumption of fecundity

In rodents, although semen is ejaculated into the vagina, both sperm and seminal plasma are swept into the uterine cavity within a few minutes (Suarez and Pacey, 2006). Therefore, in order for a spermicide to effectively exert its activity, it has to be administered into the uterine cavity. For this reason, LL-37, VCF, or PBS (negative control) was transcervically injected into female mice just before the earliest time of ovulation, which normally occurs 5 h after the start of the dark period of the estrus phase.

Tracking the estrous cycle prior to the first injection of the PBS-injected group (n = 3), LL-37-injected group (n = 3), VCF-injected group (n = 3) as well as the no-injection group (n = 3) indicated that the duration of the cycle was 4 days in almost all (10) of these mice, with only one mouse in the no-injection group and one mouse in the LL-37 group having an estrous cycle of 5 days. Before the second and third injections, all female mice injected with PBS or LL-37 or un-injected consistently had an estrous cycle of 4 days. In contrast, in the VCF-injected animals, the duration of the estrous cycle increased to 5 days and then 5.7 days before the second and third injection, respectively. The longer estrous cycles before the second and third VCF injection were significantly different from that before the first VCF injection (P < 0.001 and 0.0001, respectively) (Fig. 1). When the length of the three estrous cycles was compared between PBS-injected mice and mice with no injection and between PBS-injected and LL-37-injected mice, there was no significant difference in both comparison pairs. However, this estrous cycle length was significantly different between PBS-injected and VCF-injected animals (P < 0.0001) and between LL-37-injected and VCF-injected females (P < 0.05) (Fig. 1). The increase in the estrous cycle length in VCF-injected mice was most likely due to the action of N-9, the major component of VCF, since females transcervically injected with 12.5% N-9 following the same protocol as that used for VCF also showed similar significant increases from 4 days prior to the first injection to 5.3 and 5.3 days prior to the second and third N-9 injection, respectively (Supplementary Fig. S2).

Eggs retrieved from mice transcervically injected with LL-37 three times were fertilized by sperm from fertile males at the same rate as eggs retrieved from mice injected in parallel with PBS or those from un-injected females (Fig. 2). More than 90% of these fertilized eggs developed into 2-cell embryos in culture the following day, indicating normal in vitro fertilization (Fig. 2). Interestingly, eggs retrieved from mice transcervically injected with VCF three times had the same in vitro fertilization rates and post-fertilization embryo development as eggs collected from mice in the other 3 groups (Fig. 2), indicating that the change in the estrous cycle length in VCF-injected mice was not related to the egg quality in these animals.

Histology studies revealed that the FRT tissues (vagina, cervix, and uterus) of mice transcervically injected with 36 µM LL-37 three times retained normal cellular organizations as observed in control mice transcervically injected with PBS. Specifically, the histology of the vagina tissues showed that both PBS-injected and LL-37-injected mice were in metestrus, having multilayers of the vaginal epithelium with neutrophils recruited toward the surface and the lumen containing cornified cells (bright pink stained) (Fig. 3A). This stage of the estrous cycle was as expected since the tissues were collected about 1 day after the third transcervical injection of PBS/LL-37 in the dark phase of estrus.

In contrast to results observed in LL-37-injected or PBS-injected mice, the vaginal epithelial surface of VCF-injected females appeared drastically different (Fig. 3A). It contained a mucoid surface layer with a violet color stain, and multiple vacuoles (see the area with the yellow frame in Fig. 3A and the magnified image in Fig. 3B). The presence of vacuoles was also observed in the cervical epithelium (Fig. 3A). The nuclei of the columnar cells in the surface layer of the vaginal epithelium of the VCF-injected mice were also not lined up, indicating cell dislocalization (see the area with the blue frame in Fig. 3A and the magnified image in Fig. 3B). In addition, neutrophils were present in the surface layer of the uterine epithelium of VCF-injected mice (see the area with the green frame in Fig. 3A and the magnified image in Fig. 3B), an event not observed in mice injected with LL-37 or PBS (Fig. 3A).

LL-37 has been demonstrated for its in vivo contraceptive effect when transcervically injected with sperm from fertile males (Srakaew et al., 2014). Mice that had been transcervically injected with LL-37 three times, however, could resume their fecundity. When these mice were artificially inseminated with sperm from fertile males 1 week after the last LL-37 injection, 75% of them (6 out of 8 total mice) became pregnant. The same pregnancy rate was also observed in control mice that had been transcervically injected with PBS (Fig. 4). When the two LL-37-treated and the two PBS-treated mice that were not pregnant were artificially inseminated again 1 month after the first insemination, both of them became pregnant. This resulted in a 100% pregnancy rate in both LL-37-treated and PBS-treated mice and the litter sizes of the delivered pups were not significantly different from each other (9.1 ± 1.7 versus 9.7 ± 1.6) (Fig. 4). In contrast, only 1 of 8 VCF-treated mice became pregnant (12.5% pregnancy rate) in the first round of artificial insemination. When the seven non-pregnant VCF-treated mice were re-inseminated, only 3 of them became pregnant, resulting in a 50% total pregnancy rate. Interestingly, VCF-treated pregnant mice delivered pups with a significantly lower litter size (5.7 ± 2.4) than that of pregnant mice treated with LL-37 (9.1 ± 1.7) or PBS (9.7 ± 1.6) (P < 0.05) (Fig. 4).

Multiple vaginal exposure to LL-37 or VCF: histology of the FRT tissues and expression of cytokines/chemokines

Although LL-37 has to be transcervically injected for its spermicidal/contraceptive effects in mice, the effects of the peptide on the vaginal physiology and histology need to be established for its future clinical trials. In humans, ejaculation deposits semen as a coagulum in the vagina. Semen then becomes liquefied within 30 min, allowing motile sperm to move through the cervix into the uterine cavity. LL-37 administered into the vagina can potentially exert its spermicidal activity on these motile sperm.

The histology of the vagina, cervix, and uterus of mice intravaginally administered LL-37 did not show any changes, as compared with that of mice administered PBS (Fig. 5A). In both LL-37-treated and PBS-treated mice, the vagina showed only a few layers of epithelial cells typical of diestrus. This was as expected, since the mice were induced to be in this stage by DMPA subcutaneously injected on the first day of each round of PBS/LL-37/VCF intravaginal injection (see Supplementary Fig. S1). In contrast, the vaginal histology of VCF-treated mice was different from that of PBS-treated and LL-37-treated females. There was a mucoid layer (violet color stained) on the epithelial surface, and with areas of abnormal thickness (Fig. 5A). The nuclei of cells in this surface layer also did not line up, indicating cellular disorganization (Fig. 5A, see the area with the blue frame and the magnified image in Fig. 5B). In addition, there were tears in several areas of the epithelial surface (Fig. 5A, see the area with the red arrow and that with the yellow frame and the magnified image in Fig. 5B). On the cervical epithelial surface of VCF-treated mice, there also appeared cellular disorganization. The oval nuclei of the cervical epithelial cells, which are normally vertically oriented as seen in PBS-treated and LL-37-treated mice, were in a horizontal direction in the cervix of VCF-treated mice (Fig. 5A, see the area in the green frame and the magnified image in Fig. 5B). However, there was no abnormality in the uterus of females intravaginally administered VCF (Fig. 5A).

Seven cytokines/chemokines were selected for the determination whether multiple intravaginal injections of LL-37 induced any significant changes in their expression or not. The anti-inflammatory cytokine, IL-10, was selected because it is present in the vaginal lumen of normal mice throughout the estrous cycle (Hickey et al., 2013). The other six cytokines/chemokines (IL-1α, IL-1β, IL-6, TNF-α, MCP-1, and MIP-2) were profiled mirroring the previous study (Galen et al., 2007) analyzing whether intravaginal administration of N-9 changed the expression of these cytokines/chemokines. This selection was appropriate, since VCF (12.5% N-9) was used as a positive control treatment in our study. IL-1α, IL-1β, IL-6 and TNF-α are proinflammatory, whereas MCP-1 and MIP-2 are chemoattractants, which recruit monocytes and neutrophils, respectively, to the inflammation site (Weber et al., 2019). Of interest is that LL-37 could enhance a release of IL-8 (a human equivalent of MIP-2) from various cell types (Zuyderduyn et al., 2006; Filewod et al., 2009). Multiple intravaginal injections of LL-37 did not induce any significant changes in the levels of the seven cytokines/chemokines in the vaginal lavages as compared with those in the PBS-injected females (Fig. 6). In contrast, there was a significant increase in the MCP-1 chemokine levels in mice following the first and third injection of VCF, as compared with the corresponding levels in PBS-injected mice. In addition, the levels of IL-10 were significantly increased after the third injection of VCF as compared with those in PBS-treated mice (Fig. 6).

Discussion

We determined herein whether multiple administrations of LL-37 into the mouse FRT led to any adverse effects on topical tissues (vagina, cervix, uterus) and whether the contraceptive activity of LL-37 was reversible without permanent impairment to the female reproductive physiology. Mice were used as an experimental model, since they have been used successfully to demonstrate the contraceptive activity of LL-37 and its related peptides (Srakaew et al., 2014; Kiattiburut et al., 2018; Lee et al., 2022) as well as to determine any adverse effects of vaginal microbicides (including N-9) on the reproductive tissues (Milligan et al., 2002; Dayal et al., 2003; Galen et al., 2007; Lozenski et al., 2012; Fields et al., 2014; Barton et al., 2020). Since N-9 has been repeatedly shown to have adverse effects on the vaginal epithelium, which lead to an increase in microbial infections (Fichorova et al., 2001; Milligan et al., 2002; Dayal et al., 2003; Weber et al., 2005; Cone et al., 2006; Galen et al., 2007), this surfactant, which is still being used as a vaginal spermicide available over the counter (Raymond et al., 2004; Xu et al., 2022), was used in the form of VCF in our study as a positive control for its detrimental effects on the lower female reproductive tissues.

LL-37 was administered in two modes in our study in mice. Transcervical injection of LL-37 into the uterine lumen was performed for studies on the contraceptive activity of the peptide. This is because both mouse sperm and seminal plasma are swept into the uterus immediately after semen deposition in the vagina (Suarez and Pacey, 2006), and, therefore, the peptide should already be in the uterine lumen waiting to immobilize the incoming sperm. In contrast, since human sperm and seminal plasma (which may contain microbes) remain in the vagina for some time, the intravaginal administration of LL-37 was performed to model the human situation in which spermicides/microbicides intravaginally pre-administered can exert their activities.

We demonstrated herein that the uterus, cervix, and vagina of mice that had been transcervically administered LL-37 in the three consecutive cycles still remained histologically normal, with vaginal epithelial cell layers typical of the metestrus stage like those of mice administered PBS (Fig. 3). The number and quality of eggs retrieved after PMSG/hCG superovulation from LL-37-treated mice were also the same as those from PBS-treated mice (Fig. 2). Likewise, the length of the estrous cycle of LL-37-treated and PBS-treated mice was not significantly different from each other (Fig. 1). These two lines of results suggested that the production of sex steroid hormones under pituitary control was not altered by LL-37 multiple treatments. It was therefore not surprising that LL-37-pretreated mice resumed their fecundity 1 week after the last transcervical injection of the peptide. The degree of the resumption (75% pregnancy, 6 of 8 total mice) was the same as that in PBS-pretreated mice in the first round of artificial insemination. It could be technical imperfection in artificial insemination that failed the conception in the other two mice. These two mice each in LL-37-pretreated and PBS-pretreated groups, however, became pregnant in the second round of artificial insemination and therefore the fecundity resumption totaled to 100% (Fig. 4).

In contrast to results obtained in PBS-treated and LL-37-treated mice, VCF-treated mice showed distinct histological damage in the epithelial cell layers of the vagina, cervix, and uterus. The presence of neutrophils in the uterine epithelium of VCF-treated mice indicated tissue inflammation. The histological damage observed in the vagina and cervix suggested that there was a leakage of N-9 in VCF from the uterine lumen to the cervix and vagina. This damage in the vagina tissue may be the reason why the estrous cycle of mice transcervically administered with VCF became longer than that of females administered in parallel with PBS or LL-37. The regeneration of stratified layers of the vaginal epithelium during the progress of diestrus into proestrus and estrus would take longer in the damaged tissue in VCF-treated mice. Nonetheless, the egg number and quality obtained from superovulated females that had been multiply-treated with N-9 were still normal (Fig. 2), suggesting that VCF administered into the uterine lumen likely did not reach the ovary to alter mature egg production. The damage in the uterine epithelium, however, may be the reason why VCF-pretreated mice did not fully resume fecundity. In the first round of artificial insemination, only 1 of 8 VCF-pretreated mice became pregnant, and after the second artificial insemination with 1-month lapse time, only 3 of the remaining 7 mice became pregnant, giving an accumulated pregnancy rate of 50%. The average litter size of these 4 pregnant VCF-pre-treated mice was only about 60% of that of the eight pregnant LL-37-pretreated mice or the eight pregnant PBS-pretreated mice (Fig. 4). The histological damage to the uterus (Fig. 3) may result in impairment of embryo implantation and/or in utero fetal development in mice pretreated repeatedly with VCF.

The vagina, cervix, and uterus of mice intravaginally administered LL-37 three times were also histologically normal and appeared the same as PBS-administered mice. The profile of 7 cytokines and chemokines (IL-1α, IL-1β, IL-6, IL-10, TNF-α, MCP-1, and MIP-2) in the vaginal fluid assayed after each administration of LL-37 in the peptide-treated mice was also not different from that of PBS-treated mice (Fig. 6). In contrast, the vaginal and cervical epithelium of mice treated with VCF showed severe damage including sporadic deletion of the superficial cell layer and cell dislocalization, results in accordance with those previously described (Dayal et al., 2003; Barton et al., 2020). MCP-1 appeared to increase significantly in the vaginal fluid after the first and third VCF intravaginal injection, an event not observed in mice injected with PBS in parallel (Fig. 6). The increase in this chemokine in our study is in agreement with that observed on Day 3 and Day 7 in mice treated daily with Advantage gel (containing 3.5% N-9) (Galen et al., 2007). However, we did not see increases in other cytokines/chemokines, as observed in the previous study (Galen et al., 2007), in our VCF-treated mice. This may be due to the different formulation of N-9 form and different treatment regimens between studies.

Overall, our results indicated that multiple uterine or vaginal exposure to LL-37 at 10× concentration (36 µM) of its spermicidal concentration in medium (Srakaew et al., 2014) did not cause histological and/or physiological damage to the uterus, vagina, and cervix as well as the ovary, and after cessation of the multiple treatments with LL-37, mice fully resumed their fecundity. With its known microbicidal activity against N. gonorrhoeae, LL-37 warrants development into an MPT agent. In contrast, our results supported the previous findings that N-9 has adverse effects on the female reproductive tissues and consideration of discontinuing or decreasing the use of this spermicidal surfactant should be made.

Supplementary data

Supplementary data are available at Molecular Human Reproduction online.

Data availability

The data presented in this article are available in the article and its online Supplementary Material. Additional details related to these published data will be shared in response to a reasonable request to the corresponding author.

Acknowledgement

The authors thank Terri Van Gulik for help in manuscript preparation.

Authors’ roles

N.T. laid out the concept that spermicidal antimicrobial peptide LL-37 is safe to be administered in the FRT and she designed all experiments but with significant inputs from S.G.L., W.K., J.A., and S.C.B.S. Laboratory work and result analyses were carried out mainly by S.G.L., as well as W.K. and S.C.B.S. All authors gave comments and interpretations on the results along the way. N.T. wrote the article, whereas S.G.L. and W.K. prepared the figures. J.A. edited and gave input to the manuscript preparation. All authors approved the final version of the submitted article.

Funding

This work was funded by Canadian Institutes of Health Research (PJT 173268 to N.T.).

Conflict of interest

The authors have declared that there is no conflict of interest in the work.

References