Lubricants are commonly used to enhance pleasure while possibly reducing trauma during coitus. For example, while dyspareunia increases with the age of a woman, up to 46% of women between 18 to 45 years old experience it at some point 
. Clinicians often prescribe the use of OTC lubricants to alleviate its symptoms. Likewise, receptive anal intercourse can lead to trauma of the anus/rectum. Lubricants could reduce this trauma and are used by a significant number of men engaging in receptive anal intercourse 
. However, a comprehensive analysis of the safety of lubricant products has not been done. Our findings show that six of the 10 aqueous-based lubricants tested were hyperosmolar and reflected the cellular toxicity and damage to epithelial monolayers and explant epithelium. Two nearly iso-osmolar gels and both silicone-based gels evaluated here showed no cellular toxicity or damage to epithelial monolayers and explant epithelium. Despite the lack of increased HIV-1 infection in ectocervical tissue, these results convey the importance of using condoms in conjunction with compatible lubricants.
Preservatives are added to OTC lubricant products to ensure product shelf-life. The choice of preservative could have an impact on the microflora of the genital and GI tracts. Chlorhexidine gluconate is a bactericidal compound frequently used in pre-operative skin preparations and mouth washes for gingivitis 
. KY Jelly contains chlorhexidine gluconate and was the only lubricant that completely killed the Lactobacillus
species tested. Methyl- and propylparaben are bactericidal preservatives commonly used in cosmetics and pharmaceutical preparations. These preservatives are added to topical microbicide gel formulations to protect from microbial contamination. Nine of the 11 OTC aqueous-containing products tested contained methylparaben and/or propylparaben. However, only Gynol II and Replens significantly reduced the viability of two of the three lactobacillus species tested. It’s unknown why the other OTC lubricant products that contain these preservatives did not affect the Lactobacillus
viability, but it could be due to concentrations of the parabens used in those gels. Additionally, Gynol II contains nonoxynol-9 which has a detrimental effect on Lactobacillus
viability in women who use nonoxynol-9-containing gels 
. Collectively, changes in vaginal flora and in particular loss of Lactobacillus
species are associated with the development of bacterial vaginosis (BV) 
. KY Jelly has been used in a number of microbicide safety trials as a “placebo”. Overall, no significant increase in BV was observed in these trials. However, several women using KY Jelly developed BV during the study as compared to none in the active gel arms 
. These trials did not include a “no gel” arm that would have been helpful in defining BV incidence in the population. Our data would suggest consistent vaginal use of Gynol II and KY Jelly could lead to changes in lactobacilli populations. However, more research is needed to define the role of extended OTC lubricant product use and incident BV.
Previous testing using several in vitro
and ex vivo
models showed some aqueous-based lubricants reduced epithelial cell viability. Our evaluation of lubricants on epithelial cell lines in general showed decreasing cell viability associated with increasing osmolality of the gels. The hyperosmolar lubricants also perturbed epithelial monolayers showing a delayed reduction of the TER in the HEC-1-A cell line and a swift reduction of TER in the Caco-2 cell line. Interestingly, the slightly hypo-osmolar gel, Sliquid Organic, showed similar epithelial toxicity and loss of monolayer retention to the hyperosmolar gels. The reasons for this are not clear as there were no obvious ingredients on the package insert () which would indicate toxicity. The remaining near iso-osmolar and hypo-osmolar gels did not affect the epithelial cell viability or monolayers. Of interest, the silicone-based lubricants showed no effect on the epithelial viability or monolayer using our testing methods. The effects to the epithelial monolayers were predictive of the epithelium retention on the polarized colorectal and ectocervical tissues. Three products showed a consistent decrease in tissue viability: Astroglide, Gynol II and KY Jelly. Astroglide’s osmolality likely impacted the tissue viability while chlorhexidine affected tissue viability in KY Jelly exposed tissue. Gynol II contains 2% N9 which decrease cellular/tissue viability ex vivo
. In this study, a 24 h exposure to Gynol II reduced tissue viability by 67% to 80%. When viewed histologically, the ectocervical tissues showed fractured epithelium and were acellular while the colorectal tissues were stripped of their epithelium and were necrotic. The impact of the hyperosmolar and nonoxynol-9-containing gels on colorectal tissue was not unexpected as clinical data has shown stripping of the epithelium 
. These studies followed patients closely (within 30 min of dosing) and obtained biopsies to perform histological evaluation. Epithelial findings of the ectocervix/vagina have not been documented after vaginal use of these products. This is likely due to inconsistent observations after the last vaginal dose of gel and the lack of biopsies taken for histological testing. However, the stratified squamous epithelium of the ectocervix/vagina may not show the apparent epithelial disruption as does the columnar epithelium of the gastrointestinal tract. The other hyper- and hypo-osmolar gels had modest reductions of tissue viability with commensurate changes in mucosal epithelium. The only exception was Sliquid Organic which showed modest loss of ectocervical tissue viability and good epithelial retention of both tissues. Good Clean Love and PRÉ – iso-osmolar gels – retained tissue viability and epithelium. The silicone-based lubricants showed no alteration of tissue viability or epithelium either. Interestingly, while marketed specifically for rectal use, both Boy Butter products showed epithelial striping of the colorectal tissue with only modest changes seen in ectocervical tissue. These are both lipid-based lubricants and not intended for use with condoms. Overall, the safest lubricants, based on the testing presented here, appear to be the nearly iso-osmolar gels and silicone-based liquids.
OTC lubricants typically do not contain drugs or pharmacologically active agents known to have activity against HIV-1 (). However, they do contain a number of ingredients that contribute to the physical characteristics (viscosity, lubrication) and stabilization (antioxidants, preservatives). Two such agents are glycerin and EDTA 
. Previous work suggested that OTC lubricant products, in particular Astroglide, may have anti-HIV-1 activity based on activity afforded by inactive or excipients ingredients present in the product 
. Astroglide was shown to block cell-associated or cell-free HIV-1 infection 
. Further work showed glycerol and polyquaternium-32 were the active anti-HIV-1 compounds in Astroglide 
. However, none of their work correlated the anti-viral activity to the cellular toxicity associated with the lubricants. We show here the majority of the lubricants did not possess any anti-HIV-1 activity beyond their toxic concentrations. This was corroborated by a recent paper showing the same four OTC lubricants tested in this paper along with other iterations of them had no anti-HIV-1 activity 
. However, it was suggested by Begay et al. 
that the polyquaternium-32 could increase HIV-1 infection. Not being able to source the polyquaternium-32, a similar molecule (MADQUAT) was used in their assay. They found that MADQUAT increased HIV-1 infection. The results presented here show that in ex vivo
mucosal tissue, Astroglide did not increase HIV-1 infection. Additionally, the three aqueous-based lubricants that had minimal loss of viability, Good Clean Love, PRÉ, and Slippery Stuff, demonstrated modest anti-HIV-1 activity using our in vitro
assay. However, no reduction of HIV-1 infection was noted in ectocervical tissue by the two lubricants tested. This could be an artifact of the in vitro
assay because the viscosity of these gels could reduce the infection of the indicator cell line. It should be noted when we evaluated the hydroxyethyl cellulose (HEC) gel – the universal placebo gel – with similar viscosity, no anti-HIV-1 activity was noted 
, . The HEC gel contains 2.7% HEC polymer with other salts 
while the other gels use agar, xanthan gum, and carbomer at unknown concentrations (). These other gelling agents could impact HIV-1 infection of the TZM-bl cells at high concentrations. Sliquid Organic, which contains green tea extract and showed no in vitro
anti-HIV-1 activity, did provide modest protection against HIV-1 infection in the ectocervical tissue. The active ingredient in green tea, epigallocatechin gallate, has been shown to have anti-HIV-1 
and HSV 
properties in vitro
likely by interfering with viral attachment to its cellular targets. However, the effectiveness of this extract against HIV-1 infection in a clinical setting would likely be poor. Proven effective anti-virals such as PRO 2000 
and tenofovir 
when used as a topical microbicide have provided no to modest protection against HIV-1 acquisition in high-risk women. Consequently, the benefit of using a lubricant for HIV-1 prevention is limited.
The mechanisms of HIV-1 transmission during coitus have not been fully elucidated, but the epithelium is an important barrier to block entry 
. OTC spermicides and first generation, HIV-1 entry inhibitor microbicide gels were shown to increase susceptibility to HIV-1 infection in clinical trials 
. Subsequent analysis suggested that epithelial disruption might be a primary reason for their failure. Nonoxynol-9 and cellulose sulfate gels disrupt epithelial monolayers in vitro
which can result in increased transmission of HIV-1 across the breached epithelium 
. Astroglide, KY Jelly, and other hyperosmolar lubricant products described here and elsewhere 
show striping/facture of the epithelium with no anti-HIV-1 activity. Indeed, the HSV/mouse models suggest that these products may increase susceptibility to infection 
. Glycerin and EDTA were shown to be associated with the increased HSV infection in the mice. The glycerin likely removes the epithelium through osmotic changes while EDTA chelates the calcium and magnesium ions that are important for tight junction formation between epithelial cells and opens up the epithelium. We treated ectocervical tissue with 0.1% EDTA for 2 h and then exposed to HIV-1. We found a modest, but significant, increase in HIV-1 replication in the EDTA-exposed tissue as compared to the untreated tissue. Despite showing epithelial damage, the lubricant treated tissues did not increase HIV-1 replication. These data are supported by a paper presented at Microbicides 2008 showing little changes in gastrointestinal tract permeability after taking 20 pinch biopsies from the mucosa as compared to exposure to nonoxynol-9 
. Participants who had biopsies taken had similar plasma levels of a radiolabeled isotope, 99m
Tc-DTPA, as compared to participants who received an enema containing a saline derivative. Conversely, participants who received an enema containing nonoxynol-9 had 22-fold more radiolabeled isotope in their plasma than the saline derivative. Collectively, these data suggest that epithelial trauma alone may not be sufficient to increase HIV-1 infection, but other factors such as an inflammatory milieu may also be necessary.
There are several limitations to our evaluation of OTC lubricants. The majority of our testing was restricted to aqueous-based gels and silicone-based liquids. We attempted to evaluate lipid-based emulsions in our assays and have not had satisfactory results. For example, Boy Butter Original and Boy Butter H2O are popular lubricants used by MSM. However, their cream-like consistency does not allow them to be utilized in our assays due to our inability to wash the lubricant from the cells. This limits their thorough evaluation. However, we were able to evaluate their safety on ex vivo tissue directly. Also, this is the first time ex vivo tissue has been used to model susceptibility to HIV-1. Using EDTA as the control to increase susceptibility, a modest increase in HIV-1 replication was noted despite more p24 expressing cells consistently observed by IHC. Ex vivo tissue is not necessarily representative of the in vivo environment because of lack of tissue regeneration, lack of recruitment of immune cells, and independence from hormones.
While these data are compelling, additional research is needed to determine if hyperosmolar lubricants cause sub-clinical damage to the mucosal epithelium and lamina propria which could potentially lead to increased susceptibility to HIV-1 and other sexually transmitted diseases. Indeed, a recent paper evaluating persons who consistently use lubricants during RAI showed a higher incidence of sexually transmitted diseases 
. However, the type of lubricant (e.g. aqueous-based versus silicon-based) was not defined in this study. Taken together, condoms, irrespective of male or female, remain the best available method to prevent the acquisition of HIV-1 and sexually transmitted diseases and should be used with compatible lubricants.