Personal and Clinical Vaginal Lubricants: Impact on Local Vaginal Microenvironment and Implications for Epithelial Cell Host Response and Barrier Function

Abstract

Background: A majority of US women report past use of vaginal lubricants to enhance the ease and comfort of intimate sexual activities. Lubricants are also administered frequently in clinical practice. We sought to investigate if hyperosmolar lubricants are toxic to the vaginal mucosal epithelia.

Methods: We tested a panel of commercially available lubricants across a range of osmolalities in human monolayer vaginal epithelial cell (VEC) culture and a robust 3-dimensional (3-D) VEC model. The impact of each lubricant on cellular morphology, cytotoxicity, barrier targets, and the induction of inflammatory mediators was examined. Conceptrol, containing nonoxynol-9, was used as a cytotoxicity control.

Results: We observed a loss of intercellular connections, and condensation of chromatin, with increasing lubricant osmolality. EZ Jelly, K-Y Jelly, Astroglide, and Conceptrol induced cytotoxicity in both models at 24 hours. There was a strong positive correlation (r = 0.7326) between lubricant osmolality and cytotoxicity in monolayer VECs, and cell viability was reduced in VECs exposed to all the lubricants tested for 24 hours, except McKesson. Notably, select lubricants altered cell viability, barrier targets, and inflammatory mediators in 3-D VECs.

Conclusions: These findings indicate that hyperosmolar lubricants alter VEC morphology and are selectively cytotoxic, inflammatory, and barrier disrupting in the 3-D VEC model.

Keywords: 3-D cell culture; cytotoxicity; genital inflammation; human vaginal epithelial cells; inflammatory mediators; nonoxynol-9; vaginal lubricants.

© The Author(s) 2019. Published by Oxford University Press for the Infectious Diseases Society of America. All rights reserved. For permissions, e-mail: journals.permissions@oup.com.

Figures

Figure 1.

Exposure to hyperosmolar lubricants induces changes in cellular morphology, chromatin condensation, and loss of intercellular junctions in monolayer vaginal epithelial cells (ML VECs). ML VECs were treated with a 1:10 dilution of vaginal lubricants in cell media for 24 hours. A, Changes in cellular morphology were observed via crystal violet staining, using a ×20 brightfield objective lens to visualize. Untreated VECs are shown on the top left, and lubricant treatment groups are shown from top to bottom in increasing osmolality. Conceptrol, which contains nonoxynol-9 (N-9), was placed at the bottom as the positive control for cytotoxicity. Cellular staining density (%) can be observed at the top left of each image, quantified using Fiji (ImageJ). ML VECs were fixed with 4% paraformaldehyde. Visualization of glycoconjugates, ie, N-acetylglucosamine and N-acetylneuraminic acid (sialic acid) residues, on the cellular membrane of ML VECs was determined after staining with 10 μg/mL of wheat germ agglutinin (WGA) conjugated with Alexa Fluor 633. Nuclear material was stained using 4’,6-diamidino-2-phenylindole (DAPI), and cells were imaged by confocal microscopy at ×63 original magnification. Images were taken at the middle plane of ML VECs to highlight the cellular morphology as it relates to crystal violet images. All images were collected with the same settings. B, Fiji (ImageJ) was used to quantify crystal violet staining density. Pearson correlation coefficient analysis between ML crystal violet staining density and osmolality was calculated as r = –0.7781 (P = .0394).

Figure 2.

Treatment with hyperosmolar vaginal lubricants induces cytotoxicity and reduces monolayer vaginal epithelial cell (ML VEC) viability at 24 hours. ML VECs were treated with a 1:10 dilution of vaginal lubricants in cell media for 24 hours. A, Percentage cytotoxicity was measured using lactate dehydrogenase (LDH) assay to detect the production of the cytosolic enzyme LDH on induction of cellular toxicity. Spectrophotometric absorbance was measured at 24 hours, and each bar represents the percentage of cytotoxicity calculated relative to a ×10 lysis control. Statistical analysis was conducted by one-tailed Mann–Whitney U test, and data were collected from 3 independent experiments. B, Pearson correlation coefficient (r) analysis was used to measure the linear correlation between lubricant osmolality and percentage cytotoxicity at 24 hours. A positive correlation r value of 0.7326 was calculated with P = .0306. C, Percentage cell viability was determined using the MTT assay. MTT is produced by actively proliferating cells, and this colorimetric assay was used to measure cell viability relative to untreated ML VECs at 24 hours after treatment with vaginal lubricants. Statistical significance was determined by a one-tailed Mann–Whitney U test, and data are representative of 3 independent experiments. **P ≤ .01, ***P ≤ .001, ****P ≤ .0001.

Figure 3.

Treatment with select vaginal lubricants causes cytotoxicity and reduces cell viability in 3-dimensional (3-D) vaginal epithelial cell (VEC) aggregates at 24 hours. The 3-D VECs were treated with a 1:10 dilution of vaginal lubricants in cell media for 24 hours. A, Percentage cytotoxicity was measured via lactate dehydrogenase assay, and spectrophotometric absorbance was measured at 24 hours. Each bar represents the percentage of cytotoxicity calculated relative to a ×10 lysis control. Statistical analysis was conducted by a one-tailed Mann–Whitney U test, and data are the average of 3 independent experiments. B, Mean percentage cytotoxicity of each lubricant treatment was plotted on the left y-axis (dashed line) and lubricants were listed from left to right in order of increasing osmolality. The positive control for cytotoxicity in this study, Conceptrol, which contains 4% nonoxynol-9 (N-9), was placed on the far right of the x-axis. Cell viability of 3-D VECs was counted using Trypan blue exclusion (solid line) and displayed on the right y-axis. A strong negative correlation r value of –0.7468 was calculated with a P value of .0333. *P ≤ .05, **P ≤ .01. C, Pearson correlation analysis between monolayer (ML) and 3-D cytotoxicity was conducted including Conceptrol (4% N-9), and a strong positive coefficient (r = 0.9512, P = .0003) was calculated. Dashed lines identify quadrants of the graph. D, Pearson correlation analysis of ML vs 3-D cytotoxicity excluding Conceptrol (r = 0.6348, P = .1256).

Figure 4.

Cluster analysis identifies lubricants with potentially harmful epithelial barrier outcomes. Heat map of altered immune mediator and barrier target secretion. Three-dimensional (3-D) vaginal epithelial cells were treated with lubricants at 1:10 for 24 hours and cell supernatant was taken from 3 independent experiments. The data were mean centered, and variance was scaled for each target and is presented as log2-transformed data. Lubricants are grouped within clusters determined by hierarchical cluster analysis. Clustering of the heat map was based on Euclidean distance between rows and columns and average linkage cluster algorithm. Significance of protein induction levels as compared to untreated was conducted on log2-tranformed data using one-way analysis of variance. Results with a P value ≤.05 were considered significant. Conceptrol (4% N-9) was included as a positive control for cytotoxicity. Osmolality and 3-D cytotoxicity of corresponding lubricant treatment are shown beneath cluster analysis. Abbreviations: CCL, C-C motif chemokine ligand; EMMPRIN, extracellular matrix metalloproteinase inducer; IL, interleukin; MIF, macrophage migration inhibitory factor; MMP, matrix metalloproteinase; MUC, mucin; N-9, nonoxynol-9; RANTES, regulated on activation, normal expressed and secreted protein.