Temporal dynamics of the human vaginal microbiota

Abstract

Elucidating the factors that impinge on the stability of bacterial communities in the vagina may help in predicting the risk of diseases that affect women's health. Here, we describe the temporal dynamics of the composition of vaginal bacterial communities in 32 reproductive-age women over a 16-week period. The analysis revealed the dynamics of five major classes of bacterial communities and showed that some communities change markedly over short time periods, whereas others are relatively stable. Modeling community stability using new quantitative measures indicates that deviation from stability correlates with time in the menstrual cycle, bacterial community composition, and sexual activity. The women studied are healthy; thus, it appears that neither variation in community composition per se nor higher levels of observed diversity (co-dominance) are necessarily indicative of dysbiosis.

Fig. 1

Dynamics of vaginal community state types in 32 women over 16 weeks. (A) Heatmap showing the proportions of community state types (I, II, III, IV-A and IV-B) observed within a woman over time (color key is indicated below panel C) that were used to generate the dendogram that depicts distances between proportions of the five community state types identified. (B) Color bar indicating community classes designated DA, LC, LG, DB, and LI and as defined by clusters of proportions of community state types within a woman over time. (C) Profiles of community state types for 32 women over 16 weeks. Each dot (white or black) represents one sample in the time series. The absence of a dot indicates missing samples. The community state types in the time series for each woman are color-coded according to the schema shown below panel C (see also fig. S1). (D) Box plot of normalized Jensen-Shannon distances between all pairs of community states within each subject. Community deviation from constancy is represented by the Jensen-Shannon Index (black bar, see Supplementary Online Materials), with higher values reflecting decreased constancy. (E) Box plot of average Nugent scores for each women over 16 weeks. Panels D and E, the whiskers represent the lowest and highest datum still within 1.5 interquartile range (IQR) of the lower and upper quartile. The middle 50% of the data is represented by the height of the box.

Fig. 2

(A–D) Heatmaps (top) and interpolated bar plots (bottom) of phylotype relative abundance observed in four selected subjects over 16 weeks (heatmap color key is indicated in the lower right corner). Color codes for each phylotype represented in the interpolated bar plots are shown below the figure. See fig. S5 for heatmaps and interpolated bar plots for all subjects. Red dots below the interpolated bar graphs represent menstruation days.

Fig. 3

Temporal dynamics of vaginal bacterial communities in two women over 16 weeks. (A, C) Interpolated bar graph of phylotype relative abundance for subjects 13 (A) and 26 (C). Profiles of community state types in which Nugent scores have been superimposed (high Nugent score, 7–10, large open circles; intermediate Nugent score, 4–6, medium open circles; and low Nugent score, 1–3, small open circles) are shown below the interpolated bar graphs. Daily metadata is represented by the following: red balls, menstruation; black open square, douching; open red triangle, vaginal intercourse; open black diamond, oral sex; vertical black bar, digital penetration (insertion of finger(s) in the vagina); light blue closed circle, lubricant use. (B, D) Representation of vaginal community dynamics in 3D community space (3) for subjects 13 (B) and 26 (D). Communities dominated by species of Lactobacillus and representing community state types I (red), II (light blue), III (green), and V (purple) are shown at each of the four outer vertices of the tetrahedron, with community state type IV (yellow) at the inner vertex. The white line represents the succession of community states for each subject over time. Animations of these events are provided in the Supplementary Online Materials (movie S1 and movie S2).

Fig. 4

Modeling the dependence of the log of Jensen-Shannon divergence rate of change over the menstrual cycle. The length of menstrual cycles within and between women were normalized to 28 days (Supplementary Online Materials) with day 1 to 5 corresponding to menses (red bar). The red line shows concentrations of estradiol as a function of menstrual time (data from (21)) and the blue line shows concentrations of progesterone (data from (21)). The shaded areas around each curve show 95% point-wise confidence bands.

Fig. 5

Metabolomic analyses. ¹H-NMR metabolome profiles for 4 representative subjects at selected time points throughout the 16-week study (A–D). An interpolated bar graph of phylotype relative abundance in each community is shown above the ¹H-NMR profiles. The colored rectangles on top of the interpolated bar graph match the color of the ¹H-NMR profiles and indicate the times of collection during the 16-week study. ¹H-NMR spectra are spaced according to the matching time scale on the y-axis. Peaks for selected metabolites are labeled. Red arrow indicates lactic acid peaks, light blue arrow indicates succinate peak and green arrow indicates acetic acid peak. Red dots below the interpolated bar graphs represent menstruation days.