Yildirim et al
1
2
Supplemental Information
3
Supplementary Text S1
4
Our dataset included specimens from captive and wild species, including two
5
groups of blood-line related vervets (three to eight generations removed): one group is
6
maintained at the Wake Forest Primate Center (WFPC) and one group is free-ranging in
7
St. Kitts, respectively. The vervets housed at WFPC were fed a typical American diet
8
(TAD, lab diet 5LOP; Supplementary Figure S7) over a six-month period prior to
9
sampling. This gave us the unique opportunity to establish links among diet, captivity,
10
and vaginal microbiome composition. Despite these significant differences in diet and
11
environment, no significant difference was observed between these two groups with
12
regard to phylogenetic distribution (ANOSIM R=0.007, p=0.36; Figure 4A and 4B),
13
taxonomic composition (Figure 2), or species richness (Table S1).
14
We compared the number of shared or unique OTUs among captive and wild
15
representatives of vervets and baboons, excluding those reads that were observed less
16
than three times overall. 212 OTUs were shared among vervet representatives, more
17
than were unique to either group (wild specific = 97; captive specific = 210). Further,
18
after normalizing OTU abundances to read depth, there were no significant differences
19
in the relative abundances of any OTU between captives and wild vervets (p>0.01).
20
Captive and wild baboons shared a similar number of phyla (n=242) as seen with the
21
vervets but, in contrast, substantially more OTUs were unique to each group (wild
22
specific = 1343; captive specific = 705). Based on abundance, however, only five OTUs
23
were found to be significantly different between captive and wild baboons (p <0.01).
24
These results were not influenced by sampling depth, as no significant difference was
1
Yildirim et al
25
observed between captives and wilds for any baboon (p>0.01). Comparisons of the wild
26
and captive baboon representatives were confounded by species-distinctions and
27
substantially greater geographical separation (both latitudinal and longitudinal) relative
28
to the vervets.
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
2
Yildirim et al
Figure S1
Supplementary Figure S1. A heatmap of the relative abundance of 16S rRNA gene sequences displayed at the
taxonomic phylum level. The column z-score indicates differences between primate samples in terms of the relative
abundances of bacterial phylotypes associated with the primate samples; white color indicates relative abundance of
phylotypes having column average. Blue color tones represent relative abundances up to 4 standard deviation less than
the average abundance, thus these phylotypes are significantly underrepresented in the sample; and red color tones
representing relative abundances up to 4 standard deviation above the column average, ie, these phylotypes are
significantly enriched in the sample. Samples and bacterial phylotypes were clustered using average linkage hierarchical
clustering of a distance matrix based on Bray-Curtis distance.
3
Yildirim et al
Figure S2
Supplementary Figure S2. Clustering of pairwise binary Jaccard similarities of genus level abundance
distributions between samples. Unweighted Pair Group Method with Arithmetic Mean (UPGMA) method was used in
cluster analysis. Genus level percent abundances in Asymptomatic BV (ABV) and symptomatic BV (SBV) samples were
recently published (Ravel et al. 2013); species level assignments were binned in genus level taxonomic hierarchy (genus
level relative abundances can be found in supplementary Tables S5)
4
Yildirim et al
Figure S3
Supplementary Figure S3. Clustering of pairwise binary Bray-Curtis similarities of genus level abundance
distributions between samples. Unweighted Pair Group Method with Arithmetic Mean (UPGMA) method was used in
cluster analysis. Genus level percent abundances in Asymptomatic BV (ABV) and symptomatic BV (SBV) samples were
recently published (Ravel et al. 2013); species level assignments were binned in genus level taxonomic hierarchy (genus
level relative abundances can be found in supplementary Tables S5
5
Yildirim et al
Figure S4
Supplementary Figure S4. Kernel density plot of Pearson Correlations of phylogenetic
distance matrices obtained from primate host and vaginal microbiome (>450,000
iterations).
6
Yildirim et al
Figures S5A-S5J
7
Yildirim et al
8
Yildirim et al
Supplementary Figure S5A to 5J. Effect of socio-ecological factors on diversity of
vaginal microbiome estimated by Shannon index. Sex related factors such as female
promiscuity, mating group size and male testes mass significantly contribute to
increasing diversity (p < 4.356x10-05, p < 2.976x10-06, p < 4.4x10-04, respectively). In
contrast, group size and gestation time were found to be associated with decreased
diversity (p < 2x10-2, and p < 7.7x10-3, respectively). A- Baculum length; B- Body size;
C- Gestation time; D-Group size; E- Home range; F- Mating group size; G-Neonatal
size; H- Promiscuity, I- Swelling; J- Testes mass
9
Yildirim et al
Figure S6
Supplementary Figure S6. Variation in beta diversity among NHPs and humans
(measured as the average Bray-Curtis dissimilarity from individual samples to their
group centroid). Bw-Yellow Baboons, Bc-Olive Baboons (captive), Ch-Chimpanzees,
Hm-Humans, Lm-Lemurs, Hw-Black howler monkeys, Mb-Mangabeys (captive), RCRed Colobus, Vc-Vervets captive, Vw-Vervets wild
10
Yildirim et al
Figure S7.
Supplementary Figure S7. Ingredients in Typical American Diet (TAD) used to feed
captive primates. The diet mimics a typical American diet by containing a mixture of
animal- and plant-derived protein sources, unsaturated and saturated vegetable and
animal-derived lipids, and as a mixture of simple and complex carbohydrates.
11