This Provisional PDF corresponds to the article as it appeared upon acceptance. Fully formatted
PDF and full text (HTML) versions will be made available soon.
Pyrosequencing of 16S rRNA genes in fecal samples reveals high diversity of
hindgut microflora in horses and potential links to chronic laminitis
BMC Veterinary Research 2012, 8:231 doi:10.1186/1746-6148-8-231
Samantha M Steelman (firstname.lastname@example.org)
Bhanu P Chowdhary (email@example.com)
Scot Dowd (firstname.lastname@example.org)
Jan Suchodolski (email@example.com)
Jan E Jane¿ka (firstname.lastname@example.org)
16 June 2012
18 November 2012
27 November 2012
Like all articles in BMC journals, this peer-reviewed article can be downloaded, printed and
distributed freely for any purposes (see copyright notice below).
Articles in BMC journals are listed in PubMed and archived at PubMed Central.
For information about publishing your research in BMC journals or any BioMed Central journal, go to
BMC Veterinary Research
© 2012 Steelman et al.
This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Pyrosequencing of 16S rRNA genes in fecal samples
reveals high diversity of hindgut microflora in
horses and potential links to chronic laminitis
Samantha M Steelman1
Bhanu P Chowdhary1
Jan E Janečka1*
* Corresponding author
1 Department of Veterinary Integrative Biosciences, Texas A&M University,
College Station, TX 77843-4458, USA
2 Molecular Research LP, Shallowater 79363, TX, USA
3 Gastrointestinal Laboratory, Texas A&M University, College Station, TX
The nutrition and health of horses is closely tied to their gastrointestinal microflora. Gut
bacteria break down plant structural carbohydrates and produce volatile fatty acids, which are
a major source of energy for horses. Bacterial communities are also essential for maintaining
gut homeostasis and have been hypothesized to contribute to various diseases including
laminitis. We performed pyrosequencing of 16S rRNA bacterial genes isolated from fecal
material to characterize hindgut bacterial communities in healthy horses and those with
Fecal samples were collected from 10 normal horses and 8 horses with chronic laminitis.
Genomic DNA was extracted and the V4-V5 segment of the 16S rRNA gene was PCR
amplified and sequenced on the 454 platform generating a mean of 2,425 reads per sample
after quality trimming. The bacterial communities were dominated by Firmicutes (69.21%
control, 56.72% laminitis) and Verrucomicrobia (18.13% control, 27.63% laminitis),
followed by Bacteroidetes, Proteobacteria, and Spirochaetes. We observed more OTUs per
individual in the laminitis group than the control group (419.6 and 355.2, respectively, P =
0.019) along with a difference in the abundance of two unassigned Clostridiales genera (P =
0.03 and P = 0.01). The most abundant bacteria were Streptococcus spp., Clostridium spp.,
and Treponema spp.; along with unassigned genera from Subdivision 5 of Verrucomicrobia,
Ruminococcaceae, and Clostridiaceae, which together constituted ~ 80% of all OTUs. There
was a high level of individual variation across all taxonomic ranks.
Our exploration of the equine fecal microflora revealed higher bacterial diversity in horses
with chronic laminitis and identification of two Clostridiales genera that differed in
abundance from control horses. There was large individual variation in bacterial communities
that was not explained in our study. The core hindgut microflora was dominated by
Streptococcus spp., several cellulytic genera, and a large proportion of uncharacterized OTUs
that warrant further investigation regarding their function. Our data provide a foundation for
future investigations of hindgut bacterial factors that may influence the development and
progression of chronic laminitis.
The microflora within the gastrointestinal system directly affects energy metabolism,
digestive function, mucosal immune system development, and disease pathogenesis of its
eukaryotic host [1-4]. This is particularly true for herbivores, including the horse, which are
dependent upon fermentation by bacteria to utilize plant structural carbohydrates .
Therefore, a detailed knowledge of gut microflora is essential for understanding the
nutritional needs of horses and the contribution of gut homeostasis to equine health. Research
on bacterial communities has recently flourished with the application of next-generation
sequencing (NGS) technology . Studies incorporating NGS have led to the discovery of
thousands of novel species (i.e., Operational Taxonomic Units [OTUs]) and elucidation of
their ecological function within the gut of vertebrates [5-7]. Numerous factors including the
evolutionary history of the host, age, and diet influence the diversity of gut microbes; they in
turn have been implicated in a broad range of disorders including Crohn’s disease, chronic
diarrhea, inflammatory bowel disease, type I diabetes, obesity, and asthma [2,7-9].
Alterations in hindgut bacterial communities have also been associated with several equine
diseases [10-17]. Excess nonstructural carbohydrates (i.e., starches, fructans, or simple
sugars) that are not digested in the foregut enter the cecum and colon, where bacterial
fermentation produces byproducts including lactic acid and gas, which can cause colic
[4,16,17]. The same initiators can also lead to the development of laminitis, which often
occurs subsequent to overconsumption of grain or after feeding on lush pasture rich with
nonstructural carbohydrates [18-20]. Starch and oligofructose overload-induced models have
revealed strong associations between onset of laminitis and proliferation of Streptococcus and
Lactobacillus bacteria, with a concurrent decrease in intraluminal pH [12-15,21-23].
Numerous studies have characterized and enumerated bacteria of the equine hindgut,
primarily relying on culturing of bacteria, clone-based sequencing of Polymerase Chain
Reaction (PCR) amplicons, denaturing gradient gel electrophoresis (DGGE), fluorescence in
situ hybridization (FISH), or gene terminal restriction fragment length polymorphism (T-
RFLP) [21,23-32]. The primary microbes detected consisted of Gram-positive bacteria, many
of which were associated with the cluster XIVa of Clostridiaceae, Streptococcus spp., and
Lactobacillus spp. [14,22,24,26]. Up to 96% of all observed OTUs could not be assigned,
highlighting how little was known about this ecosystem .
Recently researchers have begun to apply 454 sequencing of 16S rRNA amplicons to
understand the equine gut microflora [10,33,34]. A total of 1,518 OTUs have been observed
in feces from just two horses, with Firmicutes, Verrucomicrobia, and Proteobacteria being the
most abundant Phyla, and Subdivision 5 Incertae sedis spp., TM7 Incertae sedis spp., and
Treponema spp. the most common genera . In a study examining colitis, Firmicutes were
found to dominate the feces of normal horses in contrast to Bacteroidetes in horses with
undifferentiated colitis . Bacterial communities in the stomach were also found to be
dominated by the Phyla Firmicutes, Proteobacteria, and Bacteroidetes, with Lactobacillus
spp., Streptococcus spp., and Moraxella spp. comprising the most abundant genera . The
stomach microflora segregated based on management (stabled versus pastured) and sampling
methods (biopsy versus post mortem) . These studies show a much more diverse
assembly of bacteria than previously described; however, the mechanisms linking bacterial
diversity to diseases such as colic, colitis, and laminitis are yet to be elucidated.
We thus explored the equine hindgut microflora by pyrosequencing bacterial 16S rRNA gene
segments present in feces of normal horses and those suffering from chronic laminitis. Our
goals were to (1) describe the level of microbial diversity and (2) compare the microflora of
healthy horses to those with chronic laminitis. We hypothesized that horses with chronic
laminitis, which had in the past experienced a bout of acute laminitis and presumably a
radical shift in bacterial flora that accompanies this disease, would harbor a different
microbial population. Our study contributes to the characterization of the equine gut
microbiome and its potential link to laminitis.
Sequencing depth and alpha diversity
The mean number of reads per sample was 5,159 (range 1,173 – 33,204). When we removed
two outliers with the highest depth (12,113 and 30,911 reads) the mean dropped to 2,425
(range 1,032 – 6,578). The 16S rRNA sequences were deposited in the NCBI Sequence Read
Archive under the Metagenome BioProject SUB135964. One of the horses was a pony and
another had recently been on antibiotics so these were not included in the study groups. We
separated the horses into 2 groups; those that did not have any history of laminitis (control, n
= 9) and those that had chronic laminitis (laminitis, n = 7). We detected a total of 4,894 OTUs
in fecal samples from all horses. Of these, 34% (1,660) were identified as chimeras by
DECIPHER and excluded from downstream analysis leaving 3,234 OTUs . After the
chimeras were removed the mean sequences per sample dropped to 2,204. The laminitis
group had a greater number of OTUs per horse than the control group (mean = 419.6 versus
355.2, respectively, P = 0.019) (Table 1). The rarefaction curve of observed OTUs did not
plateau with increasing reads suggesting that a higher number of reads per sample would
have provided a more comprehensive catalog of bacterial taxa (Figure 1A). However, the
Chao1 index of bacterial richness did start to plateau at ~600 reads indicating that the main
components of community diversity were detected with our level of depth (Figure 1B). The
Chao1 was significantly different between control and laminitis groups (P = 0.020) (Table 1).
Table 1 Diversity indices of the gastrointestinal microflora in horses
97% Similarity OTUs
OTUs per animal 381.7 (SE = 26.6)
OTUs (rfd)* 338.8.0 (SE = 52.3) 330.4 (SE = 57.5)
Chao1 795.7 (SE = 51.7)
Chao1 (rfd)* 761.3 (SE = 82.2)
Phylogenetic Distance 16.15 (SE = 0.82)
Phylogenetic Distance (rfd) 14.85 (SE = 1.50)
Shannon 5.07 (SE = 0.59)
Simpson 0.74 (SE = 0.08)
Mean per animal 53.4 ± 6.4
The OTUs (97% similarity), Chao1, Phylogenetic Distance, Shannon, and Simpson diversity
indices were estimated in QIIME. Rarified OTU, Chao1, and Phylogenetic Distance estimates
were rarefied (rfd) to a depth of 1,100 reads to reduce sampling heterogeneity. * Significant
difference, (OTUs, P = 0.019; Chao1, P = 0.020)
355.2 (SE = 26.3) 419.6 (SE = 49.3)
369.5 (SE = 51.8)
872.6 (SE = 77.4) 741.9 (SE = 66.2)
735.7 (SE = 123.4) 816.4 (SE = 65.5)
15.37 (SE = 0.85)
14.43 (SE = 1.05)
5.01 (SE = 0.71)
0.75 (SE = 0.09)
17.26 (SE = 1.50)
16.00 (SE = 2.00)
5.16 (SE = 1.06)
0.72 (SE = 0.14)
52.3 ± 6.4
55.0 ± 6.6
Figure 1 Observed bacterial OTUs and Chao1 index plots. Plots were made using data
rarefied to a depth of 1,200 reads per sample in QIIME. (A) Mean OTUs, (B) Chao1 index
The majority of OTUs belonged to Firmicutes (69.21% control, 56.72% laminitis) (Figure
2A). Verrucomicrobia was next most abundant (18.13% control, 27.63% laminitis) followed
by Bacteroidetes (5.71% control, 9.94% laminitis). The remaining 6.95% of the equine
bacterial population was either Spirochaetes (2.52%), Proteobacteria (0.95%), or belonged to
one of 11 other Phyla (0.13%). Firmicutes was always the most abundant. Verrucomicrobia
was the second most abundant in all of the horses, except for the pony and the horse that had
received antibiotics within the last 2 weeks. In both these animals, Proteobacteria was the
second most abundant Phylum.
Figure 2 The abundance of bacterial taxonomic groups in horses. The mean percentage of
reads assigned to the respective taxonomic group for control and laminitis groups.
Taxonomic assignments were based on 16S rRNA sequences using the Ribosomal Database
Project classifier in QIIME. (A) Phylum, (B) Class, (C) Order, (D) Family, (E) Genus.
*Significant difference between groups (P = 0.03 and 0.01, respectively)
Twenty-nine bacterial Classes were observed in horse feces; only 8 of these contained >1%
of all OTUs. Clostridia of the Firmicutes Phylum was the most abundant (41.75% control,
38.15% laminitis) (Figure 2B). In the controls, the second most numerous was Bacilli
(26.65%), also in Firmicutes, and third was Subdivision 5 of Verrucomicrobia (Verruco-5)
(16.81%); however, this trend was reversed in the laminitis group (17.44% Bacilli and
25.02% Verruco-5). In all but one horse, either Bacilli or Clostridia were the most common.
However, large individual variation in abundance was observed; Bacilli varied from 1.11% to
93.57%, Clostridia from 2.22% to 47.42%, and Verruco-5 from 0.86% to 38.19%.
A total of 44 Orders were detected, however, 82% of all OTUs belonged to only 3 of them;
Clostridiales (41.63% control, 37.99% laminitis), Lactobacillales (25.10% control, 16.86%
laminitis), and RFP12 of Verruco-5 (16.79% control, 25.02% laminitis) (Figure 2C).
Additional Orders with a frequency greater than 1% included Bacteroidales, Spirochaetales,
Bacillales, and Verrucomicrobiales. The most common among individuals was either
Lactobacillales (in 4 control and 3 laminitis horses) or Clostridiales (in 6 control and 4
laminitis horses). There were large amounts of individual variation in abundance of Orders
(e.g., Lactobacillales ranging from 3.18% to 93.53%). The Order Burkholderiales of
Proteobacteria was the third most common in the pony (14.64%), yet it was observed in only
two other horses at a low frequency (< 0.05%).
Eighty-two Families were detected among all horses with the most dominant being
Streptococcaceae (24.17% control, 16.11% laminitis), followed by an unassigned Family in
the RFP12 Order of Verrucomicrobia (16.79% control, 25.02% laminitis), and
Ruminococcaceae (13.15% control, 14.40% laminitis) (Figure 2D). Additional abundant
Families included Clostridiaceae, Lachnospiraceae, unassigned Bacteroidales, 2 unassigned
Clostridiales, Spirochaetaceae, Verrucomicrobiaceae, and Clostridiales Family XIII Incertae
sedi. Ninety percent of all OTUs where attributed to these 9 Families. There were also large
amounts of individual variation at this taxonomic level, with the abundance of the unassigned
RFP12 Family ranging from 1.1% to 38.19% and Streptococcaceae from 0.40% to 74.99%.
Only 0.80% of all reads were attributed to Lactobacillaceae.
A total of 140 genera were identified among the 18 sampled horses, with an average of 45.94
± 1.89 SE per horse. Nineteen genera were found in > 87% of horses; of these 11 were not
assigned to any previously described genus (Table 2). Majority of the genera were observed
in only a few of the horses. Eighty-nine were present in less than 20% of the individuals and
58 were detected in only one animal.
Table 2 The percentage of OTUs that were assigned to the 20 most abundant microbial
with Genus Present
Percent of all Horses Pooled
N = 13
N = 7
10.03%* 0.50% 0.07%
3.96% 0.64% 7.73%
5.30% 2.87% 3.88%
N = 5
9.03% 16.03% 8.71%
5.32% 25.02% 4.01%
Clostridiales XIII I. sedis (F)
The OTU assignments were made using the Ribosomal Database Project classifier in QIIME.
The mean and standard error (S.E.) of the percentage of reads that map to the respective
genus is provided. If the OTU was not assigned to a known genus its nearest available
taxonomic rank is provided. F = Family, O = Order, P = Phylum. *Significant differences
between groups (P = 0.03 and 0.01, respectively).
The dominant genera where Streptococcus (21.00% control, 16.03% laminitis), an unassigned
genus in the RFP12 Order of Verruco-5 (16.79% control, 25.02% laminitis), and an
unassigned genus in the Ruminococcaceae family (8.30% control, 9.99% laminitis) (Table 2,
Figure 2E). Streptococcus was the most abundant genus in 6 control horses and 3 laminitic
horses, while the RFP12 genus dominated most of the other horses. Differences in abundance
for the top three genera between control and laminitis groups were not significant (P > 0.21).
Twelve of the 20 most abundant genera were unassigned. Among the classified dominant
Ruminococcus, Lactobacillus, Staphylococcus, and Coprococcus. There were significantly
more OTUs attributed to two unassigned Clostridiales genera in the laminitis group compared
to the control (P = 0.03 and P = 0.01) (Table 2). Similar to all other levels of classification,
there was large individual variation in the abundance of the dominant genera; for example,
Streptococcus varied from 0.40% to 91.96%, the FRP12 genus from 2.78% to 32.60%, and
the Ruminococcaceae genus from 0.36% to 15.72%.
The short 16S sequences (< 500 bp) generated during this study did not permit reliable
species-level assignments. Nonetheless, we examined OTUs with greater than 1% abundance
to determine which described species they are most closely related to. We detected the
following taxa: Streptococcus equinus serotype 3, Rhodococcus wratislaviensis oucz59,
Prevotella ruminicola, Clostridium sardiniense, Williamsia muralis, Clostridium
chartatabidum, Clostridium orbiscindens. Clostridium had the highest number of species
relative to other genera (33).
Statistical tests dependent on taxonomic categories often fail to detect community levels
differences in diversity . Approaches that are independent of OTU assignments have thus
been developed for comparing microbiomes . We tested the control and laminitis groups
for community shifts in the microflora using UniFrac distance, which compares the
phylogenetic diversity within groups and is independent of taxonomic classification . To
visualize the differences between groups we conducted Principle Coordinate Analysis
(PCoA) of weighted and unweighted UniFrac distances and plotted the 3 factors that
explained the greatest portion of variation. Jacknifed weighted and unweighted UniFrac
distances did not show any significant differences between the two groups (Figure 3A,B,C).
Figure 3 Principle coordinate analysis of unweighted UniFrac distances. Principle
coordinate analysis (PCoA) plots were made using jackknifed UniFrac distances in QIIME.
Red data points represent control horses and the blue horses with chronic laminitis. (A) PC1
versus PC2, (B) PC1 versus PC3, (C) PC2 versus PC3
We observed more unique OTUs (3,234) than detected by Shepherd et al.  (1,510 OTUs)
despite our lower read depth (2,204 versus 28,458, respectively). This is likely because we
had a greater number of horses (16 versus 2). However, our Chao1 index of bacterial richness
(795.7) and Shannon Index of bacterial diversity (5.07) were lower than in the previous study
(2,359 and 6.7, respectively) . There was significantly higher bacterial diversity as
estimated from OTUs and the Chao1 index in the laminitis group compared to the control (P
= 0.019, P = 0.020, respectively). The only other significant differences between the control
and laminitis groups was the higher abundance of two undescribed genera of Clostridiales in
the laminar horses (P = 0.03 and P = 0.01, respectively). This suggests potential changes in
bacterial communities that should be further explored.
Our lower bacterial richness and diversity relative to what was previously reported could be
attributed to an insufficient number of reads to capture all of the diversity within each sample,
particularly for the low abundance OTUs . This is supported by our OTU rarefaction plot
that fails to plateau (Figure 1A). Future studies need to generate closer to the 5,000 reads per
sample previously recommended . We targeted this level of depth; however, because one
of our samples was over-represented (28,458) in the pooled multiplex of amplicons, it
reduced the number of reads that were generated for the other samples. Therefore, greater
attention needs to be given to DNA extraction, PCR amplification, and library construction so
that each amplicon is equally represented.
We successfully assigned a greater number of reads to Phyla (98.42% versus < 62%) than
several previous studies using 16S rRNA sequences [24,33]. This is likely because they did
not identify and exclude chimeras, which are known to inflate the number of unclassified
OTUs . We detected the same number of Phyla (16) as in Shepherd et al., including 4
that were not previously observed in horses; MVP-15, Synergistetes, Chlamydiae, and
Deferribacteres [10,33]. The most abundant Phylum we observed in horses, Firmicutes, was
also the major component of equine intestinal flora in previous studies that analyzed feces
from two adult Arabian geldings  and 6 healthy horses , stomach contents from 9
hay-fed stabled horses , and more traditional studies that used clone-based Sanger
sequencing [24,31]. Firmicutes are also common in the gut of other diverse taxa, from cats,
dogs, and polar bears to cattle [38-40]. In contrast, Bacteroidetes was the most abundant
Phylum among horses that had colitis, supporting the hypothesis that Firmicutes play an
important role in gut function .
Verrucomicrobia, Bacteroidetes, and Proteobacteria represented the next largest components
of the equine gut microbiome that we observed; a pattern similar to previous studies,
although the Phyla were not always in the same order [10,31,33]. We detected higher levels
of Verrucomicrobia than previously reported (21.78% versus < 5%) [10,24]. The abundance
of this Phylum in horses from central Texas suggests it plays a more important role in hindgut
function than previously appreciated. Our second most abundant genus among all horses was
an unknown type within the RFP12 Order of Verrucomicrobia. This is a good candidate for
culturing in order to classify it and characterize this taxa metabolic function.
The cecum and colon of the horse are important for the breakdown of structural
carbohydrates and production of volatile fatty acids . Therefore, we expected to detect
bacteria known to play such a role, these include Ruminococcus spp., Fibrobacter spp.,
Eubacterium spp., and Treponema spp. [24,41]. We indeed detected all of the above;
Ruminococcus had a mean of 1.03%, Fibrobacter 0.042%, Eubacterium 0.004%, and
Treponema 2.18%. Our values were consistent with what has been previously observed
(0.50% – 4.4%, 0.01% – 0.75%, 0.09%, 1.90% – 3.00%, respectively) [10,24,33,41].
Interestingly, among the most abundant were unassigned genera of Ruminococcaceae that
together composed 8.75% of all OTUs. These may represent important uncharacterized
cellulytic bacteria and warrant further investigation.
A vast amount of individual variation was observed in horses at all taxonomic levels. A large
portion of this likely came from environmental heterogeneity and differences in animal
history, combined with lack of sequencing depth. However, similar individual variation in the
equine gut microflora was previously observed, for example, in a study that had a mean of
4,712 reads per sample Bacteroidetes varied from 9.0% to 21.3% and Proteobacteria from
0.0% to 42.7% . Such large individual variation may be a natural trait of equine gut
communities; however, the lack of detailed studies using a large number of horse samples
limits the inferences that can be made from these patterns.
The genera previously found dominating the lower intestinal microflora in two Arabian
geldings based on 16S rRNA pyrosequencing of fecal samples included Blautia spp.,
Fibrobacter spp., Subdivision 5 Incertae sedis spp., TM7 Incertae sedis spp., Treponema
spp., and Ruminococcus spp. . In contrast, fecal analysis of a more diverse group of
horses found the primary genera Clostridium spp., Coptotermes spp., Enterococcus spp.,
Fusobacterium spp., Porphyromonas spp., Pseudomonas spp., and Prevotella spp. . A
study that used intestinal samples detected many unassigned genera affiliated with
Clostridium spp., Butyrivibrio spp., Ruminococcus spp., and Eubacterium spp. . We
detected all of the above except Coptotermes spp., Porphyromonas spp., and Pseudomonas
spp.. Among the 14 genera that we observed with > 1.0% abundance were Streptococcus
spp., Akkermansia spp., and Oscillospira spp., and 8 genera that could not be assigned to any
described genus. This large proportion of unassigned genera among highly abundant OTUs
highlights the need for more traditional studies characterizing bacteria and their phenotypic
traits to better understand the function of the equine hindgut microflora.
Within abundant genera we found evidence suggesting additional diversity. The most diverse
genus was Clostridium, which exhibits a wide range of functions and contains both beneficial
and pathogenic representatives . For example, C. botulinum causes botulism as well as
productivity problems and C. difficile leads to severe diarrhea and colitis in both humans and
livestock [43,44]. Yet, many Clostridium spp. are cellulytic and important for the digestion of
plant material [45,46]. We detected 33 species of Clostridium, including C. botulinum in one
horse. The population dynamics of bacterial species and their interactions can influence
normal gut function and the development of diseases . It is possible that some of the
bacterial shifts that affect disease states such as laminitis occur at the species level.
There are numerous lines of evidence suggesting hindgut microflora play a role in the
development of laminitis. Several studies have examined the bacterial response during
various experimental laminitis models [12-14,21-23]. An estimated 53% of acute laminitis
cases occur after overconsumption of grain or grass rich with nonstructural carbohydrates
(i.e., starch, fructans, or simple sugars) , which is also associated with an explosive
proliferation of Streptococcus spp. and Lactobacillus spp. in the cecum and a concurrent
decrease in the intraluminal pH [1,12,15]. Potentially, either of these may be a factor in
laminitis. We found remarkable variation in Streptococcus spp. among healthy horses (0.40%
to 91.96% of all OTUs); therefore the absolute abundance of Streptococcus spp. might not be
important relative to other changes disrupting hindgut equilibrium.
In the carbohydrate overload model of laminitis, Garner et al.  found that Lactobacillus
spp. increased in abundance by a factor of 105. These changes led to decreased intraluminal
pH through the production of lactic acid, which caused death and lysis of other bacterial
species including Enterobacteriaceae spp. and Bacilli spp. . Garner hypothesized that
these release endotoxins and cause mucosal damage, contributing to the development of
laminitis. Endotoxins can escape into the bloodstream and cause immune system activation,
inflammation, fever, low blood pressure, and high respiration rate; some of these symptoms
appear during the early stages of laminitis [47-49]. We found Lactobacillus spp. represented
a small portion of the bacterial communities in the horses we sampled (0.82% controls,
0.60% laminitis). However, we only obtained samples from horses that had a previous history
of this condition and not immediately after a relapse of laminitis. Therefore we would not
have detected any previous transient Lactobacillus spp. proliferation. In addition, we sampled
the microflora using feces, an approach which could potentially mask changes occurring in
the stomach, cecum, and upper colon . The affects Lactobacillus spp. and Streptococcus
spp. proliferation has on the equine gut microbiome following an increase in dietary
nonstructural carbohydrates and relapse of chronic laminitis should be explored.
The composition of the hindgut microflora also has large impacts on feed digestibility and
equine nutrition because the horse depends upon microbial fermentation to digest plant
structural carbohydrates . Similar to previous studies we found that majority of the
abundant bacterial genera were anaerobic fermenters, suggesting that the hindgut microflora
are specialized for breaking down plant material. Alterations to bacterial communities may
confer advantages to horses under certain dietary conditions . For example, gradual
addition of grain into the diet increases the ratio of propionate to acetate, presumably by
altering the bacterial microflora . Propionate can be directly converted to glucose and
thus this shift is beneficial for horses with high energy needs . However, grain also has
more simple sugars, which increase the risk of colic and laminitis [13,17,21]. Future studies
should explore how bacterial diversity and function can mediate adaptation to high-energy
diets and reduce disease risks.
Our exploration of the equine hindgut microflora revealed higher levels of bacterial diversity
in horses with chronic laminitis and identification of two Clostridiales genera that differed in
abundance from control horses. We observed large individual variation suggesting that
bacterial populations may be influenced by factors such as genetic background, age, diet,
feeding time, and body condition, which were not taken into account during this study. There
was high abundance of cellulytic bacteria, primarily Ruminococcaceae and Clostridiaceae.
We observed numerous abundant uncharacterized genera within Subdivision 5 of
Verrucomicrobia, Clostridiales, and Ruminococcaceae that warrant further investigation into
their function. Vast individual differences in Streptococcus abundance among healthy horses
suggested that this genus is likely not closely linked with chronic laminitis. We recommend
studies make efforts to reduce experimental variation by using more homogenous horse
populations and incorporating rigorous normalization during 454 library construction to
increase the sensitivity for biologically-relevant changes in bacterial communities.
Fresh fecal samples were collected from 10 healthy horses and 7 horses and 1 pony with
chronic laminitis. One of the control horses had received antibiotics 2 weeks prior to
sampling and therefore we did not include it in our control group. We also excluded the pony
from the chronic laminitis group to reduce potential breed-specific differences. Horses were
diagnosed as having chronic laminitis by a licensed veterinarian based on clinical
presentation, case history, and radiographic evidence of dorsopalmar rotation of the distal
phalanx. Horses were kept on two different farms in Brazos County, Texas. All horses had
been resident at their respective farms for at least 6 months and had not experienced any
recent changes in diet or housing conditions. All animals were maintained on a pelleted
concentrate feed containing either 12% or 16% crude protein (Producer’s Co-op, Bryan, TX)
in addition to coastal bermudagrass hay and limited amount of alfalfa hay. Detailed
information about horses and concentrate feed composition may be found in Additional file
1: Tables S1 and S2. Horses were kept in stalls, large dry lots, or a combination of both. All
horses had ad libitum access to water. Only naturally voided fecal samples were collected and
therefore did not require an IACUC Animal Use Permit. A single sample was collected from
each horse within 3 hours after the morning feeding. As feed takes approximately 28–46
hours to travel through the digestive tracks of horses, the microflora sampled from feces
would not be influenced by feeding just prior to collection. Every attempt was made to collect
samples immediately after defecation. After collection, samples were stored on wet ice for
transport to the laboratory and frozen at −20°C.
DNA extraction and pyrosequencing
The DNA was extracted from feces using the phenol:chloroform:isoamyl alcohol method
after disrupting the starting material with bead beating as described in Suchodolski et al..
The V5-V9 region of 16S rRNA gene was pyrosequenced on the Roche 454 FLX-Titanium
instrument (Roche Applied Science, Indianapolis, IN) by the Research and Testing
Laboratory (Lubbock, TX) as previously described, with Titanium chemistry modifications
[38,53]. Briefly, a 570-bp segment of 16S rRNA was PCR amplified using the HotStarTaq
Plus Master Mix Kit (Qiagen, Valencia, CA), 100 ng of template DNA, and universal
Eubacterial primers that target majority of GI bacteria: 939 F-TTGACGGGGGCCCGCAC
and 1492R-TACCTTGTTACGACTT [54,55]. The exact span of the amplicon in relation to
Streptococcus equinus strain ATCC 9812 16S rRNA complete sequence is 823 bp to 1409 bp
(GenBank Accession NR_042052.1). The thermal conditions were 94°C denaturation for 3
min, 32 cycles of 94°C for 30 sec, 60°C for 40 sec, 72°C for 1 min, and a final 5-min
elongation step at 72°C. Subsequently, a second PCR was performed on the above PCR
products using the same conditions, but with modified fusing primers that had tag sequences
added on the 5’ ends (i.e., LinkerA-Tags-939 F and LinkerB-1492R) to enable multiplexed
454 FLX amplicon pyrosequencing. This secondary PCR was used to incorporate tags and
linkers into the 16S rRNA amplicons to avoid unbalanced amplification from the DNA
samples. The final amplicons from different samples were mixed in equal volumes, purified
using Agencourt AMPure XP beads (Agencourt Bioscience Corporation, Danvers), and
sequenced on the 454 platform [38,53].
Species-level operational taxonomic unit (OTU) assignments (>97% similarity, equal to
number of matching nucleotides divided by the length of the shorter sequence) were
made after trimming positions with < Q25 quality score and discarding reads < 200 bp [56-
58]. Sequences were depleted of chimeras and assignments to putative species (>97%
similarity) were done with BlastN  against a manually curated database compiled from
NCBI by the Research and Testing Laboratory (Lubbock, TX) . However, because
species-level bacterial assignments using short, single-gene segments are not robust we only
used this information to obtain an overview of the potential species present. The main
comparisons of microbial diversity within and among horses were made for genus and
higher-level classifications as described below.
The statistical analysis of alpha and beta diversity was done from taxonomic classifications
and phylogenetic-based methods (UniFrac) not dependent on OTU assignments . The
QIIME pipeline with standard scripts and default settings was used for taxa assignments
(genus and higher), diversity estimates (OTUs, Chao1 index, phylogenetic distance index,
Shannon index, and Simpson index), and phylogeny-based analyses using UniFrac
[37,59,60]. Barcodes were removed and the reads trimmed of bases with quality score below
Q25; reads with length < 200 bp or any ambiguous bases were removed from dataset. The
remaining sequences were clustered using UCLUST with the furthest algorithm based on
>97% similarity to define OTUs . Representative sequences were selected with the most
abundant criteria. Chimeras were identified among the OTUs using DECIPHER and excluded
from all subsequent analysis . Taxonomic assignments of the OTUs were made down to
the genus level (>95% similarity) using the Ribosomal Database Project (RDP) classifier and
the Greengenes reference core set “gg97_otus_4feb2011_aligned.fasta” available from
http://greengenes.lbl.gov/cgi-bin/nph-index.cgi [61,62]. Sequences were added to this
reference alignment  with PyNAST and the alignment was optimized, then filtered to
exclude sites with only gaps and excessively variable sites [62,63]. A neighbor-joining
phylogeny was reconstructed using FastTree for UniFrac analysis of beta diversity .
Rarefied OTU tables were generated to reduce sampling heterogeneity for observed OTUs
and Chao1 index and tested for significant differences in QIIME. Unpaired t-tests were used
to compare abundance of taxonomic groups between control and laminitis groups. Beta
diversity was compared between control and laminitis groups using weighted and unweighted
UniFrac phylogenetic-based distances. Principal Coordinate Analysis (PCoA) transformed
the UniFrac distances into coordinates that explain the greatest amount of variation. The
differences were visualized with 2D and 3D PCoA plots. To minimize sampling bias we
rarified OTU matrices using the smallest number of reads observed in any one horse before
conducting the UniFrac and PCoA analyses.
The authors declare that they have no competing interests.
SMS collected samples, performed some of the data analysis, and drafted manuscript. BPC
assisted in experimental design, data interpretation, and manuscript preparation. SD
performed PCR amplification, pyrosequencing, and data analysis. JS performed DNA
extractions and contributed to development of project. JEJ developed project, performed data
analysis, and drafted manuscript. All authors read and approved the final manuscript.
The authors wish to thank Dr. David Hood of the Hoof Diagnostic and Rehabilitation Clinic
for providing access to samples. The study described herein was supported by funds from the
United States Department of Agriculture (award # 2011-67012-30685 to SMS and 2010-
65205-20446 to BPC) and the LINK Endowment.
1. O'Hara AM, Shanahan F: The gut flora as a forgotten organ. Embo Reports 2006,
2. Neish AS: Microbes in Gastrointestinal Health and Disease. Gastroenterology 2009,
3. Wardwell LH, Huttenhower C, Garrett WS: Current concepts of the intestinal
microbiota and the pathogenesis of infection. Curr Infect Dis Rep 2011, 13:28–34.
4. Hintz HF, Cymbaluk NF: Nutrition of the Horse. Ann Rev Nutr 1994, 14:243–267.
5. MacLean D, Jones JDG, Studholme DJ: Application of 'next-generation' sequencing
technologies to microbial genetics. Nature Rev Microbiol 2009, 7(4):287–296.
6. Muegge BD, Kuczynski J, Knights D, Clemente JC, Gonzalez A, Fontana L, Henrissat B,
Knight R, Gordon JI: Diet drives convergence in gut microbiome functions across
mammalian phylogeny and within humans. Sci 2011, 332(6032):970–974.
7. Turnbaugh PJ, Hamady M, Yatsunenko T, Cantarel BL, Duncan A, Ley RE, Sogin ML,
Jones WJ, Roe BA, Affourtit JP, et al: A core gut microbiome in obese and lean twins.
Nature 2009, 457(7228):480–U487.
8. Serino M, Luche E, Chabo C, Amar J, Burcelin R: Intestinal microflora and metabolic
diseases. Diabetes Metab 2009, 35(4):262–272.
9. Suchodolski JS, Xenoulis PG, Paddock CG, Steiner JM, Jergens AE: Molecular analysis
of the bacterial microbiota in duodenal biopsies from dogs with idiopathic
inflammatory bowel disease. Vet Microbiol 2010, 142(3–4):394–400.
10. Costa MC, Arroyo LG, Allen-Vercoe E, Stampfli HR, Kim PK, Sturgeon A, Weese JS:
Comparison of the fecal microbiota of healthy horses and horses with colitis by high
throughput sequencing of the V3-V5 region of the 16S rRNA gene. Plos One 2012,
11. Garrett LA, Brown R, Poxton IR: A comparative study of the intestinal microbiota of
healthy horses and those suffering from equine grass sickness. Vet Microbiol 2002,
12. Milinovich GJ, Trott DJ, Burrell PC, van Eps AW, Thoefner MB, Blackall LL, Al Jassim
RAM, Morton JM, Pollitt CC: Changes in equine hindgut bacterial populations during
oligofructose-induced laminitis. Environ Microbiol 2006, 8((5):885–898.
13. Milinovich GJ, Trott DJ, Burrell PC, Croser EL, Al Jassim RAM, Morton JM, van Eps
AW, Pollitt CC: Fluorescence in situ hybridization analysis of hindgut bacteria
associated with the development of equine laminitis. Environ Microbiol 2007, 9(8):2090–
14. Milinovich GJ, Burrell PC, Pollitt CC, Klieve AV, Blackall LL, Ouwerkerk D, Woodland
E, Trott DJ: Microbial ecology of the equine hindgut during oligofructose-induced
laminitis. Isme Journal 2008, 2(11):1089–1100.
15. Garner HE, Coffman JR, Hahn AW, Hutcheson DP, Tumbleson ME: Equine laminitis of
alimentary origin - experimental model. Am J Vet Res 1975, 36(4):441–444.
16. Shirazi-Beechey SP: Molecular insights into dietary induced colic in the horse. Equine
Vet J 2008, 40(4):414–421.
17. Durham AE: The role of nutrition in colic. Vet Clin North America-Equine Practice
18. Geor RJ: Current Concepts on the Pathophysiology of Pasture-Associated Laminitis.
Vet Clin North Am-Equine Pract 2010, 26(2):265–276.
19. Geor RJ: Pasture-associated laminitis. Vet Clin North America-Equine Practice 2009,
20. USDA: Lameness & laminitis in U.S. horses. In Edited by APHIS. Fort Collins,
Colorado: United States Department of Agriculture; 2000.
21. Milinovich GJ, Klieve AV, Pollitt CC, Trott DJ: Microbial events in the hindgut during
carbohydrate-induced equine laminitis. Vet Clin North America-Equine Practice 2010,
22. Al Jassim RA, Scott PT, Trebbin AL, Trott D, Pollitt CC: The genetic diversity of lactic
acid producing bacteria in the equine gastrointestinal tract. FEMS Microbiol Lett 2005,
23. Garner HE, Moore JN, Johnson JH, Clark L, Amend JF, Tritschler LG, Coffmann JR,
Sprouse RF, Hutcheson DP, Salem CA: Changes in cecal flora associated with onset of
laminitis. Equine Vet J 1978, 10(4):249–252.
24. Daly K, Stewart CS, Flint HJ, Shirazi-Beechey SP: Bacterial diversity within the
equine large intestine as revealed by molecular analysis of cloned 16S rRNA genes.
FEMS Microbiol Ecol 2001, 38(2–3):141–151.
25. Gronvold AMR, L'Abee-Lund TM, Strand E, Sorum H, Yannarell AC, Mackie RI: Fecal
microbiota of horses in the clinical setting: Potential effects of penicillin and general
anesthesia. Vet Microbiol 2010, 145(3–4):366–372.
26. Morita H, Nakano A, Shimazu M, Toh H, Nakajima F, Nagayama M, Hisamatsu S, Kato
Y, Takagi M, Takami H, et al: Lactobacillus hayakitensis, L-equigenerosi and L-equi,
predominant lactobacilli in the intestinal flora of healthy thoroughbreds. Anim Sci J
27. Goodson J, Tyznik WJ, Cline JH, Dehority BA: Effects of an abrupt diet change from
hay to concentrate on microbial numbers and physical-environment in the cecum of the
pony. App Environ Microbiol 1988, 54(8):1946–1950.
28. Julliand V, de Fombelle A, Drogoul C, Jacotot E: Feeding and microbial disorders in
horses: Part 3 - Effects of three hay: grain ratios on microbial profile and activities. J
Equine Vet Sci 2001, 21(11):543–546.
29. Mackie RI, Wilkins CA: Enumeration of anaerobic bacterial microflora of the equine
gastrointestinal tract. Appl Environ Microbiol 1988, 54(9):2155–2160.
30. Willing B, Voros A, Roos S, Jones C, Jansson A, Lindberg JE: Changes in faecal
bacteria associated with concentrate and forage-only diets fed to horses in training.
Equine Vet J 2009, 41(9):908–914.
31. Yamano H, Koike S, Kobayashi Y, Hata H: Phylogenetic analysis of hindgut
microbiota in Hokkaido native horses compared to light horses. Anim Sci J 2008,
32. Dougal K, Harris PA, Edwards A, Pachebat JA, Blackmore TM, Worgan WJ, Newbold
CJ: A comparison of the microbiome and the metabolome of different regions of the
equine hindgut. FEMS Microbiol Ecol 2012, 1–12.
33. Shepherd ML, Swecker WS, Jensen RV, Ponder MA: Characterization of the fecal
bacteria communities of forage-fed horses by pyrosequencing of 16S rRNA V4 gene
amplicons. FEMS Microbiol Lett 2012, 326(1):62–68.
34. Perkins GA, den Bakker HC, Burton AJ, Erb HN, McDonough SP, McDonough PL,
Parker J, Rosenthal RL, Wiedmann M, Dowd SE, et al: The equine stomach harbors 1 an
abundant and diverse mucosal microbiota. Appl Environ Microbiol 2012, 78:2522–2531.
35. Wright ES, Yilmaz LS, Noguera DR: DECIPHER, a search-based approach to
chimera identification for 16S rRNA sequences. Appl Environ Microbiol 2012, 78:712–
36. Lozupone CA, Hamady M, Kelley ST, Knight R: Quantitative and qualitative beta
diversity measures lead to different insights into factors that structure microbial
communities. Appl Environ Microbiol 2007, 73(5):1576–1585.
37. Lozupone C, Lladser ME, Knights D, Stombaugh J, Knight R: UniFrac: an effective
distance metric for microbial community comparison. Isme Journal 2011, 5(2):169–172.
38. Dowd SE, Callaway TR, Wolcott RD, Sun Y, McKeehan T, Hagevoort RG, Edrington
TS: Evaluation of the bacterial diversity in the feces of cattle using 16S rDNA bacterial
tag-encoded FLX amplicon pyrosequencing (bTEFAP). BMC Microbiol 2008, 8(125):1–
39. Shanks OC, Kelty CA, Archibeque S, Jenkins M, Newton RJ, McLellan SL, Huse SM,
Sogin ML: Community Structures of Fecal Bacteria in Cattle from Different Animal
Feeding Operations. Appl Environ Microbiol 2011, 77(9):2992–3001.
40. Ley RE, Hamady M, Lozupone C, Turnbaugh PJ, Ramey RR, Bircher JS, Schlegel ML,
Tucker TA, Schrenzel MD, Knight R, et al: Evolution of mammals and their gut microbes.
Sci 2008, 320(5883):1647–1651.
41. Julliand V, de Vaux A, Millet L, Fonty G: Identification of Ruminococcus flavefaciens
as the predominant cellulolytic bacterial species of the equine cecum. Appl Environ
Microbiol 1999, 65:3758–3741.
42. DuPont AW, DuPont HL: The intestinal microbiota and chronic disorders of the gut.
Nature Rev Gastroenterol Hepatol 2011, 8(9):523–531.
43. Sinh P, Barrett TA, Yun L: Clostridium difficile Infection and Inflammatory Bowel
Disease: A Review. Gastroenterol Res Pract 2011, 136064:1–11.
44. Musa S, Moran C, Rahman T: Clostridium difficile infection and liver disease. J
Gastrointestinal and Liver Dis 2010, 19(3):303–310.
45. Songer JG: Clostridial diseases of small ruminants. Vet Res 1998, 29(3–4):219–232.
46. Widyastuti Y, Lee WK, Suzuki K, Mitsuoka T: Isolation and characterization of rice-
straw degrading clostridia from cattle rumen. J Vet Med Sci 1992, 54(1):185–188.
47. Santos AS, Rodrigues MAM, Bessa RJB, Ferreira LM, Martin-Rosset W: Understanding
the equine cecum-colon ecosystem: current knowledge and future perspectives. Animal
48. Werners AH, Bull S, Fink-Gremmels J: Endotoxaemia: a review with implications for
the horse. Equine Vet J 2005, 37(4):371–383.
49. Opal SM: Endotoxins and Other Sepsis Triggers. In Endoxemia and Endotoxin Shock:
Disease, Diagnosis and Therapy. Volume 167, edn. Edited by Ronco C, Piccinni P, Rosner
50. Costa CC, Weese J: The equine intestinal microbiome. Ani Health Res Rev 2012,
51. Hintz HF, Argenzio RA, Schryver HF: Digestion coefficients, blood glucose levels and
molar percentage of volatile acids in intestinal fluid of ponies fed varying forage-grain
ratios. J Anim Sci 1971, 33(5):992–995.
52. Suchodolski JS, Ruaux CG, Steiner JM, Fetz K, Williams DA: Application of molecular
fingerprinting for qualitative assessment of small-intestinal bacterial diversity in dogs. J
Clin Microbiol 2004, 42(10):4702–4708.
53. Dowd SF, Sun Y, Wolcott RD, Domingo A, Carroll JA: Bacterial tag-encoded FLX
amplicon pyrosequencing (bTEFAP) for microbiome studies: Bacterial diversity in the
ileum of newly weaned Salmonella-infected pigs. Foodborne Pathogens and Disease 2008,
54. Rudi K, Skulberg OM, Larsen F, Jakobsen KS: Strain characterization and
classification of oxyphotobacteria in clone cultures on the basis of 165 rRNA sequences
from the variable regions V6, V7, and V8. Appl Environ Microbiol 1997, 63(7):2593–
55. Baker GC, Smith JJ, Cowan DA: Review and re-analysis of domain-specific 16S
primers. J Microbiol Methods 2003, 55(3):541–555.
56. Edgar RC: Extreme high-speed clustering, alignment and database search. 2010, 1:579.
57. Gontcharova V, Youn E, Wolcott RD, Hollister EB, Gentry TJ, Dowd SE: Black box
chimera check (B2C2): a windows-based software for batch depletion of chimeras from
bacterial 16S rRNA gene datasets. The Open Microbiol J 2010, 4:47–52.
58. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ: Basic local alignment search
tool. J Mol Biol 1990, 215(3):403–410.
59. Caporaso JG, Kuczynski J, Stombaugh J, Bittinger K, Bushman FD, Costello EK, Fierer
N, Pena AG, Goodrich JK, Gordon JI, et al: QIIME allows analysis of high-throughput
community sequencing data. Nature Methods 2010, 7(5):335–336.
60. Chao A, Bunge J: Estimating the number of species in a Stochastic abundance model.
Biometrics 2002, 58(3):531–539.
61. Wang Q, Garrity GM, Tiedje JM, Cole JR: Naive Bayesian classifier for rapid
assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microbiol
62. McDonald D, Price MN, Goodrich J, Nawrocki EP, DeSantis TZ, Probst A, Andersen
GL, Knight R, Hugenholtz P: An improved Greengenes taxonomy with explicit ranks for
ecological and evolutionary analyses of bacteria and archaea. Isme J 2012, 6(3):610–618.
63. Caporaso JG, Bittinger K, Bushman FD, DeSantis TZ, Andersen GL, Knight R:
PyNAST: a flexible tool for aligning sequences to a template alignment. Bioinforma
64. Price MN, Dehal PS, Arkin AP: FastTree: Computing large minimum evolution trees
with profiles instead of a distance matrix. Mol Biol Evol 2009, 26(7):1641–1650.
65. Henneke DR, Potter GD, Kreider JL, Yeates BF: Relationship between condition score,
physical measurements and body fat percentage in mares. Equine Vet J 1983, 15:371–
Additional_file_1 as DOC
Additional file 1: Table S1 Description of horses in each study group. The exact age of all
horses in the laminitis group was not known; one horse was estimated to be 7 years of age,
the others were over 15 years. None of the horses with laminitis exhibited signs of Cushing’s
syndrome, although the initial cause of laminitis was not known for all animals.
Abbreviations are as follows: BCS – body condition score , QH – Quarter horse, WB –
Warmblood, TB – Thoroughbred. Table S2 Feed analysis. Comparison of guaranteed feed
analysis of 12% concentrate pellets and 16% concentrate pellets used in this study. Pellets
were formulated by Producer’s Co-op (Bryan, TX) and further information is available
Additional files provided with this submission: Download full-text
Additional file 1: 1859908492749999_add1.doc, 51K