1939165x, 2018, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vcp.12566 by Veterinärmedizinische Universität Wien, Wiley Online Library on [03/03/2023].

See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
DOI: 10.1111/vcp.12566

ORIGINAL RESEARCH

Evaluation of microRNA stability in feces from healthy dogs

Susanna Cirera1 | Line M. Willumsen1,2 | Thea T. Johansen1,2 | Lise N. Nielsen2

1
Department of Veterinary and Animal
Sciences, University of Copenhagen,
Copenhagen, Denmark
2
Department of Veterinary Clinical
Sciences, University of Copenhagen,
Copenhagen, Denmark
Correspondence
S. Cirera, Division of Animal Genetics,
Bioinformatics and Breeding, Department of
Veterinary and Animal Sciences, Faculty of
Health and Medical Sciences, University of
Copenhagen, Frederiksberg C, Denmark
E-mail: scs@sund.ku.dk
Background: Gastrointestinal cancer accounts for approximately 8% of all canine
malignancies. Early detection of cancer may have a tremendous impact on both
treatment options and prognosis. MicroRNAs (miRNAs), a class of noncoding RNAs
that can be found stably expressed in body fluids and feces, have been suggested
as valuable human cancer biomarkers.
Objectives: The purpose of the study was to investigate the feasibility of detecting
miRNAs in canine feces and to determine the miRNA stability in fecal samples
stored at different temperatures for different duration.
Methods: The levels of 4 Canine familiaris (cfa) miRNAs (cfa-miR-16, cfa-miR-20a,
cfa-miR-21, and cfa-miR-92a) were investigated by quantitative real-time PCR(qPCR)
in fecal samples from 10 healthy dogs. Fecal samples were collected at 3 different
time points and samples from the first time point were stored at different tempera-
tures and for a different duration.
Results: A statistically significant difference was found in miRNA levels from
samples stored at room temperature compared with samples stored at 20°C for
cfa-miR-16 and cfa-miR-21. No significant difference was found in the level of the
investigated miRNAs over time.
Conclusions: Overall, miRNAs are present in dog feces at measurable levels. Some
miRNAs seem to be subject to a higher degree of degradation in samples stored at
room temperature for 24 hours compared with samples frozen after collection at
20°C. The investigated miRNAs were stably expressed over time. This study pro-
vides the basis for further research on miRNA expression profiles as biomarkers for
gastrointestinal cancer in dogs.

KEYWORDS
Biomarker, canine, diagnostics, feces, microRNA profiling, qPCR

INTRODUCTION intestinal carcinomas or leiomyosarcomas have a median survival

Cancer is one of the leading causes of death in dogs and gastroin-
testinal (GI) cancer accounts for approximately 8% of all canine
malignancies.1 Dogs with GI cancer display a variety of nonspecific
clinical signs, which complicates and potentially delays the diagnostic
process. This, in turn, may result in belated therapy and affect the
prognosis. Currently, dogs treated with surgical excision of either
time of 10 months, but if complicated by metastasis, the median sur-
vival time decreases to 3 months.2 Early disease detection and dif-
ferentiation of GI cancer from other GI diseases could have a
tremendous impact on both treatment strategies and prognosis. In
people, plasma microRNA (miRNA) profiles have been investigated in
colorectal cancer, both as diagnostic3,4 and prognostic biomarkers4,5
with encouraging results. Recently, noninvasive fecal miRNA profiles
have been investigated in human patients with colorectal cancer

Line M. Willumsen and Thea T. Johansen contributed equally to this work.

showing promising results.6 While plasma miRNA profiles have been

Vet Clin Pathol. 2018;47:115–121. wileyonlinelibrary.com/journal/vcp © 2018 American Society for | 115
Veterinary Clinical Pathology
 1939165x, 2018, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vcp.12566 by Veterinärmedizinische Universität Wien, Wiley Online Library on [03/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
116 | CIRERA ET AL.

investigated in dogs with various types of cancer7,8, to our knowl- T A B L E 1 . Breed, age, and sex of 10 healthy dogs included in the
edge, there are no studies investigating canine fecal miRNA profiles miRNA study.
alone or in association with diseases. Dog ID Breed Age Sex
MiRNAs are small noncoding RNAs that regulate gene expres- 1 Bernese Mountain dog 5 years, 10 months ME
sion at the posttranscriptional level. Their functionality includes 2 Labrador Retriever 6 years, 4 months ME
acting as proto-oncogenes or tumor suppressor genes.9 Circulating 3 Cross breed 1 year, 2 months ME
miRNAs have revealed surprising stability, which may be related
4 Labrador Retriever 1 year, 9 months FE
to the circulatory transport in association with proteins or inside
5 Cross breed 2 years, 3 months MN
vesicles.10,11
6 Miniature Bull Terrier 2 years, 2 months ME
Feces is a complex material due to the amount of bacterial and
7 Miniature Bull Terrier 7 years, 1 month FN
other contaminants that are co-purified when the RNA is isolated,
8 Labrador Retriever 1 year, 7 months FE
but is both easy to sample and a noninvasive collection method. Gut
epithelial cells and HOP homeobox (Hopx) protein-expressing cells 9 English Springer Spaniel 2 years, 1 month ME

are the main contributors to fecal miRNAs. These fecal miRNAs 10 French Bulldog 3 years, 4 months FN

encompass those included within extracellular vehicles (EV), such as
within microvesicles or exosomes, and those which are EV-free and
where the miRNAs are associated with lipoproteins or argonaute
proteins, which form part of RNA silencing processes and are essen-
tial components of the RNA-induced silencing complex.12 Moreover,
both stability of miRNAs in human fecal samples as well as the level
of fecal miRNA expression over time in cancer patients have been
previously assessed.13
We have recently shown that preanalytic aspects, including
miRNA sampling and storage protocols in canine serum and plasma,
are of the outmost importance since miRNAs degrade relatively fast
and some are more affected by degradation than others.14 Feces is
an easily accessible material that requires very little effort to collect
and would; therefore, serve as an excellent matrix for noninvasive
biomarkers. Therefore, it is of significant relevance to obtain informa-
tion regarding the stability of miRNAs in fecal material.
The objectives of this pilot study were (1) to investigate the pos-
sibility of identifying miRNAs in fecal samples from healthy dogs at
measurable levels and (2) to evaluate the stability of these miRNAs
under different storage conditions and over time.
MATERIALS AND METHODS
Recruitment of dogs
Ten healthy dogs were recruited for this study during September
and October 2016. The dogs included 3 Labrador Retrievers, 2
Miniature Bull Terriers, 2 Labrador Retriever mixed breeds, and one
each of the following: Bernese Mountain Dog, English Springer Spa-
niel, and French Bulldog. Six dogs were intact males and 4 were
females (2 spayed), with an average age of 40.3 months (Table 1).
The dogs were determined to be healthy by physical examination,
routine hematology, serum biochemistry, and urinalysis of voided
mid-stream urine. Fecal flotation for intestinal parasites was per-
formed on feces from all dogs and only dogs with a negative result
were included in the study. This study was approved by the local
ethical administrative board at the Department of Veterinary Clinical
and Animal Sciences, University of Copenhagen. Owners signed a
form of consent before participating in the study.
ME indicates male intact; MN, male chemically neutered; FE, female
intact; FN, female neutered.
Collection and storage of fecal samples
To examine the stability of miRNAs from fecal samples and the level
and availability of the miRNAs at different time points, 10 fecal
samples of around 1 g for each dog were collected into cryo tubes,
6 samples at day 1, 2 samples at day 4, and 2 samples at day 7. The
samples from day 1 of the study were stored as follows: 2 samples
were left at room temperature (22°C) for 24 h before freezing at
20°C; 2 samples were refrigerated (4°C) for 24 h before freezing
at 20°C; and 2 samples were frozen within an hour of collection at
20°C. Furthermore, the owners of the dogs collected 2 fecal sam-
ples from their dogs at day 4, and 2 samples at day 7 after the initial
sample, and froze these samples at 20°C before transport to the
laboratory on dry ice, which were stored at 20°C before further
processing (for sampling protocol see Figure 1).
Total RNA extraction
The performance of 2 kits was tested in this study following each
manufacturer’s recommendations: Stool Total RNA Purification kit
(Norgen, Biotek Corporation, Thorold, ON, Canada) and miRNeasy Mini
kit (Qiagen, Hilden, Germany). Both kits isolate total RNA including
miRNAs transported in vehicles (ie, exosomes) or attached to lipopro-
teins or argonaute proteins. In brief, the miRNeasy Mini kit (Qiagen) is
based on the extraction of total RNA using guanidine thiocyanate, phe-
nol, and chloroform followed by column-based purification, which
enriches for miRNAs. The Stool Total RNA Purification kit (Norgen) is
also column-based and uses guanidine hydrochloride and ethanol for
extraction, but is free from both phenol and chloroform and is specifi-
cally designed for extraction of total RNA from feces. Subsequently, all
samples were DNase treated (DNAse I, Qiagen) to degrade genomic
DNA. The amount and purity of the RNA samples were assessed by
OD measurements in a NanoDrop 1000 spectrophotometer (Thermo-
Fisher Scientific, Waltham, USA) . The RNA quality was further
assessed by visual inspection after agarose gel electrophoresis. RNA
stocks were kept in a 80°C freezer until further processing.
 1939165x, 2018, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vcp.12566 by Veterinärmedizinische Universität Wien, Wiley Online Library on [03/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
CIRERA ET AL. | 117

RT
RT
+4
Fecal sample, day 1
+4
-20
Dog -20
-20
Fecal sample, day 4
-20
-20
Fecal sample, day 7
-20

F I G U R E 1 . Fecal sample collection protocol from 10 healthy dogs. From each dog, 10 fecal samples were collected. Day 1, the samples
were either kept at room temperature (RT), at +4°C (+4) or frozen at 20°C ( 20) directly. After 24 hours, all samples kept at either RT or +4
were frozen to 20⁰C. The owners of the dogs collected fecal samples at day 4 and day 7 and froze the samples immediately at 20°C.

29 QuantiFast SYBRGreen PCR master mix (Qiagen), and 250 nM

cDNA synthesis
A 100 ng sample of each RNA stock was used for cDNA synthesis
according to described protocols.15 Additionally, a synthetic miRNA,
cel-miR-39-3p, a miRNA from C elegans which is not found in dogs
was used as a spike-in for cDNA synthesis to correct for the efficiency
of the cDNA reaction and subsequently to correct the raw quantita-
tive real-time PCR (qPCR) data. Two replicates were made of each
sample and negative room temperature (RT) controls lacking reverse
transcriptase were used to assess the specificity of each assay. The
cDNA stocks were diluted 8-fold before being used in qPCR.
Primer design and qPCR
Four Canine familiaris (cfa) miRNAs (cfa-miR-16, cfa-miR-20a cfa-
miR-21, and cfa-miR-92a) and one C elegans miRNA (cel-miR-39-3p)
were investigated in this study. The 4 canine assays were chosen
based on high/stable expression in the feces of human colon cancer
patients.16,17 Forward and reverse DNA primers were designed as
described previously15,18,19 using the publicly available software
“miRprimer” (for primer sequences see Table 2). qPCR was per-
formed on a MX3005P instrument (Agilent Technologies, Santa
Clara, USA) using the associated software. Two commercial kits for
RNA isolation were examined: Stool Total RNA Purification kit from
Norgen (denoted as Norgen kit) and miRNeasy Mini kit from Qiagen
(denoted as Qiagen kit). Briefly, 1 lL of cDNA diluted 8-fold, 5 lL of
of each primer were mixed in white 96-well PCR plates (ABgene;
Thermo Fisher Scientific, USA) in a total volume of 10 lL. Cycling
conditions were 95°C for 5 min and 40 cycles of 95°C for 10 s fol-
lowed by 64°C for 30 s. A melting curve analysis was performed to
evaluate the specificity of each assay. Standard curves (3- or 5-fold
dilutions of a pool of equal amounts of all cDNA samples) with 6
dilution points were also included to calculate the PCR efficiency for
each miRNA assay.
qPCR data processing and statistical analyses
The qPCR raw data (denoted as quantification cycles, Cq) were prepro-
€ teborg, Sweden). Briefly,
cessed using GenEx Pro (multiD Analyses AB, Go
interplate calibration, efficiency correction, correction of cDNA amounts
using the cel-miR-39-3p qPCR, averaging of the technical repeats, fold
change calculation, and log2 transformation were performed.
All data groups were tested for normality using D’Agostino–Pear-
son normality tests. To see if equal variances were present, F-tests
were performed. Statistical analyses of storage methods were
performed using one-way repeated measures ANOVA. If a significant
difference was present, a Tukey’s post hoc test was performed. Com-
parisons of the collections at different time points were performed
with paired Student’s t-tests. All statistical analyses were performed
using GraphPad Prism 7.0 software (GraphPad Software, San Diego,
USA) and cutoff for statistical significance was set at a P < .05.

T A B L E 2 . Mature sequences and forward and reverse primers for each miRNA tested. Primers were designed using miRprimer.18 Cel-miR-
39-3p was used as spike-in in the cDNA synthesis and was detected by qPCR in each sample.
miRNA Mature Sequence Forward Primer Reverse Primer
cfa-miR-16 UAGCAGCACGUAAAUAUUGGCG CAGTAGCAGCACGTAAATATTG CAGTTTTTTTTTTTTTTTCGCCAA
cfa-miR-20a UAAAGUGCUUAUAGUGCAGGUAG ACAGTAAAGTGCTTATAGTGCA GTCCAGTTTTTTTTTTTTTTTCTACCT
cfa-miR-21 UAGCUUAUCAGACUGAUGUUGA GCAGTAGCTTATCAGACTGATG GGTCCAGTTTTTTTTTTTTTTTCAAC
cfa-miR-92a UAUUGCACUUGUCCCGGCCUGU AGGTGTGTATAAAGTATTGCACTTGTCC CAGGTCCAGTTTTTTTTTTTTTTTACAG
cel-miR-39-3p UCACCGGGUGUAAAUCAGCUUG GTCACCGGGTGTAAATCAG CCAGTTTTTTTTTTTTTTTCAAGCTG

118 | CIRERA
RESULTS
Detection of miRNAs
All qPCR efficiencies were in the range of 80–110% except for the
cfa-miR-20a assay (71.7%). Due to the complexity of the sample
material and time limitation, the efficiency of the cfa-miR-20a assay
was accepted and included for further analyses, though suboptimal.
Additionally, a visual inspection of raw quantitation cycle (Cq) values
for the cel-miR-39-3p spike-in assay showed a total of 6 samples
with markedly higher Cq values compared with that of the other
samples: dog number 4, 6, and 7 on day 4, and dog number 2, 4,
and 5 on day 7. These samples were considered technical outliers
and were eliminated from further analysis.
Detecting the 4 investigated miRNAs in fecal samples was feasi-
ble (Figure 2A,B). The yield of total RNA was higher for the Norgen
kit but twice as much fecal material was used for RNA extraction
compared with the Qiagen kit. Nevertheless, the amount of miRNAs
isolated with the Qiagen kit resulted in lower Cq values in the qPCR.
The 260/230 ratio showing organic compound contamination was
found to be relatively high with both kits, but generally higher for
RNA extracted with Qiagen kit. Several samples processed with the
Norgen kit showed low reliability when comparing the raw Cq values
from the 2 cDNA replicates that differed by more than 1.5 cycles.
Moreover, the Norgen kit produced a larger number of unacceptable
melting curves (more than a single peak) compared with the Qiagen
kit across the 4 assays, which indicated that more contaminants
were present in the samples isolated with the Norgen kit. Paired
Student’s t-tests, in processed qPCR data, showed statistically signifi-
cant differences between the performances of the 2 kits in all assays
(Figure 1, Table 1), with the Qiagen kit showing better overall
performance. Hence, the subsequent statistical analyses were only
performed for the fecal samples processed with the Qiagen kit.
Comparison of miRNA stability between storage
conditions
The qPCR data obtained from all assays followed Gaussian distribu-
tions. The F-tests confirmed equal variances between storage methods
in all assays. Statistically significant differences were found in the
amount of extracted miRNA between storage methods for 2 miRNAs:
cfa-miR-16 (P = .047) and cfa-miR-21 (P = .019) (Figure 2A). Tukey’s
post hoc tests showed that the amounts of extracted miRNA from
samples stored at RT and 20°C was significantly different in both
assays (see Table 3) with higher amounts in samples stored at 20°C.
Comparison of miRNA profiles at different time points
The samples, frozen directly at 20°C on day 1, 4, and 7 for each
dog, were compared for all assay a (Figure 2B). Because of the
exclusion of outlier samples due to technical problems on the cel-
miR-39-3p spike-in step (see above), the sample sizes were too small
to perform a normality test. The F-tests confirmed equal variances
1939165x, 2018, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vcp.12566 by Veterinärmedizinische Universität Wien, Wiley Online Library on [03/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
ET AL.
among data from day 1, 4, and 7 in all assays. Paired Student’s t-
tests were performed in 3 separate instances for each assay: day
1 compared to day 4, day 1 compared to day 7, and day 4 com-
pared to day 7. No statistically significant differences were found
in the levels of the 4 miRNAs among any of the collection days
(Table 4).
DISCUSSION
In this study, we identified all 4 tested miRNAs in the feces of
healthy dogs. In addition, these miRNAs were stably expressed at
different time points when fecal samples were frozen at 20°C.
Levels of cfa-miR-16 and cfa-miR-21 assays differed significantly in
the samples stored at RT compared with the samples frozen at
20°C. This implies that samples stored at RT suffer some degree of
degradation and that some miRNAs are more sensitive to degrada-
tion than others, which suggests that fecal samples to be used for
miRNA profiling should be either frozen immediately or preserved in
stabilizing buffer (ie, RNA later) shortly after collection. This result is
in agreement with another study performed in the serum and plasma
of healthy dogs.14 It has been hypothesized that miRNA stability, like
mRNA, is correlated with the number of adenosine–uridine (AU)
nucleotide pairs in the miRNA sequence.20,21 Higher numbers of AU
pairs result in lower stability. Another explanation could be that the
stability of each miRNA present in feces is influenced by the pro-
teins or extracellular vesicles they are transported in.22 The existing
literature that describes miRNA stability in human feces is limited,
although an earlier study estimated that degradation of miRNA in
human feces occurs over a 72-h period and identified a 30% reduc-
tion in the original miRNA concentration during that time period and
the degradation happened primarily in the first 12 h.23 In the present
study, cfa-miR-16 showed a 37% reduction and cfa-miR-21 showed
a 49% reduction over a 24 hour time period.
In directly frozen samples, no statistically significant differences
were found between the amount of miRNA present in samples
collected on days 1, 4, or 7, for the 4 miRNAs tested, which sug-
gests that the expression levels of these miRNAs remained stable
over time in healthy dogs under these conditions. In agreement
with these findings, an earlier study17 found stable expression of 6
miRNAs in fecal samples from 8 healthy people at 2 time points.
However, miRNAs could have different behaviors, and profiling a
larger panel of miRNAs is, therefore, warranted to verify this con-
clusion.
It was expected that all miRNAs present in the samples, including
stably expressed reference miRNAs, could be affected by the degra-
dation process that occurred in the samples. Therefore, the method
chosen for correction of the qPCR data, in the present study, was
the use of a spike-in during cDNA synthesis (as suggested by the
Exiqon company24). The spike-in miRNA cel-miR-39-3p was chosen
for this purpose because it lacks homology with mammalian miRNA
sequences25 and is included in standard spike-in kits from Exiqon.
By adding the synthetic miRNA before cDNA synthesis, the
 1939165x, 2018, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vcp.12566 by Veterinärmedizinische Universität Wien, Wiley Online Library on [03/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
CIRERA ET AL. | 119

F I G U R E 2 . (A) Box plot for each

canine fecal miRNA assay including the
samples kept at room temperature for
24 hours before freezing at 20°C (RT,
N = 10 dogs), samples kept at 4°C before
freezing at 20°C (+4, N = 10 dogs) and
samples directly frozen at 20°C ( 20,
N = 10 dogs). *Statistically significantly
different P-values: miR-16 P = .047; and
miR-21 P = .019. (B) Box plots for each
canine fecal miRNA assay including the
samples directly frozen at 20°C on days
1 (N = 10), 4 (N = 7), and 7 (N = 7). Dog
number 4, 6, and 7 on day 4, and dog
number 2, 4, and 5 on day 7 were
considered outliers and were eliminated
from further analysis. There was no
significant difference in expression levels
for any of the days and any of the assays
tested.

amount of cel-miR-39-3p detected by qPCR will reflect the cDNA Feces contain several classes of inhibitors that may be preserved
levels for this particular miRNA produced during a reaction, which during the RNA isolation steps (ie, polysaccharides, glycolipids, and
is highly affected by contaminants carried over from the RNA iso- urea) along with chemical residues from the extraction procedure (ie,
lation step. phenol). All these compounds can compromise the quality of the
 1939165x, 2018, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vcp.12566 by Veterinärmedizinische Universität Wien, Wiley Online Library on [03/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
120 | CIRERA ET AL.

T A B L E 3 . Mean differences, 95% CI of differences, and adjusted qPCR, is considered acceptable. Standard software programs can
P-values for Tukey’s post hoc tests performed on cfa-miR-16 and help to choose the most stable reference genes. Alternatively, qPCR
cfa-miR-21 in canine fecal samples stored under different conditions data can be corrected using an artificial miRNA, which is spiked-in at
(room temperature [RT], 4°C [+4°C] before being frozen at 20°C,
the beginning of the RNA isolation or at the beginning of the cDNA
and frozen directly at 20°C [ 20°C]). Significant results are
highlighted in gray. synthesis (Exiqon). It is important to emphasize that synthetic miRNA
spike-ins do not necessarily behave like endogenous miRNAs, and
Mean Adjusted
cfa-miR-16 Difference 95% CI of Difference P-Value when possible, normalization to stably-expressed endogenous miR-
NAs should always be performed.
RT vs +4°C 0.54 1.32 to 0.25 .192
Extraction of miRNAs from feces can result in co-extraction of
RT vs 20°C 0.61 1.03 to 0.20 .007**
products deriving from other sources than the dog itself, such as
+4°C vs 20°C 0.08 0.72 to 0.57 .942
intestinal parasites. In fact, a recent study found that the porcine
cfa-miR-21 Mean 95% CI of difference Adjusted
whipworm, Trichuris suis, can secrete extracellular vesicles containing
difference P-value
RNA, and also small RNAs and miRNAs.28 It is not known whether
RT vs +4°C 0.79 1.77 to 0.20 .12
canine intestinal parasites are able to secrete miRNAs or if these
RT vs 20°C 0.91 1.47 to 0.34 .004**
miRNAs share sequence homology with the nematode C elegans.
+4°C vs 20°C 0.12 0.71 to 0.47 .846
Therefore, caution should be exercised when using cel-miR-39-3p to
*P < .05, **P < .01. normalize qPCR data in studies similar to this. The healthy dogs in
our study were determined to be negative for intestinal parasites by
simple fecal flotation. Although this test is a standard test for detec-
T A B L E 4 . Paired Student’s t-tests of canine fecal samples for tion of intestinal parasites, the test has a moderate sensitivity and
quantification of miRNAs on collection days. The table shows the
specificity29; and as only 2 of the dogs received anthelmintics prior
mean difference (Mean dif), 95% CI of difference, and P-values of
the 3 separated paired Student’s t-tests were performed for each of to entering the study, it may be questionable whether all the dogs in
the 4 assays: cfa-miR-16, cfa-miR-20a, cfa-miR-21, and cfa-miR-92a. this study were truly free of intestinal parasites.
No statistically significant differences were found between the Fecal matter is a complex and heterogeneous matrix with great
amounts of extracted miRNA on any of the collection days. diversity in both general composition and microbiota, and it can vary
Threshold for statistical significance was set at P < .05. 1, day 1; 4,
greatly depending on genetics and lifestyle conditions including age,
day 4; 7, day 7.
breed, and diet.30,31 The complexity of the fecal samples generated
Assay Days Mean Dif 95% CI P-value
concern that unwarranted noise and interference with the RNA extrac-
cfa-miR-16 1 vs 4 0.26 0.73 to 0.21 .221 tion process would occur, and therefore, 2 kits commonly used for RNA
1 vs 7 0.3 1.26 to 0.66 .474 isolation from feces were tested. Although both kits succeeded in iso-
4 vs 7 0.56 1.83 to 0.71 .291 lating miRNA, a statistically significant difference was found between
cfa-miR-20a 1 vs 4 0.2 0.44 to 0.05 .094 the amounts of the 4 miRNAs that were isolated with the 2 kits, which
1 vs 7 0.34 1.25 to 0.56 39 highlights the importance of choosing a good RNA extraction method,
4 vs 7 0.44 1.61 to 0.73 .356 and then, using the same method for all the samples.
cfa-miR-21 1 vs 4 0.09 1.09 to 0.90 .825 There are several limitations to be addressed in the present

1 vs 7 0.31 0.90 to 0.27 .239 study. Firstly, a sample size of 10 dogs is considered of low power,
in particular, when several samples are needed to be excluded from
4 vs 7 0.53 1.78 to 0.73 .31
the analysis due to technical issues. A larger sample size would
cfa-miR-92a 1 vs 4 0.41 0.99 to 0.17 .132
increase the possibility of detecting small variations in miRNA
1 vs 7 0.52 1.73 to 0.70 .334
expression levels over time. Secondly, it is difficult to draw general
4 vs 7 0.53 2.35 to 1.29 .466
conclusions regarding the overall stability of the miRNA population
in feces from dogs, based on only 4 miRNAs, and a larger panel of
isolated RNA and result in downstream inhibition of the enzymatic miRNAs should be tested. Thirdly, only one synthetic miRNA was
reactions such as cDNA synthesis and qPCR. Specifically, cDNA syn- used as a spike-in for normalization of technical variability. In the
thesis is sensitive to inhibitors, which results in different reaction present study, 6 out of 50 total samples had to be excluded due to
efficiencies, and subsequently, different cDNA yields between sam- technical issues that occurred during the cDNA synthesis. Several
ples. To correct for differences in the cDNA amount as input mate- synthetic miRNAs could have been used to optimize the process.
rial for the qPCR, data need to be normalized using endogenous or
exogenous reference genes. Unfortunately, there is no standardized
way to perform this correction/normalization when working with cir- CONCLUSION
culating or fecal miRNAs.26 When looking at biologic variability, nor-
malizing to at least 2 stably expressed endogenous reference This study confirmed that it is feasible to identify miRNAs from the
genes27 to correct for variability in the cDNA amounts used in feces of healthy dogs using commercial kits for extraction of total

CIRERA ET AL.
RNA. The 4 miRNAs tested (cfa-miR-16, cfa-miR-20a, cfa-miR-21,
and cfa-miR-92a) were reliably detected by qPCR. When profiling
cfa-miR-16 and cfa-miR-21 expression, a statistically significant dif-
ference was found in the miRNA levels of directly frozen samples
at 20°C and those stored at RT for 24 h. This result indicates
that some degree of degradation is happening in fecal samples
stored at RT for several hours and that some miRNAs seem to be
more affected by degradation than others. Moreover, the levels of
the 4 miRNAs tested seem to remain constant over time in
healthy dogs. Prospectively, it will be of great interest to explore
if identifying fecal miRNAs in dogs with cancer could become a
powerful new cancer biomarker.
ACKNOWLEDGMENTS
We acknowledge the veterinary nurses at the University Hospital for
Companion Animals for excellent assistance during blood sampling.
We also thank the technical staff at the Veterinary Diagnostic Labo-
ratory from the Department of Veterinary and Clinical Animal
Sciences for their help with analyses of blood and urine as well as
fecal flotation assays. We also thank the technicians at the Genetics
section at Department of Veterinary and Animal Sciences for help in
the laboratory work. Finally we thank the owners of the dogs for
participating in the study.
DISCLOSURE
The authors have indicated that they have no affiliations or financial
involvement with any organization or entity with a financial interest
in, or in financial competition with, the subject matter or materials
discussed in this article.
REFERENCES
1. Dobson JM, Samuel S, Milstein H, Rogers K, Wood JL. Canine neo-
plasia in the UK: estimates of incidence rates from a population of
insured dogs. J Small Anim Pract. 2002;43:240–246.
2. Crawshaw J, Berg J, Sardinas JC, et al. Prognosis for dogs with non-
lymphomatous, small intestinal tumors treated by surgical excision. J
Am Anim Hosp Assoc. 1998;34:451–456.
3. Zhu M, Huang Z, Zhu D, et al. A panel of microRNA signature in
serum for colorectal cancer diagnosis. Oncotarget. 2017;8:17081–
17091.
4. Kou CH, Zhou T, Han XL, Zhuang HJ, Qian HX. Downregulation of
mir-23b in plasma is associated with poor prognosis in patients with
colorectal cancer. Oncol Lett. 2016;12:4838–4844.
5. Yang J, Ma D, Fesler A, et al. Expression analysis of microRNA as prog-
nostic biomarkers in colorectal cancer. Oncotarget. 2016;8:52403–
52412.
6. Chang PY, Chen CC, Chang YS, et al. MicroRNA-223 and microRNA-
92a in stool and plasma samples act as complementary biomarkers
to increase colorectal cancer detection. Oncotarget. 2016;7:10663–
10675.
7. Borresen B, Nielsen LN, Jessen LR, Kristensen AT, Fredholm M, Cir-
era S. Circulating let-7 g is down-regulated in Bernese Mountain
dogs with disseminated histiocytic sarcoma and carcinomas - a
prospective study. Vet Comp Oncol. 2015;15:525–533.
1939165x, 2018, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/vcp.12566 by Veterinärmedizinische Universität Wien, Wiley Online Library on [03/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
| 121
8. Fujiwara-Igarashi A, Igarashi H, Mizutani N, et al. Expression profile
of circulating serum microRNAs in dogs with lymphoma. Vet J.
2015;205:317–321.
9. Zhang B, Pan X, Cobb GP, Anderson TA. microRNAs as oncogenes
and tumor suppressors. Dev Biol. 2007;302:1–12.
10. Cheng L, Sharples RA, Scicluna BJ, Hill AF. Exosomes provide a protec-
tive and enriched source of miRNA for biomarker profiling compared
to intracellular and cell-free blood. J Extracell Vesicles. 2014;3:23743.
https://doi.org/10.3402/jev.v3.23743
11. Arroyo JD, Chevillet JR, Kroh EM, et al. Argonaute2 complexes carry
a population of circulating microRNAs independent of vesicles in
human plasma. Proc Natl Acad Sci USA. 2011;108:5003–5008.
12. Liu S, da Cunha AP, Rezende RM, et al. The host shapes the gut
microbiota via fecal microRNA. Cell Host Microbe. 2016;19:32–43.
13. Xiao YF, Yong X, Fan YH, Lu MH, Yang SM, Hu CJ. microRNA
detection in feces, sputum, pleural effusion and urine: novel tools
for cancer screening (Review). Oncol Rep. 2013;30:535–544.
14. Enelund L, Nielsen LN, Cirera S. Evaluation of microRNA stability in
plasma and serum from healthy dogs. Microrna. 2017;6:42–52.
15. Balcells I, Cirera S, Busk PK. Specific and sensitive quantitative RT-
PCR of miRNAs with DNA primers. BMC Biotechnol. 2011;11:70.
16. Rotelli MT, Di Lena M, Cavallini A, et al. Fecal microRNA profile in
patients with colorectal carcinoma before and after curative surgery.
Int J Colorectal Dis. 2015;30:891–898.
17. Link A, Balaguer F, Shen Y, et al. Fecal microRNAs as novel biomark-
ers for colon cancer screening. Cancer Epidemiol Biomarkers Prev.
2010;19:1766–1774.
18. Cirera S, Busk PK. Quantification of miRNAs by a simple and specific
qPCR method. Methods Mol Biol. 2014;1182:73–81.
19. Busk PK. A tool for design of primers for microRNA-specific quanti-
tative RT-qPCR. BMC Bioinformatics. 2014;15:29.
20. Sethi P, Lukiw WJ. Micro-RNA abundance and stability in human
brain: specific alterations in Alzheimer’s disease temporal lobe neo-
cortex. Neurosci Lett. 2009;459:100–104.
21. Balzano F, Deiana M, Dei Giudici S, et al. miRNA stability in frozen
plasma samples. Molecules. 2015;20:19030–19040.
22. Diehl P, Fricke A, Sander L, et al. Microparticles: major transport
vehicles for distinct microRNAs in circulation. Cardiovasc Res.
2012;93:633–644.
23. Wu CW, Ng SS, Dong YJ, et al. Detection of miR-92a and miR-21 in
stool samples as potential screening biomarkers for colorectal cancer
and polyps. Gut. 2012;61:739–745.
24. https://www.exiqon.com/ls/Documents/Scientific/PCR-spike-in-
manual.pdf.
25. Schlosser K, McIntyre LA, White RJ, Stewart DJ. Customized internal
reference controls for improved assessment of circulating micro-
RNAs in disease. PLoS ONE. 2015;10:e0127443.
26. Schwarzenbach H, da Silva AM, Calin G, Pantel K. Data normalization
strategies for microRNA quantification. Clin Chem. 2015;61:1333–
1342.
27. Bustin SA. Why the need for qPCR publication guidelines?–The case
for MIQE. Methods. 2010;50:217–226.
28. Hansen EP, Kringel H, Williams AR, Nejsum P. Secretion of RNA-
containing extracellular vesicles by the porcine whipworm, Trichuris
suis. J Parasitol. 2015;101:336–340.
29. Dryden MW, Payne PA, Ridley R, Smith V. Comparison of common
fecal flotation techniques for the recovery of parasite eggs and
oocysts. Vet Ther. 2005;6:15–28.
30. Bermingham EN, Maclean P, Thomas DG, Cave NJ, Young W. Key
bacterial families (Clostridiaceae, Erysipelotrichaceae and Bac-
teroidaceae) are related to the digestion of protein and energy in
dogs. PeerJ. 2017;5:e3019.
31. Handl S, German AJ, Holden SL, et al. Faecal microbiota in lean and
obese dogs. FEMS Microbiol Ecol. 2013;84:332–343.