Variability in gut mucosal secretory IgA in mice along a working day
Burns et al. BMC Res Notes
Variability in gut mucosal secretory IgA in mice along a working day
Patricia Burns 0 3
Sofia Oddi 0 3
Liliana Forzani 2
Eduardo Tabacman 1
Jorge Reinheimer 0 3
Gabriel Vinderola 0 3
0 Instituto de Lactología Industrial (INLAIN, UNL-CONICET), Facultad de Ingeniería Química, Universidad Nacional del Litoral , Santiago del Estero 2829, 3000 Santa Fe , Argentina
1 Somalogic, Inc , 2945 Wilderness Pl, Boulder, CO 80301 , USA
2 Departamento de Matemática, Facultad de Ingeniería Química, Universidad Nacional del Litoral , Santiago del Estero 2829, 3000 Santa Fe , Argentina
3 Instituto de Lactología Industrial (INLAIN, UNL-CONICET), Facultad de Ingeniería Química, Universidad Nacional del Litoral , Santiago del Estero 2829, 3000 Santa Fe , Argentina
Objective: To assess the variability of secretory immunoglobulin A (S-IgA) in the lumen and feces of mice along a working day. Results: Mice were maintained under a 12 h light-dark cycle, light period starting at 8 AM. S-IgA was determined in feces and intestinal content (after one or three washes) at three points along the day: at the beginning, in the middle and at the end of the light period (ELP). Significant reduction in the content of S-IgA in the small intestine fluid and in feces was observed at the end of the light cycle, which coincides with the end of a regular working day (8 PM) in any given animal facility. It was also observed that three washes of the small intestine were more effective than one flush to recover a significant higher amount of S-IgA, with the smallest coefficient of variation observed by the ELP. A smaller CV would imply a reduced number of animals needed to achieve the same meaningful results. The results may be useful when designing animal trials for the selection of probiotic candidates based on their capacity of activating S-IgA, since it would imply a more rational use of experimental animals.
IgA; Intestine; Circadian rhythm; Mice; Variability
Probiotics are live microorganisms that when
administered in adequate amounts confer a health benefit on the
]. One criterion when selecting probiotics is their
capacity to stimulate secretion of secretory IgA (S-IgA)
], the immunoglobulin in charge of exclusion of
pathogens . Different probiotic strains have shown the
capacity of enhancing mucosal IgA in mice [
capacity has been linked to the protection against gut
]. In those assays mice are fed the strain
for different feeding periods (2, 5, 7 or 3, 6, 10 days) and
animals are sacrificed on the same day. This implies
sampling many animals along the same working day. Previous
results (not published) showed unexpected dispersion
of S-IgA values in mice of the same group, impairing
the observation of differences between groups. Immune
parameters oscillate rhythmically in the day . In
humans, salivary IgA reflect circadian rhythmicity, which
peak during sleep [
]. In rats, fecal IgA exhibited a clear
diurnal rhythm [
]. We aimed to assess the variability
of S-IgA in the lumen and feces of mice along a working
Materials and methods
Twenty-four 6-week old male BALB/c mice (20 ± 1 g)
were used (CMC-ICiVet-Litoral, CONICET–UNL).
Animals were kept for 7 days before the trial at 21 ± 2 °C,
55 ± 2% humidity and 12 h light–dark cycle. The light
period started at 8 AM. The trial was approved by the
Ethical Committee for Animal Experimentation
(FCVUNL), protocol 291/16, June 26th 2016.
Sampling of feces and intestinal content
Animals were sampled (8 animals/group in a period
of 15 min) at three time-points: at the beginning, in
the middle and at the end of the light period (named
BLP, MLP, ELP). Before sacrifice, feces were collected,
weighed, diluted 100× (1% (v/v) anti-protease cocktail
(P8340, Sigma) in PBS, homogenized (Ultra Turrax
T8, Ika Labortechnik, Staufen, Germany), centrifuged
(5000×g, 10 min, 4 °C) and the supernatant was frozen
at − 70 °C for S-IgA quantification by ELISA [
Sigma reagents (M-8769 anti-mouse IgA, Fast OPD
P-9187, M1421 IgA and A 4789, anti-mouse
Animals were anesthetized intraperitoneally
(ketamine, xylazine and acepromazine) and sacrificed by
cervical dislocation. Small intestine was removed and
flushed with 5 mL of PBS supplemented with 1% (v/v)
anti-protease cocktail. The intestinal content was
vortexed and sampled for S-IgA quantification. The
remaining intestinal content suspension was used to flush the
small intestine twice. Intestinal fluid suspensions were
centrifuged (2000×g, 30 min, 4 °C) and frozen at 70 °C
for S-IgA [
R© software (2.12.2 version) was used (R Development
Core Team, 2011). The coefficients of variation (CV) of
the logarithm of the values of S-IgA were compared for
each combination of sampling point/number of flushes
against the smallest CV observed in the intestinal fluid
or in feces. ANOVA was applied to data and the
differences between means were detected by Tukey post hoc
test. Data were considered significantly different when
p < 0.05.
Results and discussion
Need for data transformation
Many of the parametric tests and models commonly
used (linear models, t test, ANOVA) are based on a
normal distribution of data. One way to deal with this is to
transform data, usually by a log transformation, as
antibody titres do not follow normal distribution [
Box-Cox test helps determine the best transformation
procedure out of a family of power transforms (which
includes the logarithm for the power parameter of 0).
We applied this test to study S-IgA. This gives a 95%
confidence interval for the power parameter of (− 1/2,
1/2), therefore suggesting that a log transformation is
suitable. For ease of interpretation we use log in base
10. The statistical justification for log transformation is
to use the proper methodology, as analyzing data in a
transformed scale can change the significance of a test.
Let us assume for a moment that the small intestine was
flushed once and we want to compare if there is a
difference between the BLP and ELP groups. The right
methodology, under normality of data, would be to apply a
t test, then the p value obtained is 0.01173, whereas if
a log-transformation of data is applied, then p value is
Analysis of the variability of data
For detecting differences in the content of S-IgA among
groups, it is necessary to get rid of the so-called
spurious variability. S-IgA may be highly variable among
individuals of the same group [
], reducing the
effectiveness of sample size, and thus creating the
necessity of more samples to achieve a meaningful statistical
power. It is then useful to study the intrinsic variability
of the data to find out means to reduce it. In this study,
the time of the day when mice were sacrificed and the
number of flushes of the small intestine were suspected
to produce variability in the measurements of S-IgA.
The coefficient of variation (CV) is a dimensionless
parameter defined as the ratio of the standard
deviation to the mean, and is considered a useful indicator of
relative consistency of data. The sample size and power
of many common tests of differences are related to the
measurements and the variability through the CV. The
larger the CV, the smaller the power. In this work, the
smallest CV in the intestinal lumen was observed for
the group ELP-TW, whereas in feces it was observed
for the group BLP (Table 1).
It was then determined which CVs were statistically
bigger than the smallest ones in each type of sample. A
permutation test was used [
]. There are two
versions of this test (BII and BIII) but since the minimum
CV is less than 0.6, version BII is the recommended
one. Except for the first wash made in the MLP and at
the ELP, all CVs were statistically bigger (p < 0.05)
compared to corresponding smallest CV. In feces, the CVs
in the sampling points MLP and ELP were not
statistically different compared to the smallest value observed
at the BLP, but this can be due to a lower power of the
In case of a simple t test of difference of means, the
sample size needed to detect a fixed percentage change
of ratio of means depends quadratically on the
coefficient of variation [
]. For example, in the intestinal
fluid, the CV of the BLP-OW group (0.1048) was 3.2
times that of the ELP-TW group (0.0326). Therefore,
the sample size needed in a t test of difference of means
should be 10 times larger to find differences within the
same level of confidence if mice are sampled at the BLP
and making one flush of the small intestine, compared
to the fact of collecting intestinal fluid at the BLP and
flushing the small intestine three times. A smaller CV
would imply a smaller group of animals, contributing
then to the 3R’s principle (Replacement, Reduction and
In relation to the dispersion of individual values of
S-IgA in the intestinal fluid, a maximum of less than four
folds of difference were observed in this work, whereas
in the work of Grewal et al. [
] this factor reached up to
more than 30 folds. In feces, the animal that contained
more S-IgA in the BLP group (the less dispersed one),
presented a value 2.7 times higher than the mouse that
presented the smallest value, being this factor around 4.6
for the other two groups. This is in line with a previous
], where a difference of 4.8 folds was observed.
S-IgA in the intestinal fluid and in feces along a working
For S-IgA in intestine, if the data were technical and
completely independent (that is one value obtained from
each mouse), an ANOVA procedure with two factors
(sampling point and number of flushes) could be used.
However, the same mouse was used for sampling both
flushes, then it was used a mixed-effects linear model
that takes into account the dependency of the response
on the flush.
There was no interaction between sampling points
and the number of flushes (p = 0.4439), which means
that one or three flushes resulted in the same trend of
results along the day (Fig. 1, line plot). The ANOVA
coefficients of both factors were significant (p = 2.9 × 10−12
for the sampling point and 1.7 × 10−5 for the number of
flushes). A post hoc Tukey test for comparison of means
showed that three flushes resulted in a statistically higher
(p = 2 × 10−16) amount of IgA than one flush, for all
sampling points. Additionally, the amount of S-IgA
present in the intestinal lumen at the ELP was significantly
lower than in the MLP or at the BLP (p = 1 × 10−4), but
the groups MLP and BLP were not statistically different
(p = 0.253) (Fig. 1, box plot).
The content of S-IgA in feces (Fig. 2) depended on the
sampling point (p = 0.001208). The post hoc Tukey test
showed no difference between BLP and MLP groups
(p = 0.9997430), whereas S-IgA was significantly lower at
the ELP compared to the BLP (p = 0.0032681) or to the
MLP (p = 0.0031073).
Mice are nocturnal animals and the cyclicity of
hormones and several immune parameters correlates with
the pattern of the animal locomotor activity-resting. The
immune parameter that peaks at one time of day for a
diurnal species peaks about 12 h later for a nocturnal one
]. In humans, S-IgA in saliva peaks during sleep [
whereas in this work, S-IgA peaked by the middle of the
light period, where animals are quieter than during the
dark period. The mouse model of activation of S-IgA for
Salmonella infection prevention has been largely used for
assessing the probiotic potential of strains [
infected with S. Typhimurium were colonized to higher
levels and developed a higher proinflammatory response
during the early rest period for mice, showing that a
functional clock is required for optimal S. Typhimurium
colonization . The results shown here may help choosing
the proper moment along the day for challenging mice
when using the salmonellosis murine model [
It is interesting to note that the smallest CV was
observed after three flushes at the ELP, whereas in feces
the smallest CV was observed early in the morning (BLP).
Considering that the gastrointestinal transit time in mice,
from oral gavage, is close to 8 h [
], then it is likely that
the IgA present in the lumen by the ELP is also present in
feces at the BLP, coinciding that they both displayed the
lowest CVs. In order to reduce data dispersion and make
experiments more powerful, fecal samples may be
harvested at the BLP, whereas sacrifices may take place at its
end. In this sense, fecal IgA could be a suitable parameter
to monitor the activation of the gut immune response, as
S-IgA in feces correlates with those in the lavage samples
]. An ethical reduction in the number of animals used
could be achieved, as sacrifice would proceed as soon as
a peak in S-IgA is observed in feces, instead of using
several groups of mice for different feeding periods.
The influence of the circadian rhythm on S-IgA was
observed in this work. A significant reduction in the
content of S-IgA in the intestinal fluid and in feces was
noticed at the end of the light cycle, which coincides with
the end of a regular working day (8 PM) in any given
animal facility. Three washes of the small intestine were
more effective than one flush to recover S-IgA, with the
smallest CV observed by the ELP. A smaller CV implies
less animals to get the same meaningful results. These
results may be useful when designing animal trails for
the selection of probiotics based on their capacity of
activating S-IgA, since it would imply a more rational use of
experimental animals contributing to the 3R’s principles.
We were not able to measure IgA during the dark period.
The profile of certain cytokines (IL-6, IL10, IL-2, IL-12,
IFNγ and TNFα) in the small and large intestine are
parameters of interest that will be assessed in future full
S-IgA: secretory immunoglobulin A; BLP: beginning of the light period; MLP:
middle of the light period; ELP: end of the light period; OW: one wash; TW:
three washes; CV: coefficient of variation.
PB and SO performed animal experiments and S-IgA determinations. LF and
ET performed statistical analysis of data. GV and JR designed experiments,
analyzed data and wrote and reviewed the manuscript. All authors read and
approved the final manuscript.
We would like to thank Agostina Parodi for her contribution in English revision
of the manuscript.
The authors declare that they have no competing interests.
Availability of data and materials
The datasets used and/or analyzed during the current study are available from
the corresponding author on reasonable request.
Consent for publication
Ethics approval and consent to participate
The experiments with animals were approved by the Ethical Committee for
Animal Experimentation of the Facultad de Ciencias Veterinarias,
Universidad Nacional del Litoral (Esperanza, Santa Fe, Argentina), protocol 291/16
approved on June 26th 2016.
Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
1. Hill C , Guarner F , Reid G , Gibson GR , Merenstein DJ , Pot B , et al. Expert consensus document. The International Scientific Association for Probiotics and Prebiotics consensus statement on the scope and appropriate use of the term probiotic . Nat Rev Gastroenterol Hepatol . 2014 ; 11 ( 8 ): 506 - 14 .
2. Galdeano CM , de Moreno de LeBlan A , Carmuega E , Weill R , Perdigón G . Mechanisms involved in the immunostimulation by probiotic fermented milk . J Dairy Res . 2009 ; 76 ( 4 ): 446 .
3. Padmanabhan P , Grosse J , Asad ABMA , Radda GK , Golay X . Gastrointestinal transit measurements in mice with 99mTc-DTPA-labeled activated charcoal using NanoSPECT-CT . EJNMMI Res . 2013 ; 3 ( 1 ): 60 .
4. Brandtzaeg P , Bjerke K , Kett K , Kvale D , Rognum TO , Scott H , et al. Production and secretion of immunoglobulins in the gastrointestinal tract . Ann Allergy . 1987 ; 59 ( 5 Pt 2 ): 21 - 39 .
5. Generoso SV , Viana M , Santos R , Martins FS , Machado JAN , Arantes RME , et al. Saccharomyces cerevisiae strain UFMG 905 protects against bacterial translocation, preserves gut barrier integrity and stimulates the immune system in a murine intestinal obstruction model . Arch Microbiol . 2010 ; 192 ( 6 ): 477 - 84 .
6. Frece J , Kos B , Beganović J , Vuković S , Šušković J. In vivo testing of functional properties of three selected probiotic strains . World J Microbiol Biotechnol . 2005 ; 21 ( 8-9 ): 1401 - 8 .
7. Ya T , Zhang Q , Chu F , Merritt J , Bilige M , Sun T , et al. Immunological evaluation of Lactobacillus casei Zhang: a newly isolated strain from koumiss in Inner Mongolia, China . BMC Immunol . 2008 ; 9 ( 1 ): 68 .
8. Vinderola G , Matar C , Perdigón G. Milk fermented by Lactobacillus helveticus R389 and its non-bacterial fraction confer enhanced protection against Salmonella enteritidis serovar Typhimurium infection in mice . Immunobiology . 2007 ; 212 ( 2 ): 107 - 18 .
9. Benyacoub J , Pérez PF , Rochat F , Saudan KY , Reuteler G , Antille N , et al. Enterococcus faecium SF68 enhances the immune response to Giardia intestinalis in mice . J Nutr . 2005 ; 135 ( 5 ): 1171 - 6 .
10. Man K , Loudon A , Chawla A . Immunity around the clock . Science . 2016 ; 354 ( 6315 ): 999 - 1003 .
11. Wada M , Orihara K , Kamagata M , Hama K , Sasaki H , Haraguchi A , et al. Circadian clock-dependent increase in salivary IgA secretion modulated by sympathetic receptor activation in mice . Sci Rep . 2017 ; 7 ( 1 ): 8802 .
12. Eriksson E , Royo F , Lyberg K , Carlsson H-E , Hau J . Effect of metabolic cage housing on immunoglobulin A and corticosterone excretion in faeces and urine of young male rats . Exp Physiol . 2004 ; 89 ( 4 ): 427 - 33 .
13. Rodrigues AC , Cara DC , Fretez SH , Cunha FQ , Vieira EC , Nicoli JR , et al. Saccharomyces boulardii stimulates sIgA production and the phagocytic system of gnotobiotic mice . J Appl Microbiol . 2000 ; 89 ( 3 ): 404 - 14 .
14. Manikandan S. Data transformation. J Pharmacol Pharmacother . 2010 ; 1 ( 2 ): 126 - 7 .
15. Grewal HM , Karlsen TH , Vetvik H , Ahrén C , Gjessing HK , Sommerfelt H , et al. Measurement of specific IgA in faecal extracts and intestinal lavage fluid for monitoring of mucosal immune responses . J Immunol Methods . 2000 ; 239 ( 1-2 ): 53 - 62 .
16. Amiri S , Zwanzig S. An improvement of the nonparametric bootstrap test for the comparison of the coefficient of variations . Commun Stat Simul Comput . 2010 ; 39 ( 9 ): 1726 - 34 .
17. Cabras S , Mostallino G , Racugno W. A nonparametric bootstrap test for the equality of coefficients of variation . Commun Stat Simul Comput . 2006 ; 35 ( 3 ): 715 - 26 .
18. van Belle G , Martin DC . Sample size as a function of coefficient of variation and ratio of means . Am Stat . 1993 ; 47 ( 3 ): 165 .
19. Płytycz B , Seljelid R . Rhythms of immunity . Arch Immunol Ther Exp . 1997 ; 45 ( 2-3 ): 157 - 62 .
20. Castillo NA , de Moreno de LeBlanc A , Galdeano CM , Perdigón G . Probiotics: an alternative strategy for combating salmonellosis . Food Res Int . 2012 ; 45 ( 2 ): 831 - 41 .
21. Zacarías MF , Reinheimer J , Forzani L , Grangette C , Vinderola G. Mortality and translocation assay to study the protective capacity of Bifidobacterium lactis INL1 against Salmonella Typhimurium infection in mice . Benef Microbes . 2014 ; 5 ( 4 ): 427 - 36 .
22. Bellet MM , Deriu E , Liu JZ , Grimaldi B , Blaschitz C , Zeller M , et al. Circadian clock regulates the host response to Salmonella . Proc Natl Acad Sci . 2013 ; 110 ( 24 ): 9897 - 902 .
23. Curtis AM , Bellet MM , Sassone-Corsi P , O'Neill LAJ . Circadian clock proteins and immunity . Immunity . 2014 ; 40 ( 2 ): 178 - 86 .