Lactobacillus johnsonii N6.2 Mitigates the Development of Type 1 Diabetes in BB-DP Rats
Full Citation
Permanent Link:
 Material Information
Title: Lactobacillus johnsonii N6.2 Mitigates the Development of Type 1 Diabetes in BB-DP Rats
Series Title: PLoS One
Physical Description: Archival
Language: English
Creator: Valladares, Ricardo
Sankar, Dhyana
Li, Nan
Williams, Emily
Lai, Kin-Kwan
Abdelgeliel, Asmaa Sayed
Gonzalez, Claudio F.
Wasserfall, Clive H.
Larkin, Joseph III
Schatz, Desmond
Atkinson, Mark A.
Triplett, Eric W.
Neu, Josef
Lorca, Graciela L.
Publisher: Public Library of Science
Publication Date: May, 2010
Spatial Coverage:
Funding: Publication of this article was funded in part by the University of Florida Open-Access publishing Fund. In addition, requestors receiving funding through the UFOAP project are expected to submit a post-review, final draft of the article to UF's institutional repository, IR@UF, ( at the time of funding. The Institutional Repository at the University of Florida (IR@UF) is the digital archive for the intellectual output of the University of Florida community, with research, news, outreach and educational materials
 Record Information
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: 10.1371/journal.pone.0010507
System ID: UF00103213:00001

Full Text

OPEN ACCESS Freelyavilable online

Lactobacisus jonsonii N6.2 Mitigates the Development

of Type 1 Diabetes in BB-DP Rats

Ricardo Valladares'", Dhyana Sankar", Nan Li Emily Williamsl, Kin-Kwan Lail, Asmaa Sayed
Abdelgeliel', Claudio F. Gonzalez', Clive H. Wasserfall', Joseph Larkin, III', Desmond Schatzz, Mark A.
Atkinson', Eric W. Triplettl, Josef Neu Graciela L. Lorcal*
1 Department of Microbiology and Cell Science, University of Florida, Gainesville, Florida, United States of America, 2Department of Pediatrics, University of Florida,
Gainesville, Florida, United States of America, 3 Department of Pathology, Immunology, and Laboratory Medicine, University of Florida, Gainesville, Florida, United States
of America


Background: The intestinal epithelium is a barrier that composes one of the most immunologically active surfaces of the
body due to constant exposure to microorganisms as well as an infinite diversity of food antigens. Disruption of intestinal
barrier function and aberrant mucosal immune activation have been implicated in a variety of diseases within and outside
of the gastrointestinal tract. With this model in mind, recent studies have shown a link between diet, composition of
intestinal microbiota, and type 1 diabetes pathogenesis. In the BioBreeding rat model of type 1 diabetes, comparison of the
intestinal microbial composition of diabetes prone and diabetes resistant animals found Lactobacillus species were
negatively correlated with type 1 diabetes development. Two species, Lactobacillus johnsonii and L. reuteri, were isolated
from diabetes resistant rats. In this study diabetes prone rats were administered pure cultures of L. johnsonii or L. reuteri
isolated from diabetes resistant rats to determine the effect on type 1 diabetes development.

Methodology/Principal: Findings Results Rats administered L. johnsonii, but not L. reuteri, post-weaning developed type 1
diabetes at a protracted rate. Analysis of the intestinal ileum showed administration of L. johnsonii induced changes in the
native microbiota, host mucosal proteins, and host oxidative stress response. A decreased oxidative intestinal environment
was evidenced by decreased expression of several oxidative response proteins in the intestinal mucosa (Gpx1, GR, Cat). In L.
johnsonii fed animals low levels of the pro-inflammatory cytokine IFNI were correlated with low levels of iNOS and high
levels of Cox2. The administration of L. johnsonii also resulted in higher levels of the tight junction protein claudin.

Conclusions: It was determined that the administration of L. johnsonii isolated from BioBreeding diabetes resistant rats
delays or inhibits the onset of type 1 diabetes in BioBreeding diabetes prone rats. Taken collectively, these data suggest that
the gut and the gut microbiota are potential agents of influence in type 1 diabetes development. These data also support
therapeutic efforts that seek to modify gut microbiota as a means to modulate development of this disorder.

Citation: Valladares R, Sankar D, Li N, Williams E, Lai K-K, et al. (2010) Lactobacillus johnsonii N6.2 Mitigates the Development of Type 1 Diabetes in BB-DP
Rats. PLoS ONE 5(5): e10507. doi:10.1 371/journal.pone.001 0507
Editor: Krisztian Stadler, Louisiana State University, United States of America
Received February 4, 2010; Accepted April 12, 2010; Published May 6, 2010
Copyright: @ 2010 Valladares et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: These studies were supported, in part, by funds from The Brehm Coalition for Type 1 Diabetes Research and Research Innovation Award, Institute of
Food and Agricultural Sciences University of Florida (lFAS-UFL). The funders had no role in study design, data collection and analysis, decision to publish, or
preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
8 These authors contributed equally to this work.


More than 17 bacterial families encompassing 400 to 500
different microbial species can be found in human adults [1].
These commensal bacteria regulate a myriad of host processes and
provide several nutrients to their host and their symbionts within
the microbial community. In healthy individuals these relation-
ships are -l...n _1.1 to occur in equilibrium. However, disruption of
this equilibrium may contribute to a variety of conditions
including inflammatory bowel disease and atopy [2]. This
connection is gaining credibility as associations between gut
microbiota and either the risk for or presence of a variety of
specific human diseases is demonstrated [2].

a,' PLoS ONE |

Genetics undoubtedly plays a major role in the development of
type 1 diabetes (T1D), however numerous environmental factors
have been suggested that could trigger genetic susceptibility.
Interactions between the intestinal environment, barrier function,
and immune system have been shown to have a major impact in
the rate of T1D development. We previously proposed a
hypothesis involving a trio of interacting factors that may create
a "perfect storm" for T1D development [3]. These factors include
(i) an aberrant intestinal microbiota [4,,5], (ii) a 'leaky' intestinal
mucosal barrier [6], and (iii) altered intestinal immune respon-
siveness [7]. In support of this model, modulation of T1D
].~I,l.... ... ;. in animal models has proved successful -la.. ~1. early
intervention with a variety of dietary alterations [8]. Indeed, the

May 2010 | Volurne 5 | Issue 5 | e10507

PLoS one

Pre-weaning Post-weaning Onset of T1D
(PW)W} (W)~ ~ ~ ~ .........~.~

Days 2 22 69 110 141

L. johnsonii Mitigates T1D

administration of a hydrolyzed casein diet [9] or the administra-
tion of antibiotics [4,,5] has II. .. _11.. ,... 1 the hypothesis that an
aberrant microbiota could accelerate disease development.
However, studies determining the effects of antibiotics in
modulating disease development have not assessed whether
reduction of what could be considered I I .-.l ul....... .. ..~:" flora,
or alternatively, an . I _1 .. 11. of protective flora occurs.
In the BioBreeding diabetes prone (BB-DP) model of T1D, rats
spontaneously develop the automimmune disease due to genetic
susceptibility. A previous study reported a culture-independent
analysis of the bacteria in fecal samples collected from Biobreeding
diabetes resistant (BB-DR) and BB-DP rats [10]. Two genera of
bacteria, Lactobacillus and Bifidobacterzum, emerged as dominant
groups negatively correlated with the onset of T1D. Further
analysis of the Lactobacillus population within these groups
established that Lactobacillus strains with cinnamoyl esterase
activity, L. johnsonii N6.2 and L. reuterz TD1, were negatively
correlated with T1D development [l l].
This report aims to establish whether commensal bacteria
negatively correlated with T1D (i.e., present in BB-DR rats) can
delay or prevent disease onset in BB-DP rats. As proof of concept,
purified L. johnsonii isolated from BB-DR rats were fed to BB-DP
rats. The resulting change in intestinal microbiota composition
was tested for the ability to delay or prevent autoimmunity in BB-
DP rats. A general assessment of the host response to L. johnsonii
administration indicated that the microbe might target an early
signaling pathway conducive to increased levels of interepithelial
junction proteins and mucus secretion, while decreasing oxidative
status and inflammation in the intestine.


Decreased incidence of diabetes in BB-DP rats fed with L.
johnsonii N6.2
Lactobacillus strains isolated from BB-DR rats were administered
to BB-DP rats to analyze their effect on T1D development. L.
reuteri TD1 or L. johnsonii N6.2 suspensions (1 x108 CFU per
animal) were administered individually daily by oral gavage i) pre-
weaning to 1 day old BB-DP rats during mother feeding and ii)
post-weaning to 21 day old BB-DP rats (Fig. 1). The pre-weaning
administration of Lactobacillus did not modify the rate of T1D
development, however the post-weaning administration of L.
johnsonii N6.2 decreased the incidence of T1D compared to the
control animal group (BB-DP, N= 10, P<0.1). A more significant
difference was observed when comparing the L. johnsonii N6.2 fed
group to the L. reuteri TD1 fed group (P<0.04). While disease
incidence in L. johnsonii fed animals decreased, the L. reuteri fed
group showed a similar behavior as the control group (Fig. 2). The
delay in T1D onset was specific to L. johnsonii fed animals. As a
result, in the remainder of the study we compared the responses
referring to three groups: (i) L. johnsonii fed group (includes only
healthy animals), (ii) healthy controls, and (iii) diabetic animals

(includes animals from both the diabetic control and L. johnsonii fed
groups that developed T1D).

Administration of L. johnsonii N6.2 modifies the intestinal
The impact of L. johnsonii N6.2 feeding on the intestinal
microbiota was determined. Main groups of microorganisms were
cultured at the onset of diabetes in sick animals or after 141 days in
animals that remained healthy. The abundance of specific
bacterial genera was also measured by RT-qPCR. No statistically
significant differences were obtained in the stool culturable
bacterial fractions of Lactobacillus, Bacteroides, or in the total
anaerobe counts (data not shown). Interestingly, a close examina-
tion of the colony morphology on Rogosa plates showed a diversity
of morphologies in the control group while the L. johnsonii and L.
reuterz fed groups showed homogenous colony morphology. These
observations were confirmed by isolation of bacteria (10 per plate
per animal) and sequencing of the 16S RNA gene. In the control
group, the most abundant specie was L. murznus (65% of isolates)
while L. intestinalis. .I L. reuter (1 L.johnsonii(8%) were found
in lower proportions. In the L. johnsonii and L. reuterz fed groups
88% and 92%, respectively corresponded to the inoculated
bacteria in each group while the remaining colonies were
identified as L. murznus. Since these results indicate that
predominant species of Lactobacillus are uniform within groups,
observed differences in diabetes development and intestinal
environment between groups may correlate to observed differ-
ences in intestinal microbiota composition.
Similarly, no significant differences were obtained in RT-qPCR
experiments measuring the concentration of Pseudomonas, Bacteroi-
des, Staphylococcus, Bifidobacterium, Clostridium, Lactobacillus, and
enterobacteria in the ileal and colonic content. However, analysis
of the ileal mucosa unveiled a statistically significant increase in the
Lactobacillus population in all rats that did not developed diabetes,
independent of bacterial administration. On the other hand, a
statistically significant increase in concentration of enterobacteria
was found in all diabetic animals, independent of bacterial
administration (Fig. 3). Since no differences in the microbiota
were obtained in the stool samples, but were statistically significant
in the ileal mucosa, the positive effect of L. johnsonii N6.2 could be
exhibited primarily in the intestinal mucosa.

L. johnsonii N6.2 administration modifies expression of
tight junction proteins
Previous studies have reported lower levels of the major
intercellular tight junction protein claudin-1 and greater intestinal
permeability in the BB animal model before the onset of diabetes
[6]. It has been suggested that early increase in intestinal
permeability in the BB-DP rats may allow unregulated passage
of environmental antigens that could trigger the autoimmune
response culminating in T1D development. To determine if
administration of L. johnsonii N6.2 modified intestinal integrity,

Figure 1. Feeding design using BB-DP animals. Arrows in black mark the time that feeding was started. The dashed line indicates daily feeding.
The dashed box indicates the period in which rats developed T1D associated hyperglycemia.
doi:1 0.1 371/journal.pone.001 0507.g001

*, : PLoS ONE |

May 2010 | Volume 5 | Issue 5 | e10507

L. johnsonii Mitigates T1D


)5 10 15 20
Age (weeks)







Figure 2. Kaplan-Meier plot depicting development of T1D in
BB-DP rats. Rats fed A) pre-weaning or B) post-weaning with L.
johnsonii N6.2 (short dashed line), or L. reuteri TD1 (long dashed line)
compared to the PBS fed control (solid line) N= 10 per group.
doi:1 0.1 371/journal.pone.001 0507.g002

macroscopic modifications in the mucosal architecture were
examined on hematoxvlmn and easin stained slides of distal small
intestine. No morphological differences between the L. johnsonii fed
group, control healthy, or diabetic animals were found in villus
height or width or crypt depth (data not shown). Necrosis was not
observed in any samples. Interestingly, the number of goblet cells
was ;_ ..:0. ,,.11 higher in both healthy controls and L.johnsonii fed
animals when compared to the diabetic group (Fig. 4A, P<0.05).
At the molecular level we measured the expression of genes
encoding claudin-1 and occludin proteins involved in intercellular
tight junction assembly and maintenance in the intestine. The
healthy control and diabetic groups showed low expression of the
sealing claudin-1 and high levels of occludin tight junction related
transmembrane protein compared to the L. johnsonii fed group
(Fig. 4B). Specifically, the feeding of L. johnsonii upregulated the
expression of claudin-1 and decreased the expression of occludin.
Small differences were observed when comparing the healthy
control group to the diabetic animals. Consequently, L. johnsonii
N6.2 feeding could ameliorate the intestinal barrier dysfunction
observed in this animal model.

L. johnsonii decreases the host intestinal oxidative stress
Among the destructive effects of reactive oxygen species (ROS)
generated during early intestinal disease development is the
disruption of epithelial tight junctions [12]. The levels of hexanoyl-

Figure 3. Quantification using real time qPCR of lactobacilli (A)
and enterobacteria (B) from ileal mucosa. The values are
expressed as mean of the percentages from total bacteria determined
from 5 ng of DNA.*" indicates significant differences (P<0.05) between
healthy and diabetic animals (N= 6 per group).
doi:1 0.1 371/journal .pone.001 0507.g003

lysine, a biomarker for oxidative stress [13], were determined by
ELISA on ileal mucosa and were variable among the animals
tested. However, levels were ;: _.1G. ,,.11 higher (P<0.05) in
diabetic animals (53121 pM min ') when compared to L.
johnsonii group and healthy controls (14'10 I*M min ').
To determine the specific mechanisms involved, the ileal
mRNA levels of enzymes involved in ROS detoxification pathways
in the host were determined (Fig. 5A). To ease the presentation of
data we illustrated the gene levels in terms of expression in the L.
johnsonii fed group. The value of one represented expression levels
in the L. johnsonii fed group. Of the genes measured, superoxide
dismutase 2 .. .I_ catalase (C. ;I in.1 .11,.:.... reductase (GR), and
I,, .11,;..... peroxidase (Gpxl) were induced in diabetic animals
compared the L. johnsonii group. Sod2 and Gpx1 were induced in
the healthy controls compared the L. johnsonii group. Superoxide
dismutase 1 (Sodl) was the exception as it was not modified under
any condition. The expression of Sod2 and Gpx1 was induced in
the diabetic animals (~4.5 fold and ~4 fold, respectively: P<0.05)
and to a lesser extent Cat and GR (~2 and 1.8 fold, respectively).
By comparing the mRNA levels of the healthy (control and L.
johnsonii fed) with diabetic animals two primary responses were
observed: (i) genes negatively correlated with healthy status (Cat

a, : PLoS ONE |

May 2010 | Volume 5 | Issue 5 | e10507


'@ 80-




Diabetic A



2 o 0.6

Wr 0.2

Age (weeks)


L johnsonli Control Diabetic

iNOS cc ,
e-actin B

Figure 5. Assessment of the oxidative stress response in the
host. (A) RT-qPCR analysis of the expression of genes linked to the
oxidative stress response in the host. Relative amounts of iNOS, Cox2,
Sod1, Sod2, Gpx1, Cat, and GR were calculated by subtracting the
internal control (B actin) and changes in expression levels were
calculated relative to the value in L. johnsonii feed group (expression
= 1). Grey bars: relative expression in the healthy control; Black bars:
relative expression in the diabetic animals. The values are means +S.D.
(N =6); *P<0.05; oP<0.01,**P<0.0001. (B) Western blot analysis of iNOS
levels. B-actin was used as internal control.
doi:1 0.1 371/journal .pone.001 0507.g005

status of the animals. Repressed expression of Cox2 in the diabetic
group (Fig. 5A, P<0.001) and healthy control group in
comparison to the L. johnsonii group was also observed.
IFNY is an important mediator of inflammatory responses with
pleiotropic effects in the host. It was previously reported that IFNI
induces the expression of iNOS [14r] while repressing the
expression of Cox2 [15]. The aim was to determine if a negative
correlation existed between the levels of pro-inflammatory
cytokines, particularly IFNI and TNFa, and the L. johnsonii-
mediated decrease in oxidative stress response in the host. The
mRNA levels of TNFa decreased ~7 fold (P<0.05) between the
healthy and diabetic animals, but no differences between the
healthy control group and L. johnsonii fed group were observed
(Fig. 6). The results indicate that the low expression of TNFa is
correlated with healthy status and not with the administration of
bacteria. The expression of IFNI, on the contrary, was directly
linked to the administration of L. johnsonii. Diabetic animals
exhibited a ~-20 fold higher expression (P<0.005) of IFNI
compared to the L. johnsonii fed group. No significant differences
were observed between the healthy controls and diabetic animals
indicating a specific contribution of the probiotic microorganism
to the decrement of the inflammatory response.


In this study we report that the administration of L. johnsonii
N6.2 decreased the progression to T1D when administered post-
weaning. Among the possible mechanisms in which Lactobacillus
exerts beneficial effects for the host are: (i) as a physical barrier
ul.:;1.;::,. 11.. passage of inflammatory antigens, (ii) degradation of

L. johnsonii Mitigates T1D


L. johnsonii Healthy controls Diabetic

iNos cox2 Sod1 Sod2 Gpx1 Cat

u I



Figure 4. Effect of L. Johnsonii administered post weaning on
prevalence of goblet cells (A) and on mRNA levels of tight
junction genes (B). Hematoxylin and eosin stained slides of distal
small intestine were examined for morphological changes. (A)
Percentage of goblet cells in the distal small intestine by treatment
group. (B) RT-qPCR analysis of the expression of tight junction genes.
Relative amounts of claudin 1 and occludin were calculated by
subtracting the internal control (B actin) and changes in expression
levels were calculated relative to the value in the L. johnsonii fed group
(expression = 1). Grey bars: Relative expression in the healthy control,
Black bars: relative expression in the diabetic animals. The values are
moTan~s3 1SD j) r 61).po;e 5 0.001; **P<0.0001'

and GR), and (ii) genes negatively correlated with L. johnsonii
administration (Sod2 and Gpx1).
ROS lead to the synthesis of nitric oxide by inducible nitric
oxide synthase (iNOS). The mRNA levels of iNOS were
..1.,.1repressed in the L.johnsonii fed group when compared
to diabetic animals as well as healthy controls (~22 fold, Fig. 5A,
P<0.0001). However, western blot analysis showed that iNOS is
reduced in both L.johnsonii fed and healthy control groups (Fig. 5B)
indicating that low levels of iNOS are correlated with the healthy

a, : PLoS ONE |

May 2010 | Volume 5 | Issue 5 | e10507


Claudin 1

i I I

L. johnsonii Mitigates T1D



bacteria such as L. johnsonii exert beneficial effects on the host
intestines by releasing antioxidant compounds -la..n_~1. dietary
fiber hydrolysis. The release of antioxidant compounds by
probiotic bacteria is relevant since an enhanced oxidative stress
response triggered by the excessive production of reactive oxygen
species is observed in T1D and other diseases [20,21]. It has been
shown that small doses of antioxidant compounds decrease the
mncidence of diabetes in streptozotocin (STZ)-induced diabetic
mice [22]. Also, we have previously reported a negative
correlation between the presence of bacteria able to release
bioactive antioxidant components from phenolic compounds and
the BBDP model [11]. The L. johnsonii N6.2 strain used in this
study possesses two esterases that can release cinnamic acid and
other phenolic compounds with anti-inflammatory properties.
However, the direct role of these enzymes and antioxidant
compounds on T1D ]...1..._. ... ;. requires further investigation.
The oxidative status of the ileal mucosa was assessed by
I! ~ ;~ 11.. mRNA levels of genes involved in the host oxidative
stress response. Compared to the L. johnsonii fed group, genes
encoding Sod2, Gpx1, Cat, and GR were induced in diabetic
animals. However, Gpx1 and Sod2 expression was also induced in
healthy controls compared to the L. johnsonii group. Overall, the
lower level of these markers of oxidative stress in the L. johnsonii
group indicates a more favorable anti-inflammatory environment
in the ileum with lower levels of ROS.
Nitric oxide is a signaling molecule that links inflammation and
the development of T1D. An increased transcription and
translation of the inducible nitric oxide synthase (iNOS) gene
has been associated with T1D development in BB-DP rats
[23,24r]. Here, the expression level of iNOS (and its inducing
cytokine, IFNI) was downregulated in the L. johnsonii group
compared to the diabetic control group. Interestingly, the levels
of Cox-2 showed the opposite effect. Cox-2 has been reported to
be mainly induced in activated macrophages and other
inflammatory cells [25]. However, Luo et al. [26] showed that
the presence of Cox-2 in (}-cells decreased during progression of
diabetes in the NOD mouse model. In this study, the mRNA
levels of Cox-2 in the ileum increased in the healthy animals with
the highest expression in the L. johnsonii fed group. The increase
in Cox-2 expression observed also correlates with a higher
number of goblet cells in the intestine of healthy rats, in
agreement with Luo et al. [27]. However, these results suggest
that the expression of Cox-2, and its prostaglandin products, may
have a protective effect. Gilroy, Colville-Nash and others [28,29]
have shown that the synthesis of cyclopentenone prostaglandins
are determinant during inflammatory resolution.
In this study, low expression of pro-inflammatory cytokines
IFNI correlated with the administration of L. johnsonii N6.2,
whereas low expression of TNFa correlated with overall healthy
status. These results indicate that the mechanisms involved in T1D
inhibition in this study may be different from the effect mediated
by the general probiotic treatment VSL3. Previous reports in the
NOD model of diabetes have shown that the administration of the
probiotic formulation VSL3 decreased the incidence of T1D
-1.,.... _1. IL10 immunomodulation [30] and induction of pro-
inflammatory cytokine IFNI.
Is has been suggested that a highly permeable epithelium, or
'leaky gut', fails to inhibit the passage of intestinal content antigens.
This inappropriate exposure to antigens may trigger regulatory T-
cells in an autoimmune cascade [3]. We propose that presence of
L. johnsonii N6.2 in the intestinal epithelium helps maintain
epithelial barrier function by increasing mucus and tight junction
protein production while ,1. I., .~~ In _ll effects of ROS. L.johnsonii
may be targeting an early step in the inflammatory signaling







Figure 6. mRNA levels of the pro-inflammatory cytokine genes,
IFNyI and TNFa linked to the oxidative stress response in the
host. Relative expression was calculated as previously described
relative to the value in the L. johnsonii feed group (expression = 1)-
Relative expression in the L. johnsonii feed group (black bars), healthy
control (dark grey bars), and diabetic animals (grey bars). The values are
means +S.D. (N= 6); *P<0.05; #P<0.01.
doi:1 0.1 371/journal.pone.001 0507.g006

toxic components, (iii) release of nutrients, and (iv) production of
anti-inflammatory compounds. In the BB rat model a combination
of all of these mechanisms is possible due to the overly permeable
characteristic of the small intestine [6].
A higher number of goblet cells in the L. johnsonii fed and
healthy control groups was observed when compared to the
diabetic group. Goblet cells constantly produce mucus, which has
a dual role of protecting the mucosa from adhesion of certain
microorganisms while providing an initial binding site, nutrient
source, and matrix on which commensal bacteria can proliferate
[16]. In this study the observed decrease in goblet cells correlated
with the sick status of the animals, which could result in a lower
production of intestinal mucins. Of note, Mack et al. [17] showed
that L. plantarum has a direct effect on intestinal epithelial cells by
inducing secretion of mucins that diminish enteric ].,11~..._ ...
binding to mucosal epithelial cells. The sustained mucus
production present in L. johnsonii and healthy control groups could
prevent damage from enteric ].,11..._. .. While a higher number
of goblet cells was associated with all healthy specimens, claudin
expression was specifically induced following feeding of L. johnsonii.
This change in expression indicates a direct effect of the bacteria
on intestinal barrier integrity.
A puzzling result was that the T1D preventative effect of L.
johnsonii was only observed when administration began in the post-
weaning period rather than during pre-weaning. A combination of
scenarios could explain these results. Since the pre-weaning group
received continued L. johnsonii administration -1I.,....1...... the
experiment (Fig.1i), the beneficial effects of these bacteria on T1D
development may be dependent on host immune system maturity.
It is possible that the stress generated by very early intervention in
specimen development could offset the balance of microbial
intestinal composition or immune responsiveness.
The difference in prevention capabilities between L. reuteri and
L. johnsonii post-weaning may involve their response to diet
composition as this factor has been shown to be an important
contributor to development of T1D [18,19]. ,\la1...n _1. the effects
of specific diets have been reported, the mechanisms behind these
effects remain obscure. One possibility is that certain commensal

a,' PLoS ONE |

May 2010 | Volume 5 | Issue 5 | e10507


Analysis of the intestinal microflora by Real-time
DNA extractions from samples preserved at --80'C in
RNAlater* solution (Ambion, Austin, USA) were perform using
the QIAamp DNA Stool Mini kit (Qiagen Sciences, Maryland,
USA) following the manufacturer's instructions. We investigated
the most important groups of the rat fecal microbiota using DNA
extracts from each rat as a template for RT-qPCRs using the
primers described on Table 1. Quantitative PCRs were performed
in a reaction volume of 20 1*l containing lx iQ SYBR Green
Supermix (Bio-Rad, Hercules, USA), 200 nM each forward and
reverse primers, and 5 ng of DNA extracted from the stool
samples. DNA concentrations were determined with the Nano-
dropr spectrophotometer. Amplification and detection of DNA
were performed in duplicate with the i~ycler detection system
(BioRad, Hercules, USA) with optical grade 96-well PCR plates
and optical film. The reaction conditions were 50'C for 2 min and
95'C for 10 min, followed by 45 cycles of 95'C for 15 s and 62'C
for 1 min. Data analysis was conducted with the software supplied
by Bio-Rad (Hercules, USA).
DNA amplification standard curves were constructed using
purified genomic DNA in the range 10 fg to 1 ng of L. reuteri, L.
johnsonii, Staphylococcus sp., Bacteroides dorei and E. coli as previously
described in Roesch et al. [10]. The conversion of the amount
DNA of the different bacterial groups into cell numbers in the stool
samples was determined considering the genome size for each
bacteria and the copy number of the 16S RNA gene as described
by Byun et al. [31] and Matsuda et al. [32].

Intestinal morphology
Intestinal integrity was evaluated by histology. Neutral buffered
formalin (10%, V/V) -fixed ileum samples were embedded in
paraffin; cut into 4 Im sections, mounted on glass slides, and
stained with hematoxylin and eosin (H&E) according to standard

Table 1. Primer sequences based on the 16S rRNA used to
discriminate between different groups of bacteria by RT-

L. johnsonii Mitigates T1D

pathway resulting in a more tolerogenic environment which
reduces the overall oxidative stress.
In summary, this study observed a delay in the onset of T1D as
well as physiological and immunological differences in the gut
which correlated with the presence of Lactobacillus johnsonii N6.2.
Nevertheless, further research is needed to elucidate the intricacies
of the relationship between autoimmune disease, intestinal health,
and gut microbiota.

Materials and Methods

Ethics Statement
All animal work has been approved by the University of Florida
Institutional Review Board (IRB). Animal housing standards were
as prescribed by the Association for Assessment and Accreditation
of Laboratory Animal Care (AAALAC) with two male or female
rats per cage under in d..._.,,.n.. conditions. All rats were in the
same room at the same temperature and under the same light. All
rats received the same amount of water and food. The weights of
these animals were measured weekly. The analysis indicated that
all animals were gaining weight at the same pace with no
significant discrepancies when healthy.

Bacterial strains
Two bacterial strains: Lactobacillus johnsonii N6.2 or Lactobacillus
reuteri TD1 isolated from BB-DR rats [l l] were grown in MRS
broth (REMEL, Lenexa, USA) at 37'C for 16 h. Cells were
centrifuged at 3000 rpm, pellets washed with sterile PBS buffer.
Aliquots containing 10'o cells ml- were stored at --80'C until
used. Cell viability was determined by plate dilution method on
three aliquots after -1.. T u. For feeding experiments, new aliquots
were thawed immediately before administration.

BB-DP rats experimental design
Bacterial strains were administered to BB-DP rats (Biomedical
Research Models, Worcester, MA) to test whether they would delay
or inhibit the onset of T1D. L. reuteri TD1 or L. johnsonii N6.2
suspensions (108 CFU) were administered daily by oral gavage i) pre-
weaning to 1 day old BB-DP rats during mother feeding and ii) post-
weaning at 21 days old BB-DP rats. The control group was
administered PBS only. Starting at the age of 60 days, the blood
glucose levels of the animals were taken weekly using a glucose
monitor kit. If the glucose levels were ;11_1. I sl,,., 250 mg/dl for two
consecutive days, then the rat was considered to have developed
diabetes. Once a rat developed diabetes, it was sacrificed, and organs
and tissues were harvested and preserved for analysis. For anatomic
studies, the small intestine was harvested and flushed with phosphate-
buffered saline at 4'C to remove intraluminal contents. The small
intestine was divided into 3 equal sections to demarcate the
duodenum, jejunum and ileum. Small pieces of each section were
fixed in 10% (v/v) neutral buffered formalin for 24 hours for light
microscopy or preserved in RNAlater* solution (Ambion, Austin,
USA) for RT-qPCR analysis'

Analysis of the intestinal microflora by viability countS
Samples taken from colonic content were immediately placed in
5 ml of sterile PBS buffer. After dilution viable counts were
obtained using MRS (REMEL, Lenexa, USA) adjusted to pH 5.5
for lactobacilli, BBE (BD BBL, Sparks, USA) for Bacteroides and
BHI ~ ;1, : sheep blood (REMEL, Lenexa, USA) for anaerobes
incubated anaerobically (BD BBL GasPack Plus; Sparks, USA)
for 48 h at 37'C. Enterobacteria counts were determined using
McConkey agar plates (REMEL, Lenexa, USA) incubated under
aerobic conditions at 37'C for 24 h.

*,' PLoS ONE |



RProk 1492
Lactobacillus F-lacto
Bacteroides AllBac296F
Clostridium Ccoc 07
Enterobac- En-Isu-3F
Pseudomonas PSD7F
Staphylococcus STPYF
Bifidobacterium F-bifido



doi:1 0.1371 /journal.pone.001 0507.t001

May 2010 | Volurne 5 | Issue 5 | e10507

L. johnsonii Mitigates T1D

procedures. Villus height, width and crypt depth were measured
using a Nikon microscope (Universal Imaging Corp., Westchester,
PA) with an ocular micrometer without the examiner knowing the
group assignment. The intestinal injury was evaluated using a
semiquantitative scoring system ranging from 0 to 4 modified by
Arumugam et al. [33]. Normal mucosa was scored as grade 0.
Epithelial cell damages, including loss of cells and separation of the
epithelial cells from the underlying villus were scored between
grades 1-3, while loss of villus tissue was scored as grade 4.
Intestinal sections were also analyzed for goblet cells per total cells
within a villus. For each animal, counts from 6 villi for each slide in
three different regions of the slide were averaged.

Real-Time qPCR of host responses
DNA and RNA extractions from samples preserved at -80'C
in RNAlater* solution (Ambion, Austin, USA) were perform using
the Ilustra TriplePrep kit (GE Health care, UK) following the
manufacturer's instructions. cDNA was synthesized using iS-
criptTM cDNA synthesis kit (Bio-Rad, Hercules, USA) and qRT-
PCR were performed as described above. The primers used are
described in Table 2.

Western Blot analysis of iNOS expression
Protein expression was analyzed using whole cell lysates. Rat ileum
samples were weighed, minced, and disaggregated by incubation at
250 rpm and 37'C in phosphate-buffered saline (1:2, w/v) with
0.25% collagenase. Samples were immediately place on ice and
homogenized by vortexing with glass beads b a. Life Science)

containing Complete Mini Protease Inhibitor Cocktail (Roche,
Mannheim Germany). Samples were centrifuged at 1 11111 for
10 min at 4'"C. Protein concentration was determined by Bradford
method (Bio-Rad Protein Assay). 40 ug of protein per sample was
separated using sodium dodecyl sulfate-polyacrylamide electropho-
resis and transferred onto Nitroplus membranes (MSI, Flanders Ma)
using a semi-dry transfer method. Membranes were blocked for 1 h
in phosphate buffered saline with 0.075% tween 20 (T-PBS) and 5%
milk. Membranes were incubated with mouse anti-iNOS antibody
(1:1000) (Abcam, Cambridge, Ma) or mouse anti- p-actin (1:10,000)
(Abcam, Cambridge, Ma) in T-PBS at 4'OC overnight, and
subsequently washed twice in T-PBS for five minutes. Incubation
in horseradish peroxidase-conjugated anti-mouse antibody was
performed for 1 h and signal was detected using enhanced
chemiluminescent system (Amersham Pharmacia Biotech). p-actin
was utilized as an internal control.

Hexanoyl-lys enzyme-linked immunosorbent assay
Relative lipid peroxidation was determined by analyzing
hexanoyl-lys levels by ELISA on rat ileum cell suspensions.
20 mg of each rat ileum sample was finely minced on a cold glass
slide and suspended in 300 ul lX PBS +0.25% collagenase type I
(Invitrogen, Carlsbad, California). Samples were vortexed vigor-
ously at 5 min intervals while incubating at 37'C for 30 min. Free
cell suspensions were separated from remaining connective tissue
fractions after incubation. Cell concentration was determined by
optical density at 600 nm using a Syngery HT microplate reader
(BioTek Instruments, Winooski, VT) Each sample was split into

Table 2. Primer sequences used to analyze the host response by RT-qPCR.









B-actin Fw
B-actin Rv
Cldn-1 Fw
Cldn-1 Rv
Occludin Fw
Occludin Rv
TNF-a Fw
TNF-a Rv


Tumor Necrosis Factor-at

Inducible Nitric Oxide Synthase

Glutathione Peroxidase 1


Glutathione Reductase

Superoxide Dismutasel

Superoxide Dismutase 2


I doi:1 0.1 371 /journal.pone.001 0507.t002

*, : PLoS ONE |

May 2010 | Volurne 5 | Issue 5 | e10507

1ml aliquots (control and experimental sets), pelleted at
5000 rpm, and washed with 1 ml lX PBS twice. Cell pellets
were resuspended in 100 1*l of lX PBS and 100 1*l of lX PBS was
added to the control set, and 100 pl of lX PBS +2 Igg/ml
hexanoyl-lysine monoclonal antibody (JalCA, Shizuoka, Japan)
was added to the experimental set. Samples were incubated at
37'C for 1 hr, followed by two washes in lX PBS. 100 ul of lX
PBS +80 ng/ml of peroxidase labeled anti-mouse monoclonal
antibody (Amersham Pharmacia Biotech, Pittsburgh, PA) was
added to both sets and incubated for 1 hr at 37'C. Following
incubation, cells were washed twice and resuspended in 100 ul lX
PBS. A 20 mM reaction mixture of 0-Phenylenediamine 1 e. .
St. Loius, MO) was prepared in a 50 mM phosphate citrate buffer
pH = 5 and kept dark. Immediately before reading, 30% H,02
was added to the reaction mixture for a final concentration of
0.04%. 100 ul of reaction mixture was added to each sample, and
samples were read continuously at 450 nm and 570 nm for
30 min using Syngery HT microplate reader (BioTek Instruments,
Winooski, VT). Specific activity was calculated as the amount of
product (Igmol.min ') and normalized the number of cells. This
assay was performed in triplicate for each sample.


1. Yan F, Polk DB (2004) Commensal bacteria in the gut: learning who our friends
are. Curr Opin Gastroenterol 20: 565-571.
2. Sokol H, Seksik P, Furet JP, Firmesse O, Nion-Larmurier I, et al. (2009) Low
counts of Faecalibacterium prausnitzii in colitis microbiota. Inflamm Bowel Dis
15: 1183-1189.
3. Vaarala O, Atkinson MA, NeuJ (2008) The "perfect storm" for type 1 diabetes -
the complex interplay between intestinal microbiota, gut permeability, and
mucosal immunity. Diabetes 57: 2555-2562.
4. Brugman S, Klatter FA, Visser JTJ, Wildeboer-Veloo ACM, Harmsen HJM,
et al. (2006) Antibiotic treatment partially protects against type 1 diabetes in the
Bio-Breeding diabetes-prone rat. Is the gut flora involved in the development of
type 1 diabetes? Diabetologia 49: 2105-2108.
5. Schwartz RF, NeuJ, Schatz D, Atkinson MA, Wasserfall C (2007) Comment on:
Antibiotic treatment partially protects against type 1 diabetes in the Bio-
Breeding diabetes-prone rat. Is the gut flora involved in the development of type
I diabetes? Diabetologia 50: 220-221.
6. NeuJ, Reverte CM, Mackey AD, Liboni K, Tuhacek-Tenace LM, et al. (2005)
Changes in intestinal morphology and permeability in the biobreeding rat before
the onset of type 1 diabetes. J Pediatr Gastroenterol Nutr 40: 589-595.
7. Kamradt T, Mitchison NA (2001) Tolerance and autoimmunity. N EnglJ Med
344: 655-664.
8. Caicedo RA, Li N, Des Robert C, Scumpia PO, Hubsher CP, et al. (2008)
Neonatal formula feeding leads to immunological alterations in an animal model
of type 1 diabetes. Pediatr Res 63: 303-307.
9. VisserJ, Brugman S, Klatter F, Vis L, Groen H, et al. (2003) Short-term dietary
adjustment with a hydrolyzed casein-based diet postpones diabetes development
in the diabetes-prone BB rat. Metabolism 52: 333-337.
10. Roesch LF, Lorca GL, Casella G, Giongo A, Naranjo A, et al. (2009) Culture-
independent identification of gut bacteria correlated with the onset of diabetes in
a rat model. ISME J 5: 536-548.
I1. Lai KK, Lorca GL, Gonzalez CF (2009) Biochemical properties of two
cinnamoyl esterases purified from a Lactobacillus johnsonii strain isolated from
stool samples of diabetes-resistant rats. Appl Environ Microbiol 75: 5018-5024.
12. SchulzkeJD, Ploeger S, Amasheh M, Fromm A, Zeissig S, et al. (2009) Epithelial
tight junctions in intestinal inflammation. Ann N Y Acad Sci 1165: 294-300.
13. Kato Y, Yoshida A, Naito M, Kawai Y, Tsuji K, et al. (2004) Identification and
quantification of N(epsilon)-(Hexanoyl)1ysine in human urine by liquid
chromatography/tandem mass spectrometry. Free Radic Biol Med 37:
14. Teng X, Zhang H, Snead C, Catravas JD (2002) Molecular mechanisms of
iNOS induction by IL-1 beta and IFN-gamma in rat aortic smooth muscle cells.
AmJ Cell Physiol 282: C144-152.
15. Klampfer L, HuangJ, Kaler P, Sasazuki T, Shirasawa S, et al. (2007) STATI-
independent inhibition of cyclooxygenase-2 expression by IFNgamma, a
common pathway of IFNgamma-mediated gene repression but not gene
activation. Oncogene 26: 2071-2081.
16. Belley A, Keller K, Gottke M, Chadee K (1999) Intestinal mucins in
colonization and host defense against pathogens. Am J Trop Med Hyg 60:
17. Mack DR, Michail S, Wei S, McDougall L, Hollingsworth MA (1999) Probiotics
inhibit enteropathogenic E. coli adherence in vitro by inducing intestinal mucin
gene expression. AmJ Physiol 276: G941-950.

<, : PLoS ONE |

18. Elliott RB, Martin JM (1984) Dietary protein: a trigger of insulin dependent
diabetes in the BB rat? Diabetologia 26: 297-299.
19. Visser J, Hillebrands JL, Walther Boer M, Bos NA, Rozing J, et al. (2009)
Prevention of diabetes by a hydrolysed casein-based diet in diabetes-prone
BioBreeding rats does not involve restoration of the defective natural regulatory
T cell function. Diabetologia 52: 1445-1447.
20. MehtaJL, Rasouli N, Sinha AK, Molavi B (2006) Oxidative stress in diabetes: a
mechanistic overview of its effects on atherogenesis and myocardial dysfunction.
IntJ Biochem Cell Biol 38: 794-803.
21. Robertson RP, Harmon JS (2006) Diabetes, glucose toxicity, and oxidative
stress: a case of double jeopardy for the pancreatic islet ri-cell. Free Radical Biol
Med 41: 177-184.
22. Ohnishi M, Matuo T, Tsuno T, Hosoda A, Nomura E, et al. (2004) Antioxidant
activity and hypoglycemic effect of ferulic acid in STZ-induced diabetic mice
and KK-Ay mice. Biofactors 21: 315-319.
23. Kleemann R, Rothe H, Kolb-Bachofen V, Xie QW, Nathan C, et al. (1993)
Transcription and translation of inducible nitric oxide synthase in the pancreas
of prediabetic BB rats. FEBS Lett 328: 9-12.
24. Mendez II, Chung YH, Jun HS, Yoon JW (2004) Immunoregulatory role of
nitric oxide in Kilham rat virus-induced autoimmune diabetes in DR-BB rats.
J Immunol 173: 1327-1335.
25. CorbettJA, Kwon G, Marino MH, Rodi CP, Sullivan PM, et al. (1996) Tyrosine
kinase inhibitors prevent cytokine-induced expression of iNOS and COX-2 by
human islets. AmJ Physiol 270: C1581-1587.
26. Luo C, Kallajoki M, Gross R, Mulari M, Teros T, et al. (2002) Cellular
distribution and contribution of cyclooxygenase COX-2 to diabetogenesis in
NOD mouse. Cell Tissue Res 310: 169-175.
27. Luo C, Laine VJ, Ylinen L, Teros T, Makinen M, et al. (2002) Ristimaki A,
Simell O: Expression of cyclooxygenase-2 in intestinal goblet cells of pre-diabetic
NOD mice. Acta Physiol Scand 174: 265-274.
28. Gilroy DW, Colville-Nash PR (2002) New insights into the role of COX 2 in
inflammation. J Mol Med 78: 121-129.
29. Trivedi SG, Newson J, Rajakariar R, Jacques TS, Hannon R, et al. (2006)
Essential role for hematopoietic prostaglandin D2 synthase in the control of
delayed type hypersensitivity. Proc Natl Acad Sci USA 103: 5179-5184.
30. Calcinaro F, Dionisi S, Marinaro M, Candeloro P, Bonato V, et al. (2005) Oral
probiotic administration induces interleukin-10 production and prevents
spontaneous autoimmune diabetes in the non-obese diabetic mouse. Diabeto-
logia 48: 1565-1575.
31. Byun R, Byun R, Nadkarni MA, Chhour KL, Martin FE, et al. (2004)
Quantitative analysis of diverse Lactobacillus species present in advanced dental
caries. J Clin Microbiol 42: 3128-3136.
32. Matsuda K, Tsuji H, Asahara T, Matsumoto K, Takada T, et al. (2009)
Establishment of an analytical system for the human fecal microbiota, based on
reverse transcription-quantitative PCR targeting of multicopy rRNA molecules.
Appl Environ Microbiol 75: 1961-1969.
33. Arumugam TV, Arnold N, Proctor LM, Newman M, Reid RC, et al. (2003)
Comparative protection against rat intestinal reperfusion injury by a new
inhibitor of sPIA2, COX-1 and COX-2 selective inhibitors, and an LTC4
receptor antagonist. Br J Pharmacol 140: 71-80.
34. Sokol H, Pigneur B, Watterlot L, Lakhdari O, Bermuidez-Humaran LG, et al.
(2008) Faecalibacterium prausnitzii is an anti-inflammatory commensal

May 2010 | Volume 5 | Issue 5 | e10507

L. johnsonii Mitigates T1D

Statistical Analysis
Statistical analysis for significant differences was performed
according to the Student's t-test for unpaired data or by the
nonparametric Mann-Whitney. Differences with P<0.05 and
lower were considered significant. Data was analyzed by
GraphPad Prism (GraphPad Software, San Diego, USA).


We wish to thank Dr. V. Mai for scientific discussions. We wish to thank
Beverly Driver and Fernando Pagliai for their valuable technical assistance.

Author Contributions

Conceived and designed the experiments: CFG MAIAJN GLL. Performed
the experiments: RV DS NL EW ASA KKL CHW GLL. Analyzed the
Contributed reagents/materials/analysis tools: CHW DS MAA EWT JN
GLL. Wrote the paper: RV CFG JN GLL.

bacterium identified by gut microbiota analysis of Crohn disease patients. Proc
Natl Acad Sci USA 105: 16731-6.
35. Delroisse JM, Boulvin AL, Parmentier I, Dauphin RD, Vandenbol M, et al.
(2008) Quantification of Blifiobacterium spp. and Lactobacillus spp. in rat fecal
samples by real-time PCR. Microbiol Res 163: 663-670.
36. Guo X, Xia X, Tang R, ZhouJ, Zhao H, et al. (2008) Development of a real-
time PCR method for Firmicutes and Bacteroidetes in faeces and its application
to quantify intestinal population of obese and lean pigs. Lett Appl Microbiol 47:
37. Matsuda K, Tsuji H, Asahara T, Kado Y, Nomoto K (2007) Sensitive
quantitative detection of commensal bacteria by rRNA-targeted reverse
transcription-PCR. Appl Environ Microbiol 73: 32-39'
38. Zhu F, Qian C (2006) Berberine chloride can ameliorate the spatial memory
impairment ad increase the 1xpressio nho terl ukin-lb ta and inducible ni ric

39. Charoenphandhu N, Wongdee K, Tudpor K, PandaranandakaJ, Krishnamra N
(2007) Chronic metabolic acidosis upregulated claudin mRNA expression in the
duodenal enterocytes of female rats. Life Sci 80: 1729-1737.

*, : PLoS ONE |

40. Gye MC (2003) Changes in the expression of claudins and transepithelial
electrical resistance of mouse Sertoli cells by Leydig cell coculture. IntJ Androl
26: 271-8.
41. Neumann H, Schmidt H, Wilharm E, Behrens L, Wekerle H (1997) Interferon
gamma gene expression in sensory neurons: evidence for autocrine gene
regulation. J Exp Med 186: 2023-2031.
42. Jiang W, Desjardins P, Butterworth RF (2009) Direct evidence for central
proinflammatory mechanisms in rats with experimental acute liver failure:
protective effect of hypothermia. J Cereb Blood Flow Metab 29: 944-952.
43. Tam NN, Gao Y, Leung YK, Ho SM (2003) Androgenic regulation of oxidative
stress in the rat prostate: involvement of NAD(P)H oxidases and antioxidant
defense machinery during prostatic involution and regrowth. AmJ Pathol 163:
44. Ttirler AqtM re BA, Pe zonneeMA Ove haus M, KalffJC, et al. (2002) Colonic

May 2010 | Volurne 5 | Issue 5 | e10507

L. johnsonii Mitigates T1D

University of Florida Home Page
© 2004 - 2011 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Powered by SobekCM