Content uploaded by Sebastián Cogliati
Author content
All content in this area was uploaded by Sebastián Cogliati on May 04, 2020
Content may be subject to copyright.
AUTHOR COPY
Journal of Alzheimer’s Disease 73 (2020) 1035–1052
DOI 10.3233/JAD-190837
IOS Press
1035
Bacillus Subtilis Delays Neurodegeneration
and Behavioral Impairment in the
Alzheimer’s Disease Model Caenorhabditis
Elegans
Sebasti´
an Cogliati, Victoria Clementi, Marcos Francisco, Cira Crespo, Federico Arga ˜
naraz
and Roberto Grau∗
Departamento de Microbiolog´ıa, Facultad de Ciencias Bioqu´ımicas y Farmac´euticas, Universidad Nacional de
Rosario, CONICET – Rosario, Argentina
Accepted 18 November 2019
Abstract. Multiple causes, apart from genetic inheritance, predispose to the production and aggregation of amyloid-(A)
peptide and Alzheimer’s disease (AD) development in the older population. There is currently no therapy or medicine to
prevent or delay AD progression. One novel strategy against AD might involve the use of psychobiotics, probiotic gut bacteria
with specific mental health benefits. Here, we report the neuronal and behavioral protective effects of the probiotic bacterium
Bacillus subtilis in a Caenorhabditis elegans AD model. Aging and neuronal deterioration constitute important risk factors
for AD development, and we showed that B. subtilis significantly delayed both detrimental processes in the wild-type C.
elegans strain N2 compared with N2 worms colonized by the non-probiotic Escherichia coli OP50 strain. Importantly, B.
subtilis alleviated the AD-related paralysis phenotype of the transgenic C. elegans strains CL2120 and GMC101 that express,
in body wall muscle cells, the toxic peptides A3-42 and A1-42, respectively. B. subtilis-colonized CL2355 worms were
protected from the behavioral deficits (e.g., poor chemotactic response and decreased body bends) produced by pan-neuronal
A1-42 expression. Notably, B. subtilis restored the lifespan level of C. elegans strains that express Ato values similar to
the life expectancy of the wild-type strain N2 fed on E. coli OP50 cells. The B. subtilis proficiencies in quorum-sensing
peptide (i.e., the Competence Sporulation Factor, CSF) synthesis and gut-associated biofilm formation (related to the anti-
aging effect of the probiotic) play a crucial role in the anti-AD effects of B. subtilis. These novel results are discussed
in the context of how B. subtilis might exert its beneficial effects from the gut to the brain of people with or at risk of
developing AD.
Keywords: A42, Alzheimer’s disease, B. subtilis, healthy aging, neuroprotection, probiotics, psychobiotics
INTRODUCTION
Alzheimer’s disease (AD) is currently the most
prevalent neurodegenerative disease worldwide.
Every 6 seconds, a new case of AD is diagnosed, and
the total number of individuals with AD is expected
∗Correspondence to: Roberto Grau, Departamento de Micro-
biolog´
ıa, Facultad de Ciencias Bioqu´
ımicas y Farmac´
euticas,
Universidad Nacional de Rosario, CONICET – Rosario,
Argentina. E-mail: robertograu@fulbrightmail.org.
to increase to 114 million by 2050 [1, 2]. The appear-
ance of the amyloid-(A) peptide aggregation in
the central nervous system (CNS) represents the hall-
mark of AD, but its etiology is not unique but rather
multifactorial and complex [3–6]. There is no cur-
rent cure or medicine that prevents AD onset or its
progression, and currently, only acetylcholinesterase
inhibitors, and few other medicines, are being used
to alleviate AD symptoms, but not its evolution
[2, 7, 8]. There are two important lessons gained
ISSN 1387-2877/20/$35.00 © 2020 – IOS Press and the authors. All rights reserved
AUTHOR COPY
1036 S. Cogliati et al. / B. Subtilis and AD Treatment
from the more than 100 failed clinical trials directed
against CNS-localized Aaggregates. First, an effec-
tive AD therapy should include more than one target
to decrease the incidence of the multiple risk fac-
tors for the onset and progression of the disease.
Second, these strategies must be performed at very
early stages of the disease, even before neurodegen-
eration symptoms begin (i.e., as a preventive therapy)
[9, 10].
There are two main forms of the disease presen-
tation: genetic (less frequent), in which individuals
carry autosomal dominant AD-linked mutations and
present clinical symptoms during their sixth or fifth
decade of life (or earlier), and sporadic AD, which
is not inherited, but multifactorial [3–5, 11] and
appears after the seventh decade of age. Sporadic
AD represents the most abundant form of the dis-
ease (approximately 95% of all cases), and aging
constitutes the main risk factor for its onset [12–15].
Earlier, we reported that Bacillus subtilis, a human
probiotic bacterium forming robust and long-lasting
beneficial biofilms [16–18], increases the healthy
longevity of the model animal Caenorhabditis ele-
gans [19]. This B. subtilis-mediated anti-aging effect
is mainly funneled through a physiological and
reversible downregulation and upregulation of the
insulin/insulin growth factor-1 (IGF-1)-like signaling
(IILS) and dietary restriction (DR) pathways, respec-
tively [19–21]. Japanese and Jewish centenarians
harbor IILS receptor variants (i.e., IGF-1 receptor)
with decreased activity, observations that validate
the importance of insulin/IGF-1 signaling in lifespan
extension and highlight its possible participation in
human AD treatment [20, 21].
The existence of a complete neuronal connectivity
map and genetic tractability of C. elegans make this
animal model useful for studying human neurologi-
cal diseases. Analysis of multiple genetic databases
show that a considerable number of human genes
associated with AD have a significant homology to C.
elegans genes, and the genetic tools available for this
nematode have allowed the construction of predic-
tive models for studying the molecular mechanism
of AD [22, 23]. In this work, we were intrigued to
explore the possibility that B. subtilis could delay
neuronal and behavioral impairments in transgenic
C. elegans strains used as an AD model [22–27]. The
obtained results are discussed through the lens of the
possible pathways that B. subtilis could use to combat
AD onset and progression and the future implementa-
tion of this probiotic bacterium in nutraceuticals and
functional foods [28–30].
MATERIALS AND METHODS
Strains and growth media
We used the following C. elegans strains: wild-
type N2 Bristol, the AD model strains CL2006
[dvIs2 (unc-54/human Abeta peptide 1-42 minigene)
+ pRF4], CL2120 [dvIs14 (unc-54::beta 1-42 +
(pCL26) mtl-2::GFP], GMC101 [dvIs100 (unc-
54p::Abeta-1-42::unc-54 3’-UTR + mtl-2p::GFP)],
and CL2355 [pCL45 (snb-1::Abeta 1-42::3’
UTR(long) + mtl-2::GFP], and the control strain
CL2122 [(pPD30.38) unc-54(vector) + (pCL26)
mtl-2::GFP] [24–26]. The used bacterial strains
were E. coli OP50 and B. subtilis NCIB3610 [19].
The AD model nematodes were obtained from the
Caenorhabditis Genetics Center (CGC), which is
funded by the NIH Office of Research Infrastruc-
ture Programs (P40 OD010440). Nematodes were
handled according to standard methods [19, 22]. For
all worms, age-synchronized eggs were obtained by
incubating embryos from gravid hermaphrodites with
bleaching solution (1% NaOCl and 0.25 M NaOH)
for 3 min, washing three times, and storing overnight
in M9 buffer (22 mM KH2PO4,34mMK
2HPO4,
86 mM NaCl, and 1 mM MgSO4) to obtain all ani-
mals in stage L1. The L1 population was transferred
to Nematode Growth Medium (NGM) agar plates
previously seeded with the corresponding bacterial
food and incubated until they reached the young adult
stage (1-day old L4), approximately 48 h later. Most
of the C. elegans strains were maintained at 20◦Con
NGM media seeded with E. coli or B. subtilis with or
without ampicillin (100 gml
−1) supplementation,
respectively [19]. The C. elegans CL2355 strain
was maintained at 16◦C to prevent pan-neuronal A
peptide expression [22, 25]. The antifungal ampho-
tericin B (25 gml
−1; Sigma Co.) was also added to
the NGM medium; E. coli and B. subtilis were grown
in Luria-Bertani (LB) broth overnight at 37◦C [19].
Analysis of C. elegans aging-related
neurodegeneration
Plates were prepared by spreading 50 lofan
overnight culture of E. coli OP50 or B. sub-
tilis NCIB3610 over the surface of 6-cm diameter
plates prepared with NGM agar medium. These
plates were incubated overnight at 37◦C before
seeding with synchronized L1-stage N2 wild-type
worms and incubated at 20◦C throughout the entire
experiment (approximately 30 days). Every 4 days,
AUTHOR COPY
S. Cogliati et al. / B. Subtilis and AD Treatment 1037
the nematodes were labeled with 1,1´-dioctadecyl-
3,3,3´,3´,-tetramethylindocarbocyanine perchlorate
(DiI, Aldrich), a red fluorescent dye that can fill
the worm amphid neurons [31]. The DiI stock solu-
tion was 2 mg/ml in dimethyl formamide and was
stored at –20◦C in a tube wrapped in foil until use.
Briefly, OP50- and NCIB3610-fed adult worms of
the different ages were spun down, washed, resus-
pended in 1 ml M9 buffer, and incubated with 5 lof
a 1:200 dilution of DiI stock solution. Incubation was
continued on a shaker (75 rpm) for 3 h before spin-
ning, washing, and transferring the labeled worms
onto agar pads. To this end, labeled worms were
mounted onto a 2% agar pad on a glass slide using 1
M sodium azide (the azide acts as an anesthetic for
the worms) and enclosed with a coverslip. Neuron
degeneration was examined over time with an Olym-
pus FV1000 laser confocal scanning microscope, and
a semi-quantitative analysis was made. The worms
were analyzed for the absence of amphid neuron
architecture (complete loss), the presence of a com-
plete and intact set of amphid neurons (no loss), or the
presence of at least one single structural abnormality,
such as wavy, branched, or interrupted dendrites (par-
tial loss) [32, 33]. All experiments were performed at
least three times in duplicate.
Culturing bacteria from worms
The N2 C. elegans eggs were isolated using a
solution of 10% commercial bleach and 1 N NaOH,
followed by four washes with M9 buffer (22 mM
KH2PO4,42mMNa
2HPO4, 85 mM NaCl, and 1 M
MgSO4). Approximately 500 eggs were transferred
to a 60-mm plate with NGM agar and incubated
overnight at 20◦C with agitation to allow L1 lar-
vae to emerge. Then, approximately 500 L1 larvae
per experiment were grown for 48 h on NGM plates
(a time that allows worm development to reach the
L4 larvae stage) seeded with OP50 E. coli cells
(1 ×105cells/plate) or NCIB3610 B. subtilis cells
(1 ×10 5cells/plate). At different incubation times,
50 worms were transferred to Eppendorf tubes con-
taining M9 buffer and 1% Triton X-100. Worms were
treated with 25 mM levamisole to induce temporal
paralysis, superficially sterilized with 3% commer-
cial bleach for 15 min and washed three times with
M9 buffer. After the worms were surface-sterilized,
worms devoid of outside bacteria were disrupted
using a pellet pestle (Sigma Co.), centrifuged, and
resuspended in 500 l M9 buffer. Finally, 50 lof
each cell suspension were used to prepare serial
dilutions of the bacteria before counting. To this end,
100 l of the appropriate serial dilutions was spread
with a Drigalski scraper on LB Petri dishes. The num-
ber of colony-forming units (CFUs) was determined
after 24 h of incubation at 37◦C.
Octanol and diacetyl (DA) time response assays
For the behavioral experiments, C. elegans N2
worms were fed on OP50 or NCIB3610 bacterial cells
from the L1-larval stage to adulthood at 20◦C. Repul-
sion and attraction behavioral assays using octanol
(1-octanol, Sigma-Aldrich) or DA (butane-2,3-dione,
Merck) as repellent or attractant agents, respectively,
were performed as previously described [25, 26].
Briefly, OP50- or NCIB3610-fed adult worms of dif-
ferent ages were washed three times with M9 buffer
to remove any residual bacteria and placed in NGM
plates without food. One hour after food starvation,
for the repellent assay, a paintbrush hair previously
dipped in 100% undiluted octanol was placed in front
of a moving animal (care was taken to not touch the
nematode). The octanol response time was scored
as the time (s) from presentation to the initiation
of a backward or escape movement. Sterile water
was used instead of octanol as a control, and assays
were halted at 20 s to account for spontaneous rever-
sals (data not shown). For the attractant assay, 1 h
after food starvation, a 1-L drop of 0.5% DA in
ethanol was placed 1.5 cm in front of a moving ani-
mal (without touching it). The DA response time was
scored as the time (s) from presentation to the ini-
tiation of a forward movement in the direction to
DA. Ethanol was used instead of DA as a control
(data not shown). All experiments were performed in
triplicate.
Chemotaxis index (CI) assays
The OP50- or NCIB3610-fed N2 worms were col-
lected, washed three times with M9 buffer, and seeded
in NGM 10-cm plates without food for 1 h. Then,
approximately 75 worms were placed in the center
of 6-cm plates prepared with 2% agar, 1 mM CaCl2,
1 mM MgSO4, and 25 mM phosphate buffer (pH 6.0).
After all animals were transferred to the center of
the assay plates, 2 l of attractant were seeded 2 cm
from the center of the plate, and 2 l of solvent (con-
trol) in which the attractant was diluted were seeded
equidistantly. Both the attractant and the control were
added with a 1-l drop of 1 M azide. The plates were
AUTHOR COPY
1038 S. Cogliati et al. / B. Subtilis and AD Treatment
incubated for 1 h at 20 or 23◦C (as indicated in the
legend figures). Then, worms found at each end of
the plates were counted, and the CI was calculated.
The attractant compounds used for the assays were
0.5% DA diluted in ethanol and 0.1% isoamyl alco-
hol (IAA; Sigma-Aldrich) diluted in water and NaCl
150 mM. The CI is defined as the number of worms at
the attractant or repellent location – number of worms
at the control location divided by the total number of
worms on the plate [24].
Paralysis assay
The CL2122, CL2120, and GMC101 L1 larvae
were cultured at 20◦C on OP50 or NCIB3610 bac-
terial lawns until adulthood (L4 stage). Then, 1-day
old L4 adults were transferred to NGM 6-cm plates
without bacteria. After 1-h food starvation, the 6-cm
plates seeded with worms were shifted to 25◦C, and
paralysis was scored each day until the last worm
became paralyzed. Nematodes were considered para-
lyzed if they failed to complete a full body movement
or only moved their head when gently touched with
a platinum wire [26].
Behavioral assays of C. elegans AD models
The CL2122 and CL2355 L1 larvae were fed on
OP50 or NCIB3610 bacterial cells at 16◦C until they
reached the L3-larval stage (approximately 36 h). The
L3 larvae were shifted to 23◦C (to express the Apep-
tide) and incubated for another 36 h until adulthood
(L4 stage). For the chemotaxis assays, approximately
250 L4 larvae were collected, washed three times with
M9 buffer, and seeded on NGM 10-cm plates without
food for 1 h. Then, approximately 100 worms were
placed on the center of 6-cm plates prepared with
2% agar, 1 mM CaCl2, 1 mM MgSO4, and 25 mM
phosphate buffer (pH 6.0). After all animals were
transferred to the center of the assay plates, 2 l
chemoattractant (0.5% DA in 95% ethanol), along
with 1 l 1 M sodium azide, were added to the orig-
inal spot. On the opposite side of the attractant, 1 l
sodium azide and 2 l ethanol (control) were added.
Assay plates were incubated at 23◦C for 1 h, and the
CI was calculated as indicated. For the body bend
assay, OP50- and NCIB3610-fed L4 worms were
collected, washed three times with M9 buffer, and
seeded on NGM 10-cm plates without food for 1 h.
Each worm was then transferred to a single well of
a 24-well plate with 1.5 ml M9 buffer. After allow-
ing adaptation for 20 s, worms were scored for the
number of body bends generated in 30 s. A body
bend was defined as a change in the direction of
propagation along the y-axis, assuming that worms
were travelling along the x-axis [25]. Twenty worms
of each group were evaluated, and all experiments
were performed three times in duplicate.
Lifespan assays
Lifespans for C. elegans N2 and AD strains were
monitored at 20 or 23◦C as described previously
[19]. Briefly, embryos were isolated by exposing
hermaphrodite adult worms to alkaline hypochlorite
treatment for 3 min, processed as indicated above,
and synchronized eggs were allowed to develop.
In all cases, L4/young adult worms (n= 100) were
used at time zero for lifespan analysis; they were
transferred to fresh plates previously seeded with
OP50 or NCIB3610 bacterial cells each day until
the assay was completed. Worms were considered
dead when they ceased pharyngeal pumping and
did not respond to prodding with a platinum wire.
Worms with internal hatching were removed from
the plates and excluded from lifespan calculations.
All experiments were repeated at least three times in
duplicate.
Statistical analysis
All assays were performed at least three times in
duplicate. Mean survival days, standard error of the
mean (S.E.M.), intervals of mean survival days with
95% confidence, and equality pvalues to compare
averages were calculated by log-rank and Kaplan-
Meier tests using the OASIS program. The S.E.M.
values are used in the figures; p< 0.5 was considered
statistically significant.
RESULTS
Bacillus subtilis delayed age-related
neurodegeneration and cognitive damage in C.
elegans
We fed young adult N2 wild-type C. elegans (1-
day-old adult worms) with the regular bacterial food,
the OP50 E. coli strain, or the probiotic B. sub-
tilis strain NCIB3610, and investigated the in vivo
effect of B. subtilis on neural deterioration retarda-
tion throughout the life time (lifespan expectancy) of
both worm populations (Fig. 1A). Bacteria that are
AUTHOR COPY
S. Cogliati et al. / B. Subtilis and AD Treatment 1039
disrupted by the worm grinder and those that sur-
vive and reach the intestine constitute the worm food
and the worm gut flora, respectively (Fig. 1B) [19].
Age-related neurodegeneration [32, 33] was verified
by the formation of neuronal defects such as beaded,
wavy, branched, and/or interrupted dendrites or soma
branching in OP50- or NCIB3610-colonized worms.
To this end, OP50- and NCIB3610-fed worms of
different ages (Fig. 1A) were stained with the flu-
orescent dye DiI, which specifically labels amphid
(and phasmid; not shown) neurons to highlight the
chemosensory structures of live nematodes [31, 34].
The staining of OP50-colonized young adult worms
(4 days old) with Dil revealed the complete integrity
of the neuronal network (i.e., normal amphid chan-
nels and nerve ring structures, dendrites and sensory
neurons, respectively) (Figs. 1C and 2A, 4 days old,
column and left panel, respectively). As the OP50-
colonized worm population aged (up to 12 days of
cultivation), we observed different defects (partial
neural loss) in the amphid chemosensory structures
in approximately 40 and 60% of the worm popula-
tion (Figs. 1C and 2B, 8 and 12 days old, columns
and upper panel, respectively). From that age on
(12 days onwards), we started to observe signs of
total neural deterioration (i.e., neuronal death) in the
OP50-colonized worm population. The correspond-
ing percentages of partial/total neural deterioration,
at the ages of 16 and 20 days, were 60%/10% and
30%/70%, respectively (Figs. 1C and 2C, 16 and
20 days old, columns and left panels, respectively).
After 24 days of cultivation, the OP50-colonized
worms showed a complete loss of neuronal archi-
tecture, Fig. 1C (24 days old column). Interestingly,
for worms fed on the probiotic strain NCBI3610,
the 37% gained in lifespan extension when com-
pared with the lifespan of worms colonized by OP50
cells (p< 0.001, Fig. 1A) correlated with a notable
delay in neuronal deterioration. In contrast to the 40%
of the 8-day-old OP50-colonized worms, the 100%
of NCIB3610-colonized worms of the same chrono-
logical age retained a completely normal neuronal
architecture (Figs. 1D and 2A, 8 days old, column and
right panel, respectively). At longer incubation times
(12 days old and beyond), the differences in neu-
ronal architecture preservation between OP50- and
NCIB3610-colonized worms of the same chronolog-
ical age became more notorious. While 20-day-old
OP50-colonized worms showed percentages of nor-
mal, partial, and total neuronal deterioration of 0,
30, and 70%, respectively (Fig. 1C, 20 days old
column), NCIB3610-colonized worms of the same
chronological age showed percentages of 40, 50, and
10%, respectively (Figs. 1D and 2A, 20 days old, col-
umn and bottom panel, respectively). At advanced
chronological ages (i.e., 24 and 28 days), when the
neuronal architecture of OP50-colonized worms was
completely damaged, the NCIB3610-colonized pop-
ulation showed a proportion of worms with complete
neuronal loss (Fig. 2C, right panel), but still con-
tained a significant proportion of worms with normal
or partial neuronal architecture, Fig. 1D (24 and 28
days, columns). The presented results (Figs. 1 and 2)
demonstrate that the neuronal architectural decay of
worms of the same chronological age can markedly
differ in function of the type of bacterium (i.e., pro-
biotic or non-probiotic) that colonized their guts,
strongly suggesting that behavioral responses are also
affected differently.
To correlate the morphological neuroprotective
effect of B. subtilis on the functionality of the sen-
sory apparatus of C. elegans throughout adult life
(Fig. 3A), we performed behavioral chemotaxis tests
in similarly aged worms colonized by NCIB3610
or OP50 bacteria. The chemotaxis response in C.
elegans is mediated by the interplay of several
sensory neurons and interneurons to stimulate the
motor neurons so that the individual approximates
or avoids a certain chemical signal (an attractant
or a repellent, respectively) [19, 34]. As compared
with OP50-colonized worms, B. subtilis-colonized
worms displayed an enhanced behavioral response
(improved response times) when confronted with
negative and positive environmental inputs (avoid-
ance or attraction to harmful or attractant signals;
Fig. 3B and 3C, respectively). Overall, the lower
(more rapid) response times of B. subtilis-colonized
worms, compared with OP50-colonized worms, to
different external stimuli (Fig. 3B, C) correlated
well with the CIs measured at different chronolog-
ical ages (16, 20, 24, and 28 days old; Figs. 3D–F).
For instance, the CIs of 16-day-old elderly OP50-
or NCIB3610-colonized worms to DA, IAA, and
NaCl were 0.18 ±0.02 and 0.39 ±0.04; 0.16 ±0.02
and 0.33 ±0.03; 0.20 ±0.02 and 0.40 ±0.04,
respectively (n= 75, p< 0.001; Fig. 3D–F), The
improved behavioral performance (i.e., higher CIs) of
NCIB3610-colonized worms, compared with OP50-
colonized worms, remained during the complete
adult life of both compared worm populations. Even
at a very old age (i.e., 28 days old), when all
OP50-colonized worms were dead, the NCIB3610-
colonized worms showed a behavioral response
significantly better (CIs of 0.14 ±0.04, 0.12 ±0.02,
AUTHOR COPY
1040 S. Cogliati et al. / B. Subtilis and AD Treatment
Fig. 1. Age-related neuroprotection by B. subtilis. A) Life expectancy of C. elegans that harbored probiotic or non-probiotic bacteria in
the intestine. One hundred young-adult (1-day old) wild-type Bristol strain N2 worms were fed on E. coli OP50 or probiotic B. subtilis
NCIB3610 bacteria (red and green, respectively). Worms were grown on bacteria-seeded 10-cm NGM agar plates at 20◦C, and survival was
monitored as indicated until the last worm died (see Materials and Methods for details). The life expectancy of NCIB3610-colonized worms
was 37% longer than the lifespan of OP50-colonized worms (p< 0.001). A typical output of three independent experiments performed in
duplicate is presented. B) Worm intestine colonization by OP50 or NCIB3610 bacteria. L4 worms were allowed to develop at 20◦ConNGM
agar plates seeded with OP50 E. coli or NCIB3610 B. subtilis cells (red and green bars, respectively), as indicated in Material and Methods.
At each of the indicated ages, 50 worms were transferred to Eppendorf tubes, superficially sterilized, and disrupted before counting the
number of E. coli or B. subtilis cells in the worm gut. The data are representative of at least three independent experiments. Error bars show
the mean ±SEM from at least three independent experiments. See Material and Methods for details. C, D) Semi-quantification of age-related
neurodegeneration. Ten OP50- or NCIB3610-colonized N2 worms (C and D, respectively), grown on NGM plates at 20◦C, were taken at
the indicated times, processed, and labeled with DiI, as indicated in Materials and Methods, to determine the grade of age-related neuronal
deterioration (no loss, partial loss, or total loss). See Materials and Methods for details. Results are expressed as a percentage of initial worm
population (n= 100) ±S.E.M.
AUTHOR COPY
S. Cogliati et al. / B. Subtilis and AD Treatment 1041
Fig. 2. Neuronal morphological changes of aging worms colonized by OP50 or NCIB3610 bacteria. Aging N2 worms, colonized by OP50
or NCIB3610 bacterial cells (red and green rectangles in the figure, respectively) at different ages, were labeled with fluorescent DiI to
highlight amphid neuron morphology: normal morphology or no neuronal loss, A; partial neuronal alterations or partial neuronal loss, B;
and total neuronal deterioration or total neuronal loss, C. See Materials and Methods for details. Worm ages are as follow: 4 days old and
8 days old, A; 8 days old and 20 days old, B; and 20 days old and 28 days old, C; for OP50-colonized or NCIB3610-colonized worms,
respectively. The top and bottom micrographs (phase contrast and fluorescence microscopy, respectively) in A to C are representative of 10
independent worm images analyzed for each age. Arrows in A indicate the location of the chemosensory worm neurons (i.e., ASK, ADL,
ASI, ASH, ASJ, AWB), and arrows in B indicate some of the age-associated neuronal alterations.
and 0.20 ±0.02; for DA, IAA or NaCl, respectively;
n= 75, p< 0.1) than that of the 20-day-old OP50-
colonized worms (Fig. 3D–F).
The overall results (i.e., delayed aging, neuropro-
tection, improved behavioral responses, Figs. 1–3),
and the knowledge that aging and neurodegeneration
are important risk factors for AD development [13,
14, 35, 36], prompted us to use several transgenic C.
elegans strains that express the human Apeptide to
investigate whether B. subtilis might represent a new
alternative against the disease.
Bacillus subtilis alleviated the paralysis
phenotype of transgenic C. elegans expressing
the human Aβpeptide in muscle
Caenorhabditis elegans offers a valuable platform
for investigating the cellular and molecular mecha-
nisms of AD [22, 23]. The Ais believed to be the
major cause of AD pathogenesis, and its expression in
transgenic C. elegans strains produces several patho-
logical features important to better understand AD
pathology [24, 26, 37, 38]. Two of the transgenic AD
AUTHOR COPY
1042 S. Cogliati et al. / B. Subtilis and AD Treatment
Fig. 3. B. subtilis-mediated cognitive improvement during C. elegans aging. A) Schematic representation of C. elegans life cycle from
egg-laying to adult worm death. B, C) Average response times (in seconds, sec, y-axis) of OP50- and NCIB3610-colonized N2 worms (red
and green colors, respectively) of different ages (in days, x-axis) to repellent (octanol, B) and attractant (diacetyl [DA], C) exposition (see
Materials and Methods for details). Results represent the mean ±S.E.M of three independent experiments performed in duplicate. D-F)
Chemotaxis index of N2 worms of different ages exposed to different attractants: 0.5% DA (D), 0.1% isoamyl alcohol (IAA, E), and 150 mM
NaCl (F). A typical result from one of the three independent experiments performed in duplicate is presented (mean ±S.E.M). Asterisks
indicate statistical significance (***p< 0.001; **p< 0.01; and *p< 0.1; ns, no significant difference, p> 0.5).
AUTHOR COPY
S. Cogliati et al. / B. Subtilis and AD Treatment 1043
C. elegans strains are CL2120 and GMC101, which
express human Apeptides of different sizes and tox-
icities, namely A3-42 and A1-42, respectively [24,
26]. In the C. elegans strain CL2120, the A3-42 pep-
tide is constitutively expressed under the control of
the unc-54 promoter in body wall muscle cells and
produces a chronic and progressive paralysis pheno-
type [24]. In the first set of the performed paralysis
experiments, young larvae (L1) of the CL2120 strain
and its wild-type control CL2122 strain were fed on
E. coli OP50 or B. subtilis NCIB3610 cells for 48 h at
20◦C until they reached adulthood (L4 stage). They
were then washed several times and starved from bac-
terial food for 1 h (Fig. 3A). The starved young adult
worms, which contained OP50 or NCIB3610 bacteria
colonizing their guts, were shifted to 25◦C and paral-
ysis was recorded. Worms that did not move or only
moved the head (under a gentle touch with a plat-
inum loop) were scored as paralyzed (see Materials
and Methods for details). The control CL2122 worms,
maintained at 25◦C, displayed a motile (no paralysis)
phenotype for the duration of the experiment (over 1
week after adulthood), regardless of the gut bacte-
ria (OP50 or NCIB3610) they harbored (Fig. 3B).
However, the human A-peptide-expressing strain
CL2120, colonized by the OP50 E. coli strain, dis-
played an age-dependent paralysis phenotype that
started 2 days after the temperature increase from 20
to 25◦C, and 4 days after the temperature upshift,
the entire OP50-colonized CL2120 worm popula-
tion was paralyzed (Fig. 3C). Comparatively, when
the CL2120 worms were colonized by B. subtilis
NCIB3610, 100% of the worm population were
protected from paralysis and remained fully motile
during the experiment (over 1 week after adulthood;
Fig. 3C). The CL2120 strain expresses a less-toxic
form of the human Apeptide (i.e., A3-42), and
therefore, the paralysis phenotype observed in this
transgenic strain is ameliorated [24]. By contrast,
the GMC101 strain expresses the full-length human
Apeptide (A1-42), and thus, the paralysis phe-
notype displayed in this transgenic worm is more
severe [26]. In order to confirm the CL2020 strain
results, we performed a second set of paralysis exper-
iments in the GMC101 strain colonized by OP50
or NC1B3610 bacteria. As shown in Fig. 3D, the
paralysis phenotype in the OP50-colonized GMC101
strain was detected more rapidly and was more severe
than the observed paralysis displayed by the OP50-
colonized CL2120 strain (Fig. 3C). Indeed, almost
90% of the OP50-colonized GMC101 worms were
completely immotile (paralyzed) 2 days after the tem-
perature increase (Fig. 3D). Intriguingly, B. subtilis
NCIB3610 significantly delayed the start and sever-
ity of paralysis in GMC101 worms (Fig. 3D). While
the paralysis of the OP50-colonized GMC101 worm
population was almost total (100%) 2 days after
the temperature increase from 20 to 25◦C, almost
97% of the NCIB3610-colnized GMC101 worm
population were not paralyzed. Furthermore, only
15% of NCIB3610-colonized GMC101 worms were
immotile (paralyzed) after 3 days of the temperature
upshift, compared with the 100% of OP50-colonized
worms that were immotile at that time (Fig. 3D). The
PT50, the time interval from the onset of paralysis at
which 50% of the worms were paralyzed, in GMC101
worm populations was 1.7 ±0.3 days (n= 75) and
4.6 ±0.5 days (n= 75; p<0.001) for OP50- and
NCIB3610-colonized worms, respectively. Thus, at
the assayed times, there was complete paralysis
prevention or significant amelioration in transgenic
worms that express the less severe and the more toxic
forms of the human Apeptide, A3-42 or A1-42,
respectively, when B. subtilis colonized the worm
intestine (Fig. 3C, D).
Bacillus subtilis alleviated behavioral deficits of
transgenic C. elegans expressing pan-neuronal
Aβpeptide
Transgenic C. elegans individuals with neuronal
human Apeptide expression show learning-deficit
behavioral phenotypes [25]. The C. elegans strain
CL2355 employs the synaptobrevin promoter (snb-1)
to drive pan-neuronal human Apeptide expression
(snb-1::A1-42) after a temperature increase to 23◦C
[25]. We consider this transgenic AD strain to be
a useful tool to evaluate the protective effect of B
subtilis on the deteriorated behavioral performance
of transgenic worms with neuronal Aexpression.
One sensory behavior we examined in this strain
was chemotaxis. The age-synchronized wild-type
control strain (CL2122) and transgenic CL2355 C.
elegans were cultured at 16◦C from egg hatching,
using E. coli OP50 or B. subtilis NCIB3610 as
a food source up to reaching the L3 larval stage,
and then shifted to 23◦C for 36 h to induce the
production of pan-neuronal Apeptide while the
final larval stage (L4) was reached (Fig. 4A). These
young adult L4 CL2355 and CL2122 worms were
starved from food (OP50 or NCIB3610) for 1 h
before to compare their chemotactic response toward
the attractant DA (Fig. 4A). As shown in Fig. 4B,
OP50-colonized CL2355 worms exhibited a poor
AUTHOR COPY
1044 S. Cogliati et al. / B. Subtilis and AD Treatment
Fig. 4. B. subtilis protected against A-induced progressive paralysis in C. elegans AD model strains. A) A cartoon that summarizes
the paralysis assay performed on OP50- and NCIB3610-colonized AD model worms (red and green colors, respectively; see Materials and
Methods for details). B-D) Percentages of paralyzed CL2122 wild-type (wt) worms (control, B), and AD model CL2120 and GMC101worms
(C and D, respectively) fed on OP50 or NCIB3610 bacteria as indicated. PT50, in panel D, indicates the time in which 50% of the worm
population were paralyzed. Nematodes were considered paralyzed if they failed to complete a full body movement or only moved their
head when gently touched with a platinum wire. Paralysis was scored every day until the last worm became paralyzed. Panels B-D show a
representative result from three independent experiments performed in duplicate (mean±S.E.M).
chemotactic response toward DA (CI=0.19 ±0.01;
n= 100) compared with the OP50-colonized CL2122
control strain (CI = 0.55 ±0.04; n= 100, p< 0.001).
Importantly, NCIB3610-colonized CL2355 worms
displayed a chemotactic response toward the
attractant (CI = 0.62 ±0.04; n= 100) that was indis-
tinguishable from the chemotactic response of the
control wild-type CL2122 strain colonized by OP50
bacteria (CI = 0.65 ±0.04, n= 100; Fig. 4B).
We also measured whether B. subtilis improved
the slowed locomotion response (body bends) that
occurs in worms that express pan-neuronal A
peptide (Fig. 4C) [22, 25]. The OP50-colonized
CL2355 worms exhibited a low number of body
bends (50 ±4; n=100) compared with the body
bend number from the OP50-colonized control strain
CL2122 (85 ±6; n= 100, p< 0.001). Importantly,
NCIB3610-colonized CL2355 worms displayed a
AUTHOR COPY
S. Cogliati et al. / B. Subtilis and AD Treatment 1045
body bend response (98 ±6 body bends; n= 100)
also indistinguishable from the motility response of
the control wild-type CL2122 strain colonized by
OP50 cells (100 ±6 body bends; n= 100; Fig. 4D).
These results show that the cognitive impairments
(deleterious behavioral responses) for food detection
(Fig. 4B) and locomotive activity (Fig. 4D) produced
by pan-neuronal human Apeptide expression are
eliminated by B. subtilis.
Bacillus subtilis restored the healthy lifespan of
Aβpeptide-expressing C. elegans strains
The life expectancy of transgenic C. elegans
expressing human Apeptide is shortened [25, 38].
To evaluate whether the protective effect of B. sub-
tilis on neurodeterioration and behavioral impairment
of transgenic AD C. elegans was also translated to
life expectancy, we performed lifespan assays in the
AD strains CL2006, CL2120, and CL2355 colonized
by OP50 or NCIB3610 bacterial cells [19]. First,
we compared the lifespan of the transgenic CL2006
and CL2120 strains that constitutively express human
Awhen colonized by B. subtilis NCIB3610 or E.
coli OP50 (see Material and Methods for details)
[22]. In parallel, we compared the lifespan values
of the AD strains CL2006 and CL2120, colonized
by OP50 or NCIB3610 cells, with the lifespans
of the corresponding control strains wild-type N2
and CL2122, respectively. The lifespan expectancy
of the OP50-colonized CL2006 and CL2120 AD
worms decreased by 26 and 29%, respectively, com-
pared with the lifespan of OP50-colonized wild-type
worms (Fig. 6A, B, n= 100, p< 0.001). Interestingly,
B. subtilis NCIB3610 robustly extended the lifes-
pan of both AD strains, CL2006 and CL2120, to a
level indistinguishable of the life expectancy (mean
lifespan value of ∼16 days) of the corresponding
OP50-colonized N2 and CL2122 wild-type worms,
respectively (Fig. 6A, B, n= 100, p< 0.001). In the
case of the AD model strain CL2355, which pro-
duces a pan-neuronal expression of the human A
peptide, its life expectancy when colonized by OP50
cells was severely reduced (∼40% decrease) com-
pared with the lifespan level of the control wild-type
strain CL2122 (Fig. 6C, n= 100, p< 0.001). Attrac-
tively, the probiotic bacterium significantly increased
the lifespan of CL2355 worms (from ∼9 days to ∼12
days), although not exactly to the same level (∼15
days) of the corresponding OP50-colonized wild type
CL2122 strain (Fig. 6C, n= 100, p< 0.001).
DISCUSSION
Caenorhabditis elegans and mammalian neurons
are remarkably similar in terms of functionality and
connectivity, and C. elegans offers a valuable and
simple tool to unravel what might be happening
in the aging mammalian brain under normal and
pathological conditions [20, 23, 27, 34]. The results
presented in this work show that B. subtilis can delay
neuronal aging (Figs. 1–2) and improve behavioral
responses in elderly wild-type worms (Fig. 3). Since
aging is the main risk factor for AD development [2,
13, 15, 36], we investigated whether the anti-aging
effect of this bacterium [19, 39] would also protect
against AD. In particular, we measured the delete-
rious effects of Apeptide expression in transgenic
worms that harbor the probiotic bacterium in their
guts. Bacillus subtilis stopped or delayed paralysis
in the AD transgenic strains CL2120 and GMC101,
respectively (Fig. 4). The C. elegans CL2120 strain,
which constitutively expresses a less-toxic version
of the Apeptide (A3-42) in wall muscle cells,
exhibited a chronic, albeit smoother paralysis pro-
gression (Fig. 4C) than the more toxic version of
the Apeptide (A1-42) expressed by the GMC101
strain (Fig. 4D). Accordingly, B. subtilis significantly
improved the behavioral responses of transgenic
CL2355 C. elegans with pan-neuronal Apeptide
distribution (Fig. 5) and extended the lifespan in AD
model worms to levels similar to those observed in
wild-type animals (Fig. 6).
Bacillus subtilis is a probiotic member of the
human gut microbiota [40–45]. Probiotics are live
microorganisms (principally bacteria) which, when
consumed in adequate quantities, have beneficial
health effects on consumers [46, 47]. Recently, a new
probiotic category was proposed: psychobiotics (i.e.,
probiotics that benefit behavior and combat neuronal
disorders) [48]. Psychobiotics modulate brain func-
tions through the gut-brain-axis [49, 50]; they can
alter the gut microbiota composition [51], influence
immune-neuron system communication, modify
host-produced neurotransmitters, and/or synthesize
neurotransmitters de novo [52–54]. The failure of
the more than 100 AD clinical trials with drugs
that target CNS Aaggregates leads researchers and
clinicians to consider other hypotheses and thera-
pies [4–6, 55–57]. Could pro(psycho)biotics be used
in AD patients? In a recent report, probiotic lactic
bacteria (LBA) taken daily over a short time (12
weeks) produced a moderate, but significant improve-
ment in some metabolic statuses and the Mini-Mental
AUTHOR COPY
1046 S. Cogliati et al. / B. Subtilis and AD Treatment
Fig. 5. B. subtilis improved the behavioral response of C. elegans expressing pan-neuronal Aaggregates. A) and C) Cartoons that summarize
the chemotactic (B) and basal slowing movement (C) assays of OP50- and NCIB3610-colonized AD CL2355 worms (red and green colors,
respectively; see Materials and Methods for details). B and D) Chemotaxis indices for CL2122 (control or wild-type, wt) and CL2355 (AD
model) worms to 0.5% diacetyl (DA, attractant, B), and body bends of CL2122 (control) and CL2355 (AD model) worms scored for 30 s,
as shown in panels A and C, respectively. A typical result out of three independent experiments (performed in duplicate) is presented
(mean ±S.E.M). Asterisks indicate statistical significance (*p< 0.1; ***p< 0.001; ns, no significant difference, p> 0.5).
State Examination scores of elderly 60-to-95-year-
old male and female AD patients [58]. This study
showed that the gut microbiome, inhabited by tril-
lions of microorganisms, can be modulated by dietary
interventions with probiotic (psychobiotic) bacteria
to combat AD [59, 60].
Could B. subtilis be used in the future to delay or
treat human AD? If so, how might anti-AD B. sub-
tilis work? Unfortunately, our understanding of AD is
incomplete, most likely because most of the informa-
tion about the disease etiology comes from familial
(genetic-mutation-related) AD, which represents a
minor proportion of AD cases [1, 13]. However, AD
etiology is multifactorial and complex; it involves
multiple distinct and overlapping redundant path-
ways of neuronal damage [4, 8, 12, 14, 15, 61]. In
this sense, one common feature of the failed clinical
trials is that, regardless of their individual targets, all
of them were based on the belief that AD pathology
emanates from a single protein: the Apeptide (i.e.,
the amyloid cascade hypothesis) [3, 55]. Therefore,
if B. subtilis only targets A, it would not likely con-
stitute a valuable therapeutic tool for AD. However,
we envision at least three different (but overlapping
and simultaneous) possible scenarios of how B. sub-
tilis might be employed as a gut-member to delay
or treat AD in the future [5, 23, 29, 48, 49, 51, 62]
(Fig. 7).
Evidently, one weapon that B. subtilis would use to
fight AD is the anti-aging effect of the bacterium [19,
39, 63]. There are two main genetic pathways, evolu-
tionary conserved from nematodes and flies to human
beings, that control the aging process in living organ-
isms: dietary restriction (DR) and the insulin/insulin
growth factor-1 (IGF-1)-like signaling (IILS) system
[19–21, 39]. In C. elegans, the IILS pathway is under
the control of the nutrient-related signal receptor
DAF-2, that is the homologue of the human insulin-
like receptor IGFR that negatively regulates DAF-16
and HSF-1 [19–21, 39]. Here, DAF-16 (homologue
to the FOXO human transcription factor) and HSF-1
(heat shock factor) play a crucial role in the expres-
sion of numerous genes involvedin lifespan extension
[63, 64]. Dietary restriction, a condition of reduced
caloric intake [65], enhances longevity and protects
against proteotoxicity by a mechanism (distinct from
reduced IILS signaling) that requires HSF-1 activa-
tion [66]. Both longevity routes regulate genes (either
repressing or activating them in the case of IILS
or DR, respectively) involved in protection against
oxidative stress, inflammation, microbial infections,
and the production of numerous proteins with chap-
erone activity to maintain the integrity of protein
homeostasis against proteotoxicity [20, 21, 63, 66].
Bacillus subtilis has a prolongevity (anti-aging) effect
because, when colonizing the host intestine through
AUTHOR COPY
S. Cogliati et al. / B. Subtilis and AD Treatment 1047
Fig. 6. Probiotic B. subtilis produced a healthy lifespan for A-synthesizing C. elegans. Lifespan of OP50- and NCIB3610-colonized AD
model CL2006 (A), CL2120 (B), and CL2355 (C) worms. Worms were grown on bacteria-seeded 10-cm NGM agar plates at 20 or 23◦C (for
A-B and C, respectively), and survival was monitored as indicated until the last worm died (see Materials and Methods for details). Control
wild-type worm strains are N2 and Cl2122 for A and B-C, respectively. A typical output of three independent experiments performed in
duplicate is presented.
the formation of a beneficial gut-associate biofilm, it
downregulates and upregulates the IILS pathway and
the process of dietary restriction (DR), respectively
[19, 39]. The anti-aging effect of B. subtilis occurred
in 90% because to ILS inhibition and in 10% due
to DR activation [19, 39]. We envision that the pro-
longevity effect of B. subtilis would protect against
the aging-linked risk factors that are associated with
AD development (Fig. 7A).
The second pathway that B. subtilis would use to
delay or fight AD is through the production of the
quorum-sensing (QS) pentapeptide CSF (also named
AUTHOR COPY
1048 S. Cogliati et al. / B. Subtilis and AD Treatment
Fig. 7. A workable proposed model for the B. subtilis effects in AD. Cartoon that summarizes the different routes (anti-aging ILS down-
regulation/DR upregulation, A; CSF quorum-sensing peptide production, B; and A-degrading nattokinase activity, C) that probiotic B.
subtilis might use to produce beneficial effects against factors related to AD development and, therefore, in fighting against human AD (see
Discussion for details).
PhrC) [19, 67]. QS is a chemical mechanism that bac-
teria use for cell-to-cell communication with other
bacteria (inrtra-) or plants and animals (inter-specific
kingdom communication [68]. Basically, bacteria
produce small metabolites (i.e., acyl-homoserine lac-
tones and short peptides), QS molecules, which are
liberated to the surrounding environment wherein
other organisms detect and internalize them. Once
inside the host cells, the bacterial QS molecules affect
host gene expression. The CSF pentapeptide plays a
crucial intra-specific role in orchestrating cell-to-cell
communication in vital B. subtilis lifestyle processes
such as natural DNA competence, sporulation, and
biofilm formation [67]. Besides this intraspecific
(bacterium-bacterium interaction) role of CSF, there
is a reported interspecific CSF function (bacterium-
mammalian inter-kingdom interaction) [40, 69]. This
quorum-sensing pentapeptide is internalized via the
mammalian oligopeptide transporter OCTN2, where
it induces the production of the heat shock pro-
tein chaperone Hsp27 and the p38 MAPK and AKT
survival pathways [69]. This induction leads to cel-
lular protection against oxidative stress, misfolded
proteins, and loss of barrier function [69]. In vitro,
Hsp27 acts as an ATP-independent chaperone by
inhibiting protein aggregation and stabilizing mis-
folded proteins, actions that ensure refolding by the
Hsp70 complex [70]. The Hsp27 also activates the
proteasome complex to quicken the degradation of
irreversibly denatured or aberrant proteins [71, 72].
Diverse proteomic analysis showed that there is a
complex map of protein alterations in AD; these find-
ings indicate that AD is more than an Aopathy or
tauopathy: it is a proteopathy [36, 72–74]. In this
sense, the expression of protective HSPs (i.e., Hsp27
and other chaperons) and survival pathways (MAPK
and AKT) induced by probiotic quorum sensing (i.e.,
CSF) might keep Aoligomers, and other aberrant
proteins related to AD, at sub-toxic concentrations in
the brain and other body sites (Fig. 7B).
To obtain experimental support for the proposed
roles of both B. subtilis properties as novel anti-AD
weapons, we performed lifespan assays in B. sub-
tilis isogenic NCIB3610 mutant strains affected by
the prolongevity effect and CSF production (Fig. 8).
Most of the B. subtilis prolongevity effect is medi-
ated by the proficiency of B. subtilis to form a healthy
biofilm in the host gut [19, 75]. Therefore, we used
an isogenic NCIB3610 bslA mutant strain deficient
in biofilm formation [19, 75] to test the contribu-
tion of the anti-aging effect of the bacterium to the
protection against human Apeptide expression.
As shown in Fig. 8, the mean lifespan of CL2006
AD worms colonized by NCIB3610-bslA cells
decreased by 18% compared with CL2006 worms
colonized by wild-type NCIB3610 cells (n= 100,
p< 0.001). Similarly, CL2006 worms colonized by
NCIB3610-csf cells (deficient in CSF synthesis)
AUTHOR COPY
S. Cogliati et al. / B. Subtilis and AD Treatment 1049
Fig. 8. Role of the anti-aging effect and CSF production for the anti-AD effect of probiotic B. subtilis. Lifespans of the AD CL2006 C. elegans
strain colonized by different bacterial strains: OP50, wild-type NCIB3610 and isogenic NCIB3610 strains deficient in biofilm formation
(bslA) or CSF production (csf). Worms were grown on bacteria-seeded 10-cm NGM agar plates at 20◦C, and survival was monitored as
indicated until the last worm died (see Materials and Methods for details). A typical output of three independent experiments performed in
duplicate is presented.
displayed a shorter lifespan (∼9% decrease of mean
lifespan) compared with CL2006 worms colonized
by wild-type NCIB3610 cells (n= 100, p< 0.001;
Fig. 8). Because BslA and CSF are secreted to the
extracellular matrix of the biofilm [19], mixtures of
bslA and csf mutant cells complement each other
to restore full biofilm-formation and CSF-production
proficiencies [19]. Therefore, we fed CL2006 worms
on a 50:50 mixture of bslA and csf B. subtilis
cells and measured the lifespan effect produced by
this mixture. As shown in Fig. 8, the colonization
of CL2006 worms by a mixture of both B. subtilis
mutant strains (bslA and csf cells in equal propor-
tions) restored the lifespan to a level indistinguishable
of the lifespan of CL2006 worms colonized by wild-
type NCIB3610 cells (mean values of 17.17 ±0.50
days and 16.53 ±047 days, respectively, n= 100,
p< 0.001; Fig. 8). These results support the novel
roles of the anti-aging effect and CSF-synthesis pro-
ficiencies of B. subtilis against AD (Fig. 7A, B).
There is a third pathway that B. subtilis might use
against AD, that is the production of nattokinase,
a 27.7-kDa serine enzyme produced by this bac-
terium [76, 77]. This protease is found in the Japanese
fermented food natto [76–78], and nattokinase-like
proteases are probably also found in other Asian and
African functional fermented foods [79]. Nattokinase
gained tremendous popularity as a fibrin-degrading
and clot-dissolving agent [76–78]. Interestingly, in
vivo, nattokinase can be absorbed across the human
intestinal tract [80, 81], and in vitro, it can degrade
Aoligomers [82, 83]; Aoligomers (A1-42 and
A1-40) are formed and deposited in the CNS as
well as in intestinal epithelial cells and the enteric
nervous system [84–87]. Therefore, intestinally pro-
duced nattokinase (and CSF) would help decrease
Aoligomers in the gastrointestinal tract [88, 89].
This phenomenon is important because intestinal
Aoligomers would interact with immune cells and
enteric neurons to be (at least partly) responsible for
the gastrointestinal dysfunctions of elderly, includ-
ing AD patients. Moreover, intestinal Aco-localizes
with the lipoprotein ApoB, and in wild-type mice (fed
on a diet rich in saturated fatty acids), deposits of
both proteins were found in the brain [60, 90]. These
data suggest that ApoB-Acomplexes produced in
the intestine might deliver intestinal-produced Ato
the brain [60, 90]. Alternatively, intestinally absorbed
nattokinase (and/or CSF) might use the gut-brain-axis
[60] to reach the leaky blood-brain barrier of elderly
and AD patients to exert their beneficial effects in situ
in the CNS [9, 60, 90, 91, 92]. Under these differ-
ent scenarios, the beneficial role of nattokinase in the
in vivo degradation of Aoligomers deserves future
investigations (Fig. 7C).
ACKNOWLEDGMENTS
The authors thank CONICET and FONCyT from
Argentina, the Fulbright (Washington DC) and Pew
AUTHOR COPY
1050 S. Cogliati et al. / B. Subtilis and AD Treatment
(Philadelphia) foundations for their support, and
CGC for providing worm strains and technical sup-
port. This work was funded by Fondo Nacional
Ciencia y Tecnolog´
ıa (FONCyT) PICT start up 2014-
3777 to RG.
Authors’ disclosures available online (https://
www.j-alz.com/manuscript-disclosures/19-0837r1).
REFERENCES
[1] Prince M, Bryce R, Albanese E, Wimo A, Ribeiro W, Ferri
CP (2013) The global prevalence of dementia: A systematic
review and meta-analysis. Alzheimers Dement 9, 63-75.e2.
[2] Shah H (2016) Research priorities to reduce the global bur-
den of dementia by 2025. Lancet Neurol 15, 1285-1294.
[3] Fox M, Knapp L, Andrews P, Fincher C (2013) Hygiene
and the world distribution of Alzheimer’s disease: Epi-
demiological evidence for a relationship between microbial
environment and age-adjusted disease burden. Evol Med
Public Health 1, 173-186.
[4] Itzhaki RF, Lathe R, Balin BJ, Ball MJ, Bearer EL, Braak
H, Bullido MJ, Carter C, Clerici M, Cosby SL, Del Tredici
K, Field H, Fulop T, Grassi C, Griffin WS, Haas J, Hudson
AP, Kamer AR, Kell DB, Licastro F, Letenneur L, L¨
ovheim
H, Mancuso R, Miklossy J, Otth C, Palamara AT, Perry
G, Preston C, Pretorius E, Strandberg T, Tabet N, Taylor-
Robinson SD, Whittum-Hudson JA (2016) Microbes and
Alzheimer’s disease. J Alzheimers Dis 51, 979-984.
[5] Hu X, Wang T, Jin F (2016) Alzheimer’s disease and gut
microbiota. Sci China Life Sci 59, 1006-1023.
[6] Moir R, Lathe R, Tanzi R (2018) The antimicrobial protec-
tion hypothesis of Alzheimer’s disease. Alzheimers Dement
14, 1602-1614.
[7] Mehta D, Jackson R, Paul G, Shi J, Sabbagh M (2017)
Why do trials for Alzheimer’s disease drugs keep failing? A
discontinued drug perspective for 2010–2015. Expert Opin
Investig Drugs 26, 735-739.
[8] Modrego P, Lobo A (2019) A good marker does not mean
a good target for clinical trials in Alzheimer´s disease: The
amyloid hypothesis questioned. Neurodegener Dis Manag
9, 11-121.
[9] Bhatt P, Pathak S, Kumar V, Panda B (2018) Attenuation of
neurobehavioral and neurochemical abnormalities in ani-
mal model of cognitive deficits of Alzheimer’s disease by
fermented soybean nanonutraceutical. Inflammopharmacol
26, 105-118.
[10] Scott K, Antoine J, Midtvedt T,van Hemert S (2015) Manip-
ulating the gut microbiota to maintain health and treat
disease. Microb Ecol Health Dis 26, 25877.
[11] Eskelinen M, Ngandu T, Helkala E, Tuomilehto J, Nissinen
A, Soininen H, Kivipelto M (2008) Fat intake at midlife
and cognitive impairment later in life: A population-based
CAIDE study. Int J Geriatr Psychiatry 23, 741-747.
[12] Van Cauwenberghe C, Van Broeckhoven C, Sleegers K
(2016) The genetic landscape of Alzheimer disease: Clinical
implications and perspectives. Genet Med 18, 421-430.
[13] Modrego P, Lobo A (2018) Determinants of mortality and
progression in Alzheimer’s disease. A systematic review.
Neuropsychiatry (London) 8, 1465-1475.
[14] Alzheimer’s Association (2013) Alzheimer’s disease facts
and figures. Alzheimers Dement 9, 208-245.
[15] Orgeta V, Mukadam N, Sommerlad A, Livingston G (2019)
The Lancet Commission on Dementia Prevention, Interven-
tion, and Care: A call for action. Irish J Psychol Med 36,
85-88.
[16] Lombardia E, Rovetto A, Arabolaza A, Grau R (2006) A
LuxS-dependent cell to cell language regulates social behav-
ior and development in Bacillus subtilis.J Bacteriol 188,
4042-4052.
[17] Pedrido M, de O˜
na P, Ramirez W, Le˜
nini C, Go˜
ni A, Grau
R (2013) Spo0A links de novo fatty acid synthesis to sporu-
lation and biofilm development in Bacillus subtilis.Mol
Microbiol 87, 348-367.
[18] Grau R, de O˜
na P, Kunert M, Le ˜
nini C, Gallegos-Monterrosa
R, Mhatre E, Vileta D, Donato V, H¨
olscher T, Boland W,
Kuipers P, Kov´
acs ´
A (2015) A duo of potassium-responsive
histidine kinases govern the multicellular destiny of Bacillus
subtilis. Mbio 6, e00581.
[19] Donato V, Rodriguez Ayala F, Cogliati S, Bauman C, Costa
J, Le˜
nini C, Grau R (2017) Bacillus subtilis biofilm extends
Caenorhabditis elegans longevity through downregulation
of the insulin-like signaling pathway.Nat Commun 8, 14332.
[20] Kenyon C (2010) The genetics of ageing. Nature 464, 504-
512.
[21] Kenyon C (2011) The first long-lived mutants: Discovery of
the insulin/IGF-1 pathway for ageing. Philos Trans R Soc
Lond Ser B Biol Sci 366, 9-16.
[22] Griffin E, Caldwell K, Caldwell G (2017) Genetic and
pharmacological discovery for Alzheimer’s disease using
Caenorhabditis elegans.ACS Chem Neurosci 8, 2596-2606.
[23] Alexander A, Marfil V, Li C (2014) Use of Caenorhabditis
elegans as a model to study Alzheimer’s disease and other
neurodegenerative diseases. Front Genet 5, 279.
[24] McColl G, Roberts B, Gunn A, Perez K, Tew D, Masters C,
Barnham K, Cherny R, Bush A (2009) The Caenorhabditis
elegans A1-42 model of Alzheimer disease predominantly
expresses A3-42. J Biol Chem 284, 22697-22702.
[25] Dosanjh L, Brown M, Rao G, Link C, Luo Y (2010) Behav-
ioral phenotyping of a transgenic Caenorhabditis elegans
expressing neuronal amyloid-.J Alzheimers Dis 19, 681-
690.
[26] McColl G, Roberts B, Pukala T, Kenche V, Roberts C, Link
C, Ryan T,Masters C, Barnham K, Bush A, Cherny R (2012)
Utility of an improved model of amyloid-beta (A1-42)
toxicity in Caenorhabditis elegans for drug screening for
Alzheimer´s disease. Mol Neurodegener 7, 57.
[27] Chen CH, Chen Y, Jiang H, Chen CK, Pan C (2013) Neu-
ronal aging: Learning from C. elegans. J Mol Signal 8,
14.
[28] Olle B (2013) Medicines from microbiota. Nat Biotechnol
31, 309-315.
[29] Bischoff S (2016) Microbiota and aging. Curr Opin Clin
Nutr Metab Care 19, 26-30.
[30] Ayala FR, Bauman C, Cogliati S, Le˜
nini C, Bartolini M,
Grau R (2017) Microbial flora, probiotics, Bacillus sub-
tilis and the search for a long and healthy human longevity.
Microbial Cell 4, 133-136.
[31] Honing M, Huma R (1989) Dil and Dio: Versatile fluores-
cent dyes for neuronal labeling and pathway tracing. Trends
Neurosci 12, 333-335.
[32] Toth M, Melentijevic I, Shah L, Bhatia A, Lu K, Talwar
A, Naji H, Ibanez-Ventoso C, Ghose P, Jevince A, Xue
J, Herndon L, Bhanot G, Rongo C, Hall D, Driscoll M
(2012) Neurite sprouting and synapse deterioration in the
aging Caenorhabditis elegans nervous system. J Neurosci
32, 8778-8790.
AUTHOR COPY
S. Cogliati et al. / B. Subtilis and AD Treatment 1051
[33] Offenburger S, Ho X, Tachie-Menson T, Coakley S, Hilliard
M, Gartner A (2016) 6-OHDA-induced dopaminergic neu-
rodegeneration in Caenorhabditis elegans is promoted by
the engulfment pathway and inhibited by the transthyretin-
related protein TTR-33. PLoS Genet 14, e1007125.
[34] White J, Southgate E, Thomson J, Brenner S (1986) The
structure of the nervous system of the nematode Caenorhab-
ditis elegans.Philos Trans R Soc Lond B Biol Sci 314,
1-340.
[35] Maloney B, Lahiri D (2016) Epigenetics of dementia:
Understanding the disease as a transformation rather than a
state. Lancet Neurol 15, 760-774.
[36] Robinson J, Lee E, Xie S, Rennert L, Suh E, Bredenberg
C, Caswell C, Van Deerlin V, Yan N, Yousef A, Hurtig H,
Siderowf A, Grossman M, McMillan C, Miller B, Duda J,
Irwin D, Wolk D, Elman L, McCluskey L, Chen-Plotkin
A, Weintraub D, Arnold S, Brettschneider J, Lee V, Tro-
janowski J (2018) Neurodegenerative disease concomitant
proteinopathies are prevalent, age-related, and APOE4-
associated. Brain 141, 2181-2193.
[37] Saharia K, Arya U, Kumar R, Sahu R, Kumar Das C, Gupta
K, Dwivedi H, Subramaniam J (2012) Reserpine modulates
neurotransmitter release to extend lifespan and alleviate age-
dependent Aproteotoxicity in Caenorhabditis elegans.
Exp Gerontol 47, 188-197.
[38] Keowkase R, Shoomarom N, Bunargin W, Sitthithaworn W,
Weerapreeyakul N (2018) Sesamin and sesamolin reduce
amyloid-toxicity in a transgenic Caenorhabditis elegans.
Biomed Pharmacother 107, 656-664.
[39] Gusarov I, Gautier L, Smolentseva O, Shamovsky I, Erem-
ina S, Mironov A, Nudler E (2013) Bacterial nitric oxide
extends the lifespan of C. elegans. Cell 152, 818-830.
[40] Williams P (2007) Bacillus subtilis: A shocking message
from a probiotic. Cell Host Microbe 1, 248-249.
[41] Seo W,Nam S, Lee C, Cho K (2011) Identification of poten-
tial Bacillus subtilis probiotics from Korean soybean paste
and their antimicrobial and immune activities. Prev Nutr
Food Sci 16, 37-44.
[42] Hanifi A, Culpepper T,Mai V,Anand A, Ford AL, Ukhanova
M, Christman M, Tompkins T, Dahl W (2015) Evalua-
tion of Bacillus subtilis RO179 on gastrointestinal viability
and general wellness: A randomized, double-blind, placebo-
controlled trial in healthy adults. Benef Microbes 6, 19-27.
[43] Lee S, Lee J, Jin Y, Jeong J, Chang Y,Lee Y, Jeong Y, Kim M
(2017) Probiotic characteristics of Bacillus strains isolated
from Korean traditional soy sauce. LWT - Food Sci Tech 79,
518-524.
[44] Lefevre M, Racedo S, Denayrolles M, Ripert G, Des-
fougeres T, Lobach A, Simon R, Pelerin F, J¨
usten P, Urdaci
M (2017) Safety assessment of Bacillus subtilis CU1 for use
as a probiotic in humans. Regulatory Toxicol Pharmacol 83,
54-65.
[45] Piewngam P, Zheng Y, Nguyen T, Dickey S, Joo H, Villaruz
A, Glose K, Fisher E, Hunt R, Li B, Chiou J, Pharkjaksu
S, Khongthong S, Cheung G, Kiratisin K, Otto M (2018)
Pathogen elimination by probiotic Bacillus via signalling
interference. Nature 562, 532-537.
[46] Thomas C, Versalovic J (2010) Probiotics-host communi-
cation: Modulation of signaling pathways in the intestine.
Gut Microbes 1–3, 148-163.
[47] VidyaLaxme B, Rovetto A, Grau R, Agrawal R (2014) Syn-
ergistic effects of probiotic Leuconostoc mesenteroides and
Bacillus subtilis in malted ragi (Eleucine corocana) food for
antagonistic activity against V. cholerae and other beneficial
properties. J Food Sci Technol 51, 3072-3082.
[48] Sarkar A, Lehto S, Harty S, Dinan T, Cryan J, Burnet P
(2016) Psychobiotics and the manipulation of bacteria-gut-
brain signals. Trends Neurosci 39, 763-781.
[49] Sherwin E, Rea K, Dinan T, Cryan J (2016) A gut (micro-
biome) feeling about the brain. Curr Opin Gastroent 32,
96-104.
[50] Kim N, Yun M, Oh Y, Choi H (2018) Mind-altering with
the gut: Modulation of the gut-brain axis with probiotics.
J Microbiol 56, 172-182.
[51] Saulnier D, Ringel Y, Heyman M, Foster J, Bercik P, Scgul-
man R, Versalovic J, Verdu E, Dinan T, Hecht G, Guarner F
(2017) The intestinal microbiome, probiotics and prebiotics
in neurogastroenterology. Gut Microbes 4, 17-27.
[52] Barrett E, Ross R, O’Toole P, Fitzgerald G, Stanton C
(2012) Gamma-Aminobutyric acid production by culturable
bacteria from the human intestine. J Appl Microbiol 113,
411-417.
[53] Reigstad C, Salmonson C, Rainey J, Szurszewski J, Lin-
den D, Sonnenburg J, Farrugia G, Kashyap P (2015) Gut
microbes promote colonic serotonin production through an
effect of short-chain fatty acids on enterochromaffin cells.
FASEB J 29, 1395-1403.
[54] Yano J, Yu K, Donaldson G, Shastri G, Ann P, Ma L, Nagler
C, Ismagilov R, Mazmanian S, Hsiao E (2015) Indigenous
bacteria from the gut microbiota regulate host serotonin
biosynthesis. Cell 161, 264-276.
[55] Hardy J, Higgins G (1992) Alzheimer’s disease: The amy-
loid cascade hypothesis. Science 256, 184-185.
[56] Karran E, De Strooper B (2016) The amyloid cascade
hypothesis: Are we poised for success or failure? J Neu-
rochem 139, 237-252.
[57] Wu S, Cao Z, Chang K, Juang J (2017) Intestinal microbial
dysbiosis aggravates the progression of Alzheimer´s disease
in Drosophila.Nat Comunn 8, 24.
[58] Akbari E, Asemi Z, Kakhaki R, Bahmani F, Kouchaki E,
Tamtaji O, Hamidi G, Salami M (2016) Effect of probiotic
supplementation on cognitive function and metabolic status
in Alzheimer´s disease: A randomized, double-blind and
controlled trial. Front Aging Neurosci 8, 256.
[59] Leblhuber F, Egger M, Schuetz B, Fuchs D (2018) Com-
mentary: Effect of probiotic supplementation on cognitive
function and metabolic status in Alzheimer’s disease: A
randomized, double-blind and controlled trial. Front Aging
Neurosci 10, 54.
[60] Kowalski K, Mulak A (2019) Brain-gut microbiota axis
in Alzheimer’s disease. J Neurogastroenterol Motil 25,
48-60.
[61] Lahiri D (2019) Lessons from Alzheimer’s disease (AD)
clinical trials: Instead of “A-drug”,AD-D prevention to avert
AD. Curr Alzheimer Res 16, 279-280.
[62] Robertson R, Manges A, Finlay B, Prendergast A (2019)
The human microbiome and child growth – first 1000 days
and beyond. Trends Microbiol 27, 131-147.
[63] Cohen E, Bieschke J, Perciavalle R, Kelly J, Dillin A (2006)
Opposing activities protect against age-onset proteotoxicity.
Science 313, 1604-1610.
[64] Hsu A, Murphy C, Kenyon C (2003) Regulation of aging
and age-related diseases by DAF-16 and heat-shock factor.
Science 300, 1142-1145.
[65] Patel N, Gordon M, Connor K, Good R, Engelman R, Mason
J, Morgan D, Morgan T, Finch C (2005) Caloric restric-
tion attenuates a beta-deposition in Alzheimer transgenic
models. Neurobiol Aging 26, 995-1000.
[66] Steinkraus K, Smith E, Davis C, Carr D, Pendergrass W, Sut-
phin G, Kennedy B, Kaeberlein M (2008) Dietary restriction
AUTHOR COPY
1052 S. Cogliati et al. / B. Subtilis and AD Treatment
suppresses proteotoxicity and enhances longevity by an hsf-
1-dependent mechanism in Caenorhabditis elegans.Aging
Cell 7, 394-404.
[67] Lazazzera B, Solomon J, Grossman A (1997) An exported
peptide functions intracellularly to contribute to cell density
signaling in B. subtilis.Cell 89, 917-925.
[68] Abisado R, Benomar S, Klaus J, Dandekar A, Chandlera J
(2018) Bacterial quorum sensing and microbial community
interactions. mBio 9, e02331-e02317.
[69] Fujiya M, Musch M, Nakagawa Y, Hu S, Alverdy J, Kohgo
Y, Schneewind O, Jabri B, Chang E (2007) The Bacillus sub-
tilis quorum-sensing molecule CSF contributes to intestinal
homeostasis via OCTN2, a host cell membrane transporter.
Cell Host Microbe 1, 299-308.
[70] Arrigo A, Virot S, Chaufour S, Firdaus W, Kretz-Remy
C, Diaz-Latoud C (2005) Hsp27 consolidates intracellular
redox homeostasis by upholding glutathione in its reduced
form and by decreasing iron intracellular levels. Antioxid
Redox Signal 7, 414-422.
[71] Voisine C, Pedersen J, Morimoto R (2010) Chaperone net-
works: Tipping the balance in protein folding diseases.
Neurobiol Dis 40, 12-20.
[72] Fonte V, Kipp R, Yerg III J, Merin D, Forrestal M, Wagner
E, Roberts C, Link C (2008) Suppression of in vivo beta-
amyloid peptide toxicity by overexpression of the HSP-16.2
small chaperone protein. J Biol Chem 283, 784-791.
[73] Drummond E, Nayak S, Faustin A, Pires G, Hickman R,
Askenazi M, Cohen M, Haldiman T, Kim C, Han X, Shao
Y, Safar J, Ueberheide B, Wisniewski T (2017) Proteomic
differences in amyloid plaques in rapidly progressive and
sporadic Alzheimer’s disease. Acta Neuropathol 133, 933-
954.
[74] Li K, Ganz A, Smit A (2018) Proteomics of neurodegen-
erative diseases: Analysis of human postmortem brain. J
Neurochem 151, 435-445.
[75] Smolentseva O, Gusarov I, Gautier L, Shamovsky I, De
Francesco A, Losick R, Nudler E (2017) Mechanism of
biofilm-mediated stress resistance and lifespan extension
in C. elegans. Sci Rep 7, 7137.
[76] Sumi H, Hamada H, Tsushima H, Mihara H, Muraki H
(1987) Novel fibrinolytic enzyme (nattokinase) in the veg-
etable cheese Natto; a typical and popular soybean food in
the Japanese diet. Experientia 43, 1110-1111.
[77] Urano T, Ihara H, Umemura K, Suzuki Y, Oike M, Akita S,
Tsukamoto Y, Suzuki I, Takada A (2001) The profibrinolytic
enzyme subtilisin NAT purified from Bacillus subtilis
cleaves and inactivates plasminogen activator inhibitor type
1. J Biol Chem 276, 24690-24696.
[78] Suzuki Y, Kondo K, Matsumoto Y, Zhao B, Otsuguro K,
Maeda T, Tsukamoto Y, Urano T, Umemura K (2003)
Dietary supplementation of fermented soybean, natto, sup-
presses intimal thickening and modulates the lysis of mural
thrombi after endothelial injury in rat femoral artery. Life
Sci 73, 1289-1298.
[79] Go J, Kim J, Kwak M, Koh E, Song S, Sung J, Kim D, Hong
J, Hwang D (2015) Neuroprotective effects of fermented
soybean products (Cheonggukjang) manufactured by mixed
culture of Bacillus subtilis MC31 and Lactobacillus sakei
383 on trimethyltininduced cognitive defects mice. Nutr
Neurosci 19, 247-259.
[80] Sumi H, Hamada H, Nakanishi K, Hiratani H (1990)
Enhancement of the fibrinolytic activity in plasma by oral
administration of nattokinase. Acta Haematol 84, 139-143.
[81] Fujita M, Hong K, Ito Y, Misawa S, Takeuchi N, Kariya K,
Nishimuro S (1995) Transport of nattokinase across the rat
intestinal tract. Biol Pharm Bull 18, 1194-1196.
[82] Hsu R, Lee K, Wang J, Lee L, Chen R (2009) Amyloid-
degrading ability of nattokinase from Bacillus subtilis natto.
J Agric Food Chem 57, 503-508.
[83] Ji H, Yu L, Liu K, Yu Z, Zhang Q, Zou F, Liu B (2014) Mech-
anisms of Nattokinase in protection of cerebral ischemia.
Eur J Pharmacol 745, 144-151.
[84] Joachim C, Mori H, Selkoe D (1989) Amyloid beta-protein
deposition in tissues other than brain in Alzheimer’s disease.
Nature 341, 226-230.
[85] Arai H, Lee V, Messinger M, Greenberg B, Lowery D,
Trojanowski J (1991) Expression patterns of beta-amyloid
precursor protein (beta-APP) in neural and nonneural
human tissues from Alzheimer’s disease and control sub-
jects. Ann Neurol 30, 686-693.
[86] Cabal A, Alonso-Cortina V, Gonzalez-Vazquez L, Naves F,
Del Valle M, Vega J (1995) -Amyloid precursor protein
(APP) in human gut with special reference to the enteric
nervous system. Brain Res Bull 38, 417-423.
[87] Kendra L, Manocha D, Combs C (2015) Amyloid precursor
protein mediated changes in intestinal epithelial phenotype
in vitro. PLoS One 10, e0119534.
[88] Fadl N, Ahmed H, Booles H, Sayed A (2013) Serrapeptase
and nattokinase intervention for relieving Alzheimer´s dis-
ease pathophysiology in rat model. Hum Exp Toxicol 37,
721-735.
[89] Wang Y, Zhou H, Zhou X (2006) Clearance of amyloid-beta
in Alzheimer’s disease: Progress, problems and perspec-
tives. Drug Discov Today 11, 931-938.
[90] TakechiR, Galloway S, Pallebage-Gamarallage M, Welling-
ton C, Johnsen R, Dhaliwal S, Mamo J (2010) Differential
effects of dietary fatty acids on the cerebral distribution of
plasma-derived apo B lipoproteins with amyloid-beta. Br J
Nutr 103, 52-62.
[91] Nelson AR, Sagare AP, Zlokovic BV (2016) Blood–brain
barrier transport of Alzheimer’s amyloid -peptide. In
Developing Therapeutics for Alzheimer’s Disease.Progress
and Challenges, Wolfe MS, ed. Academic Press, pp. 251-
263.
[92] Braniste V, Al-Asmakh M, Kowal C, Anuar F, Abbaspour
A, TothM, Korecka A, Bakocevic N, Ng L, Kundu P, Guly ´
as
B, Halldin C, Hultenby K, Nilsson H, Hebert H, Volpe B,
Diamond B, Pettersson S (2014) The gut microbiota influ-
ences blood-brain barrier permeability in mice. Sci Transl
Med 6, 263-158.
