Eur J Nutr
DOI 10.1007/s00394-013-0649-9
ORIGINAL CONTRIBUTION
Lactobacillus strains isolated from infant faeces possess potent
inhibitory activity against intestinal alpha- and beta-glucosidases
suggesting anti-diabetic potential
Harsh Panwar • Danielle Calderwood •
Irene R. Grant • Sunita Grover • Brian D. Green
Received: 17 September 2013 / Accepted: 27 December 2013
Ó Springer-Verlag Berlin Heidelberg 2014
Abstract
Purpose Inhibitors of intestinal alpha-glucosidases are
used therapeutically to treat type 2 diabetes mellitus.
Bacteria such as Actinoplanes sp. naturally produce potent
alpha-glucosidase inhibitor compounds, including the most
widely available drug acarbose. It is not known whether
lactic acid bacteria (LAB) colonising the human gut pos-
sess inhibitory potential against glucosidases. Hence, the
study was undertaken to screen LABs having inherent
alpha- and beta-glucosidase inhibitory potential.
Methods This study isolated, screened, identified and
extracted Lactobacillus strains (Lb1–15) from human
infant faecal samples determining their inhibitory activity
against intestinal maltase, sucrase, lactase and amylase.
Lactobacillus reference strains (Ref1–7), a Gram positive
control (Ctrl1) and two Gram negative controls (Ctrl2–3),
were also analysed to compare activity.
Results Faecal isolates were identified by DNA
sequencing, with the majority identified as unique strains of
Lactobacillus plantarum. Some strains (L. plantarum, L.
fermentum, L. casei and L. rhamnosus) had potent and
broad spectrum inhibitory activities (up to 89 %;
p \ 0.001; 500 mg/ml wet weight) comparable to acarbose
Electronic supplementary material The online version of this
article (doi:10.1007/s00394-013-0649-9) contains supplementary
material, which is available to authorized users.
H. Panwar  D. Calderwood  I. R. Grant  B. D. Green (&)
Institute for Global Food Security, School of Biological
Sciences, Queen’s University Belfast, David Keir Building,
Stranmillis Road, Belfast BT9 5AG, UK
e-mail: b.green@qub.ac.uk
H. Panwar  S. Grover
Molecular Biology Unit, Dairy Microbiology Division,
National Dairy Research Institute, Karnal, Haryana, India
(up to 88 %; p \ 0.001; 30 mg/ml). Inhibitory activity was
concentration-dependent and was freely available in the
supernatant, and was not present in other bacterial genera
(Bifidobacterium bifidum and Escherichia coli or Salmo-
nella typhimurium). Interestingly, the potency and spec-
trum of inhibitory activity across strains of a single species
(L. plantarum) differed substantially. Some Lactobacillus
extracts had broader spectrum activities than acarbose,
effectively inhibiting beta-glucosidase activity (lactase) as
well as alpha-glucosidase activities (maltase, sucrase and
amylase). Anti-diabetic potential was indicated by the fact
that oral gavage with a L. rhamnosus extract (1 g/kg) was
able to reduce glucose excursions (Area under curve;
22 %; p \ 0.05) in rats during a carbohydrate challenge
(starch; 2 g/kg).
Conclusion These results definitively demonstrate that
Lactobacillus strains present in the human gut have alpha-
and beta-glucosidase inhibitory activities and can reduce
blood glucose responses in vivo. Although the potential use
of LAB such as Lactobacillus as a dietary supplement,
medicinal food or biotherapeutic for diabetes is uncertain,
such an approach might offer advantages over drug therapies
in terms of broader spectrum activities and fewer unpleasant
side effects. Further characterisation of this bioactivity is
warranted, and chronic studies should be undertaken in
appropriate animal models or diabetic subjects.
Keywords Lactobacillus  Alpha-glucosidase 
Beta-glucosidase
Abbreviations
LAB Lactic acid bacteria
GRAS Generally recognised as safe
SE Sonicated extract
SN Supernatant
123

Introduction
Glucosidase enzymes are located in the brush border of the
intestine and are primarily responsible for the breakdown
of complex oligosaccharides and disaccharides into glu-
cose for subsequent absorption by the gut [1, 2]. The
concentration and kinetics of intestinal glucosidases con-
tribute significantly to post-prandial blood glucose levels
[3] which are kept in check by several glucoregulatory
mechanisms, most notably by the secretion of pancreatic
insulin. In humans with diabetes mellitus, glucose homo-
eostasis is severely compromised due to impaired insulin
secretion/action. The inhibition of alpha-glucosidase
enzymes is an efficacious way of curbing post-prandial
hyperglycaemia, acting to delay carbohydrate digestion and
preventing excessive glucose absorption. Indeed, intestinal
alpha-glucosidases are targeted by commercially available
alpha-glucosidase inhibitors such as acarbose, voglibose
and miglitol, which competitively bind and inhibit alpha-
glucosidase enzymes [4]. Studies demonstrate that acar-
bose controls the release/absorption of glucose after a meal
and is effective in preventing post-prandial hyperglycaemia
in diabetic patients [5, 6]. Aside from diabetic therapy, the
study of alpha-glucosidase activity has relevance for car-
bohydrate uptake disorders, obesity, dental caries, peri-
odontal diseases and anti-viral therapy [1, 7].
Although alpha-glucosidase inhibitors have established
efficacy and safety [8–10], they have common side effects
such as flatulence and diarrhoea [11], and alternative
inhibitors are sought. For example, natural plant sources
have been recently screened for inhibitory activity [12–17].
It is noteworthy that the most frequently used inhibitor
(acarbose), clinically used for more than 20 years, is a
fermentation product of a Gram positive aerobic bacte-
rium: Actinoplanes sp. [18]. This bacterium, unlike probi-
otic species such as lactobacilli, is not common and does
not normally interact with humans or human foodstuffs.
Earlier, it was reported that a filtrate of lactic acid bacteria
(LAB) hydrolysate of skimmed milk can inhibit alpha-
glucosidase [19], but it is not known whether LAB can
directly inhibit alpha-glucosidase. Administration of LAB
is a potentially novel and useful lifestyle intervention for
controlling hyperglycaemia and in the future could act as
an adjunct to diabetes treatment [20]. Some LABs have
recognised anti-inflammatory effects on the intestine and
are used clinically [21]. For example, VSL#3 is a probiotic
bacterial preparation containing 450–900 billion live bac-
teria that has been classified by the FDA as a ‘medicinal
food’ designated for the dietary management of three major
gastrointestinal conditions: ulcerative colitis, ileal pouch
and irritable bowel syndrome [22, 23]. Clinically, these
probiotics have proven safety and tolerability and any side
123
Eur J Nutr
effects exhibited are uncommon [22]. The potential use of
probiotic organisms in diabetes treatment is understudied,
and anti-diabetic effects of bacteria colonising the human
gut should be investigated.
The aim of this study was to probe the anti-diabetic
potential of one LAB genus (Lactobacillus) by measuring
the ability of a range of bacterial extracts to inhibit alpha-
and beta-glucosidase enzymes. Using Lactobacillus strains
isolated from infant faeces, we evaluated their inherent
ability to inhibit a spectrum of intestinal carbohydrate-
digesting enzymes: maltase, sucrase, lactase and amylase.
Isolates were subsequently identified, and their activities
compared against reference and control strains. Activities
of extracts were further characterised by studying con-
centration-dependent effects and by monitoring acute
in vivo glucose responses following an oral carbohydrate
challenge.
Materials and methods
Chemicals and reagents
De Man, Rogosa and Sharpe broth (MRS) (M369) was
obtained from HiMedia Laboratories (Mumbai, India).
Mueller–Hinton broth (CM0405) from Oxoid (Hampshire,
UK). L-cysteine hydrochloride, acarbose, PBS (109), rat
intestinal acetone powder, maltose, sucrose, lactose and
starch were obtained from Sigma (Poole, Dorset, UK).
Isolation and identification of Lactobacillus strains
Faecal samples were collected from five healthy breast-fed
infants \9 months in age living in Shamli, Uttar Pradesh,
India. In each case, full parental consent was obtained. The aim
of study was to screen for bioactivity and therefore only single
samples were collected. Intra-individual day to day variability
and the inter-individual variability of microorganisms were not
assessed. Lactobacillus cultures were isolated from faecal
samples of healthy human infants (Lb1–15; Table 1). Lacto-
bacillus reference strains (Ref1–7; Table 1) and a Gram
positive control (Bifidobacterium bifidum; Ctrl1; Table 1)
were obtained from the National Collection of Industrial, Food
and Marine Bacteria (Aberdeen, UK). Escherichia coli K12
(Ctrl 2; Table 1) was procured from National Collection of
Type Cultures (NCTC) (Colindale, London). Salmonella
typhimurium (Ctrl 3; Table 1) was isolated from a pig carcass
swab at Agri-Food and Biosciences Institute for Northern
Ireland (Newforge Lane, Belfast) and kindly provided by Dr
Robert Madden. Identity of Lactobacillus isolates was deter-
mined to genus level by PCR using genus-specific primer pair
LbLMA1/R-161 (forward 50 -ctcaaaactaaacaaagtttc-30 ; reverse
50 -ctcgtacttgtacacaccgcccgtca-30 ) [24] targeting 16SrRNA
Eur J Nutr

Table 1 List of bacterial strains examined in the study

Sample Strain code Type Identification % Sequence Accession or culture
similarity collection no.

1 Lb1 Faecal isolate Strain of Lactobacillus plantarum 98 % Unknown

2 Lb2 Faecal isolate Strain of Lactobacillus plantarum 99 % Unknown
3 Lb3 Faecal isolate Lactobacillus plantarum subsp. Argentorotensis 99 % KC491380
4 Lb4 Faecal isolate Strain of Lactobacillus plantarum 99 % KF678450
5 Lb5 Faecal isolate Strain of Lactobacillus plantarum 95 % Unknown
6 Lb6 Faecal isolate Strain of Lactobacillus plantarum 96 % Unknown
7 Lb7 Faecal isolate Strain of Lactobacillus fermentum 98 % Unknown
8 Lb8 Faecal isolate Strain of Lactobacillus plantarum 99 % KF678451
9 Lb9 Faecal isolate Strain of Lactobacillus plantarum 99 % KF678452
10 Lb10 Faecal isolate Strain of Lactobacillus plantarum 99 % KF678453
11 Lb11 Faecal isolate Strain of Lactobacillus plantarum 99 % Unknown
12 Lb12 Faecal isolate Lactobacillus sp. 99 % Unknown
13 Lb13 Faecal isolate Strain of Lactobacillus fermentum 97 % KC866340
14 Lb14 Faecal isolate Strain of Lactobacillus plantarum 99 % Unknown
15 Lb15 Faecal isolate Strain of Lactobacillus acidophilus 99 % Unknown
16 Ref1 Reference culture Lactobacillus acidophilus n/a NCIMB701748
17 Ref2 Reference culture Lactobacillus casei n/a NCIMB4114
18 Ref3 Reference culture Lactobacillus fermentum n/a NCIMB2797
19 Ref4 Reference culture Lactobacillus johnsonii n/a NCIMB8795
20 Ref5 Reference culture Lactobacillus paracasei n/a NCIMB1407
21 Ref6 Reference culture Lactobacillus plantarum n/a NCIMB1406
22 Ref7 Reference culture Lactobacillus rhamnosus n/a NCIMB6375
23 Ctrl1 Gram positive control Bifidobacterium bifidum n/a NCIMB702715
24 Ctrl2 Gram negative control Escherichia coli K12 n/a NCTC 10538
25 Ctrl3 Gram negative control Salmonella typhimurium n/a Unknown
Bacterial strains Lb1–15 were isolated from faeces from healthy human infants. Reference strains (Ref1–7) were obtained from NCIMB
n/a not applicable

region. Sequencing was carried out using 16SrRNA (forward
50 -ccagagtttgatcmtggctcag-30 ; reverse 50 -cggttaccttgttacgactt-
cacc-30 ) [25, 26] and Phe (forward 50 -tatttcaaaattgcraaacgr-30 ;
reverse 50 -cccwgcwcgtgatatgca-30 ) specific primers [27].
Amplified products (1,400 bp for 16SrRNA and 600 bp for
Phe) were sequenced using an external DNA sequencing ser-
vice (University of Dundee, UK).
Sample preparation
Bacterial extracts were assessed for their ability to inhibit
glucosidase enzymes under in vitro conditions. All Lacto-
bacillus isolates and reference cultures were inoculated in
MRS broth; B. bifidum under anaerobic conditions in MRS
fortified with 0.5 g/l L-cysteine hydrochloride (Sigma,
Poole, Dorset, UK), and Gram negative cultures, Salmo-
nella typhimurium and E. coli, in Mueller–Hinton broth
and incubated overnight at 37 °C. In order to achieve
maximum activity, all isolates and reference cultures were
sub-cultured three times under optimum conditions before
use for inhibition studies. Bacterial culture (1 %) was
inoculated into 50 ml of broth and incubated overnight
under optimum conditions (37 °C). Overnight broth cul-
tures were harvested and washed twice in 19 PBS
(12,000 g/15 min) followed by re-suspension in nuclease-
free water on weight/volume basis. Bacterial pellets of
500 mg (wet weight) were re-suspended in 1 ml nuclease-
free water and mixed by pipetting and vortexing. Minor
variations were observed in optical density of diluted
samples at 600 nm. Bacterial suspensions were heat-killed
(65 °C/30 min) in a water bath, sonicated (20 kHz,
3 9 30 s pulse) and stored at -70 °C until assayed for
their glucosidase inhibition activity. These heat-killed
sonicated extracts (SE) of bacteria were screened for their
glucosidase inhibitory potential in presence of different
sugars and rat intestinal acetone powder.

123

Measurement of glucosidase inhibition
Glucosidase inhibition was assessed by monitoring the rate
of glucose formation from carbohydrates incubated with rat
intestinal extract (Sigma, Dorset, UK) in the presence of
each bacterial extract. Reaction mixtures contained 300 ll
of each carbohydrate (30 mg/ml) and 150 ll of either PBS
(pH 7.2) (control), acarbose (30 mg/ml, as a positive con-
trol) or heat-killed sonicated bacterial extract (500 mg/ml
wet weight). These were pre-incubated in a water bath
(37 °C; 10 min), and reactions were initiated by the addi-
tion of rat intestinal acetone extract. The optimal amount of
intestinal extract required for maltose, sucrose, lactose and
starch incubations was determined by a series of pre-
liminary experiments. Glucose concentrations were mea-
sured at 0, 30, 60 and 90 min incubation time points using
an enzymatic assay kit (Analox Ltd. London, UK) and
detected on a PGM7 Micro-Stat Analyser (Analox Instru-
ments Ltd; London, UK).
In order to assess the concentration-dependent effect of
heat-killed sonicated bacteria over their glucosidase
inhibitory potential, various concentrations were tested
in vitro. Heat-killed sonicated bacterial extracts were pre-
pared as described previously with bacterial pellets sus-
pended in increasing amounts of nuclease-free water to
generate final concentrations of 500, 350, 300, 250 and
200 mg/ml, and glucosidase inhibition was determined as
described above. To assess the localisation of glucosidase
inhibitory activity, cell SE was compared to the aqueous
cell-free supernatant (SN) from three strains: Lb3, Lb4 and
Lb5. Each sonicated cell extract was clarified of cell debris
by centrifugation (10 min; 12,000 g; 4 °C).
In vivo responses of oral Lactobacillus rhamnosus
to a maltose/starch challenge
Acute inhibition of glucosidase activity in vivo was
assessed by maltose and starch tolerance tests. Sprague-
Dawley rats aged 6–8 weeks were housed in an air-con-
ditioned room at 22 ± 2 °C with a 12-h light cycle
(0600–1800 hour): 12-h dark cycle (1800–0600 hour).
Drinking water and a standard rodent maintenance diet
(Teklad global rodent diet, Harlan, UK) were freely
available. All animal experiments were carried out in
accordance with the UK Animals (Scientific Procedures)
Act 1986. No adverse effects were observed following
administration of any of the compounds. Rats were fasted
for 16 h before oral gavage with maltose or starch (2 g/kg),
or maltose or starch in combination with heat-killed soni-
cated L. rhamnosus extract (1 g/kg). Blood glucose was
analysed immediately before administration and 15, 30, 60
and 105 min post-administration using a Free Style Blood
Glucose Monitor (Abbott Laboratories Ireland Ltd.,
123
Eur J Nutr
Dublin, Ireland). L. rhamnosus was selected for in vivo
experiments because the reference cultures had the highest
inhibitory activities (albeit very marginally higher than L.
casei) against maltose and amylase.
Data analysis
Results were expressed as mean ± SEM. Percentage
inhibition of maltase, lactase, sucrase and amylase was
calculated by measuring area under curve (AUC0-90 min)
and expressing it as a percentage of control. Plasma glu-
cose data were compared using the unpaired Student’s
t test. Area under the curve values, data were compared
using repeated measures one-way analysis of variation
(ANOVA) followed by the Student–Newman–Keuls post
hoc test. Incremental areas under plasma glucose curves
(DAUC0-105 min) were calculated using a computer-
generated programme employing the trapezoidal rule [28]
with baseline subtraction. Groups of data were considered
to be significantly different if p \ 0.05.
Results
Identification and selection of isolates
DNA isolated from Gram positive rod-shaped bacteria
obtained from infant faeces was amplified with genus-
specific primers, and individual isolates generating a single
PCR amplicon of around 250 bp were identified as genus
Lactobacillus (Supplementary Figure 1). The identity of
Lactobacillus isolates to species level, as determined by
DNA sequencing, is presented in Table 1.
Inhibitory effects of Lactobacillus extracts
on glucosidase enzymes
Glucosidase inhibition in Lactobacillus isolates
Extracts from a majority of Lactobacillus isolates signifi-
cantly inhibited maltase (15–76 %, p \ 0.001, n = 3;
Fig. 1a) sucrase (21.3–73.3 %, p \ 0.001, n = 3; Fig. 1b),
lactase (23.6–54.6 %, p \ 0.001, n = 3; Fig. 1c), amylase
(19.4–89.4 %, p \ 0.001, n = 3; Fig. 1d). Acarbose
(positive control) significantly inhibited maltase, sucrose
and amylase by 70–88 %. Acarbose was completely inef-
fective in inhibiting beta-glucosidase (lactase) activity.
Glucosidase inhibition in bacterial reference cultures
All reference culture extracts except L. paracasei signifi-
cantly inhibited maltase (Fig. 2a; 12–75 %, p \ 0.01,
n = 3). L. casei (52 %, p \ 0.001, n = 3) and L.
Eur J Nutr

(A) Maltase (B) Sucrase

100
*** 80
*** *** *********
80 *** *** *** *** ***
*** *** *** *** ***
******
% Inhibition
60 ***
60 *** *** ***

% Inhibition
*** ***
40 ***
40
***
20 *** 20
***

0 0
rb l

1
2
3
4
5
6
7
8
Lb 9
e

Lb 0
Lb 1
Lb 2
Lb 3
Lb 4
15
ca ro

rb l

1
2
3
4
5
6
7
8
Lb 9
e

10
Lb 1
Lb 2
Lb 3
Lb 4
15
os
Lb
Lb
Lb
Lb
Lb
Lb
Lb
Lb
Lb
1
1
1
1
1

ca ro
os
Lb
Lb
Lb
Lb
Lb
Lb
Lb
Lb
Lb

1
1
1
1
A ont

Lb
A ont
-20
C

C
(C) Lactase 100
(D) Amylase
60 *** *** ************
*** ***
***
***
****** 80 ************
*** ***

% Inhibition
***
% Inhibition

40 60 *** ***

***
*** 40
***
20
20 ***

0
0

rb ol

Lb1
Lb2
Lb3
Lb4
Lb5
Lb6
Lb7
L8
Lbb9
e

Lb10
Lb11
Lb12
Lb13
Lb14
15
os
Lb
rb l

1
2
3
4
5
6
7
8
Lb 9
Lb 0
Lb 1
Lb 2
13
Lb 4
15
e

ca r
ca ro
os

A ont
Lb
Lb
Lb
Lb
Lb
Lb
Lb
Lb
Lb
1
1
1

1
Lb
A ont

C
C

(E) Maltase (F) Sucrase

Glucose Concentration (mM)
Glucose Concentration (mM)

10 15
PBS Control
8 Acarbose Acarbose
Lb3 10 Lb3
6 Lb4 Lb4
Lb5 Lb5
4 5 Lb14
Lb14
2

0 0
0 20 40 60 80 100 0 20 40 60 80 100

Time (min) Time (min)

(G) Lactase (H) Amylase

Glucose Concentration (mM)

Glucose Concentration (mM)

5 15
Control Control
4 Acarbose Acarbose
Lb3 Lb3
10
3 Lb4 Lb4
Lb5 Lb5
2
Lb14 5 Lb14
1

0 0
0 20 40 60 80 100 0 20 40 60 80 100
Time (min) Time (min)

Fig. 1 Inhibitory effects of extracts from Lactobacillus isolates on activity, respectively. A vehicle control (PBS) and an established
glucosidase enzymes. Heat-killed sonicated bacterial extracts were glucosidase inhibitor (acarbose) are also shown. e–h Typical rates of
incubated with different carbohydrates (maltose, sucrose, lactose and glucose production occurring in selected incubations: Lb3, Lb4, Lb5
starch) in presence of rat intestinal extract and glucose measured at and Lb14. The majority of isolates had inhibitory potential. Acarbose
different time intervals. a–d Inhibitory effect of 15 extracts from significantly inhibited maltase, sucrose and amylase but was ineffec-
Lactobacillus isolates (Lb1–15) on maltase, sucrase, lactase, amylase tive against lactase

rhamnosus (51 %, p \ 0.001, n = 3) were found to sig- glucosidase (lactase) activity (Fig. 2c; 18–64 %, p\0.001,
nificantly inhibit sucrase activity (Fig. 2b). L. rhamnosus, n = 3). L. casei, L. rhamnosus, L. johnsonii, B. bifidum, L.
L. casei and L. paracasei significantly inhibited beta- acidophilus and L. plantarum extracts significantly

123
 Eur J Nutr

(A) Maltase (B) Sucrase

100 80
*** ***
80 *** *** 60 ***
% Inhibition ***
60

% Inhibition
*** 40
40
***
*** 20
20
*** ** **
0 0
op se

L. ra ii

bi s
L. L us

jo um

m
L. ent ei

L. ant sei
am um

S. E. c dum

im 12
ac ar l

m
L. Ac tro

op e

L. ntu i

L. rac i

S. col m
m

am m
s

L. anta ei

bi s
L. Ac trol
B. osu
pa on

e e

pa oni
lu

su

im 2
rm as

id os
iu

iu
id bo
l

ph i K

rm s

pl as
rh ru

u
p l ca
hi

ph 1
rh ar

fe . ca
n

-20

hi
L. hns

ur
fe . c

ur
ac arb

fid
no

ty i K
n
n

s
Co

f
ty ol

hn
Co
-20

L. L

jo

B.
E.
L.
(C) Lactase (D) Amylase
80 100
***
***
80 *** ***
60
***
% Inhibition

% Inhibition
60
40
40

*** ***
20
20 ***
***
** **
0 0
m
im 12
op e

L. rac i
h m

am m

bi s
s

ei

L. ant ei

S. . co um
L. Ac trol

m
op e

L. rac i

bi s
s

m
L. entu i

L. anta ei
am m

S. . c um

im 12
L. Ac trol
pa ni

B. osu
lu

pa oni
e
id os

B. osu
lu
id os
iu
rm s

pl as
L. ntu

rh aru

iu
rm s

pl as
rh ru
ty li K
L. nso
fe . ca

ph K
hi

fe . ca
ac arb

E id

hi
n

ac arb

E id
ur

ur
s
n

n
Co

ty oli
f

hn
Co
e

f
L. L

L. L
jo

jo
ph

L.
Glucose Concentration (mM)

(E) Maltase (F) Sucrase

Glucose Concentration (mM)

15 15
PBS Control
Acarbose Acarbose
10 L. fermentum L. fermentum
10
L. johnsonii L. johnsonii
L. rhamnosus L. rhamnosus
5 5

0
0 20 40 60 80 100 0
0 20 40 60 80 100
Time (min)
Time (min)

(G) Lactase (H) Amylase

Glucose Concentration (mM)

Glucose Concentration (mM)

5 15
Control Control
4 Acarbose Acarbose
L. fermentum 10 L. johnsonii
3 L. johnsonii L. rhamnosus
L. rhamnosus
2
5
1

0 0
0 20 40 60 80 100
0 20 40 60 80 100
Time (min) Time (min)

Fig. 2 Inhibitory effects of extracts from bacterial reference cultures lactase, amylase activity, respectively. A vehicle control (PBS) and an
on glucosidase enzymes. Heat-killed sonicated bacterial extracts were established glucosidase inhibitor (acarbose) are also shown. e–h Dem-
incubated with different sugars (maltose, sucrose, lactose and starch) onstrate the rate of glucose production occurring in selected
in presence of rat intestinal enzyme and glucose measured at different incubations: L. fermentum, L. johnsonii and L. rhamnosus. L. casei
time intervals. a–d Inhibitory effect of seven Lactobacillus reference and L. rhamnosus most significantly inhibited all four enzymes:
cultures, one Gram positive control (B. bifidum) and two Gram neg- maltase, sucrose, lactase and amylase. E. coli and S. typhimurium had
ative controls (E. coli and S. typhimurium) on maltase, sucrase, no inhibitory activities

123
Eur J Nutr

(A) (B)

Glucose Concentration (mM)

100 15
*** PBS
80 Acarbose
*** 10
500mg/ml
60 333mg/ml
% Inhibition

40
*** 285mg/ml
250mg/ml
*** 5
20 200mg/ml
***
0
0
S

se

l
0 20 40 60 80
g/m

g/m

g/m

g/m

g/m
PB

bo

-20
ar

0m

3m

5m

0m

0m
Time (min)
Ac

50

33

28

25

20
100
(C) 8
(D)

Glucose Concentration (mM)

*** Control
80 *** Acarbose
*** 6
*** Lb3 SE
% Inhibition

60 Lb3 SN
4 Lb4 SE
40 ***
Lb4 SN
***
2 Lb5 SE
20 ***
Lb5 SN
0 0
0 20 40 60 80 100
ol

SE

SN

SE

SN

SE

SN
os
tr

rb

Time (min)
3

5
3

5
on

Lb

Lb

Lb
Lb

Lb

Lb
ca
C

Fig. 3 Inhibitory activity is concentration-dependent and is found in and their supernatants (SN) were compared for maltase inhibition
cell-free supernatant. Different concentrations of heat-killed sonicated (c and d). For Lb3, 4 and 5 glucosidases significant inhibitory activity
bacterial extract were tested for concentration-dependent inhibition of was observed both in the SE and in the SN
maltase activity (a and b). Heat-killed sonicated bacteria extracts (SE)

inhibited amylase (Fig. 2d; 6–74 %, p \ 0.01, n = 3). The presence of both SE and SN of selected Lactobacillus
Gram negative pathogens Salmonella typhimurium and isolates.
E. coli either possessed very weak inhibitory activity or no
inhibitory activity against maltase, sucrose, amylase and In vivo effects of oral Lactobacillus to a carbohydrate
lactase. challenge

Inhibitory activity is concentration-dependent Oral administration of L. rhamnosus (Ref7) extract did not
and not solely confined to cell pellets significantly lower glycaemic excursions in rats receiving a
maltose challenge (Fig. 4a, b). However, when rats were
Glucosidase inhibitory activity from bacterial extracts challenged with starch, L. rhamnosus extract significantly
appeared to be highly concentration-dependent. Different lowered responses compared with the saline vehicle
concentrations of SE (500, 350, 300 and 250 mg) diluted in (Fig. 4c, d). Analysis by area under the curve (DAUC0-
1 ml of nuclease-free water ranged in activity from *70 to 105) revealed that L. rhamnosus extract reduced starch
*20 %. There was no observable inhibition at the lowest challenge responses by 22 % (p \ 0.05; Fig. 4d). In both
concentration of 200 mg/ml (Fig. 3a, b). these studies, acarbose was effective in reducing glycaemic
To assess whether inhibitory activity occurred in extract excursions (33–49 %; p \ 0.01).
SNs, three isolates (Lb3, Lb4 and Lb5) were randomly
selected, and SE were prepared (as described earlier) fol-
lowed by centrifugation (12,000g; 10 min; 5 °C). The cell- Discussion
free SN was used in enzyme assay using maltose as a
carbohydrate source (Fig. 3c, d). Enzyme assay revealed Historically bacteria have been the source of many human
that although glucosidase inhibitory activity was more therapeutics and potential exists for their use in developing
concentrated within cells (SE), it was quite labile with novel diabetes therapies or biotherapeutics. There is an
significant amounts of activity also occurring in SN ongoing interest in the application of gut microbiota as
(18–38 %). Figure 3c shows % inhibition of maltase in ‘probiotics’ for alleviating disease, and this stems from

123
 Eur J Nutr

Fig. 4 Effects of oral L. 7 (A) 8 (C)

rhamnosus extract on glucose
6 7
excursions in rats after
challenge with maltose or 6
5

Blood glucose

Blood glucose
(mmol/L)

(mmol/L)
starch. Rats were orally gavaged 5
4
with either (a) maltose (2 g/kg) 4
or (c) starch (2 g/kg), in 3
3
combination with saline 2 Maltose + saline 2 Starch + saline
(0.9 %), acarbose (80 mg/kg) or Maltose + acarbose Starch + acarbose
L. rhamnosus extract (1 g/kg) 1 Maltose + L.rhamnosus 1 Starch + L.rhamnosus
and glucose responses were 0 0
monitored at 0, 15, 30, 60 and 0 15 30 45 60 75 90 105 0 15 30 45 60 75 90 105
105 min. Incremental areas Time (min) Time (min)
under plasma glucose curves
(DAUC0-105) for (b) maltose or Starch + saline
Maltose + saline Starch + acarbose
(d) starch challenges were Maltose + acarbose (D)
calculated by the trapezoidal 250
(B) Maltose + L.rhamnosus 200
Starch + L.rhamnosus

rule with baseline subtraction. ns

*

AUC (mmol/l.min)
AUC (mmol/l.min)

Results are mean ± SEM 200

150

Blood glucose
Blood glucose

(n = 4); *p \ 0.05; **p \ 0.01,

compared with saline control; 150
* 100
**
ns—not significant
100

50
50

0 0

their rich diversity, ‘GRAS’ status, use as biocatalysts and
also the wide range of secondary metabolites which they
produce. The application of probiotic species within the
context of diabetes remains largely unexplored, and this
study set out to establish whether LAB have inhibitory
effects on intestinal glucosidases.
Intestinal alpha-glucosidase inhibition is an established
strategy for controlling post-prandial hyperglycaemia [3].
Although clinical inhibitors of alpha-glucosidase are
widely available, alternative alpha-glucosidase inhibitors
are continually sought in the hope of minimising side
effects and reducing drug costs. There is also the potential
for functional foods containing inhibitory activity, and
many natural, plant and food-based sources have been
screened [12–17]. Bacteria have also been screened for
inhibitory activity; in one case, this led to the discovery of
acarbose in Actinoplanes sp. [18]. In another case, micro-
bial strains were screened for their ability to produce hy-
droxycitric acid. Hydroxycitric acid was previously
identified as an alpha-glucosidase inhibitor in tropical
plants, Garcinia cambogia and Hibiscus subdariffa [17],
and two Streptomyces spp. were found to produce it [17].
Alpha-glucosidase inhibition has also been documented in
a Streptomyces strain (PW638) from which the inhibitor,
acarviostatin, has been isolated [29]. Another inhibitor,
1-deoxynojirimycin (DNJ), has been recently identified and
isolated from Bacillus subtilis B4 which has DNJ produc-
tion in presence of sorbitol [30]. Chaudet et al. (2012) [31]
reported the presence of two alpha-glucosidase inhibitors
(BT_0339 and BT_3299) in Bacteroides thetaiotaomicron
that possess demonstrated activity against maltase and iso-
maltase but these were less active against lactase (a beta-
glucosidase) and sucrase.
Despite these previous findings, the fundamental issue is
that none of the bacterial strains mentioned above are LAB
and would not under normal circumstances come into
contact with alpha- or beta-glucosidases of the human
intestine. There has been one unconfirmed report that a
skimmed milk-LAB hydrolysate has alpha-glucosidase
inhibitory activity [19]. The current study is the first to
actually demonstrate that LAB produce significant alpha-
glucosidase inhibitory activity. We focused on the enzymes
digesting the most abundant food carbohydrates (maltose,
starch, lactose and sucrose), and specifically on Lactoba-
cilli isolated from the human intestine. Heat-killed SEs of
LAB were used to ensure that any observed effects were
not the result of bacterial metabolism or fermentation.
The current investigation presents definitive evidence
that certain strains of Lactobacillus produce significant
inhibitory potential against a range of alpha-glucosidases
and one beta-glucosidase. It is still unclear whether
the lactobacilli play any physiological role in modulating
these enzymes, however, the potential importance of the
findings is underlined by the following observations.
Firstly, the Lactobacillus strains tested here were isolated
from human infant faecal samples, and their subsequent
identification clearly assigns them as common and rela-
tively abundant gut inhabitants. Lactobacilli are present in

123
Eur J Nutr
the small intestine where carbohydrate digestion takes
place but cell densities (104–108cfu/ml) are much lower
than those in the large intestine (1012-14 cfu/ml) [32, 33].
Secondly, at the concentrations tested (up to 500 mg/ml
wet weight), some strains had comparable in vitro activities
to acarbose (30 mg/ml). Further purification and identifi-
cation of the bioactive(s) are required to reveal their true
potency. Thirdly, one selected Lactobacillus sp. (L.
rhamnosus) was able to curb glucose excursions in vivo
following a starch challenge (although admittedly it was
unable to achieve this with a maltose load. This may be due
to the fact that maltose hydrolysis is more rapid and
therefore more difficult to curb in a physiological situa-
tion.) The glycaemic effect was modest but it clearly
impacted on blood glucose levels, and even small
improvements in blood glucose in healthy individuals can
improve longevity and reduce risk of disease. Fourthly, one
strain of Bifidobacterium (serving as a Gram positive
control) did not significantly inhibit glucosidases, nor did
the Gram negative pathogens E. coli or S. typhimurium.
Other LAB genera should be screened to see whether
activity is confined to Lactobacillus. Fifthly, several Lac-
tobacillus strains had a broader spectrum of activity than
even acarbose; they effectively inhibited lactase activity in
addition to maltase, sucrase and amylase activity. Acarbose
is known to be active against several alpha-glucosidase
enzymes, including brush border maltase, sucrase, gluco-
amylase, dextrinase and pancreatic amylase, but a potential
weakness is that it has minimal activity against isomaltase
and zero inhibition for beta-glucosidases such as lactase
[3]. Lastly, the inhibitory activity observed was clearly
concentration-dependent and was also present in cell-free
SNs, indicating that cytoplasmic contents or products of
bacterial metabolism may be responsible for this activity.
The majority of faecal isolates were identified as strains
of Lactobacillus plantarum and all had potent inhibitory
activities, including one identified as L. plantarum subsp.
argentorotensis. DNA sequencing of the isolates indicated
that all 11 L. plantarum strains were different, with no
duplicates recorded. It was clear that the spectrum of
inhibitory activity amongst the L. plantarum isolates dif-
fered substantially. For example, some were virtually
devoid of activity (i.e. Lb14), some had broad activity with
no lactase inhibitory activity (i.e. Lb1 and 2), or with
modest lactase inhibitory activity (i.e. Lb3 and 4), or with
high lactase inhibitory activity (i.e. Lb5–6 and 8–11). This
demonstrated that even within a single species there may be
subspecies with substantially different activities. Amongst
the reference strains, L. casei (Ref2) and L. rhamnosus
(Ref7) were the most potent inhibitors (50–70 %) and had
broad spectrum activities. Both the L. acidophilus isolate
(Lb15) and L. acidophilus reference stain (Ref1) had similar
profiles both inhibiting maltase and amylase but failing to
inhibit sucrase and lactase. In a previously published study,
three days of oral L. acidophilus administrations were
shown to have no effect on alpha-glucosidase activity in
fasted Wistar rats [34].
It is also noteworthy that in vitro assay incubations of
Lb14 and L. johnsonii with maltose and sucrose produced
more glucose than the PBS buffered control which suggests
that these bacteria contain maltase and sucrase enzymes. It
is well established that many bacteria possess intrinsic
glucosidase enzymes [35]. For example, B. adolescentis
has broad activity hydrolysing glycosidic bonds of 10
different carbohydrates [36]. Some species contain ther-
mostable alpha-glucosidase which exclusively hydrolyse
alpha-1, 4-glycosidic linkages [37]. The reasons why some
bacterial strains produce compounds with inhibitory
activity are not fully understood, however, it could be a
means of out-competing other bacterial species. There
remains the possibility that the true inhibitory potential of
some of the Lactobacillus strains tested here is masked by
the presence of bacterial glucosidase enzymes. Only
through activity-guided purification will the actual effec-
tiveness of inhibitors be known.
In conclusion, our results show definitively that Lacto-
bacillus strains from human infant faecal samples have
considerable glucosidase inhibitory activity and in vivo this
can reduce blood glucose responses to a carbohydrate
challenge. Their potential use as a dietary supplement,
medicinal food or biotherapeutic for diabetes is still
uncertain, but LAB such as Lactobacillus sp. offer poten-
tial advantages over drug therapies, perhaps having a
broader spectrum activity and fewer unpleasant side
effects. A better understanding of the source of the bio-
activity is needed and chronic studies using live cultures in
diabetic patients or animal models should be undertaken
with candidate Lactobacillus strains identified by this
study.
Acknowledgments Harsh Panwar was funded by a PhD studentship
(INCN-2011-43) from Commonwealth Scholarship Commission, UK
to conduct studies at Queen’s University Belfast (UK).
Conflict of interest On behalf of all authors, the corresponding
author states that there is no conflict of interest.
References
1. Sales PM, Souza PM, Simeoni LA, Magalhaes PO, Silveira D
(2012) a-Amylase inhibitors: a review of raw material and isolated
compounds from plant source. J Pharm Pharm Sci 15(1):141–183
2. Kazeem MI, Dansu TV, Adeola SA (2013) Inhibitory effect of
Azadirachta indica A. Juss leaf extract on the activities of a-
amylase and a-glucosidase. Pak J Biol Sci 16(21):1358–1362
3. Lebovitz HE (1997) Alpha-glucosidase inhibitors. Endocrinol
Metab Clin 26(3):539–551
123

4. Derosa G, Maffioli P (2012) a-Glucosidase inhibitors and their
use in clinical practice. Arch Med Sci 8(5):899–906
5. Creutzfeldt W (1999) Effects of the a-glucosidase inhibitor
acarbose on the development of long term complications in dia-
betic animals: pathophysiological and therapeutic implications.
Diabetes Metab Res Rev 15:289–296
6. Standl E, Schnell O (2012) Alpha-glucosidase inhibitors 2012—
cardio vascular considerations and trial evaluation. Diabetes Vasc
Dis Res 9(3):163–169
7. Chang J, Block TM, Guo JT (2013) Antiviral therapies targeting
host ER alpha-glucosidases: current status and future directions.
Antivir Res 99(3):251–260
8. Rosak C, Mertes G (2012) Critical evaluation of the role
of acarbose in the treatment of diabetes: patient considerations.
Diabetes Metab Syndr Obes 5:357–367
9. Derosa G, Maffioli P (2012) Efficacy and safety profile evalua-
tion of acarbose alone and in association with other antidiabetic
drugs: a systematic review. Clin Ther 34(6):1221–1236
10. Tschope D, Hanefeld M, Meier JJ, Gitt AK, Halle M, Bramlage
P, Schumm-Draeger PM (2013) The role of co-morbidity in the
selection of antidiabetic pharmacotherapy in type-2 diabetes.
Cardiovasc Diabetol 12:62
11. Chiasson JL, Josse RG, Gomis R, Hanefeld M, Karasik A, Laakso
M (2002) Acarbose for prevention of type 2 diabetes mellitus: the
STOP-NIDDM randomised trial. Lancet 359:2072–2077
12. Al-Zuhair S, Dowaidar A, Kamal H (2010) Inhibitory effect of
dates-extract on a-amylase and b-glucosidase enzymes relevant
to non-insulin dependent diabetes mellitus. J Biochem Tech
2(2):158–160
13. Tundis R, Loizzo MR, Menichini F (2010) Natural products as
alpha-amylase and alpha-glucosidase inhibitors and their hypo-
glycaemic potential in the treatment of diabetes: an update. Mini
Rev Med Chem 10(4):315–331
14. Mohamed ELH, Siddiqui MJA, Ang LF, Sadikun A, Chan SH,
Tan SC, Asmawi MZ, Yam MF (2012) Potent a-glucosidase and
a-amylase inhibitory activities of standardized 50% ethanolic
extracts and sinensetin from Orthosiphonstamineus Benth as anti-
diabetic mechanism. BMC Complement Altern Med 12:176
15. Nair SS, Kavrekar V, Mishra A (2013) In vitro studies on alpha
amylase and alpha glucosidase inhibitory activities of selected
plant extracts. Eur J Exp Biol 3(1):128–132
16. Katekhaye SD, Nagmoti DM (2013) a-Glucosidase and a-amy-
lase inhibitory activities of Pithecellobium dulce bark and leaves.
Phytopharmacology 4(1):123–130
17. Yamada T, Hida H, Yamada Y (2007) Chemistry, physiological
properties, and microbial production of hydroxycitric acid. Appl
Microbiol Biotechnol 75(5):977–982
18. Schwientek P, Szczepanowski R, Ruckert C, Kalinowski J, Klein
A, Selber K, Wehmeier UF, Stoye J, Puhler A (2012) The
complete genome sequence of theacarbose producer Actinoplanes
sp. SE50/110. BMC Genomics 13:112
19. Ramchandran L, Shah NP (2008) Proteolytic profiles and
angiotensin-1 converting enzyme and alpha-glucosidase inhibi-
tory activities of selected lactic acid bacteria. J Food Sci
73(2):M75–M81
20. Panwar H, Rashmi HM, Batish VK, Grover S (2013) Probiotics
as potential biotherapeutics in the management of type 2 diabe-
tes—prospects and perspectives. Diabetes Metab Res Rev
29:103–112
123
Eur J Nutr
21. Ritchie ML, Romanuk TN (2012) A meta-analysis of probiotic
efficacy for gastrointestinal diseases. PLoS One 7(4):e34938
22. Chapman TM, Plosker GL, Figgitt DP (2007) Spotlight on
VSL#3 probiotic mixture in chronic inflammatory bowel dis-
eases. Bio Drugs 21(1):61–63
23. Anonymous or Sigma Tau Pharmaceuticals Inc. (2011) VSL#3
double strength product information sheet. http://www.vsl3.com/
pdf/VSL3_DS_PL.pdf
24. Dubernet S, Desmasures N, Gueguen M (2002) A PCR based
method for identification of lactobacilli at the genus level. FEMS
Microbiol Lett 214(2):271–275
25. Turner S, Pryer KM, Miao VPW, Palmer JD (1999) Investigating
deep phylogenetic relationships among cyanobacteria and plas-
tids by small subunit rRNA sequence analysis. J Eukaryot
Microbiol 46:327–338
26. Rogall TJ, Wolters TF, Bottger EC (1990) Towards a phylogeny
and definition of species at the molecular level within the genus
Mycobacterium. Int J Syst Bacteriol 40:323–330
27. Naser SM, Dawyndt P, Hoste B, Gevers D, Vandemeulebroecke
K, Cleenwerck I, Vancanneyt M, Swings J (2007) Identification
of lactobacilli by pheS and rpoA gene sequence analyses. Int J
Syst Evol Microbiol 57:2777–2789
28. Burington RS (1973) Handbook of mathematical tables and for-
mulas. McGraw-Hill, New York
29. Meng P, Xie C, Geng P, Qi X, Zheng F, Bai F (2013) Inhibitory
effect of components from Streptomyces species on alpha-glu-
cosidase and alpha-amylase of different origin. Prikl Biokhim
Mikrobiol 49(2):181–189
30. Onose S, Ikeda R, Nakagawa K, Kimura T, Yamagishi K, Hig-
uchi O, Miyazawa T (2013) Production of the a-glycosidase
inhibitor 1-deoxynojirimycin from Bacillus species. Food Chem
138(1):516–523
31. Chaudet MM, Allen JL, Rose DR (2012) Expression and purifi-
cation of two Family GH31 a-glucosidases from Bacteroides
thetaiotaomicron. Protein Expr Purif 86(2):135–141
32. Ley RE, Peterson DA, Gordon JI (2006) Ecological and evolu-
tionary forces shaping microbial diversity in the human intestine.
Cell 124:837–848
33. Walter J, Ley R (2011) The human gut microbiome: ecology and
recent evolutionary changes. Annu Rev Microbiol 65:411–429
34. Mountzouris KC, Kotzampassi K, Tsirtsikos P, Kapoutzis K,
Fegeros K (2009) Effects of Lactobacillus acidophilus on gut
microflora metabolic biomarkers in fed and fasted rats. Clin Nutr
28(3):318–324
35. Perez-Martı́n F, Sesena S, Izquierdo PM, Martı́n R, Palop ML
(2012) Screening for glycosidase activities of lactic acid bacteria
as a biotechnological tool in oenology. World J Microbiol Bio-
technol 28(4):1423–1432
36. Holt SM, Teresi JM, Cote GL (2008) Influence of alternansucr-
ase-derived oligosaccharides and other carbohydrates on alpha-
galactosidase and alpha-glucosidase activity in Bifidobacterium
adolescentis. Lett Appl Microbiol 46(1):73–79
37. Park JE, Park SH, Woo JY, Hwang HS, Cha J, Lee H (2013)
Enzymatic properties of a thermostable a-glucosidase from aci-
dothermophilic crenarchaeon Sulfolobus tokodaii strain 7.
J Microbiol Biotechnol 23(1):56–63