Bioscience, Biotechnology, and Biochemistry

ISSN: 0916-8451 (Print) 1347-6947 (Online) Journal homepage: https://www.tandfonline.com/loi/tbbb20

Coffee and Caffeine Improve Insulin Sensitivity

and Glucose Tolerance in C57BL/6J Mice Fed a
High-Fat Diet

Yuji MATSUDA, Misato KOBAYASHI, Rie YAMAUCHI, Makoto OJIKA, Masanori

HIRAMITSU, Takashi INOUE, Takao KATAGIRI, Atsushi MURAI & Fumihiko
HORIO

To cite this article: Yuji MATSUDA, Misato KOBAYASHI, Rie YAMAUCHI, Makoto OJIKA,
Masanori HIRAMITSU, Takashi INOUE, Takao KATAGIRI, Atsushi MURAI & Fumihiko HORIO
(2011) Coffee and Caffeine Improve Insulin Sensitivity and Glucose Tolerance in C57BL/6J
Mice Fed a High-Fat Diet, Bioscience, Biotechnology, and Biochemistry, 75:12, 2309-2315, DOI:
10.1271/bbb.110452

To link to this article: https://doi.org/10.1271/bbb.110452

Published online: 22 May 2014.

Submit your article to this journal

Article views: 777

View related articles

Citing articles: 34 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=tbbb20
Biosci. Biotechnol. Biochem., 75 (12), 2309–2315, 2011

Coffee and Caffeine Improve Insulin Sensitivity and Glucose Tolerance

in C57BL/6J Mice Fed a High-Fat Diet
Yuji M ATSUDA,1 Misato K OBAYASHI,1 Rie Y AMAUCHI,1 Makoto O JIKA,1 Masanori H IRAMITSU,2
Takashi I NOUE,2 Takao K ATAGIRI,2 Atsushi M URAI,1 and Fumihiko H ORIO1; y
1
Department of Applied Molecular Bioscience, Graduate School of Bioagricultural Sciences, Nagoya University,
Nagoya 464-8601, Japan
2
Product Research and Development Department, Pokka Corporation, Aichi 481-8515, Japan

Received June 10, 2011; Accepted August 4, 2011; Online Publication, December 7, 2011
[doi:10.1271/bbb.110452]

We have previously demonstrated that coffee and
caffeine ameliorated hyperglycemia in spontaneously
diabetic KK-Ay mice. This present study evaluates the
antidiabetic effects of coffee and caffeine on high-
fat-diet-induced impaired glucose tolerance in C57BL/
6J mice. C57BL/6J mice fed a high-fat diet were given
regular drinking water (control group), or a 2.5-fold-
diluted coffee or caffeine solution (200 mg/L) for 17
weeks. The ingestion of coffee or caffeine improved
glucose tolerance, insulin sensitivity, and hyperinsuline-
mia when compared with mice in the control group.
The adipose tissue mRNA levels of inflammatory
adipocytokines (MCP-1 and IL-6) and the liver mRNA
levels of genes related to fatty acid synthesis were lower
in the coffee and caffeine groups than those in the
control group. These results suggest that coffee and
caffeine exerted an ameliorative effect on high-fat-diet-
induced impaired glucose tolerance by improving
insulin sensitivity. This effect might be attributable in
part to the reduction of inflammatory adipocytokine
expression.
Key words: coffee; caffeine; diet-induced impaired
glucose tolerance; antidiabetic effect; insu-
lin resistance
Type 2 diabetes is a chronic disease associated with
high rates of morbidity and premature mortality.
Although the pathogenesis of type 2 diabetes is known
to be complex, previous studies have shown that its
process involves both polygenetic and environmental
factors.1–3) The importance of the primary prevention of
type 2 diabetes has been increasingly recognized in
recent years, and this has stimulated interest in the role
of dietary factors in the etiology of type 2 diabetes.
Coffee is one of the world’s most popular beverages.
The numerous beneficial health effects of coffee
consumption have recently received significant scientific
attention; epidemiological studies have suggested that
regularly drinking coffee prevents chronic diseases, and
especially metabolic disorders such as type 2 diabe-
tes.4–7) The potential to suppress hyperglycemia through
a popular beverage such as coffee could be exploited as
a comfortable strategy to prevent type 2 diabetes. An
experimental study using a diabetic animal model would
therefore be effective for evaluating the antidiabetic
potency of coffee. However, few such studies have been
performed, and the mechanism underlying the antidia-
betic effect of coffee and its main antidiabetic com-
pounds are not fully understood.
We have previously investigated the preventive effect
of coffee on the development of hyperglycemia in KK-
Ay mice which spontaneously develop type 2 diabetes.8)
The KK-Ay mouse is a polygenic model for type 2
diabetes; it exhibits obesity due to hyperphagia and
develops hyperglycemia at 6 weeks of age. Hyper-
insulinemia, hypertriglyceridemia, and fatty liver are
also apparent. Our previous results have shown that
coffee and caffeine had antidiabetic effects, and sug-
gested that coffee and caffeine could ameliorate hyper-
glycemia by improving the insulin resistance in KK-Ay
mice. Coffee or caffeine ingestion also reduced the
inflammatory adipocytokine expression in adipose tissue
and decreased fat accumulation in the liver.8) It was
expected from these results that coffee and caffeine
would exhibit antidiabetic effects in other diabetic
animal models with insulin resistance.
Recent epidemiological studies have indicated that
the consumption of a high-calorie diet and the ensuing
obesity were two of the principal causes of type 2
diabetes.3) An obesity-induced experimental model of
type 2 diabetes induced by a high-calorie diet would be
useful for evaluating the antidiabetic action of coffee
and caffeine. C57BL/6 mice fed a high-fat diet (HFD)
have been shown to be a representative model of
obesity-induced diabetes.9,10) HFD causes insulin resist-
ance in the skeletal muscle, liver, and adipose tissue, and
the HFD-induced increase in the serum concentration
of free fatty acids (FFAs) provokes the impairment of
insulin signaling in several tissues.11)

y
To whom correspondence should be addressed. Fax: +81-52-789-4077; E-mail: horiof@agr.nagoya-u.ac.jp
Abbreviations: HFD, high-fat diet; FFA, free fatty acid; 200CA, 200 mg of caffeine/L; GTT, glucose tolerance test; ITT, insulin tolerance test;
TG, triglycerides; TC, total cholesterol; WAT, white adipose tissues; BAT, brown adipose tissue; PL, phospholipids; AUC, area under the curve;
SREBP-1, sterol regulatory element binding protein-1; FAS, fatty acid synthase; ACC, acetyl-CoA carboxylase; PPAR-
, peroxisome proliferator
activator receptor-alpha; CPT-1, carnitine palmitoyltransferase-1; IRS-1, insulin receptor substrate-1; JNK, c-jun N-terminal kinase; IKK, inhibitor
of NF-B kinase ; ERK, extracellular signal-related kinase; SOCS, suppressor of cytokine signaling; JAK-STAT, Janus kinase-signal transduction
and activator of transcription; DAG, diacylglycerol; AMPK, AMP-activated protein kinase
2310 Y. M ATSUDA et al.
We examined in this study whether or not the
ingestion of coffee and caffeine would prevent HFD-
induced impaired glucose tolerance in C57BL/6J mice.
The results show that the ingestion of coffee or caffeine
improved insulin sensitivity and glucose tolerance.
These effects suggest that the reduced mRNA levels of
inflammatory adipocytokines in adipose tissue and the
reduced mRNA levels of genes related to fatty acid
synthesis in the liver both contributed to the antidiabetic
effect of coffee or caffeine.
Materials and Methods
Chemicals. Regular canned black coffee was presented by Pokka
Corporation (Aichi, Japan) and used after 2.5-fold dilution with water.
The respective concentrations of caffeine, chlorogenic acid, and
trigonelline in the diluted coffee used in this experiment were 250, 120,
and 180 mg/L. Caffeine was purchased from Wako Pure Chemical
Industries (Osaka, Japan).
Animals. Seven-week-old male C57BL/6J mice (SLC Japan,
Shizuoka, Japan) were used for this experiment. They were maintained
at a controlled temperature of 23
2  C and 55
5% humidity with a
12-h light/dark cycle. The mice were allowed free access to water and
a standard CE-2 laboratory diet (Clea Japan, Tokyo, Japan) for 5 d
before the experiment was begun.
The acclimatized eight-week-old mice were assigned to three
groups and respectively given water (control group, 5 mice), diluted
black coffee (black coffee/water = 1:1.5; coffee group, 6 mice), or a
caffeine solution (200 mg caffeine/L; 200CA group, 6 mice) for
drinking. The coffee and caffeine solutions were prepared every second
day, the undiluted coffee being stored at 20  C until being diluted.
The mice were allowed free access to drinking water and a Quick Fat
high-fat diet (Clea) for 17 weeks. The composition of the high-fat diet
was as follows: protein, 242 g/kg; fat, 136 g/kg (beef tallow was the
primary source of fat); non-nitrogenous substances, 465 g/kg; crude
fiber, 30 g/kg, crude ash, 52 g/kg; energy 4055 kcal/kg; and sufficient
minerals and vitamins to maintain the health of the mice.
Blood samples were collected from the tail vein 2, 5, 8, 11, and 17
weeks after the start of the experiment for measuring the serum
glucose concentration. The blood samples were collected at 10:00
a.m., 1 h after the food had been removed. A glucose tolerance test
(GTT) and insulin tolerance test (ITT) were performed during the
experiment. The mice were killed by decapitation at the end of the
experiment (between 10:00 a.m. and 12:00 p.m.), and the blood was
collected. These blood samples were used to measure the serum
insulin, triglyceride (TG), and total cholesterol (TC) concentrations.
The liver, white adipose tissues (WAT; subcutaneous, epididymal,
mesenteric, and retroperitoneal fat tissues), and interscapular brown
adipose tissue (BAT) were removed and weighed. Subcutaneous fat
tissue is defined as the fat pads below the root of the forefoot on
the side of the body. The tissue samples were then immediately
frozen in liquid nitrogen and kept at 80  C until needed. The
collected blood was kept at room temperature for 15 min for
coagulation. The serum was then obtained from the coagulated blood
by centrifugation at 2;430  g for 10 min at 4  C and then kept at
30  C prior to its use.
The animal care and experimental procedures were approved by
the Animal Research Committee of Nagoya University and were
conducted according to the Regulations for Animal Experiments at
Nagoya University.
Analysis of the metabolic parameters. Serum glucose was measured
with a Glucose C-test assay kit (Wako Pure Chemical Industries)
by the glucose oxidase method. The serum triglyceride and cholesterol
concentrations were respectively measured with a Triglyceride-E Test
assay kit (Wako Pure Chemical Industries) by the glycerol-3-phosphate
oxidase method, and with a Cholesterol-E test kit (Wako Pure
Chemical Industries) by the cholesterol oxidase method. A commer-
cially available ELISA kit was used to determine the serum
concentration of insulin (Morinaga Seikagaku, Kanagawa, Japan).
Glucose tolerance test (GTT). GTT was performed 3, 9, and 16
weeks after the start of the experiment. After 14 h of fasting, blood
samples were collected from the tail vein (the fasting blood sample was
the 0 min sample in GTT). A glucose solution (Otsuka Seiyaku, Tokyo,
Japan) was then intraperitoneally injected (2 g/kg of BW). Blood
samples were again collected from the tail vein 30, 60, and 120 min
after the glucose injection, and the serum glucose concentration was
measured. The area under the curve (AUC; mg.min/dL) for the
glucose concentration was calculated according to the trapezoidal rule
from the glucose measurements at fasting (0 min), 30, 60, and 120 min.
Insulin tolerance test (ITT). ITT was performed 15 weeks after the
start of the experiment. After a 14-h fast, blood samples were collected
from the tail vein (the fasting blood sample was the 0 min sample in
ITT). A human insulin solution (Humulin, 0.2 units/kg of BW; Eli
Lilly, Kobe, Japan) was then intraperitoneally injected. Blood samples
were collected from the tail vein 30, 60, and 120 min after the
injection, and the serum glucose concentration was measured.
Hepatic lipid analysis. Frozen livers were homogenized with
chloroform-methanol (2:1), and the liver lipids were extracted into an
organic solvent. A portion of this extract was dried, and the hepatic
contents of TG, TC, and phospholipids (PL) were respectively
measured by a Triglyceride E-Test, Cholesterol E-Test, and phospho-
lipid C-Test (Wako Pure Chemical Industries). This extract was also
used to measure the total lipid content according to the method of
Folch, Lees, and Sloane Stanley.12)
RNA preparation and expression level analysis. Total RNA was
extracted from frozen tissues by using the TRIzol reagent (Invitrogen,
Carlsbad, CA, USA), before being treated with DNase by using a
Turbo DNA-free kit (Ambion, Austin, TX, USA). cDNA was
synthesized by using a high-capacity cDNA Reverse Transcription
kit (Applied Biosystems, Foster City, CA, USA). Gene expression was
quantified by real-time PCR, using an ABI 7300 real-time PCR system
with the PCR Master Mix reagent and probes (TaqMan gene
expression assays; Applied Biosystems). The level of each mRNA
was normalized to that of corresponding 18S rRNA.
Statistical analysis. Each result is expressed as the mean
SEM.
Phenotypic data were statistically analyzed by one-way ANOVA and
the Tukey-Kramer multiple-comparison test. Values of p < 0:05 were
considered statistically significant (StatView; SAS Institute, Cary,
NC, USA).
Results
Effects of coffee or caffeine ingestion on blood
glucose, glucose tolerance, and insulin resistance in
HFD-fed C57BL/6J mice
Eight-week-old C57BL/6J mice ingested the coffee
solution, caffeine solution, or water as their drinking
liquid for 17 weeks. The body weight, food intake, water
intake, and tissue weight are shown in Table 1. The final
body weight did not differ among the control, coffee,
and 200CA groups. The food intake and water intake
(measured after 4, 10, and 16 weeks) were also no
different among these three groups. Likewise, the liver
weights of the coffee and 200CA groups did not differ
from that of the control group.
C57BL/6J mice were fed in this study with a mild
high-fat diet containing 13.6% of fat; the non-fasting
blood glucose concentration therefore did not increase in
the control group during the course of the experiment
and did not differ among the three groups at any time
during the experiment (Fig. 1).
Figure 2 shows that the glucose tolerance in the
control group was gradually impaired by HFD. The
blood glucose concentrations 9 weeks later after 60 and
 Coffee and Caffeine Suppress Diet-Induced Glucose Intolerance in Mice 2311

120 min, and AUC during GTT in the 200 CA group and 200 CA groups than in the control group, and
were significantly lower than the respective values in the the blood glucose concentrations 120 min after the
control group (Fig. 2B and D). The blood glucose insulin injection were significantly lower in the coffee
concentrations 16 weeks later after 60 and 120 min, and 200 CA groups than in the control group (Fig. 3A).
and AUC during GTT were significantly lower in the The non-fasting serum insulin concentrations at the
coffee and 200 CA groups than in the control group end of the experiment were significantly lower in the
(Fig. 2C and D). coffee and 200CA groups than in the control group
The effects of coffee and caffeine ingestion on the (Fig. 3B).
insulin sensitivity were confirmed by performing ITT 15 Table 1 shows the serum TG and TC concentrations,
weeks after the start of the experiment. The glucose- liver total lipid content, liver TG content, liver TC
lowering effect of insulin was greater in the coffee content, and liver PL content. The serum TG concen-
tration was lower in the 200CA group than in the control
group, although the serum TC concentration did not
250 differ among the three groups. The mean values for the
liver total lipids and TG contents in the coffee and
200CA groups tended to be lower than those in the
200
control group. The liver TC and PL contents also did not
differ among the three groups.
( mg/dL )

150
Changes to the adipocytokine mRNA levels in the
100 epididymal fat tissue due to coffee and caffeine ingestion
Control We investigated the mechanism by which coffee and
Coffee caffeine improved insulin sensitivity by measuring the
50
200CA epididymal fat tissue mRNA levels of adipocytokines
that are known to influence insulin sensitivity. The
0 results are shown in Fig. 4A. MCP-1, TNF-
and IL-6
0 2 4 6 8 10 12 14 16 18 are adipocytokines that cause insulin resistance in
(weeks) peripheral tissues. The mRNA level of MCP-1 in the
200CA group was significantly lower than that in the
Fig. 1. Time-Course Characteristics of the Non-Fasting Blood
Glucose Concentration in C57BL/6J Mice Fed with HFD for the control group. The MCP-1 mRNA level in the coffee
Control (n ¼ 5), Coffee (n ¼ 6), and 200 CA (n ¼ 6) Groups. group tended to be lower than that in the control group,
Data are expressed as the mean
SEM. although not significantly (p ¼ 0:08). The mRNA level

A B C
600 b b
ab
a a
500 b
b a
400
( mg/dL)

b
a
300 b a
a
ab
200 a
100
0
0 30 60 90 120 0 30 60 90 120 0 30 60 90 120
( min ) ( min ) ( min )
Control Coffee 200CA

D
60000 b
a a
50000 b
b
( mg min/dL )

a Control
40000
Coffee
30000
200CA
20000

10000

0
3w 9w 16w

Fig. 2. Glucose Tolerance Test in C57BL/6J Mice Fed with HFD for the Control (n ¼ 5), Coffee (n ¼ 6), and 200CA (n ¼ 6) Groups.
3 weeks (A), 9 weeks (B), and 16 weeks (C) after starting the experiment. D, AUC (area under the curve) of GTT during the course of the
experiment. Data are expressed as the mean
SEM. Means not sharing a common letter (a and b) are significantly different by the Tukey-
Kramer test (p < 0:05).
2312 Y. M ATSUDA et al.

A B
3.0
b Control
b 2.5 b Coffee
a
ab a 200CA
2.0

( mg/dL )

( ng/mL )
a

1.5 a
Control a
Coffee 1.0
200CA
0.5

0.0

( min )

Fig. 3. Insulin Tolerance Test and Serum Insulin Concentration (non-fasting).

A, ITT of C57BL/6J mice fed with HFD for the control (n ¼ 5), coffee (n ¼ 6), and 200CA (n ¼ 6) groups after 15 weeks of the experiment.
B, Serum insulin concentration (non-fasting) at the end of the experiment. Data are expressed as the mean
SEM. Means not sharing a common
letter (a and b) are significantly different by the Tukey-Kramer test (p < 0:05).

Table 1. Initial and Final Body Weights, Food Intake, Water Intake, control group. Adiponectin and leptin are adipocyto-
Tissue Weights, Serum Components, and Liver Lipid Contents of kines that improve insulin sensitivity. The mRNA levels
C57BL/6J Mice Fed with HFD in the Control, Coffee, and 200CA
Groups
of adiponectin did not differ among these three groups,
although the leptin mRNA level was significantly lower
Control Coffee 200CA in the 200CA group than that in the control group
Parameter
(n ¼ 5) (n ¼ 6) (n ¼ 6) (p ¼ 0:15 for coffee).
Body weight (g)
Initial 21:9
0:6 21:5
0:5 21:9
0:5 Change to gene expression in the liver by coffee and
Final 34:2
1:6 31:9
1:3 31:8
1:1 caffeine ingestion
Both coffee and caffeine ingestion tended to decrease
Food intake (g/100gBW/day) the liver TG content when compared to the control
4 weeks 12:0
1:3 12:2
0:3 11:9
0:5
10 weeks 11:3
1:0 12:7
1:0 11:7
0:6
group (Table 1). We measured the hepatic levels of
16 weeks 11:3
0:6 11:4
0:4 11:0
0:4 mRNA of the genes related to fatty acid synthesis and
fatty acid oxidation, the results being shown in Fig. 4B.
Water intake (g/100gBW/day) Sterol regulatory element binding protein-1 (SREBP-1)
4 weeks 17:0
1:5 17:5
0:5 16:2
0:4 has been found to upregulate the expression of fatty acid
10 weeks 13:5
0:5 15:8
1:0 12:7
1:8 synthase (FAS) and acetyl-CoA carboxylase (ACC)
16 weeks 13:3
0:5 12:5
0:6 12:5
0:7
which are regulatory enzymes for fatty acid synthesis.13)
Tissue weights (g/100gbody weight)
Carnitine palmitoyltransferase-1 (CPT-1) is the regula-
Liver 4:55
0:30 4:82
0:11 4:61
0:13 tory enzyme for fatty acid -oxidation, the expression of
Subcutaneous fat 2:42
0:28 2:17
0:14 1:88
0:08 CPT-1 being positively regulated by peroxisome pro-
Epididymal fat 3:64
0:32 3:45
0:29 3:16
0:28 liferator activator receptor-alpha (PPAR-
). As shown
Retroperitoneal fat 0:94
0:08 0:90
0:08 0:76
0:06 in Fig. 4B, the SREBP-1, FAS, and ACC mRNA levels
Mesenteric fat 1:39
0:14 1:27
0:06 1:16
0:04 in the coffee group were significantly lower than those
Interscapular BAT 0:41
0:05 0:35
0:04 0:29
0:03
in the control group. The SREBP-1 and FAS mRNA
Serum components levels in the 200CA group were also significantly lower
Non-fasting TG 121:6
5:6b 111:5
9:2ab 89:6
3:1a than those in the control group. However, the hepatic
(mg/dL) PPAR-
and CPT-1 mRNA levels did not differ among
Non-fasting TC 157:6
13:6 156:3
8:1 146:1
7:3 the three groups.
(mg/dL)
Discussion
Liver lipids (mg/gliver)
Total lipids 66:2
6:4 60:6
2:3 60:5
2:4
TG 25:6
4:1 21:5
1:6 20:4
1:6 We have demonstrated in this study that coffee and
TC 2:95
0:11 2:81
0:10 3:01
0:14 caffeine ingestion both improved glucose tolerance and
PL 18:2
0:4 18:4
0:2 18:6
0:3 hyperinsulinemia in C57BL/6J mice fed with HFD.
Data are expressed as the mean
SEM. Means not sharing a common letter
These effects of coffee and caffeine ingestion were not
(a and b) are significantly different by the Tukey-Kramer test (p < 0:05). accompanied by either a reduction in food intake or any
Interscapular BAT, brown adipose tissue; TG, triglyceride; TC, total body weight loss. The result of the insulin tolerance test
cholesterol; PL, phospholipid. shows that coffee and caffeine ingestion both enhanced
the insulin sensitivity of C57BL/6J mice fed with HFD.
of TNF-
tended to be lower in the 200CA group than in It is considered that the improved insulin resistance
the control group, but again not significantly (p ¼ conferred by coffee or caffeine ingestion contributed to
0:105). The mRNA levels of IL-6 in the coffee and the suppressive effect on HFD-induced impaired glucose
200CA groups were significantly lower than that in the tolerance.
 Coffee and Caffeine Suppress Diet-Induced Glucose Intolerance in Mice
A 1.4
Epididymal fat
1.2 b b
( mRNA/18S rRNA)
1.0
ab
0.8
0.6 a
0.4
0.2
0.0
MCP-1 TNF-α
B 1.4
Liver
1.2 b b
( mRNA/18S rRNA)
1.0
0.8
a a
0.6 a a
0.4
0.2
0.0
SREBP-1 FAS
Fig. 4. Epididymal Fat Tissue and Liver mRNA Levels.
A, Epididymal fat tissue mRNA levels of the indicated genes determined by real-time RT-PCR (results are expressed as number-fold changes
vs. the control group set to 1 unit). B, Liver mRNA levels of the indicated genes determined by real-time RT-PCR (results are expressed
as number-fold changes vs. the control group set to 1 unit). The mice in the control, coffee, and 200CA groups were respectively treated for
17 weeks with water, coffee, and a 200 mg/L caffeine solution. Data are expressed as the mean
SEM. Means not sharing a common letter
(a and b) are significantly different by the Tukey-Kramer test (p < 0:05).
HFD consumption is the principal cause of type 2
diabetes, and the HFD-fed C57BL/6J mouse is well
recognized as a model of obesity-induced diabetes.9,10)
HFD has increased the FFA concentration in serum, and
this elevation inhibited insulin signaling in the periph-
eral tissues.11) HFD has also caused hypertrophy of
adipocytes which induced the production and secretion
of such macrophage chemoattractants as MCP-1.14) The
macrophages recruited by MCP-1 subsequently infiltrate
adipose tissues, leading to a pro-inflammatory state. The
infiltrating macrophages and hypertrophied adipocytes
extensively secrete MCP-1 and such other adipocyto-
kines as TNF-
and IL-6. These adipocytokines in the
circulation have been reported to induce insulin resist-
ance in peripheral tissues.15–17) TNF-
has been indi-
cated to phosphorylate the serine/threonine residue of
insulin receptor substrate-1 (IRS-1) through the activa-
tion of either c-jun N-terminal kinase (JNK), the
inhibitor of NF-B kinase  (IKK), or extracellular
signal-related kinase (ERK).18) Phosphorylation of IRS-
1 by TNF-
interferes with the subsequent insulin-
stimulated tyrosine phosphorylation of IRS-1, leading to
insulin resistance. IL-6 upregulates the suppressor of
cytokine signaling (SOCS) expression by activating the
pathway for Janus kinase-signal transduction and tran-
scription (JAK-STAT) in the liver and adipose tissues,
and inhibits insulin signaling.19,20) IL-6 expression is
also induced by TNF-
.
The results of this study show that neither coffee nor
caffeine ingestion affected the white adipose tissue
weight (subcutaneous, epididymal, mesenteric, and
retroperitoneal fat tissues). Measurement of the adipo-
2313
Control
b
Coffee
ab 200CA
a
a a
IL-6 Adiponectin Leptin
b
a ab
ACC PPAR-α CPT-1
cytokine mRNA levels in adipose tissues revealed that
the IL-6 mRNA levels were significantly lower in the
coffee and 200CA groups than in the control group
(Fig. 4A). The MCP-1 mRNA level in the 200CA group
was significantly lower than that in the control group,
while the MCP-1 mRNA level in the coffee group also
tended to be lower than that in the control group.
The TNF-
mRNA level tended to be lower in the
200CA group than in the control group. These results all
suggest that either coffee or caffeine ingestion lowered
the production of inflammatory adipocytokines and
restrained adipose tissue inflammation, resulting in an
amelioration of the whole-body insulin resistance.
Although we plan to clarify how coffee acts on insulin
signaling in the peripheral tissues, it is possible that
adipose tissue is one of the targets for the direct action of
both coffee and caffeine. This hypothesis is supported by
the results of a previous study in which caffeine
suppressed TNF-
expression in a primary culture of
human adipocytes.21) Fukushima et al. have also
reported that the inflammatory cytokine gene expression
in adipose tissue and liver were reduced by ingesting
decaffeinated or caffeine-containing instant coffee in
C57BL/6J mice fed with HFD.22)
It has been elucidated that the increase of hepatic
lipids caused insulin resistance in the liver.23) The
hepatic accumulation of such lipid metabolites as
diacylglycerol (DAG) is known to activate PKC", this
activated PKC" binding to the insulin receptor and
inhibiting its tyrosine kinase activity. The activation of
PKC" by lipid metabolites may thereby also interfere
with the insulin action to phosphorylate tyrosine
2314 Y. M ATSUDA et al.
residues of IRS-1 and IRS-2, resulting in insulin
resistance in the liver.24,25) The respective hepatic total
lipid contents in this study in the control, coffee, and
200CA groups were 66:2
6:4, 60:6
2:3, and 60:5

2:4 mg/g (Table 1), while the respective hepatic TG
contents were 25:6
4:1, 21:5
1:6, and 20:4
1:6
mg/g. Although the mean values of these contents in the
coffee and 200CA groups were lower than the control
values, we could not detect any significant differences.
We subsequently examined alterations in the hepatic
expression of genes involved in fatty acid synthesis and
oxidation. Interestingly, our results show that the
expression of the genes related to fatty acid synthesis
(SREBP-1, FAS, and ACC) was decreased in the mice
ingesting either coffee or caffeine (Fig. 4B). Expression
of the FAS and ACC genes is known to be upregulated
by transcription factor SREBP-1.13) As the hepatic
SREBP-1 mRNA level was lowered by coffee or
caffeine ingestion when compared to the control group,
this lowering could have contributed to the reduction in
FAS and ACC mRNA levels in the liver. In contrast,
coffee or caffeine ingestion did not affect the hepatic
expression of genes related to fatty acid oxidation
(PPAR-
and CPT-1). It is thus presumed that ingesting
either coffee or caffeine suppresses fatty acid synthesis
and led to an improvement of the fatty liver. Interest-
ingly, Murase et al. have reported that ingesting coffee
polyphenols reduced the liver mRNA level of SREBP-1
and the liver TG content in C57BL/6J mice fed with
HFD.26) The present results suggest that the amelioration
of hyperinsulinemia caused a decrease in the expression
of SREBP-1 genes. It has previously been revealed
that insulin stimulated SREBP-1 gene expression.27) As
the non-fasting serum insulin concentration was signifi-
cantly lower in both the coffee and 200CA groups than
in the control group (Fig. 3B), this reduction could have
contributed to the decrease in SREBP-1 gene expres-
sion. It has also been reported that the expression of
IRS-2 was downregulated by SREBP-1;28) however, the
hepatic IRS-2 mRNA level in the present study was
not altered by either coffee or caffeine ingestion (data
not shown). We observed in our previous study that
coffee or caffeine ingestion ameliorated the fatty liver in
KK-Ay mice which spontaneously developed steatohe-
patitis and hyperinsulinemia. It is speculated that the
ameliorative effect of coffee and caffeine on fatty liver
would occur in both KK-Ay mice and C57BL/6J mice
fed with HFD.
Coffee and caffeine were equally effective for HFD-
induced impaired glucose tolerance in this study, so
caffeine is suggested to be one of the most effective
antidiabetic compounds in coffee. Caffeine has been
shown to have such biological functions as acting as an
adenosine receptor antagonist,29) activating AMP-acti-
vated protein kinase (AMPK) in skeletal muscle,30) and
protecting pancreatic beta-cells against streptozotocin
toxicity.31) Further investigation is needed to determine
whether or not these functions of caffeine were involved
in the antidiabetic effect of this compound.
Previous epidemiological studies have demonstrated
that ingesting both coffee and decaffeinated coffee was
correlated with a reduction in diabetes incidence.6) It has
also been reported that decaffeinated coffee ingestion
might improve the insulin sensitivity of the skeletal
muscle in rats,32) and that coffee components other than
caffeine had a beneficial effect on glucose metabo-
lism.33–38) These results lead us to speculate that there are
antidiabetic compounds in coffee other than caffeine. We
are currently searching for these unidentified compounds.
In conclusion, we have demonstrated that both coffee
and caffeine had an antidiabetic effect on HFD-fed
C57BL/6J mice. The ingestion of either coffee or
caffeine improved the insulin sensitivity and glucose
tolerance in this mouse model. The reduction in the
adipose tissue mRNA levels of inflammatory adipocy-
tokines and in the hepatic mRNA levels of genes
relating to fatty acid synthesis might have contributed to
the antidiabetic effect of coffee and caffeine. These
actions of coffee and caffeine were similar to their
actions observed in spontaneously diabetic KK-Ay
mice.8) As HFD-induced type 2 diabetes occurs fre-
quently in humans, the present results suggest that
coffee consumption would be effective as a preventive
measure against human type 2 diabetes.
Acknowledgment
We thank the All Japan Coffee Association for
financial support for this study.
References
1) Sacks DB and McDonald JM, Am. J. Clin. Pathol., 105, 149–
156 (1996).
2) Malecki MT, Diabetes Res. Clin. Pract., 68 (Suppl1), S10–S21
(2005).
3) Parillo M and Riccardi G, Br. J. Nutr., 92, 7–19 (2004).
4) van Dam RM and Feskens EJ, Lancet, 360, 1477–1478 (2002).
5) van Dam RM and Hu FB, JAMA, 294, 97–104 (2005).
6) Huxley R, Lee CM, Barzi F, Timmermeister L, Czernichow S,
Perkovic V, Grobbee DE, Batty D, and Woodward M, Arch.
Intern. Med., 169, 2053–2063 (2009).
7) Goto A, Song Y, Chen BH, Manson JE, Buring JE, and Liu S,
Diabetes, 60, 269–275 (2011).
8) Yamauchi R, Kobayashi M, Matsuda Y, Ojika M, Shigeoka S,
Yamamoto Y, Tou Y, Inoue T, Katagiri T, Murai A, and Horio
F, J. Agric. Food Chem., 58, 5597–5603 (2010).
9) Surwit RS, Kuhn CM, Cochrane C, McCubbin JA, and Feinglos
MN, Diabetes, 37, 1163–1167 (1988).
10) Rebuffé-Scrive M, Surwit R, Feinglos M, Kuhn C, and Rodin J,
Metabolism, 42, 1405–1409 (1993).
11) Boden G and Shulman GI, Eur. J. Clin. Invest., 32 (Suppl 3),
14–23 (2002).
12) Folch J, Lees M, and Sloane Stanley GH, J. Biol. Chem., 226,
497–509 (1957).
13) Shimano H, Yahagi N, Amemiya-Kudo M, Hasty AH, Osuga J,
Tamura Y, Shionoiri F, Iizuka Y, Ohashi K, Harada K, Gotoda
T, Ishibashi S, and Yamada N, J. Biol. Chem., 274, 35832–
35839 (1999).
14) Sartipy P and Loskutoff DJ, Proc. Natl. Acad. Sci. USA, 100,
7265–7270 (2003).
15) Xu H, Barnes GT, Yang Q, Tan G, Yang D, Chou CJ, Sole J,
Nichols A, Ross JS, Tartaglia LA, and Chen H, J. Clin. Invest.,
112, 1821–1830 (2003).
16) Hotamisligil GS, Shargill NS, and Spiegelman BM, Science,
259, 87–91 (1993).
17) Kern PA, Ranganathan S, Li C, Wood L, and Ranganathan G,
Am. J. Physiol. Endocrinol. Metab., 280, E745–E751 (2001).
18) Lorenzo M, Fernández-Veledo S, Vila-Bedmar R, Garcia-Guerra
L, de Alvaro C, and Nieto-Vazquez I, J. Anim. Sci., 86, E94–
E104 (2008).
19) Senn JJ, Klover PJ, Nowak IA, Zimmers TA, Koniaris LG,
Furlanetto RW, and Mooney RA, J. Biol. Chem., 278, 13740–
13746 (2003).
 Coffee and Caffeine Suppress Diet-Induced Glucose Intolerance in Mice 2315
20) Shi H, Tzameli I, Bjørbaek C, and Flier JS, J. Biol. Chem., 279, 29) Huang ZL, Qu WM, Eguchi N, Chen JF, Schwarzschild MA,
34733–34740 (2004). Fredholm BB, Urade Y, and Hayaishi O, Nat. Neurosci., 8, 858–
21) Dray C, Daviaud D, Guigné C, Valet P, and Castan-Laurell I, 859 (2005).
J. Physiol. Biochem., 63, 329–336 (2007). 30) Egawa T, Hamada T, Kameda N, Karaike K, Ma X, Masuda S,
22) Fukushima Y, Kasuga M, Nakao K, Shimomura I, and Iwanaka N, and Hayashi T, Metabolism, 58, 1609–1617
Matsuzawa Y, J. Agric. Food Chem., 57, 11100–11105 (2009).
(2009). 31) Kagami K, Morita H, Onda K, Hirano T, and Oka K, J. Pharm.
23) Postic C and Girard J, J. Clin. Invest., 118, 829–838 (2008). Pharmacol., 60, 1161–1165 (2008).
24) Samuel VT, Liu ZX, Qu X, Elder BD, Bilz S, Befroy D, 32) Shearer J, Sellars EA, Farah A, Graham TE, and Wasserman
Romanelli AJ, and Shulman GI, J. Biol. Chem., 279, 32345– DH, Can. J. Physiol. Pharmacol., 85, 823–830 (2007).
32353 (2004). 33) Hemmerle H, Burger HJ, Below P, Schubert G, Rippel R,
25) Samuel VT, Liu ZX, Wang A, Beddow SA, Geisler JG, Kahn Schindler PW, Paulus E, and Herling AW, J. Med. Chem., 40,
M, Zhang XM, Monia BP, Bhanot S, and Shulman GI, J. Clin. 137–145 (1997).
Invest., 117, 739–745 (2007). 34) McCarty MF, Med. Hypotheses, 64, 848–853 (2005).
26) Murase T, Misawa K, Minegishi Y, Aoki M, Ominami H, 35) Jung UJ, Lee MK, Park YB, Jeon SM, and Choi MS,
Suzuki Y, Shibuya Y, and Hase T, Am. J. Physiol. Endocrinol. J. Pharmacol. Exp. Ther., 318, 476–483 (2006).
Metab., 300, E122–E133 (2011). 36) Jung EH, Kim SR, Hwang IK, and Ha TY, J. Agric. Food
27) Fleischmann M and Iynedjian PB, Biochem. J., 349, 13–17 Chem., 55, 9800–9804 (2007).
(2000). 37) Mishkinsky J, Joseph B, and Sulman FG, Lancet, 2, 1311–1312
28) Ide T, Shimano H, Yahagi N, Matsuzaka T, Nakakuki M, (1967).
Yamamoto T, Nakagawa Y, Takahashi A, Suzuki H, Sone H, 38) Yoshinari O and Igarashi K, Curr. Med. Chem., 17, 2196–2202
Toyoshima H, Fukamizu A, and Yamada N, Nat. Cell Biol., 6, (2010).
351–357 (2004).