A transgenic dwarf rat model as a tool for the study of calorie
restriction and aging
Haruyoshi Yamaza, Toshimitsu Komatsu, Takuya Chiba, Hiroaki Toyama, Kazuo To,Yoshikazu Higami, Isao Shimokawa*
Department of Pathology and Gerontology, Nagasaki University Graduate School of Biomedical Sciences, 1-12-4 Sakamoto, Nagasaki City 852-8523, Japan
Received 20 July 2003; received in revised form 29 October 2003; accepted 4 November 2003
We have previously reported a long-lived transgenic dwarf rat model, in which the growth hormone (GH)-insulin like growth factor (IGF)-
1 axis was selectively suppressed by overexpression of antisense GH transgene. Rats heterozygous for the transgene (tg/2) manifest
phenotypes similar to those in calorie-restricted (CR) rats. To further characterize the transgenic rat in comparison with CR rats, the present
study evaluated glucose and insulin tolerance in tg/2 and control Wistar (2 /2 ) rats at 69 months of age. Rats were fed ad libitum (AL) or30% CR from 6 weeks of age. In CR rats, glucose disposal after glucose load was facilitated without any significant surge of serum insulin,
and insulin tolerance test also indicated increased insulin sensitivity. In transgenic rats, similar findings were observed after glucose and
insulin load, and CR in tg/2 rats further facilitated glucose disposal during glucose and insulin tolerance tests. These findings suggest thepresence of both common and separate mechanisms regulating the glucoseinsulin system between CR and the reduced GHIGF-1 axis
paradigms. The transgenic rat model is, therefore, a useful one for studies of CR and aging.
q 2004 Elsevier Inc. All rights reserved.
Keywords: Growth hormone; IGF-1; Calorie restriction; Glucose; Insulin; Transgenic rat
Utilization of spontaneously mutated or genetically
engineered rodent models that mimic physiological states
induced by calorie restriction (CR) progresses our under-
standing of the aging process and assists in developing anti-
aging interventions in humans. We previously reported a
long-lived transgenic dwarf rat model, in which the growth
hormone (GH)-insulin like growth factor (IGF)-1 axis was
selectively suppressed by overexpression of an antisense
GH transgene (Shimokawa et al., 2002). These rats share
some phenotypes with CR rats, including longer lifespan,
some pathologies, reduced body size and food intake, and
lower plasma levels of insulin, glucose, and IGF-1
(Shimokawa et al., 2003).
In lower organisms such as nematodes and fruit flies, in
which insulin and IGF-1 systems are not clearly separated,
functional mutations in insulin- or IGF-1-signaling expand
lifespan (Strauss, 2001). In rodents, genetic mutations in this
signaling also increase lifespan, although most direct
manipulations of insulin signaling induce metabolic impair-
ments such as diabetes and shorten lifespan (Baudry et al.,
2002). Nonetheless, the glucose-insulin system appears
important for regulation of aging and longevity in mammals;
recent studies demonstrate that a reduced GHIGF-1 axis
concomitantly modulates the glucose-insulin system, simi-
lar to CR (Longo and Finch, 2003), and that a fat-specific
knock out of the insulin receptor gene increases lifespan in
mice (Bluher et al., 2003).
Comparative studies using CR and rodent models with
the reduced GHIGF-1 axis could facilitate our under-
standing of the molecular mechanisms of aging and
longevity. In this short communication, we described a
transgenic dwarf rat with glucose and insulin tolerance and
indicated the suitability of the model for future CR and
0531-5565/$ - see front matter q 2004 Elsevier Inc. All rights reserved.
Experimental Gerontology 39 (2004) 269272
* Corresponding author. Tel.: 81-95-849-7050; fax: 81-95-849-7052.E-mail address: email@example.com (I. Shimokawa).
2. Materials and methods
The details of the rats and their husbandry in our
laboratory are described elsewhere (Shimokawa et al., 2003,
2002). The transgenic rats (Jcl: Wistar-TgN (ARGH-
GEN)1Nts) were kindly provided by Nippon Institute for
Biological Science (Oume City, Tokyo, Japan) and the
present rat colony has been established in a barrier facility in
the Laboratory Animal Center at Nagasaki University
School of Medicine since 1997. The transgene consisted
of four copies of the thyroid hormone response element, the
rat GH promoter, and antisense cDNA sequences for rat GH
(Matsumoto et al., 1993). Male rats heterozygous for the
transgene (tg/2 ) were used, because those rats manifestedphenotypes similar to those in control non-transgenic Wistar
(2 /2 ) rats subjected to CR (Shimokawa et al., 2003).Control male Wistar rats were purchased from Japan Clea,
Inc. (Tokyo, Japan).
At 4 weeks of age, weanling male rats were transferred to
a barrier facility (temperature 2225 8C; 12-h light/darkcycle), kept separately under a specific pathogen-free
condition, and fed ad libitum (AL) with Charles River-
LPF diet (Oriental Yeast Co. Ltd. Tsukuba, Japan). At
6 weeks of age, 30% CR was started by providing 140% of
the mean food intake for 2 days in each group AL every
other day. General data for each rat group at 6 months of age
are presented as a reference (Table 1).
All experiments reported here were performed in accord
with provisions of the Ethics Review Committee for Animal
Experimentation at Nagasaki University.
2.2. Glucose tolerance test
Glucose tolerance tests (GTT) were performed on rats at
67 months of age. After a 15-h overnight fast, rats were
injected intraperitoneally with D-glucose (1.0 g/kg body
weight; 50% solution) and blood samples were withdrawn
from tail veins at 0, 15, 30, 60, 90, and 120 min after glucose
load without anesthesia using a 24-gauge needle. When
collected a blood sample at each time point, a rat was placed
into a body-sized-adjusted plastic box restrainer and blood
was obtained in less than 60 s from the contact with the rat.
In this experiment, blood samples were taken from each rat
at only two time points because of the difficulty of repetitive
samplings in the same animals. Group 1 rats were subjected
to blood samplings at 0 and 60 min, while group 2 at 15 and
90 min, and group 3 at 30 and 120 min. Blood glucose
concentration was immediately measured with ACCU-
CHEKw Active (Roche Diagnostics GmbH, Tokyo,
Japan). Serum samples were also prepared by centrifugation
of blood samples, and then stored at280 8C until assays forinsulin concentrations. Serum insulin levels were measured
at 0, 15, 30, and 60 min with a rat insulin enzyme-
immunoassay system (Amersham Pharmacia Biotech, Little
2.3. Insulin tolerance test
The rats used for GTT were also subjected to insulin
tolerance tests (ITT) at 89 months of age, and were fasted
again for a 15-h before ITT. Blood samples were withdrawn
from tail veins after intraperitoneal injection of human
insulin (0.75 units/kg body weight; 1 unit/ml solution,
Sigma Chemical Co., St Louis, MO) without anesthesia
following the same procedure for GTT. Blood glucose
concentration was also measured with ACCU-CHEKw
Active (Roche Diagnostics GmbH).
Data were presented as means ^ SD Blood glucose and
serum insulin concentrations were analyzed for the main
effect of transgene (Tg; 2 /2 or tg/2 ), diet (Diet; AL orCR), time (Time; 0, 15, 30, 60, 90, or 120 min), and their
interaction (Tg Diet, Tg Time, Diet Time, Tg Diet Time) by three-factor analyses of variance (3-fANOVA) after logarithmic transformation of data. Fishers
protected least significant difference (PLSD) test was also
2 /2 tg/2
AL CR AL CR
Body weight (g) 481.4 ^ 23.9 (21) 336.9 ^ 23.3 (21) 308.7 ^ 14.3 (19) 209.1 ^ 14.8 (20)
Food intake for 2 days (g) 45.3 ^ 2.7 (10) 31.7 31.7 ^ 2.4 (10)* 22.2
IGF-1 (ng/ml) 1058.3 ^ 127.2 (12) 818.3 ^ 82.3 (12) 626.5 ^ 89.6 (5)** 345.6 ^ 39.8 (5)
Glucose (mg/dl) 126.1 ^ 33.9 (5) 107.1 ^ 10.2 (8) 105.5 ^ 18.0 (5) 90.2 ^ 13.1 (8)
Insulin (ng/ml) 101.8 ^ 48.8 (5) 16.0 ^ 9.3 (8) 21.6 ^ 17.7 (5) 23.5 ^ 26.0 (8)
Values represent the mean ^ SD (the number of rats examined). All data was measured at 6 months of age. The data of IGF-1, glucose, and insulin are
cited from the paper of Shimokawa et al (2003). Results of 2-f ANOVA on body weight are (1) body weight: Tg effect, p , 0:0001; CR effect, p , 0:0001;
Tg CR, p , 0:0001; (2) IGF-1: Tg effect, p , 0:0001; CR effect, p , 0:0001; Tg CR, not significant (ns), (3) Glucose: Tg effect, p , 0:05; CR effect,p , 0:05; Tg CR, ns, (4) Insulin: Tg effect, p , 0:05; CR effect, p , 0:01; Tg CR, p , 0:05; *p , 0:0001 vs 2 /2 (AL) by Fishers PSLD test after 1-fANOVA. **p , 0:05 versus 2 /2 (CR) by Fishers PSLD test after 1-f ANOVA.
H. Yamaza et al. / Experimental Gerontology 39 (2004) 269272270
performed as a post hoc test. One-factor ANOVA and the
post hoc test were also carried out as needed for multiple
comparisons. The level of significance was set at p , 0:05:
3.1. General data
The body weight and food intake in tg/2 (AL) rats werecomparable with those in 2 /2 (CR) rats (Table 1). Plasmaconcentrations of IGF-1, glucose, and insulin in the fed-state
had been previously determined (Shimokawa et al., 2003).
3.2. Glucose tolerance test
As a whole, blood glucose concentration was increased
to a peak value at 15 min after glucose load, and gradually
returned to basal (0 min) level (Fig. 1(a); T effect,
p , 0:0001: Blood glucose decreased in tg/2 rats (Tgeffect, p , 0:0001; although the time-dependent alterationwas not affected (Tg Time, not significant). Bloodglucose was also reduced similarly by CR in both 2 /2and tg/2 rats (Diet effect, p , 0:0001; Tg Diet, notsignificant). The time-dependent changes in glucose con-
centrations were significantly affected by CR (Diet Time,p , 0:05: In AL rats, the blood glucose concentrationgradually decreased between 15 and 90 min; however, in
CR rats, it quickly returned to basal level at 30 min.
Serum insulin concentration during GTT was affected by
CR, Tg, and Time (Fig. 1(b); Tg effect, p , 0:0001; Dieteffect, p , 0:0001; Time effect, p , 0:0001); however,there were also significant interactions between and among
the factors (Tg Diet, p , 0:008; Diet Time, p , 0:007;Tg Diet Time, p , 0:0006: The concentration of
insulin was transiently increased at 15 min in 2 /2 (AL)rats, and the level reduced precipitously to basal level at
30 min. There was no similar surge of insulin in the other
three groups of rats.
3.3. Insulin tolerance test
Blood glucose concentration decreased gradually
between 15 and 90 min after insulin injection and stayed
constant until 120 min (Fig. 2; Time effect, p , 0:0001:CR reduced blood glucose concentration in 2 /2 and tg/2rats (Diet effect, p , 0:0001; Tg Diet, not significant;Tg Diet Time, not significant), and Tg also reduced it inAL and CR rats (Tg effect, p , 0:0001:
Fig. 1. (A) Blood glucose concentration during glucose tolerance testing. Data represent means ^ SD of four to seven rats. *p , 0:05 vs each correspondent
group AL at 30 min. **p , 0:005 vs tg/2 (AL) and 2 /2 (AL), and p , 0:05 vs 2 /2 (CR) by multiple comparisons at 60 min. (B) Serum insulin
concentration during glucose tolerance testing. Data represent means ^ SD of three to six rats. #p , 0:0001 vs the other three groups by multiple comparisons
at 15 min. ##p , 0:05 vs 2 /2 (AL) by multiple comparisons at 30 min.
Fig. 2. Blood glucose concentration during insulin tolerance testing. Data
represent means ^ SD of five to six rats. *p , 0:05 vs each correspondent
group AL, **p , 0:0001 vs tg/2 (AL), #p , 0:005 vs tg/2 (AL) and
2 /2 (CR), and ##p , 0:05 vs 2 /2 (AL) by multiple comparisons at each
H. Yamaza et al. / Experimental Gerontology 39 (2004) 269272 271
The present results were comparable to those in previous
studies indicating that CR improves glucose tolerance and
enhances insulin sensitivity in rodents (Escriva et al., 1992).
Blood glucose concentration returned quickly to basal level
without an insulin surge in response to exogenous glucose;
and ITT also confirmed increased insulin sensitivity under
CR conditions. Interestingly, transgenic rats with the
reduced GHIGF-1 axis, which were fed AL, manifested
similar findings. The property of the glucoseinsulin system
in the transgenic rat was different from those in long-lived
Ames dwarf and GHR-KO mice; as those mice models show
glucose intolerance, although insulin sensitivity is increased
(Coschigano et al., 1999; Dominici et al., 2002). In this
respect, our transgenic rat model mimics the physiological
state induced by CR more closely than the dwarf mice
In the present model, CR further augmented glucose
disposal without any significant change of serum insulin.
These findings suggest not only that CR modulates glucose
metabolism independently from the GHIGF-1 axis, but
also that CR enhances insulinotropic or non-insulin
dependent mechanisms for glucose disposal. Further
analyses will be needed to elucidate differences between
CR and the reduced GHIGF-1 axis in the mechanisms that
underlie increased insulin sensitivity and glucose
Considering previously presented data on longevity and
pathology, and together with the general data presented here
in Table 1, we conclude that our transgenic dwarf rat is
suitable for molecular analyses on the anti-aging effects of
CR; as well as for assessing the relationship between
longevity and insulin/IGF-1 signalings.
We thank Yutaka Araki and the staff in the laboratory
animal center at Nagasaki University School of Medicine
for their excellent technical support. We also thank Nippon
Institute for Biological Science for providing the transgenic
rat. This work was supported by the Research Grant for
longevity Sciences (grants 11-C) from the Ministry of
Health, Welfare, and Labor of Japan.
Baudry, A., Leroux, L., Jackerott, M., Joshi, R.L., 2002. Genetic
manipulation of insulin signaling, action and secretion in mice. Insights
into glucose homeostasis and pathogenesis of type 2 diabetes. Eur. Mol.
Biol. Org. Rep. 3 (4), 323328.
Bluher, M., Kahn, B.B., Kahn, C.R., 2003. Extended longevity in mice
lacking the insulin receptor in adipose tissue. Science (5606), 572574.
Coschigano, K.T., Riders, M.E., Bellush, L.L., Kopchick, J.J., 1999.
Glucose Metabolism in Growth Hormone Receptor/Binding Protein
Gene Disrupted Mice, Eighty First Annual Meeteing of the Endocrine
Society, San Diego, CA, USA, p. 553.
Dominici, F.P., Hauck, S., Argentino, D.P., Bartke, A., Turyn, D., 2002.
Increased insulin sensitivity and upregulation of insulin receptor,
insulin receptor substrate (IRS)-1 and IRS-2 in liver of Ames dwarf
mice. J. Endocrinol. 173 (1), 8194.
Escriva, F., Rodriguez, C., Cacho, J., Alvarez, C., Portha, B., Pascual-
Leone, A.M., 1992. Glucose utilization and insulin action in adult