94 | ACTA NATURAE | VOL. 9 № 2 (33) 2017
RESEARCH ARTICLES
INTRODUCTION
In the several decades that have elapsed since it was
discovered that neurotrophic factors play a key role in
the development and maintenance of the viability of
neurons [1], facts showing that they exhibit a similar
regulatory activity at the level of non-neuronal systems
have been obtained [2]. An understanding of the role of
neurotrophins in the development of pancreatic β-cells
was one of the important results of these discoveries.
The data provide grounds to believe that the similarity
between the growth factors and differentiation is re-
sponsible for the similarity between pancreatic β-cells
and neurons, which form via the same fundamental
development program, although they originate from
different cell lineages [3]. The regulatory role of neuro-
trophins in pancreatic β-cells has been confirmed in a
number of studies [4, 5]. The effect of the nerve growth
factor (NGF) on pancreatic β-cells was found to be me-
diated by TrkA, the high-affinity neurotrophin recep-
tor [6]. NGF ensures β-cell neogenesis not only during
the fetal and neonatal periods, but also in adult organ-
isms [7]. The removal of NGF from a β-cell culture
medium [8] and administration of antibodies against
this neurotrophic factor [9] enhances β-cell apoptosis.
Convincing evidence has been obtained showing that a
reduced NGF level in type 2 diabetes mellitus (T2DM)
decreases the proliferation of and/or enhances β-cell
apoptosis [10–12].
Meanwhile, the comorbidity of T2D and cognitive
deficit (reduced information-processing speed, re-
duced verbal memory and conceptualization), whose
risk in T2DM is much higher than in healthy individu-
als, is well-known. According to epidemiological data,
the degree of increase in risk ranges between 50 and
150% [13, 14]. Post-mortem studies have revealed a
decreased NGF level in the frontal cortex of patients
in the phase that precedes Alzheimer’s disease (AD)
[15]. A reduced activity of choline acetyltransferase,
the enzyme whose activity in cholinergic neurons of the
basal brain structures is regulated by NGF, is already
in fact in this phase. It has been demonstrated that the
level of TrkA receptors in the hippocampus, the brain
Low-Molecular-Weight NGF Mimetic
Corrects the Cognitive Deficit and
Depression-like Behavior in Experimental
Diabetes
R. U. Ostrovskaya*, S. S. Yagubova, T. A. Gudasheva, S. B. Seredenin
V.V. Zakusov Institute of Pharmacology, Baltijskaya Str., 8, Moscow, 125315, Russia
*E-mail: rita.ostrovskaya@gmail.com
Received September 28, 2016; in final form, February 20, 2017
Copyright © 2017 Park-media, Ltd. This is an open access article distributed under the Creative Commons Attribution License,which permits
unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
ABSTRACT Based on the comorbidity of diabetes, depression, and dementia and recognizing that a deficiency of
the nerve growth factor (NGF) is involved in all of these kinds of pathologies, we studied the effect of the mimet-
ic of dimeric dipeptide NGF loop 4, GK-2, on a model of streptozotocin-induced type 2 diabetes in C57Bl/6 mice.
GK-2 [hexamethylenediamide bis-(N-monosuccinyl-glutamyl-lysine)] was synthesized at the V.V. Zakusov Sci-
entific Research Institute of Pharmacology. The study revealed the ability of GK-2 to ameliorate hyperglycemia
induced by streptozotocine (STZ 100 mg/kg i.p.) in C57Bl/6 mice, to restore learning ability in the Morris Water
Maze test, and to overcome depression after both intraperitoneal (0.5 mg/kg) and peroral (5 mg/kg) long-term
administration. The presence of the listed properties and their preservation in the case of peroral treatment
determines the prospects of research. Taking into account the previous findings on the ability of GK-2 to selec-
tively activate PI3K/Akt, these data suggest that Akt-signaling is sufficient for pancreatic beta cell function.
GK-2 has been shown to exhibit pronounced neuroprotective activity. The coexistence of neuroprotective and
antidiabetic effects is in agreement with the fundamental concept holding that the function of neurons and
pancreatic beta cells is controlled by similar mechanisms.
KEYWORDS depression, diabetes, dipeptide NGF mimetic, learning.
ABBREVIATIONS AD – Alzheimer’s disease; BDNF – brain-derived neurotrophic factor; i.p. – intraperitoneal;
NGF – nerve growth factor; per os – peroral; STZ – streptozotocin; T2D – type 2 diabetes.
RESEARCH ARTICLES
VOL. 9 № 2 (33) 2017 | ACTA NATURAE | 95
structure responsible for the main cognitive functions
and memory, in particular, is reduced in patients with
mild cognitive impairment [16]. Hippocampal atrophy
is an important prognostic sign of an aggravation of the
cognitive pathology and a transition from mild cogni-
tive impairment to AD [17]. Deficiency in NGF plays an
important role in it, since this neurotrophin prevents
the formation of β-amyloid peptide (Aβ1–42) [18]. The
decrease in the NGF level accompanying a cognitive
deficit is associated with an increased level of its pre-
cursor (proNGF) that suppresses the proliferation and
differentiation of the basal brain and hippocampal
structures [19]. A shift in the proNGF/NGF ratio to-
wards precursor prevalence is regarded as the main
reason for cholinergic deficit, leading to cognitive im-
pairment [20].
The risk of depression and depressive-like behavior
in T2DM is at least twice as high as that in individuals
without resistance to insulin [21]. The bilateral comor-
bidity of these disorders (depression aggravates the
course of diabetes and vice versa) has been studied [22,
23]. In addition to the convincing data on the role of a
deficiency in the brain-derived neurotrophic factor
(BDNF) in the pathogenesis of depressive states of dif-
ferent etiologies, including in patients with diabetes [24],
it has been demonstrated that the activity of NGF drops
both in depression and in diabetes, which is considered
to be an important factor that determines their comor-
bidity. A meta-analysis of 21 publications [25] confirmed
a statistically significant decrease in the blood level of
NGF in depression, which correlated with impairment
intensity. It has been suggested that the reduced level of
NGF in blood serum should be regarded as a biomarker
for major depression [26]. Such a reduction is also ob-
served in patients with bipolar disorder [27] and senile
depressions [28]. Post-mortem examinations of brain tis-
sues from suicide victims have revealed an almost two-
fold decrease in NGF expression and a more than three-
fold decrease in TrkA density [29].
A combination of the reported data demonstrates
that NGF could be used in patients with type 2 diabetes
mellitus because of its ability to maintain β-cell func-
tion, stimulate insulin secretion, and simultaneously
impede the development of diabetes mellitus and its
comorbidities. However, in their attempts to use na-
tive NGF, researchers have faced a problem associated
with the unsatisfactory pharmacokinetic properties of
this protein molecule (low biological stability and in-
ability to pass through biological barriers when admin-
istered systemically) and pleiotropicity of NGF activity,
which may result in such side effects as weight loss and
hyperalgesia. Meanwhile, the effectiveness of topical
administration of NGF in trophic ulcers of diabetic gen-
esis has been reported [30]. As for systemic administra-
tion of NGF, phase I/II clinical trials of recombinant
NGF have revealed a tendency towards a favorable
effect in patients with diabetic neuropathy; however,
side effects and the lack of a therapeutic effect were
observed when a broader patient population was used
in phase III trials [31].
One of the strategies used to overcome the draw-
backs of native neurotrophins involves the design of
low-molecular-weight agents that can induce NGF-like
therapeutic effects upon systemic administration with-
out the side effects typical of native NGF. Several com-
pounds of this type have been reported; in particular,
NGF mimetic of nonpeptide structure, compound MT-2
[32], and peptide NGF-mimetic BB14 [33, 34]. However,
the effects of these compounds have been studied only
in in vitro systems.
A dimeric dipeptide NGF mimetic GK-2 (hexameth-
ylenediamide-bis-(N-monosuccinyl-glutamyl-lysine))
has been designed at the V.V. Zakusov Research Insti-
tute of Pharmacology on the basis of the structure of
the NGF loop 4 β-turn. It exhibited a high neuroprotec-
tive activity in in vitro experiments, as well as in vivo in
models of stroke, Alzheimer’s and Parkinson’s diseases,
and had none of the side effects typical of native NGF.
GK-2 was shown to activate TrkA receptors [35–37].
Preliminary experiments in rats demonstrated that
GK-2 exhibits anti-hyperglycemic activity [38]. On
the basis of the comorbidity of diabetes and cognitive
impairment and depression, we modeled streptozoto-
cin-induced diabetes in mice and studied the effect
of GK-2, the original NGF mimetic, on the cognitive
impairment and depressive-like behavior in these ani-
mals.
EXPERIMENTAL
Animals
Male C57Bl/6 mice with an initial body weight of 23–
28 g purchased from the Stolbovaya breeding farm
were used in the experiments. The animals were kept
under standard vivarium conditions, with unrestricted
access to food (except for 16 h prior to streptozotocin
administration) and water. The guidelines for ethical
rules in the care and use of animals in research summa-
rized in the European Communities Council Directive
86/609/ЕЕС were followed.
Experiment design
Type 2 diabetes mellitus was induced by intraperito-
neal (i.p.) administration of streptozotocin (STZ, Sigma,
USA) at a dose of 100 mg/kg, which was effective for
C57Bl/6 mice [39].
The mice were randomly divided into four groups:
group 1 (passive control, n = 10), group 2 (active con-
96 | ACTA NATURAE | VOL. 9 № 2 (33) 2017
RESEARCH ARTICLES
trol, n = 11), and experimental groups 3 (n = 11) and 4
(n = 12). The mice in the passive control group received
saline, either i.p., or perorally (per os), for 31 days
1
. The
animals from the active control group received saline
i.p. for 14 days; a single dose of STZ (100 mg/kg) was
administered i.p. on day 15 after 16-hour fasting; then,
mice continued to receive saline for 16 days.
The low molecular weight (831 Da) of GK-2 makes
it reasonable to study the effects of both i.p. and the
peroral route of administration. The effect of GK-2 ad-
ministered per os needs to be studied, since this com-
pound is intended to be used as a drug for long-term
clinical application. The freshly prepared GK-2 solu-
tion (in 0.9% NaCl) was administered once a day during
14 days: in study group 3, i.p. at a dose of 0.5 mg/kg;
in study group 4, per os at a dose of 5 mg/kg. On day
15 (30 min after the animals had received the final dose
of GK-2), they were i.p. treated with STZ (100 mg/kg)
on an empty stomach; then, both groups of mice contin-
ued to receive GK-2 for 16 days.
The glucose level in the blood collected from the tail
vein was measured using a One Touch Ultra glucom-
eter (USA). The dynamics of the effect of GK-2 was
assessed using the indicator of relative antihyperglyce-
mic activity (Ag) according to the formula
Ag = gl.STZ – gl.(STZ + GK-2) / gl.STZ – gl.saline × 100%,
where gl.STZ is the blood glucose level in the active
control group (group 2); gl.STZ + GK-2 is the blood
glucose level in the study group 3 or 4; and gl.saline
is the blood glucose level in the passive control group
(group 1).
Studying the effect of GK-2 on learning
ability in the Morris water maze
Spatial learning and memory were assessed 24 h after
the mice had received the final dose of GK-2 (day 17
after administration of STZ) using the Morris water
maze [40]. The experimental device consisted of a pool
150 cm in diameter with 60-cm-high walls filled with
water (23–25°С). The pool was imaginatively divided
into four quadrants. A platform 9 cm in diameter, 1 cm
higher than the water level, was placed in the center of
one quadrant.
During day 1, the animals were allowed to find the
visible platform. If the mouse did not find the plat-
form during the 60 s cut-off, it was placed on the
platform and allowed to stay there for 20 s before
returning to its home cage. Four trials (one per each
quadrant) were used. After 24 h, a platform sub-
1
No significant differences between i.p. or per os administration of saline
during the entire experiment were revealed, so these animals were merged
into one group.
merged 1 cm below the water level was placed onto
the same spot as in day 1, but water was preliminarily
whitened with milk. Identically to day 1, four trials
were used, one for each quadrant. The same proce-
dure was repeated on days 3, 4, 5, and 8. The number
of animals that found the platform within the 60 s cut-
off was recorded.
Studying the effect of GK-2 using
the depression model
The depressive-like behavior (the behavioral despair)
was assessed using the modified forced swim test on
days 45 and 46 after discontinuation of GK-2 [41, 42].
Cylindrically shaped vessels 10 cm in diameter and
30 cm high (OOO Research and Production Company
Open Science) were filled with water (23–25°С) to the
level of 20 cm from the bottom. On day 1, the animal
was placed into the vessel for 10 min and its behavior
was video-recorded in the interval between the 2
nd
and
the 6
th
minute. The test was repeated for 6 min after
24 h. Active swimming and immobilization durations
in both sessions were determined using the RealTimer
software. According to the definition given by the au-
thors of the test, active swimming implied the periods
when the forelimbs moved upward along the cylinder
walls, while immobilization implied remaining com-
pletely motionless or making the minor movements
necessary to maintain the head above water. The total
duration of immobilization episodes was the key pa-
rameter of the severity of depressive-like behavior in
this test.
Exploratory behavior, as well as the general loco-
motor activity, was assessed using the open field test
2 days prior to performing the Morris water maze. The
animals were placed in the center of the open field, and
the horizontal motor activity and the numbers of holes
and vertical bars were measured during 5 min.
The animals’ body weight was measured every 3
days.
Figure 1A shows the order in which the compounds
were administered and behavioral tests were per-
formed.
Statistical analysis
The experimental data are shown as mean values, with
the mean error and the standard error of the mean
(M ± SЕM) indicated. The statistical analysis was per-
formed using the Statistica 8.0 software. The statistical
significance of intergroup differences was assessed us-
ing the nonparametric method, the Mann–Whitney U
test. The χ
2
test was used for the parameters measured
in %. The results were considered to be statistically sig-
nificant at p ≤ 0.05.
RESEARCH ARTICLES
VOL. 9 № 2 (33) 2017 | ACTA NATURAE | 97
mmol/l
А
B
GK-2 or
saline
treatment
STZ GK-2 or
saline
treatment
After
discontinuation
of treatment
Morris
Water
Maze
Forced
swim
test
days
1 14 1 16 17 24 61 62
25
20
15
10
5
0
Before STZ
(14 days of
treatment)
day 5
(20 days of
treatment)
day 19 (3 days
after treatment
discontinuation)
day 60
(44 days after
treatment
discontinuation)
Passive control (saline)
Active control (STZ)
GK-2, i.p. + STZ
GK-2, per os + STZ
Fig. 1. Design of the experiment (A) and the dynamics of the blood glucose level (mmol/l) in C57Bl/6 mice (B) in
the following groups: passive control (Saline + Saline), active control (Saline + STZ 100 mg/kg, i.p. + Saline), GK-2
treated group 3 (GK-2 0.5 mg/kg, i.p. + STZ 100 mg/kg, i.p. + GK-2 0.5 mg/kg, i.p.), GK-2 treated group 4
(GK-2 5 mg/kg, per os + STZ 100 mg/kg, i.p. + GK-2 5 mg/kg, per os). Data are presented as M ± SEM. The sta-
tistical significance of the differences was calculated using the Mann–Whitney U-test: * р<0.05 compared to passive
control; # р<0.05 compared to active control (STZ).
98 | ACTA NATURAE | VOL. 9 № 2 (33) 2017
RESEARCH ARTICLES
RESULTS
Data on the dynamics of the blood glucose level in dif-
ferent groups are presented in Fig. 1B. While the glu-
cose level in the peripheral blood of mice in the passive
control group was 6–7 mmol/l, administration of STZ
at a dose of 100 mg/kg to С57Bl/6 mice increased that
blood glucose level to 16–20 mmol/l, which is close to
the values obtained earlier in the experiments with rats
[38]. In full compliance with the antihyperglycemic ef-
fect of GK-2 observed in the experiments with rats, we
revealed the antihyperglycemic effect of GK-2 in mice.
It is important to emphasize that the antihyperglyce-
mic effects were similar for rats and mice: e.g., the cal-
culated Ag parameter on day 17 after administration of
STZ to rats was 80%, being 90% on day 19 in mice.
Assessment of the cognitive function performed 24
h after the final dose of GK-2 had been injected dem-
onstrated (table) that, whereas the number of animals
that found the platform within 60 s in repeated tests
significantly increased in the passive control group, this
occurred muchmore slowly in the active control group
(the differences between the two groups were statisti-
cally significant on days 4 and 8). These results agree
with the data on cognitive impairment in STZ-induced
diabetes [43]. Intraperitoneal administration of GK-2
caused a statistically significant increase in the number
of animals that found the platform on days 2, 4, and 8 of
training compared to the animals in the active control
group. Upon administration per os, the learning ability
significantly increased only on test day 2. It should be
mentioned that in the beginning of the experiment, the
learning ability of mice for both administration routes
was even higher than that in the passive control group.
The intergroup differences were significant on test
days 3 and 5 as well (except for day 5 in the group that
received GK-2 per os, when the differences between
the active control and the study group failed to reach
the level of statistical significance).
The effect of GK-2 on the severity of the depression-
like behavior was assessed in a long-term period after
STZ administration (day 45), since the duration of the
depressive-like behavior in the diabetes model was re-
ported to be rather long [25].
Comparison of active swim test parameters and the
immobilization duration in different groups revealed
the following regularities (Fig. 2): In mice in the active
control group, immobilization duration increased, while
the duration of active swimming decreased compared
to the parameters in the passive control group, while
i.p. administration of GK-2 reduced the immobilization
duration and increased the active swimming duration,
making them as high as the control values. The inten-
sity of the effect of GK-2 administered per os was the
same as upon i.p. administration.
Similar regularities were observed on day 2: in-
creased immobilization duration and reduced active
swimming duration in the active control group, where
GK-2 reduced the severity of depression when admin-
istered both i.p. and per os.
In order to interpret the results, we needed to under-
stand whether the streptozotocin-induced behavioral
disorders were related to the overall wellbeing of the
animals (reduced motor activity and body weight loss).
In order to answer this question, we performed the
open field test 2 days prior to the Morris water maze,
where changes in neither the orientational nor explor-
atory activity and overall mobility were observed in
the animals treated with STZ. GK-2 upon both admin-
istration routes had no effect on these indicators. It was
demonstrated that, unlike the passive control group
where animal body weight increased during the en-
tire experiment (10.5% with respect to the initial weight
by the time the Morris water maze was performed and
16.7% by the time the forced swim test was formed), a
slight decrease in body weight by the time of Morris
water maze study (–6.7%) and body weight gain by the
time of the forced swim test (1.8%) were observed in
the active control group. GK-2 reduced this effect of
STZ administered both i.p. (–2 and 4.6%, respectively)
and per os (1 and 10%, respectively). Therefore, the
resulting data allow one to rule out the changes in the
overall wellbeing of animals as the reason for the STZ-
induced behavioral disorders and their normalization
due to the administration of NGF mimetic.
Learning ability of mice in the Morris water maze (the
percentage of animals that found the platform within the
60 s cut-off time)
Group day 2 day 4 day 8
Group 1
Passive control (saline)
14.3% 85.7% 100%
Group 2
Active control (STZ, 100
mg/kg)
9.09% 54.54%* 72.7%*
Group 3
GK-2, 0.5 mg/kg
i.p. + STZ
27.3%*
#
72.7%*
#
90.9%*
#
Group 4
GK-2, 5 mg/kg
per os + STZ
50%*
#
50%* 100%
#
The statistical significance of differences was assessed us-
ing the χ
2
test.
*р < 0.05 compared to passive control group (saline).
#
р < 0.05 compared to active control group (STZ).
![Fig. 3. NGF is synthesized from the precursor, pro-NGF. NGF binds to the TrkA receptor and this interaction induces the activation of the signaling pathway of β-cell survival. Diabetes-induced hyperglycemia is known to cause the oxidative stress that decreases protease activity, thus provoking pro-NGF accumulation resulting in β-cell apoptosis (modified from [19, 48])](/figures/fig-3-ngf-is-synthesized-from-the-precursor-pro-ngf-ngf-3vc9o5vb.webp)