FROM:
Environ Mol Mutagen. 2016 (Jun); 57 (5): 382–404 ~ FULL TEXT
J.A. Lemon, V. Aksenov, R. Samigullina, S. Aksenov, W.H. Rodgers, C.D. Rollo, D.R. Boreham
Department of Medical Physics and Applied Radiation Sciences,
McMaster University,
1280 Main Street West,
Hamilton ON, Canada, L8S 4K1.
Transgenic growth hormone mice (TGM) are a recognized model of accelerated aging with characteristics including chronic oxidative stress, reduced longevity, mitochondrial dysfunction, insulin resistance, muscle wasting, and elevated inflammatory processes. Growth hormone/IGF-1 activate the Target of Rapamycin known to promote aging. TGM particularly express severe cognitive decline. We previously reported that a multi-ingredient dietary supplement (MDS) designed to offset five mechanisms associated with aging extended longevity, ameliorated cognitive deterioration and significantly reduced age-related physical deterioration in both normal mice and TGM. Here we report that TGM lose more than 50% of cells in midbrain regions, including the cerebellum and olfactory bulb.
This is comparable to severe Alzheimer's disease and likely explains their striking age-related cognitive impairment. We also demonstrate that the MDS completely abrogates this severe brain cell loss, reverses cognitive decline and augments sensory and motor function in aged mice. Additionally, histological examination of retinal structure revealed markers consistent with higher numbers of photoreceptor cells in aging and supplemented mice. We know of no other treatment with such efficacy, highlighting the potential for prevention or amelioration of human neuropathologies that are similarly associated with oxidative stress, inflammation and cellular dysfunction.
Keywords aging; multi-ingredient dietary supplement; neurodegeneration; neuroprotectant; oxidative stress
From the FULL TEXT Article:
INTRODUCTION
The brain is particularly vulnerable to free radical damage
with its high content of unsaturated fatty acids, high
oxygen metabolism (20% of total body consumption) and
relatively low levels of endogenous antioxidants. Accumulation
of oxidative damage in post-mitotic neurons is a crucial
factor in normal brain aging, and contributes directly to
cognitive, motor and sensory impairments [Antier et al.,
2004; Mattson and Magnus, 2006; Wang and Michaelis,
2010; Yin et al., 2014]. This appears to be exacerbated in
many neurodegenerative diseases, many of which are also
associated with aging [Markebury and Carney, 1999, Butterfield
et al., 2001; Ma et al., 2003; Butterfield, 2014].
Mitochondrial dysfunction and oxidative metabolism
are principal sources of oxidative stress leading to neurodegeneration,
although NAD(P)H oxidase and other sources
of free radicals also contribute [Rollo, 2002; Sonta
et al., 2004; Cui et al., 2012; Gandhi and Abramov,
2012]. Neurons have high energy demands (ATP consumption)
associated with membrane ionic pumps, channel
activity, and synaptic transmission [Erecinska et al.,
2004; Du et al., 2008], which can lead to increased free
radical production. Elevation in free radicals can increase
levels of glutathione disulfide (GSSG) which inhibits the
thiol-dependent enzyme NADH dehydrogenase, causing a
mitochondrial complex I defect [Cohen et al., 1997].
Mitochondrial dysfunction can lead to neuronal degeneration
via impaired production of ATP (through disruption
of the electron transport chain (ETC)), increased generation
of reactive oxygen species (ROS), altered calcium
homeostasis and excitotoxicity [Shigenaga et al., 1994,
Lenaz, 1998; Sastre et al., 2003; Reddy and Beal, 2008;
Hattingen et al., 2009; Bratic and Larsson, 2013; Yan
et al., 2013]. Brains of aged animals have significantly
higher levels of oxidized proteins and lipids, and reduced
protease activities compared to young animals [Keller
et al., 1997; Dei et al., 2002; Grimm et al., 2011].
Age-related functional loss is not limited to cognition
and physical activity. Aging is associated with rapid deterioration
of sensory and somatosensory functions including
vision, olfaction, and motor coordination [Wallace
et al., 1980; Cain and Stevens, 1989; Spear, 1993; Foster
et al., 1996; Bickford et al., 1999; Nakayasu et al., 2000;
Rawson, 2006]. While not inherently fatal these conditions
are associated with substantially elevated morbidity
and mortality [Schiffman et al., 1990; Struble and Clark,
1992; Ter Laak et al., 1994; Devanand et al., 2000; Girardi
et al., 2001; Schiffman et al., 2002; Tan et al., 2008;
Li et al., 2010; Baba et al., 2012; Doty, 2012], and
obvious impacts on quality of life.
Age-related deterioration of motor coordination is a
major cause of falls in the elderly, exacerbated by balance,
arthritis, and declining physical strength [Girardi
et al., 2001]. Olfaction is of unique clinical importance,
as olfactory deficits precede and accurately predict onset
of debilitating neurodegenerative conditions such as
dementia, Alzheimer’s and Parkinson’s [Struble and
Clark, 1992; Devanand et al., 2000; Schiffman et al.,
1990, 2002; Li et al., 2010; Ter Laak et al., 1994; Baba
et al., 2012; Doty, 2012].
Figure 1
|
Regardless of the current debate regarding the free radical
theory of aging, we predicted a priori that our transgenic
rat growth hormone mouse (TGM) would express
elevated free radical processes and that these would be
correlated to aging rates. Subsequent examination of lipid
peroxidation (LP) and superoxide radical (SO) found elevated
free radical processes in all tissues examined, particularly
the brain [see Rollo et al., 1996; Carlson et al.,
1999; Hauck and Bartke, 2001]. Levels of SO and LP
also increased with age and was highly correlated to longevity
[Rollo et al., 1996]. TGM are substantially larger
than normal mice (Figure 1A), have reduced lifespan (50%
of normal siblings), early mitochondrial deterioration,
accelerated kidney and liver disease, and early onset of
symptoms resembling normal murine aging (Figure 1B)
[Steger et al., 1993; Wolf et al., 1993; Meliska et al.,
1997; Ogueta et al., 2000]. These include reduced cellular
replicative potential, early reproductive senescence,
increasing tissue oxidative and nitrosative damage, arthritis,
reduced motor activity, cataracts, sarcopenia, kyphosis,
and altered fur quality [Bartke et al., 2002; Bartke,
2003; Lemon et al., 2005; Aksenov et al., 2010, 2013;
Long et al., 2012].
Young TGM (<7 mo.) have vastly superior cognition
than normal mice, learning an eight-choice radial maze in
roughly half the trials required by age-matched normal
mice (many of which did not learn at all) [Rollo et al.,
1999; Lemon et al., 2003]. However, cognitive abilities
of TGM rapidly deteriorated and by 11 months of age,
most were unable to learn the maze. The performance of
normal mice is essentially unchanged across the same age
range [Lemon et al., 2003; Long et al., 2012; Aksenov
et al., 2013]. Increasing levels of ROS, low-level inflammation,
impaired energy supply, loss of membrane fluidity
and mitochondrial dysregulation are all common
features of normal brain aging and neuropathologies
[Lyras et al., 1997; Mattson et al., 1999; Butterfield et al.,
2001; Mutlu-Turkoglu et al., 2003; Sastre et al., 2003;
Antier et al., 2004]. We hypothesized that an intervention
targeting multiple cellular processes might ameliorate premature
aging in TGM and associated pathologies. Specific
development and testing of multipurpose, multi-ingredient
supplements is lacking.
Most scientific studies examine
one or only a few ingredients at a time (usually antioxidants
or anti-inflammatories), which precludes obtaining
the benefits of synergistic or interactive effects which
may emerge in more complex formulations. We employed
a multiple ingredient dietary supplement (MDS) designed
to specifically target five critical processes associated
with aging [Lemon et al., 2003, Aksenov et al., 2010,
2013, Long et al., 2012]. Ingredients and dosages have
been described [Lemon et al., 2003, 2005]. The dietary
supplement completely abolished the age-related cognitive
decline in TGM. Remarkably, MDS supplemented
older TGM had significantly better maze performance
than younger control TGM and more than 2-fold faster
task learning than normal mice. TGM cognitive decline
was not only prevented, but augmented by the MDS. The
MDS supplement also benefited older normal mice (>24
mo.) [Lemon et al., 2003, Long et al., 2012, Aksenov
et al., 2013].
Increases in mean and maximal longevity
were also observed in supplemented TGM and normal
mice [Lemon et al., 2005]. Even in advanced ages, MDS
supplemented mice exhibited youthful cognitive and
motor abilities [Lemon et al., 2003, 2005, Aksenov et al.,
2010, 2013]. While intensity of physical activity was
reduced with age, the duration of daily locomotion was
unchanged from youth well into oldest ages (Aksenov
et al., 2010). Thus, mice appear to retain the capacity for
youthful cognitive functionality, and to a large degree,
youthful motor abilities into advanced ages. Improved
cognition and physical abilities in supplemented animals
were attributed to reduced oxidative and nitrosative damage
and enhanced mitochondrial activity in MDS treated
animals [Lemon et al., 2003; Aksenov et al., 2010; Long
et al., 2012].
In previous work, we showed that age-related cognitive
and motor declines could be prevented or delayed by dietary
supplementation. We postulated that, due to the
chronic oxidative stress experienced by TGM, increased
brain cell loss, likely through apoptosis, was a likely
mechanism underlying dramatic age-related cognitive
decline. Given that the MDS abolishes the early cognitive
decline in TGM, we speculated there would be a reduction
in apoptotic cell loss in older TGM on the dietary
supplement. In the present work, we also examined
whether benefits of the MDS extended to sensory and
behavioral function in aging mice. Mice were subjected
to a battery of tests to assess motor coordination, vision,
olfaction, emotionality and contextual discrimination.
Remarkably, MDS supplemented mice showed significant
improvements on virtually all aspects examined.
MATERIALS AND METHODS
Animals
Heterozygous TGM males were mated to normal (Nr) females yielding
equal numbers of TGM and Nr offspring of similar genetic background.
Heterozygote TGM were distinguishable from their normal
littermates by their larger size and phenotypic alterations at 28 d of age.
Mice were maintained in standard housing cages on a 12 : 12 h light:-
dark photoperiod at 22 ± 2°C. Food and water were supplied ad libitum.
Lifespan of our untreated Nr and TGM mice are consistent with published
lifespan of this mouse strain [Hauke et al., 2001; Bartke et al.,
2003; Lemon et al., 2008]. All procedures and protocols were approved
by McMaster University Animal Research Ethics Board and adhered to
the Canadian Council on Animal Care. Genders, genotypes, ages and
numbers of mice used in various studies are indicated as appropriate.
Multi-Ingredient Dietary Supplement
The MDS was designed to simultaneously ameliorate key processes
implicated in aging (oxidative stress, inflammatory processes, insulin resistance,
and membrane and mitochondrial deterioration). Materials were chosen
based on documented effectiveness for one or more of the targeted
features and could be safely taken orally. Dosages for the mice were reformulated
based on amounts commonly recommended for humans. Dosages
were adjusted for the smaller body size of the mice and increased by a factor
of 10 based on the higher gram-specific metabolic rate (and consequently
faster utilization and turnover) of mice compared to humans
[Calder, 1984]. Biological actions of individual components have been previously
summarized [Lemon et al., 2008]. The supplement was prepared in
liquid form and a 0.4ml volume was soaked into a 1 cm x 1.5 cm x 1 cm
piece of bagel and allowed to dry. Each mouse received 1 piece of dried
bagel with or without MDS, daily (midway through the photoperiod). The
bagel pieces were rapidly eaten by the mice within 20 min, ensuring mice
obtained full and equivalent doses. The formulation of the supplement has
been previously published [Lemon et al., 2003, 2005] and was maintained
for the duration of the study. At weaning mice were randomly assigned to
either MDS supplemented or untreated group. MDS mice were treated
daily from weaning throughout the lifespan of the animals.
Somatosensory Tests
The severity of age-related losses in motor coordination and overall
mobility of older TGM mice makes quantification of somatosensory deficits
difficult to delineate from impaired motor function [Lemon et al.,
2005], as such all somatosensory tests were conducted on Nr males and
females up to 2 years old.
Landing Response
Mice were held by the base of the tail and lowered onto a flat surface,
3cm short of contact. Mice with intact vestibular function extend
their forepaws in anticipation of landing.
Negative Geotaxis
Similar to the landing response, this test assesses vestibular function
[Szechtman, 1988]. Mice were placed on a 30° inclined plane facing
down. Mice with uncompromised vestibular function quickly orient their
body position to face upward.
Visual Placing
This test has been used previously to grossly assess vision [Chaudhry
et al., 2008]. Mice held by the base of the tail were lowered past a black
table edge, far enough to prevent vibrissae contact. Mice with intact
vision reach for the table edge with their front paws.
Visual Acuity
Based on the principle underlying the visual placing test, a more precise
assay for measuring visual acuity was developed. Instead of a single
visual cue, this test uses an array of increasingly more challenging visual
cues. Mice were suspended by the base of the tail and repeatedly lowered
past a black horizontal wire at a distance of 5cm (to avoid vibrissae
contact). If able to see the wire, mice reached towards it with the forepaws.
Lack of a reaching response on five consecutive attempts indicated
that mice were unable to see the wire. The next set of challenges
consisted of decreasing wire thickness, background contrast (i.e., white
or black background) and room lighting.
Specifically, mice were challenged in bright illumination with:
(a) 5 mm wire suspended over white background,
(b) 5 mm wire suspended over black background,
(c) 1 mm wire suspended over black background,
(d) 0.5 mm wire suspended over black background,
(e) 0.5 mm wire suspended over black background in dim light, and
(f) 0.5 mm wire suspended over black background in near darkness. Mice were tested in order of least to most challenging test with 30 min inter-trial intervals.
Olfactory Sensitivity
Test protocols were adopted from Witt et al., [2009]. Three serial
dilutions were prepared by diluting 1.25, 2.50, and 5.00 mg of peanut
butter in mineral oil to a final volume of 100 ml. A fourth blank dilution
(mineral oil only) was used as a negative control. Mice were placed
individually placed in a 28 3 16 3 12 cm plastic enclosure and acclimated
for 30 min. A camera was set up to record behavior. Just prior to
starting a trial, 1 ml of a peanut butter dilution was pipetted onto a 3 3
3 cm filter paper square. The filter paper was sealed inside a small Petri
dish with several round holes to allow dispersal of scent and placed in
the enclosure. Each mouse was tested on all four peanut butter dilutions
in a random order with 2–3 days in between trials. Olfactory sensitivity
was assessed by scoring the time spent exploring the Petri dish (e.g.,
sniffing, licking, biting) in a 10 min time interval. All behavior was
video recorded from above.
Pinch Re£ex
Mice were lightly pinched by the hind paw. Mice with normal pain
sensation and reflexes immediately withdraw the paw.
Behavioral Experiments
Behavioral experiments were performed on both Nr and TGM mice
with ages representing the lifespan of the both groups of mice.
Open Field
The bright open field test was previously applied to assess emotionality
[Chaudhry et al., 2008]. A square arena 70 cm (w)370 cm
(l)345 cm (h) constructed from white Plexiglas was illuminated by two
100 W white lightbulbs. A 50 3 50 cm square in the middle and the
10cm wide outside border constituted “central” and “peripheral” zones,
respectively. Mice were placed individually in the central zone and videotaped
from above for 5 min. Arena was cleaned with ethanol between
runs to remove any scent trails from previous trials. Image tracking software
(Noldus Ethovision®) was used to score variables:
(a) latency to exit central zone,
(b) distance traveled in central zone
(c) distance traveled in peripheral zone, and
(d) mean running velocity.
Step-Down Test
Mice were placed individually on top of a circular 10cm diameter
platform, 7 cm high. Initially, trials were conducted in bright illumination
(two 100 W white lights), the test was repeated in dim lighting (single
25 W red light), with one week between tests. Latency to step down
was recorded with a 10-min time limit.
Circle Run
This test was developed in our lab and was used to assess emotionality
in mice [Chaudhry et al., 2008]. Mice have an aversive response to
bright, open areas, eliciting anxiety-like behavior; whereas dim illumination
allows for observation of active or exploratory behavior [Trullas
and Skolnick, 1993; Chaudhry et al., 2008]. A circular white arena 1 m
in diameter was surrounded by a shaded overhang 5 cm high and 20 cm
wide. A 7 cm diameter, 5 cm deep depression was fitted in the center of
the arena. Each mouse was placed in the depression and videotaped
from above, latency to climb out of cup with four paws, followed by
latency to reach the shaded overhang were scored. Mice were tested on
two separate trials, two weeks apart, first in bright, and then in dim
illumination.
Rotarod
This standard test assesses motor coordination and balance [Carter
et al., 2001]. Mice were individually placed on a horizontal 6cm diameter
plastic cylinder, rotating at 12 rev/min. Latency to fall onto a cushioned
landing pad was scored on three consecutive trails. Improvement
was assessed by subtracting latency to fall on first trial from the latency
on the last trial.
Cerebellum, Olfactory Bulb, and Retinal Histology
In mid-photophase, untreated (n = 6) and MDS supplemented (n = 6)
mice, aged 14–17 months were decapitated and brains and eyes were
removed. Cerebellum and olfactory bulbs were separated. All tissues
were placed in 10% formalin solution, processed overnight and embedded
onto paraffin blocks. Tissues were sliced at 5 lm on a microtome
and stained with H&E and Nissl stain.
Cerebellum
Thickness of molecular layer and granule cell layer was assessed in
Lobules II and III using ImageJ software. Number of Purkinje cells in
Lobules II and III were counted. Some studies estimated total cerebellar
cellularity by interpolation of counts from an array of sections spanning
the entire width of the cerebellum [Woodruff-Pak, 2006]. However, we
were only interested in the relative difference between treatment groups;
hence, a single representative inter-hemispheric sagittal section was sufficient
[Rogers et al., 1984].
Retina
Mid-sagittal sections of eye were used to measure the thickness of
the outer nuclear layer and the outer fragment layer in retina. Agerelated
loss of photoreceptors is most prominent in the central retinal
portions; therefore, measurements were performed on retinal crosssections 1,000 µm inferior and superior to the position of optical nerve
attachment.
Olfactory Bulb
One olfactory bulb from each animal was randomly chosen. Thickness
of the glomerular layer, external plexiform layer and counts of
mitral cells were performed. Strict criteria were employed in the identification
of mitral cells in the olfactory bulb (refer to Fig. 13 caption).
Brain Collection and Preparation for Apoptosis, Cell Counts
Fifty-two mice were used for brain weight assessment. Mice were
assigned to one of 16 experimental groups. The mice were divided by
the following criteria (n = 3-4/group); age: consisting of mice aged 3-4
months or mice aged 11-12 months (old based on TGM lifespan of 12
months, but middle-aged for normal mice), genotype: TGM or normal
isogenic control mice, and treatment: standard rodent chow or standard
chow plus MDS. The brains used for the brain weight study were harvested
and immediately weighed on a precision balance (AB204-S, Mettler
Toledo, Mississauga ON), placed in a cryovial (Nalge Nunc
International, Rochester NY), flash frozen in liquid nitrogen and stored
at –80°C. To determine if the apparent weight discrepancy in old
untreated TGM was due to cell loss by increased apoptosis, the brains of
48 mice were harvested for apoptosis and brain cell density study, n = 3
in the same 16 experimental groups. The brains were harvested, cut in
half along the longitudinal axis and placed in 10% neutral buffered formalin
within 2 min of dissection. The brain tissue was left in the fixative
and stored at room temperature for a minimum of one week prior to
embedding in paraffin. The brains were sectioned sagittally from the longitudinal
cerebral fissure in 10 µm thick sections using a rotary microtome
(Riechert Scientific Instruments, Buffalo NY). Sequential tissue
slices were placed on microscope slides, from one side of the longitudinal
axis for each mouse covering lateral 0.15 to 0.85.
Apoptosis and Cell Density
Slides were prepared following the protocol provided in the
ApoptagVR Fluorescein kit (Serologicals, Temecula CA). Briefly, the tissue
sections were deparaffinized using xylene, followed by an ethanol
rehydration and 1 wash in room temperature phosphate buffered solution
(PBS; 137 mM NaCl, 2.7 mM KCl, 4.3 mM Na2HPO4, 1.4 mM
KH2PO4). The tissue was then pre-treated with 20mg/ml proteinase K
(Sigma-Aldrich, Oakville ON) for 10 min at room temperature, followed
by a wash in 2 changes of PBS. Excess liquid was removed from the tissue
sections by blotting and an equilibration buffer (supplied) was
applied directly to the slide and incubated for 30 sec. The excess liquid
was removed and the terminal deoxynucleotidyl transferase enzyme with
nucleotides and a digoxigenated cytosine (supplied) was applied directly
to the tissue and covered with a plastic coverslip.
The slides were incubated in a humidified chamber for 1 h at 37°C.
Slides were then immersed in stop/wash buffer (supplied) for 10 min
and washed with 3 changes of PBS. The excess liquid was removed and
anti-digoxigenin-fluorescein (FITC) conjugate was applied to the tissue
sections, covered with a plastic coverslip and incubated at room temperature
in a humidified chamber for 30 min. Slides were washed in 4
changes of PBS, the excess moisture was removed and the slides were
allowed to air dry. The tissue sections were counterstained with 0.4µg/
ml 4’,6-diamidino-2-phenylindole (DAPI), a fluorochrome that binds to
double-stranded DNA, in AntifadeVR (Serologicals, Temecula CA) and
mounted under a glass coverslip. The slides were analyzed on the Zeiss
Axiophot 2 image analysis system, using a 63x oil immersion objective
(total magnification: 630X). Slides were scored using a dual bandpass
filter designed to simultaneously view FITC (apoptotic DNA) and DAPI
(total DNA). Apoptotic nuclei were identified as positive for both the
DAPI counter-stain and FITC with the dual bandpass filter. Apoptotic
cells were confirmed by viewing the positive cells individually with separate
FITC (ex. 490nm & em. 520nm) and DAPI (ex. 365nm & em
480nm) filters. Apoptotic positive cells were also confirmed morphologically
having a nucleus that was homogenously staining and crescent
shaped and overall smaller cell size.
The cell density in the brains of TGM and age-matched normal mice
was scored by manually counting cell nuclei stained with DAPI. The
slides were counted at a magnification of 630X to aid in the morphological
identification of nuclei. Total cell numbers per field of view were
counted for the entire tissue section.
SPECT and PET Imaging
Twelve male mice ages 11–12 months were used for imaging in
four experimental groups; untreated Nr, untreated TGM, MDS-Nr,
MDS-TGM (n = 3) Mice used for SPECT imaging were injected via tail
vein with approximately 15 MBq 99mTc-HMPAO. The compound was
allowed to circulate for 15 min, at which time the animals were sacrificed
and imaged overnight to maximize image resolution. SPECT scans
were acquired on an X-SPECT system (Gamma Medica, Northridge,
CA) using dual sodium iodide crystals in combination with low energy
pinhole collimators with 1 mm aperture and a radius of rotation of
3.5 cm. The SPECT scan consisted of thirty-two 15-min projections and
was followed immediately by the collection of four rotations of 1024 Xray
projections for CT, also acquired on the X-SPECT system with Xray
tube characteristics of 75 kVp and 220 µA.
Mice used for PET imaging were injected with 11-14 MBq 18F-FDG
via the tail vein. The compound was allowed to circulate for 60 min, at
which time, the animals were sacrificed and imaged overnight to maximize
image resolution. PET scans were acquired on a MOSAIC small
animal PET scanner (Philips, Andover MA). PET images were acquired
over an 8-hr period, followed immediately by the collection of four rotations
of 1024 X-ray projections for CT, acquired on the X-SPECT system.
All imaging work was completed at the McMaster Centre for
Preclinical and Translational Imaging (MCPTI) at McMaster University
(Hamilton, ON, Canada).
Statistical Analyses
Statistical tests applied are described in corresponding figure and
table captions. Briefly, where age-ranges were available, effects of treatment,
genotype, gender, and other independent variables were first
assessed with ANCOVA (covariate = age). If main effects or interaction
effects were not resolved for a given predictor variable, groups (or ages)
were pooled. Effects of the remaining variables were discriminated with
post-hoc SNK or Duncan’s tests. In the absence of other predictor variables,
effects of treatment were resolved with a t-test. A two-tailed (Fisher’s)
chi-square test was used when comparing number of animals
eliciting positive responses in each treatment group on behavioral tests
of visual acuity and somatosensory function. Where applicable, agerelated
effects were analysed described with linear regressions.
RESULTS
Three-month-old Nr and TGM mice (Fig. 1A), illustrate
the difference in adult body size in the mice. Average
body masses for female Nr and TGM mice at 4
months of age were 30.1 ± 2.36 g and 46.8 ± 0.33g
respectively [Lemon et al., 2003]. Twelve-month old control
and MDS supplemented TGM were shown in Figure
1B to illustrate the dramatic improvement in body
condition of MDS supplemented TGM compared to agematched
untreated TGM (coat colour has been previously
determined as an irrelevant variable).
Behavioral Analysis
Table I
Figure 2
Table II
Figure 3
Table III
Figure 4
Figure 5
Table IV
|
Open Field
Nr and TGM mice aged between 4–18 months were
scored on four parameters in a brightly illuminated open
field (n = 56); (a) Latency to exit the central zone was
not affected by age, genotype or treatment (data not
shown). (b) Distance traveled in peripheral zone was
affected by age and genotype but not treatment (Table I).
Oldest mice (15–18 months) traveled ~15% less (data not
shown), and TGM covered ~40% less distance than Nr
mice (Figure 2). Slightly greater distances traveled by supplemented
mice were not significantly resolved (Fig. 2).
All mice traveled greater distances in peripheral compared
central zone (Fig. 2). (c) Distance traveled in central
zone was reduced across age-range (~30% from 4 to
18 months; Table I). A significant effect of genotype was
resolved, but this reflected a genotype*treatment interaction
mainly associated with changes in Nr behavior
(Table I). MDS Nr mice traveled more than double the
distance covered by untreated Nr (P<0.003; Fig. 2). Supplementation
had no effect on distance traveled in the
central zone in TGM mice (Fig. 2). (d) Mean running
velocity was not affected by treatment (Table I). TGM
were slower compared to Nr mice by 24% and mean running
velocity was negatively correlated with age (Table I). Parameters (a–d) showed significant age-related regressions
when genotypes and treatments were pooled (Table I). For each parameter, we assessed whether MDS supplemented
and untreated Nr and TGM mice resolved different
slopes for age-related regressions. Differences in
slopes were not resolved with ANCOVA (Table I) indicating
that impacts of age were similar for all groups and
were unaffected by genotype or treatment.
Step-Down Test
Latency to step down from an elevated platform was
first measured under bright illumination, and following a
rest period, a repeat assessment was performed in neardarkness.
A small number of animals used in the dark trials
were different than those tested in bright illumination.
Latency to step down was significantly affected by MDS
treatment and level of illumination; however age and
genotype did not significantly influence latency (Table II).
To increase statistical power, ages and genotypes were
pooled. For untreated animals, the latency to step down
did not significantly differ between bright (n = 67) and
dark (n = 39) illumination (Figure 3). However, in supplemented
mice differences between bright (n = ) and
dark (n = 21) illumination resulted in a significant 38%
decrease in step-down latency (Fig. 3). In bright conditions,
supplemented mice showed a non-significant
decrease (P>0.151) in latency compared to untreated
controls. Latency to step down in dim illumination was
significantly reduced in MDS supplemented mice
(P<0.038; Fig. 3).
Circle Run
Effects of age, genotype, treatment, and illumination
on latency to (a) emerge from cup and (b) escape to
shade are presented in Table III. The same group of mice
was tested in bright and dark conditions, with one week
between trials. Effects of age and genotype were not significantly
resolved (Table III), to increase statistical
power Nr and TGM mice were combined across ages. (a)
Latency to emerge from cup. Untreated mice took the
same amount of time to emerge from the cup regardless
of illumination (Figure 4). Compared to untreated mice,
MDS supplemented mice took 34% longer (P<0.0001)
to emerge in bright illumination (Fig. 4). Alternatively,
when tested in near-darkness, supplemented mice
emerged from the cup significantly faster (P<0.0001)
than in bright lighting (Fig. 4). (b) Latency to escape to
shade. Latency to escape to shade was unchanged in
untreated mice regardless of level of illumination (Fig. 4).
In bright conditions, MDS mice spent 17% (P<0.048)
longer in the open zone before entering the escape overhang
compared to controls. However, when tested in dim
illumination, escape time of MDS mice was 18% faster
relative to controls (P<0.019; Fig. 4). This result was
also demonstrated in the significantly different escape
latency of MDS mice going from bright to dark illumination
(P<0.006; Fig. 4).
Rotarod
Latency to fall off a rotating cylinder was measured on
three consecutive trials. Mean (3–trial average) latency
was nearly 4 times lower in TGM compared to Nr mice;
but effects of treatment on mean latency were unresolved
in either genotype (Figure 5A). A key aspect of this test,
however, is rate of improvement. Improvement was
scored by subtracting latency to fall on the first trial from
that on the last attempt. Improvement was not affected by
genotype (P>0.390); hence we combined Nr and TGM
mice for better statistical power (Fig. 5B). Untreated mice
Simple Somatosensory Tests
Results of the (a) landing response, (b) visual placing,
(c) negative geotaxis, and (d) pinch reflex are summarized
for Nr (~2-year-old) mice (Table IV). Virtually all mice
showed normal function in these simple somatosensory
tests and further improvement from supplementation was
not possible (Table IV).
Visual Acuity
Table V
|
A newly developed protocol was used to assess visual
acuity of 1.5- to 2-year-old Nr mice. Animals were challenged
with six visual cues of increasing difficulty. A
reaching response indicated that mice were able to see
the cue. All mice showed a reaching response on the least
difficult challenge (#1; Table V). As difficulty increased
(challenges: #2–5), progressively fewer mice successfully
located the visual cue. The number of animals maintaining
positive responses was consistently greater in the
MDS group (Table V). On challenge #4, 63% of MDS
mice displayed a reaching response compared to 23% in
untreated mice (P<0.014; Table V). A greater than 60%
drop-off in reaching responses occurred for untreated
mice between challenge #3 to #4.
In MDS mice, a comparable drop was observed, however
it occurred between challenges #4 to #5. No mice
showed reaching responses on challenge #6 (Table V)
which employed nearly complete darkness, providing a
negative control, confirming that responses on the easier
challenges were indeed due to visual acuity.
Olfactory Sensitivity
Figure 6
|
Olfactory sensitivity was measured by quantifying the
duration of exploration in response to varying concentrations
of an attractive scent (peanut butter). Both supplemented
and untreated mice spent little time (~130 sec)
investigating filter papers with zero or low (0.125 mg/10
ml) concentrations of peanut butter. When presented with
higher (0.250 mg/10 ml, and 0.500 mg/10 ml) concentrations,
supplemented mice increased exploratory behavior
duration by 68% and 105%, respectively. Plotting the
mean duration of exploratory behavior (Y-axis) over peanut
butter dilution (X-axis) for supplemented mice
returned a strong positive correlation (i.e., dose response)
(r2 = 0.976; P<0.012; Figure 6). Conversely, untreated
mice showed nearly no change (<10% increase) in exploration
going from zero to high peanut butter concentrations
(r2 = 0.069; P>0.738; Fig. 6). Slopes of regressions
significantly differed for treatments (ANCOVA:
P<0.036).
Brain Cell Number and Weight
Figure 7
Figure 8
|
There is significant difference in brain cell density in
12-month-old untreated TGM (Figure 7A) and MDS TGM
(Fig. 7B). Brain cell number was determined by manually
counting the number of DAPI stained nuclei in each brain
slice at 630X magnification (same tissue sections as those
used for the Apoptag® assay), and totalling the number
of cells from all brain sections. There was no significant
difference in total brain cell number between 3-month-old
untreated and MDS Nr mice (94,646.7 ± 9,098.4 and
100,816.0 ± 7,944.6, respectively, P>0.636; Figure 8A).
Three-month-old untreated Nr mice had 77.4% of the
brain cells found in age-matched untreated TGM. Threemonth
old TGM also demonstrated no significant differences
in brain cell number when MDS supplemented animals
(118,194.7 ± 6,337.2) were compared to control
mice (122,358.7 ± 2,553.3; P>0.575). Control and MDS
12-month-old Nr mice were not significantly different
with respect to brain cell numbers with mean values of
78,545.3 ± 7,165.2 and 85,312 ± 6,013.1, respectively
(P>0.509). Untreated 12-month old TGM had a mean of
53,513.362,151.9 brain cells, significantly lower than
age-matched MDS TGM with mean brain cell number of
120,514.6 ± 1,570.7 (P<0.00001). 12-month old MDS
TGM did not differ significantly in brain cell number
from either 3-month-old untreated or MDS supplemented
TGM (P>0.740 and P>0.571, respectively). Analysis of
variance (ANOVA) detected a significant impact of genotype
on mean brain weight (P<0.000001), no other significant
main effect was detected for any other variable.
ANOVA did reveal a strong interactive effect between
genotype and supplement (P<0.06) and between genotype
and age (P<0.05), with only the latter reaching significance.
Analysis of mouse brain weight (Fig. 8B)
demonstrates that 3-month-old Nr mouse brains were only
57.3% (483 ± 34 mg) of the mass of young TGM controls
(843 ± 79 mg; P<0.0100). Three-month old supplemented
Nr (487 ± 40 mg) and supplemented TGM (851 ± 47 mg) do not differ from their controls in brain weight
(P<0.942 and P<0.934, respectively).
Three-month old control TGM had significantly greater
brain mass than 12-month-old control TGM at 541 ± 67mg (P<0.011) and 12-month-old Nr mice at
534 ± 73 mg (P<0.013). Twelve-month old control TGM brain
mass was 64.2% of young control TGM brains. Twelvemonth
old control TGM did not differ significantly in
brain weight from either age-matched control Nr
(P>0.949) or 3-month old control Nr mice (P>0.590).
The brain weight of 12-month-old supplemented TGM
(873 ± 51 mg) was significantly greater than 12-monthold
control TGM (P<0.003), but did not differ significantly
from 3-month-old control TGM (P>0.770) or supplemented
TGM (P>0.763). Twelve-month-old
supplemented Nr mice (522 ± 38 mg) did not differ from
age-matched control Nr mice in brain weight (P>0.896).
Apoptosis
Apoptotic cells were identified with Apoptag®, a modified
TUNEL assay, used to detect fragmented DNA associated
with apoptotic nuclei. Since the entirety of each
brain slice was scored for apoptotic cells, the total number
of cells scored varied for each experimental group
(Fig. 8C), apoptotic cells are identified as a percentage of
the total number of cells in the brain slices. The results
from all groups (3-month-old and 12-month old, control
and supplemented) showed that only 3-month-old control
Nr mice had an elevated level of apoptosis (P<0.05;
Fig. 8C). Although the level of apoptosis in young Nr
controls is significantly elevated compared to all other
groups it was still below 0.1%, so the biological significance
of this finding is undetermined.
Figure 9
Figure 10
Figure 11
Figure 12
Figure 13
Figure 14
Figure 15
Figure 16
|
Tissue Morphology
Cerebellum Histology
Purkinje cells (PC) in inter-hemispheric sagittal cerebellar
sections of lobules II and III (Figure 9A) were
counted in Nr mice aged 14–17 months (narrow agerange
precluded analysis of age-related trends; supplemented:
n = 6; untreated: n = 6). PC are restricted to the
Purkinje cell layer (Figure 9B) and length of the Purkinje
cell line were accounted for when counting cells [Rogers
et al., 1984; Woodruff-Pak, 2006]. Supplemented mice
showed a 20% increase in PC number per mm of cell
line (P<0.05; Figure 10A). Figure 9D shows a close up of
PC from supplemented and untreated mice dyed with
Nissl stain. PC in a sample from an untreated mouse
show visibly reduced PC number compared to supplemented
mice. Average thickness of the molecular layer
(ML) and granular layer (GL) in Lobules II and III were
calculated using ImageJ software according to schematic
shown in Figure 10C. Supplemented mice had a 16%
(P<0.05; Fig. 10B) and 18% (P<0.01; Fig. 10C)
increase in thickness of the ML and GL, respectively,
compared to untreated mice. Cerebella were collected
from three male and three female mice in each treatment
group. An effect of gender or gender*treatment interactive
effects were not resolved.
Retina Histology
Mid-sagittal sections of eyes were stained with H&E in
supplemented (n = 4) and untreated (n = 4) 2-year-old Nr
females (Figure 11A). In supplemented mice, thickness of
ONL and OS averaged over the length of proximal retinal
sections (Fig. 11) were 26% and 29% greater, respectively,
compared to controls (Figure 12). Close-ups of
retinal sections from supplemented and control mice
showed visible differences (Figs. 11B–11E).
Olfactory Bulb Histology
Cross-sections of olfactory bulbs were obtained from
supplemented (n = 6) and untreated (n = 5) mice (ages:
14–17 months). ImageJ was used to calculate mean thickness
of glomerular layer (GLM) and external plexiform
layer (EPL) (organization of neuronal layers shown in: Figure 13A). GLM and EPL were significantly reduced in supplemented
mice by 25% (P<0.026; Figure 14B) and 28%
(P<0.001; Fig. 14C) respectively, compared to controls.
Mitral Cell Layer
For counts of mitral cells strict criteria were applied
including cell size, shape, position, and visibility of
clearly defined nuclear envelope and nucleolus (Figs. 13B
and 13C). Although this underestimates total cell numbers
it provides maximal accuracy for comparisons between
samples. Mitral cell counts in a 1 mm representative section
were 29% higher in supplemented mice compared to
controls (P<0.030; Fig. 14A). Despite this significant
resolution, greater samples numbers, more representative
sections and additional staining techniques could reinforce
these preliminary findings.
Functional Brain Imaging
18F-FDG uptake correlates to glucose utilization, which
provided an indicator of brain metabolic rate (Figure 16A).
Unsupplemented 12-month-old TGM showed a significant
reduction in metabolic rate compared to age-matched Nr
mice (P<0.027; Fig. 16A). Supplemented TGM had significantly
higher brain metabolic activity than untreated
TGM (P<0.013; Fig. 16A), which did not differ from
control or supplemented Nr mice (P>0.087 and
P>0.0.385, respectively; Fig. 16A). Dietary supplementation
had little effect on the brain metabolic rate of 12-
month-old Nr mice (P>0.210; Fig. 16A).
99mTc-HMPAO binds to neutrophils in the peripheral
blood, providing a biomarker of blood perfusion through tissue
and an important indicator of general brain function.
Normal mice did not differ in brain perfusion between treatments
(P>0.454; Fig. 16B), however, both groups had significantly
higher activity compared to unsupplemented 12-
month-old TGM (P<0.047 and P<0.023, respectively; Fig.
16B). Untreated TGM had significantly reduced blood perfusion
compared to all groups. The most striking difference
was in supplemented TGM with a 2-fold increase in perfusion
compared to untreated TGM (P<0.034; Fig. 16B).
DISCUSSION
Overall Brain Function
It has been established that TGM experience significant
oxidative stress and inflammatory processes that result in
an accelerated aging phenotype, with dramatic negative
consequences on cognitive and motor function [Rollo
et al., 1996, Lemon et al., 2003, 2005, Aksenov et al.,
2010, 2013, Long et al., 2012]. Severe cognitive decline
was abolished by the MDS supplement formulated to target
key physiological mechanisms associated with aging
[Lemon et al., 2003, 2005, Aksenov et al., 2010, 2013,
Long et al., 2012]. Here we show that there is a concomitant
loss in brain cells, deterioration in sensory function
and reductions in cerebral metabolic rate and blood perfusion
in old TGM. Together these strongly contribute to
age-related cognitive impairment. Brain deterioration was
equivalent to advanced Alzheimer’s disease in humans,
with >50% losses on a cellular level, 36% loss in brain
mass and at least 2-fold reductions in metabolism and
blood flow. Not only did the MDS restore cognitive function
in old TGM, but brain cell density and brain mass
were maintained at levels comparable to young mice. To
our knowledge, this is the first dietary supplement to
abolish such severe loss of brain cells and maintain cognitive
function with such efficacy. The mechanisms yielding
such benefits are not fully elucidated; however
previous research has verified MDS efficacy on improvements
in all targeted mechanisms [Lemon et al., 2003,
2005; Aksenov et al., 2010, 2013; Long et al., 2012].
The complete amelioration of brain cell loss in supplemented
old TGM suggests that the mechanisms targeted
by the supplement (reactive oxygen species, inflammation,
maintenance of cell membranes, insulin sensitivity,
and mitochondrial function) were well chosen. Long et al.
[2012] added nitrosative stress as another factor modulated
by the MDS which is also highlighted as a factor in
neurodegeneration [Nakamura and Lipton, 2007; Maurya
et al., 2016]. Notably, one of the most acellular regions
found in the brains of old TGM corresponds to a region
(Lateral 0.675mm; Bregma 20.25 mm to 20.65 mm, and
4.00 mm24.50 mm) encompassing the bed nuclei of the
stria terminalis (BST) and reticular nucleus of the thalamus
(RT). Activity within the BST correlates with anxiety
in response to threat monitoring, acting as a relay site
to regulate hypothalamic-pituitary-adrenal axis activity in
response to acute stress [Choi et al., 2007]. The RT
nucleus receives input from the cerebral cortex and dorsal
thalamic nuclei. Most input comes from collaterals of
fibers passing through the thalamic reticular nucleus.
Primary thalamic reticular nucleus efferent fibers project
to dorsal thalamic nuclei, modulating information from
other thalamic nuclei, playing a role in disinhibition of
thalamic cells, an essential function for initiation of
movement [Beierlein, 2014; Mitchel, 2015]. Protection of
these brain regions may provide clues to the anxiolytic
effect and improved motor function of MDS treated aging
animals. Additionally, the apparent heterogeneity of cell
loss indicates that there may be differential losses among
brain regions and neurotransmitter systems in TGM, however
further study will be required to confirm.
Untreated young normal mice had significantly higher
brain cell apoptosis, compared to other experimental
groups, although their average apoptosis was less than
0.1%. Apoptosis is a normal part of brain development
which continues for several months after birth in mice.
One mechanism of obtaining larger brains in IGF-1 mice
is reduced apoptosis [Popkin et al., 2004], this could also
apply to TGM. Old unsupplemented normal mice do not
experience significant brain cell loss. A lack of significant
apoptosis in old non-supplemented TGM brains was
unexpected, given the chronically elevated endogenous
production of ROS characteristic of TGM, and the vulnerability
of the brain to oxidative damage, however it is
possible that the most susceptible cells had already been
lost at the advanced age of the mice used in this study.
Alternatively, cells may have deteriorated rapidly via
necrosis. All groups expressed very low levels of apoptotic
cells, but these samples represent a single moment in
time, low numbers of cell loss at constant rates still have
large impacts across normal developmental and aging
time scales. Brain cells from young animals are also more
resistant to oxidative stress and excitotoxicity compared
to adult animals [Ikeyama et al., 2002, Savory et al.,
1999]. When exposed to glutamate, mitochondria in the
brain slices of adult rats showed greater changes in ATP/
ADP ratio, NAD/NADH ratio and ROS formation compared
to young rats [Kannurpatti et al., 2004].
The ‘Old’
normal mice in our study were age-matched to TGM, but
were middle-aged, based on average life-span of a normal
mouse in our breeding colony. Consequently it would be
unlikely that they would show increased brain apoptosis
or significant changes brain weight or cell number. Interestingly
there were non-significant increases in both metabolic
rate and blood perfusion, which may indicate
beneficial effects of the supplement in normal mice that
could be resolved given a larger sample size. Examining
the brains from senescent Nr mice in both treatment
groups demonstrated that the supplement provides similar
protective effects for normal aging brain [Aksenov et al.,
2010, 2013]. There are a few probable explanations for
the extremely low cell numbers and the apparent lack of
apoptosis in old control TGM. Apoptotic cell loss is
likely a significant process in untreated TGM, but if
occurring at a slightly elevated homeostatic rate, would
contribute to minimal differences at any one time point,
but causing significant quantities of cell loss over the lifetime
of the animal, with the concomitant possibility of
accelerated cell loss as the animal approaches senescence.
Alternatively, and less likely, there could have been a
short term, large apoptotic event during middle age, when
enough oxidative damage had accumulated in the brain to
trigger the apoptotic response.
This event may coincide
with the age-related escalation in the cognitive decline
observed in control TGM starting at 7 months of age
[Lemon et al., 2003; Aksenov et al., 2013]. Sampling
TGM brains late in senescence may have been too late to
observe the mechanisms of cell loss. By 12 months of
age, control TGM had only 44% of brain cells remaining
compared to old diet supplemented TGM and young
TGM (both control and supplemented). Harvesting brains
of non-supplemented TGM at across the age spectrum
may clarify the mechanism and rate of cell loss. Additional
analysis on the brain tissue has provided some
clarity in the mechanisms behind the profound protective
effect of this dietary supplement [Aksenov et al., 2010,
2013]; however further studies will determine if there is
differential cell loss in the populations of cells that comprise
the brain. It will also be important to determine if
specific neurotransmitter pathways are differentially
affected by the substantial cell loss in control TGM.
Motor Coordination
Aging impacts motor behavior and coordinated control
of motor function (i.e., balance) [Wallace et al., 1980;
Foster et al., 1996; Bickford et al., 1999; Aksenov et al.,
2010]. When motor coordination was tested, 12-month
old untreated TGM mice had severely compromised
motor coordination compared to age-matched untreated
Nr, consistent with previous reports [Long et al., 2012].
Initial balance performance of Nr mice on the rotarod
was not affected by MDS treatment; however MDS supplemented
animals showed significant improvement in
performance after three trials. Untreated Nr mice did not
demonstrate such improvements, suggesting there may be
a learned component to the motor coordination testing
from which the MDS mice were able to benefit.
Reductions in locomotor abilities are closely linked to
oxidative stress [Foster et al., 1996; Aksenov et al., 2010]
and may reflect reduced mitochondrial activity [Aksenov
et al., 2010]. Motor coordination has been negatively correlated
with increasing levels of protein carbonyls in motor
control regions of the brain including the cerebellum [Foster
et al., 1996; Bickford et al., 1999]. MDS treatment significantly
reduced protein carbonyls in brains of aging
mice and was associated with improved locomotor behavior
in MDS supplemented Nr and TGM [Aksenov et al.,
2010]. Brain mitochondrial and neurotransmitter functions
were also augmented [Aksenov et al., 2010, 2013], with
this study indicating that positive impacts of MDS on brain
physiology manifest in better motor coordination.
While the cerebellum does not initiate movement, it is
critically responsible for control, coordination and correction
of body movement including sensory analysis of consequences
of movement [Paulin 1993]. Animals with
cerebellar defects show impaired motor coordination
[Kashiwabuchi et al., 1995, Colucci-Guyon et al., 1999;
Barski et al., 2003; Weimer et al., 2009]. The cerebellum
has also been highlighted in motor and spatial learning
[Dahhaoui et al., 1992, Le Marec et al., 1997, Bickford
et al., 1999, Ito 2000, Wulff et al., 2009]. Purkinje cells
(PC) are the dominant neurons involved in cerebellar
information processing [Rogers et al., 1984]. They are
one of the largest neurons in the mammalian brain with
very intricate dendritic projections and great numbers of
dendritic spines. PC provide the sole output pathway of
the cerebellar cortex [Barski et al., 2000; Ito 2000] and
participate in motor coordination [Barski et al., 2000,
2003] and motor learning [Bickford et al., 1999, Ito 2000,
Woodruff-Pak 2006, Barski et al., 2000, Wulff et al.,
2009]. Loss of PC can result in impairment of coordinated
movement [Kashiwabuchi et al., 1995, Sakaguchi
et al., 1996, Barski et al., 2003, Weimer et al., 2009],
decreased cerebellum-dependent learning, sensory processing
and other cerebellum-associated behaviors [Barski
et al., 2003; Woodruff-Pak 2006, 2010]. PC are highly
vulnerable neurons [Woodruff-Pak 2006] and show significant
decreases in normal aging in humans and rodents
[Rogers et al., 1984; Doulazmi et al., 1999; Larsen et al.,
2000; Andersen et al., 2003; Woodruff-Pak 2006]. The
cerebellum has been shown to express markers of senescence
earlier than other brain regions, with loss or dysfunction
of PC identified as a primary contributor to this
phenomenon [Woodruff-Pak, 2010].
Beyond the role of the cerebellum in motor coordination,
spatial processing and motor learning [Dahhaoui
et al., 1992; Paulin, 1993; Kashiwabuchi et al., 1995;
Sakaguchi et al., 1996; Bickford et al., 1999; Colucci-
Guyon et al., 1999; Ito, 2000; Barski et al., 2000, 2003;
Woodruff, 2006; Weimer et al., 2009; Wulff et al., 2009;
Woodruff et al., 2010], this brain region is also implicated
in perception of time [Monfort et al., 1998], autonomic
functions [Tong et al., 1993; Ghelarducci et al.,
1996] and emotional behavior [Bobee et al., 2000;
Schmahmann and Caplan, 2006; Balaban, 2002; Schutter
and Van Honk, 2005]. Several reports have documented
co-morbidity of emotional and balance disorders [Balaban,
2002; Balaban et al., 2011]. In recent years, the role
of the cerebellum in modulation of anxiety has been highlighted
[Bobee et al., 2000; Balaban, 2002; Schutter and
Van Honk, 2005; Schmahmann and Caplan, 2006; Secchetti
et al., 2009], prompted largely by the discovery
neuronal connections between the cerebellum and other
regions of the brain involved in emotional control [Bobee
et al., 2000; Balaban, 2002; Balaban et al., 2011].
Reported decades ago, abnormal cerebellar development
or degeneration of cerebellar neurons produced nervous
or anxious phenotypes in rodents [Sidman and Green,
1970; Landis, 1973]. To a large degree, this was attributed
to loss or dysfunction of PC [Sidman and Green,
1970; Landis, 1973]. Attenuation of age-related PC loss
in supplemented mice suggests that neurons involved in
modulation of anxiety-related behaviors should also be
intact. This is presently supported by behavioral testing
and may constitute an additional anxiolytic mechanism of
the MDS.
Granule cells are the smallest and most numerous neurons
in the mammalian brain [D’Angelo and De Zeeuw,
2009], but relatively little is known about their specific
roles. Granule cells and their synapses possess nonlinear
transmission properties and are connected to allow them
to operate complex transformations of input signals in the
spatiotemporal domain including forms of long-term
potentiation and depression. These characteristics may
provide a significant impact on cerebellar learning abilities
and computational power (D’Angelo 2011). Genetic
mutations affecting granule cell development are associated
with poor cognitive function and motor coordination
[Schiffmann et al., 1999; Weimer et al., 2009]. However,
isolating the effects of granule cell depletion is problematic
as this is usually coupled with loss of neighboring
neurons (e.g. Purkinje cells) [Le Marec et al., 1997;
Schiffmann et al., 1999; Weimer et al., 2009], which can
confound results. A recently proposed hypothesis argues
that cerebellar granule cells compute stimuli from adjacent
neurons to generate operational time-windows which
set a temporal framework for integrating sensory information
with motor domains [D’Angelo and De Zeeuw
2009]. This may also set the tone for long-term potentiation
(LTP), effecting learning capacities.
Thickness (or
volume) of the granule cell layer (GL) reflects the number
of granule cells [Larsen et al., 2000, Weimer et al.].
Granule cells send parallel fibers extending to the molecular
layer (ML) where they interact with the dendritic
arbors of Purkinje cells (PC) [Apps and Garwicz, 2005].
The ML also contains interneurons (stellate and basket
cells) that provide GABAergic input to PC [Apps and
Garwicz, 2005]. The ML interneurons play a role in sensory
information processing and motor coordination
[Apps and Garwicz 2005, Chu et al., 2012]. Aging rats
lost 60% of the parallel fiber length and up to 80% of PC
synapses [Huang et al., 1999], reflected by a 30% reduction
of ML thickness [Huang et al., 1999]. MDS supplementation
resulted in significantly increased thickness of
the GL and ML, and greater numbers of PC in older
mice, suggesting multiple positive benefits on cerebellar
morphology. Improved motor coordination and enhanced
spatial learning in older MDS supplemented mice appear
to be a consequence of the maintenance of a more youthful
cerebellar morphology.
Emotionality
Variants of the open field test have been used to assess
emotionality in rodents [Trullas and Skolnick, 1993, Belzung
and Griebel, 2001, Voikar et al., 2001; Carola et al.,
2002; De Oliveira et al., 2007; Chaudhry et al., 2008].
However, depending on protocol variations and parameters
scored, results may be more indicative of exploratory
behavior than anxiety-driven responses [Trullas and Skolnick,
1993; Carola et al., 2002; Berry et al., 2007]. Variables
measured in tests for anxiety (e.g. open field,
elevated maze) were more consistent with general activity
levels when carried out in dim light when mice are normally
active [Trullas and Skolnick, 1993]. To assess emotionality
we employed bright illumination and white
Plexiglas construction creating aversive environments
[Chaudhry et al., 2008]. Collectively, the tests performed
in the current study suggest that the MDS reduced anxiety
in mice, allowing supplemented mice to more freely
explore an ‘unsafe/novel’ environment. Although the
MDS was not intentionally designed to target emotionality,
current findings suggest it demonstrated significant
anxiolytic properties.
Essentially all variants of behavioral tests rely on
movement of animals to infer emotionality levels [Trullas
and Skolnick, 1993, Carola et al., 2002; Chaudhry et al.,
2008]. Intrinsic differences in activity levels between
treatment groups (if present) may bias interpretation. The
MDS was previously shown to augment physical activity
[Aksenov et al., 2010]. This could influence movement of
animals in the experimental apparatus irrespective of
emotionality. However, in the open field test, distance
traveled by mice in the peripheral zone and mean running
velocity were unaffected by MDS treatment. This indicates
that increased exploration of the central zone by
supplemented Nr mice was not a consequence of upregulated
motor activity. In the circle run test, increased physical
activity of supplemented mice would predict faster
escape; however, the opposite result was observed, suggesting
differences in intrinsic activity levels between
treatment groups did not confound results.
Aging alone is not a significant risk factor for emotional
and anxiety disorders in humans [Beekman et al.,
1998]; however, some mechanisms associated with emotionality
involve pathways implicated in aging. Oxidative
stress has been linked to elevated anxiety in aging rodents
[Berry et al., 2007; Salim et al., 2010] and reactive oxygen
species (ROS) in brain positively correlate with anxiety
behavior [De Oliveira et al., 2007; Bouayed et al.,
2009]. Both ROS and inflammatory processes are strongly
implicated in aging and age-related diseases [Finkel and
Holbrook, 2000; Chung et al., Nemat et al., 2009]. Furthermore,
ROS and inflammation appear to be costimulatory
[Janssen et al., 1993; Lavarosky et al., 2000;
Droge, 2001; Reuter et al., 2010]. Elevated ROS via
NADPH oxidase activity induced anxiety which was
reversed by apocynin, a NADPH oxidase inhibitor
[Masood et al., 2008]. In addition, chronic inflammation
and inflammatory cytokines have been shown to induce
anxiety in mice and rats [Bercik et al., 2010; Song et al.,
2003]. There are multiple mechanisms linking ROS and
cytokines to modulation of anxiety. Neuronal losses in
stress-regulating regions of the brain, altered neurotransmitter
levels and dysregulation of the hypothalamicpituitary-
adrenocortical (HPA) axis, including glucocorticoid
receptors have been linked to ROS, pharmacological
or genetic manipulations causing anxiety also impact
memory and cognition [Izquierdo and Medina, 1999; Park
et al., 2001; O’Shea et al., 2004; Venero et al., 2005].
Neuroinflammatory damage can prevent mice from eliciting
normal responses to light/dark stimuli [Pascual et al.,
2011]. Neuroinflammation and other neurodegenerative
conditions are commonly associated with aging [McGeer
and McGeer, 2004; Rollo, 200]. Therefore, behavioral
regulation in response to changing environmental contexts
(such as illumination) may be impaired in old mice. After
testing in bright light, we repeated the circle run test and
step-down test in dim illumination. In both tests, aging
untreated mice showed virtually identical behaviors
regardless of lighting. Conversely, behavioral responses
of supplemented mice were significantly different in dim
light compared to bright illumination. These results suggest
that supplemented mice had better contextual discrimination
implying stronger cognitive function. Indeed,
our MDS was already shown to improve cognition of
aging mice [Lemon et al., 2003, Aksenov et al., 2013].
Impaired light/dark discrimination was also associated
with poor locomotor ability [McGeer and McGeer, 2004].
Our MDS was shown to upregulate locomotion in mice
[Aksenov et al., 2010]. Taken together, it appears that our
treatment has a general impact on brain sensorimotor
neurocircuitry.
Visual Acuity
The principle behind tests of visual acuity involves
challenging mice with increasingly more difficult visual
cues and determining when correct responses are extinguished
[Prusky et al., 2000]. Based on this, we developed
a novel test from a less sensitive version [Chaudhry
et al., 2008], which uses the animal’s natural tendency to
reach for perceived objects or surfaces to escape when
suspended by the tail. We found that ~2-year-old Nr
mice had largely intact vision as all animals successfully
responded to repeated presentation of a distinct visual cue
on high contrast background, irrespective of treatment.
However, untreated mice showed a progressive loss of
reaching responses as increasingly more subtle visual
cues were presented. In comparison, responses of supplemented
mice extinguished at slower rates. Significant differences
between supplemented and untreated mice were
resolved on a moderately difficult visual challenge. These
findings suggest that the MDS may offset visual acuity
decline seen in ageing.
Histological examination of the retinas from 2-year-old
Nr mice indicated markers of retinal morphology were
found to be consistent with those of age-related retinal
degeneration. The MDS was designed to target key mechanisms
associated with aging which include: oxidative
stress, inflammation, and mitochondrial function [Lemon
et al., 2003]. The robust efficacy of our MDS on these
parameters has been experimentally confirmed [Lemon
et al., 2008a,b, Aksenov et al., 2010, 2013; Long et al.,
2012]. Thicknesses of the retinal outer nuclear layer
(ONL) and outer segment (OS) were increased by 26 and
29%, respectively in MDS mice, compared to untreated
mice. Unfortunately due to the limited age-range of samples,
it was not possible to determine the magnitude of
retinal degeneration and the extent to which the retinal
atrophy was salvaged by MDS supplementation. However,
despite this limitation, the reduced retinal atrophy in
MDS supplemented mice was likely the result of attenuated
oxidative and inflammatory processes and boosted
mitochondrial function. Irregular organization and degeneration
of the ONL and OS organization are closely associated
with age-related macular degeneration (AMD) in
humans [Gartner and Henkind, 1981]. In mouse models
for AMD, geographic atrophy of the ONL and OS are
regarded as chief biomarkers of retinal degeneration consistent
with human AMD [Kim et al., 2002, Rakoczy
et al., 2002, Karan et al., 2005, Rakoczy et al., 2006, Justilien
et al., 2007].
The ONL contains photoreceptor
nuclei; thinning of the ONL occurs as a result of cell
depletion with a reduction in associated photoreceptors,
resulting in diminished visual acuity [Gartner and Henkind,
1981, Kim et al., 2002, Kashani et al., 2009]. Likewise,
the OS reflect rod and cone abundance [Carter-
Dawson and LaVail, 1979]. In this fashion, improved visual
acuity in old supplemented mice is consistent with
reduced retinal degeneration. Given that similar geographic
atrophy is observed in AMD, it is arguable that
our MDS may offset this pathology in humans. Although
murine retinas do not possess macula, patterns of retinal
degeneration in mice correspond to those in human macular
degeneration, thus useful inferences can be drawn to
human visual acuity [Rakoczy et al., 2002; Karan et al.,
2005; Rakoczy et al., 2006; Edwards and Malek, 2007;
Justilien et al., 2007; Chu et al., 2013]. Retinal degeneration
results largely from dysregulation in multiple factors,
however inflammatory processes and oxidative damage
play key roles in the development of this pathology [Winkler
et al., 1999; Hogg and Chakravarthy, 2004; Edwards
and Malek, 2007, Hollyfield et al., 2008; Saski et al.,
2009; Barot et al., 2011]. Studies investigating the use of
antioxidants to prevent AMD had variable, yet promising
success [Hogg and Chakravarthy, 2004; Van Leeuwen
et al., 2005].
However, a recent review of clinical and
experimental data concluded that singly administered antioxidants
(such as vitamin E and β-carotene) did not
improve AMD prognosis [Evans, 2008]. Recently, mitochondrial
dysfunction was additionally implicated in
development of AMD [Barot et al., 2011]. Considering
that this pathology involves dysregulation of multiple cellular
systems, greater therapeutic benefits should be possible
through complex multi-targeted interventions.
Olfaction
Morphological changes in olfactory bulbs of aging
rodents are somewhat debatable with respect to laminar
features. Volume (or thickness) of GLM and EPL are normally
used as measures of aging, however results have
been inconsistent [Hinds and McNelly, 1977; Mirich
et al., 2002; Richard et al., 2010]. Some report no significant
changes in volumes with age [Richard et al., 2010],
while others show that thickness of these layers (as well
as the entire bulb volume) increases between 6 and 24
months [Mirich et al., 2002]. Increases appeared to be
equally proportional in each layer [Mirich et al., 2002].
Hinds and McNelly [1977] observed that volumes of all
laminar layers in rat olfactory bulbs significantly
increased from 3 to 24 months of age, followed by a
sharp decline thereafter. At oldest ages (~30 months)
layer volumes were identical to those in 3–12 month old
rats. Collectively, it appears that laminar volumes
increase with age, if extremely old animals were excluded
[Hinds and McNelly, 1977; Mirich et al., 2002]. The
mice in the current study were aged 14–17 months, with
MDS treatment resulting in significantly reduced volumes
of the GLM and EPL. This suggested that MDS treatment
offset age-related morphological changes, which was supported
by the improved olfactory sensitivity in MDS
treated aging mice.
Mitral cells are the primary output neurons in the olfactory
bulb that process olfactory sensory input prior to
activating higher-order processing [Meister and Bonhoeffer,
2001; Rawson, 2006]. Age-related (or otherwise
caused) loss of odor perception may occur due to complications
afferent to mitral cells, but mitral cell dysfunction
is ultimately implicated [Meisami et al., 1998; Rawson,
2006]. Loss of mitral cells with age is documented in
rodents [Hinds and McNelly 1977; Mirich et al., 2002]
and sizable declines are also observed in humans [Bhatnager
et al., 1987; Meisami et al., 1998]. Our MDS treated
mice showed a 29% increase in mitral cells, suggesting
MDS treatment protected these neurons from age-related
atrophy. However, due to difficulty in morphological
discrimination of mitral cells from sections stained only
with H&E, these preliminary results should be interpreted
with caution. Mitral cells are particularly susceptible to
oxidative and nitrosative stress [Vaishnav et al., 2007,
Yang et al., 2013]. Amount of 3-NT (a marker for protein
nitration) was significantly elevated in olfactory bulbs of
old mice [Vaishnav et al., 2007; Yang et al., 2013]. 3-NT
was evident in the mitral cell layer, but was also delocalized
to the EPL and GLM in advanced ages [Vaishnav
et al., 2007]. Previous studies on MDS treatment demonstrated
significantly lowered 3-NT levels in brains of
aging mice [Long et al., 2012] and protein carbonyls
were also reduced [Aksenov et al., 2010], emphasizing
the potential protective capacity of the MDS on the mitral
cell population.
Olfaction and Neurodegenerative Disease in Humans
Humans, like other animals, display age-related
declines in olfactory function [Cain and Stevens, 1989;
Nakayasu et al., 2000; Rawson, 2006]. Losing the sense
of smell per se, does not pose a major threat for humans,
although quality of life may be impacted [Rawson, 2006].
However, greater concerns emerged when loss of olfaction
was found to be strongly correlated with the risk of
developing severe neurodegenerative conditions [Wilcock
and Esiri, 1982; Ohm and Braak, 1987]. Alzheimer’s
(AD) patients show deterioration of mitral cells, with substantial
mitral cell loss apparent before clinical AD symptoms
emerged [Struble and Clark, 1992]. Combined with
previous reports of severe olfactory deficits in dementia,
AD and Parkinson’s disease (PD), authors suggested that
olfactory dysfunction may be an early manifestation of
neurodegenerative disease [Doty et al., 1989; Koss et al.,
1989; Schiffman et al., 1990; Struble and Clark, 1992;
Devanand et al., 2000; Schiffman et al., 2002; Li et al.,
2010; Baba et al., 2012; Doty, 2012]. Neuronal loss and
altered olfactory morphology was later shown in young
AD patients [Ter Laak et al., 1994], suggesting that loss
of olfaction and olfactory neurons is not simply an agerelated
condition that parallels cognitive impairment in
old AD patients.
In fact some have even proposed that
pathogenesis in AD and PD may be catalyzed by agents
that enter the brain via olfactory pathways [Doty, 2008].
The accuracy of olfactory tests in early detection of neural
pathology suggests that individuals with intact sense
of smell are at lower risk of developing neurodegenerative
conditions in the near future. Higher counts of mitral
cells in the olfactory bulb also reflect higher neuronal
populations in adjacent brain regions. Given that aging
supplemented mice showed better olfactory sensitivity
and higher number of mitral cells in the olfactory bulb,
may suggest that the MDS could be offsetting neurodegeneration
throughout the brain.
CONCLUSION
Effects of treatment with a multi-ingredient dietary
supplement designed to ameliorate key mechanisms of
aging showed treatment was associated with reduced
anxiety-like behaviors, augmented discrimination of environmental
context, improved motor balance, and
improved visual and olfactory acuity. This was correlated
with positive morphological changes and higher neuronal
populations in the cerebellum and olfactory bulb,
increased overall brain cell numbers and improved brain
function. Intact olfaction is strongly indicative of suppression
of neuronal degeneration. Retinal atrophy (associated
with AMD) was also diminished in supplemented mice.
Given that MDS treatment has been shown to significantly
reduce oxidative damage, boost mitochondrial
function [Lemon et al 2008a,b; Aksenov et al., 2010;
Aksenov et al., 2013] and alleviate symptoms of inflammation
[Lemon et al., 2005], suggests that neuronal protection
and sensory function are likely attributed to
diminishing oxidative/inflammatory stress and improved
energy balance. The extent of functional benefits attained
by our MDS here and in earlier studies [Lemon et al.,
2003, 2005, 2008a,b; Aksenov et al., 2010, 2013; Long
et al., 2012; Hutton et al., 2015] strongly suggests that
aging animals retain the capacity to support youthful phenotypes
and that powerful impacts can be achieved
through multi-ingredient dietary supplementation that
addresses the multifactorial nature of aging organisms.
ACKNOWLEDGMENTS
The authors would like to recognize that JA Lemon
and V Aksenov contributed equally to both writing of this
manuscript and collecting/analyzing data presented in this
work. Olfactory sensitivity was tested by R Samigullina,
while S Aksenov contributed majorly to histological analysis
of retina, cerebellum and olfactory bulb. Special
thanks to Ted A. Aristilde for providing technical histology
support and Dr. Henry Szechtman for providing the
image tracking system for behavioral studies and the
expert technical assistance
AUTHOR CONTRIBUTIONS
JAM, VA, CDR, DRB and RS designed the study.
CDR and DRB provided funding and obtained Ethics
Board approval. JAM, VA, RS, SA, and WHR collected
and synthesized the data. JAM, VA, WHR, and SA analyzed
the data and prepared draft figures and tables. JAM
and SA prepared the manuscript draft with important
intellectual input from CDR and DRB. All authors
approved the final manuscript. JAM, VA, CDR and DRB
had complete access to the study data.
REFERENCES
Aksenov V, Long J, Lokuge S, Foster JA, Liu J, Rollo CD. 2010. Dietary
amelioration of locomotor, neurotransmitter and mitochondrial
aging. Exp Biol Med 235:66–76.
Aksenov V, Long J, Liu J, Szechtman H, Khanna P, Matravadia S,
Rollo CD. 2013. A complex dietary supplement augments spatial
learning, brain mass, and mitochondrial electron transport chain
activity in aging mice. Age 35:23–33.
Andersen BB, Gundersen HJG, Pakkenberg B. 2003. Aging of the
human cerebellum: A stereological study. J Compar Neurol 466:
156–365.
Antier D, Carswell HV, Brosnan MJ, Hamilton CA, Macrae IM, Groves
S, Jardine E, Reid JL, Dominiczak AE. 2004. Increased levels of
superoxide in brains from old female rats. Free Radic Res 38:
177–182.
Apps R, Garwicz M. 2005. Anatomical a physiological foundations of
cerebellar information processing. Nat Rev Neurosci 6:297–311.
Baba T, Kikuchi A, Hirayama K, Nishio Y, Hosokai Y, Kanno S,
Hasegawa T, Sugeno N, Konno M, Suzuki K, et al., 2012. Severe
olfactory dysfunction is a prodromal syndrome of dementia associated
with Parkinson’s disease: A 3 year longitudinal study.
Brain 135:161–169.
Balaban CD. 2002. Neural substrates linking balance control and anxiety.
Physiol Behav 77:469–475.
Balaban CD, Jacob RG, Furman JM. 2011. Neurologic bases for comorbidity
of balance disorders, anxiety disorders and migraine: neurotheraputic
implication. Expert Rev Neurotheraputics 11:379–394.
Barot M, Gokulgandi R, Mitra AK. 2011. Mitochondrial dysfunction in
retinal diseases. Curr Eye Res 36:1069–1077.
Barski JJ, Dethleffsen K, Meyer M. 2000. Cre recombinase expression
in cerebellar Purkinje cells. Genesis 28:93–98.
Barski JJ, Hartmann J, Rose CR, Hoebeek F, M€orl K, Noll-Hussong M,
De Zeeuw CI, Konnerth A, Meyer M. 2003. Calbindin in the cerebellar
Purkinje cells is a critical determinant of the precision of
motor coordination. J Neurosci 23:3469–3477.
Bartke A, Chandrashekar V, Bailey B, Zaczek D, Turyn D. 2002. Consequences
of growth hormone (GH) overexpression and GH
resistance. Neuropep 36:201–208.
Bartke A. 2003. Can growth hormone (GH) accelerate aging? Evidence
from GH-transgenic mice. Neuroendocrinol 78:210–216.
Beekman ATF, Bremmer MA, Deeg DJH, Van Balkom AJLM, Smit JH,
De Beurs E, Dyck RV, Van Tilburg W. 1998. Anxiety disorders
in later life: A report from the longitudinal aging study Amsterdam.
Int J Geriat Psych 13:717–726.
Beierlein M. 2014. Synaptic mechanisms underlying cholinergic control
of thalamic reticular nucleus neurons. J Physiol 592:4137–4145.
Belzung C, Griebel G. 2001. Measuring normal and pathological
anxiety-like behavior in mice: A review. Behav Brain Res 125:
141–149.
Berry A, Capone F, Giorgio M, Pelicci PG, De Kloet ER, Alleva E,
Minghetti L, Cirulli F. 2007. Deletion of the life span determinant
p66Sch prevents age-related dependant increase in emotionality
and pain sensitivity in mice. Exp Geront 42:37–45.
Bercik P, Verdu EF, Foster JA, Marci J, Potter M, Huang X,
Malinowski P, Jackson W, Blennerhassett P, Neufeld KA, et al.,
2010. Chronic gastrointestinal inflammation induces anxiety-like
behavior and alters nervous system biochemistry in mice. Gastroenterol
139:2102–2112.
Bhatnagar KP, Kennedy RC, Baron G, Greenberg RA. 1987. Number of
mitral cells and the bulb volume in the aging human olfactory
bulb: A quantitative morphological study. Anat Record 218:73–78.
Bickford PC, Shukitt-Hale B, Joseph J. 1999. Effects of aging on cerebellar
noradrenergic function and motor learning: Nutritional
interventions. Mech Ageing Dev 111:141–154.
Bobee S, Mariette E, Tremblay-Leveau H, Caston J. 2000. Effects of
early midline cerebellar lesion on cognitive and emotional functions
in the rat. Behav Brain Res 112:107–117.
Bouayed J, Rammal H, Soulimani R. 2009. Oxidative stress and anxiety:
Relationship and cellular pathways. Oxid Med Cell Longev 2:63–67.
Bratic A, Larsson NG. 2013. The role of mitochondrial in aging. J Clin
Invest 123:951–957.
Butterfield DA. 2014. The 2013 SFRBM discovery award: Selected discoveries
from the butterfield laboratory of oxidative stress and its
sequela in brain in cognitive disorders exemplified by Alzheimer
disease and chemotherapy induced cognitive impairment. Free
Radic Biol Med 74:157–174.
Butterfield DA, Howard BJ, LaFontaine MA. 2001. Brain oxidative
stress in animal models of accelerated aging and the age-related
neurodegenerative disorders, Alzheimer’s disease and Huntington’s
disease. Curr Med Chem 8:815–828.
Cain WS, Stevens JC. 1989. Uniformity of olfactory loss in aging. Ann
NY Acad Sci 561:29–38.
Calder WA. 1984. Size, Function and Life History. Cambridge: Harvard
University Press. 431 p.
Carlson JC, Bharadwaj R, Bartke A. 1999. Oxidative stress in hypopituitary
dwarf mice and in transgenic mice overexpressing human
and bovine GH. Age 22:181–186.
Carola V, D’Olimpio F, Brunamonti E, Mangia F, Renzi P. 2002. Evaluation
of the elevated plus-maze and open-field tests for the
assessment of anxiety-related behaviour in inbred mice. Behav
Brain Res 134:49–57.
Carter RJ, Morton J, Dunnett SB. 2001. Motor coordination and balance
in rodents. Curr Protoc Neurosci. Chapt 8:Unit 8.12.
Carter-Dawson LD, LaVail MM. 1979. Rods and cones in the mouse
retina. J Comp Neur 188:245–262.
Chaudhry AM, March-Rollo SE, Aksenov V, Rollo CD, Szechtman H.
2008. Modifier selection by transgenesis: The case of growth hormone
transgenic and hyperactive circling mice. Evol Biol 35:
267–286.
Choi D, Furay A, Evanson N, Ostrander M, Ulrich-Lai Y, Herman J.
2007. Bed nucleus of the stria terminalis subregions differentially
regulate hypothalamic–pituitary–adrenal axis activity: Implications
for the integration of limbic inputs. J Neurosci 27:2025–
2034.
Chung HY, Sung B, Jung KJ, Zou Y, Yu BP., 2006 The molecular
inflammatory processes in aging. Antioxid Redox Sign 8:572–581.
Cohen G, Farooqui R, Kesler N. 1997. Parkinson disease: A new link
between monoamine oxidase and mitochondrial electron flow.
Proc Natl Acad Sci USA 94:4890–4894.
Colucci-Guyon E, Ribotta MGY, Maurice T, Babinet C, Privat A. 1999.
Cerebelar defect and impaired motor coordination in mice lacking
vimentin. Glia 25:33–43.
Cui H, Kong Y, Zhang Y. 2012. Oxidative stress, mitochondrial dysfunction,
and aging. J Signal Transduct 2012:646354
Dahhaoui M, Lannou J, Stelz T, Caston J, Guastavina JM. 1992. Role of
the cerebellum in spatial orientation of the rat. Behav Neural
Biol 58:180–189.
D’Angelo E. 2012. Cerebellar Granule Cell. In: Manto M, Schmahmann
JD, Rossi F, Gruol DL, Koibuchi N, editors. Handbook of the
Cerebellum and Cerebellar Disorders. Netherlands: Springer. pp
765–791.
D’Angelo E, De Zeeuw CI. 2009. Timing and plasticity on the cerebellum:
Focus on the granular layer. Trends Neurosci 32:30–40.
Dei R, Takeda A, Niwa H, Li M, Nakagomi Y, Watanabe M, Inagaki T,
Washimi Y, Yasuda Y, Horie K, et al., 2002. Lipid peroxidation
and advanced glycation end products in the brain in normal aging
and in Alzheimer’s disease. Acta Neuropathol (Berl) 104:113–122.
De Oliveira MR, Silvestrin RB, Mello E, Souza T, Moreira JC. 2007.
Oxidative stress in the hippocampus, anxiety-like behavior and
decreased locomotory and exploratory activity of adult rats:
Effects of subacute vitamin A supplementation at therapeutic
doses. Neurotoxicol 6:1191–1199.
Devanand DP, Michaels-Marston KS, Liu X, Pelton GH, Padilla M,
Marder K, Bell K, Stern Y, Mayeux R. 2000. Olfactory deficits
in patients with mild cognitive impairment predict Alzheimer’s
disease at follow-up. Am J Psychiatry 157:1399–1405.
Doty RL. 2008. The olfactory vector hypothesis of neurodegenerative
disease: is it viable? Ann Neurol 63:7–15.
Doty RL. 2012. Olfactory dysfunction in Parkinson’s disease. Nat Rev
Neurol 8:329–339.
Doty RL, Riklan M, Deems DA, Reynolds C, Stellar S. 1989. The olfactory
and cognitive deficits of Parkinson’s disease: Evidence for
independence. Ann Neurol 25:166–171.
Doulazmi M, Frederic F, Lemaigre-Debreuil Y, Hadj-Sahraoui N,
Delhaye-Bouchard N, Mariani J. 1999. Cerebellar purkinje cell
loss during life span of the heterozygous Staggerer mouse (Rora1/
Rorasg) is gender related. J Compar Neurol 411:267–273.
Droge W. 2002. Free radicals in the physiological control of cell function.
Physiol Rev 82:47–95.
Du F, Zhu XH, Zhang Y, Friedman M, Zhang N, Ugurbil K, Chen W.
2002. Tightly coupled brain activity and cerebral ATP metabolic
rate. Proc Natl Acad Sci USA 105:6409–6414.
Edwards AO, Malek G. 2007. Molecular genetics of AMD and current
animal models. Angiogen 10:119–132.
Erecinska M, Cherian S, Silver IA. 2004. Energy metabolism in mammalian
brain during development. Prog Neurobiol 73:397–445.
Evans J. 2008. Antioxidant supplements to prevent or slow down the
prognosis of AMD: A systemic review and meta-analysis. Eye
22:751–760.
Finkel T, Holbrook NJ. 2012 Oxidants, oxidative stress and the biology
of ageing. Nature 408:239–247.
Foster MJ, Dubey A, Dawson KM, Stutts WA, Lal H, Sohal RS. 1996.
Age-related losses of cognitive function and motor skills in mice
are associated with oxidative protein damage in the brain. Proc
Natl Acad Sci USA 93:4765–4769.
Gandhi S, Abramov AY. 2012. Mechanism of oxidative stress in neurodegeneration.
Oxid Med Cell Longev 2012:428010
Gartner S, Henkind P. 1981. Aging and degeneration of the human macula.
I. Outer nuclear layer and photoreceptors. Brit J Ophthalmol
65:32–28.
Ghelarducci B, Salamone D, Simoni A, Sebastiani L. 1996. Effects of
early cerebellar removal on the classically conditioned bradycardia
of adult rabbits. Exp Brain Res 111:417–423.
Girardi M, Konrad HR, Amin M, Hughes LF. 2001. Predicting fall risks
in an elderly population: Computer dynamic posturography versus
electronystagmography test results. Laringoscope 111:1528–
1532.
Grimm S, Hoehn A, Davies KJ, Grune T. 2011. Protein oxidative modifications
in the ageing brain: Consequence for the onset of neurodegenerative
disease. Free Radic Res 45:73–88.
Hattingen E, Magerkurth J, Pilatus U, Mozer A, Seifried C, Steinmetz
H, Zanella F, Hilker R. 2009. Phosphorus and proton magnetic
resonance spectroscopy demonstrates mitochondrial dysfunction
in early and advanced Parkinson’s disease. Brain 132:3285–3297.
Hauck SJ, Bartke A. 2001. Free radical defenses in the liver and kidney
of human growth hormone transgenic mice: Possible mechanisms
of early mortality. J Geront 56:B153–B162. A:
Hinds JW, McNelly NA. 1977. Aging of the rat olfactory bulb: growth
and atrophy of constituent layers and changes in size and number
of mitral cells. J Comp Neurol 171:345–368.
Hogg R, Chakravarthy U. 2004. AMD and micronutrient antioxidants.
Curr Eye Res 29:387–401.
Hollyfield JG, Bonihla VL, Rayborn ME, Yang X, Shadrach KG, Lu L,
Urfet RL, Salomon RG, Perez VL. 2008. Oxidative damageinduced
inflammation initiates age-related macular degeneration.
Nat Med 14:194–198.
Huang C, Brown N, Huang RH. 1999. Age-related changes in the cerebellum:
Parallel fibers. Brain Res 840:148–152.
Ikeyama S, Kokkonen G, Shack S, Wang XT, Holbrook NJ. 2002. Loss
in oxidative stress tolerance with aging linked to reduced extracellular
signal-regulated kinase and Akt kinase activities. FASEB
16:114–116.
Ito M. 2000. Mechanisms of motor learning in the cerebellum. Brain
Res 886:237–245.
Izquierdo I, Medina JH. 1999 GABAA receptor modulation of memory:
The role of endogenous benzodiazepines. Trends Pharmacol Sci
12:260–265.
Janssen YM, Houten B, Borm PJ, Mossman BT. 1993. Cells and tissue
responses to oxidative damage. Lab Invest J Tech Method Pathol
69:261–274.
Justilien V, Pang JJ, Renganathan K, Zhan X, Crabb JW, Kim SR,
Sparrow JR, Hauswirth WW, Lewin AS. 2007. SOD2 knockdown
mouse model of early AMD. Invest Ophthalmil Vis Sci 48:4407–
4420.
Kannurpatti SS, Sanganahalli BG, Mishra S, Joshi PG, Joshi NB. 2004.
Glutamate-induced differential mitochondrial response in young
and adult rats. Neurochem Int 44:361–369.
Karan G, Lillo C, Yang Z, Cameron DJ, Locke KG, Zhao Y,
Thirumalaichary S, Li C, Birch DG, Vollmer-Snarr HR, et al.,
2005. Lipofuscin accumulation, abnormal electrophysiology and
photoreceptor degeneration in mutant ELOVL4 transgenic mice:
a model for macular degeneration. Proc Natl Acad Sci 102:4161–
4169.
Kashani AH, Keane PA, Dustin L, Walsh AC, Sadda SR. 2009. Quantitative
subanalysis of cystoid spaces and outer nuclear layer using
optical coherence tomography in age-related macular degeneration.
Invest Ophthalmil Vis Sci 50:3366–3373.
Kashiwabuchi N, Ikeda K, Araki K, Hirano T, Shibuki K, Takayama C,
Inoue Y, Kutsuwada T, Yagi T, Kang Y, et al., 1995. Impairment of
motor coordination, Purkinje cell synapse formation, and cerebellar
long-term depression in GluRd2 mutant mice. Cell 81:245–252.
Keller JN, Pang Z, Geddes JW, Begley JG, Germeyer A, Waeg G,
Mattson MP. 1997. Impairment of glucose and glutamate transport
and induction of mitochondrial oxidative stress and dysfunction in
synaptosomes by amyloid b-peptide: Role of the lipid peroxidation
product 4-hydroxynonenal. J Neurochem 69:273–284.
Kim SY, Sadda S, Pearlman J, Humayun MS, de Huan E, Jr., Melia
BM, Green WR. 2000. Morphometric analysis of the macula in
eyes with discoform age-related macular degeneration. Retina 22:
471–477.
Koss E, Weiffenbach JM, Haxby JV, Friedland RP. 1988. Olfactory
detection and identification performance in early Alzheimer’s disease.
Neurology 38:1228–1232.
Landis S. 1973. Ultrastructural changes in the mitochondria of cerebellar
Purkinje cells of nervous mutant mice. J Cell Biol 57:782–797.
Larsen JO, Skalicky M, Viidik A. 2000. Does long-term physical exercise
counteract age-related Purkinje cell loss? A stereological
study of rat cerebellum. J Compar Neurol 428:213–222.
Lavrovsky Y, Chatterjee B, Clark RA, Roy AK. 2000. Role of redoxregulated
transcription factor in inflammation, aging and agerelated
diseases. Exp Gerontol 35:521–532.
Lemon JA, Boreham DR, Rollo CD. 2003. A dietary supplement abolishes
age-related cognitive decline in transgenic mice expressing
elevated free radical processes. Exp Biol Med 228:800–810.
Lemon JA, Boreham DR, Rollo CD. 2005. A complex dietary supplement
extends longevity of mice. J Gerontol A Biol Sci Med Sci
60:275–279.
Lemon JA, Rollo CD, McFarlane NM, Boreham DR. 2008a. Radiationinduced
apoptosis in mouse lymphocytes is modified by a
complex dietary supplement: The effect of genotype and gender.
Mutagenesis 23:465–471.
Lenaz G. 1998. Role of mitochondria in oxidative stress and ageing.
Biochim Biophys Acta 1366:53–67.
Li W, Howard JD, Gottfried JA. 2010. Disruption of odor quality coding
in piriform cortex mediates olfactory deficits in Alzheimer’s disease.
Brain 133:2714–2726.
Long J, Aksenov V, Rollo CD, Liu J. 2012. A complex dietary supplement modulates nitrative stress in normal mice and in a new mouse model of nitrative stress and cognitive aging. Mech Ageing
Devel 133:523–529.
Lyras L, Cairns NJ, Jenner A, Jenner P, Halliwell B. 1997. An assessment
of oxidative damage to proteins, lipids, and DNA in brains from
patients with Alzheimer’s disease. J Neurochem 68:2061–2069.
Ma L, Morton AJ, Nicholson LF. 2003. Microglia density decreases
with age in a mouse model of Huntington’s disease. Glia 43:
274–280.
Markesbury WR, Carney JM. 1999. Oxidative alterations in Alzheimer’s
disease. Brain Pathol 9:133–146.
Masood A, Nadeem A, Mustafa SJ, O’Donnell JM. 2008. Reversal of
oxidative stress-induced anxiety by inhibition of
phosphodiesterase-2 in mice. J Pharm Exp Ther 326:369–379.
Mattson MP, Pedersen WA, Duan W, Culmsee C, Camandola S. 1999.
Cellular and molecular mechanisms underlying perturbed energy
metabolism and neuronal degeneration in Alzheimers and Parkinson’s
diseases. Ann NY Acad Sci 893:154–175.
Mattson MP, Magnus T. 2004 Ageing and neuronal vulnerability. Nat
Rev Neurosci 7:278–294.
Maurya PK, Noto C, Rizzoa LB, Riosa AC, Nunes SOV, Barbosa DS,
Sethi S, Zeni M, Mansur RB, et al., 2016. The role of oxidative
and nitrosative stress in accelerated aging and major depressive
disorder. Prog Neuropsychopharmacol Biol Psychiatry 65:134–144.
McGeer PL, McGeer EG. 2004. Inflammation and the degenerative diseases
of aging. Ann NY Acad Sci 1035:104–116.
Meisami E, Mikhail L, Baim D, Bhatnagar KP. 1998. Human olfactory
bulb: Aging of glomeruli and mitral cells and a search for the
accessory olfactory bulb. Ann NY Acad Sci 855:708–715.
Meister M, Bonhoeffer T. 2001. Tuning and topography in an odor map
on the rat olfactory bulb. J Neurosci 21:1351–1360.
Meliska CJ, Burke PA, Bartke A, Jensen RA. 1997. Inhibitory avoidance
and appetitive learning in aged normal mice: comparison with
transgenic mice having elevated plasma growth hormone levels.
Neurobiol Learn Mem 68:1–12.
Mirich JM, Williams NC, Berlau DJ, Brunjes PC. 2002. Comparative
study of aging in the mouse olfactory bulb. J Comp Neurol 454:
361–372.
Mitchell AS. 2015. The mediodorsal thalamus as a higher order thalamic
relay nucleus important for learning and decision-making. Neurosci
Biobehav Rev 54:76–88.
Monfort V, Chapillon P, Mellier D, Lalonde R, Caston J. 1998. Time
active avoidance learning in lurcher mutant mice. Behav Brain
Res 91:165–172.
Mutlu-Turkoglu U, Ilhan E, Oztezcan S, Kuru A, Aykac-Toker G, Uysal
M. 2003. Age-related increases in malondialdehyde and protein
carbonyl levels and lymphocyte DNA damage in elderly patients.
Clin Biochem 36:397–400.
Nakayasu C, Kanemura F, Hirano Y, Shimizu Y, Tonosaki K. 2000. Sensitivity
of olfactory sense declines with the aging in senescenceaccelerated
mouse (SAM-P1). Physiol Behav 70:135–139.
Nakamura T, Lipton SA. 2007. Molecular mechanisms of nitrosative
stress-mediated protein misfolding in neurodegenerative diseases.
Cell Mol Life Sci 64:1609–1620.
NK, Yadollah A. MM, 2009. Chronic inflammation and oxidative stress
as a major cause of age-related diseases and cancer. Recent Pat
Inflamm Allergy Drug Discov 3:73–80.
Ogueta S, Olazabal I, Santos I, Delgado-Baeza E, Garcia-Ruiz JP. 2000. Transgenic mice expressing bovine GH develop arthritic disorder and self-antibodies. J Endocrinol 165:321–328.
Ohm TG, Braak H. 1987. Olfactory bulb changes in Alzheimer’s disease.
Acta Neuropathol 73:365–369.
O’Shea M, Singh ME, McGregor IS, Mallet PE. 2004. Chronic cannabinoid exposure produces lasting memory impairment and increased anxiety in adolescent but not adult rats.
J Psychopharmacol 18:502–508.
Palmiter RD, Brinster RL, Hammer RE. 1982. Dramatic growth of mice
that develop from eggs microinjected with metallothionein–
growth hormone fusion genes. Nature 300:611–615.
Park CR, Campbell AM, Diamond DM. 2001. Chronic psychological stress impairs learning and memory and increases sensitivity to Yohimbine in adult rats. Biol Psychiatry 50:994–1004.
Parrish-Aungst S, Shipley MT, Erdelyi F, Szabo G, Puche AC. 2007. Quantitative analysis of neuronal diversity in the mouse olfactory bulb. J Comp Neurol 501:825–836.
Pascual M, Bali~no P, Alfonso-Loeches S, Aragon CMG, Guerri C. 2011. Impact of LTR4 on behavioral and cognitive dysfunctions associated with alcohol-induced neuroinflammatory damage. Brain
Behav Immun 25:S80–S90.
Paulin MG. 1993. The role of the cerebellum in motor control and perception.
Brain Behav E 41:39–50.
Prusky GT, West PWR, Douglas RM. 2000. Behavioral assessment of
visual acuity in mice and rats. Vision Res 40:2201–2209.
Rakoczy PE, Yu MJT, Nusinowitz S, Chang B, Heckenlively JR. 2006.
Mouse models of age-related macular degeneration. Exp Eye Res
82:741–752.
Rakoczy PE, Zhang D, Robertson T, Barnet NL, Papadimitriou J,
Constable IJ, Lai CM. 2002. Progressive age-related changes similar
to age-related macular degeneration in a transgenic mouse
model. Am J Pathol 161:1515–1524.
Rawson NE. 2006. Olfactory loss in aging. Sci Aging Knowl Environ pe6
Reddy PH, Beal MF. 2008. Amyloid beta, mitochondrial dysfunction
and synaptic damage: implications for cognitive decline in aging
and Alzheimer’s disease. Trends Mol Med 14:45–53.
Reuter S, Gupta SC, Chaturvedi MM, Aggarwal BB. 2010. Oxidative stress, inflammation and cancer: how are they linked? Free Radic Biol Med 49:1603–1616.
Richard MB, Taylor SR, Greer CA. 2010. Age-induced disruption of
selective olfactory bulb synaptic circuits. Proc Natl Acad Sci USA 107:15613–15618.
Rogers J, Zornetzer AF, Bloom FB, Mervis RE. 1984. Senescent microstructural
changes in rat cerebellum. Brain Res 292:23–32.
Rollo CD, Carlson J, Sawada M. 1996. Accelerated aging of giant transgenic growth hormone mice is associated with elevated free radical processes. Can J Zoo 74:606–620.
Rollo CD, Ko CV, Tyerman JGA, Kajiura L. 1999. The growth hormone axis and cognition: empirical results and integrated theory derived from giant transgenic mice. Can J Zoo 77:1874–1890.
Rollo CD. 2002 Growth negatively impacts the life span of mammals. Evol Dev 4:55–61.
Sakaguchi A, Katamine S, Nishida N, Moriuchi R, Shigematsu K,
Sugimoto T, Nakatani A, Kataoka Y, Houtani T, Shirabe S,
et al., 1996. Loss of cerebellar Purkinje cells in aged mice homozygous
for a disrupted PrP gene. Nature 380:528–531.
Salim S, Sarraj N, Taneja M, Saha K, Tajeda-Simon MV, Chugh G.
2010. Moderate treadmill exercise prevents oxidative stressinduced
anxiety-like behavior in rats. Behav Brain Res 208:545–552.
Sasaki M, Ozawa Y, Kurihara T, Noda K, Imamura Y, Kobayashi S,
Ishida S, Tsubota K. 2009. Neuroprotective effect of an antioxidant,
lutein, during retinal inflammation. Invest Ophthalmol Vis
Sci 50:1433–1439.
Sastre J, Pallardo FV, Vina J. 2003. The role of mitochondrial oxidative
stress in aging. Free Rad Biol Med 35:1–8.
Savory J, Rao JKS, Huang Y, Letada PR, Herman MM. 1999. Agerelated
hippocampal changes in Bcl-2:Bax ratio, oxidative stress,
redox-active iron and apoptosis associated with aluminiuminduced
neurodegeneration: increased susceptibility with aging.
Neurotox 20:805–818.
Schiffmann SN, Cheron G, Lohof A, d’Alcantara P, Meyer M,
Parmentier M, Schurmans S. 1999. Impaired motor coordination
and Purkinje cell excitability in mice lacking calretinin. Proc
Natl Acad Sci USA 96:5257–5262.
Schiffman SS, Clark CM, Warwick ZS. 1990. Gustatory and olfactory
dysfunction in dementia: not specific to Alzheimer’s disease.
Neurobiol Aging 11:597–600.
Schiffman SS, Graham BG, Sattely-Miller EA, Zervakisa J, Welsh-
Bohmera K. 2002. Taste, smell and neuropsychological performance
of individuals at familial risk for Alzheimer’s disease. Neurobiol
Aging 23:397–404.
Schmahmann JD, Caplan D. 2006. Cognition, emotion and the cerebellum. Brain 129:290–292.
Schutter DJLG Van Honk J. 2005. The cerebellum on the rise in human emotion. Cerebellum 4:290–294.
Shigenaga MK, Hagen TM, Ames BN. 1994. Oxidative damage and mitochondrial decay in aging. Proc Natl Acad Sci USA 91: 10771–10778.
Sidman RL, Green MC. 1970. Nervous, a new mutant mouse with cerebellar
disease. In: Sabourdy M, editor. Les mutants pathologiques
chez l’animal, leur inter^et dans la recherche biomedicale. CNRS,
Paris:CNRS.
Song C, Li X, Leonard BE, Horrobin DF. 2003. Effects of dietary n-3
or n-6 fatty acids on interleukin-1b-induced anxiety, stress, and
inflammatory responses in rats. J lipid Res 44:1984–1991.
Sonta T, Inoguchi T, Tsubouchi H, Sekiguchi N, Matsumoto S, Utsumi
H, Nawata H. 2004. Evidence for contribution of vascular
NAD(P)H oxidase to increased oxidative stress in animal models
of diabetes and obesity. Free Radic Biol Med 37:115–123.
Spear PD. 1993. Neural bases of visual deficits during aging. Vis Res
33:2589–2609.
Steger RW, Bartke A, Cecim M. 1993. Premature aging in transgenic
mice expressing different growth hormone genes. J Reprod Fert
46:61–75.
Struble RG, Clark HB. 1992 Olfactory bulb lesions in Alzheimer’s disease.
Neurobiol Aging 13:469–473.
Szechtman H. 1988. Effects of dopamine receptor agonist apomorphine
on sensory input. Naunyn-Schmiedeberg’s Arch Pharmacol 338:
489–496.
Tan JSL, Wang JJ, Liew G, Rochtchina E, Mitchell P. 2008. Age-related macular degeneration and mortality from cardiovascular disease or stroke. Br J Ophthalmol 92:509–512.
Ter Laak HJ, Renkawek K, Van Workum FPA. 1994. The olfactory bulb in Alzheimer’s disease: A morphologic study of neuron loss, tangles, and senile plaques in relation to olfaction. Alz. Dis Assoc Dis 8:38–48.
Tong G, Robertson LT, Brons J. 1993. Climbing fiber representation of
the renal afferent nerve in the vermal cortex of the cat cerebellum. Brain Res 607:65–75.
Trullas R, Skolnick P. 1993. Differences in fear motivated behaviors
among inbred mouse strains. Psychopharm 111:323–331.
Vaishnav RA, Getchell ML, Poon HF, Barnett KR, Hunter SA, Pierce
WM, Klein JB, Butterfield DA, Getchell TV. 2007. Oxidative
stress in the aging murine olfactory bulb: Redox proteomics and
cellular localization. J Neurosci Res 85:373–385.
Van Leeuwen R, Boekhoorn S, Vingerling JR, Witteman JCM, Klaver
CCW, Hofman A., De Jong PTVM 2005. Dietary intake of antioxidants
and age-related macular degeneration. JAMA 294:3101–3107.
Venero C, Guada~no-Ferraz A, HAI, Nordstr€om K, Manzano J, De
Escobar GM, Bernal J., Vennstr€om B, 2005. Anxiety, memory
impairment, and locomotor dysfunction caused by a mutant thyroid
hormone receptor a1 can be ameliorated by T3 treatment.
Genes Dev 19:2152–2163.
Voikar V, K~oks S, Vasar E, Rauvala H. 2001. Strain and gender differences
in the behavior of mouse lines commonly used in transgenic
studies. Physiol Behav 72:271–281.
von Bohlen und Halbach O, Unsicker K. 2002. Morphological alterations
in the amygdala and hippocampus of mice during ageing. Eur J
Neurosci 16:2434–2440.
Wallace JE, Krauter EE, Campbell BA. 1980. Motor and reflective
behavior in the aging rat. J Gerontol 35:364–370.
Wang X, Michaelis EK. 2010. Selective neuronal vulnerability to oxidative
stress in the brain. Front Aging Neurosci 2:12
Weimer JM, Benedict JW, Getty AL, Pontikis CC, Lim MJ, Cooper JD,
Pearce DA. 2009. Cerebellar defects in a mouse model of juvenile
neuronal ceroid lipofuscinosis. Brain Res 1266:93–107.
Wilcock GK, Esiri MM. 1982. Plaques, tangles and dementia. A quantitative
study. J Neurol Sci 56:342–356.
Winkler BS, Boulton ME, Gottsch JD, Sternberg P. 1999. Oxidative
damage and age-related macular degeneration. Mol Vis 5:32
Witt RM, Galligan MM, Despinoy JR, Segal R. 2009. Olfactory behavioral
testing in the adult mouse. J Vis Exp 23:949.
Wolf E, Kahnt E, Ehrlein J, Hermanns W, Brem G, Wanke R. 1993.
Effects of long-term elevated serum levels of growth hormone on
life expectancy of mice: Lessons from transgenic animal models.
Mech Ageing Dev 68:71–87.
Wolf OT. HPA axis and memory. Best Pract Res Cl En 17:287–299.
Woodruff-Pak DS. 2006. Stereological estimation of Purkinje neuron
number in C57BL/6 mice and its relation to associative learning.
Neuroscience 141:233–243.
Woodruff-Pak DS, Foy MR, Akopian GG, Lee KH, Zach J, Nguyen
KPT, Comalli DM, Kennar JA, Agelan A, Thompson RF. 2010.
Differential effects and rates of normal aging in cerebellum and
hippocampus. Pnas 107:1624–1629.
Wulff P, Schonewille M, Renzi M, Viltono L, Sassoe`-Pognetto M,
Badura A, Gao Z, Hoebeek FE, van Dorp S, Wisden W, et al.,
2009. Synaptic inhibition of Purkinje cells mediates consolidation
of vestibule-cerebellar motor learning. Nature Neurosci 12:1042–
1049.
Yan MH, Wang X, Zhua X. 2013. Mitochondrial defects and oxidative
stress in Alzheimer disease and Parkinson disease. Free Radic
Biol Med 62:90–101.
Yang MJ, Sim S, Jeon JH, Jeong E, Kim HC, Park YJ, Kim IB. 2013.
Mitral and tufted cells are potential cellular targets of nitration in
the olfactory bulb of age mice. PLoS One 8:e59673
Yin F, Boveris A, Cadenas E. 2014. Mitochondrial energy metabolism
and redox signaling in brain aging and neurodegeneration. Antiox
Redox Signal 20:353–371.
Return to NUTRITION
Since 8–06-2018
|