Diffusion-weighted MRI and quantitative biophysical modeling of hippocampal neurite loss in chronic stress.
ABSTRACT Chronic stress has detrimental effects on physiology, learning and memory and is involved in the development of anxiety and depressive disorders. Besides changes in synaptic formation and neurogenesis, chronic stress also induces dendritic remodeling in the hippocampus, amygdala and the prefrontal cortex. Investigations of dendritic remodeling during development and treatment of stress are currently limited by the invasive nature of histological and stereological methods. Here we show that high field diffusion-weighted MRI combined with quantitative biophysical modeling of the hippocampal dendritic loss in 21 day restraint stressed rats highly correlates with former histological findings. Our study strongly indicates that diffusion-weighted MRI is sensitive to regional dendritic loss and thus a promising candidate for non-invasive studies of dendritic plasticity in chronic stress and stress-related disorders.
[show abstract] [hide abstract]
ABSTRACT: Stress is a biologically significant factor that, by altering brain cell properties, can disturb cognitive processes such as learning and memory, and consequently limit the quality of human life. Extensive rodent and human research has shown that the hippocampus is not only crucially involved in memory formation, but is also highly sensitive to stress. So, the study of stress-induced cognitive and neurobiological sequelae in animal models might provide valuable insight into the mnemonic mechanisms that are vulnerable to stress. Here, we provide an overview of the neurobiology of stress memory interactions, and present a neural endocrine model to explain how stress modifies hippocampal functioning.Nature reviews. Neuroscience 07/2002; 3(6):453-62. · 30.44 Impact Factor
Article: Stress, memory and the amygdala.[show abstract] [hide abstract]
ABSTRACT: Emotionally significant experiences tend to be well remembered, and the amygdala has a pivotal role in this process. But the efficient encoding of emotional memories can become maladaptive - severe stress often turns them into a source of chronic anxiety. Here, we review studies that have identified neural correlates of stress-induced modulation of amygdala structure and function - from cellular mechanisms to their behavioural consequences. The unique features of stress-induced plasticity in the amygdala, in association with changes in other brain regions, could have long-term consequences for cognitive performance and pathological anxiety exhibited in people with affective disorders.Nature Reviews Neuroscience 07/2009; 10(6):423-33. · 26.48 Impact Factor
Article: Stress and cognitive function.[show abstract] [hide abstract]
ABSTRACT: Stress affects cognition in a number of ways, acting rapidly via catecholamines and more slowly via glucocorticoids. Catecholamine actions involve beta adrenergic receptors and also availability of glucose, whereas glucocorticoids biphasically modulate synaptic plasticity over hours and also produce longer-term changes in dendritic structure that last for weeks. Prolonged exposure to stress leads to loss of neurons, particularly in the hippocampus. Recent evidence suggests that the glucocorticoid- and stress-related cognitive impairments involving declarative memory are probably related to the changes they effect in the hippocampus, whereas the stress-induced catecholamine effects on emotionally laden memories are postulated to involve structures such as the amgydala.Current Opinion in Neurobiology 05/1995; 5(2):205-16. · 7.44 Impact Factor
Diffusion-Weighted MRI and Quantitative Biophysical
Modeling of Hippocampal Neurite Loss in Chronic Stress
Peter Vestergaard-Poulsen1*, Gregers Wegener2, Brian Hansen1, Carsten R. Bjarkam5, Stephen J.
Blackband4, Niels C. Nielsen3, Sune N. Jespersen1
1Center for Functionally Integrative Neuroscience, Aarhus University, Aarhus, Denmark, 2Center for Basic Psychiatric Research, Aarhus Psychiatric University Hospital,
Risskov, Denmark, 3Center for Insoluble Protein Structures and Interdisplinary Nanoscience Center, Department of Chemistry, Aarhus University, Aarhus, Denmark,
4Department of Neuroscience, McKnight Brain Institute and The National High Magnetic Field Laboratory, University of Florida, Gainesville, Florida, United States of
America, 5Department of Neurosurgery, Aarhus University Hospital, Aarhus, Denmark
Chronic stress has detrimental effects on physiology, learning and memory and is involved in the development of anxiety
and depressive disorders. Besides changes in synaptic formation and neurogenesis, chronic stress also induces dendritic
remodeling in the hippocampus, amygdala and the prefrontal cortex. Investigations of dendritic remodeling during
development and treatment of stress are currently limited by the invasive nature of histological and stereological methods.
Here we show that high field diffusion-weighted MRI combined with quantitative biophysical modeling of the hippocampal
dendritic loss in 21 day restraint stressed rats highly correlates with former histological findings. Our study strongly indicates
that diffusion-weighted MRI is sensitive to regional dendritic loss and thus a promising candidate for non-invasive studies of
dendritic plasticity in chronic stress and stress-related disorders.
Citation: Vestergaard-Poulsen P, Wegener G, Hansen B, Bjarkam CR, Blackband SJ, et al. (2011) Diffusion-Weighted MRI and Quantitative Biophysical Modeling of
Hippocampal Neurite Loss in Chronic Stress. PLoS ONE 6(7): e20653. doi:10.1371/journal.pone.0020653
Editor: Joseph Najbauer, City of Hope National Medical Center and Beckman Research Institute, United States of America
Received December 29, 2010; Accepted May 6, 2011; Published July 1, 2011
Copyright: ? 2011 Vestergaard-Poulsen et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which
permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This research was funded by the Danish National Research Foundation (95093538-2458, project 100297) and The Danish Biotechnological Instrument
Centre. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: firstname.lastname@example.org
The adverse effects of chronic stress on physiology, learning and
memory are well described  and are known to be involved in
the development of anxiety disorders such as post-traumatic stress
disorder (PTSD) and major depressive illness . The stress
response of the body acts via a glucocorticoid-mediated negative
feedback on the hypothalamus-pituitary-adrenal axis upon which
the hippocampus has a major regulatory role . Being of central
importance in spatial learning and memory, the hippocampus is a
highly vulnerable brain structure susceptible to the damaging
effects of chronic stress and circulating adrenal steroids .
Animal studies of increased glucocorticoid levels induced either
by stressful environments or by exogenous administration have
identified three major effects on neural plasticity in the hip-
pocampus: First, modified intrinsic excitability and activity-
dependent synaptic plasticity have been reported in several
studies. For instance, a distinct reduction in long-term potentiation
(LTP) coupled with increased long-term depression (LTD) in the
CA1 pyramidal cell region was found in rats exposed to even mild
levels of stress . Second, high levels of glucocorticoids or stress
can result in inhibition or cessation of neurogenesis in the dentate
gyrus (DG) of the hippocampus [6,7]. Third, there is profound
evidence that chronic high level stress or glucocorticoid admin-
istration in rats and primates is associated with loss of apical
dendritic material of pyramidal neurons and even neuronal death
[8,9], especially in CA3 [9,10], but also in the CA1 region . In
moderate durations of stress or glucocorticoid administration, this
effect has been reported in CA3 pyramidal neurons and correlates
with a reduced performance in both short-term spatial and other
types of memory tasks in rats [12–18]
dendritic length and branching numbers of neurons visualized
by histology . Hence, further in vivo investigation of dendritic
remodeling during development, monitoring or treatment of
stress and stress-related diseases such as depression and PTSD
is limited by the absence of a sensitive and non-invasive
Magnetic Resonance Imaging (MRI) sensitized to water self
diffusion (DWI) has proven to be uniquely sensitive to subtle
changes in brain tissue microstructure in a number of reports
[19,20], notably early in vivo detection of regional cerebral
ischemia . The intrinsic water diffusion anisotropy of white
matter can be measured in vivo by diffusion tensor MR imaging
(DTI) and used for in vivo axonal fiber tracking in the brain ,
thus proving useful in studies of a number of white matter diseases
. Diffusion anisotropy derived from the diffusion tensor model
– which simply models the water diffusion in each voxel as an
ellipsoid - has been used to explore the effects of stress on white
matter development. For instance, in monkeys exposed to
intermittent social separation stress, it was found that the
ventromedial prefrontal white matter diffusion anisotropy was
greater than in controls . Early life stress was associated with
reduced fractional diffusion anisotropy in the anterior internal
capsule in monkeys . In humans, children with PTSD were
shown to have reduced diffusion anisotropy in the medial and
PLoS ONE | www.plosone.org1 July 2011 | Volume 6 | Issue 7 | e20653
posterior corpus callosum which may be attributed to reduced
myelination or changes in axonal structure .
In gray matter, cortical diffusion anisotropy has been shown to
be sensitive to the level of neuronal migration in developing ferret
brains . While having an extreme sensitivity to changes
occurring in the underlying tissue microstructure, the parameters
that describe the diffusion-weighted signal, such as the diffusion
coefficient or the diffusion anisotropy, suffer from a lack of
specificity to the microstructural geometry of neuronal tissue .
Many attempts have been made to develop a biophysical model
of the diffusion-weighted signal capable of quantifying the
microstructure of the neuronal system in terms of physically
both gray and white matter brain tissue  was recently validated
towards both quantitative light- and electron microscopy through
demonstrating a very strong correlation with the neurite density
obtained from DWI in several brain regions . We hypothesize
that appropriate DWI and this specific type of biophysical modeling
of neuronal tissue may offer the ability to detect and quantify the
underlying regional dendritic remodeling observed in standardized
studies of chronic stress. In this study, a validated model of neurite
density  of high field DWI data were used to detect the regional
microstructural changes that occur in the rat hippocampus after a
21 day period of standardized chronic stress.
Materials and Methods
Animals and Stress paradigm
Ten adult male Wistar rats aged 9-10 weeks (300 grams,
Taconic MB, Denmark) were randomly and evenly divided into a
group receiving exposure to stress and a control group. All animal
experiments were approved in accordance with all guidelines and
regulations of The Danish National Committee for Ethics in
Animal Experimentation (Approved by the Local Erthics comittee
for Aarhus County, Denmark, authorization number 2007/561-
1378). The animals were housed in groups of two with ad libitum
access to food and water. All animals were maintained in a
temperature controlled room, with a light/dark cycle of 12/
12 hours (lights on at 06.00 a.m.). During the three week stress
period, the control rats remained in their home cages with daily
handling, i.e. each rat was taken out of the cage, gently handled for
approximately 20 seconds, and subsequently returned to the home
cage. In a separate room, the stress group was subjected to a
6 hour daily restraint stress schedule (09.00 a.m. to 3.00 p.m.) for
21 days in transparent acrylic restrainers secured at the head and
tail, with an intensive light source above (1000 Lux).
Animals were euthanized by an injection of 5 ml Sodium-
pentobarbital (20 mg/ml). When deep reflexes were no longer
present, the animals were exsanguinated during intraaortic
perfusion with isotonic saline containing heparin (10 IU/mL),
followed by perfusion-fixation using 4% paraformaldehyde
dissolved in phosphate-buffered saline (pH 7.4). The brains were
removed and immersion-fixed in a fresh 4% formaldehyde
solution at room temperature and stored for two weeks. The
brains were bisected mid-sagittally, and the right hippocampi
removed by gross dissection using the lateral ventricle as a
reference . The samples were then immersion-fixed in fresh
4% formaldehyde solution until DWI data collection.
DWI data collection
Prior to MRI, the brains were washed for 48 hours in a
phosphate buffered saline solution (pH=7.4) in order to remove
formalin and reduce signal loss . Each specimen was then
placed in a thin-walled, standard 5 mm diameter NMR glass tube
with the longitudinal direction of the excised hippocampus parallel
to the tube axis, and positioned in a 16.4 T (700 MHz for1H)
vertical, wide bore Bruker Avance II NMR spectrometer (Bruker
BioSpin GmbH, Rheinstetten, Germany), equipped with a
gradient system capable of up to 300 Gauss/cm. All experiments
were performed at a controlled temperature of 21uC.
A standard spin echo Stejskal-Tanner diffusion-weighted
sequence was used to acquire a total of 54 diffusion directions
chosen from a 54 point spherical 9-design . A total of 9 shells
(b-values=0, 2000, 3000, 4000, 5000, 6000, 8000, 10000,
15000 s/mm2) were acquired with 6 unique directions on each
shell. The remaining diffusion and imaging parameters were as
follows: TR=2.5 s, TE=14.3 ms, data matrix=64664, field of
view =4.5 mm 64.5 mm, axial slice thickness =0.32 mm, and
D/d =8/2 ms. Four averages were acquired per direction. Total
acquisition time was 9 h 36 min.
Biophysical model of water-self diffusion in neuronal
The model is based on a biophysical description of brain
microstructure  with the fundamental assumption that water
diffusion can be described in terms of two non-exchanging
components. One component is associated with diffusion in
cylindrically symmetric structures, such as dendrites and axons
(collectively called neurites) with exchange of water being
sufficiently slow to be considered impermeable on the time scale
of the diffusion imaging experiment. The net signal from this
component then arises as a sum of the signal from all neurites
weighted by an orientation distribution function, i.e. a probability
density function specifying the number of neurites in every
direction. The second component of the diffusion signal accounts
for diffusion everywhere else, in particular in cell bodies,
extracellular space, and glia cells. Here, diffusion is hindered
and approximated by Gaussian isotropic diffusion with an effective
diffusion constant. Several cytoarchitectural parameters can be
extracted from this framework . Here, we estimate the voxel-
wise neurite density from the volume fraction of the net signals
from modeled neurites. The voxel-wise mean diffusivity of the
extra-cylindrical space was estimated for use only in defining the
hippocampal subregions (see next section).
Analysis and statistics
An experienced neuroanatomist (CRB), blinded to which group
the rat belonged, defined regions of interest (ROI) in the stratum
oriens, pyramidal cell layer, stratum radiatum and the stratum
lacunosum moleculare of the central CA1 and CA3 regions, as
well as the molecular and the granule cell layer of the DG. The
definition of the ROI was performed on the mean diffusivity maps
with very high signal in cell layers at a position of the rostral-
caudal axis of approximately –2.9 mm relative to bregma .
Two data sets were discarded, one from each group. One data set
contained a tear in the hippocampus at the position of the rostral-
caudal axis where we applied the analysis, and the other data set
contained image artifacts of unknown origin in a portion of the
diffusion directions. Despite efforts to perform an analysis without
these directions, a reliable analysis was not obtained.
A Wilcoxon rank sum test was applied to the group mean of all
pixels in the defined ROIs obtained from the modeled neurite
density maps. Differences were considered significant at a level of
P,0.05 (uncorrected for multiple comparisons between the
different regions). The relative difference of the modeled neurite
density of the groups was compared to formsimilar literature light
Dendritic Remodeling by Diffusion MRI
PLoS ONE | www.plosone.org2July 2011 | Volume 6 | Issue 7 | e20653
microscopy studies of the rat hippocampus after 21 days of
standardized chronic stress.
The modeled neurite density of the apical CA3 and CA1
regions was lower in the stressed animals than in the controls as
can be seen on maps of the hippocampal normalized neurite
density (Figure 1A,B). The modeled neurite density in both the
CA3 stratum radiatum and the stratum lacunosum moleculare of
the stressed group were significantly lower than in the controls
(Table 1, Figure 2). A similar reduction was seen in the CA1
(Table 1, Figure 2). However, no statistical difference in modeled
neurite density of the pyramidal cell layers in both CA3 and CA1
could be detected between groups (Table 1, Figure 2). Statistical
difference in modeled neurite density of the stratum oriens layer
could be detected between groups in CA1 but not CA3 (Table 1,
The modeled neurite density in the DG granule cell layer of the
stressed rats was lower than in the controls (Table 1, Figure 1 and
3). The DG molecular layer neurite density of the stressed group
was likewise significantly lower than in the controls (Table 1,
Figure 1 and 3). Interestingly, the variance of the modeled neurite
density in the stressed group, in regions where there were larger
differences from the control group, was generally lower than the
control group. This is seen especially in the granule and molecular
layers of the DG (see Figure 3).
In a 21 day restraint-stressed rat model we found significantly
lower DWI measured neurite density in the hippocampal CA3
apical areas compared to controls. No differences were found in
the stratum oriens region or the pyramidal cell layer.
Our findings in CA3 are consistent with earlier histological
studies in rats exposed to similar 6 hour/21 day restraint stress
models. These studies all showed that while the total dendritic
length of the basal dendrites in stratum oriens did not change, the
apical dendrites of these neurons were reduced by approximately
18 to 36% [13,15–17,35–38]; an average reduction of 27%.
The general histological findings of a reduced dendritic length
confined to only the apical part of hippocampal pyramidal
neurons due to stress were also reflected in a reduced neurite
density estimated by DWI in our study. We note that the dendritic
length cannot be directly compared to the neurite density (which is
the volume fraction of dendrites and axons in the biophysical
model employed) obtained in this study. However, under the
reasonable assumption that the dendritic length correlates quasi-
linearly to the volume fraction of dendrites in the dendritic tree,
our reported 24% average reduction of apical dendrite density in
the CA3 region is strikingly similar to the findings of former
histological studies which found an average dendritic length
reduction of 27% when employing similar 21 day restraint stressed
models in rats.
Previous 21 day restraint stressed rat studies did not include the
CA1 region [13,15–17,35,37,39] or found no changes in the
length of the apical dendritic tree CA1 . We have found a
reduction in CA1 modeled neurite density in the stratum radiatum
of 33%. A more recent study supports CA1 dendritic retraction by
demonstrating a 33% decrease of the terminal segment length of
the dendritic tree in CA1 and a 25% reduction of the total
dendritic length in the CA3 region following 30 days restraint
stress . Also, only 6 days of activity stress in rats caused a 33%
reduction of the total dendritic length in CA1 apical dendrites
Figure 1. Neurite density maps of stressed and control rat hippocampi. (A) stressed rats, (B) control rats. The color bar shows the
normalized neurite density. Note: the highest red intensity on the color bar refers to lowest neurite density.
Table 1. Normalized hippocampal modeled neurite density
in stressed and control groups of rats.
CL0.2460.05 0.3360.05 0.11
LM 0.3160.050.4160.04 0.06
CA3 SO0.3660.06 0.4160.120.69
ML0.2860.03* 0.3660.03 0.03
All values are mean 6 standard deviation of the subregion (n=4 for both
groups). Subregion abbreviations are: stratum oriens (SO), pyramidal cell layer
(CL), stratum radiatum (SR), stratum lacunosum moleculare layer (LM), granule
cell layer (GL), molecular layer (ML).
*Significant difference between stressed and control group (Wilcoxon rank sum
test P,0.05 uncorrected for multiple comparisons).
Dendritic Remodeling by Diffusion MRI
PLoS ONE | www.plosone.org3 July 2011 | Volume 6 | Issue 7 | e20653
Subtle differences in the restraint stress induction models may
affect the stress response considerably and could be part of the
explanation for these apparent differences . Further, despite
being invented a century ago, the mechanisms that are responsible
for Golgi-Cox staining have not yet been completely elucidated
suggesting that differences in fixation procedures might affect the
staining of neurons and thereby the measurement of total dendritic
length and branching numbers .
Future studies should include both histology and DWI on the
same specimens to more directly compare the specificity of these
methods in the estimation of regional neurite density. This would
also shed more light on the potential confounds from the possible
sensitivity of DWI to other structural differences than dendritic
retraction between groups of stressed and control animals. Its
known that i.e. the apical, but also the basal dendritic spine density
in CA1 are increased due to stress induction  which might
affect water diffusion dynamics in addition to the changed water
diffusion dynamics due to dendritic retraction. Also changes in the
dendritic spine density of the CA3 (the socalled thorny
excrecences) shown in induced stress could principly have an
effect on the DWI. It is still controversial in the histolgy literature
whether this is the case in stress [11,44]. A close relationship
between neurite structure and hippocampus-dependent learning
and memory using a water maze has been demonstrated in rats
, an indication that such measures should be included in future
correlative studies of DWI and histological and stereological
analysis of brains from stressed animals.
There was a decreased - but not significant (P,0.06) - of
modeled neurite density in the DG molecular layer of 25% in our
21 day study, while an earlier study found a 38% reduction of total
dendritic length in a 30 day restraint stress rat model .
In the DG granule layer of the the stressed group a 26%
reduction of the modeled neurite density was found in this study.
However, a reduction of the estimated normalized neurite density
within the employed model of diffusion must be interpreted as a
reduced volume fraction of cylindrical segments (dendrites or
processes) and thus, an increased volume fraction of the second
compartment. In the cell layer, this second compartment is the
sum of the extracellular space and the cell somas: both of which
exhibit hindered, but approximately Gaussian isotropic diffusion.
Thus, this model which was developed to estimate neurite density
Figure 2. CA3 and CA1 normalized neurite density of stratum oriens (SO), cell layer (CL), stratum radiatum (SR) and stratum
lacunosum moleculare layer (LM). Individual data from controls rats (green dots) and stressed rats (red) are shown in conjunction with the mean
values and standard deviation (black).*P,0.05.
Figure 3. DG normalized neurite density of granule cell layer
(GL) and molecular layer (ML). Individual data from controls rats
(green dots) and stressed rats (red) are shown in conjunction with the
mean values and standard deviation (black).*P,0.05.
Dendritic Remodeling by Diffusion MRI
PLoS ONE | www.plosone.org4 July 2011 | Volume 6 | Issue 7 | e20653
cannot distinguish between these two subcompartments in the cell
layers, and we conclude only that the total volume fraction of
Gaussian isotropic diffusion is increased after 21 day stress.
Nevertheless, the finding that DWI detects a change in the
water diffusion properties in the cell layers of the DG is highly
interesting because it indicates sensitivity to the structural changes
associated with neurogenesis. Earlier reports using corticosterone
injections of rats or psycho-social stress (in tree shrews) have
demonstrated a reduced rate of progenitor cell birth in the hilus
[6,45]. Progenitor cells have been suggested to migrate into the
granule cell layer within a 21 day period . This could indicate
that there is a decrease of cellular density in the granule layer
causing the extracellular space to expand which could be the basis
for the higher volume fraction of the Gaussian diffusion observed
here. As no changes have been detected in the pyramidal cell
layers of CA1/CA3 in this study, matching the general findings of
no neuronal loss in 21-day stress studies, this suggests that DWI is
a marker of structural changes in the cellular region, although not
a specific one. Further research on biophysical models that include
the cellular layer is needed to further explore the specificity to the
structural changes in cell layers.
Chronic stress selectively reduces the total hippocampal volume
by 3% in rats after a 21 day restraint stress period . It is
currently not known how regional shrinkages generate a reduction
of the total hippocampal volume or how the intra- and extra-
cellular fractions are modulated as a result of dendritic retraction.
A combination of mapping the regional neurite density by DWI in
the entire hippocampus and total hippocampal volume estimation
might explain this and should be included in further investigations.
In this in vitro study of a standard restraint stressed rat model, we
have demonstrated the sensitivity of DWI to underlying micro-
structural changes, probably to a loss in dendritic material, that
agree very well with former histological findings. We predict that a
number of stress research fields can benefit from an imaging
method that is sensitive to changes in dendritic structure. First,
studies of how the type of stressor, context, duration, gender, age
and genes are related to neurite structure would be more feasible
with a non-invasive neuroimaging technique. Next, the reversibil-
ity of chronic stress has been the subject of many discussions .
Most studies show that the dendritic retraction in the hippocam-
pus after a 21 day period of restraint stress seems to be reversible
within approximately 10 days [36,47]. However, brain structures
involved in the control of the HPA-axis are interconnected, so the
prefrontal cortex and the hippocampus are both affected by
remodeling of neurons in the basolateral complex of the amygdala
(BLA) which is an important substrate in integrating the influence
of hormonal and neurotransmitter systems on memory consolida-
tion . Interestingly, and in contrast to the effects on the
hippocampus, chronic stress produces dendritic growth in the BLA
that does not recover even after a longer period in stress-free
environments . The continued hypertrophy of neurons in the
BLA was also reflected in a persistent state of heightened anxiety
and thus may be a consequence of a cellular substrate promoting
high anxiety levels. An in vivo neuroimaging method sensitive to
dendritic remodeling would be an important tool to understand
these mechanisms and develop treatment of anxiety disorders like
PTSD, but would also be applicable to numerous other fields
including development, aging, seizures, ischemia, rehabilitation
and depression treatment.
However, there are a number of technical challenges to be
addressed before this method can be applied in vivo. First, much
shorter examination times are required. This implies that the DWI
acquisition must be simplified (i.e. reducing the number of b-
values, directions or averages) in a manner that does not preclude
reliable estimation of biophysical model parameters. Second,
the lower field strengths (typically 5–11 Tesla for animal magnets,
1.5–3 Tesla for human magnets) and lower performance of the
gradient systems in MRI systems for live animal and especially for
human studies will result in limited signal-to-noise ratio and spatial
resolution. The limited range of diffusion times and thereby echo
times for DWI in clinical scanners compared to experimental high
field systems with high performance gradient systems needs further
attention due to the sensitivity to the characteristic diffusion length
of water molecules and tissue subcompartmental T2relaxation.
Further studies, beginning with in vivo animal studies, are needed
to explore the ramifications and possible solutions to these issues.
In summary, this study shows that there are strong indications
that DWI is sensitive to the dendritic retraction of rat hippocampal
neurons that undergo a 21 day restraint stress. The regional degree
of hippocampal neuritic loss found by this method was in
agreement to neuritic loss measured using light microscopy in
earlier studies of 21 day restraint stress. Thus, DWI might support
or even in some cases substitute histology in a number of in vitro
applications as well as being the only present candidate for a non-
invasive in vivo neuroimaging method – apparently sensitive to
dendritic remodeling - in studies of anxiety disorders, depression,
several neuro-degenerative disorders and development and aging
of the brain. Further validation studies and technical developments
are needed to fully elucidate this potential.
We thank J. Skewes, A. Møller, J. Frandsen, T. Vosegaard, and T. Shepherd
for their support with the design of this study. We thank C. Frith, Kim
Mouridsen, and J. Flint for their valuable comments on this manuscript.
Conceived and designed the experiments: PV-P GW SJB SNJ. Performed
the experiments: PV-P GW BH NCN SNJ. Analyzed the data: PV-P CRB
SJB SNJ. Contributed reagents/materials/analysis tools: GW CRB NCN.
Wrote the paper: PV-P GW BH SJB SNJ.
1. Kim JJ, Diamond DM (2002) The stressed hippocampus, synaptic plasticity and
lost memories. Nat Rev Neurosci 3: 453–462.
2. Roozendaal B, McEwen BS, Chattarji S (2009) Stress, memory and the
amygdala. Nat Rev Neurosci 10: 423–433.
3. McEwen BS, Sapolsky RM (1995) Stress and cognitive function. Curr Opin
Neurobiol 5: 205–216.
4. McEwen BS (1999) Stress and hippocampal plasticity. Annu Rev Neurosci 22:
5. Xu L, Anwyl R, Rowan MJ (1997) Behavioural stress facilitates the indu-
ction of long-term depression in the hippocampus. Nature 387: 497–
6. Cameron HA, Gould E (1994) Adult neurogenesis is regulated by adrenal
steroids in the dentate gyrus. Neuroscience 61: 203–209.
7. Gould E, McEwen BS, Tanapat P, Galea LA, Fuchs E (1997) Neurogenesis in
the dentate gyrus of the adult tree shrew is regulated by psychosocial stress and
NMDA receptor activation. J Neurosci 17: 2492–2498.
8. Uno H, Tarara R, Else JG, Suleman MA, Sapolsky RM (1989) Hippocampal damage
associated with prolonged and fatal stress in primates. J Neurosci 9: 1705–1711.
9. Kerr DS, Campbell LW, Applegate MD, Brodish A, Landfield PW (1991)
Chronic stress-induced acceleration of electrophysiologic and morphometric
biomarkers of hippocampal aging. J Neurosci 11: 1316–1324.
10. Sapolsky RM, Krey LC, McEwen BS (1985) Prolonged glucocorticoid exposure
reduces hippocampal neuron number: implications for aging. J Neurosci 5:
11. Sousa N, Lukoyanov NV, Madeira MD, Almeida OF, Paula-Barbosa MM
(2000) Reorganization of the morphology of hippocampal neurites and synapses
Dendritic Remodeling by Diffusion MRI
PLoS ONE | www.plosone.org5 July 2011 | Volume 6 | Issue 7 | e20653
after stress-induced damage correlates with behavioral improvement. Neurosci-
ence 97: 253–266.
12. McEwen BS, Albeck D, Cameron H, Chao HM, Gould E, et al. (1995) Stress
and the brain: a paradoxical role for adrenal steroids. Vitam Horm 51: 371–402.
13. Magarinos AM, McEwen BS (1995) Stress-induced atrophy of apical dendrites
of hippocampal CA3c neurons: involvement of glucocorticoid secretion and
excitatory amino acid receptors. Neuroscience 69: 89–98.
14. Woolley CS, Gould E, McEwen BS (1990) Exposure to excess glucocorticoids
alters dendritic morphology of adult hippocampal pyramidal neurons. Brain Res
15. Watanabe Y, Gould E, Daniels DC, Cameron H, McEwen BS (1992)
Tianeptine attenuates stress-induced morphological changes in the hippocam-
pus. Eur J Pharmacol 222: 157–162.
16. Watanabe Y, Gould E, McEwen BS (1992) Stress induces atrophy of apical
dendrites of hippocampal CA3 pyramidal neurons. Brain Res 588: 341–345.
17. Watanabe Y, Gould E, Cameron HA, Daniels DC, McEwen BS (1992)
Phenytoin prevents stress- and corticosterone-induced atrophy of CA3
pyramidal neurons. Hippocampus 2: 431–435.
18. Magarinos AM, McEwen BS, Flugge G, Fuchs E (1996) Chronic psychosocial
stress causes apical dendritic atrophy of hippocampal CA3 pyramidal neurons in
subordinate tree shrews. J Neurosci 16: 3534–3540.
19. Katz M, Liu C, Schaer M, Parker KJ, Ottet MC, et al. (2009) Prefrontal
plasticity and stress inoculation-induced resilience. Dev Neurosci 31: 293–299.
20. Jackowski AP, Douglas-Palumberi H, Jackowski M, Win L, Schultz RT, et al.
(2008) Corpus callosum in maltreated children with posttraumatic stress
disorder: a diffusion tensor imaging study. Psychiatry Res 162: 256–261.
21. Moseley ME, Cohen Y, Mintorovitch J, Chileuitt L, Shimizu H, et al. (1990)
Early detection of regional cerebral ischemia in cats: comparison of diffusion-
and T2-weighted MRI and spectroscopy. Magn Reson Med 14: 330–46.
22. Basser PJ, Pierpaoli C (1996) Microstructural and physiological features of tissues
elucidated by quantitative-diffusion-tensor MRI. J Magn Reson B 111: 209–219.
23. Le Bihan D (2003) Looking into the functional architecture of the brain with
diffusion MRI. Nat Rev Neurosci 4: 469–480.
24. Coplan JD, Abdallah CG, Tang CY, Mathew SJ, Martinez J, et al. (2010) The
role of early life stress in development of the anterior limb of the internal capsule
in nonhuman primates. Neurosci Lett 480: 93–96.
25. Kroenke CD, Taber EN, Leigland LA, Knutsen AK, Bayly PV (2009) Regional
patterns of cerebral cortical differentiation determined by diffusion tensor MRI.
Cereb Cortex 19: 2916–2929.
26. Stanisz GJ, Szafer A, Wright GA, Henkelman RM (1997) An analytical model of
restricted diffusion in bovine optic nerve. Magn Reson Med 37: 103–11.
27. Vestergaard-Poulsen P, Hansen B, Ostergaard L, Jakobsen R (2007)
Microstructural changes in ischemic cortical gray matter predicted by a model
of diffusion-weighted MRI. J Magn Reson Imaging 26: 529–540.
28. Ong HH, Wright AC, Wehrli SL, Souza A, Schwartz ED, et al. (2008) Indirect
measurement of regional axon diameter in excised mouse spinal cord with q-
space imaging: simulation and experimental studies. Neuroimage 40:
29. Jespersen SN, Kroenke CD, Ostergaard L, Ackerman JJ, Yablonskiy DA (2007)
Modeling dendrite density from magnetic resonance diffusion measurements.
Neuroimage 34: 1473–1486.
30. Jespersen SN, Bjarkam CR, Nyengaard JR, Chakravarty MM, Hansen B, et al.
(2010) Neurite density from magnetic resonance diffusion measurements at
ultrahigh field: comparison with light microscopy and electron microscopy.
Neuroimage 49: 205–216.
31. Alger BE (1984) Characteristics of a slow hyperpolarizing synaptic potential in
rat hippocampal pyramidal cells in vitro. J Neurophysiol 52: 892–910.
32. Shepherd TM, Thelwall PE, Stanisz GJ, Blackband SJ (2009) Aldehyde fixative
solutions alter the water relaxation and diffusion properties of nervous tissue.
Magn Reson Med 62: 26–34.
33. Hardin RH, Sloane NJA (1996) McLaren’s improved snub cube and other new
spherical designs in three dimensions. Discrete & Computational Geometry 15:
34. Paxinos, G (2005) The rat brain in stereotactical coordinates. Elsevier Academic
35. Galea LA, McEwen BS, Tanapat P, Deak T, Spencer RL, et al. (1997) Sex
differences in dendritic atrophy of CA3 pyramidal neurons in response to
chronic restraint stress. Neuroscience 81: 689–697.
36. Conrad CD, LeDoux JE, Magarinos AM, McEwen BS (1999) Repeated
restraint stress facilitates fear conditioning independently of causing hippocam-
pal CA3 dendritic atrophy. Behav Neurosci 113: 902–913.
37. McLaughlin KJ, Gomez JL, Baran SE, Conrad CD (2007) The effects of chronic
stress on hippocampal morphology and function: an evaluation of chronic
restraint paradigms. Brain Res 1161: 56–64.
38. McLaughlin KJ, Baran SE, Wright RL, Conrad CD (2005) Chronic stress
enhances spatial memory in ovariectomized female rats despite CA3 dendritic
retraction: possible involvement of CA1 neurons. Neuroscience 135: 1045–1054.
39. Conrad CD, Galea LA, Kuroda Y, McEwen BS (1996) Chronic stress impairs
rat spatial memory on the Y maze, and this effect is blocked by tianeptine
pretreatment. Behav Neurosci 110: 1321–1334.
40. Lambert KG, Buckelew SK, Staffiso-Sandoz G, Gaffga S, Carpenter W, et al.
(1998) Activity-stress induces atrophy of apical dendrites of hippocampal
pyramidal neurons in male rats. Physiol Behav 65: 43–49.
41. Buynitsky T, Mostofsky DI (2009) Restraint stress in biobehavioral research:
Recent developments. Neurosci Biobehav Rev 33: 1089–1098.
42. Pasternak JF, Woolsey TA (1975) On the ‘‘selectivity’’ of the Golgi-Cox method.
J Comp Neurol 160: 307–312.
43. Shors TJ, Chua C, Falduto J (2001) Sex differences and opposite effects of stress
on dendritic spine density in the male versus female hippocampus. J Neurosci 21:
44. Sunanda, Rao MS, Raju TR (1995) Effect of chronic restraint stress on dendritic
spines and excrescences of hippocampal CA3 pyramidal neurons—a quantita-
tive study. Brain Res 694: 312–317.
45. Gould E, Woolley CS, Cameron HA, Daniels DC, McEwen BS (1991) Adrenal
steroids regulate postnatal development of the rat dentate gyrus: II. Effects of
glucocorticoids and mineralocorticoids on cell birth. J Comp Neurol 313:
46. Lee T, Jarome T, Li SJ, Kim JJ, Helmstetter FJ (2009) Chronic stress selectively
reduces hippocampal volume in rats: a longitudinal magnetic resonance imaging
study. Neuroreport 20: 1554–1558.
47. Luine V, Villegas M, Martinez C, McEwen BS (1994) Repeated stress causes
reversible impairments of spatial memory performance. Brain Res 639:
48. Dyrby TB, Baare WF, Alexander DC, Jelsing J, Garde E, et al. (2011) An ex vivo
imaging pipeline for producing high-quality and high-resolution diffusion-
weighted imaging datasets. Hum Brain Mapp 32: 544–563.
Dendritic Remodeling by Diffusion MRI
PLoS ONE | www.plosone.org6 July 2011 | Volume 6 | Issue 7 | e20653