Directory UMM :Data Elmu:jurnal:B:Brain Research:Vol887.Issue1.Dec2000:

Brain Research 887 (2000) 34–45
www.elsevier.com / locate / bres

Research report

Mild hypothermia decreases the incidence of transient ADC reduction
detected with diffusion MRI and expression of c-fos and hsp70
mRNA during acute focal ischemia in rats
Anthony Mancuso a,e , *, Nikita Derugin a,c , Kazushi Hara a , Frank R. Sharp a,b,d ,
Philip R.a Weinstein
a

Department of Neurological Surgery, University of California at San Francisco and the Department of Veterans Affairs Medical Center,
San Francisco, CA 94121, USA
b
Department of Neurology, University of California at San Francisco and the Department of Veterans Affairs Medical Center, San Francisco,
CA 94121, USA
c
Department of Radiology, University of California at San Francisco and the Department of Veterans Affairs Medical Center, San Francisco,
CA 94121, USA
d

Department of Neurology, University of Cincinnati, Cincinnati, OH 45267, USA
e
Department of Radiology, University of Pennsylvania, Philadelphia, PA 19104, USA
Accepted 12 September 2000

Abstract
The effects of mild hypothermia on the apparent diffusion coefficient of water (ADC) and expression of c-fos and hsp70 mRNA were
examined during acute focal cerebral ischemia. Young adult rats were subjected to 60-min middle cerebral artery occlusion under either
normothermia (37.58C) or hypothermia (338C). Diffusion-weighted echo-planar magnetic resonance imaging was used to monitor changes
in ADC throughout the ischemic period. Perfusion MRI with dysprosium contrast was used at the end of the ischemic period to verify that
the occlusion was successful. C-fos and hsp70 mRNA expression were examined with in situ hybridization at the end of the ischemic
period. The results indicate that the size of the region that exhibited reduced ADC was smaller during hypothermia than during
normothermia. Hypothermia also decreased the frequency of occurrence of transient ADC reductions, especially in dorsal aspects of
cortex. Expression of both c-fos and hsp70 mRNA were markedly reduced by hypothermia. Transient ADC reduction and c-fos
expression are associated with spreading depression, which is believed to contribute to lesion expansion during acute focal ischemia. The
results suggest that part of the neuroprotective effect of hypothermia may be due to a reduced incidence of spreading depression.
 2000 Elsevier Science B.V. All rights reserved.
Theme: Disorders of the nervous system
Topic: Ischemia
Keywords: Hypothermia; Diffusion MRI; Apparent diffusion coefficient; Focal cerebral ischemia; C-fos; Hsp70

1. Introduction
Hypothermia is highly effective in reducing injury
during cerebral ischemia. Deep to moderate hypothermia

*Corresponding author. Department of Radiology / 4283, B1 StellarChance Laboratory, 422 Curie Boulevard, University of Pennsylvania,
Philadelphia, PA 19104, USA. Tel.: 11-215-898-1805; fax: 11-215-5732113.
E-mail address: mancuso@rad.upenn.edu (A. Mancuso).

(20 to 308C) has been found to be useful clinically
following head trauma and during surgical procedures that
disrupt the cerebral circulation [23]. However, it is also
associated with numerous adverse affects such as cardiac
dysfunction [45,62]. Mild hypothermia (|338C), while less
protective than deep hypothermia [4,24], has gained considerable interest recently because it provides significant
neuroprotection with far fewer complications [36]. The
effect of mild hypothermia on clinical ischemic stroke has
not yet been examined in large clinical trials [21,36].
In animal models of global ischemia, mild hypothermia

0006-8993 / 00 / $ – see front matter  2000 Elsevier Science B.V. All rights reserved.
PII: S0006-8993( 00 )02963-2

A. Mancuso et al. / Brain Research 887 (2000) 34 – 45

can reduce infarct size and the extent of ischemic neuronal
damage in non-infarcted regions [8,11,15,17]. For focal
ischemia, mild hypothermia can reduce the size of the
necrotic region [32,35,50] and the extent of induction of
apoptosis in non-necrotic regions [46]. It can also help to
prevent neuronal and endothelial damage associated with
reperfusion injury, if it is initiated during [28,35] or after
ischemia [41]. Applying hypothermia during the acute
phase of prolonged (3 to 4 day) permanent ischemia can
reduce the rate of pathophysiological changes, but it
generally does not reduce the final extent of injury [53].
Hypothermia has many effects on the underlying
pathophysiological changes that occur during cerebral
ischemia. It reduces cerebral metabolic rate (CMR) as
evidenced by decreased glucose and oxygen consumption

rates [4], which delays the onset of energy failure and
terminal ischemic depolarization. Direct measurement of
ATP levels by 31 P NMR and invasive methods has
confirmed that hypothermia slows the rate of energy
depletion [42,63]. However, this does not completely
account for its neuroprotective capacity since anesthetics
that diminish CMR to the same extent as hypothermia do
not provide the same degree of neuroprotection [52,61].
Hypothermia also reduces neurotransmitter release during ischemia, including glutamate [3,12,28,64], aspartate
[7], glycine, gamma-aminobutyric acid [51], and dopamine
[12]. Glutamate is a potent excitotoxin that can hasten
ischemic depolarization in regions where energy depletion
is incomplete [57]. In focal ischemia, hypothermia reduces
intra-ischemic increases of glutamate in penumbral [64]
but not core regions [28]. However, recent results suggest
that the reduction in glutamate release does not fully
account for the neuroprotective effect of hypothermia [66].
Diffusion-weighted imaging is sensitive to changes in
the apparent diffusion coefficient of water (ADC). Reduced ADC during ischemia is believed to be, at least in
part, the result of cellular edema, which occurs with

ischemic depolarization [60]. Such depolarization is primarily caused by energy failure [6,13,20] or by excitotoxic
injury from neurotransmitters [6]. Reduced ADC also
occurs during spreading depression due to abnormal ion
fluxes that produce temporary cellular edema [10,43].
Spreading depression is believed to be an important
component of lesion size expansion during focal ischemia
[27]. Additional phenomena that are not as yet completely
understood may also contribute to reduced ADC during
ischemia [25]. Diffusion-weighted imaging has been
shown previously to be sensitive to reduced water diffusion during hypothermic ischemia [31].
Focal ischemia induces immediate early genes including
c-fos [26,38] and stress genes including hsp70 [49] that
may be important in protecting the brain from ischemic
injury. We previously observed that during acute focal
ischemia, c-fos expression was associated with brief
transient ADC reduction [47], which was probably caused
by transient cellular depolarization [10]. Hsp70 expression

35

occurred in regions exhibiting both transient and permanent ADC reduction [47].
Hypothermia can affect postischemic expression of both
c-fos and hsp70 mRNA following temporary occlusion.
C-fos expression following temporary forebrain ischemia
is hastened by moderate (308C) hypothermia [33,39].
Cerebral hsp70 expression can be induced by 8-h sustained
mild hypothermia without ischemia [18]. Expression following brief forebrain ischemia is reduced by moderate
hypothermia [16,40]. The effect of hypothermia on c-fos
and hsp70 expression during acute ischemia has not been
reported.
In this work, the effect of intra-ischemic mild (338C)
hypothermia on both transient and persistent ADC reduction and expression of c-fos and hsp70 mRNA was
examined. The hypotheses tested were that hypothermia
would reduce the extent of persistent ADC reduction, the
incidence of transient ADC reduction and c-fos mRNA
expression.

2. Methods

2.1. Animal model

Male Sprague–Dawley rats weighing 270–325 g were
used. They were given free access to food and water prior
to use in experiments. Anesthesia was induced with 3%
isoflurane in a closed chamber and subsequently maintained with 1.5% isoflurane (both in 70 / 30 N 2 O / O 2 )
through an intratracheal 16-gauge Teflon  tube. A rodent
ventilator (Harvard Instruments, South Natick, MA) was
used for mechanical ventilation; respiratory adjustments
were made to maintain normal arterial blood gas levels.
Just before imaging, pancuronium bromide (0.2 mg in 0.2
ml of 0.9% saline) was administered intraperitoneally to
induce paralysis and thereby minimize motion artifacts.
Permanent focal ischemia was produced by occluding
the MCA as described previously [34]. A 3-0 monofilament nylon suture with a heat-rounded tip was introduced
into an internal carotid artery (ICA), through the stump of
a transected external carotid artery (ECA). It was advanced
22 mm past the bifurcation of the common carotid artery
(CCA) to block the origin of the MCA. The CCA proximal
to the occluded MCA was left patent. The duration of the
occlusion was 60 min.
Rats were randomly assigned to two groups, either

normothermic, 37.5618C (n57) or hypothermic, 33618C
(n511). Body temperature was monitored with a rectal
thermistor encased in a grounded electrical shield. A
thermostatically controlled warm water pad under the
animal was used to maintain rectal temperature within the
specified ranges. During imaging, the head was held in a
polycarbonate stereotactic frame constructed in our laboratory. The frame helped to eliminate brain motion and
reduce heat loss from the head of the rat. In this device, the

36

A. Mancuso et al. / Brain Research 887 (2000) 34 – 45

rectal temperature accurately represented the brain temperature [47]. Blood gas samples were taken from a
femoral artery catheter before and during the ischemic
period. Arterial blood pressure was monitored with the
same catheter.

2.2. Magnetic resonance methods
Diffusion-weighted echo-planar images were acquired

with an MBEST pulse sequence on a 2-Tesla Omega CSI
system (Bruker Medical Systems, Fremont, CA), equipped
with 20 gauss / cm self-shielded gradients. A 5.5-cm diameter low-pass birdcage coil was used for both pulse
transmission and signal reception. Half sine diffusionsensitizing gradients of up to 6 gauss / cm were used to
produce b-values of up to 1710 s / mm 2 . The diffusion
gradients were applied parallel to the long axis of the
brain. Other image acquisition parameters were: FOV550
mm, slice thickness52 mm, and acquisition matrix size5
1283128. The relaxation delay between images was 3.0 s.
Images were acquired for a single coronal slice in each rat,
centered at the level of bregma. A series of 8 to 11 images
were used to construct each ADC map. To verify that the
occlusion of the MCA was successful, perfusion imaging
was performed with dysprosium DTPA–BMA contrast at
the end of the 60-min ischemic period. Additional details
of the imaging methods were described previously [47].
ADC maps were constructed with an IDL (Research
Systems Inc., Boulder, CO) program written in our laboratory [47]. The program performed pixel-by-pixel leastsquares regression to the line: lnsS /Sod 5 2 bsADCd, where
S is the diffusion-gradient attenuated pixel intensity, So is
the unattenuated pixel intensity, and b is the attenuation

factor [44]. Values of b were calculated from the equation:
b 5s2Gdg /pd 2sD 2 d / 4d, where g is the gyromagnetic ratio
of the hydrogen nucleus, G is the diffusion gradient
magnitude, and d and D are the duration and inter-pulse
delay, respectively, for the diffusion gradients. The program was also used to compute ADC threshold maps and
to interpolate the results to 2563256 resolution. A reduction of 20%, relative to the average value of the nonischemic hemisphere, was assumed for distinguishing the
ADC lesion from the rest of the brain.
A second IDL program was used to identify regions
with transient reduction in ADC (TR-ADC). The program
initially masked the region with persistently reduced ADC
on all maps in each time series. Persistent reduction in
ADC (PR-ADC) was determined by averaging the final
three ADC maps and identifying the pixels with reductions
of 20% or more. The program was then used to determine
the size of the ADC lesions (as a function of time) within
the five regions of interest (ROI — 1, 2, 3, 4, and 5)
shown in Fig. 1. A region was considered positive for
TR-ADC if at least 30% of the region was transiently
included in the ADC lesion. Similarly, the same five
regions of interest were used to analyze the c-fos auto-

Fig. 1. Regions of interest (ROI) used to evaluate the association of
transient ADC reduction and c-fos expression and to determine the
frequency of significant transient ADC reduction.

radiographs. A 30% area threshold was assumed for
classifying a region as c-fos-positive.

2.3. In situ hybridization for c-fos and hsp70 mRNA
Frozen brains were cut with a cryostat (Hacker Instruments, Fairfield, NJ) into 20-mm-thick coronal sections.
The sections were analyzed for expression of c-fos and
hsp70 mRNA with modified versions [37] of methods
described by Schalling [54]. The probe used to detect
hsp70 was a synthetic oligonucleotide that corresponds to
highly conserved amino acids (122–129) near the 59-end
of the human hsp70 coding sequence [29]. For c-fos, the
sequence complementary to bases 142–186 was used [19].
Additional details were described previously [47].

2.4. Statistical analysis
A single-tailed t-test was used to determine differences
in ADC lesion size between the two groups. Differences in
the frequency of transient ADC reduction between the
normothermic and hypothermic groups in the regions of
interest of Fig. 1 and in c-fos-positive regions were
determined with the Mann–Whitney Rank Sum Test. The
Fisher Exact Test was used to determine if c-fos and hsp70
expression occurred more frequently during normothermia
than hypothermia. The association of TR-ADC reduction
with c-fos expression was evaluated with the Chi-Squared
Test. All statistical calculations were performed with
Sigma Stat 2.03 (SPSS, Chicago, IL). Values presented in
Tables 1, 2 and 4 are mean6S.D.

3. Results

3.1. Physiological data
Blood pressure and blood gas results for the pre- and

A. Mancuso et al. / Brain Research 887 (2000) 34 – 45

37

Table 1
Physiologic blood gas data during pre and intra-ischemic periods a
Normothermic group

Pre-isch
Intra-isch
a

Hypothermic group

MABP
(mmHg)

pH

PaCO 2
(mmHg)

PaO 2
(mmHg)

MABP
(mmHg)

pH

PaCO 2
(mmHg)

PaO 2
(mmHg)

113617
131617

7.4860.06
7.4460.06

3066
2967

143614
132616

105613
10364

7.5060.05
7.3760.07

3167
42619

158618
144620

All values are mean6S.D.

intra-ischemic periods are summarized in Table 1. The
blood pressure, pH, and PO2 values are all in the normal
physiological range. However, the PCO2 levels before
ischemia for both groups and during ischemia for the
normothermic group were slightly low. The potential
significance of this observation will be presented in the
discussion section.

3.2. Normothermic group
Fig. 2 shows typical ADC maps and autoradiographs for
rats in both the normothermic and hypothermic groups.
Regions with ADC reductions greater than 20% of the
mean value of the contralateral hemisphere (ADC lesion)
are shown in color. For the normothermic rat in Fig. 2A,
ischemia initially produced significant ADC reductions of
20–45% in the lateral basal ganglia. The region with
reduced ADC spread during the following 57 min and
eventually encompassed almost all of the MCA territory.
The second ADC map in the figure represents the average
of the final three ADC maps; the colored region for this
averaged result was defined as the persistently ischemic
region. On the third ADC map, the persistently ischemic
region is shown in dark gray. By digitally superimposing
this region on all of the ADC maps acquired during the
60-min ischemic period, regions exhibiting TR-ADC were
readily identified. For example, the ADC maps from 42 to
45 min after the onset of ischemia (lower part of Fig. 2A)
indicate that TR-ADC occurred in region 1 (defined in Fig.
1). The duration of ischemia (min:sec) is shown below
each ADC map.
The complete time-course for the ADC lesion size for
the rat in Fig. 2A (as a percentage of the ischemic
hemisphere) is shown in Fig. 3A. The data indicate that
much of the expansion of the ADC lesion occurred during

the first 20 min of ischemia. Transient expansion of the
ADC lesion into dorsal medial cortex occurred repeatedly.
The time-course for the percentage of ROI-1 (see Fig. 1)
included in the ADC lesion is shown in Fig. 3B. Three
significant transient increases occurred with a duration of
approximately 3 min. The maximum extent of the transient
increases was slightly more than 50% of ROI-1.
Results very similar to those shown in Fig. 2A were
observed in two other rats for the normothermic group,
which also showed transient ADC reduction and c-fos
expression in ROI-1. Smaller final ADC lesions were
observed in the remaining four rats. Two rats had ADC
lesions that encompassed all of the basal ganglia and a
small part of cortex. For both of these rats, TR-ADC
occurred repeatedly in much of cortex. The remaining two
rats had persistent ADC reduction in only part of the basal
ganglia; in these rats TR-ADC occurred over a large part
of the MCA territory.
For the rat in Fig. 2A, c-fos mRNA was induced in the
region where TR-ADC was observed. Hsp70 mRNA was
weakly induced over a large part of the MCA distribution.
It was induced to a much greater extent at the dorsomedial
edge of the MCA territory in cortex, where TR-ADC was
observed, but not in cingulate cortex.
C-fos expression was observed in all seven rats of the
normothermic group, almost exclusively in regions with
TR-ADC. Three rats exhibited c-fos in only medial frontal
and cingulate cortex, similar to the results shown in Fig.
2A. The remaining four rats exhibited c-fos in a larger
region of cortex, which included medial frontal and
cingulate cortex. Two rats with c-fos expression throughout cortex also showed c-fos in the lateral basal ganglia.
The time-course for ADC changes within the c-fos-positive
region of two rats is shown in Figs. 4A and B. The
ordinate of these graphs represents the percent of the

Table 2
Average ADC lesion size following 15, 30, 45 and 60 min of ischemia a
Time
(min of ischemia)

Normothermic
ADC lesion size
(% of hemisphere)

Hypothermic
ADC lesion size
(% of hemisphere)

P

15
30
45
60

32615
38615
38619
38617

1968
19611
20610
22613

,0.05
,0.01
,0.01
,0.05

a

All values are mean6S.D.

38

A. Mancuso et al. / Brain Research 887 (2000) 34 – 45

Fig. 2. ADC maps and autoradiographs for c-fos and hsp70 expression in rats subjected to 1 h permanent MCA occlusion. Typical normothermic results are
shown in A. The color scale indicates the extent of ADC reduction relative to the mean value of the non-ischemic hemisphere. ADC reductions ranging from 20 to
25% are shown in red; each further 5% increment in ADC reduction is indicated with a different color. The second ADC map is the average of the final three ADC
maps and the entire colored region was defined as the persistently ischemic region. This region is shown in dark gray in the third ADC map and in the time series
of ADC maps in the bottom of A. Transient ADC reduction was observed in medial frontal cortex, where c-fos was also observed. Hsp70 was weakly expressed
throughout the region with persistently reduced ADC and strongly expressed at the dorsomedial edge of this region in cortex. Many rats of the hypothermic group
showed little transient ADC reduction (TR-ADC), c-fos or hsp70 expression (B). However, four rats in the hypothermic group showed a single incident of
TR-ADC that was associated with c-fos expression (C). Small bands of hsp70 expression were observed for these rats.

A. Mancuso et al. / Brain Research 887 (2000) 34 – 45

39

Fig. 3. Representative ADC lesion size changes for a normothermic and a hypothermic rat. The top graphs represent the results for the entire hemisphere
and the bottom graphs represent the results for ROI-1 (defined in Fig. 1).

Fig. 4. Representative changes in the size of the region with significant ADC reduction in c-fos-positive regions of two normothermic and two hypothermic
rats.

40

A. Mancuso et al. / Brain Research 887 (2000) 34 – 45

c-fos-positive region that exhibited significant ADC reduction. The results in Fig. 4A are for the rat that was shown
in Fig. 2A. The graph indicates that three significant peaks
in ADC lesion size occurred in the c-fos-positive region.
Four significant peaks were observed for the rat represented by Fig. 4B. Two peaks were observed for all of the
other rats of the normothermic group except one, which
had only one peak.
All seven rats of the normothermic group exhibited
hsp70 mRNA expression. Expression was weak in central
portions and strong along the periphery of the PR-ADC
region. Hsp70 was generally limited to the MCA territory;
almost no hsp70 was observed in medial frontal or
cingulate cortex where TR-ADC occurred. Strong hsp70
expression in much of the MCA territory was observed in
one rat. For this rat, the ADC lesion initially expanded
very rapidly to cover a large part of the MCA territory, but
subsequently contracted to encompass only about half of
the basal ganglia.

3.3. Hypothermic group
ADC, c-fos and hsp70 results for a hypothermic rat are
shown in Fig. 2B. The initial ADC lesion was very small.
It expanded gradually to eventually include much of the
basal ganglia and a small portion of cortex. The timecourse for the size of the ADC lesion is shown in Fig. 3C.
The five ADC maps at the bottom of Fig. 2B indicate that
TR-ADC did not occur during early ischemia. For the
normothermic group, TR-ADC was most commonly observed during early ischemia and in ROI-1. The timecourse for ADC changes in ROI-1 is shown in Fig. 3D; no
significant TR-ADC was observed in this or any other
ROI. Also, no c-fos or hsp70 expression was detected. Six
other rats of the hypothermic group exhibited no c-fos or
hsp70 and almost no TR-ADC. Three of these rats had
PR-ADC regions that involved all of the basal ganglia and
part of cortex; one had a PR-ADC region that was limited
to the basal ganglia. Three rats had PR-ADC regions that
covered less than 10% of the ischemic hemisphere, despite
exhibiting large perfusion deficits. This contrasts sharply
with the findings for the normothermic group, where TRADC, c-fos and hsp70 were observed in all seven rats.
Four rats of the hypothermic group showed a single
incident of TR-ADC at the beginning of the ischemic
period. Results for one such rat are shown in Fig. 2C. No
significant ADC lesion was observed in the first map. In
subsequent maps (not shown), a small region with reduced
ADC appeared in lateral basal ganglia and lateral cortex.
As shown at the bottom of Fig. 2C, this region expanded
abruptly to include most of the dorsal half of cortex,
including cingulate cortex. It contracted during the next 5
min and eventually was limited to the basal ganglia at the
end of the ischemic period. The complete time-course for
the size of the region with reduced ADC is shown in Fig.
5B. TR-ADC did not recur in this or any hypothermic rat.

Fig. 5. Results for a normothermic rat and a hypothermic rat with a large
expansion in ADC lesion size that reversed spontaneously during early
ischemia.

One normothermic rat also showed a single TR-ADC that
was fairly large (Fig. 5A). C-fos was induced throughout
the region that exhibited TR-ADC in Fig. 2C. Two typical
time-courses for the ADC lesion size within the c-fospositive region are shown in Figs. 4C and D. Transient
ADC reductions in the hypothermic group were longer
than they were in the normothermic group (Figs. 4A and
B) and did not recur.
Moderate hsp70 mRNA expression was observed in a
small band of lateral cortex for the rat in Fig. 2C. The faint
hsp70 expression normally observed in the persistently
ischemic region of normothermic rats was not observed in
this or any other hypothermic rat. Three rats exhibited
moderate hsp70 expression in a thin band along the dorsal
periphery of the PR-ADC region. One rat exhibited strong
hsp70 expression in a large part of cortex within the MCA
distribution, where a single transient reduction in ADC
occurred during the first 10 min of ischemia.

3.4. Statistical comparisons between the normothermic
and hypothermic groups
Mean ADC lesion sizes for the normothermic and
hypothermic groups following 15, 30, 45 and 60 min of
ischemia are summarized in Table 2. The values for the
normothermic group were significantly larger than the
corresponding values for the hypothermic group at all four
time-points.
Table 3 summarizes the association of TR-ADC with

A. Mancuso et al. / Brain Research 887 (2000) 34 – 45

41

Table 3
Association of c-fos mRNA expression and transient ADC reduction
Region a

1
2
3
4
5
a

Normothermic (n57)

Hypothermic (n511)

c-fos

TR-ADC

c-fos
1TR-ADC

c-fos

TR-ADC

c-fos
1TR-ADC

7
4
3
1
1

7
4
3
0
2

7
4
2
0
0

4
4
4
2
3

4
5
3
3
4

4
4
3
2
2

Regions used for this analysis are shown in the anatomical map in Fig. 1.

c-fos expression, based on the regions of interest shown in
Fig. 1. For the normothermic group, all seven rats exhibited both c-fos and TR-ADC in region 1. In regions 2
and 3, the incidence of c-fos expression was lower, but
nearly all c-fos-positive regions exhibited TR-ADC. Little
c-fos or TR-ADC was observed in regions 4 and 5. Of the
total 35 regions of interest for the normothermic group,
only five exhibited c-fos without TR-ADC or TR-ADC
without c-fos. Statistically, the association between c-fos
and TR-ADC was highly significant (P,0.001). For the
hypothermic group, the frequency of c-fos expression in
regions 1, 2 and 3 was less than it was for the normothermic group (on a percentage basis). Of the 11 rats, c-fos was
observed in only four rats for regions 1, 2 and 3. For the
hypothermic group, only six of the 55 regions of interest
did not show correspondence between c-fos expression and
TR-ADC; the association between these two parameters
was again highly significant (P,0.001).
Table 4 summarizes the frequency of TR-ADC in the
regions of interest shown in Fig. 1. For region 1, the mean
number of incidents of TR-ADC for the normothermic
group was 1.960.6, which is significantly higher than the
0.560.7 for the hypothermic group (P,0.01). The frequency of TR-ADC for the normothermic group was also
higher in regions 2 and 5, but the differences were not
significant (P.0.05).
The frequency of transient ADC reduction within the
c-fos-positive regions was much lower in the hypothermic
group than in the normothermic group. For the c-fospositive regions of the normothermic group, the average
number of peaks in ADC lesion size was 2.361.0, which
Table 4
Average frequency of occurrence of transient ADC reduction in the five
regions of interest shown in Fig. 1 a
Region

Normothermic
Frequency
of TR-ADC

Hypothermic
Frequency
of TR-ADC

P

1.0
2.0
3.0
4.0
5.0

1.960.6
1.361.5
0.460.5
0.060.0
0.761.3

0.560.7
0.560.7
0.460.7
0.260.4
0.460.5

,0.01
.0.05
.0.05
–
.0.05

a

All values are mean6S.D.

was significantly higher than the 1.060.0 for the hypothermic group (P,0.05). For the hypothermic rats, the
duration of the single ADC lesion size peak was much
longer than it was for the normothermic group, generally
lasting for about 10 min.
All seven rats of the normothermic group exhibited
moderate hsp70 expression in at least some part of the
ischemic hemisphere. Only four of the 11 hypothermic rats
exhibited any hsp70, which was significantly less than for
the normothermic group (P,0.05).

4. Discussion
The most significant findings of this work were that
during acute focal cerebral ischemia, mild hypothermia
diminished the incidence of transient ADC reduction and
c-fos mRNA expression. TR-ADC was observed much
more frequently in dorsal cortex for the normothermic
group than it was in the hypothermic group. Also, c-fospositive regions of rats in the normothermic group had a
higher number of significant peaks in ADC lesion size,
2.361.0, than the hypothermic group, 1.060.0. Both of
these findings support the second hypothesis of this work,
that hypothermia decreases the incidence of transient ADC
reduction. C-fos expression was observed in all seven rats
of the normothermic group. For the hypothermic group, a
markedly reduced incidence of c-fos expression was
observed (five of 11 rats), which is consistent with the
third hypothesis. We also previously reported a very high
incidence of c-fos expression during normothermic MCA
occlusion [47]. Both TR-ADC [10,43] and c-fos expression
[26] are believed to be associated with spreading depression. Thus, the results of this work suggest that mild
hypothermia reduces the incidence of spreading depression
during acute focal ischemia.
Hypothermia has been shown to reduce the incidence of
transient recurring DC potential shifts measured with
cortical electrodes during MCA occlusion [14]. Such shifts
are believed to be caused by transient cellular depolarization that occurs during spreading depression. Moderate
hypothermia reduced the average number of DC potential
shifts from 3.75 (at 378C) to 2.0 (at 308C). These results
are consistent with our observation that hypothermia

42

A. Mancuso et al. / Brain Research 887 (2000) 34 – 45

attenuated changes believed to be associated with spreading depression. However, in a recent study done with
electrode measurement of extracellular potassium in cortical penumbra, hypothermia did not affect the number of
peaks in potassium concentration during acute focal ischemia [56]. Such peaks are believed to be caused by
spreading depression. The findings of Sick et al. contrast
with those of this work and the work of Chen et al.
However, the potassium measurements of Sick et al. were
only made at a single point in penumbra. Further studies,
with electrodes at multiple penumbral cites, would be
useful for clarifying this discrepancy.
Spreading depression during acute focal ischemia is
initiated by high concentrations of interstitial excitatory
amino acids and / or potassium [27]. During MCA occlusion in rats, measurements with dialysis probes have
shown that mild hypothermia significantly reduces interstitial glutamate accumulation in cortical regions of the MCA
territory. After approximately 1 h of MCA occlusion,
interstitial glutamate levels in dorsolateral parietal cortex
were 2.5–3 times higher at 378C than they were at 338C
[3,64]. A more recent study has confirmed that mild
hypothermia (328C) markedly reduces cortical but not
striatal glutamate levels [28]. Similar effects were reported
for interstitial accumulation of glycine [28]. Thus, the
reduced incidence of TR-ADC and c-fos expression observed in this work may have been caused by diminished
levels of interstitial glutamate in cortical regions.
During MCA occlusion, mild hypothermia only slightly
reduces steady state extracellular potassium levels in
penumbral and core regions [55,56]. In addition, mild
hypothermia does not inhibit the initiation of spreading
depression by cortical application of potassium chloride
[59]. These results suggest that the neuroprotective effect
of hypothermia during focal ischemia is not mediated by
changes in potassium-induced spreading depression.
Reduced incidence of spreading depression during ischemia could be an important component of the neuroprotective effect of hypothermia, since spreading depression is believed to increase the size of the necrotic region
during focal ischemia [9,58]. Blood perfusion levels in a
region depolarized by spreading depression increase in
response to the increased energy demand necessary to
restore ionic gradients [27]. Spreading depression in the
absence of marked perfusion deficit does not produce
permanent injury. However, in the partially ischemic
region where blood flow is restricted, perfusion levels can
not increase to meet the increased energy demand. This
hastens energy failure, which allows calcium to cross the
cell membrane. High levels of intracellular calcium activate enzymes that cause necrosis [57].
A number of studies with cortical electrodes have
demonstrated that spreading depression contributes to
lesion size growth during focal ischemia. Inhibition of
spreading depression by the NMDA receptor antagonist
MK-801 (dizocilpine maleate) [22,30] reduces the eventual

infarct size. Conversely, increased incidence of spreading
depression during MCA occlusion, caused by electrical
stimulation, increases neuronal damage and infarct size [2].
The slowing of ADC lesion expansion by hypothermia
is also consistent with the first hypothesis of this study and
with observations described in a recent preliminary report
[67]. In addition to reduced incidence of spreading depression, the slowed rate may have been caused directly by
reduced rates of ATP depletion or excitotoxin release.
Consistent with both of these mechanisms, DC potential
measurements during acute ischemia have shown that
hypothermia delays the onset of terminal ischemic depolarization [5].
In addition to being caused by spreading depression,
some of the transient ADC reductions observed in this
work may have been caused by transient ischemia. Following the initial blood flow reduction with occlusion, an
ischemia-induced increase in collateral blood flow may
have occurred. For example, the results in Figs. 5A and 5B
may have been caused by this mechanism. This would
most likely occur in rats with only a limited reduction in
blood flow, during the earliest part of the ischemic period.
Four hypothermic rats and one normothermic rat exhibited
a single TR-ADC during early ischemia. Thus, hypothermia may have preserved energy metabolism sufficiently to allow improved recovery from the brief ischemia. However, not enough rats exhibited these patterns
to statistically evaluate this hypothesis. We previously
reported that blood flow reduction in our model of focal
ischemia can be somewhat variable [48]. The variability is
likely a consequence of leaving the CCA proximal to the
occluded MCA patent. This model is used in our work
because it is a better model for clinical stroke, where some
residual blood flow in the ischemic region is usually
present. MCA occlusion models with an intraluminal
suture and CCA ligation may give more reproducible
lesion sizes, but they may also be too severe to accurately
represent the pathophysiological changes that occur clinically.
Whether caused by spreading depression or transient
ischemia, an important finding of this work is that during
both normothermic and hypothermic acute ischemia, c-fos
expression is associated with TR-ADC. This observation is
consistent with our previous findings for normothermic rats
subjected to 30 or 60 min MCA occlusion [47].
Intra-ischemic hypothermia has been shown to hasten
c-fos and fos-B protein expression following temporary
global ischemia [39]. Gerbils subjected to 10-min bilateral
common carotid artery (CCA) occlusion exhibited C-FOS
protein in the CA2–CA4 regions of hippocampus within 1
h after moderate hypothermic (308C) ischemia. This was
significantly earlier than for normothermic control animals.
Similarly, hippocampal c-fos mRNA expression during
early reperfusion was hastened by mild hypothermia
(338C) following 15-min transient forebrain ischemia in
rats [33]. The authors of both studies hypothesized that the

A. Mancuso et al. / Brain Research 887 (2000) 34 – 45

faster expression may have been the result of faster
recovery of intracellular signaling with hypothermia. These
results indicate that induction of c-fos by transient cellular
depolarization is not blocked by hypothermia. In our work,
the reduced incidence of c-fos expression may have
resulted from hypothermic attenuation of the stimulus for
expression, i.e., spreading depression.
The effect of intra-ischemic hypothermia on HSP70
protein expression has been examined following transient
forebrain ischemia [40]. Normothermic gerbils subjected to
10-min bilateral CCA occlusion exhibited HSP70 protein
in hippocampus and neocortex 24 h after reperfusion. With
moderate hypothermia during ischemia (308C), almost no
HSP70 was observed. Similarly, rats subjected to 8-min
bilateral CCA and hypotension at 378C showed strong
HSP70 protein expression in hippocampus 48 h after
ischemia, but very little HSP70 expression when the intraischemic temperature was reduced to 308C [16]. These
results, and the results of this study, indicate that the
ischemic stimulus for hsp70 expression is diminished by
hypothermia. Hsp70 induction is believed to be controlled
by the formation of denatured proteins [1]. Hence, hypothermia may play an important role in reducing the
denaturation of proteins during ischemia. The results of
this work also suggest that the neuroprotective effect of
hypothermia is not related to HSP70 protein expression, at
least not during acute focal ischemia.
A limitation of this study is that transient ADC reduction was generally not observed in 100% of the c-fospositive regions. Three possible reasons for this are: (1)
the temporal resolution of the ADC maps was not sufficient to consistently capture complete waves of transient
ADC reduction, (2) partial volume effects may have
obscured some of the ADC changes, and (3) an ADC
reduction threshold of 20% was too high to detect all
occurrences of transient ADC reduction. Nevertheless, the
results demonstrate a statistically significant association of
hypothermia with reduced TR-ADC and c-fos expression.
The low PCO2 values during ischemia for the normothermic group may have resulted in slightly reduced cerebral
blood flow levels, relative to the hypothermic group.
However, because the degree of hypocapnia was very
slight (|10 mmHg), the extent of reduction in perfusion
was likely very small [65], especially since isoflurane was
used as an anesthetic [68]. The reduced perfusion rates
could potentially reduce the incidence of spreading depression in the normothermic group. However, we observed
that the incidence of spreading depression was much
higher for the normothermic group than it was for the
hypothermic group. Thus, the slight hypocapnia probably
did impact our primary conclusions.
In summary, the observations presented in this work
indicate that mild hypothermia reduced the incidence of
TR-ADC and c-fos expression. Both of these phenomena
are believed to be associated with spreading depression.
Therefore, part of the neuroprotective affect of hypo-

43

thermia during focal ischemia may be due to a reduced
incidence of spreading depression. Hypothermia also significantly reduced the ADC lesion size. Hsp70 expression
was markedly diminished by hypothermia, indicating that
the neuroprotective affect of hypothermia is not mediated
by expression of this stress protein.

Acknowledgements
This work was supported by grants from the National
Institute for Neurological Disorders and Stroke, NS22022
(PRW), NS14543 (FRS), and NS28167 (FRS). We thank
Mr Leo Benjamin and Ms Theresa A. Marsh for performing the in situ hybridization studies and Dr Michael F.
Wendland for assisting with MRI data acquisition.

References
[1] J. Ananthan, A.L. Goldberg, R. Voellmy, Abnormal proteins serve as
eukaryotic stress signals and trigger the activation of heat shock
genes, Science 232 (1986) 522–524.
[2] T. Back, M.D. Ginsberg, W.D. Dietrich, B.D. Watson, Induction of
spreading depression in the ischemic hemisphere following experimental middle cerebral artery occlusion: effect on infarct
morphology, J. Cereb. Blood Flow Metab. 16 (1996) 202–213.
[3] C.J. Baker, A.J. Fiore, V.I. Frazzini, T.F. Choudhri, G.P. Zubay, R.A.
Solomon, Intraischemic hypothermia decreases the release of glutamate in the cores of permanent focal cerebral infarcts, Neurosurgery
36 (1995) 994–1001.
[4] F.C. Barone, G.Z. Feuerstein, R.F. White, Brain cooling during
transient focal ischemia provides complete neuroprotection, Neurosci. Biobehav. Rev. 21 (1997) 31–44.
[5] R.D. Bart, S. Takaoka, R.D. Pearlstein, F. Dexter, D.S. Warner,
Interactions between hypothermia and the latency to ischemic
depolarization: implications for neuroprotection, Anesthesiology 88
(1998) 1266–1273.
[6] H. Benveniste, L.W. Hedlund, G.A. Johnson, Mechanism of detection of acute cerebral ischemia in rats by diffusion-weighted
magnetic resonance microscopy, Stroke 23 (1992) 746–754.
[7] F. Boris-Moller, T. Wieloch, Changes in the extracellular levels of
glutamate and aspartate during ischemia and hypoglycemia. Effects
of hypothermia, Exp. Brain Res. 121 (1998) 277–284.
[8] A. Buchan, W.A. Pulsinelli, Hypothermia but not the N-methyl-Daspartate antagonist, MK-801, attenuates neuronal damage in gerbils
subjected to transient global ischemia, J. Neurosci. 10 (1990) 311–
316.
[9] E. Busch, M.L. Gyngell, M. Eis, M. Hoehn-Berlage, K.A. Hossmann, Potassium-induced cortical spreading depressions during
focal cerebral ischemia in rats: contribution to lesion growth
assessed by diffusion-weighted NMR and biochemical imaging, J.
Cereb. Blood Flow Metab. 16 (1996) 1090–1099.
[10] E. Busch, M. Hoehn-Berlage, M. Eis, M.L. Gyngell, K.A. Hossmann, Simultaneous recording of EEG. DC potential and diffusionweighted NMR imaging during potassium induced cortical spreading depression in rats, NMR Biomed. 8 (1995) 59–64.
[11] R. Busto, W.D. Dietrich, M.Y. Globus, I. Valdes, P. Scheinberg, M.D.
Ginsberg, Small differences in intraischemic brain temperature
critically determine the extent of ischemic neuronal injury, J. Cereb.
Blood Flow Metab. 7 (1987) 729–738.
[12] R. Busto, M.Y. Globus, W.D. Dietrich, E. Martinez, I. Valdes, M.D.
Ginsberg, Effect of mild hypothermia on ischemia-induced release

44

[13]

[14]

[15]

[16]

[17]

[18]

[19]
[20]

[21]

[22]

[23]

[24]

[25]

[26]
[27]
[28]

[29]

[30]

[31]

A. Mancuso et al. / Brain Research 887 (2000) 34 – 45
of neurotransmitters and free fatty acids in rat brain, Stroke 20
(1989) 904–910.
A.L. Busza, K.L. Allen, M.D. King, B.N. van, S.R. Williams, D.G.
Gadian, Diffusion-weighted imaging studies of cerebral ischemia in
gerbils. Potential relevance to energy failure, Stroke 23 (1992)
1602–1612.
Q. Chen, M. Chopp, G. Bodzin, H. Chen, Temperature modulation
of cerebral depolarization during focal cerebral ischemia in rats:
correlation with ischemic injury, J. Cereb. Blood Flow Metab. 13
(1993) 389–394.
M. Chopp, H. Chen, M.O. Dereski, J.H. Garcia, Mild hypothermic
intervention after graded ischemic stress in rats, Stroke 22 (1991)
37–43.
M. Chopp, Y. Li, M.O. Dereski, S.R. Levine, Y. Yoshida, J.H.
Garcia, Hypothermia reduces 72-kDa heat-shock protein induction
in rat brain after transient forebrain ischemia, Stroke 23 (1992)
104–107.
G.L. Clifton, W.C. Taft, R.E. Blair, S.C. Choi, R.J. DeLorenzo,
Conditions for pharmacologic evaluation in the gerbil model of
forebrain ischemia, Stroke 20 (1989) 1545–1552.
K.E. Cullen, K.D. Sarge, Characterization of hypothermia-induced
cellular stress response in mouse tissues, J. Biol. Chem. 272 (1997)
1742–1746.
T. Curran, J.I. Morgan, Fos: an immediate-early transcription factor
in neurons, J. Neurobiol. 26 (1995) 403–412.
C. Decanniere, S. Eleff, D. Davis, P.C. van Zijl, Correlation of rapid
changes in the average water diffusion constant and the concentrations of lactate and ATP breakdown products during global
ischemia in cat brain, Magn. Reson. Med. 34 (1995) 343–352.
G. Devuyst, J. Bogousslavsky, Clinical trial update: neuroprotection
against acute ischaemic stroke, Curr. Opin. Neurol. 12 (1999)
73–79.
R. Gill, P. Andine, L. Hillered, L. Persson, H. Hagberg, The effect
of MK-801 on cortical spreading depression in the penumbral zone
following focal ischaemia in the rat, J. Cereb. Blood Flow Metab. 12
(1992) 371–379.
M.D. Ginsberg, L.L. Sternau, M.Y. Globus, W.D. Dietrich, R. Busto,
Therapeutic modulation of brain temperature: relevance to ischemic
brain injury, Cerebrovasc. Brain Met. Rev. 4 (1992) 189–225.
Y. Goto, N.F. Kassell, K. Hiramatsu, S.W. Soleau, K.S. Lee, Effects
of intraischemic hypothermia on cerebral damage in a model of
reversible focal ischemia, Neurosurgery 32 (1993) 980–984.
N.G. Harris, E. Zilkha, J. Houseman, M.R. Symms, T.P. Obrenovitch, S.R. Williams, The relationship between the apparent
diffusion coefficient measured by magnetic resonance imaging,
anoxic depolarization, and glutamate efflux during experimental
cerebral ischemia, J. Cereb. Blood Flow Metab. 20 (2000) 28–36.
D.G. Herrera, H.A. Robertson, Activation of c-fos in the brain, Prog.
Neurobiol. 50 (1996) 83–107.
K.A. Hossmann, Periinfarct depolarizations, Cerebrovasc. Brain
Metab. Rev. 8 (1996) 195–208.
F.P. Huang, L.F. Zhou, G.Y. Yang, Effects of mild hypothermia on
the release of regional glutamate and glycine during extended
transient focal cerebral ischemia in rats, Neurochem. Res. 23 (1998)
991–996.
C. Hunt, R.I. Morimoto, Conserved features of eukaryotic hsp70
genes revealed by comparison with the nucleotide sequence of
human hsp70, Proc. Natl. Acad. Sci. USA 82 (1985) 6455–6459.
T. Iijima, G. Mies, K.A. Hossmann, Repeated negative DC deflections in rat cortex following middle cerebral artery occlusion are
abolished by MK-801: effect on volume of ischemic injury, J.
Cereb. Blood Flow Metab. 12 (1992) 727–733.
Q. Jiang, M. Chopp, Z.G. Zhang, J.A. Helpern, R.J. Ordidge, J.
Ewing, P. Jiang, B.A. Marchese, The effect of hypothermia on
transient focal ischemia in rat brain evaluated by diffusion- and
perfusion-weighted NMR imaging, J. Cereb. Blood Flow Metab. 14
(1994) 732–741.

[32] A. Kader, M.H. Brisman, N. Maraire, J.T. Huh, R.A. Solomon, The
effect of mild hypothermia on permanent focal ischemia in the rat,
Neurosurgery 31 (1992) 1056–1060.
[33] F. Kamme, K. Campbell, T. Wieloch, Biphasic expression of the fos
and jun families of transcription factors following transient forebrain
ischaemia in the rat. Effect of hypothermia, Eur. J. Neurosci. 7
(1995) 2007–2016.
[34] H. Karibe, J. Chen, G.J. Zarow, S.H. Graham, P.R. Weinstein,
Delayed induction of mild hypothermia to reduce infarct volume
after temporary middle cerebral artery occlusion in rats, J. Neurosurg. 80 (1994) 112–119.
[35] H. Karibe, G.J. Zarow, S.H. Graham, P.R. Weinstein, Mild intraischemic hypothermia reduces postischemic hyperperfusion, delayed
postischemic hypoperfusion, blood–brain barrier disruption, brain
edema, and neuronal damage volume after temporary focal cerebral
ischemia in rats, J. Cereb. Blood Flow Metab. 14 (1994) 620–627.
[36] K. Kataoka, H. Yanase, Mild hypothermia–a revived countermeasure against ischemic neuronal damages, Neurosci. Res. 32 (1998)
103–117.
[37] H. Kinouchi, F.R. Sharp, M.P. Hill, J. Koistinaho, S.M. Sagar, P.H.
Chan, Induction of 70-kDa heat shock protein and hsp70 mRNA
following transient focal cerebral ischemia in the rat, J. Cereb.
Blood Flow Metab. 13 (1993) 105–115.
[38] K. Kogure, H. Kato, Altered gene expression in cerebral ischemia,
Stroke 24 (1993) 2121–2127.
[39] K. Kumar, X. Wu, A.T. Evans, Expression of c-fos and fos-B
proteins following transient forebrain ischemia: effect of hypothermia, Mol. Brain Res. 42 (1996) 337–343.
[40] K. Kumar, X. Wu, A.T. Evans, F. Marcoux, The effect of hypothermia on induction of heat shock protein (HSP)-72 in ischemic
brain, Metab. Brain Dis. 10 (1995) 283–291.
[41] A.R. Laptook, R.J. Corbett, R. Sterett, D.K. Burns, D. Garcia, G.
Tollefsbol, Modest hypothermia provides partial neuroprotection
when used for immediate resuscitation after brain ischemia, Pediatric Res. 42 (1997) 17–23.
[42] A.R. Laptook, R.J. Corbett, R. Sterett, D. Garcia, G. Tollefsbol,
Quantitative relationship between brain temperature and energy
utilization rate measured in vivo using 31P and 1H magnetic
resonance spectroscopy, Pediatric Res. 38 (1995) 919–925.
[43] L.L. Latour, Y. Hasegawa, J.E. Formato, M. Fisher, C.H. Sotak,
Spreading waves of decreased diffusion coefficient after cortical
stimulation in the rat brain, Magn. Reson. Med. 32 (1994) 189–198.
[44] D. Le Bihan, E. Breton, D. Lallemand, P. Grenier, E. Cabanis, M.
Laval-Jeantet, MR imaging of intravoxel incoherent motions: application to diffusion and perfusion in neurologic disorders, Radiology
161 (1986) 401–407.
[45] J. Maher, V. Hachinski, Hypothermia as a potential treatment for
cerebral ischemia, Cerebrovasc. Brain Met. Rev. 5 (1993) 277–300.
[46] C.M. Maier, K. Ahern, M.L. Cheng, J.E. Lee, M.A. Yenari, G.K.
Steinberg, Optimal depth and duration of mild hypothermia in a
focal model of transient cerebral ischemia: effects on neurologic
outcome, infarct size, apoptosis, and inflammation, Stroke 29 (1998)
2171–2180.
[47] A. Mancuso, N. Derugin, Y. Ono, K. Hara, F.R. Sharp, P.R.
Weinstein, Transient MRI-detected water apparent diffusion coefficient reduction correlates with c-fos mRNA but not hsp70 mRNA
induction during focal cerebral ischemia in rats, Brain Res. 839
(1999) 7–22.
[48] A. Mancuso, H. Karibe, W.D. Rooney, G.J. Zarow, S.H. Graham,
M.W. Weiner, P.R. Weinstein, Correlation of early reduction in the
apparent diffusion coefficient of water with blood flow reduction
during middle cerebral artery occlusion in rats, Magn. Res. Med. 34
(1995) 368–377.
[49] S.M. Massa, R.A. Swanson, F.R. Sharp, The stress gene response in
brain, Cerebrovasc. Brain Metab. Rev. 8 (1996) 95–158.
[50] D.J. Moyer, F.A. Welsh, E.L. Zager, Spontaneous cerebral hypothermia diminishes focal infarction in rat brain, Stroke 23 (1992)
1812–1816.

A. Mancuso et al. / Brain Research 887 (2000) 34 – 45
[51] K. Nakashima, M.M. Todd, Effects of hypothermia on the rate of
excitatory amino acid release after ischemic depolarization, Stroke
27 (1996) 913–918.
[52] K. Nakashima, M.M. Todd, D.S. Warner, The relation between
cerebral metabolic rate and ischemic depolarization. A comparison
of the e