Growth Hormone Improves Cognitive Function

After Experimental Stroke
Lin Kooi Ong, PhD; Wei Zhen Chow, MMedSc; Clifford TeBay, BBiomedSc;
Murielle Kluge, MSc; Giovanni Pietrogrande, MSc; Katarzyna Zalewska, MSc;
Patricia Crock, MBBS, FRACP, MD; N. David Åberg, MD, PhD; Andrew Bivard, PhD;
Sarah J. Johnson, PhD; Frederick R. Walker, PhD; Michael Nilsson, MD, PhD;
Jörgen Isgaard, MD, PhD

Background and Purpose—Cognitive impairment is a common outcome for stroke survivors. Growth hormone (GH) could
represent a potential therapeutic option as this peptide hormone has been shown to improve cognition in various clinical
conditions. In this study, we evaluated the effects of peripheral administration of GH at 48 hours poststroke for 28 days
on cognitive function and the underlying mechanisms.
Methods—Experimental stroke was induced by photothrombotic occlusion in young adult mice. We assessed the associative memory
cognitive domain using mouse touchscreen platform for paired-associate learning task. We also evaluated neural tissue loss,
Downloaded from http://stroke.ahajournals.org/ by guest on June 6, 2018

neurotrophic factors, and markers of neuroplasticity and cerebrovascular remodeling using biochemical and histology analyses.
Results—Our results show that GH-treated stroked mice made a significant improvement on the paired-associate learning
task relative to non–GH-treated mice at the end of the study. Furthermore, we observed reduction of neural tissue loss in
GH-treated stroked mice. We identified that GH treatment resulted in significantly higher levels of neurotrophic factors
(IGF-1 [insulin-like growth factor-1] and VEGF [vascular endothelial growth factor]) in both the circulatory and peri-
infarct regions. GH treatment in stroked mice not only promoted protein levels and density of presynaptic marker (SYN-1
[synapsin-1]) and marker of myelination (MBP [myelin basic protein]) but also increased the density and area coverage
of 2 major vasculature markers (CD31 and collagen-IV), within the peri-infarct region.
Conclusions—These findings provide compelling preclinical evidence for the usage of GH as a potential therapeutic tool in
the recovery phase of patients after stroke.
Visual Overview—An online visual overview is available for this article.   (Stroke. 2018;49:1257-1266. DOI: 10.1161/
STROKEAHA.117.020557.)
Key Words: cognition ◼ growth hormone ◼ insulin-like growth factor-1 ◼ neuronal plasticity ◼ stroke
◼ vascular endothelial growth factor A ◼ vascular remodeling

C ognitive impairment is a well-documented and problematic

outcome for up to 80% stroke survivors.1,2 The cognitive
domains that are commonly effected by stroke include memory,
learning, and executive functions, which can lead to negative
impacts on quality of life and the functional independence of
individual stroke survivors. Currently, there are no recognized
therapeutic options for cognitive impairment poststroke. If con-
sideration is given to available options more broadly, a signifi-
cant body of work suggests that growth hormone (GH) could
represent a potential therapy. GH is a peptide hormone released
from the pituitary gland. In addition to its classical actions on
longitudinal body growth and intermediate metabolism, GH has
been shown to improve cognition in both rodents and humans.3–5
These procognitive effects are linked to many significant
changes within the central nervous system, including enhanced
neurogenesis and vasculogenesis.5–7 A distinct advantage of GH
is its approval profile and safety. Further, a recent retrospective
combined registry analysis of published safety data from 1988 to
2016 on GH identified that there was no evidence to indicate that
GH treatment increased mortality, stroke, or new malignancy.8

Received December 21, 2017; final revision received March 7, 2018; accepted March 14, 2018.
From the Priority Research Centre for Stroke and Brain Injury (L.K.O., F.R.W., M.N., J.I.), School of Biomedical Sciences and Pharmacy (W.Z.C.,
C.T., M.K., G.P., K.Z.), Department of Paediatric Endocrinology and Diabetes, Priority Research Centre Grow Up Well, John Hunter Children’s Hospital
(P.C.), School of Electrical Engineering and Computing (S.J.J.) and Department of Neurology, John Hunter Hospital (A.B.), University of Newcastle,
Australia; Hunter Medical Research Institute, Australia (L.K.O., A.B., F.R.W., M.N., W.Z.C., C.T., M.K., G.P., K.Z., S.J.J.); National Health and Medical
Research Council Centre of Research Excellence Stroke Rehabilitation and Brain Recovery, Australia (F.R.W., M.N., L.K.O.); and Centre for Brain Repair
and Rehabilitation, Institute of Neuroscience and Physiology and Department of Internal Medicine (J.I.), Sahlgrenska University Hospital, University of
Gothenburg, Sweden (N.D.A.).
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.
117.020557/-/DC1.
Correspondence to Lin Kooi Ong, PhD, Hunter Medical Research Institute, Kookaburra Circuit, New Lambton Heights, NSW 2305, Australia, E-mail
lin.ong@uon.edu.au, or Jörgen Isgaard, MD, PhD, Department of Internal Medicine, Sahlgrenska University Hospital, University of Gothenburg, Sweden,
E-mail jorgen.isgaard@medic.gu.se
© 2018 American Heart Association, Inc.
Stroke is available at http://stroke.ahajournals.org DOI: 10.1161/STROKEAHA.117.020557

1257
 1258  Stroke  May 2018
Several studies have shown the beneficial effects of GH in
promoting cognition in several experimental contexts includ-
ing brain injury.9,10 Although modest in scope, GH has been
examined for its ability to stimulate cognition poststroke.
Pathipati et al,11, using an endothelin model of ischemia, exam-
ined the influence of GH delivered directly into the lateral ven-
tricle of the brain starting 4 days poststroke for 42 days. Here,
the authors identified the ability of GH to enhance aspects of
spatial memory, using the Morris water maze. Specifically,
they observed an improved performance in the probe trial with
the GH-treated stroked animals spending more time in the plat-
form zone than the non–GH-treated mice. Although promising,
the results of Pathipati et al,11 are limited in 2 significant ways.
First, it is not clear whether peripheral GH delivery, as would
be required clinically, could have the same procognitive ben-
efits as when delivered intracerebroventricularly. Second, there
is now broad consensus that classical tests of spatial working
memory including the Morris water maze are subject to sig-
nificant interpretational challenges and where possible newer
Downloaded from http://stroke.ahajournals.org/ by guest on June 6, 2018
assessments using more translationally relevant platforms,
such as touchscreen testing are preferred.12,13
In this study, we evaluated the effects of peripheral admin-
istration of GH on cognitive function poststroke. Specifically,
our primary hypothesis was that stroked mice treated with
commercially available recombinant human GH (r-hGH) at 1.4
mg/kg body weight per day subcutaneously via mini-osmotic
pumps for 28 days starting 48 hours poststroke would result
in improved performance in cognition tasks relative to stroked
non–GH-treated mice. Our secondary hypotheses were that
r-hGH–treated stroked mice would show increased expression
of neurotrophic factors, enhancement of neural plasticity and
cerebrovascular remodeling, and reduced cerebral tissue loss.
Methods
The data that support the findings of this study are available from the
corresponding author on reasonable request. A detailed methods sec-
tion is available in the online-only Data Supplement.
Ethical Statements
Animal research was undertaken in accordance to the ARRIVE
guidelines (Animal Research: Reporting of In Vivo Experiments).
Experiments were approved by the University of Newcastle Animal
Care and Ethics Committee (A-2014-432) and conducted in accor-
dance with the New South Wales Animals Research Act and the
Australian Code of Practice for the use of animals for scientific
purposes.
Experimental Design
All the experimental groups are randomized, and all outcome analy-
ses were performed by independent study team members blinded to
the treatment condition. A total of 56 mice (C57BL/6, male, 8 weeks
old) were used in this study. Details on animal numbers for each
experiment and inclusion/exclusion criteria are included in Figure I in
the online-only Data Supplement. The first cohort of mice was allo-
cated for pilot study to investigate the bioactive effects of r-hGH in
sham mice (Figure IIA in the online-only Data Supplement). The sec-
ond cohort of mice was used to investigate the effectiveness of r-hGH
in facilitating cognitive recovery and the underlying mechanisms
after experimental stroke (Figure 1A). Day 0, mice were subjected
to photothrombotic occlusion or sham surgery. Day 2, mice received
r-hGH or saline at 1.4 mg/kg body weight per day subcutaneously via
mini-osmotic pumps for 28 days. To evaluate the effect of r-hGH on
cognitive function, stroked mice received r-hGH or saline treatment
were subjected to mouse touchscreen platform for paired-associate
learning (PAL) task for 20 days (days 9–29 poststroke). Brains and
blood samples were collected at 30 days poststroke.
Animal Surgery
Photothrombotic occlusion was performed as previously
described.14–18 Mini-osmotic pump (model 2004, Alzet) placement
was performed as previously described.19
Assessment of Cognitive Deficits
The stroked mice were subjected to a mouse touchscreen platform
for PAL task to assess the associative memory cognitive domain.20,21
Biochemical Analysis
Protein homogenates from the peri-infarct samples were obtained as
described.17 The serum and peri-infarct supernatants were measured
for IGF-1 (insulin-like growth factor-1) and VEGF (vascular endo-
thelial growth factor) levels using commercially available mouse/rat
IGF-1 Quantikine ELISA (MG100; R&D systems, MN) and mouse
VEGF Quantikine ELISA (MMV00), respectively, according to the
manufacturer’s instruction. Western blotting was performed as previ-
ously described.22
Histology Analysis
Cresyl violet staining was performed as described.23 For immuno-
peroxidase labeling, free-floating sections were immunostained as
described.17 Images were acquired at ×20 using Aperio AT2 (Leica,
Germany). ImageJ (1.50, National Institutes of Health) and Matlab
(R2015a, MathWorks) were used to estimate tissue loss and to ana-
lyze intensity and area coverage of immunolabeling.15–18
Sample Size Calculation
Sample size was estimated using the following formula:
( )
2
2SD2 z1−– + z1−†
SS = 2
d2
Using preliminary data from PAL task (% correct rate; sham versus
stroke), we obtained, SD=8.5 and an effect size of d=12. Allowing a
type 1 error of 5%, α=0.05 with the power of 80%, and β=0.2, we
calculated a sample size of 8 animals per group. More than 8 animals
per group would ensure that a treatment effect will be detected, if
present, with a >80% chance.
Statistical Analysis
All data were expressed as mean±SD and were analyzed using Prism
for Windows Version 7.02, GraphPad Software. The primary outcome
measurement was differences between Stroke+Saline and Stroke+r-
hGH. Data from ELISA, PAL task (final session analysis) Western
blotting, and immunoperoxidase labeling were analyzed using
2-tailed t test. Delta weight change, PAL task (20 sessions temporal
analysis), tissue loss, and cumulative threshold were analyzed using
2-way ANOVA, followed by Sidak multiple comparisons. Pearson
correlation was used to determine the association between serum
neurotrophic factors and cognitive performance. A P value <0.05 was
considered statistically significant.
Results
GH Treatment Increases Weight Gain
and Serum Neurotrophic Factors
First, we confirmed in sham mice that the commer-
cially available r-hGH has effects on mice physiology,
 Ong et al   Growth Hormone Promotes Cognition Poststroke   1259
Downloaded from http://stroke.ahajournals.org/ by guest on June 6, 2018

Figure 1. The effects of recombinant human growth factor (r-hGH) treatment poststroke on physiological parameters. A, Experimental
design. All mice received a focal photothrombotic stroke. Fourty-eight hours poststroke, mice were randomly treated with either saline or
r-hGH via mini-osmotic pumps subcutaneously for 28 days, which were randomly distributed for either measurement of changes in body
weight or assessment of cognitive function. B, r-hGH treatment poststroke significantly increased the body weight poststroke. Mice were
euthanized 30 days poststroke. C, r-hGH treatment poststroke significantly increased serum IGF-1 (insulin-like growth factor-1) and VEGF
(vascular endothelial growth factor) levels. Mean±SD. *P<0.05, **P<0.01, and ***P<0.001. PAL indicates paired-associate learning.

when delivered subcutaneously via mini-osmotic pumps
(sham+saline, n=6; sham+r-hGH, n=8; Figure II in the
online-only Data Supplement). We measured the body
weight weekly and presented the data as changes (delta, Δ)
of body weight from baseline. The baseline body weight
was 27.52±1.00 g (minimum=25.10 g; maximum=28.60 g).
There was a significant interaction between the effect of
time and r-hGH treatment on body weight (F(5, 60)=4.088;
P=0.0029). We found a significant increase in serum IGF-1
(20.7%; P=0.0388) and VEGF (49.5%; P=0.0214) levels in
r-hGH–treated mice compared with that in saline-treated
mice. We also found a significant 17.8% decrease in blood
glucose (P=0.0175) levels but no changes in the liver to
body weight ratio.
Next, we investigated the effects of r-hGH treatment after
permanent focal cerebral ischemia (Figure 1). The baseline
body weight was 26.95±1.99 g (minimum=23.40 g; maxi-
mum=31.10 g). There was a significant interaction between
the effect of time and r-hGH treatment on body weight
(F(5, 75)=7.435; P<0.0001). Post hoc analysis indicated a sig-
nificant increase in body weight between 3 and 4 weeks after
r-hGH treatment poststroke. We found a significant increase
in serum IGF-1 (54.2%; P<0.0001) and VEGF (21.3%;
P=0.0072) levels in stroked mice treated with r-hGH. There
 1260  Stroke  May 2018

was no significant change in blood glucose levels and liver to
body weight ratio.
GH Treatment Improves Poststroke Cognitive
Impairment and Reduces Tissue Loss
Mice were treated with saline (n=11) or r-hGH (n=9) 48 hours
poststroke, followed by 7 days recovery. Then, mice were
Downloaded from http://stroke.ahajournals.org/ by guest on June 6, 2018
assessed using mouse touchscreen platform for PAL task for
20 consecutive days. We found a significant 22.4% increase in
percentage correct rate (P=0.0092) in r-hGH–treated stroked
mice compared with saline at final PAL session (Figure 2A).
A range of metrics from the PAL task were also collected for
temporal analysis. The data were presented as 4 blocks of
mean of 5 consecutive sessions (Figure III in the online-only

Figure 2. Illustration of the paired-associate learning (PAL) task and tissue loss. A, Top, Correct stimulus and location (flower left, plane
center, and spider right). Bottom, Stroked mouse cognition assessed using the touchscreen chamber apparatus (Campden Instruments,
Ltd). Recombinant human growth factor (r-hGH) treatment poststroke significantly increased percentage correct rate at final PAL ses-
sion. B, Representative images of Cresyl violet staining at different Bregma levels. Bar=1 mm. Tissue loss was calculated as contralateral
(CL) hemisphere area−ipsilateral (IL) hemisphere area. r-hGH treatment poststroke significantly decreased tissue loss (mm2). Mean±SD.
*P<0.05 and **P<0.01.
 Ong et al   Growth Hormone Promotes Cognition Poststroke   1261

Data Supplement). There was a significant effect of r-hGH
treatment (F(1, 18)=5.606; P=0.0293) and time (F(3, 54)=17.83;
P<0.0001) on percentage correct rate. Post hoc analysis indi-
cated a significant 18% increase in percentage correct at fourth
block after r-hGH treatment poststroke (P=0.0292). There was
no significant effect in number of trials completed within 60
minutes, time to complete 36 trials, repeated tasks per trials,
correct trials completed, and correct touch latency between
stroked mice treated with r-hGH compared with saline.
Rostral caudal analysis of tissue loss revealed a signifi-
cant effect of r-hGH treatment (F(1, 50)=5.601; P=0.0219) and
at bregma levels (F(4, 50)=5.601; P=0.0219). Post hoc analysis
Downloaded from http://stroke.ahajournals.org/ by guest on June 6, 2018
indicated a significant 56.8% decrease in tissue loss at Bregma
0.0 after r-hGH treatment poststroke (P=0.0094; Figure 2B).
GH Treatment Promotes IGF-1 and
VEGF Levels Within the Peri-Infarct
The protein homogenates from the peri-infarct region of
stroke+saline (n=8) and stroke+r-hGH (n=9) were analyzed
along with sham+saline (n=6) using Western blotting. IGF-1
and VEGF (and other protein markers discussed in below sec-
tions) were normalized to β-actin and were expressed as a
fold change of mean±SD for each group relative to the mean
of the sham+saline group. We found a significant increase

Figure 3. Protein homogenates for the peri-infarct were analyzed biochemically for neurotrophic factors. A, IGF-1 (insulin-like growth fac-
tor-1) and (B) VEGF (vascular endothelial growth factor) levels were expressed as a fold change of mean±SD for each group relative to the
mean of the sham+saline group (dotted line). Representative blots for IGF-1 and VEGF. Loading controls were performed by analysis of
β-actin. We found a significant increase in IGF-1 and VEGF levels within the peri-infarct in stroked mice treated with recombinant human
growth factor (r-hGH) compared with saline. ELISA further confirmed that r-hGH treatment poststroke significantly increased (C) IGF-1 and
(D) VEGF levels within the peri-infarct region. Mean±SD. *P<0.05, **P<0.01, and ***P<0.001.
 1262  Stroke  May 2018

in IGF-1 (43.6%; P=0.0054; Figure 3A) and VEGF (23.8%;
P<0.001; Figure 3B) levels within the peri-infarct in stroked
mice treated with r-hGH compared with saline. We further
confirmed using ELISA that r-hGH treatment poststroke sig-
nificantly increased IGF-1 (95.2%; P=0.0070; Figure 3C)
and VEGF (77.8%; P=0.0115; Figure 3D) levels within the
peri-infarct.
GH Treatment Promotes the Expression
of Presynaptic and Myelin Markers
Within the Peri-Infarct
We analyzed the protein homogenates for markers of mature
neurons (NeuN [neuronal nuclei]), PSD-95 (postsynaptic
density protein-95), presynaptic (SYN-1 [synapsin-1]), and
structural component of myelin (MBP [myelin basic protein])
using Western blotting. There was no significant change in
NeuN (P=0.5424) and PSD-95 (P=0.0836) levels (Figure 4A).
r-hGH treatment poststroke significantly increased SYN-1
(10.1%; P=0.0354) and MBP (45.4%; P=0.0414) levels
Downloaded from http://stroke.ahajournals.org/ by guest on June 6, 2018
peri-infarct was assessed using threshold analysis. In the peri-
infarct region, we found a significant increase in material
thresholded for SYN-1 (757.1%; P=0.0032, at pixel intensity
150; Figure 5A) and MBP (33.6%; P=0.0386, at pixel intensity
110; Figure 5B) in stroked mice treated with r-hGH compared
with saline. Sections were also counterstained with Cresyl
violet for qualitative observation. For cumulative threshold
analysis, see Figure IV in the online-only Data Supplement.
GH Treatment Promotes the Expression of
Vascular Markers CD31 and Collagen-IV
We also analyzed the protein homogenates for markers of
vasculature (CD31 and collagen-IV) using Western blotting.
r-hGH treatment poststroke significantly increased expres-
sion levels of CD31 (56.8%; P=0.0033) but not collagen-IV
(P=0.1240; Figure 6A).
To assess the distribution and density of vasculature, we
performed vessel digital reconstruction and threshold analysis
on CD31 (Figure 6B) and collagen-IV (Figure 6C) immunos-

(Figure 4B). tained images of the peri-infarct. r-hGH treatment poststroke

We further confirmed the protein data using immunohis- significantly increased percentage area covered by CD31+
tochemistry staining. Optical density of SYN-1 and MBP (39.4%; P=0.0345) and collagen-IV+ (42.9%; P=0.0048) vas-
(and other staining discussed in below sections) within the cular cells. We also found a significant increase in thresholded

Figure 4. Detection of neuronal markers within the peri-infarct. A, There was no significant change in NeuN (neuronal nuclei; marker
of mature neurons, quantified as optical density of 2 bands for each sample) and PSD-95 (postsynaptic density protein-95) levels. B,
Recombinant human growth factor (r-hGH) treatment poststroke significantly increased synapsin-1 (phosphoprotein specific to the cyto-
plasmic surface of synaptic vesicles of the presynaptic) and MBP (myelin basic protein, a structural component of myelin, quantified as
optical density of 4 bands for each sample). Mean±SD. *P<0.05.
 Ong et al   Growth Hormone Promotes Cognition Poststroke   1263
Downloaded from http://stroke.ahajournals.org/ by guest on June 6, 2018

Figure 5. Immunohistochemistry staining of presynaptic and myelin markers within the peri-infarct. A, Synapsin-1 and (B) MBP (myelin
basic protein) immunostaining (left) and the material thresholded at specific pixel intensity (PI; middle). Sections were also counter-
stained with Cresyl violet (right) for qualitative observation. Recombinant human growth factor (r-hGH) treatment poststroke significantly
increased thresholded staining of synapsin-1 (at PI 150) and MBP (at PI 110) levels. Mean±SD. *P<0.05 and **P<0.01.

CD31 (81.4%; P=0.0345, at pixel intensity 130) and collagen-
IV (50%; P=0.0048, at pixel intensity 190) immunoreactive
levels in stroked mice treated with r-hGH compared with
saline. For cumulative threshold analysis, see Figure IV in the
online-only Data Supplement.
Serum IGF-1 Positively Correlates With
Cognitive Performance Poststroke
Pearson correlation analysis shows a significant positive corre-
lation between serum IGF-1 levels and cognitive performance
poststroke, as indexed by percentage correct rate at final PAL
session (r=0.5914; P(Y=0.04493X+44.24)=0.0076). There was no sig-
nificant correlation between serum VEGF levels and cognitive
performance poststroke (r=0.398; P(Y=0.2271*X+46.01)=0.0915).
Discussion
In this study, we have shown that r-hGH delivered subcuta-
neously over a 28-day window starting 48 hours poststroke
significantly improved performance in the PAL task, when
compared with non–GH-treated stroked mice. We also
provided support for our secondary hypotheses, including
the ability of r-hGH to reduce loss of neural tissue, enhance
expression of neurotrophic factors (IGF-1 and VEGF), and
several markers of neural plasticity and vascular remodeling.
We placed a particular focus on translational therapeutic con-
siderations, most notably the route of GH administration and
mode of cognitive testing. Collectively, these positive results
from our study reinforce the concept of using GH as a poten-
tial treatment for promoting cognition poststroke.
Clinically, cognitive impairment is a significant concern for
stroke survivors, and much effort has been directed toward
the development of therapeutic strategies for preventing
cognitive decline after stroke. The evidence supporting the
potential use of GH as a safe and effective agent to promote
cognitive function has continued to build over recent years.
Already, promising small clinical studies have identified the
therapeutic use of GH in patients who have experienced trau-
matic brain injury or stroke.24–28 In the current study, we were
motivated to extend the existing preclinical literature that
has considered the potential procognitive effects of GH after
stroke.11 We used the photothrombotic stroke model, useful
for its ability to produce highly consistent and repeatable isch-
emic injury.29 Importantly, the photothrombotic stroke model
has been reported to consistently elicit learning and memory
deficits.30,31 To assess these deficits, we turned to touchscreen-
based assessments as these are recognized to avoid many of
 1264  Stroke  May 2018
Downloaded from http://stroke.ahajournals.org/ by guest on June 6, 2018

Figure 6. Detection of vascular markers within the peri-infarct region. A, Representative blots for CD31 (endothelial cell adhesion mol-
ecule) and collagen-IV (structural component of basement membranes). Recombinant human growth factor (r-hGH) treatment poststroke
significantly increased expression levels of CD31 but not collagen-IV. B, CD31 and (C) collagen-IV immunostaining (left), digital recon-
struction of positive labeled cells (middle), and the material thresholded at specific pixel intensity (PI; right). r-hGH treatment poststroke
significantly increased area covered by CD31 and collagen-IV–positive staining. There were also significant increases in thresholded stain-
ing of CD31 (at PI 130) and collagen-IV (at PI 190) levels in stroked mice treated with r-hGH compared with saline. Mean±SD. *P<0.05,
**P<0.01, and ***P<0.001.

the limitations associated with classical assessments.12 For
instance, Morris water maze requires animal to swim to a hid-
den platform that is located beneath the surface of the water,
which could be perceived as stressful situation. Chronic stress
can have a significant impact on brain recovery and functional
outcomes poststroke.15,17,32 Clinically, stroke is associated with
poorer performance in multiple cognitive domains, including
associative memory.33 Specifically, we used the PAL task that
has been extensively studied in clinical assessment of cogni-
tion,34 and the task available for rodents is analogous to clini-
cal research.35 Beyond this clinical translation aspect, PAL
task is recognized to assess associative memory domain. With
respect to our results, we compared the number of correct tri-
als performed by GH- and non–GH-treated mice poststroke at
the final session. This comparison indicated that GH-treated
stroked mice made a significant 22.4% more correct choices
on the PAL task relative to non–GH-treated mice, thus con-
firming our primary hypothesis.
Extending from our observations of the procognitive effects
of GH, we also examined the impact on the neurorestorative
effects, such as neural plasticity and cerebrovascular remod-
eling. As expected, stroke reduced NeuN, a marker of mature
neurons, SYN-1, a presynaptic marker, PSD-95, a postsynaptic
marker, and MBP, a marker of myelination. In support of our
secondary hypotheses, GH treatment in stroked mice resulted
in significantly higher protein levels of SYN-1 and MBP within
the peri-infarct in comparison to saline treatment. These results
were confirmed using immunohistochemistry, which revealed
GH treatment significantly increased density of SYN-1–
positive terminals by ≈7.5-fold and a moderate 33.6% increase
in MBP-positive structures within the motor and somatosensory
cortices. The presence of synaptogenesis and remyelination
poststroke after GH treatment is a critical finding, as neuroplas-
ticity is believed to play a critical role in the stroke recovery.
The effect of GH treatment poststroke on cerebrovascular
remodeling has not previously been investigated. In the current
 Ong et al   Growth Hormone Promotes Cognition Poststroke   1265
study, we considered 2 major vasculature markers. First, CD31,
a cell adhesion molecule secreted by endothelial cells and modu-
lates cell adhesion, endothelial cell migration, and angiogenesis.
Second, collagen-IV, a basal lamina protein secreted by endothe-
lial cells. Stroked mice in comparison to shams showed no change
in both CD31 and collagen-IV vasculature protein markers within
the peri-infarct 30 days poststroke. Studies have shown that stroke
induces significant reduction in cerebrovascular density within the
peri-infarct at 1 day poststroke. Interestingly, spontaneous biologi-
cal recovery occurs by gradual increase in revascularization over
several weeks after the initial stroke.36,37 Angiogenic response is
essential for ischemic brain repair and in poststroke brain may
indicate a higher demand for vascular support in regions under-
going active tissue reorganization. Importantly, GH treatment
in stroked mice increased the density and area coverage of both
CD31 and collagen-IV–positive cells, suggesting the possibility
of GH-induced revascularization beyond what occurs with spon-
taneous recovery. There were no significant changes in average
diameter of vessels in GH treatment (data not shown), suggesting
Downloaded from http://stroke.ahajournals.org/ by guest on June 6, 2018
that the changes of density and area coverage observed was driven
by an increase in the total number of vessels and increases in their
length. It should be pointed out that the relative increase in area
coverage for these vascular markers is ≈40%, and these are com-
paratively high figures, in terms of angiogenesis.
GH treatment resulted in several biological effects that
are completely consistent with those reported previously
in the literature. Notably, mice treated with GH gained a
significant amount of weight across the treated window.
We also identified that GH resulted in significantly higher
circulating levels of IGF-1 and VEGF, both of which are
known to be elevated after GH treatment. The enhancement
of IGF-1 and VEGF apart from confirming the efficacy of
GH treatment in this model is also of interest as higher cir-
culatory levels of these growth factors straight after stroke
correlates with better long-term outcome in patients.38,39
Our data suggested that higher circulatory levels of IGF-1
is associated with better cognition in stroked mice. We also
found a robust increase in IGF-1 and VEGF levels within
the peri-infarct region of the brain in GH-treated stroked
mice. One potential concern with GH treatment is its abil-
ity to promote diabetogenic effects. In the current study, we
observed no stimulatory effects on glucose between GH-
and non–GH-treated mice, which we think is of potential
clinical importance.
One observation from the current study that requires con-
sideration is that stroked mice treated with GH exhibited
significantly less tissue loss from the ipsilateral hemisphere.
Although we cannot definitely account for these tissue gains,
we would not be inclined to attribute them to GH-promoting
neuronal survival within the infarct, as treatment of GH com-
menced at 48 hours poststroke at a time when most, if not
all, neurons within the infarcted territory would have died.
Instead, we would speculate that the greater tissue volume
seen in GH-treated mice could be attributable to an increase
in synaptogenesis, myelination, and vessel growth occurring
within the ipsilateral hemisphere. However, we would also
give consideration to glia, including astrocytes and microglia,
as there is some evidence that these cells have significantly
greater oxidative tolerance when compared with neurons.40,41
It will be of considerable interest to examine in future studies,
using an approach such as bromodeoxyuridine tagging,42 to
specifically identify which populations are most responsible
for the less severe levels of tissue loss observed in mice treated
with GH and to determine the time frame in which this occurs.
There are multiple avenues to consider in terms of advanc-
ing the findings of the current study. First, with respect to
the effects of GH on cognition, we note that we have only
examined the associative memory cognitive domain. We
were largely constrained to one assessment because of the
significant time required to perform testing. Nevertheless, it
would certainly be worthwhile for future studies to consider
examining long-term memory and attention. Further dosing
and timing studies on the procognitive effects of GH, such as
how long after the cessation of the treatment do the cognitive
enhancements persist for and specifically address whether a
one-off or repeated administration would be required to pro-
mote cognition, are warranted.
Conclusions
We have shown that peripheral GH treatment poststroke stim-
ulates cognitive recovery, which could make GH a useful ther-
apeutic tool in the recovery phase of patients after stroke. Our
findings of increased expression of neurotrophic factors, such
as IGF-1 and VEGF, markers of synaptogenesis and myelina-
tion, and the formation of cerebrovascular networks confirm
that GH has a wide range of neurorestorative effects that could
have a potential in a clinical setting. Encouragingly, some
small and pilot studies in humans indicate a positive outcome
when GH treatment is used together with specific rehabilita-
tion after traumatic brain injury or stroke.24–28 The results from
this study, supported by previous studies, suggest a rational
for clinical intervention studies with GH after stroke.
Acknowledgments
We thank Hunter Medical Research Institute Core Histology Facility
for assistance with the histology images. We also thank Leeanne
Carey for the insights in article preparation.
Sources of Funding
This study was supported by the Swedish Government (ALFGBG
74 390), University of Gothenburg, James S. McDonnell Foundation,
Hunter Medical Research Institute, The Brawn Bequest, Priority
Research Centre for Stroke and Brain Injury Research Support
Grant, Mary Costello Alzheimer’s Pilot Grant, and The University of
Newcastle, Australia.
Disclosures
None.
References
1. Levine DA, Galecki AT, Langa KM, Unverzagt FW, Kabeto MU,
Giordani B, et al. Trajectory of cognitive decline after incident stroke.
JAMA. 2015;314:41–51. doi: 10.1001/jama.2015.6968.
2. Sun JH, Tan L, Yu JT. Post-stroke cognitive impairment: epidemiol-
ogy, mechanisms and management. Ann Transl Med. 2014;2:80. doi:
10.3978/j.issn.2305-5839.2014.08.05.
3. Nyberg F, Hallberg M. Growth hormone and cognitive function. Nat Rev
Endocrinol. 2013;9:357–365. doi: 10.1038/nrendo.2013.78.
4. Falleti MG, Maruff P, Burman P, Harris A. The effects of growth hormone
(GH) deficiency and GH replacement on cognitive performance in adults:
 1266  Stroke  May 2018
a meta-analysis of the current literature. Psychoneuroendocrinology.
2006;31:681–691. doi: 10.1016/j.psyneuen.2006.01.005.
5. Waters MJ, Blackmore DG. Growth hormone (GH), brain development
and neural stem cells. Pediatr Endocrinol Rev. 2011;9:549–553.
6. Aberg ND, Brywe KG, Isgaard J. Aspects of growth hormone and
insulin-like growth factor-I related to neuroprotection, regeneration, and
functional plasticity in the adult brain. Sci World J. 2006;6:53–80. doi:
10.1100/tsw.2006.22.
7. Devesa J, Almengló C, Devesa P. Multiple effects of growth hormone
in the body: is it really the hormone for growth? Clin Med Insights
Endocrinol Diabetes. 2016;9:47–71. doi: 10.4137/CMED.S38201.
8. Stochholm K, Kiess W. Long-term safety of growth hormone-a com-
bined registry analysis. Clin Endocrinol (Oxf). 2018;88:515–528. doi:
10.1111/cen.13502.
9. Zhang H, Han M, Zhang X, Sun X, Ling F. The effect and mecha-
nism of growth hormone replacement on cognitive function in
rats with traumatic brain injury. PLoS One. 2014;9:e108518. doi:
10.1371/journal.pone.0108518.
10. Enhamre-Brolin E, Carlsson A, Hallberg M, Nyberg F. Growth hormone
reverses streptozotocin-induced cognitive impairments in male mice.
Behav Brain Res. 2013;238:273–278. doi: 10.1016/j.bbr.2012.10.036.
11. Pathipati P, Surus A, Williams CE, Scheepens A. Delayed and
chronic treatment with growth hormone after endothelin-induced
stroke in the adult rat. Behav Brain Res. 2009;204:93–101. doi:
Downloaded from http://stroke.ahajournals.org/ by guest on June 6, 2018
10.1016/j.bbr.2009.05.023.
12. Balkaya MG, Trueman RC, Boltze J, Corbett D, Jolkkonen J. Behavioral
outcome measures to improve experimental stroke research [published
online ahead of print July 29, 2017]. Behav Brain Res. doi: 10.1016/j.
bbr.2017.07.039. https://www.sciencedirect.com/science/article/pii/S01
66432817307325.
13. Schaar KL, Brenneman MM, Savitz SI. Functional assessments in
the rodent stroke model. Exp Transl Stroke Med. 2010;2:13. doi:
10.1186/2040-7378-2-13.
14. Kluge MG, Kracht L, Abdolhoseini M, Ong LK, Johnson SJ, Nilsson
M, et al. Impaired microglia process dynamics post-stroke are specific
to sites of secondary neurodegeneration. Glia. 2017;65:1885–1899. doi:
10.1002/glia.23201.
15. Ong LK, Zhao Z, Kluge M, Walker FR, Nilsson M. Chronic stress expo-
sure following photothrombotic stroke is associated with increased levels
of Amyloid beta accumulation and altered oligomerisation at sites of tha-
lamic secondary neurodegeneration in mice. J Cereb Blood Flow Metab.
2017;37:1338–1348. doi: 10.1177/0271678X16654920.
16. Zalewska K, Ong LK, Johnson SJ, Nilsson M, Walker FR. Oral admin-
istration of corticosterone at stress-like levels drives microglial but not
vascular disturbances post-stroke. Neuroscience. 2017;352:30–38. doi:
10.1016/j.neuroscience.2017.03.005.
17. Zhao Z, Ong LK, Johnson S, Nilsson M, Walker FR. Chronic stress
induced disruption of the peri-infarct neurovascular unit following
experimentally induced photothrombotic stroke. J Cereb Blood Flow
Metab. 2017;37:3709–3724. doi: 10.1177/0271678X17696100.
18. Zalewska K, Pietrogrande G, Ong LK, Abdolhoseini M, Kluge M,
Johnson SJ, et al. Sustained administration of corticosterone at stress-
like levels after stroke suppressed glial reactivity at sites of thalamic sec-
ondary neurodegeneration. Brain Behav Immun. 2018;69:210–222. doi:
10.1016/j.bbi.2017.11.014.
19. Walser M, Hansén A, Svensson PA, Jernås M, Oscarsson J, Isgaard J, et
al. Peripheral administration of bovine GH regulates the expression of
cerebrocortical beta-globin, GABAB receptor 1, and the Lissencephaly-1
protein (LIS-1) in adult hypophysectomized rats. Growth Horm IGF Res.
2011;21:16–24. doi: 10.1016/j.ghir.2010.11.002.
20. Shepherd A, Tyebji S, Hannan AJ, Burrows EL. Translational
assays for assessment of cognition in rodent models of Alzheimer’s
disease and dementia. J Mol Neurosci. 2016;60:371–382. doi:
10.1007/s12031-016-0837-1.
21. Horner AE, Heath CJ, Hvoslef-Eide M, Kent BA, Kim CH, Nilsson
SR, et al. The touchscreen operant platform for testing learning
and memory in rats and mice. Nat Protoc. 2013;8:1961–1984. doi:
10.1038/nprot.2013.122.
22. Ong LK, Zhao Z, Kluge M, TeBay C, Zalewska K, Dickson PW, et al.
Reconsidering the role of glial cells in chronic stress-induced dopami-
nergic neurons loss within the substantia nigra? Friend or foe? Brain
Behav Immun. 2017;60:117–125. doi: 10.1016/j.bbi.2016.10.001.
23. Tynan RJ, Beynon SB, Hinwood M, Johnson SJ, Nilsson M, Woods
JJ, et al. Chronic stress-induced disruption of the astrocyte network is
driven by structural atrophy and not loss of astrocytes. Acta Neuropathol.
2013;126:75–91. doi: 10.1007/s00401-013-1102-0.
24. Devesa J, Reimunde P, Devesa P, Barberá M, Arce V. Growth hor-
mone (GH) and brain trauma. Horm Behav. 2013;63:331–344. doi:
10.1016/j.yhbeh.2012.02.022.
25. Moreau OK, Cortet-Rudelli C, Yollin E, Merlen E, Daveluy W, Rousseaux
M. Growth hormone replacement therapy in patients with traumatic brain
injury. J Neurotrauma. 2013;30:998–1006. doi: 10.1089/neu.2012.2705.
26. High WM Jr, Briones-Galang M, Clark JA, Gilkison C, Mossberg KA,
Zgaljardic DJ, et al. Effect of growth hormone replacement therapy on
cognition after traumatic brain injury. J Neurotrauma. 2010;27:1565–
1575. doi: 10.1089/neu.2009.1253.
27. Reimunde P, Quintana A, Castañón B, Casteleiro N, Vilarnovo Z, Otero
A, et al. Effects of growth hormone (GH) replacement and cognitive
rehabilitation in patients with cognitive disorders after traumatic brain
injury. Brain Inj. 2011;25:65–73. doi: 10.3109/02699052.2010.536196.
28. Song J, Park K, Lee H, Kim M. The effect of recombinant human growth
hormone therapy in patients with completed stroke: a pilot trial. Ann
Rehabil Med. 2012;36:447–457. doi: 10.5535/arm.2012.36.4.447.
29. Labat-gest V, Tomasi S. Photothrombotic ischemia: a minimally inva-
sive and reproducible photochemical cortical lesion model for mouse
stroke studies. J Vis Exp. 2013. doi: 10.3791/50370. https://www.jove.
com/video/50370/photothrombotic-ischemia-minimally-invasive-
reproducible.
30. Diederich K, Schmidt A, Strecker JK, Schäbitz WR, Schilling M,
Minnerup J. Cortical photothrombotic infarcts impair the recall of previ-
ously acquired memories but spare the formation of new ones. Stroke.
2014;45:614–618. doi: 10.1161/STROKEAHA.113.001907.
31. Schmidt A, Diederich K, Strecker JK, Geng B, Hoppen M, Duning
T, et al. Progressive cognitive deficits in a mouse model of recur-
rent photothrombotic stroke. Stroke. 2015;46:1127–1131. doi:
10.1161/STROKEAHA.115.008905.
32. Walker FR, Jones KA, Patience MJ, Zhao Z, Nilsson M. Stress as
necessary component of realistic recovery in animal models of experi-
mental stroke. J Cereb Blood Flow Metab. 2014;34:208–214. doi:
10.1038/jcbfm.2013.211.
33. Weinstein G, Preis SR, Beiser AS, Au R, Kelly-Hayes M, Kase CS, et al.
Cognitive performance after stroke–the Framingham Heart Study. Int J
Stroke. 2014;9(suppl A100):48–54. doi: 10.1111/ijs.12275.
34. Barnett JH, Blackwell AD, Sahakian BJ, Robbins TW. The paired
associates learning (pal) test: 30 years of cantab translational neurosci-
ence from laboratory to bedside in dementia research. In: Robbins TW,
Sahakian BJ, eds. Translational Neuropsychopharmacology. Cham:
Springer International Publishing; 2016:449–474.
35. Nithianantharajah J, McKechanie AG, Stewart TJ, Johnstone M,
Blackwood DH, St Clair D, et al. Bridging the translational divide: iden-
tical cognitive touchscreen testing in mice and humans carrying muta-
tions in a disease-relevant homologous gene. Sci Rep. 2015;5:14613. doi:
10.1038/srep14613.
36. Hayashi T, Noshita N, Sugawara T, Chan PH. Temporal pro-
file of angiogenesis and expression of related genes in the brain
after ischemia. J Cereb Blood Flow Metab. 2003;23:166–180. doi:
10.1097/01.WCB.0000041283.53351.CB.
37. Yanev P, Seevinck PR, Rudrapatna US, Bouts MJ, van der Toorn A,
Gertz K, et al. Magnetic resonance imaging of local and remote vascu-
lar remodelling after experimental stroke. J Cereb Blood Flow Metab.
2017;37:2768–2779. doi: 10.1177/0271678X16674737.
38. Åberg D, Jood K, Blomstrand C, Jern C, Nilsson M, Isgaard J, et al.
Serum IGF-I levels correlate to improvement of functional outcome after
ischemic stroke. J Clin Endocrinol Metab. 2011;96:E1055–E1064. doi:
10.1210/jc.2010-2802.
39. Lee SC, Lee KY, Kim YJ, Kim SH, Koh SH, Lee YJ. Serum VEGF levels
in acute ischaemic strokes are correlated with long-term prognosis. Eur
J Neurol. 2010;17:45–51. doi: 10.1111/j.1468-1331.2009.02731.x.
40. Lee DR, Helps SC, Gibbins IL, Nilsson M, Sims NR. Losses of NG2 and
NeuN immunoreactivity but not astrocytic markers during early reperfusion
following severe focal cerebral ischemia. Brain Res. 2003;989:221–230.
41. Thoren AE, Helps SC, Nilsson M, Sims NR. Astrocytic function
assessed from 1-14C-acetate metabolism after temporary focal cerebral
ischemia in rats. J Cereb Blood Flow Metab. 2005;25:440–450. doi:
10.1038/sj.jcbfm.9600035.
42. Kuhn HG, Eisch AJ, Spalding K, Peterson DA. Detection and phenotypic
characterization of adult neurogenesis. Cold Spring Harb Perspect Biol.
2016;8:a025981. doi: 10.1101/cshperspect.a025981.
 Growth Hormone Improves Cognitive Function After Experimental Stroke
Lin Kooi Ong, Wei Zhen Chow, Clifford TeBay, Murielle Kluge, Giovanni Pietrogrande,
Katarzyna Zalewska, Patricia Crock, N. David Åberg, Andrew Bivard, Sarah J. Johnson,
Frederick R. Walker, Michael Nilsson and Jörgen Isgaard
Downloaded from http://stroke.ahajournals.org/ by guest on June 6, 2018

Stroke. 2018;49:1257-1266; originally published online April 10, 2018;

doi: 10.1161/STROKEAHA.117.020557
Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 2018 American Heart Association, Inc. All rights reserved.
Print ISSN: 0039-2499. Online ISSN: 1524-4628

The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://stroke.ahajournals.org/content/49/5/1257

Data Supplement (unedited) at:

http://stroke.ahajournals.org/content/suppl/2018/04/10/STROKEAHA.117.020557.DC1

Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Stroke can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.
Once the online version of the published article for which permission is being requested is located, click
Request Permissions in the middle column of the Web page under Services. Further information about this
process is available in the Permissions and Rights Question and Answer document.

Reprints: Information about reprints can be found online at:

http://www.lww.com/reprints

Subscriptions: Information about subscribing to Stroke is online at:

http://stroke.ahajournals.org//subscriptions/
 SUPPLEMENTAL MATERIAL

Title: Growth hormone improves cognitive function after experimental stroke.

Authors:
Lin Kooi Ong1,2,3, Wei Zhen Chow1,2,3, Clifford Tebay1, Murielle Kluge1,2,3, Giovanni
Pietrogrande1,2,3, Katarzyna Zalewska1,2,3, Patricia Crock4, N. David Åberg5,6, Andrew
Bivard2,3,7, Sarah J Johnson2,8, Frederick R Walker1,2,3, Michael Nilsson2,3, Jörgen Isgaard2,6

Author affiliations:
1
School of Biomedical Sciences and Pharmacy and Priority Research Centre for Stroke and
Brain Injury, University of Newcastle, Australia.
2
Hunter Medical Research Institute, Australia.
3
NHMRC Centre of Research Excellence Stroke Rehabilitation and Brain Recovery,
Australia.
4
Department of Paediatric Endocrinology and Diabetes, John Hunter Children's Hospital and
Priority Research Centre Grow Up Well, University of Newcastle, Australia.
5
Centre for Brain Repair and Rehabilitation, Institute of Neuroscience and Physiology,
Sahlgrenska University Hospital, University of Gothenburg, Sweden.
6
Department of Internal Medicine, Sahlgrenska University Hospital, University of Gothenburg,
Sweden.
7
Departments of Neurology, John Hunter Hospital and University of Newcastle, Australia.
Hunter Medical Research Institute, Australia.
8
School of Electrical Engineering and Computing, University of Newcastle, Australia.

Corresponding Authors:
Lin Kooi Ong, PhD, School of Biomedical Sciences and Pharmacy, University of Newcastle,

Australia. Phone: +61249215736 Email: lin.ong@uon.edu.au

Jörgen Isgaard, MD, PhD, Department of Internal Medicine, Sahlgrenska University Hospital,

University of Gothenburg, Gothenburg, Sweden. Tel: +46317411714 Email:

jorgen.isgaard@medic.gu.se

1
 ONLINE SUPPLEMENT

Supplemental Methods
Animals. C57BL/6 male mice (eight weeks old) were obtained from the Animal Services Unit
at the University of Newcastle. Mice were maintained in a temperature (21˚C±1) and humidity
controlled environment with food and water available ad libitum. Lighting was on a 12:12 h
reverse light–dark cycle (lights on 19:00 h) with all procedures conducted in the dark phase.
Mice were habituated for a minimum of seven days prior to the start of the experiment.

Randomization and Blinding. In all instances mice were randomly allocated to experimental
groups. All outcome analyses were performed by independent study team members blinded to
the treatment condition.

Photothrombotic occlusion. Photothrombotic occlusion was performed as described

previously.1-5 Mice were anaesthetized by 2% isoflurane during surgical procedure on a
temperature controlled (37˚C ± 1) stereotaxic frame. The skull was exposed by incision of the
skin along the midline of the scalp. Rose Bengal (200 µl, 10 mg/ml solution in sterile saline,
Sigma-Aldrich, USA) was injected intraperitoneally. After 8 min, the skull was illuminated for
15 min by a 4.5 mm diameter cold light source positioned at 2.2 mm left lateral of Bregma,
targeting the left motor and somatosensory cortices. For the sham group, the same surgical
procedure was applied except Rose Bengal was replaced with 200 µl of sterile saline (0.9%
NaCl, Pfizer, Australia).

Mini-osmotic pump placement. Mini-osmotic pump (Model 2004, Alzet, USA) placement was
performed as described.6 Mice were allowed to recover for 48 hours before mini-osmotic pump
placement was performed. Mice were anaesthetized by 2% of isoflurane during surgical
procedure on a temperature controlled (37˚C±1) stereotaxic frame. An incision was made in
the skin between the scapulae and a pocket was formed by spreading the subcutaneous
connective tissues apart. The mini-osmotic pumps were filled with 200 µl of either recombinant
human growth hormone (r-hGH, Somatropin 10mg/1.5 mL, SciTropin A, SciGen, Australia)
or sterile saline. The pumps deliver 0.25 µl/hr for 28 days (0.04 mg r-hGH per day). The pump
was inserted into the pocket and skin incision was closed with sutures. Pumps were weighed
before placement and after mice were sacrificed.

2
Mouse touchscreen platform for paired-associate learning task. Mouse touchscreen operant
chambers (Fig. 2A, Campden Instruments Ltd, England) were used in the cognitive testing as
described with modifications 7, and were conducted in a blinded and randomised manner. Mice
were calorie restricted overnight before cognitive testing (allowed access to food after
cognitive testing for 6 hours, between 12:00 – 18:00 h). Mice were introduced to a series of
habituation and training tasks where they were placed in the operant chambers and trained to
associate nosepoke touchscreen with reward liquid food delivery (strawberry milkshake). Over
7 days all mice learnt to perform the task with a minimum correct rate of 70%. Mice then
underwent photothrombotic occlusion surgery, followed by mini-osmotic pump placement.
Following a recovery period of seven days, mice were introduced to the PAL task where they
learn to associate novel objects (flower, plane and spider) with their specific locations (left,
middle and right). At any one time point, there was two images on the screen with only one
image presented at its correct location (S+, rewarded stimulus and S-, unrewarded stimulus)
(Fig. 2A). When the mouse made a correct choice, S+, a tone will be triggered, the food
receptacle was illuminated and the food reward was delivered. A period of inter-trial interval
(ITI) of 10 seconds occurred between each task. For S- response, there was a 10 seconds time
out and the house light was deactivated (lights on). The food receptacle was illuminated and a
poke initiated the correction trial which displayed the same stimuli as in the previous session.
In each PAL session, mice were required to complete 36 trials within a maximum time of 60
minutes per day, whichever occurred first. All mice completed a total of 20 sessions. The
primary outcome measure was percentage correct rate at final PAL session. A range of metrics
were also collected for temporal analysis and data was presented as four blocks of mean of five
consecutive sessions (Figure III).

Tissue processing. Mice were deeply anaesthetised with sodium pentabarbitol. Blood samples
(approximately 0.5 mL) were collected transcardially. Blood glucose was measured
immediately using Accu-check performa glucometer (Roche, Mannheim, Germany). Blood
samples were incubated for 30 min at room temperature and then centrifuged 5000g for 5 min
at 4°C, serum was collected and frozen in -80°C for future analyses. Following blood
collection, mice were perfused via the ascending aorta with ice cold 0.9% saline followed by
ice cold 4% paraformaldehyde (pH 7.4) for histological analysis or with ice cold 0.1%
diethylpyrocarbonate in 0.9% saline for biochemical analysis. For histological analysis, brains
were dissected, post-fixed in 12.5% sucrose in 4% paraformaldehyde for 4 h and transferred to

3
12.5% sucrose in PBS for storage and cryoprotection. Coronal sections of the brains were
sectioned with a freezing microtome (Leica, North Ryde, NSW, Australia) at 30 µm. For
biochemical analysis, brains were dissected and rapidly frozen in -80˚C isopentane. Frozen
brains were sliced using the cryostat at the thickness of 200 µm and peri-infarct territory (2
mm2 around infarct core) were punched using a 1 mm tissue punch (Bregma +1.0 to -1.0).
Samples were stored frozen in -80˚C until further analysis.

Protein extraction. Protein homogenates from the peri-infarct samples were obtained as
described.4 Peri-infarct samples were sonicated in 300 µl lysis buffer (50 mM TRIS buffer pH
7.4, 1 mM EDTA, 1 mM DTT, 80 µM ammonium molybdate, 1 mM sodium pyrophosphate,
1 mM sodium vanadate, 5 mM b-glycerolphosphate, 1 protease inhibitor cocktail tablet, 1
phosphatase inhibitor cocktail tablet, final concentration) and centrifuged 14,000g for 20 min
at 4°C. Next, supernatants were collected and protein levels were estimated by using
commercially available Pierce BCA (Thermo Fisher Scientific, USA) protein assay kit
according to the manufacturer’s instructions. Samples were diluted to equalize protein
concentrations (1.5 mg/mL), aliquoted and stored at -80˚C until further analysis.

Western blotting. Western blotting was performed as described.8 Peri-infarct supernatants were
mixed with sample buffer (2% sodium dodecyl sulfate, 50 mM Tris, 10% glycerol, 1% DTT,
0.1% bromophenol blue, pH 6.8). 15 µg of total tissue protein samples were electrophoresed
to Biorad Criterion TGC Stain-Free 4–20% gels and gels were transferred to PVDF
membranes. PVDF membranes were washed in Tris-buffered saline with tween (TBST) (150
mM NaCl, 10 mM Tris, 0.075% Tween-20, pH 7.5) and incubated in 5% skim milk powder in
TBST for 1 h at 25°C. Membranes were incubated with primary antibodies overnight at 4°C
and secondary antibody for 1 h at 25°C (see Table II for antibodies concentration). In between
each incubation step, membranes were washed in TBST. Membranes were visualized on an
Amersham Imager 600 using Luminata Classico western blotting detection reagents. The
density of the bands was measured using Amersham Imager 600 analysis software.

Cresyl Violet staining. Cresyl Violet staining was performed as described.9 Free floating
sections (5 sections per brain, Bregma 1.5 to -1.5) were mounted on glass slides and air-dried.
Sections were defatted in chloroform: ethanol solution for 8 min, followed by rehydration in a
series ethanol solutions (100, 95 and 70%). Sections were stained in Cresyl Violet solution for

4
15 min. Sections were washed in a series of ethanol solutions (70, 95 and 100%), cleared in
xylene and cover slipped.

Immunohistochemistry. For immunoperoxidase labelling, free-floating sections were

immunostained as described.4 All reactions for labels marker were run at the same time, with
same reagents, at the same concentrations. Peri-infarct sections (Bregma 0.0) were incubated
with 1% hydrogen peroxidase for 30 min at 25°C and followed by 3% horse serum for 30 min
at 25°C. Brain sections were incubated with primary antibodies for 72 h at 4°C and followed
by secondary antibodies for 1 h at 25°C (Table II for antibodies concentration). Next, brain
sections were incubated for 2 h at 25°C with avidin–biotin-peroxidase complex and finally
developed using DAB peroxidase substrate. Brain sections were washed with PBS in between
each incubation step. After processing was complete, brain sections were mounted onto chrome
alum-coated slides and cover slipped.

Images were acquired at 20 X using Aperio AT2 (Leica, Germany). The estimated tissue loss
area [area of contralateral hemisphere – area of ipsilateral hemisphere] was determined on
Cresyl Violet stained sections with ImageJ software 1.50 a, NIH. The quantitative analysis was
undertaken specifically in the peri-infarct territory as defined by 0.01 mm from infarct, the
regions were 0.5 mm by 0.5 mm in size. For the purpose of analysis, pixel intensity level
considered to be optimal for detecting genuine differences in immunoreactive signal was
determined using ImageJ software to visualise thresholding of cropped regions at individual
pixel intensities. This threshold level was used to investigate group differences for all labels.
10, 11
Figure IV shows cumulative threshold analysis performed using Matlab functions . The
number of pixels occurring at each of the pixel intensities was determined. The pixel intensity
values are then rank ordered 0-255 along with the corresponding number of pixel that occur at
each value. The cumulative threshold analysis shows the consistency of differences across the
threshold range.

The vessel digital reconstruction was performed as described 3, 5. CD31 or Collagen IV positive
cells were isolated from the background using multi-level Otsu’s thresholding method which
calculates the threshold that minimizes the interclass pixel intensity variance between groups.
Using Matlab functions we determined percentage area covered by the labelling.

5

Table I. Checklist of Methodological and Reporting Aspects for Articles Submitted to Stroke Involving Preclinical Experimentation
Methodological and Reporting Aspects
Experimental groups and study
timeline
Inclusion and exclusion criteria
Randomization
Blinding
Sample size and power
calculations
Data reporting and statistical
methods
Experimental details, ethics,
and funding statements
Stroke Online Supplement
Description of Procedures
□
✔ The experimental group(s) have been clearly defined in the article, including number of animals in each
experimental arm of the study.
□
✔ An account of the control group is provided, and number of animals in the control group has been reported. If no
controls were used, the rationale has been stated.
□
✔ An overall study timeline is provided.
□
✔ A priori inclusion and exclusion criteria for tested animals were defined and have been reported in the article.
□
✔ Animals were randomly assigned to the experimental groups. If the work being submitted does not contain
multiple experimental groups, or if random assignment was not used, adequate explanations have been provided.
□
✔ Type and methods of randomization have been described.
□
✔ Methods used for allocation concealment have been reported.
□
✔ Blinding procedures have been described with regard to masking of group/treatment assignment from the
experimenter. The rationale for nonblinding of the experimenter has been provided, if such was not feasible.
□
✔ Blinding procedures have been described with regard to masking of group assignment during outcome assessment.
□
✔ Formal sample size and power calculations were conducted based on a priori determined outcome(s) and
treatment effect, and the data have been reported. A formal size assessment was not conducted and a
rationale has been provided.
□
✔ Number of animals in each group: randomized, tested, lost to follow-up, or died have been reported. If the
experimentation involves repeated measurements, the number of animals assessed at each time point is provided,
for all experimental groups.
□
✔ Baseline data on assessed outcome(s) for all experimental groups have been reported.
□
✔ Details on important adverse events and death of animals during the course of experimentation have been provided,
for all experimental arms.
□
✔ Statistical methods used have been reported.
□
✔ Numeric data on outcomes have been provided in text, or in a tabular format with the main article or as
supplementary tables, in addition to the figures.
□
✔ Details on experimentation including stroke model, formulation and dosage of therapeutic agent, site and
route of administration, use of anesthesia and analgesia, temperature control during experimentation, and
postprocedural monitoring have been described.
□
✔ Different sex animals have been used. If not, the reason/justification is provided.
□
✔ Statements on approval by ethics boards and ethical conduct of studies have been provided.
□
✔ Statements on funding and conflicts of interests have been provided.
6
Table II: List of antibodies used for western blot and immunohistochemistry analyses.
Target Sources of antibodies Application Dilution
IGF-1 Santa Cruz Biotechnology, IGF-1 (H-70), sc-9013 WB 1:500
VEGF Abcam, Anti-VEGFA, ab46154 WB 1:2000
NeuN Cell Signaling, Anti-NeuN (D3S31), #12943 WB 1:2000
PSD-95 Cell Signaling, PSD95 Antibody, #2507 WB 1:1000
Synapsin-1 Cell Signaling, Anti-Synapsin-1 (D12G5), #5297 WB 1:10000
Millipore, Anti-Synapsin-1, clone 10.22, IHC 1:1000
MABN894
MBP Cell Signaling, Anti-Myelin Basic Protein WB 1:1000
(D8X4Q), #78896
IHC 1:1000
CD31 Cell Signaling, Anti-CD31(PECAM-1) (D8V9E), WB 1:1000
#77699
IHC 1:1000
Collagen-IV Abcam, Anti-Collagen IV, ab6587 WB 1:1000
IHC 1:1000
² -actin Sigma-Aldrich, Monoclonal Anti-² -actin-HRP WB 1:50000
antibody, A3854
Rabbit IgG Biorad, Anti-Rabbit-HRP antibody, #170-6515 WB 1:5000
Jackson ImmunoReseach, Anti-Rabbit-Biotin, IHC 1:500
#111-065-003
Mouse IgG Biorad, Anti-Mouse-HRP antibody, #170-6516 WB 1:10000
Jackson ImmunoReseach, Anti-Mouse-Biotin, IHC 1:500
#115-065-003
WB, western blot; IHC, immunohistochemistry.

7
Supplemental Figures and Figure Legends

Figure I. Diagram for inclusion and exclusion of mice in this study. Mice were removed
from the study if we histologically identified that the stroke had not occurred or malfunction
of the mini-osmotic pumps. A total of 14 sham mice were used in pilot study (Sham + Saline,
n=6; Sham + r-hGH, n=8). A total of 17 stroke mice were used for biochemistry analysis of the
brains (Stroke + Saline, n=8; Stroke + r-hGH, n=9). A total of 20 stroke mice were used for
paired-associate learning (PAL) task (Stroke + Saline, n=11; Stroke + r-hGH, n=9). Blood
samples were analysed in all included mice.

8
Figure II. Pilot study to validate the efficacy of r-hGH treatment in sham mice. A.
Experimental design. All mice received a sham surgery. 48 hrs later mice were randomly
treated with either saline or r-hGH via mini-osmotic pumps subcutaneously for 28 days. B. r-
hGH treatment significantly increased the body weight. C. r-hGH treatment significantly
increased serum IGF-1 and VEGF levels. r-hGH treatment significantly decreased blood
glucose levels. No significant change in percentage liver to total body weight. Mean±SD. *
p<0.05.

9
Figure III. Metrics of the paired-associate learning (PAL) task. r-hGH treatment
significantly increased percentage correct rate. While there is no statistically significant
between treatments (saline vs r-hGH), stroked mice treated with r-hGH have the trend of
improved cognition indexed by number of trials completed within 60 min, time to complete 36
trials, repeated tasks per trials, number of correct trials completed and correct touch latency.
Mean±SD. * p<0.05.

10
Figure IV. Cumulative threshold analysis. The percentage of material thresholded of A.
Synapsin 1, B. MBP, C. CD31 and D. Collagen IV immunolabels at different levels of pixel
intensity (PI). Dotted line are PI level considered to be optimal for detecting genuine
differences in immunoreactive signal (Fig. 5 and 6). Graphs showed as mean.

11
Figure V. Peripheral growth factors as prognostic biomarker of cognitive outcome. A.
Pearson correlation analysis shows a significant positive correlation of serum IGF-1 levels and
cognitive performance (r=0.5914, p (Y = 0.04493X + 44.24) = 0.0076). B. Pearson correlation analysis
shows no correlation of serum VEGF levels and cognitive performance (r=0.398, p (Y = 0.2271*X
+ 46.01) = 0.0915).  Stroke + Saline;  Stroke + r-hGH.

Supplemental References

1. Kluge MG, Kracht L, Abdolhoseini M, Ong LK, Johnson SJ, Nilsson M, et al. Impaired microglia
process dynamics post-stroke are specific to sites of secondary neurodegeneration. Glia.
2017;65:1885-1899
2. Ong LK, Zhao Z, Kluge M, Walker FR, Nilsson M. Chronic stress exposure following
photothrombotic stroke is associated with increased levels of amyloid beta accumulation
and altered oligomerisation at sites of thalamic secondary neurodegeneration in mice. J
Cereb Blood Flow Metab. 2017;37:1338-1348
3. Zalewska K, Ong LK, Johnson SJ, Nilsson M, Walker FR. Oral administration of corticosterone
at stress-like levels drives microglial but not vascular disturbances post-stroke. Neuroscience.
2017;352:30-38
4. Zhao Z, Ong LK, Johnson S, Nilsson M, Walker FR. Chronic stress induced disruption of the
peri-infarct neurovascular unit following experimentally induced photothrombotic stroke. J
Cereb Blood Flow Metab. 2017:271678X17696100
5. Zalewska K, Pietrogrande G, Ong LK, Abdolhoseini M, Kluge M, Johnson SJ, et al. Sustained
administration of corticosterone at stress-like levels after stroke suppressed glial reactivity
at sites of thalamic secondary neurodegeneration. Brain Behav Immun. 2017
6. Walser M, Hansen A, Svensson PA, Jernas M, Oscarsson J, Isgaard J, et al. Peripheral
administration of bovine gh regulates the expression of cerebrocortical beta-globin, gabab
receptor 1, and the lissencephaly-1 protein (lis-1) in adult hypophysectomized rats. Growth
Horm IGF Res. 2011;21:16-24
7. Horner AE, Heath CJ, Hvoslef-Eide M, Kent BA, Kim CH, Nilsson SR, et al. The touchscreen
operant platform for testing learning and memory in rats and mice. Nat Protoc.
2013;8:1961-1984

12
8. Ong LK, Zhao Z, Kluge M, TeBay C, Zalewska K, Dickson PW, et al. Reconsidering the role of
glial cells in chronic stress-induced dopaminergic neurons loss within the substantia nigra?
Friend or foe? Brain Behav Immun. 2017;60:117-125
9. Tynan RJ, Beynon SB, Hinwood M, Johnson SJ, Nilsson M, Woods JJ, et al. Chronic stress-
induced disruption of the astrocyte network is driven by structural atrophy and not loss of
astrocytes. Acta Neuropathol. 2013;126:75-91
10. Johnson SJ, Walker FR. Strategies to improve quantitative assessment of
immunohistochemical and immunofluorescent labelling. Sci Rep. 2015;5:10607
11. Kongsui R, Johnson SJ, Graham BA, Nilsson M, Walker FR. A combined cumulative threshold
spectra and digital reconstruction analysis reveal structural alterations of microglia within
the prefrontal cortex following low-dose lps administration. Neuroscience. 2015;310:629-
640

13