doi:10.1093/brain/awl139 (2006), 129, 2158–2176 changes in ... · immediately frozen in liquid...

doi:10.1093/brain/awl139 Brain (2006), 129, 2158–2176

Changes in hyaluronan production and metabolismfollowing ischaemic stroke in man

Ahmed Al’Qteishat,1 John Gaffney,1 Jerzy Krupinski,4,5 Francisco Rubio,4,5 David West,3 Shant Kumar,2

Patricia Kumar,1 Nicholas Mitsios1 and Mark Slevin1

1Department of Biology, Chemistry and Health Science, Manchester Metropolitan University, 2University of Manchesterand Christie Hospital, Manchester, 3Department of Biochemistry, University of Liverpool, Liverpool, UK, 4Department ofNeurology, Stroke Unit, Hospital Universitario de Bellvitge (HUB) and 5Insitut D’Investigacio Biomedica de Belvitge(IDIBELL), Barcelona, Spain

Correspondence to: Mark Slevin, Department of Biology, Chemistry and Health Science, Manchester Metropolitan University,Manchester M1 5GD, UKE-mail: [email protected]

The extent of recovery from stroke is dependent on the survival of neurons, particularly in peri-infarctedregions. Angiogenesis is critical for the development of newmicrovessels and leads to re-formation of collateralcirculation, reperfusion and better recovery. Hyaluronan (HA) is an important component of the brain extra-cellular matrix and a regulator of cellular differentiation, migration, proliferation and angiogenesis. We havefound that the production of total HA and low molecular mass 3–10 disaccharides of HA (o-HA) was increasedin post-mortem tissue and in the serum of patients 1, 3, 7 and 14 days (peaking at 7 days) after ischaemic stroke.Hyaluronidase activity was also increased in serum samples (peaking after 3 days), which might explain thesubsequent increase in o-HA. Affinity-histochemical staining was performed using a HA-specific biotinylatedbinding protein, and it showed enhanced deposition of HA in blood vessels and intracellularly as well as in thenuclei of peri-infarcted neurons. Western blotting and immunohistochemistry demonstrated upregula-tion of HA synthases (HAS1 and 2) and hyaluronidases (HYAL1 and 2) in inflammatory cells from bothstroke and peri-infarcted regions of the brain. HYAL1 was upregulated in microvesssels and intracellularlyin neurons, whilst HAS2 became translocated into the nuclei of neurons in peri-infarcted areas. Receptor forHA-mediated motility was observed intracellularly and in the nuclei of neurons, in the tunica media of largerblood vessels and in the endothelial cells of microvessels in stroke-affected tissue, whilst expression of otherreceptors for HA, CD44 and tumour necrosis factor-stimulated gene 6 (TSG-6) were mainly increasedin infiltrating mononuclear cells from inflammatory regions. The data presented here demonstrate thatHA breakdown is a feature of the acute stage of stroke injury. Increased o-HA production soon after strokemay be detrimental through enhancement of the inflammatory response, whilst activation of HA and/or o-HA-induced cellular signalling pathways in neurons and microvessels may impact on the remodelling processby stimulating angiogenesis and revascularization, as well as the survival of susceptible neurons.

Keywords: RHAMM; hyaluronan; hyaluronidase; hyaluronan synthase; ischaemic stroke

Abbreviations: BSA = bovine serum albumin; EC = endothelial cells; ECM = extracellular matrix; HA = hyaluronan;o-HA = oligosaccharides of hyaluronan; HABP = hyaluronan binding protein; HAS = hyaluronan synthase; HYAL = hyaluronidase;IS= ischaemic stroke; PBS=phosphate-buffered saline; RHAMM= receptor for hyaluronan-mediatedmotility; RT–PCR= reversetranscription–polymerase chain reaction; SDS–PAGE = sodium dodecyl sulphate–polyacrylamide gel electrophoresis;SSS = Scandinavian Stroke Scale; TSG-6 = tumour necrosis factor-stimulated gene 6

Received January 20, 2006. Revised April 24, 2006. Accepted April 26, 2006. Advance Access publication May 26, 2006

IntroductionStroke is a leading cause of death and disability in the

Western world and usually occurs as a result of progressing

atherothrombosis, cardiac or artery–artery embolism (Rubio

et al., 2005) resulting in loss of membrane integrity, proteo-

lysis, the ability to synthesize proteins and altered signal

transduction (Lipton, 1999). The survival of neurons,

# The Author (2006). Published by Oxford University Press on behalf of the Guarantors of Brain. All rights reserved. For Permissions, please email: [email protected]

particularly in peri-infarcted regions, determines the

extent of patient recovery (Slevin et al., 2005). Stroke

patients with a higher density of blood vessels appear to

have reduced morbidity and survive longer (Krupinski

et al., 1993, 1994, 1996, 1997). It has also been suggested

that restoration of cerebral microvascular circulation follow-

ing angiogenesis/revascularization in peri-infarcted regions

may help salvage tissue and enhance functional recovery

after stroke (Coyle and Heistad, 1987; Sbarbarti et al.,

1996; Krupinski et al., 2003; Slevin et al., 2005).

Hyaluronan (HA) is abundant in the brain extracellular

matrix (ECM) together with hyaluronan binding proteins

(Delpech et al., 1989; Sherman et al., 2002); therefore, we

hypothesized that HA and associated enzymes [hyaluroni-

dases (HYALs); and hyaluronan synthases (HASs)] may be

mobilized following injury caused by stroke. This may

impact upon both neuronal survival (Nagy et al., 1995)

and revascularization, especially in the peri-infarcted

region. HA is a member of the glycosaminoglycan family

and is a polymer of repeating disaccharide units of N-

acetylglucosamine and D-glucuronic acid, linked by alter-

nating b1-3 and b1-4 bonds. HA has a strong affinity for

water and occupies a large hydrodynamic volume acting

as a lubricant, cushioning agent and supporting matrix

(Tammi et al., 2002). HA is synthesized by HASs on the

inner surface of the plasma membrane and is extruded across

the cell membrane into the ECM (Prehm, 1984; Itano

and Kimata, 2002). HA levels are maintained through

regulation of the enzymatic activity of HAS isozymes

(HAS1, HAS2 and HAS3) and degradation, by HYALs

(Stern, 2003).

The biological effects of HA are size-dependent, are

mediated through cell surface receptors including receptor

for HA-mediated motility (RHAMM) and CD44 (Slevin

et al., 1998, 2002, 2004) and may result in (i) modulation

of endothelial cell (EC) function and angiogenesis (West

et al., 1985, 1995; West and Kumar, 1989; Slevin et al.,

1998, 2002, 2004); (ii) production of an inflammatory

response via macrophage activation after injury (Noble,

2002); and (iii) neuronal differentiation, extension, motility

and synaptic signalling (Nagy et al., 1995; Lynn et al., 2001).

Moreover, RHAMM is expressed by developing and adult

neurons and oligodendrocytes, and has a role in astrocyte

motility, neurite migration and axonal growth (Lynn et al.,

2001). Tumour necrosis factor-stimulated gene 6 (TSG-6) is

upregulated in many cell types in response to proinflamma-

tory stimulation and growth factors (Milner and Day, 2003).

It is activated during blood vessel injury and can modulate

cell proliferation and, interestingly, CD44 expression, where

it increases the binding affinity of HA (Lesley et al., 2002).

We hypothesize that HA turnover increases in stroke

tissue, and our aim in the present study was to investigate

the serum and tissue levels of HA and associated enzymes

together with the tissue expression of HA receptors CD44,

RHAMM and TSG-6 in human subjects after ischaemic

stroke (IS).

Material and methodsPatient detailsThis study was approved by the local Ethical Committee, and

consent was obtained according to the Declaration of Helsinki.

Serial blood samples were collected from indwelling intravenous

catheters on admission (Day 1) and on Days 3, 7 and 14, from

patients aged between 50 and 93 years, admitted with IS (n = 54; 27

male and 27 female). A full clinical examination was carried out on

admission, and the patients were scored according to the 58-point

Scandinavian Stroke Scale (SSS) (1985) (Slevin et al., 2000). Lesions

were evaluated by CT or MRI and classified as being large infarct

(LI; infarct > 4 cm; SSS < 30, n = 14), moderate infarct (MI; >1.5 cm

and <4 cm; SSS > 30, n = 8) or small infarct (SI; <1.5 cm; SSS > 40,

n = 32). Routine blood parameters including peripheral leucocy-

tosis, cholesterol, fibrinogen, urea and glucose levels were also

determined on admission. Excluded from the study were patients

with previous transient ischaemic attack or stroke; recent history of

head trauma; major cardiac, renal, hepatic, or cancerous disease;

and obvious signs of infection after admission.

A set of healthy, age-matched control serums (n = 24) from

patients who had no recent infections or history of serious illness

or recent head trauma, and who were subsequently shown to be

disease-free, were obtained from the National Blood Service,

Manchester, UK.

Tissue samples were obtained from nine patients who died 3–37

days after stroke following middle cerebral artery occlusion

(Table 1). Immediately after death the bodies were routinely refri-

gerated at 4�C and tissue was collected within 8 h. Tissue samples

were taken from grey and white matter and from infarcted and

peri-infarcted areas. The peri-infarcted region was defined as the

area of tissue adjacent to the ischaemic core, which has previously

been found to exhibit stroke-related changes including neuronal

apoptosis and angiogenesis and was characterized by tissue oedema

and discolouration, morphology of the neurons and maintenance

of structural integrity (Krupinski et al., 2000). Sections were

stained with 2,3,5-triphenyltetrazolium chloride, which stains active

mitochondria pink; negatively stained areas represent the stroke

area. The contralateral hemisphere served as a control from

which tissue samples were removed at the same time. Tissue was

immediately frozen in liquid nitrogen, and following dissection,

Table 1 Primer sequences used in PCR

HAS1-F 50-GCG GGC TTG TCA GAG CTA-30

HAS1-R 50-AGA GCG ACA GAA GCA CCA-30

HAS2-F 50-GTG ATG ACA GGC ATC TCA-30

HAS2-R 50-GCG GGA AGT AAA CTC GA-30

HAS3-F 50-CAG CCT GCA CCA TCG A-30

HAS3-R 50-AGA GGT GGT GCT TAT GGA-30

HYAL1-F 50-CCA GGA TTT CTG GTG GAA GA-30

HYAL1-R 50-CCA CAG GAC TTG CTC TAG-30

HYAL2-F 50-CCT CTG GGG CTT CTA CCT CT-30

HYAL2-R 50-CTG AAC ACG GAA GCT CAC AA-30

HYAL3-F 50-TCA GCT GCC ACT GTT ACT GG-30

HYAL3-R 50-TGA CTC TCC CTG GAA CTG CT-30

CD44-F 50-GTG GGC CAA CAA AGA ACA CT-30

CD44-R 50-TGG AGC AGG CCC AAA TAT AG-30

RHAMM-F 50-GAA AGG GAA GAAGGC TGA AC-30

RHAMM-R 50-TGC CAA AAT CTG ATG CTG AA -30

TSG-6-F 50-TCA CAT TTC AGC CAC TGC TC-30

TSG-6-R 50-AGA CCG TGC TTC TCT GTG GT-30

Hyaluronan production following stroke Brain (2006), 129, 2158–2176 2159

at �70�C. A sample of each tissue was processed for histology and

stained with haematoxylin and eosin to determine tissue

morphology. The value of post-mortem samples in studies invol-

ving measurement of RNA and protein expression has been

previously identified (Castennson et al., 2000; Love et al., 2003;

Mitsios et al., 2006).

Measurement of HA in serum and tissueThe method was based on that of West et al. (1995) with minor

modifications. Serum and tissue samples were digested with

pronase E [0.1 cm3; 1 mg/cm3 in phosphate-buffered saline

(PBS)], boiled for 30 min and centrifuged (3000 g, 20 min). The

supernatant was divided into two, and one sample was used to

determine total HA concentration. The second sample was centri-

fuged through a Sephacryl S100 Filtron column (6000 g, 30 min)

with a 50 kDa cut-off and oligosaccharides of hyaluronan (o-HA)

(<50 kDa) were collected.

HA concentration was measured by enzyme-linked immuno-

sorbent assay (ELISA) (West et al., 1997). A 96-well ELISA plate

was coated overnight at room temperature (RT) with 200 ml of HA

solution (100 mg/cm3), and after washing in PBS the wells were

blocked with bovine serum albumin (BSA) (2% in PBS containing

0.05% Tween-20; PBST) for 2 h, and washed again with PBS.

Duplicate 100 ml aliquots of samples and standards were applied

and 100 ml of biotinylated-HA binding protein (B-HABP; 1 mg/cm3

in PBST, prepared and characterized as described elsewhere; West

et al., 1997) was added to all wells except the blank. Plates were

sealed and incubated overnight at RT, washed in PBS (·4) and

incubated with alkaline phosphatase–streptavidin (diluted 1 : 3000

in PBST containing 1% w/v BSA). Plates were washed (·4) with

PBST and developed with p-nitrophenylphosphate (1 mg/cm3 in

100 mM diethanolamine, pH 9.8, containing 2 mM CaCl2).

Absorbance was read at 405 nm using an ELISA plate reader

with Titresoft software.

Measurement of serum HYALThe method of Natowicz et al. (1996) was used. Briefly, serum

(10 ml) was incubated with 250 ml buffered substrate solution

(0.1 M sodium formate, 0.1 M NaCl, 250 mg/l HA and 1.5 mM

saccharic acid 1,4-lactone; 4 h at 37�C). The reaction was termi-

nated by the addition of 50 ml of borate buffer, the samples

were heated for 20 min at 37�C and the absorbance was read at

585 nm. A standard curve was prepared with concentrations of

N-acetylglucosamine from 0 to 1 mM. One unit of hyaluronidase

activity was defined as the amount of enzyme producing 1 mmol of

N-acetylglucosamine per minute at 37�C.

RNA isolationRNA extraction was performed using standard Clontech procedures

(BD Biosciences, Hertfordshire, UK). RNA quality was measured

using a spectrophotometer.

Reverse transcription–polymerase chainreaction (RT–PCR)Gene expression was examined by RT–PCR, in samples of contral-

ateral, peri-infarcted and infarcted tissue, using primers designed

against the gene of interest. PCR conditions were as follows: an

initial denaturation step of 10 min at 95�C, followed by 35 cycles

of 1 min at 94�C, 1 min at 60�C and 1 min at 72�C, and a final

extension step of 10 min at 72�C. The products were visualized

following agarose gel electrophoresis (1.5%) and DNA stained with

ethidium bromide (10 mg/ml). Samples without cDNA were used

as negative controls.

All experiments were carried out twice and at 20, 30 and 40

cycles to ensure equal loading of cDNA and proportional cycling.

Results for 30 cycles are shown. S16 housekeeping gene was used to

normalize the results (forward: 50-GGC AGA CCG AGA TGA ATC

CTC A-30 and reverse: 50-CAG GTC CAG GGG TCT TGG TCC-30)Sense and antisense oligonucleotide primers containing 27 nt

based on previously reported mRNA sequences in the GenBank

depository were designed with the aid of the Primer3 Output

Program (Version 0.2). Invitrogen plc. (Paisley, UK) synthesized

the primer sets.

Western blottingProteins were extracted from tissue samples (�50 mg) and 20 mg

from each sample (determined by Biorad protein estimation)

was separated by SDS–PAGE gel electrophoresis as described

elsewhere (Slevin et al., 1998). Proteins were electroblotted onto

nitrocellulose filters (Hoefer, San Francisco, CA, USA), blocked

overnight with 1% BSA in Tris-buffered saline-Tween (TBS-

Tween) and incubated for 4 h at RT with fully characterized

antibodies to the following proteins: rabbit polyclonal HAS1 and

HAS2 (gifts of Dr P. Heldin, Ludwig Institute for Cancer Research,

Uppsala, Sweden; Usui et al., 2000), mouse monoclonal HYAL1

(a gift from Dr B. Triggs-Raine, Department of Biochemistry

and Medical Genetics, University of Manitoba, Canada;

Shuttleworth et al., 2002), rabbit polyclonal HYAL2 (a gift from

Professor R. Stern, Department of Pathology, University of

California, San Francisco, USA; Bourguignon et al., 2004), rabbit

polyclonal RHAMM (a gift from Dr E. Turley, Department of

Physiology, University of Manitoba, Winnipeg, Canada; Lynn

et al., 2001), rat monoclonal CD44 (which binds the membrane

N-terminal portion of the receptor; Calbiochem, CA, USA)

and mouse monoclonal TSG-6 (a gift from Dr Katalin Mikecz,

Department of Orthopedic Surgery, Rush University Medical

Centre, Chicago, USA; Lesley et al., 2002). All antibodies were

used at a dilution of 1 : 1000 except anti-CD44 (1 : 500).

Filters were washed in TBS-Tween before staining with anti-

bodies to peroxidase-conjugated secondary antibody. Proteins

were detected using the ECL western blotting detection system

(Amersham, UK). The intensity of staining relative to a-actin

was measured with an LKB densitometer. Results are semi-

quantitative and presented as a fold change compared with the

contralateral tissue, which was given an arbitrary value of 1.0.

All experiments were performed at least twice and a representative

example is shown. If a protein was not detected in the control,

the weakest band on the gel, visible by eye, was assigned a value of

1.0. Specificity of the antibodies was confirmed by measuring

expression in human tissue cell lysates (HeLa, T98G and Jurkat

cells) and bands of the expected mass (HYAL1, 57 kDa; HYAL2,

60 kDa; HAS1, 75 kDa; HAS2, 65 kDa; RHAMM, 55 kDa; CD44,

80 KDa; and TSG-6, 36 kDa) were observed.

Affinity-histochemical staining of HA usingbiotinylated HABPSections were deparaffinized in xylene, rehydrated, washed in PBS,

pH 7.4, and blocked in 1% H2O2 for 5 min. Sections were incubated

2160 Brain (2006), 129, 2158–2176 A. Al’Qteishat et al.

with 1% BSA for 30 min and then overnight with biotinylated

HABP (3 mg/cm3 in 1% BSA in PBS; a gift from the Seikagaku

Corporation, Tokyo, Japan). Sections were washed in PBS, treated

with avidin–biotin peroxidase complex (ABC, CA, USA; 1 : 200

dilution, 1 h at RT) and then incubated with 3,3-diaminobenzidine

(DAB, Sigma, Poole, UK; 0.5% w/v) containing 0.03% H2O2 for

5 min. Slides were counterstained with haematoxylin and then

dehydrated and mounted in DPX. Positive staining was indicated

by the presence of a characteristic brown end-product at the site

of the target antigen. Controls pre-treated with Streptomyces hya-

luronidase (100 turbidity reducing units/cm3 in 50 mM sodium

acetate, pH 5.0, for 3 h) in the presence of protease inhibitors

were prepared in parallel.

ImmunohistochemistrySections were deparaffinized and rehydrated as above. Sections

were boiled for 1 min in concentrated citric acid, pH 6.0, to

unmask the antigens. Slides were then incubated in 0.5% H2O2

in methanol for 10 min followed by addition of the primary anti-

bodies HAS1/2, HYAL1/2 and RHAMM (1 : 50) or CD44 (1 : 25) in

Tris–HCl, pH 7.2, for 4 h. Sections were then incubated with

streptavidin-peroxidase-labelled polymer for 30 min and the colour

developed by exposure to Vector DAB solution for 3–20 min.

For double labelling, sections were subsequently stained with

rabbit polyclonal antibodies to b-tubulin (1 : 50; Autogen Bioclear,

Wiltshire, UK) or glial fibrillary acidic protein (GFAP; 1 : 50; Sigma

Chemicals, Dorset, UK), recognizing neurons and glia, respectively,

and developed with Vector blue. Tissue was counterstained with

haematoxylin (except in double-labelled sections). Slides were

dehydrated and mounted in DPX. Positive staining was indicated

by the presence of a characteristic brown end-product at the

site of the target antigen. Control slides in which the primary

antibody was replaced with PBS were run in parallel (data not

included).

Statistical analysis of serum HA/HYAlStatistical analysis was based on the assumption that the data were

not normally distributed (Slevin et al., 2000) and was performed

using non-parametric tests for paired groups (Wilcoxon) and

unpaired groups (Mann–Whitney U-test). The analysis was two-

tailed unless otherwise specified. Data are expressed as means 6

standard deviation. Values were considered statistically different

when P < 0.05. The relationships between o-HA, HA, other pro-

teins and all measured blood parameters were determined by linear

regression analysis (Spearman’s p). The SPSS statistical package was

used for all analyses.

ResultsHA and o-HA concentrations are increasedin the serum after strokeThe concentration of HA was significantly increased in

serum at all time points after IS (P < 0.01 in all cases;

Fig. 1A, i). The highest measured levels occurred 7 days

after stroke (145.6 6 97 ng/cm3). Mean HA level in the

age-matched controls (n = 24) was 21.6 6 19.1 ng/cm3.

The concentration of o-HA was significantly increased

in serum at all time points after IS (P < 0.01 in all cases;

Fig. 1B, i). Highest levels were 7 days after stroke (54.0 6

47.6 ng/cm3). The mean o-HA concentration in the control

group was 7.4 6 4.6 ng/cm3.

HYAL activity is increased in the serumafter strokeTotal serum HYAL activity was significantly increased in

IS patients at all time points (P < 0.01 in all cases), reaching

a maximum of 4.3 6 2.5 mM/cm3/min 3 days after stroke

(P < 0.05 in all cases; Fig. 1C, i). The mean activity in the

control group was 1.8 6 1.0 mM/cm3/min. HYAL activity

was still elevated compared with controls, after 14 days.

Correlation studiesThere was no relationship between patient HA or o-HA

serum levels and blood parameters (markers of inflamma-

tion), including peripheral leucocyte number (Fig. 1; A–C, ii)

and fibrinogen concentration (Fig. 1 A–C, iii). HYAL activity

correlated with leucocyte number after 3 days (Fig. 1C, iii;

P = 0.011). The expression of HA also correlated with FGF-2

expression in these patients (Fig. 1D; P = 0.014). There was

no significant difference in hyaluronidase activity between

patients with large or small vessel disease or in association

with clinical recovery based on SSS measurement.

HA and o-HA are increased instroke-affected tissueThe production of HA and o-HA was investigated in

infarcted, peri-infarcted and normal-looking contralateral

tissue samples in three groups containing three samples

surviving 2–6 days, 10–17 days and 26–37 days after stroke,

Fig. 1 Serum expression of HA and HYAL after stroke. (A) (i) Serum levels of HA in patients 1–14 days after IS. Data are expressed asmean 6 standard deviation. *P < 0.01 compared with the control. (ii) Negative correlation with peripheral leucocyte concentration and(iii) circulating fibrinogen (data show analysis 3 days after stroke; Spearman’s P > 0.05). (B) Serum levels of o-HA in patients 1–14 days afterIS. Data are expressed as mean 6 standard deviation. *P < 0.01 compared with the control. (ii) Negative correlation with peripheralleucocyte concentration and (iii) circulating fibrinogen (data show analysis 3 days after stroke; Spearman’s P > 0.05). (C) Serum levels ofHYAL in patients 1–14 days after IS. Data are expressed as mean 6 standard deviation. *P < 0.01 compared with the control. (ii) Positivecorrelation with peripheral leucocyte concentration after 3 days (P = 0.011) but not with (iii) circulating fibrinogen (data show analysis 3 daysafter stroke; Spearman’s P > 0.05). (D) Positive correlation of FGF-2 expression with o-HA in the serum of patients 3 days after IS(P = 0.014; Spearman’s p). (E) Tissue levels of HA in patients at different times after IS. Data are expressed as mean 6 standard deviation.*P < 0.01 compared with the control. (F) Tissue levels of o-HA in patients at different times after IS. Data are expressed as mean6 standarddeviation. *P < 0.01 compared with the control.


which approximately correspond to the periods of

inflammation, angiogenesis and revascularization and tissue

remodelling after stroke. Two to six days after stroke, the

mean HA concentration was 17.6 6 2.0 mg/g in the con-

tralateral tissue, 50.0 6 3.8 mg/g in the infarcted core and

45.6 6 1.7 mg/g in the peri-infarcted area. HA levels

remained increased in the 9–16 days survival group and

reached a maximum of 65.3 6 1.8 mg/g in stroke tissue

and 55.1 6 4mg/g in peri-infarcted tissue, compared with

16.0 6 1.4 in the control, in the longest surviving group

(Fig. 1D). At all time points there was a significant increase

in HA deposition in stroke and peri-infarcted tissue (P < 0.05

in all cases).

There was a significant elevation of o-HA (P < 0.05 in all

cases) in tissue from the same time points, with the

maximum values occurring after 26–37 days. The values

were 52.6 6 2 mg/g in stroke tissue, 40.6 6 2.5 mg/g in

peri-infarcted tissue and 8.1 6 1.4 mg/g in the contralateral

tissue (Fig. 1E).

HA is increased in peri-infarcted neuronsand blood vessels after ischaemic strokeThe localization of HA in affected brain tissue from selected

patients listed in Table 2 was investigated. Patchy staining

for HA was evident in grey matter from the contralateral

tissue and was more pronounced around blood vessels

(Fig. 2A, Patient 20). Increased staining was seen in infarcted

and peri-infarcted grey matter 3 days after stroke, and was

particularly noticeable in the microvessels in this region

(Fig. 2B, Patient 20). Strong staining of microvessels

together with nuclear staining of neurons near to the

infarcted core was also observed in peri-infarcted regions,

in a patient who survived for 5 days after stroke (Fig. 2C

and D; Patients 13 and 20, respectively). Increased staining

persisted in the grey matter up to 37 days after stroke.

Staining was abolished following pre-treatment of a control

section by hyaluronidase (Fig. 2E, Patient 13).

HYAL expression is increased in theinflammatory phase of strokeRT–PCR and western blotting showed upregulation of

HYAL1 gene and protein expression, respectively, in the

majority of patients in both peri-infarcted and infarcted

regions (Fig. 3A and B). Western blotting demonstrated

increased expression of HYAL1 protein in infarcted grey

matter in all nine patients examined compared with the

contralateral control tissue (range: 2.4- to 10.0-fold). The

increase was more modest in peri-infarcted tissue (4 out

of 9 patients; range: 2.0- to 7.5-fold). In white matter,

there was an increase in 6 out of 9 samples in peri-infarcted

tissue (range: 1.8- to 10.2-fold) and in all samples from the

stroke tissue (range: 1.5- to 4.2-fold). All the changes are

reported in Table 3. In HYAL1-stained western blots, a weak

band was sometimes observed at around 50 kDa. This may

represent a proteolytic cleavage fragment, which could help

release it from its GPI anchor (Stern, 2003), or most likely,

could be the result of a slight amount of tissue degeneration

in our post-mortem tissue samples, since it was expressed in

equal proportions in each sample, and our unpublished data

demonstrated staining of only one band in fresh rat MCAO

tissue (data submitted elsewhere). Immunohistochemistry

showed that HYAL1 was expressed intracellularly, although

only weakly, in normal-looking neurons in all layers in the

contralateral tissue samples. There was weak but consistent

staining in non-neuronal cells, possibly perineuronal oligo-

dendrocytes (Fig. 3C, i; Patient 20). Astrocytes showed an

absence of detectable labelling (Fig. 3C, ii). There was no

observable staining in the white matter. In peri-infarcted grey

matter HYAL1 staining was increased both intracellularly

and in the axons of abnormal-looking neurons 3–6 days

after stroke (Fig. 3D and E) as well as in microglia and

mononuclear inflammatory cells (Fig. 3F, Patient 20). At

later time points (10–37 days after stroke) HYAL1 expression

was still observed in inflammatory peri-infarcted and

infarcted regions.

Table 2 Clinical details of patients with IS

Patient* Age Sex Survivaltime (days)

Diabetes Hypercholesterolaemia† Stroke type(Oxfordshire classification)

Cause of death§

20 84 M 3 No Yes TACI Malignant stroke13 71 M 5 No No TACI Brain oedema14 84 M 6 No No PACI Cardiac failure15 84 M 10 No Yes PACI Respiratory infection8¶ 77 M 15 No Yes PACI Heart attack24 58 M 17 No Yes PACI Bronchial aspiration22 73 M 26 No No TACI Heart attack23 84 F 29 Yes No PACI Bronchial aspiration12 75 M 29 No Yes PACI Bronchial aspiration21 69 F 37 yes No PACI Pulmonary embolism

*Prior to first stroke, patients were not on anti-platelet or statin treatment; †serum cholesterol concentration > 5.5 moll�1; PACI,partial anterior circulation infarct; TACI, total anterior circulation infarct; LACI, lacunar infarct; §according to the hospital death certificateand clinical history, the main cause of death was extensive stroke. The intermediate causes of death are listed in the Table; ¶Patient 8was used only in the HA tissue analysis study.


Fig. 2 Distribution of HA in the stroked brain. (A) Normal grey matter (Patient 20) showing weak staining for HA around blood vessels[·100; insert: ·200, arrows; (i) blood vessel; (ii) neurons]. (B) The boundary zone between infarcted and peri-infarcted grey matter(Patient 20) showing more intense staining in the peri-infarcted region and in infarcted blood vessels (·40; insert: ·200, arrows).(C) Infarcted grey matter showing strong staining around blood vessels in Patient 13 (·100; insert: ·200, arrows). (D) Infarcted grey matterfrom Patient 20 showing increased nuclear and intracellular staining of neurons (·100; insert: ·200, arrows). (E) Staining was abolishedfollowing pre-treatment with hyaluronidase (Patient 13). Staining was carried out with biotinylated-HA and sections were developed usingavidin–biotin peroxidase.


Fig. 3 HYAL1 expression after IS. (A) RT–PCR showing increased gene expression of HYAL1 in peri-infarcted and infarcted samples.Example shows the grey matter from Patient 20 who survived for 3 days after stroke. Lane 1, contralateral hemisphere; Lane 2,peri-infarcted region; Lane 3, infarcted region; Lane 4, positive control; Lane 5, external control; and Lane 6, negative control (-cDNA).(B) Western blots of HYAL1 expression in Patients 20 and 23. (i) Antibody specificity was demonstrated using HeLa cell lysate. (ii) Increasedexpression was seen in peri-infarcted and infarcted regions. Lanes 1–3 are grey matter and Lanes 4–6 are white matter; C, control; PI,peri-infarcted; and I, infarcted region. a-Actin loading gels are also shown. Each experiment was carried out at least twice and arepresentative example is shown. (iii) Bar charts show relative protein expression compared with the control, given an arbitrary value of 1.0.(C–F) Immunohistochemical analysis of HYAL1 distribution in stroke tissue. (C) Normal-looking grey matter showing weak intracellularstaining in neurons (i, arrow) and possibly oligodendrocytes (i, broken arrow), but an absence of detectable labelling of astrocytes indouble-labelled sections [ii, arrow; DAB for GFAP and Vector blue for HYAL1; broken arrow shows a weakly labelled neuron(·100; inserts: · 200)]. (D–E) Strong intracellular and axonal staining of abnormal-looking neurons from the peri-infarcted region ofPatient 20 (·100; insert: · 200). (F) General positive staining of infiltrating inflammatory cells and microglia in peri-infarcted grey mattertissue from Patient 24 (·100; insert: · 200).


HYAL2 was increased in the majority of patients mainly in

grey matter infarcted regions (2 out of 9 in peri-infarcted

tissue, 2.5- to 8.5-fold; 5 out of 9 in stroke tissue, 1.6- to

9.4-fold). Representative RT–PCR results and western blots

are shown (Fig. 4A and B). Weak intracellular HYAL2

staining was observed mainly in neurons in normal-looking

grey matter (Fig. 4C), but extensive intracellular and

nuclear neuronal staining as well as an increase in positively

stained mononuclear inflammatory cells was seen in both

peri-infarcted and infarcted tissue in patients 3–10 days

after stroke (Fig. 4D and E). In infarcted grey matter,

staining for HYAL 2 was also associated with microvessels

after 3 days, and this pattern of staining persisted in

patients surviving for 25 days (Fig. 4E).

HAS expression is increased in neuronsand inflammatory cells after strokeHAS1 was upregulated in 3 out of 9 samples (range: 1.5- to

10.8-fold) in grey matter peri-infarcted tissue and 3 out of

9 samples in grey matter stroke tissue (range: 1.5- to 12.6-

fold). In white matter, 3 out of 9 samples were upregulated in

peri-infarcted tissue (range: 2.1- to 2.4-fold) and the same

number in stroke tissue (range: 1.5- to 6.6-fold). Represen-

tative RT–PCR and western blots are shown (Fig. 5A and B).

Immunohistochemistry demonstrated that HAS1 was weakly

expressed intracellularly in occasional normal-looking neu-

rons. In peri-infarcted and infarcted grey matter, HAS1

intracellular clumping occurred in neurons and increased

expression was also observed in mononuclear inflammatory

cells and cells with the appearance of oligodendroglia in

patients who survived from 3 to 29 days after stroke

(Fig. 5D and E).

HAS2 was upregulated to a greater extent in grey matter

(5 out of 9 in infarcted tissue, range: 1.6- to 9.5-fold; 6 out

of 9 in peri-infarcted tissue, range: 1.8- to 7.1-fold) than

white matter (3 out of 9 in stroke tissue, range: 2.0- to

5.0-fold, and 3 out of 9 in peri-infarcted tissue, range:

1.7- to 6.0-fold). Representative RT–PCR and western

blots are shown (Fig. 6A and B). Several smaller very weak

bands were observed in some tissue samples stained by

HAS2, which most likely represented degradation products

of HAS2 occurring in our post-mortem samples. There was

no observable staining of HAS2 in normal-looking grey

matter (Fig. 6C, Patient 20). In peri-infarcted grey matter,

staining for HAS2 was increased in neurons, where it was

often translocated to the nucleus (Fig. 6D and E), and in

mononuclear inflammatory cells from peri-infarcted and

infarcted regions. Double labelling was used to demonstrate

co-localization of HAS2 (brown) with neurons (b-tubulin;

blue) in peri-infarcted grey matter of Patient 20 (Fig. 6F,

i and ii).

RHAMM expression was upregulated inneurons and blood vessels after strokeWe were unable to detect the presence of RHAMM gene

expression with the primers listed. RHAMM protein was

Table 3 Relative protein expression in white matter compared with contralateral tissue (given an arbitrary value of 1.0)

Grey matter peri-infarct Grey matter infarct

S* HAS1 HAS2 HYAL1 HYAL2 RHAMM CD44 TSG-6 HAS1 HAS2 HYAL1 HYAL2 RHAMM CD44 TSG-6

20 3 10.8 17.0 0.2 1.0 12.5 5.0 1.1 22.6 25.0 4.8 1.3 11.4 7.5 3.413 5 1.0 1.0 1.0 0.4 3.0 0.5 12.5 1.0 0.4 2.4 12.0 0.5 12.0 1.014 6 1.5 1.0 1.3 25 0.4 0.4 2.0 1.0 1.6 2.4 24.9 0.3 0.4 1.315 10 1.0 1.0 4.3 0.6 3.6 11.8 11.2 1.2 0.5 5.1 0.6 4.0 12.5 2.424 17 1.0 5.0 1.1 0.9 1.5 1.0 0.3 12.5 1.0 12.5 20.0 12.0 20.5 20.322 26 1.0 2.0 1.0 0.8 5.5 5.0 1.5 9.5 10 13 13.0 20.2 12.2 12.312 29 1.0 1.8 11.5 0.5 12.4 0.7 0.8 1.5 1.0 8.3 0.5 11.8 0.0 0.123 29 1.7 8.8 2.0 2.5 2.0 0.6 3.1 1.0 12.6 9.6 9.6 3.0 11.9 6.121 37 1.0 1.0 12.1 0.0 1.2 0.7 1.5 2.0 2.0 10.0 0.5 0.5 0.6 0.3

White matter peri-infarct White matter infarct

S* HAS1 HAS2 HYAL1 HYAL2 RHAMM CD44 TSG-6 HAS1 HAS2 HYAL1 HYAL2 RHAMM CD44 TSG-6

20 3 1.0 11.0 2.2 1.5 1.7 1.0 0.2 1.0 10.0 1.5 1.0 12.4 2.0 0.613 5 2.1 1.0 0 1.0 1.7 2.0 12.0 1.0 0.3 2.0 1.0 0.8 7.5 4.514 6 1.0 1.0 1.5 1.3 0.6 0.4 1.0 0.7 0.2 2.1 1.4 0.0 1.7 0.815 10 0.3 1.2 2.1 1.0 2.2 9.7 8.9 0.0 0.6 2.0 1.3 12.1 11.8 12.124 17 1.0 1.1 1.8 1.2 2.0 1.0 1.6 1.0 0.5 4.6 1.0 7.5 2.5 3.122 26 2.4 1.0 6.7 0.6 1.5 9.0 3.5 11.6 2.0 8.0 0.8 9.5 12.6 11.612 29 2.0 10.1 10.2 1.0 10.0 0.3 3.7 1.0 10.5 9.3 0.8 7.5 0.2 5.123 29 1.0 1.7 0.7 0.7 0.7 0.6 1.6 10.2 0.7 1.9 1.9 0.5 1.2 12.221 37 0.5 0.1 0.6 0.6 0.3 0.4 2.5 1.5 1.0 2.5 0.5 1.5 0.4 1.5

*Survival time after stroke in days.


upregulated mainly in grey matter (7 out of 9 in peri-

infarcted tissue, range: 1.5- to 8.2-fold; and 6 out of 9 in

infarcted tissue, range: 3.0- to 5.6-fold) and to a lesser extent

in the white matter (5 out of 9 in peri-infarcted tissue, range:

1.5- to 12.0-fold; 5 out of 9 in infarcted tissue, range: 1.5- to

5.4-fold). Representative western blots are shown (Fig. 7A).

Weak intracellular staining of RHAMM was evident in

neurons, and some non-neuronal cells, possibly oligo-

dendrocytes, although astrocytes and blood vessels showed

an absence of detectable labelling in normal grey matter

Fig. 4 HYAL2 expression in stroked tissue. (A) RT–PCR showing increased gene expression of HYAL2 in peri-infarcted and infarctedsamples. Example shows the grey matter from Patient 23. Lane 1, contralateral hemisphere; Lane 2, peri-infarcted region; Lane 3,infarcted region; Lane 4, positive control; Lane 5, external control; and Lane 6, negative control (-cDNA). (B) Western blots of HYAL2expression in Patients 23 and 24. (i) Antibody specificity was demonstrated using HeLa cell lysate. (ii) Increased expression was seen inperi-infarcted and infarcted regions. Lanes 1–3 are grey matter and Lanes 4–6 are white matter; C, control; PI, peri-infarcted; and I, infarctedregion. a-Actin loading gels are also shown. Each experiment was carried out at least twice and a representative example is shown. (iii) Barcharts show relative protein expression compared with the control, given an arbitrary value of 1.0. (C–E) Immunohistochemical analysis ofHYAL2 distribution in stroke tissue. (C) Normal-looking grey matter showing even intracellular expression in neurons (·100; insert: · 200,arrows). (D) Increased staining intensity and clumping, together with nuclear staining in neurons in peri-infarcted regions, and (E), a generalincrease in HYAL2 positive cells including a positively stained blood vessel [from Patient 23 (·100; insert: · 200, arrows)].


(Fig. 7B, Patient 20). In peri-infarcted grey matter, there was

much stronger intracellular as well as nuclear staining of

neurons, and more prominent staining of non-neuronal

cells (Fig. 7C and D, Patient 13), whilst both the EC and

tunica media of blood vessels were strongly stained in peri-

infarcted and infarcted regions in patients surviving

from 3 to 29 days after stroke (Fig. 7E and F, respectively,

Patient 21; arrows).

Fig. 5 HAS1 expression in stroke tissue. (A) RT–PCR showing increased gene expression of HAS1 in peri-infarcted and infarcted samples.Example shows the grey matter from Patient 20. Lane 1, contralateral hemisphere; Lane 2, peri-infarcted region; Lane 3, infarcted region;Lane 4, positive control; Lane 5, external control; and Lane 6, negative control (-cDNA). (B) Western blots of HAS1 expression in Patients20 and 22. (i) Antibody specificity was demonstrated using T98G cell lysate. (ii) Increased expression was seen in peri-infarcted and infarctedregions. Lanes 1–3 are grey matter and Lanes 4–6 are white matter; C, control; PI, peri-infarcted; and I, infarcted region. a-Actin loading gelsare also shown. Each experiment was carried out at least twice and a representative example is shown. (iii) Bar charts show relative proteinexpression compared with the control, given an arbitrary value of 1.0. (C–E) Immunohistochemical analysis of HAS1 distribution in stroketissue. (C) Normal-looking grey matter showing weak intracellular staining of occasional neurons (arrow); increased expression andclumping of HAS1 in neurons (D; arrow) and a general increase in mononuclear inflammatory cell and possibly oligodendroglia expression(E, arrow) in peri-infarcted regions from Patient 22 (·100; insert: · 200).


Fig. 6 HAS2 expression in stroke tissue. (A) RT–PCR showing increased gene expression of HAS2 in peri-infarcted and infarcted samples.Example shows the grey matter from Patient 12. Lane 1, contralateral hemisphere; Lane 2, peri-infarcted region; Lane 3, infarcted region;Lane 4, positive control; Lane 5, external control; and Lane 6, negative control (-cDNA). (B) Western blots of HAS2 expression in Patients12 and 20. (i) Antibody specificity was demonstrated using T98G cell lysate. (ii) Increased expression was seen in peri-infarcted andinfarcted regions. Lanes 1–3 are grey matter and Lanes 4–6 are white matter; C, control; PI, peri-infarcted; and I, infarcted region. a-Actinloading gels are also shown. Each experiment was carried out at least twice and a representative example is shown. (iii) Bar charts showrelative protein expression compared with the control, given an arbitrary value of 1.0. (C–F) Immunohistochemical analysis of HAS2distribution in stroke tissue. (C) Normal-looking grey matter showing negative staining in neurons (·100; insert: ·200, arrows). (D and E)Strong staining of neuronal nuclei from the peri-infarcted regions of damaged tissue (Patient 20 and 24, respectively; ·100; insert: ·200).[F (i and ii)] Shows HAS2 positive (brown) peri-infarcted neurons identified by co-labelling with b-tubulin (Vector blue; ·200).


Fig. 7 RHAMM expression in stroke tissue. (A) Western blots of RHAMM expression in Patients 22 and 24. (i) Antibody specificity wasdemonstrated using Jurkat cell lysate. (ii) Increased expression was seen in peri-infarcted and infarcted regions. Lanes 1–3 are grey matterand Lanes 4–6 are white matter; C, control; PI, peri-infarcted; and I, infarcted region. a-Actin loading gels are also shown. Eachexperiment was carried out at least twice and a representative example is shown. (iii) Bar charts show relative protein expression comparedwith the control, given an arbitrary value of 1.0. (B–F) Immunohistochemical analysis of RHAMM distribution in stroke tissue.(B) Normal-looking grey matter showing weak intracellular staining in neurons (insert i, arrow) and possibly oligodendroglia (insert i,broken arrow). There was an absence of detectable labelling in astrocytes. Insert (ii) shows a GFAP positive (brown) astrocyte with nodetectable labelling of RHAMM (arrow) and a weakly positive neuron (Vector blue; broken arrow) (·100; inserts: ·200, arrows).(C) Strong intracellular staining of peri-infarcted neurons from Patient 20 and (D) in the nucleus (·100; insert: ·200, arrows). (E) IncreasedRHAMM expression in peri-infarcted microvessels (·100; insert: ·200), and (F) in the tunica media of larger blood vessels fromstroke-affected regions (·100) from Patient 24.


CD44 was upregulated in stroke tissueCD44 was upregulated in the majority of samples from

infarcted tissue (6 out of 9 in both grey and white matter,

range: 3.5- to 10.0-fold and 2.5- to 8.0-fold, respectively) but

only a minority of samples from peri-infarcted regions (3 out

of 9 from both grey and white matter, range: 5- to 8.1-fold

and 2- and 10-fold, respectively). Representative RT–PCR

and western blots are shown in Fig. 8A and B. CD44 staining

was not evident in normal contralateral white matter (data

not shown), and sporadic weak CD44 staining of neurons

occurred in normal grey matter (Fig. 8C). Strong surface

staining for CD44 was evident in dead or dying neurons

from peri-infarcted/infarcted tissue as well as in inflamma-

tory mononuclear cells 3–17 days after stroke (Fig. 8D and E;

arrows).

TSG-6 expression in stroke tissueExpression of TSG-6 was increased in the majority of stroke-

affected tissue. In white matter, 7 out of 9 samples were

increased in both peri-infarcted and infarcted tissue

(range: 1.6- to 8.8-fold and 1.5- to 4.8-fold, respectively).

In grey matter, TSG-6 increased in 6 out of 9 samples in peri-

infarcted tissue (range: 1.5- to 8.0-fold) and 5 out of 9 in

infarcted tissue (range: 2.4- to 9.0-fold). Representative RT–

PCR and western blots are shown for TSG-6 (Fig. 9A and B).

There was little TSG-6 staining in normal-looking grey or

white matter (Fig. 9C); however, increased TSG-6 staining

was associated with inflammatory mononuclear cells and

damaged neurons from infarcted regions in patients surviv-

ing from 3 to 29 days after stroke (Fig. 9D; arrows).

DiscussionUsing a HA-specific biotinylated probe we have demon-

strated increased staining for HA in stroke-affected tissue

of human brain. HA accumulation in tissue and serum

was confirmed using biochemical assays up to 37 days

after stroke. Although HA has been demonstrated to form

coats around neurons, especially in the cerebral cortex and

hypothalamus in normal human and rat brain (Yasahara

et al., 1994), the pathological brain has not been studied

in detail. The HA concentration in cerebral tissues decreases

from the foetus to the adult. In the rat embryo, HA forms an

extracellular component of the migration and prolifera-

tion areas of the cerebral cortex, suggesting an important

role in the modelling process (Delpech et al., 1989). The

concentration of HA was increased in the CSF of patients

with head injury and cerebral infarction, concomitant with

accumulation in the superficial layer of the cerebral cortex

(Laurent et al., 1996). HA was consistently increased in

tissue and serum over the period of our study, suggesting

continuous HA synthesis.

Rapid increases in circulating levels of HA have been

demonstrated in patients with burns, septicaemia and

shock, and are also increased in those with rheumatoid

arthritis and vasculitis (Mio et al., 2000). The increase in

HA serum production may have been at least partly as a result

of the systemic inflammatory effects of the acute

phase response; however, our evidence for increased expres-

sion in the damaged brain tissue, together with a lack

of correlation with markers of inflammation including

leucocyte concentration and fibrinogen, and a strong

correlation with increased FGF-2 expression, which was

also increased in brain tissue (Issa et al., 2005), suggests a

relationship to ECM changes in the brain.

The synthesis of HA by stromal cells usually accompanies

the re-establishment of an ECM during wound repair fol-

lowed at a later stage by cell migration and cell division

(Rauch, 2004). It has been suggested that HAS1 maintains

a basal level of HA synthesis, whereas HAS2 stimulates

morphogenesis and cell migration and invasion during

embryonic development (Adamia et al., 2005). In this

study, immunohistochemistry demonstrated increased

HAS1 and HAS2 staining in inflammatory cells from peri-

infarcted regions. Of particular interest was the nuclear

staining of peri-infarcted neurons by HAS2, in combination

with HA nuclear staining. HA is normally synthesized by

HASs on the inner surface of the plasma membrane and

is extruded across the cell membrane into the ECM (Itano

and Kimata, 2002). In vitro data have suggested that HAS1

and HAS2 produce high molecular weight HA (RMM

3.9 · 106 Da), and HAS3 a lower molecular weight form

(RMM 0.12–1 · 106 Da) (Jonas and Paraskevi, 1999).

Although antibodies to HAS3 are not available, analysis of

gene expression by RT–PCR showed no changes in stroke-

affected areas (data not included).

To our knowledge, no previous studies have identified

HAS expression in cell nuclei; however, HA has been

observed in rat neuronal nuclei and nucleoli as well as in

other cells (Ripellino et al., 1988), and appears to be involved

in chromatin condensation and cell mitosis (Hascall et al.,

2004). Our results suggest a preferential synthesis of high

molecular weight HA in stroke tissue, perhaps in an attempt

to re-establish the ECM that accompanies tissue remodell-

ing after IS. HA from the matrix may be taken up by

receptor-mediated mechanisms in neurons or may represent

newly synthesized material (Stern, 2003). Increased HA

expression enhanced peripheral nerve regeneration and

axon myelination in sciatic nerve-transected rats, suggesting

a protective role after stroke (Wang et al., 1998). Recently,

the structural and cytochemical characteristics as well as the

molecular composition of ECM perineuronal nets, rich in

proteoglycans including HA, and surrounding the soma

and proximal parts of dendrites have been described

(Bruckner et al., 2006). These may have the ability to

stabilize the neuronal micro-environment maintaining

synaptic plasticity and the generation of action potentials.

Net-bearing neurons in the cerebellar cortex of rats

expressed HAS, which might be important in retaining

HA on the cell surface (Carulli et al., 2006). Therefore,

reduction in HA within these perineuronal nets after stroke


Fig. 8 CD44 expression after IS. (A) RT–PCR showing increased gene expression of CD44 in peri-infarcted and infarcted samples.Example shows the grey matter from Patient 20 who survived for 3 days after stroke. Lane 1, contralateral hemisphere; Lane 2,peri-infarcted region; Lane 3, infarcted region; Lane 4, positive control; Lane 5, external control; and Lane 6, negative control (-cDNA).(B) Western blots of CD44 expression in Patients 20 and 22. (i) Antibody specificity was demonstrated using Jurkat cell lysate. (ii) Increasedexpression was seen in peri-infarcted and infarcted regions. Lanes 1–3 are grey matter and Lanes 4–6 are white matter; C, control; PI,peri-infarcted; and I, infarcted region. a-Actin loading gels are also shown. Each experiment was carried out at least twice and arepresentative example is shown. (iii) Bar charts show relative protein expression compared with the control, given an arbitrary value of 1.0.(C–E) Immunohistochemical analysis of CD44 distribution in stroke tissue. (C) Normal-looking grey matter showing weak surfacestaining of neurons (insert, arrow) and some blood vessels (·100; insert: ·200). Staining of inflammatory mononuclear cells (D) and dead ordying neurons (E) in peri-infarcted regions from Patient 22 (·100; insert: ·200, arrows).


might reduce the electrical activity and stability of neurons

in these regions.

HYALs are involved in tissue remodelling during embryo-

nic development, promotion of tumour invasion and wound

healing. We have demonstrated increased expression of

HYAL in the serum of patients with ICH and IS. HYAL1

and 2 were also upregulated in stroke-affected tissue, sug-

gesting that breakdown of high molecular weight HA is a

feature of stroke pathophysiology. HYAL2, a GPI-anchored

membrane protein, generates a 20-kDa HA fragment

internalized by receptor-mediated endocytosis and delivered

to lysosomes, where further degradation by HYAL1 to

potentially angiogenic oligosaccharides of HA occurs

(Slevin et al., 2002; Stern, 2003). HYAL1 is a universally

distributed lysosomal enzyme, but although HYAL2 is also

present in many tissues it is not normally found in the

adult brain where the gene is inactivated by hypermethy-

lation (Lepperdinger et al., 2001). Mouse astrocytoma cells

over-expressing HYAL2 become invasive and form highly

vascularized tumours (Novak et al., 1999). Our results sug-

gest that the HYAL2 gene may be re-activated after stroke.

HYAL1 was expressed by non-neuronal cells, possibly

oligodendrocytes in normal-looking grey matter tissue.

Degradation of HA following spinal cord injury in rats

Fig. 9 TSG-6 expression after IS. (A) RT–PCR showing increased gene expression of TSG-6 in peri-infarcted and infarcted samples. Exampleshows the grey matter from Patient 24. Lane 1, contralateral hemisphere; Lane 2, peri-infarcted region; Lane 3, infarcted region; Lane 4,positive control; Lane 5, external control; and Lane 6, negative control (-cDNA). (B) Western blots of TSG-6 expression in Patients 13 and24. (i) Antibody specificity was demonstrated using Jurkat cell lysate. (ii) Increased expression was seen in peri-infarcted and infarctedregions. Lanes 1–3 are grey matter and Lanes 4–6 are white matter; C, control; PI, peri-infarcted; and I, infarcted region. a-Actin loading gelsare also shown. Each experiment was carried out at least twice and a representative example is shown. (iii) Bar charts show relativeprotein expression compared with the control, given an arbitrary value of 1.0. (C and D) Immunohistochemical analysis of TSG-6distribution in stroke tissue. (C) Normal-looking grey matter with no TSG-6 detectable labelling of neurons (·100; insert: ·200, arrow).(D) Increased staining of inflammatory mononuclear cells and abnormal-looking neurons (arrow) in infarcted regions from Patient 13(·100; insert: ·200, arrow).


increased astrocyte proliferation, and addition of HYAL

to quiescent astrocyte cultures induced proliferation, sug-

gesting potential involvement in gliosis (Struve et al.,

2005). Oligodendrocytes could be a potential source of

HYAL secretion after stroke. Increased expression of

HYAL1 and 2 was observed in infiltrating mononuclear

inflammatory cells, whilst HYAL1 was also seen intra-

cellularly in peri-infarcted neurons, perhaps indicating the

possibility of intracellular signalling pathway activation via

o-HA in ischaemic cells.

HA signalling through RHAMM and CD44 receptors

mediates EC and neuronal migration and angiogenesis

(Savani et al., 2001; Slevin et al., 2002). RHAMM is expressed

by normal neurons, where it regulates neurite migration and

axon growth, and also by oligodendrocytes and astrocytes,

where it affects motility and proliferation (Nagy et al., 1995;

Lynne et al., 2001). Binding of o-HA to RHAMM in human

EC cell lines in vitro activated MAP kinase (ERK1/2) and

initiated mitogenesis (Savani et al., 2001). We were unable to

demonstrate RHAMM gene expression in any of our post-

mortem samples using several different primer designs.

Other tissues and positive controls showed RHAMM gene

product using the same primers; therefore, we presume that

the gene product must have been substantially degraded in

our tissues. The general measurement of gene expression in

post-mortem samples has been demonstrated to be repre-

sentative and proportional to that of fresh samples (Mitsios

et al., 2006 and references therein), and so this result was

surprising. RHAMM protein was expressed by some non-

neuronal cells, possibly oligodendrocytes in normal-looking

grey matter tissue. RHAMM expression was increased in

peri-infarcted and infarcted tissue of stroke patients and

was expressed both intracellularly and in the nuclei of

peri-infarcted neurons, as well as in the intima and EC of

microvessels. Increased RHAMM expression might enhance

calmodulin-mediated signalling to cytoskeletal elements in

neurons (Lynn et al., 2001), but has also been observed in the

mitotic spindles of smooth muscle cell nuclei, suggesting a

role in mitosis (Hascall et al., 2004). It could also be involved

in stimulation of angiogenesis, improving collateral circula-

tion in ischaemic regions after stroke.

CD44 is involved in multiple responses to inflammation

including leucocyte recruitment, cell–matrix interactions and

matrix remodelling (Pure and Cuff, 2001). CD44 expression

is regulated by HA fragments, interleukin-1b and tumour

necrosis factor-alpha (TNF-a), all of which are increased

in stroke (Slevin et al., 2005). Previous studies have

shown that CD44-deficient mice were protected against

cerebral ischaemia injury, in association with reduced

expression of interleukin-1-beta (Wang et al., 2002). In

this study, CD44 was upregulated in stroke tissue, mainly

associated with inflammatory cells. The concurrent induc-

tion of CD44 and HA in the ischaemic area may potentiate

the inflammatory effects of hypoxia. We have shown

increased neuronal expression in larger neurons, whilst

increased CD44 has been previously demonstrated in

growth factor-treated ischaemic rat cortical cultures, which

might impact upon survival via modulation of growth

factor-induced signalling pathways (Yoshida et al., 2004).

TSG-6 expression is also associated with inflammation

and tissue remodelling (Milner et al., 2003; Mahoney et al.,

2005), and modulates the interaction between CD44

and HA (Lesley et al., 2004). We have shown increased

expression of TSG-6 in the majority of tissue samples

and localized it mainly to inflammatory cells. Since tissue

remodelling is a known feature of most sites of TSG-6

expression, it may have the same function in stroke tissue

(Mahoney et al., 2005).

In summary, the results presented here show a significant

change in the enzymes responsible for HA synthesis and

degradation together with upregulation of HA receptors.

This may reflect tissue remodelling after stroke and be par-

tially responsible for increased angiogenesis and neuronal

migration. One limitation of this study was that we were

unable to determine the exact size of the HA molecules

expressed in the infarcted regions of stroke tissue and in

the serum. Antibodies recognizing HA of specific molecular

weight are not available and other methods of isolation and

analysis are not yet sensitive enough to detect such small

quantities. Owing to the small numbers of patient tissue

samples examined, it was not possible to carry out statistical

analysis of protein concentration in relation to clinical para-

meters, and so these data are mostly qualitative rather

than quantitative analysis. Within the nine samples studied,

there was a notable variation in expression of investigated

proteins, partly because the original dissected material, for

example, peri-infarcted region, was subsequently shown to

contain some areas of lesser damaged tissue at the peri-

meters, as well as infarcted zones closest to the point of

stroke origin. However, the greater sensitivity and ability

of IHC to cellular sub-type has allowed us to identify the

particular zonal expression of our test proteins.

One study of recovery from cortical injury used HA- and

laminin-impregnated gels to demonstrate HA scaffolding

into which collateral blood vessels grew, resulting in inhibi-

tion of glial scarring, suggesting that its manipulation could

be beneficial to the recovery after stroke (Hou et al., 2005).

It will be important to identify the mechanisms through

which this occurs and the effects on cell survival and growth.

Further studies should attempt to identify potential thera-

peutic benefits of HA/receptor manipulation after stroke.

References

Adamia S, Maxwell CA, Pilarski LM. Hyaluronan and hyaluronan synthases:

potential therapeutic targets in cancer. Curr Drug Targets Cardiovasc

Haematol Disord 2005; 5: 3–14.

Bourguignon LY, Singleton PA, Diedrich F, Stern R, Gilad E. CD44

interaction with Na+- H+ exchanger (NHEI) creates acidic micro-

environments leading to hyaluronidase-2 and cathepsin B activation and

breast tumour cell invasion. J Biol Chem 2004; 278: 26991–7007.

Bruckner G, Szeoke S, Pavlica S, Grosche J, Kacza J. Axon initial segment

ensheathed by extracellular matrix perineuronal nets. Neuroscience 2006;

138: 365–75.


Carulli D, Rhodes KE, Brown DJ, Bonnert TP, Pollack SJ, Oliver K, et al.

Composition of perineuronal nets in the adult rat cerebellum and the

cellular origin of their components. J Comp Neurol 2006; 494: 559–77.

Castensson A, Emilsson L, Preece P, Jazin EE. High-resolution quantification

of specific mRNA levels in human brain autopsies and biopsies. Genome

Res 2000; 10: 1219–29.

Coyle P, Heistad D. Blood flow through cerebral collateral vessels one month

after middle cerebral artery occlusion. Stroke 1987; 18: 407–11.

Delpech B, Delpech A, Bruckner G, Girard N, Maingonnat C. Hyaluronan

and hyaluronectin in the nervous system. Ciba Found Symp 1989;

143: 208–20.

Hascall VC, Majors AK, de la Motte CA, Evanko SB,Wang A, Drazba JA, et al.

Intracellular hyaluronan: a new frontier for inflammation. Biochim

Biophys Acta 2004; 1673: 3–12.

Hou S, Tian W, Xu Q, Cui F, Zhang J, Lu Q, et al. The enhancement

of cell adherence and inducement of neurite outgrowth of dorsal root

ganglia co-cultured with hyaluronic acid hydrogels modified with

Nogo-66 receptor antagonist in vitro. Neuroscience 2006; 137: 519–29.

Issa R, AlQteishat A, Mitsios N, Saka M, Krupinski J, Tarkowski E, et al.

Expression of basic fibroblast growth factor mRNA and protein in the

human brain following ischaemic stroke. Angiogenesis 2005; 8: 53-62.

Itano N, Kimata K. Mammalian hyaluronan synthases. Life 2002; 54: 195–9.

Jonas B, Paraskevi H. Expression of recombinant hyaluronan synthase (HAS)

isoforms in CHO cells reduces cell migration and cell surface CD44.

Exp Cell Res 1999; 252: 342–51.

Krupinski J, Kaluza J, Kumar P, WangM, Kumar S. Prognostic value of blood

vessel density in ischaemic stroke. Lancet 1993; 341: 925–7.

Krupinski J, Kaluza J, Kumar P, Kumar S, Wang JM. Role of angiogenesis in

patients with ischaemic cerebral stroke. Stroke 1994; 25: 1794–8.

Krupinski J, Kumar P, Kumar S, Kaluza J. Increased expression of TGF-b1 in

brain tissue after ischaemic stoke in humans. Stroke 1996; 27: 852–7.

Krupinski J, Issa R, Bujny T, Slevin M, Kumar P, Kumar S, et al. A putative

role for platelet derived growth factor in angiogenesis and neuroprotection

after ischaemic stroke in humans. Stroke 1997; 28: 564–73.

Krupinski J, Lopez E, Marti E, Ferrer F. Expression of caspases and their

substrates in the rat model of focal cerebral ischaemia. Neurobiol Dis 2000;

7: 332–42.

Krupinski J, Stroemer P, Slevin M, Marti E, Kumar P, Rubio F. Three-

dimensional structure of newly formed blood vessels after focal cerebral

ischaemia in rat. Neuroreport 2003; 14: 1171–6.

Laurent UB, Laurent TC, Hellsing LK, Persson L, Hartman M, Lilja K.

Hyaluronan in human cerebrospinal fluid. Acta Neurol Scand 1996 Sep;

94:194-206.

Lepperdinger G, Muellegger J, Kreill G. Hyal-2, less active but more versatile.

Matrix Biol 2001; 20: 509–14.

Lesley J, English NM, GaI I, Kikecz K, Day AJ, Hyman R. Hyaluronan binding

properties of a CD44 chimera containing the link module of TSG-6.

J Biol Chem 2002; 277: 26600–8.

Lesley J, Gal I, Mahoney DJ, Cordell MR, Rugg MS, Hyman R, et al. TSG-6

modulates the interaction between hyaluronan and cell surface CD44.

J Biol Chem 2004; 279: 25745–54.

Lipton P. Ischaemic cell death in brain neurons. Physiol Rev 1999; 79: 431–68.

Love S. Apoptosis and brain ischaemia. Prog Neuropsychopharmacol Biol

Psychiatry 2003; 27: 267–82.

Lynn BD, Turley EA, Nagy JI. Subcellular distribution, calmodulin

interaction, and mitochondrial association of the hyaluronan-binding

protein RHAMM in rat brain. J Neurosci Res 2001; 65: 6–16.

Mahoney DJ, Mulloy B, Forster MJ, Blundell CD, Fries E, Milner CM, et al.

Characterization of the interaction between tumour necrosis factor-

stimulated gene-6 and heparin: implications for the inhibition of

plasmin in extracellular matrix microenvironments. J Biol Chem 2005;

280: 27044–55.

Milner CM, Day AJ. TSG-6 a multifunctional protein associated with

inflammation. J Cell Sci 2003; 116: 1863–73.

Mio K, Carrette O, Maibach HI, Stern R. Evidence that the serum inhibitor

of hyaluronidase may be a member of the inter-a-inhibitor family.

J Biol Chem 2000; 275: 13–21.

Mitsios N, Gaffney J, Kumar P, Kumar S, Slevin M. Apoptosis after

ischaemic stroke. Cell Biochem Biophys 2006. In press.

Nagy JI, Hacking J, Frankenstein UN, Turley EA. Requirement of the

Hyaluronan receptor RHAMM in neurite extension and motility as

demonstrated in primary neurons and neuronal cell lines. J Neurosci 1995;

15: 241–52.

Natowicz MR, Wang Y. Human serum hyaluronidase: characterisation of a

clinical assay. Clin Chim Acta 1996; 245: 1–6.

Noble PW. Hyaluronan and its catabolic products in tissue injury and

repair. Matrix Biol 2002; 21: 25–9.

Novak SS, Stylii AH, Kaye AH, Lepperdinger G. Hyaluronidase-2 over-

expression accelerates intracerebral but not subcutaneous tumor formation

of murine astrocytoma cells. Cancer Res 1999; 59: 6246–50.

Prehm P. Hyaluronan is synthesized at plasma membranes. Biochem J 1984;

220: 597–600.

Pure E, Cuff CA. A crucial role for CD44 in inflammation. Trends Mol Med

2001; 7: 213–21.

Rauch U. Extracellular matrix components associated with remodelling

processes in brain. Cell Mol Life Sci 2004; 61: 2031–45.

Ripellino JA, Bailo M, Margiolis RU, Margolis RK. Light and electron

microscopic studies on the localization of hyaluronic acid in developing

rat cerebellum. J Cell Biol 1988; 106: 845–55.

RubioF,Martinez-YelamosS,CardonaP,KrupinskiJ.Carotidendarterectomy:

is it still the gold standard? Cerebrovasc Dis 2005; 20: 119–22.

Savani RC, Cao G, Pooler PM, Zaman A, Zhou Z, DeLisser HM. Differential

involvement of the receptor of the hyaluronan (HA) receptors CD44 and

receptor for HA-mediated motility in endothelial cell function and

angiogenesis. J Biol Chem 2001; 276: 36770–8.

Sbarbarti A, Pietra C, Baldassari AM, Guerrini U, Ziviani L, Reggiani A, et al.

The microvascular system in ischaemic cortical lesions. Acta Neuropathol

(Berlin) 1996; 92: 56–63.

Scandinavian Stroke StudyGroup (1985),Multicentre trial of hemodilution in

ischaemic stroke: background and study protocol. Stroke 1985; 16: 885–90.

Sherman LS, Struve JN, Rangwala R, Wallingford NM, Tuohy MF, Kuntz C

IV. Hyaluronate-based extracellular matrix: keeping glia in their place.

Glia 2002; 38: 93–102.

Shuttleworth TL, Wilson MD, Wicklow BA, Wilkins JA, Triggs-Raine BL.

Characterization of the murine hyaluronidase gene region reveals

complex organization and cotranscription of HYAL1 with down-stream

genes Fus2 and HYAL3. J Biol Chem 2002; 277: 23008–18.

Slevin M, Krupinski J, Kumar S, Gaffney J. Angiogenic oligosaccharides

of hyaluronan induce protein tyrosine kinase activity in endothelial cells

and activate a cytoplasmic signal transduction pathway resulting in

proliferation. Lab Invest 1998; 78: 987–1003.

Slevin M, Krupinski J, Slowik A, Kumar P, Szczudlik A, Gaffney J. Serial

measurement of vascular endothelial growth factor and transforming

growth factor beta1 in serum of patients with acute ischaemic stroke.

Stroke 2000; 31: 1863–70.

Slevin M, Kumar S, Gaffney J. Angiogenic ologosaccharides of hyaluronan

induce multiple signal transduction pathways affecting vascular endo-

thelial cell mitogenic and wound healing responses. J Biol Chem 2002;

277: 41046–59.

SlevinM,West D, Kumar S. Hyaluronan, angiogenesis and malignant disease.

Int J Cancer 2004; 109: 793–4.

Slevin M, Krupinski J, Kumar P, Gaffney J, Kumar S. Gene activation and

protein expression following ischaemic stroke: strategies towards neuro-

protection. J Cell Mol Med 2005; 9: 85–102.

Stern R. Devising a pathway for hyaluronate catabolism: are we there yet?

Glycobiology 2003; 12: 105R–15R.

Struve J, Maher PC, Li YQ, Kinney S, Fehlings MG, Kuntz C IV, et al.

Disruption of the hyaluronan-based extracellular matrix in spinal cord

promotes astrocyte proliferation. 2005; 52: 16–24.

TammiMI, Day AJ, Turley EA. Hyaluronan and homeostasis: a balancing act.

J Biol Chem 2002; 277: 4581–4.

Usui T, Amano S, Oshika T, Suzuki K, Miyata K, Arai M, et al. Expression

regulation of hyaluronan synthase in corneal endothelial cells. Invest

Ophthalmol Vis Sci 2000; 41: 3261–7.


Wang KK, Nemeth IR, Seckel BR, Chakalis-Haley DP, Swann DA, Kuo JW,

et al. Hyaluronic acid enhances peripheral nerve regeneration in vivo.

Microsurgery 1998; 18: 270–5.

Wang X, Xu L, Wang H, Zhan Y, Pure E, Feuerstein GZ. CD44 deficiency

in mice protects brain from cerebral ischemia injury. J Neurochem 2002;

83: 1172–9.

West DC, Hampson IN, Arnold F, Kumar S. Angiogenesis induced by

degradation products of hyaluronic acid. Science 1985; 228: 1324–6.

West DC, Kumar S. The effect of hyaluronan and its oligosaccharides on

endothelial cell proliferation and monolayer integrity. Exp Cell Res 1989;

183: 179–86.

WestDC,LeesVC,FanTP.Angiogenesis inadelayedrevascularisationmodel is

accelerated by oligosaccharides of hyaluronan. Lab Invest 1995; 73: 259–66.

Yasahara O, Akiyama H, McGeer EG, McGeer PL. Immunohistochemical

localisation of hyaluronic acid in rat and human brain. Brain Res 1994;

28: 269–82.


doi:10.1093/brain/awl139 (2006), 129, 2158–2176 changes in ... · immediately frozen in liquid...

Documents