Increased Plasma Matrix Metalloproteinase-9 Levels
Keiko Imamura,MD,PhD;Takao Takeshima,MD,PhD;Emi Fusayasu,MD,PhD;
Background and Objective.—Cortical spreading depression and neurogenic inflammation have been hypothesized to be
key steps in the development of migraine headache. Recent studies have highlighted matrix metalloproteinase-9 (MMP-9) in
cortical spreading depression, neurogenic inflammation, and cerebral ischemia. To seek their possible association, we investi-
gated plasma MMP-9 levels in migraineurs during headache-free periods.
Methods.—Plasma MMP-9 levels in 84 migraine subjects and 61 controls were determined by enzyme-linked immunosor-
bent assay. In addition, 23 patients with tension type headache were included in the study as comparative subjects.
Results.—The MMP-9 levels in migraineurs (42.5 ? 4.6 ng/mL, mean ? SE) were significantly higher than those in con-
trols (25.4 ? 2.7 ng/mL, P < .005). Those levels in tension type headache subjects (24.6 ? 4.8 ng/mL) did not differ from those
in controls.There was no significant difference between subjects having migraine with aura and those without aura.The MMP-9
levels did not correlate with age, duration of illness, frequency of migraine attack, duration of headache attack, or medication
for headache.Mean plasma MMP-9 levels were the highest in subjects from whom blood samples were taken 2-4 days after their
Conclusions.—The degradation of extracellular matrix showing the increase of MMP-9 in migrainurs may be associated
with an abnormality in their blood vessel permeability. MPP-9 plays some role in migraine pathophysiology. Further studies of
MMPs are necessary to elucidate their role.
Key words: cortical spreading depression, gelatinase, ischemia, neurogenic inflammation, migraine
Abbreviations: BBB blood–brain barrier, CSD cortical spreading depression, ECM extracellular matrix, IL-1b interleukin-1b,
MMP matrix metalloproteinase, MMP-9 matrix metalloproteinase-9, TNF-a tumor necrosis factor-a
Migraine headache, a highly prevalent and dis-
abling disease, is characterized by the peripheral and
central sensitization of pain perceptive systems.
Although the pathophysiology is not fully under-
stood, cortical spreading depression (CSD) and neu-
rogenic inflammation have been hypothesized to be
major steps in the development of migraine head-
ache.1,2Recent studies suggests possible association of
migraine and ischemic brain lesions, especially pos-
terior lobes and cerebellum.3The matrix metallopro-
teinases (MMPs) are a family of enzymes with more
than 20 members identified to date that are all extra-
cellular endopeptidases requiring Zn2+. The major
From the Department of Neurology, Institute of Neurological
Sciences, Faculty of Medicine, Tottori University, Yonago,
Address all correspondence to Dr.TakaoTakeshima,Associate
Professor of Neurology, Department of Neurology, Institute of
Neurological Sciences, Faculty of Medicine,Tottori University,
36-1 Nishicho, Yonago, 683-8504, Japan.
Accepted for publication June 19, 2007.
Conflict of Interest: None.
Published by Blackwell Publishing
© 2007 the Authors
Journal compilation © 2008 American Headache Society
targets are the proteins of the extracellular matrix
(ECM).4The role for MMPs has also been suggested
in the pathogenesis of both acute and chronic neuro-
logical disorders such as stroke,Alzheimers’s disease,
HIV-associated dementia, and multiple sclerosis.5-7
Matrix metalloproteinase-9 (MMP-9 or gelati-
nase B) is a relevant member of MMPs and has a
broad range of specific substrates such as gelatine,
collagens, and fibronectin.6Significant pathological
roles of MMP-9 have been reported in cerebral
ischemia and neural inflammation.8-10
Cortical spreading depression is a propagating
wave of neuronal and glial depolarization in cerebral
cortex and has been implicated in disorders of
neurovascular regulation such as stroke,head trauma,
and migraine. Gursoy-Ozdemir et al11demonsrated
CSD increased the cortical tissue MMP-9 level, in
animal models and altered the levels of inflammatory
cytokines. We have reported alteration of plasma
transforming growth factor-b, a relevant cytokine, in
interictal migrainures.12Alterations of plasma inter-
leukins have also been reported in migrainures.13
Changes of plasma MMPs in patients have been
reported in some neurological disorders such as
ischemic stroke14and multiple sclerosis.15However,
there has been no study concerning plasma MMPs in
migraine subjects.To explore the possible association
of the MMPs in migraine, we measured plasma
MMP-9 in migraineurs during the headache-free
SUBJECTS AND METHODS
Eighty-four volunteer migraineurs, who visited a
headache clinic atTottori University Hospital,partici-
pated in this study.All subjects were given a general
physical and neurological examination.The diagnosis
of headache type was established according to the
diagnostic criteria of the International Classification
of Headache Disorders II.16Twenty-one subjects suf-
fered from migraine with aura and 63 from migraine
without aura. All headache sufferers were generally
normal except for their headaches. We recruited
61 healthy subjects without headache as controls.
They comprised 58 volunteers recruited from hospital
workers, 2 students of the University, and 1 family
member of a patient, who were generally normal and
received no medication. In addition, 23 patients with
tension type headache were included in the study as
comparative subjects. Mean age and male to female
ratio of the subjects are shown in Table 1. Clinical
migraine subjects are summarized in Table 2.All par-
ticipants gave their informed consent following their
full understanding of the nature and aim of the study.
We obtained venous blood samples from each partici-
pant during the headache-free period, at least 2 days
after the last headache attack. Sample blood was
cooled immediately in EDTA-containing tubes. We
obtained plasma by centrifugal separation. The
plasma samples were frozen and stored at -30°C
until assay. MMP-9 levels in plasma were determined
by enzyme-linked immunosorbent assay (MMP-9,
human, Biotrak ELISA System; Amersham Bio-
science, Piscataway, NJ, USA). Statistical analyses
were performed using SPSS software version 11.01-J
(Tokyo, Japan). The data were evaluated using the
non-parametric Kruskal-Wallis test followed by
Mann–Whitney U-test.The Spearman coefficient was
used to evaluate correlations between variables.
The plasma MMP-9 levels in controls, migraine,
and tension type headache subjects are presented in
Table 3. The MMP-9 levels in migraineurs were sig-
nificantly higher than those in controls.Those levels in
tension type headache subjects did not differ from
those in controls. There was no significant difference
between MMP-9 levels in migraine subjects with aura
and those having migraine without aura.The MMP-9
Table 1.—Mean Age and Sex Ratio of the Subjects
nAge (years)Male : female
Migraine with aura
Migraine without aura
32.5 ? 11.1
33.6 ? 13.0
27.0 ? 8.2
36.0 ? 13.7
51.3 ? 18.2
Mean ? SD.
levels did not correlate with age, duration of illness,
frequency of migraine attack, duration of headache
attack, or medication for headache. Mean plasma
MMP-9 levels were the highest in subjects from
whom blood samples were taken 2-4 days after their
latest attack (Figure).
This study demonstrated that plasma MMP-9
levels were increased in migraineurs during the
headache-free periods.Although there might be some
possibility that high MMP-9 is a preparative state for
a migraine attack or that migraine attacks cause an
increase of MMP-9, the elevation of plasma MMP-9
levels seemed to be a consequence of migraine attack,
because they were associated with the period from
the latest headache attack in our findings.
Matrix metalloproteinases have been shown to
contribute to: (1) the breakdown of the blood–brain
barrier (BBB); (2) cytokine production and the pro-
pagation of an inflammatory response including
Table 2.—Headache Characteristics and Medication of Migraine Subjects
Migraine with aura
(n = 22)
Migraine without aura
(n = 62)
(n = 84)
Age of onset (years)
Duration of illness (years)
Frequency of headache (per month)
Duration of headache (hours)
Medication for headache (%)
No acute medication
17.1 ? 8.7
9.2 ? 8.6
3.4 ? 2.8
17.8 ? 11.3
23.0 ? 8.6
12.9 ? 10.7
4.5 ? 4.1
23.6 ? 15.9
21.5 ? 9.0
12.0 ? 10.3
4.2 ? 3.8
22.1 ? 15.0
Mean ? SD.
NSAID = non-steroidal anti-inflammatory drug.
Table 3.—Plasma MMP-9 Levels
Migraine with aura
Migraine without aura
25.4 ? 2.7
42.5 ? 4.6*
34.9 ? 6.0**
45.1 ? 5.8**
24.6 ? 4.8
*P < .005 compared with controls, **P < .05 compared with
Mean ? SE;Kruskal-Wallis
MMP-9 = matrix metalloproteinase-9.
P < .01; Mann-Whitney
Figure. —Plasma MMP-9 levels in migraineurs and their latest
headache attack. Plasma MMP-9 levels in migraineurs whose
latest headache attack occurred 2-4 days previously were sig-
nificantly higher than in other groups. Bars represent
mean ? SE. Mann–Whitney U-test, *P < .05 compared with
other 2 groups. MMP-9 = matrix metalloproteinase-9.
demyelination; (3) tumor invasion, metastasis, and
ECM leading to an alteration of structural integrity in
various diseases.6Since neurogenic inflammation is a
key feature of migraine, increased MMP-9 levels in
migraineurs might play an important role in its patho-
physiology.Acute attack of multiple sclerosis,which is
a demyelinating disorder associated with neurogenic
inflammation, causes an elevation of plasma MMP-9
Stimulation with lipopolysaccharide is
known to mimic the neurogenic inflammation in
experimental animal models.18Lipopolysaccharide
can enhance MMP-9 mediated by activation for
inflammatory cytokines such as tumor necrosis
factor-a (TNF-a) and interleukin-1b (IL-1b).
Cortical spreading depression is a neural phe-
nomenon associated with the migraine aura, such as
fortification spectra. CSD in rat models increases the
production of TNF-a and IL-1b in the brain tissure.19
Gursoy-Ozdemiret al demonstrated
upregulates MMP-9 in the brain, and then alters the
permeability of BBB in murine models.11Mean
plasma MMP-9 level in migraine with aura was sig-
nificantly higher than in the controls.It was unexpect-
edly lower than in migraine without aura, although
the difference was not statistically significant. Plasma
MMP-9 levels in migraine with aura were not higher
than those in migraine without aura in our clinical
study.There are some opinions that CSD might even
occur in the brains of patients with migraine without
aura.Premonitory symptoms such as change of mood
or feeling of hunger – the so called clinically silent
aura – often appear before migraine attack, when a
change of neuronal activity occurs in areas that do not
present any neuronal symptoms. If we hypothesize
that silent CSD occurs in migraine without aura, this
facilitates an understanding of the high MMP-9 levels
seen in subjects with migraine without aura in this
Previous reports have shown that plasma MMP-9
is increased in patients with ischemic stroke,9and that
MMP-9 promoted injury of BBB.8In an animal model
of brain ischemia, activation of MMP-9 appeared in
endothelial cells 48 hours after reperfusion following
the release of TNF-a and IL-1b. Activated MMP-9
causes the disruption of BBB.20Some recent reports
have claimed that migraine is a risk factor for cerebral
ischemia. It has been reported that subjects with
migraine show a higher prevalence of white matter
lesions in their brain as demonstrated by MRI.21
Elevated plasma MMP-9 levels in migraine may be
related to proteolytic degradation of the BBB.
Increased MMP-9 in migraineurs may support the
hypothesis that migraine may share a pathophysiol-
ogy with ischemic stroke.
The origin of MMP-9 in migraineurs is unclear.In
ischemic stroke, endothelial cells in cerebral vessels
seem to produce MMP-9.20Astrocytes and microglia
are demonstrated to be associated with the induction
of MMP-9 in a culture model of inflammatory stimu-
ration.10From the point of view of the relationship
between migraine and neurogenic inflammation or
CSD, it may be hypothesized that increased plasma
MMP-9 in migraineurs could be derived from the
brain tissue, but systemic ECM. However, this view
As far as we are aware, this is the first report
investigating increased plasma MMP-9 concentration
in migraine subjects. We found that plasma MMP-9
headache-free periods.The increased level of MMP-9
in migraineurs may be associated with the character-
istics of their blood vessels. Our results are an impor-
tant contribution to the understanding of migraine
headache. Further studies of MMPs in migraine are
necessary to fully elucidate their role.
Acknowledgment: This study was partly supported
by a Grant-in-Aid for Scientific Research from the Min-
istry of Education, Science, and Culture of Japan (K.I.
1. Goadsby PJ, Edvinsson L. The trigeminovascular
system and migraine: Studies characterizing cere-
brovascular and neuropeptide changes seen in
humans and cats. Ann Neurol. 1993;33:48-56.
2. Moskowitz MA, Macfarlane R. Neurovascular and
molecular mechanisms in migraine headaches.Cere-
brovasc Brain Metab Rev. 1993;5:159-177.
3. Kruit MC, Launer LJ, Ferrari MD, et al. Infarcts in
the posterior circulation territory in migraine. The
population-based MRI CAMERA study. Brain.
4. Dzwonek J, Rylski M, Kaczmarek L. Matrix metal-
loproteinases and their endogenous inhibitors in
neuronal physiology of the adult brain. FEBS Lett.
5. Lukes A,Mun-Bryce S,Lukes M,et al.Extracellular
matrix degradation by metalloproteinases and
central nervous system diseases. Mol Neurobiol.
6. Yong VW, Krekoski CA, Forsyth PA, et al. Matrix
metalloproteinases and diseases of the CNS. Trends
7. Yong VW,Power C,Forsyth P,et al.Metalloprotein-
ases in biology and pathology of the nervous system.
Nat Rev Neurosci. 2001;2:502-511.
8. Asahi M, Wang X, Mori T, et al. Effects of matrix
metalloproteinase-9 gene knock-out on the pro-
teolysis of blood-brain barrier and white matter
components after cerebral ischemia. J Neurosci.
9. Reynolds MA, Kirchick HJ, Dahlen JR, et al. Early
biomarkers of stroke. Clin Chem. 2003;49:1733-
10. RosenbergGA. Matrix
neuroinflammation. Glia. 2002;39:279-291.
11. Gursoy-Ozdemir Y, Qiu J, Matsuoka N, et al. Corti-
cal spreading depression activates and upregulates
MMP-9. J Clin Invest. 2004;113:1447-1455.
12. Ishizaki K, Takeshima T, Fukuhara Y, et al.
Increased plasma transforming growth factor-beta1
in migraine. Headache. 2005;45:1224-1228.
13. Perini F, D’Andrea G, Galloni E, et al. Plasma
14. Heo JH, Kim SH, Lee KY, et al. Increase in plasma
matrix metalloproteinase-9 in acute stroke patients
with thrombolysis failure. Stroke. 2003;34:e48-
15. Lee MA,Palace J,Stabler G,et al.Serum gelatinase
B, TIMP-1 and TIMP-2 levels in multiple sclerosis.
A longitudinal clinical and MRI study. Brain.
16. Headache Classification of the International Head-
ache Society. The International Classification of
17. Waubant E, Goodkin DE, Gee L, et al. Serum
MMP-9 andTIMP-1 levels are related to MRI activ-
ity in relapsing multiple sclerosis. Neurology. 1999;
and B (matrix metalloproteinase 2 and 9) activity in
cultured rat astrocytes. J Neurochem. 1995;64:1513-
19. Jander S, Schroeter M, Peters O, et al. Cortical
cytokine gene expression in the rat brain. J Cereb
Blood Flow Metab. 2001;21:218-225.
20. Rosenberg GA, Cunningham LA, Wallace J, et al.
Immunohistochemistry of matrix metalloprotein-
ases in reperfusion injury to rat brain: Activation
croglia in cell cultures. Brain Res. 2001;893:104-
21. Kruit MC, van Buchem MA, Hofman PA, et al.
Migraine as a risk factor for subclinical brain lesions.
stromelysin-1 and mi-