вход по аккаунту


Antigen-specific adaptive immune responses in fingolimod-treated multiple sclerosis patients.

код для вставкиСкачать
Antigen-Specific Adaptive
Immune Responses in
Fingolimod-Treated Multiple
Sclerosis Patients
Matthias Mehling, MD,1,2 Patricia Hilbert, MD,1
Stefanie Fritz, MSc,2 Bojana Durovic, MD, PhD,2
Dominik Eichin, BSc,2 Olivier Gasser, PhD,2
Jens Kuhle, MD,1 Thomas Klimkait, PhD,3
Raija L.P. Lindberg, PhD,1 Ludwig Kappos, MD,1
and Christoph Hess, MD, PhD2
T cells exit secondary lymphoid organs along a sphingosine1-phosphate (S1P) gradient and, accordingly, are
reduced in blood upon fingolimod-mediated S1P-receptor (S1PR)-blockade. Serving as a model of adaptive immunity, we characterized cellular and humoral immune
responses to influenza vaccine in fingolimod-treated
patients with multiple sclerosis (MS) and in untreated
healthy controls. Although the mode of action of fingolimod might predict reduced immunity, vaccine-triggered
T cells accumulated normally in blood despite efficient
S1PR-blockade. Concentrations of anti–influenza A/B
immunoglobulin (Ig)M and IgG also increased similarly in
both groups. These results indicate that fingolimodtreated individuals can mount vaccine-specific adaptive
immune responses comparable to healthy controls.
ANN NEUROL 2011;69:408–413
n multiple sclerosis (MS) lesions lymphocytes mediate
inflammation, demyelination, and axonal damage.1
Lymphocytes that express the chemokine receptor CCR7
are able to migrate to secondary lymphoid organs (SLO)
such as lymph nodes, whereas cells not expressing CCR7
(CCR7-negative) do not recirculate to SLO on a regular
basis. In order to egress from SLO to the peripheral
blood circulation cells migrate along a sphingosine 1phosphate (S1P) gradient.2 The oral S1P receptor
(S1PR)-agonist fingolimod—which has shown efficacy in
the treatment of MS3–5—blocks this egress, thereby
reducing peripheral lymphocyte counts, in a dose-dependent manner, to 25% to 40% of baseline values.3,6–8
As a consequence, CCR7-negative cells represent the
major T cell population circulating in the blood of fingolimod-treated patients. Despite severe lymphopenia only
few infectious complications were observed in fingolimod-treated patients with MS. However, a case of fatal
disseminated varicella zoster infection and a case of
C 2011 American Neurological Association
408 V
herpes simplex virus type 1 encephalitis nonetheless raise
concern with regard to the immunological competence
vis-à-vis viral pathogens in fingolimod-exposed individuals.5,9 So far only animal data are available on the effect
of S1PR-blockade on adaptive immune responses following viral antigen exposure. In simian human immunodeficiency virus (SHIV)-infected rhesus macaques treatment
with fingolimod did not result in deviations from the natural pattern of viral control.10 Treatment with fingolimod
also had no effect on the disease course and T cell exhaustion in mice infected with lymphocytic choriomeningitis
virus (LCMV).11 In contrast, treatment with fingolimod
lead to a significant reduction of influenza-antigen specific CD8þ T cells in lungs of animals infected with
influenza.12 Fundamental in this context, yet never experimentally addressed in humans, is how blocking S1PR
impacts on the presence of bulk vs recently antigen-activated T cells in the peripheral circulation. Here we
sought to define in a prospective observational study the
effect of fingolimod-mediated S1PR-blockade on the development of antigen-specific immune responses in
patients with MS.
Patients and Methods
Study Subjects and Procedures
We conducted an open-label, observational, prospective study
to assess the adaptive immune response induced by influenzavaccine in fingolimod-treated patients with MS and in healthy
controls (HC). The trial was conducted during the influenza
vaccination periods 2008/2009 and 2009/2010 (for inclusion
From the 1Department of Neurology and Clinical Neuroimmunology
Laboratory/Department of Biomedicine, 2Immunobiology Laboratory/
Department of Biomedicine and Medical Outpatient Department, and
University Hospital Basel, Basel, Switzerland; and 3Institute of Medical
Microbiology, University of Basel, Basel, Switzerland.
Address correspondence to Prof Hess, MD, PhD, Medical Outpatient
Department and Department of Biomedicine, University Hospital Basel,
4 Petersgraben, CH-4031 Basel, Switzerland. E-mail:; and
Ludwig Kappos, MD, Department of Neurology and Department of
Biomedicine, University Hospital Basel, 4 Petersgraben, CH-4031 Basel,
Switzerland, E-mail:, Phone: þ41 (0)61 265 44 64, Fax:
þ41 61 265 41 98
Additional Supporting Information can be found in the online version of
this article.
Received May 12, 2010, and in revised form Nov 26, 2010. Accepted for
publication Dec 3, 2010.
View this article online at DOI: 10.1002/ana.22352
Mehling et al: Vaccine-Triggered T Cells in Fingolimod Patients
TABLE: Characteristics of the Study Population and Tolerability of Influenza Vaccination and Incidence of
Influenza-Like Illnesses
Healthy Controls
MS Fingolimod
Median age, yr (range)
37 (19–46)
44 (31–60)
Median disease duration, yr (range)
12.3 (3–20)
Median EDSS (range)
2.6 (1.0–4.0)
Median therapy duration, mo (range)
36 (7–42)
Injection-site reactions day 0–3 postvaccination
12/18 (66%)
7/14 (50%)
General symptoms day 0–3 postvaccination
2/18 (11%)
4/14 (29%)
MS relapses
0/14 (0%)
Incidence of influenza-like illness
2/18 (11%)
2/14 (14%)
Baseline characteristics
Fingolimod dosage 0.5mg/1.25mg
Tolerability of vaccine/incidence of influenza-like illness
EDSS ¼ Expanded Disability Status Scale; MS ¼ multiple sclerosis; MS fingolimod ¼ fingolimod-treated patients with multiple
sclerosis; NA, not applicable.
and exclusion criteria see the Supporting Information Methods).
The institutional review board of both Basels approved the
study. After written informed consent, blood samples from
study subjects were obtained before and 7, 14, and 28 days after seasonal influenza vaccination with Mutagrip (Sanofi Pasteur
SA, Lyon, France). Clinical assessments are described in the
Supporting Information Methods.
Flow Cytometry
T cells were analyzed for expression of CD3, CD4, and CD8
using a CyAn cytometer (DakoCytomation, Glostrup, Denmark) according to standard procedures (used antibodies listed
in Supporting Information Methods).
Anti–Influenza A and Anti–Influenza
B Enzyme-Linked Immunosorbent Assay
Concentrations (given as virotech [VE] units/ml) of anti–influenza A and anti–influenza B immunoglobulin (Ig)M and IgG
were determined using a quantitative enzyme-linked immunosorbent assay (ELISA) according to the manufacturer’s instructions (Genzyme Virotech, Ruesselsheim, Germany). As recommended by the manufacturer, seroprotection was defined as an
anti–influenza A/B IgG concentration of 10 VE/ml.
Statistical Analyses
See Supporting Information Methods.
Enzyme-Linked Immunospot Assay
Enzyme-linked immunospot (ELISpot) was done as described13
with the modification that we used with Inflexal (Berna Biotech,
Kuesnacht, Switzerland) as the source of antigen (year-adjusted)
(see also Supporting Information Methods).
Virus-Specific Antigen Presentation Assay
Epstein-Barr virus (EBV)-specific T cell responses were characterized in the presence or absence of fingolimod as described14
(see also Supporting Information Methods).
Reverse-Transcription–Polymerase Chain
Reaction Analysis
For reverse-transcription–polymerase chain reaction (RT-PCR)
analysis, see Supporting Information Methods.
February 2011
Characteristics of the study population are summarized
in the Table. Rates of local injection site reactions and
general tolerability of the vaccine, as monitored by clinical assessments and a patient diary, were comparable in
fingolimod-treated patients and HC (see Table). Flow
cytometric analyses revealed a reduction of mean lymphocyte counts in fingolimod-treated patients by 64%
compared to the lower limit of the reference range
(CD4þ by 76–83%, CD8þ by 42–63%) (Fig 1A), an
observation that is in line with our previous findings.6
The frequency of T cells producing interferon (IFN)-c in
response to influenza-antigen was assessed by ELISpot.
Before vaccination, frequencies of influenza-specific IFNc–secreting T cells were comparable in fingolimodtreated patients and HC, as was the number of
of Neurology
FIGURE 1: Lymphocyte counts, cellular immune response after influenza-vaccination in fingolimod-treated patients vs healthy
controls, and virus-specific T cell responses in the presence or absence of fingolimod. (A) Mean lymphocyte counts in MS fingolim. and in HC as assessed by flow cytometry (6SEM). (B) The frequency of influenza-specific cells in MS fingolim. and HC as
detected by SFC in equal amounts of PBMC is shown before (day 0) and at days 7, 14, and 28 after influenza vaccination (median and interquartile ranges). (C) Ex vivo assessment of EBV-specific and (D) CMV-specific CD41 T cell-dependent IFN-c production as detected by SFC in equal amounts of PBMC from HC in the absence (EBV/CMV lysate and medium) or presence of
fingolimod P (EBV/CMV lysate and fingolimod P). (E) Frequencies of EBV-specific T cells before (day 0, CD31 T cells) and after
expansion of IFN-c secreting cells (day 7) by autologous EBV-transformed B cell clones in the absence (medium alone) and
presence of fingolimod and fingolimod P. (F) S1P1 mRNA levels were quantified in EBV-specific CD31 T cells after 3 and 7
days of control-culture and related to S1P1 mRNA levels in T cells after 3 and 7 days culture in the presence of fingolimod and
fingolimod P. *p < 0.05. CMV 5 cytomegalovirus; EBV 5 Epstein-Barr virus; fingolimod P 5 phosphorylated (or biologically
active) fingolimod; HC 5 healthy controls; IFN 5 interferon; mRNA 5 messenger RNA; MS 5 multiple sclerosis; MS fingolim
5 fingolimod-treated patients with MS; PBMC 5 peripheral blood mononuclear cells; SEM 5 standard error of the mean; SFC
5 spot-forming cells. [Color figure can be viewed in the online issue, which is available at]
individuals with no detectable influenza-specific response.
By day 7 postvaccination, frequencies significantly
increased in both groups and reached comparable levels
(see Fig 1B). Numbers of influenza-specific T cells
remained increased and comparable until day 14 postvaccination in both study groups. By day 28 postvaccination, frequencies of IFN-c–secreting cells contracted to
prevaccination levels in both groups. No individual
mounting a very high frequency of influenza-specific cells
was contained in the fingolimod group—a finding
which, however, did not reach statistical significance. No
statistically significant correlation between lymphocyte
counts and vaccine-specific immune responses was found
in HC or fingolimod-treated individuals (data not
shown). To assess whether fingolimod alters antigen-specific triggering of circulating virus-specific T cells or their
in vitro induction, the effect of active (ie, phosphorylated) fingolimod on ex vivo triggering and on in vitro
expansion of antigen-specific T cells was assessed. In the
presence vs absence of fingolimod (in its active, ie, phosphorylated form) no differences were detected in either
experimental system (see Fig 1C–E). Also, the presence
of fingolimod did not impact S1P1 mRNA levels in this
system (see Fig 1F).
To investigate how S1PR-blockade influences antibody responses, we quantified in these same patients and
HC influenza-specific IgM and IgG antibody production
by ELISA. Prevaccination levels of anti–influenza A and
Volume 69, No. 2
Mehling et al: Vaccine-Triggered T Cells in Fingolimod Patients
FIGURE 2: Antibody-response after influenza-vaccination in fingolimod-treated patients and in healthy controls. The concentration of (A) anti–influenza A and (B) anti–influenza B IgM is shown as detected before (day 0) and at days 7, 14, and 28 after
influenza vaccination in MS fingolim. and HC (median and interquartile ranges). The percentage of patients fulfilling IgG seroprotection criteria for (C) influenza A and (D) influenza B is shown before (day 0) and at days 7, 14 and 28 after influenza vaccination in MS fingolim. and HC. The percentage of initially seronegative patients converting to seroprotection for (E) influenza
A and (F) influenza B following vaccination (days 7–28). **p < 0.001. HC 5 healthy controls; Ig 5 immunoglobulin; MS 5 multiple sclerosis; MS fingolim. 5 fingolimod-treated patients with MS. [Color figure can be viewed in the online issue, which is
available at]
anti–influenza B IgM were comparably low in fingolimodtreated patients and HC. Following vaccination, concentrations of anti–influenza A and anti–influenza B IgM
increased significantly and comparably in both study
groups, and remained increased at comparable levels until
day 28 postvaccination (Fig 2A, B). Before vaccination,
71% of the fingolimod-treated patients and 50% of the
HC fulfilled the predefined seroprotection criteria (IgG
10 VE/ml) for influenza A (p ¼ 0.41), 71% of the fingolimod-treated patients and 44% of the HC for influenza B (p ¼ 0.38), indicating previous contact with antigen from these viruses in a substantial proportion of
study participants (see Fig 2C, D). At day 7 after vaccination the proportion of individuals fulfilling seroprotection criteria was comparably increased in both fingolimod-treated patients and HC (influenza A: p ¼ 0.64;
influenza B: p ¼ 0.53), and remained increased at days
14 and 28 postvaccination in both groups (day 14 and
28: influenza A and B; p ¼ 1.0). The proportion of individuals converting from undetectable to protective antiFebruary 2011
body levels was also similar in fingolimod-treated
patients and HC (see Fig 2E, F). Thus, the vaccine-specific production of IgM and, more importantly, IgG in
fingolimod-treated individuals was not impaired when
compared to levels in HC.
The key observation of this study was that fingolimodtreated patients with MS—despite severe peripheral lymphopenia—could mount a vaccine-specific adaptive immune
response that is comparable to the response observed in
HC. Finding a similar postvaccination frequency of influenza-specific peripheral blood T cells in fingolimod-treated
patients and HC—in spite of fingolimod-mediated lymphopenia—was unexpected. This observation indicates that in
humans lymphocyte egress from SLO is controlled differentially between lymph node–homing T cells interacting with
cognate antigen, as opposed to T cells screening for—but
not interacting with—cognate antigen. Recent animal data
are in line with such a model.15
of Neurology
An impaired antibody response in fingolimod-treated
individuals is a concern, as the drug directly impacts germinal center reactions and B cell migration.16,17 Again,
the vaccine-specific production of IgM and IgG in fingolimod-treated individuals was, however, not detectably
impaired when compared to levels in healthy controls, a
finding in line with some,18 but not all,17 data obtained
in animal models.
Our study has limitations, both from an immunological and from a clinical point of view. The vaccination model
we used does not take into account the complexity brought
by an influenza infection or any other virus infection, and
our study was underpowered to evaluate clinical endpoints
such as protection from influenza infection. Likewise, our
experiments detecting unchanged EBV-specific immune
responses in vitro cannot directly be extrapolated to indicate
intact immune control of other virus infections in fingolimod-treated individuals. However, the data serve as definite
proof-of-principle, demonstrating that blocking S1P-dependent lymphocyte migration in humans does not hinder
the appearance of antigen-activated T cells in the peripheral
circulation, nor does it affect the antibody response quantitatively. Clinical conditions with T cell lymphopenia comparable to that induced by fingolimod (human immunodeficiency virus [HIV] infection, myelotoxic chemotherapy)
are associated with a high risk for opportunistic infections.19,20 Our data indicate that fingolimod-treated
patients in principle can mount a virus-specific immune
response. It remains unclear, however, to what extent these
findings in the context of vaccine responses allow extrapolation to immunological competence vis-à-vis infectious
pathogens. The molecular basis of the observed bypass of
S1PR-dependent SLO egress by vaccine-triggered T cells in
humans remains to be determined. For clinicians these data
are informative when weighing the grade of immunosuppression inflicted on individuals treated with fingolimod,
and they permit a more rational interpretation of infectious
complications if they occur.
This research was supported by a grant from Novartis
AG, Basel, Switzerland.
We thank Gabriela Zenhaeusern for help with
establishing the influenza-specific ELISpot assay, and
Nicole Koch and Michaela Scherer for support with organization of study logistics.
Potential Conflicts of Interest
C.H. is supported by the Swiss National Science Foundation (PP00B-114850) and L.K. by the Swiss MS Society.
L.K. acted as a consultant, member, or chair of Steering
Committees, Data Safety Monitoring Boards, or Advisory
Boards in MS clinical trials sponsored by Accorda, Actelion,
Allergan, Allozyne, Bayer Schering, Biogen Idec, BiogenDompé, Boehringer Ingelheim, Genmab, GlaxoSmithKline,
Medicinova, Merck Serono, Novartis, Roche, Sanofi
Aventis, Santhera, Teva Pharmaceuticals, UCB Pharma,
and Wyeth, and has received lecture fees from Biogen Idec,
Helvea, GlaxoSmithKline, Mediservice, and Merck Serono.
Payments and consultancy fees have been exclusively used
for the support of research activities. L.K. discloses that he
has received grant support from Bayer Schering, Biogen
Idec, CSL Behring, the European Community Research
Fund, Genmab, Genzyme, GlaxoSmithKline, Medicinova,
Merck Serono, Novartis, Novartis Foundation, the Rubato
Foundation, Roche, Santhera, Sanofi Aventis, and UCB
Pharma. Payments and consultancy fees have been
exclusively used for the support of research activities.
M.M., P.H., S.F., J.K., T.K., R.L.P.L., and C.H. have no
competing financial interests to declare.
Goverman J. Autoimmune T cell responses in the central nervous
system. Nat Rev Immunol 2009;9:393–407.
Matloubian M, Lo CG, Cinamon G, et al. Lymphocyte egress from
thymus and peripheral lymphoid organs is dependent on S1P
receptor 1. Nature 2004;427:355–360.
Kappos L, Antel J, Comi G, et al. Oral fingolimod (FTY720) for
relapsing multiple sclerosis. N Engl J Med 2006;355:1124–1140.
Kappos L, Radue EW, O’Connor P, et al. A placebo-controlled
trial of oral fingolimod in relapsing multiple sclerosis. N Engl J
Med 2010;362:387–401.
Cohen JA, Barkhof F, Comi G, et al. Oral fingolimod or intramuscular interferon for relapsing multiple sclerosis. N Engl J Med.
Mehling M, Brinkmann V, Antel J, et al. FTY720 therapy exerts differential effects on T cell subsets in multiple sclerosis. Neurology
Vaessen LM, van Besouw NM, Mol WM, et al. FTY720 treatment
of kidney transplant patients: a differential effect on B cells, naive
T cells, memory T cells and NK cells. Transpl Immunol 2006;15:
Mehling M, Lindberg R, Raulf F, et al. Th17 central memory T cells
are reduced by FTY720 in patients with multiple sclerosis. Neurology 2010;75:403–410.
Garber K. Infections cast cloud over Novartis’ MS therapy. Nat
Biotechnol 2008;26:844–845.
Kersh EN, Luo W, Adams DR, et al. Evaluation of the lymphocyte
trafficking drug FTY720 in SHIVSF162P3-infected rhesus macaques. J Antimicrob Chemother 2009;63:758–762.
Walsh KB, Marsolais D, Welch MJ, et al. Treatment with a sphingosine analog does not alter the outcome of a persistent virus
infection. Virology 2010;397:260–269.
Patricia Hilbert and Stefanie Fritz, as well as Christoph
Hess and Ludwig Kappos contributed equally to this work.
Volume 69, No. 2
Eikermann-Haerter et al: CADASIL and CSD
McGill J, Legge KL. Cutting edge: contribution of lung-resident T
cell proliferation to the overall magnitude of the antigen-specific
CD8 T cell response in the lungs following murine influenza virus
infection. J Immunol 2009;183:4177–4181.
mutations in its pathogenesis. Our results link vascular
smooth muscle Notch 3 mutations to enhanced spreading depression susceptibility, implicating the neurovascular unit in the development of migraine aura.
Gasser O, Bihl FK, Wolbers M, et al. HIV patients developing primary CNS lymphoma lack EBV-specific CD4þ T cell function irrespective of absolute CD4þ T cell counts. PLoS Med 2007;4:e96.
ANN NEUROL 2011;69:413–418
Hill AB, Lee SP, Haurum JS, et al. Class I major histocompatibility
complex-restricted cytotoxic T lymphocytes specific for EpsteinBarr virus (EBV)-transformed B lymphoblastoid cell lines against
which they were raised. J Exp Med 1995;181:2221–2228.
Pham TH, Okada T, Matloubian M, et al. S1P1 receptor signaling
overrides retention mediated by G alpha i-coupled receptors to
promote T cell egress. Immunity 2008;28:122–133.
Cinamon G, Zachariah MA, Lam OM, et al. Follicular shuttling of
marginal zone B cells facilitates antigen transport. Nat Immunol
Han S, Zhang X, Wang G, et al. FTY720 suppresses humoral immunity by inhibiting germinal center reaction. Blood 2004;104:
Pinschewer DD, Ochsenbein AF, Odermatt B, et al. FTY720 immunosuppression impairs effector T cell peripheral homing without
affecting induction, expansion, and memory. J Immunol 2000;164:
Masur H, Ognibene FP, Yarchoan R, et al. CD4 counts as predictors of opportunistic pneumonias in human immunodeficiency virus (HIV) infection. Ann Intern Med 1989;111:223–231.
Mackall CL. T-cell immunodeficiency following cytotoxic antineoplastic therapy: a review. Stem Cells 2000;18:10–18.
Cerebral Autosomal
Dominant Arteriopathy with
Subcortical Infarcts and
Syndrome Mutations Increase
Susceptibility to Spreading
Katharina Eikermann-Haerter, MD,1
Izumi Yuzawa, MD,1 Ergin Dilekoz, DVM, PhD,1
Anne Joutel, MD, PhD,3,4
Michael A. Moskowitz, MD,1 and Cenk Ayata, MD1,2
Migraine with aura is often the first manifestation of
cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy syndrome
(CADASIL), a disorder caused by NOTCH3 gene mutations expressed predominantly in vascular smooth muscle. Here, we report that cortical spreading depression
(CSD), the electrophysiological substrate of migraine
aura, is enhanced in mice expressing a vascular Notch 3
CADASIL mutation (R90C) or a Notch 3 knockout mutation. The phenotype was stronger in Notch 3 knockout
mice, implicating both loss of function and neomorphic
February 2011
erebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy syndrome
(CADASIL) is a systemic nonamyloid, nonatherosclerotic
vasculopathy, with characteristic subcortical lacunar
infarcts and white matter lesions. CADASIL is the most
common monogenic inherited form of adult-onset stroke
and vascular dementia, linked to mutations in Notch 3
expressed predominantly in vascular smooth muscle cells.1
The incidence of migraine with aura, typically the first disease symptom, is 5 times greater compared to the general
population.1–3 Classical visual and sensory auras are most
common in CADASIL, although 50% of the patients also
experience atypical attacks with basilar, hemiplegic, or prolonged aura, and even coma.4,5 The mechanisms of this
association are unknown. Among more than 150 CADASIL mutations identified to date, the Notch 3 R90C mutation produces a strong phenotype in humans with early
onset of disease symptoms.6,7
Animal experiments as well as human imaging
studies suggest that migraine aura is caused by cortical
spreading depression (CSD), an intense depolarization of
neuronal and glial membranes slowly propagating by way
of gray matter contiguity. Evoked when extracellular Kþ
concentrations ([Kþ]e) exceed a critical threshold, CSD is
associated with massive Kþ and glutamate efflux, depolarizing adjacent neurons and glia to facilitate its spread.
Neuronal mutations in Cav2.1 channels causing familial
hemiplegic migraine type 1 augment CSD susceptibility
as a mechanism for the severe migraine with aura phenotype in this syndrome.8 Using transgenic mice expressing
the human archetypal CADASIL mutation Notch 3
From the 1Stroke and Neurovascular Regulation Laboratory, Department
of Radiology, and 2Stroke Service and Neuroscience Intensive Care Unit,
Department of Neurology, Massachusetts General Hospital, Harvard Medical
School, Charlestown, MA; 3Institut National de la Santé et de la Recherche
Médicale (INSERM), U740, Paris, France; and 4Université Paris7-Denis
Diderot, Faculté de Médecine, Site Lariboisière, Paris, France.
Address correspondence to Dr Ayata, Stroke and Neurovascular
Regulation Laboratory, Massachusetts General Hospital, 149 13th Street,
Room 6403, Charlestown, MA 02129. E-mail:
Current address for Ergin Dilekoz: Department of Pharmacology, Gazi
University Faculty of Medicine, Besevler, Ankara, Turkey 06500.
Additional Supporting Information can be found in the online version of
this article.
Received May 15, 2010, and in revised form Sep 14, 2010. Accepted for
publication Sep 24, 2010.
View this article online at DOI: 10.1002/ana.22281
Values are mean 6 SD. The amplitude and the duration at one-half amplitude of only the first CSD were measured in each mouse. Systemic physiological parameters were averaged
over the entire experiment.
BP ¼ mean arterial blood pressure; CSD ¼ cortical spreading depression; pCO2 ¼ partial pressure of CO2; pO2 ¼ partial pressure of O2; SD ¼ standard deviation; WT ¼ wild-type.
130 6 18
32 6 2
7.37 6 0.04
81 6 9
110 6 131
WT carotid stenosis
16 6 5
164 6 22
30 6 3
7.40 6 0.03
82 6 9
40 6 8
WT sham
20 6 5
147 6 22
33 6 2
7.39 6 0.04
76 6 8
31 6 3
30 6 4
151 6 20
32 6 4
7.36 6 0.06
76 6 6
54 6 20
30 6 2
136 6 25
31 6 3
7.40 6 0.03
86 6 6
33 6 6
18 6 2
136 6 17
29 6 2
7.41 6 0.02
86 6 5
32 6 8
19 6 4
128 6 15
32 6 2
7.39 6 0.03
83 6 7
42 6 12
43 6 12
19 6 3
133 6 18
31 6 3
7.39 6 0.02
87 6 3
CSD Amplitude
CSD Duration
Age (mo)
Experimental Group
TABLE: Electrophysiological Measures of CSD and Systemic Physiological Parameters
22 6 2
pCO2 (mmHg)
pO2 (mmHg)
of Neurology
Systemic Physiology
R90C that exhibit cerebrovascular dysfunction and most
of the arterial pathological hallmarks of CADASIL, we
now show that a CADASIL-associated vascular Notch 3
mutation also augments susceptibility to CSD, implicating CSD susceptibility as one common determinant of
migraine with aura phenotype in human syndromes.
Materials and Methods
Experimental procedures strictly followed institutional guidelines for animal care and use for research purposes, and were
approved by the institutional review committee. A total of 84
mice were used in this study. Experimental groups and the
numbers of mice in each group are shown in the Table. Male
and female (6–7 months old) wild type (Notch3WT) and transgenic mice expressing human CADASIL mutation Arg90Cys
under the control of SM22a promoter (TgNotch3R90C), were
studied.9 TgNotch3R90C mice overexpress mutant Notch 3, specifically in arterial smooth muscle cells, on a WT Notch 3
background (ie, normal endogenous expression of WT mouse
Notch 3); therefore, we compared TgNotch3R90C to transgenic
mice overexpressing the human WT Notch 3 (TgNotch3WT;
n ¼ 5) in addition to nontransgenic WT controls (Notch3WT).
We did not observe a difference between Notch3WT and
TgNotch3WT in any parameter, and therefore pooled the data
from these 2 groups into a single WT group. We also studied adult
(5 months old) male Notch 3 knockout mice (Notch3/)10 and
compared them to their WT littermates (Notch3WT). Although
Notch3/ mice do not carry a CADASIL mutation per se, they
model CADASIL mutations located in the ligand-binding domain
that have been shown to abolish Notch 3 receptor function.1,7,11
To address the importance of brain perfusion, we tested
the impact of chronic cerebral hypoperfusion on CSD susceptibility in WT mice after bilateral common carotid stenosis (steel
coils 0.18mm internal diameter, n ¼ 6), as described.12 CSD
susceptibility was tested 9 6 1 weeks later, and compared to
age-matched sham-operated controls (n ¼ 5).
The femoral artery was catheterized, and the trachea was intubated for mechanical ventilation under isoflurane anesthesia (2.5%
induction, 1% maintenance, in 70% N2O/30% O2). Arterial blood
gases and pH were measured every 20 minutes and maintained
within normal limits by adjusting ventilation. Systemic physiological
parameters did not differ among groups (see Table). For analysis of
CSD susceptibility, 3 burr holes were drilled at the coordinates
described.8 Two glass capillary microelectrodes were placed to record
extracellular steady (DC) potential and electrocorticogram. On each
hemisphere, first the electrical CSD threshold was determined by
escalating intensity cathodal square pulses (10–8000lC) via a bipolar electrode placed on the occipital cortex, and then a 1mm cotton
ball soaked in 300mM KCl was topically applied onto occipital cortex for 1 hour and the frequency of evoked CSDs was recorded.
The amplitude, propagation speed, and duration at one-half amplitude of the first CSD in each hemisphere were also measured. The
data obtained from the 2 hemispheres were averaged to serve as a
single data point for each animal. In separate experiments using
Volume 69, No. 2
Eikermann-Haerter et al: CADASIL and CSD
FIGURE : Increased CSD susceptibility in TgNotch3R90C mutant mice. (A) Left panel: Representative tracings showing that the electrical threshold for CSD induction was significantly lower in TgNotch3R90C compared to WT mice. Stepwise escalating cathodal
stimulation was used to determine the CSD threshold. Right panel: Representative 30-minute tracings showing that TgNotch3R90C
mutant mice developed significantly higher number of CSDs compared to WT during continuous topical KCl (300mM) application.
Calibration bars 5 10mV, 5 minutes. (B) Left panel: Electrical threshold for CSD induction in WT and TgNotch3R90C mutant mice.
Each data point represents the threshold for 1 animal (square, male; circle, female). Medians and interquartile ranges are also
shown. Please note the logarithmic ordinate scale. *p < 0.001 vs WT. Right upper panel: The frequency of CSDs triggered during
60-minute topical KCl application in WT and TgNotch3R90C mutant mice. Topical continuous KCl application (300mM) induced repetitive CSDs with higher frequency in TgR90C than in WT. Data are mean 6 SD. *p < 0.001 vs WT. Right lower panel: CSD propagation speed was faster in TgNotch3R90C mutant mice compared to WT in both males and females (*p < 0.05 vs WT), although
there was a trend toward even faster propagation speeds in female vs male TgNotch3R90C (p 5 0.07).
TgNotch3R90C mice and WT controls (n ¼ 6 each), we measured
and compared blood flow changes during KCl-induced CSD
(300mM) by laser speckle flowmetry, as described.13
February 2011
All experiments were done in a blinded fashion. Data
were analyzed with SPSS (version 11.0). Using a general linear
model of covariance analysis (ANCOVA), we tested for an
of Neurology
effect of the independent variables genotype and age on the
dependent variables cortical SD threshold, frequency, and
propagation speed. Other electrophysiological measures of
CSD and systemic physiological data were compared among
groups using 1-way ANOVA. Blood flow changes during CSD
were compared between groups by 2-way ANOVA for
repeated measures. A p value of <0.05 was considered statistically significant.
Direct cortical cathodal stimulation with stepwise escalating intensities triggered a CSD in all mice. The electrical
threshold for CSD was approximately 10-fold lower in
TgNotch3R90C compared to WT mice (Fig A,B). Female
TgNotch3R90C tended to exhibit a lower CSD threshold
compared to males (mean ¼ 25 [range ¼ 21–55] vs 55
[range ¼ 53–55] lC; p ¼ 0.066). Continuous epidural
KCl application evoked repetitive CSDs in all mice. Consistent with increased CSD susceptibility, TgNotch3R90C
mice developed approximately 40% more CSDs during
KCl application (see Fig A,B). Propagation speed of CSD
was also higher in TgNotch3R90C mice compared to WT
(see Fig B); however, CSD duration and amplitude did
not differ among groups (see Table). Moreover, Notch3/
mice developed even higher numbers of CSDs in response
to KCl (15 6 2 vs 9 6 1 CSDs/hour, in Notch3/ and
WT littermates; p < 0.001).
CADASIL patients, as well as mouse mutants, develop chronic vascular dysfunction and cerebral hypoperfusion.14,15 To test whether chronic hypoperfusion might
explain enhanced CSD susceptibility, we induced bilateral
common carotid stenosis in WT mice. However, we
found that both the frequency of KCl-induced CSDs
and their propagation speed were lower compared to
sham controls (6.0 6 1.4 CSDs/hour vs 9.2 6 1.8
CSDs/hour, 1.8 6 0.5mm/minute vs 2.5 6 0.2mm/
minute, n ¼ 6 and 5, respectively; p < 0.05; Supporting
Information Fig 1). Moreover, setting another contrast
with the TgNotch3R90C phenotype, CSD durations
tended to be prolonged in the stenosis group, probably
reflecting reduced cerebral perfusion pressure16 (see
Table). To test an alternative hypothesis that enhanced
CSD susceptibility might be caused by altered vascular
response to CSD, we measured CSD-induced blood flow
changes using laser-speckle flowmetry and did not find a
significant difference between TgNotch3R90C and WT
mice (n ¼ 6 each; Supporting Information Fig 2).
Genetic and epidemiological evidence suggests a strong
association between migraine with aura and adult-onset
genetic vasculopathies, such as CADASIL.1,17 Using 2 independent but complementary techniques (electrical and
KCl stimulation), we found that 2 different Notch 3
mutations (TgNotch3R90C and Notch3/) enhance cortical susceptibility to spreading depression as a potential
mechanism for this clinical association. Enhanced CSD
susceptibility in CADASIL may also explain more frequent and severe aura features (eg, coma), similar to familial hemiplegic migraine type 1 associated with mutations in the neuronal Cav2.1 channels.8
Enhanced CSD susceptibility in both the Notch3/
and the TgNotch3R90C appears paradoxical, because
Notch3/ mutation is a loss-of-function, whereas
Notch3R90C is probably neomorphic, as it retains Notch 3
function and is expressed on a WT Notch 3 background.18 However, consistent with our data in mutant
mice, both neomorphic and loss-of-function CADASIL
mutations are known to be associated with a migraine
phenotype in affected individuals.1,11
Cerebrovascular dysfunction is a common phenotype in both mouse mutants as well as in CADASIL
patients regardless of Notch 3 gain-of-function or loss-offunction,10,15,19–22 implicating vascular mechanisms for
the enhanced CSD susceptibility phenotype in Notch 3
mutants. TgNotch3R90C mice do develop vascular changes
characteristic of CADASIL, including disruption of
smooth muscle anchorage to adjacent extracellular matrix
and cells, cytoskeletal changes, and later smooth muscle
degeneration.9 Even before the morphological changes
become conspicuous, TgNotch3R90C mice display abnormal myogenic and flow-mediated vascular responses,
impaired autoregulation and diminished hypercapnic hyperemia.19,20 Further supporting a vascular link, mutations
in TREX1, another vascular locus linked to the retinal
vasculopathy with cerebral leukodystrophy syndrome, are
also frequently associated with a migraine phenotype.23
The mechanisms linking vascular dysfunction to
enhanced CSD susceptibility, however, are not clear. Disturbances in the microcirculation caused by platelet dysfunction, inflammation, microemboli, and intermittent
hypoperfusion have been implicated.24 It has been speculated that ischemic neuronal injury could facilitate the
occurrence of CSD and migraine attacks in CADASIL.25
However, migraine with aura is often the first symptom
in CADASIL and typically precedes ischemic events by
more than a decade. Moreover, migraine attacks gradually
recede at later stages of disease when ischemic injury manifests.17 One possible mechanism is that blood flow dysregulation in CADASIL brains may limit the ability of cerebral vasculature to maintain energy and ionic homeostasis
during intense brain activation, facilitating [Kþ]e to rise
above the CSD threshold. However, in separate
Volume 69, No. 2
Eikermann-Haerter et al: CADASIL and CSD
experiments we found that chronic forebrain hypoperfusion alone, induced by bilateral common carotid artery
stenosis, did not enhance KCl-induced CSD susceptibility,
and that CSD-induced blood flow changes did not differ
between WT and CADASIL mutants, both arguing
against a primary vascular mechanism for enhanced CSD
susceptibility. Additional mechanisms are therefore needed
to explain enhanced CSD susceptibility in CADASIL
Importantly, whether Notch 3 CADASIL mutations
impact other cell types in the neurovascular unit has not been
investigated in detail. For example, astrocyte end feet are in
close proximity to cerebral blood vessels, and abnormal
Notch 3 signaling may disrupt normal astrocyte-smooth
muscle communication. Indeed, impaired functional neurovascular coupling was recently reported in a CADASIL
mouse model expressing the Notch 3 R169C mutation.14
Alternatively, Notch 3 mutations expressed in neural progenitor cells and transiently in newly-born neurons may lead to
enhanced CSD susceptibility phenotype later in life.2,10,17
In summary, we found that Notch 3 mutations
modulate cortical susceptibility to spreading depression,
the electrophysiological substrate of migraine aura. These
data are the first to link a monogenic vasculopathy to a
spreading depression phenotype, and provide an explanation for the frequent and severe migraine with aura
phenotype in CADASIL patients.
This research was supported by grants from the American
Heart Association (10SDG2610275 to K.E.-H.) and the
National Institutes of Health (NIH) (NS061505 to C.A.;
NS35611 to M.M; NS054122 to A.J.).
Notch 3 knockout mice were kindly provided by
Dr. Spyros Artavanis-Tsakonas and Dr. Joseph F. Arboleda-Velasquez at Harvard Medical School.
Potential Conflicts of Interest
Feuerhake F, Volk B, Ostertag CB, et al. Reversible coma with
raised intracranial pressure: an unusual clinical manifestation of
CADASIL. Acta Neuropathol 2002;103:188–192.
Schon F, Martin RJ, Prevett M, et al. ‘‘CADASIL coma’’: an underdiagnosed acute encephalopathy. J Neurol Neurosurg Psychiatry
Utku U, Celik Y, Uyguner O, et al. CADASIL syndrome in a large
Turkish kindred caused by the R90C mutation in the Notch3 receptor. Eur J Neurol 2002;9:23–28.
Monet-Lepretre M, Bardot B, Lemaire B, et al. Distinct phenotypic
and functional features of CADASIL mutations in the Notch3
ligand binding domain. Brain 2009;132:1601–1612.
Eikermann-Haerter K, Dilekoz E, Kudo C, et al. Genetic and hormonal factors modulate spreading depression and transient hemiparesis in mouse models of familial hemiplegic migraine type 1.
J Clin Invest 2009;119:99–109.
Ruchoux MM, Domenga V, Brulin P, et al. Transgenic mice expressing mutant Notch3 develop vascular alterations characteristic of
cerebral autosomal dominant arteriopathy with subcortical infarcts
and leukoencephalopathy. Am J Pathol 2003;162:329–342.
Arboleda-Velasquez JF, Zhou Z, Shin HK, et al. Linking Notch signaling
to ischemic stroke. Proc Natl Acad Sci U S A 2008;105:4856–4861.
Arboleda-Velasquez JF, Lopera F, Lopez E, et al. C455R notch3
mutation in a Colombian CADASIL kindred with early onset of
stroke. Neurology 2002;59:277–279.
Shibata M, Ohtani R, Ihara M, Tomimoto H. White matter lesions
and glial activation in a novel mouse model of chronic cerebral
hypoperfusion. Stroke 2004;35:2598–2603.
Ayata C, Dunn AK, Gursoy OY, et al. Laser speckle flowmetry
for the study of cerebrovascular physiology in normal and ischemic mouse cortex. J Cereb Blood Flow Metab 2004;24:
Joutel A, Monet-Lepretre M, Gosele C, et al. Cerebrovascular dysfunction and microcirculation rarefaction precede white matter
lesions in a mouse genetic model of cerebral ischemic small vessel
disease. J Clin Invest;120:433–445.
Chabriat H, Pappata S, Ostergaard L, et al. Cerebral hemodynamics in CADASIL before and after acetazolamide challenge assessed
with MRI bolus tracking. Stroke 2000;31:1904–1912.
Sukhotinsky I, Yaseen MA, Sakadzic S, et al. Perfusion pressuredependent recovery of cortical spreading depression is independent of tissue oxygenation over a wide physiologic range. J Cereb
Blood Flow Metab 2010;30:1168–1177.
Dichgans M, Mayer M, Uttner I, et al. The phenotypic spectrum of
CADASIL: clinical findings in 102 cases. Ann Neurol 1998;44:731–739.
Monet M, Domenga V, Lemaire B, et al. The archetypal R90C
CADASIL-NOTCH3 mutation retains NOTCH3 function in vivo.
Hum Mol Genet. 2007.
Lacombe P, Oligo C, Domenga V, et al. Impaired cerebral vasoreactivity in a transgenic mouse model of cerebral autosomal
dominant arteriopathy with subcortical infarcts and leukoencephalopathy arteriopathy. Stroke 2005;36:1053–1058.
Dubroca C, Lacombe P, Domenga V, et al. Impaired vascular
mechanotransduction in a transgenic mouse model of CADASIL
arteriopathy. Stroke 2005;36:113–117.
Pfefferkorn T, von Stuckrad-Barre S, Herzog J, et al. Reduced
cerebrovascular CO(2) reactivity in CADASIL: a transcranial Doppler sonography study. Stroke 2001;32:17–21.
Domenga V, Fardoux P, Lacombe P, et al. Notch3 is required for
arterial identity and maturation of vascular smooth muscle cells.
Genes Dev 2004;18:2730–2735.
Richards A, van den Maagdenberg AM, Jen JC, et al. C-terminal
truncations in human 30 -50 DNA exonuclease TREX1 cause autosomal dominant retinal vasculopathy with cerebral leukodystrophy.
Nat Genet 2007;39:1068–1070.
A.J. has one or more patents with Athenadiagnostics for
diagnostic assay of CADASIL disease.
Chabriat H, Joutel A, Dichgans M, et al. Cadasil. Lancet Neurol
Chabriat H, Vahedi K, Iba-Zizen MT, et al. Clinical spectrum of
CADASIL: a study of 7 families. Cerebral autosomal dominant
arteriopathy with subcortical infarcts and leukoencephalopathy.
Lancet 1995;346:934–939.
Vahedi K, Chabriat H, Levy C, et al. Migraine with aura and brain
magnetic resonance imaging abnormalities in patients with CADASIL. Arch Neurol 2004;61:1237–1240.
February 2011
of Neurology
Dalkara T, Nozari A, Moskowitz MA. Migraine aura pathophysiology: the role of blood vessels and microembolisation. Lancet
Neurol 2010;9:309–317.
Ferrari M. In: Silberstein SD, Lipton RB, Dodick DW, eds. Wolff’s
headache and other head pain. 8th ed. Oxford, UK: Oxford University Press, 2008:133–149.
strong luminescence of Gaussia luciferase (GL) to detect protein-protein interactions with high sensitivity. Using this
method, we evaluated the titers and pathogenic properties of
serum antibodies to the extracellular portion of Lrp4 in
patients with AChR antibody-negative MG.
Materials and Methods
Autoantibodies to
Low-Density Lipoprotein
Receptor–Related Protein
4 in Myasthenia Gravis
Osamu Higuchi, PhD,1 Johko Hamuro, PhD,1
Masakatsu Motomura, MD,2
and Yuji Yamanashi, PhD1
Myasthenia gravis (MG) is an autoimmune disease of the
neuromuscular junction, where acetylcholine receptor
(AChR), muscle-specific kinase (MuSK), and low-density
lipoprotein (LDL) receptor-related protein 4 (Lrp4) are
essential. About 80% and 0% to 10% of patients with
generalized MG have autoantibodies to AChR and
MuSK, respectively, but pathogenic factors are elusive
in others. Here we show that a proportion of AChR antibody-negative patients have autoantibodies to Lrp4.
These antibodies inhibit binding of Lrp4 to its ligand
and predominantly belong to the immunoglobulin G1
(IgG1) subclass, a complement activator. These findings
together indicate the involvement of Lrp4 antibodies in
the pathogenesis of AChR antibody-negative MG.
ANN NEUROL 2011;69:418–422
he neuromuscular junction (NMJ) is a synapse between
the motor nerve terminal and the skeletal muscle endplate.1 Postsynaptic clustering of the neurotransmitter receptor acetylcholine receptor (AChR) is controlled by musclespecific kinase (MuSK) and low-density lipoprotein (LDL)
receptor-related protein 4 (Lrp4), which form an essential
postsynaptic receptor complex for its ligand, neural agrin.2–4
Myasthenia gravis (MG) is an autoimmune disease of the
NMJ.5 About 80% of patients with generalized MG have
AChR antibodies, which is a causative factor for the disease,
and a variable proportion of the remaining patients (0–50%
throughout the world) have MuSK antibodies.6–12 However,
diagnosis and clinical management remain complicated for
patients who are negative for MuSK and AChR antibodies,
giving rise to a need for unveiling the hidden causative factors
in MG. Given the essential role and postsynaptic localization
of Lrp4 in the NMJ, we hypothesized that Lrp4 autoantibodies might be a pathogenic factor in MG. In this study, we
developed a simple technique termed luciferase-reporter
immunoprecipitation (LUCIP), which takes advantage of the
We studied serum samples from 300 patients with AChR antibody-negative MG diagnosed by typical clinical features, the
edrophonium test, and/or the repetitive nerve stimulation test.
Sera from 100 healthy volunteers, 100 patients with AChR
antibody-positive MG, and 101 patients with Lambert-Eaton
myasthenic syndrome (LEMS) were also studied. All MG sera
were tested for AChR and MuSK antibodies using the standard
radioimmunoprecipitation assay (RIA).7,13
For the Lrp4-LUCIP assay, HEK293 cells were transfected with expression plasmid for Lrp4-GL, in which the entire
extracellular portion of Lrp4 was fused to GL,14 and the fusion
protein was purified from the culture supernatant (Supporting
Information Methods). The specific luciferase activities of Lrp4GL were 1.25 108 relative light units [RLU]/pmol using the
BioLux Gaussia luciferase assay kit (BioLabs) and a Lumat LB
9507 luminometer (Berthold Technologies).
To titrate Lrp4 antibodies, 5ll serum was added to
24fmol Lrp4-GL in 700ll phosphate-buffered saline with
0.05% Tween 20 and 3% bovine serum albumin for overnight
incubation at 4 C. Immunoglobulin G (IgG)-bound Lrp4-GL
was precipitated with 15ll protein G-Sepharose (GE Healthcare). The precipitates were washed and their luciferase activities
were determined to calculate the amount of Lrp4-GL protein,
whose value was used to represent the titer of Lrp4 antibodies.
For subclass-specific titration of IgG antibodies to Lrp4,
1ll serum was incubated with Lrp4-GL as described above in
the presence of 10, 5, 2, and 2lg biotinylated anti-human
IgG1, IgG2, IgG3, and IgG4 antibodies (Binding Site), respectively.10 Immune complexes comprised of these subclass-specific
antibodies, serum IgGs to Lrp4, and Lrp4-GL were precipitated
with 15ll NeutrAvidin-Agarose (Thermo Scientific). The precipitates were washed and their luciferase activities were determined to calculate subclass-specific antibody titers.
From the 1Division of Genetics, Department of Cancer Biology, the
Institute of Medical Science, the University of Tokyo, Tokyo, Japan; and
First Department of Internal Medicine, Graduate School of Biomedical
Sciences, Nagasaki University, Nagasaki, Japan.
Address correspondence to Dr Yamanashi, Division of Genetics,
Department of Cancer Biology, Institute of Medical Science, University
of Tokyo, 4-6-1 Shirokanedai, Minato-ku, Tokyo 108-8639, Japan.
Additional Supporting Information can be found in the online version of
this article.
Received May 4, 2010, and in revised form Sep 28, 2010. Accepted for
publication Oct 15, 2010.
View this article online at DOI: 10.1002/ana.22312
Volume 69, No. 2
Higuchi et al: Lrp4 Antibodies in MG
To evaluate agrin-Lrp4 interaction, HEK293 cells were
transfected with expression plasmid for 3FLAG-agrin in
Opti-modified Eagle medium (Opti-MEM) and the culture supernatant containing secreted 3FLAG-agrin was harvested.
The culture supernatant containing 11fmol 3FLAG-agrin was
incubated with 6fmol Lrp4-GL, and then 3FLAG-agrin was
immunoprecipitated with anti-FLAG antibody-conjugated agarose (Sigma). The precipitates were washed and luciferase activities were determined to calculate the amount of Lrp4-GL
bound to 3FLAG-agrin (see Supporting Information Methods
for details). To further investigate the effects of serum or serum
IgGs on agrin-Lrp4 interaction, 11fmol 3FLAG-agrin and
different volumes of serum or corresponding amounts of IgGs
were subjected to this assay. Serum IgGs were purified using
the IgG Purification Kit-G (Dojindo), or IgGs were depleted
from serum using an excess amount of protein G-Sepharose.
To validate the significance of the observed differences,
we analyzed simple pairwise comparisons with the Student t
test (2-tailed distribution with 2-sample equal variance).
We tested for Lrp4 autoantibodies in sera from patients
with AChR antibody-negative MG by the Lrp4-LUCIP
assay, which uses a fusion protein of the entire extracellular portion of Lrp4 and GL (Lrp4-GL) as a reporter
(Supporting Information Data; Supporting Information
Figs 1 and 2). From a cohort of 300 patients, 9 patients
were positive for antibodies to the extracellular portion of
Lrp4, where the cutoff value (0.015nM) was determined
based on the mean þ 4 standard deviations (SDs)
obtained with 100 healthy control sera (Fig 1A). The control GL-LUCIP assay confirmed that these patients were
negative for serum antibodies to GL (Supporting Information Table 1). Titers of Lrp4 antibodies in the 9 patient
sera ranged from 0.019nM to 2.07nM (median, 0.65nM;
Table). These titers are statistically significant because the
minimum titer value (0.019nM) in the Lrp4 antibodypositive sera occurs with a probability of less than 6.312
1010 under the null hypothesis that the AChR antibody-negative MG data follow the same distribution as
healthy controls.
Next, we tested sera from 100 MG patients positive for
AChR antibodies and all were negative for the Lrp4 antibodies, suggesting that the autoantibodies are mutually exclusive
in MG (see Fig 1A). Furthermore, Lrp4-LUCIP testing of
sera from 101 patients with LEMS, a different form of NMJ
autoimmune disease,16 revealed that the patients were negative
for Lrp4 antibodies aside from one who showed weak positivity (see Fig 1A). In addition, among 28 patients with MuSK
antibody-positive MG in the cohort of 300 AChR antibodynegative patients, 3 patients were also positive for Lrp4 antibodies (index case nos. 6, 8, and 9; see Table).
February 2011
FIGURE 1: Serum autoantibodies to the extracellular portion of
Lrp4 were found in patients with AChR-Ab2 MG and recognized
native Lrp4. (A) Scatter plot for the calculated titers (nM) of Lrp4
autoantibodies in sera from patients and the HC as indicated.
The red line indicates the cutoff value, calculated as the mean 1
4 SDs of the healthy control values. (B) IP of full-length Lrp4 protein with sera from patients (index cases nos. 1–3). Immunoprecipitates or WCL of HEK293 cells that had been transfected with
expression plasmid for Lrp4 (1) or the empty vector (2) were
subjected to IB with rat antiserum to the cytoplasmic portion of
Lrp4 (aL4C) (Supporting Information Methods). The blank lane
contained size markers (sm). (C) Immunostaining of cell surface
Lrp4 with sera from patients (index case nos. 1 and 3). Intact
(nonpermeabilized) HEK293 cells that had been transfected with
expression plasmid for Lrp4 or the empty vector (mock) were
stained with sera from patients or the HC (see Supporting Information Methods). Anti-human IgG-Alexa488 was used as a secondary antibody to visualize cell surface Lrp4 (arrowheads).
Scale bars 50lm. AChR 5 acetylcholine receptor; AChR-Ab1 5
AChR antibody–positive; AChR-Ab2 5 AChR antibody–negative; HC 5 healthy control; IB 5 immunoblotting; IgG 5 immunoglobulin G; IP 5 immunoprecipitation; Lrp4 5 low-density
lipoprotein receptor-related protein 4; LEMS 5 Lambert-Eaton
myasthenic syndrome; MG 5 myasthenia gravis; SD 5 standard
deviation; WCL 5 whole cell lysates. [Color figure can be viewed
in the online issue, which is available at]
of Neurology
TABLE: Clinical Features of Lrp4 Antibody-Positive Patients with Generalized MG
Case No.
Antibody (nM)a
Antibody (nM)b
Data from the Lrp4-LUCIP assay.
Data from the MuSK-RIA assay.
Disease severity was graded according to the MGFA classification as described.15
Lrp4 ¼ low-density lipoprotein receptor-related protein 4; LUCIP ¼ luciferase-reporter immunoprecipitation; MG ¼ myasthenia gravis;
MGFA ¼ Myasthenia Gravis Foundation of America; MuSK ¼ muscle-specific kinase; RIA ¼ radioimmunoprecipitation assay.
We next examined whether the serum antibodies to
Lrp4 present in these MG patients recognize the native
form of Lrp4. HEK293 cells transfected with full-length
Lrp4 expression plasmid and C2C12 myotubes expressing
endogenous Lrp4 were subjected to immunoprecipitation
with Lrp4 antibody-positive sera (index case nos. 1–3) or
healthy control sera (see Fig 1B and Supporting Information Fig 3A). In Supporting Information Figure 3A, index
case no. 2 was excluded due to paucity. Sera from these
patients but not serum from the healthy control precipitated Lrp4 proteins. Likewise, cell surface Lrp4 ectopically
expressed in HEK293 cells could be visualized by immunostaining of intact cells with sera from these patients, but
not the healthy control or antiserum to the cytoplasmic
portion of Lrp4 (see Fig 1C and Supporting Information
Figs 4 and 5), demonstrating that the serum antibodies to
Lrp4 can recognize its native form. However, sera from the
remaining patients (index case nos. 4–9) failed to visualize
cell surface Lrp4 (data not shown), and these sera, aside
from index case no. 4, also failed to immunoprecipitate
Lrp4 (see Supporting Information Fig 3B). Unlike the
healthy control, however, sera from these patients reacted
with the LA domain of the extracellular portion of Lrp4 in
immunoblots (see Supporting Information Fig 3C).
The clinical features of 9 patients with Lrp4 antibody-positive MG (index case nos. 1–9) are summarized
in Table. Generalized MG was diagnosed in these
patients, who showed severe limb muscle weakness or
progressive bulbar palsy or both. Thymoma was not
observed in any of these patients, unlike the situation in
patients with AChR antibody-positive MG.17
Because Lrp4 is the agrin-binding subunit of the
Lrp4:MuSK receptor complex,3,4 and serum antibodies
to Lrp4 in MG patients bound to the molecule’s extracellular portion, we speculated that those antibodies might
compete with agrin for binding with Lrp4. Indeed, sera
from Lrp4 antibody-positive patients, but not the healthy
control, inhibited interaction of Lrp4-GL with neural
agrin (Fig 2A). We confirmed that serum and IgGs prepared from the same patient (index case no. 3) showed
comparable inhibition (see Fig 2B). Conversely, when
IgGs were depleted from the patient’s serum, it lost its inhibitory activity (Supporting Information Fig 6). Thus, autoantibodies to Lrp4 could exert pathogenicity through
their potential to inhibit agrin and Lrp4:MuSK signaling
required for NMJs. Furthermore, to assess potential
involvement of the complement system in Lrp4 antibodypositive MG, we investigated the IgG subclass composition
of Lrp4 antibodies in patients (see Fig 2C and Supporting
Information Fig 7). The Lrp4-LUCIP assay in combination with subclass-specific immunoprecipitation of IgGs
revealed that Lrp4 autoantibodies were predominantly
comprised of IgG1, a complement activator, in each
patient, suggesting the potential for these antibodies to
cause complement-mediated impairment of NMJs.
We identified the novel antigen Lrp4 as a target for autoantibodies in AChR antibody-negative MG. These antibodies are mainly IgG1 and have the potential to inhibit
interaction between neural agrin and the extracellular
Volume 69, No. 2
Higuchi et al: Lrp4 Antibodies in MG
FIGURE 2: Autoantibodies to Lrp4 in MG patients have pathogenic properties. (A, B) Inhibition of the interaction between neural agrin and Lrp4 by serum autoantibodies to Lrp4. (A) Binding of neural agrin (33FLAG-agrin) to Lrp4-GL was inhibited by
sera from patients (index case nos. 1–3), but not by the HC, in a dose-dependent manner. (B) Interaction between 33FLAGagrin and Lrp4-GL was comparably inhibited by serum and the corresponding amount of serum IgGs purified from a patient
(index case no. 3), but not by the HC. Student t test, *p < 0.01. (C) Determination of IgG subclasses composing Lrp4 autoantibodies. Biotinylated anti-human IgG subclass (G1–G4) antibodies were used instead of protein G in the Lrp4-LUCIP assay to
evaluate the subclass-specific titer of serum antibodies from patients (index case nos. 1–4) to the extracellular portion of Lrp4.
Data are means 6 SDs, n 5 3 for each experimental group. HC 5 healthy control; IgG 5 immunoglobulin G; Lrp4 5 low-density lipoprotein receptor-related protein 4; LUCIP 5 luciferase-reporter immunoprecipitation; MG 5 myasthenia gravis; SD 5
standard deviation. [Color figure can be viewed in the online issue, which is available at]
portion of Lrp4. Therefore, these findings suggest pathogenic involvement of the complement system and
reduced agrin:Lrp4:MuSK signaling in Lrp4 antibodypositive MG. However, it is important to carefully evaluate contributions of these antibodies to myasthenia, especially those of the antibodies with lower titers (index case
nos. 4–9; see Table), which failed to visualize cell surface
Lrp4 likely due to their lower titers. Interestingly, Lrp4
antibodies were found in 3 of 28 patients with MuSK
antibody-positive MG and 1 of 101 patients with
LEMS. Because MuSK antibodies predominantly belong
to the IgG4 subclass, which does not activate complement, and Lrp4 is a postsynaptic protein, antibodies to
Lrp4 might contribute differently to pathogenesis than
antibodies to MuSK or the P/Q-type presynaptic Ca2þ
channel, a target for autoantibodies in LEMS.16 However, since MuSK antibodies, though predominantly
IgG4, are partially IgG1 subclass capable of activating
complement,18 MuSK and Lrp4 antibodies might also
contribute similarly to pathogenesis in a complement-dependent manner. Given that titers of Lrp4 antibodies were
relatively low in sera from patients with MuSK antibodypositive MG or LEMS, again contributions of Lrp4 antibodies to each myasthenia must be carefully evaluated. The
LUCIP assay developed in this study is a simple in vitro
system using no radioisotope. Moreover, the MuSK-LUCIP
assay, in which MuSK-GL was used as a reporter, showed
roughly a 50-fold lower cutoff value than that determined
with the conventional RIA for MuSK autoantibodies, indicating greater sensitivity for the LUCIP assay (Supporting
Information Fig 8). Therefore, this assay system could be
February 2011
used for routine diagnosis and clinical management of various autoimmune disorders, including MG.
It should be noted that the proportion of MuSK
antibody-positive patients within an AChR antibody-negative MG cohort varies from 0% to 50% throughout the
world.12 Given the number and narrow ethnic origins of
patients in the current study, further clinical and experimental data on greater numbers of patients worldwide
are required to fully understand the etiology and pathology of Lrp4 antibody-positive MG.
This research was supported by a grant (20247024) from the
Ministry of Education, Culture, Sports, Science and Technology of Japan and from Ministry of Health, Labour and
Welfare of Japan grants (200936024A and 200936129A)
(Grants-in-Aid for Scientific Research to Y.Y. and M.M.).
We thank S. Imoto, A. Niida, R.F. Whittier, and
A. Vincent for helpful suggestions.
The study was approved by the ethics committees
of the Graduate School of Biomedical Sciences, Nagasaki
University (no. 09031864), and the Institute of Medical
Science, the University of Tokyo (no. 20-60-210403).
Potential Conflicts of Interest
Nothing to report.
M.M. and Y.Y. have received grants from Grantsin-Aid for Scientific Research from the Ministry of Health,
Labour and Welfare of Japan.
of Neurology
Sanes JR, Lichtman JW. Induction, assembly, maturation and
maintenance of a postsynaptic apparatus. Nat Rev Neurosci 2001;
Hoch W, McConville J, Helms S, et al. Auto-antibodies to the receptor tyrosine kinase MuSK in patients with myasthenia gravis without
acetylcholine receptor antibodies. Nat Med 2001;7:365–368.
Weatherbee SD, Anderson KV, Niswander LA. LDL-receptorrelated protein 4 is crucial for formation of the neuromuscular
junction. Development 2006;133:4993–5000.
McConville J, Farrugia ME, Beeson D, et al. Detection and characterization of MuSK antibodies in seronegative myasthenia gravis.
Ann Neurol 2004;55:580–584.
Vincent A, Leite MI. Neuromuscular junction autoimmune disease:
muscle specific kinase antibodies and treatments for myasthenia
gravis. Curr Opin Neurol 2005;18:519–525.
Matthews I, Chen S, Hewer R, et al. Muscle-specific receptor tyrosine kinase autoantibodies-a new immunoprecipitation assay. Clin
Chim Acta 2004;348:95–99.
Kim N, Stiegler AL, Cameron TO, et al. Lrp4 is a receptor for
agrin and forms a complex with MuSK. Cell 2008;135:334–342.
Zhang B, Luo S, Wang Q, et al. LRP4 serves as a coreceptor of
agrin. Neuron 2008;60:285–297.
Vincent A. Unravelling the pathogenesis of myasthenia gravis. Nat
Rev Immunol 2002;2:797–804.
Almon RR, Andrew CG, Appel SH. Serum globulin in myasthenia
gravis: inhibition of alpha-bungarotoxin binding to acetylcholine
receptors. Science 1974;186:55–57.
Tannous BA, Kim DE, Fernandez JL, et al. Codon-optimized Gaussia luciferase cDNA for mammalian gene expression in culture and
in vivo. Mol Ther 2005;11:435–443.
Jaretzki A III, Barohn RJ, Ernstoff RM, et al. Myasthenia gravis: recommendations for clinical research standards. Task Force of the
Medical Scientific Advisory Board of the Myasthenia Gravis Foundation of America. Neurology 2000;55:16–23.
Lindstrom JM, Seybold ME, Lennon VA, et al. Antibody to acetylcholine receptor in myasthenia gravis. Prevalence, clinical correlates, and diagnostic value. Neurology 1976;26:1054–1059.
Toyka KV, Brachman DB, Pestronk A, et al. Myasthenia gravis:
passive transfer from man to mouse. Science 1975;190:397–399.
Newsom-Davis J. Lambert-Eaton myasthenic syndrome. Rev Neurol (Paris) 2004;160:177–180.
Engel AG, Lambert EH, Howard FM. Immune complexes (IgG and
C3) at the motor end-plate in myasthenia gravis: ultrastructural
and light microscopic localization and electrophysiologic correlations. Mayo Clin Proc 1977;52:267–280.
Vernino S, Lennon VA. Autoantibody profiles and neurological
correlations of thymoma. Clin Cancer Res 2004;10:7270–7275.
Leite MI, Jacob S, Viegas S, et al. IgG1 antibodies to acetylcholine receptors in ‘seronegative’ myasthenia gravis. Brain 2008;131:
Volume 69, No. 2
Без категории
Размер файла
1 464 Кб
patients, response, specific, immune, antigen, adaptive, sclerosis, multiple, treated, fingolimod
Пожаловаться на содержимое документа