Serum amyloid P component: A novel potential player ...

Journal of the Neurological Sciences 379 (2017) 69–76

Contents lists available at ScienceDirect

Journal of the Neurological Sciences journal homepage: www.elsevier.com/locate/jns

Serum amyloid P component: A novel potential player in vessel degeneration in CADASIL Akihito Nagatoshi a, Mitsuharu Ueda a, Akihiko Ueda a, Masayoshi Tasaki a,b, Yasuteru Inoue a, Yihong Ma a, Teruaki Masuda a, Mayumi Mizukami a, Sayaka Matsumoto a, Takayuki Kosaka a, Takayuki Kawano c, Takaaki Ito d, Yukio Ando a,⁎ a

Department of Neurology, Graduate School of Medical Sciences, Kumamoto University, Kumamoto 860-0811, Japan Department of Morphological and Physiological Sciences, Graduate School of Health Sciences, Kumamoto University, Kumamoto 862-0976, Japan c Department of Neurosurgery, Graduate School of Medical Sciences, Kumamoto University, Kumamoto 860-0811, Japan d Department of Pathology and Experimental Medicine, Graduate School of Medical Sciences, Kumamoto University, Kumamoto 860-0811, Japan b

a r t i c l e

i n f o

Article history: Received 19 December 2016 Received in revised form 24 April 2017 Accepted 16 May 2017 Available online 26 May 2017 Keywords: CADASIL Proteomic Serum amyloid P component Periostin NOTCH3 Granular osmiophilic material

a b s t r a c t In cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL), granular osmiophilic material (GOM) may play some roles in inducing cerebrovascular events. To elucidate the pathogenesis of CADASIL, we used laser microdissection and liquid chromatography–tandem mass spectrometry to analyze cerebrovascular lesions of patients with CADASIL for GOM. The analyses detected serum amyloid P component (SAP), annexin A2, and periostin as the proteins with the largest increase in the samples, which also demonstrated NOTCH3. For the three proteins, anti-human SAP antibody had the strongest reaction in the lesions where the anti-human NOTCH3 antibody showed positive staining. Moreover, immunofluorescence staining with the two antibodies clearly showed co-localization of SAP and NOTCH3. mRNA analyses indicated no positive SAP expression in the brain materials, which suggested that the source of SAP found in the GOM was only the liver. A solid phase enzyme-linked immunosorbent assay confirmed the binding of SAP with NOTCH3. Serum SAP concentrations were neither up-regulated nor down-regulated in CADASIL patients, when compared with those in control subjects. SAP may play an important role in GOM formation although precise mechanisms remain to be elucidated. © 2017 Published by Elsevier B.V.

1. Introduction Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) is an intractable brain disorder for which no effective therapies have been developed. CADASIL is one of the most well-known inherited small vessel diseases caused by a single amino acid substitution in the NOTCH3 gene [1,2]. The primary clinical manifestations of the disease have been well documented as migraine with aura, ischemic stroke, mood disturbances, apathy, Abbreviations: BSA, bovine serum albumin; CAA, cerebral amyloid angiopathy; CADASIL, cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy; CPHPC, (R)-1-[6-[(R)-2-carboxy-pyrrolidin-1-yl]-6-oxohexanoyl]pyrrolidine-2-carboxylic acid; EDTA, ethylenediaminetetraacetic acid; ELISA, enzyme-linked immunosorbent assay; FAP, familial amyloid polyneuropathy; GOM, granular osmiophilic material; LC-MS/MS, liquid chromatography–tandem mass spectrometry; LMD, laser microdissection; PBST, phosphate-buffered saline with Tween 20; RT-PCR, reverse transcription-polymerase chain reaction; SAP, serum amyloid P component; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; TBST, Tris-buffered saline plus Tween 20; TGF-β, transforming growth factor-β. ⁎ Corresponding author at: Department of Neurology, Graduate School Medical Sciences, Kumamoto University, 1-1-1 Honjo, Chuo-ku, Kumamoto 860-0811, Japan. E-mail address: [email protected] (Y. Ando).

http://dx.doi.org/10.1016/j.jns.2017.05.033 0022-510X/© 2017 Published by Elsevier B.V.

motor disability, and vascular dementia [3]. As pathological findings, granular osmiophilic material (GOM) deposition, small granular degeneration in the tunica media of the arteries, and loss of smooth muscle cells are specific characteristics of the disease [4]. However, the precise pathogenic mechanism of the disease remains to be elucidated. The major component of GOM has been thought to be the ectodomain of NOTCH3 [5], but detailed mechanisms of how GOM causes vessels to degenerate remain unclear. That other coexisting proteins, in addition to NOTCH3, may also play key roles in the pathogenesis of the vessel wall degeneration in CADASIL is possible. Tissue inhibitor of metalloproteinase 3 and vitronectin were also thought to be components in NOTCH3 aggregates in homogenized whole human brain [6]. CADASIL model mice possessing a NOTCH3 mutation showed that those two proteins were involved in the vascular function or white matter lesions of the brain [7]. Another group demonstrated that latent transforming growth factor β (TGF-β) binding protein 1 was also one of the GOM components and suggested that TGF-β signal was involved with formation of GOM [8]. In addition, Wang and colleagues reported that various types of collagen and extracellular matrix proteins such as decorin and biglycan (BGN) were increased in CADASIL vessels [9–11]. Moreover, clusterin/apolipoprotein J was thought to be a

70

A. Nagatoshi et al. / Journal of the Neurological Sciences 379 (2017) 69–76

protein related to small vessel diseases [12]. However, the pathogenesis of CADASIL has not yet been fully explained, and the methods that have been used in the analyses may have been insufficient. In addition to the above-mentioned proteins, other candidate proteins may help elucidate the pathogenesis and therapy of the disease and are worthy of investigation. An approach combining laser microdissection (LMD) and liquid chromatography–tandem mass spectrometry (LC-MS/MS) has been useful for detecting amyloid components in tissues of patients with amyloidosis and degenerated proteins in patients with several neurodegenerative disorders [13–15]. In our studies reported here, to clarify the key molecules related to vascular degeneration or formation of GOM in CADASIL, we analyzed perivascular lesions obtained from patients with CADASIL and observed GOM by means of LMD and LC-MS/ MS. We also discuss the possible role of candidate molecules as related to NOTCH3 and GOM in CADASIL. 2. Materials and methods

Table 2 Dilution ratios of antibodies used for immunohistochemistry and immunofluorescence staining. No.

Antibodya

Dilution

1 2 3 4 5 6 7 8

Mouse monoclonal anti-human NOTCH3 Rabbit monoclonal anti-human SAP Rabbit polyclonal anti-human annexin A2 Rabbit polyclonal anti-human periostin HRP rabbit anti-mouse immunoglobulin antibody HRP goat anti-rabbit immunoglobulin antibody Donkey anti-mouse IgG H&L (Alexa Fluor 594) Goat anti-rabbit IgG H&L (Alexa Fluor 488)

1:100 1:200 1:100 1:100 1:50 1:100 1:200 1:200

Abbreviation: HRP: horseradish peroxidase. a Antibodies were obtained from Abnova (#1), Abcam (#2, 4), Cosmo Bio Co., Ltd. (#3), Dako (#5, 6), and Thermo Fisher Scientific (#7, 8).

acid (EDTA), and 0.002% Zwittergent 3–16 (Calbiochem, San Diego, CA). The collected tissues were heated at 98 °C for 90 min and then sonicated for 60 min in a water bath. Samples were digested overnight at 37 °C with 1.5 μL of 1 mg/mL trypsin (Promega, Madison, WI).

2.1. Patients and controls 2.4. Proteomic analysis We studied two autopsied tissues and one biopsied tissue obtained from patients with CADASIL and five autopsied tissues and one biopsied tissue obtained from control patients. The genetic mutation of NOTCH3 was determined as described in Ueda et al. [16]. We checked all patients with CADASIL having GOM in their vessel walls by means of immunohistochemistry and electron microscopy. Table 1 provides the age, sex, diagnosed diseases, duration of CADASIL, and vascular risk factors, which were limited to information available from clinical records of each patient. CADASIL patients 1 and 2 [4,17,18] and control patients 1–5 were autopsied cases. We used surgical specimens of the superficial temporal artery from CADASIL patient 3 [19] and control patient 6. We selected as control autopsied patients those patients whose death was not caused by a central nervous system disorder. We studied serum samples from five patients with CADASIL and six healthy control subjects (Table 3). 2.2. Materials and antibodies All chemical agents used were of analytical grade. Table 2 provides the antibodies used in the study. 2.3. Tissue collection and protein extraction We collected brain vessels, mainly leptomeningeal arteries and arterioles, by using the LMD7000 system (Leica Microsystems, Wetzlar, Germany). We placed the dissected vessels into 0.5-mL microcentrifuge tube caps containing 30 μL of 10 mM Tris, 1 mM ethylenediaminetetraacetic

The samples digested by trypsin were reduced with dithiothreitol, vacuum dried, and dissolved again with 40 μL of MS-grade water containing 0.1% trifluoroacetic acid and 2% acetonitrile. Samples of peptide mixtures processed from formalin-fixed paraffin-embedded tissues were used for nano-flow reversed-phase LC-MS/MS (LTQ Velos Pro; Thermo Fisher Scientific, San Jose, CA). We used a capillary reversedphase LC-MS/MS system that consisted of an Advance Splitless NanoCapillary LC dual solvent delivery system (Bruker-Michrom, Auburn, CA), the HTS-xt PAL autosampler (CTC Analytics, Zwingen, Switzerland), and LTQ Velos Pro with an XYZ nanoelectrospray ionization source (AMR, Tokyo, Japan). We injected these samples into a peptide L-trap column (Chemical Evaluation Research Institute, Tokyo, Japan). We separated the peptides by using a capillary reversed-phase C18 column (Chemical Evaluation Research Institute) with gradient elution and an ion spray into the mass spectrometer, with a spray voltage of 2.3 kV. The peptide mass tolerance was 2.0 Da, and the fragment mass tolerance was 0.8 Da. The normalized spectral abundance factor (NSAF) for a protein is the number of spectral counts (SpC, the total number of MS/MS spectra) identifying a protein, divided by the protein's length (L), divided by the sum of SpC/L for all proteins in the experiment. 2.5. Immunohistochemistry and immunofluorescence staining We used formalin-fixed brain, superficial temporal artery, and skin tissues obtained from CADASIL patients and control patients for

Table 1 Overview of patients. Patienta

Sex

Age (y)

Diagnosis

Specimen

Duration of diseaseb

Vascular risk factorb

CAD 1 CAD 2 CAD 3 Ctrl 1 Ctrl 2 Ctrl 3

M M M M M F

62 62 54 80 50 80

Leptomeningeal arteries Leptomeningeal arteries, skin Superficial temporal artery Leptomeningeal arteries Leptomeningeal arteries Leptomeningeal arteries

24 years 21 years 3 years

HT, SM SM SM HT, DL, SM HT, DL HT

Ctrl 4 Ctrl 5 Ctrl 6

M M M

53 75 48

CADASIL (R133C)c CADASIL (R449C) CADASIL (R75P) ALS, AMI Cardiac embolism Dissecting aortic aneurysm, cerebellar hemorrhage, PD AML AMI, AS Cerebral aneurysm

Leptomeningeal arteries Leptomeningeal arteries Superficial temporal artery

SM HT, DL

Abbreviations: ALS: amyotrophic lateral sclerosis, AMI: acute myocardial ischemia, AML: acute myeloid leukemia, AS: aortic stenosis, CAD: CADASIL patient, Ctrl: control patient, DL: dyslipidemia, F: female, HT: hypertension, M: male, PD: Parkinson disease, SM: smoking. a CAD 1 and 2 and Ctrl 1–5 were autopsy cases. b Duration of disease and vascular risk factors were determined according to information available from clinical and autopsy records. c Genetic mutations in CADASIL patients were determined by PCR as described in the text.


immunostaining (Table 1). Deparaffinized sections were incubated in 0.5% periodic acid for 15 min to inactivate endogenous peroxidase, after which 10% animal serum of each secondary antibody in 0.5% bovine serum albumin (BSA) was used for 30 min to block nonspecific background staining. For immunohistochemistry and immunofluorescence staining of NOTCH3, we activated antigens by using 1 mM EDTA buffer and autoclaving. The primary antibodies were mouse monoclonal anti-human NOTCH3, clone 1G5 (Abnova, Taipei, Taiwan), rabbit monoclonal anti-human serum amyloid P component (SAP) (Abcam, Cambridge, UK), rabbit polyclonal anti-human periostin (Abcam), and rabbit anti-human annexin A2 (Cosmo Bio Co., Ltd., Tokyo, Japan) (Table 2). For immunohistochemistry, we used horseradish peroxidase-conjugated rabbit anti-mouse or goat anti-rabbit immunoglobulin antibodies (Dako, Glostrup, Denmark) as secondary antibodies (Table 2). Reactivity was visualized by using diaminobenzidine (Dojindo, Mashiki, Japan) and NaN3 (Nacalai Tesque, Kyoto, Japan). Sections were counterstained with Mayer's hematoxylin (Muto Pure Chemicals Co., Ltd., Tokyo Japan). For immunofluorescence staining, the secondary antibodies were Alexa Fluor 594 donkey anti-mouse IgG H&L (Thermo Fisher Scientific, Rochester, NY) and Alexa Fluor 488 goat anti-rabbit IgG H&L (Thermo Fisher Scientific) (Table 2). Sections were analyzed by means of confocal microscopy (TCS STED CW) (Leica Microsystems). 2.6. Binding analysis via an enzyme-linked immunosorbent assay (ELISA) Zhang et al. used solid phase binding assay to confirm interaction between NOTCH3 and BGN [11]. We coated wells in a 384-well plate (Thermo Fisher) with 5 μg/mL recombinant human NOTCH1 (amino acids 19-526) Fc chimera protein, NOTCH2 (amino acids 26-530) Fc chimera protein, and NOTCH3 (amino acids 40-467) Fc chimera protein (R&D Systems, Minneapolis, MN) respectively, in coating buffer (8.4 g NaHCO3, 3.56 g Na2CO3, and H2O to 1.0 L, pH 9.5), followed by overnight incubation. All these peptides were small fragments of the extracellular domain of NOTCH proteins. After the plate was washed three times with 20 mM phosphate-buffered saline plus Tween 20 (PBST), it was incubated for 1 h at room temperature with Blocking One (Nacalai Tesque) at a 1:5 dilution. We applied various concentrations of native human SAP (Merck Millipore, Billerica, MA) in 20 mM PBS (pH 7.4) to each well, followed by incubation for N 24 h in 4 °C. We used a rabbit monoclonal anti-human SAP antibody (Abcam) as the primary antibody at a 1:2000 dilution in 1% BSA with PBST, and sample was incubated for 1 h at room temperature. We then used horseradish peroxidase-conjugated goat anti-rabbit immunoglobulin antibody (Dako) as the secondary antibody, which was used at a 1:2000 dilution in 1% BSA with PBST, after which the sample was incubated for 1 h at room temperature. Between each step, we washed each well three times with PBST. Reactivity was visualized by using the SureBlue TMB (KPL, Gaithersburg, MD), and the reaction was stopped by adding 1 N HCl (Nacalai Tesque). Absorption was analyzed by means of a microplate spectrophotometer (BioRad, Hercules, CA). 2.7. RT-PCR We extracted cerebral arteries from four autopsied brain samples (two CADASIL patients and two controls) according to a previously reported method [20]. We extracted mRNA from each tissue and human liver tissue by using QIAGEN RNeasy Mini Kit or RNeasy Micro Kit (QIAGEN, Hilden, Germany). We then performed reverse transcription (RT) by using PrimeScript RT Master Mix (Perfect Real Time) (Takara Bio Inc., Kusatsu, Japan). We applied SYBR Premix DimerEraser (Perfect Real Time) (Takara Bio Inc.), for the forward primer and reverse primer of SAP and GAPDH, respectively, to white LightCycler 480 Multiwell Plates 96 (Roche, Indianapolis, IN), performed RT-polymerase chain reaction (RT-PCR), and used the LightCycler 480 System (Roche) for analysis.

71

Table 3 Serum samples from patients with CADASIL and healthy control subjects. No.a

Samples from

Sex

Age (y)

1 2 3 4 5 6 7 8 9 10 11 12

CADASIL (R133C)

M

CADASIL (R133C) CADASIL (R133C) CADASIL (R75P) CADASIL (R182C) Healthy control 1 Healthy control 2 Healthy control 3 Healthy control 4 Healthy control 5 Healthy control 6

F F M M M M M F F F

52 62 (autopsy) 56 55 61 69 62 67 68 54 60 63

Abbreviations: F: female, M: male. a Six serum samples from five patients with CADASIL and six serum samples from six healthy control subjects were used. Serum samples 1 and 2 were obtained from the patient labeled CAD 1 in Table 1.

2.8. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDSPAGE) and Western blotting We diluted CADASIL and control serum samples (Table 3) at 1:100 by using 5% β-mercaptoethanol (Nacalai Tesque) in 2× Laemmli sample buffer (Bio-Rad). After we boiled samples at 95 °C for 5 min, we applied each sample to 4–20% Mini-PROTEAN TGX Gel (Bio-Rad) and performed SDS-PAGE. We transferred each band to a polyvinylidene fluoride blotting membrane (GE Healthcare UK Ltd., Little Chalfont, England) via the wet transfer method, and we used 5% skim milk with Tris-buffered saline plus Tween 20 (TBST) to block nonspecific reactions. We used the rabbit monoclonal anti-human SAP as the primary antibody at a 1:5000 dilution in 5% skim milk with TBST and incubated samples at 4 °C overnight. We then used the horseradish peroxidase-conjugated goat anti-rabbit immunoglobulin antibody as the secondary antibody at a 1:2000 dilution in 5% skim milk with TBST and incubated samples at room temperature for 1 h. Between each step, we washed the membrane three times with TBST. Reactivity was visualized by using the ECL Prime Western Blotting Detection Reagent (GE Healthcare UK Ltd); LAS 4000 EPUV Mini (GE Healthcare UK Ltd) was then used for analysis. The intensity of each band was analyzed by using NIH ImageJ software. 2.9. Statistics All data are expressed as means ± SD. We evaluated data with Student's t-test. All analyses were performed with Excel 2011 (Microsoft, Redmond, WA). P values of b 0.05 were regarded as statistically significant. 2.10. Ethics The Human Ethics Review Committee of Kumamoto University approved the study protocol, and signed consent forms were obtained from families of subjects. 3. Results 3.1. Proteomic analysis of vascular components via LMD and LC-MS/MS To investigate the components promoting or preventing GOM formation, we used LMD to collect GOM-enriched leptomeningeal arteries from two autopsied CADASIL patients, and superficial temporal artery from one biopsied CADASIL patient. The extracted samples were digested as described previously, after which they were subjected to LC-MS/MS. The analyses revealed that, in addition to NOTCH3, the proteins SAP, annexin A2, and periostin, in that order, exhibited the largest

72


Table 4 Proteomic analysis of vascular components by using LMD and LC-MS/MS. No. Accession number

GN

1

P02743

APCS

2 3 4 5 6 7

P07355 Q15063 Q08431 P01024 P04004 Q99715

8 9 10

Q15149 P21810 P06396 Q9UM47

Description

Serum amyloid P component ANXA2 Annexin A2 POSTN Periostin MFGE8 Lactadherin C3 Complement C3 VTN Vitronectin COL12A1 Collagen alpha-1 (XII) chain PLEC Plectin BGN Biglycan GSN Gelsolin NOTCH3 Neurogenic locus notch homolog protein 3

In the absence of NOTCH3, no obvious binding of SAP was observed. No significant interaction was detected in either NOTCH1 (amino acids 19-526) Fc chimera protein or NOTCH2 (amino acids 26-530) Fc chimera protein (Fig. 3B and C).

CADASIL average NSAF

Control average NSAF

0.0152786

N.D.

3.4. RT-PCR of SAP mRNA

0.0046881 0.0035755 0.0081538 0.0005939 0.0063174 0.0007535

N.D. 0.0001936 0.0013153 0.0000996 0.0011800 0.0001466

0.0001584 0.0120185 0.0022669 0.0003609

0.0000322 0.0025856 0.0004967 N.D.

Because anti-human SAP antibody reacted strongly with the vessel walls of CADASIL materials, we suspected that SAP mRNA might be synthesized in the brain. To clarify this issue, we performed RT-PCR of SAP mRNA by using extracted cerebral vessels and liver tissue. Although the SAP primer employed in this experiment detected SAP mRNA in liver tissue, mRNA expression of SAP was below detectable level in the CADASIL samples or the control samples (Suppl. Fig. 1).

Abbreviations: GN: gene name, NSAF: normalized spectral abundance factor, N.D.: not detectable level. Major proteins in the CADASIL samples were analyzed with LC-MS/MS after being processed by using LMD, as described in the text. All proteins (No. 1–10) in this table were only those detected in all three CADASIL patients. Data are expressed as an average NSAF of each protein. Supplementary Table 1 provides a more inclusive list of proteins.

increase. Samples from all three patients showed the same tendency with regard to detection of proteins. Table 4 provides the average NSAF of major proteins as measured by LC-MS/MS and calculated as described earlier. Neither SAP nor annexin A2 and only trace amounts of periostin were detected in materials from control patients. 3.2. Histopathological findings To further characterize the proteins SAP, annexin A2, and periostin, as well as NOTCH3, in the vessels from the CADASIL patients, we performed immunohistochemical staining with anti-human SAP, antihuman annexin A2, anti-human periostin, and anti-human NOTCH3 antibodies. As Fig. 1 illustrates, only anti-human SAP antibody reacted strongly with lesions where the anti-human NOTCH3 antibody staining was positive. Although lesions from the CADASIL patients demonstrated positive results, the same vascular lesions from the control patients showed no positive staining. Because SAP is a plasma protein, the perivascular region in the control patients was weekly positive. Antihuman periostin antibody reacted strongly with proliferative intima and tunica adventitia, but no co-localization with NOTCH3-positive staining lesions was observed. Although annexin A2 in leptomeningeal arteries and in small arteries in the brain parenchyma of CADASIL samples stained slightly, CADASIL and controls samples did not differ significantly. The anti-annexin A2 antibody and the anti-periostin antibody respectively reacted with endothelial cells (Fig. 1Y) and neuronal cells (Fig. 1Z) in which annexin and periostin were detected in a tissue atlas database (The human protein atlas; http://www.proteinatlas.org/). Immunofluorescence staining revealed that NOTCH3 and SAP antibodies co-localized predominantly in the tunica media of the vessels in a manner independent of the vessel diameter (Fig. 2). In contrast, both annexin A2 and periostin immunofluorescence staining revealed no co-localization. In the skin sample from a CADASIL patient (CAD 2 in Table 1), subcutaneous blood vessels had a pattern of NOTCH3 and SAP staining similar to that found in the leptomeningeal arteries. 3.3. Interaction between NOTCH3 and SAP To investigate whether SAP would bind biochemically to NOTCH3, we used solid phase ELISA, as described earlier. In the NOTCH3 (amino acids 40-467) Fc chimera protein-fixed well of a 384-well plate, the added SAP bound to NOTCH3 in a dose-dependent manner (Fig. 3A).

3.5. Serum SAP concentrations in CADASIL patients Because SAP is a serum protein produced by the liver, serum SAP concentrations may be consumed or up-regulated during GOM formation. As Fig. 4 illustrates, serum SAP concentrations of CADASIL patients and healthy controls did not differ statistically. 4. Discussion We demonstrated here, by means of immunohistochemical and immunofluorescence methods, that SAP co-localized with NOTCH3 in the small vessel perivascular areas in the brains of patients with CADASIL (Figs. 1 and 2). In formalin-fixed sections of the brain and superficial temporal arteries, we identified GOM-rich vessels and subjected the digested and dissolved samples to LC-MS/MS. Table 4 lists the three major proteins—SAP, annexin A2, and periostin—that we found. Two proteomics analyses for GOM have been performed previously: Arboleda-Velasquez et al. [15] reported several co-localized proteins, however, NOTCH3, a major component of GOM, was not included as a candidate protein. Monet-Leprêtre et al. [6] also performed proteomic analysis using post-mortem brain tissues. They used whole human brain homogenate from only one patient with CADASIL and one control disease patient. Moreover, they did not use LMD, which could selectively collect GOM rich lesions. In contrast, we could perform the quantitative analyses by means of LMD and LC-MS/MS to focus on GOM rich lesions and obtain the reasonable proteins. The total number of spectral counts and proteins identified in CADASIL and control samples were presented in the attached table (Suppl. Table 1). Although NOTCH3 could not be detected in CAD 1 sample, the protein was present in those of CAD 2 and CAD 3 patients. Proteomic analyses have the limitation to especially aggregated proteins in tissues. However, we believe SAP may play an important role in GOM formation in CADASIL because in all the samples analyzed in our study, the results were consistent: especially SAP was always detected with the largest increase. Immunohistochemical studies demonstrated that only SAP was present around the small vessels from CADASIL patients and that it co-localized clearly with NOTCH3. We observed a similar staining pattern of SAP and NOTCH3 in a skin sample, which is important for understanding the occurrence of GOM in patients with CADASIL. Materials from control patients manifested no such co-localization. To determine the biochemical affinity of SAP for NOTCH3, we used the solid phase ELISA system, after fixing NOTCH3 on the well surface in the 384-well plate. As Fig. 3 illustrates, SAP showed significant affinity for NOTCH3 but with the NOTCH3-free ELISA, SAP showed no obvious affinity for the ELISA plate. These results suggest that NOTCH3, a key molecule in the onset and progression of CADASIL, may have a pathological role together with SAP in CADASIL. To determine the source of the protein found in cerebral vessels of CADASIL patients, we performed RT-PCR of the homogenized samples from CADASIL patients and control patients. We detected no SAP mRNA expression in both CADASIL and control samples. This result clearly indicated that, like the association of SAP with amyloid fibrils


73

Fig. 1. Immunohistochemistry of CADASIL and control arteries. To detect NOTCH3, SAP, annexin A2, and periostin in the perivascular areas of samples from CADASIL patients (CAD) and control patients (Ctrl), paraffin-embedded tissue sections were stained with anti-human NOTCH3, SAP, annexin A2, and periostin antibodies as described in Table 2. Granular deposits of NOTCH3 are observed in all CADASIL samples (arrowheads). (A–H, M–T) Leptomeningeal arteries. (I–L, U–X) Superficial temporal arteries. (Y, Z) Cerebral cortex tissues (positive controls for annexin A2 and periostin). Sections were counterstained with hematoxylin (blue) and analyzed by using bright-field microscopy. Scale bars: 20 μm (A–H), 100 μm.(I–Z).

in senile plaques and brain vessels in patients with Alzheimer disease and cerebral amyloid angiopathy (CAA) [21,22], the origin of SAP in CADASIL patients is likely the liver. We determined serum SAP concentrations in both CADASIL patients and control patients, because SAP is a serum protein and may be consumed or up-regulated during GOM formation in the skin as well as the brain. However, we found no clear difference in serum SAP concentrations in both groups. Although SAP is an acute phase protein in several kinds of rodents, SAP does not have altered serum concentrations even in inflammation and infection in humans. Moreover, the amount of SAP associated with GOM may be small and the plasma half-life of SAP may be too short (t1/2 = 30 h) to compensate for the serum SAP concentrations in CADASIL patients [23].

SAP has been well known as one of the important co-localized proteins of amyloid fibrils [24], in addition to apolipoprotein E, several proteases, cathepsin G, human neutrophil elastase, vitronectin, and clusterin in various amyloidosis such as Alzheimer disease, CAA, familial amyloid polyneuropathy (FAP), and AL amyloidosis. Pepys et al. [25] reported that SAP played an important part in amyloid formation by strongly binding to amyloid fibrils. Hawkins and colleagues also demonstrated that radioisotope-labeled SAP scintigraphy had diagnostic value by binding amyloid fibrils in various types of systemic amyloidosis, such as FAP, AL amyloidosis, and AA amyloidosis [26]. In FAP patients with liver transplantations, the accumulation of radioisotope-labeled SAP decreased in systemic organs [27]. These results clearly indicate that SAP may help stabilize aggregates of proteins such as amyloid. Similar to

74


Fig. 2. Immunofluorescence staining of CADASIL arteries. Immunofluorescence was used to study the co-localization of NOTCH3 with SAP, annexin A2, or periostin. Granular deposits of NOTCH3 are observed in all CADASIL samples (arrowheads). Upper panels (A–I) show specimens of leptomeningeal arteries from CADASIL patient (CAD) 1, and lower panels (J–L) show arteries from CAD 2 skin samples. (A, J) SAP. (D) Annexin A2. (G) Periostin. (B, E, H, K) NOTCH3. (C, F, I, L) Merged images of SAP and NOTCH3 (C, L), annexin A2 and NOTCH3 (F), and periostin and NOTCH3 (I). Arrows indicate NOTCH3 and SAP staining-positive areas. Scale bars: 100 μm (upper panels), 50 μm (lower panels).

Fig. 3. Solid-phase binding assay of SAP with NOTCH proteins. SAP binding with NOTCH3 (A), NOTCH1 (B), and NOTCH2 (C) in each NOTCH protein-fixed well. Data are expressed as means + SD of three independent examinations. NS: not significant. *P b 0.05, **P b 0.01, ***P b 0.001.


Fig. 4. Measurement of serum SAP concentrations. (A) Serum SAP concentrations in CADASIL and control samples were determined by means of SDS-PAGE and Western blotting, as described in the text. Lane numbers correspond to the serum sample numbers. (B) Data are expressed as means + SD. CADSIL and control samples did not differ significantly.

the situation in amyloidosis, SAP may play an important role in stabilizing GOM in CADASIL. Although annexin A2 exhibited no reaction in lesions with positive NOTCH3 staining, periostin was stained in tissues around NOTCH3-positive lesions where fibrotic changes were severe. Periostin has been well known as a fibrosis-promoting protein [28,29], and CADASIL vessels have shown fibrotic changes during progression of the disease [30,31]. Periostin may thus play some role in promoting vascular fibrotic changes in the later stage of CADASIL. Additional studies are needed to clarify this issue. Removal of SAP by using the low-molecular-weight SAP analogue (R)-1-[6-[(R)-2-carboxy-pyrrolidin-1-yl]-6-oxo-hexanoyl]pyrrolidine2-carboxylic acid (CPHPC) was suggested to suppress amyloid fibril formation and to have a therapeutic potential by means of competitive binding with amyloid fibrils [32,33]. In a murine model, CPHPC also had a therapeutic effect on amyloid deposition in CAA [34]. Because the primary damaged lesions in CAA and CADASIL are the small vessels of the brain, an effect of CPHPC on CADASIL may be possible. Moreover, combined administration of anti-human SAP antibody and CPHPC [35] is now undergoing a Phase II study in FAP patients. Combined administration of CPHPC and SAP antibody may also suppress CADASIL progression. Additional information and investigations are needed to determine whether this approach will be effective. In conclusion, SAP may play an important role in GOM formation although precise mechanisms remain to be elucidated. In the future, the effect of SAP should be studied by cross-breeding SAP knockout mice and CADASIL model mice. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.jns.2017.05.033. Conflicts of interest None. Acknowledgments The authors' work was supported by Grants-in-Aid for Scientific Research (B) 15-H04841 and (A) 24249036 and by a grant from the Ministry of Education, Science, Sports, Culture and Technology of Japan, Tokyo, Japan. We thank Ms. H. Katsura and Ms. M. Oka for technical support with the histopathological analyses, and Ms. Judith B. Gandy for providing professional English editing of the manuscript. References [1] E. Tournier-Lasserve, A. Joutel, J. Melki, et al., Cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy maps to chromosome 19q12, Nat. Genet. 3 (1993) 256–259.

75

[2] A. Joutel, C. Corpechot, A. Ducros, et al., Notch3 mutations in CADASIL, a hereditary adult-onset condition causing stroke and dementia, Nature 383 (1996) 707–710. [3] H. Chabriat, A. Joutel, M. Dichgans, E. Tournier-Lasserve, M.G. Bousser, Cadasil, Lancet Neurol. 8 (2009) 643–653. [4] A. Ueda, T. Hirano, K. Takahashi, et al., Detection of granular osmiophilic material of cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy by light microscopy in frozen sections, Neuropathol. Appl. Neurobiol. 35 (2009) 618–622. [5] Y. Yamamoto, L.J. Craggs, A. Watanabe, et al., Brain microvascular accumulation and distribution of the NOTCH3 ectodomain and granular osmiophilic material in CADASIL, J. Neuropathol. Exp. Neurol. 72 (2013) 416–431. [6] M. Monet-Lepretre, I. Haddad, C. Baron-Menguy, et al., Abnormal recruitment of extracellular matrix proteins by excess Notch3ECD: a new pathomechanism in CADASIL, Brain 136 (2013) 1830–1845. [7] C. Capone, E. Cognat, L. Ghezali, et al., Reducing Timp3 or vitronectin ameliorates disease manifestations in CADASIL mice, Ann. Neurol. 79 (2016) 387–403. [8] J. Kast, P. Hanecker, N. Beaufort, et al., Sequestration of latent TGF-β binding protein 1 into CADASIL-related Notch3-ECD deposits, Acta Neuropathol. Commun. 2 (2014) 96. [9] H. Dong, M. Blaivas, M.M. Wang, Bidirectional encroachment of collagen into the tunica media in cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy, Brain Res. 1456 (2012) 64–71. [10] S.J. Lee, X. Zhang, M.M. Wang, Vascular accumulation of the small leucine-rich proteoglycan decorin in CADASIL, Neuroreport 25 (2014) 1059–1063. [11] X. Zhang, S.J. Lee, M.F. Young, M.M. Wang, The small leucine-rich proteoglycan BGN accumulates in CADASIL and binds to NOTCH3, Transl. Stroke Res. 6 (2015) 148–155. [12] L. Craggs, J. Taylor, J.Y. Slade, et al., Clusterin/apolipoprotein J immunoreactivity is associated with white matter damage in cerebral small vessel diseases, Neuropathol. Appl. Neurobiol. 42 (2016) 194–209. [13] M. Tasaki, M. Ueda, K. Obayashi, et al., Effect of age and sex differences on wild-type transthyretin amyloid formation in familial amyloidotic polyneuropathy: a proteomic approach, Int. J. Cardiol. 170 (2013) 69–74. [14] Y. Ando, T. Coelho, J.L. Berk, et al., Guideline of transthyretin-related hereditary amyloidosis for clinicians, Orphanet J. Rare Dis. 8 (2013) 31. [15] J.F. Arboleda-Velasquez, J. Manent, J.H. Lee, et al., Hypomorphic Notch 3 alleles link Notch signaling to ischemic cerebral small-vessel disease, Proc. Natl. Acad. Sci. U. S. A. 108 (2011) E128–E135. [16] A. Ueda, M. Ueda, A. Nagatoshi, et al., Genotypic and phenotypic spectrum of CADASIL in Japan: the experience at a referral center in Kumamoto University from 1997 to 2014, J. Neurol. 262 (2015) 1828–1836. [17] W.A. Hassan, N. Udaka, A. Ueda, Y. Ando, T. Ito, Neoplastic lesions in CADASIL syndrome: report of an autopsied Japanese case, Int. J. Clin. Exp. Pathol. 8 (2015) 7533–7539. [18] E. Uyama, M. Tokunaga, A. Suenaga, et al., Arg133Cys mutation of Notch3 in two unrelated Japanese families with CADASIL, Intern. Med. 39 (2000) 732–737. [19] D. Muta, T. Kawano, N. Shinojima, J. Kuratsu, Superficial temporal artery-to-middle cerebral artery bypass surgery for middle cerebral artery stenosis in a patient with cerebral autosomal-dominant arteriopathy with subcortical infarcts and leukoencephalopathy, Spring 4 (2015) 609. [20] A.C. Boulay, B. Saubamea, X. Decleves, M. Cohen-Salmon, Purification of mouse brain vessels, J. Vis. Exp. (2015), e53208. [21] L.S. Perlmutter, E. Barron, M. Myers, D. Saperia, H.C. Chui, Localization of amyloid P component in human brain: vascular staining patterns and association with Alzheimer's disease lesions, J. Comp. Neurol. 352 (1995) 92–105. [22] R.N. Kalaria, P.G. Galloway, G. Perry, Widespread serum amyloid P immunoreactivity in cortical amyloid deposits and the neurofibrillary pathology of Alzheimer's disease and other degenerative disorders, Neuropathol. Appl. Neurobiol. 17 (1991) 189–201. [23] M.R. Dillingh, B. van den Blink, M. Moerland, et al., Recombinant human serum amyloid P in healthy volunteers and patients with pulmonary fibrosis, Pulm. Pharmacol. Ther. 26 (2013) 672–676. [24] D. Ozawa, R. Nomura, P.P. Mangione, et al., Multifaceted anti-amyloidogenic and pro-amyloidogenic effects of C-reactive protein and serum amyloid P component in vitro, Sci. Rep. 6 (2016) 29077. [25] M.B. Pepys, R.F. Dyck, F.C. de Beer, M. Skinner, A.S. Cohen, Binding of serum amyloid P-component (SAP) by amyloid fibrils, Clin. Exp. Immunol. 38 (1979) 284–293. [26] P.N. Hawkins, J.P. Lavender, M.B. Pepys, Evaluation of systemic amyloidosis by scintigraphy with 123I-labeled serum amyloid P component, N. Engl. J. Med. 323 (1990) 508–513. [27] A. Rydh, O. Suhr, S.O. Hietala, K.R. Ahlstrom, M.B. Pepys, P.N. Hawkins, Serum amyloid P component scintigraphy in familial amyloid polyneuropathy: regression of visceral amyloid following liver transplantation, Eur. J. Nucl. Med. 25 (1998) 709–713. [28] V. Lindner, Q. Wang, B.A. Conley, R.E. Friesel, C.P. Vary, Vascular injury induces expression of periostin: implications for vascular cell differentiation and migration, Arterioscler. Thromb. Vasc. Biol. 25 (2005) 77–83. [29] K. Izuhara, S.J. Conway, B.B. Moore, et al., Roles of periostin in respiratory disorders, Am. J. Respir. Crit. Care Med. 193 (2016) 949–956. [30] H. Dong, H. Ding, K. Young, M. Blaivas, P.J. Christensen, M.M. Wang, Advanced intimal hyperplasia without luminal narrowing of leptomeningeal arteries in CADASIL, Stroke 44 (2013) 1456–1458. [31] R. Okeda, K. Arima, M. Kawai, Arterial changes in cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) in relation to pathogenesis of diffuse myelin loss of cerebral white matter: examination of cerebral medullary arteries by reconstruction of serial sections of an autopsy case, Stroke 33 (2002) 2565–2569.

76


[32] M.B. Pepys, J. Herbert, W.L. Hutchinson, et al., Targeted pharmacological depletion of serum amyloid P component for treatment of human amyloidosis, Nature 417 (2002) 254–259. [33] J.D. Gillmore, G.A. Tennent, W.L. Hutchinson, et al., Sustained pharmacological depletion of serum amyloid P component in patients with systemic amyloidosis, Br. J. Haematol. 148 (2010) 760–767.

[34] R. Al-Shawi, G.A. Tennent, D.J. Millar, et al., Pharmacological removal of serum amyloid P component from intracerebral plaques and cerebrovascular Aβ amyloid deposits in vivo, Open Biol. 6 (2016) 150202. [35] D.B. Richards, L.M. Cookson, A.C. Berges, et al., Therapeutic clearance of amyloid by antibodies to serum amyloid P component, N. Engl. J. Med. 373 (2015) 1106–1114.