Previous studies have shown that lipopolysaccharide-induced inflammation in the lung results in tyrosine nitration. The objective of this study was to evaluate the contribution of myeloperoxidase and peroxynitrite pathway to the tyrosine nitration in lipopolysaccharide-administered lungs of rats that were otherwise untreated or leukocyte-depleted by cyclophosphamide or received inhaled nitric oxide (NO).
The authors analyzed the immunoreactivity of inducible nitric oxide synthase (iNOS), nitrotyrosine (a product of the myeloperoxidase or peroxynitrite pathway), and chlorotyrosine (a byproduct of the myeloperoxidase pathway) by use of specific antibodies. The number of neutrophils in bronchoalveolar lavage fluid (BALF) and levels of myeloperoxidase activity in lung homogenates were also measured.
Lipopolysaccharide enhanced the immunoreactivity of iNOS, nitrotyrosine, and chlorotyrosine in alveolar wall cells, alveolar macrophages, and neutrophils. Leukocyte depletion by cyclophosphamide and inhibition of leukocyte accumulation in the lungs by NO inhalation did not eliminate the increase in iNOS immunoreactivity in alveolar macrophages after lipopolysaccharide treatment, but nitrotyrosine and chlorotyrosine were not produced in these cells. Tyrosine nitration in response to lipopolysaccharide was associated with increases in neutrophil count in BALF and in myeloperoxidase activity in lung homogenates, whereas NO inhalation suppressed the neutrophil count in BALF and reduced tyrosine nitration and chlorination.
These findings suggest that myeloperoxidase pathway has a role in tyrosine nitration in the lungs of lipopolysaccharide-treated rats, and that NO inhalation during early phase of inflammation does not increase but rather decreases tyrosine nitration and chlorination, possibly by reducing neutrophil sequestration.
NITROTYROSINE is a product of the peroxynitrite pathway (fig. 1) that has been used as a probe to detect nitric oxide (NO)-mediated oxidative reactions in vivo 1and has been found in the lungs of patients with the acute respiratory distress syndrome (ARDS). 2–4If nitrotyrosine is produced via the peroxynitrite pathway in the injured lung, inhaled NO will increase the reaction between superoxide and NO to enhance tissue nitration and exacerbate injury. In a previous study of lung injury induced by lipopolysaccharide, however, we found otherwise: inhaled NO (20 ppm) decreased leukocyte accumulation in the lungs and inhibited tyrosine nitration in response to lipopolysaccharide instillation. 5
Recently, it has been indicated that leukocyte myeloperoxidase may contribute to nitrotyrosine formation during the accumulation and activation of these cells. Neutrophil myeloperoxidase can use nitrite and hydrogen peroxide or hypochlorous acid as substrates to catalyze tyrosine nitration in proteins;6this nitration via a myeloperoxidase pathway is enhanced by the addition of NO2−or by flux of NO. 7During this reaction, a part of tyrosine is chlorinated to become chlorotyrosine as a byproduct of the myeloperoxidase pathway (fig. 1). It has been shown that inhaled NO increases the formation of nitrotyrosine and chlorotyrosine in bronchoalveolar lavage fluid (BALF) of patients with ARDS. 2
Our aim of this study was to evaluate the contribution of myeloperoxidase and peroxynitrite pathway to the tyrosine nitration in lipopolysaccharide-administered lungs of rats that were otherwise untreated or leukocyte-depleted by cyclophosphamide or received inhaled NO. We analyzed the immunoreactivity of inducible nitric oxide synthase (iNOS), nitrotyrosine, and chlorotyrosine by use of specific antibodies, and the number of neutrophils in BALF and levels of myeloperoxidase activity in lung homogenates were also measured.
Materials and Methods
General Protocol
Animal care was in accordance with the guidelines of the Animal Care Committee of Kitasato University, and the conduct of these studies conformed to the Guide for the Care and Use of Laboratory Animals published by the National Academy of Science. Male (200–250 g) Sprague-Dawley rats (Clea Japan, Tokyo, Japan) were assigned to one of four groups: untreated controls (control group), lipopolysaccharide-treated (LPS group), leukocyte-depleted and lipopolysaccharide-treated (CPA + LPS group), and lipopolysaccharide-treated + NO inhalation (LPS + NO group). There were no deaths associated with lipopolysaccharide treatment, leukocyte depletion, or NO exposure.
Lipopolysaccharide Administration
During halothane anesthesia, 100 μl of endotoxin-free saline (control) or 10 μg lipopolysaccharide (Escherichia coli 0127:B8, phenol extract, L-3129, Sigma, St. Louis, MO) in 100 μl saline was sprayed into the trachea close to the bifurcation using a miniaturized nozzle (Model IA-1b, Penn-Century, Philadelphia, PA) to optimize the dose and assure uniform distribution of the solution throughout the lung. 8,9After treatment, the rats were housed in a filtered-air chamber for 6 h. Following anesthesia (intraperitoneal sodium pentobarbital, 50 mg/kg body weight), the abdominal aorta was severed. We chose to study a 6-h period after saline- or lipopolysaccharide-treatment based on a preliminary study showing that after lipopolysaccharide treatment, the leukocyte count in BALF (see below) and myeloperoxidase activity in lung homogenates (see below) had increased over the baseline values (3.00 ± 0.71 × 106, mean ± SEM, per half lung for neutrophil count in BALF of five animals, and 0.121 ± 0.015, mean ± SEM, optical density unit/g protein for myeloperoxidase activity in lung homogenate of five animals) at 3 h (7.30 ± 1.71 × 106and 0.535 ± 0.086, respectively, for n = 5), peaked at 6 h (7.96 ± 3.14 × 106and 1.42 ± 0.096, respectively, for n = 5), and decreased by 18 h (4.81 ± 9.85 × 106and 0.61 ± 0.12, respectively, for n = 5). Similarly in the human lung after bronchial endotoxin instillation, a high neutrophil count associated with elevated cytokines and chemokines in BALF was found after 6 h, whereas the pulmonary alveolar macrophage (PAM) count did not increase until between 24 and 48 h. 10
Leukocyte Depletion
Rats received intraperitoneal cyclophosphamide (100 mg/kg, Cytoxan, Bristol-Myers, Syracuse, NY) 6 days before lipopolysaccharide treatment and on the day of the lipopolysaccharide treatment (50 mg/kg). 5
Nitric Oxide Inhalation
Rats inhaled 20 ppm NO for 6 h after exposure to lipopolysaccharide. A recent study has demonstrated that this concentration reduces pulmonary inflammation. 5NO was supplied from a nitrogen-balanced NO gas cylinder (20,000 ppm ultrahigh-pure grade, Sumitomo Seika Chem. Co., Osaka, Japan) through thermal mass-flow controllers (SEC-4400, ESTEC, Kyoto, Japan), and the NO gas was diluted with filtered air (25°C, relative humidity 50–55%) just before the exposure chamber (20 l) to obtain a concentration of 20 ppm NO. The flow rate in each chamber was set at 10 l/min. Three or four animals were housed per chamber. The NO and NO2levels were monitored by electrochemical sensors (TM-100 and TM-1002, respectively, Saan, Osaka, Japan), and the NO2level in the chamber was found to always be less than 1.0 ppm. Only filtered 10-l/min air was introduced into the chamber of the animals not exposed to NO inhalation.
Tissue Fixation and Immunohistochemistry (iNOS, Nitrotyrosine, and Chlorotyrosine)
Three rats were studied in each group. A polyclonal antibody was prepared in collaboration with Nippon Biotest Laboratories, Tokyo, Japan, by injecting rabbits with chlorotyrosine haptens using bovine serum albumin (BSA) as a carrier protein and with Freund complete adjuvant. Chlorotyrosine was conjugated with BSA by using N-(6-maleimidocaproyloxy) succinimide (Dojindo, Kumamoto, Japan) as a bridge between amino-residue of chlorotyrosine and the SH-residue of BSA. 11The antibody titer in the immunized rabbit's serum was checked as follows. The serum (diluted to 1:1 × 102to 2,048 × 102) was incubated for 60 min at 37°C in micro–enzyme-linked immunosorbent assay (ELISA) plates (Falcon 3912 Assay Plate, Becton Dickinson, Franklin Lakes, NJ) coated with chlorotyrosine. The plates were washed in phosphate-buffered solution (PBS) (× 4), and 50 μl alkaline phosphate-conjugated goat F(ab′)2antirabbit IgG (Biosource Technical Service, Sunnyvale, CA) diluted 1:1,000 in PBS was placed in each well; the wells incubated at 37°C. They were then washed with PBS (× 5) and incubated at 37°C with 200 μl of the phosphatase substrate in PBS (pH 9.5) to develop. Absorbance at 405 nm was measured on a microplate reader (EL307, BioTek Instruments, Winooski, VT). The titer was checked weekly, and after obtaining the plateau titer, all the serum of the immunized animal was obtained and purified through a BSA column (Amersham Pharmacia Biotech, Buckinghamshire, England). No cross-reaction with BSA was observed in the immunoprecipitation method after the purification of the antibody. A polyclonal antibody to iNOS and its blocking peptide and a polyclonal antibody to nitrotyrosine were obtained from commercial sources (Biomol Research Laboratories Inc., Plymouth Meeting, PA, and Upstate Biotechnology, Charlottesville, VA, respectively).
Both lungs were fixed with 4% paraformaldehyde via the trachea at 23 cm H2O 12and embedded in paraffin wax. De-paraffinized 4-μm sections were incubated overnight at 4°C with the primary antibody (anti-iNOS, antinitrotyrosine, or antichlorotyrosine antibody) diluted to 1:100. The sections were treated with a secondary antibody (biotinylated goat antirabbit IgG) and an enzyme complex (Histostain SP kit, Zymed Laboratories Inc., South San Francisco, CA); reactive sites were visualized with aminoethylcarbazole (producing red reaction product), and the sections were counterstained with hematoxylin.
Although alveolar type 1 and type 2 cells form the alveolar surface, it was not possible to clearly differentiate these two cell types in our (4-μm) sections; for this reason, in this study, we refer to them as “alveolar wall cells.”
Quantitative Analysis of the Counts of Immunoreactive Neutrophils and the Degree of Immunostaining in Lung Sections
The numbers of neutrophils and the numbers of neutrophils immunoreactive for nitrotyrosine and chlorotyrosine were counted in three randomly chosen 50-μm square areas of lung sections from three animals in each group (i.e. , nine optical fields in each group).
A quantitative analysis of the degree of immunostaining was also performed. The immunostaining of nitrotyrosine and chlorotyrosine in three randomly chosen 50-μm square areas of lung sections from three animals in each group (i.e. , nine images in each group) was captured and saved as a RGB digital image with 8-bit density resolution in each color with image software (Adobe Photoshop v. 5.5, Adobe Systems Incorporated, San Jose, CA), and the mean density of 334,572-pixel red image of each RGB image was measured using NIH image (Scion Image v. 1.62, Scion Corporation, Frederick, MD) because the reactive sites were visualized with red reaction product.
Validation of Antibody Specificity
The iNOS antibody does not cross-react with neuronal (NOS I) or endothelial NOS (NOS III) by Western analysis (see Manufacturers Data Sheets). Preadsorption experiments confirmed its specificity for the NOS isoform, i.e. , no positive reaction sites were detected after competitive binding of the iNOS antibody with 20 mm of peptide antigen (Biomol Research Laboratories, Inc., Plymouth Meeting, PA).
Negative controls for nitrotyrosine and chlorotyrosine staining were (1) antibody preadsorption with authentic nitrotyrosine or chlorotyrosine (20 mm) and (2) antibody substitution with nonimmune IgG. To exclude the possibility of (artifactual) tyrosine nitration during quenching of endogenous peroxidase activity by H2O2in the presence of intrinsic NO2−, 13tissue sections were flooded with three washes (20 s each) of 1 m sodium hydrosulfite (adjusted to pH 9.5 with 2 N NaOH), and the sections were quenched (3% H2O2) and stained for nitrotyrosine as described above.
To provide positive controls for antinitrotyrosine staining, sections were quenched (3% H2O2), washed in PBS (× 3), and incubated in 1 mm peroxynitrite (Upstate Biotechnology, Charlottesville, VA) for 2 h at 37°C; for antichlorotyrosine staining, sections were incubated in PBS with 150 mm chloride ion and then in 500 μm HOCl for 1 h to generate chlorotyrosine.
To exclude cross-reactivity between chlorotyrosine and the antinitrotyrosine antibody and between nitrotyrosine and the antichlorotyrosine antibody, the antinitrotyrosine antibody was incubated with 20 mm authentic chlorotyrosine and used on the nitrotyrosine-rich tissue treated with peroxynitrite; the antichlorotyrosine antibody was incubated with 20 mm authentic nitrotyrosine and used in the chlorotyrosine-rich tissue treated with HOCl.
Bronchoalveolar Lavage
Ten animals were studied in each group. The left main bronchus was ligated, and the right lung lavaged (× 5) with endotoxin-free saline (Otsuka Pharm, Naruto, Japan). Each wash volume was 0.0175 (ml/g) multiplied by body weight (g). The recovered amount was always more than 90%. A nucleated cell count and differential cell count were performed, and the nitrite–nitrate content was measured by the Griess method using an automated analyzer (TCI-NOx 1000 and S-3200, Tokyo Kasei Kogyo, Tokyo, Japan).
Myeloperoxidase Activity in the Lung
Ten animals were studied in each group. Myeloperoxidase activity in left lung homogenates was assayed as described by Hirano 14at an optical density of 460 nm (OD460). Myeloperoxidase activity was then normalized to protein content, which was measured using a bicinchoninic acid protein kit according to the manufacturer's instruction (Pierce, Rockford, IL).
Statistical Analysis
A one-way analysis of variance (ANOVA) with Scheffépost hoc test was used to detect the difference between groups. P values less than 0.05 were considered significant.
Results
Neutrophil Number in Blood
After lipopolysaccharide treatment, the number of neutrophils circulating in peripheral blood was significantly increased above the control value (P < 0.001), and NO inhalation did not alter this response. Leukocyte depletion decreased the number of neutrophils circulating in response to lipopolysaccharide (compared with lipopolysaccharide treated group, P < 0.001;table 1).
Validation of Antibody Specificity
The negative controls for nitrotyrosine and chlorotyrosine staining, i.e. , (1) antibody preadsorption with authentic nitrotyrosine or chlorotyrosine (20 mm;fig. 2) and (2) antibody substitution by nonimmune IgG, did not stain. There was no (artifactual) tyrosine nitration during quenching of endogenous peroxidase activity by H2O2in the presence of intrinsic nitrite.
Positive Controls for Nitrotyrosine and Chlorotyrosine Staining
No cross-reactivity between chlorotyrosine and the antinitrotyrosine antibody or between nitrotyrosine and the antichlorotyrosine antibody was observed as a result of nitrotyrosine staining after preadsorption of antinitrotyrosine antibody with authentic chlorotyrosine or as a result of chlorotyrosine staining after preadsorption of antichlorotyrosine antibody with authentic nitrotyrosine (fig. 2).
iNOS, Nitrotyrosine, and Chlorotyrosine in Lung Cells
In the lungs of rats in the untreated control group, iNOS protein was weakly expressed by some alveolar wall cells but not by PAMs or by the rare neutrophils present. Lipopolysaccharide treatment enhanced the iNOS immunoreactivity of alveolar wall cells, PAMs, and neutrophils. Most of the large numbers of neutrophils present were immunopositive. Leukocyte depletion and NO inhalation decreased neutrophil accumulation in the lungs of lipopolysaccharide-treated rats. In these animals, iNOS immunoreactivity was present in PAMs and alveolar wall cells (figs. 2 and 3).
In the control group, no nitrotyrosine-positive cells were detected. In the lungs of lipopolysaccharide-treated rats, however, many positive alveolar wall cells, PAMs, and neutrophils were present. After leukocyte depletion and NO inhalation, no nitrotyrosine-positive cells were detected (figs. 2 and 3).
Chlorotyrosine immunoreactivity was similar to that of nitrotyrosine. In control rats, no chlorotyrosine-positive cells were detected, and in lipopolysaccharide-treated rats, many chlorotyrosine-positive alveolar wall cells, PAMs, and neutrophils were present. After leukocyte depletion and after NO inhalation, no chlorotyrosine-positive cells were detected in the lungs of rats treated with lipopolysaccharide (figs. 2 and 3).
Quantitative analysis of tissue sections also showed that lipopolysaccharide exposure resulted in neutrophil accumulation with nitrotyrosine-positive and chlorotyrosine-positive neutrophils and that the photometric red-color density of immunostained images of lung sections had increased, whereas leukocyte depletion and NO inhalation had decreased neutrophil accumulation and red-color density in the lungs of lipopolysaccharide-exposed rats (table 2).
Neutrophil and PAM Number and Nitrite–Nitrate Levels in BALF
Lipopolysaccharide treatment increased the neutrophil count (P < 0.001) in BALF, whereas leukocyte depletion prevented this increase and NO inhalation suppressed it (P < 0.05). The number of PAMs did not change after lipopolysaccharide treatment, leukocyte depletion, or NO inhalation. Nitrite–nitrate levels increased after lipopolysaccharide treatment (P < 0.05), were unchanged by leukocyte depletion, and were further increased by NO inhalation (fig. 4).
Myeloperoxidase Activity in Lung Homogenate
Lipopolysaccharide treatment increased myeloperoxidase activity in lung homogenates (P < 0.001). This activity was decreased by leukocyte depletion (P < 0.001) and suppressed by NO inhalation (P < 0.05;fig. 5).
Discussion
The present study shows that instillation of lipopolysaccharide via a nozzle enhances iNOS immunoreactivity and the production of nitrotyrosine and chlorotyrosine in lung cells and is associated with high numbers of neutrophils in blood and BALF and an increase in myeloperoxidase activity in lung tissue. It further demonstrates that neither leukocyte depletion nor NO inhalation influences iNOS immunoreactivity in PAMs and alveolar wall cells in response to lipopolysaccharide treatment, but that each eliminates nitrotyrosine and chlorotyrosine production in lung cells and suppresses the influx of neutrophils into BALF and lung myeloperoxidase activity. These data indicate that neutrophil-derived myeloperoxidase contributes to tissue nitration in lipopolysaccharide-induced lung injury, and that during the early phase of inflammation, when neutrophils sequester and infiltrate the lung, inhaled NO has an inhibitory effect on the nitration response.
Lipopolysaccharide-induced Tissue Nitration
Clearly, a reduced number of sequestrated neutrophils in the lung results in reduced levels of iNOS and myeloperoxidase generation. Although neutrophil depletion reduces levels of tissue nitration and chlorination in response to lipopolysaccharide via reduced neutrophil-derived myeloperoxidase activity and iNOS expression (fig. 6), our finding of the continued expression of lipopolysaccharide-induced iNOS in PAMs suggests that, combined with superoxide production, this could be a source of peroxynitrite generation. While cyclophosphamide treatment has been shown to stimulate superoxide production in rat lungs in vivo , 15in our study, the lack of nitrotyrosine production in PAMs in neutrophil-depleted rats indicated that the contribution of these cells to this pathway and to tissue nitration was minor. Cyclophosphamide is known to decrease NO production by PAMs and nitrotyrosine formation in mice 72 h after infection with Mycoplasma pulmonis , 16but in our study, the nitrite–nitrate levels in BALF increased 6 h after lipopolysaccharide treatment and were unchanged after leukocyte depletion by cyclophosphamide, suggesting that NO production by lung cells was not significantly influenced by cyclophosphamide treatment in our study. However, in this study, the possibility that the cytotoxicity of cyclophosphamide may also have modified enzyme function, resulting in reduced production of nitrotyrosine and chlorotyrosine, cannot be ruled out.
Similar to cyclophosphamide, inhaled NO inhibits leukocyte sequestration and reduces lung damage in animals administered lipopolysaccharide either intravenously or by inhalation, 5,17enhancing leukocyte deformability and inhibiting the expression of adhesion molecules by pulmonary vascular endothelium and leukocytes. 18An increase in NO levels in tissue can be expected in response to NO inhalation and is indicated in the present study by our finding of an increased content of nitrite–nitrate in BALF. Despite any enhancement by inhaled NO on the reaction between NO and superoxide that may have occurred in the lipopolysaccharide-injured lung, we found that it did not result in nitrotyrosine production. It appears that the inhibitory effect of NO on neutrophil sequestration in the lung supervened its effects in enhancing tissue nitration via PAMs and the peroxynitrite pathway (fig. 6).
Myeloperoxidase Pathway in Tyrosine Nitration
A number of peroxidases, such as lactoperoxidase in airway cells 13and eosinophil peroxidase, are found in the lung, in addition to myeloperoxidase. Loss of nitrotyrosine and chlorotyrosine production in lung cells after leukocyte depletion suggests, however, that among the possible peroxidases, the contribution of myeloperoxidase in lipopolysaccharide-induced lung injury is significant. Our finding of nitrotyrosine and chlorotyrosine generation at the intracellular level demonstrates that the neutrophils and the myeloperoxidase pathway play a role in tissue nitration. We were able to evaluate the contribution of this pathway by the use of a specific antibody to chlorotyrosine, which was raised for this study. To our knowledge, this is the first report of the use of such an antibody to dissect the contribution of the myeloperoxidase pathway to tissue nitration.
Although analysis of lung homogenates demonstrated an increase in myeloperoxidase activity, ideally our data for chlorotyrosine in lung cells would be supported by quantitative measurement of chlorotyrosine levels by high-performance liquid chromatography (HPLC) 2or by gas chromatography combined with mass spectrometry. 19
The cellular pathways of nitrotyrosine production via the myeloperoxidase pathway will be further clarified by administering a selective myeloperoxidase inhibitor or by using genetically manipulated mice. However, because any inhibitor has nonspecific effects and it is likely that other radical systems, such as the peroxynitrite pathway, would be activated to compensate for loss of the myeloperoxidase system in genetically manipulated animals, these studies await the availability of conditional knockout mice to disclose the relative role of myeloperoxidase in tyrosine nitration in the injured lung.
Increased expression of iNOS in PAMs in response to lipopolysaccharide indicates a further cellular source for nitrotyrosine production, i.e. , NO for peroxynitrite pathway and nitrite for myeloperoxidase pathway. Although this expression persisted after leukocyte depletion and NO inhalation, nitrotyrosine and chlorotyrosine production in PAMs was suppressed. It is possible that in response to lipopolysaccharide, nitrotyrosine and chlorotyrosine were generated directly via the myeloperoxidase pathway in these cells and that they contained myeloperoxidase released by neutrophils undergoing phagocytosis. Alternatively, myeloperoxidase derived from neutrophils in the tissue could have indirectly induced nitrotyrosine and chlorotyrosine production in these cells. We cannot exclude the possibility that cyclophosphamide induced a change in the population or function of type 2 cells, 15but lipopolysaccharide treatment induced iNOS in alveolar macrophages in the presence of cyclophosphamide or inhaled NO.
Clinical Relevance of NO Inhalation
Clinically, inhaled NO can be expected to decrease the number of sequestered leukocytes in the lung and to reduce the myeloperoxidase activity, thereby reducing tyrosine nitration and chlorination. This beneficial effect can offset tyrosine nitration and chlorination in the injured lung that can alter protein function 20,21and affect the activity of the enzymes that have tyrosine at their active site, 22–24resulting in enzyme malfunction.
Nitric oxide inhalation, however, likely has contradictory effects in lung, one being to increase tissue nitration as a result of increased amounts of NO and nitrite, thereby enhancing the production of peroxynitrite and the reaction via peroxidase as a source of nitrite, 2and the other being to decrease nitration by inhibiting leukocyte accumulation in the lungs, thereby reducing the production of NO and superoxide (i.e. , peroxynitrite) and the amount of myeloperoxidase. 5,17,18
In a previous study in which we measured nitrotyrosine quantitatively by HPLC, we confirmed that inhaled NO decreases the production of nitrotyrosine. 5In this study, we further confirmed that inhaled NO eliminated nitrotyrosine together with chlorotyrosine in the lungs of lipopolysaccharide-treated rats, indicating that the myeloperoxidase pathway mainly contributed to the production of nitrotyrosine, and, because the inhibitory effect of inhaled NO on leukocyte accumulation in the lung supervened the effect of NO delivery to the lung tissue, that NO inhalation did not enhance the reaction of myeloperoxidase-dependent tyrosine nitration.
Instillation of lipopolysaccharide was used as a model of clinical lung injury caused by severe gram-negative bacterial pneumonia or by inhalation of massive amounts of environmental endotoxin; however, the lung injury in clinical settings is not so simple as this animal model, and the possible adverse effect of inhaled NO during the later phase of inflammation remains to be evaluated. Nevertheless, our data indicate that inhaled NO has an inhibitory effect on nitration and chlorination of cellular components during the early phase of inflammation caused by intrapulmonary lipopolysaccharide, possibly by reducing neutrophil sequestration in the lungs.
The authors thank Tomoyuki Tomita, M.D., Ph.D. (Professor Emeritus, Department of Medicine, Kitasato University School of Medicine, Sagamihara, Kanagawa, Japan) for his advice and encouragement in their study. The authors also thank Hisashi Mitsufuji, M.D., Ph.D., Kenya Honda, M.D., Ph.D., and Masumi Tanaka, Technician, (Department of Medicine, Kitasato University School of Medicine, Sagamihara, Kanagawa, Japan) for their excellent technical support.