| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Departments of 1Pharmacology and Molecular Therapeutics, 2Immunogenesis, 3Tumor Genetics and Biology, and 4Immunology, Graduate School of Medical Sciences, Kumamoto University, Kumamoto, Japan; 5Molecular Medical Center for Cardiovascular Diseases, Cardiovascular Institute and Fuwai Hospital, Peking Union Medical College and Chinese Academy of Medical Sciences, Beijing; 6The 2nd Hospital and Institute of Cardiovascular Disease Research Affiliated to Nanchang University, Donghu, Nanchang, People's Republic of China; 7Department of Internal Medicine, Sendai Shakaihoken Hospital, Sendai; 8Departments of Cardiovascular Medicine and Clinical Bioinformatics, University of Tokyo Graduate School of Medical Sciences, Tokyo; and 9Department of Cardiovascular Science and Medicine, Chiba University Graduate School of Medical Sciences, Chiba, Japan
Submitted 11 March 2008 ; accepted in final form 5 June 2008
 |
ABSTRACT |
|---|
 TOP ABSTRACT EXPERIMENTAL PROCEDURESRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
-actinin-positive cells. TNNI3K improved cardiac function by enhancing beating frequency and increasing the contractile force and epinephrine response of spontaneous action potentials without an increase of the single-cell size. TNNI3K suppressed phosphorylation of cardiac troponin I, annexin-V+ cells, Bax protein, and p38/JNK-mediated apoptosis. Intramyocardial administration of TNNI3K-overexpressing P19CL6 cells in mice with myocardial infarction improved cardiac performance and attenuated ventricular remodeling compared with injection of wild-type P19CL6 cells. In conclusion, our study clearly indicates that TNNI3K promotes cardiomyogenesis, enhances cardiac performance, and protects the myocardium from ischemic injury by suppressing p38/JNK-mediated apoptosis. Therefore, modulation of TNNI3K activity would be a useful therapeutic approach for ischemic cardiac disease.
cardiomyogenesis; apoptosis
The domain structure of TNNI3K is homologous to that of integrin-linked kinase (ILK), a protein serine/threonine kinase (8) that interacts with the cytoplasmic domains of integrin β1- and β3-subunits, thereby linking cell-matrix interactions to signals regulating cytoskeletal remodeling and cellular growth, proliferation, survival, and differentiation (7, 8). ILK is localized in cytoplasm and links integrin-mediated signaling to downstream pathways involved in the suppression of apoptosis, promotion of cell cycle progression (4, 5), and cardiac remodeling (28, 29), whereas TNNI3K is distributed in both the cytoplasm and nucleus (20, 34) of cardiac myocytes. Because TNNI3K shares more than 50% homological structural similarity with ILK (Fig. 1A), it also likely may possess similar biological functions, including the mediation of integrin-mediated signaling inducing cellular growth, proliferation, survival, and differentiation of cardiomyocytes.
|
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
P19CL6-derived cardiomyocytes with overexpression of TNNI3K.
TNNI3K was cloned from an adult heart cDNA library, and the vector pcDNA6-flag/TNNI3K was obtained and amplified in the Eco line, as described previously (34). P19CL6 cells were cultured as described previously (6). In brief, the cell were grown in a 100-mm tissue culture-grade dish under adherent conditions with -minimal essential medium (GIBCO BRL) supplemented with 10% fetal bovine serum, 100 U/ml penicillin, and 100 µl/ml streptomycin (growth medium) and were maintained in a 5% CO2 atmosphere at 37°C. To achieve stable transformations incorporating TNNI3K vs. the vector control, the vectors pcDNA6-flag (flag-only) and pcDNA6-flag/TNNI3K were transferred into P19CL6 cells with the use of an electropore 2000 instrument (Eurogentec, Liege, Belgium) and were maintained in -MEM containing 500 µg/ml blasticidin (Invitrogen). After 30 days, 28 clones with stable overexpression of TNNI3K were identified using Western blotting methods. To induce differentiation under adherent conditions, we plated P19CL6 cells at a density of 4 x 105 cells in a 60-mm bacterial culture-grade dish with growth medium containing an inducer, 1% dimethyl sulfoxide (DMSO), for 4 days. The medium was changed every 2 days. Days of differentiation were numbered consecutively after the first day of the DMSO treatment (day 0). The morphology of beating masses was investigated under an inverted microscope (x200 magnification, total of 20 view fields), and the beating frequency was calculated. Spontaneous beating activities of single cardiomyocytes isolated using collagenase were investigated with the whole cell patch-clamp technique (15, 17) under current-clamp conditions (EPC-7; List Electronic, Darmstadt, Germany). The contraction force of beating masses was investigated using computer software to collect and analyze the frequency and amplitude of optical signals obtained from video recordings in the beating masses. The shortening length of the beating mass was calculated by measuring the mass diameter difference obtained from the mass longitudinal axis between the maximal contraction period and dilatation period.
Immunohistochemistry and Western blotting.
Morphological studies for differentiation of P19CL6 cells to cardiomyocytes were assessed by expression of -actinin, which was detected using an indirect immunofluorescence method with monoclonal antibodies against -actinin (Sigma) followed by confocal microscopic imaging. In brief, cells were isolated enzymatically from culture wells with beating masses at day 16 after exposure to the induction agent, centrifuged, resuspended, divided into 10 small wells (35 mm), and recultured for 24 h. The cells were fixed in a 35-mm dish with acetone-methanol (1:1) for 2 min, washed, and then blocked with 3% BSA in PBS for 15 min. Primary antibody (anti--actinin antibodies diluted with 0.2% Triton X-PBS) was then added for 1 h, and then secondary antibody (FITC-conjugated goat anti-mouse IgG) diluted with 0.2% BSA-PBS was added. After cells were washed and fixed with 2.5% DABCO (1,4-diazabicyclo[2.2.2]octane)-80% glycerol, confocal microscopic images were obtained and analyzed (Fluoview FV300; Olympus). To detect phosphorylation of p38, ERK1/2, and JNK in the cells, Western blotting was performed as described previously (12) using antibodies to detect phosphorylated proteins. The expression levels of phosphorylated proteins were calibrated using β-actin expression as an internal standard. Antibody against phosphorylation of cTnI (5E6; Fitzgerald) and anti-total cTnI isoform antibody (P14-14G5; HyTest, Turku, Finland) were used to detect phosphorylated cTnI.
Intracellular Ca2+ concentration. Measurement of the intracellular Ca2+ concentration ([Ca2+]i) with fura-2 AM (Dojindo Laboratories, Kumamoto, Japan) in single cells isolated using collagenase (1.5 mg/ml) was performed with a video imaging system (ACOUScosmos; Hamamatsu Photonics, Hamamatsu, Japan) as described previously (3).
Experiment II
Myocardial infarction model in mice.
Thirty-two male C57BL6/J mice (Charles River Laboratories Japan), weighing 27.0–32.0 g at 10–12 wk of age received humane care in compliance with the "Role of the Animal Experimentation Committee, Kumamoto University, Graduate School of Medical Sciences" (approval no. C17-036). Myocardial infarction (MI) was induced by ligation of the left anterior descending branch of the coronary artery (LAD) at 
2 mm distant from its origin with a 8-0 polypropylene suture as previously described (10, 16, 30). Ten minutes later, surviving animals were randomly assigned to three experimental groups to receive intramyocardial injection at four points, with 20 µl per point, in the border zone surrounding the infarcted area of the medium-only group (n = 12), flag-only transfected cells induced 4 days in 1% DMSO at 4 x 105 cells/ml (n = 8), or TNNI3K-overexpressing P19CL6 cells induced 4 days at 4 x 105 cells/ml (n = 6). The other six animals received sham operation as a control group. After completion of all protocols, the chest was closed and the animals were fed for further experiments.
Transthoracic echocardiographic studies. Transthoracic echocardiography was performed with an echocardiographic system equipped with a 12.0-MHz phase-array transducer (SONOS-4500; Philips) as described previously (30).
Histological studies. Animals were killed at 2 wk after MI for histological studies. Infarcted tissues were positively stained with 1% triphenyltetrazolium chloride, and the area at risk (AAR) was determined as described previously (9, 23). Histological studies were performed using hematoxylin and eosin staining under a light microscope.
Measurement of Apoptotic Cells
Apoptotic cells in P19CL6-derived cells were detected at 5, 10, 15, and 20 days after induction with DMSO. To amplify the apoptotic response of the cells, we performed heat treatment or UV irradiation before analysis for FACScan and Bax protein, respectively. In brief, the beating masses were harvested from the well, the single cells were isolated by trypsinization and washed, and the cells were recultured for 2 h. The cells were then incubated in a water bath at 45°C for 90 min or exposed to UV irradiation for 20 s and stained with FITC-conjugated annexin V (Pharmingen). The percentage of annexin V-positive cells was determined using FACScan analysis by counting fluorescent cell numbers per 10,000 cells in each sample. In infarcted tissues, apoptotic cells were determined using terminal deoxynucleotide transferase-mediated dUTP nick-end labeling (TUNEL) assays with an in situ apoptosis detection kit (MK500; Takara Bio, Tokyo, Japan). The number of TUNEL-positive cardiomyocytes was counted in eight fields of the border zone at x400 magnification.
Reagents and Statistics
All salts used to prepare solutions and anti--actinin antibodies were purchased from Sigma (St. Louis MO). Data are means ± SD. Statistical significance was determined by ANOVA and Student's t-test for paired or unpaired data. P values <0.05 were considered to be statistically significant.
|
RESULTS |
|---|
|
TOPABSTRACTEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
P19CL6 cells transfected with pcDNA6-TNNI3K plasmid showed stable overexpression of TNNI3K, whereas cells transfected with the flag-only plasmid did not (Fig. 1B). Cardiomyocyte differentiation and cardiomyogenesis were investigated in a P19CL6 cell model by adding an inducer (1% DMSO). After 4 days of culture with the inducer, both the TNNI3K-overexpressing P19CL6 cells and the flag-only transfected cells formed embryoid bodies, and certain numbers of cells became beating masses displaying beating activity at days 8–15 after induction was initiated. There was almost no difference in the number of wells with beating cells after induction, as shown in Fig. 2Ba, between the TNNI3K-overexpressing group (30/32 wells, 93.8%) and the flag-only group (33/35 wells, 94.3%). However, the number of total beating masses was significantly larger in the TNNI3K-overexpressing group than in the flag-only group (Fig. 2Bb, P < 0.05). This quantitative difference in beating masses was observed through almost all stages of differentiation, when both groups showed a culture time-dependent increase in the number of beating masses from day 8 to day 24 after induction.
|
Furthermore, our data also showed that the contractile force and frequency of beating cells in the TNNI3K group were respectively stronger and higher than in the flag-only group (Fig. 3A). Since our previous study showed that TNNI3K might interact with cTnI to affect the contractile force (34), we examined phosphorylation of cTnI using Western blotting. Expression of phosphorylated cTnI was less in TNNI3K-overexpressing cells than in the flag-only transfected cells, suggesting that suppressive effects of TNNI3K protein on cTnI phosphorylation occurred (Fig. 3B).
|
Spontaneous action potentials (SAPs) were recorded in the single cardiomyocytes that were enzymatically isolated from beating masses of flag-only or TNNI3K-overexpressing cells using current-clamped patch-clamp techniques. The pattern of SAPs in cardiomyocytes differentiated from TNNI3K-overexpressing P19CL6 cells was similar to that in cells differentiated from flag-only transfected P19CL6 cells, except the slope rate in phase 4 of SAPs in the TNNI3K group was steeper than that in the flag-only group (Fig. 3A). The action potential duration at 90% repolarization in the TNNI3K group (164.2 ± 9.7 ms) was longer than that in the flag-only group (156.1 ± 2.5 ms, n = 6, P < 0.05). On the other hand, as shown in Fig. 3C, epinephrine at 10–8–10–6M concentrations dose-dependently increased the beating rate in both groups, whereas more elevation in beating rates was observed at concentrations of 10–8 and 10–7M in the TNNI3K group, indicating that TNNI3K increased the epinephrine-response of SAPs (Fig. 3C).
Caffeine(10 mM) transiently increased [Ca2+]i in both the flag-only and TNNI3K-overexpressing P19CL6 cells (Fig. 3D), but the amplitude of the Ca2+-response was higher in the TNNI3K-overexpressing myocytes than in flag-only myocytes (P < 0.05). Similar results were obtained when ryanodine (20 µM) was added. On the other hand, KCl (40 mM) triggered a sustained elevation of [Ca2+]i in both groups. Similar results were obtained when L-type Ca2+ currents were measured with voltage-clamped patch-clamp techniques; there was no significant difference between the two groups (P > 0.05; data not shown).
TNNI3K Increases the -Actinin-Positive Cell Population
Figure 4A shows representative single cardiomyocytes isolated from beating masses of the flag-only and TNNI3K-overexpressing groups at day 16 after differentiation induction. There were no significant differences in cell size and cell area of -actinin-positive single cardiomyocytes between the two groups (P > 0.05). However, the number of -actinin-positive cells increased in a time-dependent manner in both the flag-only and TNNI3K-overexpressing groups but was significantly higher in the TNNI3K-overexpressing group than in the flag-only group from day 8 through day 24 (P < 0.05; Fig. 4B). Furthermore, total -actinin expression levels in the TNNI3K-overexpressing cells were also higher than in the flag-only cells from day 4 through day 24 (Fig. 4C). The increased protein observed on the Western blot is likely due to the increased number of -actinin-positive cells but not to an increase in protein at the single-cell level.
|
When differentiation of flag-only and TNNI3K-overexpressing P19CL6 cells was induced with 1% DMSO in increased density cultures at 1, 1.5, 2, 2.5 and 3.0 times the original culture density (4.0 x 105 cells/well), total mass numbers, including beating and nonbeating masses, at 12, 16, 20, and 24 days were significantly higher in the TNNI3K-overexpressing group than in the flag-only group (Fig. 5A). Furthermore, in the flag-only group, when the original cell densities were >10 x 105 cells/well, mass formation was undetectable. However, in the TNNI3K-overexpressing group, beating masses were detected in the cell densities examined.
|
TNNI3K Improves Left Ventricular Performance in a MI Model
Vehicle medium-only, flag-only, or TNNI3K-overexpressing P19CL6 cells at day 4 after induction were intramyocardially injected into mice suffering from MI by LAD ligation. Transthoracic ultrasound cardiography before MI (control) and 2 wk after MI (Fig. 6, B and C) showed that impaired ejection fraction and left ventricular end-diastolic and end-systolic dimensions were significantly ameliorated in the TNNI3K-overexpressing group compared with the other two groups (P < 0.05). Histological examination at week 2 revealed that infarct size was significantly smaller in the TNNI3K-overexpressing cell-injected mice than in the rest of the groups (Fig. 6, D and E). There were many necrotic cells observed in the infarct area in the medium-only group. Apoptotic cells were present in less abundance in the TNNI3K-overexpressing cell-injected animals compared with the flag-only group. Increased numbers of TUNEL-positive apoptotic cardiomyocytes were detected in the infarct area, border zone, and noninfarcted tissues in all three groups at 2 wk after ligation of LAD. However, the occurrence of TUNEL-positive cells in the border zone was significantly fewer in the TNNI3K-overexpressing group (P < 0.01) compared with the other groups (Fig. 6, F and G). When the cells at day 16 or 20 after induction were injected into the heart, there were no cardiac protection effects of TNNI3K-overexpressing cells (data not shown). Although the reason is unclear, it may be necessary to integrate early differentiated donor cells efficiently and effectively into host heart tissues compared with differentiated cells.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
The main findings in the present study are 1) overexpression of TNNI3K promotes P19CL6-derived cardiomyogenesis by suppressing apoptosis and accelerating maturation during the early stage of differentiation in a P19CL6 model; 2) TNNI3K increases the cell beating frequency and contractile force by increasing -actinin-positive cells in the beating masses, suppressing phosphorylation of cTnI, and enhancing the Ca2+ response in P19CL6-derived cardiomyocytes; and 3) in a mouse MI model, transplantation of the P19CL6 pluripotent progenitor cells with TNNI3K overexpression suppressed MI-induced injury, prevented MI-induced ventricular remodeling, and improved myocardial performance.
TNNI3K time-dependently increased the incidence of beating masses from day 8 to day 24 after induction (P < 0.05, Fig. 2B). However, there was no statistical difference in the appearance time of the first beating activity in culture wells between the TNNI3K-overexpressing group and the flag-only group (Fig. 2A). This result indicates that TNNI3K may be not an initial inducer but is a promoting factor for the early differentiation of the P19CL6-derived cardiac cells. Indeed, in our preliminary experiments, no cardiac differentiation occurred and no beating mass formed when DMSO was not added. To answer why TNNI3K promotes P19CL6-derived cardiomyogenesis, one explanation is that TNNI3K enhanced the formation of beating masses through its antiapoptotic effects during development of cardiomyocytes. At first, when the original cell density used for differentiation was increased 1.5- to 3-fold from 4 x 105 cells/well (Fig. 5A), almost all cell densities above 1.5-fold in the flag-only group did not continuously differentiate but died from day 5 after induction. However, the cardiac differentiation in the TNNI3K group continued at all cell densities used, and the beating mass numbers were significantly larger than in the flag-only group (P < 0.05). This result indicated that TNNI3K might favor contact growth and cardiac mass formation by suppressing cell death. Second, data in Fig. 5B show that the percentage of annexin V-positive cells, Bax protein expression, and phosphorylation of p38 and JNK were significantly decreased during the early differentiation period in the TNNI3K group compared with the flag-only group. Therefore, TNNI3K, at least, promotes cardiomyogenesis by 1) suppressing apoptosis formation by inhibiting the phosphorylation of p38 and JNK and 2) decreasing the annexin V-positive cells and suppressing Bax protein expression to result in a decrease of cell death and an increase in beating mass formation. As shown by our recent unpublished data, TNNI3K suppressed the hypoxia-induced expression of Bcl-2 protein and caspase-3 activity in cultured rat neonatal ventricular myocytes. Finally, suppression of the apoptosis was also found in an MI model of mice injected with TNNI3K-overexpressing cells (Fig. 6), indicating that TNNI3K can attenuate ischemia-induced injury by both promoting cardiac myogenesis and suppressing development of apoptosis.
TNNI3K Promotes the Contractility of P19CL6-Derived Beating Masses
Our data clearly showed that TNNI3K increased the beating frequency of SAPs, promoting the contractility and beating frequency of the beating masses (Figs. 2 and 3A). As a basis for the contraction, the number of beating masses and expression of -actinin protein were increased in the TNNI3K-overexpressing group (Fig. 4B). For the functional mechanisms, TNNI3K enhanced the epinephrine response of the spontaneous activities and increased the caffeine- and ryanodine-triggered intracellular Ca2+ release (Fig. 3, C and D). These findings correspond with our previous study (34) in which TNNI3K was known to interact with -actin, myosin binding protein, and cTnI.
One question is how TNNI3K interacts with cTnI, since cTnI is phosphorylated by at least three protein kinases, Ca2+/phospholipid-dependent protein kinase C (PKC), cGMP-dependent protein kinase (PKG), and cAMP-dependent kinase (PKA). Even though we found that the cTnI phosphorylation was downregulated in TNNI3K-overexpressing cells (Fig. 3B), there is no direct evidence to answer this question, because phosphorylation of cTnI at different sites may induce different physiological effects, and abnormal activation or absence of these sites results in abnormal cardiac contractility and Ca2+ responses. However, previous reports showed that inhibition of PKC-mediated cTnI phosphorylation can increase cardiac performance in vivo (26), whereas activation of PKC-mediated cTnI phosphorylation inhibits actomyosin interaction (1, 13, 32), decreases tension development, and impairs cell shortening in vitro (2, 21, 25). By contrast, stimulation of PKA-dependent cTnI phosphorylation increases myocardial contractility, accelerates the rate of force development and relaxation, and amplifies the positive inotropic and lusitropic effects of higher stimulation frequency (27). Therefore, as one presumptive result to correspond with the finding that TNNI3K increased the contractility and downregulation of expression of cTnI protein, TNNI3K may regulate the cTnI phosphorylation to induce an increase in the Ca2+ responsiveness of myofilaments by inhibiting PKC-mediated phosphorylation (26) but not by enhancing PKA-mediated phosphorylation as indicated (24, 27).
Furthermore, inhibition of ASK1-modulated apoptosis by TNNI3K overexpression also may be involved. The ASK1 pathway plays a pivotal role in the progression of apoptosis (31), and overexpression of ASK1 induces cTnT phosphorylation and inhibits cell shortening, Ca2+ transients, and contractile function in cardiomyocytes (11). Therefore, overexpression of TNNI3K may suppress ASK1-induced cTnT phosphorylation to promote cardiac differentiation and increase contractile function. This is supported by the observation that expression of p38, JNK, and Bax protein and the number of annexin V-positive cells were lower in the TNNI3K-overexpressing cells than in the flag-only cells (Fig. 5). The possibility that TNNI3K directly inhibits the cTnI activity also cannot be excluded.
Modulation of TNNI3K May Be a Useful Therapeutic Approach for Cardiac Disease
Clinical application of stem cells for the treatment of heart diseases has been limited by the failure of embryonic stem (ES) cells to fuse with resident cardiomyocytes and by the lack of clear data demonstrating that this strategy can improve the perfusion and contractile performance of the injured heart. In this study, it is clear that in addition to its promoting the differentiation of P19CL6 cells into cardiomyocytes, TNNI3K also has beneficial effects on the cardiac contraction function. Because TNNI3K promotes both cardiomyogenesis and cardiac contractility in vitro, it is possible that TNNI3K also may promote the transplantation of pluripotent embryonic cell- or ES cell-induced cardiomyogenesis and cardiac contraction in myocardial infarction disease in vivo.
Finally, TNNI3K may be a promising tool for the treatment of cardiac diseases. Enhancement of TNNI3K activity in cardiomyocytes by genomic techniques or by artificially synthesized molecules may be able to recover impaired cardiac contraction force of the disease heart and may be a useful approach to treat ischemic cardiac diseases.
Clinical Perspective
Recent research on the therapeutic approach for ischemic cardiac diseases using stem cell transplantation has been a great success. However, for the clinical application, it still has many problems and questions needed to be answered. The most feared complications from cell therapy are arrhythmogenesis, tumor formation, and inflammation (35). Another serious problem is that the cardiomyocytes derived from transplanted ES cells may have not enough ability to perform their contractile function or to synchronize their beating with host cardiomyocytes since MI would increase the cTnI activity as reported (36). In the present study, we demonstrated that TNNI3K, a cardiac-specific MAP kinase, can promote the cardiomyogenesis with increasing beating masses and -actinin-positive cells, enhancing beating frequency and contractile function, attenuating the MI-induced left ventricular remodeling through suppression of the p38- and JNK-modulated apoptosis, and decreasing phosphorylation of cTnI. Although P19CL6 is not a good donor line for clinical transplantation, the project of enhancing TNNI3K activity in the engraftment ES or bone marrow cells to promote the cardiomyogenesis, beating frequency, and contractile function may be a new therapeutical strategy for the treatment of ischemic cardiac diseases (14).
|
GRANTS |
|---|
|
TOPABSTRACTEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
* Z.-F. Lai and Y.-Z. Chen contributed equally to this work. 
|
REFERENCES |
|---|
|
TOPABSTRACTEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
q and PLC-β protein abundance in cardiac hypertrophy and failure. Am J Physiol Heart Circ Physiol 277: H2298–H2304, 1999.-Adrenergic response and myofilament activity in mouse hearts lacking PKC phosphorylation sites on cardiac TnI. Am J Physiol Heart Circ Physiol 282: H2397–H2405, 2002.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
