AJP - Heart Fuel your research with LabChart
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Am J Physiol Heart Circ Physiol 286: H2314-H2321, 2004; doi:10.1152/ajpheart.01137.2003
0363-6135/04 $5.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (5)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hirsch, J. C.
Right arrow Articles by Metzger, J. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hirsch, J. C.
Right arrow Articles by Metzger, J. M.

Comparative analysis of parvalbumin and SERCA2a cardiac myocyte gene transfer in a large animal model of diastolic dysfunction

Jennifer C. Hirsch,1 Andrea R. Borton,1,3 Faris P. Albayya,3 Mark W. Russell,2 Richard G. Ohye,2 and Joseph M. Metzger3

1Departments of Surgery, 2Pediatrics, and 3Molecular and Integrative Physiology, University of Michigan, Ann Arbor, Michigan 48109

Submitted 1 December 2003 ; accepted in final form 26 January 2004


   ABSTRACT
TOP
 ABSTRACT
METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Diastolic dysfunction results from impaired ventricular relaxation and is an important component of human heart failure. Genetic modification of intracellular calcium-handling proteins may hold promise to redress diastolic dysfunction; however, it is unclear whether other important aspects of myocyte function would be compromised by this approach. Accordingly, a large animal model of humanlike diastolic dysfunction was established through 1 yr of left ventricular (LV) pressure overload by descending thoracic aortic coarctation in canines. Serial echocardiography documented a progressive increase in LV mass. Diastolic dysfunction with preserved systolic function was evident at the whole organ and myocyte levels in this model, as determined by hemispheric sonomicrometric piezoelectric crystals, pressure transducer catheterization, and isolated myocyte studies. Gene transfer of the sarco(endo)plasmic reticulum calcium-ATPase (SERCA2a) and parvalbumin (Parv), a fast-twitch skeletal muscle Ca2+ buffer, restored cardiac myocyte relaxation in a dose-dependent manner under baseline conditions. At high Parv concentrations, sarcomere shortening was depressed. In contrast, during {beta}-adrenergic stimulation, the expected enhancement of myocyte contraction (inotropy) was abrogated by SERCA2a but not by Parv. The mechanism of this effect is unknown, but it could relate to the uncoupling of SERCA2a/phospholamban in SERCA2a myocytes. Considering that inotropy is vital to overall cardiac performance, the divergent effects of SERCA2a and Parv reported here could impact potential therapeutic strategies for human heart failure.

heart failure; calcium; contraction; inotropy; sarco(endo)plasmic reticulum Ca2+-ATPase 2a; canine

DIASTOLIC HEART FAILURE results from impairment in myocardial relaxation and accounts for significant morbidity and mortality in developed countries (16). In the United States alone, 5,000,000 people have heart failure; of those, ~40% have diastolic failure exclusive of systolic contractile abnormalities (21). A variety of disease states cause diastolic dysfunction (DD), including hypertensive, ischemic, or valvular heart disease as well as hypertrophic and restrictive cardiomyopathies. The incidence of diastolic heart failure increases with age and is associated with such risk factors as female gender, diabetes, and increased body mass index (1, 10). At the intracellular level, DD is caused, at least in part, by a delay in Ca2+ sequestration causing abnormally slow cardiac muscle relaxation. In myocytes from human and animal hearts with both systolic dysfunction and DD, there are prolonged calcium transients, decreased levels of the sarco(endo)plasmic reticulum Ca2+-ATPase pump (SERCA2a), and increased diastolic [Ca2+] (8, 15, 24, 25, 27, 28). Despite increasing knowledge about the cellular underpinnings of DD, current medical therapy remains directed toward alleviating symptoms with no modalities clinically available to directly alter intracellular Ca2+ handling. The primary treatment for end-stage heart failure remains transplantation.

Improved understanding of the cellular defects in heart failure enables the consideration of new therapeutic options to directly enhance Ca2+ handling in failing myocytes. Parvalbumin (Parv) and SERCA2a are two key proteins involved in Ca2+ handling in a variety of striated muscle types. SERCA2a is an ATP-dependent Ca2+ pump that actively sequesters Ca2+ into the sarcoplasmic reticulum (SR) during cardiac relaxation. Parv is a soluble Ca2+ buffer that is naturally expressed in the cytoplasm of fast-twitch skeletal muscles but not in the heart. In fast-twitch skeletal muscles, Parv is thought to facilitate rapid muscle relaxation through the transient buffering of myoplasmic Ca2+. Recently, SERCA2a and Parv gene transfer have been shown to improve cardiac muscle myocyte relaxation via a more rapid decay rate in the calcium transient (5, 13, 30). However, to date, there has been no direct comparison of the effects of Parv and SERCA2a on normal or diseased cardiac myocyte function under varied physiological conditions.

One aim of this study was to determine whether the overexpression of SERCA2a or the de novo expression of Parv would result in improved relaxation in individual myocytes from a large animal model of DD. Canine myocytes have Ca2+-handling properties that closely resemble those of human myocytes (3). Accordingly, we established a model of pure DD in canines by ventricular pressure overload over a 1-yr time period that allowed for analysis of functional alterations with Parv and SERCA2a in a highly controlled heart failure system. Progressive left ventricular (LV) hypertrophy ensued, and, after 1 yr, DD with maintained systolic function was evident at both the whole organ and individual myocyte levels. In adult canine myocytes with DD, we report here the convergent, dose-dependent effects of Parv and SERCA2a to restore normal relaxation.

A second aim of the study was to assess SERCA2a- and Parv-transduced myocyte function under altered physiological conditions. Unexpectedly, we found that the characteristic marked increase in myocyte contractility by {beta}-adrenergic stimulation was abrogated in SERCA2a-transduced myocytes, whereas it was retained in Parv-transduced myocytes. In concert with this finding, and in agreement with previous work, SERCA2a overexpression, but not Parv gene transfer, affects the stoichiometry between SERCA2a and phospholamban (PLB), a key inhibitor of SERCA2a activity. The differential outcomes of these Ca2+-handling proteins, particularly in terms of varied altered adrenergic reserve, may have implications for possible translation to human failing myocytes.


   METHODS
TOP
 ABSTRACT
 METHODS
RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Recombinant adenoviruses. The Parv and {beta}-galactosidase (LacZ) recombinant adenovirus vectors were constructed as previously described (23). The SERCA2a recombinant adenovirus vector contained the human cDNA for SERCA2a and a green fluorescent protein (GFP) reporter (the kind gift of Dr. Roger Hajjar, Harvard Medical School). In each vector, gene expression was under control of the cytomegalovirus promoter and polyadenylation signal provided by simian virus 40. High-titer, plaque-purified adenoviral stocks were produced and purified as described previously (23). Viral aliquots were stored at –80°C.

Animal model and hemodynamics. The procedures used in this study were in agreement with the guidelines of the Internal Review Board of the University of Michigan and approved by the University of Michigan Committee on the Use and Care of Animals. Veterinary care was provided by the University of Michigan Unit for Laboratory Animal Medicine. The University of Michigan is accredited by the American Association of Accreditation of Laboratory Animal Health Care, and the animal care use program conforms to the standards of the National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH Pub. No. 86-23).

In the DD group, DD was induced in five bred-for-purpose juvenile beagles (male, 5 kg). Nine age and size-matched control beagles served as comparisons [the noncoarctation (NC) group]. Progressive LV pressure overload was initiated by the placement of a descending thoracic aortic coarctation via a Gore-Tex interposition disc with a central 4-mm aperture via a left thoracotomy. Animals underwent serial transthoracic echocardiography every 2–4 wk for 1 yr. Systolic and diastolic blood pressure (obtained by pneumatic pressure cuff) for the DD dogs was 151.9 ± 3.86 and 106.0 ± 2.62 mmHg, respectively. Light sedation with 0.07 ml/kg BAG (100 mg butorphanol, 25 mg acepromazine, and 5 mg glycopyrolate/50 ml saline) was administered intramuscularly before evaluation. M-mode and two-dimensional echocardiography was performed transthoracically for the serial studies and epicardially during the final instrumentation. Short-axis LV views at the midpapillary level were used to measure septal and posterior wall thickness, transmitral E-to-A (E/A) ratio, and fractional shortening. The long-axis view was then obtained to complete LV mass determination. When maximal growth of the animal was achieved, the animals were instrumented via a median sternotomy with sonomicrometric piezoelectric crystals for hemodynamic monitoring during inflow occlusion (18). Animals were induced with thiopental sodium, were maintained with isoflurane on a mechanical ventilator, and underwent a median sternotomy. Snares were placed around the superior and inferior vena cava. Hemispheric sonomicrometric piezoelectric crystals were placed on the long and short axis of the LV. A high-fidelity pressure transducer catheter (Millar Instruments; Houston, TX) was introduced through the LV apex for ventricular pressure monitoring. Fluid-filled arterial catheters were placed in the femoral artery and innominate artery for systemic pressure monitoring. Continuous pressure-volume loops were generated during repeated complete inflow occlusion. The end-diastolic pressure-volume relationship was determined for each animal. At the termination of the experiment, the animals were heparinized (500 U/kg) and euthanized with 1 ml/5 kg pentobarbital (390 mg/ml). The heart was explanted, immediately rinsed, and cooled in Krebs-Henseleit buffer (KHB) with calcium. Age- and weight-matched control animals (n = 4) underwent identical hemodynamic monitoring before the heart was harvested.

Myocyte isolation and gene transfer. After 1 yr of pressure overload, the hearts were explanted, and KHB with supplemented calcium (118 mM NaCl, 4.8 mM KCl, 1.2 mM KH2PO4, 1.2 mM MgSO4·7H2O, 1.0 mM CaCl2·2H2O, 25 mM HEPES, and 11 mM glucose) was used to rinse the myocardium. The LV free wall was excised along with a cuff of the ascending aorta. The endocardium and epicardium were reapproximated to ligate the perforating vessels, and the left anterior descending coronary artery was cannulated. The ventricle was initially perfused on a modified Langendorf apparatus with KHB + Ca2+ until the heart started contracting and the effluent was clear. The perfusate was then changed to modified Tyrode (MT) solution (118 mM NaCl, 4.8 mM KCl, 1.2 mM KH2PO4, 1.2 mM MgSO4·7H2O, 25 mM HEPES, 11 mM glucose, 0.68 mM glutamine, 5 mM sodium pyruvate, 2 mM mannitol, and 10 mM taurine) with 15 mg/ml hyaluronidase (type I-S, Sigma; St. Louis, MO) and 0.4 mg/ml collagenase (type 2, Worthington Biochemical; Lakewood, NJ) for tissue digestion. Individual cells were then isolated through a series of morselizations, digestions, and triturations. Adult myocytes were resuspended in incubation solution [MT solution with 2% BSA (fraction V, Boehringer-Mannheim; Indianapolis, IN)] and made calcium tolerant over 1 h (1.8 mM Ca2+). The average number of rod-shaped myocytes per heart was 6.87 x 106 (±1.2 x 106) with 51.4 ± 2.6% viability in NC animals and 7.95 x 106 (±1.4 x 106) myocytes with 55.7 ± 2.7% viability in DD animals. There were no significant differences in number or viability between NC and DD animals (t-test). Aliquots of myocytes (2 x 104 myocytes) were plated on laminin-coated coverslips and incubated at 37°C in DMEM containing 5% fetal bovine serum, 50 U/ml penicillin, and 50 µg/ml streptomycin for 2 h. After gentle aspiration, myocytes were incubated with recombinant adenovirus in serum-free DMEM plus penicillin/streptomycin (P/S). A total of four treatment groups were analyzed: control (media alone), LacZ [multiplicity of infection (MOI) 200], Parv (MOI 250), and SERCA2a (MOI 200). These optimal titers were derived from pilot studies demonstrating a high efficiency (at 48 h: LacZ staining 97%, GFP fluorescence 95%) without cytotoxicity. Serum-free medium was changed the day after viruses were added and then every 2 days for up to 4 days in culture.

Western blots and immunofluorescence. Myocytes were collected in sample buffer at days 0, 2, 3, and 4 postisolation. Samples were separated on SDS-polyacrylamide gels (3.5% acrylamide in the stacking gel, 12% acrylamide in the separating gel). The gel was then transblotted onto polyvinylidene difluoride membranes for 2,000 V·h. Immunodetection was carried out on blots fixed in glutaraldehyde. Western blot analysis was performed as described (31) with a Parv antibody (PARV19, Sigma) titer of 1:1,000, a SERCA2a antibody (SERCA2a, Upstate Biotechnology) titer of 1:1,000, a PLB (Upstate Biotechnology) titer of 1:1,000, or an anti-actin (5c5, Sigma) titer of 1:10,000. Individual blots were analyzed with Multi-analyst software (Bio-Rad) using actin immunolabeling to assess relative loading. Indirect immunofluorescence labeling of myocytes was performed as described (30) with primary [PARV19, 1:500; SERCA2a, 1:1,000; and PLB 1:500] and secondary (anti-mouse IgG-Texas red, 1:100, Molecular Probes) antibodies. Samples were examined on a Leitz Aristplan microscope outfitted with a Sony digital camera.

Laser diffraction and mechanical measurements. A diffraction chamber, as previously described (30), was filled with M199+ (medum 199 stock, Sigma) supplemented with 10 mM glutathione, 26.2 mM sodium bicarbonate, 0.02% BSA, and 50 U/ml P/S (pH adjusted to 7.4 with NaOH), with 1.8 or 5 mM CaCl2. Temperature was set at 37°C. In each experimental group, an equal number of myocytes were evaluated at each calcium concentration. A supramaximal, 5-ms square-wave pulse was used to stimulate the myocytes. A myocyte was positioned in the laser beam, and the first-order diffraction line was focused with a cylindrical lens onto a linear position detector (LSC 5D or 30D, United Detector Technology). The output of the detector was amplified and recorded on a digital oscilloscope at 5,000 Hz. The average of 10 twitches/myocyte was used for subsequent analysis. After the above protocol, the detector was removed and a viewing screen was inserted in the beam path. The distance from the zero to the first-order diffraction pattern was used to calculate the initial sarcomere length (SL) and the amplitude of the SL change after electrical stimulation. The time from stimulation (Ts) to maximum shortening/relengthening and from maximum shortening (Tp) to 1/4, 1/2, and 3/4 relengthening and the time to 1/2 width (T 1/4 R, T 1/2 R, and T 3/4 R, respectively) were measured from the averaged trace for each myocyte using Igor Pro wave-analysis software (Wavemetrics) The maximum rate of shortening and relengthening (+dL/dtmax and –dL/dtmax, respectively) were calculated and normalized to the maximum shortening amplitude in each myocyte. {beta}-Adrenergic stimulation was accomplished by adding 100 nM of isoproterenol to M199. All experiments involving isoproterenol were performed at 1.8 mM CaCl2.

Statistics. Statistical analysis was performed using GraphPad Instat software. For each dog, there were typically four experimental groups (control, LacZ, Parv, and SERCA2a), and data were assessed on days 2–4. On average, from each animal, 18 myocytes were studied x3 time points x4 experimental groups. This translates to a total of 216 myocytes for each dog on average. Thus, for the DD group, there were 5 dogs studied x18 myocytes averaged/point = an average of 90 myocytes/point. For the NC group, four dogs were studied. Observations per point vary somewhat between experiments due to some differences in the number of myocytes studied from day to day and from dog to dog. Results from dogs within the same group (NC or DD) were similar, as reflected by the errors reported in each group. Values for each group are expressed as means ± SE. Significant differences between groups (P < 0.05) were tested with ANOVA and a post hoc Student-Newman-Keuls multiple-comparison test or by a Student's t-test if only two groups were compared.


   RESULTS
TOP
 ABSTRACT
 METHODS
 RESULTS
DISCUSSION
 GRANTS
 REFERENCES
 
Canine DD model. After 1 yr of LV pressure overload via descending thoracic aortic coarctation in canines, the animals maintained normal levels of activity without overt signs of heart failure (i.e., tachypnea, basal tachycardia, or exercise intolerance). Echocardiography indicated an increase in LV mass and preserved LV fractional shortening (Table 1). Hypertrophy at the completion of the study was confirmed at the time of final instrumentation by increased LV mass (g/m2) on epicardial echocardiography and increased LV posterior wall thickness (Table 1). Systolic function was preserved, with the percent shortening fraction not different between control and DD animals (Table 1). Hypertrophy was evident at the isolated myocyte level, where cardiac myocyte width and cross-sectional area were greater in DD [31.5 ± 3.6 µm and 3,455 ± 232 µm2 (n = 21)] vs. NC [17.1 ± 0.7 µm and 2,121 ± 124 µm2 (n = 21)] myocytes (P < 0.0001). Once developmental maturity and maximal growth had been attained, the control and DD animals underwent instrumentation with sonomicrometric piezoelectric crystals for hemodynamic monitoring during inflow occlusion challenges (Table 2). The end-diastolic pressure-volume relation demonstrated an increase in the y-intercept between control (1.0 ± 4.17 mmHg) and DD (37.92 ± 9.31 mmHg, P = 0.008) animals and an increase in slope between NC (1.0 ± 8.98 mmHg/ml) and DD (20.88 ± 7.4 mmHg/ml, P = 0.027) animals. These results provide evidence of decreased compliance and LV DD. Over the course of the study, all animals maintained normal levels of activity without evidence of overt heart failure.


View this table:
[in this window]
[in a new window]
  		Table 1. Noninvasive hemodynamic data

 

View this table:
[in this window]
[in a new window]
  		Table 2. Invasive hemodynamic monitoring data

 
To directly assess cellular function, adult myocytes were isolated from age-matched NC and DD animals. Impaired relaxation in DD versus NC myocytes was observed using laser diffraction analysis of SL shortening (Fig. 1A). Measurement of Ts 1/4 R indicated a significant decrease in the rate of relaxation [NC 209.6 ± 10.5 ms (n = 45) and DD 251.0 ± 5.9 ms (n = 90)] that was also evident at later time points in relaxation (Ts 1/2 R and Ts 3/4 R; Fig. 1B). Contractile function was preserved at the cellular level with no difference in the relative SL shortening between groups [NC 0.07 ± 0.01 µm (n = 45) vs. DD 0.06 ± 0.002 µm (n = 90)]. At the molecular level, the level of SERCA2a/actin expression in DD myocytes was 66% of the expression in NC myocytes [actin expression remained stable over time (Western blot not shown) and the time-dependent increase in SERCA2a expression is demonstrated in Fig. 3A].



View larger version (17K):
[in this window]
[in a new window]
  		Fig. 1. Characterization of cellular diastolic dysfunction in adult canine cardiac myocytes. A: representative traces of sarcomere length (SL) shortening in individual canine cardiac myocytes from noncoarctation (NC) and diastolic dysfunction (DD) animals. B: comparison of the time from stimulus to 1/4, 1/2, and 3/4 relaxation (Ts 1/4 R, Ts 1/2 R, and Ts 3/4 R) in myocytes from NC and DD animals on day 2 in primary culture. C: summary of vector control [{beta}-galactosidase (LacZ) vector] studies on relaxation performance in canine myocytes. Myocytes are from NC animals on day 2. Values are means ± SE; n = 45 NC and 90 DD myocytes. *P < 0.0001, DD > NC.

 


View larger version (46K):
[in this window]
[in a new window]
  		Fig. 3. Parv and SERCA2a expression dose-response relationships in DD myocytes. A: summary of increased SERCA2a expression as a function of days postgene transfer. To control for loading, expression of SERCA2a is represented as a ratio of SERCA2a to actin. The baseline ratio of SERCA2a to actin was established in nontransduced DD myocytes (control), and this value was arbitrarily set as 1.0, to which all other values were normalized. The solid line through the data is from a linear regression analysis, r2 = 0.77 (values are means ± SE; n = 8 per point). Inset, representative Western blot demonstrating SERCA2a overexpression at day 4. Control samples at days 2, 3, and 4 are C2, C3, and C4; LacZ (L) and SERCA2a (S) samples are shown at the corresponding days 2, 3, or 4 postgene transfer. B: summary of increased Parv expression as a function of days postgene transfer. Expression of Parv is represented as a ratio of Parv to actin. The solid line through the data is from a linear regression analysis, r2 = 0.994 (values are means ± SE; n = 4 per point). Inset, Western blot probed with Parv antibody. Parv is first detected at day 2 and increases with time postgene transfer. Lane E, extensor digitorum longus (rat fast skeletal muscle positive control); Lanes C, control; lanes L, lacZ; Lanes P, Parv; days 2, 3, and 4 postgene transfer are as labeled. For A and B, values are means ± SE; n = 4. C: immunofluorescence labeling of SERCA2a in nontransduced DD (control) and SERCA2a gene transfer DD myocytes, with SERCA2a antibody, exhibiting increased fluorescence at identical exposures. Insets, enlarged view. D: immunofluorescence labeling of Parv in nontransduced DD (control) and Parv gene transfer DD myocytes with Parv antibody. E: summary of Ts 1/4 R as a percentage of control (control Ts 1/4 R values were 251.02 ± 5.94 ms on day 2, 196.08 ± 5.6 ms on day 3, and 179.62 ± 5.5 ms on day 4). Parv and SERCA2a Ts 1/4 R values on days 2–4 were different from control (*P < 0.01). SERCA2a-transduced myocytes relaxed faster than Parv-transduced myocytes on days 2 and 3 (+P < 0.01), with no difference by day 4. F: change in SL as a percentage of control myocytes (control data: 0.055 ± 0.002 µm, day 2; 0.053 ± 0.003 µm, day 3; 0.057 ± 0.003 µm, day 4). Parv-transduced myocytes demonstrate a progressive decrease in amplitude of sarcomere shortening from days 2 to 4 (*P < 0.01). No significant difference in sarcomere shortening was demonstrated for SERCA2a at days 3 and 4. G and H: summary of Ts 1/2 R and Ts 3/4 R (G) and minimum and maximum dL/dt (H) at day 4 postgene transfer in DD myocytes. *P < 0.001, Parv and SERCA2a different than control; +P < 0.001, SERCA2a different from both control and Parv. For E–H, values are means ± SE; n = 81–90 per point.

 
Effects of Parv and SERCA2a on contractile performance. Contractile properties of DD myocytes were assessed initially at day 2 postgene transfer. Results from control (in all figures, "control" refers to nontransduced myocytes) NC myocytes and LacZ-transduced NC myocytes indicated that gene transfer per se had no significant effect on myocyte relaxation performance (Fig. 1C). Similar results were exhibited with control and LacZ DD myocytes (results not shown). Thus all groups were subsequently compared with nontransduced control DD myocytes. At day 2 postgene transfer, alterations in relaxation were evident in SERCA2a- and Parv-expressing myocytes. Individual traces of Parv and SERCA2a myocyte shortening showed marked enhancement in the rate of relaxation, with restoration of relaxation to NC levels (Fig. 2A). The Ts 1/4 R values for Parv [207.2 ± 6.9 ms ( n = 90)] and SERCA2a [177.9 ± 5.0 ms (n = 90)] were restored back to or below the NC baseline [209.6 ± 10.5 ms (n = 45)], with SERCA2a demonstrating significantly faster relaxation than control DD (P < 0.01) and Parv-expressing myocytes (P < 0.01). These findings are in keeping with earlier results demonstrating that SERCA2a and Parv hasten relaxation by facilitating a more rapid decay in the intracellular calcium transient (5, 13, 30). Most evident at later time points in relaxation, SERCA2a was significantly more effective than Parv at accelerating relaxation (Fig. 2B).



View larger version (26K):
[in this window]
[in a new window]
  		Fig. 2. Mechanical performance of DD myocytes transduced with parvalbumin (Parv) and sarco(endo)plasmic reticulum Ca2+-ATPase (SERCA)2a recombinant vectors. A: representative traces of sarcomere shortening in control (nontransduced DD) and Parv- and SERCA2a-expressing DD myocytes (day 2 postgene transfer). B: enhanced rate of relaxation in myocytes expressing Parv and overexpressing SERCA2a compared with control at day 2 postgene transfer. *P < 0.001, Parv < control; +P < 0.001, SERCA2a < control and Parv. Values are means ± SE; n = 90.

 
In general, the enhancement of the myocyte SL relengthening time course obtained by gene transfer of SERCA2a and Parv to DD myocytes was qualitatively similar in myocytes from NC animals (Table 3, http://ajpheart.physiology.org/cgi/content/full/01137.2003/DC1).

Dose-response relationships for Parv and SERCA2a. The high efficiency of cardiac gene transfer allowed construction of the contractile dose-response relationships for Parv and SERCA2a. From days 2 to 4 postgene transfer, both SERCA2a and Parv expression steadily increased (linear regression analysis: r2 was 0.77 and 0.99 for SERCA2a and Parv, respectively; Fig. 3, A–C), and indirect immunofluorescence demonstrated that >90% of myocytes expressed Parv (Fig. 3D). Western blot analysis showed that Parv increased from nondetectable levels at day 0 to Parv-to-actin ratios of 0.007 ± 0.05 by day 2, 0.129 ± 0.03 by day 3, and 0.169 ± 0.01 by day 4. With the use of the rat fast-twitch extensor digitorium longus muscle as a positive control [Parv concentration ~0.33 mM (11)] and a transduced cardiac myocyte Parv-to-actin ratio of 3.62 ± 0.7, the Parv concentration in myocytes at day 4 was estimated to be ~19 µM. In SERCA2a-transduced myocytes there also was a time-dependent, progressive increase in expression (Fig. 3A). By day 4 postgene transfer, SERCA2a-transduced myocytes had a 40% increase in the SERCA2a-to-actin ratio compared with nontransduced controls. PLB expression remained stable from days 2 to 4. Therefore, the SERCA2a-to-PLB ratio increased over time, in agreement with previous studies (13). Immunofluorescence imaging showed the expected striated pattern of SERCA2a expression in these myocytes, providing evidence that the overexpressed SERCA2a was localized to the SR (Fig. 3C).

Because Parv and SERCA2a increased as a function of time postgene transfer, it is possible to construct a dose-response relationship for contractile function in DD myocytes by calculating the percent change from nontransduced DD myocytes at each corresponding day in primary culture. In this analysis, both Parv- and SERCA2a-transduced myocytes demonstrated a progressive decrease in Ts 1/4 R from days 2 to 4, with values in transduced myocytes being significantly faster than the day-matched control DD myocytes (P < 0.01; Fig. 3A and Table 3). At days 2 and 3, SERCA2a-transduced myocytes had faster relaxation properties compared with those transduced with Parv (P < 0.01). However, owing to the progressive increase in Parv concentration by day 4, there was a marked acceleration in relaxation in Parv-transduced myocytes such that Ts 1/4 R and Ts 1/2 R were no longer different between the Parv and SERCA2a groups, although differences were still apparent in Ts 3/4 R and ±dL/dt between groups (Fig. 3, G and H). These results are consistent with Parv serving as a delayed buffer for calcium early in diastole and a source for calcium late in diastole. Figure 4 demonstrates that the absolute relaxation times were not different between NC and DD myocytes after Parv gene transfer. In contrast, with SERCA2a gene transfer, the relaxation times were significantly slower in DD compared with NC myocytes (Fig. 4B). In association with increased Parv expression, there was a significant decrease in the relative sarcomere shortening that was not generally exhibited in SERCA2a myocytes (Fig. 3F). As discussed recently, Parv, at high enough concentrations, can buffer systolic calcium and attenuate sarcomere shortening amplitude (5).



View larger version (14K):
[in this window]
[in a new window]
  		Fig. 4. Comparative analysis of Parv (A) and SERCA2a (B) gene transfer on relaxation performance between NC and DD canine myocytes. All results are on day 4 after gene transfer. Values are means ± SE; n = 52–85 per group. *P < 0.0001, DD significantly greater than NC.

 
Studies in NC canine cardiac myocytes demonstrated qualitatively similar alterations in relaxation and relative sarcomere shortening with similar dose-response curves (Table 3 online). In agreement with previous work in rodents (5, 26, 27), the decay in the intracellular calcium transient was hastened by both Parv and SERCA2a (see Table 3). The decrease in shortening amplitude by increased Parv expression in isolated myocytes can be attributed to the heightened buffering of calcium, as discussed in detail previously (5).

{beta}-Adrenergic stimulation. During {beta}-adrenergic stimulation, both the amplitude of contraction (positive inotropy) and rate of relaxation are dramatically increased. We tested, under physiological Ca2+ concentration (1.8 mM), whether Parv or SERCA2a gene transfer would affect the myocyte response to {beta}-adrenergic stimulation (Fig. 5). In nontransduced DD myocytes, {beta}-adrenergic stimulation resulted in the characteristic enhancement of cellular contraction. In both Parv- and SERCA2a-transduced myocytes, {beta}-adrenergic stimulation did not cause a further enhancement in relaxation (Fig. 5). SERCA2a myocytes also were nonresponsive to {beta}-adrenergic stimulation in terms of amplitude of contraction, whereas the Parv-transduced myocyte response, both qualitatively and quantitatively, was indistinguishable from control (Fig. 5). Thus, under {beta}-adrenergic stimulation, the absolute magnitude of change in SL was no longer different between Parv and control and was markedly greater than that obtained with SERCA2a (Fig. 5B). Qualitatively similar findings were obtained in NC myocytes transduced with Parv and SERCA2a vectors. To ascertain whether in SERCA2a-transduced myocytes this was caused by dysfunction in the {beta}-adrenergic signaling cascade, or rather caused by a more generalized alteration in excitation-contraction coupling, myocytes were examined in the presence of heightened extracellular Ca2+ (5 mM). Normally, heightened extracellular Ca2+ produces enhanced Ca2+-induced Ca2+ release in cardiac myocytes (Fig. 5D). This mimics the positive inotropy seen with {beta}-adrenergic stimulation but is independent of subcellular protein phosphorylation. In nontransduced DD myocytes, SL shortening increased to 211% of baseline by increasing Ca2+ from 1.8 to 5 mM; however, in SERCA2a- and Parv-transduced myocytes the increase in SL amplitude was reduced (130% and 116%, respectively).



View larger version (25K):
[in this window]
[in a new window]
  		Fig. 5. Effects of {beta}-adrenergic stimulation on SL shortening amplitude and rate of relaxation in Parv- and SERCA2a-transduced myocytes from canines with DD at day 4 postgene transfer. A: representative traces of sarcomere shortening in individual myocytes at baseline and with the addition of isoproteronol (Iso; 100 nM). Baseline was normalized to 1.00 for all myocytes, with each myocyte serving as its own control. B: summary of Iso-mediated enhancement of absolute (µm) SL shortening in control (n = 5), Parv-transduced (n = 5), and SERCA2a-transduced (n = 5) myocytes. Nontransduced DD (control) and Parv-transduced myocytes, but not SERCA2a-transduced myocytes, demonstrated an increase in the amplitude of sarcomere shortening with Iso (*P < 0.01). C: summary of Iso-mediated effects on the time from peak contraction to 1/4 relaxation (Tp 1/4 R) in control (n = 5), Parv-transduced (n = 5), and SERCA2a-transduced (n = 5) myocytes. Nontransduced DD (control) myocytes demonstrated enhanced relaxation with Iso (*P < 0.05), the rates were comparable to the Parv- and SERCA2a-transduced myocytes with and without Iso. D: summary of absolute SL shortening in control (n = 40), Parv-transduced (n = 40), and SERCA2a-transduced (n = 40) myocytes at 1.8 and 5.0 mM Ca2+. Control myocytes had a significant augmentation in amplitude of contraction (*P < 0.01), whereas Parv- and SERCA2a-transduced myocytes had a blunted response. In B–D, values are means ± SE.

 

   DISCUSSION
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
GRANTS
 REFERENCES
 
This study provides a direct comparison of SERCA2a and Parv on cardiac myocyte mechanical performance. A large animal model of pure hypertrophic DD with preserved systolic function was developed to provide a reproducible and controlled environment for assessing the functional consequence of genetic modification in Ca2+-handling proteins. In addition, myocytes from large mammals, as opposed to those from rodents, have intracellular Ca2+ handling properties highly comparable with human cardiac muscle (3). Parv and SERCA2a gene transfer corrected relaxation abnormalities in failing myocytes in a similar dose-dependent manner. In contrast, whereas Parv-transduced and control myocytes responded to {beta}-adrenergic stimulation with the typical marked increase in contractility (positive inotropy), SERCA2a-transduced myocytes were nonresponsive. As inotropy is a central effecter of overall cardiac performance, the divergent outcomes of these intracellular calcium handling proteins will likely be important in possible therapeutic translation to human heart failure.

Cardiac myocyte calcium handling is vital to overall heart function (3). In human heart failure, it is estimated that 40% of patients have pure DD (21). Diastolic heart failure results in part by a decreased expression of SERCA2a; consequently, the intracellular calcium transient is prolonged and mechanical relaxation is slowed (27). Considerable attention has been directed at the mechanisms of calcium handling in the normal and failing heart. Recent work has clearly demonstrated that modifications in either SERCA2a or PLB, an inhibitor of SERCA2a activity, markedly influence cardiac relaxation and may therefore have therapeutic potential in the failing heart. Decreased expression of PLB by gene ablation in mice or by antisense gene transfer has demonstrated the important role of the rate of calcium removal on heart and myocyte relaxation (7, 22). Similarly, SERCA2a overexpression by transgenesis or by gene transfer can hasten relaxation kinetics (2, 13). Collectively, these studies provide strong evidence that the SR SERCA2a-to-PLB ratio is an important determinant of myocardial relaxation (19).

The present findings, while supporting a role for SERCA2a in facilitating relaxation, provide evidence that an elevated SERCA2a-to-PLB ratio by SERCA2a gene transfer abrogates the contractile response of the myocyte to {beta}-adrenergic stimulation. SERCA2a gene transfer results in increased SERCA2a content without alterations in PLB or other calcium-handling proteins, thus specifically increasing the SERCA2a-to-PLB ratio (27). Our finding of disabled {beta}-adrenergic-mediated inotropy in SERCA2a-transduced myocytes is comparable to that reported in myocytes from PLB knockout mice and in SERCA2a-overexpressing mice in vivo, which also have an increased SERCA2a-to-PLB ratio (20, 22). The mechanism by which an increased SERCA2a-to-PLB ratio, achieved by SERCA2a overexpression or by PLB gene ablation, disrupts molecular inotropy is not known. One possibility is that increased SR calcium pump activity, caused by uncoupling of SERCA2a from its negative regulator PLB (4, 14), alters the enhanced excitation-contraction coupling normally associated with {beta}-adrenergic stimulation. Recent studies on myocytes from PLB knockout mice support this possibility. Calcium release events by SR calcium release channels [ryanodine receptors (RyR)] decayed more rapidly and produced increased occurrences of localized RyR calcium release failures in PLB knockout myocytes compared with the wild type (17). That study concluded that enhanced local buffering by SERCA2a caused an increase in subcellular calcium gradients. SERCA2a overexpression, through a similarly heightened local buffering by the increased SR calcium uptake, may also affect signaling from diad to diad in the SR terminal cisternae, as apparent from myocytes from PLB knockout mice (17). In this context, it is not immediately clear why the enhanced calcium buffering by Parv preserved molecular {beta}-adrenergic-mediated inotropy, whereas SERCA2a did not (Fig. 5). One possibility could relate to differences in their spatial localization within the myocyte. SERCA2a pumps are fixed spatially in the SR membrane in the relative vicinity of RyRs (3), whereas Parv is diffuse in the myoplasm (5). It is possible that the relative proximity of the overexpressed SERCA2a pumps to the RyRs could account for this observation. In addition, the newly expressed SERCA2a pumps are probably not all under regulation by PLB and, therefore, are already maximally functioning. With adrenergic stimulation, these pumps would not increase activity to further load the SR with Ca2+. In contrast, with Parv, the SERCA2a-to-PLB ratio is not altered so that enhanced SR loading and calcium release are retained with adrenergic stimulation. The preceding discussion of calcium buffering may also help explain why in SERCA-transduced myocytes heightened extracellular Ca2+ (from 1.8 to 5 mM in the absence of adrenergic stimulation) did not augment contraction. It is not until extracellular Ca2+ is raised to ~8 mM or higher or frequency of stimulation is increased that contractility is enhanced in SERCA2a-transduced myocytes (7). It is unclear why Parv-transduced myocytes responded to adrenergic stimulation but not to elevated extracellular calcium in terms of positive inotropy (Fig. 5). Further experimentation will be required to resolve these issues.

The present findings may have implications for human heart failure. Diastolic heart failure in particular is increasing in incidence and is the leading cause of hospitalization in the elderly. There are no direct clinical treatments for diastolic heart failure. Efforts to improve calcium handling in diseased myocardium hold promise toward the remediation of slow relaxation in human heart failure (6, 9, 12, 16, 29). In this light, the divergent effects of Parv and SERCA2a in canine myocytes in the presence of inotropes may have direct applicability to failing human myocytes. In particular, the distinguishing properties of Parv and SERCA2a gene transfer shown here are likely important given the clinical relevance of inotropic reserve and interventions and therapies for the normal and failing heart. Under the present experimental conditions, SERCA2a was found to affect molecular inotropy. In comparison, at high concentrations, Parv altered contraction in the absence but not presence of inotropes. Future organ level studies in higher mammalian species, and under various pathophysiological conditions, will be important to complement these results from canine isolated myocytes and hearts from transgenic mice. Collectively, the results of this study provide a foundation to further the advancement of experimental therapeutics focused on restoring normal Ca2+ characteristics for patients with heart failure.


   GRANTS
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
 GRANTS
REFERENCES
 
This study was supported by the National Institutes of Health.


   ACKNOWLEDGMENTS
 
We acknowledge Katie Evans Hong, Dr. Phil Wahr, Dr. Margaret Westfall, Dr. Edward Bove, Dr. Steve Bolling, and Dr. Pierre Coutu for assistance, advice, and/or support during the course of this study.


   FOOTNOTES
 

Address for reprint requests and other correspondence: J. M. Metzger, Dept. of Molecular and Integrative Physiology, 7730 Medical Science II, Univ. of Michigan, Ann Arbor, MI 48109 (E-mail: metzgerj{at}umich.edu).

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.


   REFERENCES
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 

  1. American Heart Associaiton. Heart and Stroke Statistical Update. American Heart Association, 1997.
  2. Baker DL, Hashimoto K, Grupp IL, Ji Y, Reed T, Loukianov E, Grupp G, Bhagwhat A, Hoit B, Walsh R, Marban E, and Periasamy M. Targeted overexpression of the sarcoplasmic reticulum Ca2+-ATPase increases cardiac contractility in transgenic mouse hearts. Circ Res 83: 1205–1214, 1998.[Abstract/Free Full Text]
  3. Bers DM. Cardiac excitation-contraction coupling. Nature 415: 198–205, 2002.[CrossRef][Medline]
  4. Brittsan AG, Kiss E, Edes I, Grupp IL, Grupp G, and Kranias EG. The effect of isoproterenol on phospholamban-deficient mouse hearts with altered thyroid conditions. J Mol Cell Cardiol 31: 1725–1737, 1999.[CrossRef][ISI][Medline]
  5. Coutu P and Metzger JM. Optimal range for parvalbumin as relaxing agent in adult cardiac myocytes: gene transfer and mathematical modeling. Biophys J 82: 2565–2579, 2002.[Abstract/Free Full Text]
  6. Dash R, Kadambi V, Schmidt AG, Tepe NM, Biniakiewicz D, Gerst MJ, Canning AM, Abraham WT, Hoit BD, Liggett SB, Lorenz JN, Dorn GW, and Kranias EG. Interactions between phospholamban and beta-adrenergic drive may lead to cardiomyopathy and early mortality. Circulation 103: 889–896, 2001.[Abstract/Free Full Text]
  7. Del Monte F, Harding SE, Dec GW, Gwathmey JK, and Hajjar RJ. Targeting phospholamban by gene transfer in human heart failure. Circulation 105: 904–907, 2002.[Abstract/Free Full Text]
  8. Del Monte F, Harding SE, Schmidt U, Matsui T, Kang ZB, Dec GW, Gwathmey JK, Rosenzweig A, and Hajjar RJ. Restoration of contractile function in isolated cardiomyocytes from failing human hearts by gene transfer of SERCA2a. Circulation 100: 2308–2311, 1999.[Abstract/Free Full Text]
  9. Dillmann WH. Calcium regulatory proteins and their alteration by transgenic approaches. Am J Cardiol 83: 89H–91H, 1999.[ISI][Medline]
  10. Gaasch W. Diagnosis and treatment of heart failure based on left ventricular systolic and diastolic dysfunction. JAMA 271: 1276–1280, 1994.[Abstract]
  11. Green HJ, Klug GA, Reichmann H, Seedorf U, Wiehrer W, and Pette D. Exercise-induced fibre type transitions with regard to myosin, parvalbumin, and sarcoplasmic reticulum in muscles of the rat. Pflügers Arch 400: 432–438, 1984.[CrossRef][ISI][Medline]
  12. Hajjar RJ, del Monte F, Matsui T, and Rosenzweig A. Prospects for gene therapy for heart failure. Circ Res 86: 616–621, 2000.[Abstract/Free Full Text]
  13. Hajjar RJ, Kang JX, Gwathmey JK, and Rosenzweig A. Physiological effects of adenoviral gene transfer of sarcoplasmic reticulum calcium ATPase in isolated rat myocytes. Circulation 95: 423–429, 1997.[Abstract/Free Full Text]
  14. Hajjar RJ, Muller FU, Schmitz W, Schnabel P, and Bohm M. Molecular aspects of adrenergic signal transduction in cardiac failure. J Mol Med 76: 747–755, 1998.[CrossRef][ISI][Medline]
  15. Hajjar RJ, Schmidt U, Matsui T, Guerrero JL, Lee KH, Gwathmey JK, Dec GW, Semigran MJ, and Rosenzweig A. Modulation of ventricular function through gene transfer in vivo. Proc Natl Acad Sci USA 95: 5251–5256, 1998.[Abstract/Free Full Text]
  16. Hunter JJ and Chien KR. Signaling pathways for cardiac hypertrophy and failure. N Engl J Med 341: 1276–1283, 1999.[Free Full Text]
  17. Huser J, Bers DM, and Blatter LA. Subcellular properties of [Ca2+]i transients in phospholamban-deficient mouse ventricular cells. Am J Physiol Heart Circ Physiol 274: H1800–H1811, 1998.[Abstract/Free Full Text]
  18. Kilgore KS, Shwartz CF, Gallagher MA, Steffen RP, Mosca RS, and Bolling SF. RSR13, a synthetic allosteric modifier of hemoglobin, improves myocardial recovery following hypothermic cardiopulmonary bypass. Circulation 100: II351–II356, 1999.
  19. Koss KL, Grupp IL, and Kranias EG. The relative phospholamban and SERCA2 ratio: a critical determinant of myocardial contractility. Basic Res Cardiol 92, Suppl 1: 17–24, 1997.
  20. Li L, Chu G, Kranias EG, and Bers DM. Cardiac myocyte calcium transport in phospholamban knockout mouse: relaxation and endogenous CaMKII effects. Am J Physiol Heart Circ Physiol 274: H1335–H1347, 1998.[Abstract/Free Full Text]
  21. Lorell BH. Significance of diastolic dysfunction of the heart. Annu Rev Med 42: 411–436, 1991.[CrossRef][ISI][Medline]
  22. Luo W, Grupp IL, Harrer J, Ponniah S, Grupp G, Duffy JJ, Doetschman T, and Kranias EG. Targeted ablation of the phospholamban gene is associated with markedly enhanced myocardial contractility and loss of beta-agonist stimulation. Circ Res 75: 401–409, 1994.[Abstract/Free Full Text]
  23. McGrory WJ, Bautista DS, and Graham FL. A simple technique for the rescue of early region I mutations into infectious human adenovirus type 5. Virology 163: 614–617, 1988.[CrossRef][ISI][Medline]
  24. Mercadier J, Lompre A, Duc P, Boheler K, Fraysse J, Wisnewsky C, Allen P, Komajda M, and Schwartz K. Altered sarcoplasmic reticulum calcium ATPase gene expression in the human ventricle during end-stage heart failure. Am J Cardiol 92: 305–309, 1995.
  25. Meyer M, Schillinger W, Pieske B, Holubarsch C, Heilman C, Posival H, Kuwajima G, Mikoshiba K, Just H, and Hasenfuss G. Alterations of sarcoplasmic reticulum proteins in failing human dilated cardiomyopathy. Circulation 92: 778–784, 1995.[Abstract/Free Full Text]
  26. Miyamoto MI, del Monte F, Schmidt U, DiSalvo TS, Kang ZB, Matsui T, Guerrero JL, Gwathmey JK, Rosenzweig A, and Hajjar RJ. Adenoviral gene transfer of SERCA2a improves left-ventricular function in aortic-banded rats in transition to heart failure. Proc Natl Acad Sci USA 97: 793–798, 2000.[Abstract/Free Full Text]
  27. Schmidt U, del Monte F, Miyamoto MI, Matsui T, Gwathmey JK, Rosenzweig A, and Hajjar RJ. Restoration of diastolic function in senescent rat hearts through adenoviral gene transfer of sarcoplasmic reticulum Ca2+-ATPase. Circulation 101: 790–796, 2000.[Abstract/Free Full Text]
  28. Schmidt U, Hajjar RJ, Helm PA, Kim CS, Doye AA, and Gwathmey JK. Contribution of abnormal sarcoplasmic reticulum ATPase activity to systolic and diastolic dysfunction in human heart failure. J Mol Cell Cardiol 30: 1929–1937, 1998.[CrossRef][ISI][Medline]
  29. Szatkowski ML, Westfall MV, Gomez CA, Wahr PA, Michele DE, DelloRusso C, Turner II, Hong KE, Albayya FP, and Metzger JM. In vivo acceleration of heart relaxation performance by parvalbumin gene delivery. J Clin Invest 107: 191–198, 2001.[ISI][Medline]
  30. Wahr PA, Michele DE, and Metzger JM. Parvalbumin gene transfer corrects diastolic dysfunction in diseased cardiac myocytes. Proc Natl Acad Sci USA 96: 11982–11985, 1999.[Abstract/Free Full Text]
  31. Westfall MV, Rust EM, and Metzger JM. Slow skeletal troponin I gene transfer, expression, and myofilament incorporation enhances adult cardiac myocyte contractile function. Proc Natl Acad Sci USA 94: 5444–5449, 1997.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 Physiol. Rev.Home page
J. Davis, M. V. Westfall, D. Townsend, M. Blankinship, T. J. Herron, G. Guerrero-Serna, W. Wang, E. Devaney, and J. M. Metzger
Designing Heart Performance by Gene Transfer
Physiol Rev, October 1, 2008; 88(4): 1567 - 1651.
[Abstract] [Full Text] [PDF]

Home page
 Circ. Res.Home page
L. E. Vinge, P. W. Raake, and W. J. Koch
Gene Therapy in Heart Failure
Circ. Res., June 20, 2008; 102(12): 1458 - 1470.
[Abstract] [Full Text] [PDF]

Home page
 Physiol. GenomicsHome page
S. M. Day, P. Coutu, W. Wang, T. Herron, I. Turner, M. Shillingford, N. C. LaCross, K. L. Converso, L. Piao, J. Li, et al.
Cardiac-directed parvalbumin transgene expression in mice shows marked heart rate dependence of delayed Ca2+ buffering action
Physiol Genomics, May 1, 2008; 33(3): 312 - 322.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Physiol. Heart Circ. Physiol.Home page
D. W. Rodenbaugh, W. Wang, J. Davis, T. Edwards, J. D. Potter, and J. M. Metzger
Parvalbumin isoforms differentially accelerate cardiac myocyte relaxation kinetics in an animal model of diastolic dysfunction
Am J Physiol Heart Circ Physiol, September 1, 2007; 293(3): H1705 - H1713.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (5)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hirsch, J. C.
Right arrow Articles by Metzger, J. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hirsch, J. C.
Right arrow Articles by Metzger, J. M.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Visit Other APS Journals Online
Copyright © 2004 by the American Physiological Society.