during the development of hypertensive heart disease, heart failure (HF) is often preceded by concentric cardiac hypertrophy (11). Although the molecular mechanisms of this transition are incompletely understood (9), defects in excitation-contraction (E-C) coupling are thought to play a central role in cardiac hypertrophy and HF (12, 43). In HF, increased phosphorylation of ryanodine receptors (RyR₂) at Ser²⁸⁰⁸ (RyR₂-Ser²⁸⁰⁸) has been proposed to enhance RyR₂ channel sensitivity to cytosolic Ca²⁺ concentration, resulting in Ca²⁺ leakage from the sarcoplasmic reticulum (SR) and, thus, reduction of the SR Ca²⁺ load (23, 24). Reduced SR Ca²⁺ load is expected to decrease E-C coupling gain, leading to depressed myocardial contractility and pump failure (3, 40). However, the concept that RyR₂-Ser²⁸⁰⁸ is hyperphosphorylated in failing hearts has been a subject of debate in the literature.

Hyperphosphorylation of RyR₂ channels at Ser²⁸⁰⁸ in failing hearts has been reported in multiple clinical and experimental studies (1, 23, 31, 42, 45). However, other studies reported no obvious change in RyR₂-Ser²⁸⁰⁸ phosphorylation in some animal models of HF (15, 48). The discrepancy could be due to differences in the etiology or the severity of heart disease in different models. It has become clear that a better understanding of the temporal profile of RyR₂-Ser²⁸⁰⁸ phosphorylation during the progressive development of cardiac hypertrophy and HF is warranted to resolve this controversy. Therefore, we carefully examined the phosphorylation level of RyR₂-Ser²⁸⁰⁸ at various stages of hypertensive heart disease from the development of cardiac hypertrophy to HF. Moreover, we measured changes in RyR₂-Ser²⁸⁰⁸ phosphorylation in different regions of the heart, inasmuch as regional heterogeneity in Ca²⁺-handling proteins might also affect cardiac contractility.

It has been proposed that defects in E-C coupling in HF may result not only from dysregulation of Ca²⁺-handling proteins, but also from restructuring of the spatial organization of RyR₂ clusters that form Ca²⁺ release units (CRUs) within the cardiomyocyte (7, 14). Although changes in CRU organization are expected to affect the Ca²⁺ dynamics, such changes have yet to be demonstrated in HF. Also unknown is the correlation between the spatial organization and the modulation (i.e., PKA phosphorylation levels) of RyR₂ in heart disease development.

The progression of concentric hypertrophy to HF has been well demonstrated in the spontaneously hypertensive rat (SHR) model, which mimics human essential hypertension and heart disease (4). We hypothesized that in the SHR model the transition from cardiac hypertrophy to HF might influence the level of RyR₂ phosphorylation, as well as the subcellular organization of CRUs. The purpose of this study is to determine whether the progressive development of hypertensive heart disease in SHR is accompanied by 1) changes in the phosphorylation of RyR₂-Ser²⁸⁰⁸ and 2) changes in the spatial distribution of CRUs in ventricular myocytes.

MATERIALS AND METHODS

Animals and tissue preparation.

Male SHR, normotensive Wistar-Kyoto rats (WKY), and Sprague-Dawley rats were purchased from Charles River ( http://www.criver.com). Blood pressure was monitored weekly in nonanesthetized SHR and WKY by the tail-cuff method. For tissue preparation, rats were anesthetized with pentobarbital sodium (100 mg/kg ip with 4,000 U/kg heparin). Isoproterenol (1 mg/kg ip) was injected before cardiac explantation for the β-adrenergic receptor stimulation experiments. After suppression of spinal cord reflexes, the heart was exposed via a midline thoracotomy, and the chambers were rinsed with Ca²⁺-free PBS, which was injected into the left ventricle (LV) and vented through a small incision in the right atrial free wall. The heart was subsequently removed, and laterally dissected to expose the ventricular walls and chambers. Tissue freezing medium (OCT compound) was injected into the chambers to preserve cardiac morphology during the subsequent freezing process. The tissue was flash frozen by submersion in chilled isopentane for 10–20 s, placed on dry ice, and then stored in a −80°C freezer. The frozen tissue was cut into 20-μm-thick sections in a cryostat (model 2800 Frigocut-E, Reichert, Bannockburn, IL). All chemicals and reagents were purchased from Sigma-Aldrich if not specified otherwise. All animals were handled strictly in accordance to the National Institutes of Health guidelines and protocols approved by our Institutional Animal Care and Use Committee at the University of Kentucky.

Western blot.

Western blot was used to measure the amount of phosphorylation of RyR₂-Ser²⁸⁰⁸ relative to the total amount of RyR₂ in LV tissue according to a previously published protocol (32). Phosphorylated RyR₂-Ser²⁸⁰⁸ (RyR₂-pSer²⁸⁰⁸) was stained using a phosphorylated epitope-specific antibody [affinity-purified polyclonal rabbit antibody raised against phosphorylated peptide sequence C-RTRRI-(pS)-QTSQV]; total RyR₂ was stained using a polyclonal anti-RyR₂ antibody.

Semiquantitative immunohistochemistry.

Tissue sections were incubated in a blocking solution containing 5% goat serum and 3% BSA in PBS for 30 min and rinsed twice in PBS; then the sections were incubated in primary antibody solution (1:200 dilution) for 1.5 h, rinsed twice in PBS, and incubated in secondary antibody solution (1:200 dilution) for 1.5 h. The same phosphorylated epitope-specific (polyclonal, rabbit) antibody was used in semiquantitative immunohistochemistry (SQ-IHC) and quantitative Western blot to label RyR₂-pSer²⁸⁰⁸, and a monoclonal pan anti-RyR₂ antibody (mouse IgG1; clone C3-33, Affinity BioReagents) was used to label total RyR₂ in SQ-IHC. Phosphorylated and pan RyR₂ labeling were visualized using secondary antibody-conjugated fluorophores: anti-rabbit IgG-conjugated Alexa Fluor 569 and anti-mouse IgG-conjugated Alexa Fluor 488, respectively. The antibody-labeled tissue sections were covered with Antifade and sealed under a glass coverslip (no. 1) for imaging.

Confocal imaging and image analysis for SQ-IHC.

Confocal images were obtained using a confocal microscope (Radiance 2000, Bio-Rad) with a water immersion objective (×63, NA 1.2) corrected for the thickness of the no. 1 glass coverslip. To obtain confocal images of antibody labeling, we placed the focal plane in the middle of the 20-μm-thick tissue slice to avoid the interface between the tissue and the glass. The imaging areas were chosen by random scanning of the tissue section, and various areas in the tissue section were used to obtain an average value. For quantitative analysis and comparison between the control group and the test group, we strictly used identical antibody labeling conditions and confocal imaging parameters for all the tissue samples in the groups of comparison.

To quantitatively measure the labeling intensity, we recorded the fluorescence emission from a defined area (using a ×63 objective and 1,024 × 1,024 pixels in the x-y plane) with a defined depth (pinhole size optimized to confocal z resolution of ∼1 μm). Histograms of the optical intensity [i.e., fluorescence intensity (FI)] of each pixel were plotted. The background signal was subtracted from the total histogram before the average FI was calculated. The average FI was then calculated for each image. To calculate the relative phosphorylation of RyR₂-Ser²⁸⁰⁸, we double labeled RyR₂-Ser²⁸⁰⁸ and the total RyR₂ in the tissue using specific primary antibodies and secondary antibody-conjugated Alexa Fluor 568 and Alexa Fluor 488, respectively. The ratio of the average FI of Alexa 568 to that of Alexa 488 reflects the amount of RyR₂-Ser²⁸⁰⁸ relative to the total amount of RyR₂.

The images used to calculate the antibody labeling intensity were obtained without digital zoom (gain = 1), and the images used to measure the spacing between RyR₂ clusters (CRUs) were obtained with maximum digital zoom (gain = 10). The spacing between RyR₂ clusters along the longitudinal axis of the cells was measured using our previously described method (7).

Cell isolation.

The rats were anesthetized with pentobarbital sodium (Nembutal; 100 mg/kg ip). Hearts were tested for the suppression of reflexes and then explanted via a midline thoracotomy. A standard enzymatic technique was used to isolate the ventricular myocyte. Briefly, the heart was mounted on a Langendorff system and perfused with a modified Tyrode solution containing (in mmol/l) 135 NaCl, 4 KCl, 1.0 MgSO₄, 0.34 NaH₂PO₄, 15 glucose, 10 HEPES, and 10 taurine, with pH adjusted to 7.25 with NaOH; the perfusion solution was prewarmed to 37°C and bubbled with 100% O₂. Then collagenase B (∼1 mg/ml; Hoffmann-La Roche, Basel, Switzerland), protease type XIV (∼0.1 mg/ml), 0.1% BSA, and 20 μM Ca²⁺ were added into the perfusion solution, and the heart was enzymatically digested for 15–20 min. The ventricular tissue was cut down and minced, the remaining tissue was further incubated in the enzyme solution at 37°C for 15–45 min and minced again, and the ventricular myocytes were collected.

Immunocytochemistry.

Freshly isolated ventricular myocytes were labeled using anti-RyR₂ monoclonal antibody (mouse IgG1; clone C3-33, Affinity BioReagents) as described preciously in detail (7, 14).

Statistical analysis.

All samples were coded using randomized codes for blinded processing during experiments and data analysis. Values are means ± SD. Unpaired Student's t-test with equal variance and two tails was used to compare the age-matched SHR vs. WKY; the difference in the mean values is deemed significant if P < 0.05. Two-way ANOVA was used to evaluate the differences in the longitudinal study of the age-related changes in WKY and in the disease-related changes in SHR.

RESULTS

Progressive development of hypertrophy and HF in SHR.

Hypertensive heart disease develops progressively in SHR through distinctive stages of hypertension, cardiac hypertrophy, and HF. The history of blood pressure development is shown in Fig. 1A. SHR were prehypertensive during their first 4–6 wk of life, rapidly developed hypertension at 8–12 wk of age, and remained hypertensive thereafter. In comparison, WKY controls remained normotensive throughout their ∼2-yr life span. After the onset of hypertension, cardiac hypertrophy gradually developed in SHR at 3–18 mo of age, and HF typically occurred at 1.5–2 yr of age. The history of hypertrophy development in SHR is shown in Fig. 1B. As a normotensive control, WKY displayed a slight decrease in heart weight-to-body weight ratio (HW/BW), as the increase in body weight outpaced the increase in heart weight as the animals aged. HW/BW was the same in 5-wk-old (prehypertensive) and 11-wk-old (onset of hypertension) SHR as in age-matched WKY controls. HW/BW was significantly higher in 1-yr-old SHR than in age-matched WKY, demonstrating cardiac hypertrophy: 5.38 ± 0.77 (n = 10) vs. 3.84 ± 0.51 (SD) mg/g (n = 7; P < 0.05, t-test). During HF, HW/BW was higher in SHR than in age-matched WKY [6.19 ± 1.68 (n = 5) vs. 3.63 ± 1.13 mg/g (n = 18)], but the individual variation was greater during HF, most likely due to cardiac cachexia (P < 0.1, t-test). Consistent with previous studies showing echocardiographic evidence for HF in >1.5-yr-old SHR (6, 8, 34, 35), we found clinical signs of HF (e.g., severe hypertrophy or chamber dilation, ascites, pericardial effusion, chest edema, and lung edema) in SHR at postmortem examination.

Western blot measurement of RyR₂-pSer²⁸⁰⁸ phosphorylation.

We used the standard Western blot technique to measure RyR₂-pSer²⁸⁰⁸. Figure 1C shows Western blot measurement of RyR₂-pSer²⁸⁰⁸ relative to total RyR₂ in the LV of SHR at four distinct disease stages: 5-wk-old prehypertensive SHR, 11-wk-old SHR after the onset of hypertension but before hypertrophy, 1-yr-old SHR with overt hypertrophy, and >1.5-yr-old SHR at HF. The results show that RyR₂-Ser²⁸⁰⁸ phosphorylation drastically increased in >1.5-yr-old SHR [3.00 ± 0.94 in SHR vs. 1.44 ± 0.49 in WKY (n = 4, P < 0.05 by t-test and P < 0.001 by 2-way ANOVA with Bonferroni's posttest)] after the progression from hypertrophy to HF.

SQ-IHC method development.

To provide another independent measure for RyR₂-Ser²⁸⁰⁸ phosphorylation levels, we developed an SQ-IHC method by combining immunohistochemistry and confocal microscopy techniques (Fig. 2A). The specificity of antibody labeling was verified by preincubation of the antibody with its phosphorylated epitope peptide, which also served as the background image for subtraction of the nonspecific labeling. Basal RyR₂-pSer²⁸⁰⁸ in the control tissue was relatively low (Fig. 2B). In contrast, intense labeling of RyR₂-pSer²⁸⁰⁸ was observed in cardiac tissue obtained from the heart pretreated with isoproterenol, which induced PKA phosphorylation of RyR₂ due to β-adrenergic stimulation (Fig. 2B). The punctate staining in a striated pattern is also consistent with the known RyR₂ clustering and localization on Z disks (7).

The original SQ-IHC method used a digital camera to image and quantify the molecules in tissue samples (25). We improved on the original SQ-IHC method by employing confocal microscopy to image optical sections inside the tissue (avoiding tissue surface artifacts) and strictly controlled the amount of fluorescence emission by using a fixed area and a fixed depth in the optical section [providing normalization for averaging across different sections, so the labeling intensity in each image is measured from a unit volume (normalized) and, hence, can be used to compare different tissue sections]. FI values of all pixels in a representative image are shown in Fig. 2C. The background image showed very low (<34) FI, in accordance with low background labeling. The antibody labeling image showed well-distributed FI values across the digitizing range for intensity without saturation. To subtract the background, we truncated the FI histogram of each image by removing all points <34 FI. This subtraction also digitally removed “holes” in the tissue section caused by physiological structures (e.g., capillary, nerve, and extracellular space). The mean FI of each image was then calculated from the background-subtracted histogram. We used identical treatment for all tissue samples throughout the process, from antibody labeling (reagents and protocol) to confocal imaging (optical and parameter settings), to eliminate variations.

To measure the RyR₂-pSer²⁸⁰⁸ relative to the total amount of RyR₂ protein in a tissue section, we double labeled sections using a phosphorylated epitope-specific antibody recognizing RyR₂-pSer²⁸⁰⁸ and a pan RyR₂ antibody tagged with secondary antibody-conjugated Alexa Fluor 568 and Alexa Fluor 488, respectively (Fig. 2D, pseudocolored red for RyR₂-pSer²⁸⁰⁸ and green for pan RyR₂). The SQ-IHC method was used to obtain mean FI values for each antibody labeling; then we calculated the ratio of mean FI values of RyR₂-pSer²⁸⁰⁸ to the total RyR₂ labeling, which provided a measure for the relative amount of RyR₂-pSer²⁸⁰⁸ in the tissue. A great benefit of this ratiometric method is that it eliminates the variations in different tissue regions (e.g., cell density, orientation, and accessibility) and, hence, enables us to average across different tissue sections. The ratiometric method was used to measure relative RyR₂-pSer²⁸⁰⁸ in control and epinephrine-treated hearts, and the results are shown in Fig. 2E. Infusion of epinephrine into the heart caused β-adrenergic stimulation and increased PKA phosphorylation of RyR₂-Ser²⁸⁰⁸.

The advantage of the SQ-IHC method is that it bypasses several protein-processing steps of the Western blot method, including tissue homogenization, protein extraction, total protein calibration, and protein transfer from gel to membrane, which should help reduce potential artifacts associated with these procedures. The disadvantage of the SQ-IHC method is that molecular weights cannot be separated and very specific antibodies are required. Although we treated all samples using identical conditions throughout the antibody labeling and image acquisition procedures and employed the ratiometric method, the SQ-IHC measurement remains semiquantitative. However, the SQ-IHC method can be used in combination with Western blot to verify the data.

SQ-IHC measurement of RyR₂-Ser²⁸⁰⁸ phosphorylation.

Sample images depicting double labeling of RyR₂-pSer²⁸⁰⁸ and total RyR₂ in the SHR failing heart and age-matched WKY control are shown in Fig. 2D. Although total RyR₂ labeling (green) was comparable in SHR and WKY, RyR₂-pSer²⁸⁰⁸ labeling (red) was more intense in SHR than WKY, demonstrating an increase in RyR₂-pSer²⁸⁰⁸ relative to total RyR₂ in SHR.

Relative RyR₂-Ser²⁸⁰⁸ phosphorylation in various regions of the heart at three distinct stages of heart disease is shown in Fig. 3. In 11-wk-old SHR at the onset of hypertension before hypertrophy, RyR₂-pSer²⁸⁰⁸ levels were similar to those in age-matched WKY across LV, septum, and right ventricle [RV; Fig. 3A; P > 0.05 for strain difference at each region (t-test) and P > 0.05 for strain difference across all regions (2-way ANOVA)]. In 1-yr-old SHR with overt cardiac hypertrophy, RyR₂-pSer²⁸⁰⁸ remained normal in LV and septum but was significantly increased in RV [Fig. 3B; P = 0.01 for strain difference across all regions (2-way ANOVA) and P < 0.05 for RV (Bonferroni's posttest)]. In >1.5-yr-old SHR during HF, however, RyR₂-pSer²⁸⁰⁸ was significantly increased across all regions of the heart, including LV, septum, and RV [Fig. 3C; P = 0.002 for strain difference across all regions (2-way ANOVA) and P > 0.05 possibly due to small sample numbers (Bonferroni's posttest)]. Interestingly, RyR₂-pSer²⁸⁰⁸ was significantly higher in LV in >1.5-yr-old SHR (P < 0.05, t-test) but higher in RV 1-yr-old SHR [P < 0.05 (t-test) and P < 0.05 (Bonferroni's posttest)] than in age-matched WKY controls. Figure 3D shows the relative RyR₂-pSer²⁸⁰⁸ levels in SHR normalized to age-matched WKY. The changes in LV measured using the SQ-IHC method are in agreement with those measure using Western blot (Fig. 1C).

SQ-IHC measurement of spatial distribution of RyR₂ clusters.

The SQ-IHC method enabled us to measure not only the phosphorylation levels of RyR₂ but also the spatial localization of RyR₂ clusters (each contains ∼100 RyR₂ molecules and serves as a CRU). We used a previously described method (7) (spatial resolution ∼0.25 μm) to measure the longitudinal spacing between the neighboring RyR₂ clusters in the tissue cross sections (Fig. 4A). The histogram of CRU spacing showed a Gaussian distribution (Fig. 4B). The shorter CRU spacing in the failing heart of the SHR is manifested by a leftward shift of the distribution from that of WKY. To confirm the change of CRU spacing in the failing heart, we also used 2-mo-old healthy Sprague-Dawley rats as an additional control. As shown in Fig. 4C, CRU spacing is 1.72 ± 0.34 (SD) μm in the failing heart of the SHR, 9.5% shorter than in the age-matched WKY [1.90 ± 0.32 μm, P < 0.01 (t-test)] and 11.3% shorter than in the Sprague-Dawley control (1.94 ± 0.31 μm, P < 0.01). Since increased collagen expression in failing hearts (40) might constrain the cell length or shrink the tissue more on processing, we used collagenase and protease to enzymatically digest the heart to isolate ventricular myocytes. We conducted immunocytochemistry experiments to label RyR₂ clusters in isolated cells and measured the longitudinal spacing between the antibody-labeled CRUs (Fig. 4C). Again, CRU spacing was shorter in cells from the failing heart of the SHR (1.77 ± 0.15, n = 130) than in cells from the Sprague-Dawley control [1.87 ± 0.18, n = 786, P < 0.001 (t-test)]. Furthermore, average CRU spacing in WKY and Sprague-Dawley control hearts is consistent with normal resting sarcomere length in ventricular myocytes (1.8–2.2 μm; data not shown), which demonstrates preserved cellular ultrastructure in the tissue section. (Since RyR₂ are localized in Z disks, the longitudinal spacing between CRUs corresponds to the sarcomere length.) Given that the identical protocol was used to prepare all tissues, the shorter CRU spacing in failing hearts of SHR cannot be attributed to artifacts that might change the cellular ultrastructure or to differential collagen expression. Therefore, it is most likely that shortening of the CRU longitudinal spacing in the failing heart is a real phenomenon of pathological significance.

DISCUSSION

Hyperphosphorylation of RyR₂-Ser²⁸⁰⁸ after progression from hypertrophy to HF in SHR.

In the present study, we investigated two important factors that may contribute to changes in Ca²⁺ signaling during cardiac hypertrophy and HF: 1) the phosphorylation status of the cardiac RyR (Ser²⁸⁰⁸) and 2) the spatial distribution of RyR₂ clusters within the cardiomyocytes. First, we conducted a longitudinal study to measure RyR₂-pSer²⁸⁰⁸ in SHR at the four distinct stages in the development of hypertensive heart disease: prehypertension, hypertension before hypertrophy, cardiac hypertrophy, and finally HF. Western blot studies revealed that RyR₂-pSer²⁸⁰⁸ in the LV of SHR was unaltered during the prehypertrophic and hypertrophic stages of hypertensive heart disease and became significantly elevated during HF. Immunohistochemistry studies further showed that RyR₂-pSer²⁸⁰⁸ was normal during the prehypertrophic stage of hypertensive heart disease, whereas RyR₂-pSer²⁸⁰⁸ became elevated in the RV of SHR with cardiac hypertrophy. After the development of overt HF in SHR, RyR₂-pSer²⁸⁰⁸ significantly increased across the LV, septum, and RV in >1.5-yr-old rats. Overall, there was good agreement between the data obtained using Western blot and SQ-IHC. Hence, a major finding of this study is that hyperphosphorylation of RyR₂-Ser²⁸⁰⁸ occurs late during the development of hypertensive heart disease, after the development of cardiac failure.

Our data provide new insights into the controversy in the literature regarding increased phosphorylation of RyR₂-Ser²⁸⁰⁸ in structural heart disease. The finding that phosphorylation of RyR₂-Ser²⁸⁰⁸ is only increased in the LV of SHR with overt HF suggests that this posttranslational modification of the RyR may occur late during the development of HF. Our findings are consistent with enhanced PKA phosphorylation of RyR₂ in patients with end-stage HF (24, 31). Hyperphosphorylation of RyR₂ has also been reported in several animal models of HF, including a canine model of pacing-induced HF (30, 49), a rabbit aortic-banding model (1), and rat (26, 31) and mouse models of ischemic HF (45). Nonetheless, other studies found no changes in RyR₂-Ser²⁸⁰⁸ phosphorylation in patients and animals with HF. In light of the findings of our studies in SHR, it is likely that the degree of RyR₂-Ser²⁸⁰⁸ phosphorylation varies with the etiology and severity of heart disease. Indeed, Ward et al. (41) reported that SR Ca²⁺ load and RyR₂ function were not affected in SHR with mild symptoms of HF.

Factors that regulate RyR₂-Ser²⁸⁰⁸ phosphorylation: local control by kinases and phosphatases in the macromolecular complex.

Phosphorylation of RyR₂-Ser²⁸⁰⁸ is regulated by multiple kinases and phosphatases in the RyR₂ macromolecular channel complex. Anchored via leucine/isoleucine zipper motifs on the cytoplasmic NH₂-terminal domain of RyR₂ are PKA [via its targeting protein, myocardial A kinase-anchoring protein (mAKAP)], Ca²⁺-calmodulin-dependent protein kinase II (CaMKII), phosphatase PP1 (via protein spinophilin), phosphatase PP2A (via PR130), phosphodiesterase 4D3 (PDE4D3, via mAKAP), and several others, including calmodulin, FKBP12.6 (calstabin2), and sorcin (2, 18, 22, 24). Consequently, an increase in RyR₂-pSer²⁸⁰⁸ could result from several events, e.g., increased PKA activity (31); downregulation of PDE4D3, which increases the local cAMP level and the PKA activity (18); or downregulation of PP1 and PP2A, which has been observed in patients with HF (24, 30, 32).

Furthermore, the RyR₂-Ser²⁸⁰⁸ phosphorylation site has been proposed to be a consensus phosphorylation site not only for PKA (RRXS), but also for CaMKII (RXXS) and PKG (RR/KXS/T) (16). Western blot experiments showed that RyR₂-Ser²⁸⁰⁸ could be phosphorylated, at least in vitro in SR vesicles or using purified RyR₂, by PKA and, possibly, by CaMKII (33, 47) or PKG (48). Using knock-in mice in which Ser²⁸⁰⁸ has been mutated to Ala, it has been demonstrated that Ser²⁸⁰⁸ is the major, physiologically active PKA phosphorylation site on RyR₂ (45). It remains to be seen whether CaMKII or PKG indeed phosphorylate RyR₂-Ser²⁸⁰⁸ in vivo. Finally, it has recently been suggested that Ser²⁰³⁰ on RyR₂ constitutes a second PKA phosphorylation site, although the physiological importance of this site remains controversial (48).

Functional consequences of RyR₂-Ser²⁸⁰⁸ phosphorylation: implications in HF.

The functional consequences of PKA hyperphosphorylation of RyR₂ are the subject of scientific debate. Although PKA phosphorylation of RyR₂ increased the single RyR₂ channel activity in vitro in the planar lipid bilayers (24), the Ca²⁺ spark activity (each spark is generated by cooperative opening of a number of RyR₂ channels in a CRU) may not be changed by PKA phosphorylation in vivo (19). Plausible explanations for this controversy include 1) incomplete PKA phosphorylation of RyR₂ in isolated cardiomyocytes and 2) an increase in the opening of uncoupled RyR₂ channels (“rogue RyRs”) by PKA phosphorylation of RyR₂ (36), which might cause small amounts of Ca²⁺ leak that could not be detected by the current technique.

It has been demonstrated that, in failing hearts, RyR₂ channels are more prone to abnormal SR Ca²⁺ release during diastole, leading to SR Ca²⁺ leak. One plausible explanation of enhanced SR Ca²⁺ release includes PKA hyperphosphorylation of RyR₂-Ser²⁸⁰⁸ (44). However, in failing rabbit hearts, SR Ca²⁺ leak could not be reduced using the PKA inhibitor H89 but was significantly blocked by CaMKII inhibition, suggesting that abnormal CaMKII phosphorylation of RyR₂ may also cause RyR₂ dysregulation in HF (1). Consistent with this notion, CaMKII phosphorylation of RyR₂ was shown to increase the Ca²⁺ spark activity in vivo (17, 13, 21), as well as the RyR₂ single-channel activity in vitro (46). Given that it has been proposed that CaMKII might also phosphorylate RyR₂-Ser²⁸⁰⁸ (33), the functional consequence of RyR₂-Ser²⁸⁰⁹ hyperphosphorylation (alone or in combination with other phosphorylation sites on RyR₂) remains to be determined.

Shortened RyR₂ cluster spacing is expected to promote spontaneous Ca²⁺ wave generation in the failing heart.

It has been proposed that derangement of Ca²⁺ dynamics during HF could result not only from altered RyR₂ phosphorylation, but also from restructuring in the cellular organization that alters the coupling of RyR₂ with the neighboring RyR₂ or with other Ca²⁺-handling molecules (14, 37). In failing hearts of SHR, spatial dispersion of the transverse t tubule system was found to disrupt the junctional coupling between CRUs and L-type Ca²⁺ channels, causing asynchronous Ca²⁺ spark activity (37). In a previous study, we predicted that the spatial distribution of RyR₂ clusters (CRUs) could also greatly affect Ca²⁺ wave generation (14). In the present study, we measured the distance between the RyR₂ clusters and found ∼10% shortening of the CRU longitudinal spacing in failing hearts of SHR. This shortening of CRU spacing might be caused by several pathological factors in failing hearts. 1) Elevated diastolic Ca²⁺ would cause impaired relaxation. 2) Decreased phosphorylation levels of contractile proteins, including cardiac troponin I, troponin T, and myosin light chain 2, in SHR failing hearts (20) are expected to increase myofilament sensitivity to Ca²⁺ and, therefore, shorten sarcomere length and RyR₂ cluster longitudinal spacing (29, 38, 39). 3) Increased collagen expression in failing hearts (40) might constrain the cell length or shrink the tissue more upon processing. However, experiments using cells enzymatically digested by collagenase and protease also showed shortening of the longitudinal CRU spacing in the cells from failing hearts. Hence, the CRU spacing change in the failing heart cannot be entirely attributed to collagen and is most likely caused by the elevated diastolic Ca²⁺ and increased myofilament sensitivity to Ca²⁺.

Quantitative model simulations show that 10% shortening of CRU longitudinal spacing would greatly increase the probability of Ca²⁺ wave generation (14). Wave generation is further magnified when combined with an increased RyR₂ activity (14), and spontaneous Ca²⁺ waves are known to induce abnormal electrical activity and contribute to arrhythmias (for review see Ref. 5). In support of this scenario, increased arrhythmogenic activity has been associated with the development of cardiac hypertrophy and HF in SHR (10, 27, 28).

Summary.

Our longitudinal study in rats with hypertensive heart disease provides new insights into the pathogenesis of abnormal Ca²⁺ handling in HF. In particular, our data demonstrate that phosphorylation of RyR₂-Ser²⁸⁰⁸ is increased only in SHR with end-stage HF. Moreover, our data suggest shortening of the spatial distance of CRUs in SHR failing hearts, which is expected to promote Ca²⁺-induced Ca²⁺ release from neighboring CRUs (14) and contractile dysfunction. Additional studies are required to delineate the contribution of various modifications of RyR₂ and other Ca²⁺-handling proteins during the development of hypertensive heart disease.

GRANTS

This work was funded by American Heart Association National Scientist Development Grants 0335250N (Y. Chen-Izu) and 0535310N (X. H. T. Wehrens) and National Heart, Lung, and Blood Institute Grants K25 HL-068704 (L. T. Izu) and R01 HL-071865 (C. W. Balke and L. T. Izu).

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.