Protein Kinase G Modulates Human Myocardial Passive Stiffness by Phosphorylation of the Titin Springs
The sarcomeric titin springs influence myocardial distensibility and passive stiffness. Titin isoform composition and protein kinase (PK)A-dependent titin phosphorylation are variables contributing to diastolic heart function. However, diastolic tone, relaxation speed, and left ventricular extensibility are also altered by PKG activation. We used back-phosphorylation assays to determine whether PKG can phosphorylate titin and affect titin-based stiffness in skinned myofibers and isolated myofibrils. PKG in the presence of 8-pCPT-cGMP (cGMP) phosphorylated the 2 main cardiac titin isoforms, N2BA and N2B, in human and canine left ventricles. In human myofibers/myofibrils dephosphorylated before mechanical analysis, passive stiffness dropped 10% to 20% on application of cGMP-PKG. Autoradiography and anti-phosphoserine blotting of recombinant human I-band titin domains established that PKG phosphorylates the N2-B and N2-A domains of titin. Using site-directed mutagenesis, serine residue S469 near the COOH terminus of the cardiac N2-B–unique sequence (N2-Bus) was identified as a PKG and PKA phosphorylation site. To address the mechanism of the PKG effect on titin stiffness, single-molecule atomic force microscopy force–extension experiments were performed on engineered N2-Bus–containing constructs. The presence of cGMP-PKG increased the bending rigidity of the N2-Bus to a degree that explained the overall PKG-mediated decrease in cardiomyofibrillar stiffness. Thus, the mechanically relevant site of PKG-induced titin phosphorylation is most likely in the N2-Bus; phosphorylation of other titin sites could affect protein–protein interactions. The results suggest that reducing titin stiffness by PKG-dependent phosphorylation of the N2-Bus can benefit diastolic function. Failing human hearts revealed a deficit for basal titin phosphorylation compared to donor hearts, which may contribute to diastolic dysfunction in heart failure.
Myocardial and chamber diastolic function are influenced by chamber geometry, hypertrophy, the extracellular matrix and the sarcomeric titin springs. Titins are giant proteins which exist in the heart in 2 main isoforms coexpressed in sarcomeres: a shorter, stiffer N2B-titin (3.0 MDa) and longer, more compliant N2BA isoforms (3.2 to 3.7 MDa). Differential expression of these isoforms is related to alternate gene splicing affecting the functionally elastic titin region, which is confined to the sarcomeric I-band.1 The springy titin segment comprises regions of serially linked immunoglobulin-like (Ig) domains separated by a cardiac-specific N2-B domain and a so-called PEVK segment (rich in proline, glutamate, valine, and lysine residues). The N2BA isoforms additionally have an N2-A domain and contain more Ig domains and PEVK-rich modules compared to the N2B isoform.
Differential expression of titin isoforms determines passive stiffness of the sarcomere.2,3 A low ratio of N2BA:N2B isoforms is found in sarcomeres with relatively high passive tension (PT), a high ratio in those with low PT. There is evidence for species-, age, transmural-, and chamber-specific differences in titin isoform expression.2,3 Patients with systolic heart failure (SHF) caused by coronary artery disease or dilated cardiomyopathy (DCM) express increased proportions of compliant N2BA isoforms in the left ventricles (LVs) and have lower-than-normal passive myofibrillar stiffness,4–6 possibly in an attempt to rescue diastolic function. In contrast, a recent study pooled endomyocardial biopsy specimens from patients with diastolic heart failure (DHF) and found reduced N2BA:N2B ratios as compared to SHF, whereas myocyte passive stiffness was high.7
Titin-based stiffness can also be modulated by posttranslational modification. Phosphorylation of a unique 572 amino acid sequence within the N2-B domain (N2-Bus) by cAMP-dependent protein kinase (PK)A lowers passive stiffness in rat, bovine, and human myocardium via yet unknown mechanisms.8–11 Interestingly, incubating skinned human cardiomyocytes with PKA produced a substantial reduction in passive stiffness in both SHF and DHF but more so in DHF,7 suggesting there may be a deficit for PKA-mediated titin phosphorylation which could explain part of the elevated passive stiffness of failing human hearts. The PKA effect on titin adds to the many other alterations in diastolic, as well as systolic function triggered by β-adrenergic stimulation.12,13
The influence of cAMP signaling on cardiac function can in part be opposed by cGMP,14,15 which is generated by guanylyl cyclases in response to nitric oxide (NO) and natriuretic peptides. Cross-talk exists between cAMP and cGMP signaling with regard to the activity of the cyclic nucleotide-degrading enzymes, phosphodiesterases (PDEs), in that cGMP stimulated (PDE-2) or inhibited (PDE-3) PDEs hydrolyze both cGMP and cAMP.15,16 Importantly, inhibition of cGMP-hydrolyzing PDE-5 by sildenafil ameliorated cardiac hypertrophy, fibrosis and systolic dysfunction in a murine model of pressure overload.17 NO- or cGMP-enhancing therapies increased resting diastolic cell length in isolated myocytes and downward-shifted the diastolic pressure–volume relationship during filling of intact hearts, indicating decreased myocardial stiffness and enhanced left ventricular distensibility.18–21 Extensive in vitro and in vivo studies have demonstrated antihypertrophic, antifibrotic, and prolusitropic effects of natriuretic peptides.22–24 Thus, substantial evidence suggests that beneficial effects of cGMP can be augmented by exogenous natriuretic peptides or NO donors or inhibition of cGMP metabolism by a variety of PDEs. Although activating the cGMP effector PKG clearly affects mechanical properties known to be determined in part by the titin springs, such as diastolic tone (lowered by PKG), left ventricular extensibility (increased by PKG), and relaxation speed (accelerated by PKG), to date, data on a possible PKG-mediated influence on titin function have been lacking.
We wanted to know whether cGMP-activated PKG can phosphorylate titin in mammalian, particularly human, heart and whether this posttranslational modification can improve diastolic function by reducing titin-based stiffness. Indeed, we detected phosphorylation sites for PKG in human cardiac titin and found a PKG-dependent decrease in myofibrillar stiffness. This mechanical effect is most likely mediated by phosphorylation of a serine residue in the cardiac-specific N2-Bus. Thus, cGMP-enhancing therapy could benefit diastolic function via increasing titin phosphorylation, thereby reducing titin stiffness.
Materials and Methods
Left ventricular tissue from human donor and failing DCM hearts was procured, deep-frozen in liquid nitrogen (LN2), and stored at −80°C (see the expanded Materials and Methods section in the online data supplement, available at http://circres.ahajournals.org). Left ventricular tissue of normal adult dogs was sampled as described.25 All procedures were conducted in full accordance with the institutional guidelines and were approved by the respective ethics committees.
Skinned Fiber Preparation
Left ventricular muscle strips were prepared from unfrozen human hearts and skinned in relaxing solution (for composition, see the expanded Materials and Methods section) supplemented with 1% Triton-X-100.4,26 The tissue was washed in relaxing solution and fiber bundles were dissected (diameter, 200 to 300 μm; length, 1.0 to 2.5 mm).
Skinned human fibers were studied at 26°C using a muscle mechanics workstation (Scientific Instruments, Heidelberg, Germany).4,6,26 Force and sarcomere length (SL) were recorded on stepwise stretching the samples from slack SL (average, 1.9 μm) to a maximum SL of 2.6 to 2.7 μm. Another set of experiments was done after 30-minute incubation with 1.68×10−5 U/μL purified PKG and 300 μmol/L 8-pCPT-cGMP (cGMP) (Calbiochem). Alternatively, fibers were dephosphorylated by alkaline phosphatase (AP) (New England Biolabs) or protein phosphatase (PP)2a (Sigma-Aldrich, recombinant catalytic subunit, 0.3 U/μL) before incubation with PKG and cGMP in presence of phosphatase-inhibitor cocktail (Sigma-Aldrich).
Myofibril Force Measurements
Myofibrils isolated from left ventricular tissue of human donor heart as described6 were stretched-released in relaxing solution from slack SL by 50% at a frequency of 1 Hz for 10 seconds (room temperature). The force amplitudes in each cycle were measured and averaged to obtain passive stiffness. Recordings were taken before and after addition of cGMP-PKG.
Recombinant Titin Fragments
Human cardiac titin fragments were expressed in Escherichia coli and purified by the GST-Fusion System (Amersham Biosciences): I2 to I4, N2-B (I24/25–N2-Bus–I26), I24/25, N2-Bus, N terminus of N2-Bus (N1–189), middle part of N2-Bus (M190–410), C terminus of N2-Bus (C411–572), N-terminally truncated C terminus of N2-Bus (C420–572), C-terminally truncated C terminus of N2-Bus (C411–545), C terminus of N2-Bus with mutation at position 469 (CS469A), I26, N2-A, and PEVK (N2B-PEVK).
SDS-PAGE and 32P-Autoradiography
Phosphorylation by PKG and PKA was probed by autoradiography following 15% or 2% SDS-PAGE as reported.26 Recombinant titin fragments were analyzed after incubation with 1.68×10−5 U/μL PKG, cGMP, and [γ32P]ATP (250 μCi/μM) for 20 minutes at 36°C. Some constructs were incubated with catalytic subunit of PKA (Biaffin, 1 U/μL) instead of cGMP-PKG. Skinned fibers were either directly phosphorylated or were first dephosphorylated by AP or PP2a (0.3 U/μL) before incubation with the respective kinase.
Dot Blot Analysis
Recombinant titin fragments were dephosphorylated by AP before incubation with cGMP and PKG, solubilized, and dotted onto a poly(vinylidene fluoride) membrane. The membrane was probed by monoclonal antibodies against phosphoserine residues (phosphoDetect anti-mouse mAB, clone 16B4, Calbiochem).
Atomic Force Microscopy Force Spectroscopy
Single-molecule force–extension traces were recorded (pulling rate, 500 nm sec−1) at 22°C with an MFP-3D atomic force microscope (Atomic Force, Mannheim, Germany) using the engineered constructs I25-I26-N2-Bus-I26-I27 or N2-A (see the online data supplement), stretched under the same buffer conditions as in the fiber/myofibril measurements. Data were analyzed by Igor procedures (WaveMetrics, Portland, Ore) using the worm-like chain (WLC) model27: equation
where F is entropic force; Lp, persistence length; x, end-to-end distance; Lc, contour length; kB, Boltzmann constant; and T, absolute temperature.
Modeling Titin-Based Force
Titin-based force-SL curves were predicted using a force–extension curve generated from the weighted sum of 3 WLC force–extension relations corresponding to the different extensible regions in titin.6,28
Significant differences were probed using the unpaired Student’s t test. probability values <0.05 were taken as indicating significant differences.
PKG Phosphorylates Cardiac Titin
Titin phosphorylation was probed in skinned cardiac fiber bundles of adult dog LV and human LV incubated with PKG and activator cGMP in the presence of radioactively labeled ATP. Back-phosphorylation was studied on fibers incubated with PKG only or with PKG following initial dephosphorylation by PP1, PP2a, or AP (all revealing similar results). The titin isoforms N2BA and N2B were both phosphorylated by PKG in dog hearts (n=3; example in Figure 1A, top gels). In human nonfailing LVs, phosphatase pretreatment substantially increased the phosphorylation signals (examples in Figure 1A, middle and bottom gels). There was considerable variability in signal strength in PKG-only–treated human fibers; some showed very low intensity (example in Figure 1A, bottom right). Assuming the signal intensity following phosphatase pretreatment always represented the respective maximum phosphorylation, the relatively low back-phosphorylation signals obtained with PKG only in donor hearts suggest these hearts have rather high inherent (basal) levels of titin phosphorylation.
PKG Reduces PT in Skinned Human Cardiac Fibers
To test for possible functional consequences of PKG-mediated titin phosphorylation, we measured the passive force–SL relation of skinned fiber bundles from human donor LVs (n=4 nonfailing hearts, 3 fibers/tissue). Again, experiments were performed either on fibers incubated with PKG alone or on samples dephosphorylated before incubation with PKG (PP+PKG; Figure 1B). Data were normalized to those obtained for control fibers incubated without kinase (dashed lines), and mean values were fit by linear regression (Figure 1B). Incubation with PKG only did not cause a significant reduction in average PT, although a trend was apparent (Figure 1B, left). In phosphatase-pretreated fibers, PT decreased significantly by ≈10% to 20% within the physiological SL-range (Figure 1B, right). The maximum average decrease was 13% at 2.2 μm SL.
PKG-Mediated PT Reduction Occurs at the Level of the Sarcomere
Titin springiness dominates the PT development in mammalian heart in the SL range from 1.9 to ≈2.2 μm, whereas collagen stiffness dominates at higher physiological SLs.29 This fact could potentially explain our observation of a significant PKG-mediated PT reduction below 2.4 μm SL only (Figure 1B, right). To test whether the PT decrease is attributable to a direct PKG effect on titin-based stiffness, passive force in response to repetitive stretch-release cycles (bursts of ten 1-Hz oscillations between slack SL and 150% slack length) was studied in collagen-free myofibril preparations isolated from human donor LV (Figure 2). Passive stiffness on application of cGMP-activated PKG to myofibrils in relaxing buffer was compared to that before phosphorylation (Figure 2A). In myofibrils treated with cGMP-PKG passive stiffness remained unchanged, but in myofibrils dephosphorylated before mechanical analysis incubation with cGMP-PKG dropped stiffness significantly, on average by 12% (Figure 2B). This decrease is similar to that observed at shorter physiological SLs in skinned fibers (Figure 1B). Thus, PKG reduces passive stiffness via a direct effect on the sarcomeres, presumably on titin.
PKG Phosphorylates Several Sites Within the Spring Region of Titin
We hypothesized that PKG phosphorylates titin within its springy segment. To probe this hypothesis, we generated recombinant fragments of various domains along the human I-band-titin region (Figure 3A) and tested these constructs for cGMP-activated PKG-mediated phosphorylation. For comparison, constructs were also incubated with PKA (final concentration, 1 U/μL relaxing solution), which was previously shown to phosphorylate the cardiac N2-Bus.8,11 Autoradiography detected PKG-dependent phosphorylation in the N2-B and N2-A regions, but not in the PEVK segment (Figure 3B, top gels). A relatively weak phosphorylation signal was obtained with the Ig domains I2 to I4. In contrast, PKA phosphorylated only the N2-B segment. The N2-B segment was split up into its constituting subfragments, I24/25, N2-Bus, and I26, and these constructs were again analyzed for PKG- or PKA-mediated phosphorylation (Figure 3B, bottom gels). Both kinases phosphorylated the N2-Bus, whereas PKG additionally phosphorylated I24/25. No signal was detected for I26. These results were supported by dot blots probing the PKG-phosphorylated recombinant fragments using monoclonal antibodies against phosphoserine residues (Figure 3C): cGMP-PKG again phosphorylated N2-B and I24/25, but not I26. Unlike in the autoradiography tests, Ig domains I2 to I4 showed no phosphorylation signal. Analysis of 3 subfragments of the N2-Bus (Nus, Mus, Cus) demonstrated the presence of PKG phosphorylation site(s) in the C-terminal part, but not in the other sections, of this unique sequence (Figure 3C).
Both PKG and PKA Phosphorylate Serine S469 of the Human N2-Bus
Comparing the amino acid sequence of the N2-Bus with online databases (NetphosK and Scansite Motif scanner), we found 5 potential consensus sequences for PKG and PKA phosphorylation, 1 each in the N-terminal (N1–189) and the middle portion (M190–410) and 3 in the C-terminal part (C411–572) (Figure 4A). However, using autoradiography, we readily confirmed the dot blot results (Figure 3C): PKG phosphorylated C411–572 (Cus) but not N1–189 (Nus) or M190–410 (Mus) (Figure 4B and 4C). We further generated 3 different mutants of the C411–572 construct, 1 with a truncation of the N-terminal potential phosphorylation site (C420–572), a second with a truncation of the C-terminal potential phosphorylation site (C411–545), and a third with the serine at position 469 of the N2-Bus exchanged by an alanine (CS469A) to inactivate this potential phosphorylation site. Of these 3 constructs, only the CS469A mutant was no longer phosphorylatable by PKG. These findings establish that PKG targets the N2-Bus of cardiac titin at position S469. Notably, autoradiography of normal and mutation constructs showed that serine 469 is also phosphorylated by PKA (Figure 4B and 4C).
PKG-Mediated Phosphorylation Alters the Elasticity of the N2-Bus
The N2-Bus extends in the physiological SL range, in addition to Ig domain regions and the PEVK domain of titin.2 We wanted to know whether the elastic properties of the N2-Bus can be altered by PKG-mediated phosphorylation. Hence we generated a recombinant fragment in which the N2-Bus is bordered by the 2 naturally flanking Ig domains on either side, I24/25 and I26/27, respectively. The force–extension relationship of this N2-B(+I27) construct was measured by single-molecule atomic force microscopy (AFM)27 (Figure 5A, top) and a possible mechanical effect owing to phosphorylation by cGMP-activated PKG was studied. The Ig domains unfold under forces of ≈200 pN and provide characteristic sawtooth-like “unfolding” peaks (Figure 5A, bottom). At least 3 regularly spaced Ig-unfolding peaks must appear in the force trace to make sure the whole N2-Bus is stretched; recordings not fulfilling this requirement were not analyzed (see the expanded Materials and Methods section).
From the force–extension traces we could parameterize the mechanical properties of the N2-Bus (Figure 5A, bottom). Modeling the initial trace leading up to the first unfolding peak (corresponding to the stretching of the N2-Bus27) using the WLC model (Equation) allowed extraction of the persistence length (Lp), a measure of the bending rigidity of the entropic spring. Histogram analyses showed that LpN2-Bus was 0.39±0.26 nm (mean±SD) under control conditions (relaxing buffer) (Figure 5B, top). In the presence of cGMP-PKG, the majority of recordings revealed a significantly increased Lp averaging 0.67±0.11 nm (Figure 5B, bottom). The contour length (LcN2-Bus), another fitted parameter in the WLC model (Equation), was unchanged at ≈205 nm under all experimental conditions, and also the unfolding forces of the Ig domains remained unaltered (data not shown).
In another experimental series, we also stretched a recombinant N2-A construct by AFM. Although the N2-A domain is phosphorylatable by cGMP-PKG (Figure 3B), its mechanical properties were not altered by the kinase (Figure I in the online data supplement).
Altered Bending Rigidity of the N2-Bus Explains the PKG-Induced Decrease in Titin-based Stiffness
Using the 2 average Lp values measured for the N2-Bus (Figure 5B), we modeled the elastic titin force versus SL relationship (Figure 5C) as that of 3 independent WLCs corresponding to the titin Ig domain regions, the PEVK segment, and the N2-Bus.6 The prediction used the human titin sequence information and mechanical parameters of cardiac titin domain function established by single-molecule AFM force spectroscopy (see the expanded Materials and Methods section).28 The model, which also considered the 35:65 ratio of N2BA:N2B isoforms expressed in human donor heart,6 demonstrated that increasing the persistence length of the N2-Bus from 0.35 to 0.67 nm, as seen in the presence of cGMP-PKG, is sufficient to lower the average force per titin by 17% (Figure 5C). Thus, the 10% to 20% decrease in passive stiffness observed under the influence of PKG in both skinned myofibers and isolated myofibrils can be explained solely by a PKG effect on the elasticity of the N2-Bus of titin.
Failing Human Hearts Have a Deficit in Basal Titin Phosphorylation
To address a possible pathophysiological significance of the PKG-mediated titin phosphorylation, we compared the titin back-phosphorylation signals on autoradiograms using human left ventricular tissue obtained from 3 nonfailing donors and 3 patients with end-stage DCM. Interestingly, PKG alone typically produced stronger back-phosphorylation signals in DCM than in donor hearts (Figure 6A; compare with Figure 1A). Measuring the difference in signal intensity between PKG-only and PP pretreated+PKG samples, the inherent (basal) phosphorylation levels were inferred (weighted for protein loading). There was a clear trend toward decreased basal titin phosphorylation in DCM compared to donor hearts (Figure 6B).
Cyclic GMP signaling, which is centrally involved in functional regulation of the cardiovascular system, can be promoted by natriuretic peptides (atrial natriuretic peptide and brain natriuretic peptide) activating the guanylyl cyclase A receptor and by stimulating NO synthase.16,22–24,30 The second messenger NO produces biphasic contractile effects in heart tissue, with augmentation at low levels and depression at high levels. NO participates not only in the control of contractility and heart rate but also limits cardiac remodeling after an infarction and contributes to the protective effect of ischemic pre- and postconditioning.16,30,31 Growing evidence suggests that activation of PKG (cGKI) via cGMP signaling can lower diastolic tone and increase LV extensibility.18–21 Our work now demonstrates that cGMP-activated PKG phosphorylates specific sites within the titin spring segment and decreases titin-based stiffness by altering the elastic properties of a cardiac-specific titin region, the N2-Bus. These PKG-dependent mechanical effects on titin could account for at least part of the observations that NO decreases myocardial stiffness, increases myocyte resting length, and downward shifts the diastolic filling curve in PV loops.18,19,21 Thus, PKG-mediated regulation of titin-based stiffness appears to be important in adjusting ventricular filling and regulating diastolic function in physiological and pathophysiological settings.
Apart from NO, myocardial diastolic stiffness can also be modulated acutely by endothelin-1,32 β-adrenoceptor agonists that activate PKA,7–9,11,33,34 and angiotensin II,21 agents better known for their stimulating effects on cardiac contractility. These acute effects on diastolic function are distinct from the well-established long-term influences related to variables such as changes in chamber geometry, hypertrophy, extracellular matrix remodeling, and transitions in titin isoform composition. To date, mechanistic explanations for the acute effects on diastolic stiffness are sparse. Angiotensin II and endothelin-1 have been proposed to decrease myocardial passive stiffness through little understood mechanisms requiring the activation of G protein–coupled receptors and involving PKC and the Na+/H+ exchanger21,32; also, NO has been implicated in mediating these effects.21 In contrast, the PKA-induced PT reduction observed in rat,8 bovine,9 and human11 heart, including cardiomyocytes from patients with SHF or DHF,7,34 has been explained by phosphorylation of the N2-Bus of titin,8,9,11 a suggestion supported by the present results (Figures 3 and 4⇑). Although titin has long been known to contain phosphorylation sites and to be phosphorylatable,2,3,35 the PKA-dependent phosphorylation of the N2-Bus was the only posttranslational modification of titin previously shown to alter titin stiffness.
The novel mechanism of diastolic stiffness regulation established here is mediated by phosphorylation of the human cardiac N2-Bus by PKG. The relative drop in passive stiffness induced by PKG was about half that seen elsewhere with PKA.7–9,11 We demonstrated a similar percentage of PKG-induced passive stiffness decrease (12% to 17%) in human cardiomyofibers (Figure 1B), single myofibrils (Figure 2), and single titin molecules (Figure 5), the latter being brought about solely by increasing the bending rigidity of the N2-Bus. Because the N2-Bus is part of the N2-B domain, which is not expressed in the skeletal muscle N2A titin isoforms, the mechanical effect should be cardiac-specific. The PKG effect was significant in phosphatase pretreated samples but almost disappeared when the dephosphorylation step was omitted, an observation that we attribute to the high basal levels of phosphorylation seen in the human donor hearts (Figure 6B). However, the phosphorylation state of the human donor samples might have been altered by the pharmacological treatment of these hearts before transplantation surgery. In this context, inherent titin phosphorylation was found to be lower in healthy dog LV than in human donor LV. Follow-up studies on a larger cohort of human hearts and on animal models should help clarify this issue.
Using mutagenesis of recombinant constructs, we identified serine S469 of the human N2-Bus as the PKG phosphorylation site within that titin sequence. Our evidence suggests that it is most likely phosphorylation of S469 that alters N2-B–titin elasticity. We further demonstrated that S469 can also be phosphorylated by PKA. As the length and amino acid composition of the N2-Bus differ among mammalian species,36 it remains to be seen whether PKG (and PKA) phosphorylates the N2-Bus at the same and/or different sites in other species. By screening the N2-Bus sequence of rat cardiac titin,36 we found a sequence motif (QKTS) at position 556 to 559 of the N2-Bus that is very similar to the sequence motif (AKTS) at residues 466 to 469 of the human N2-Bus. Taken together with the observation that PKG phosphorylated both the N2BA and N2B isoforms in adult dog heart, we propose that PKG-mediated phosphorylation is a more general mechanism to regulate diastolic stiffness in mammalian heart.
We provide evidence that the PKG phosphorylation site relevant for the mechanical effect resides within the N2-Bus; however, PKG phosphorylated other domains in human titin as well, both in the N2-B (Ig domains I24/25) and N2-A segments (Figure 3). Because we found no influence of PKG on the mechanical properties of titin Ig domains in single-molecule AFM stretch experiments with engineered N2-B (Figure 5) or N2-A constructs (supplemental Figure I), we exclude that the PKG-mediated PT decrease in myofibrils results from an effect on these domains. The proximal Ig domains, I2 to I4, showed weak, if any, propensity to be phosphorylated by PKG, whereas no signal was detected for the Ig domain I26, which was suggested to be a substrate for PKG in rat uterus titin.37 The functional role(s) of the additional posttranslational modifications detected are currently unknown, but we speculate that phosphorylation at sites outside the N2-Bus could be important for regulating protein-protein interactions, because the titin springs interact with multiple structural and signaling molecules.2
Because PKG phosphorylates various sites on human titin, a correlation between the titin phosphorylation level and the magnitude of PKG-induced mechanical effect was deemed not meaningful. This limitation notwithstanding, we observed a trend toward reduced basal levels of PKG-mediated titin phosphorylation in failing human DCM hearts compared to donor hearts (Figure 6). Thus, failing human hearts could have a titin phosphorylation deficit which includes the mechanically relevant phosphorylation site (although, as discussed above, the donor hearts could be hyperphosphorylated, which would make the observed low titin phosphorylation state in DCM less clinically relevant). One can then speculate that a titin phosphorylation deficit increases passive stiffness in failing human hearts. Indeed, skinned cardiomyocytes isolated from failing hearts of SHF or DHF patients showed elevated PT levels, which were substantially reduced by administration of PKA.7 The present findings suggest the intriguing possibility that a pathologically increased passive myocyte stiffness could also be normalized by PKG-mediated phosphorylation. Future work on this issue should consider the crosstalk between the PKA and PKG signaling systems in light of the fact that both kinases phosphorylate titin at site S469 of the N2-Bus. In any case, our results led us to suggest the following possible scenario: the cGMP-hydrolyzing PDE5 is present in the Z-disk and might regulate local pools of cGMP, which could then activate PKG.14 This kinase would reduce myofilament Ca2+ sensitivity and depress contraction but also lower titin-based stiffness. In the failing heart, this mechanism may be compromised and deficits in PKG-mediated titin phosphorylation could add to an increased passive stiffness. In turn, PKG-mediated regulation of titin stiffness could be a handle for pharmacological intervention, eg, via inhibition of PDE514 by sildenafil.17 Finally, nitrates are already commonly used in the treatment of both SHF and DHF and their beneficial effects could well involve keeping titin stiffness low.
In summary, this study demonstrates that PKG reduces titin-based stiffness via phosphorylation of a serine residue (S469) within the N2-Bus of titin, thereby acting to improve diastolic function in human hearts.
We are grateful to Prof Peter Macdonald (St Vincent’s Hospital Heart and Lung Transplant Unit) for providing clinical data on the failing human hearts.
Sources of Funding
This work was supported by the German Research Foundation (Li 690/7-1; SFB 629).
Original received July 29, 2008; revision received October 21, 2008; accepted November 6, 2008.
Freiburg A, Trombitas K, Hell W, Cazorla O, Fougerousse F, Centner T, Kolmerer B, Witt C, Beckmann JS, Gregorio CC, Granzier H, Labeit S. Series of exon-skipping events in the elastic spring region of titin as the structural basis for myofibrillar elastic diversity. Circ Res. 2000; 86: 1114–1121.
Linke WA. Sense and stretchability: the role of titin and titin-associated proteins in myocardial stress-sensing and mechanical dysfunction. Cardiovasc Res. 2008; 77: 637–648.
Granzier HL, Labeit S. The giant protein titin: a major player in myocardial mechanics, signaling, and disease. Circ Res. 2004; 94: 284–295.
Neagoe C, Kulke M, del Monte F, Gwathmey JK, de Tombe PP, Hajjar RJ, Linke WA. Titin isoform switch in ischemic human heart disease. Circulation. 2002; 106: 1333–1341.
Nagueh SF, Shah G, Wu Y, Torre-Amione G, King NM, Lahmers S, Witt CC, Becker K, Labeit S, Granzier HL. Altered titin expression, myocardial stiffness, and left ventricular function in patients with dilated cardiomyopathy. Circulation. 2004; 110: 155–162.
Makarenko I, Opitz CA, Leake MC, Neagoe C, Kulke M, Gwathmey JK, del Monte F, Hajjar RJ, Linke WA. Passive stiffness changes caused by upregulation of compliant titin isoforms in human dilated cardiomyopathy hearts. Circ Res. 2004; 95: 708–716.
van Heerebeek L, Borbely A, Niessen HW, Bronzwaer JG, van der Velden J, Stienen GJ, Linke WA, Laarman GJ, Paulus WJ. Myocardial structure and function differ in systolic and diastolic heart failure. Circulation. 2006; 113: 1966–1973.
Yamasaki R, Wu Y, McNabb M, Greaser M, Labeit S, Granzier H. Protein kinase A phosphorylates titin’s cardiac-specific N2B domain and reduces passive tension in rat cardiac myocytes. Circ Res. 2002; 90: 1181–1188.
Fukuda N, Wu Y, Nair P, Granzier HL. Phosphorylation of titin modulates passive stiffness of cardiac muscle in a titin isoform-dependent manner. J Gen Physiol. 2005; 125: 257–271.
Bers DM. Altered cardiac myocyte Ca regulation in heart failure. Physiology. 2006; 21: 380–387.
Kass DA, Solaro RJ. Mechanisms and use of calcium-sensitizing agents in the failing heart. Circulation. 2006; 113: 305–315.
Zaccolo M, Movsesian MA. cAMP and cGMP signaling cross-talk: role of phosphodiesterases and implications for cardiac pathophysiology. Circ Res. 2007; 100: 1569–1578.
Kass DA, Takimoto E, Nagayama T, Champion HC. Phosphodiesterase regulation of nitric oxide signaling. Cardiovasc Res. 2007; 75: 303–314.
Shah AM, Spurgeon HA, Sollott SJ, Talo A, Lakatta EG. 8-bromo-cGMP reduces the myofilament response to Ca2+ in intact cardiac myocytes. Circ Res. 1994; 74: 970–978.
Paulus WJ, Vantrimpont PJ, Shah AM. Acute effects of nitric oxide on left ventricular relaxation and diastolic distensibility in humans. Assessment by bicoronary sodium nitroprusside infusion. Circulation. 1994; 89: 2070–2078.
Heymes C, Vanderheyden M, Bronzwaer JG, Shah AM, Paulus WJ. Endomyocardial nitric oxide synthase and left ventricular preload reserve in dilated cardiomyopathy. Circulation. 1999; 99: 3009–3016.
Kuhn M. Structure, regulation, and function of mammalian membrane guanylyl cyclase receptors, with a focus on guanylyl cyclase-A. Circ Res. 2003; 93: 700–709.
Shapiro BP, Lam CS, Patel JB, Mohammed SF, Kruger M, Meyer DM, Linke WA, Redfield MM. Acute and chronic ventricular-arterial coupling in systole and diastole: insights from an elderly hypertensive model. Hypertension. 2007; 50: 503–511.
Kruger M, Kohl T, Linke WA. Developmental changes in passive stiffness and myofilament Ca2+ sensitivity due to titin and troponin-I isoform switching are not critically triggered by birth. Am J Physiol Heart Circ Physiol. 2006; 291: H496–H506.
Hofmann F, Feil R, Kleppisch T, Schlossmann J. Function of cGMP-dependent protein kinases as revealed by gene deletion. Physiol Rev. 2006; 86: 1–23.
Leite-Moreira AF, Bras-Silva C, Pedrosa CA, Rocha-Sousa AA. ET-1 increases distensibility of acutely loaded myocardium: a novel ETA and Na+/H+ exchanger-mediated effect. Am J Physiol Heart Circ Physiol. 2003; 284: H1332–H1339.
Yamamoto K, Masuyama T, Sakata Y, Nishikawa N, Mano T, Yoshida J, Miwa T, Sugawara M, Yamaguchi Y, Ookawara T, Suzuki K, Hori M. Myocardial stiffness is determined by ventricular fibrosis, but not by compensatory or excessive hypertrophy in hypertensive heart. Cardiovasc Res. 2002; 55: 76–82.
Borbely A, van der Velden J, Papp Z, Bronzwaer JG, Edes I, Stienen GJ, Paulus WJ. Cardiomyocyte stiffness in diastolic heart failure. Circulation. 2005; 111: 774–781.