Cholesterol regulates numerous cellular processes. Depleting its synthesis in skeletal myofibers induces vacuolization and contraction impairment. However, little is known about how cholesterol reduction affects cardiomyocyte behavior. Here, we deplete cholesterol by incubating neonatal cardiomyocytes with methyl-beta-cyclodextrin. Traction force microscopy shows that lowering cholesterol increases the rate of cell contraction and generates defects in cell relaxation. Cholesterol depletion also increases membrane tension, Ca2+ spikes frequency and intracellular Ca2+ concentration. These changes can be correlated with modifications in caveolin-3 and L-Type Ca2+ channel distributions across the sarcolemma. Channel regulation is also compromised since cAMP-dependent PKA activity is enhanced, increasing the probability of L-Type Ca2+ channel opening events. Immunofluorescence reveals that cholesterol depletion abrogates sarcomeric organization, changing spacing and alignment of α-actinin bands due to increase in proteolytic activity of calpain. We propose a mechanism in which cholesterol depletion triggers a signaling cascade, culminating with contraction impairment and myofibril disruption in cardiomyocytes.
Cholesterol plays a fundamental role in regulating plasma membrane fluidity, integrity, and in compartmentalizing intracellular signaling events1. In non-muscle cells, such as fibroblasts and lymphoblasts, cholesterol depletion reorganizes the actin cytoskeleton architecture and reduces lateral mobility of membrane-bound proteins in a phosphatidylinositol 4,5- biphosphate dependent way2. Cholesterol-depleted endothelial cells3 and fibroblasts4 also show an increase in cell stiffness due to actin remodeling and that is probably due to an enhancement in membrane-cytoskeleton adhesion energy5. In skeletal muscle cells, cholesterol depletion impairs contraction, causes vacuolization and morphological abnormalities and uncouples a transmembrane glycoprotein, β-dystroglycan that connects the extracellular matrix to the cytoskeleton of those cells6,7,8. However, less is known about how cholesterol depletion might affect cardiomyocyte physiology and contraction dynamics.
Flask-shaped and cholesterol enriched domains, known as caveolae, are fundamental for cardiomyocyte functioning9. Caveolins are proteins that reside in caveolae and caveolin-3 is the most abundant muscle isoform10,11. These proteins are also important for mechanosensing12,13 and membrane tension regulation14 in muscle cells. Aside from modulating cellular mechanics, caveolae also regulate Ca2+ dynamics in cardiomyocytes being responsible for gathering and regulating the activity of important ion channels, such as L-Type Ca2+ channel (LTCC)15. Besides, caveolae localization guarantees proper β1 and β2-adrenergic receptors (β-ARs) signaling in cardiomyocytes16,17,18 that play a major role in the initiation of Ca2+ cycling in those cells.
Extracellular Ca2+ entry is an important step for Ca2+ cycling within the cardiomyocyte and it is mediated by LTCC. Regulation of LTCC opening and closing events is fundamental to guarantee a synchronous Ca2+ signaling19 that ultimately leads to cardiomyocyte contraction. β1 and β2-adrenergic receptors (β-ARs) are G-protein coupled receptors that initiate the cascade of events that culminate with the opening of a LTCC. Under agonist stimulation, β-ARs are activated and trigger cAMP production by adenylyl cyclase. cAMP activates protein kinase A (PKA) which in turn phosphorylates Cav1.2, the pore-forming subunit of LTCC, keeping it open and promoting extracellular Ca2+ entry to the sarcoplasm20. It is known that extracellular Ca2+ plays a pivotal role in neonatal cardiomyocyte contraction since those cells lack t-tubules11 that are fundamental structures for the Ca2+-induced-Ca2+-release mechanism for contraction in adult cardiomyocytes21. Besides, neonatal cardiomyocytes also show a more pronounced increase in β2-AR-dependent cAMP concentration that activates PKA in those cells22.
Not only Ca2+ but also myofibrillar organization and stability are critical for muscle cell contraction. Myocytes have an extremely organized and stable actomyosin network that is responsible for uniform and long-range contraction23. This actomyosin cytoskeleton has a striated pattern that arises from sarcomere sequential positioning being each sarcomere the contractile unit of the myofibril24. Each sarcomere is delimited by two z-lines where actin barbed ends localize. The z-lines also contain α-actinin that is fundamental for organizing the myofibrils during development25. The contraction of the myofibers happen when Ca2+ binds to troponin-C changing tropomyosin conformation and releasing the site where myosin binds to actin allowing the filaments to slide past each other25. Buffering of Ca2+ and release of this ion to either the sarcoplasmic reticulum or to the extracellular milieu is important for proper relaxation of the myofibers since Ca2+ mishandling can cause cardiac issues26. Moreover, excess in sarcoplasmic Ca2+ can trigger activation of proteases like calpains27 that can degrade sarcomeric proteins and abrogate myofibrillar organization and contractility28.
To investigate the importance of cholesterol in regulating cardiomyocyte physiology we incubated the cells with a cyclodextrin called methyl-beta cyclodextrin (MβCD) to acutely deplete plasma membrane cholesterol in isolated primary neonatal rat cardiomyocytes. MβCD is a cyclic oligosaccharide that has a hydrophobic core (with a high affinity for cholesterol molecules) and a hydrophilic surface that facilitates its solubilization in aqueous media29 and does not have the pleiotropic effects that other cholesterol lowering drugs have, such as depletion of isoprenoids30. MβCD takes advantage of the cholesterol efflux dynamics in order to trap the molecule inside its cage: once the cyclodextrin gets closer to the plasma membrane, cholesterol migrates from the membrane to the MβCD core without having to be in direct contact with the aqueous media31. Cyclodextrins similar to MβCD can be used in vivo as a way of facilitating the reverse cholesterol transport in which accumulated cholesterol is removed from the vessel walls and transported to the liver for degradation decreasing the incidence of atherosclerosis in hyperlipidemic rabbits32.
Traction Force Microscopy (TFM) experiments revealed that cells with lower cholesterol levels had higher and variable contraction rates and relaxation defects. MβCD treatment increased the cardiomyocytes cortical mechanical properties values, the frequency of Ca2+ spikes and intracellular Ca2+ concentration. We correlated these results with the fact that caveolin-3 and Cav1.2 change distribution across the sarcolemma upon cholesterol depletion which, in turn, might impact the regulation of the LTCC. In order to test this hypothesis we measured cAMP-dependent PKA activation and we saw an increase in the enzymatic activity consequent to cholesterol depletion. This result corroborates the increase in intracellular Ca2+ after MβCD treatment since PKA activity is directly involved in Cav1.2 opening events. Immunofluorescence revealed that cholesterol depletion abrogated the organization of myofibrils, changing the spacing and alignment of α-actinin bands and this change was correlated with increase in proteolytic activity of calpain probably due to enhancement in cytoplasmic Ca2+ concentration.
This work suggests for the first time a complete mechanism by which cholesterol depletion, due to MβCD incubation, impairs contraction in neonatal cardiomyocytes.
Materials and Methods
For more details in the methods section visit the supplemental information that is available online.
All animals were maintained at The University of Chicago Animal Resources Center. Animal husbandry and all experimental procedures were performed in compliance with the guidelines of the Institutional Animal Care and Use Committee (IACUC). Experimental protocols were approved by the University of Chicago IACUC (protocol number 72347).
Primary culture of neonatal rat cardiomyocytes
Neonatal rat hearts (p0-p3) were submitted to enzymatic digestion steps that were adapted from a previously stablished protocol33. Purified cardiomyocytes were collected, plated either on round glass coverslips coated with fibronectin alone or containing fibronectin crosslinked to polyacrylamide (PAA) gels.
To deplete cholesterol the cardiomyocytes were incubated with either 5.0 or 7.5 mmol/L MβCD, for 45 minutes, at 37 °C.
After fixation, cardiomyocytes were incubated with the following primary antibodies: mouse monoclonal anti-sarcomeric α-actinin, rabbit polyclonal anti-caveolin-3, mouse monoclonal Cav1.2. Next, the cells were rinsed and incubated with secondary fluorescent antibodies. Images of cells were taken on Ti-E Nikon inverted confocal microscope using either 60 × 1.49 NA ApoTIRF or 40 × 1.30 Plan Fluor oil-immersion objectives.
Traction Force Microscopy
Cardiomyocytes were plated on polyacrylamide gels with a shear modulus of 8.64 KPa embedded with fluorescent beads as previously described34,35,36. Images of the fluorescent beads with and without the cells on top were acquired and analyzed using MATLAB scripts previously developed34,35,37,38,39,40.
Tether extraction with optical tweezers
An infrared optical tweezers (OT) setup was used to extract tethers from the surface of the cardiomyocytes. The OT setup and force calibration were performed as previously described41,42. Scanning Electron Microscopy (SEM) was used to measure the tether radii4,41,42.
Imaging of calcium sparks
Epifluorescence signal from the Fluo-4 AM Ca2+ probe (Thermo Fisher Scientific) was acquired for control and treated cardiomyocytes and cytoplasmic Ca2+ sparks were analyzed using ImageJ. Conversion of fluorescence values into Ca2+ concentration was based on the following equation43:
K is the dissociation constant of the Ca2+ dye used, R is the fluorescence ratio (F′0 is the minimum fluorescence recorder for a specific cell analyzed), [Ca2+]rest is approximately 140 nM44.
Measurement of PKA and calpain enzymatic activities
To measure cAMP-mediated PKA or calpain activities, cell extracts were submitted to either enzymatic assay using PKA activity kit (Enzo Life Sciences) or using calpain Activity Fluorometric Assay Kit (BioVision Inc) according to manufacturer’s instructions.
Statistical analysis and data quantification
Student’s t-test was used to compare between control and treated groups. Data was represented by mean values ± standard error unless otherwise stated. The experimental errors for F0 and R were propagated by taking the derivatives of the equations 2 and 3 with respect to both F0 and R. Statistical differences were labeled with asterisks. Fourier analysis of contraction and Ca2+ sparks as well as quantification of α-actinin bands were done using MATLAB scripts described in the supplemental material.
Cholesterol depletion impairs contraction in rat neonatal cardiomyocytes
To deplete plasma membrane cholesterol we used a cyclodextrin called MβCD. To directly measure how cholesterol depletion affects contractility of neonatal cardiomyocytes, we used time-resolved Traction Force Microscopy (TFM). Cell contractility is defined as the ability of the cell to generate contractile forces via actomyosin interactions within myofibrils. For the case of cells attached to a flat substrate, we can measure their contractile behavior using TFM. Traction stress maps were obtained for 40 seconds to capture the spontaneous beating of cardiomyocytes plated on fibronectin coated gels. The maximal and minimal traction stresses obtained during this time lapse are shown in Fig. 1a and b, respectively, and correspond to the maximally contracted and relaxed states of the cardiomyocyte. The strain energy is a measure of the total mechanical work done by the cell in deforming the substrate. The ratio of strain energy to cell area defines a characteristic contractility measurement that is independent of cell shape and size39. The quantification of the strain energy per cell area from the traction stress maps demonstrates that control cardiomyocytes exhibit periodic contraction-relaxation patterns (Fig. 1c), with a majority of the power exerted at a frequency of 0.3 Hz (Fig. 1d). The contraction becomes more irregular in cholesterol depleted cells (Fig. 1a–d). Treatment with 7.5 mmol/L MβCD increased both the energy of the peaks (Fig. 1e) and the troughs (Fig. 1f) which suggests impairment in cell relaxation. In order to evaluate the resultant strain energy per cell area we plotted the difference between peak and trough strain energy/cell area (Fig. 1g) and we found that the overall contractility decreases. Moreover, the power spectra show that cholesterol depletion yielded noisier spectra (Fig. 1d) and increased the beating frequency (Fig. 1h). Thus, cholesterol depletion severely affects neonatal cardiomyocytes contractility.
Cholesterol depletion increases cortical mechanical properties values of neonatal rat cardiomyocytes
To determine how cholesterol depletion altered cortical mechanical properties of neonatal rat cardiomyocytes, tethers were extracted from the surfaces of these cells using OT. In this assay, OT were used to trap and control the position of an uncoated polystyrene bead that forms a tight bond with the cell surface when it is in close proximity (Fig. 2a and b, left). As the position of the bead and the cell are varied, a tether is formed (Fig. 2a and b, right). By using OT calibration, it is possible to obtain the force as a function of distance between the cell and bead (Fig. 2c). When the bead starts to be pulled away from the cell surface, the applied force increases up to a maximum value (Fm) and then abruptly decreases to a constant value F0. Fm corresponds to the moment that precedes the tether formation whereas F0 corresponds to the force applied to form and elongate the tether. We measured F0 for all tested conditions (Fig. 2d) and saw a trend in increasing this value for 5.0 mmol/L MβCD treated cells (45 ± 4 pN) and a statistical increase for cells treated with 7.5 mmol/L MβCD (54 ± 6 pN) in comparison to control cardiomyocytes (35 ± 4 pN).
In order to calculate the membrane tension σ and bending rigidity κ we used equations (2) and (3):
by substituting the values for F0 and R in the equations (2) and (3) we obtained σ and κ for the tested groups (Table 1).
Based on those results it is possible to affirm that cholesterol depletion increases both cortical mechanical properties values (σ and κ).
In order to uncouple contributions from the plasma membrane and from the cytoskeleton to cortical mechanical properties we decided to induce plasma membrane vesicles (PMVs) (depicted in Fig. S1A) in the cardiomyocytes (for technical information, please see supplementary information section). By doing that, we were able to measure specifically the contribution of the plasma membrane since the PMVs are devoid of cytoskeletal components. In order to calculate σ and κ for the tethers extracted from PMVs it is necessary to obtain the tethers’ radii. However, since PMVs are fragile and cannot resist to the fixation process and sample preparation for SEM, we used equation (4) to calculate the PMVs tethers’ radii45:
where Fm is the force to extract the tether from the PMV, F0 is the force to elongate the tether and Rp is the patch radius, defined in Fig. S1D. Mean values and their respective standard errors for Fm and F0 are shown in Figs S1B and S1C. PMVs’ membrane tension and bending modulus, amongst other values, are listed on Table S1.
Values for PMVs’ membrane tension are higher when cholesterol is depleted whereas there are no differences in bending moduli values between control and treated groups. Thus, contribution for enhancement in bending rigidity in cardiomyocytes comes almost exclusively from the cytoskeleton component.
Cholesterol depletion disorganizes neonatal cardiomyocytes myofibers
To investigate whether altered contractile properties coincided with changes in the myofibril architecture, cells were fixed and stained for actin and sarcomeric α-actinin. Figure S2, left panel, shows actin cytoskeleton of control (Fig. S2A), 5.0 mmol/L (Fig. S2B) and 7.5 mmol/L (Fig. S2C) MβCD treated cardiomyocytes. Cholesterol depletion impacts significantly the organization of thin filaments within the myofibers. α-actinin labeling shows that lack of cholesterol also makes the z-bands distorted in comparison to control cells which have equally spaced and parallel z-bands (Fig. S2A).
To quantify changes in myofibrillar α-actinin organization, image analyses were performed to measure the separation distance and angle between z-bands. In control cardiomyocytes, the spacing between z-bands is quite regular, with a mean value of (1.75 ± 0.09) μm (Fig. 3a,c). In contrast, cholesterol depleted cardiomyocytes show a reduction in mean band spacing ((1.49 ± 0.29) and (1.47 ± 0.31) μm for 5.0 and 7.5 mmol/L MβCD treated cells respectively) and with a much larger distribution around the mean value for both conditions (Fig. 3c). Figure 3d shows that control cardiomyocytes have more parallel aligned z-bands within a myofibril and the angle between them is small (mean value (6.8 ± 2.3)°), as expected. As an illustrative output of the MATLAB script, the superposition of neighboring z-bands within a myofibril (Fig. 3d, inset on the left) also shows a well-defined line that represents the degree of band alignment for the control group (Fig. 3b shows a cartoon depicting how the distances and angles between z-bands were calculated). In contrast, cholesterol depleted cardiomyocytes have less aligned z-bands and that is shown in Fig. 3a (mid and bottom panels). Illustrative outputs show a broader and smeared intensity distribution for the superposition of the z-bands and that is quantified in the graph (Fig. 3d) ((20 ± 10)° for the 5.0 mmol/L MβCD case and (22 ± 10)° for the 7.5 mmol/L MβCD case) (insets in the middle and on the right show superposition of z-bands from selected myofibers, in Fig. 3a).
Cholesterol depletion changes caveolin-3 distribution on cardiomyocytes
Caveolin-3, localized in cholesterol invaginated domains known as caveolae, has been previously shown to be important for adult cardiomyocytes to adapt to sudden changes in membrane tension14. Moreover, actin cytoskeleton also modulates caveolin organization46,47. Since cholesterol depletion changes both membrane tension and actin cytoskeleton architecture, we decided to investigate whether caveolin-3 distribution was affected. Cardiomyocytes were fixed and labeled for actin (Fig. 4a), α-actinin (Fig. 4b) and caveolin-3 (Fig. 4c), a caveolae-specific scaffolding protein highly expressed in muscle cells10. Control cardiomyocytes show a punctate distribution of caveolin-3 that is mostly localized along the z-bands (Fig. 4c,d and e, left panels). After cholesterol depletion with either 5.0 (Fig. 4c,d and e, middle panels) or 7.5 mmol/L MβCD (Fig. 4c,d and e, right panels), caveolin-3 distribution becomes more homogeneous and those punctate domains, that appear in control cells, disappear. Quantification of caveolin-3 redistribution is shown in Fig. S3. Cholesterol depletion disorganizes not only cytoskeleton architecture but also caveolin-3 distribution along the z-bands.
Cholesterol depletion changes cytoplasmic calcium spikes dynamics in cardiomyocytes
So far, we demonstrated that cholesterol depletion impaired actin cytoskeleton organization and contractility of the cardiomyocytes. We also showed that cholesterol depletion changed the cortical mechanical properties as well as caveolin-3 distribution in cardiomyocytes. Since Ca2+ is another fundamental piece of the contractile machinery in muscle cells, we sought to investigate whether cholesterol depletion was also changing sarcoplasmic Ca2+ cycling in cardiomyocytes. For that, control and cholesterol depleted cardiomyocytes were incubated with Fluo-4 AM and imaged in order to track changes in intracellular calcium spike dynamics. In Fig. 5a the heat maps show Ca2+ fluorescence for control and MβCD treated cardiomyocytes in the contracted and relaxed states. By comparing contracted states, we see that cytoplasmic Ca2+ fluorescence intensity increases when cholesterol is depleted. For the relaxed states we also noticed that cholesterol depleted cells cannot buffer the cytoplasmic Ca2+ efficiently during relaxation. Quantification of intracellular Ca2+ fluorescence (Fig. 5b) reveals that the oscillation in Ca2+ spikes is periodic in the control case, and during the relaxation period the fluorescence signal drops almost to zero. On the other hand, cardiomyocytes with lower cholesterol levels, especially the cells treated with 7.5 mmol/L MβCD, have a higher baseline for calcium intensity and during relaxation the calcium is not buffered as efficiently as it is in control cells. Moreover, we observe noisier spectra for calcium spikes in cholesterol-depleted cells and this feature is quantified in the power spectra graphs (Fig. 5c). We not only see an increase in contraction rate (Fig. 1) but we also observe an increase in calcium peak frequency as cholesterol is depleted from the cardiomyocytes (Fig. 5d). By quantifying the cytoplasmic Ca2+ concentration (see equation 1) we confirmed that cholesterol depleted cardiomyocytes presented more events where Ca2+ concentration was higher in comparison to control cells (Fig. 5e). In accordance with TFM results, the Ca2+ spike data show that cholesterol depletion changes cytoplasmic Ca2+ dynamics leading to enhancement in both Ca2+ concentration and frequency of Ca2+ spikes and impairment in cell buffering of cytoplasmic Ca2+ when the cells undergo relaxation.
Cholesterol depletion impairs L-type Ca2+ channel localization at plasma membrane of cardiomyocytes triggering PKA and calpain activation
Since extracellular Ca2+ is critical for neonatal cardiomyocyte contraction44,48,49,50,51, the interaction between the sarcolemma and extracellular milieu is crucial to guarantee homeostasis during contraction. We saw that cholesterol depletion increases membrane tension and disrupts caveolin-3 distribution. Since caveolar domains contain Ca2+ channels, such as LTCC, and regulate their activity52, we decided to investigate whether cholesterol depletion was also changing localization of LTCC in the sarcolemma of neonatal cardiomyocytes. To do that, cardiomyocytes were fixed and labeled for LTCC subunit Cav1.2 (also known as α1c pore-forming subunit, highly expressed in cardiac muscle53) (Fig. 6, left pannel) and caveolin-3 (Fig. 6, mid panel). Control cardiomyocytes (Fig. 6a) exhibit a homogenous distribution of the Cav1.2 labeling (Fig. 6a, left panel, inset). However, after cholesterol depletion with either 5.0 (Fig. 6b) or 7.5 mmol/L (Fig. 6c) MβCD the distribution of the Cav1.2 subunit becomes less uniform with areas of the cells that are devoid of labeling, especially at perinuclear regions (compare Fig. 6a,b and c, left panels. See Fig. S4 for quantification of Cav1.2 distribution). It is known that PKA phosphorylates Cav1.2 in order to control the opening of the LTCC pore54. In order to correlate changes in localization of LTCC with its function we performed an enzymatic assay to measure cAMP-dependent PKA activity in control and cholesterol depleted cardiomyocytes. Figure 6d shows that 7.5 mmol/L MβCD elicits enhancement in cAMP-dependent PKA activity in cardiomyocytes in comparison to control cells leading to more pore-opening events, which allow Ca2+ entry. In conjunction with our previous results, this evidence suggests that cholesterol depletion activates cAMP-dependent PKA activity leading to enhancement in Ca2+ entry, which in turn deregulates contraction dynamics in neonatal cardiomyocytes. In addition, the clear disruption of myofibers in cells treated with MβCD raised the question whether some Ca2+ -dependent cytoplasmic protease might have been activated in the process. Since we saw an increase in cytoplasmic Ca2+ concentration we decided to test activity of calpain, one of the main proteases in muscle tissue that is activated by intracellular Ca2+ 27. Normalized values for calpain activity showed that MβCD treatment increased calpain activity in neonatal cardiomyocytes in comparison to control cells. This result corroborates the idea that this Ca2+-dependent protease might be degrading the myofibers due to mishandling of cytoplasmic Ca2+ caused by cholesterol depletion.
It is widely known that cholesterol plays a pivotal role in controlling permeability, diffusion and receptor clustering at the plasma membrane level55. In non-muscle cells, cholesterol regulates actin cytoskeleton architecture and cortical mechanical properties2,3,4,5 whereas in smooth and skeletal muscle cells, cholesterol is important for modulating the contraction of these cells6,7,8,56. However, less is known about how cholesterol affects cardiomyocyte physiology and how its depletion affects cell contractility.
The contractility of cardiomyocytes is controlled by two different actin cytoskeletal architectures: sarcomeric myofibrils and the sub-sarcolemmal actin. There is substantial evidence that the sub-sarcolemmal actin contractility is regulated in a similar manner as in non-muscle cells57. By using MβCD incubation, an acute and very specific method for depleting cholesterol from the plasma membrane31,58, this work showed important features acquired by neonatal ventricular rat cardiomyocytes when submitted to cholesterol depletion. As expected, similarly to what was previously observed for immortalized fibroblasts4, OT-mediated tether extraction, revealed an increase in both membrane tension and bending rigidity of the cardiomyocytes submitted to MβCD incubation. Tether extraction on PMVs also showed an increase in membrane tension but not in bending rigidity pointing out to an exclusive contribution from the sub-sarcolemmal actin cytoskeleton to bending rigidity increase in cholesterol-depleted cardiomyocytes.
TFM experiments demonstrated that even though cardiomyocytes increased both peak and trough strain energies upon cholesterol depletion, the resultant strain energy (peak minus trough) shows a decrease in cell contractility. In addition to that, the increase in trough strain energy and decrease in overall contractility suggests impairment in cell shortening and increase in isometric force generation in low cholesterol scenarios. Interestingly, isometric cardiac contraction is correlated with myofibrillar disarray in left ventricles obtained from patients with idiopathic hypertrophic subaortic stenosis59. Moreover, our TFM data also showed that more rapid and less rhythmic contractions were observed in low-cholesterol scenarios. It is known that cardiomyocyte contraction is a multifactorial process that is regulated in three main interdependent levels: Ca2+ handling, adrenergic signaling cascade, sarcomeric integrity and caveolae domains are thought to be important to all of them (Fig. 7).
Caveolae are flask-shaped and cholesterol-enriched plasma membrane invaginations that regulate many cellular processes including lipid homeostasis and adaptation to membrane tension47,60,61. Caveolins are proteins that are responsible not only for scaffolding caveolae but also are important for cell signaling62 and caveolin-3 is the most abundant isoform in striated muscle10,63. By doing immunostaining we saw that cholesterol depletion changed drastically the distribution of caveolin-3 in the cardiomyocyte plasma membranes with the control cells having a more punctate distribution of caveolin-3 along the myofibers whereas MβCD treated cells showed a disperse labeling of that protein. This result can also be corroborated by the increase in membrane tension that we saw on our OT-tether pulling experiments since cells tend to disassemble caveolae under mechanical stress61.
Calcium is one of the three main pillars that regulate cardiomyocyte contraction. For the neonatal cardiomyocyte case, extracellular Ca2+ plays a fundamental role in contraction44,48,49,50,51,64 since those cells lack T-tubules11. T-tubules are membranous cisternae networks that couple extracellular Ca2+ entry to calcium-induced-calcium-release that is essential for adult ventricular cardiomyocytes contraction65,66. Since neonatal heart cells depend mostly on extracellular Ca2+ to contract (revised by ref. 67) and we were doing perturbations in the plasma membrane cholesterol we sought to investigate whether those perturbations were affecting Ca2+ channels, membrane-bound Ca2+ channel regulators and Ca2+ influx.
Some members of the G-protein Coupled Receptors (GPCRs) family, known as adrenergic receptors (β1 and β2-ARs, e.g), regulate cardiac Ca2+ signaling, at the plasma membrane level. Upon agonist binding to the β-ARs, adenylyl-cyclase (AC) is activated and accelerates the production of cAMP, which in turn activates PKA. PKA phosphorylates the pore-forming subunit of LTCC, Cav1.2, promoting opening of the channel and Ca2+ entry. It had been shown via immunoprecipitation that caveolin-3, Cav1.2, β2-AR (but not β1-AR), AC and PKA are closely associated in neonatal murine cardiomyocytes18 and it is also established that caveolin binding inhibits AC activity68. Experiments performed in murine neonatal cardiomyocytes show that Filipin, a polyene antibiotic that binds to unsterified cholesterol and disorganizes caveolae domains, increased by a factor of 2 the β2-AR stimulated myocyte contraction rate upon isoproterenol incubation69. There is also evidence that 1% hydroxypropyl-β-cyclodextrin treatment increased cAMP accumulation in rat neonatal cardiomyocyes upon agonist binding (isoproterenol, zinterol or forskolin)70. These two above-mentioned studies increase the body of evidence that demonstrates that caveolar localization dampens β2-AR signaling. Our immunostaining results show that when the cells are submitted to cholesterol depletion, Cav1.2 subunit of LTCC changes distribution within the cells being excluded from the perinuclear region. We also demonstrated, by enzymatic assay, that cAMP-dependent PKA activity increases when cholesterol is depleted even in the absence of β2-AR agonists, which might increase the number of Cav1.2 opening events leading to increase of Ca2+ influx. In fact, we saw that cholesterol reduction led to an increase in the Ca2+ probe fluorescence in the cytoplasm of the cardiomyocytes. Moreover, the Ca2+ spikes were more frequent and less synchronized in the MβCD treated cells, which correlates to our TFM results and corroborates our hypothesis that cholesterol sequestration impairs both contraction dynamics and Ca2+ handling by the cardiac cells.
The last pillar that governs cardiac contraction is sarcomeric integrity. Since we measured impairment in cardiomyocyte contraction dynamics and also changes in membrane tension in those cells we hypothesized that MβCD treatment was somehow perturbing the sarcomeric actin cytoskeleton that plays an important role in cardiomyocyte contraction71. By doing actin and α-actinin staining we demonstrated that cholesterol chelated cells showed sarcomeres that were not equally spaced throughout the myofibers with curvy z-bands as opposed to the control cells that show equally spaced and straight z-bands. Image analysis showed that, on average, z-bands get closer and the average angles between z-bands get bigger when cholesterol is removed. Impairment in myofibrillar architecture can also be correlated with decrease in cell contractility and defects in cell relaxation in cholesterol depleted cardiomyocytes, as observed in our TFM experiments. It is suggestive that impairment in cardiomyocyte shortening might give rise to myofibrillar disarray, as previously observed59.
In addition, we decided to investigate whether the disruption of the myofibers was caused by Ca2+-dependent proteolytic activity in the sarcoplasm. In order to test that, we measured cytoplasmic calpain enzymatic activity. It is known that cytoplasmic Ca2+ triggers calpain activation which in turn can degrade sarcomeric proteins such as troponins, tropomyosins and titins leading to abrogation of the z-bands organization28. Since we saw an increase in cytoplasmic Ca2+ upon cholesterol depletion we hypothesized that it might be activating calpain degradation of myofibers impacting the normal contraction behavior of those cells. In fact, we saw an increase in calpain activity due to cholesterol depletion and that correlates with the disorganization of the z-bands that we measured.
Overall this work points out to a new mechanism in which sarcoplasmic cholesterol depletion by MβCD incubation leads to impairment in cardiomyocyte contractility and calcium handling increasing cortical mechanical properties values, cytoplasmic Ca2+ concentration, calpain proteolytic activation and disorganization of myofibers.
How to cite this article: Hissa, B. et al. Cholesterol depletion impairs contractile machinery in neonatal rat cardiomyocytes. Sci. Rep. 7, 43764; doi: 10.1038/srep43764 (2017).
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
We acknowledge Yvonne Beckham for critical reading of the manuscript. This work received funding from NIH Director’s Pioneer Award and NIH RO1-GM104032 to M.L.G. BH was funded by Science without Borders (Capes-Brazil).