CaMKII activation in early diabetic hearts induces altered sarcoplasmic reticulum-mitochondria signaling

Prediabetic myocardium, induced by fructose-rich diet (FRD), is prone to increased sarcoplasmic reticulum (SR)-Ca2+ leak and arrhythmias due to increased activity of the Ca2+/calmodulin protein kinase II (CaMKII). However, little is known about the role of SR-mitochondria microdomains, mitochondrial structure, and mitochondrial metabolisms. To address this knowledge gap we measured SR-mitochondrial proximity, intracellular Ca2+, and mitochondrial metabolism in wild type (WT) and AC3-I transgenic mice, with myocardial-targeted CaMKII inhibition, fed with control diet (CD) or with FRD. Confocal images showed significantly increased spontaneous Ca2+ release events in FRD vs. CD WT cardiomyocytes. [3H]-Ryanodine binding assay revealed higher [3H]Ry binding in FRD than CD WT hearts. O2 consumption at State 4 and hydrogen peroxide (H2O2) production rate were increased, while respiratory control rate (RCR) and Ca2+ retention capacity (CRC) were decreased in FRD vs. CD WT isolated mitochondria. Transmission Electron Microscopy (TEM) images showed increased proximity at the SR-mitochondria microdomains, associated with increased tethering proteins, Mfn2, Grp75, and VDAC in FRD vs. CD WT. Mitochondria diameter was decrease and roundness and density were increased in FRD vs. CD WT specimens. The fission protein, Drp1 was significantly increased while the fusion protein, Opa1 was unchanged in FRD vs. CD WT hearts. These differences were prevented in AC3-I mice. We conclude that SR-mitochondria microdomains are subject to CaMKII-dependent remodeling, involving SR-Ca2+ leak and mitochondria fission, in prediabetic mice induced by FRD. We speculate that CaMKII hyperactivity induces SR-Ca2+ leak by RyR2 activation which in turn increases mitochondria Ca2+ content due to the enhanced SR-mitochondria tethering, decreasing CRC.


Western blotting
Hearts were freeze-clamped, pulverized, and processed as previously described 4

Mitochondria isolation
Animals were anesthetized, and hearts were immediately excised. Heart mitochondrial purified fractions were obtained as described earlier 6  Homogenates were then centrifuged at 8000 g for 10 min to discard the excess of proteinase.
Pellets were resuspended in 4 mL STE buffer and centrifuged at 700 g for 10 min to discard nuclei and cell debris. The supernatant was then centrifuged at 8000 g for 10 min to isolate the mitochondrial fraction. Finally, the pellet was washed and resuspended in the same isolation buffer obtaining a final protein concentration of 10 mg/mL. The whole procedure was carried out at 0 -4°C. The isolated fraction consisted of mitochondria able to carry out oxidative phosphorylation. Protein concentration was measured by the Lowry assay using BSA as a standard 7 .

Mitochondrial O2 consumption
Mitochondrial O2 consumption was measured using a Clark-type O2 electrode for high- Respiratory control ratio (RCR) was calculated as the ratio between state 3 and state 4 respiration rates 6,8 .

Mitochondrial ATP production rate
ATP production was measured in freshly isolated mitochondria using the luciferin/luciferase assay in a microplate reader (Varioskan® LUX, Thermo Scientific, MA, USA as % of the CD ATP production rate 9 .

Mitochondrial hydrogen peroxide (H2O2) production rate
To study ROS generation from freshly isolated mitochondria, H2O2 production was measured by fluorescence spectroscopy using the Amplex Red/HRP system. Grids were examined by TEM (JEOL JEM 1200; Zeiss) using a digital camera (Erlangshen ES1000).
Morphometry analysis. TEM images were used to measure different morphometric parameters as previously described 14 . Mitochondria roundness and diameter were measured on 12000x magnification images. Roundness reports how round is an object (mitochondria) as compared to a circle. Circular objects will have a roundness = 1; other shapes will have a roundness different to 1. It is determined by the following formula: where Pm is the perimeter and A is the area. SR-mitochondria distance was measured on 80000x magnification images. For the latter, thresholded images were skeletonized to establish the centerline of the SR membrane and of the OMM. At least ten perpendicular lines were then traced between both skeletonized lines and their average size was represented as the distance between both organelles. Moreover, the space between organelles was colored for better discrimination of the microdomains.
Mitochondrial density quantification. Samples were fixed as described before to obtain the TEM images. Mitochondria density quantification was carried out on at least nine 12000x and/or 30000x magnification images. Quantification was normalized per µm 2 of the image to express density as number/µm 2 (nº/µm 2 ).

Mitochondrial holes quantification.
Lesser electrodense areas (holes area) present in the mitochondria were quantified in TEM photographs in 30000x and 12000x magnification.
Either the holes or total mitochondrial areas were skeletonized using the program ImageProPlus. The holes areas were normalized by the total mitochondrial area. Figure S1. Disarrangement tissue in prediabetic heart mice. Representatives

Figure S2. Full-length western blots from WT heart mice.
Full-length western blot of figures 2B and 5. The red squares, delimitate the representative blots selected to be shown in the figure of the main text. The blue squares, delimited blots of samples that were disregarded for visible experimental reasons (broken gel, tail effect, bubbles, smeared bands, repeated samples, etc). The blots that are not delimitated were used for average data. Full gels are not available since once running and transferred, the membranes were cut according to the molecular weight of the primary antibody to search.

Figure S3. Full-length western blots from AC3-I heart mice.
Full-length western blot of figures 6B and 9. The red squares delimitate the representative blots selected to be shown in the main figure.