Light sheet-guided analysis of cardiac perfusion in reperfused acute myocardial infarction

Background
Myocardial infarction (MI) is the leading cause of death worldwide. Upon reperfusion some capillary vessels remain dysfunctional, leading to the phenomenon of microvascular obstruction (MVO). MVO size may serve as a desirable therapeutic target to salvage damaged myocardium. Light sheet-based MI analysis has been established recently, based on staining of vascular damage, which was speculated to also show MVO. We set out to quantify cardiac perfusion in the whole murine heart and incorporate this analysis into an established pipeline.

Methods & Results
We subjected mice to an in vivo ischemia/reperfusion (I/R) protocol with 45 min of ischemia and 24 h to 28 d of reperfusion. To determine endothelial integrity, mice were intravenously injected with fluorophore labelled CD31 antibody, as well as 4% thioflavin S solution. Mice were sacrificed 10 min after injection and hearts were extracted. Hearts were then cut into horizontal slices and stained in 1% triphenyl tetrazoliumchloride (TTC) solution. Images of the slices were taken under UV and white light to visualize thioflavin S and TTC staining. Following, slices were fixated in formaldehyde, dehydrated with ethanol, bleached with peroxide and stained with eosin. Slices were again dehydrated with ethanol and cleared with ethyl cinnamate. Heart slices were imaged using a LaVision Biotec Blaze light sheet microscope. 3D modelling was done using Imaris software and custom python scripts were used for HE-pseudocolor conversion. TTC^neg marked areas of damaged tissue, which were similar in size and localisation as CD31^neg volumes, but thioflavin S^neg areas were 66% smaller. We also established two models for measurement of cardiac perfusion by injecting either fluorophore-labeled gelatine or microbeads via the aorta. Here we found that perfusion inside CD31^neg volumes was drastically decreased in comparison with remote and baseline conditions. We could further show that eosin staining in our model truthfully depicted cardiac damage at various time points during healing.

Conclusion
CD31^neg volumes following myocardial I/R seem to not indicate sites of total occlusion, but rather show distinct vascular damage apart from the no reflow phenomenon. This results from obstruction of said vessels, rather than loss of CD31 signal or endothelial cells. We could further establish a model for quantification of cardiac tissue damage via eosin staining, which correlates with vascular damage shown by CD31^neg volumes. This highlights the significance for research of the no reflow phenomenon for MI therapy.