The heart is composed of cardiomyocytes (CM) and non-myocytes (NM). The latter include fibroblasts (FB), endothelial, and immune cells. FB and macrophages (MΦ) are electrically coupled to CM in the native heart [1,2]. We are interested in how FB and MΦ affect the electrical activity during myocardial remodeling in response to injury. In order to study NM-CM interactions, we developed an optogenetic approach based on cell-type specific expression of the light-gated cation channel Channelrhodopsin-2 (ChR2). This allows us to not only study cell-specific contributions to overall cardiac electrophysiology, but also enables 3D reconstruction of FB and MΦ microarchitecture in scar tissue, border zone and remote myocardium.

We used Cre-Lox tamoxifen inducible recombination to target ChR2-eYFP to NM in the mouse heart. More specifically, inducible Tcf21 and Cx3cr1-Cre lines were used for ChR2 targeting to FB and cardiac resident MΦ, respectively. Heart from adult Tcf21- and Cx3cr1-ChR2 mice with and without left ventricular cryo- or ischemia reperfusion injury to generate a non-transmural scar were used. We performed electrical and optical pacing of isolated Langendorff-perfused hearts to evaluate the effects of FB or MΦ depolarization on CM electrophysiology when ChR2 is activated in the scar and/or border zone. To visualize the fluorescently labelled cells in intact tissue, we optically cleared hearts using X-CLARITY and imaged them with confocal microscopy. This allowed us to reconstruct 3D models of FB and MΦ, and to assess their morphology, distribution, surface area and fractional volume in near-native tissue.

Experiments showed that depolarization of FB and MΦ by ChR2 activation in cryoinjury hearts can alter cardiac conduction by modulating action potential duration and restitution in the scar and scar border. Additionally, we studied the morphology and distributions of FB and MΦ in healthy and post-injury cleared hearts. We found in healthy hearts that FB have elongated shapes and thin branches, which form an interconnected network that follows the orientation of CM. Surprisingly, resident cardiac MΦ show a very similar morphology, although they are arranged as solitary cells. Volume and surface area of FB and MΦ show no significant difference (FB volume 2192±247 µm3, FB surface area 1768±192 µm2, n= 36 networks containing at least 126 cells; and MΦ volume 1859±157 µm3 and surface area 1457±102 µm2, n= 36 individual cells). In post-injury hearts, we found larger populations of FB and MΦ in the scar, scar border zone and even in remote tissue. In the scar, FB form a complex and dense network that surrounds other NM populations. Abundant MΦ are found, forming small networks of MΦ in the scar. The fractional volumes of FB and MΦ in healthy tissue were 2.5±0.3 %, n= 10 and, 1.6±0.2 %, respectively, n= 8 in left ventricular regions (mean±SE, n= regions). For injured hearts, FB and MΦ occupy a volume of 12.5±1.1 %, n= 12 and 6.86±0.5 %, n= 5 in the scar, and 7.8±0.3%, n= 16 and 2±0.1%, n= 5 in remote tissue from cryoinjured hearts. Our study showed the electrophysiological and structural relevancies of FB and MΦ in situ in post-injury hearts. Future work will be focused on FB and MΦ roles in arrhythmogenesis using optogenetic tools.

 

References

[1] Quinn TA et al. Proc Natl Acad Sci 2016/113(51): 14852–7 [PMID: 27930302]

[2] Hulsmans M et al. Cell 2017/169: 510-522 [PMID: 28431249