Project 7

Background and preliminary work: Different cardiac diseases such as myocardial infarction (MI) and heart failure are accompanied by strong sympathetic activation. MI goes along with acute elevation of circulating catecholamines as well as non-esterified fatty acids (NEFAs), as catecholamines activate peripheral adipocyte lipolysis via β₃-adrenergic receptors. This activation is transient as both, catecholamines and NEFA levels return to baseline within 24 h. Of note, we recently showed that next to this acute stimulation, especially the subcutaneous depot of peripheral white adipose tissue is affected chronically (up to 28 days) post MI. In a mouse model of cardiac ischemia, we found reduced adipocyte size and elevated expression of the main lipolytic enzyme adipose triglyceride lipase (ATGL) 28 d after ischemia ¹. 

Peripheral adipose tissue lipolysis is mainly regulated by G-protein-coupled-receptor signaling and insulin signaling. While the β-adrenergic G_αs- coupled receptor activation in response to catecholamines are considered the primary regulators of lipolysis, additional paracrine and autocrine factors contribute to the fine tuning of lipolysis as well. Adenosine is generated in adipose tissue in the context of adipose tissue inflammation, which we could show to occur after MI ¹, or from ATP released as co-transmitter of noradrenaline from sympathetic neurons ². Adenosine receptors are expressed in adipose tissue and regulate lipolysis in a diverse manner, as the different isoforms are coupled to either G_αi (A₁, A₃) or G_αs (A_2A, A_2B). It was recently shown in the Harris-group that differential expression of the A₁ and A_2b-isoform plays important adaptive roles in the regulation of lipolysis during fasting and feeding cycles, with food intake causing a switch in adenosine receptor signaling from primarily A₁-mediated suppression of lipolysis to A_2B-mediated activation of lipolysis ³. Stress–mediated release of adenosine occurs under dramatically different metabolic conditions; either fed (anabolic) conditions, or fasted (catabolic) conditions. The response of adipose tissue to stress is thus very dependent on the predominant adenosine receptor present, either A₁ in fasted conditions or A_2b in the postprandial state. We therefore hypothesize that adipocyte A_2b receptor signaling may play a major role in the lipolytic activation during the post-ischemic cardiac remodeling phase.

We and others have shown before that inhibition of lipolysis improves cardiac function in different cardiac disease models ^4-7. Interestingly the outcome after myocardial infarction strongly depends on the timing of lipolysis inhibition. While inhibition during ischemia has a protective effect ⁴, chronic inhibition using an adipocyte-specific ATGL-KO mouse model leads to a rather adverse remodeling with reduced cardiac function (unpublished).

FB activation is an indispensable step in cardiac healing after MI and FB undergo major transcriptional and phenotypic changes throughout the remodeling process, indicating a high cellular plasticity. Single-cell sequencing experiments in the past years have identified numerous transcriptionally different states of cardiac FB after cardiac ischemia. From resting FB with typical markers such as Col1a1, fibroblasts get activated to early “injury response”-FB (markers Mt1-2, Ccl2, Cxcl1) at day 1 post-ischemia, and further to “myofibroblasts” (markers Acta2, Cthrc1, Postn) around day 3. At later stages further transcriptional states led to the identification of “matrifibrocytes” (markers Comp, Chad, Cilp2) around day 14 and “late-resolution” FB (markers Col8a1, Meox1) from day 14-28 ⁸. 

Interestingly, in different models of cardiac disease cardiac fibrosis was affected by inhibition of lipolysis. In the context of isoproterenol-induced cardiac damage, inhibition of lipolysis reduced cardiac fibrosis and collagen deposition as well as Postn-expression in cardiac FB ⁷, possibly via reduced secretion of Galectin-3 from white adipose tissue. In contrast, we could show in adipocyte-specific ATGL-KO mice increased scaring after MI. 

Ongoing spatial transcriptomic analyses of hearts of these ATGL-KO and control mice after 24h of reperfusion revealed differential expression of markers indicative for injury response- and myofibroblasts. Early FB activation marker as Ccl2, Postn and Acta2 were downregulated in the ischemic zone and border zone of hearts in ATGL-KO animals. Interestingly, this differential expression was reversed in the remote zone of these hearts, with an upregulation of several FB marker genes as Mt2, Ccl2, Timp1, different collagens (Col1a1, 3a1, 4a1, 4a2, 5a3) as well as Tgfb2 and Bgn, indicating a specific regulation of FB activation depending on localization within the ischemic and reperfused heart. 

Hypothesis: It is well known that metabolic reprograming of cardiac FB is able to alter their activation, as for example the switch to glycolysis promotes collagen production ⁹. Therefore, we hypothesize that lipolytic products from white adipose tissue such as free fatty acids, but also other adipose tissue-derived factors, such as adipo- and cytokines, alter cardiac FB trans-differentiation and substrate metabolism during the remodeling process after MI.

The overarching aim of this project is the identification of key factors and functional interrelationships between lipolysis and adipocyte A_2B signaling on fibroblast plasticity after myocardial infarction. Specifically, we propose to investigate:

1. the role of lipolysis and adipose tissue derived-factors in fibroblast plasticity and infarct healing. 

2. the role of adenosine signaling in post MI lipolysis and fibroblast differentiation.

Work program: Two genetically modified mouse models will be used: a tamoxifen-inducible adipocyte-specific KO of ATGL, the first and rate limiting enzyme of lipolysis and a tamoxifen-inducible adipocyte-specific KO of adenosine A_2B receptor. All mouse models will undergo closed chest cardiac ischemia followed by different reperfusion times from 30 minutes up to 28 days. In the closed chest-model of experimentally induced MI mice first undergo surgery to introduce a ligature around the left coronary artery without ligating it. This is followed by cardiac ischemia, without reopening the chest, 3-7 days later. This allows the discrimination of lipolytic activation from the surgery itself from the lipolytic activation due to cardiac ischemia. The cardiac remodeling process will then be analyzed with respect to hemodynamics via echocardiography and morphology via histology. A special focus will be thereby be on scar formation after cardiac ischemia, which can be monitored via histological staining for collagen (Gomoris and Sirius Red) and immunostaining for markers of FB activation (Postn, Ctrhc1, Acta2, Col1a1, Col3a1) and flow cytometric analysis of immune cell subsets. 

To identify the specific effect of lipolysis on FB differentiation and phenotype we plan to use next-generation sequencing methods such as single-cell RNA sequencing and spatial transcriptomics. scRNA seq will allow to identify transcriptionally distinct fibroblast states and how their abundance and gene expression profile is altered due to inhibition of lipolysis. Using additional spatial transcriptomic analysis of cardiac tissue slices enables then also the spatial localization of the different FB cell types and their interaction with surrounding cardiac cells such as cardiomyocytes, pericytes and immune cells. In preliminary experiments single cell sequencing and spatial transcriptomics using 10X platforms was established. For bioinformatic analysis Seurat packages and R are being used and analysis pipelines already exist which will allow efficient training of graduate students.

To further identify specific adipose tissue derived factors, possibly mediating fibroblast differentiation, we plan to perform multiplex analysis and lipidomics. Multiplex analysis will allow to identify adipose tissue-derived cyto- and adipokines, and how their secretion profile is altered in response to lipolysis inhibition. With lipidomic analysis of circulating and tissue lipid species, we will analyze alterations in lipotoxic lipid species due to lipolytic inhibition and identify specific lipid species, which determine FB differentiation and phenotype. 

The in vivo experiments will be accompanied by in vitro analyses using isolated primary cardiac FB to investigate specific mechanisms and signaling pathways in response to the earlier identified adipose tissue-derived factors. We will investigate proliferation, activation, and fibroblast substrate metabolism using extracellular flux measurements in response to different lipid concentrations, specific lipid species as well as adipo- and cytokines. 

Added value of the collaboration: The groups of Thurl Harris and Bottermann/Fischer have already a successful history of collaborative work ^{4, 5, 10, 11}. Prof. Harris will contribute his expertise in the field of adipose tissue lipolysis, adenosine signaling ^{3, 11, 12} and metabolic research. Dr. Bottermann and Prof. Fischer combine their expertise in lipolysis, myocardial infarction, and cardiac fibroblast biology ^{1, 4, 5, 13-15}. During the internship in the Harris Lab the student will perform MS-based lipidomic measurements in serum and tissue samples, which were collected during his/her time in the Institute of Pharmacology in Düsseldorf. Vice versa, the doctoral researcher from the Harris lab will work in Düsseldorf on cardiac cell isolation to cultivate and characterize primary cardiac fibroblasts from adipocyte specific A_2B-receptor KO mice. These will be analyzed with respect to myofibroblast differentiation, extracellular matrix synthesis as well as cellular metabolism and mitochondrial function.