Myocardial slices to study cardiovascular disease in vitro

The use of living animals in cardiovascular research is currently necessary due to the lack of appropriate and relevant in vitro models. Current in vitro models suffer from being reductionists (cell lines and stem cell-derived cardiac cells, often immature, no 3D structure) and/or are subjected to rapid changes induced by culture (can only be used for acute studies). The development of the living myocardial slice (LMS) model with its preserved tissue structure and ability to remain stable in culture for several days/months, therefore has the potential to have a significant impact upon the 3Rs. Several slices can be prepared from the same heart so that, for instance, 1 rat heart used in LMSs experiments can provide the same number of experimental repeats of 6-8 rat hearts used for experiments with other preparations. In addition, LMSs are ideal to study both structure and function and, with one LMS, one can perform several experiments compared with other models which need several hearts. Because slices are prepared by pure mechanical dissociation, the same slices analysed for contractility experiments can then be dissected in smaller slices and used for biochemistry, imaging and other studies. This ability to optimise the use of tissue can bring about a minimum of 50% reduction of the number of animals needed.

In our project we will test the feasibility to induce injury on slices in vitro and not in the living animals. This aspect is likely to have a profound impact on cardiovascular research in many academic labs and industry. The ability to induce disease "in a dish" means that laboratories can avoid performing severe regulated procedures aimed at inducing cardiovascular disease in living animals.

Furthermore, by using the LMS model it is now possible to optimise the use of available human heart tissue. While this is limited, transplant centres perform approximately 20-30 heart transplants per year. Another expanding source of human heart tissue is the implantation of left ventricular assist devices (LVADs), now performed routinely in many cardiac surgery centres (20-30 patients per year/per centre). The implantation of LVADs is accompanied by the removal of a core of ventricular tissue that is amenable to the preparation of the LMS model. In addition, unused donor human hearts are available through dedicated research programmes, making the use of human LMSs not unrealistic and particularly relevant.

We will test several kinds of disease in vitro using LMSs, and we will validate the use of this technique to study the effects of drugs that are toxic for the heart (some treatments for cancer bring about cardiac disease) and drugs that can be used to treat cardiac disease. Overall, we expect that the method developed by this study will became a major tool in cardiovascular research with important implications for the use of living animals in research.

Living myocardial slices (LMSs) are a new in vitro model of intermediate complexity and with the great potential to optimise and reduce the use of living animals in cardiovascular research. We have recently obtained methods that prevent or substantially reduce culture-induced remodelling of LMSs for several days. In this project, we will test the hypothesis that LMSs kept in stable conditions for several days can be subjected to relevant injury in vitro, and can be used to reduce/replace animals in cardiovascular research. We will prepare slices from rat and rabbit hearts without subjecting the animals to regulated procedures and human heart biopsies obtained from heart transplant programmes. We will apply cryoinjury to LMSs in culture, to induce localised and reproducible myocardial necrosis, induce global ischemia/reperfusion, and mechanically overload the LMSs using electromechanical simulation protocols involving high preload/afterload, to simulate the processes that lead to heart failure. From the resulting injured LMSs, we will measure contractility, parameters of calcium regulation, tissue structure, ultrastructure of cardiac cells and biochemical assays already established in our laboratory. We will validate the effects of toxic substances (antineoplastic drugs with cardiotoxicity, e.g. sunitinib) and beneficial treatments (ACE-inhibitors, beta-blockers, ivabradine amongst others) on the contractile and electrical properties of the model. The validated LMSs models of injury will have a clear role in the cardiovascular research pipeline, bridging the gap between in vitro and in vivo models, with the potential to substantially reduce the number of, and/or partially replace, living animals undergoing high severity regulated procedures and utilised in cardiovascular discovery.

Reduction: the number of animals used in studies of cardiac pathophysiology is substantial. A Pubmed search using the terms "heart ischemia reperfusion rat" indicated that 2,080 studies on this topic were published in the last 5 years. While it is difficult to estimate exactly how many animals were used in these studies, as the total value is rarely reported and depends on the type of investigation, if we assume that 50 rats were used in each study (which is in the lower range), we can estimate that at least 20,000 rats per years are used for published studies on ischemia reperfusion injury alone. If we consider all the studies which are not published, including toxicology studies performed by the pharmaceutical industry, this number can be easily twice as large, bringing the number to 40,000 rats per year. Several LMSs can be prepared from the same heart and therefore, in a rat experiment (for each rat heart used for the LMS method) 6-8 rats are used for experiments with Langendorff-perfused hearts or wedge preparations. Even if we assume that only 50% of the I/R studies can be replaced using LMSs, a net reduction of 17,000 rats per year is expected. When searching Pubmed for "heart failure in rats", 2413 studies were published in the last 5 years. With the same caveats made for the I/R experiments, at least another 17,000 rat procedures per year can be avoided using LMSs.
In addition, LMSs are ideal for structure- function correlation experiments and, with one LMS, one can perform several experiments compared with other models which need several hearts. For example, if experiments involve isolated myocytes, the hearts undergoing enzymatic digestions are unlikely to be suitable for other experiments. Because slices are prepared by pure mechanical dissociation, the same slices analysed for contractility experiments can then be dissected into smaller slices and used for biochemistry, imaging and other studies. This ability to optimise the use of tissue alone can bring about a minimum of 50% further reduction of the number of animals needed and is, of course, very important when using human myocardial biopsies.
Replacement: Our model can provide replacement for several reasons. With the LMS model, it is now possible to optimise the use of available human heart tissue. While the supply of this tissue is limited, transplant centres perform approximately 20-30 heart transplants per year. Another expanding source of human heart tissue is the implantation of left ventricular assist devices (LVADs)20, now performed routinely in many transplant centres (20-30 patients per year/per centre). The implantation of LVADs is accompanied by the removal of a core of ventricular tissue that is amenable to the preparation of LMSs8. In addition, unused donor human hearts are available through NHSBT or dedicated research programmes, making the use of human LMSs realistic and particularly relevant. During 2018, we received 4 donor hearts and this number is likely to increase in 2019, due to optimisation of the logistical aspects of the collection and transport. These experiments replace the use of animals, particularly those subjected to regulated procedures aimed at generating severe heart failure.
Our method is also an example of partial replacement, where hearts are taken from animals without the need of regulated procedures, as we will induce injury in vitro and not in the living animals. This aspect is likely to have a profound impact on cardiovascular research in many academic labs and industry. In our lab we perform several regulated procedures, including coronary artery ligation, heterotopic abdominal transplantation and aortic banding, all classified as moderate/severe. The ability to perform injury directly in vitro is already replacing many of these experiments with great impact on the 3Rs.