​ Cardiology

Bioprinting 3D heart pumps

A concept that is gaining traction

By Molly Kupfer, PhD, and Brenda Ogle, PhD

eart disease is the leading cause of death worldwide, due in large part to the low regenerative capacity of the heart. With recent advances in stem cell biology, cardiac tissue engineering with human cells has emerged as an avenue to replace lost muscle after a cardiac event and to produce human models in vitro that can be used for disease modeling and testing of drugs and medical devices.

Early engineered heart tissues, pioneered in the late 1990s and early 2000s, consisted of geometrically simple structures (strips or rings) made by casting cardiomyocytes in a protein-based gel. While such tissues can recapitulate the contractility of cardiac muscle, their lack of geometric complexity limits their capacity to reflect clinically relevant characteristics of the heart. That is, while they can generate force, they possess no internal chambered structure with which to pump fluid.

“Printing” cardiac tissue

3D bioprinting, wherein structures are fabricated layer-by-layer utilizing a cell-laden “bio-ink” as a substrate, has been proposed as a means to generate more geometrically complex tissues from the bottom up. The concept is gaining traction, as the ability to print tissues composed entirely of native proteins, cells, and/or biocompatible synthetic components is possible and accessible to many laboratories. Further, robust protocols have been developed for differentiating human-induced pluripotent stem cells (hiPSCs) into a variety of cell types, making it relatively easy to obtain cardiomyocytes ex vivo. However, while researchers have demonstrated the capacity to 3D-print entire heart organ models using biological materials, no one has yet demonstrated electromechanical function of cardiomyocytes within such a tissue.

​The end result is a living pump that mimics the chambers and large vessel conduits.

The fact that macroscale contractile function has not yet been achieved in a 3D-printed, perfusable, chambered heart model reflects the challenges associated with handling mature cardiac muscle cells. More specifically, cardiomyocytes do not proliferate or migrate readily. For this reason, it is challenging to achieve the high cell density required for the formation of functional cell-cell junctions while maintaining the structural support needed for an enclosed chamber. Macroscale cardiac function relies on the electromechanical coupling of individual cardiomyocytes to form an organized, synchronously contracting tissue.

Traditionally, researchers have taken the approach of differentiating hiPSCs into cardiomyocytes in a tissue culture dish, and then collecting the differentiated cardiomyocytes and 3D printing with them. However, when hiPSCs are differentiated into cardiomyocytes this way, they tend to couple to each other and form a beating monolayer. To collect the cells from such an environment for further downstream applications typically requires one to break up these connections. Hence, to incorporate these cells into an engineered tissue, it is necessary to place them in a context where they can reform these interrupted connections. This is feasible in smaller, millimeter-scale tissues, but it becomes challenging in larger, centimeter-scale tissues where the physical distance between cardiomyocytes after printing is too large to overcome.

A new strategy

Our alternative approach is to print stem cells, which are highly proliferative, and then induce differentiation of cardiomyocytes in situ following cell expansion. To enable this approach, we sought to develop a bio-ink formulation that:

1. Promotes hiPSC viability;
2. Enables hiPSC proliferation and subsequent differentiation into cardiomyocytes; and
3. Is amenable to printing complex structures.

​Building on our understanding of how native extracellular matrix proteins modulate cell behavior, we developed an optimized bioink formulation composed of native proteins found in the heart. Some of these proteins were chemically modified to enable photo-crosslinking of the printed construct in order to maintain its geometric shape and structural integrity.

To generate the printing template, an MRI scan of a human heart was obtained and scaled to the size of a mouse heart such that the longest axis was approximately 1.3 centimeters. In addition, the septum between ventricles was partially removed to provide a throughway such that unidirectional flow could be propagated through the printed structure for ease of nutrient delivery. The structure was further modified to limit the vascular connections to two major vessels extending from the top of the structure, corresponding to the aorta and vena cava from the digital template.

Prior to printing, hiPSCs were mixed into the bio-ink, which was then loaded into a syringe. The bio-ink was extruded from the nozzle using a commercial 3D printer and deposited layer-by-layer according to the print template. The tissues were printed into a gelatin support bath so that the relatively low-viscosity bio-ink would maintain its shape prior to photo-crosslinking with blue light. After crosslinking, the gelatin bath was washed away, and the structures were cultured for two weeks to allow the stem cells to proliferate and fill the tissue gaps. The stem cells were subsequently differentiated into cardiomyocytes using a previously developed small molecule-based protocol.

This in situ differentiation approach enables the cells to form connections to each other as they differentiate, similar to what would happen in human development. The end result is a living pump that mimics the chambers and large vessel conduits of a native heart while housing viable, densely packed, and functional cardiomyocytes. These human chambered muscle pumps (hChaMPs) exhibit robust macroscale contraction. The cellular makeup is primarily cardiomyocytes (approximately 88%), but there are also other cardiac cell types present, specifically endothelial cells and smooth muscle cells. Importantly, the combined cardiac cell cocktail often fully circumvented the hChaMP, and the thickness of the wall was typically between 100 µm and 500 µm. However, at its thickest regions, we show that the muscularized region can exceed 500 µm, which is much higher than any previously reported values for engineered cardiac tissues.

​It will also be useful for testing drug toxicity.

Cells of the hChaMP robustly express protein markers of cardiomyocyte structural and functional maturation, including gap junctions, ion channels, and intracellular machinery associated with the sarcolemma and sarcoplasmic reticulum. These proteins are necessary for the efficient trafficking of ions, which enables contiguous impulse propagation through the tissue. Optical mapping enabled visualization of electrical signal propagation throughout the hChaMP in real time. The average spontaneous APD80 was 499.9 ± 83.5 milliseconds, and action potentials detected on the surface of the hChaMP reflected a dramatic and predicted response to altered pacing frequency and drug stimulation.

The location of the structure from which the activity was propagated was stochastic, sometimes from the large vessels, sometimes from a region near the large vessels, and sometimes near the apex. This outcome likely reflects the accumulation of pacemaker cells or immature cardiomyocytes with the capacity for spontaneous membrane depolarization in a given region that dominates and therefore initiates the response. However, in some cases the spontaneous source of depolarization could be overcome, and the directionality of propagation altered via electrical point stimulation at another location within the hChaMP.

​

Potential applications

The primary benefit of a chambered tissue like the hChaMP is that it can replicate the pump function of the heart, allowing future researchers to trace and track what is happening at the cellular and molecular levels, introduce disease and damage to the model, and study the effects of medications and other therapeutics.

To determine pressure volume dynamics as a clinically relevant comparator for this new model system, a conductance catheter harboring a pressure transducer was inserted into one chamber of the hChaMP. The coupling of the pressure transducer with the conductance catheter enabled us to plot both pressure and volume simultaneously as a function of time, which was done for spontaneously contracting and isoproterenol-treated hChaMPs. Pressure-volume vs. time plots were used to generate pressure-volume loops, and from these stroke work could be determined despite the fact that there are no valves to resist emptying and filling. Using the pressure-volume setup, we were able to detect changes in beat rate corresponding to multiple concentrations of isoproterenol. The usual volume moved through the chambers was 0.5 µL and maximum volume moved through the chambers was 5.0 µL, which is approximately 25% that of the average stroke volume of an adult murine heart. Based on these values, we calculated an ejection fraction of 0.7% on average, with a maximum value of 6.5%.

Summing up

This advance represents a critical step toward generating macroscale tissues, akin to aggregate-based organoids, but with the critical advantage of harboring geometric structures essential to the pump function of cardiac muscle. The utility of this technology for the field of cardiology is access to a human model system that can sustain flow profiles and exhibit pressure-volume dynamics characteristic of the native heart. This model will therefore be useful for understanding remodeling associated with cardiac disease progression imposed by mechanical insult or genetic predisposition. It will also be useful for testing drug toxicity or efficacy and, given the scale, is amenable to the testing of medical devices, implantation to the heterotopic position in mice, and perhaps, one day, clinical transplantation.

Molly Kupfer, PhD, completed her doctorate in biomedical engineering at the University of Minnesota under the mentorship of Brenda Ogle, PhD. She has utilized human stem cells and 3D printing to generate living, contractile cardiac tissue for therapeutic use and in vitro modeling.

Brenda Ogle, PhD, is professor and head of biomedical engineering, professor of pediatrics, and director of the Stem Cell Institute at the University of Minnesota. Her research team investigates the impact of extracellular matrix proteins on stem cell behavior, especially in the context of the cardiovascular system.