Skip to Main Content

“Oh crap,” Kevin Costa thought. A biomechanical engineer, he had labored for months to create miniature hearts, starting with a slurry of stem cells from rats and then coaxing them with chemicals in comfy little lab dishes not only to turn into specialized cardiac cells but also to gallivant around until they formed a three-dimensional, mini-me organ.

As he peered into a jury-rigged beaker held together with binder clips, however, Costa didn’t need a microscope — heck, he didn’t even need reading glasses — to see that something had gone crushingly wrong. The little heart had a hole at one end.


As his own heart sank — well, this didn’t work, Costa remembers thinking — he kept looking at the light-pink organoid. Then he had a thought. He carefully pipetted red dye into the nutrient broth that was keeping the organoid alive. Seconds later, he saw a little puff of color come from the hole. Then another puff … puff … puff, like a cigar lover blowing smoke from a stogie.

The heart organoid was pulsing out small billows of dye that had seeped into it. The mini-heart was pumping.

Ever since Costa and his colleagues, then at Columbia University, created rat heart organoids able to mimic that defining property of actual hearts, he has been barreling ahead to do the same thing for human hearts. Now Novoheart, the company Costa co-founded in 2014, has succeeded, creating the only known heart organoid — it’s about half an inch across — with a pumping chamber.

It’s all part of a growing effort to overcome shortcomings of using animals or human cell cultures in research, by building more realistic human organ models. Such models have a range of uses: By observing organoids created from the cells of people with a genetic disease, for instance, scientists are trying to understand what goes wrong and when, in hopes of finding ways to prevent or fix the problem. More immediately, they are using organoids to screen drugs for safety and efficacy, ideally identifying problems before a company takes an experimental compound into expensive clinical trials — but definitely before people are harmed, as happened with Merck’s arthritis drug Vioxx.


Most heart organoids are two-dimensional, basically a layer of cells with no true structure. A handful of labs have 3-D human heart organoids, like one created at Clemson University, but without chambers and other characteristic properties of actual human hearts.

Heart organoid still
Human heart organoids created from stem cells mimic properties of the real organ, including pumping fluid. NovoHeart

“The 3-D geometry of a heart is important to its function,” Costa said during an interview this month at his lab at New York’s Icahn School of Medicine at Mount Sinai. “With our heart organoids, we’ve recapitulated what makes a heart a heart: electrophysiology and pumping.”

As described in an upcoming paper in the journal Biomaterials, the organoids are “biomimics” of human hearts, albeit with only a single pumping chamber. The “human hearts in a jar,” as Costa calls them, share many of the heart’s mechanical and electrical properties, including ejecting fluid from a ventricle-like chamber, developing calcium and other ion channels, and throbbing with electrical signals called action potentials. They are studded with receptors which, when stimulated with adrenaline, make them beat faster, just like an actual heart that’s experienced fear or excitement. And like a human heart that beats harder when its owner exercises, the chambered heart organoid beats harder when stressed.

To make heart organoids, Costa and his colleagues start with adult skin cells. Using a Nobel-Prize-winning technique discovered in 2007, they turn the cells into stem cells, which can differentiate into any kind of cell. They add biochemicals that direct these induced pluripotent stem cells to become cardiomyocytes, the main component of heart muscle, as well as crucial support cells. Without the latter, Costa was surprised to discover, the organoids “didn’t form structures,” he said. “You need accessory cells like fibroblasts and stromal cells to form a realistic heart organoid.”

That also requires putting the induced pluripotent stem cells into collagen gel in custom-made molds — basically, the organoid version of suspending pineapple chunks in barely set Jell-O. Over a few days, the cells spontaneously form 3-D structures that mimic a heart. The micro-hearts punch above their weight: Their “twitch force,” the strength of their muscle contractions, is one-tenth that of an adult heart, said Costa, who switched his research from lungs to hearts when a beloved grandfather died of cardiovascular disease.

The Novoheart team, which includes co-founder Ronald Li of the Karolinska Institute, recently created its first 3-D heart organoids from people with genetic mutations that cause one or another heart disease, such as familial cardiomyopathy. One is a rare condition that scientists haven’t been able to make lab animals develop, so the organoids could allow researchers to study the development of the disease, from nascent fetal-like heart to a defective adult one, for the first time and, if all goes well, develop drugs or other interventions to prevent or reverse the disease.

Human hearts have four chambers, so the single-chambered version is only a start. But there are hints that might be sufficient for several purposes, including screening the safety of drug candidates. In another study scheduled for publication, Costa and his colleagues tested 17 drugs on a simpler version of the organoids to see if the compounds disrupted their function. For six weeks the scientists dripped minuscule doses onto 3-D strips of heart muscle, periodically measuring the structures’ “twitch force.”

“In some cases it was obvious” when a drug either sent the micro-heart into overdrive, which in real life might cause arrhythmia, or depressed its contractions, which could cause heart failure, Costa said.

The scientists were blinded to which drugs they were testing. Once all the results were in, it was time for the big reveal. An executive at the pharmaceutical company that asked Novoheart to do the test (Novoheart declined to name it) showed Costa and his colleagues, via Skype, which drugs had caused heart problems in real-world patients. The organoids perfectly predicted the presence or absence of heart toxicity for all the drugs in two drug classes, and got two wrong in the third class, for a score of 15 correct calls out of the 17 drugs.

Investors, however, haven’t been impressed. Novoheart is a penny stock, trading on a Toronto stock exchange.

That might be a consequence of regulatory uncertainty. Drug maker GlaxoSmithKline has “an active interest in organoid models given their potential to be a more human-relevant test system” and reduce the use of lab animals, said spokesperson Mary Anne Rhyne. But the Food and Drug Administration has “not yet formally adopted the use of organoids as an alternative to meet toxicology testing requirements,” she added, “and further research is needed to replicate the complete physiology of the heart.”

The lack of investor enthusiasm might also reflect a hunch that other systems, though less realistic than heart organoids, will be good enough to test experimental drugs for toxicity. For decades, the standard approach has been cells in a dish, though that failed to flag the cardio-toxicity of Vioxx and many compounds that damaged the hearts of patients in clinical trials or of lab animals in preclinical research. The biggest competitor to organoids are “organs on a chip,” including one that Kit Parker of Harvard’s Wyss Institute for Biologically Inspired Engineering unveiled last month. Its cells organize into tissue that contracts like heart muscle, but without fully 3-D structures such as a chamber.

“The underlying biology is fundamentally different in 2-D vs. in 3-D,” Costa argued. “For many uses you need a functional, pumping cardiac chamber.”

And, yes, when he gets a quiet moment and thinks about how closely science can mimic creation, he does sit and watch the organoids silently beating, pumping, and tingling with the electricity that gives human hearts life.