Hesperos: A CRO For Body-On-A-Chip Enthusiasts

November 22, 2019

By Deborah Borfitz

November 22, 2019 | A subset of the engineered microphysiology systems being employed in preclinical drug development are physically coupling multiple organ models on a single device. Researchers recently succeeded in using such a system to examine how a drug and its chemical by-products target cells and other tissue at the same time, as reported (DOI: 10.1126/scitranslmed.aav1386) in Science Translational Medicine. The five-chamber system accommodates culture liver cells, two cancer cell lines and a pair of cardiac function chips, through which both blood-mimicking fluid and chemotherapy drugs can circulate.

The National Institutes of Health’s (NIH’s) Helping to End Addiction Long-term Initiative, launched in April 2018, will be supporting research involving a four-organ, human-on-a-chip system designed to concurrently predict potential toxicity and efficacy of drugs, including opioids and opioid antagonists such as Narcan. The system was developed at the University of Central Florida (UCF) and Cornell University, which licensed the technology to Hesperos.

Cornell University biomedical engineering professor Michael Shuler is co-founder of Hesperos and serves as its president and CEO. The company’s specialty is providing integration assistance to drug developers interested in using a multiorgan model for their preclinical efficacy and toxicity studies, he says. Each build-to-suit system is a tad bigger than a microscope slide, featuring a silicon-based compartment housing different types of tissue constructs and a pumpless, gravity-based system for moving liquid.

Each of the tissue constructs, relative to one another, corresponds to the organ’s actual size in the body, says Shuler. The only exception has been when Cornell built a 13-organ system and the fat and muscles were so large they had to be incorporated at one-eighth their physiological size.

Operationally, Hesperos works much like a contract research organization (CRO) that delivers study results back to sponsoring companies. Emulate is also a service organization but primarily employs sophisticated single organs and, unlike Hesperos, sells devices. “We focus mainly on cardiac and neuro tissues, the liver and some of the barriers, like the blood-brain barrier and GI [gastrointestinal] tract,” Shuler says. “But we can also build fat modules, bone marrow, and models for the kidney and lung.”

The platform has a proprietary serum-free medium that allows Hesperos to work with all types of organ models and a technique for taking functional measurements of electrical activity and force generation of cells, says Shuler. Both inventions have been patented by James Hickman, a professor at UCF’s NanoScience Technology Center and chief scientist for Hesperos.

Serum-containing media sometimes make it difficult to understand what is happening mechanistically with cells and can also prevent cells from fully differentiating toward phenotypes, says Shuler. Hickman’s serum-free medium uses only chemically defined components, giving the Hesperos platform better control over multi-organ systems and assists tissue constructs to obtain a mature phenotype.

Hesperos is the first spin-off company of the tissue chip program developed by the NIH and Defense Advanced Research Projects Agency, says Shuler. The company combines Shuler’s device-related patents with Hickman’s analytics-related ones and executed its first contract in June 2015.

Many Models

Human-on-a-chip systems are built by Hesperos based on a mathematical model of the body, divided into compartments representing different organs, says Shuler. One of the chambers, reserved for “other tissue,” models the dynamics of how drugs and metabolites get distributed through the body. The systems generally comprise three to six organs, which almost always include the liver where metabolic processes are concentrated and the heart where drug side effects are generally seen.

Neuromuscular junctions are also important because they often respond to drugs, resulting in muscle weakness in many patients, Shuler adds. A lot of late-stage failures in drug development are caused by renal toxicity.

When the experimental compound will be ingested as a pill, topically applied, or involves inhalation, multiorgan systems frequently include barrier tissues, Shuler continues. That might involve modeling of those constructs in the kidney, GI tract, blood-brain barrier, skin, or lung. “Occasionally, we work with single barrier tissue just to look at its properties, but always with the intention to integrate it into a multiorgan system.”

The platform’s tissue chambers can alternately hold single, isolated cell types or three-dimensional structures, which sometimes include organoids, says Shuler. But in general, the tissue constructs are engineered rather than self-assembling. Organoids, including intestinal models, are often “inside out” relative to their position in the body and can create an orientation issue.

Hesperos has done a limited amount of work on organ models with vasculature since most of its tissue constructs are small enough that they don’t need to have nutrients ferried in and waste carried away, Shuler says. An exception is when the company is studying the interaction of circulating tumor cells and immune cells, where the vasculature plays a key role.

One goal of Hesperos is to build systems that are cheaper than working with animals, says Shuler. Actual cost depends on what a sponsor company wants to do with a human-on-a-chip model. Which organs are part of the model can make a big difference, and the inclusion of mechanical force measurement can also be somewhat expensive. The price point has been progressively coming down, he adds, largely as a result of switching from manual to automated measurements and assembly of devices.

The systems Hesperos builds will operate for up to 28 days—a key time point in drug development for seeing chronic and acute effects, Shuler notes. Creating a human-on-a-chip system can vary widely from days to weeks for the different tissue constructs and the components are sequentially built to accommodate the timeline for project kickoff.

State of the Art

Shuler is a proponent of the mathematical modeling approach to designing multiorgan systems, interpreting results and properly mimicking what’s happening in the human body, as covered in a recent APL Bioengineering article. The techniques, all of which have been validated and in use for several decades, seem to be a neglected aspect of many of the systems currently being assembled, he says.


Drug developers are often interested in using a human-on-a-chip model to better understand the mechanistic actions of candidate compounds and use the results to extrapolate what will happen in the human body, Shuler says. None of the in vitro models Hesperos is building are perfectly mimicking humans, but “the formality of a mathematical model forces you to think about how you’re putting a system together in terms of the flow rate and the size of the organs, to make it more physiological.”

Recent advances (DOI: 10.1021/acs.analchem.8b05293) in body-on-a-chip systems include a trend toward the use of plastic materials with lower adsorption and absorption rates of drugs and development of more on-chip sensing of parameters such as temperature, pH, oxygen, glucose, cardiac response and liver toxicity. There have also been attempts to capture the dynamics of multiorgan interaction on pharmacokinetics and toxicity, and the interaction of different cells types within organs and the skin, says Shuler. More advanced systems simulate absorption, distribution, metabolism, and elimination processes.

The remaining challenges, as the paper points out, are coming up with a blood surrogate and suitable biomaterials for multiple cell types, creating a physiologically realistic fluidic microenvironment and correctly scaling systems.

“This is a new way of looking at the body that is potentially better than animal models at predicting what is going to actually happen in humans,” Shuler says. “If we can [more accurately] predict the response to drugs and chemicals we have a better chance of understanding environmental toxicology and which are the right drugs to give people.” Body-on-a-chip systems can also be used for studies, including drug-drug interactions, which are difficult to conduct with animals due to differences in body chemistry and physiology.

Personalizing Medicine

Body-on-a-chip technology still has a “ways to go” before it becomes the more economical choice for preclinical drug testing, says Shuler. But the systems can deliver potentially better results and savings to both drug developers and the broader healthcare system over the long term.

In the future, multiorgan models could be an enabler of personalized medicine, Shuler adds. Currently, drugs are often developed using inbred, genetically identical mice—not at all like humans whose biochemistry, and response to drugs, has tremendous variability.

Using a body-on-a-chip device and cells from a patient’s own body, doctors could learn if ancestry predisposes a person to respond well or poorly to various medications before prescribing the best one, he says. That may be a cost worth bearing for people with cancer.

The main danger would be in doing a biopsy on multiple organs, adding to the risk of side effects from the anesthesia, Shuler says. Another option, at least for people with chronic diseases who can wait months for answers, would be to make induced pluripotent stem cells—adult cells that have been genetically reprogrammed to an embryonic stem cell-like state—and generate enough tissue constructs to do a whole range of tests.

Regulatory Status

Many drug manufacturers are interested in the potential of body-on-a-chip technology and using it to some degree, as it facilitates preclinical efficacy and toxicity studies, says Shuler. The chief motivation are drugs that fail late, after sailing through animal studies and early rounds of clinical trials.

Hesperos works primarily with medium- and small-sized pharmaceutical companies that want to better anticipate the failures, says Shuler, noting that big pharma tends to build their own technology in-house. The NIH and American Institute for Medical and Biological Engineering recently hosted six workshops on the technology and pharmaceutical companies and the U.S. Food and Drug Administration (FDA) were invited.

European authorities five years ago accepted one of the first widely used non-animal tests to predict toxicity at the first stage in the skin sensitivity pathway, Shuler points out. A scientist at the Procter & Gamble Company received a 2015 Black Box prize from Lush Prize judges in the U.K. for his Direct Peptide Reactivity Assay. The FDA still requires animal testing for medical products, but says it is working toward an alternative approach.