Will Organ-On-A-Chip Models Be Viable Alternatives to Animal Testing?

By Deborah Borfitz

November 20, 2019 | Organ-on-a-chip models could fundamentally change the way drugs make their journey to clinical trials and reduce the need for animals in laboratory experiments. The new testing technology is of great interest to pharmaceutical companies looking to better understand the effects of medicines—especially the unintended consequences of their administration—and will facilitate study of not only toxicity, but also the mechanism by which drugs work and the timing of their delivery and physiological response, according to Vanderbilt University professor John Wikswo, a biological physicist and founding director of the Vanderbilt Institute for Integrative Biosystems Research and Education.

As Wikswo defines them, organs-on-chips are microfluidic devices that are populated with living cells, most often human, to create two-dimensional (2D) or three-dimensional (3D) microphysiological systems that recapitulate human physiology better than cells grown on flat plastic. They’re in the same class with, but more complex than, either spheroids (typically formed from cancer cell lines or tumor biopsies) and organoids (self-organized, organ-specific cultures typically derived from stem cells). The coupling together of different organ-on-chip models allows researchers to mimic human physiology better than any single chip could do individually.

The human body has roughly 200 organs and they don’t yet all have a chip counterpart, says Wikswo. Around two dozen of them have been micro-engineered to date—including the blood-brain barrier, gut, kidney, lung, parts of the female and male reproductive systems, the mammary gland, bone marrow, cardiac and skeletal muscle, and the bone-cartilage interface—as well as multiple liver-on-a-chip models.

Wikswo holds 18 patents and has multiple patents pending related to the instrumentation and control of cells and the support hardware for organs-on-chips. These include the MicroFormulator, an innovative platform for controlling the concentration of drugs in each well of a multiwell plate commonly used in biomedical and clinical research. Organs-on-chips, at their core, are all microfluidic cell culture devices.

The organ chips themselves are transparent and roughly the size of an AA battery, each with its own instrumentation and software. They vary by purpose as well as size, shape and what they grow on, says Wikswo. One of the new economy models coming out of Vanderbilt is shaped like a puck, and others have been engineered to resemble bendy posts or curling levers.

At the University of Washington’s (UW) Kidney Research Institute, the focus for the past seven years has been on bioengineering kidney tubules, the kidney filtration barrier and kidney microvasculature on chips for several purposes, says Director Jonathan Himmelfarb. One is to model disease and another is to understand how the kidneys metabolize drugs. A third driver is to learn how the kidneys gets injured by xenobiotics (including drugs) and metabolites. The bulk of research funding comes from the tissue chip program of the National Center for Advancing Translational Sciences (NCATS).

The research team includes vascular biologists, biomedical engineers, pharmaceutical scientists, pharmacologists, and stem cells experts, Himmelfarb says. Among them is Benjamin Freedman, one of the first to develop human kidney organoids from induced pluripotent stem (iPS) cells and to use CRISPR CAS-9 genome editing in organoids to create a disease model of polycystic kidney disease.

Many Challenges

“This is a very young field and, academically, the challenge has been to design, build, and validate the chips,” Wikswo says. The chief obstacle to overcome is their isolation from other organs during testing, although functional coupling of up to 10 organs has been successfully demonstrated.

Cornell has physically coupled as many as 13 organs, but mostly to demonstrate the capabilities of its human-on-a-chip technology, reports Hesperos President and CEO Michael Shuler, a biomedical engineering professor at Cornell University. “From a practical point of view, three to five organs is usually enough to address most of the questions a pharmaceutical company might be interested in.”

Two additional problems are engineering systems with fluid flow and tissue constructs that approximate what happens in the human body, says Shuler. Imperfect systems can still be tremendously useful, given that only 10% of drugs going into clinical trials are currently coming out as approved, useful products. The odds might be increased to 30% by using the technology in late-stage preclinical development to choose which compounds to pursue—and lower the opportunity cost of choosing wrong.

These systems can likewise be helpful for clinical toxicology testing, although it has been “less explored” for that possibility, Shuler says. Standard toxicity testing is a 1930s methodology using experimental animals, but “whether a drug kills a rat or not doesn’t say much about what happens in humans.” Animals models accurately predict human response to drugs only about 3% to 8% of the time.

Among the multiple other challenges in using organ-on-a-chip systems for drug research, top on the list for Himmelfarb is finding cells that will respond appropriately to stimuli in the bioengineered environment. Reproducibility is also an issue, at least if the systems are going to start screening libraries for drug discovery, especially by the pharmaceutical industry, he says. Building the right vasculature into the system is also technically challenging.

“This field is moving very fast,” Himmelfarb continues. “It was almost nonexistent 10 years ago, and now… every few months there is huge progress.”

Outside of its work with NCATS, funded by the National Institutes of Health (NIH), UW’s Kidney Research Institute deals primarily with smaller, agile biotechnology companies interested in making therapeutics for kidney diseases, Himmelfarb says. Many of the drugs currently under development are biologics or antibodies that aren’t well suited to testing in animal models.

Pharmaceutical companies are hoping organ-on-a-chip systems will prove to be a better fit for toxicology testing than animal models, which are known to give false positives and false negatives, says Himmelfarb. On the discovery side, there’s also a lot of interest in both organoids and organs-on-chips for research purposes.

NIH-funded tissue chip researchers are partnering with the nonprofit International Consortium for Innovation and Quality in Pharmaceutical Development (IQ Consortium) to test and develop the tissue chip devices, Himmelfarb notes. GlaxoSmithKline, Pfizer, and AstraZeneca compounds are being used to perform functionality tests on microphysiological platforms.

Roche and Genentech are among the other big-name drug makers that have invested heavily in organ-on-a-chip technology, Wikswo says, driven by their interest in reducing the unanticipated toxicity of drugs while also ensuring their effectiveness.

While Wikswo concerns himself with the theoretical—including how to engineer and scale the systems—Himmelfarb is a pure pragmatist. “If these systems are not better than animal models or other in vitro systems for disease modeling or testing new therapeutics out, then I will not be all that interested. My focus in on translatability of these systems to clinical drug development.”

Better Biology

Any microphysiological system is a step up from traditional 2D biology on plastic. As was pointed out in a 2017 paper Wikswo co-authored for Experimental Biology and Medicine (DOI: 10.1177/1535370217732765), immortalized cells cultured on stiff plastic that gorge on sugar, don’t exercise or sleep, and live in the dark in their own excrement are not physiologically relevant models of human tissue function.

“Despite the limitations of traditional cell culture, we have learned a tremendous amount of biology this way,” says Wikswo. “The only question is how relevant those self-consistent studies are to a particular human disease. You can study something to death, and it makes perfect sense, but it may or may not translate to the intact human.”

Organoids can be a good alternative, Shuler says. “Organoid culture techniques are simpler in some sense and don’t require as much engineering. They self-assemble.”

Wikswo agrees, noting that organoids can replicate much of the complexity of an organ or express selected aspects of it. His Vanderbilt colleagues Lisa McCawley and Dmitry Markov succeeded in growing mammary epithelial cells into small spheres that produce milk protein, creating a “powerful tool for studying the effects of drugs.”

In recent months, there has been a steady flow of organoid-related news. Researchers in Singapore developed miniature kidneys to better understand how kidney disease develops in individual patients. Scientists in California have used retinal organoids to investigate the cause of eye diseases and test potential treatments, as well as mini-brains to observe brain waves resembling those of preterm babies.

On the International Space Station, astronaut-scientists have also studied the effects of microgravity on human brain organoids. And Cincinnati Children’s Hospital reports that it has created the world’s first three-organoid system comprising a liver, pancreas and biliary ducts—which could hasten the arrival of precision medicine and transplantable, lab-grown tissues.

Hype and Hope

University of Washington researchers have been working with NCATS to use human kidney organoids for high-throughput screening of drugs that inhibit cyst formation specific to polycystic kidney disease, says Himmelfarb. “We’re now putting those organoids into microphysiological systems and shown the rate of cyst growth changes considerably, giving us insights into the mechanism of cyst formation in a genetic disease.”

Use of organoids in regenerative medicine is still many years away, Himmelfarb says, although the idea has inspired a fair amount of hype. “We have to be careful not to give patients too much false hope that this is around the corner.”

Among the many problems yet to be solved are how to turn iPS-derived cells into mature organ cells and giving them a vasculature so they can grow to a larger size and not become necrotic, says Himmelfarb. The predominant focus of UW researchers with organoids, as well as kidneys-on-chips, is thus on drug discovery leading to safe and effective therapeutics.

But UW researchers are also “very interested” in regenerative medicine, he quickly adds. They’re currently exploring the off-target effects of CRISPR gene editing in microphysiology systems and have a funding request out with the NIH to also design personalized clinical trials—which would be a first.

A few years ago, UW’s Kidney Research Institute began accepting urine samples from people around the world with various kidney diseases. Scientists extract cells in the urine that came from the kidney and programmed them to become iPS cells for growing person- and disease-specific kidney organoids, Himmelfarb explains. “We’re making a library of stem cells from these individuals.”

As an extension of that effort, the UW researchers are now hoping to learn if the chips could be used to randomize people into clinical trials based on molecular analysis of their own kidney and predicted response to various drugs, says Himmelfarb. Their partner would be an NIH-funded consortium using molecular interpretation of biopsies from people with kidney disease to design studies.

Limitations

Organoids need the proper microenvironment to grow and researchers have gotten quite creative, says Wikswo, including growing them in hanging droplets on the side of a petri dish using special, low-adhesion well plates.

“If you can develop an organoid for drug testing, clearly it is best to do it,” he says. “It’s easy to manipulate and relatively inexpensive.” The connection between cells is good, and organoids can be transplanted to avoid the problem of tissue rejection.

The difficulty with organoids is their limited lifespan. “They keep growing and once they become more than about three times the diameter of a human hair [300 microns] the inside becomes necrotic and dies from lack of food and lack of oxygen and no place to pee,” says Wikswo. The only countermeasure is to break them up into small pieces after a few days and let them reform.

That makes organoids impractical for studying long-term effects over weeks or months, Wikswo continues. It will be another few years before researchers figure out how to give organoids capillaries that are functionally connected to the human arterial system.

This shortcoming was exposed years ago when researchers began to perform pancreatic islet transplants for type 1 diabetes using deceased donor islets, originally injected into the abdominal cavity and more recently into a large vein entering the liver. The harvesting of the islets from the donor destroyed the original connections to the perfusion capillaries within each islet, Wikswo notes. The survival and insulin-producing capabilities of the implanted islets were determined in part by how well they revascularized and hence could receive nutrients and oxygen, deliver insulin and remove waste products.

Several groups are working to vascularize organoids and get these small vessels connected to an external perfusion system, says Wikswo. He expects that a clean demonstration of an externally perfused organoid is a year or two away.

One limitation of organoids is that many of them would be required to provide enough tissue to do sophisticated, untargeted, multi-omic screening, Wikswo says. Also, since the interior of each organoid is difficult to access, it’s hard to use organoids to study biological barriers—for example, the blood-brain barrier, the blood-testis barrier and the epithelial cells that protect the lungs and the gut from pathogens. “The interesting substances might be inside, but you can’t yet get readily connected to it.”

And unlike engineered systems, Shuler says, organoid systems have also been unable to sustain bacteria in the gut.

None of these limitations apply to organs on chips, which have different strengths and weaknesses.

Improved Tooling

The MicroFormulator, developed by Wikswo and his group at the request of a researcher at AstraZeneca, is addressing another downside of traditional testing of compounds in well plates—no liver to metabolize the drugs and no kidney to transport them away. The result is a mismatch with what drug concentrations will look like over time in a well plate (constant in time) as compared to mice and humans (rising and then more slowly decreasing), he says.

The platform, now licensed to CN Bio Innovations, supports both biology on plastic and organs on chips, says Wikswo. Researchers can use it to impose a different pharmacokinetic profile on each organ chip or well of a multiwell plate.

The pharmaceutical industry is interested in the MicroFormulator because the behavior of cancer drugs is predicated on a sustained dose above a certain threshold, but not so high that it starts killing people, says Wikswo. “If you’re trying to figure out which of 10 drugs on a well plate are better than the others and the concentrations are flat, you’re going to misestimate their behavior by not having a time-dependent drug exposure profile; you’re either going to have too high or too low of a dose for too long.”

Wikswo and his team are now working on an upgrade to the MicroFormulator that will give cells and tissue chips a circadian rhythm, enabling studies to factor in the role of the body’s cyclical hormone control system on organ function. Drugs have different pharmacokinetic and pharmacodynamic properties, depending on the time of day, which can differ by a factor of two to 10 over 24 hours, he notes.

The pump and valve technologies that Wikswo developed may simplify the move to coupled organs, or so-called “homunculi,” he points out. That might better recreate the complex physiological signals that regulate the endocrine, metabolic, and digestive systems.

The liver, for example, has three zones with different oxygen levels that can affect drug metabolism, Wikswo says. No worries. D. Lansing Taylor, director of the University of Pittsburgh Drug Discovery Institute, has successfully built a microphysiological device with three liver chips in a row showing how drugs are metabolized in the different compartments.

Over the near term, drug toxicity testing is probably not where organ-on-a-chip technology will flourish, Wikswo adds. “Pharmaceutical companies have done a pretty good job of checking for toxicity, and you don’t want to come up with an assay you say is safe until you’ve proven that you haven’t missed something.”

The bar is “justifiably quite high,” says Wikswo. “What you don’t want is a false negative.” Companies need to know if a drug they’re making has the “desired efficacy and pharmacokinetics.”

Wikswo admits that his pumps and valves may appear complicated to biologists. His advice? “If you can use gravity and a rocking plate, or a pressurized reservoir, that’s a good way to go today.” But Wikswo’s focus is on overcoming the shortcomings of these technologies in recreating important biological processes that involve recirculation, pharmacokinetics, circadian rhythms and coupled organs.

“Our ultimate goal is to make our pumps and valves so simple, reliable and inexpensive that people love to use them,” Wikswo says. “We’re not there yet, but we are moving closer every day.”