Pioneering work to create organs-on-chips could revolutionise the future of drug development
Aug 07, 2014 • News Articles

Lawrence, J. The Pharmaceutical Journal. Aug. 7, 2014

Four years ago scientists observed how a white blood cell reacts when it senses an infection. They watched the leukocyte as it wriggled through capillary cells — the cells that line blood vessels — and then through the cells that line the lung, to then engulf an invading bacterium. But this was not happening inside a patient, it was happening on a microchip[1]. “It mimics the human response, it’s amazing to watch,” says Geraldine Hamilton, senior staff scientist at the Wyss Institute for Biologically Inspired Engineering.

Researchers at the Wyss Institute, which is based at Harvard University in Boston, Massachusetts, are pioneering the development of a whole pipeline of human ‘organs-on-chips’, including the lung, gut, heart, liver, skin, bone marrow, pancreas, kidney, eye and even a system that mimics the blood-brain barrier. The idea is to recreate the smallest functional unit of any particular organ in a micro-environment that closely imitates the human body, explains Hamilton.

To the naked eye, these ‘organs’ look nothing like the human body. A clear, flexible polymer is used to form the rectangular chips, which are about the size of a memory stick. Tiny channels pass through the chips like miniature ribbons, lined with living human cells, which vary depending on the organ being modelled. Other elements of the chip can also be customised.

So, for example, to create a lung-on-chip, the channels are split horizontally down the middle with a porous membrane. On top of the membrane, human lung cells grow and below it grow capillary cells. Air is passed over the top of the lung tissue and blood cells flow underneath the layer of capillary cells — mimicking the interface between the alveoli, the pockets of lung into which air is inhaled, and the blood vessels that carry oxygen away, all within the same tiny channels.

The flexibility of the chips means the mechanical forces that cells experience inside the human body can also be recreated. In the case of the lung-on-chip, the motion of breathing is simulated by applying a vacuum to side channels on the left and right of the main channel, stretching and relaxing the lung tissue inside.

This is important. Hamilton explains that Donald Ingber, who leads the organ-on-chip work at Wyss, showed that mechanical forces are key drivers of how a cell functions[2].

“We are trying to create an environment where they can function like they would in the body,” says Hamilton.

Organs-on-chips can be used to determine whether a particular protein is a suitable target for drug development, to identify drug toxicity or even to assess efficacy, says Hamilton.

The technology is still in its infancy, but there is a need for it; at the moment the techniques used to discover and develop drugs — animal models and human cell lines — are too often failing to predict what will happen in humans. Of the drugs that make it to phase I trials, only about 10% successfully reach the market[3].

Lost in translation

In 1938, testing drugs on animals became routine, after a drugs scandal prompted the US Food and Drug Administration (FDA) to clamp down on licensing. But researchers say that they have pushed animal models as far as they can.

“Human disease is not faithfully recreated in animals,” says Alan Wells, vice chair of pathology at the University of Pittsburgh, Pennsylvania. “Animals such as the rodent are only 30–70% predictive of human toxicity,” adds Lans Taylor, director of the drug discovery unit at the University of Pittsburgh (both men are working on developing a micro-liver using human cells). “We know how to cure cancer in a mouse, it’s humans that are the problem — the challenge is whether we can build something as good as an animal or better,” says Taylor.

Researchers began to culture human cells in vitro for drug development in the early 1970s. By the late 1980s, the technique went mainstream when the US National Cancer Institute replaced its mouse model with a human cell line.

However, human cells grown in a two-dimensional environment — such as on the classical Petri dish — do not respond in the same way as cells in the body, says Hamilton. Greater understanding of cell biology has highlighted the importance of the dynamic environments in which cells function in the body, where they grow and interact with their surroundings in all three dimensions. Researchers began to try and replicate that, developing cells in ‘three-dimensional cultures’; however, these also have their drawbacks. “With these techniques you can’t see what is happening,” says Hamilton. “Our cells are in a 3D environment too, but engineered in a way that allows us to get cellular and molecular visualisation.”

How to build a human

Hamilton believes that organs-on-chips could revolutionise the pharmaceutical industry by making drug development faster and cheaper, and by producing more successful therapies. Different disease states can be modelled on the chips and, crucially, drugs can be added to the channels and their effects on the tissue examined.

The team at Wyss have already tested this out. They delivered interleukin-2, which is used to treat certain types of cancer, to the ‘blood vessel space’ of their lung on a chip. As in humans, they found that it caused fluid to leak into the ‘lung space’, mimicking pulmonary oedema, a common and life-threatening side effect of the treatment[4]. But researchers were able to prevent this fluid leakage by using an experimental drug, currently being developed for this purpose by GlaxoSmithKline (GSK), headquartered in London.

“Very exciting science is happening at the Wyss Institute,” says Anthony Holmes, programme manager for technology development at the NC3Rs — the National Centre for the Replacement, Refinement and Reduction of Animals in Research in London. The NC3Rs was established as an independent scientific organisation by the UK government in 2004 and, in 2012, awarded the 3Rs prize to the Wyss Institute’s Ingber for his work on the lung-on-chip model. However, Holmes warns that the organ-on-chip research is still in its early stages.

The eventual goal? To link multiple organs together to create a ‘human-on-chip’. That original work in 2010, observing lymphocytes respond to the presence of bacteria in the lung-on-chip model, was a proof of concept that spurred on the US Defence Advanced Research Projects Agency (DARPA) in Arlington, Virginia, to award a grant to Ingber and his team of US$37m to do just that (additional funding has come from the FDA). “The DARPA grant requires us to connect a system of 10 organs, but we’re doing 12,” says Hamilton.

Another team, based at the Massachusetts Institute of Technology (MIT) in Cambridge, has also been awarded a hefty chunk of funding to develop a human microphysiological system: $26.3m from DARPA and a further $6.25m from the National Institute of Health (NIH) in Bethesda, Maryland. In total, the NIH has contributed $75m to researchers across universities in the United States to do their own work on developing micro-organs, in parallel to the organ-on-chip research.

A start-up company has now been formed to commercialise organ-on-chip technology developed at Wyss. Emulate launched in July 2014 with US$12m in Series A funding, led by the venture capital firm NanoDimension, alongside investment by Cedars-Sinai Medical Centre and private investor Hansjörg Wyss, the founder of Wyss Institute.

One of the team’s key goals is to get the technology out into the real world. “If it’s going to be used by drug companies it needs to be simple and reproducible; this is key for drug testing,” says Hamilton.

To address these problems, they are exploiting computer manufacturing techniques. The technology giant Sony has been brought on board to help manufacture the chips on a larger scale. Already, the researchers can ‘plug and play’ the chips, by inserting them like a CD into a prototype instrument that can link different organs together or multiple organs of the same type. “It has a really simple user interface so you don’t need to worry about the complex technology behind it,” says Hamilton.

Eventually, the system will allow researchers to, for example, analyse how a drug for asthma is inhaled in the lungs, goes into circulation, is metabolised by the liver, and then how the metabolites have an effect on heart rate and are excreted by the kidney, she says.

At the moment, only the lung and liver chips have been successfully connected. And although measurements such as liver enzyme and albumin levels can be taken, Hamilton says that, “sometimes cell-based data is not easy to translate into data that is relevant in the clinic”. Nonetheless, the Wyss team has already tested some well-characterised compounds in the lung and is starting to work with drug companies to test more drugs in more organs.

In-vitro takeover

Holmes, from the NC3Rs, thinks that a human-on-chip could be useful as a screening tool and to make decisions about whether to advance a drug through development towards clinical trials. But he is sceptical that any single in-vitro model will have the ability to completely replace animals in drug development. “We will need a whole suite of approaches,” he says.

The FDA’s position is that it is supportive of micro-organs but believes more work needs to be done. The FDA argues that tests would need to be developed to the point where they can provide answers to the various questions posed by regulators. For example, is the drug safe for use during pregnancy? “Some day in-vitro models may generally replace animals for regulatory purposes, but complete replacement is unlikely in the next few years,” a spokesperson said.

Holmes says that international regulatory agencies, including the FDA and the European Medicines Agency (EMA), are open to replacing animal models with in vitro models “but they are conservative, as their primary objective is to protect human health”. The key is to involve these agencies from the start so that they can feed into discussions about alternative testing strategies, he says. “To achieve this in practice, we try and include regulators such as the Medicines and Healthcare Product Regulatory Agency (MHRA), the EMA, the FDA and [agencies] from Asia on our working groups wherever possible.” 

But the regulatory agencies are not the ones doing the testing — the needs of the pharmaceutical industry also need to be addressed. The NC3Rs has created CRACK-IT, which links up academia, the small and medium enterprise (SME) sector and industry. The idea is that the pharmaceutical industry sets challenges aimed at replacing, refining and reducing animal use in research (the 3Rs) and, if academia or SMEs can come up with solutions, the NC3Rs will fund the research, says Holmes. “We also act as an honest broker, facilitating the exchange of preclinical information between drug companies to generate an evidence base to support change.”

Holmes believes that the NC3Rs’ work is appealing to the pharmaceutical industry because companies want to know whether a drug will fail as early in the development process as possible. “Drug development costs are spiralling and productivity has dropped — once a drug makes it to clinical trials it’s a very expensive time for it to fail. It’s about making sure that the models used early on are as predictive of humans as possible: fail fast, fail cheap,” he says.

For example, human lymph node models could predict immune responses, although at the moment these models need much more development, says Holmes. But the pay-offs would be high, potentially avoiding situations such as the one that occurred in 2006, with the TGN1412 monoclonal antibody. Having successfully made it through animal testing, TGN1412 was trialled for the first time in human volunteers and it caused a number of them to have a catastrophic immune response.

Join the dots

The question is whether these new human micro-organs could predict events like this better than the current methods.

Around 45% of cases of a drug being withdrawn from the market are because the drug is toxic to the liver or heart, says Gordana Vunjak-Novakovic, director of the laboratory for stem cells and tissue engineering at Columbia University in New York. She is one of the NIH-funded researchers working to create micro-organs to be used in drug discovery. She is developing a microphysiological heart-liver-vascular ‘organoid’ — the model has been nicknamed HeLiVa[5].

Vunjak-Novakovic highlights the importance of interactions between the liver and the heart. “For example, often when a medicine is metabolised by the liver before entering the heart you see more toxicity because the metabolic products can be more toxic than the original drug,” she says.

The approach taken by the team that created HeLiVa is to grow each organoid in a chamber connected by channels. So far, 12 drugs have been tested in the device to discover whether the response mimics that in humans. Drug response is measured using IC50 and EC50 values – the concentration of a drug required for 50% inhibition or stimulation, respectively.

“We can do this by plotting the beat frequency of our heart organoid as we increase the dose of proarrhythmic or antiarrhythmic drugs, for example,” she says. So far, the IC50/EC50 values are within the human range, says Vunjak-Novakovic. Her team has just been awarded another three years of funding by the NIH, after the initial two-year phase was deemed a success.

Spreading the knowledge

The NIH has granted extended funding to 12 teams working on microphysiological systems. One of these teams is led by Linda Griffith at MIT, who is creating a model of how cancer metastasises to the liver (although the model should provide wider clues to the process of cancer spread). Collaborating on the project is Wells at the University of Pittsburgh. “You always hear about the cost of a drug, but most of this is on failures,” he says. ”In cancer, only 4.7% of drugs work in humans once they get to clinical trials — all of these worked spectacularly in animals.”

Wells explains that cancer in animals seems to be different to cancer in humans. “It is more aggressive in animals,” he says. “We’re good at dealing with primary tumours in humans, but the problem is once cancer metastasises we can push it back, but we almost never cure it.” He explains that, at the metastasis site, the tumour will change back to a less aggressive form, making it harder to distinguish from healthy cells and therefore harder to kill. “Or, it can be dormant and come back 15 years later. We need a new approach because animal models don’t go into dormancy.”

About six years ago the team started experimenting with putting tumour cells in its liver micro-organ model. The liver is one of the main sites of solid tumour metastasis, explains Wells. Now, using cells from patients, the team can create tiny liver organoids less than a millimetre thick and seed them with a small number of cancer cells. “Some will grow out and proliferate and some will become dormant. The NIH grant will fund research to find molecules that control that process — will some drugs wake them from dormancy while others keep them dormant or kill them?”

The researchers will next look at linking their liver model to other organs; collaborations have already been established. “Liver function is affected by sex hormones so we are working with a group that is creating a mini ovary that we can connect to,” Wells says.

But linking the organs together presents a whole new set of challenges. “If you want the organs to talk to each other accurately you need the right amount of fluid,” says John Wikswo, a biomedical engineer at Vanderbilt University in Nashville, Tennessee. He is working on how to scale micro-organs, both in terms of their size relative to one another and how to make them in large numbers.

“These organoids are tiny,” says Wikswo. “The problem is once you get too small the fluid you can take to analyse is not enough to do tests.”

Taylor, who is working on a liver model with Wells at the University of Pittsburgh, admits that Wikswo’s work to scale the micro-organs presents one of the biggest challenges. Despite this, he believes that there will be a first-generation system in place within five years and that in a decade the system may start to convince the FDA. “But a human-on-chip is further down the line than that.”

Taylor says that there is a lot of excitement about the new technology. “It could be over-enthusiasm, but it’s coming and it’s coming fast — it will be the same as the acceleration of the human genome project and I think we could have the same level of success,” he predicts. 

Making it personal

But the ultimate dream for micro-organ researchers is to be able to test drugs on different genetic backgrounds, argues Taylor, explaining that a patient’s genetic profile can make the difference between a drug producing toxic effects or effecting a cure.

The advent of induced pluripotent stem cell (iPSC) technology — in which skin cells, for example, can be taken from a patient and turned into any cell type in the body — presents exciting opportunities for researchers working with micro-organs. Vunjak-Novakovic at Columbia already uses iPSCs in her HeLiVa organoid and a team at Harvard, in collaboration with the Wyss Institute, has experimented with a model of heart-disease-on-chip using iPSCs from a patient with a genetic heart condition.

“iPS cells are super interesting because you can take cells from a diseased population, or the elderly, children, or those with a different transporter protein that affects drug uptake. You can help identify at-risk populations,” says Hamilton.

Thinking about personalised medicine keeps the team at Wyss inspired, she says, adding: “I can see the technology developing into clinical-trials-on-chips or individual-patients-on-chips. We could put you on a chip.”

Read original article here...


[1]Huh D, Matthews BD, Mammoto A, Montoya-Zavala M, Hsin HY, Ingber DE. Reconstituting organ-level lung functions on a chip. Science 2010;328(5986): 1662–1668.

[2]Ingber DE & Jamieson JD (1985). Cells as tensegrity structures: Architectural regulation of histodifferentiation by physical forces transduced over basement membrane. In: Andersson L.C., Gahmberg C.G., Ekblom P., eds. Gene Expression During Normal and Malignant Differentiation. Orlando: Academic Press, 13–32.


[4]Huh D, Leslie DC, Matthews BD, Fraser JP, Jurek S, Hamilton GA, Thorneloe KS, McAlexander MA, Ingber DE. A human disease model of drug toxicity-induced pulmonary edema in a lung-on-a-chip microdevice. Sci Transl Med 2012;4(159): 159ra147

[5]Vunjak-Novakovic G, Bhatia S, Chen C, Hirschi K. HeLiVa platform: integrated heart-liver-vascular systems for drug testing in human health and disease. Stem Cell Research & Therapy 2013; 4:S8.