Professor Donald Ingber, Founding Director of the Wyss Institute for Biologically Inspired Engineering at Harvard University

Date: 23rd July 2021

Professor Donald Ingber was interviewed by Nomthandazo Ziba who is completing a master's degree in pharmacology at Coventry University. This transcript has been edited for clarity and brevity.

Photograph of Professor Professor Donald Ingber, credit: Wyss Institute.

Born in East Meadow New York in 1956, Ingber did an undergraduate and master’s degree in molecular biophysics and biochemistry and then did a combined MD and PhD at Yale University. A renowned cell biologist and bioengineer, in 2009 Ingber founded and became the director of the Wyss Institute for Biologically Inspired Engineering, a partnership between Harvard University and its affiliated hospitals. Launched with a $125 million gift from the Swiss billionaire Hansjörg Wyss, the aim of the Wyss Institute is to support high-risk research and leverage new engineering techniques for medicine, including the development of new medical devices. Professor Ingber is also Judah Folkman Professor of Vascular Biology, Harvard Medical School & Boston Children’s Hospital; and Professor of Bioengineering, Harvard John A. Paulson School of Engineering and Applied Sciences.

Nomthandazo

Thank you very much for taking the time to talk to me. To start off I just want to get a better understanding of how you got into the organ-on-a-chip area.

Donald

It's a long story. So briefly, I've been interested in the role of mechanical forces in biology for 40 years. In the early 1980s, I presented the idea that mechanical forces are as important as chemicals and genes for both development and health and disease, which now forms the core of the field we call Mechanobiology. The path to Organ Chips started with my exploring development of different model systems to control cell shape - how far the cell stretches – to demonstrate the importance of mechanical distortion. In around 1990, we're talking 30 years ago, we developed the best system we came up with by adopting computer microchip manufacturing techniques to micro engineer cell culture environments, because it gave you control over features at the nanometre to micrometre scale that living cells and tissues live at. With George Whitesides at Harvard, we developed a version of what is called soft lithography or microcontact printing, where we could print little adhesive islands on the scale of individual cells coated with extracellular matrix molecules, which are what cells adhere to. So we had a series of Science papers where we showed that we could control cell functions, including growth, differentiation, apoptosis and motility, by modulating cell distortion. Then a couple years later, George’s team started to develop microfluidic systems by adapting the same soft lithographic manufacturing approach. This is where you create little hollow channels in which you can have controlled flow. This approach has been used widely to miniaturise instrumentation in industry for diagnostics and analytical systems. But I've worked on vascular biology for years and to me, these are like little engineered microvascular networks. The idea that birthed Organ Chips was basically to combine all this together to replicate organ level structures and functions. We started to culture cells in these microfluidic channels in the late 1990s and a postdoc who was co-mentored by George and me, Shuichi Takayama, published a Nature paper in the early 2000s in which we cultured living cells in microfluidic devices (Takayama Ostuni LeDuc). This was not yet an Organ Chip; it was simply culturing cells within a microfluidic channel. However, a few years later after Shuichi moved to set up his own independent lab at U. Michigan, I heard him present on what he called a lung-on-a-chip in a small meeting. At that time, it did not have cells in it. It just had little channels shaped like the small airways of the lung. I'll never forget when he showed how putting a little droplet of fluid through these channels, like a mucus plug in your lung, and the chip made a noise. That noise was exactly the noise I was taught to listen for through a stethoscope when I went to medical school, it was called a ‘crackle’. I was amazed. I thought that was the coolest thing because when I was in medical school, all of the students would ask our professors what caused that sound, and they’d say ‘I don't know; it’s fluid in the lungs’. What he basically showed was this sound was due to little mucus plugs going through bronchioles, small airway size conduits. A year or two later, the graduate student in Shu’s lab who did that work, Dan Huh, applied to my lab for a postdoc, and when he interviewed, I shared how much I had been impressed with his past work, but then I challenged him by saying that if he joined my lab, why don't we create a real living human lung-on-a-chip. My first idea was to leverage the fact that if you have two inlets in a single channel microfluidic device and flow two different liquids through these inlets, there's no mixing when the liquids merge into one channel; it's only laminar flow because there’s new turbulence due to the small width of the channel and so the liquids they just flow by each other. George had done work precipitating metals at the interface, using two different chemical solutions. I thought maybe we could precipitate an extracellular matrix, like a basement membrane, and we could culture lung epithelial lining cells on one side of this matrix, capillary cells on the other, and we'd have a living human alveolar-capillary interface. That was the original idea I suggested to Dan. Now Dan and other members of my team pursued that, and we made some headway with this precipitation, but it was not robust. So then Dan came up with integrating a flexible polymeric membrane with pores and coating it matrix, and that was the real beginning of organs-on-chips. The other thing was that I believed that mechanical forces are critically important for cell, tissue, and organ function, as I mentioned earlier. That made me want to integrate breathing motions in this Lung Chip, and Dan was able to accomplish this by integrating side chambers to which we applied cyclic suction, and because the entire device was made of flexible material, this caused the engineered alveolar-capillary interface to stretch and relax to the same degree and at the same rate as occurs in the breathing lung. It is essentially a living three dimensional cross section through the lung alveolus, which includes the lung lining cells maintained under an air-liquid interface, the basement membrane above, and then a capillary endothelium below with flowing nutrient medium through it, or even blood for short times. That's basically how Organ Chips came about. I called these Organ Chips, because in organs two or more tissues come together and interface so that new functions emerge, and that is what we accomplished in these chips. This is different than what we and others had done previously, which was basically culturing cells or tissues in chips. However, in recent years, the term organ-on-a-chip or tissue-on-a-chip has come to be used by many to describe any microfluidic culture system.

Nomthandazo

Do you mean the ones that are tissue-on-a chip are just two dimensional?

Donald

What I mean is that a tissue-on-a-chip is one cell type that forms a tissue-like structure in a microfluidic culture system. An Organ Chip has at least two channels so that you have at least two or more tissues interacting. Usually you want a vascular channel, a blood vessel channel because that's how organs are normally fed. But it’s an Organ Chip if it is a microfluidic culture device that has multiple tissue types and it replicates organ level structures and functions.

Nomthandazo

When did you come up with the term organ-on-a-chip?

Donald

That’s interesting. I first used the term in a couple of my grant applications on an unrelated project years before the Lung Chip came about. I was working on a microfluidic device that was cleansing blood by removing pathogens as a sepsis treatment device and I called that a spleen-on-a-chip because it was functioning like a spleen, but there were no living cells in it. By the way, a modified version of that technology is now in clinical trials. But the Lung Chip paper in Science in 2010 was really the first time I think people saw what an organ-on-a-chip really could be. That's when we first really published and used that term.

Nomthandazo

When you speak about the spleen-on-a-chip project being now in clinical trials, is the work being done by a specific Institute or is it a big pharma company?

Donald

That's a start-up that we formed called Boa Biomedical Inc., but the technology eventually transformed into one that is not microfluidic based; it leverages a dialysis-like device setup. That's in clinical trials now for cleansing the blood of pathogens.

Nomthandazo

What is Boa?

Donald

It is a company I founded. So I head the Wyss Institute for Biologically Inspired Engineering at Harvard. We're academic and we're part of Harvard, but we are largely a Translation Institute, which has been very successful. We now enable almost 25% of all of Harvard's intellectual property and start-ups each year. I've actually done multiple start-ups. BOA is developing sepsis therapeutic devices and diagnostics. Emulate Inc. is another company I founded which is commercialising the Organ Chip technology. I have other ones, but they're not relevant to this discussion.

Nomthandazo

Have you ever had the opportunity to collaborate with pharmaceutical companies?

Donald

Yes. I've collaborated with big pharmaceutical companies for over 35 years. In terms of Organ Chip collaborations, we've published with Merck, AstraZeneca, and Janssen. I'm also working with Fulcrum and with GlaxoSmithKline. So yes, we collaborate with big pharma. I'm very proud that I've had two papers where two competing big pharmas were joint authors on the same papers with us; this is because they are excited about the technology.

Nomthandazo

How is this technology going to be beneficial to patients?

Donald

When it comes to drug development, the goal is to shorten the time in research, decrease the cost, and increase the likelihood of success because animal models are wrong 75 to 95% of the time. Plus, it can help with personalised medicine, where you could potentially make your own chips for individual patients to look for the best and least toxic drug. And also for rare genetic disorders, where you could collect cells from patients around the world and essentially do testing on chips to find the right drug, and then maybe be able to distribute that drug to do real clinical trials at a distance. With recent advances in human patient-derived organoids, and stem cell technologies, like iPS, you can imagine finding a genetic subgroup or rare genetic disorder, getting 50 or 100 patients’ cells, make the chips, develop a drug for them, and then use those patients for a focused clinical trial. That would increase the likelihood of success, decrease costs and shorten time. Organ Chips also could help with designing clinical trials, for example shortening the time, because we've been able to show that it is possible to predict drug pharmacokinetics, how drug levels go up and down in humans. We did that by linking multiple organ chips to create a sort of a human body on chips and with computational modelling. With that approach, you might be able to determine what dose administration regimen to start a clinical trial with, rather than starting really low and going higher and higher and higher, which would shorten times. Also it could save animal lives.

Nomthandazo

To what extent do you think this technology is going to cut out the use of animals in clinical studies?

Donald

I hope it's going to reduce them progressively more and more over time. We probably will always have some need for animal studies, but I'm hoping that different animal models will be dropped one by one. There are a couple of things that I was not aware of when we started, which we discovered over time. For example, something like 40% of all the drugs that are currently in the development pipeline are biologics, such as therapeutic monoclonal antibodies or RNA- or CRISPR-based therapeutics. Their targets are human specific, and so they often don't exist in animal models, and don’t even cross react in non-human primates sometimes. So there's no way to test those other than something like a human organ-on-a-chip. That's another place where the technology can help people. Finally, there's an emerging area called ‘One Health’, which involves many diseases that start in animal populations. SARS and COVID-19 are great examples. They think SARS started in bats, Ebola in primates and monkeys. You can't have every animal species in a lab to study this. But you could make chips. We've actually already made chips from different species. We've made liver chips from dogs, rats, and humans, because pharmaceutical companies have to do preclinical, liver toxicity testing in rats and dogs. And we were able to show that we could replicate species specific differences in liver toxicity, which is important because pharma companies often get conflicting results in rats and dogs as well as human cultured cells, and therefore often don't know what to do in humans, and they often fail in the clinic. So that's another side where I think we all can benefit that I never thought of in the beginning.

Nomthandazo

And it's quite expensive to use the animals to get drugs into clinical trials, which would also cut down the cost.

Donald

Yes, hopefully. The other side is that we have found that sometimes we can do some sorts of clinical studies better than with patients, especially where you're trying to find cause and effect. For example, we could do transcriptomics analysis to measure gene expression changes in response to lung chips being exposed to cigarette smoke. And we could do that in lung chips from different patients. We can see which genes actually are triggered by cigarette smoke exposure. There have been clinical studies where they've done this. But what they do is to look at people who smoke cigarettes that are otherwise healthy, they have different smoking histories, they have different working histories, they have different family home environments. So I think for challenges like that we can actually get much more directly to cause and effect, because we are carrying out match comparative modelling. So there are things that you can do with these chips, because they're human, that we never even thought of when we started.

Nomthandazo

How affordable is the technology going to be?

Donald

Right now, it's expensive, but every new technology is expensive in the beginning. I remember when gene microarrays came out they were so expensive. They had to have hospital core facilities and you could do one, maybe, and that was it, but now they're reasonably priced. So I believe it will decrease in cost over time. Relative to animal models it'll be cheaper. But these are not replacements for 96 well plates. These are replacements for animal studies, like primate studies. Non-human primates are incredibly expensive and with organ chips it will be much less expensive. I should mention that Emulate has sold their chips and instruments to something like 19 out of the top 25 biopharma and biotech companies and FDA has them on site too. So this is no longer in the future, this is beginning of a transformation.

Nomthandazo

Okay, that's good. What are the key challenges so far with the technology trying to move it forward and where do you think the solutions might lie?

Donald

In the lab, the biggest problem was bubbles, literally bubbles. Every time you change a fluid you have a chance of getting a bubble, and that would kill the cells. Emulate and other companies are needed to automate more and more of the process, but there's a long way to go. Right now, you can automate the culture, but you have to do the plating by hand and then you take them out to a microscope to image them. Some of the challenges are automating all of the steps, so that it's really plug and play. It also would be nice to have inline sensors, and we have done this in my lab for oxygen and barrier function. Basically we need to automate more and more inline analysis. But the key thing is robustness and reproducibility. Another big challenge is cell sourcing. The downside with human cells is patient to patient variability. The upside is that variability is similar to what you're going to see in clinical trials. So one has to figure out how to balance those two. This involves finding donor-derived cells that produce reproducible results that you can use in your initial work. But then when you're thinking about doing your later stage work, you might want to include cells from different genetic subpopulations, men versus women, different ethnic groups, etc. It's a different way of thinking about doing in vitro studies as now people often just use an established cell line because they know they'll get the same results each time. But the results often don't look anything like what you would see in human patients.

Nomthandazo

You mentioned that there is a need for reproducibility and robustness. I was reading somewhere that it's important that organs on-a-chip are standardised. Do you think that's going to be possible since there are a lot of companies that are already around and all doing it differently?

Donald

I don’t think standardized is the right goal; the need is for Organ Chips that are validated to replicate human organ functions and to be reproducible. Emulate has done that with liver, colon and kidney. They do that by finding a good cell source, the right medium, the right chips, the right matrix coating, and basically selling a package to people so they know they've got all that. But that takes a lot of work. Most of the work on chips now is still you get an empty chip, you put your cells in it, your medium, and you work that out. Of course, that is an advantage for people who, for example, have their own patient derived organoids and they want to work with their populations. But it depends what you're talking about. If you're talking about it for drug development programmes, for basic research, then you want people to be able to adapt it for their needs. If you're talking about clinical trials, you have the challenge that you may want to represent different ethnic groups or be able to define one group that is most responsive on chips, just like you would do clinically and then do a targeted trial. So it can't be generalised to everybody. There are different applications that have different needs.

Nomthandazo

And I know that clinical trials and the drug regulatory process are highly regulated. So what legislation would you change to improve this technology?

Donald

It's funny you asked that because my last email was about something similar. In the United States they're just going to Congress with what is called the FDA Modernization Act, which I was asked to speak to a Congressional subcommittee about when they were trying to get interest. Right now the FDA requires preclinical testing in animals and they're trying to change a few words to change the verbiage to broaden it to include preclinical testing using human relevant models. Then you could use these Organ Chip systems instead of animal studies, and that would be a huge, huge plus.

Nomthandazo

It's called human relevant models?

Donald

I forget the exact terminology they use, but they're making it more general to include human preclinical models rather specifically require testing in animals.

Nomthandazo

What has been your favourite aspect so far in your research?

Donald

There's two levels to that. In terms of the most novel aspect, I think, is our ability to study complex human microbiome interactions with human living cells in an organ relevant context, which is not really possible any other way. You know, the microbiome has been a major paradigm shift in medicine over the last 20 years. But almost everything we know is based on genomic or meta genomic analysis. Because if you put hundreds of different microbes on your human cell culture we call it contamination and the cells die. But because we have flow and we recreate the microenvironment, we have been able to culture hundreds of different bacteria from stool specimens on our human intestine chips and we've done multiple bugs in vagina and cervix chips, and we've cultured clinically relevant microbes in cystic fibrosis lung chips as well. We can do this for many days. One of the most gratifying, surprising, and important advances on a higher level has been our ability to develop over 15 different chips here at the Wyss Institute. People ask me have you had a chip that you couldn't do? And the answer is no, we have not yet had a chip that we couldn't do. There were some that we couldn't get to work for a few years because we couldn't get the cells and then we developed a way to obtain the cells using iPS technology, and then it worked. What's been gratifying is the ability to replicate not only normal physiology, but pathophysiology and disease states with high fidelity. Then the other is the critical importance of mechanical forces, flow peristaltic-like motions in intestine, breathing motions in lung, one distortion per second like with each heartbeat in the kidney, which is really validating all the work I've done in my whole life, confirming that mechanical forces really are important.

Nomthandazo

I think the blood brain barrier was the most interesting one for me.

Donald

That is interesting for us, because we're now leveraging that technology to develop shuttles that will take drugs across the blood brain barrier. I mean they're all really interesting. It's amazing, most people in science like to work in one area. I've always worked across many scientific fields and now I work across many clinical fields. I speak at the kidney meeting, the liver meeting, the lung meeting. I speak at all of those because everyone is interested in these things. One of the newer ones we have that's not really published is a lymphoid follicle chip that replicates vaccination responses. I think that's going to be a really important one. And we're beginning to link it to other organs to get mucosal immunity. So I think that Immune Chips are going to be of great value for the field.

Nomthandazo

Do you think that one of these could perhaps be used in any way in assisting with the COVID-19 pandemic?

Donald

Well, for example, the Lymphoid Follicle Chip could potentially replace non-human primates for vaccine development, which has been really limiting for the field. We also have leveraged our human Lung Chip models of viral infection to identify an existing drug that is currently in clinical trials across many sites in Africa.

Nomthandazo

You mentioned that you've got more than 20 organ chips now. So how many are you able to put together on one chip? Do you call it ‘human-on-a-chip’?

Donald

I think scientists across the Wyss Institute have developed about 20 and my group has about 15. So there are many of us doing this. We've linked multiple Organ Chips, but they’re not on one chip, we link multiple chips together. We've linked up to 10 chips to create what we call a ‘human body on chips’. We published that, I think with eight chips, but we've done 10. But to predict drug pharmacokinetics, we simplified the system because it's extremely hard to do 10. Each one of these has to be cultured for days to weeks to differentiate them. Then you have to stagger them so that they're at the right level of differentiation to join them together. There's a huge amount of analysis, such as quantification of cytokine levels and secretion of particular factors, which needs to be carried out on samples from every chip, and then when they are connected, we have to carry out mass spec to measure drug levels at each point in the connected system. And it's not yet automated so this all requires an army of people. When we did the drug for pharmacokinetic prediction, we just made what is called a ‘first pass model’ in the drug development field, which includes liver, intestine, and kidney. And we also had a mixing chamber that would mimic the heart pumping the blood so you can take a sample independent of the effect of any single Organ Chip, like you were sampling from peripheral blood.

Nomthandazo

Is Emulate only based in America? Are you guys planning to be established in other countries?

Donald

Yeah, it's all over the world now. I mean, their head executive VP for research currently lives in Cambridge, England.

Nomthandazo

What about the collaboration with Queen Mary University of London?

Donald

That is an interesting model where Queen Mary has set up a site with multiple Emulate instruments that has provided a way to get a larger community familiar with the technology. But that was just one site where they started. They are now selling all over Europe and I think it's going to begin in Asia as well in Japan. They are definitely selling in Europe. I know the Pasteur Institute has them. I know they have them in Switzerland, and in multiple sites in England. Public Health England has them in a BSL3 lab.

Nomthandazo

Is there anything else you'd like to share that I didn't cover?

Donald

I just looked up the FDA Modernization Act. They're changing it from animals to ‘non-clinical’ tests so that could be a study which can involve organ chips, microphysiological systems, or other cell based human relevant test.

Nomthandazo

All right. Thank you very much. I really appreciate all the work that you're doing. Thank you so much.

Reference

Takayama, S, Ostuni, E, LeDuc, P, et al (28 June 2001) 'Subcellular positioning of small molecules', Nature, 111, 1016. Back

Respond to or comment on this page on our feeds on Facebook, Instagram or Twitter.