Additional content from James Dahlman on Gene Therapy and the incredible work being done at Georgia Tech.
[analog radio tuner scanning stations]
[big band swing rendition of Ramblin' Wreck from Georgia Tech]
[interposed voices of Steve McLaughlin] ...sounds incredibly complex...it sounds like...to have abilities that span...I'm really geeking out here.
[applause and laugh track]
[big band swing rendition of Ramblin' Wreck from Georgia Tech]
[boiling, bubbling liquid]
Steve McLaughlin: I'm really curious about why an engineer? We think of these therapies as being the domain of doctors and hospitals and patients. I'm an electrical engineer. You know, I work on cell phones and computer chips. And we think of engineers as bridge builders and things like that. Can you say a little bit about why engineering is so important to some of these new and novel therapies
James Dahlman: A lot of the most important problems, or hurdles I should say, that are facing gene therapies are actually engineering hurdles. To give one easy-to-understand example, we have a professor at Georgia Tech named Krish Roy who's developing what we call CAR-T therapies. These are therapies that use T cells to treat disease. So you take your own T cells out of your bloodstream, you change their genetics inside a dish, and you put him back in and train them to go after a tumor or something like this. They've actually been pretty promising, but there's one simple problem with them: It's hard to make a lot of them. And so you can run these small clinical trials right now, but if you want to scale up and treat thousands of patients, people, even the biggest companies in the world don't know how to make enough T cell therapies to treat thousands of people. That is, at its most core, an engineering problem, and that's what Professor Roy is doing.
Steve McLaughlin: I know that you have students in your lab, both undergraduate and graduate students. Can you say a little bit more about what your students do and the kind of interests and impact that they're having?
James Dahlman: Gladly. I will talk about my students all day. The first thing I will say is that we are very lucky here at Georgia Tech to have really excellent students at the undergraduate level and the graduate level. A special shout out to Cory Sago, Collina, and Melissa. Our students are excellent, and they do perform a lot of different types of experiments. So if you're a student in my lab, you learn a lot of different techniques. This includes designing the nanoparticles, which is a chemistry-based technique. You know, you may want to make a nanoparticle that is 50 nanometers, and then you want to make a nanoparticle that's 100 nanometers—that involves different chemistry.
Once you make the nanoparticle, you need to package the drug inside of it. This involves what we call “microfluidics” which is another engineering technique. You then want to administer the drugs to cells to see if they work. And that involves some biomedical engineering as well. And then on the back end, we actually do this analysis called “deep sequencing.” It's the same thing that you do when you send your, you know, when you spit in a tube and send it to Ancestry.com or 23andMe. And that involves a bunch of molecular biology. And so our biomedical engineers here at Georgia Tech learn chemistry, chemical engineering, biology, biomedical engineering, and molecular biology all at once in the lab.
Steve McLaughlin: So you now have developed a technology that allows you to effectively test many thousands of potential therapies all at one time. So not only does that seem to be an incredible breakthrough in terms of drug delivery, but it also sounds like it has huge implications for all kinds of therapies, and maybe even starting a new business or or a new company that tries to commercialize it. Can you say a little bit about—you know, it spans so many dimensions, can you say a little bit more about the things that have you most excited?
James Dahlman: We're very excited right now. The way I kind of divide the work up is in one bucket we have this simple question which is let's find drug delivery vehicles, let’s find nanoparticles for non-liver tissues—that's the real clinical need right now. And we do this pretty simply; we test thousands of things and we try to learn from the experiments. You know, we'll do Experiment 1 with particles 1 through 250, and then we'll use bioinformatics to learn from those data, then use those data to inform the design of particles 251 to 500. And we just kind of iteratively improve the delivery and try to find things that target new tissues.
The second bucket is also interesting; it’s more scientific. The second bucket is, right now, we don't really understand why nanoparticles work or why they don't. And we don't understand the factors that affect whether a particle works or not. So we can perform really interesting biological studies to understand which genes actually alter how well the nanoparticle works itself.
Steve McLaughlin: So you have a nanoparticle that has inside of it, you know, a drug that you want to deliver to a cell, and now you've attached kind of a barcode to it. So first of all, what a fantastic engineering problem. But the engineer in me might say, “Well, what about if by putting that barcode on the nanoparticle, does that change or affect how effective that that nanoparticle can be?” What's the interaction between the barcode? Could that somehow inhibit or prevent the therapy from being there?
James Dahlman: I think in some cases the answer will be “Yes.” So if you asked, “Does this DNA barcode affect how well a nanoparticle will deliver aspirin?” The answer is probably “Yes” because aspirin is a very small molecule that is not charged. However, if you said, “Does this barcode affect the delivery of an RNA therapy?” The answer is almost definitely “No” because the barcode in the RNA look exactly the same. And so it really depends on the type of drug that you're trying to deliver. It will be interesting to try and design barcoding systems for small molecules like aspirin. I do think it is possible but, you know, it could take another two years of work before we get anything that's robust.
Steve McLaughlin: So it sounds like your experiments generate a huge amount of data, and you need to have the ability—you know, if you're going to run, I assume in the future you're going to have the ability to run not just hundreds or thousands, but even many more, that will result in incredibly rapid advances and development of new drugs and therapies. But you're going to be generating tons and tons of data. So it sounds like you need to know a little bit of computer science and a little bit of mathematics and your students are kind of learn all trades. So is that kind of where all this is headed?
James Dahlman: So that speaks to the interdisciplinary aspect of my team. So my team has—my students have backgrounds in biology, biomedical engineering, chemical engineering, chemistry. And we also have two bioinformatics students from the Georgia Tech master’s bioinformatics program here whose sole job is to develop the mathematical framework and the bioinformatics pipeline on the back end for that exact reason. So, one thing I'll say is, if you're a student out here listening and you're thinking, “Oh man, I don't know if I have what it takes to become a biomedical engineer. You know, I studied math as an undergraduate.” You can become a biomedical engineer. It's a very collaborative discipline and it's a discipline that needs people with different skill sets. And so some of the best biomedical engineers in my lab have no formal training in biomedical engineering.
I grew up in Dayton, Ohio, which is in southern Ohio near Kentucky. I was not born, but raised, in Ohio, and I stayed in Ohio all the way through my undergraduate where I got that at Wright State University, which is the local university. I studied biomedical engineering, but at that time I was actually doing materials science research at the Wright-Patterson Air Force Base, so I was basically an Air Force contractor for four years as an undergraduate. And this was really traditional material science; we were working on what are called “bulk metallic glasses” with the express purpose of making a corrosion-resistant material that was lightweight enough to be shot up into space. It's pretty far afield of what we're doing now.
Well, after I graduated from with my Ph.D.—I’d done my Ph.D. in a lab that focuses primarily on chemical engineering, so I had chemical engineering skill set. What interested me at that time wasn't chemical engineering. What interested me at that time was genomics. And so I jumped ship and joined a genomics lab that was called the Broad Institute, which is a dual organization set up between Harvard and MIT. And you can think of it as the human genome building, basically. it's where some of the world's best genomics scientists reside. I loved that work at the Broad, and I learned a lot about molecular biology and all these genomic skill sets. And I just had this gut feeling that I could I could use these genomic skill sets somehow to improve chemical engineering. And lo and behold, that's actually exactly what happened. All of the barcoding stuff that we've been talking about today that my lab does, all of that was learned at the Broad Institute.
My lab lives at the interface of chemical engineering and molecular biology and genomics. And that's because I think, luckily, I followed my gut and just did what I liked. And that happened to be chemical engineering in genomics. I've been very lucky to have excellent mentors. I did my Ph.D. with a guy named Robert Langer who's just a storied biological engineer, and I did my postdoc with Feng Zhang who is a young, rising superstar in the field of gene editing. I am not Robert Langer; I'm not Feng Zhang. However, I do my best to pass on their mentoring styles to my students because I really sincerely appreciate what they did for me.
One of the best parts about being in Georgia Tech is the quality of our students. I love mentoring these students. It's probably the best part of my day is interacting with my students in the lab—it is the best part of my day. The other parts of my day are spent grant writing, so I guess, you know, it definitely is the best part of my day.
Steve McLaughlin: I remember from high school biology, you know, the double helix of DNA, and you're saying CRISPR has the ability to go in, kind of identify parts of that DNA strand and literally edit it out to change it.
James Dahlman: So every cell in your body has three billion base pairs in it. Let's say the gene that's causing trouble is located at base pair 1003. How on earth are you going to find base pair 1003 from a string that’s three billion base pairs long? Well, CRISPR does that. You can go in, find base pair 1003, and perturb base pair 1003 very specifically. You may ask how on earth did this happen, and I will tell you that CRISPR evolved as a bacterial defense mechanism. So think about it this way: Bacteria and viruses have been battling it out since life began and there are viruses that infects bacteria; they're called bacteriophages—just another name for a virus that gobbles up bacteria. The virus attacks the bacteria, and the way it works is it injects its DNA into the bacteria, and this makes more virus. What CRISPR is a bacterial defense system to remember what viral DNA is—so to remember that it got attacked by this virus, and then to delete the viral DNA when it gets attacked again. All we did was we took that DNA recognition and DNA deleting system, and we repurposed it for human cells, and this is what CRISPR is. That's basically a bacterial defense system that we've hijacked and used to edit DNA now.
We use CRISPR in two ways in my lab. The first is we develop CRISPR-based drugs to go after cystic fibrosis or to go after a hemophilia or, you know, whatever. There are a lot of genetic diseases that would benefit from CRISPR therapies. And the second way that we use CRISPR is we use it to study biology. You know, it is now far easier than it's ever been to knock out this gene or knock out that gene and see what happens. So if I want to figure out if gene A or gene B is affecting how well my nanoparticle works, because I want to understand how my nanoparticle works, I can just CRISPR gene A and I can CRISPR gene B, and I can see if the nanoparticle works. So CRISPR, in addition to being therapeutic, potentially, is already accelerating the rate that we can make scientific discoveries in the biological space.
[Ramblin' Wreck from Georgia Tech marching band rendition]