Dr. Andre Ghetti, CEO of AnaBios, joins to discuss how his company helps de-risk drug development by using ethically sourced, functional human tissues to generate preclinical data. He shares how AnaBios bridges the translational gap between animal models and clinical trials through precision assays, AI-powered analytics, and validated human tissue platforms.
In this episode of Speaking of Mol Bio, Dr. Andre Ghetti, CEO of AnaBios, offers a deep dive into the world of translational preclinical research. AnaBios is redefining early human insights by using ethically sourced, functional human tissues and cells to generate actionable data before compounds ever enter clinical trials. Ghetti walks us through the company’s approach of offering human-relevant safety and efficacy data, validating drug targets, and supporting everything from small startups to major pharma groups.
We learn how AnaBios engages with clients to customize assays, especially in high-need areas like non-opioid pain therapies, fibrosis, and cardiac safety, and how they use a blend of standardized and novel functional assays, some of which required building their own hardware. He also discusses their integration of RT-PCR, RNA-seq, and calcium imaging, including genetically encoded sensors to monitor neuronal activity at scale.
From their use of machine learning to analyze massive data sets, to collaborations with the FDA, to their unique ability to preserve tissue viability across the U.S., AnaBios offers a powerful glimpse into the future of translational biology. Dr. Ghetti also shares advice for young scientists and reflects on what’s next for AnaBios, including oncology and stem-cell model integration.
Steve Lewis00:10
Welcome to Speaking of Mol Bio, a podcast series about molecular biology and its training applications in life sciences. I'm Steve Lewis. Today, we welcome Dr. Andre Ghetti to the show. Andre is the CEO of AnaBios, which focuses on drug development using ethically sourced functional human tissues and cells to generate preclinical data. We cover a wide range of topics, from translational research to the evolving landscape of biotechnology. It's a great conversation, and I hope you enjoy.
Andre Ghetti, PhD 00:48
I'm Andre Ghetti. I'm just Chief Executive Officer of AnaBios Corporation. I originally joined the founding team as Chief Scientific Officer. I do have a science background, and I've been with AnaBios pretty much since inception, for the last 15 years.
Steve Lewis01:05
That's wonderful. And how did you come to be one of the founders of AnaBios?
Andre Ghetti, PhD 01:10
Yeah, so I already knew the other two co-founders and we wanted to establish a new company that had very specific capabilities that would enable pharmaceutical companies to obtain early human data, even prior to clinical studies. And so we wanted to focus on what is generally known as translational research, and enable this kind of research at scale, introducing new capabilities that were not on the market at the time.
Steve Lewis01:44
And can you tell our audience a bit more what you mean by the preclinical stage?
Andre Ghetti, PhD 01:49
Yeah, so this is a stage of research in drug discovery in which molecules are typically tested in in vitro systems. Can be cell lines. Later, there can be extensive testing in animal models, different, using a variety of species. And this work is essentially aimed at generating a lot of data from which one could hope to infer what would be the effect in the clinic of this kind of novel compound. And so there is a lot of guessing and prediction that is based on, as I mentioned, cell lines and animal models. And unfortunately, oftentimes the data generated with this type of research are not very accurately predictive of the clinical responses. And see, this was the market need that we identified, and that constitutes the niche for AnaBios’ business.
Steve Lewis02:44
In a way, kind of de-risking drug development for customers?
Andre Ghetti, PhD 02:49
That's right. So clinical trials can be extremely time-consuming and expensive, and so prior to introducing a molecule into those types of studies, one would want to have sufficient assurances and data supporting the possible success of that program.
Steve Lewis03:08
This is a really interesting area, the kind of translational approach at the at the preclinical stage. Tell me about that flow. What does it look like when a customer engages your organization?
Andre Ghetti, PhD 03:22
Yeah. So customers come to us, typically already having some pretty good idea about their molecular target or their intended or HOPE mechanism of actions for the drug, although in some cases, they may come to us specifically to validate or identify a target. And then we drill down into specifically the therapeutic space, what type of assays and tissues would be necessary to try and generate data supporting the possible efficacy for that particular molecule. When it comes to safety, we’re generally focused on cardiac safety. So we conduct work using human primary cardiomyocytes. Regarding efficacy, we have been doing and continue to do a lot of work in the pain space, the search for analgesic compounds that are non-addictive. So we've been part of the big effort to try and identify drugs that are potent analgesic but without the side effects and the issues of opioids. Other areas in which we're very active are fibrosis. And so in particular lung fibrosis, we've been doing a lot of work in this particular space, but also kidney fibrosis. And so these are all spaces in which companies come to us hoping to further validate their compound and therefore justify additional investment and possibly clinical studies.
Steve Lewis04:55
That's really interesting. Can you tell me a bit about your tissues, or cell lines, that might be a bit novel to your company?
Andre Ghetti, PhD 05:04
Right. So these are all primary tissues that we recover through an extensive network of collaborations with hospitals and transplant centers across the U.S.. These are all tissues that have been consented for research. So these are non-transplantable tissues that have received consent from the donors or the donor’s families, for use in research. And we are very specific as to the type of use we do. Also it's fully disclosed that this is in the context of commercial utilization, essentially helping pharmaceutical companies. But one thing that is generally appreciated by the donor and the partnering hospitals is the fact that AnaBios is not focused on any particular drug or therapeutic area or pharmaceutical company. We really help the industry at large. So it's really an open platform, widely accessible to pharmaceutical companies, and we've been instrumental in helping many programs succeed, reach clinical stage, and advanced compounds to drug approval.
Steve Lewis06:17
I imagine that's a really valuable service offering as a whole. Do you customize for every single company in that regard?
Andre Ghetti, PhD 06:28
Yeah, there is a certain level of customization which is required. You know, some companies come and they have drug candidates that are biologics or that are small molecules, different physical, chemical properties, tissue penetration dynamics. So there is some level of customization. However, you know, we also strive for a level of standardization that allows us to compare the results we obtain with a particular drug with the results obtained with standards and in other molecules. So it's a combination of customization and standardizing the assets as well.
Steve Lewis07:05
This is such an interesting interface between R&D, right, the development, and then translating into clinical. You're almost like that very final step before a company makes that large investment in the clinical trials?
Andre Ghetti, PhD 07:22
That's exactly right. So as I said, we can engage early on in terms of identifying targets. So we do quite a bit of transcriptomics, genomic works early on, even proteomics with some of our clients. But then typically engagement is late, preclinical stage, prior to, you know, the finalization of the IND or prior to the selection of a clinical candidate for the ultimate validation of compounds.
Steve Lewis07:52
This is a very fascinating area of the market, and I think you might be the first company who I've spoken with in in this particular segment. You mentioned the omics, right, proteomics and genomics, and I imagine that there's a lot of data that needs to be generated to prove that safety and efficacy for the customers that you're doing work for. How many human samples do you need, in general?
Andre Ghetti, PhD 08:23
That is a really interesting question, and in fact, it was one of the first questions that a few years ago the FDA asked us. So pretty early on in the history of the company, we did proactively engage the FDA to get their feedback and also to familiarize them with our platform and capabilities. And so they asked us to conduct a study to look exactly into that, what is the variability, donor to donors, and how many samples one would have to evaluate to have a conclusive research. In general terms, if you're looking at a set of healthy donors, the number is surprisingly low. So the variability in drug response that is observed, typically on the market or in subjects in a clinical studies, is largely related to variability in how drugs are metabolized, primarily by the liver. Past the metabolism hurdles, the response of a particular tissue to a given drug is significantly robust and consistent from one individual to the next. And so with the FDA studies, for example, the conclusion was that for assessing cardiac safety, taking three tissue samples from three different donor hearts would be sufficient to have robust conclusion. And this is a surprisingly small number, which makes this type of study extremely practical, and even from a cost perspective, very affordable for pharmaceutical companies.
Steve Lewis10:01
With such a lower number of samples, that is lower than I expected for the answer as well, how long does it take to do a study?
Andre Ghetti, PhD 10:12
Typically, weeks to probably a couple of months. Depends on the complexity of the study and exactly what the type of assays that need to be conducted. It can take a little bit longer. In recent days, many companies are actually using genetic-type of approaches, or gene therapy type of strategies, and so in this case, we need to treat the tissues or treat the cells. Let the changes in gene expression take place, or suppression of RNA expression happen, and then we can run the test. But generally we're within weeks or up to a month or two at the most. It's a fast timeline, especially if you compare to any timeline typical of clinical studies. Okay. So first of all, let me be clear, we absolutely do not replace clinical studies. But as you mentioned earlier, this is a way to de-risk the clinical studies. And so by the time a compound is put in first in human studies, has a higher probability of success and also a much lower probability of resulting in toxicities in the early volunteers that are involved in the phase one studies.
Steve Lewis11:24
Sometimes on this show, we talk about the startup world. With the focus being de-risking, can companies that come to you utilize your data to help make that step toward perhaps fundraising for that next stage?
Andre Ghetti, PhD 11:42
Yeah, we see, we see a lot of that. In fact, our clients go from very small startup teams that essentially have virtual companies still in fundraising mode, all the way to the very large pharmaceuticals. And you're absolutely right. In fact, we've seen even in some cases, venture capital groups come to us asking to do some tests on a drug that they are evaluating for potential investment.
Steve Lewis12:09
So that seems like a really, a really valuable service. I think risk, especially in today's world, the more you can reduce it, the better. Now when you're designing your studies, do you develop novel kind of functional assays too?
Andre Ghetti, PhD 12:27
Yeah, that is another great question. So the gold standard the way to validate these assays is really to first have an idea of what the drug, at least the reference molecule has been reported to do in the clinic. So all the assays that we run are validated using compounds that have been in man and for which toxicities or efficacy are well known. And so then we go back to the laboratory and we identify specific test and endpoint measurements that are highly correlated to the clinical outcomes. And so in some cases, it's a pretty straightforward methodology. In other cases, we have to develop sort of customized assays and even customized methodologies. For example, with cardiac safety, a few years back we developed an instrument and the software to be able to measure a specific element of cardiac contractility for which there was no instrument of the market and no way to make those measurements using off the shelf technologies. So again, for the purpose of creating tests that are maximally predictive of clinical outcomes.
Steve Lewis13:48
Did you expect that you would ever have to build hardware when you started the company?
Andre Ghetti, PhD 13:53
No, absolutely not. And we try to avoid that as much as possible, but sometimes really there is no there's no way around it.
Steve Lewis14:01
When we're talking about this area of preclinical, kind of looking forward toward clinical, tell me how characterized these biomarkers are? Or alternatively, does AnaBios really focus on that, maybe selectivity or specificity in their studies to support the customers?
Andre Ghetti, PhD 14:25
Yeah, we establish selectivity and specificity primarily using these reference compounds, again, molecules that have been used in in humans for many years. And then try, essentially, to position the new drugs on the spectrum of activity of the known compounds and report back to the client and say, "Yes, it looks like your compound is behaving similar to this set of molecules which have been shown to be highly toxic or highly efficacious". And then they use this information to make their own internal decision making.
Steve Lewis15:03
Now with the data analysis, how has that changed in light of the recent development of generative AI?
Andre Ghetti, PhD 15:12
That's another interesting question. We do use quite a bit software that is essentially relying heavily on elements of machine learning. So there's a couple of challenges that we saw with those strategies. The first one is the detection of very small signals within, for example, a continuous recording of electrical activity or contractility in cardiac cells. And so when you feed thousands of traces into an AI, they are surprisingly good at learning to distinguish even tiny changes in the in the patterns of activity that can then very accurately predict the outcome for a given compound. So, there is quite a bit of our analytical tools that rely on AI. And the other one is when we utilize imaging methods, and again, looking at hundreds or thousands of human sensory neurons, for example, in a preparation the ability to automatically identify all the cell bodies and analyze the responses and activation or inhibition of all those individual neurons, again, using machine learning type of tools is extremely helpful. And so we've had, again, to create a lot of custom tools to accomplish this, but once you have those in place, the work and then the analysis is very, very fast. In the early days, when we had to do everything manually, we had situation where it may take a day for us to perform the experiment and the measurements and then would be two weeks to analyze the data. And so now we're down to minutes for running those type of analysis.
Steve Lewis17:05
We're excited to be in season three of Speaking of Mol Bio, and we know that we have you our loyal listeners to thank for the growing success of our podcast series. As a thank you, we're offering a free portable wireless speaker so you can listen to the podcast or your music anywhere. I have one on my desk, and I love how easily it connects to my phone. It's nice when I want a break from my headphones or want to share what I'm listening to with others. I hope you'll visit thermofisher.com/molbiopodcast to request yours today. Please note this item is only available in some regions and only while supplies last. Again, visit thermofisher.com/molbiopodcast to request yours. And now back to our interview.
Steve Lewis17:53
Moving more into the molecular biology space. Can you tell us a bit about technologies that you might use in the laboratory?
Andre Ghetti, PhD 18:01
Yeah, so we use quite a bit molecular cloning, RT-PCR. We also do utilize RNA sequencing. And so there is a few different activities in which these technologies are utilized. So as I mentioned, there is obviously the genomic and transcriptomic characterization of the tissue. And what is interesting, by the way, about this platform is that we have access to tissues from healthy donors, but we also have access to tissue from diseased donors. And so for example, if you're studying pain, having access to nerve preparations from individuals that were affected by chronic pain can be extremely useful, because you can then identify particular molecular signatures that are associated with chronic pain, and then helps identifying potential molecular targets. Another area where we use molecular biology extensively is in the generation and implementation of what are called genetically encoded sensors. So for example, one measurement that for us is very important, is the detection of intracellular calcium dynamics. And there's a few different ways to do this. One is to introduce into the cells small molecules that become fluorescent when the calcium concentration in the cell increases. And these are great for sort of very short-term type of experiment transient measurements. But if you actually want to track calcium transient over a period of days or weeks in the same cell preparation, then genetically encoded sensors are a much better way to do it. And so one strategy is to again, clone these genetically encoded sensor which oftentimes are variants of, or include the fluorescent molecules, like GFAP, into construct which are then transfected into the cells or infected into the cells. And then the last area where we and in oftentimes in collaboration with our clients, use molecular genetic methods is for applications related to gene therapy. So cloning related to expression of small inhibitory RNAs, introducing exogenous genes into cells, or working to do RNA knockdown. And again, a lot of the validation for these methodologies relies on RT-PCR preparations, or even on single cell.
Steve Lewis20:38
Why would somebody want to track the movement of calcium?
Andre Ghetti, PhD 20:42
So the movement of calcium, especially in neuronal cells, is an indicator of neuronal activation. So, sensory neurons signal pain signals, for example, to the brain using electrical impulses in the form of action potentials. But every time a cell and neuronal cell fires an action potential, there is also the activation of voltage gated calcium channels, which result in influx of calcium into the cell. And so, using calcium imaging is a methodology that allows for easier scaling and higher throughput of the measurement, as opposed to the traditional voltage clamp electrophysiology, which we also perform, but is essentially measuring one cell at a time, impaling one neuron at a time, and it's very time consuming and complicated. With calcium imaging, you can track the activation or inhibition of hundreds or thousands of neurons in a single experiment in a matter of minutes.
Steve Lewis21:48
Moving into you mentioned that the gene therapy and cloning, can you tell us a bit about the type of cloning that you would do for that kind of study?
Andre Ghetti, PhD 21:57
Yeah, so this is primarily adapting into different, different plasmid and different constructs, exogenous genes, or in some cases, even construct containing small inhibitory RNAs for the purpose of either expressing a protein in the cells that is believed to provide a therapeutic effect or inhibiting the expression of an endogenous protein that is believed to be associated with the pathology. And so primarily the cloning has to do with the preparation of the of the plasmid and the construct that then is then transfected into the cell. And transfections can take a few different forms, from lipid preparations, lipofectamine-like or LNPs, or in some cases, there is a viral transaction. So we have a viral construct, and then the viruses is used to infect the cell preparation.
Steve Lewis22:58
Are there any gaps in the technologies that you use?
Andre Ghetti, PhD 23:01
No, I think in these days we, you know, there are ways to do the cloning with really high efficiency, and so these are relatively straightforward steps. And again, as I mentioned, there are a variety of methodologies and vectors that can be used to then insert the exogenous DNA construct into the cell. So typically, the choice of vector is related to the specific cell types. Some cells can be more easily transfected with the lipofectamines or LNPs. For others that may be a little bit more challenging, we resort to viruses.
Steve Lewis23:43
What number of, I guess, base pairs do you do most often assemble?
Andre Ghetti, PhD 23:48
Yeah, typically is in the is in the few thousands. So the constructs are, you know, generally below 10 Kb, yeah. But that includes, so that includes all the other elements that are on the on the plasmids that are necessary for expression. So the target proteins themselves, or the construct that are driving the RNA knock down, those are typically less than 3 Kb.
Steve Lewis24:15
Got it. And what kind of throughput? Like do you all do that daily in the laboratory, or is that a bi-weekly thing?
Andre Ghetti, PhD 24:23
No, that's probably more of a monthly or quarterly. And again, it's often collaborative between us and our clients. So each construct can then be used for extended, sophisticated experiments. So with each one of these constructs, we can do a lot of work and generate a lot of data.
Steve Lewis24:46
Zooming out a bit to speak more about AnaBios, I've seen on your website and heard you mention early human insights. What does that mean to your company?
Andre Ghetti, PhD 24:58
Yeah, so as we mentioned earlier, the idea is to enrich preclinical research, which is typically focused on cell line, in vitro preparations as well as animal models, enrich this stage of research with human relevant data. It is not a replacement for clinical studies, absolutely not. But having data that is generated in primary cells or even primary tissue preparations from human subjects can provide really critical translational data, translational information.
Steve Lewis25:37
Do you see this area of translational work growing in the coming years in light of the new technologies like gene therapy and cell therapy becoming more popular?
Andre Ghetti, PhD 25:52
Yeah, so that's one element. The other one is that also regulatory bodies as well as government science funding bodies like the NIH, even in recent months, have really issued statements where there is a very strong push towards the replacement of a lot of animal-based work and animal-based research with more human centered type of research. And so this can be a combination of in vitro model organoids that are derived from stem cell type of preparations. It could be cellular component of in silico investigations, as well as the type of research we do on primary human cell and human tissues. And so all of these three areas I think are going to be really the next wave of translational research, trying to replace as much as possible working animals.
Steve Lewis26:50
What's next for AnaBios?
Andre Ghetti, PhD 26:53
We are continuing to expand looking at a larger and larger number of therapeutic areas. And so what we have learned in terms of recovery of human cells and tissues were, I have to mention one of the greatest challenges is to maintain the full viability and functionality of these preparations as they are shipped from hospitals all over the country to our laboratory in San Diego. We now have learned how to do this at scale with a very high success rate. And so there's a lot of lessons that can be applied to a growing number of tissues and therapeutic areas. And so one areas where we're looking to grow now is oncology, but there is also other areas with significant unmet medical need. We're also especially interested in areas where animal models have proven particularly problematic and not very good at predicting clinical outcomes. This is where we want to focus, and this is where I think we'll have to, the opportunity to contribute the most interesting data.
Steve Lewis28:05
Going back to one of the areas of research that you mentioned, can you tell me a bit how you are using PCR?
Andre Ghetti, PhD 28:14
PCR is used in a variety of ways. One is to, as I said, when there are experiments aimed at gene or messenger RNA knockdown of a specific transcript, we use RT-PCR to track the efficacy of the knockdown. It can also be used for detecting expression of what are believed to be pathological forms of specific messengers, and so sometimes in tissue from disease donors, it is very useful to actually measure which variant of a particular RNA is being expressed. And this helps our client decides whether or not that particular donor is a good target for testing their drug.
Steve Lewis29:03
In today's conversation, has there been any topic that I may not have touched on that you want to share with our listeners?
Andre Ghetti, PhD 29:14
No, I think we've touched upon, yeah, the challenges in using, you know, tissue. And this is actually very important, because just having access to human tissue doesn't guarantee quality actionable information. You know, you always have to check and validate that the human tissue is preserving the pharmacological and physiological properties of, of the original, native sample. And then again, always referencing molecules that have been in the clinic that have known effects. You know, I see sometimes companies trying to use human cells or human preparations, but without, you know, adequate validation and reference to known clinical responses. This is very important to create sort of a standard of reference in the field.
Steve Lewis30:11
Do you see an opportunity there in maybe inducible pluripotent stem cells for supporting your work?
Andre Ghetti, PhD 30:21
So there is some room for those type of studies. At the same time, you know, we have seen and published on differences between stem cell derived cells and tissues and the actual primary cells and tissues. And so there are elements well known, related to immaturity, oftentimes, of cells or tissues derived from stem cells. But with that said, I see the utility of stem cells, especially in the earlier stages of screening, where you need the quantities of plates and cells. That's, frankly, the primary cells that we use cannot, cannot provide. And so one could envision a pipeline in which there is an earlier screening stage with a very large number of compounds in the thousands or hundreds of thousands, performed on stem cell-derived preparations, and then later on, as the funnel shrinks, moving into primary cells and primary tissues.
Steve Lewis31:28
For somebody who might be earlier in their career, and they wanted to kind of follow in your footsteps. What advice would you give to them?
Andre Ghetti, PhD 31:37
Yeah, so probably the place to start is team up and surround yourself with great people. And great people means not just fun folks you would want to go out and have a beer with, but people that really complement your skills, people that are going to you know, speak openly and truthfully to you and tell you at your face when they think you're making a mistake. That's first step, and the second one is, is never stop learning. Basically, continue to be hungry for knowledge, and always remember that you can and should learn from pretty much everybody around you. The people above, below, you know, the colleagues, that there is always something to learn pretty much everywhere and from everyone. And it's critical because biotech, especially in the startup, is a highly dynamic environment, and you're going to have to continuously evolve the skills that you have in order to, you know, stay on top of the game.
Steve Lewis32:43
What has been the key to your success, aside from surrounding yourself with people?
Andre Ghetti, PhD 32:52
Yeah, people, the other one is probably the other two elements are focus and persistence. Yeah, I think there is oftentimes, especially in science, for a scientist, you know, it's very tempting to go down many different alleys. There's so many interesting things that come your way and could essentially end up being distractions. So really having clarity on what you want and need to focus on is important. And then persistence, because before things go the right way, there's going to go a million times the wrong way. So if you're not persistent, you're not going to make it to the finish line.
Steve Lewis33:37
I really appreciate your time today, Andre. This has been such a just wonderful conversation. And getting to learn a bit about what I'll call the tip of the spear, or really that tipping point where it's right before going into clinical studies, has just been so tremendous. So thank you very much for your time today, and I wish all the best to AnaBios and your continued success.
Andre Ghetti, PhD 34:02
Thank you, Steve. I did appreciate the opportunity to speak with you.
Steve Lewis34:09
That was Dr. Andre Ghetti, CEO of AnaBios in San Diego, California. To learn more about Andre and his work, visit the Speaking of Mol Bio page to download his guest profile at thermofisher.com/molbiopodcast. There you can explore all three seasons of this podcast series, which includes great conversations that really stand the test of time. Again, that URL is thermofisher.com/molbiopodcast.
Thanks for tuning in, and until next time, cheers and good science. Speaking of Mol Bio is produced by Matt Ferris, Sarah Briganti, and Matthew Stock.