Interactions of Biotechnology with Human Physiology and Function Along the Gut-Brain Axis
Proceedings of a Workshop—in Brief
The microbial communities that reside within the human gut have profound effects on our health. Recent research has revealed that their sphere of influence extends well beyond their long-recognized role in digestion and nutrient absorption to include interactions with the immune system, the brain, and other systems. New biotechnologies are tapping into interconnections with commensal microbes along the gut-brain axis, opening exciting opportunities to prevent and treat neurological and other disorders, discover new therapeutic modalities, and improve health throughout the lifespan.
The National Academies of Sciences, Engineering, and Medicine hosted a workshop on Interactions of Biotechnology with Human Physiology and Function on December 15-16, 2022. The workshop was organized under the auspices of the National Academies’ Standing Committee on Biotechnology Capabilities and National Security Needs.
Presenters and attendees from government, academia, and industry gathered for discussions and presentations addressing recent research into the connections and mediators along the gut-brain axis, key challenges and limitations in this research, transdisciplinary technologies being pursued for potential applications, and a vision for future developments in this field. This Proceedings of a Workshop—in Brief provides the rapporteurs’ high-level overview of the event.
Workshop Chair Elliot Chaikof (Harvard Medical School) set the stage with introductory remarks. The microbes that dwell on and within the human body are known to have profound effects on human physiology, and Chaikof highlighted how recent insights have brought an increasing appreciation for the ways in which microbial communities mediate communication across the gut-brain axis. For example, he explained, commensal microbes are integral to the sensory systems we use to monitor our internal and external environments, and they are also instrumental in immune systems, with important implications for neurological development in early life and immune functioning throughout the lifespan.
Chaikof described how coevolution between humans and our gut microbes has spanned our entire history as a species, yet the conditions in which people live today are dramatically different from those of even the relatively recent past. Cultural and technological developments in the past few centuries, including our
physical environments, diets, and behaviors, he said, have implications for health and the human body’s interactions with the microbes with which people have co-evolved. One challenge in dissecting these interactions, he added, is the once-overlooked vast source of genetic variation stemming from microbes both within and among humans, including a diversity in gene expression and metabolic pathways.
While this unique point in human history poses challenges, understanding more about microbiome activity and interactions along the gut-brain axis also reveals tremendous opportunities. Chaikof noted an emerging body of research suggests that we may be able to leverage this connection to treat symptoms of brain disorders, modulate neuronal signals, curb maladaptive inflammatory responses, tune the immune system, or develop new strategies to influence health through the foods people eat.
To explore and realize these opportunities, Chaikof said, will require not just deep expertise in one field but bringing together facts and theory across multiple disciplines to create a common knowledge base—something that can only be achieved with a balanced workforce of specialists and generalists with the support necessary to foster innovation and problem-solving. By convening experts from a wide range of areas to examine the emerging opportunities in this space, this workshop provided a forum for envisioning the path toward a greater understanding of the gut-brain axis to enhance fundamental knowledge and spur technological advances.
INTEGRATION ACROSS SYSTEMS
In the first session, speakers examined how gut microbiota influence the immune and nervous systems and drive molecular and sensory inputs that contribute to behavior. Speakers offered examples of opportunities to modulate the two-way transmission of signals along the gut-brain axis with electronic devices, bacteria-based therapeutics, and other approaches.
Sriram Chandrasekaran (University of Michigan) discussed how genome-scale metabolic models can help to inform precision medicine approaches by deciphering interactions between microbiome metabolites and the epigenome. He described how metabolites produced by bodily processes, including by the activity of our microbiome, result in changes in gene expression.1 The complexity of the molecular signals mediating the crosstalk between metabolism and the epigenome makes it challenging to untangle these pathways, but researchers have found that combining engineering models with machine learning can help. Using this approach, Chandrasekaran and colleagues have discovered new metabolic-epigenetic interactions and developed ways to integrate across biological scales, bridging from macroscale inputs such as diet and environment to molecular-level changes within cells. For example, they traced how different nutrients alter cells’ sensitivity to certain drugs, illustrating that metabolites can, in effect, determine a drug’s potency.2 Chandrasekaran noted that the work highlights how linking siloed biological data across different scales of space and time can help researchers predict cellular behavior and inform new approaches for precision medicine.
One way to influence the signals between the nervous system and the immune system is through electronic impulses. Kevin Tracey (Feinstein Institutes for Medical Research and Northwell Health) shared his work in this area, known as bioelectronic medicine, with a focus on vagus nerve stimulation. Tracey described how the vagus nerve is a structure composed of two paired nerves (left and right) which are commonly referred to in the singular, and together collectively comprising approximately 160,000 nerve fibers that transmit information between the body and the brain. It has long been known to play a key role in the systems that maintain homeostasis such as heart rate, respiration rate, and insulin secretion. Over the past two decades, Tracey and colleagues have uncovered its role in the immune system and inflammatory processes,3 using a variety of experimental approaches aided by machine learning. They used this information to develop vagus nerve
__________________
1 Campit, S. E., A. Meliki, N. A. Youngson, and S. Chandrasekaran. 2020. Nutrient sensing by histone marks: Reading the metabolic histone code using tracing, omics, and modeling. BioEssays 42(9):2000083.
2 Shen, F., L. Boccuto, R. Pauly, S. Srikanth, and S. Chandrasekaran. 2019. Genome-scale network model of metabolism and histone acetylation reveals metabolic dependencies of histone deacetylase inhibitors. Genome Biology 20(1):49.
3 Tracey, K. 2002. The inflammatory reflex. Nature 420:853-859.
stimulation technologies that tap into key nerve circuits to better control inflammation, with demonstrated success in relieving the symptoms of rheumatoid arthritis.4
Another way to influence the gut-brain axis is through the manipulation of microbial communities. Mauro Costa-Mattioli (Altos Labs) discussed opportunities to use microbes to treat neurological disorders, supported by a growing body of evidence that microbes modulate behaviors. Through work examining the relationship between maternal obesity and autism spectrum disorders (ASD), Costa-Mattioli and colleagues discovered that the bacterium Lactobacillus reuteri is instrumental in the neurological pathway through which social interactions engender feelings of reward—a finding, he said, that is relevant to ASD since people with autism often do not find social interaction rewarding in the same way that neurotypical people do. He continued that the research suggests that the presence of these bacteria in the gut can influence behavior by altering the production of molecules in cells. These chemical signals can then be transmitted through the vagus nerve and ultimately alter pathways in the brain. Costa-Mattioli described how in animal studies and small clinical trials in humans, L. reuteri has been found to help reverse some of the social deficits associated with ASD.5 The work offers early evidence that precision probiotics could represent a promising new strategy to treat neurological disorders, though Costa-Mattioli said further research is needed.
MODIFICATION OF MICROBES TO PROMOTE HEALTH
A deeper understanding of the gut microbiome leads to the possibility of manipulating these microbes to influence their roles to the advantage of the host. In the second workshop session, speakers highlighted work focused on developing microbiome-based therapeutic interventions.
Elaine Hsiao (University of California, Los Angeles) studies fundamental questions about how gut microbes interact with the nervous system, including their influence on the development of the nervous system in early life, their impact on chemosensory signaling, their place in the gene–environment interface, and the ways in which these roles influence microbial fitness and evolution.6 Applying microbiome research to inform new therapeutic approaches, Hsaio’s team has used molecular approaches to identify bacterial species and molecules that predict protection against seizures in people with epilepsy, and is exploring how these might be exploited to enhance the protective effects of a ketogenic diet.
Sarkis Mazmanian (California Institute of Technology) discussed his research on the role of the microbiome in Parkinson’s disease. While Parkinson’s is thought of as a brain disease affecting motor signaling, studies have shown linkages with other organ systems, with gastrointestinal features including constipation often preceding motor symptoms by many years. After studying pathways involving alpha-synuclein, a protein that can promote neural dysfunction when misfolded,7 the researchers tested dietary interventions that increase short chain fatty acids, finding that a high-fiber, Mediterranean-style diet limited neuroinflammation and attenuated motor deficits in animal models. Further studies pointed to microglia activation states as a likely mediator of this effect. Together, Mazmanian noted that this body of work suggests that dysbiosis in the gut microbiome could contribute to Parkinson’s disease and that a high-fiber diet may limit neuroinflammation through its impacts on microglia.
SIGNALING FROM GUT TO BRAIN
The gut-brain axis has multiple mechanisms for detecting and transducing signals, including neuronal signaling, metabolite production, and interactions with the immune system. To start the workshop’s second day,
__________________
4 Koopman, F. A., S. S. Chavan, S. Miljko, S. Grazio, S. Sokolovic, P. R. Schuurman, A. D. Mehta, Y. A. Levine, M. Faltys, R. Zitnik, K. J. Tracey, and P. P. Tak. 2016. Vagus nerve stimulation inhibits cytokine production and attenuates disease severity in rheumatoid arthritis. Proceedings of the National Academy of Sciences of the United States of America 113(29):8284-8289.
5 Sgritta, M., S. W. Dooling, S. A. Buffington, E. N. Momin, M. B. Francis, R. A. Britton, and M. Costa-Mattioli. 2019. Mechanisms underlying microbial-mediated changes in social behavior in mouse models of autism spectrum disorder. Neuron 101(2):246-259.
6 Fung, T. C., H. E. Vuong, C. D. G. Luna, G. N. Pronovost, A. A. Aleksandrova, N. G. Riley, A. Vavilina, J. McGinn, T. Rendon, L. R. Forrest, and E. Y. Hsiao. 2019. Intestinal serotonin and fluoxetine exposure modulate bacterial colonization in the gut. Nature Microbiology 4(12):2064-2073.
7 Sampson, T. R., J. W. Debelius, T. Thron, S. Janssen, G. G. Shastri, Z. E. Ilhan, C. Challis, C. E. Schretter, S. Rocha, V. Gradinaru, M. F. Chesselet, A. Keshavarzian, K. M. Shannon, R. Krajmalnik-Brown, P. Wittung-Stafshede, R. Knight, and S. K. Mazmanian. 2016. Gut microbiota regulate motor deficits and neuroinflammation in a model of Parkinson’s disease. Cell 167(6):1469-1480.
speakers highlighted recent findings, emerging tools, and conceptual frameworks for understanding and leveraging these signaling mechanisms to improve health.
Mark Lyte (Iowa State University) discussed neurochemicals as an evolutionarily conserved mechanism for communication between microbiota and their hosts. Neurochemicals are ubiquitous, found not only in organisms with complex nervous systems but also in plants and microbes. He described how there is mounting evidence that commensal bacteria are not passively relying on their hosts as a source of nutrition and energy, but rather are active players in the host’s health, by both producing and responding to neurochemicals.8 This mechanism can be exploited to design probiotics to influence the gut-brain axis, but to advance this work will require traditional microbiology studies to better understand microbes’ capacity to make and use neurochemicals and will require addressing key challenges in bioinformatics to enhance predictive capabilities, Lyte said.
Melody Zeng (Cornell University) highlighted insights into the role of the microbiome in immune and neuronal development during pregnancy, birth, and early life. The maternal microbiome has been found to influence susceptibility to infections, asthma, obesity, and neurodevelopmental disorders in offspring.9 Zeng and colleagues have traced the activity of gut microbiota-specific immunoglobulin G (IgG) antibodies as a mediator in the gut-placenta axis.10 They also found that commensal bacteria in the neonatal intestine produce 5-hydroxytryptamine receptors (serotonin receptors), which in turn promote immune tolerance to commensal gut bacteria.
Michael Fischbach (Stanford University) described how new research platforms can accelerate advances in microbiome research, and he discussed a model developed by his lab. To overcome the drawbacks of fecal transplant-based studies, which offer the benefit of a microbial community-based approach but lack specificity, and single organism-based approaches, which offer specificity but fail to capture the community context, Fischbach and colleagues developed a complex model gut microbiome.11 This collection of microbes reproducibly and reliably colonizes germ-free mice and can be used to trace how individual microbial species modulate the immune system by profiling T-cell specificity to each strain. Using this model microbiome, researchers found that sets of T-cell receptors were stimulated by multiple microbes at the same time, rather than each bacterial species interacting with immune cells in a one-to-one fashion, demonstrating the model’s utility in uncovering how microbial communities interact with immune cells and identifying the genes that define such relationships. The researchers are making the model freely accessible for others to use and adapt, Fischbach noted.
Diego Bohórquez (Duke University) highlighted the signaling pathways through which the body conveys information from the intestine to the brain, creating feedback that informs eating behavior. While the chemistry of digestive processes has been well studied, far less is known about the neurological processes involved. Studies of cellular behavior,12,13 spread of a monosynaptic rabies virus from the gut to the brain,14 and activity of microbes in regulating eating behavior15 have elucidated how the body quickly sorts nutrients and informs the brain so that behavior can be adjusted
__________________
8 Lyte, M. and D. R. Brown. 2018. Evidence for PMAT- and OCT-like biogenic amine transporters in a probiotic strain of Lactobacillus: Implications for interkingdom communication within the microbiota-gut-brain axis. PLoS One 13(1):0191037.
9 McDonald, B. and K. D. McCoy. 2019. Maternal microbiota in pregnancy and early life. Science 365(6457):984-985.
10 Sanidad, K. Z., M. Amir, A. Ananthanarayanan, A. Singaraju, N. B. Shiland, H. S. Hong, N. Kamada, N. Inohara, G. Núñez, and M. Y. Zeng. 2022. Maternal gut microbiome-induced IgG regulates neonatal gut microbiome and immunity. Science Immunology 7(72):3816.
11 Cheng, A. G., P. Y. Ho, A. Aranda-Díaz, S. Jain, F. B. Yu, X. Meng, M. Wang, M. Iakiviak, K. Nagashima, A. Zhao, P. Murugkar, A. Patil, K. Atabakhsh, A. Weakley, J. Yan, A. R. Brumbaugh, S. Higginbottom, A. Dimas, A. L. Shiver, A. Deutschbauer, N. Neff, J. L. Sonnenburg, K. C. Huang, and M. A. Fischbach. 2022. Design, construction, and in vivo augmentation of a complex gut microbiome. Cell 185(19):3617-3636.
12 Kaelberer, M. M., K. L. Buchanan, M. E. Klein, B. B. Barth, M. M. Montoya, X. Shen, and D. V. Bohórquez. A gut-brain neural circuit for nutrient sensory transduction. Science 361(6408):5236.
13 Buchanan, K. L., L. E. Rupprecht, M. M. Kaelberer, A. Sahasrabudhe, M. Klein, J. Villalobos, W.W. Liu, A. Yang, J. Gelman, S. Park, P. Anikeeva, and D. V. Bohórquez 2022. The preference for sugar over sweetener depends on a gut sensor cell. Nature Neuroscience. 25:191-200.
14 Bohórquez, D. V., R. A. Shahid, A. Erdmann, A. M. Kreger, Y. Wang, N. Calakos, F. Wang, and R. A. Liddle. 2015. Neuroepithelial circuit formed by innervation of sensory enteroendocrine cells. Journal of Clinical Investigation 125(2):782-786.
15 Vijay-Kumar, M., J. D. Aitken, F. A. Carvalho, T. C. Cullender, S. Mwangi, S. Srinivasan, S. V. Sitaraman, R. Knight, R. E. Ley, A. T. Gewirtz. Metabolic syndrome and altered gut microbiota in mice lacking Toll-like receptor 5. Science 328(5975):228-231.
accordingly—as well as the potential consequences when this process is dysregulated or possibly co-opted during infection. In the future, Bohórquez suggested it will be useful to study how commensal bacteria might influence behavior in ways that encourage people to eat foods that would benefit the microbes, but not necessarily their human host.
Vanessa Ridaura (Bill & Melinda Gates Foundation) discussed how her organization is investing in microbiome research as part of a broader effort to improve the health of women and children globally. In the near term, the foundation aims to build a body of work that elucidates the role of the maternal microbiome in fetal and infant development and informs products and mechanisms of action needed to impact clinical outcomes in vulnerable populations. Looking forward, she said it will be valuable to study immune adaptations through pregnancy, interactions between the vaginal and gut microbiomes, and impact of the gut microbiome on placental development.
BIOTECHNOLOGIES FOR NEUROMODULATION
Advances in understanding how the nervous system maintains homeostasis suggest that devices or drugs targeting the central nervous system and peripheral neural circuits can be exploited to learn about, and potentially modify, the underlying pathology of some diseases. In the second session of the workshop’s second day, speakers examined opportunities and challenges in this space.
Todd Coleman (Stanford University) discussed methods for leveraging electrical activity in the digestive system as a tool to tap into the gut-brain axis. He described a series of lab-based, portable, and wearable devices he developed to measure electrical activity in the digestive system. These devices shed light on the features and causes of digestive disorders and can even be used to entrain digestive system components in a target direction, representing a potential novel therapeutic modality.16 Coleman and colleagues are also bringing these approaches together with neuroscience to further probe the gut-brain axis, with particular focus on a brain structure known as the anterior insula, through simultaneous electrophysiologic monitoring of the stomach and brain.17
Hubert Lim (University of Minnesota) described ultrasound-based methods for influencing neuronal signaling and immune response. He described how the field is in a new era of brain-machine interfacing and neural prosthetics, from implanted devices such as spinal cord and vagal nerve stimulators to external devices such as transcranial magnetic stimulation and transcutaneous electrical nerve stimulation (TENS) devices, each of which has benefits and drawbacks. Within this technology landscape, ultrasound neuromodulation has emerged as a promising way to achieve targeted modulation of the brain and peripheral nerves.18 Lim noted that ultrasound has the potential to be used as a noninvasive tool to probe or modulate the function of other cell types and end-organs in the body, including the liver, gut, pancreas, and celiac plexus for applications in a range of health conditions. In particular, Lim’s group studied the application of ultrasound to the spleen, an organ instrumental in interfacing between the nervous system and immune system.19 Studies in animals and humans suggest this approach can help suppress a hyperactive innate immune response, pointing to possible clinical applications for diseases like rheumatoid arthritis and COVID-19.
Stewart Campbell (Axial Therapeutics) discussed his company’s development of small molecule drug candidates that target the microbiome. One benefit of this approach, he said, is that drugs can affect neurological outcomes without needing to cross the
__________________
16 Perley, A., M. Roustaei, M. Aguilar-Rivera, D. C. Kunkel, T. K. Hsiai, T. P. Coleman, and P. Abiri. 2021. Miniaturized wireless gastric pacing via inductive power transfer with non-invasive monitoring using cutaneous electrogastrography. Bioelectronic Medicine 7(1):12.
17 Balasubramani, P. P., A. Walke, G. Grennan, A. Perley, S. Purpura, D. Ramanathan, T. P. Coleman, and J. Mishra. 2022. Simultaneous gut-brain electrophysiology shows cognition and satiety specific coupling. Sensors 22(23):9242.
18 Cotero, V., Y. Fan, T. Tsaava, A. M. Kressel, I. Hancu, P. Fitzgerald, K. Wallace, S. Kaanumalle, J. Graf, W. Rigby, T. J. Kao, J. Roberts, C. Bhushan, S. Joel, T. R. Coleman, S. Zanos, K. J. Tracey, J. Ashe, S. S. Chavan, and C. Puleo. 2019. Noninvasive sub-organ ultrasound stimulation for targeted neuromodulation. Nature Communications 10:952.
19 Zachs, D. P., Offutt, S. J., Graham, R. S., Kim, Y., Mueller, J., Auger, J. L., Schuldt, N. J., Kaiser, C. R. W., Heiller, A. P., Dutta, R., Guo, H., Alford, J. K., Binstadt, B. A., and Lim, H. H. 2019. Noninvasive ultrasound stimulation of the spleen to treat inflammatory arthritis. Nature communications 10(1):951.
blood-brain barrier. He added that small molecule drugs can have an easier path to implementation compared to other emerging microbiome-based therapeutic modalities since they fit into existing pathways for pharmaceutical developers and regulators. As an example, Campbell described the development of a small molecule drug targeting a microbiome metabolite, 4-ethylphenyl sulfate, that is dysregulated in ASD as a means of addressing irritability, a manifestation of ASD that poses particular challenges for patients and families.20 Animal studies and early phase trials showed that ingestion of the experimental agent results in a host of changes in microbial metabolites as well as reductions in irritability and anxiety,21 and it is currently being tested in a phase two trial.
CHALLENGES TO APPLICATION
In open discussion, speakers explored challenges to advancing basic research at the intersection of the microbiome, brain, and immune system along with considerations for translating research insights into practical applications.
Examples of Basic Research Priorities
Even as some technologies are already being used clinically or rapidly moving in that direction, some panelists emphasized the importance of basic research. They identified examples of several areas for further exploration, including how to ensure interventions based on manipulating processes along the gut-brain axis are appropriately specific in order to avoid side effects or unintended consequences; how interventions might be tailored to a person’s background or genetics; and what factors affect the timing and persistence of effects, which could impact how experimental findings are interpreted and how therapies are used in practice. For example, Hsiao noted it will be important to understand which interventions may have more short-term effects (such as a probiotic that does not lead to colonization of the gut microbiome) and which may have longer-term effects (such as interventions during prenatal development that could have lifelong ramifications).
Speakers also discussed different ways therapies could be developed and used. Hsiao noted that researchers are examining the use of both endogenous and engineered microbes. Mazmanian commented on the distinction between prophylactic and therapeutic applications and added that it will be useful to further examine the microbiome’s influence on how the body accesses and responds to drugs, which could help explain why some people respond to certain drugs or experience certain side effects while others do not. Building on this point, Hsiao suggested that the effects of the microbiome could be incorporated into screening approaches used for pharmacology and even the regulation of environmental chemicals.
Given the complexity of the microbiome and its connections across multiple organs and systems, Hsiao said researchers need to focus more on integrated, multi-system modeling and less on reductionist, trial- and-error approaches. Chandrasekaran encouraged experimenting with combined therapies, for example, incorporating dietary changes with interventions to stimulate the microbiome or reduce inflammation. Mazmanian agreed, noting that microbiome-based therapeutics may require a framework of thinking that goes beyond particular molecules hitting particular receptors and triggering a discrete, finite effect. He commented that metagenomic methods are a cost-effective approach for a more integrated view, but a lack of knowledge about gene functions limits its application.
Drivers and Roadblocks for Translational Science
Throughout the discussions, some speakers pointed to a tension between the drive to develop medical innovations and the importance of understanding the biological mechanisms through which they work. Costa-Mattioli stressed the value of uncovering mechanisms to the greatest extent possible; for example, in the case of precision probiotics, to identify the specific gene in a bacterium that results in the metabolite that affects brain signals. Knowing the mechanism, he said, can inform refinements of interventions to increase their
__________________
20 Needham, B. D., M. Funabashi, M. D. Adame, Z. Wang, J. C. Boktor, J. Haney, W. L. Wu, C. Rabut, M. S. Ladinsky, S. J. Hwang, Y. Guo, Q. Zhu, J. A. Griffiths, R. Knight, P. J. Bjorkman, M. G. Shapiro, D. H. Geschwind, D. P. Holschneider, M. A. Fischbach, and S. K. Mazmanian. 2022. A gut-derived metabolite alters brain activity and anxiety behaviour in mice. Nature 602:647-653.
21 Campbell, A. S., B. D. Needham, C. R. Meyer, J. Tan, M. Conrad, G. M. Preston, F. Bolognani, S. G. Rao, H. Heussler, R. Griffith, A. J. Guastella, A. C. Janes, B. Frederick, D. H. Donabedian, and S. K. Mazmanian. 2022. Safety and target engagement of an oral small-molecule sequestrant in adolescents with autism spectrum disorder: An open-label Phase 1b/2a trial. Nature Medicine 28:528-534.
benefits and reliability, for example, from a 10-20 percent improvement in outcomes to a 60-70 percent improvement. He added it can also lower the barrier to applying new interventions. Tracey noted that to advance the use of bioelectronic devices for treating a broader variety of conditions (such as anxiety or post-traumatic stress disorder, as some workshop participants suggested), focusing on diseases for which the molecular basis is well understood is likely to be most fruitful.
Trickier questions arise when the technology gets ahead of the basic science such that clinical studies demonstrate that an intervention works before researchers can fully describe the mechanism involved. Tracey noted that patient advocacy drives innovation and adoption of new therapeutic modalities, particularly with therapies for which mechanisms are not yet known. Given the urgent need for effective interventions and societal challenges such as medical debt, he suggested that researchers should not let a lack of mechanistic knowledge hamper advances that could improve lives today. Mazmanian agreed, noting that in his view a lack of mechanistic knowledge shouldn’t impede clinical application of a microbiome-based therapeutic if it is shown to be safe and there are people suffering from ailments that could be addressed. On the other hand, adequately ensuring safety can be a challenge when, as Chaikof noted, some biotechnologies are racing ahead of the regulatory oversight structures such that they are being adopted directly by consumers before elucidation and communication of potential benefits and risks.
Tracey noted that clinical trials can help to fully understand risks and benefits, but they are costly. In the life sciences, venture capitalists are accustomed to investing in technologies that may take a decade or more to make it to market, but they do still have some expectation of return. Rather than the question of “will this help patients,” these investment decisions are driven by expectations regarding how many people may use the product, how much they will pay for it, and how long it will take to penetrate the market, he said. As an example, the misalignment of incentives is why TENS devices have not been well studied as companies are not likely to make a high enough return on investment to support clinical trials for this inexpensive and widely available technology. Given this incentive structure, he said, one potential solution might be greater data sharing among researchers across diseases and mechanisms to enable more efficiency in, and prioritization of, clinical trials. To extract more value from the clinical trials that do occur, he suggested researchers and investors pay greater attention to clinical trials that fail; for example, by assessing whether there is a subset of patients who do derive benefits and focusing further assessments of that population. Drawing lessons from COVID-19 vaccine trials, Lim added that a general cultural shift toward volunteering for clinical trials could help trials move forward more quickly. Tracey added that building nonprofit research groups that join multiple health systems or sponsors could also help to increase trial enrollment and speed studies. Chandrasekaran noted that sharing of genomic data could also help.
From a practical standpoint, Mazmanian noted that while the securing of intellectual property (IP) and the ability to make a profit play a role in investment decisions for technologies in this space and getting therapies to market, manufacturing could also pose as a barrier. Jorge Santiago-Ortiz (Apertura Gene Therapy) stressed the importance of addressing product development and scalability, including manufacturing capabilities, early in the process to avoid insurmountable barriers later.
Envisioning the 10-Year Horizon
Asked what may be achieved by leveraging insights into the gut-brain axis over the next decade, panelists expressed their hope that research findings can translate into effective clinical interventions for various conditions. To achieve this, Mazmanian and Costa-Mattioli stressed the need to resolve regulatory questions with the U.S. Food and Drug Administration to create a path to approval for therapies that are safe and effective. In moving forward with basic research and translational efforts simultaneously, Hsiao and Chandrasekaran said it will be important to account for the complexity of the microbiome, which means not giving up after the first failed attempt, continuing to characterize mechanisms and molecules, and experimenting with combinatorial approaches that may
achieve greater impacts than individual interventions in isolation. Tracey expressed hope that neuromodulation could reduce the need for pharmaceutical interventions for a variety of conditions, increasing safety and driving costs down. To get there, he said it will be important to better align incentives and increase data sharing based on a drive to address patient needs rather than simply seek profit.
Panelists also noted that advancing translational research in this field will likely benefit from effective interdisciplinary teams. Some participants said this requires people from diverse fields to be willing to work across disciplines and break out of the traditional academic career structure, which may require rethinking the higher-level organization of institutes within universities as well as a willingness to cross from academia into industry. Lim said that it is important to build deep expertise within a field, but also to create incentives to foster team science. Under current structures, he said, the work required to sustain one’s own laboratory and advance one’s own career can disincentivize large multi-lab interdisciplinary teams and studies for which credit is more dispersed. Chandrasekaran added that training the next generation of researchers to have basic knowledge of key terminology across fields will help researchers speak the same language and collaborate more effectively.
REIMAGINING THE INNOVATION ECOSYSTEM
To envision a path forward for the research insights and biotechnologies presented at the workshop, speakers discussed how various research strategies and funding models might influence the innovation ecosystem to support or pose barriers to advancements and applications.
Kicking off the discussion, William Bonvillian (Massachusetts Institute of Technology) offered remarks on models for accelerating convergence to advance opportunities for insight and intervention at the gut-brain axis. Placing convergence alongside molecular biology and genomics—which he described as two previous interdisciplinary “revolutions” in biology—Bonvillian said this approach merges science and engineering to build a new knowledge base while developing new tools and new therapies. Bonvillian noted that the merger of talent bases of biology, engineering, and physical sciences could create new fields of scientific research. Convergence science offers great opportunities but also faces headwinds, he described, including lack of common language across fields, lack of interdisciplinary training to enable biologists to take advantage of emerging toolsets, entrenched siloing in funding and institutional operations despite efforts to increase cross-disciplinary collaboration, and limitations and barriers inherent to peer review models. He suggested that the Advanced Research Projects Agency for Health (ARPA-H) initiative could overcome some of these challenges and become a new mechanism for enabling convergence research.
Lessons from COVID-19 Vaccine Development
Bonvillian suggested that lessons from Operation Warp Speed, the program that accelerated development of COVID-19 vaccines, could inform approaches to accelerate progress in emerging technology at the gut-brain axis. Features of that program that proved instrumental to its success included a willingness to “pick winners” by identifying a small number of technologies to invest in; the establishment of guaranteed contracts; and a variety of other measures to assist with practicalities that often bog down technology development and deployment—including certification; contracting; regulatory compliance; and supply chain, scale-up, and distribution.
Of course, he noted, that program was undertaken in response to a unique set of circumstances, and one key advantage was that several vaccine platforms were already “in range” for translation into rapid vaccine development when the pandemic began. While that may not necessarily be the case for technologies in every field, Bonvillian emphasized that the mechanism of guaranteed contracts is something that could be leveraged in various contexts and could result in a cost savings for the government if the technology is poised to benefit (and reduce health care costs for) a large group of people. Chaikof raised the point that without sufficient checks and balances, guaranteed contracts could result in a
command economy. In response, Lim noted that models such as the Defense Advanced Research Projects Agency (DARPA) provide sufficient funding to enable working prototypes, but ultimately private investment is required for commercialization.
Lim noted that another difference with COVID-19 vaccines is that the risk-benefit assessment for clinical trials was calculated against the backdrop of a global pandemic and large numbers of people were willing to volunteer to participate. He suggested that future clinical trials could move forward more quickly by shifting toward a model more centered around patient safety in a way that better accounts for the true societal burdens of the condition being studied, rather than what he described as a liability-centric model that characterizes current practice.
Barriers and Opportunities for Translational Science
Throughout the workshop, speakers noted that a challenge in many translational research efforts is the difficulty of moving between animal models and humans. For example, Zeng noted that the short gestation period of mice makes them useful for studies of the microbiome during pregnancy and early development, but the types of bacteria found in the microbiomes of human infants do not reliably colonize germ-free mice, creating a challenge for translational studies. Bohórquez noted another challenge is the heterogeneity of gut microbiota, both between individuals and throughout the long course of the human digestive tract, which adds complexity to translating laboratory studies to practical applications intended for diverse populations such as those Ridaura’s team is working to benefit.
Speakers also highlighted opportunities to translate basic science and emerging biotechnologies relevant for the gut-brain axis to other areas of the body. For example, Coleman said the electrophysiologic devices his team has developed for the digestive system could be adopted for use elsewhere; his team is already working on applying it to the neck for monitoring of nervous structures, including the vagus nerve. Similarly, Lim said ultrasound modulation is being tested on various parts of the body; the key challenge is developing ways to target the intervention to specific structures and then measure the effects in a manner that is sufficiently localized. For small molecule drug development, Campbell said the microbial metabolites under investigation for irritability with ASD could potentially be evaluated for relevance to other mood disorders such as depression or schizophrenia.
To facilitate transformative technology in the coming years, Coleman and Lim pointed to a need for synergistic, multimodal approaches, such as developing ways to measure microbiome activity and physiology simultaneously or establishing biomarkers to provide a read-out of what exactly is being modulated when ultrasound is applied. Campbell said that there is a continuing need to better integrate datasets, many of which exist in separate silos, to enable a predictive, personalized approach to drug development. Costa-Mattioli noted that DARPA and other funders have explored creative models for providing more lasting funding streams to push truly innovative work that may require a longer timeframe than the typical National Institutes of Health funding cycle, balancing the drive for rapid development with the deep basic research investments needed to make major breakthroughs.
Scale-Up and Manufacturing Considerations
The ability to scale up and manufacture biomedical products is crucial to their ultimate success, yet Chaikof pointed out that there is very little attention afforded to this area in science education and academia. Lim and Bonvillian pointed to DARPA as a good model for funding the intermediary work necessary to bridge academic research to implementation. In particular, Bonvillian said that ARPA-H and other organizations could help to facilitate scale-up by cultivating collaborative communities of thinkers (which informs the selection of technologies to prioritize for investment) and by creating roadmaps to chart a path forward and anticipate challenges early in the process.
Manufacturing microbe-based products raises some special considerations. Ridaura speculated that microbial products derived from foods are likely easier to manufacture and scale up, but thinking beyond individual
bacterial species to leverage microbial communities adds complexity. She also noted that World Health Organization guidelines specify that such products should be made in facilities with a Good Manufacturing Practice (GMP) designation, but this is not happening in practice and will be logistically difficult to implement while keeping costs low enough to be suitable for widespread use in under-resourced regions. Lyte added that, as living organisms, microbes will continue to divide and evolve as they are grown in a fermenter, and studies have shown that probiotics often have very different properties depending on how they are manufactured. He said it will be important to go beyond simply documenting a product’s genus and species and establish functional assays to ensure products will behave as intended when given to patients.
Enabling Collaboration
Recognizing that advancing this field likely involves effective cross-disciplinary collaboration, speakers reflected on how their organizational structure influences the way they work as individuals and in teams. Zeng, working in a traditional university structure, said that she wants her work to ultimately find applications—not end in the form of publications—but finding that path (including partners in industry) can be challenging in academia. Lim commented that moving from academia to industry can be refreshing because the structure of having investor expectations and milestones to meet forces a more practical viewpoint and can speed progress. Costa-Mattioli said that while his company is still new, it is attempting to forge a model that allows researchers to tackle difficult questions in a new way outside of the traditional government funding structure. While most of the researchers came to the company from careers in academia (and, accordingly, with a mental model in which rewards are earned based on individual achievements), he said the goal is to keep the focus on the discoveries rather than the individuals behind them. The benefit of giving up the expectation of individual acclaim, he said, is the opportunity to make scientific contributions that are likely larger than what any contributor could achieve alone.
Science education and training structures also influence the capacity for effective team science. Bonvillian said being educated in silos creates a barrier to convergence and suggested a focus on developing “T-shaped learners”, who have a depth of knowledge in one field but broad exposure to concepts, terminology, and tools in other fields. Lim agreed, adding that nudging people to learn about other fields without sacrificing depth of knowledge in their own field will require a culture shift and incentives that are aligned with team science, which may require a rethinking of metrics for success used in academia.
CLOSING THOUGHTS
Chaikof closed the workshop with reflections on the opportunities and challenges raised by the technologies highlighted at the workshop (summarized in Table 1). Elucidating the intricate interactions along the gut-brain axis opens new opportunities to understand food and microbes as medicines, with exciting potential to develop new treatments for debilitating disorders and to improve health throughout the lifespan. Work in this field has advanced the ability to track and influence how our bodies interact with microbes to send and receive signals and modulate pathways affecting functioning from the level of cells to the level of human behavior. Workshop discussions also surfaced possible areas for future investments; including understanding mechanisms to refine therapeutic approaches, new tools to modulate signaling along the axis, streamlining regulatory pathways to make these tools clinically available, and integrating data and building bridges across fields and between academia and industry.
Chaikof speculated that therapies targeting the gut-brain axis may be approaching a tipping point at which decades of work will culminate in transformative medical advances, likening biotechnologies in this area to gene therapies, RNA-based medicines, and cell-based cancer therapies in that, only a few years ago, these innovations were still unproven but have now become reality in clinics around the world.
TABLE 1
Overview of Example Technologies Discussed at the Workshop
| EXAMPLE TECHNOLOGY | POTENTIAL APPLICATIONS | ADVANTAGES | LIMITATIONS | MATURITY/ECOSYSTEM |
| Genome-scale metabolic models (Chandrasekaran) |
Basic science discoveries to inform precision medicine | Integrates biological data from across scales of space and time to predict cellular behavior | Understanding metabolite-epigenome crosstalk remains a complex problem | Basic research approach ready for refinement and use |
| Model gut microbiome (Fischbach) |
Basic science discoveries to inform microbiome-based interventions | Enables tracing of individual species’ activities in a community context | Provides a model system, but does not reflect real-world variation | Basic research tool available for broader use |
| Gut electrophysiologic monitoring and modulation devices (Coleman) |
Monitoring to study and diagnose gastrointestinal disorders and the coupling of gut and brain; modulation as novel treatment modality | New opportunities to observe digestive processes and signaling; accessible alternative to brain-monitoring technologies; potential minimally invasive therapeutic option | Technology is still early-phase; limited exploration of research and clinical applications to date | Emerging technology with limited clinical evidence |
| Ultrasound modulation (Lim) |
Modulate immune-neural axis and potentially multiple end-organ targets to treat a range of conditions | Potential portable, low-cost, noninvasive new therapeutic option for inflammatory disorders with minimal side effects | Can modulate neural and non-neural cells; limited exploration of research and clinical applications to date | Emerging technology with ongoing pilot clinical trials to establish clinical evidence |
| Vagus nerve stimulation/bioelectronic medical devices (Tracey) |
Targeted stimulation of vagus nerve to treat a range of inflammation-related and other conditions | Potential to intervene with minimal side effects in inflammatory processes with known molecular basis; demonstrated initial success with rheumatoid arthritis; some treatment effect in some patients in other conditions | Some devices require surgical implantation; varying levels of evidence on effectiveness for some conditions; unclear whether effects persist in absence of active stimulation; difficult to establish molecular basis of connection to neural circuits for conditions with unclear molecular mechanisms | Commercially available but limited use for some conditions and mixed evidence on efficacy; ongoing clinical trials to validate efficacy for other conditions |
| Precision probiotics or microbiome-targeting dietary interventions (Costa-Mattioli, Hsiao, Mazmanian, Ridaura) |
Augmenting microbiome in targeted ways to treat neurological conditions or affect immune, metabolic, or neurocognitive outcomes in infants | Potential new therapies for conditions such as autism and epilepsy that have limited treatment options currently; could reduce barriers to treatment and increase access to effective therapeutic approaches; likely minimal safety concerns | Important to pinpoint mechanisms at work to improve reliability and efficacy | Pre-clinical cell and animal studies; small clinical trials |
| Small molecule therapies targeting the microbiome (Campbell) |
Drugs to treat neurological and other conditions | Little systemic absorption; capability to influence neurological outcomes without crossing the blood-brain barrier; fits with established drug development pathways for pharmaceutical industry and regulators | Requires well-elucidated pathways and mechanisms of action | Pre-clinical studies underway for several agents; two in phase 1-2 clinical trials |
NOTE: This table summarizes examples of technologies shared by workshop participants and should not be interpreted as consensus conclusions or recommendations of the National Academies.
DISCLAIMER This Proceedings of a Workshop—in Brief was prepared by Anne Johnson, Kanya Long, and Andrew Bremer as a factual summary of what occurred at the workshop. The statements made are those of the rapporteurs or individual workshop participants and do not necessarily represent the views of all workshop participants; the planning committee; or the National Academies of Sciences, Engineering, and Medicine.
WORKSHOP PLANNING COMMITTEE Elliot L. Chaikof (Chair), Harvard Medical School; Yasmine Belkaid, National Institute of Allergy and Infectious Diseases; Todd P. Coleman, Stanford University; Mauro Costa-Mattioli, Altos Labs; and Diane DiEuliis, National Defense University.
REVIEWERS To ensure that it meets institutional standards for quality and objectivity, this Proceedings of a Workshop—in Brief was reviewed by Elliot L. Chaikof, Harvard Medical School; Hubert Lim, University of Minnesota, Twin Cities; and Harris Wang, Columbia University. Lauren Everett, National Academies of Sciences, Engineering, and Medicine, served as the review coordinator.
SPONSORS The Standing Committee on Biotechnology Capabilities and National Security Needs, under which this workshop was organized, is supported by the U.S. government.
For additional information regarding the workshop, visit https://www.nationalacademies.org/our-work/interactions-of-biotechnology-with-human-physiology-and-function-a-workshop.
SUGGESTED CITATION National Academies of Sciences, Engineering, and Medicine. 2023. Interactions of Biotechnology with Human Physiology and Function Along the Gut-Brain Axis: Proceedings of a Workshop—in Brief. Washington, DC: The National Academies Press. https://doi.org/10.17226/27014.
