In a first-of-its-kind study, researchers demonstrate that functional nervous systems can form within self-organized living cellular robots, conferring complex movement patterns and distinct gene expression profiles
By Benjamin Boettner
(BOSTON) — Biobots, whose growing line of variants started with Xenobots , are fascinating tiny self-powered living robots built exclusively using frog embryonic cells. Originally developed in the laboratories of Wyss Institute Associate Faculty member and Tufts University Professor Michael Levin , Ph.D. and his collaborators at University of Vermont , biobots are remarkably motile, moving autonomously through aqueous environments. Since then, the team has shed light on many exciting properties of biobots, including their ability for kinematic self-replication , and responding to sound stimuli .
Biobots can similarly be constructed using human cells in the form of Anthrobots , which have the ability to heal neural wounds in vitro. Thus, a vision emerged that biobots, made out of patients' own cells, could one day be deployed to repair spinal cord or retinal nerve damage, clear plaques from the arteries, locally deliver pro-regenerative drugs, and perform other vital tasks in the human body. More fundamentally, Levin says, "Such novel beings, exhibiting both new morphology and behavior, despite their wild-type unmodified genome, can reveal important aspects of multicellular plasticity, of relevance to evolutionary biology, bioengineering, and regenerative medicine. They uniquely enable us to investigate questions like 'What is the origin of anatomical and physiological properties in living forms that have no history of selection for those traits?' and 'What determines the range of possible forms, functions, and lifestyles that a given genome can facilitate?'"
Despite their remarkable journey, one important feature that biobots were lacking so far was a "nervous system," which could potentially endow them with novel behavioral phenotypes and additional capabilities.
Now, Levin's team have made an important step in this direction by creating the first "neurobots," which essentially are biobots that with the help of a micro-surgical technique are integrated with neuronal precursor cells and allowed to grow unperturbed. The study shows that novel types of nervous systems self-organize within neurobots with neuronal processes extending in between neurons as well as towards non-neuronal cells lining the surface of the bots. Target cells include multiciliated cells (MCCs) that allow biobots to be motile, mucus-secreting goblet cells which among other functions facilitate ciliary beating, ionocytes that regulate the balance of ions, and small secretory cells (SSCs) that produce MCC-stimulating molecules. The study is published in Advanced Science .
"Importantly, integration of a nervous system reshapes neurobot shape (morphology) and function. Relative to biobots, neurobots are more elongated, exhibit distinct MCC expression patterns, display increased activity and more complex spontaneous behaviors, and undergo substantial changes in global gene expression," said first-author Haleh Fotowat , Ph.D., who spearheaded the development of neurobots with Levin, and is a Senior Scientist at the Wyss Institute.
"This all plays into very fundamental questions that we asked with Haleh at the beginning, namely can a nervous system develop at all in a completely novel context that is not the product of millions of years of natural selection and, if yes, how does it relate to and function within this synthetic biological environment, or even change and augment its responses and behaviors," explained Levin. "Finding answers to these questions has major implications for neuroscience, as well as the bioengineering of organs and tissues, and entirely novel biological entities with programmable functions." At Tufts University , Levin is a Distinguished Professor, Vannevar Bush Chair at the Department of Biology, and Director of the Allen Discovery Center.
The making of a neurobot
To construct neurobots, the team developed an experimental procedure for implanting biobots with neuronal precursor cells during the first minutes of their formation. Biobots are generated from undifferentiated skin tissue excised from embryos of the frog species Xenopus laevis, which has been used for developmental and cell biology research for decades and enabled many fundamental discoveries. During a 30-minute healing period, the excised tissue morphs from a bowl-like structure into a spherical form. The researchers used this transition as a window of opportunity to introduce into the interior of healing biobots undifferentiated neuronal precursor cells that they derived from a separate set of donor embryos. Following the implant procedure, biobot-typical spherical structures formed, which after a day, had completely healed up. After another day, MCCs emerged at their surface and the newly created neurobots started to dance and move around.
"The implanted neuronal precursor cells differentiated into mature neurons with defined cell bodies and axonal and dendritic projections. They connected to one another and extended processes to cells at the surface of the neurobot," said Fotowat. "This all happened spontaneously in a completely novel biological context that we created, one that was different from the way the nervous system normally develop in frogs."
Powered by neurons
The observation of neuronal development in neurobots immediately opened a number of questions. Since biobots' first distinguishing feature was the ability to freely move, they wanted to know whether the added nervous system had an impact on their motility by directly or indirectly stimulating higher ciliary beating frequencies of MCCs. Indeed, neurobots had a more elongated shape and tended to move more actively than their non-neuronal counterparts and, interestingly, exhibited more complex movement patterns that differed considerably from each other while exhibiting repeated motifs of motion.
To investigate whether neural activity could influence spontaneous movement patterns in neurobots, the researchers treated neurobots and biobot controls with a drug that triggers seizures in animals. The drug, called pentylenetetrazole (PTZ), inhibits a type of receptor known as GABAA (Gamma-aminobutyric acid- type A) receptor, that, when activated, dampens the activity of neurons. Thus, inhibiting GABAA shifts neuronal activity into overdrive. "To our surprise, PTZ treatment made non-neuronal biobots less motile suggesting that the drug can impact non-neuronal outer body cells in biobots, which are in fact the same as those making up the body of neurobots. Neurobots on the other hand could either increase or decrease their movement complexity. This finding suggested that at least in some neurobots, removing inhibition can lead to increased activity, which should have overwritten the decreased activity observed in biobots, as neurobots are biobots plus neurons," said Fotowat. "Dissecting how this happens, what the identities of the neurons are, and how neuronal activity affects target cell types on the neurobot surface, will be part of an exciting much deeper future investigation."
As an entry point into this and other questions, the team analyzed the gene expression programs of neurobots, biobots, and an additional non-neuronal control bots made with undifferentiated implanted cells (shams). These comparisons showed that expression of many genes was upregulated in neurobots compared to biobots and shams. As expected, many of these upregulated genes consisted of those important for development of the nervous system. However, much more surprisingly, among the upregulated genes was a large group that encoded significant parts of the molecular machinery that is required for developing the visual system in the eyes of Xenopus frogs and enabling the perception and processing of visual stimuli. "Although we have to validate these observations on the level of proteins, and map them across individual cells, they could mean that some kind of visual system may be developing in neurobots. If so, they could display visually-evoked behaviors such as light-controlled motility, which could be a powerful way to guide their behavior for useful applications, and learn about the evolutionary origin of behavioral competencies." said Levin. Assessing presence of light- and other sensory-evoked behaviors in neurobots is an important next step.
"Biobots, and now neurobots, are the kind of advances that defy scientific thinking and all previously existing paradigms. They present a new frontier in biomedical research with potential for gaining insights into fundamental biology and developing solutions to problems in medicine that can't even be fathomed yet," said Wyss Founding Director Donald Ingber , M.D., Ph.D., who is also the Judah Folkman Professor of Vascular Biology at Harvard Medical School and Boston Children's Hospital, and the Hansjörg Wyss Professor of Biologically Inspired Engineering at Harvard John A. Paulson School of Engineering and Applied Sciences.
Other authors on the study are Laurie O'Neill, Léo Pio-Lopez, Megan Sperry, Patrick Erickson, and Tiffany Lin. The study was supported by the Department of Defense (under awards #HR0011-18-2-0022 and #W911NF1920027), and grants from the John Templeton Foundation and Northpond Ventures.
