Brain in a dish – the potential of organoid intelligence and biological computing

Intelligence in the dish – is science fiction becoming reality?

In their 2023 article, the authors describe the concept of “intelligence in the dish”, which represents the most promising platform for cultivating and studying cognitive capabilities in a controlled environment outside of a living organism. This innovative platform integrates a high cell density brain organoid onto a cutting-edge three-dimensional (3D) microelectrode array (MEA), that facilitates the signalling and recording of brain organoid activities. The information gathered can then be used to explore cognitive functions within the brain organoid, focusing on learning, long-term memory and decision-making processes.

The overarching vision of this approach is to understand the cognitive capabilities of brain organoids while harnessing their computational potential through their connection to computer systems and networks. This integration could enable researchers to capture, monitor, analyse and interpret the complex data generated by these biological models. By doing so, they could uncover patterns, corelations and insights crucial for understanding how the human brain processes information and responds to stimuli.

To realise this vision, researchers are developing intricate, networked interfaces that connect brain organoids with a variety of sensors and output devices. These future interfaces would allow for real-time monitoring of organoid activities and would involve training brain organoids using feedback mechanisms, where the organoids receive feedback on their activity and adjust their responses accordingly. This iterative learning process helps to refine and enhance the cognitive abilities of the organoids over time.

Since their article was published, the authors have laid the groundwork for establishing a new scientific community dedicated to organoid intelligence as a novel scientific discipline, to advance science, technology and medicine. If the industry is to truly realise the potential of organoid intelligence, it will require a collaborative, multidisciplinary team to achieve success in such an interactive and highly complex area.

Performing tests on lab-grown brain tissue to assess biological cognitive function may seem like something bordering on science fiction. Certainly, this is something the authors of the article consider themselves, comparing the concept to a scene from the 1966 Star Trek TV show, where brains take centre stage as the core of computers. Nevertheless, the model of “intelligence in the dish” described in their article demonstrates that this technology can already be used as a research tool in neuroscience for diagnosing and treating brain diseases.

We spoke to Thomas Hartung, a Professor from Department of Environmental Health Sciences in John Hopkins University, and a leading voice in this field, to understand his perspective on the potential future impact of OI technology and the steps needed to get there.

“As OI is a combination of several disruptive technologies, i.e., bioengineering, AI and sensors, I expect extraordinary fast progress. We should be clear, however, that we are also at very humble beginnings. OI is already allowing us to study aspects of the cellular physiology of learning. We are in reach for testing its usefulness to model diseases and then test candidate drugs or toxicants. Any actual biological computation right now is science fiction, but so were many technologies we enjoy today for a long time”.
Prof. Thomas Hartung

The potential applications of organoid intelligence in healthcare

Organoid intelligence is an incredibly exciting area, with many potential applications in the medical world. This includes as a platform for drug development as well as toxicology testing; as a tool in personalised medicine; and as a human-relevant platform for use in preclinical trials, potentially eliminating the need for animal testing.

Organoid intelligence in drug development

According to Professor Thomas Hartung, the most realistic use for organoid intelligence in the near future will be in drug development.

“The most realistic use is certainly drug development. 25% of all clinical trials are around brain diseases. Having a human cell-based system to help respective drug development or identify problems of brain effects early on is an enormous value proposition. There are also needs to test large numbers of chemicals for risks, such as the impact on the developing brain associated with Autism and Attention Deficit Hyperactivity Disorder (ADHD). Similarly, endocrine disruption of the thyroid system is an unmet testing need and OI-based models could be more sensitive as they employ a very sensitive biological function.”
Professor Thomas Hartung 

Organoid intelligence as an alternative to animal testing

The adaptation of OI research models to neurodegenerative diseases holds tremendous potential in advancing our understanding and treatment of these debilitating conditions. Neurodegenerative diseases, such as Alzheimer’s, Parkinson’s and Huntington’s, present complex challenges due to their multifaceted nature and the limited availability of accurate preclinical models that realistically recapitulate human biology.

Traditional animal models have been valuable tools in studying these diseases, but they often fail to fully translate the complexity of human neurobiology, leading to limitations in the translatability of prospective treatments to human patients. Human-based brain organoid OI models offer a viable alternative to address this limitation, by creating miniature, simplified versions of organs or specific brain regions in the laboratory. These organoids closely mimic the cellular architecture, functionality and pathological features of their respective tissues, providing a more physiologically relevant platform for studying disease mechanisms. Subsequently, these models provide a unique opportunity to study the complexities of the brain’s structure and function in a controlled and non-invasive manner.

By understanding the underlying cellular mechanisms involved in brain function and dysfunction, researchers can identify potential targets for therapeutic interventions. This can lead to the development of new drugs and treatments that can alleviate symptoms, slow disease progression, or even reverse the effects of certain neurological and psychiatric disorders.

Organoid intelligence models can also be used to test the efficacy and safety of novel drug candidates, gene therapies, cell-based therapies and other modes of treatments in a high-throughput manner. By identifying promising therapeutic candidates and optimising treatment strategies in human-derived organoids, researchers can advance the translation of preclinical findings into clinical trials, potentially accelerating the development of effective therapies for neurodegenerative diseases.

Organoid intelligence for personalised medicines

In addition to disease modelling, drug discovery and development, organoid intelligence models hold significant promise for advancing personalised medicine. One of the most compelling aspects of these models is their ability to generate patient-specific organoids from individuals affected by neurodegenerative diseases. These patient-derived organoids can be used to study disease progression, identify personalised therapeutic targets and tailor treatment approaches to individual patients’ genetic backgrounds and disease profiles. This personalised approach allows for a more refined understanding and treatment of these complex conditions, which may not be evident in traditional cell culture or animal models.

“Indeed, the implications of this research extend far beyond the confines of laboratory experimentation.”
Prof. Thomas Hartung.

The science behind brain organoids

In the past decade, there has been a transformative shift in the culture of cells, departing from conventional 2D monolayer techniques to more advanced 3D cultures. The adoption of 3D cultures marks a significant leap forward, allowing researchers to explore the complexities of neural development, connectivity and function in a manner that closely mirrors the workings of the human brain.

Current advancements in 3D cell culture are mainly attributed to the breakthrough in the field of cell reprogramming, which allows any somatic cells (usually skin or blood cells) to be transformed back into induced pluripotent stem cells (iPSC). This was a pivotal moment in the field of organoid research as it removed the confines of limited embryonic stem cell availability. With this barrier removed, scientists could explore a broader range of avenues and accelerate progress in understanding organ development and disease modelling.

With the right ‘cocktail’ of soluble growth factors, the iPSCs can be differentiated into three primary germ layers – the endoderm, mesoderm and ectoderm. To create a brain organoid, scientists first grow a layer of cells called ectoderm in a 3D environment that mimics the natural structure of the brain. Within this setup, the cells multiply and organise themselves into a small, spherical tissue structure called an organoid. The resulting 3D multicellular brain organoid structure bears resemblance to an in vivo embryonic brain, meticulously replicating the essential functional, biological and structural attributes that are characteristic of an early human brain development.

Current limitations to brain organoid development

To date, brain organoid models remain avascular, meaning they lack a functional blood network and predominantly depend on passive diffusion as the primary mechanism for nutrient delivery, as well as cellular waste removal. Over an extended culture period, the absence of a vascular system impedes the delivery of oxygen and nutrients to the innermost regions of the brain organoid, leading to apoptosis (cell death) within the inner zone.

To safeguard against this cell death, microfluidic systems are employed. The microfluidic system processes tiny amounts of liquid by facilitating the delivery of oxygen, nutrients and growth factors to the growing organoid, while efficiently removing waste products. Brain organoids cultured in this integrated microfluidic system allow for the generation of larger size organoids comprised of high-density brain cells, with an increased level of axon myelination. The process of myelination is crucial for the proper functioning of the nervous system. It not only increases the speed of signal transmission, but also provides metabolic support to the axon – the part of the neuron that conducts electrical impulses. The rapid transmission that myelination supports is crucial for supporting and enhancing the efficiency of fundamental cognitive processes, including learning and memory. The longer brain organoids are cultivated in these integrated microfluidic systems, the more they begin to feature greater diversity of mature brain cell types.

Despite these advancements, current brain organoids are not yet able to represent the entire cellular complexity of the developed brain. However, the intricate interplay of this enhanced growth system has still opened up a new avenue for neuroscience research, offering a more sophisticated platform to study and understand the complexities of neural function and communication. With the appropriate tools and the advanced integrated microfluidic system to grow 3D brain organoid cultures, the researchers at Johns Hopkins University initiated the beginning of organoid intelligence and biological computing exploration.

Organoid intelligence and biological computing

“Organoid Intelligence is an interface between living tissue and computer technology whereby brain cell cultures grown into 3D structures, also known as organoids, are integrated into organ-on-chip systems and the resulting output data is interpreted.” 

One of the first challenges for this emerging field lies in defining the appropriate terminology – an important one being biological computing. Biological computing, according to the authors of the Frontiers in Science article, is defined as tasks that are typically carried out by computers being carried out by biological systems instead.

The brain-computer interface device proposed in the article – a form of biological computing – establishes a connection between the brain organoid and the computer through a custom-built 3D microelectrode assay (MEA). The 3D MEA system is configured to integrate the microfluidic system, which not only maintains the homeostasis and viability of the organoid (as discussed above) but also facilitates the delivery of chemicals to support a signalling input. Such an integrated system demonstrates the capability to stimulate and capture electrical signals generated by the brain organoid in response to a variety of stimuli – ranging from chemical to electrical and optical signals.

In this system, the human brain organoid is encapsulated within a flexible, ultra-soft-coated, self-folding electroencephalogram-like shell. This shell is densely covered with tiny electrodes, enabling the recording of the organoid’s activity and the transmission of electric signals across its entire surface, which leads to a higher signal resolution and a superior signal-to-noise ratio. To further enhance signal resolution, the authors also explore the possibility of growing future organoids directly around the implantable electrode. However, they acknowledge the need to balance the effectiveness of these systems against their invasiveness, as potential damage or separation to neuronal networks could influence the behaviour of the organoid.

Companies like MaxWell Biosystems, 3D Brain and Axion Biosystems are currently developing a range of electrode-based 3D MEA systems to study both physiological and pathological functional activities of various type of organoids.

Organoid intelligence and Artificial Intelligence (AI)

Integrating neurons into a digital system is a fascinating and challenging undertaking, that merges neuroscience with computer science. A recent achievement in the field of neuroscience and stem cell research has shown that both rodent and human neuronal culture exhibits a form of biological intelligence, when cultured on high-density (HD) MEA integrated with in silico computing.

In another study, these neuronal cultures showcased their ability to learn and adapt within a stimulated gameplay environment, mimicking the arcade game “Pong” through feedback-driven learning. In this setup, the culture received immediate feedback in the form of white noise whenever the controlled paddle failed to hit the ball. This real-time feedback mechanism aimed to reinforce correct responses and discourage errors, mimicking the principles of reinforcement learning observed in human biological neural networks. Despite this progress within individual sessions, the cultures were unable to retain their learning overnight, requiring them to start fresh each day. This phenomenon suggests limitations in their long-term memory or ability to consolidate learned behaviours. It also highlights the complexity of neuronal learning and memory processes, which are still not fully understood even in simplified neuronal cultures.

During these brain organoid-HD-MEAs experiments, a vast amount of data is recorded from organoid cultures. This extensive recorded dataset, comprised of electrical, chemical and/or optical signals, presents challenges in data handling, storage and analysis due to its complexity. To make sense of these complex datasets and extract meaningful insights, the authors of “Organoid Intelligence (OI): The New Frontier in Biocomputing and Intelligence-in-a-Dish” discuss the possibility of using artificial intelligence (AI), in combination with statistical analysis and machine learning techniques.

AI has the capacity for sifting through and analysing the large datasets generated by organoid-HD-MEAs experiments, to identify hidden patterns, correlations and insights that might be challenging for human researchers to discern. An AI algorithm could also be used to iteratively refine the culture conditions for organoids based on real-time monitoring and feedback. This dynamic optimisation could help to enhance the reproducibility and reliability of organoid-based experiments.

Breakthroughs in interface technology will likely be needed to enhance the ability to interact with and understand organoids. By developing advanced input, output and feedback interfaces, we can pave the way for groundbreaking research in developmental biology, neuroscience and regenerative medicine. These interfaces will not only expand scientific knowledge but also accelerate the development of therapies and treatments for a wide range of diseases and conditions.

The future of organoid intelligence & biocomputing

The concept of “intelligence in the dish” explores the possibility of studying cognitive capabilities in brain organoids within controlled environments. This involves integrating brain organoids with microelectrode arrays, allowing researchers to monitor and analyse their activities. While still in the early stages of research, this approach could lead to significant insights into brain function and behaviour.

What is clear is that there is currently rapid progress and many potential applications for organoid intelligence, particularly in drug development for neurological diseases and toxicity testing. Despite these advancements, many challenges remain, such as the avascular nature of brain organoids and the need for sophisticated interface technologies. Once these limitations are overcome, the industry could soon unlock the full potential of organoid intelligence in advancing neuroscience and healthcare.