These particles, called extracellular vesicles, are considered attractive vehicle models for new drug therapies. But until now, researchers haven’t had the complete picture of how they work. 

In a new study, a team led by medical researchers at The Ohio State University determined that these vesicles contain an ion channel – a protein that opens a corridor allowing electrical charges to pass through the protective outer membrane, a necessary step to keep contents and conditions stable inside. 

Animal experiments also showed the ion channel influences the cargo, meaning the protein is important not just to the structure of extracellular vesicles (EVs), but also their function. Researchers compared the effects of RNA molecules delivered by EVs with and without the membrane protein to mice with ailing hearts. Only molecules carried by EVs with ion channels were able to repair the heart damage. 

Harpreet Singh, professor of physiology and cell biology, and Mahmood Khan, professor of emergency medicine, both in Ohio State’s College of Medicine, co-led the study. 

“We have not only discovered ion channels in these vesicles. We have recorded functional ion channels for the first time ever,” Singh said. “From forming a simple fundamental hypothesis that these vesicles should have ion channels all the way to showing that these vesicles will contain different cargo that can either protect or harm your cells – in this case, the heart – we have told the whole story.” 

The paper was published Jan. 2 in Nature Communications. 

Extracellular vesicles carry proteins and other molecules from donor to recipient cells to alter physiological and biological responses. In addition to facilitating cellular communication and maintaining cellular balance, the particles have been linked to immune responses, viral infectiousness, and cardiovascular disease, cancer and neurological disorders. 

Based on his specialization in the study of ion channels, Singh predicted that EVs must have ion channels to safely transport molecules from cellular interiors to the extracellular environment and back into another type of cell. Otherwise, their membranes would be subject to bursting – caused by a rush of water triggered by osmotic stress or shock – as positive and negative electrical charges of ions in those varying environments ebb and flow. 

“We know from our experience and from all this great work done in the last hundred years that ion channels are really, really important to maintain any structure which has a membrane,” Singh said. 

Take the electrolyte potassium, for example. It is the most abundant positively charged ion inside cells, but its concentration is 30-fold lower in the extracellular environment. 

“Suddenly an extracellular vesicle is coming from a huge potassium concentration to a low potassium concentration. What is going to happen if you can’t maintain ionic balance? You are going to feel the osmotic shock,” he said. 

For this work, researchers isolated mouse EVs provided by Khan, also director of basic and translational research in the Department of Emergency Medicine, whose lab focuses on repairing damaged heart muscle with stem-cell therapy. 

Because these particles are extremely small, the scientists created a technique they called near-field electrophysiology to record currents in the EV membranes. The method established the presence of a calcium-activated large-conductance potassium channel (BKCa). 

They followed by isolating EVs from normal mice and knockout mice lacking the gene that encodes the BK potassium channel, and found the cargo in EVs from the knockout mice were very different in number and size – suggesting a functional role for the BKCa channel. 

Several small RNA segments that regulate gene activation that were found among the cargo in the normal mouse vesicles were known to help protect the heart against oxidative stress, Khan said. EVs from the mice lacking the BK channel gene contained a different set of these segments, called microRNAs. 

This finding led to the animal experiments in Khan’s lab, where EVs from normal mice and mice lacking the BK gene were injected into mice with diseased hearts. 

“EVs from the wild-type animals protected the heart,” Singh said. “EVs that came out of the knockout mice could not protect the heart properly and, in fact, made things worse. Bad microRNAs were enriched in the vesicles that don’t have the channel. 

“Is the cargo different because of different packaging, or is it because the vesicles without the channels are not surviving? That is an open question, and we are trying to address that.” 

Another chief open question is identifying proteins, called transporters, that enable vesicles to maintain ionic balance as they transition from the extracellular environment back into a cell with a high potassium concentration. 

Besides increasing fundamental knowledge about extracellular vesicles, Singh said, this work has potential to advance development of their use as therapeutics.   

“People talk about loading these vesicles with charged molecules – whether it’s a drug, RNA proteins, or something else. If you’re loading them with charged molecules and you’re not managing ion homeostasis, you will have some sort of consequences,” he said. “That’s our big point, that if you are bioengineering EVs, you have to have the right combination of ion channels and transporters.”

This work was supported by an Ohio State President’s Predoctoral Fellowship, the Department of Physiology and Cell Biology, and a Graduate School Alumni Grant; the American Heart Association; the National Heart, Lung, and Blood Institute; the National Institute of Arthritis and Musculoskeletal and Skin Diseases; and the National Center for Advancing Translational Sciences. 

Additional co-authors are Shridhar Sanghvi, Divya Sridharan, Parker Evans, Julie Dougherty, Kalina Szteyn, Denis Gabrilovich, Mayukha Dyta, Jessica Weist and Lianbo Yu of Ohio State; Sandrine Pierre of Marshall University; Shubha Gururaja Rao of Ohio Northern University; Dan Halm of Wright State University; and Tingting Chen, Panagiotis Athanasopoulos and Amalia Dolga of the University of Groningen.

Traditionally, most drugs are assembled using molecular fragments called alkyl building blocks, organic compounds that have a wide variety of applications. However, because of how difficult it can be to combine different types of these compounds into something new, this method of creation is limited, especially for complex medicines.  

To help solve this issue, a team of chemists report the discovery of a particular type of stable nickel complex, a chemical compound that contains a nickel atom.

Since this compound can be made directly from classic chemical building blocks and is easily isolated, scientists can blend them with other building blocks in a manner that promises access to a new chemical space, said Christo Sevov, the principal investigator of the study and an associate professor in chemistry and biochemistry at The Ohio State University.

“There are really no reactions that can very reliably and selectively construct the bonds that we are now constructing with these alkyl fragments,” Sevov said. “By attaching the nickel complexes to them as temporary caps, we found that we can then stitch on all sorts of other alkyl fragments to now make new alkyl-alkyl bonds.”

The study was published in Nature.

On average, it can take a decade of research and development before a drug can successfully be brought to market. During this time, scientists also create thousands of failed drug candidates, further complicating an already extremely expensive and time-intensive process. 

Despite how elusive nickel alkyl complexes have been for chemists, by relying on a unique merger of organic synthesis, inorganic chemistry and battery science, Sevov’s team found a way to unlock their astonishing capabilities. “Using our tool, you can get much more selective molecules for targets that might have fewer side effects for the end user,” said Sevov. 

According to the study, while typical methods to construct a new molecule from a single chemical reaction can take much time and effort, their tool could easily allow researchers to make upwards of 96 new drug derivatives in the time it would normally take to make just one. 

Essentially, this ability will reduce the time to market for life-saving medicines, increase drug efficacy while lowering the risk of side effects, and reduce research costs so chemists can work to target severe diseases that impact smaller groups, the researchers say. Such advances also pave the way for scientists to study the bonds that make up the fundamentals of basic chemistry and discover more about why these challenging bonds work, said Sevov. 

The team is also already collaborating with scientists at numerous pharmaceutical companies who hope to use their tool to see how it impacts their workflow. “They’re interested in making thousands of derivatives to fine-tune a molecule’s structure and performance, so we teamed up with the pharmaceutical companies to really explore the power of it,” Sevov said. 

Ultimately, the team hopes to keep building on their tool by eventually turning their chemical reaction into a catalytic process, a method that would allow scientists to speed up other chemical reactions by providing an energy-saving way to do so.

“We’re working on making it so much more efficient,” Sevov said. 

Other co-authors include Samir Al Zubaydi, Shivam Waske, Hunter Starbuck, Mayukh Majumder and Curtis E. Moore from Ohio State, as well as Volkan Akyildiz from Ataturk University and Dipannita Kalyani from Merck & Co., Inc. This work was supported by the National Institutes of Health and the Camille and Henry Dreyfus Teacher Scholar Award. 

The field of imageomics aims to help explore fundamental questions about biological processes on Earth by combining images of living organisms with computer-enabled analysis and discovery. 

Wei-Lun Chao, an investigator at The Ohio State University’s Imageomics Institute and a distinguished assistant professor of engineering inclusive excellence in computer science and engineering at Ohio State, gave an in-depth presentation about the latest research advances in the field last month at the annual meeting of the American Association for the Advancement of Science. 

Chao and two other presenters described how imageomics could transform society’s understanding of the biological and ecological world by turning research questions into computable problems. Chao’s presentation focused on imageomics’ potential application for micro to macro-level problems.

“Nowadays we have many rapid advances in machine learning and computer vision techniques,” said Chao. “If we use them appropriately, they could really help scientists solve critical but laborious problems.” 

While some research problems might take years or decades to solve manually, imageomics researchers suggest that with the aid of machine and computer vision techniques – such as pattern recognition and multi-modal alignment – the rate and efficiency of next-generation scientific discoveries could be expanded exponentially. 

“If we can incorporate the biological knowledge that people have collected over decades and centuries into machine learning techniques, we can help improve their capabilities in terms of interpretability and scientific discovery,” said Chao. 

One of the ways Chao and his colleagues are working toward this goal is by creating foundation models in imageomics that will leverage data from all kinds of sources to enable various tasks. Another way is to develop machine learning models capable of identifying and even discovering traits to make it easier for computers to recognize and classify objects in images, which is what Chao’s team did. 

“Traditional methods for image classification with trait detection require a huge amount of human annotation, but our method doesn’t,” said Chao. “We were inspired to develop our algorithm through how biologists and ecologists look for traits to differentiate various species of biological organisms.”

Conventional machine learning-based image classifiers have achieved a great level of accuracy by analyzing an image as a whole, and then labeling it a certain object category. However, Chao’s team takes a more proactive approach: Their method teaches the algorithm to actively look for traits like colors and patterns in any image that are specific to an object’s class – such as its animal species – while it’s being analyzed. 

This way, imageomics can offer biologists a much more detailed account of what is and isn’t revealed in the image, paving the way to quicker and more accurate visual analysis. Most excitingly, Chao said, it was shown to be able to handle recognition tasks for very challenging fine-grained species to identify, like butterfly mimicries, whose appearance is characterized by fine detail and variety in their wing patterns and coloring. 

The ease with which the algorithm can be used could potentially also allow imageomics to be integrated into a variety of other diverse purposes, ranging from climate to material science research, he said.

Chao said that one of the most challenging parts of fostering imageomics research is integrating different parts of scientific culture to collect enough data and form novel scientific hypotheses from them. 

It’s one of the reasons why collaboration between different types of scientists and disciplines is such an integral part of the field, he said. Imageomics research will continue to evolve, but for now, Chao is enthusiastic about its potential to allow for the natural world to be seen and understood in brand-new, interdisciplinary ways. 

“What we really want is for AI to have strong integration with scientific knowledge, and I would say imageomics is a great starting point towards that,” he said. 

Chao’s AAAS presentation, titled “An Imageomics Perspective of Machine Learning and Computer Vision: Micro to Global,” was part of the session “Imageomics: Powering Machine Learning for Understanding Biological Traits.”

The findings in the new study significantly broaden the scope of what scientists can do using artificial intelligence to analyze images of plants, animals and fungi to answer new questions, said Samuel Stevens, lead author of the study and a PhD student in computer science and engineering at Ohio State.

“Our model will be useful for tasks spanning the entire tree of life,” Stevens said. “Researchers will be able to do studies that wouldn’t have been possible before.”

Stevens and his colleagues first curated and released the world’s largest and most diverse machine learning-ready image dataset, TreeOfLife-10M, which contains over 10 million images of plants, animals and fungi covering more than 454,000 taxa in the tree of life. In comparison, the previous largest database ready for machine learning contains only 2.7 million images covering 10,000 taxa. The diversity of this data is one of the key enabling features of their algorithm. 

They then developed BioCLIP, a new machine learning model released to researchers in December and designed to learn from the dataset by using both visual cues in the images with various types of text associated with the images, such as taxonomic labels and other information.

The researchers tested BioCLIP by seeing how well it could classify images as to where they belonged in the tree of life – including a rare species dataset that it did not see during training.  Results showed that it performed 17% to 20% better than existing models on the task.

The study was published on the open-access preprint server arXiv. The BioCLIP model is publicly accessible here. Its demo, said Stevens, can also accurately discern the species of an arbitrary organism image, be it from the Serengeti Savannah, your local zoo or your backyard.

Traditional computational approaches used to organize abundant biology image databases are typically designed for specific tasks and aren’t as capable of addressing new questions, contexts and datasets, Stevens said.

Additionally, because the model can be widely applied to the entire tree of life, their AI is more supportive of biologists whose real-world research is more broadly focused, instead of those studying specific niches, he added.

What makes this team’s approach so effective, said Yu Su, co-author of the study and an assistant professor of computer science and engineering at Ohio State, is their model’s ability to learn fine-tuned representations of images, or being able to tell the difference between similar-looking organisms within the same species and one species mimicking their appearance.

Whereas general computer vision models are useful for comparing common organisms like dogs and wolves, previous studies have revealed that they can’t take note of the subtle differences between two species of the same plant genus. 

Because of its better grasp of nuance, said Su, the model in this paper is also uniquely qualified to make determinations on rare and unseen species as well. 

“BioCLIP covers many orders of magnitude more species and taxa than the previously publicly available for general vision models,” he said. “Even when it has not seen a certain species before, it can come to a reasonable conclusion about how if this organism looks similar to this, then it’s likely that.”

As AI continues to advance, the study concludes, machine learning models like this one could soon become important tools for unraveling biological mysteries that would otherwise take much longer to understand. And while this first iteration of BioCLIP relied heavily on images and information from citizen science platforms, Stevens said future models could be upgraded by including more images and data from scientific labs and museums. Because labs are able to collect richer textual descriptions of species that detail their morphological features and other subtle differences between closely related species, such resources will provide a bevy of important information for the AI model. 

In addition, many scientific labs have information on the fossils of extinct species, which the team expects will also broaden the model’s usefulness.

“Taxonomies are always changing as we update names and new species, so one thing we’d like to do in the future is leverage existing work much more heavily on how to integrate them,” he said. “In AI, when you throw more data at a problem, you’re going to get better results, so I think there’s a bigger version we can continue to train into a larger, stronger model.”

The study was supported by the National Science Foundation, the Ohio Supercomputer Center, and the NSF Imageomics Institute. Other Ohio State co-authors include Jiaman Wu, Matthew J. Thompson, Elizabeth G. Campolongo, Chan Hee Song, David Edward Carlyn, Tanya Berger-Wolf and Wei-Lun Chao. Li Dong from Microsoft Research, Wasila M Dahdul from the University of California, Irvine, and Charles Stewart from the Rensselaer Polytechnic Institute also contributed.