Dear Aventine Readers,

This issue we look at a promising new cancer treatment that targets individual cancer cells, reducing some of the collateral damage caused by traditional radiation treatments. Yes, there are hurdles — it involves injecting radioactive material into the body — but early results are promising and advocates believe the treatment could significantly reduce some of the side effects of traditional radiation therapy. 

Also in this issue: An AI system that can conceive of, conduct and write up its own scientific experiments; new building materials that adjust to temperature and the potential to use CRISPR to tweak and perfect us as we age. 

Thanks for reading! 

Danielle Mattoon 
Executive Director, Aventine

Subscribe

Subscribe to our newsletter and be kept up to date on upcoming Aventine projects

There’s a new frontier for cancer treatment: Small, targeted, custom radiation therapies, called radioligands. 

Radiation treatment has been part of the standard toolkit for cancer care since the early nineteenth century. Now, after decades of experimentation, an approach that can deliver radiation with cellular precision is gaining traction as a viable means of treating several forms of cancer in a way that minimizes radiation exposure to the rest of the body. 

The targeted therapies were first tested in the early 2000s and gained traction treating lymphomas (like Non-Hodgkin’s lymphoma). More recently, radioligands have shown success treating prostate and neuroendocrine cancers, raising hopes that the technique can be further developed to treat many other forms of the disease. 

The therapy involves giving a formula through IV injection that contains both radioactive material and a biological marker that enables the radioisotopes — the name for radioactive forms of certain chemical elements — to identify and kill the cancer cells in question. The precision with which radioligands can target cancer cells means that far fewer (though not zero) healthy cells are exposed directly to radiation, an improvement over traditional radiation therapy, which typically exposes several millimeters of tissue to radiation, leading to more so-called toxicity in the body.

"Traditional radiation cancer therapy can be highly effective and has been improved to increase its accuracy, but is still effectively "blind" to what cells it kills, cancerous or otherwise." Radioligands operate differently because they target specific cells, not areas of the body. And because different cancers involve different cells with different surfaces and identifiers, it takes different marker molecules and radioactive materials to target and kill them. This means that specific radioligands need to be created for each and every kind of cancer. 

Early on, when radioligands were available only to treat lymphomas, there was an uphill battle for adoption despite promising early results. This was in no small part because — at $24,000 a dose —  it was the most expensive single-dose drug on the market at the time, and not covered by insurance. Patients were also leery of being injected with a radioactive substance. Now, decades later, new radioligands have been approved by the FDA, including Novartis’s Lutathera, for neuroendocrine cancer, which gained approval in 2018 and its Pluvicto drug for prostate cancer, approved in 2022. Manufacturers cite studies showing that the drugs significantly improve the quality of life for patients due to reduced side effects. Proponents of the treatment hope that such studies, along with the new FDA approvals and the fact that some treatments are now insured, will encourage more uptake in the future. 

There is also hope that some of the limitations around how the treatments have been administered might relax. So far, for example, radioligands have almost exclusively been approved for use after patients have already undergone several other treatments, including rounds of chemotherapy, but there are hopes of gaining approval to administer them earlier in the course of treatment. There is also hope the treatment could be administered orally, without the need for an IV. 

Yet there remain many open questions around the future of radioligands. Will approval to administer them earlier in treatment be granted? Can we find radioligands that successfully treat a variety of cancers?  And can the approach overcome the issues of cost and acceptance that plagued it the first time around? Aventine spoke with five experts about how they view the current and future applications of the therapy in terms of cancer care. 

 Image by Ian Lyman/Midjourney.

A surprisingly simple way to peer inside living creatures. Understanding what’s happening inside a laboratory mouse usually involves medical imaging scans of living creatures or the dissection of dead ones. But a team of researchers affiliated with Stanford University and the University of Texas at Dallas has shown that a new technique allows light to penetrate skin tissue more deeply and thus allow observers to see through to inside organs. The research, published research in Science, describes how rubbing tartrazine, a food dye also known as E102 or Yellow 5 (a common food coloring used in Cheetos and Doritos), makes the skin of mice temporarily transparent. The technique, which involves  massaging the dye into the skin of a hairless mouse, enabled the researchers to easily observe the muscle movements that push food through the body and also identify blood vessels on the surface of the brain. The work will not be particularly useful in humans: It makes tissue transparent only to a depth of about three millimeters, and much of what is interesting to study in humans lies far deeper beneath the surface. Yet the technique is safe, reversible and suitable for use on living creatures, which will likely make it a popular approach for more easily and affordably understanding what happens inside lab animals.

The soup-to-nuts AI scientist conducting its own experiments. There are many, varied steps in the scientific process. Forming hypotheses, designing experiments to test them, performing those experiments, processing data, interpreting the results and writing up papers to describe the work are all essential components, and that’s far from an exhaustive list. Now, a team that includes researchers from the University of British Columbia and the University of Oxford have built an AI system, underpinned by large language models (LLMs), that attempts to perform that process from start to finish. The AI Scientist, as it’s known, uses LLMs to come up with ideas, checks the novelty of those ideas, and — if they pass muster — writes code based on them. It tests the new approaches, records the results of those tests, plots charts and then uses LLMs to write an academic paper describing the results. The system is, admittedly, limited in its scope: So far it conducts research only on machine learning and the results it comes up with are, in the words of the lead researcher, Jeff Clune, “not breakthrough ideas.” It also goes without saying that it can’t conduct any physical experiments right now. But it is nevertheless a tantalizing glimpse at how artificial intelligence could assist scientists in the future, potentially allowing humans to offload simple research to machines while they focus on more complex ideas. 

New types of materials could help keep cities cool. The world is getting hotter, and cities — with their abundance of human-made heat and heat-trapping material — feel the impact more than their rural surroundings. Nature reports on the progress being made in developing a variety of new materials that could help keep urban environments cooler. At their most complex, some of these materials can change their physical properties depending on the temperature: One example, developed at the University of Melbourne in Australia, is a building material that can change its composition at a microscopic level from insulating zigzags at low temperatures to a heat-reflective flat surface at higher temperatures. Other materials are more straightforward, simply being highly reflective of solar radiation (referred to as cool materials) or doing that while also emitting a lot of their own thermal radiation (referred to as supercool materials). Such materials can remain at least several degrees Celsius cooler than the air around them, so they can be used to prevent  buildings, asphalt and other urban surfaces from absorbing and retaining heat. Some materials that exhibit these properties already exist and are low in cost, but uptake is still relatively slow — something that will likely only change when their efficacy is proven in cities, such as Riyadh, which have become early adopters of the technology.

Making ‘Food Out Of Thin Air,’ from Noema
4,200 words, or about 27 minutes

Take a vat full of hydrogen. Add some ammonium, mix in some oxygen and garnish liberally with carbon dioxide. Then feed it all to a carefully selected organism known as a hydrogenotroph, Et voilà: With a little processing, you can create Solein, an all-natural, protein-rich ingredient which, its maker Solar Foods hopes, could help reduce humanity’s reliance on animal protein. This article takes a close look at the company and the nascent industry in which it operates, where precision fermentation is being used to create new types of food from sources such as methane or hydrogen. While you may never chomp down on a Solein steak, the business plan behind these processes is to sneak the ingredient into mass-produced foods like ice cream or mayonnaise to increase the protein content of non-meat foods. Some regulators around the world — notably in Italy and in some U.S. states — discourage the production of  synthetically modified foods, but if the results prove safe, the method represents a compelling route toward a food industry with a lighter carbon footprint. 

A day in the life of the world’s fastest supercomputer, from Nature
2,700 words, or about 18 minutes

How do you simulate the behavior of a human cell from the atoms up, predict changes in weather over the course of a 50-year period or study how entire galaxies evolve over time? For research questions like these, scientists turn to supercomputers, the fastest computers in the world, which crunch through calculations at a ferocious pace. These devices are stupefying in their scale: Frontier, the world’s fastest such computer, located in Oak Ridge, Tennessee, is made up of 50,000 processors, consumes as much energy as 10,000 homes and can perform a quintillion — that’s a billion billion — operations per second. This story attempts to put some of these figures into perspective, digs into the kinds of research being enabled by supercomputers and reveals the fierce competition among researchers to get a chance at using the devices.

Beyond gene-edited babies: the possible paths for tinkering with human evolution, from MIT Technology Review
4,300 words, or about 28 minutes

Back in 2018, a Chinese scientist used CRISPR to edit the genomes of embryos that grew to full-term babies and were born. The researcher went to prison as a result, and the saga served as a stark warning about the ethical swamp surrounding the genetic engineering of humans. But gene editing doesn’t need to be done at the embryo stage of life — it can be administered at any point, including during adulthood — and this story digs into how the technique could be used to provide humans with all sorts of semi-superpowers, from stronger bones to immunity from Alzheimer’s disease. Some of the experts who spoke with MIT Technology Review can see a future in which humans regularly have their genomes tweaked and tinkered with to make improvements to their bodies, speeding up the path of human evolution along the way. There are plenty of  hurdles and ethical questions to contend with — not least of which are the safety and expense of such treatments, as well as the fact that many of the most useful tweaks are currently difficult to deliver. But the prospect of imbuing us all with these kinds of life-enhancements is intoxicating, and not as far off as you might expect.