Ancient Firefly Genes Resurrected & Protein Folding Problem Solved?
The Protein Folding Problem…Solved?
DeepMind, based in beautiful London, not far from King’s Cross, announced that their artificial intelligence program, called AlphaFold, had provided “a solution to a 50-year-old grand challenge in biology,” namely, protein folding. That claim is based on results from the 14th Critical Assessment of protein Structure Prediction, or CASP, challenge. The AlphaFold team measured the accuracy of their AI’s predictions via a Global Distance Test, or GDT, that ranges from 0–100.
“In simple terms, GDT can be approximately thought of as the percentage of amino acid residues (beads in the protein chain) within a threshold distance from the correct position,” wrote the AlphaFold team on their blog. “According to Professor Moult, a score of around 90 GDT is informally considered to be competitive with results obtained from experimental methods.” AlphaFold achieved a median score of 92.4 GDT.
For more details on this news, check out their blog post and accompanying ~8 minute video. I also provide news coverage on these findings at the bottom of this newsletter — including some healthy skepticism regarding the new findings, as reported by Business Insider and Futurism. Link to blog & Link to video
Ancient Firefly Genes
Watching fireflies, on warm summer nights, was one of the best parts of my childhood. I loved catching them in jars, but was always careful to let them go before too long — I couldn’t stand the thought of waking up to dead bugs on my night stand. The genes responsible for a firefly’s glow is luciferase, which can emit light in multiple different colors, ranging from green to orange-yellow. In a new study, researchers ‘resurrected’ 100 million year old luciferase genes, from seven ancestral fireflies, and studied their properties in the laboratory. They found that the last common ancestor of modern fireflies emitted a warm, green glow. Science Advances (Open Access). Link
A DNA-Based Vaccine Protects Monkeys
mRNA vaccines may be the sexiest bit of science in 2020 (the protein folding stuff is cool, too), but did you know that DNA vaccines are also a thing? For a new study, researchers gave Cynomolgus macaques (monkeys) a DNA-based vaccine to protect them against Crimean-Congo haemorrhagic fever virus, which is transmitted by ticks and has a case fatality rate up to 40 percent, according to the WHO. The study found that macaques that were given the vaccine “were significantly protected” relative to a control group. Nature Microbiology (Open Access). Link
Fatty Yeast Make Fuel
Yarrowia lipolytica, a “fatty” type of yeast, were engineered to produce both alkanes and alkenes, which are the chemicals used as fuel for jets, boats, trains and cars. The burning of these compounds is responsible for an estimated 24 percent of “hard to remove” greenhouse gas emissions, according to a 2018 study in Science, so making them with synthetic biology — in a renewable manner — is one way to lower global carbon emissions. A new study found that acyl-CoAs are the preferred substrate to produce these chemicals in the yeast. Researchers used that finding to “redirect 89% of acyl-CoAs…to [alkanes and alkenes] and reach titers of 1.47 g/L from glucose.” The researchers also successfully used wheat straw and acetate as carbon sources to produce the fuels. Nature Communications (Open Access). Link
🧫 Rapid-Fire Highlights
More research & reviews worth your time
A brief letter to the editor, at Nature Biotechnology, argues that, as DNA synthesis becomes more common, it could be exploited to create pathogenic organisms. The authors argue that such biothreats are possible because “there are no comprehensive databases of pathogenic sequences, and the guidelines — unenforced outside of US National Institutes of Health (NIH) grantees — are outdated.” Nature Biotechnology (Open Access). Link
CRISPR-Cas9 often basks in the limelight, and for good reason. But have you considered the historical events that propelled the discovery of guide RNAs? An intriguing article, by André Schneider, takes a closer look at these crucial CRISPR components. EMBO Reports. Link
An opinion piece argues that the scientific quest to build synthetic life — that is, a replicating cell, built from chemical components — should be carefully considered and regulated. By creating life in the lab, the authors say that: “This blurring of boundaries may in turn change our views and values towards life.” That statement, at least, is not a new opinion. Trends in Biotechnology (Open Access). Link
Baker’s yeast (S. cerevisiae) were engineered to manufacture “bikaverin, a tetracyclic polyketide with antibiotic, antifungal and anticancer properties” at a titer of 202.75 mg/L using glucose or galactose as the carbon source. Nature Communications (Open Access). Link
Engineers have created a genetic circuit, in photosynthetic cyanobacteria, that can oscillate a gene’s expression with the cell’s circadian rhythm. That is, a gene was turned on and off with the rising and setting of the sun. To build it, the researchers used dCas9 and a guide RNA to construct a NAND genetic logic gate that was hooked up to a circadian promoter (“the purF gene; peak expression at dawn”). ACS Synthetic Biology. Link
George Church’s group has unveiled the ‘Human TFome’, which is a collection of 1,564 transcription factor genes found in human cells. They tested the library of transcription factors in three different types of human pluripotent stem cells, discovering that “290 TFs, including 241 that were previously unreported” were able to induce the cells to differentiate in 4 days. They also used some of the transcription factor genes to program the stem cells to differentiate into neurons, fibroblasts, and other types of cells. Nature Biotechnology. Link
The liver is built from two types of cells: hepatocytes and ductal cells. A new paper explains how to turn human hepatocyte cells into organoids, and also includes instructions for knocking out or modifying multiple genes at once, with CRISPR, in these organoids. Be warned: “The protocols to genome engineer human liver ductal organoids and human fetal hepatocyte organoids take 2–3 months.” Nature Protocols. Link
In a brilliant new study, researchers used a modified version of the FISH method (which stands for fluorescence in situ hybridization) to study the spatial networks of microbes in mouse guts. They found that, when a mouse is treated with antibiotics, the spatial networks of bacteria in the gut are profoundly disrupted. The researchers also used the method to study the microbiome living on human teeth. Nature. Link
Synthetic silks could, one day, be turned into printable ink and loaded into 3D printers at home. Imagine creating an ultra-sturdy, but lightweight, shirt in your living room. That’s (kind of) the topic of this compelling review. Trends in Biotechnology. Link
A huge CRISPR activation screen experiment, using deactivated Cas9, was performed to study which transcription factors — proteins that control a gene’s expression — are involved in determining how a stem cell changes into a neuron. Duke University also published a press release on this study. Cell Reports (Open Access). Link
Researchers have engineered E. coli bacteria to produce dihydroquercetin — a flavonoid found in onions, milk thistle and tamarind seeds — in a one-step enzymatic reaction from glycerol. To do that, they took a gene from Arabidopsis thaliana, a type of weed, and expressed it in the bacterial cells. bioRxiv (Open Access). Link
Cytokines are itty-bitty proteins that act as signaling, or ‘immunomodulation’, compounds. That is, they play a role in the immune system, and tweak how it responds to foreign bacterial or viral invaders, among other things. A new Perspective article takes a look at how synthetic biology can be used to tune and tweak cytokine signaling and, thus, control an immune response. Science (Open Access). Link
Peroxisomes are organelles that float in the cytoplasm of eukaryotic cells. For a new study, researchers engineered peroxisomes, in yeast, to produce geranyl diphosphate — a critical starting point for manufacturing loads of different chemicals, like cannabinoids, camphor, and menthol. PNAS. Link
Take a dozen yeast cells, and let them divide. If you watch very carefully (with a microscope), you will notice that each cell in the population divides at a different time; their division is asynchronous. For a new study, researchers placed genetically-engineered yeast cells into a microfluidic device, attached the device to a microscope, and carefully controlled their growth medium to determine when each cell divides. In other words, they built an automated, robotic system to synchronize yeast cell divisions. Cyber-Yeast! bioRxiv (Open Access). Link
A new Perspective article discusses the myriad biological problems that can be solved, or studied, by constructing genomes from scratch. This piece comes at an intriguing time, especially as researchers wait to hear about final results from the Synthetic Yeast 2.0 project, which aims to build a modified version of the Baker’s yeast genome from scratch. Nature Communications (Open Access). Link
For a new study, researchers used a deep learning model — trained on 20,000 mRNA datasets, taken from seven organisms, including bacteria and humans — to predict the number of mRNA transcripts from a DNA sequence alone. That is, given only a DNA sequence, the researchers can predict how many mRNA copies it will produce. Nature Communications (Open Access). Link
When engineering a cell, researchers often want to pack proteins together, really closely, to speed up a chemical reaction. For a new preprint, researchers created light-inducible cellular compartments. When they shine light on E. coli cells, a desired protein can very quickly form a compartment in the cytosol, enhancing the reaction efficiency. When they stop shining the light, the proteins disperse back into the cytosol after about 15 minutes. bioRxiv (Open Access). Link
📰 #SynBio in the News
There was also a lot of media coverage on AlphaFold’s protein folding announcement. It has been intriguing to see how various news outlets covered the release.
- John Timmer wrote “DeepMind AI handles protein folding, which humbled previous software” for Ars Technica. Link
- Ewen Callaway wrote “‘It will change everything’: DeepMind’s AI makes gigantic leap in solving protein structures” for Nature. Link
- Cade Metz wrote “London A.I. Lab Claims Breakthrough That Could Accelerate Drug Discovery” for The New York Times. Link
- Robert F. Service wrote “‘The game has changed.’ AI triumphs at solving protein structures” for Science. Link
- Ian Sample wrote “DeepMind AI cracks 50-year-old problem of protein folding” for The Guardian. Link
- Will Douglas Heaven wrote “DeepMind’s protein-folding AI has solved a 50-year-old grand challenge of biology” for MIT Technology Review. Link
- Casey Ross wrote “DeepMind’s protein-folding AI stuns with a solution to one of biology’s biggest challenges” for STAT. Link
I slogged through the Forbes 30 Under 30 and it looks like many great scientists made the list. Zibo Chen, from Michael Elowitz’s lab at Caltech, is on the list. So is Maddie Hall, co-founder at Living Carbon (genetically-engineered poplar trees), as well as Basem Al-Shayeb from Jennifer Doudna’s lab at Berkeley.
An amazing result for the Moderna vaccine: Science reports that “No one who got Moderna’s vaccine in trial developed severe COVID-19.” Science. Link
Moderna has submitted its vaccine for emergency approval. Forbes. Link
A participant in the AstraZeneca-Oxford vaccine trial claimed that the vaccine caused neurological issues, according to a report from Forbes. The Serum Institute of India has disputed those claims and is threatening a legal response. Forbes. Link
The UK has already granted emergency approval for the Pfizer/BioNTech COVID-19 vaccine. MIT Technology Review. Link
The Bill & Melinda Gates Foundation has been backing biotechnology companies that aim to use CRISPR to cure sickle cell disease. Future Human. Link
Can synthetic biology help combat aging, and promote longevity? A long explainer from Labiotech.eu. Link
🐦 Tweet of the Week
The DeepMind news is indeed exciting, and it does appear to be an important advancement in predicting folded protein structures from an amino acid sequence. However, a full paper on the findings have not yet been released or critically evaluated by the scientific community. Check out some other opinions on the announcement, and diversify your news intake. 👇
Also this week, a massive new study analyzed all of the mutations in the SARS-CoV-2 virus that could help it ‘escape’ from antibody treatments. Read the brilliant Twitter thread from the Bloom lab. 👇