🧬 New DNA Auctocorrector

Hey there,

We are so back. After a minor hiatus—the first episode of our new season is out on YouTube.

Things got pretty gnarly in the final stages of rendering that episode—so I had to briefly pause the newsletter side of operations while I ironed out those last details.

The good news is—we’ll have a much smoother time rendering future episodes. Episode 1 came out 2 months late so that we can enjoy consistent, monthly new episodes moving forward.

And—incredibly—that new video about DNA replication basically became outdated 48 hours after release when a brilliant new paper came out describing some of the error correction processes inside DNA Polymerase Epsilon.

There’s still a lot of incredible science to cover, so let’s explore the big findings here in the world of biochemistry:

new video (finally)

Meet Your Replisome

In our latest episode, we unpack the incredible complexity of copying your entire genome every single cell division

 🧬 Eukaryotes sure like to make things complicated.

Every time your cells need to divide—they need to copy your entire genome first. We usually learn the basics of how cells do this in school, but that usually focuses on prokaryotic DNA replication.

Eukaryotes like us have a way more complicated replication story for a variety of reasons. Let’s break down a few here.

SPLITTING THAT DNA

I was forced to cut so much from this video in order to get it published in a reasonable amount of time. We glossed over a lot of incredible details to focus on the more entry-level topics. One of my favorite structures I’ve ever animated is the core of the eukaryotic replisome: CMG Helicase.

CMG has it all. It has support structures that help guide parent DNA into its core motor—and channels that simultaneously unwind and split parent DNA into two independent strands. New research shows that splitting the leading and lagging strands might be a more gentle and gradual process than using a single point to break the complementary strands apart.

YOU HAVE THREE POLYMERASES

The biggest change eukaryotes make here is relying on specific variants of DNA polymerase for specific jobs.

DNA Polymerase Epsilon (POLE) docks directly to the replisome’s CMG helicase and just spits out the leading strand with no issues. That simplicity is huge—but scroll down a bit to see how our understanding of POLE is getting stronger and stronger.

Meanwhile—DNA Polymerase Alpha exists entirely to extend the RNA primer made by Primase every time replication needs to start. This only happens once on the leading strand, but has to be constantly repeated on the lagging strand. POLA only prints ~30 nucleotides per primer and therefore does not have any proofreading exonuclease domains. POLE and POLD have the ability to proofread the DNA they are replicating while they synthesize their respective strands. POLA is (potentially) physically restrained by the primosome during primer extension, so that it can’t print errors while making a big enough hook for POLE or POLD to attach to.

In prokaryotes—both leading and lagging strand polymerases are physically linked by a clamp loader complex. While eukaryotes have clamp-loaders as well, there isn’t a lot of evidence suggesting that DNA Polymerase Delta has a physical interaction with the replisome. Instead—POLD independently synthesizes the lagging strand after the primosome adds a primer to it. Some experiments have shown that POLD loses efficiency if it is not close to the core replisome—so there’s still a lot of research to be done as we unravel the complexities of eukaryotic DNA replication.

  | watch the video here  |

thank you parasite, very cool

Maybe Brain Parasites Make Good Drug-Delivery Systems

Some folks will resort to just about anything to beat the blood-brain barrier

 🧠 Gene editing just entered a bold new chapter

One of the toughest problems in medicine right now is getting therapeutic compounds across the blood-brain barrier. Your body is extremely proficient at blocking suspect chemicals from getting into your brain. Large molecule treatments that can help alleviate specific conditions are basically impossible to get across.

So, scientists have been cooking up newer and wilder delivery systems to try and consistently get across this barrier.

And after decades of development—one team appears to have landed on a wild new possibility. Engineering the single-celled brain parasite Toxoplasmosis gondii into being less of a parasite and more of a drug-delivery system.

A new paper in Nature has some pretty wild results from a new trial testing this system. Let’s get into it:

BESPOKE DRUG FACTORY

So yeah—Toxoplasmosis Gondii is a classic protozoan parasite that can infect the nervous tissue of any warm-blooded animal. Toxo exists in the popular imagination as the parasite that can infect you if you mishandle cat litter or any other mammalian waste.

Toxo can cross the blood-brain barrier in its granular form—which is how it ends up being a successful parasite. Once inside nerve cells, Toxo sets up shop and bombards its host with a whole mess of proteins designed to basically allow it to survive. Those proteins can do anything from affecting the metabolism of their host cell—to even crossing the nuclear envelope.

So—if you can figure out a way to make the Toxo parasite benign—then you might just have a solid way to deliver treatments beyond the blood-brain barrier. But how can you get therapeutic compounds inside Toxo in the first place?

DIY DRUG SYNTHESIS

The main large-molecule treatments that get blocked by the blood-brain barrier are chunky proteins. Since Toxoplasmosis Gondii is a living cell—researchers engineered its genome to manufacture therapeutic compounds.

But that’s not far enough. Those proteins also need to get to the correct part of a patient’s nerve cells to actually receive treatment. So—this team of researchers engineered Toxo to synthesize its own proteins with therapeutic ones attached to them. That way, the therapy actually gets to where it needs to go.

Toxo infiltrates nerve cells during a larval stage—so you can’t just fill it up with small-molecule drugs and release it. This method only works with peptide drugs that can be manufactured inside cellular machinery.

The team targeted two proteins and two delivery systems—the GRA16 dense granule protein that can infiltrate a host cell’s nucleus and a common fiber called Toxofilin. They engineered Toxo to produce MePC2 attached to these proteins.

MePC2 is a solid therapy for Rett Syndrome—a debilitating neurodegenerative condition of the brain and neural tissue. Rett Syndrome is a non-inherited genetic disease stemming from a single mutation of a single gene on the X chromosome in early life. Folks afflicted with Rett Syndrome can’t build MePC2 in the right shape—and that causes a cascade of issues that impair development over time.

With this engineered Toxo delivery system—folks afflicted with Rett Syndrome can get access to a constant supply of MePC2 and therefore mitigate that condition.

LONG WAY TO GO

The researchers behind this paper demonstrated that GRA16 did a solid job of delivering MePC2 to brain cells in mice. Of course—this treatment platform is barely entering into the ‘proof of concept’ phase. There’s a lot to refine in order to make sure that engineered Toxo cells don’t also generate some of the negative side effects of wild-type Toxoplasmosis infection.

For now—this simply stands as a triumph for synthetic biology. With gene-editing technology pretty much ubiquitous and relatively cheap as we race through the second decade of the CRISPR era—it’s genuinely exciting to see what new platforms we can develop to safely deliver highly targeted therapies like this.

BETTER OUTLETS FOR THIS SAME STORY

After I finished writing this article—Julia Bauman produced a far stronger video summary of this paper.  Bauman is a PhD student at Stanford and easily one of the top science communicators in the current vertical video meta. Her TikTok page is invaluable for folks trying to keep up with the cutting edge of life sciences research.

| Read the paper here |

no more errors

Human DNA Polymerase Has Autocorrect

A new paper demonstrates how POLE removes mistakes automatically during DNA replication

🎯 How can our cells target and remove DNA errors?

Every time your cells divide—they have to fully copy your entire genome. Your cells divide legitimately billions of times a day and they have to copy ~3 billion base pairs each time they do. With quadrillions of opportunities for catastrophic error—how on Earth does this operation run smoothly?

While there are a lot of well-understood mechanisms that correct errors in DNA replication after the fact—it turns out that our polymerases also have an autocorrect feature that helps prevent DNA errors from happening in the first place.

This is incredibly well-illustrated in a new paper from the Yeeles Lab at the MRC. This is the same team that resolved the initial structures of a human replisome used in my clockwork video.

This new paper shows how DNA Polymerase Epsilon changes conformations to guide new nucleotides into the emerging leading strand of DNA. It also shows how POLE can block RNA nucleotides from entering the active site and remove DNA errors automatically.

Let’s get into it:

REFINING OUR VIEW OF POL-E

The first big finding here is a set of states that show how DNA Polymerase Epsilon opens and closes in order to help new nucleotides join the template DNA of the leading strand. POLE has an ‘open’ and ‘closed’ state it moves between to help push individual nucleotides into the proper shape for binding to template DNA.

The team at the Yeeles Lab also found a third conformation—the ajar state—which appears to help block RNA nucleotides from mistakenly getting added to the growing DNA strand.

The key actor here is the ‘finger’ domain of POLE. Polymerases have three main regions we learn about in school: the ‘fingers’ the ‘thumb’ and the catalytic ‘palm’ where DNA synthesis actually happens.

In POLE—the fingers swivel on a hinge, opening and closing over the active site to guide new nucleotides into the template strand.

But that’s not the only shape this team found.

AUTOCORRECT EVERYTHING

POLE also has an automatic set of conformation changes that kick off when the wrong nucleotide gets added to the template strand.

In this model—a Cytosine nucleotide joined the strand where a Guanine should have instead. Incredibly, the team resolved subsequent structures that show how that erroneous Cytosine letter is pushed away from the template strand and cleaved from the daughter strand by POLE’s exonuclease domain.

Basically—Cytosine really has no business bonding with Thymine. Those two bases can form a kind of hydrogen bond—but since they are both pyrimidines (y’know, the short DNA letters), they form a little kink in the synthesized DNA.

This shape change is enough to push the daughter strand off to the side a little, where it can get picked up by other residues and handed off to the space between the thumb and exonuclease domains. There—the DNA error can get snipped off, allowing POLE to get back to regular DNA synthesis.

This is another one of those beautiful illustrations of how balanced things are at the molecular level. Protein structures are so delicate and malleable that a single, minor shape change can directly set off a cascade of other conformational changes that end up correcting an error and looping back to the original structure. It takes incredible effort to master your understanding of how all these different residues interact with an overall structure—but it is so satisfying once you start making intuitive sense of precisely how a protein’s structure helps to determine its function.

 | Read the paper here  | Animated loop here |