The intricate design of a fundamental filter protein

To truly understand a disease, scientists often must pin down the individual atoms of the proteins involved. A mutation that replaces one amino acid with another can often make the difference between good health and medical problems.

For instance, LDL receptor related protein 2 (LRP2) is a gargantuan 4660 amino acid protein expressed across the animal kingdom, from microscopic eukaryotes to humans. Nicknamed “megalin” because of its mammoth size, LRP2 interests scientists because of its role in blood filtering in the kidney and implications in Alzheimer’s disease outcomes (1,2). But how many mutations in this protein actually cause disease remained unknown until researchers from Columbia University mapped the protein’s atomic structure. Their work led by the nephrologist Jonathan Barasch and the biochemists Anthony Fitzpatrick and Lawrence Shapiro was published recently in the journal Cell  (3). 

LRP2 is a multiligand receptor expressed throughout the body, binding upwards of 75 different ligands, an astonishing variety for a single receptor. It dots the surface of cells by the thousands and is a central player in cellular endocytosis, transporting proteins, hormones, and other biomolecules from the cell surface into a cell’s interior. Megalin works like a claw machine game: The claw, in a closed conformation, lowers down into a basket of stuffed animals, opens up to grasp one, changes shape once again to drop the prize into a chute, then returns to its original shape and position. 

“[LRP2] has a transit, a cycle,” said Stephen Blacklow, a biochemist at Harvard Medical School who wasn’t involved in the research. “Every 25 minutes or so, it goes into the cell and then back out to the surface. This happens whether or not a ligand is bound.” 

Previously, other researchers had tried to determine LRP2’s structure by expressing and purifying the protein in E. coli. No one was able to produce a complete structure this way, and Shapiro theorized that purifying from bacteria caused the protein to misfold. 

Shapiro and Andrew Beenken, a nephrologist in Barasch’s group, decided to purify the protein directly. “Andrew killed hundreds and hundreds of mice in order to recover their kidneys. Each prep that he did was made from about 500 kidneys, so the mice are terrified of him,” said Shapiro. “For a protein of this size, if you try to do it with recombinant overexpression, you're going to get a lot of junk protein. Using the natural source was key. Even though that's an old style, ‘grind-and-bind’ kind of biochemistry, that was the right way to be successful in the project.” 

Using the natural source was key. Even though that's an old style, ‘grind-and-bind’ kind of biochemistry, that was the right way to be successful in the project.
- Stephen Blacklow, Harvard Medical School 

Using the laboriously purified proteins, the team determined the protein’s structure in multiple conformations using cryo electron microscopy. They found that LRP2 switches from open to closed conformations depending on the pH of its environment. When it resides on the cell surface, LRP2 dangles into the extracellular space, which has a pH of about 7.5. At that pH, LRP2 forms a homodimer in a ligand-friendly conformation that can grab molecules that float by. When it grabs ahold of a ligand, LRP2 shuttles into an endosome and transports its cargo into the interior of the cell. 

In an endosome, the pH drops to an acidic 5.2. That pH shift changes megalin’s shape, and the amino acids that tethered the ligand to the receptor burrow into the receptor’s interior, making them inaccessible. The ligand simply releases from the receptor. LRP2 then travels back to the surface for another round of ligand shuttling. 

Critically, the pH-sensitive regions of LRP2 responsible for dropping ligands off are the same ones that cause the most severe diseases when mutated. For example a single point mutation, R3192Q, obliterates all of LRP2’s endocytic functions, resulting in Donnai-Barrow syndrome, a genetic disease that causes proteinuria, forebrain malformations, deafness, and myopia (4). The researchers identified other mutations that happen at pH insensitive parts of LRP2 and noted that those mutations tend to result in what they called nonsyndromic phenotypes such as autism. 

The structure excited the researchers for its implications in potential drug development. One of LRP2’s most bedeviling activities is in the brain. There, LRP2 transports amyloid-β, a misfolded protein implicated in Alzheimer’s disease, out of the brain and into the circulatory system, where it can be filtered out. “That's the positive aspect of what it does in the brain,” said Shapiro. On the other hand, LRP2 also transports tau, another protein involved in Alzheimer’s disease progression, from cell to cell, potentially enhancing tau’s damaging effects. With the details of LRP2’s ligand binding functions now known in exquisite detail, the individual amino acids responsible for targeting tau are targetable for small molecule intervention, according to the researchers. 

For Shapiro, the most satisfying part of the project was curing him of his skepticism. “As a biochemist, often when I hear that something binds 75 different ligands, I just don't believe it. In the case of megalin, I believe it, and I think that our structure gives at least a plausible way that this kind of specific binding could happen.”

References

1. Saito, A., Pietromonaco, S., Loo, A. K. & Farquhar, M. G. Complete cloning and sequencing of rat gp330/"megalin," a distinctive member of the low density lipoprotein receptor gene family. Proc Natl Acad Sci USA  91, 9725–9729 (1994).
2. Chambers, J. C. et al. Genetic loci influencing kidney function and chronic kidney disease. Nat Genet  42, 373–375 (2010).
3. Beenken, A. et al. Structures of LRP2 reveal a molecular machine for endocytosis. Cell  186, 821-836.e13 (2023).
4. Flemming, J. et al. Induced pluripotent stem cell-based disease modeling identifies ligand-induced decay of megalin as a cause of Donnai-Barrow syndrome. Kidney International  98, 159–167 (2020).