From its perch on the cell surface, low-density lipoprotein (LDL) receptor-related protein 2 (LRP2) somehow manages to nab ligands, bring them to the endosome, drop them off, and return to the cell surface every 10 to 30 minutes. How does the endocytic receptor do this? According to a cryo-electron microscopy study published January 28 in Cell, pH-dependent shape-shifting drives this endocytic shuttle.

  • Cryo-EM structure of LRP2 receptor solved at different pHs.
  • At neutral pH, LRP2 homodimers receive ligand with open arms.
  • At acidic pH, they give it the cold shoulder.

Researchers led by Anthony Fitzpatrick, Jonathan Barasch, and Lawrence Shapiro at Columbia University in New York reported that on the cell surface at neutral pH, the receptor forms homodimers with an extensive interface that hoists binding domains outward, welcoming passing ligands. In the acidic environs of the endosome, the homodimers ball up, retracting their ligand binding sites and brushing off cargo. The authors propose that these pH-driven acrobatics drive LRP2’s catch and release of ligands, such as apolipoproteins, Aβ, and tau. Other members of this large receptor family might work the same way.

LRP2, aka megalin, is the largest of a family of single-pass transmembrane receptors composed of similar domains (see diagram below). First identified in the kidney, where it nabs small proteins that escape glomerular filtration, LRP2 is also expressed in the CNS. It works as an endocytic receptor, binding upwards of 75 different ligands and internalizing them via clathrin-mediated endocytosis.

LRP2 and its siblings, LRP1 and LDLR, have been implicated in Alzheimer's disease pathogenesis. They bind to ApoE, Aβ, and/or tau (Apr 2013 news; Mar 2020 news; Cooper et al., 2021; Jul 2021 news). Genetic variants near LRP1, LRP2, and other LDLR family member genes have been tied to AD risk (Vargas et al., 2010; Wang et al., 2011). 

All in the Family. Each LDLR member contains different combinations of the same domains. Domain organization between LRP1 and LRP2 is highly conserved. Mouse and human LRP2 are organized similarly. [Courtesy of Beenken et al., Cell, 2023.]

LRP2 comprises a hefty ectodomain, a single transmembrane region, and a short cytoplasmic tail that may aid in recycling of the receptor between different compartments. The ectodomain boasts 36 LDL receptor type A (L) repeats, which are grouped into four clusters bookended by several epidermal growth factor (EGF)-like repeats and a YWTD motif-containing β-propellers. L repeats bind ligands, and the EGF-like repeats and β-propeller regions may coordinate ligand binding as well as shedding.

How does LRP2 reconfigure these dozens of domains to transition from a ligand recruiter on the cell surface to a ligand dumper within the endosome? Suspecting that changes in pH could drive the switch, first author Andrew Beenken and colleagues used cryo-EM to solve and compare the structures of LRP2 isolated from mouse kidney samples at a relatively neutral extracellular pH of 7.5, or an acidic endosomal pH of 5.2. Both structures emerged as homodimers with twofold symmetry.

Open and Shut. LRP2 forms homodimers with twofold symmetry (one homodimer is rainbow-colored; the other is beige). At extracellular pH (left), the homodimer is open for ligand binding, while at acidic pH, it clams up, shunning them. [Courtesy of Beenken et al., Cell, 2023.]

At neutral pH, each molecule of LRP2 within a homodimer was arranged like a U-shaped hairpin, which could be roughly divided into three sections—two “legs” and a central canopy (image above). Starting at the N-terminus, which is positioned close to the membrane but not embedded within it, both ascending arms consist of the first two clusters of L repeats and β-propellers. They rise into the central canopy, where EGF domains and β-propellers meet up, forming the major homodimer interface. Then, descending arms, which comprise the third and fourth cluster of L-repeats and β-propellers, form a column that dives back to the membrane insertion site, where both subunits enter the membrane about 25Å apart.

Ligand Lassos. At neutral pH, the extensive homodimer interface formed by EGF repeats and β-propellers (P) in the central canopy creates a scaffold that allows clusters of L-repeats (labeled R1-R4) to loop (dashed lines) out into solvent in search of ligands. These free-flowing portions of the protein were too disordered to be resolved by cryo-EM. [Courtesy of Beenken et al., Cell, 2023.]

At endosomal pH, the entire structure balls up like a hedgehog under threat. While each subunit still maintains a hairpin shape, most of the homodimer interfaces are drastically smaller, such that the L-repeat clusters fold inward and become tightly packed within the protein, hidden away from would-be ligands. There, the packed clusters of L-repeats buddy up with β-propellers and EGF-like domains within their respective subunits, displacing ligands and turning away from new suiters.

That’s not all. At neutral pH, some parts of the homodimers assumed a free-wheeling structure, with portions of each L-repeat extending like loops out into the solvent (image above). As such, the structure of these dangling portions were murky on cryo-EM, while the rest of the structure was resolved down to 2.83 Å. In contrast, because LRP2 homodimers formed under acidic conditions packed into a more compact, rigid structure, the entire dimer was resolved to 2.97Å.

From Open to Exclusive. At neutral pH (left), LRP2 homodimers form extensive homodimer interfaces (red dashed lines) facilitated by symmetric calcium binding sites (green dots). This interface supports the exposure of ligand binding sites. At endosomal pH, intra-molecular contacts (blue dashed lines) predominate, excluding extracellular ligands. [Courtesy of Beenken et al., Cell, 2023.]

How does pH trigger this radical shift? One mechanism could relate to changes in ion binding in the canopy. Contacts between symmetrically positioned pairs of calcium-binding sites on each subunit cement the dimer together. Changes in pH alter that ion binding, leading to reduction of the homodimer interface (image above). “This liberated surface area on the canopy’s β-propellers can now bind L repeats to displace ligands,” the authors wrote.

Another prominent pH-driven shift in the homodimer structure occurs at the membrane insertion point, where the distance between the two LRP2s expands from 25Å at neutral pH to 140Å at low pH. The authors speculated that this repositioning of the cytoplasmic domains could change how LRP2 binds adaptor proteins that control protein trafficking, and thus could regulate endosomal recycling of the protein.

To explore whether this pH-driven switch matters functionally, the scientists mapped the location of 40 LRP2 mutations known to be associate with Donnai-Barrow syndrome, a disease of both the kidney and brain. Among the 40, the researchers identified a half-dozen missense mutations predicted to disrupt pH-sensitive sites within LRP2, suggesting that this switch is critical for proper LRP2 function.

The authors proposed that similar mechanisms could regulate the function of other LDLR family members, given their related domain organization. Because these receptors bind key culprits in the AD pathogenic cascade, the authors think their findings have implications for understanding this much more common disease.—Jessica Shugart


  1. We’ve known for a long time that members of this receptor family play a critical role in biology. What has been missing is a detailed understanding of how they accomplish their function at the molecular level. Solving the structure of LRP2 represents a significant achievement and offers critical insight into important biological questions.

    In addition to providing mechanistic insight into how ligand uncoupling occurs in the low-pH environment of endosomal compartments, this study reveals two unexpected findings. First, while studies performed on isolated ligand binding clusters from LRP2 have similar ligand binding properties, the data reveal structural-based control of each ligand-binding region, indicating a level of regulation that was not anticipated.

    Second, the pH-induced conformational change results in a dramatic separation of the intracellular domains in the dimer, from the 25 Å at extracellular pH to 140 Å at endosomal pH. This could impact assembly of adaptor molecules involved in trafficking and signaling events. Finally, it is likely that information learned from the structure of LRP2 will be applicable to LRP1 as well, as many of the critical structural elements are also found in LRP1.

  2. In this landmark paper, Andrew Beenken et al. used cryoEM to solve the structure of LRP2, which is part of the LDL-receptor-related family of proteins, at both extracellular and endosomal pH. This family of proteins is able to bind ApoE isoforms, which greatly affect the risk for late-onset AD. Competition between ApoE and other ligands such as tau and Aβ for binding to these receptors is implicated in Aβ plaque deposition and tau seeding and spreading.

    The proper recycling of receptors like LRP1, LRP2, and LDLR is important for clearing Aβ and tau from the ISF. In fact, overexpression of LDLR has been shown to be protective in a mouse model of tau pathology. Similarly, polymorphisms in the promoter region of LRP2, which lowers LRP2 expression, are associated with higher AD risk, possibly due to the decreased clearance of proteins like Aβ and tau. However, despite the importance of these receptors to AD pathology, the molecular mechanisms underlying their function and ligand binding had not been well described until now.

    The structure resolved in this paper greatly alters the model predicted by Alphafold, whereby the unique hairpin folding and dimerization states “overturn” the textbook depiction of lipoprotein receptor family as “tube men” at the cell surface. This study provides a mechanistic description of how the pH changes LRP2 experiences while trafficking from the plasma membrane to the endosome regulate ligand binding and release. The structural model shows how LRP2 mutations associated with Donnai-Barrow Syndrome disrupt pH-sensitive sites that would interfere with the conformational change that LRP2 undergoes when at endosomal pH, preventing intramolecular ligands to bind EGF-like domains in order to release LRP2’s cargo.

    This study now gives us a framework to try to better understand how mutations in other LDL receptor family proteins could be playing a role in AD pathology. Interestingly, SORL1, which is a risk factor for late-onset AD, is an endosomal-sorting protein containing an EGF-like domain and 11 LDLR class A repeats. SORL1 has been shown to be involved in APP processing, and it is intriguing to think that mutations that disrupt cargo release throughout trafficking, in a manner similar to LRP2, could have implications for APP processing and trafficking.

    A critical question that remains is how extramolecular ligands bind receptors like LRP2. This would be important for understanding how ApoE competes with Aβ or tau for binding to LDL-receptor-related receptors. Future studies characterizing this family of proteins and their ligands will expand our understanding of the molecular mechanisms underlying disease risk variants and how LDL-receptor-related proteins are contributing to AD pathogenesis.

    Beyond the AD-related ligands, the structural dependence on pH of those lipoprotein receptors may imply new roles for themselves as well as ApoE and other lipoproteins and molecules involved in lysosomal storage disorders. This paper also provides new thoughts for basic science in this area. One important question is how different lipoprotein receptors modulate their interactions with ligands by changing numbers and arrangement of those conserved domains. This further raises the question how exactly the pH change drives such a huge tertiary structure movement from ligand-binding conformation into the closed conformation at pH 5.2. In line with this paper, the conserved domains of L repeats and β-propellers with Ca2+-coordinating interactions might be the answer. The interactions between L repeats and β-propellers in this structure indicate a universal pH-dependent domain-level sliding that governs the ligand-catch-and-release mechanism in lipoproteins. Different numbers of conserved domains among different lipoprotein receptors seem to have a more profound impact on a variety of superdomain movements that eventually determine the capture-and-release efficiency of subtype lipoproteins. Some of the essential Ca2+ ions would contribute to the electron transfers across different residues. Molecular dynamics paired with CryoEM 3D classification at intermediate pH between 5.2 and 7.4 could be important in  future studies to fully understand the process.

    As the authors discussed, one limitation of this study is the lack of determinations of the intracellular domain. However, it is observable that the last two EGF-like domains experience considerable spatial movements at different pH conditions, which leads to a reasonable prediction that the cytoplasmic domain undergoes substantial confirmational changes because of the former movements. Although the authors encountered technical challenges to visualize the disordered cytoplasmic domain, it would be interesting to complex LRP2 protein with known endosomal sorting proteins such as WASH/CCC family to investigate the pH-dependent sorting system.

Make a Comment

To make a comment you must login or register.


News Citations

  1. ApoE Does Not Bind Aβ, Competes for Clearance
  2. Tau Receptor Identified on Cell Surface
  3. Taming ApoE Via the LDL Receptor Calms Microglia, Slows Degeneration

Paper Citations

  1. . Regulation of tau internalization, degradation, and seeding by LRP1 reveals multiple pathways for tau catabolism. J Biol Chem. 2021 Jan-Jun;296:100715. Epub 2021 Apr 28 PubMed.
  2. . A megalin polymorphism associated with promoter activity and Alzheimer's disease risk. Am J Med Genet B Neuropsychiatr Genet. 2010 Jun 5;153B(4):895-902. PubMed.
  3. . A single nucleotide polymorphism in LRP2 is associated with susceptibility to Alzheimer's disease in the Chinese population. Clin Chim Acta. 2011 Jan 30;412(3-4):268-70. PubMed.

Further Reading

No Available Further Reading

Primary Papers

  1. . Structures of LRP2 reveal a molecular machine for endocytosis. Cell. 2023 Feb 16;186(4):821-836.e13. Epub 2023 Feb 6 PubMed.