. Crystal structure of a rhomboid family intramembrane protease. Nature. 2006 Nov 9;444(7116):179-80. PubMed.


Please login to recommend the paper.


  1. The 2.1A° resolution crystal structure of a truncated but proteolytically active version of the E. coli GlpG rhomboid protease (named GlpG core domain) provides the first atomic-scale representation of an intramembrane protease. While the requirement for water to accomplish peptide bond hydrolysis is expected, Wang et al. provide the first experimental evidence that the active site, found in a central cavity that contains all the conserved polar residues (including the Ser-his catalytic dyad), accepts a number of water molecules. In this context, the study by Wang et al. is striking.

    Interestingly, a large V-shaped opening between two transmembrane helices (S1 and S3) and facing laterally towards the lipid is proposed to be the route (substrate docking site? substrate binding site?) by which substrate enters the active site. Importantly, the crystal structure described in this report (and resolved in the absence of any substrate) shows that this lateral opening is blocked by a membrane-embedded loop structure (called L1). The authors postulate that L1 functions as a “lateral gate,” which may control substrate access to the active site. The latter observation is consistent with the model proposed by Brunkan et al. in which the membrane-embedded domain that contains the presenilin (PS) endoproteolysis site controls substrate access to the catalytic aspartates of γ-secretase by occluding the active site (1). However, a major difference lies on the observation that γ-secretase activity depends on the processing of full-length PS (FL-PS) into PS-NTF and PS-CTF domains (resulting in accessibility for the substrates to the active site), whereas the rhomboid L1 loop does not require this additional maturation step. Overall, it seems that γ-secretase with the naturally occurring FAD mutation ΔE9, which cannot be matured to PS-NTF and -CTF, better reflects the proposed mechanism of rhomboid-catalyzed intramembrane proteolysis.

    Despite some functional similarities, γ-secretase probably uses different and additional structural motifs to accomplish substrate selectivity and intramembrane proteolysis. For example, the nicastrin large extracellular domain has recently been shown to be essential for recognition of substrate by the γ-secretase complex (2). The structure (and sequence) of GlpG do not display a similar domain. Next, the 3D electron microscopic structure of the purified, proteolytically active γ-secretase revealed that the nicastrin ectodomain covers the top of the large aqueous intramembrane chamber, suggesting a type of flexible lid that could regulate 1) the entry of water molecules into the central chamber and/or 2) the exit of hydrophilic ectodomain products (3). When taken together, it seems unlikely, though it cannot yet be definitively excluded, that the short loop (L5) which tightly caps the GlpG active site from above fulfills similar functions. Co-crystallization of the GlpG core domain with a recombinant substrate would certainly provide a structural explanation for how a substrate can modulate the conformation of the proposed lateral gate, providing a good starting model for how substrates can be differentially handled in the catalytic site of γ-secretase.

    Intriguingly, GlpG forms a trimer in the crystal, leading in the formation of a globular structure with a large aqueous intramembrane chamber similar to the one observed in the 3D structure of γ-secretase (3). If the γ-secretase structure was confirmed at higher resolution, it could suggest that rhomboids can develop an additional strategy for getting water to the catalytic site. Investigating the physiological relevance of the trimer is essential to better understand the stability, function, and regulation of GlpG.

    Finally, and as mentioned by Matthew Freeman (4), γ-secretase, site-2 protease (S2P) and the signal peptide peptidase (SPP) are unrelated to rhomboids by mechanism, sequence, or evolution. Awaited high-resolution structures of these intramembrane proteases will solve the mystery of how they are similar and different.


    . Presenilin endoproteolysis is an intramolecular cleavage. Mol Cell Neurosci. 2005 May;29(1):65-73. PubMed.

    . Nicastrin functions as a gamma-secretase-substrate receptor. Cell. 2005 Aug 12;122(3):435-47. PubMed.

    . Electron microscopic structure of purified, active gamma-secretase reveals an aqueous intramembrane chamber and two pores. Proc Natl Acad Sci U S A. 2006 May 2;103(18):6889-94. PubMed.

    . Structural biology: enzyme theory holds water. Nature. 2006 Nov 9;444(7116):153-5. PubMed.

    View all comments by Patrick Fraering

Make a Comment

To make a comment you must login or register.

This paper appears in the following:


  1. First Crystal Structure of an Intramembrane Protease Revealed