In the 1960s, thesequence similarity of several proteases indicated that they were evolutionarily related.[8] These were grouped into thechymotrypsin-like serine proteases[9] (now called theS1 family). As the structures of these, and other proteases were solved byX-ray crystallography in the 1970s and 80s, it was noticed that several viral proteases such asTobacco Etch Virus protease showedstructural homology despite no discernible sequence similarity and even a different nucleophile.[2][10][11] Based on structural homology, asuperfamily was defined and later named the PA clan (by theMEROPS classification system). As more structures are solved, more protease families have been added to the PA clan superfamily.[12][13]
TheP refers toProteases of mixed nucleophile. TheA indicates that it was the first such clan to be identified (there also exist the PB, PC, PD and PE clans).[1]
Above, sequence conservation of 250 members of the PA protease clan (superfamily). Below, sequence conservation of 70 members of the C04 protease family. Arrows indicatecatalytic triad residues. Aligned on the basis of structure byDALI
Surface structure of TEV protease. The C-terminal extension only present in viral members of the PA clan of chymotrypsin-like proteases as(a) surface with loop in blue(b) secondary structure and(c)b-factor putty (wider regions indicate greater flexibility) for the structure of TEV protease. Substrate in black, active site triad in red. The final 15 amino acids (222-236) of the enzyme C-terminus are not visible in the structure as they are too flexible. (PDB:1lvm,1lvb)
Despite retaining as little as 10% sequence identity, PA clan members isolated from viruses, prokaryotes and eukaryotes showstructural homology and can bealigned by structural similarity (e.g. withDALI).[3]
PA clan proteases all share a core motif of twoβ-barrels with covalent catalysis performed by an acid-histidine-nucleophilecatalytic triad motif. The barrels are arranged perpendicularly beside each other with hydrophobic residues holding them together as the core scaffold for the enzyme. The triad residues are split between the two barrels so thatcatalysis takes place at their interface.[14]
In addition to the double β-barrel core, some viral proteases (such asTEV protease) have a long,flexible C-terminal loop that forms a lid that completely covers the substrate and create a binding tunnel. This tunnel contains a set of tight binding pockets such that each side chain of the substrate peptide (P6 to P1’) is bound in a complementary site (S6 to S1’) and specificity is endowed by the large contact area between enzyme and substrate.[11] Conversely, cellular proteases that lack this loop, such astrypsin have broaderspecificity.
Structural homology indicates that the PA clan members are descended from a common ancestor of the same fold. Although PA clan proteases use a catalytic triad perform 2-stepnucleophilic catalysis,[7] some families useserine as thenucleophile whereas others usecysteine.[2] The superfamily is therefore an extreme example ofdivergent enzyme evolution since during evolutionary history, the corecatalytic residue of the enzyme has switched in different families.[15] In addition to their structural similarity,directed evolution has been shown to be able to convert a cysteine protease into an active serine protease.[16] All cellular PA clan proteases areserine proteases, however there are both serine andcysteine protease families of viral proteases.[7] The majority areendopeptidases, with the exception being the S46 family ofexopeptidases.[17][18]
In addition to divergence in their core catalytic machinery, the PA clan proteases also show wide divergent evolution in function. Members of the PA clan can be found ineukaryotes,prokaryotes andviruses and encompass a wide range of functions. In mammals, some are involved inblood clotting (e.g.thrombin) and so have high substrate specificity as well asdigestion (e.g.trypsin) with broad substrate specificity. Severalsnake venoms are also PA clan proteases, such aspit viperhaemotoxin and interfere with the victim's blood clotting cascade. Additionally, bacteria such asStaphylococcus aureus secreteexfoliative toxin which digest and damage the host's tissues. Many viruses express theirgenome as a single, massive polyprotein and use a PA clan protease to cleave this into functional units (e.g.polio,norovirus, andTEV proteases).[19][20]
There are also severalpseudoenzymes in the superfamily, where the catalytic triad residues have been mutated and so function as binding proteins.[21] For example, theheparin-binding proteinAzurocidin has a glycine in place of the nucleophile and a serine in place of the histidine.[22]
Within the PA clan (P=proteases of mixednucleophiles), families are designated by their catalytic nucleophile (C=cysteine proteases, S=serine proteases). Despite the lack of sequence homology for the PA clan as a whole, individual families within it can be identified by sequence similarity.
^de Haën C, Neurath H, Teller DC (February 1975). "The phylogeny of trypsin-related serine proteases and their zymogens. New methods for the investigation of distant evolutionary relationships".Journal of Molecular Biology.92 (2):225–59.doi:10.1016/0022-2836(75)90225-9.PMID1142424.
^Lesk AM, Fordham WD (May 1996). "Conservation and variability in the structures of serine proteinases of the chymotrypsin family".Journal of Molecular Biology.258 (3):501–37.doi:10.1006/jmbi.1996.0264.PMID8642605.
^Dougherty WG, Parks TD, Cary SM, Bazan JF, Fletterick RJ (September 1989). "Characterization of the catalytic residues of the tobacco etch virus 49-kDa proteinase".Virology.172 (1):302–10.doi:10.1016/0042-6822(89)90132-3.PMID2475971.
^Iversen LF, Kastrup JS, Bjørn SE, Rasmussen PB, Wiberg FC, Flodgaard HJ, Larsen IK (April 1997). "Structure of HBP, a multifunctional protein with a serine proteinase fold".Nature Structural Biology.4 (4):265–8.doi:10.1038/nsb0497-265.PMID9095193.S2CID19949043.
