Ribosomal protein
S6--L-glutamate ligase (5ZCT) from Escherichia
coli (strain K12)
Created by: Robert
Fisher
Ribosomal protein S6--L-glutamate ligase (RimK) is a ligase derived from Escherichia coli strain K12 (1). RimK (PDB
ID: 5ZCT) catalyzes the post-translational polyglutamylation of ribosomal
protein S6 (RpsF) in bacteria and also catalyzes the synthesis of poly-α-L-glutamic acid
in an ATP-dependent manner by using L-glutamic acid as a substrate (1). Since
it can catalyze the synthesis of poly-α-L-glutamic acid, scientists can mutate it
to a form which could produce poly-α-L-glutamic acid or RpsF (PDB ID: 6AWB) on
an industrial scale (1, 2). Poly-α-glutamic acid is used in a wide variety of
surgical and pharmaceutical supplies for purposes such as enhancing solubility
and controlling the half-life of drugs (2).
RimK
contains 26-30% β-sheets, 32-34% α-helices, and 10-20% random coils, varying
based on subunit (3). Since RimK contains more α-helices than β-sheets, it is
flexible in places near its active site, allowing it to easily change between
open and closed conformations. The large number of α-helices and β-sheets
also create many hydrophobic interactions, causing RimK to easily associate
into tetramers.
The
asymmetric unit of the crystal structure of RimK consists of two tetramers (1).
Each tetramer is a homodimer of heterodimers, including four subunits which
adopt alternating open and closed conformations, with subunit A in closed
conformation, subunit B in open conformation, and so on (4). The open and closed forms have dramatic
structural differences in region 1, residues 138-154, region 2, residues 157-174, and region 3, residues 204-218 (1). Region 1 is hidden while RimK is in its tetrameric form and does not contact neighboring tetramers, which means
that crystal-packing effects do not play a role in the differences between the
open and closed forms (1). The open form of region 1 has an antiparallel β-sheet
structure. This extended β-sheet expands the ATP-binding and substrate-binding
groove such that it is wider than that of the closed form, regulating the
substrate-binding and catalytic activities of RimK (1). The α-helix in the open form of region 2 is
slightly longer than that of the closed form (1). Additionally, when the two
structures are overlapped, the region 2 α-helices of the open and closed forms are
2 Å apart. The region 3 loop of the open
form contains a short, helix-like structure not present in the closed form (1).
The closed form of region 1 forms a loop in residues 142-151 (1). The region 1 loop interacts with the region 3 loop in the closed form, potentially
stabilizing the structure of the closed form (1).
The
ligands associated with RimK were predominantly used to aid in crystallization
and precipitation of the protein. Sulfate ions form hydrogen-bond with various
amino acids in RimK. Magnesium ions form
metal-coordination bonds with AMP-PNP and various amino acids in RimK. These
charged ions help RimK to precipitate out of solution and to more easily form a
regular crystal structure. As is seen in RimK bound to ADP, (PDB ID: 4IWX) RimK
requires ATP hydrolysis to catalyze poly-α-L-glutamic acid synthesis (4).
AMP-PNP is a nonhydrolyzable substitute for ATP, which is used to prevent RimK
from beginning the poly-α-L-glutamic acid synthesis reaction, thereby keeping
it in a steady, asymmetric conformation that is helpful for crystallizing the
active form of RimK.
RimK
adds an unprotected Glu residue to the C-terminal sequence of L-aspartic acid-L-serine-L-glutamic
acid-L-glutamic acid on RpsF up to four times, although the role of this
modification remains unclear (2). RimK catalyzes ligation by phosphorylating
the carboxyl group of the carboxy-terminal glutamic acid, resulting in
nucleophilic action of the α-amino group of the glutamic acid residue with the
phosphorylated carboxy-terminal glutamic acid on RpsF, and thereafter, the
release of phosphate (2).
Comparing the crystal structure of RimK bound to AMP-PNP to that of RimK bound to ADP reveals how RimK functions at its binding sites. The active site of RimK involves several variable residues important to bind poly-α-glutamate: Ile-270, Glu-268, Tyr-13, and Thr-67 (4). None of these residues move significantly when RimK changes from an open form to a closed form – only the ATP-binding groove is changed significantly (1). Resultingly, the change in binding site availability for ATP does not significantly affect the binding site availability for poly-α-glutamate, so the conformational changes are more likely for regulation than for catalytic activity.
At the ATP binding site, four
amino acids play a significant role in binding ATP. Glu-178 forms a hydrogen bond
with the N6 nitrogen of ATP (3). Asn-213 forms a hydrogen bond with an oxygen
in ATP (3). Asn-213 also forms a metal coordination with a magnesium ion which
is bound to ATP (3). Lys-100 forms a hydrogen bond between its r-group amine
and an oxygen in ANP (3). Ser-212 forms three hydrogen bonds, one with an ATP
oxygen using its amino group, and two with ATP oxygens using its hydroxide
group (3). Several residues, including Arg-203, Arg-189, Asp-248, Asn-262, and
Ser-264 interact with the product of the polymerization reaction as the ATP
binding site changes conformation, pulling the product away from the binding
site in the open conformation, and releasing it in the closed conformation (1, 4).
The
molecular mass of RimK was expected to be 137.2 kDa (2). However, the Protein
Data Bank (PDB) found that the mass of the crystallized structure of RimK is
270.42294 kDa. Expasy, a bioinformatics server consisting of an assortment of
tools for biochemical analysis, instead determined that the molecular weight of
RimK is 264.35586 kDa, slightly less than double the literature provided
estimate (3, 5). The isoelectric point of RimK is theoretically 8.53, which
means that RimK functions at its highest rate while nearly neutral, at around
pH 9 (2, 5).
To
get a meaningful indication for how the structure of RimK determines its
function, it can be compared with putative acetylornithine deacetylase (PDB ID:
3VPB), an enzyme that catalyzes the post-translational alteration of OrfF, (PDB
ID: 3WWL) otherwise known as LysW, which consists of subunits E and F in the structure of putative acetylornithine deacetylase. Much like RimK, putative acetylornithine
deacetylase (ArgX) adds a glutamic acid residue to the carboxy-terminus of a
protein, specifically LysW (1).
PSI-BLAST
allows comparison between the amino acid sequence of one protein and another
similar protein in a quantitative manner, assigning an “E value” based on how
similar the structures are. An E value below 0.05 indicates that two proteins
are similar enough to be compared meaningfully. The E value between the A chain
of RimK and ArgX, is 7e-96, whereas for RimK bound to ADP instead of AMP-PNP it
is 9e-159 (6). Unsurprisingly, RimK bound to ADP is identical to RimK bound to AMP-PNP. However, RimK bound to AMP-PNP is only 29% identical to ArgX 48%
positively matched and 5% of the comparison involve gaps (6). These differences are significant, but still indicate a remarkable degree of similarity – the
large number of positives indicates functional similarity, which is important around
the active sites of the proteins (6).
The
Dali server calculates the difference between proteins based on their tertiary
structure and calculates the differences in intramolecular distances between
them, thereby assigning a “Z-score” based on how their structures differ. A
Z-score above 2 is an indication that two proteins are similar enough to be
compared meaningfully. The Z-score comparing RimK to putative acetylornithine
deacetylase is 31.1, whereas for RimK bound to ADP instead of AMP-PNP it is
41.3 (7). Since RimK bound to ADP is nearly identical to RimK bound to AMP-PNP,
it is unsurprising to see such a high Z-score between them. For RimK and ArgX,
the high Z-score indicates that the two proteins share many structural features – including tetrameric associations and binding sites for ADP.
ArgX
possesses the ability to conjugate acidic amino acids to the carboxy-terminus
of a protein through a similar mechanism to that of RimK. ArgX is bound to zinc
ions, phosphate ions, sulfate ions, ADP, and magnesium ions. Zinc, sulfate, and
magnesium all function to precipitate RimK and to crystallize it (4). The spontaneous
hydrolysis of the ATP which was in the initial incubation solution for ArgX
produces ADP and phosphate. These hydrolysis products are important for the
identification and analysis of the ATP binding site on ArgX (4, 8).
There
are various places in which ArgX and RimK differ significantly in terms of
their structure. ArgX has a singular, short β-strand between residues 35 and 48, whereas RimK has multiple β-strands from residues 35 to 58, corresponding
to the same general area. This results from how an ArgX tetramer must bind two
LysW monomers to function in solution, whereas RimK can function without
binding a separate protein (1, 3). The spatial accommodations ArgX makes to be able to bind LysW include the compression of the area between residues 35 and 48.
While RimK fits three β-strands into this area, ArgX fits only one, so that
LysW can bind and induce a conformational change at the active site which
affects its substrate binding site.
Additionally,
the region between residues 125 and 136 on ArgX includes five binding residues,
whereas the same binding site for RimK between residues 138 and 148 has only
two binding residues (3). The difference of binding residue number is due to
the relative location of the substrate and ATP binding sites in RimK and ArgX.
ArgX has its ATP and substrate binding site close together, so it has more
binding residues in the region between residues 138 and 148 (8). Conformational
changes caused by binding to LysW result in changes to the ATP binding site,
affecting enzyme regulation, as well as changes to the substrate binding site,
affecting enzyme activity. RimK has a larger distance between its ATP and substrate binding sites, such that its conformational changes do not affect
catalytic activity, only regulation (1, 8).