A recurrent magnesium-binding motif provides a framework for the ribosomal peptidyl transferase center

Abstract

The ribosome is an ancient macromolecular machine responsible for the synthesis of all proteins in all living organisms. Here
we demonstrate that the ribosomal peptidyl transferase center (PTC) is supported by a framework of magnesium microclusters
(Mg2+-μc's). Common features of Mg2+-μc's include two paired Mg2+ ions that are chelated by a common bridging phosphate group in the form Mg(a)2+–(O1P-P-O2P)–Mg(b)2+. This bridging phosphate is part of a 10-membered chelation ring in the form Mg(a)2+–(OP-P-O5′-C5′-C4′-C3′-O3′-P-OP)–Mg(a)2+. The two phosphate groups of this 10-membered ring are contributed by adjacent residues along the RNA backbone. Both Mg2+ ions are octahedrally coordinated, but are substantially dehydrated by interactions with additional RNA phosphate groups.
The Mg2+-μc's in the LSU (large subunit) appear to be highly conserved over evolution, since they are unchanged in bacteria (Thermus thermophilus, PDB entry 2J01) and archaea (Haloarcula marismortui, PDB entry 1JJ2). The 2D elements of the 23S rRNA that are linked by Mg2+-μc's are conserved between the rRNAs of bacteria, archaea and eukarya and in mitochondrial rRNA, and in a proposed minimal
23S-rRNA. We observe Mg2+-μc's in other rRNAs including the bacterial 16S rRNA, and the P4–P6 domain of the tetrahymena Group I intron ribozyme. It
appears that Mg2+-μc's are a primeval motif, with pivotal roles in RNA folding, function and evolution.

Full text

3134–3142 Nucleic Acids Research, 2009, Vol. 37, No. 10 Published online 11 March 2009
doi:10.1093/nar/gkp119
A recurrent magnesium-binding motif provides
a framework for the ribosomal peptidyl
transferase center
Chiaolong Hsiao and Loren Dean Williams*
School of Chemistry and Biochemistry, Parker H. Petit Institute of Bioengineering and Bioscience,
Georgia Institute of Technology, Atlanta, GA 30332-0400, USA
Received October 23, 2008; Revised February 7, 2009; Accepted February 11, 2009
ABSTRACT
The ribosome is an ancient macromolecular
machine responsible for the synthesis of all proteins
in all living organisms. Here we demonstrate that the
ribosomal peptidyl transferase center (PTC) is sup-
ported by a framework of magnesium microclusters
(Mg
2+
-kc’s). Common features of Mg
2+
-kc’s include
two paired Mg
2+
ions that are chelated by a common
bridging phosphate group in the form Mg2+
(a)–(O1P-P-
O2P)–Mg2+
(b). This bridging phosphate is part of a
10-membered chelation ring in the form Mg2+
(a)–
(OP-P-O5’-C5’-C4’-C3’-O3’-P-OP)–Mg2+
(a). The two
phosphate groups of this 10-membered ring are con-
tributed by adjacent residues along the RNA back-
bone. Both Mg
2+
ions are octahedrally coordinated,
but are substantially dehydrated by interactions with
additional RNA phosphate groups. The Mg
2+
-kc’s in
the LSU (large subunit) appear to be highly con-
served over evolution, since they are unchanged in
bacteria (Thermus thermophilus, PDB entry 2J01)
and archaea (Haloarcula marismortui, PDB entry
1JJ2). The 2D elements of the 23S rRNA that are
linked by Mg
2+
-kc’s are conserved between the
rRNAs of bacteria, archaea and eukarya and in mito-
chondrial rRNA, and in a proposed minimal
23S-rRNA. We observe Mg
2+
-kc’s in other rRNAs
including the bacterial 16S rRNA, and the P4–P6
domain of the tetrahymena Group I intron ribozyme.
It appears that Mg
2+
-kc’s are a primeval motif, with
pivotal roles in RNA folding, function and evolution.
INTRODUCTION
Recent structures of large RNAs, such as the P4–P6
domain of the tetrahymena ribozyme (1–5), and larger
RNAs such as ribosomes (6–14) continue to reveal impor-
tant new principles of RNA structure and folding.
Globular cores of large RNAs are characterized by
base–base tertiary interactions and ‘buried’ phosphate
groups (15,16). During RNA-folding cations are seques-
tered from bulk solvent, and held in close proximity to the
polymer. Folding increases proximities of phosphate
groups, and the electrostatic repulsion among them.
Thus phosphate–phosphate repulsion must be offset by
attraction between phosphates and cations.
Mg
2+
, since the beginning of life, has been closely asso-
ciated with some of the central players in biological
systems—phosphates and phosphate esters (17). Mg
2+
shares a special geometric, electrostatic, thermodynamic
relationship with phosphates (18). The ionic radius of
Mg
2+
is small (0.65 A
˚), the charge density is high, the
coordination geometry is octahedral (the AOCN—aver-
age observed coordination number—is 5.98), the preferred
ligands are charged or neutral oxygens, and the hydration
enthalpy is large (458 kcal/mol) (19–22). Ligand–ligand
crowding is one of the hallmarks of Mg
2+
complexes,
leading to highly restrained ligand–Mg
2+
–ligand geome-
try, and strong ligand–ligand repulsive forces. In compar-
ison with group I ions, Ca
2+
or polyamines, Mg
2+
has a
much greater affinity for phosphate oxygens, and binds to
them with well-defined geometry. Unlike other cations,
Mg
2+
brings phosphate oxygens in its first shell into
direct contact with each other.
Chelation effects and topology influence interactions of
nucleic acids with ions. For example magnesium forms a
mononuclear motif with ADP/ATP in which one Mg
2+
ion is chelated by a six-membered ring consisting
of atoms Mg2þ
ðaÞ–O
a
P-P-O-P-O
b
P–Mg2þ
ðaÞ(18). Here we
observe that a framework of dinuclear magnesium com-
plexes Mg2þ
ðaÞ–RNA–Mg2þ
ðbÞflanks the peptidyl transferase
center (PTC) in LSU ribosomal structures (Figure 1).
In these dinuclear clusters, two Mg
2+
ions are chelated
by a common bridging phosphate group. Additional
*To whom correspondence should be addressed. Tel: +1 404 894 9752; Fax: +1 404 894 7452; Email: loren.williams@chemistry.gatech.edu
ß2009 The Author(s)
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/
by-nc/2.0/uk/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

structural features are conserved among these clusters,
which are also identified in other RNAs, indicating that
dinuclear RNA magnesium clusters compose by distinctive
yet recurrent motif. This dinuclear Mg2þ
ðaÞ–RNA–Mg2þ
ðbÞ
motif is referred to here as the magnesium microcluster
(Mg
2+
-mc).
MATERIALS AND METHODS
Molecular interactions
Ribosomal and other RNA structures were obtained
from the PDB (23) and subject to automated (24–28)
and manual structural analysis and decomposition.
Because its resolution is greatest, the Haloarcula maris-
mortui LSU is used as the benchmark and the basis for
comparison. First, shell Mg
2+
–ligand interactions are
defined by distances less than 2.4 A
˚(18). The Mg
2+
posi-
tions in 23S-rRNA
HM
are, as determined by various geo-
metric criteria, highly credible, with a few exceptions. The
Mg
2+
to oxygen distance is an excellent metric in that the
predicted and observed (in 23S-rRNA
HM
) frequencies
reach distinct maxima at 2.1 A
˚and fall to nearly zero by
2.6 A
˚. The thermal factors of the Mg
2+
atoms in the Mg
clusters do not suggest partial occupancy (i.e. they are not
in general significantly higher than nearby atoms.) As indi-
cated by relative OP (phosphate oxygen atom) positions,
essentially all relevant Mg
2+
ions were added correctly to
Figure 1. Four Mg
2+
-mc’s are observed in the LSU of H. marismortui (PDB entry 1JJ2). (A) View into the Peptidyl Transfer Center. The four Mg
2+
-
mc’s are represented as solid surfaces. The RNA atoms lining the polypeptide exit tunnel are accented in black. Mg
2+
-mc’s D1, D2 and D4 encircle
the PTC. Mg
2+
-mc’s are colored: D1, purple; D2, yellow; D3, gray; D4, green. Ribosomal proteins and the 5S rRNA are omitted for clarity. (B) This
view, looking across the polypeptide exit tunnel, is rotated by 908relative to that of panel A. (C) The secondary structures of LSU rRNAs of H.
marismortui [23S rRNA (7), dashed black line] and the mitochondrion of B. taurus [16S rRNA (51), red line]. Phosphate groups that are linked by
magnesium ions within Mg
2+
-mc’s are indicated by broad colored lines. The secondary structural elements that interact with Mg
2+
-mc’s are
conserved in these widely divergent LSUs. In the C. elegans LSU (not shown), the rRNA that binds to D3 is absent (52). The question mark
indicates the portion of the mitochondrial B. Taurus LSU rRNA for which the secondary structure is unknown.
Nucleic Acids Research, 2009, Vol. 37, No. 10 3135

the 1JJ2 model. An exhaustive survey of all Mg
2+
–RNA
interactions in the database was conducted with the
MeRNA database (29). Hydrogen bond distances are
defined by distances less than 3.4 A
˚between hydrogen
bond donating and accepting heavy atoms.
Sequence homology and alignment
The fragment of ribosomal protein L2 that associates with
the H. marismortui D2 complex (PNVRGVAMNA
VDHPFGGG, peptide L2/D2) was determined by inspec-
tion of the 3D structure (1JJ2). That sequence was used as
a search string in Blast (30), finding matches in all cyto-
plasmic and chloroplast ribosomes and in fungal mitori-
bosomes, but not in other mitoribosomes or in other
chloroplast ribosomes. Exhaustive searching failed to
find L2/D2 homologs in nonfungal mitoribosomes.
Protein alignment was performed with Clustalw (31).
Structural alignment
To date, structural alignment of very large RNAs remains
challenge due to the large size, intricate backbone chore-
ography and tertiary interactions. We developed a heur-
istic method, using a ‘Divide and Conquer’ strategy for
performing whole body superimposition of 23s rRNAs.
With this method (to be described elsewhere), the align-
ment and superimposition of the 23s rRNAs of Thermus
thermophilus and H. marismortui gives an overall RMSD
of atomic positions of 1.2 A
˚. This superimposition utilizes
73% of RNA backbone atoms (around 2129 residues).
This accurate superimposition allows one to identify con-
served Mg
2+
ions.
RESULTS
Forty-one Mg
2+
ions in H. marismortui LSU were deter-
mined by geometric analysis to be highly coordinated by
rRNA phosphate groups [more than one phosphate group
in the Mg
2+
first shell, also see (11)]. Visual inspection of
each of these Mg
2+
ions revealed that four pairs are dis-
tinctive environments, leading to our classification of them
as Mg
2+
-mc’s.
The basic motif of the Mg
2+
-mc is defined by several
features (Figure 2A). Two paired Mg
2+
ions are chelated
by a common bridging phosphate group in the form
Mg2þ
ðaÞ–(O1P-P-O2P)–Mg2þ
ðbÞ. This bridging phosphate is
part of a 10-membered chelation ring in the form Mg2þ
ðaÞ–
(OP-P-O50-C50-C40-C30-O30-P-OP)–Mg2þ
ðaÞ. The two phos-
phate groups of this 10-membered ring are contributed by
adjacent residues along the RNA backbone. Both Mg
2+
ions are octahedrally coordinated, but are substantially
dehydrated by interactions with additional RNA phosphate
groups.
The conformation of the RNA and the positions of the
Mg
2+
ions are substantially conserved in a Mg
2+
-mc. The
position and conformation of the RNA is constrained by
the Mg
2+
ions. The relative positions of the Mg
2+
ions
are constrained by the Mg2þ
ðaÞ–(O1P-P-O2P)–Mg2þ
ðbÞbridges.
The superimposition of four Mg
2+
-mc’s is shown in
Figure 2B. The RMSD of the atomic positions of this
superimposition is 0.59 A
˚.
Although the basic motif is conserved in all four
clusters, Mg
2+
-mc’s are elaborated by additional RNA
ligands. RNA phosphate groups that are not part of the
bridging phosphate or the 10-membered ring can enter the
Mg
2+
coordination sphere at various positions. Each
Mg
2+
-mc is defined in toto by two paired Mg
2+
ions
plus all RNA residues that engage in first shell interactions
with the paired Mg
2+
ions. Therefore each Mg
2+
-mc has a
unique shape. Specific descriptions of the four clusters
are provided here.
Figure 2. The Mg
2+
-mc motif. (A) A schematic diagram illustrating the
Mg2þ
ðaÞ–(O1P-P-O2P)–Mg2þ
ðbÞbridge (outlined in blue), the 10-membered
chelation ring (yellow), an unstacked base and the additional
RNA phosphate groups that enter the Mg
2+
first shell at variable
positions. Carbon is green, oxygen is red and phosphorous is orange.
Magnesium (a) is cyan while magnesium (b) is brown. (B)
Superimposition of four Mg
2+
-mc’s (D1 purple, D2 yellow, D3 gray
and D4 green) from the H. marismortui LSU. All of the atoms shown
here were used in the superimposition of the clusters except for the
RNA bases.
3136 Nucleic Acids Research, 2009, Vol. 37, No. 10

Mg
2+
-kcD2
In this cluster, the Mg2þ
ðaÞ–(O1P-P-O2P)–Mg2þ
ðbÞbridge is
composed of the phosphate group of G877 along with
Mg
2+
ions 8003 and 8013 (Figure 3A). The 10-membered
Mg2þ
ðaÞ–(OP-P-O50-C50-C40-C30-O30-P-OP)–Mg2þ
ðaÞsystem is
composed of the phosphate group and ribose atoms of
A876 along with the phosphate group of G877. Both
phosphates from that segment of RNA chelate Mg
2+
8003.
The two Mg
2+
ions 8003 and 8013 of Mg
2+
-mcD2
jointly link RNA segments that are remote in the linear
sequence and in the secondary structure (32). These lin-
kages take the form of first shell OP
(i)
–Mg
2+
-OP
(j)
, where
j>> i.Mg
2+
8003 links the phosphates of A876 and G877
to the phosphate of A2624. Mg
2+
8013 links the phos-
phate of G877 to the phosphate of G2623. The Mg
2+
-mc
D2 linkages are indicated by the long yellow strips in the
secondary structure in Figure 1C.
The first shell RNA–Mg
2+
interactions unstack the
bases (Figure 3A, bottom panel). Residue A876 is
unstacked from G877, with a C10–C10distance (A876 to
G877) of 7.2 A
˚. Residue A2622 is unstacked from G2623
(r
C1
0
–C1
0
= 7.4 A
˚). This unstacking is important for assem-
bly. The unstacked face of G877 forms part of the cavity
for protein L2 (below). A2622 (unstacked in Mg
2+
-mc D2)
intercalates between U1838 and A1839 (Mg
2+
-mc D4,
below).
Mg
2+
-kcD4
In this cluster the Mg2þ
ðaÞ–(O1P-P-O2P)–Mg2þ
ðbÞbridge is
composed of the phosphate group of U1839 along with
Mg
2+
ions 8005 and 8007 (Figure 3B). There are two
10-membered ring systems. One Mg2þ
ðaÞ–(OP-P-
O50-C50-C40-C30-O30-P-OP)–Mg2þ
ðaÞsystem is composed of
the phosphate group and ribose atoms of U1838 plus the
phosphate group of A1839, along with Mg
2+
8005, which
is also chelated by the phosphate of A1836. A second
Mg2þ
ðaÞ–(OP-P-O50-C50-C40-C30-O30-P-OP)–Mg2þ
ðaÞsystem is
composed of the phosphate group and ribose atoms of
A1839 along plus the phosphate group of A1840. This
RNA segment chelates Mg
2+
8007.
These Mg
2+
–RNA interactions induce an extended
array of unstacked bases (Figure 3B, bottom panel).
Residue U1835 is unstacked from A1836 (r
C1
0
–C1
0
= 9.6 A
˚). Residue A1836 is unstacked from G1837
(r
C1
0
–C1
0
= 6.6 A
˚). Residue G1837 is unstacked from
U1838 (r
C1
0
–C1
0
= 8.8 A
˚). Residue U1838 is unstacked
from A1839 (r
C1
0
–C1
0
= 7.7). Residue A1839 is unstacked
from A1840 (r
C1
0
–C1
0
= 7.8 A
˚). These unstacked bases of
Mg
2+
-mc D4 provide a complementary docking surface
for Mg
2+
-mc D2 (Figure 4). As noted above A2622
(of Mg
2+
-mc D2) intercalates between U1838 and A1839
(of Mg
2+
-mc D4).
Mg
2+
ion 8007 of Mg
2+
-mc D4 links RNA segments
that are remote in the linear sequence and in the secondary
structure. This Mg
2+
ion links the phosphates of A1839
and A1840 and to the phosphate of U832. This linkage is
indicated by the long green strip in the secondary structure
in Figure 1C.
Mg
2+
-kcD1
This cluster has a double bridge. One Mg2þ
ðaÞ–(O1P-P-
O2P)–Mg2þ
ðbÞbridge is composed of the phosphate group
of C2534 along with Mg
2+
ions 8001 and 8002
(Figure 3C). A second Mg2þ
ðaÞ–(O1P-P-O2P)–Mg2þ
ðbÞlinkage
is composed of the phosphate group of A2483, which
joins the same two Mg
2+
ions. The double bridge is
associated with a decreased Mg
2+
–Mg
2+
distance, as
illustrated by the shift of one Mg
2+
relative to its homo-
logs in Figure 2B. The 10-membered Mg2þ
ðaÞ–(OP-P-
O50-C50-C40-C30-O30-P-OP)–Mg2þ
ðaÞsystem is composed of
the phosphate group and ribose atoms of C2533 along
with the phosphate group of C2534. This segment of
RNA chelates Mg
2+
8001.
Both Mg
2+
ions link RNA segments that are not adja-
cent in the linear sequence or secondary structure. The
OPs of A2483 are linked to the phosphates of C2533
(one OP) and C2534 (both OPs) in a joint interaction
involving both Mg
2+
ions. These RNA–Mg
2+
interac-
tions unstack G2482 from A2483 (Figure 3C, bottom
panel; r
C1
0
–C1
0
= 8.3 A
˚). In addition, A2532 is unstacked
from C2533 (r
C1
0
–C1
0
= 6.9 A
˚).
Mg
2+
-kcD3
The Mg2þ
ðaÞ–(O1P-P-O2P)–Mg2þ
ðbÞbridge in this cluster is
composed of the phosphate group of C1679 along with
Mg
2+
ions 8016 and 8029 (Figure 3D). The 10-membered
Mg2þ
ðaÞ–(OP-P-O50-C50-C40-C30-O30-P-OP)–Mg2þ
ðaÞsystem is
composed of the phosphate group and ribose atoms of
A1678 along with the phosphate group of C1679. This
segment of RNA chelates Mg
2+
8016.
Both Mg
2+
ions link RNA segments that are not adja-
cent in the linear sequence or secondary structure. An OP
of A1678 and an OP of C1679 are linked by Mg
2+
8016 to
an OP of A1504. An OP of C1679 is linked via Mg
2+
8029
to an OP of U1503. These RNA–Mg
2+
interactions
unstack the bases (Figure 3D, bottom panel). U1677 is
unstacked from A 1678 (r
C1
0
–C1
0
= 8.0 A
˚). Residue A1502
is unstacked from U1503 (r
C1
0
–C1
0
= 7.7)
Role of Mg
2+
-kc’s in activity
Three Mg
2+
-mc’s (D1, D2 and D4) flank the PTC. These
Mg
2+
-mc’s are not directly involved in catalysis, and do
not form the innermost layer of the PTC, but provide
the framework and supporting structure for RNA that
does. Mg
2+
-mc’s lend critical support to function by form-
ing convoluted binding surfaces and providing rigid fra-
meworks for attachment and buttressing of catalytic
residues.
The fourth cluster (D3) is located near the exit site of
the polypeptide exit tunnel. Previously Steitz and Moore
(11,32) showed that within the LSU, the concentration of
Mg
2+
ions is greatest near PTC. They also described one
‘magnesium cluster’ (part of Mg
2+
-mc D2).
Mg
2+
-kc—ribosomal protein interactions
Mg
2+
-mc D2 is associated with ribosomal protein L2.
Inspection of the structures reveal an 18 amino-acid
loop of L2 (loop-L2/D2, Table 1) forms a binding
Nucleic Acids Research, 2009, Vol. 37, No. 10 3137

Figure 3. Atomic level representations of the four Mg
2+
-mc’s of H. marismortu LSU (PDB entry 1JJ2). The bridging phosphates group are outlined
in blue and the 10-membered chelation rings are shaded in yellow. (A) Top: Mg
2+
-mc D2 with bases, riboses (except C30,C4
0,C5
0) and protein
sidechains omitted (Mg
2+
ions 8003 and 8013). A three residue fragment of ribosomal protein L2 contains universally conserved asparagine (i) and
methione (h). Bottom: Mg
2+
-mc D2 with bases, riboses and protein sidechains included. (B) Top: Mg
2+
-mc D4 with bases and riboses omitted (Mg
2+
ions 8005 and 8007). This Mg
2+
-mc has two 10-membered chelation rings. Bottom: Bases and riboses are included. (C) Top: Bases and riboses
omitted (Mg
2+
ions 8001 and 8002). This Mg
2+
-mc has two Mg2þ
ðaÞ-O1P-P-O2P-Mg2þ
ðbÞbridges (phosphates of residues C2534 and A2483). Bottom:
Bases and riboses are included. (D) Top: Mg
2+
-mc D3 with bases and riboses omitted (Mg
2+
ions 8016 and 8029). Bottom: Bases and riboses are
included. First-shell Mg
2+
contacts are black solid lines. Hydrogen bonds are dashed lines. Carbon is green, oxygen is red, phosphorous is orange
and magnesium is yellow. Residue labels in parentheses are from T. Thermophilus, using the E. coli numbering scheme.
3138 Nucleic Acids Research, 2009, Vol. 37, No. 10

pocket for Mg
2+
-mc D2. As shown in Figures 3A and 4,
methionine (h) and valine (c) along with the backbone
carbonyl of an alanine (g) form a cleft for A875
(A782
TT
). Asparagine (i) and the carbonyl oxygen of ala-
nine (g) bind to the first shell water molecules of the Mg
2+
ions. Histidine (m) forms part of the tightly packed core
of loop-L2/D2 (data not shown). The conformation of
loop L2/D2 is highly conserved between T. thermophilus
and H. marismortui (Figure 4). The RMSD of atomic
positions of loop-L2/D2 is 0.6 A
˚in the globally
Figure 4. The complex formed by Mg
2+
-mc’s D4 and D2 and the loop-L2/D2 of ribosomal protein L2. Structures of both H. marismortui and
T. Thermophilus are shown. Magnesium ions are indicated by spheres. When the 23S rRNAs of H. marismortui and T. thermophilus are superimposed
using 73% of rRNA backbone atoms, the RMSD of atomic positions of the eight Mg
2+
ions within the four Mg
2+
-mc’s is very small, only 0.4 A
˚.
Table 1. Mg
2+
-mc D2-binding loop of ribosomal protein L2 (loop-L2/D2)
a
Amino-acid sequence
b
Species Code NNNNNNNNNNNNNNNNabcdefghijklmnopqrNNNN
c
Homo sapiens EAW82048.1 KAGRAYHKYKAKRNCWPRVRGVAMNPVEHPFGGG-NHQ
Caenorhabditis elegans NP_507940.1 KAGRSYHKYKAKRNSWPRVRGVAMNPVEHPHGGG-NHQ
Saccharomyces cerevisiae P05736.3 KAGRAFHKYRLKRNSWPKTRGVAMNPVDHPHGGG-NHQ
Haloarcula marismortui
d
AAA86862.1 KAGNKHHKMKARGTKWPNVRGVAMNAVDHPFGGG-GRQ
Escherichia coli BAE77974.1 KAGAARWRGVR- - - –PTVRGTAMNPVDHPHGGGEGRN
Thermus thermophilus
e
AAS81667.1 KAGRSRWLGRR- - - –PHVRGAAMNPVDHPHGGGEGRA
Arabidopsis thaliana-chloro
f
NP_051123.1 RAGSKCWLGKR- - - –PVVRGVVMNPVDHPHGGGEGRA
Saccharomyces cerevisiae-mito
f
NP_010864.1 KAGRSRWLGIR- - - –PTVRGVAMNKCDHPHGGGRGKS
Homo sapiens-mito
g
NP_057034.2 KAGRNRWLGKR- - - –PNSGRWHRKGGWAGRKIRPLPP
a
The eukaryotic equivalent of ribosomal protein L2 is L8. The mitochondrial equivalent of L2 is rml2. Blue text indicates conserved sequences in all
ribosomes including mitoribosomes. Red text indicates conserved sequences in all cytoplasmic and chloroplast ribosomes and in fungal mitoribosomes,
but not in other mitoribosomes. The ordering of this table was obtained from the complete L2/L8/rml2 sequence alignment with ClustalW (31).
b
Loop-L2/D2 is bold. Observed sequence changes of loop-L2/D2 are conservative.
c
Positions of loop-L2/D2 are defined by amino acid positions a–r.
d
Loop-L2/D2 contains ribosomal protein L2 residues 187–204 in H. marismortui.
e
Loop-L2/D2 contains ribosomal protein L2 residues 219–226 in T. thermophilus.
f
Loop-L2/D2 contains ribosomal protein rml2 residues 331–348 in S. cerevisiae. Mitochondrial and chloroplast ribosomes are thought to have
undergone major remodeling (51,53,54) and are the most divergent from other ribosomes.
g
The H. sapiens mitoribosome lacks loop-L2/D2 as do other non-fungal mitoribosomes.
Nucleic Acids Research, 2009, Vol. 37, No. 10 3139

superimposed LSUs (H. marismortui versus T. thermophi-
lus, using all atoms of 18 residues except for differing
sidechains for the RMSD calculation, 110 atoms total).
Mg
2+
-kc’s in other RNAs
As shown in the Supplementary Data, Mg
2+
-mc’s are
observed not only in the 23S rRNA, but also in the bacte-
rial 16S rRNA (8), the P4–P6 domain of the tetrahymena
Group I intron ribozyme (2). A related magnesium cluster
is found in Group II intron ribozyme (33).
DISCUSSION
Mg
2+
-kc structure
Four Mg
2+
-mc’s (Microclusters D1, D2, D3 and D4;
Figure 1) are found within the LSU of H. marismortui
[an archaebacterium, PDB entry 1JJ2 (7)]. These four
Mg
2+
-mc’s are also found within the LSU of T. thermo-
philus [a bacterium, PDB entry 2J01 (13)]. The Mg
2+
-mc’s
are highly conserved in position, conformation and inter-
actions between the two LSUs. Mg
2+
-mc’s are defined by
common features (Figure 2A) including (i) chelation of
two Mg
2+
ions by a common bridging phosphate in
the form of Mg2þ
ðaÞ–(O1P-P-O2P)–Mg2þ
ðbÞ, (ii) chelation of
one of these Mg
2+
ions by phosphate groups of adja-
cent residues in the form of Mg2þ
ðaÞ–(OP-P-
O50-C50-C40-C30-O30-P-OP)–Mg2þ
ðaÞ, (iii) unpaired and
unstacked bases with non-canonical RNA conformations
and (iv) and close proximity to regions of function.
Mg
2+
-mc’s are unique structural entities with rigidity
and forced dispositions of functional groups that would
be difficult to achieve by RNA alone or by RNA in asso-
ciation with group 1 cations. Mg
2+
-mc’s by nature of their
Mg2þ
ðaÞ–(O1P-P-O2P)–Mg2þ
ðbÞlinkages impose unusual con-
straints on RNA conformation and force unstacking of
bases. The bridging phosphate is, like all phosphates,
restricted to tetrahedral geometry. The ligands of Mg
2+
ions are restricted to octahedral geometry. Mg
2+
-mcis
composed of two octahedra linked by a tetrahedral phos-
phorous atom. Therefore, the core of each cluster is geo-
metrically rigid and tightly packed. In addition, it can be
seen that the RNA of Mg
2+
-mc’s form intricate and con-
voluted surfaces (Figure 4), in part facilitating their asso-
ciation with other RNA and with protein.
Ordinarily, the conformational space accessible to RNA
is rather limited, and is driven by stacking interactions
and double-strand formation. Noller (34) has suggested
that small molecules can extend the repertoire of RNA
structure, and may have performed that role during
early evolution. Magnesium, like a small molecule, can
drive RNA into unusual conformation states (11,18).
Mg
2+
-mc’s demonstrate not only Mg
2+
-driven deviation
from canonical stacked conformations, but show how
these altered states increase surface undulation, and facil-
itate highly specific interactions, such as those between
Mg
2+
-mc D2 and Mg
2+
-mc D4 and those between
Mg
2+
-mc D2 and loop-L2/D2 of ribosomal protein L2
(Figure 4).
RNA–protein interactions
Mg
2+
-mc D2 is associated with ribosomal protein L2 in
the LSU. Mutations in ribosomal protein L2 degrade ribo-
some activity (35–38). The amino-acid sequence of the
N-terminal region of ribosomal protein L2 appears to be
among the most highly conserved in the phylogenic tree
(Table 1). An 18 amino-acid loop of L2 (loop-L2/D2)
forms a binding pocket for Mg
2+
-mc D2. Ten amino-
acid residues of loop-L2/D2 are universally conserved in
all cytoplasmic and chloroplast ribosomes and in fungal
mitoribosomes. Mutations of other residues of loop-L2/
D2 are infrequent and are between analogous amino acids
such as aspartic acid and glutamic acid or between valine
and threonine.
Prior to the availability of ribosome structures in 2000,
a direct role for rL2 in catalysis appeared to be consistent
with phylogenic, mutagenesis and biochemical data. But a
catalytic role for L2 is ruled out by the realization that the
ribosome is a ribozyme (10,39,40). The strict sequence
conservation of loop-L2/D2 appears to arise from a
requirement for complementarily of the L2 protein surface
with that of Mg
2+
-mc D2 (Figures 3A and 4). Because its
mutation knocks out PTC activity, His(m) (i.e. histidine
residue m of loop L2/D2, Table 1) was previously thought
to be part of a serine-protease like charge-relay system. It
now appears that the importance of His(m) arises from
stabilization of loop-L2/D2 in the appropriate conforma-
tion for association with Mg
2+
-mc D2. The conformation
of loop L2/D2 is highly conserved between T. thermophi-
lus and H. marismortui (Figure 4).
Microclusters in other RNAs
Mg
2+
-mc’s are observed in other ribozymes. A Mg
2+
-mc,
shown in the Supplementary Data, is observed in
the P4–P6 domain of the tetrahymena Group I intron
ribozyme (2). A Mg
2+
-mc in the 16S rRNA of T. thermo-
philus [PDB entry 1FJG (41)], appears to be disrupted
upon ribosomal assembly [PDB entry 2J00 (13)]. The
SARS s2m RNA described by Scott contains a pair of
Mg
2+
ions linked by a single phosphate (42), but those
Mg
2+
are otherwise fully hydrated, with no additional
RNA ligands and so do not constitute a Mg
2+
-mc.
Reasonable coordination geometry in a Mg
2+
-mc
results in Mg
2+
-Mg
2+
distances of 5.3–5.6 A
˚(except
for the doubly bridged D1 cluster, with a Mg
2+
–Mg
2+
distance of 4.7 A
˚). Mg
2+
-mc’s differ from the Mg
2+
com-
plexes proposed in the two-Mg
2+
catalyzed phosphoryl-
transfer mechanism (43,44). In those complexes, a phos-
phate oxygen (not a phosphate group) bridges two Mg
2+
ions, with a Mg
2+
–Mg
2+
distance of 3.9 A
˚.
Evolution
Our discussion here makes use of the basic elements of the
comparative approach (45) where history is reconstructed
by degree of similarity between homologous structures
or sequences. Biologists have long used macroscopic
structure (skeletal, cellular, etc.) to infer phylogenetic rela-
tionships. The logic of that approach extends to the
level of macromolecular structure. Comparison between
3140 Nucleic Acids Research, 2009, Vol. 37, No. 10

homologous macromolecular structures can provide infor-
mation on very distant evolutionary events, because
macromolecular structure changes more slowly than
sequence over evolutionary time (46,47).
The Mg
2+
-mc’s in the LSU appear to be highly con-
served over evolution, since they are unchanged in bacte-
ria (T. thermophilus) and archaea (H. marismortui), which
are thought to have diverged at the LUCA, several billions
of years ago (48). Mg
2+
-mc’s are found in association with
the most conserved rRNA secondary structures
(Figure 1C). Mg
2+
-mc’s link these conserved secondary
structural elements via ‘electrostatic tertiary interactions’,
which are composed of phosphate-Mg
2+
-phosphate inter-
actions. As shown in Figure 1C, the 2D elements of the
23S rRNA that are linked by Mg
2+
-mc’s are conserved
between the rRNAs of bacteria, archaea and eukarya
(49) and the LSU rRNA of mitochondria (50,51), and in
a proposed minimal LSU rRNA (52). The exception is
Mg
2+
-mc D3, which has been dispensed of in some mitor-
ibosomes (such as that of Caenorhabditis elegans) by con-
version of the RNA-based polypeptide exit tunnel to a
protein-based tunnel (51).
We suggest that Mg
2+
-mc’s represent a primeval motif,
with roles in RNA folding and catalysis that have found
utility over deep evolutionary history. The Mg
2+
-mc’s are
located at functionally important regions of the ribosome,
and are associated with unusual conformational states of
RNA, forming intricate binding surfaces.
Mg
2+
-mc’s D1, D2 and D4 but not loop-L2/D2 are
conserved in all mitoribosomes. Mitoribosomes have
been substantially remodeled over time, as can be seen
by comparison of extant mitoribosomes with those of
the ancestral endosymbiont (51,53,54). Mitoribosomes
have twice the protein and half the rRNA of the bacterial
ribosome. Although rRNA secondary elements that con-
tain Mg
2+
-mc’s D1, D2 and D4 are conserved all mito-
ribosomes, loop-L2/D2 appears to be absent from
mitoribosomes other than those of fungi (Table 1). In
the human mitoribosome protein L8 (the mammalian
equivalent of L2), the N-terminus has been replaced by
a sequence that diverges widely from loop-L2/D2
(Table 1). It may be that loop-L2/D2 has been structurally
replaced by a nuclear encoded protein with unrelated
sequence.
SUMMARY
The ribosomal peptidyl transferase center (PTC) is sup-
ported by a framework of magnesium microclusters
(Mg
2+
-mc’s). Mg
2+
-mc’s are characterized by direct phos-
phate chelation of two magnesium ions in the form of
Mg2þ
ðaÞ–(O1P-P-O2P)–Mg2þ
ðbÞ, phosphate groups of adjacent
RNA residues as first-shell ligands of a common Mg
2+
ion, and undulated RNA surfaces with unpaired and
unstacked bases.
SUPPLEMENTARY DATA
Supplementary Data are available at NAR Online.
ACKNOWLEDGEMENTS
The authors thank Drs Roger Wartell, Steve Harvey and
Nicholas Hud for helpful discussions.
FUNDING
This work was funded in part by the NASA Astrobiology
Institute. Funding for open access charge: NASA
Astrobiology Institute.
Conflict of interest statement. None declared.
REFERENCES
1. Basu,S., Rambo,R.P., Strauss-Soukup,J., Cate,J.H., Ferre-
D’Amare,A.R., Strobel,S.A. and Doudna,J.A. (1998) A specific
monovalent metal ion integral to the AA platform of the RNA
tetraloop receptor. Nat. Struct. Biol.,5, 986–992.
2. Cate,J.H., Hanna,R.L. and Doudna,J.A. (1997) A magnesium ion
core at the heart of a ribozyme domain. Nat. Struct. Biol.,4,
553–558.
3. Cate,J.H., Gooding,A.R., Podell,E., Zhou,K., Golden,B.L.,
Kundrot,C.E., Cech,T.R. and Doudna,J.A. (1996) Crystal structure
of a group I ribozyme domain: principles of RNA packing. Science,
273, 1678–1685.
4. Cate,J.H., Gooding,A.R., Podell,E., Zhou,K., Golden,B.L.,
Szewczak,A.A., Kundrot,C.E., Cech,T.R. and Doudna,J.A. (1996)
RNA tertiary structure mediation by adenosine platforms. Science,
273, 1696–1699.
5. Juneau,K., Podell,E., Harrington,D.J. and Cech,T.R. (2001)
Structural basis of the enhanced stability of a mutant ribozyme
domain and a detailed view of RNA–solvent interactions. Structure,
9, 221–231.
6. Cate,J.H., Yusupov,M.M., Yusupova,G.Z., Earnest,T.N. and
Noller,H.F. (1999) X-ray crystal structures of 70S ribosome
functional complexes. Science,285, 2095–2104.
7. Ban,N., Nissen,P., Hansen,J., Moore,P.B. and Steitz,T.A. (2000)
The complete atomic structure of the large ribosomal subunit at
2.4 A
˚resolution. Science,289, 905–920.
8. Wimberly,B.T., Brodersen,D.E., Clemons,W.M. Jr., Morgan-
Warren,R.J., Carter,A.P., Vonrhein,C., Hartsch,T. and
Ramakrishnan,V. (2000) Structure of the 30S ribosomal subunit.
Nature,407, 327–339.
9. Harms,J., Schluenzen,F., Zarivach,R., Bashan,A., Gat,S.,
Agmon,I., Bartels,H., Franceschi,F. and Yonath,A. (2001) High
resolution structure of the large ribosomal subunit from a meso-
philic eubacterium. Cell,107, 679–688.
10. Yusupov,M.M., Yusupova,G.Z., Baucom,A., Lieberman,K.,
Earnest,T.N., Cate,J.H. and Noller,H.F. (2001) Crystal structure of
the ribosome at 5.5 A resolution. Science,292, 883–896.
11. Klein,D.J., Moore,P.B. and Steitz,T.A. (2004) The contribution of
metal ions to the structural stability of the large ribosomal subunit.
RNA,10, 1366–1379.
12. Berk,V., Zhang,W., Pai,R.D. and Cate,J.H. (2006) Structural basis
for mRNA and tRNA positioning on the ribosome. Proc. Natl
Acad. Sci. USA,103, 15830–15834.
13. Selmer,M., Dunham,C.M., Murphy,F.V.t., Weixlbaumer,A.,
Petry,S., Kelley,A.C., Weir,J.R. and Ramakrishnan,V. (2006)
Structure of the 70S ribosome complexed with mRNA and tRNA.
Science,313, 1935–1942.
14. Voss,N.R., Gerstein,M., Steitz,T.A. and Moore,P.B. (2006) The
geometry of the ribosomal polypeptide exit tunnel. J. Mol. Biol.,
360, 893–906.
15. Tinoco,I. Jr and Bustamante,C. (1999) How RNA folds. J. Mol.
Biol.,293, 271–281.
16. Brion,P. and Westhof,E. (1997) Hierarchy and dynamics of RNA
folding. Annu. Rev. Biophys. Biomol. Struct.,26, 113–137.
17. Westheimer,F.H. (1987) Why nature chose phosphates. Science,
235, 1173–1178.
18. Hsiao,C., Tannenbaum,M., VanDeusen,H., Hershkovitz,E.,
Perng,G., Tannenbaum,A. and Williams,L.D. (2008) In Hud,N.
Nucleic Acids Research, 2009, Vol. 37, No. 10 3141

(ed.), Nucleic Acid Metal Ion Interactions. The Royal Society of
Chemistry, London, pp. 1–35.
19. Brown,I.D. (1988) What factors determine cation coordination
numbers. Acta Crystallogr. Sect. B.,44, 545–553.
20. Brown,I.D. (1992) Chemical and steric constraints in inorganic
solids. Acta Crystallogr. Sect. B.,48, 553–572.
21. Bock,C.W., Katz,A.K., Markham,G.D. and Glusker,J.P. (1999)
Manganese as a replacement for magnesium and zinc: functional
comparison of the divalent ions. J. Am. Chem. Soc.,121,
7360–7372.
22. Rashin,A.A. and Honig,B. (1985) Reevaluation of the born model
of ion hydration. J. Phys. Chem.,89, 5588–5593.
23. Berman,H.M., Westbrook,J., Feng,Z., Gilliland,G., Bhat,T.N.,
Weissig,H., Shindyalov,I.N. and Bourne,P.E. (2000) The protein
data bank. Nucleic Acids Res.,28, 235–242.
24. Apostolico,A., Ciriello,G., Guerra,C., Heitsch,C.E., Hsiao,C. and
Williams,L.D. (2009) Finding 3D motifs in ribosomal RNA struc-
tures. Nucleic Acids Res.,21, 21.
25. Richardson,J.S., Schneider,B., Murray,L.W., Kapral,G.J.,
Immormino,R.M., Headd,J.J., Richardson,D.C., Ham,D.,
Hershkovits,E., Williams,L.D. et al. (2008) RNA backbone:
Consensus all-angle conformers and modular string nomenclature
(an RNA Ontology Consortium contribution). RNA,14, 1–17.
26. Hsiao,C., Mohan,S., Hershkovitz,E., Tannenbaum,A. and
Williams,L.D. (2006) Single nucleotide RNA choreography. Nucleic
Acids Res.,34, 1481–1491.
27. Hershkovitz,E., Sapiro,G., Tannenbaum,A. and Williams,L.D.
(2006) Statistical analysis of the RNA backbone. IEEE/ACM Trans.
Comp. Biol. Bioinformatics,3, 33–46.
28. Hershkovitz,E., Tannenbaum,E., Howerton,S.B., Sheth,A.,
Tannenbaum,A. and Williams,L.D. (2003) Automated identification
of RNA conformational motifs: theory and application to the HM
LSU 23S rRNA. Nucleic Acids Res.,31, 6249–6257.
29. Stefan,L.R., Zhang,R., Levitan,A.G., Hendrix,D.K., Brenner,S.E.
and Holbrook,S.R. (2006) MeRNA: a database of metal ion bind-
ing sites in RNA structures. Nucleic Acids Res.,34, D131–D134.
30. Altschul,S.F., Madden,T.L., Schaffer,A.A., Zhang,J., Zhang,Z.,
Miller,W. and Lipman,D.J. (1997) Gapped BLAST and PSI-
BLAST: a new generation of protein database search programs.
Nucleic Acids Res.,25, 3389–3402.
31. Larkin,M.A., Blackshields,G., Brown,N.P., Chenna,R.,
McGettigan,P.A., McWilliam,H., Valentin,F., Wallace,I.M.,
Wilm,A., Lopez,R. et al. (2007) Clustal W and Clustal X version
2.0. Bioinformatics,23, 2947–2948.
32. Hansen,J.L., Schmeing,T.M., Klein,D.J., Ippolito,J.A., Ban,N.,
Nissen,P., Freeborn,B., Moore,P.B. and Steitz,T.A. (2001) Progress
toward an understanding of the structure and enzymatic mechanism
of the large ribosomal subunit. Cold Spring Harb. Symp. Quant.
Biol.,66, 33–42.
33. Toor,N., Keating,K.S., Taylor,S.D. and Pyle,A.M. (2008) Crystal
structure of a self-spliced group II intron. Science,320, 77–82.
34. Noller,H.F. (2004) The driving force for molecular evolution of
translation. RNA,10, 1833–1837.
35. Uhlein,M., Weglohner,W., Urlaub,H. and Wittmann-Liebold,B.
(1998) Functional implications of ribosomal protein L2 in protein
biosynthesis as shown by in vivo replacement studies. Biochem. J.,
331, 423–430.
36. Cooperman,B.S., Wooten,T., Romero,D.P. and Traut,R.R. (1995)
Histidine 229 in protein L2 is apparently essential for 50S peptidyl
transferase activity. Biochem. Cell. Biol.,73, 1087–1094.
37. Diedrich,G., Spahn,C.M., Stelzl,U., Schafer,M.A., Wooten,T.,
Bochkariov,D.E., Cooperman,B.S., Traut,R.R. and Nierhaus,K.H.
(2000) Ribosomal protein L2 is involved in the association of the
ribosomal subunits, tRNA binding to A and P sites and peptidyl
transfer. EMBO J.,19, 5241–5250.
38. Meskauskas,A., Russ,J.R. and Dinman,J.D. (2008) Structure/func-
tion analysis of yeast ribosomal protein L2. Nucleic Acids Res.,7,7.
39. Nissen,P., Hansen,J., Ban,N., Moore,P.B. and Steitz,T.A. (2000)
The structural basis of ribosome activity in peptide bond synthesis.
Science,289, 920–930.
40. Noller,H.F., Hoffarth,V. and Zimniak,L. (1992) Unusual resistance
of peptidyl transferase to protein extraction procedures. Science,
256, 1416–1419.
41. Carter,A.P., Clemons,W.M., Brodersen,D.E., Morgan-Warren,R.J.,
Wimberly,B.T. and Ramakrishnan,V. (2000) Functional insights
from the structure of the 30S ribosomal subunit and its interactions
with antibiotics. Nature,407, 340–348.
42. Robertson,M.P., Igel,H., Baertsch,R., Haussler,D., Ares,M. Jr and
Scott,W.G. (2005) The structure of a rigorously conserved RNA
element within the SARS virus genome. PLoS Biol.,3, e5.
43. Beese,L.S. and Steitz,T.A. (1991) Structural basis for the 30-50exo-
nuclease activity of Escherichia coli DNA polymerase I: a two metal
ion mechanism. EMBO J.,10, 25–33.
44. Steitz,T.A. and Steitz,J.A. (1993) A general two-metal-ion
mechanism for catalytic RNA. Proc. Natl Acad. Sci. USA,90,
6498–6502.
45. Martins,E.P. and Hansen,T.F. (1997) Phylogenies and the com-
parative method: a general approach to incorporating phylogenetic
information into the analysis of interspecific data. Am. Nat.,149,
646–667.
46. Rost,B. (1999) Twilight zone of protein sequence alignments.
Protein Eng.,12, 85–94.
47. Heinz,D.W., Baase,W.A., Zhang,X.J., Blaber,M., Dahlquist,F.W.
and Matthews,B.W. (1994) Accommodation of amino acid inser-
tions in an alpha-helix of T4 lysozyme. Structural and thermody-
namic analysis. J. Mol. Biol.,236, 869–886.
48. Olsen,G.J. and Woese,C.R. (1997) Archaeal genomics: an overview.
Cell,89, 991–994.
49. Cannone,J.J., Subramanian,S., Schnare,M.N., Collett,J.R.,
D’Souza,L.M., Du,Y., Feng,B., Lin,N., Madabusi,L.V.,
Muller,K.M. et al. (2002) The comparative RNA web (CRW) site:
an online database of comparative sequence and structure infor-
mation for ribosomal, intron, and other RNAs. BMC
Bioinformatics,3,2.
50. Mears,J.A., Sharma,M.R., Gutell,R.R., McCook,A.S.,
Richardson,P.E., Caulfield,T.R., Agrawal,R.K. and Harvey,S.C.
(2006) A structural model for the large subunit of the mammalian
mitochondrial ribosome. J. Mol. Biol.,358, 193–212.
51. Sharma,M.R., Koc,E.C., Datta,P.P., Booth,T.M., Spremulli,L.L.
and Agrawal,R.K. (2003) Structure of the mammalian mitochon-
drial ribosome reveals an expanded functional role for its compo-
nent proteins. Cell,115, 97–108.
52. Mears,J.A., Cannone,J.J., Stagg,S.M., Gutell,R.R., Agrawal,R.K.
and Harvey,S.C. (2002) Modeling a minimal ribosome based on
comparative sequence analysis. J. Mol. Biol.,321, 215–234.
53. Smits,P., Smeitink,J.A., van den Heuvel,L.P., Huynen,M.A. and
Ettema,T.J. (2007) Reconstructing the evolution of the mitochon-
drial ribosomal proteome. Nucleic Acids Res.,35, 4686–4703.
54. O’Brien,T.W. (2003) Properties of human mitochondrial ribosomes.
IUBMB Life,55, 505–513.
3142 Nucleic Acids Research, 2009, Vol. 37, No. 10