A unique cysteine-rich zinc finger domain present in a majority of class II ribonucleotide reductases mediates catalytic turnover

Article abstract of DOI:10.1074/jbc.M117.806331

Ribonucleotide reductases (RNRs) catalyze the reduction of ribonucleotides to the corresponding deoxyribonucleotides, used in DNA synthesis and repair. Two different mechanisms help deliver the required electrons to the RNR active site. Formate can be used as reductant directly in the active site, or glutaredoxins or thioredoxins reduce a C-terminal cysteine pair, which then delivers the electrons to the active site. Here, we characterized a novel cysteine-rich C-terminal domain (CRD), which is present in most class II RNRs found in microbes. The NrdJd-type RNR from the bacterium Stackebrandtia nassauensis was used as a model enzyme. We show that the CRD is involved in both higher oligomeric state formation and electron transfer to the active site. The CRD-dependent formation of high oligomers, such as tetramers and hexamers, was induced by addition of dATP or dGTP, but not of dTTP or dCTP. The electron transfer was mediated by an array of six cysteine residues at the very C-terminal end, which also coordinated a zinc atom. The electron transfer can also occur between subunits, depending on the enzyme's oligomeric state. An investigation of the native reductant of the system revealed no interaction with glutaredoxins or thioredoxins, indicating that this class IIRNRuses a different electron source. Our results indicate that the CRD has a crucial role in catalytic turnover and a potentially new terminal reduction mechanism and suggest that the CRD is important for the activities of many class II RNRs.

Full text of DOI:10.1074/jbc.M117.806331

JBC Papers in Press. Published on October 2, 2017 as Manuscript M117.806331
The latest version is at http://www.jbc.org/cgi/doi/10.1074/jbc.M117.806331
Catalytic turnover in class II ribonucleotide reductase
A unique cysteine-rich Zn-finger domain present in a majority of class II ribonucleotide reductases
mediates catalytic turnover 1
‡
§
‡#
‡
Christoph Loderer , Venkateswara Rao Jonna , Mikael Crona , Inna Rozman Grinberg ,
‡
§
‡
‡
Margareta Sahlin , Anders Hofer , Daniel Lundin* , Britt-Marie Sjöberg*
‡
From the Department of Biochemistry and Biophysics, Stockholm University and Department of Medical
§
Biochemistry, Umeå University, Sweden ; Current address: Swedish Orphan Biovitrum AB, Solna,
Sweden#
Running title: Catalytic turnover in class II ribonucleotide reductase
To whom correspondence should be addressed: Daniel Lundin, telephone: +46-8-52481520, fax: +46-
8
-155597, daniel.lundin@dbb.su.se; Britt-Marie Sjöberg, telephone: +46-8-164150, fax: +46-8-155597,
britt-marie.sjoberg@dbb.su.se
Keywords: ribonucleotide reductase, oligomerization, metal ion-protein interaction, oxidation-reduction
(
redox), phylogenetics, thioredoxin
ABSTRACT
Ribonucleotide reductases (RNRs) catalyze the
also coordinated a zinc atom. The electron transfer
can also occur between subunits, depending on the
enzyme’s oligomeric state. An investigation of the
native reductant of the system revealed no interac-
tion with gluta- or thioredoxins, indicating that this
class II RNR uses a different electron source. Our
results are indicating that the CRD has a crucial role
in catalytic turnover and a potentially new terminal
reduction mechanism and suggest that the CRD is
important for the activities of many class II RNRs.
reduction of ribonucleotides to the corresponding
deoxyribonucleotides, used in DNA synthesis and
repair. Two different mechanisms help deliver the
required electrons to the RNR active site. Formate
can be used as reductant directly in the active site, or
gluta- or thioredoxins reduce a C-terminal cysteine
pair, which then delivers the electrons to the active
site. Here, we characterized a novel cysteine-rich
C-terminal domain (CRD), which is present in most
class II RNRs found in microbes. The NrdJd-type
RNR from the bacterium Stackebrandtia nassauen-
sis was used as a model enzyme. We show that the
CRD is involved in both, higher oligomeric state for-
mation and electron transfer to the active site. The
CRD-dependent formation of high oligomers, such
as tetramers and hexamers, was induced by addition
of dATP or dGTP, but not of dTTP or dCTP. The
electron transfer was mediated by an array of six
cysteine residues at the very C-terminal end, which
Ribonucleotide reductases (RNRs)2 catalyze the re-
duction of ribonucleotides to the corresponding de-
oxyribonucleotides, which are required for DNA
synthesis and repair. Since this reaction is the only
known biological pathway for de novo deoxyribonu-
cleotide synthesis, RNRs are considered to be essen-
tial for every living organism. Today, three different
classes of RNRs are known, which all share a com-
mon fold and a general reaction mechanism based
1
The work was supported by the Swedish Research Council (2016-01920), Swedish Cancer Society (CAN 2016/670), Wenner-
Gren Foundations, Carl Trygger’s Foundation and Kempe Foundations.
2
The abbrevations used are: AdoCbl, adenosylcobalamin; AdoMet, S-adenosyl-L-methionine; CRD, cysteine-rich C-terminal
domain; GEMMA, gas phase electrophoretic macromolecule analysis; HED, 2-hydroxyethyl disulfide; ICP, inductively coupled
plasma analysis; ICP-OES, inductively coupled plasma optical emission spectrometry; IMAC, metal ion affinity chromatogra-
phy; ITC, isothermal titration calorimerty; NrdJd-wt, the full-length version of the Stackebrandtia nassauensis class II RNR;
NrdJd∆CRD, the C-terminally deleted version of the S. nassauensis class II RNR; RNR, ribonucleotide reductase; SEC, size
exclusion chromatography; TCEP, tris(2-carboxyethyl)phosphine hydrochloride.
1
Copyright 2017 by The American Society for Biochemistry and Molecular Biology, Inc.

Catalytic turnover in class II ribonucleotide reductase
on a thiyl radical (1). All known RNRs, except a
few viral RNRs, also share an allosteric substrate-
specificity regulation, where binding of a dNTP or
ATP molecule modulates the specificity of the ac-
tive site of the enzyme (1). The three classes share
a common ancestor (2) and the main difference be-
tween them is their mechanism of radical generation
been observed for the full length NrdJ from Thermo-
toga maritima, while a truncated variant utilized for
X-ray crystallography crystallized as a dimer (13).
The terminal reductant is poorly studied for NrdJ
enzymes. The monomeric NrdJ from L. leichman-
nii was shown to receive electrons from thioredoxin
(14), but for other NrdJ enzymes reduction by thiore-
doxin or glutaredoxin has not been studied.
(
1, 3). With the different radical generating mech-
anisms follow differences in dependence of oxygen
and oligomeric composition.
In the monomeric NrdJ, as well as in class I RNRs,
the interaction with thioredoxin or glutaredoxin is
mediated via a conserved C-terminal cysteine pair.
Due to its position on a flexible C-terminal tail, this
reduced cysteine pair is believed to have access to
the active site and to re-reduce the disulfide that
is oxidized during the nucleotide reduction reaction
in turn forming a disulfide that is subsequently re-
duced by the terminal reductant (4, 5). This study
reports that most non-monomeric class II RNRs con-
tain a so far undescribed cysteine-rich C-terminal
domain (CRD). We show that the CRD is involved
in oligomerization and the terminal reduction of the
class II RNR from Stackebrandtia nassauensis and
that the CRD hence plays a crucial role in the cat-
alytic cycle of class II RNRs.
Class I RNRs (NrdAB/NrdEF) generate the radical
in a separate subunit (β, NrdB/F) via a dimetal center
that is oxidized by molecular oxygen. The radical is
then transferred to the catalytic subunit (α, NrdA/E)
and back after the catalytic cycle. Thus, this class is
oxygen-dependent and forms active α β heterote-
2
2
tramers. Terminal reduction of the oxidized active
site in α is performed by thioredoxin or glutaredoxin
systems at the end of each turnover (4, 5).
Class III RNRs (NrdD) also require a separate sub-
unit (NrdG) to generate the radical. An iron-sulfur-
cluster in NrdG confers homolytic cleavage of the co-
factor S-adenosyl-L-methionine (AdoMet) that gen-
erates a stable glycyl radical in the catalytic subunit.
This glycyl radical is extremely sensitive towards
oxygen, leading to protein backbone cleavage at oxy-
gen exposition. NrdDs are described to be active as
α2 homodimers and NrdG is only needed to gen-
erate the glycyl radical in NrdD (6, 7). Formate
was shown to be a terminal reducing agent for some
NrdD enzymes (8). Recently, class III RNRs that
use a thioredoxin system for terminal reduction have
been described (9).
RESULTS
Phylogenetic analysis — We estimated a
maximum likelihood phylogenetic tree from an
alignment of sequences representing the full di-
versity of NrdJ sequences at the 75% identity
level (Figure 1A; Supporting information). The
resolution increased after exclusion of sequences
from the monomeric class II enzymes, subclass
NrdJm, allowing us to define two well sup-
ported, likely monophyletic subclasses (NrdJd,
NrdJa+b) with at least one enzymatically character-
ized member (Figure 1B; Supporting information;
http://rnrdb.pfitmap.org). In addition, four likely
monophyletic, well supported, candidate subclasses
and a number of unclassified sequences were iden-
tified. Subclass NrdJm contains the monomeric
enzyme from L. leichmannii (12), subclass NrdJd
contains the enzyme from Streptomyces clavuligerus
(15) and subclass NrdJa+b contains the split enzyme
from Pseudomonas aeruginosa (16, 17). Other well
characterized NrdJ enzymes such as the one from T.
maritima and T. acidophilum are among the unclas-
sified sequences (11).
The RNR that is the focus of the current study,
class II (NrdJ), generates the radical via homolytic
cleavage of the vitamin B12 coenzyme (AdoCbl)
within the enzyme, close to the active site. Af-
ter the catalytic cycle, AdoCbl is reconstituted (10).
This mechanism makes NrdJ enzymes independent
of oxygen, but instead dependent on B12. The
oligomeric organisation of the NrdJ enzymes is more
diverse than in the other classes. Some representa-
tives, such as the enzyme from Thermoplasma aci-
dophilum, have been shown to be active as α ho-
2
modimers (11) whereas the NrdJ from Lactobacillus
leichmannii is active as monomer, where an addi-
tional domain mimics the dimer interface, required
for allosteric regulation (12). Higher oligomers have
2

Catalytic turnover in class II ribonucleotide reductase
An alignment of all NrdJ sequences shows major
differences in the C-terminal region. A cysteine-
rich C-terminal domain (CRD) is present in 70 %
of the investigated sequences. The CRD is 150-220
amino acids long and contains hydrophobic regions
with single conserved hydrophilic residues (Figure
not CMP or CTP.
Allosteric regulation — An important char-
acteristic of all RNRs known so far, with the ex-
ception of class I RNRs from herpes viruses (18),
is the allosteric regulation of substrate specificity.
Thereby, effector binding in the specificity site in
the dimer interface influences the substrate speci-
ficity of the active site. The potential effectors ATP,
dATP, dCTP, dGTP and dTTP were tested for their
influence on the reduction of ADP, CDP, GDP and
UDP by NrdJd-wt and NrdJd∆CRD. All four po-
tential substrates were present in the assay together
with one of the five potential effector molecules at a
time and conversion of all substrates was measured.
The effectors ATP and dATP activated the wild type
enzyme for CDP reduction (Figure 3A). dCTP and
dTTP enhance the activity for GDP reduction, while
dGTP activates both enzymes for ADP and GDP
reduction. UDP conversion was not observed with
any of the tested effector nucleotides. Except for a
low activity with ATP as effector, the experiment
yielded similar results for NrdJd∆CRD (Figure 3B).
To investigate the affinity of NrdJd-wt and
NrdJd∆CRD towards the preferred effectors, activ-
ity assays with a single substrate were performed in
presence of increasing effector concentrations. For
NrdJd-wt all effectors yielded nearly maximal activ-
ity already at the lowest tested effector concentration
1C). The most striking feature of this domain is an
array of six fully conserved cysteine residues at the
C-terminal end of the sequence (Figure 1C). No
sequence similarity between the class I C-terminal
sequence, which contains a single pair of conserved
cysteines, and the NrdJ CRD was detected.
The CRD is predominant in subclass NrdJd and in
three out of the four candidate subclasses as well
as in a majority of the unclassified sequences, but
not at all in NrdJm. The NrdJa+b subclass consists
of enzymes where the CRD is not genetically fused
to the catalytic domain, but encoded by a separate
gene, nrdJb. In some Frankia species and a few Al-
phaproteobacteria, two NrdJd enzymes are present
where only one contains the CRD (Suppl. Table S1;
http://rnrdb.pfitmap.org).
Recombinant expression of candidate en-
zymes from subclass NrdJd — To study the proper-
ties and the function of the NrdJ subclass in gen-
eral and the CRD in particular, a set of candidate
proteins were identified using a bioinformatic pre-
screen to estimate successful expression of soluble
protein. The NrdJd enzymes from Streptomyces
clavuligerus, Coraliomargarita akajimensis and S.
nassauensis were selected and recombinantly ex-
pressed in Escherichia coli. Only the S. nassauensis
NrdJd showed a sufficient amount of soluble expres-
sion of active enzyme with an intact CRD (Figure
−1
of 4 µmol L (Figure 3C). This concentrations rep-
resents only a two fold molar excess over the applied
−1
enzyme concentration of 2 µmol L . This indicates
−1
very low K -values (< 4 µmol L ), but it also means
L
that the actual values cannot be determined based
on activity at these conditions. For NrdJd∆CRD
the effector titrations yielded saturation curves with
−1
2).
K -values between 50 and 150 µmol L (Figure
L
In addition, both the isolated catalytic domain
of the S. nassauensis enzyme (positions 1-714;
NrdJd∆CRD), as well as the CRD alone (positions
3D). In conclusion, the affinity of the C-terminally
truncated enzyme towards each tested effector is
reduced. However, the overall pattern of activation
or inactivation of the enzyme for different substrates
by the effectors is intact.
7
15-947) were cloned and expressed in E. coli sep-
arately (Figure 2). The expression of the CRD
yielded negligible amounts of soluble protein and
aggregation occurred quickly. From the full length
enzyme NrdJd-wt and the isolated catalytic domain
NrdJ∆CRD, soluble protein was obtained in high
purity and sufficient quantity. Initial activity assays
showed that the enzyme was able to reduce CDP but
Oligomeric state — The oligomeric state of
S. nassauensis NrdJd was studied using size exclu-
sion chromatography (SEC) at high protein concen-
−1
tration (1.0 mg mL ) and gas phase electrophoretic
macromolecule analysis (GEMMA) at a lower con-
3

Catalytic turnover in class II ribonucleotide reductase
−
1
centration (0.075 mg mL ). The latter is compa-
rable to the enzyme concentration in the activity
assays. The full length NrdJd yielded a SEC peak at
database (http://www.rcsb.org/pdb/home/home.do)
suggested similarity to a zinc binding site from
tRNA(Ile2) 2-agmatinylcytidine synthetase (PDB:
4RVZ) from Archaeoglobus fulgidus (20). Based
on this structure, a homology model of the S. nas-
sauensis C-terminal 35 residues was created apply-
ing Swiss Model (Figure 6A).
4
50 kDa, corresponding to a tetramer that was unaf-
fected by dATP addition (Figure 4A). In GEMMA,
NrdJd-wt shows a mixture of monomers and dimers,
that shifts to a tetramer only in the presence of 100
−
1
µmol L dATP (Figure 4C). NrdJd∆CRD, on the
other hand, is present as a monomer between 90-100
kDa both in SEC and in GEMMA (Figure 4B,D).
The NrdJd∆CRD monomer and the small fraction
of dimer visible in GEMMA were not affected by
the addition of dATP (Figure 4D). dGTP yielded
results similar to dATP, whereas addition of dTTP
The model suggested that a zinc is coordinated by
the cysteine residues C920, C923, C936 and C939
(Figure 6B). The remaining two cysteine residues
C933 and C945 are in close proximity in the model
and could form the site of terminal reduction. This
homology model was applied as a starting model
for subsequent experimental confirmation. Accord-
ingly, six different cysteine to alanine substitution
variants of S. nassauensis NrdJd were constructed
and their enzyme activity tested in assays applying
either DTT or TCEP as reductants. All variants
retained at least 50% of their specific activity with
the reductant DTT (Figure 6C). However, none of
the variants besides the wild type showed any en-
zyme activity with the reductant TCEP, suggesting
all cysteines are essential for active site reduction.
In an attempt to assess the four cysteine residues that
are suggested to coordinate zinc, inductively cou-
pled plasma (ICP) measurements were performed
for the wild type enzyme and all six cysteine to
alanine mutants (Figure 6C). The wild-type enzyme
as well as the variants C933A and C939A showed
a zinc occupation of more than 60% per monomer.
The variants C920A, C923A and C936A yielded
reduced occupancy to between 20% and 40%, while
C945A was fully occupied. Thus, the exchange of
the two cysteine residues predicted to form the ter-
minal cysteine pair had comparable or higher Zn2
occupation than the wild type whereas of the four
cysteine residues predicted to coordinate the zinc,
three showed a reduced zinc content, while one was
equal to the wild type.
(
Figure 4C) and dCTP induced dimerization but not
tetramerization in NrdJd-wt. In titrations with dATP,
as well as dGTP, NrdJd-wt was in a monomer-dimer
equilibrium at lower concentrations of the effector
and gradually changed to tetramers and even hexam-
ers when the effector concentration was increased
(
Figure 4E,F). By interpolation of the quantified
GEMMA results based on mass concentration, half
the enzyme is present as tetramer or hexamer at a
−
1
concentration of 17 µmol L dATP.
Electron transfer — Since two C-terminal
cysteine residues are known to be the acceptors of
electrons from glutaredoxins and thioredoxins in
the class I and NrdJm enzymes, the role in electron
transfer of the S. nassauensis NrdJd CRD and its six
cysteine residues was investigated. Earlier studies
with alternative reducing systems have demonstrated
that the phosphine TCEP specifically reduces the C-
terminal cysteine residues of class I RNR, while
DTT can reduce the enzyme active site directly
+
(
19). In this study S. nassauensis NrdJd-wt and
NrdJd∆CRD were tested with both reductants (Fig-
ure 5). NrdJd-wt activity is supported by both DTT
and TCEP, with four-fold higher specific activity
using the latter. NrdJd∆CRD exhibits a comparable
activity with the reductant DTT as the wild-type
enzyme, but is completely inactive with TCEP. An
attempted reconstitution of the full length protein
with equimolar amounts of NrdJd∆CRD and the
separately expressed CRD did not show any activity
with TCEP (data not shown).A PROSITE search
did not fit the S. nassauensis CRD to any known
cysteine-rich motif such as zinc fingers. On the other
hand, a sequence similarity search against the PDB
Subunit crosstalk — The CRD is responsible
for electron delivery to the active site but as shown
above, the NrdJd can adopt different oligomeric
states. This leads to the question whether the C-
terminal cysteine residues can only reach the active
site of their own subunit or if they can also deliver
electrons to other subunits within the oligomer. To
investigate this, an active site cysteine deficient mu-
4

Catalytic turnover in class II ribonucleotide reductase
tant (C370A) was mixed in equimolar amounts with
a C-terminal mutant protein. In control experiments
the C370A mutant showed no activity with any of
the reductants, TCEP or DTT, while the C-terminal
mutant C933A was active with DTT but not with
TCEP (Figure 7A). Yet the mixture of the two mutant
proteins exhibited high enzyme activity with both
DTT or TCEP, indicating electron transfer from the
intact C-terminus of the active site mutant C370A
to the active site of the CRD mutant C933A.
Native electron donor — To further investi-
gate the role of the C-terminal cysteine residues as
terminal reducing sites of the enzyme, we attempted
to define the native reductant. In many class I RNRs
as well as for the NrdJm from L. leichmannii, thiore-
doxins and glutaredoxins serve as native reductants
(1, 14). Therefore, we performed a sequence sim-
ilarity search for thio- and glutaredoxins in the S.
nassauensis proteome. The organism encodes two
thioredoxins, Snas 2647 (thioredoxin 1) and Snas
6430 (thioredoxin 2), as well as one glutaredoxin,
Snas 1785. The three redoxins, together with the
thioredoxin reductase (TR) Snas 6431, were cloned
and recombinantly expressed in E. coli. All con-
structs yielded soluble protein in high amounts with
high specific activities on artificial substrates (Fig-
ure 8A). All thio- and glutaredoxins were tested
in the RNR activity assay together with their cor-
responding electron source (thioredoxin reductase
+ NADPH for the thioredoxins, and glutathione +
glutathione reductase + NADPH for glutaredoxin).
However, no RNR activity was detected in presence
of any of the redoxin systems (Figure 8B). To test
whether the conditions for the combined assay are
suitable for RNR and the redoxin systems, both reac-
tion systems were monitored separately. Addition of
the artificial reductant DTT to the full reaction mix
resulted in high specific activity of the RNR (Figure
8B). Addition of the artificial electron acceptors in-
sulin (thioredoxins) and HED (glutaredoxin) yielded
high specific activities for the redoxin systems (Fig-
ure 8A). Both the RNR and the redoxin systems,
exhibited high activity at the combined assay condi-
tions, but no transfer of electrons could be observed
between them.
Due to the complex oligomeric organization of
NrdJd-wt, the influence of the oligomeric state on
the subunit crosstalk was investigated. Activity as-
says were performed with the wild type enzyme and
the mixture of the CRD mutant C933A and active
site mutant C370A. To test for crosstalk in the dimer,
the substrate/effector pair GDP/dTTP was applied,
where no formation of higher oligomers was ob-
served in the previous experiments. Crosstalk in
higher oligomers was investigated, using the sub-
strate/effector pair CDP/dATP, where formation of
tetramers and hexamers was shown in the GEMMA
experiments (Figure 7B). In both cases TCEP was
applied as reductant. NrdJd-wt had a low K -value
L
−
1
<
4.0 µmol L for dATP. For the C933A/C370A
mixture, an increased K -value of 15.0 ± 4.5 µmol
L
−1
L
was determined. The Vmax value in the variant
mixture was reduced to 30% of the wild type, which
is close to the theoretical maximum value of 50%
considering that half of the subunits in the mixture
has a defective active site. For the effector dTTP,
NrdJd-wt showed maximum activity already at the
lowest dTTP concentration, indicating a high affin-
ity. The C933A/C370A mixture exhibited very low
overall activity between 3 and 7% of NrdJd-wt with
no significant effect of increasing dTTP concentra-
tions. To further assess the effect of dATP, assays
were performed with mixtures of dATP and dTTP
In another attempt to identify the native electron
donor, RNR assays were performed with different
types of S. nassauensis cell lysates (Figure 8C).
Tests were performed with protein-free lysate (1),
cleared lysate (2) and complete lysate (3). None
of the lysates increased activity beyond the level
measured in the negative control.
(
Figure 7C). For both the C933A/C370A mixture
and NrdJd-wt, activity with the substrate CDP was
lower in the presence of dATP plus dTTP compared
to dATP alone. However, the addition of 250 µmol
−1
−1
L dATP to 1 mmol L dTTP lead to an increase in
GDP reduction for the C933A/C370A mixture and
NrdJd-wt. The effect was especially pronounced
in the C933A/C370A mixture, where activity in-
creased 18-fold compared to 4-fold in the wild-type
enzyme.
DISCUSSION
We characterized a unique C-terminal
cysteine-rich domain, present in a majority of class
5

Catalytic turnover in class II ribonucleotide reductase
II RNRs, which is involved in oligomerization and
terminal reduction during catalysis. This domain is
identical to the NrdJb subunit in the NrdJa+b split
enzyme described for P. aeruginosa (16, 17) and the
CRD that was spontaneously cleaved from T. mar-
itima NrdJ, used for crystallization (13). Sequence
analysis revealed the presence of the CRD in 70%
of the non redundant NrdJ sequences, excluding the
monomeric NrdJm which never contains the CRD
periments also showed that this oligomerization is
stimulated by addition of the purine nucleotides
dATP and dGTP, but not by the pyrimidine nu-
cleotides dTTP and dCTP.
While the actual mechanism behind the CRD’s role
in oligomerization and how this affects activity of
the enzyme still remains to be described, the role
of the CRD in the re-reduction of the active site
cysteine residues is much clearer. Although se-
quence similarity between the CRD of most NrdJs
on one side and the C-termini of Class I RNRs and
monomeric NrdJs on the other side is limited to
a single pair of cysteines, the mechanism appears
comparable (4, 5, 14). This is supported by the
inability of NrdJd∆CRD to accept TCEP as termi-
nal reductant while activity is retained with DTT,
the latter being small enough to directly reduce the
active site whereas the former is not (19).
(
Figure 1B; Supplementary Table S1). Among the
non-monomeric NrdJs, the presence of the CRD
does not follow a simple phylogenetic pattern (Fig-
ure 1A,B). The domain is present in the majority of
subclasses and candidate subclasses. Absences are
spread throughout the tree, suggesting independent
losses of the domain. In NrdJd, the only absences of
the CRD we could detect, were from genomes also
encoding another NrdJd with a CRD.
The high occurrence of the CRD-domain in NrdJ
enzymes, as well as its genetic separation from the
catalytic subunit in NrdJa+b, indicate not only the
general importance of the domain but also that it is
a separately folded independent domain. Previous
studies of enzymes containing the CRD have suf-
fered from the aggregation-promoting properties of
this domain as well as its susceptibility to cleavage
The TCEP experiments with the six cysteine-to-
alanine exchanges showed that all six cysteine
residues are necessary for the functionality of the
electron transfer. Homology modeling indicated that
four of these cysteine residues (C920, C923, C936,
C939) could be involved in zinc binding while the
remaining two (C933, C945) appear to form the
redox-active cysteine pair. However, the ICP mea-
surement only shows reduced zinc content in three
of the four predicted positions. A zinc occupancy
between 0.2 and 0.4 in the variants, compared to 0.6
in the wild type enzyme, is in good agreement with
values measured in comparable studies (21, 22). It
is noteworthy, that both C933A and C939A show
the same zinc content as the wild type, while it is
reduced in C936A. Possibly, the zinc can be coor-
dinated either by C933-C936 or C936-C939. Both
these pairs have the common sequence C-X-X-C
that is typical of zinc ligating cysteines. A certain
promiscuity in the coordination of the zinc would
also explain why no single mutant showed a com-
plete loss of zinc as observed in the full truncation.
Taking the potential swap between C933 and C939
into account, the results are consistent with the ho-
mology model. We thus propose, that the residues
C920, C923, C936 and C939 coordinate a zinc
residue, structuring the C-terminal end and bringing
C933 and C945 in close proximity to each other.
This pair can then act as an electron shuttle between
the electron donor and the active site. Coordination
(
11, 17). For the genetically separate CRD, NrdJb
from P. aeruginosa, it was not possible to obtain
soluble protein by heterologous expression in E.
coli (17). Recombinant expression of the full length
unclassified NrdJ from T. maritima yielded soluble
protein but the CRD was partially cleaved during
expression (13). Fortunately, the NrdJd from S. nas-
sauensis did not show any sign of cleavage during
expression or further experiments.
Our functional analysis of the CRD suggests an im-
portant role in oligomerization and electron transfer.
The CRD-dependend formation of higher oligomers
is well in line with observations made for the NrdJ
from T. maritima, which was first characterized as
a higher oligomer with about 450 kDa (11). The
C-terminally truncated version was later character-
ized as a dimer (13). In this study, we show that
the C-terminally truncated NrdJ enzyme from S.
nassauensis is in a monomer-dimer equilibrium.
Tetramers and hexamers were only observed in the
full length enzyme, showing the importance of the
CRD for the formation of higher oligomers. Ex-
6

Catalytic turnover in class II ribonucleotide reductase
of the zinc by C933 in case of the C939A variant
seems possible, in which case a restructured metal
binding site could still bind the zinc ligand, but C945
would be deprived of its partner, prohibiting the role
of terminal reduction.
A C-terminal zinc binding site was previously de-
scribed for NrdD enzymes (23, 24). It was proposed
to be a structure forming element, which keeps the
catalytically crucial glycyl radical in correct orien-
tation. Due to the low sequence similarity and the
very different location of the zinc binding site within
the sequences, there is no evidence for homology
between the sites in class II and class III RNRs.
Cysteine-coordinated zincs with a structural role are
commonly found in proteins (25, 26) and are often
the result of convergent evolution.
the described allosteric regulation, addition of dATP
to a GDP/dTTP assay instead led to an increase in
activity. However, dATP does not activate GDP
reduction via the described allosteric mechanism.
Instead, it suggests that dATP binding to another site
may mediate the increase in activity by formation of
higher oligomers.
To assess the physiological role of the CRD, we
attempted to find the native electron donor of the
system. The three tested gluta- and thioredoxins
from S. nassauensis were not able to supply the
RNR with electrons under the tested conditions.
The observation that all systems separately, the
RNR and the gluta-/thioredoxin systems, are active
but cannot exchange electrons, questions whether
RNRs with the CRD are dependent on reduction
by a redoxin system at all. Attempts to find the
native electron donor in S. nassauensis cell lysates
did, however, not yield any positive results. Since
the organism also contains a class I RNR, the na-
tive electron donor system might not be expressed
during the employed culture conditions or might be
inactivated in the lysis process. As shown by Berardi
and Bushweller in NMR studies, the very flexible,
disordered C-terminal tail of class I and NrdJm en-
zymes plays a vital role in the interaction of the
C-terminal cysteine residues with the redoxins (28).
The highly ordered C-terminus in CRD-containing
NrdJ enzymes might be unsuitable for an interaction
with redoxins in a fashion comparable to class I or
NrdJm RNRs. Our so far unproductive attempts to
find the reducing system for S. nassauensis together
with the recent discovery that some class III RNRs
are indeed reduced by redoxins (9), and not by for-
mate that was long believed to be the sole reductant
of these enzymes, show the plurality of ultimate
reductants that can be involved in ribonucleotide
reduction.
With the help of mutant proteins defective in the
active site (C370A) and in the CRD (C933A), we
demonstrated electron transfer from the CRD of one
subunit to the active site of another subunit. The
effects of oligomerization on this crosstalk were
investigated by addition of the effectors dATP and
dTTP with their different oligomerization effects.
The results indicate that an efficient transfer of
electrons between the subunits can occur only in
tetramers or higher oligomers. The measured K -
L
−
1
value for dATP of 15 µmol L compares well with
the GEMMA results where, at an estimated concen-
−
1
tration of 17 µmol L , half of the enzyme is present
as tetramer or higher oligomer. Crosstalk between
terminal reduction sites has been observed before in
the class I RNR from S. cerevisiae and suggestded
to occur within the dimer. (27). Higher oligomer-
ization that allows for tuning of electron transfer
between the subunits could represent a novel layer
of activity regulation.
In GEMMA experiments and activity assays all four
dNTPs were shown to induce dimerization and act
as effectors via the specificity site but the formation
of higher oligomers was induced only by dATP and
dGTP. We therefore sought to study whether there
is an additional effector binding site, different from
the specificity site. Attempted direct binding assays
and ITC experiments did not yield conclusive results
EXPERIMENTAL PROCEDURES
Bioinformatics — All full-length NrdJ se-
quences in RefSeq were downloaded from RNRdb
(http://rnrdb.pfitmap.org) (2017-02-17) and clus-
tered at 75% identity with USEARCH (29) to create
a selection of representative sequences. Sequences
were aligned with Probcons (30) and reliable align-
ment positions were selected with BMGE (31) using
(
data not shown), but experiments with mixtures of
effectors were indicative of a third nucleotide bind-
ing site. While addition of dTTP to a CDP/dATP
assay inhibited enyzme activity in accordance with
7

Catalytic turnover in class II ribonucleotide reductase
the BLOSUM30 model. A phylogeny was estimated
with RAxML (32) in rapid bootstrapping mode with
automatic bootstopping (extended majority rule) us-
ing the PROTGAMMAAUTO model. An HMM
profile for the CRD domain was constructed from
the CRD domain found in 217 sequences in the
alignment used for phylogenetic estimation. The
Skylign server was used to graphically represent the
profile as a sequence logo (33).
of 0.05-0.1 and incubated at 37°C and 150 RPM.
After reaching an OD600 of 0.8 the temperature
was decreased to 15°C (NrdJds and truncations) or
20°C (glutaredoxin, thioredoxins and thioredoxin
reductase) After induction with IPTG with a final
concentration of 0.1 mmol L , the cultures were
incubated for 4 h (NrdJds and truncations) or 18 h
(glutaredoxin, thioredoxins and thioredoxin reduc-
tase). Cells were then harvested by centrifugation.
Cell lysis was performed in HisWash buffer (50
−1
Secondary structure prediction was performed with
the Chou & Fasman Secondary Structure Prediction
Server (CFSSP) (34, 35). The homology model
of the C-terminus was constructed via the Swiss
model server (36–39). After automated template
search, a model was built based on the template
with the highest sequence identity. For the model
discussed above, the 35 amino acid C-terminus of
tRNA(Ile2) 2-agmatinylcytidine synthetase from A.
fulgidus (PDB: 4RVZ) with a sequence identity of
−
1
−1
mmol L Tris pH 7.5, 300 mmol L NaCl, 20
−1
−1
mmol L imidazole, 1 mmol L DTT) by high
pressure homogenization three times at 15 MPa.
Purification was performed via Immobilized metal
ion affinity chromatography (IMAC). The super-
natant from the cell lysis was applied on a HisTrap
column washed with HisWash buffer and eluted
−1
with HisElute buffer (50 mmol L Tris pH 7.5, 300
−1
−1
mmol L NaCl, 500 mmol L imidazole, 1 mmol
−1
31% was applied as template.
L
DTT).
In a second purification step, elution fractions from
the IMAC were loaded on Highload™ 16/600 Su-
perdex 200 pg gel filtration column and run with
Molecular cloning — The genes from the
NrdJd from S. nassauensis (Snas 4560; full length
and CRD) and S. clavuligerus (AJ224870) as well
as the genes from the thioredoxins (Snas 6430, Snas
−1
SEC buffer (50 mmol L Tris pH 7.5, 300 mmol
−
1
−1
L
NaCl, 1 mmol L DTT). If not used imme-
2
647), the glutaredoxin (Snas 1785) and the thiore-
diately, purified proteins were frozen in N2
stored at -80°C.
and
(liq)
doxin reductase (Snas 6431) were amplified from
genomic DNA via PCR. All genes were cloned into
a pET22b(+) vector via NdeI and HindIII restric-
tion sites adding a N-terminal His-tag connected
via a thrombin restriction site. For recombinant
expression, the vectors were transformed in E. coli
BL21(DE3).
Analytical size exclusion chromatography
(SEC) — SEC for analytical purposes were run on
a Superdex™ increase 200 3.2 column. 25 µL of
−1
protein sample with a concentration of 1.0 mg mL
were injected into the column and run with SEC
−
1
−1
buffer (50 mmol L Tris pH 7.5, 300 mmol L
−1
Mutagenesis — The C-terminal truncation
of the NrdJd from S. nassauensis, NrdJd∆CRD,
as well as all cysteine to alanine mutations were
performed via site-directed mutagenesis using mis-
matched oligonucleotides. For NrdJd∆CRD the
GAG bases coding for Glu715 were exchanged by
a TAA stop codon. For the cysteine to alanine mu-
tations the TGC and TGA codons for cysteine were
replaced by GCA coding for alanine.
NaCl, 1 mmol L DTT) in the presence or absence
of 100 µmol L effector.
−1
Gas-phase electrophoretic mobility molec-
ular analyzer (GEMMA) — GEMMA experiments
were performed on TSI3480 electrospray aerosol
generator, TSI3080 electrostatic classifier and
TSI3025A ultrafine condensation particle counter.
The buffer of the protein samples was changed to
−1
4
0 mmol L ammonium acetate (pH 7.5) via a de-
Recombinant expression and purifica-
salting column. Substrates and effectors were added
to the protein sample with an equimolar amount of
tion — Expression was performed in LB-medium in
6
00 mL scale in baffled shake flasks. The medium
MgCl . The samples were run in the final condi-
2
−1
was inoculated from a preculture to a starting OD600
tions: 40 mmol L ammonium acetate (pH 7.8),
8

Catalytic turnover in class II ribonucleotide reductase
−
1
0
.005% Tween 20, 0.2 mmol L DTT, 100 µmol
buffer was prepared with KH PO by setting the
2
4
−1
−1
L
effector, 100 µmol L substrate. The final
pH value with KOH. The analytes were eluted in a
linear gradient of methanol from 10 - 30% (v/v).
Photometric assays for thioredoxins and glutaredox-
ins based on the artificial electron acceptors Insulin
and HED were performed as described in earlier
studies (40, 41). The thioredoxin assay contains 50
concentrations of the proteins applied were 0.075
−1
mg mL for S. nassauensis NrdJd full length and
−
1
0
,025 mg mL for NrdJd∆CRD. The samples were
run at a pressure of 1.7 PSI, a current of 160-200
nA and a voltage of 1.8 - 2.0 kV. For data collection,
five runs per sample were conducted.
−
1
−1
mmol L Tris (pH 8.0), 20 µmol L thioredoxin,
−
1
−1
1
µmol L thioredoxin reductase, 750 µg mL
−1
Enzyme activity assay — Specific enzyme
activities were determined by measurement of the
conversion of NDP to dNDP, catalyzed by the en-
zyme in the presence of different dNTPs as allosteric
effectors. All enzymatic assays were performed in
triplicates. Under standard conditions the assay
insulin and 250 µmol l NADPH. The glutaredoxin
−1
assay contains 50 mmol L Tris (pH 8.0), 20 µmol
−1
−1
L glutaredoxin, 7.5 U mL glutathione reductase,
−
1
−1
1 mmol L glutathione, 750 µmol l HED and 250
−
1
µmol l NADPH. All assays were run in triplicates
for 5 min at room temperature.
−
1
1
−1
contains 50 mmol L Tris (pH 8.0), 20 mmol L
In the combined redoxin/RNR assays all reaction
components for the single RNR and redoxin assays
were combined with the following adjustments: 5
−
−1
MgCl , 50 mmol L DTT, 1.0 mmol L dATP, 2.0
2
−
1
−1
mmol L CDP, 5.0 µmol L AdoCbl and between
−
1
−1
−1
−1
2
and 5 µmol L enzyme ( 0.2-0.5 mg mL ). In
µmol L RNR, 5 mmol L glutathione, 20 µmol
−1
effector titration experiments, a concentration range
L thioredoxin/glutaredoxin. The assays were incu-
bated for 30 min at room temperature and analyzed
photometrically as well as by HPLC as described
above.
−
1
from 4 to 1000 µmol L was applied, including a
measurement without effector. In the four substrate
−
1
assays, 0.5 mmol L of each substrate (ADP, CDP,
−1
GDP, UDP) were applied with 0.5 mmol L of the
tested effector. The assays, performed in 50 µL
scale, were incubated 15 or 30 min at room temper-
ature and stopped by addition of 50 µL methanol.
After addition of 200 µL water, the samples were an-
alyzed via HPLC. 20 µL of the sample were injected
onto a C18 column and run with the following buffer:
Inductively coupled plasma mass spectrom-
etry (ICP) — ICP measurements were performed at
the Z.B.M. - Analytical Laboratory at the Hebrew
University of Jerusalem. ICP-OES analysis was
performed after an acidic hot block digestion. ICP-
OES measurements were performed in triplicates
applying yttrium as internal standard.
−1
5
0
0 mmol L KP (pH 7.5), 10% (v/v) methanol,
.25% (v/v) tetrabutylammonium hydroxide. The
i
Acknowledgments: We thank Fredrik Tholander for generously sharing his HPLC-based RNR assay method
and equipment with us, and Vasiliy Rosen for the performance of the ICP experiments at Hebrew University
of Jerusalem. We also thank the two anonymous revierwers for their constructive revision of our manuscript.
The work was supported by the Swedish Research Council (2016-01920), Swedish Cancer Society (CAN
2016/670), Wenner-Gren Foundations, Carl Trygger’s Foundation and Kempe Foundations.
Conflict of interest: No conflict of interest declared.
Author contributions: The concept of the manuscript and the writing were performed by CL, DL and BMS.
CL performed a large part of the experimental work. DL and BMS performed the phylogenetic investigation
of class II RNRs. VRJ and AH helped with the performance and interpretation of the GEMMA experiments.
IRG and MS contributed in the production and characterization of the studied enzymes. MC established and
performed HPLC analytics and helped interpreting the results. All authors critically reviewed the manuscript
and approve with its publication.
9

Catalytic turnover in class II ribonucleotide reductase
REFERENCES
1
.
.
Nordlund, P. and Reichard, P. (2006) Ribonucleotide reductases. Annu. Rev. Biochem. 75, 681–706.
Lundin, D., Berggren, G., Logan, D. T., and Sjöberg, B. M. (2015) The origin and evolution of
ribonucleotide reduction. Life 5(1), 604–636.
2
3.
4.
5.
Stubbe, J., Nocera, D. G., Yee, C. S., and Zhang, M. C. Y. (2003) Radical initiation in the
class I ribonucleotide reductase: long-range proton-coupled electron transfer. Chem. Rev. 103,
2167–2201.
Åberg, A., Hahne, S., Karlsson, M., Larsson, Å., Ormö, M., Åhgren, A., and Sjöberg, B. M.
(
1989) Evidence for two different classes of redox-active cysteines in ribonucleotide reductase of
Escherichia coli. J. Biol. Chem. 264(21), 12249–12252.
Mao, S. S., Yu, G. X., Chalfoun, D., and Stubbe, J. (1992) Characterization of C439SR1, a mutant
of Escherichia coli ribonucleotide diphosphate reductase: evidence that C439 is a residue essential
for nucleotide reduction and C439SR1 is a protein possessing novel thioredoxin-like activity.
Biochemistry 31(40), 9752–9759.
6
.
.
Logan, D. T., Andersson, J., Sjöberg, B. M., and Nordlund, P. (1999) A glycyl radical site in the
crystal structure of a class III ribonucleotide reductase. Science 283, 1499–1504.
Torrents, E., Buist, G., Liu, A., Eliasson, R., Kok, J., Gibert, I., Gräslund, A., and Reichard, P.
7
(
2000) The anaerobic (class III) ribonucleotide reductase from Lactococcus lactis. J. Biol. Chem.
275(4), 2463–2471.
8
9
1
.
Mulliez, E., Ollagnier, S., Fontecave, M., Eliasson, R., and Reichard, P. (1995) Formate is the
hydrogen donor for the anaerobic ribonucleotide reductase from Escherichia coli. Proc. Natl. Acad.
Sci. USA 92, 8759–8762.
.
Wei, Y., Li, B., Prakash, D., Ferry, J. G., Elliott, S. J., and Stubbe, J. (2015) A ferredoxin disul-
fide reductase delivers electrons to the Methanosarcina barkeri class III ribonucleotide reductase.
Biochemistry 54(47), 7019–28.
0.
Lawrence, C. C. and Stubbe, J. (1998) The function of adenosylcobalamin in the mechanism of
ribonucleoside triphosphate reductase from Lactobacillus leichmannii. Curr. Opin. Chem. Biol. 2,
6
50–655.
1
1.
2.
Eliasson, R., Pontis, E., Jordan, A., and Reichard, P. (1999) Allosteric control of three B12-
dependent (class II) ribonucleotide reductases. J. Biol. Chem. 274(11), 7182–7189.
Sintchak, M. D., Arjara, G., Kellogg, B. A., Stubbe, J., and Drennan, C. L. (2002) The crystal
structure of class II ribonucleotide reductase reveals how an allosterically regulated monomer
mimics a dimer. Nat. Struct. Biol. 9(4), 293–300.
1
1
1
1
3.
4.
5.
Larsson, K. M., Jordan, A., Eliasson, R., Reichard, P., Logan, D. T., and Nordlund, P. (2004)
Structural mechanism of allosteric substrate specificity regulation in a ribonucleotide reductase.
Nat. Struct. Mol. Biol. 11(11), 1142–1149.
Booker, S., Licht, S., Broderick, J., and Stubbe, J. (1994) Coenzyme B12-dependent ribonucleotide
reductase: evidence for the participation of five cysteine residues in ribonucleotide reduction.
Biochemisty 33, 12676–12685.
Borovok, I., Schreiber, R., Holmgren, A., Kreisberg-Zakarin, R., Myslovati, M., Cohen, G.,
Yanko, M., Åslund, F., and Aharonowitz, Y. (2002) Streptomyces spp. contain class Ia and class II
ribonucleotide reductases: expression analysis of the genes in vegetative growth. Microbiology
148, 391–404.
1
6.
7.
Torrents, E., Poplawski, A., and Sjöberg, B. M. (2005) Two proteins mediate class II ribonucleotide
reductase activity in Pseudomonas aeruginosa: expression and transcriptional analysis of the
aerobic enzymes. J. Biol. Chem. 280(17), 16571–16578.
1
Crona, M., Hofer, A., Astorga-Wells, J., Sjöberg, B. M., and Tholander, F. (2015) Biochemical
characterization of the split class II ribonucleotide reductase from Pseudomonas aeruginosa. PLoS
10

Catalytic turnover in class II ribonucleotide reductase
One 10(7), e0134293.
18.
19.
20.
Averett, D. R., Lubbers, C., Elion, G. B., and Spector, T. (1983) Ribonucleotide reductase induced
by Herpes simplex type 1 virus. J. Biol. Chem. 258(16), 9831–9838.
Domkin, V. and Chabes, A. (2014) Phosphines are ribonucleotide reductase reductants that act via
C-terminal cysteines similar to thioredoxins and glutaredoxins. Sci. Rep. 4, 5539.
Li, F., Dong, J., Hu, X., Gong, W., Li, J., Shen, J., Tian, H., and Wang, J. (2015) A covalent
approach for site-specific RNA labeling in mammalian cells. Angew. Chem. Int. Ed. Engl. 54(15),
4597–4602.
2
2
2
2
1.
2.
3.
4.
Bracey, M. H., Christiansen, J., Tovar, P., Cramer, S. O., and Bartlett, S. G. (1994) Spinach car-
bonic anhydrase: investigation of the zinc-binding ligands by site-directed mutagenesis, elemental
analysis, and EXAFS. Biochemistry 33(44), 13126–13131.
Zhou, Z. S., Peariso, K., Penner-Hahn, J. E., and Matthews, R. G. (1999) Identification of
the zinc ligands in cobalamin-independent methionine synthase (MetE) from Escherichia coli.
Biochemistry 38(48), 15915–15926.
Logan, D. T., Mulliez, E., Larsson, K. M., Bodevin, S., Atta, M., Garnaud, P. E., Sjöberg, B. M.,
and Fontecave, M. (2003) A metal-binding site in the catalytic subunit of anaerobic ribonucleotide
reductase. Proc. Natl. Acad. Sci. USA 100(7), 3826–3831.
Luttringer, F., Mulliez, E., Dublet, B., Lemaire, D., and Fontecave, M. (2009) The Zn center of the
anaerobic ribonucleotide reductase from E. coli. J. Biol. Inorg. Chem. 14(6), 923–933.
Bertini, I., Luchinat, C., and Monnanni, R. (1985) Zinc enzymes. J. Chem. Educ. 62(11), 924–927.
Pace, N. J. and Weerapana, E. (2014) Zinc-binding cysteines: diverse functions and structural
motifs. Biomolecules 4(2), 419–434.
2
2
5.
6.
2
7.
Zhang, Z., Yang, K., Chen, C. C., Feser, J., and Huang, M. (2007) Role of the C terminus of the
ribonucleotide reductase large subunit in enzyme regeneration and its inhibition by Sml1. Proc.
Natl. Acad. Sci. USA 104(7), 2217–2222.
2
2
3
3
8.
9.
0.
1.
Berardi, M. J. and Bushweller, J. H. (1999) Binding specificity and mechanistic insight into
glutaredoxin-catalyzed protein disulfide reduction. J. Mol. Biol. 292, 151–161.
Edgar, R. C. (2010) Search and clustering orders of magnitude faster than BLAST. Bioinforma.
26(19), 2460–2461.
Do, C., Mahabhashyam, M. S. P., and Batzoglou, S. (2005) ProbCons: Probabilistic consistency-
based multiple sequence alignment. Genome Res. 15(2), 330–340.
Criscuolo, A. and Gribaldo, S. (2010) BMGE (Block Mapping and Gathering with Entropy): a
new software for selection of phylogenetic informative regions from multiple sequence alignments.
BMC Evol. Biol. 10, 210.
3
2.
3.
Stamatakis, A. (2014) RAxML version 8: a tool for phylogenetic analysis and post-analysis of large
phylogenies. Bioinforma. 30(9), 1312–1213.
3
Wheeler, T. J., Clements, J., and Finn, R. D. (2014) Skylign: a tool for creating informative,
interactive logos representing sequence alignments and profile hidden Markov models. BMC
Bioinfo. 15(7).
3
3
3
3
3
4.
5.
6.
7.
8.
Chou, P. Y. and Fasman, G. D. (1974) Conformational parameters for amino acids in helical,
β-sheet, and random coil regions calculated from proteins. Biochemistry 13(2), 211–222.
Chou, P. Y. and Fasman, G. D. (1974) Prediction of protein conformation. Biochemistry 13(2),
222–245.
Arnold, K., Bordoli, L., Kopp, J., and Schwede, T. (2006) The SWISS-MODEL workspace: a
web-based environment for protein structure homology modelling.. Bioinforma. 22(2), 195–201.
Kiefer, F., Arnold, K., Kunzli, M., Bordoli, L., and Schwede, T. (2009) The SWISS-MODEL
repository and associated resources. Nucleic Acids Res. 37(Database issue), 387–92.
Guex, N., Peitsch, M. C., and Schwede, T. (2009) Automated comparative protein structure mod-
11

Catalytic turnover in class II ribonucleotide reductase
eling with SWISS-MODEL and Swiss-PdbViewer: a historical perspective. Electrophoresis 30
Suppl 1, S162–173.
Biasini, M., Bienert, S., Waterhouse, A., Arnold, K., Studer, G., Schmidt, T., Kiefer, F., Gallo Cas-
sarino, T., Bertoni, M., Bordoli, L., and Schwede, T. (2014) SWISS-MODEL: modelling protein
tertiary and quaternary structure using evolutionary information. Nucleic Acids Res. 42(Web Server
issue), W252–8.
39.
4
4
0.
1.
Holmgren, A. (1979) Thioredoxin catalyzes the reduction of insulin disulfides by dithiothreitol and
dihydrolipoamide. J. Biol. Chem. 254(19), 9627–9632.
Vlamis-Gardikas, A., Åslund, F., Spyrou, G., Bergman, T., and Holmgren, A. (1997) Cloning,
overexpression, and characterization of glutaredoxin 2, an atypical glutaredoxin from Escherichia
coli. J. Biol. Chem. 272(17), 11236–11243.
12

Catalytic turnover in class II ribonucleotide reductase
FIGURES
Figure 1: Phylogenetic analysis of NrdJ and sequence analysis of NrdJd. A) Maximum likelihood
phylogenetic tree of a representative selection of NrdJ sequences. Subclasses with characterized members
(
NrdJm (green), NrdJd (red) and NrdJa+b (yellow)) and candidate subclasses (cyan, purple, orange and light
blue) are indicated with color. Lactobacillus leichmannii RNR (Ll) is located in subclass NrdJm, while
the RNR from Thermotoga maritima (Tm) is not part of a defined subclass B) Appearance of the Cysteine
Rich Domain (CRD) mapped on a phylogenetic tree of a selection of non-monomeric NrdJ sequences. The
selected subset is marked by the dotted line in panel A. Subclasses with characterized members (NrdJd (red)
and NrdJa+b (yellow)) and candidate subclasses (cyan, purple, orange and light blue) are indicated with
color. The presence of a CRD is indicated with a blue circle, absence with a white circle. The NrdJa+b
suclass has the encoded CRD as a separate gene. C) Logo of the CRD derived from 217 non-redundant
sequences from subclasses NrdJd and unclassified NrdJ.
13

Catalytic turnover in class II ribonucleotide reductase
Figure 2: SDS-PAGE of the final purification of heterologously expressed S. nassauensis enzymes.
Electrophoresis was performed with a 4-12% Bis-Tris Protein Gel. NrdJd-wt (Snas 4560): full length NrdJd
enzyme (1-947); NrdJd∆CRD: C-terminal truncation without CRD (1-714), CRD: CRD of NrdJd (715-947),
glutaredoxin (Snas 1785), thioredoxin 1 (Snas 2647), thioredoxin 2 (Snas 6430), TR: thioredoxin reductase
(
Snas 6431)
14

Catalytic turnover in class II ribonucleotide reductase
Figure 3: Activity of NrdJd-wt and NrdJd∆CRD in the presence of effector. A/B) Four substrate assay
for NrdJd-wt (A) and NrdJd∆CRD (B) with or without effector. Each reaction contains the four potential
substrates ADP, CDP, GDP and UDP plus one of the potential effectors dATP dCTP, dGTP, dTTP or ATP.
C/D) Influence of the effector concentration on enzyme activity of NrdJd-wt (C) and NrdJd∆CRD (D) with
one of the potential effectors dATP, dCTP, dGTP, dTTP or ATP. Titrations were performed in a concentration
−
1
range from 4 to 1000 µmol L , including a measurements without effector. Data where obtained in three
independent experiments. Error bars indicate the standard deviation from the mean value.
15

Catalytic turnover in class II ribonucleotide reductase
Figure 4: Oligomeric states of NrdJd and NrdJd∆CRD determined by SEC and GEMMA. Size
−
1
exclusion chromatography of A) NrdJd-wt and B) NrdJd∆CRD with 1 mg mL protein each. C) GEMMA
−
1
−1
experiments with 0.075 mg mL NrdJd-wt and 100 µmol L of one of the potential effectors. D) GEMMA
−
1
−1
experiment with 0.025 mg mL NrdJd∆CRD and and 100 µmol L of one of the potential effectors. E)
−
1
GEMMA experiment with 0.075 mg mL NrdJd-wt and a titration of the dATP concentration. F) GEMMA
−
1
experiment with 0.075 mg mL NrdJd-wt and a titration of the dGTP concentration. All GEMMA
experiments represent 5 separate measurements for each condition.
16

Catalytic turnover in class II ribonucleotide reductase
Figure 5: Reductant dependence of NrdJd and NrdJd∆CRD. Enzyme activity of NrdJd-wt (red) and
NrdJd∆CRD (green) in the presence of different concentrations of the reductants A) DTT and B) TCEP.
Assays were performed with CDP as substrate and dATP as effector. All experiments were performed in
triplicates. Error bars indicate the standard deviation from the mean value.
17

Catalytic turnover in class II ribonucleotide reductase
Figure 6: Homology model and biochemical characterization of the C-terminal zinc binding site. A)
Homology model of the last 30 amino acids of the CRD from S. nassauensis NrdJd, based on the crystal
structure of tRNA(Ile2) 2-agmatinylcytidine synthetase (PDB: 4RVZ) (20). B) Logo of the C-terminal end
of S. nassauensis NrdJd-wt. C) Activities with the reductant TCEP (narrow bars; black) and DTT (narrow
bars; green) and zinc content (broad bars) of NrdJd-wt and the cysteine to alanine variants. The colors red
and blue refer to the predicted location of the respective cysteine residue. Data where obtained in three
independent experiments. Error bars indicate the standard deviation from the mean value.
18

Catalytic turnover in class II ribonucleotide reductase
Figure 7: Enzyme activity of mixtures of mutant proteins. A) Specific activity per monomer of the
C933A and C370A proteins tested individually and in an equimolar C933A/C370A mixture (mutant mix).
B) Relative activities of NrdJd-wt and the mutant mix in effector titrations with the substrate effector pairs
−
1
−1
CDP/dATP (100% equals kcat = 15.3 min ) and GDP/dTTP (100% equals k
= 3.1 min ) and TCEP
cat
as reductant. C) Activities of NrdJd-wt and the mutant mix with the substrates CDP or GDP and effector
−
1
−1
mixtures. TCEP was applied as reductant. Concentrations: dATP 250 µmol L , dTTP 1 mmol L . Data
where obtained in three independent experiments. Error bars indicate the standard deviation from the mean
value.
19

Catalytic turnover in class II ribonucleotide reductase
Figure 8: Activity assays in presence of potential native electron donor systems. A) Specific activity of
glutaredoxin and the thioredoxins in the presence or absence of NrdJd-wt (+RNR and -RNR), respectively.
The assays were performed with or without additional oxidant used as substrate (+Ox or -Ox). The additional
oxidants were insulin for the thioredoxins and HED for glutaredoxin. B) Specific activity of NrdJd-wt in
combination with glutaredoxin and the thioredoxins with the substrate CDP. C) Activity assays with NrdJd-wt
−1
in the presence of cell lysate preparations from S. nassauensis with (green) or without (red) 250 µmol L
NADPH addition: protein-free lysate (Prep 1), cleared lysate (Prep 2) and complete lysate (Prep 3). No
artificial reductant such as DTT was added. As positive and negative controls, RNR-assays with (+DTT)
and without (-) the reductant DTT were performed. RNR activity in the preparations and controls was
determined by HPLC measurement of CDP consumption and dCDP synthesis. All data where obtained in
three independent experiments. Error bars indicate the standard deviation from the mean value.
20

A unique cysteine-rich Zn-finger domain present in a majority of class II
ribonucleotide reductases mediates catalytic turnover
Christoph Loderer, Venkateswara Rao Jonna, Mikael Crona, Inna Rozman Grinberg,
Margareta Sahlin, Anders Hofer, Daniel Lundin and Britt-Marie Sjöberg
J. Biol. Chem. published online October 2, 2017
Access the most updated version of this article at doi: 10.1074/jbc.M117.806331
Alerts:
•
•
When this article is cited
When a correction for this article is posted
Click here to choose from all of JBC's e-mail alerts
Supplemental material:
http://www.jbc.org/content/suppl/2017/10/02/M117.806331.DC1
This article cites 0 references, 0 of which can be accessed free at
http://www.jbc.org/content/early/2017/10/02/jbc.M117.806331.full.html#ref-list-1