Towards a functional identification of catalytically inactive [Fe]-hydrogenase paralogs

Article abstract of DOI:10.1111/febs.13351

[Fe]-hydrogenase (Hmd), an enzyme of the methanogenic energy metabolism, harbors an iron-guanylylpyridinol (FeGP) cofactor used for H₂ cleavage. The generated hydride is transferred to methenyl-tetrahydromethanopterin (methenyl-H₄MPT⁺). Most hydrogenotrophic methanogens contain the hmd-related genes hmdII and hmdIII. Their function is still elusive. We were able to reconstitute the HmdII holoenzyme of Methanocaldococcus jannaschii with recombinantly produced apoenzyme and the FeGP cofactor, which is a prerequisite for in vitro functional analysis. Infrared spectroscopic and X-ray structural data clearly indicated binding of the FeGP cofactor. Methylene-H₄MPT binding was detectable in the significantly altered infrared spectra of the HmdII holoenzyme and in the HmdII apoenzyme-methylene-H₄MPT complex structure. The related binding mode of the FeGP cofactor and methenyl-H₄MPT⁺ compared with Hmd and their multiple contacts to the polypeptide highly suggest a biological role in HmdII. However, holo-HmdII did not catalyze the Hmd reaction, not even in a single turnover process, as demonstrated by kinetic measurements. The found inactivity can be rationalized by an increased contact area between the C- and N-terminal folding units in HmdII compared with in Hmd, which impairs the catalytically necessary open-to-close transition, and by an exchange of a crucial histidine to a tyrosine. Mainly based on the presented data, a function of HmdII as Hmd isoenzyme, H₂ sensor, FeGP-cofactor storage protein and scaffold protein for FeGP-cofactor biosynthesis could be excluded. Inspired by the recently found binding of HmdII to aminoacyl-tRNA synthetases and tRNA, we tentatively consider HmdII as a regulatory protein for protein synthesis that senses the intracellular methylene-H₄MPT concentration. Database Structural data are available in the Protein Data Bank under the accession numbers 4YT8; 4YT2; 4YT4 and 4YT5. The genome of some methanogens contains besides the [Fe]-hydrogenase (Hmd) a related protein termed HmdII. We showed by X-ray and infrared spectroscopic data that HmdII binds iron-guanylylpyridinol, the Hmd specific cofactor, and methylene-tetrahydromethanopterin, the substrate of Hmd. However, the HmdII holoenzyme did not catalyze the Hmd reaction. We propose HmdII as a regulatory protein that senses the intracellular methylene-tetrahydromethanopterin concentration.

Full text of DOI:10.1111/febs.13351

Towards a functional identification of catalytically
inactive [Fe]-hydrogenase paralogs
Takashi Fujishiro1, Kenichi Ataka2, Ulrich Ermler3 and Seigo Shima1,4
€
1 Max-Planck-Institut fur terrestrische Mikrobiologie, Marburg, Germany
€
2 Department of Physics, Freie Universitat Berlin, Germany
€
3 Max-Planck-Institut fur Biophysik, Frankfurt/Main, Germany
4 PRESTO, Japan Science and Technology Agency (JST), Saitama, Japan
Keywords
[Fe]-hydrogenase (Hmd), an enzyme of the methanogenic energy metabo-
lism, harbors an iron-guanylylpyridinol (FeGP) cofactor used for H2
cleavage. The generated hydride is transferred to methenyl-tetrahy-
[Fe]-hydrogenase; Hmd paralogs;
methanogenic archaea; sensor protein;
tetrahydromethanopterin
dromethanopterin
(methenyl-H4MPT+).
Most
hydrogenotrophic
methanogens contain the hmd-related genes hmdII and hmdIII. Their
function is still elusive. We were able to reconstitute the HmdII holoen-
zyme of Methanocaldococcus jannaschii with recombinantly produced
apoenzyme and the FeGP cofactor, which is a prerequisite for in vitro
functional analysis. Infrared spectroscopic and X-ray structural data
clearly indicated binding of the FeGP cofactor. Methylene-H4MPT
binding was detectable in the significantly altered infrared spectra of the
HmdII holoenzyme and in the HmdII apoenzyme–methylene-H4MPT
complex structure. The related binding mode of the FeGP cofactor and
methenyl-H4MPT+ compared with Hmd and their multiple contacts to
the polypeptide highly suggest a biological role in HmdII. However,
holo-HmdII did not catalyze the Hmd reaction, not even in a single
turnover process, as demonstrated by kinetic measurements. The found
inactivity can be rationalized by an increased contact area between the
C- and N-terminal folding units in HmdII compared with in Hmd, which
impairs the catalytically necessary open-to-close transition, and by an
exchange of a crucial histidine to a tyrosine. Mainly based on the
presented data, a function of HmdII as Hmd isoenzyme, H2 sensor,
FeGP-cofactor storage protein and scaffold protein for FeGP-cofactor
biosynthesis could be excluded. Inspired by the recently found binding of
HmdII to aminoacyl-tRNA synthetases and tRNA, we tentatively con-
sider HmdII as a regulatory protein for protein synthesis that senses the
intracellular methylene-H4MPT concentration.
Correspondence
€
S. Shima, Max-Planck-Institut fur
terrestrische Mikrobiologie,
Karl-von-Frisch-Straße 10, 35043 Marburg,
Germany
Fax: +49 6421 178199
Tel: +49 6421 178122
E-mail: shima@mpi-marburg.mpg.de
(Received 20 April 2015, revised 15 June
2015, accepted 17 June 2015)
doi:10.1111/febs.13351
Database
Structural data are available in the Protein Data Bank under the accession numbers 4YT8;
4YT2; 4YT4 and 4YT5.
Abbreviations
apo-Hmd, Hmd apoenzyme; apo-HmdII, HmdII apoenzyme; C-unit, central globular unit; FeGP, iron-guanylylpyridinol; H4MPT,
tetrahydromethanopterin; holo-Hmd, Hmd holoenzyme reconstituted with the FeGP cofactor; holo-HmdII, HmdII holoenzyme reconstituted
with the FeGP cofactor; N-unit, N-terminal domain unit; SeMet, L-selemomethionine.
3412
FEBS Journal 282 (2015) 3412–3423 ª 2015 FEBS

T. Fujishiro et al.
Potential function of an [Fe]-hydrogenase paralog
The genomes of many hydrogenotrophic methano-
genic archaea harbor genes related to the hmd gene. The
encoding proteins HmdII and HmdIII have 20–30%
amino acid sequence identity with Hmd [17]; those of
M. jannaschii (HmdII, MJ1338 and HmdIII, MJ0715)
show ~ 20% identity to Hmd (MJ0784) [18]. Some
hydrogenotrophic methanogens, e.g. Methanococ-
cus voltae (US DOE Joint Genome Institute, Walnut
Creek, CA, USA) and Methanobrevibacter smithii [19],
have an hmd gene and produce active Hmd [20] but their
genomes do not harbor genes encoding HmdII and
HmdIII. The hmdII gene has been recently identified in
the genome of the hyperthermophilic bacterium Desul-
furobacterium thermolithotrophum, but this non-metha-
nogen lacks the hmd gene [21].
Introduction
[Fe]-hydrogenase, also known as H2-forming methylene-
tetrahydromethanopterin (methylene-H4MPT) dehydro-
genase (Hmd), is one of three known hydrogenases [1–5].
Hmd is involved in the methanogenic pathway of many
hydrogenotrophic methanogenic archaea and reversibly
catalyzes the stereoselective hydrogenation of methenyl-
H4MPT+ to methylene-H4MPT using H2 as a hydride
donor (Fig. 1A) [6,7].
Hmd occurs as a homodimer consisting of a central
globular unit (termed C-unit) formed by the C-termi-
nal domains of each subunit and two N-terminal
Rossmann-like domains (termed N-unit). Hmd was
found in an ‘open’ or ‘closed’ conformation [3],
defined by the orientation of the C- and N-units
relative to each other. The active site is located
between them and the open-to-close transition is
considered to play a crucial role in the catalytic cycle
[2,6]. Each N-unit of Hmd harbors one iron-guany-
lylpyridinol (FeGP) cofactor as a prosthetic group
(Fig. 1B) [4,8]. The iron center is coordinated by two
cis-CO, one cysteine-thiolate, one solvent, and the
pyridinol-nitrogen and the acyl-carbon of the
guanylylpyridinol ligand [9–13]. Based on the X-ray
crystal structure of Hmd complexed with isocyanide
inhibitors, the solvent-binding site of the organometal-
lic iron–carbonyl complex has been proposed to serve
as the H2-binding site [14,15]. The FeGP cofactor can
be extracted from the Hmd holoenzyme (holo-Hmd)
[13,16] and subsequently used to reconstitute a fully
active holo-Hmd from an Hmd apoenzyme
(apo-Hmd). The X-ray structure of the reconstituted
holo-Hmd of Methanocaldococcus jannaschii revealed
binding of the intact FeGP cofactor [1,2,4].
HmdII and HmdIII from various methanogens have
a sequence identity of ~ 60–80% relative to each other
(those from M. jannaschii 65%), which indicates that
HmdII and HmdIII have similar structures and func-
tions. After their identification in genomes, HmdII and
HmdIII were tentatively considered as Hmd homologs
that catalyze the same reaction as Hmd [22]. Later on,
preliminary data suggested that HmdII can bind the
FeGP cofactor [23,24]; however, neither HmdII nor
HmdIII have been reconstituted with the FeGP cofactor
and kinetically analyzed. Therefore, their function
remains elusive.
Here, we present the reconstitution of the HmdII
holoenzyme (holo-HmdII) from the HmdII apoenzyme
(apo-HmdII) and the FeGP cofactor, its infrared spec-
troscopic characterization, kinetic studies and X-ray
structure analysis of the apo-HmdII, holo-HmdII and
apo-HmdII with bound methylene-H4MPT. Based on
the obtained results, we discuss potential functions of
the Hmd-related proteins.
A
Fig. 1. Hmd-catalyzed reaction and
structure of the FeGP cofactor. (A) Hmd-
catalyzed reversible hydrogenation of
methenyl-H4MPT+. R indicates the residual
part of the H4MPT derivative. (B) Chemical
B
structure of the FeGP cofactor. Upon
reaction, H2 replaces the solvent at the
FeGP cofactor iron, is heterolytically
cleaved, and the generated hydride is
subsequently transferred from the pro-R
side to the carbocationic C14a of
methenyl-H4MPT+ [7].
FEBS Journal 282 (2015) 3412–3423 ª 2015 FEBS
3413

Potential function of an [Fe]-hydrogenase paralog
T. Fujishiro et al.
X-Ray structure analysis of HmdII of
Methanocaldococcus jannaschii
Results
Infrared analysis of holo-HmdII
Overall structure
In the infrared spectrum of holo-HmdII, the peaks
of the antisymmetric and symmetric stretching band
of the iron-ligating CO were split and have frequen-
The crystal structures of apo-HmdII and reconstituted
holo-HmdII from M. jannaschii have been determined
ꢀ
at 1.9 and 2.2 A resolution, respectively, using the
cies of 2015 + 1994 cmꢀ1 and 1946 + 1930 cmꢀ1
,
Se-methionine-based multiple anomalous dispersion
method for initial phase determination. The structures
of apo-HmdII and reconstituted holo-HmdII are
almost identical. Like Hmd, HmdII is a homodimer
composed of a C-unit formed by the intertwined C-ter-
minal segments (residues 205–355) of the two subunits
and two N-units both binding one FeGP cofactor
(residues 1–204) [3]. Despite related global architec-
respectively (Fig. 2). These frequencies resembled
those of holo-Hmd (2011 and 1944 cmꢀ1), which is
indicative of similar environments for the iron com-
plexes. The peak splitting argued for at least two
distinct FeGP-binding modes. Addition of methylene-
H4MPT to the holo-HmdII solution resulted in a
minor frequency shift and a substantial intensity
change of the CO bands (Fig. 2). Moreover, the CO
peaks are sharper in holo-HmdII than in holo-Hmd,
implying that the CO ligands are more encapsulated
after binding methylene-H4MPT. By contrast,
methenyl-H4MPT+ changed the infrared spectrum of
holo-HmdII only slightly, but appears also to bind
in the vicinity of the CO ligands. The greater
influence of methylene-H4MPT on the environment
of CO might reflect larger conformational changes of
the protein.
ꢀ
tures reflected in an rmsd of 2.9 A (271 of 353 resi-
dues), HmdII and Hmd reveal remarkable structural
changes (Fig. 3A,B) of potential functional relevance.
The structure of the loop region after the strand
13 : 17 directly involved in the FeGP cofactor binding
is largely conserved, but adjacent regions that pack
against it as the two-helical segment (residues 22–40)
and the helical region (residues 68–115) of Hmd are
truncated in HmdII. Apparently, the FeGP cofactor
binding site is more loosely anchored with the rest of
the protein in HmdII than in Hmd.
The C terminus of each subunit is prolonged in
HmdII by ~ 50 residues (residues 302–353) compared
with Hmd. This dominantly helical segment increases
not only the interface between the C-terminal segments
of both subunits, but also the interface between the C-
and N-units. Helix 336 : 351 contacts helix 131 : 140,
and residues 315–319 the loop after strand 167 : 171,
whose prolongation of nine residues in HmdII affords
this contact. Consequently, the transition between
open and closed states, as postulated for Hmd [2,3,25]
might be impaired in HmdII. In addition, the rigid-
body movement of the N-unit relative to the C-unit is
much smaller between apo- and holo-HmdII than
between apo- and holo-Hmd. In principle, this finding
supports the limited open-to-close capabilities but
might also be triggered by crystal lattice effects. Inter-
estingly, the conformation of apo- and holo-HmdII
structures rather resembles that of the closed apo-
Hmd.
We analyzed the charge distribution of the surface
of HmdII and Hmd to determine differences between
them for identifying potential binding sites for other
proteins present in HmdII, but not in Hmd. However,
no attractive hot spot on the HmdII surface could be
identified. A positively charged patch was detectable
around the binding site of the FeGP cofactor in
Fig. 2. Infrared spectra of holo-HmdII and holo-Hmd. Black line,
holo-HmdII; red line, holo-Hmd; blue line, holo-HmdII in the
presence of methenyl-H4MPT+; magenta line, holo-HmdII in the
presence of methylene-H4MPT.
3414
FEBS Journal 282 (2015) 3412–3423 ª 2015 FEBS

T. Fujishiro et al.
Potential function of an [Fe]-hydrogenase paralog
A
B
C
D
Fig. 3. X-Ray crystal structure of (A) holo-HmdII and (B) holo-Hmd, both from Methanocaldococcus jannaschii [1,2]. The FeGP cofactor is
represented as a gray stick model. (C) FeGP cofactor binding site in holo-HmdII. (D) FeGP cofactor binding site in holo-Hmd. Residues in
contact to the FeGP cofactor are represented as yellow and wheat stick models, respectively. Polar interactions are shown as black dashed
lines.
apo-Hmd and even more extended in apo-HmdII
(data not shown), which was partly neutralized after
FeGP binding.
Methylene-H4MPT binding to apo-HmdII
Crystals of apo-HmdII were soaked with methenyl-
H4MPT+ and methylene-H4MPT and their structures
were determined by the molecular replacement method
(Table 1). In contrast to methenyl-H4MPT+, methy-
lene-H4MPT could be bound to apo-HmdII with a
high occupancy. Extra electron density interpretable as
methylene-H4MPT was clearly visible in the active site
cleft next to Cys127, a ligand of the FeGP cofactor
iron in holo-HmdII (Fig. 4). Its conformation resem-
bles that of methylene-H4MPT bound to holo-Hmd
C176A (PDB code 3H65) (Fig. 4D) [2]. In both Hmd
and HmdII, the electron density of methenyl-H4MPT
after the benzyl group is weak and beyond the ribose
group is largely blurred [2]. In HmdII, the pterin ring
of methylene-H4MPT of apo-HmdII forms polar
interactions with Ala212, Ser216 and Ser279, the latter
residue originates from the counter monomer
(Fig. 4C). In addition, methylene-H4MPT is flanked
by several nonpolar and well-distributed residues
including Leu24, Tyr25, Val159, Met214, Ile237,
Ala239, Met243 and Met283, suggesting that its bind-
ing to HmdII has a biological function. Moreover,
Binding of the FeGP cofactor
FeGP cofactor is bound in holo-HmdII at the same
position and in a related conformation as in holo-Hmd
(Fig. 3C,D). Several cofactor–protein interactions, i.e.
Cys127 to Fe as well as Asp86 and Glu61 (HmdII
nomenclature) to guanosine, conserved between HmdII
and Hmd argue for HmdII as a natural FeGP cofactor
binding protein. Nevertheless, notable differences exist.
The pyridinol ring is sandwiched between Phe99 and
Phe102 in HmdII and by the more polar Trp148 and
Lys151 in Hmd. Trp148 forms, in addition, a hydro-
gen bond to the phosphate group of the FeGP cofac-
tor, which is compensated by a hydrogen bond with
Tyr25 in HmdII. Tyr25 is replaced by His14 in Hmd,
which possibly serves as an acceptor for the proton
released after heterolytic cleavage of H2; substitution
of His14 with alanine in Hmd leads to a dramatic
decrease in enzymatic activity [2,4].
FEBS Journal 282 (2015) 3412–3423 ª 2015 FEBS
3415

Potential function of an [Fe]-hydrogenase paralog
T. Fujishiro et al.
Table 1. Data collection and refinement statistics.
HmdII with
SeMet-HmdII
SeMet-apo-HmdII SeMet-apo-HmdII SeMet-apo-HmdII SeMet-apo-HmdII
bound FeGP with bound
(MAD peak)
(MAD inflection)
(MAD remote)
(For refinement)
apo-HmdII
cofactor
methylene-H4MPT
Data collection
Temperature (K)
100
100
100
100
100
100
100
ꢀ
Wavelength (A)
0.9791
P212121
50–2.3
(2.6–2.3)
0.9798
P212121
50–1.9
(2.0–1.9)
0.9720
P212121
50–2.3
(2.6–2.3)
0.9791
P212121
50–1.9
(2.0–1.9)
1.000
1.000
1.000
Space group
P212121
50–1.65
(1.7–1.65)
P41212
50–2.2
(2.3–2.2)
P212121
50–1.9
(2.0–1.9)
ꢀ
Resolution (A)
Cell dimensions
ꢀ
a, b, c (A)
52.6, 78.0,
153.7
52.6, 78.0,
153.7
52.6, 78.0,
153.7
52.6, 78.0,
153.7
52.5, 77.8,
153.5
52.6, 78.0,
153.7
53.0, 78.0,
153.9
a, b, c (°)
Completeness
(%)a
90, 90, 90
100.0 (100.0)
90, 90, 90
99.4 (95.9)
90, 90, 90
99.7 (99.9)
90, 90, 90
99.3 (95.7)
90, 90, 90
90, 90, 90
90, 90, 90
99.0 (98.7)
99.7 (100.0) 99.5 (96.6)
Rsym (%)a,b
I/rIa
Redundancya
Refinement
statistics
11.4 (34.7)
16.2 (7.7)
5.1 (5.2)
11.9 (90.5)
7.9 (1.3)
12.3 (63.1)
8.6 (2.7)
7.8 (58.6)
11.8 (2.2)
5.3 (3.9)
6.4 (56.5)
12.1 (2.9)
4.9 (5.1)
8.4 (31.5)
17.1 (3.7)
10.1 (3.2)
5.5 (83.6)
18.8 (2.9)
5.5 (5.7)
4.4 (3.3)
4.4 (4.6)
ꢀ
Resolution (A)
43.6–1.9
(1.94–1.9)
2
43.5–1.65
(1.71–1.65)
2
42.8–2.2
(2.3–2.2)
1
43.8–1.9
(1.94–1.90)
2
No. monomers/
asymmetric unit
No. ligands/
1
4
1
3
asymmetric unit
No. waters/
225
320
108
176
asymmetric unit
work/Rfree (%)c,d
R
18.5/21.5
(30.5/37.3)
0.008
18.6/22.1
(28.0/32.4)
0.011
18.6/22.3
(28.9/32.4)
0.014
18.4/21.3
(24.6/29.0)
0.013
ꢀ
rmsd bond length (A)
rmsd bond angle (°)
1.193
1.362
1.913
1.371
a
b
c
d
Values in parentheses are for the highest resolution shell.
Rsym = ΣΣ|I – <I>|/ΣI, where I is the intensity of each reflection.
Rwork = Σ||Fo|–k|Fc||/Σ|Fo|, where Fo and Fc are the observed and calculated structure factor amplitudes, respectively.
Rfree was calculated as the Rwork for 5% of the reflections that were not included in the refinement. MAD, multiple anomalous dispersion.
methylene-H4MPT has many more contacts with the
polypeptide in the closed apo-HmdII than in the open
holo-Hmd C176A structures (Fig. 4D).
holo-Hmd C176A–methylene-H4MPT complex and
the apo-Hmd structure (Fig. 4E,F) [2]. The same cat-
alytic capabilities for HmdII and Hmd appear to be
conceivable.
Despite various attempts, crystals of the
holo-HmdII–methenyl-H4MPT+ or the holo-HmdII–
methylene-H4MPT complexes were not obtained. To
analyze the arrangement of the FeGP cofactor relative
to methylene-H4MPT, apo-HmdII with bound methy-
lene-H4MPT and holo-HmdII were superimposed. In
the resulting model, the C14a atom of methylene-
Activity assays of holo-HmdII
The recombinantly produced, pure and successfully
reconstituted holo-HmdII now allowed serious kinetic
studies to find out its catalytic activity. To probe
Hmd-like reactions, we assayed the oxidation of
methylene-H4MPT and methylene-tetrahydrofolate
(tetrahydrofolate is an analog of H4MPT), and the
reduction of methenyl-H4MPT+ and methenyl-tetrahy-
drofolate with H2 both at 40 °C under the conditions
described in the Materials and Methods. Despite
repeated attempts using different preparations, we
ꢀ
H4MPT is placed 2.4 A from the iron site of the
FeGP cofactor (Fig. 4E). Only small rearrangements
in the holo-HmdII–methylene-H4MPT complex are
required to provide enough space for H2 binding and
an optimal active site geometry. The obtained
FeGP cofactor–methylene-H4MPT geometry is similar
to that proposed for Hmd modeled from the open
3416
FEBS Journal 282 (2015) 3412–3423 ª 2015 FEBS

T. Fujishiro et al.
Potential function of an [Fe]-hydrogenase paralog
A
B
C
D
E
F
Fig. 4. Crystal structure of apo-HmdII and holo-Hmd C176A with bound methylene-H4MPT. (A) Structure of apo-HmdII with bound
methylene-H4MPT (purple stick model). (B) Structure of holo-Hmd C176A from Methanocaldococcus jannaschii with bound methylene-
H4MPT (violet stick model). (C) Methylene-H4MPT binding site in apo-HmdII. Amino acid residues with blue labels belong to the partner
monomer of the homodimer. The black dashed lines indicate hydrogen bonds between methylene-H4MPT and the amino acid residues. The
2Fo–Fc electron density (blue mesh) is contoured at 1r. (D) Methylene-H4MPT binding site in holo-Hmd C176A [2]. Hmd and HmdII are
shown in the same orientation. (E) Modeled ternary HmdII–FeGP–methylene-H4MPT–FeGP complex. For modeling apo-HmdII with bound
methylene-H4MPT (magenta and pink cartoons for the two HmdII subunits) and holo-HmdII (green and yellow cartoons) were
superimposed. The distances between methylene-H4MPT and the FeGP cofactor (black dashed lines) are indicated. (F) Modeled ternary
Hmd–FeGP–methylene-H4MPT–FeGP complex. Apo-Hmd and the holo-Hmd C176A–methylene-H4MPT complex were superimposed [2].
could not detect any catalytic activity of holo-HmdII.
Following the proposal of Reeve et al. [22], we also
tried to measure the reduction of methylene-H4MPT
with H2 to methyl-H4MPT. Again, no catalytic activity
was found. With the aim of recovering the Hmd activ-
ity, an exchange of Tyr25 of HmdII with a histidine
was performed because the histidine at this position in
Hmd was considered as a proton acceptor after H2
cleavage [2]. However, the Tyr25His HmdII variant
was catalytically inactive.
Previous UV/vis and CD spectroscopic measure-
ments indicated that Hmd with bound methenyl-
H4MPT+ can reversibly accept hydride from H2 and
produce methylene-H4MPT under single turnover
conditions at pH 6.0 [25]. To probe the single turnover
Hmd reaction of holo-HmdII at the same conditions,
FEBS Journal 282 (2015) 3412–3423 ª 2015 FEBS
3417

Potential function of an [Fe]-hydrogenase paralog
T. Fujishiro et al.
we used equimolar concentrations (0.1 mM) of enzyme
and substrate, but could not detect any activity (data
not shown).
or homolog of Hmd but a paralog. We can, of course,
not exclude that HmdII uses completely different
compounds as substrates but the unambiguous binding
of the FeGP cofactor and methylene-H4MPT does not
favor this possibility.
Discussion
A possible H2-sensor function of HmdII could be
ruled out due to the performed single turnover hydride
transfer assays, which exhibited no activity. A single
turnover event would be still achievable in the absence
of a specific proton uptake and an open-to-closed
transition.
Over the years, several functional studies on HmdII
and HmdIII have been performed. The hmdII and
hmdIII genes are upregulated at low and high concen-
trations of H2, respectively [22] when expressed in
Methanothermobacter thermautotrophicus cells grown
under laboratory conditions. Likewise, expression of
hmdII in Methanococcus maripaludis is upregulated
under syntrophic growth conditions [26]. The amounts
of HmdII and HmdIII in Methanothermobacter
marburgensis grown under normal conditions are 1%
and < 0.1%, respectively, of the soluble proteins in the
cell extract based on western blotting [27]. Binding of
the FeGP cofactor was predicted by homology models
[28] and inhibition of the holoenzyme reconstitution
with apo-HmdII [23,24]. HmdII and HmdIII of
M. jannaschii binds to proyl-tRNA synthetase and
tRNA inferred from the copurification of HmdII from
cell extracts with prolyl-tRNA synthetase [29,30] and
in vitro binding studies [31].
Because some methanogenic archaea lack the genes
for Hmd paralogs, but still produce active Hmd
[20,32] and because hmdII gene-deletion experiments
with Methanococcus maripaludis had no impact on the
biosynthesis of the FeGP cofactor [33], a scaffold
function for the cofactor biosynthesis can also be ruled
out. The affinity between HmdII and methylene-
H4MPT is also not compatible with this idea. The
latter argument also makes a storage function of
HmdII for the FeGP cofactor unlikely.
The only plausible hypothesis remains a function of
HmdII as a regulatory protein in protein synthesis based
on the previously found binding of holo-HmdII to
aminoacyl-tRNA synthetase and tRNA [29–31].
According to infrared spectroscopic data, HmdII can
distinguish between methenyl-H4MPT+ and methylene-
H4MPT, and binding of the latter induces marked
conformational changes around the Fe complex. There-
fore, we suggest that HmdII senses the intracellular
methylene-H4MPT concentration, implying that the
postulated conformational changes upon methylene-
H4MPT binding alters the affinity of HmdII to potential
cellular binding proteins or nucleic acids. Preliminary
comparative surface profile analyses of HmdII and
Hmd of M. jannaschii and protein-docking calculations
of HmdII and prolyl-tRNA synthetase from M. jan-
naschii (PDB code: 1NJ8) [34], inconclusive to date,
cannot replace future experimental studies. The impact
of methylene-H4MPT on interactions between
holo-HmdII and aminoacyl-tRNA synthetases and
tRNA (or nucleic acids) has to be biochemically
explored. From the perspective of the hydrogenotrophic
methanogen, a strict control of protein synthesis under
H2-limiting conditions by the amount of reducing equiv-
alents is beneficial as proposed previously [22]. Methy-
lene-H4MPT is a metabolite produced in the midst of
the methanogenic pathway, and a regulation of protein
synthesis by its intracellular concentration would satisfy
this need.
Based on this information, various functions for the
Hmd-related proteins have been considered: (a)
catalytic Hmd isoenzyme [22], (b) H2 sensor [27], (c)
scaffold protein involved in FeGP cofactor biosynthe-
sis [24,28], (d) storage protein for the FeGP cofactor
[24,28], and (e) regulation of protein biosynthesis due
to the interactions with aminoacyl-tRNA synthetase
and tRNA [29–31].
The potential functions for HmdII and HmdIII are
predicted based on primary structures, modeled
tertiary structures, heterologously produced apoen-
zymes or cell extracts. The provided pure apo- and
holo-HmdII allow a detailed structural and kinetic
analysis and open new perspectives for functional
studies. Holo-HmdII did not exhibit any catalytic
activities with H4MPT and tetrahydrofolate derivatives
and/or H2 as substrates, despite similar FeGP cofactor
and methylene-H4MPT-binding properties and active
site geometries in HmdII and Hmd. A more careful
structural inspections reveals two features that
rationalizes these findings: (a) the increased interface
between N- and C-units, mainly due to a C-terminally
prolonged polypeptide chain, impairs the mobility
between them which is required for the catalytic cycle
of Hmd; and (b) the exchange of a histidine in Hmd
to a tyrosine in HmdII. The histidine is most likely
involved in proton transfer from buried H2 to bulk
solvent in Hmd [2]. Thus, HmdII is not an isoenzyme
We have previously successfully predicted the cat-
alytic function of several proteins based on the crystal
structure and on subsequent in vitro activity studies
3418
FEBS Journal 282 (2015) 3412–3423 ª 2015 FEBS

T. Fujishiro et al.
Potential function of an [Fe]-hydrogenase paralog
for verification (structure-to-function approach)
[35,36]. In this study, structural together with infrared
spectroscopic and kinetic data clearly excluded most of
the postulated functions of HmdII and we present a
hypothesis about its cellular role of methylene-H4MPT
sensor. This hypothesis is consistent with the available
data and experimentally verifiable in future on the
basis of the established protocol for preparing pure
holo-HmdII. This study teaches us that a functional
annotation exclusively based on sequence or even con-
vincing structural data still requires a direct experi-
mental validation. Despite similar overall structures,
potentially identical prosthetic groups and substrates
as well as active site geometries, HmdII is not an Hmd
isoenzyme.
Amino acid sequences were aligned using CLUSTAL W2 [38].
The figure of the alignment was generated using ESPRIPT3 [39].
Heterologous production and purification of
HmdII
Escherichia coli BL21 (DE3) cells were transformed with
the expression plasmid containing M. jannaschii apo-HmdII
with a C-terminal His6 tag and grown at 37 °C in Luria–
Bertani medium supplemented with 100 lgꢁmLꢀ1 ampicillin
to an optical density 600 (OD600) = 1.0. Protein expres-
sion was induced by the addition of 1 mM isopropyl b-D-
thiogalactopyranoside. After incubation for 4–6 h the cells
were harvested by centrifugation and stored at ꢀ80 °C
before
further
use.
The
apo-
HmdII Y25H mutant was produced in the same way as
apo-HmdII using 50 lgꢁmLꢀ1 kanamycin instead of
100 lgꢁmLꢀ1 ampicillin. All procedures for the purification
of apo-HmdII and the apo-HmdII Y25H mutant were
performed on ice or at 4 °C. The frozen cells containing
apo-HmdII and the apo-HmdII Y25H mutant were resus-
pended in 50 mM potassium phosphate buffer (pH 7.0)
containing 0.5 M KCl, 20 mM imidazole and 1 mM dithio-
threitol (buffer A), and disrupted by sonication. The super-
natant was collected by centrifugation and loaded onto a
Ni2+-charged HiTrap chelating column (GE Healthcare,
Uppsala, Sweden) equilibrated with buffer A. The column
was washed with buffer A and eluted with a linear gradient
of imidazole from 20 to 500 mM in 50 mM potassium
phosphate buffer (pH 7.0) containing 0.5 M KCl and 1 mM
dithiothreitol. The protein fractions were concentrated with
a Centrifugal Filter Unit (Merck Millipore, Darmstadt,
Germany) and loaded onto a HiPrep sephacryl S-200 col-
umn (GE Healthcare) equilibrated with 50 mM potassium
phosphate buffer (pH 7.0) containing 0.3 M KCl and 1 mM
dithiothreitol. The purity of the final sample was estimated
by SDS/PAGE. Protein concentration was determined by
the Bradford method. For the SeMet-labeled apo-HmdII
preparation, E. coli B834 (DE3) (Novagen) cells trans-
formed with the expression vector carrying the hmdII gene
were cultured in M9 medium supplemented with 2.5 mM
MgSO4, 2% (w/v) D-glucose, 0.01% (w/v) thiamine,
0.025 mM FeCl3, 50 lgꢁmLꢀ1 L-selenomethionine, and
100 lgꢁmLꢀ1 ampicillin at 37 °C to an OD600 = 0.5
before induction of expression with 1 mM isopropyl b-D-
thiogalactopyranoside as described above. SeMet-apo-
HmdII was purified as described for apo-HmdII.
Materials and methods
Materials
All chemicals were used without further purification. Poly
(ethylene glycol) 300, nickel sulfate, 3- (2-pyridyl)-5,6-di(2-
furyl)-1,2,4-triazine-50500-disulfonic acid disodium salt (Fer-
ene), and potassium hydroxide were purchased from
Sigma-Aldrich (Taufkirchen, Germany). L-Selemomethio-
nine (SeMet) was purchased from Acros Organics (Geel,
Belgium). Magnesium chloride, 32% ammonia solution,
poly(ethylene glycol) 200 and thiamine were purchased
from Merck (Darmstadt, Germany). The following
reagents were obtained from Roth (Karlsruhe, Germany):
methanol, kanamycin sulfate, ampicillin sodium salt, Mops,
imidazole, ammonium chloride, L-methionine, sodium
chloride, sodium dihydrogen phosphate, disodium hydro-
gen phosphate, hydrochloric acid, potassium chloride,
potassium dihydrogen phosphate, dipotassium hydrogen
phosphate, D-glucose and FeCl3ꢁ6H2O. Isopropyl b-D-thio-
galactopyranoside was purchased from Fermentas (St.
Leon-Rot, Germany).
The FeGP cofactor was isolated from holo-Hmd from
Methanothermobacter marburgensis [16]. Apo-Hmd was
purified from recombinant Eschericha coli BL21(DE3) har-
boring an expression vector, as described previously [8,16].
The expression plasmid with the M. jannaschii hmdII gene
in pET22b (Novagen, Darmstadt, Germany) at its NdeI–
BamHI sites was a kind gift of Y.M. Hou [29]. The M. jan-
naschii hmdII Y25H mutant gene with codon optimization
for Escherichia coli host cells was purchased from Gen-
Script USA Inc. (Piscataway, NJ, USA) as hmdII Y25H-
pET24b (Novagen). Methenyl-H4MPT+ and methylene-
H4MPT were prepared, as reported previously [37]. (6R,S)
-5,10-methenyl-5,6,7,8-tetrahydrofolic acid chloride and
(6R,S) -5,10-methylene-5,6,7,8-tetrahydrofolic acid calcium
salt were purchased from Schircks Laboratories (Jona,
Switzerland).
Reconstitution of the holoenzymes
To reconstitute the holoenzymes, apo-HmdII and apo-Hm-
dII Y25H mutant were mixed with the FeGP cofactor
under a N2/H2 (95%/5%) atmosphere at 8 °C in an anae-
robic chamber (Coy Laboratories, Grass Lake, MI, USA)
under red or yellow light [16]. The buffers of purified
FEBS Journal 282 (2015) 3412–3423 ª 2015 FEBS
3419

Potential function of an [Fe]-hydrogenase paralog
T. Fujishiro et al.
apo-HmdII and apo-HmdII Y25H mutant were exchanged
with anoxic 10 mM Mops/KOH (pH 7.0) buffer containing
1 mM dithiothreitol using a PD-10 column (GE Health-
care). Apo-HmdII and the apo-HmdII Y25H mutant were
then incubated with at least 1.5 eq. of the FeGP cofactor
in 0.1 M Tris/HCl buffer pH 8.0 for 1 h at 8 °C under
anoxic conditions. The mixture of apo-HmdII and apo-
HmdII Y25H mutant with FeGP cofactor was loaded onto
a PD-10 column equilibrated with 10 mM Mops/KOH buf-
fer containing 1 mM dithiothreitol (pH 7.0) to separate
HmdII and the apo-HmdII Y25H mutant both reconsti-
tuted with FeGP cofactor from unbound FeGP cofactor.
with the FeGP cofactor were obtained in a month.
SeMet-apo-HmdII crystals were incubated in 0.1 M
phosphate-citrate buffer (pH 4.2) containing 34% (v/v)
poly(ethylene glycol) 300 and 1.4 mM methenyl-H4MPT+
or 1.4 mM methylene-H4MPT at 8 °C for 1 day.
X-ray data collection and refinement
Crystals of apo-HmdII, SeMet–apo-HmdII, HmdII
reconstituted with the FeGP cofactor and SeMet–apo-HmdII
incubated with various ligands were frozen under a cryo-
stream of N2 at 100 K without adding a cryoprotectant.
Diffraction data were collected at 100 K on beamline X10SA
equipped with a PILATUS 6M detector at the Swiss Light
Source (Villigen, Switzerland). Data were processed using XDS
[40]. To determine the crystal structure of apo-HmdII, multi-
ple anomalous dispersion data were measured at the selenium
edge of the SeMet–apo-HmdII crystals. Selenium atom sites
were detected with SHELX C/D [41]. Phases were determined
using the program SHARP and improved by the solvent flatten-
ing procedure of SOLOMON implemented in SHARP [42]. Auto-
mated model building was performed using ARP/WARP [43].
The SeMet–apo-HmdII model was manually completed and
refined using COOT [44], REFMAC5 [45] and PHENIX [46]. Crystal
structures of other crystal forms of HmdII were solved by
molecular replacement with MOLREP [47] or PHASER [48] using
the SeMet–apo-HmdII structure as a search model. The TLS
option was used for all structures in the final stage of refine-
ment [49]. The final models were validated using MOLPROBITY
[50]. The resulting structures were established at a resolution
Infrared spectroscopy
Attenuated total reflection-infrared spectra were recorded
using a Bruker VERTEX 70V FTIR spectrometer equipped
with an MCT detector. Attenuated total reflection optical
configuration (DuraSampleIRTM
,
Sensir Technologies,
Watford, UK) was employed with a silicon prism. The mea-
surements were handled under red light to avoid
photoinduced decomposition of the FeGP cofactor in the
samples. Infrared spectra were continuously recorded with
averaging of 512 scans at 4 cmꢀ1 resolution under an argon
stream; the sample was gradually concentrated on the optical
element of the attenuated total reflection unit due to an evap-
oration of the water solute. The spectrum was chosen when
the sample concentration was sufficiently high to resolve the
CO band from the FeGP cofactor; the water band was sub-
tracted and the spectrum was baseline corrected.
ꢀ
ꢀ
ꢀ
of 1.7 A for apo-HmdII, 1.9 A for SeMet–apo-HmdII, 2.2 A
ꢀ
for holo-HmdII, 1.9 A for methylene-H4MPT-bound SeMet–
apo-HmdII, and 1.9 A for methenyl-H4MPT+-bound
ꢀ
Crystallization
SeMet-apo-HmdII. Data collection and refinement statistics
are summarized in Table 1. Atomic coordinates and structure
factors are available in the Protein Data Bank (www.pdb.org)
under the accession numbers; 4YT8 for SeMet-labeled apo-
HmdII; 4YT2 for apo-HmdII; 4YT4 for holo-HmdII; 4YT5
for apo-HmdII with bound methylene-H4MPT. Crystal struc-
tures were superimposed using DALILITE [51] or SUPERPOSE [52]
in CCP4 [53]. Charge distributions on the surfaces of the struc-
All proteins were crystallized at 8 °C using the sitting drop
vapor diffusion method. A 1 lL aliquot of apo-HmdII
(12 mgꢁmLꢀ1
) in 50 mM potassium phosphate buffer
(pH 7.0) with 0.3 M KCl was mixed with 1 lL of a reser-
voir solution composed of 40% (v/v) poly(ethylene
glycol) 300 and 100 mM phosphate-citrate buffer (pH 4.2);
crystals appeared within 1 week. Crystals of SeMet–apo-
HmdII were obtained within 1 week when adding a 1 lL
€
tures were calculated by PYMOL (v. 1.3r1, Schrodinger, LLC).
aliquot of SeMet–apo-HmdII (20 mgꢁmLꢀ1
) in 10 mM
€
Figures 3 and 4 were generated by PYMOL (v. 1.3r1, Schrodin-
Mops/KOH buffer (pH 7.0)/1 mM dithiothreitol to 1 lL of
a reservoir solution composed of 35% (v/v) poly(ethylene
glycol) 300 and 100 mM phosphate-citrate buffer (pH 4.2).
For crystallization of HmdII reconstituted with the FeGP
cofactor, all procedures were carried out in a Coy chamber
with a N2/H2 (95%/5%) atmosphere under red or yellow
light. An 1 lL aliquot of HmdII reconstituted with the
FeGP cofactor (9 mgꢁmLꢀ1) in anoxic 10 mM Mops/KOH
buffer (pH 7.0)/1 mM dithiothreitol was mixed with 1 lL
ger, LLC).
Enzyme activity assay
To assay the conversion of methenyl-H4MPT+ to methy-
lene-H4MPT, holo-HmdII or the holo-HmdII Y25H
mutant was added to 20lM methenyl-H4MPT+ in 120 mM
potassium phosphate buffer (pH 7.5) in a 1-cm-path quartz
cuvette under 100% H2 at 40 °C. To assay Hmd-catalyzed
conversion of methylene-H4MPT to methenyl-H4MPT+,
holo-HmdII or the holo- HmdII Y25H mutant was added
of
a
reservoir solution composed of 50% (v/v)
poly(ethylene glycol) 200, 100 mM Na/K-phosphate buffer
(pH 6.2), and 0.2 M NaCl. Crystals of HmdII reconstituted
3420
FEBS Journal 282 (2015) 3412–3423 ª 2015 FEBS

T. Fujishiro et al.
Potential function of an [Fe]-hydrogenase paralog
to 20lM methylene-H4MPT in 120 mM potassium phos-
phate buffer (pH 6.0) inside a 0.3-cm-path quartz cuvette
under 100% N2. Methenyl-tetrahydrofolate and methylene-
tetrahydrofolate were tested as the Hmd-catalyzed reactions
described above. The change in absorbance at 336 nm was
monitored. For the activity under single-turn-over condi-
tions, UV/vis spectra of 0.1 mM HmdII in 50 mM Mes/
NaOH (pH 6.0) in the presence of 0.1 mM methenyl-
H4MPT+ under N2, and H2 were measured as described
previously [25].
5 Lubitz W, Ogata H, Rudiger O & Reijerse E (2014)
Hydrogenases. Chem Rev 114, 4081–4148.
6 Shima S & Ermler U (2011) Structure and function of
[Fe] -hydrogenase and its iron-guanylylpyridinol (FeGP)
cofactor. Eur J Inorg Chem 2011, 963–972.
7 Thauer RK, Kaster AK, Goenrich M, Schick M,
Hiromoto T & Shima S (2010) Hydrogenases from
methanogenic archaea, nickel, a novel cofactor, and H2
storage. Annu Rev Biochem 79, 507–536.
8 Buurman G, Shima S & Thauer RK (2000) The metal-
free hydrogenase from methanogenic archaea: evidence
for a bound cofactor. FEBS Lett 485, 200–204.
9 Korbas M, Vogt S, Meyer-Klaucke W, Bill E, Lyon EJ,
Thauer RK & Shima S (2006) The iron-sulfur cluster-free
hydrogenase (Hmd) is a metalloenzyme with a novel iron
binding motif. J Biol Chem 281, 30804–30813.
10 Lyon EJ, Shima S, Boecher R, Thauer RK, Grevels
FW, Bill E, Roseboom W & Albracht SPJ (2004)
Carbon monoxide as an intrinsic ligand to iron in the
active site of the iron-sulfur-cluster-free hydrogenase
H2-forming methylenetetrahydromethanopterin
dehydrogenase as revealed by infrared spectroscopy.
J Am Chem Soc 126, 14239–14248.
11 Shima S, Lyon EJ, Sordel-Klippert MS, Kauss M,
Kahnt J, Thauer RK, Steinbach K, Xie XL, Verdier L
& Griesinger C (2004) The cofactor of the iron-sulfur
cluster free hydrogenase Hmd: structure of the light-
inactivation product. Angew Chem Int Ed 43,
Acknowledgements
We thank Prof. Dr Rolf Thauer for discussions and
helpful suggestions, Prof. Dr Hartmut Michel for con-
tinuous support, and the staff of the PXII beamline at
the Swiss Light Source (Villigen) and Ulrike Demmer
for help during data collection and X-ray structure
analysis. We thank Prof. Ya-Ming Hou for kindly pro-
viding the expression plasmid for M. jannaschii
HmdII. This work was supported by a grant of the
Max Planck Society to R. Thauer and a grant of the
PRESTO program from the Japan Science and
Technology Agency to SS.
Author contributions
SS directed the research. TF and SS designed the
research. TF prepared, crystallized and characterized
proteins. KA performed the infrared spectroscopy. TF
and UE performed the X-ray crystallographic analysis.
TF, UE and SS interpreted the data and wrote the
manuscript.
2547–2551.
12 Shima S, Lyon EJ, Thauer RK, Mienert B & Bill E
€
(2005) Mossbauer studies of the iron-sulfur cluster-free
hydrogenase: the electronic state of the mononuclear Fe
active site. J Am Chem Soc 127, 10430–10435.
13 Shima S, Schick M, Kahnt J, Ataka K, Steinbach K &
Linne U (2012) Evidence for acyl-iron ligation in the
active site of [Fe] -hydrogenase provided by mass
spectrometry and infrared spectroscopy. Dalton Trans
41, 767–771.
References
1 Hiromoto T, Ataka K, Pilak O, Vogt S, Stagni MS,
Meyer-Klaucke W, Warkentin E, Thauer RK, Shima S &
Ermler U (2009) The crystal structure of C176A mutated
[Fe] -hydrogenase suggests an acyl- iron ligation in the
active site iron complex. FEBS Lett 583, 585–590.
2 Hiromoto T, Warkentin E, Moll J, Ermler U & Shima
S (2009) The crystal structure of an [Fe] -hydrogenase-
substrate complex reveals the framework for H2
activation. Angew Chem Int Ed 48, 6457–6460.
3 Pilak O, Mamat B, Vogt S, Hagemeier CH, Thauer
RK, Shima S, Vonrhein C, Warkentin E & Ermler U
(2006) The crystal structure of the apoenzyme of the
iron-sulphur cluster-free hydrogenase. J Mol Biol 358,
798–809.
14 Shima S & Ataka K (2011) Isocyanides inhibit [Fe] -
hydrogenase with very high affinity. FEBS Lett 585,
353–356.
15 Tamura H, Salomone-Stagni S, Fujishiro T, Warkentin
E, Meyer-Klaucke W & Ermler U (2013) Crystal
structures of [Fe] -hydrogenase in complex with
inhibitory isocyanides: implications for H2-activation
site. Angew Chem Int Ed 52, 9656–9659.
16 Shima S, Schick M & Tamura H (2011) Preparation of
[Fe] -hydrogenase from methanogenic archaea. Method
Enzymol 494, 119–137.
17 Smith DR, DoucetteStamm LA, Deloughery C, Lee
HM, Dubois J, Aldredge T, Bashirzadeh R, Blakely D,
Cook R, Gilbert K et al. (1997) Complete genome
sequence of Methanobacterium thermoautotrophicum
DH: functional analysis and comparative genomics.
J Bacteriol 179, 7135–7155.
4 Shima S, Pilak O, Vogt S, Schick M, Stagni MS, Meyer-
Klaucke W, Warkentin E, Thauer RK & Ermler U (2008)
The crystal structure of [Fe] -hydrogenase reveals the
geometry of the active site. Science 321, 572–575.
FEBS Journal 282 (2015) 3412–3423 ª 2015 FEBS
3421

Potential function of an [Fe]-hydrogenase paralog
T. Fujishiro et al.
18 Bult CJ, White O, Olsen GJ, Zhou LX, Fleischmann
RD, Sutton GG, Blake JA, FitzGerald LM, Clayton
RA, Gocayne JD et al. (1996) Complete genome
sequence of the methanogenic archaeon, Methanococcus
jannaschii. Science 273, 1058–1073.
19 Samuel BS, Hansen EE, Manchester JK, Coutinho
PM, Henrissat B, Fulton R, Latreille P, Kim K,
Wilson RK & Gordon JI (2007) Genomic and
metabolic adaptations of Methanobrevibacter smithii
to the human gut. Proc Natl Acad Sci USA 104,
10643–10648.
with a putative metabolic protein in archaea.
Biochemistry 42, 7487–7496.
30 Lipman RSA, Sowers KR & Hou YM (2000) Synthesis
of cysteinyl-tRNA(Cys) by a genome that lacks the
normal cysteine-tRNA synthetase. Biochemistry 39,
7792–7798.
31 Oza JP, Sowers KR & Perona JJ (2012) Linking energy
production and protein synthesis in hydrogenotrophic
methanogens. Biochemistry 51, 2378–2389.
32 Schick M, Xie XL, Ataka K, Kahnt J, Linne U &
Shima S (2012) Biosynthesis of the iron-
€
20 Schworer B & Thauer RK (1991) Activities of
guanylylpyridinol cofactor of [Fe] -Hydrogenase in
methanogenic archaea as elucidated by stable-isotope
labeling. J Am Chem Soc 134, 3271–3280.
formylmethanofuran dehydrogenase,
methylenetetrahydromethanopterin dehydrogenase,
methylenetetrahydromethanopterin reductase, and
heterodisulfide reductase in methanogenic bacteria.
Arch Microbiol 155, 459–465.
33 Lie TJ, Costa KC, Pak D, Sakesan V & Leigh JA
(2013) Phenotypic evidence that the function of the [Fe]
-hydrogenase Hmd in Methanococcus maripaludis
requires seven hcg (hmd co-occurring genes) but not
hmdII. FEMS Microbiol Lett 343, 156–160.
34 Kamtekar S, Kennedy WD, Wang JM, Stathopoulos
C, Soll D & Steitz TA (2003) The structural basis of
cysteine aminoacylation of tRNA(Pro) by prolyl-
tRNA synthetases. Proc Natl Acad Sci USA 100,
1673–1678.
21 Goker M, Daligault H, Mwirichia R, Lapidus A, Lucas
S, Deshpande S, Pagani I, Tapia R, Cheng JF,
Goodwin L et al. (2011) Complete genome sequence of
the thermophilic sulfur-reducer Desulfurobacterium
thermolithotrophum type strain (BSAT) from a deep-sea
hydrothermal vent. Stand Genomic Sci 5, 407–415.
€
22 Reeve JN, Nolling J, Morgan RM & Smith DR (1997)
Methanogenesis: genes, genomes, and who’s on first?
J Bacteriol 179, 5975–5986.
23 Pilak O (2004) Synthese von Hydrogenase Hmd und
von zwei vermuteten Isoenzymen in
35 Fujishiro T, Kahnt J, Ermler U & Shima S (2015)
Protein-pyridinol thioester precursor for biosynthesis of
the organometallic acyl-iron ligand in [Fe] -hydrogenase
cofactor. Nat Commun 6, 6895.
Methanothermobacter marburgensis bei Wachstum unter
Fe-, NH3-oder Ni-Limitierung, Diplomarbeit Phillipps-
Universita€t Marburg.
36 Fujishiro T, Tamura H, Schick M, Kahnt J, Xie XL,
Ermler U & Shima S (2013) Identification of the HcgB
enzyme in [Fe] -hydrogenase-cofactor biosynthesis.
Angew Chem Int Ed 52, 12555–12558.
24 Shima S & Thauer RK (2007) A third type of
hydrogenase catalyzing H2 activation. Chem Rec 7,
37–46.
25 Shima S, Vogt S, Gobels A & Bill E (2010) Iron-
chromophore circular dichroism of [Fe] -hydrogenase:
the conformational change required for H2 activation.
Angew Chem Int Ed 49, 9917–9921.
26 Walker CB, Redding-Johanson AM, Baidoo EE,
Rajeev L, He ZL, Hendrickson EL, Joachimiak MP,
Stolyar S, Arkin AP, Leigh JA et al. (2012) Functional
responses of methanogenic archaea to syntrophic
growth. ISME J 6, 2045–2055.
37 Shima S & Thauer RK (2001)
Tetrahydromethanopterin-specific enzymes from
Methanopyrus kandleri. Methods Enzymol 331,
317–353.
38 Larkin MA, Blackshields G, Brown NP, Chenna R,
McGettigan PA, McWilliam H, Valentin F, Wallace
IM, Wilm A, Lopez R et al. (2007) Clustal W and
clustal X version 2.0. Bioinformatics 23, 2947–2948.
39 Robert X & Gouet P (2014) Deciphering key features
in protein structures with the new ENDscript server.
Nucleic Acids Res 42, W320–W324.
27 Afting C, Kremmer E, Brucker C, Hochheimer A &
Thauer RK (2000) Regulation of the synthesis of H2-
forming methylenetetrahydromethanopterin
dehydrogenase (Hmd) and of HmdII and HmdIII in
Methanothermobacter marburgensis. Arch Microbiol 174,
225–232.
28 Goldman AD, Leigh JA & Samudrala R (2009)
Comprehensive computational analysis of Hmd
enzymes and paralogs in methanogenic archaea. BMC
Evol Biol 9, 199.
40 Kabsch W (2010) Xds. Acta Crystallogr D 66, 125–132.
41 Sheldrick GM (2008) A short history of SHELX. Acta
Crystallogr A 64, 112–122.
42 dela Fortelle E & Bricogne G (1997) Maximum-
likelihood heavy-atom parameter refinement for
multiple isomorphous replacement and multiwavelength
anomalous diffraction methods. Methods Enzymol 276,
472–494.
43 Langer G, Cohen SX, Lamzin VS & Perrakis A (2008)
Automated macromolecular model building for X-ray
crystallography using ARP/wARP version 7. Nat
Protoc 3, 1171–1179.
29 Lipman RSA, Chen J, Evilia C, Vitseva O & Hou YM
(2003) Association of an aminoacyl-tRNA synthetase
3422
FEBS Journal 282 (2015) 3412–3423 ª 2015 FEBS

T. Fujishiro et al.
Potential function of an [Fe]-hydrogenase paralog
44 Emsley P & Cowtan K (2004) Coot: model-building tools
for molecular graphics. Acta Crystallogr D 60, 2126–2132.
45 Murshudov GN, Vagin AA & Dodson EJ (1997)
Refinement of macromolecular structures by the
maximum-likelihood method. Acta Crystallogr D 53,
240–255.
46 Afonine PV, Grosse-Kunstleve RW, Chen VB, Headd
JJ, Moriarty NW, Richardson JS, Richardson DC,
Urzhumtsev A, Zwart PH & Adams PD (2010)
phenix.model_vs_data: a high-level tool for the
calculation of crystallographic model and data
statistics. J Appl Crystallogr 43, 669–676.
in macromolecular refinement. Acta Crystallogr D 57,
122–133.
50 Chen VB, Arendall WB, Headd JJ, Keedy DA,
Immormino RM, Kapral GJ, Murray LW,
Richardson JS & Richardson DC (2010)
MolProbity: all-atom structure validation for
macromolecular crystallography. Acta Crystallogr D
66, 12–21.
51 Holm L, Kaariainen S, Rosenstrom P & Schenkel A
(2008) Searching protein structure databases with
DaliLite v. 3. Bioinformatics 24, 2780–2781.
52 Krissinel E & Henrick K (2004) Secondary-structure
matching (SSM), a new tool for fast protein structure
alignment in three dimensions. Acta Crystallogr D 60,
2256–2268.
47 Vagin A & Teplyakov A (1997) MOLREP: an
automated program for molecular replacement. J Appl
Crystallogr 30, 1022–1025.
48 Mccoy AJ, Grosse-Kunstleve RW, Adams PD, Winn
MD, Storoni LC & Read RJ (2007) Phaser
53 Winn MD, Ballard CC, Cowtan KD, Dodson EJ,
Emsley P, Evans PR, Keegan RM, Krissinel EB, Leslie
AGW, McCoy A et al. (2011) Overview of the CCP4
suite and current developments. Acta Crystallogr D 67,
235–242.
crystallographic software. J Appl Crystallogr 40, 658–674.
49 Winn MD, Isupov MN & Murshudov GN (2001) Use
of TLS parameters to model anisotropic displacements
FEBS Journal 282 (2015) 3412–3423 ª 2015 FEBS
3423