Y. Ni et al. / Journal of Molecular Catalysis B: Enzymatic 103 (2014) 52–55
53
Table 1
Comparison between H2 and CO as reductants for the P. furiosus-catalyzed reduction
of cinnamic acid.a
Product
Yield [%]
pH 5.5
pH 6.5
pH 8.0
H2 as reductant
0
40.3 1.8
29.1 1.3
8.2 0.8
2.7 0.2
0
10.4 0.7
67.7 5.9
Scheme 1. Suggested P. furiosus-mediated reduction of carboxylic acids with H2
or CO as reductant. AOR: aldehyde oxidoreductase; ADH: alcohol dehydrogenase;
Hase: hydrogenase; CODH: CO dehydrogenase; Fdred/Fdox: reduced/oxidized ferre-
doxin.
3.7 0.1
CO as reductant
2.1.1. Preparation of P. furiosus cells
P. furiosus (DSM 3638) was grown in a 100 L fermenter at 90 ◦C,
under anaerobic conditions with potato starch as carbon source as
previously described [6]. Cells were harvested by crossflow filtra-
tion and centrifugation and stored at −80 ◦C for further use.
0
0
37.2 1.6
16.1 1.0
8.3 1.3
0.7 0.1
1.7 0.1
78.3 3.5
0.6
0
2.2. General reaction conditions
Reaction conditions were: 100 mM sodium phosphate buffer (pH 6.5), T = 40 ◦C,
c(substrate) = 10 mM, p(H2 or CO, respectively) = 5 bar, c(P. furiosus) = 0.15 g mL−1
For our investigations we utilized the reaction conditions
previously identified to be suitable for the P. furiosus-catalyzed
hydrogenation of carboxylic acids [6]. In short: reaction mixtures
of 2 ml in 16 ml autoclaves containing 0.3 g P. furiosus frozen cells,
10 mM carboxylic acid substrate and 100 mM sodium phosphate
buffer (pH 6.5) were flushed with N2 and pre-purged with H2
(p = 5 bar). A photograph of the experimental setup is shown in
the supporting information. The reactions were incubated at 40 ◦C
with magnetic agitation at 100 rpm for 24 h. The reaction mixture
was acidified to pH 2.0 with 5 N HCl, extracted twice with distilled
ethyl acetate or diethyl ether containing 1-octanol or n-decane as
an internal standard, and analyzed by GC. In the case of some aro-
matic acids, the samples were centrifuged for 15 min at 13,000 rpm
after adding equal volumes of acetonitrile, and the supernatant was
analyzed by HPLC.
.
hydrogenoformans Ni-dependent CO-dehydrogenase [15]. Details
can be found in the supporting information.
3. Results and discussion
To approach the question whether P. furiosus might utilize CO as
a sacrificial electron donor we used cinammic acid as a model sub-
strate and compared the product distribution using H2 and CO as
reductants under otherwise identical reaction conditions (Table 1).
We found that P. furiosus not only tolerated CO but also utilized
it as sacrificial electron donor. As shown in Table 1 conversion and
product distributions were essentially identical using either H2 or
version was observed in the absence of either CO or H2 (but under
5 bars of N2). Currently, we can only speculate about the electron
transport chain enabling P. furiosus to utilize CO as reductant. Possi-
bly a CO dehydrogenase (CODH) as suggested for C. thermoaceticum
[4d,11] might also be present in P. furiosus. However, no indications
for homologues of the known Mo- or Ni-dependent CODHs (PDB:
1ZXI E and 1SU6 A) were found in the genome of P. furiosus [13]. It
should be noted that the Ni-CODH has sequence homology to parts
of the coding sequence of the Hybrid cluster protein (HCP), which
exists in P. furiosus [16]. The only established enzymatic activ-
ity of HCP is hydroxylamine reduction to ammonia. Further work
in our laboratory will concentrate on isolating and characterizing
the putative CODH of P. furiosus. In any case, these findings point
towards the usage of much cheaper synthesis gas instead of highly
purified H2 as reductant. Also we like to interpret the identical
results using H2 and CO by assuming that the thermodynamically
challenging reduction of carboxylic acids is overall rate-limiting,
which is also in line with the finding that the intermediate aldehyde
was not observed.
2.3. Analytical procedures
The reaction progress as well as the optical purity of the
products was determined using GC analysis or HPLC. GC was per-
formed using a CP-Sil 5 CB column (50 m × 0.53 mm × 1.0 m)
or a CP-Wax 52 CB (50 m × 0.53 mm × 2.0 m) with N2 as car-
rier gas and flame ionization detector. HPLC analysis was carried
out with a Waters Xterra column (C18, 5 m, 4.6 × 150 mm) with
CH3CN:H2O:HCOOH (20:80:1, v/v/v) as eluent. The samples were
isocratically eluted at a flow rate of 1 ml/min.
2.4. BLAST searches
The genome of P. furiosus [13] was searched for the follow-
ing sequences to identify putative (1) enoate reductases or (2)
CO-dehydrogenases. To scan for enoate reductases three rep-
resentatives of the two subfamilies of the Old Yellow Enzyme
(OYE) family and two representatives of the Clostridial enoate
reductases were used: Saccharomyces pastorianus OYE1 as
a
and Thermus scotoductus SA-01 Chromate reductase as rep-
resentatives of the thermophilic-like OYE, and Clostridium
tyrobutyricum and Clostrium ljungdahlii as representatives of
the Clostridial enoate reductases [3a,14] Furthermore, in order
to scan for putative CO-dehydrogenases two different CO-
dehydrogenase representatives that have been structurally and
biochemically characterized have been used: Oligotropha carboxi-
dovorans Mo-dependent CO-dehydrogenase and Carboxydothermus
Another interesting finding from these experiments is the
comparably poor chemoselectivity as significant amounts of the
saturated carboxylic acid and alcohol were observed. In previ-
ous experiments we found that non-conjugated C C-double bonds
were not converted by P. furiosus [6].
Enoate reductases (ERs, E.C. 1.3.1.31) are well-known to catalyze
the reduction of conjugated C C-double bonds to the correspond-
ing saturated carbonyl compounds (especially aldehydes) [3],