Biotechnology Advances xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Biotechnology Advances
journal homepage: www.elsevier.com/locate/biotechadv

Research review paper

Improved strategies for electrochemical 1,4-NAD(P)H2 regeneration: A new

era of bioreactors for industrial biocatalysis
Clifford S. Morrisona, William B. Armigere, David R. Doddse, Jonathan S. Dordicka,b,c,d,
Mattheos A.G. Koffasa,b,⁎
a
Department of Chemical and Biological Engineering, Rensselaer Polytechnic Institute, 110 8th Street, Troy, NY 12180, United States
b
Department of Biological Sciences, Rensselaer Polytechnic Institute, 110 8th Street, Troy, NY 12180, United States
c
Department of Materials Science and Engineering, Rensselaer Polytechnic Institute, 110 8th Street, Troy, NY 12180, United States
d
Department of Biomedical Engineering, Rensselaer Polytechnic Institute, 110 8th Street, Troy, NY 12180, United States
e
BioChemInsights, Inc., Malvern, PA 19355, United States

A R T I C L E I N F O A B S T R A C T

Keywords: Industrial enzymatic reactions requiring 1,4-NAD(P)H2 to perform redox transformations often require con-
Cofactors voluted coupled enzyme regeneration systems to regenerate 1,4-NAD(P)H2 from NAD(P) and recycle the co-
NADH factor for as many turnovers as possible. Renewed interest in recycling the cofactor via electrochemical means is
NADPH motivated by the low cost of performing electrochemical reactions, easy monitoring of the reaction progress, and
Biocatalysis
straightforward product recovery. However, electrochemical cofactor regeneration methods invariably produce
Electrochemical bioreactors
adventitious reduced cofactor side products which result in unproductive loss of input NAD(P). We review
Cofactor regeneration
Industrial biotechnology various literature strategies for mitigating adventitious product formation by electrochemical cofactor re-
Renalase generation systems, and offer insight as to how a successful electrochemical bioreactor system could be con-
structed to engineer efficient 1,4-NAD(P)H2-dependent enzyme reactions of interest to the industrial biocatalysis
community.

1. Introduction
Nicotinamide adenine dinucleotide, commonly written as NAD+/
NADH, and its phosphorylated variants nicotinamide adenine dinu-
cleotide phosphate, NADP+/NADPH, are biochemical redox cofactors
responsible for the transfer of electrons and protons in biological redox
reactions. Formally, this allows the transfer of a molecule of hydrogen
between metabolic intermediates. The oxidized form of the cofactor
accepts electrons, balanced by protons, and in doing so can assume a
number of isomeric structures as well as a dimer. These are shown in
Fig. 1.
The naming convention for the species in Fig. 1, however, is not
consistent in the literature, and the common NAD(P)+terminology
causes confusion when considering electrochemical reactions. As ex-
plained further in Section 2.1, the oxidized form will be termed NAD
(P), and the biologically active reduced form will be termed 1,4-NAD(P)
H2. The C-2 and C-6 isomers of the reduced form will be termed 1,2-
NAD(P)H2 and 1,6-NAD(P)H2, respectively, and the more general term
NAD(P)H2 will be used when specifying a particular isomer of the re-
duced cofactor is not necessary.
Nearly 20% of known oxidoreductases require cofactors to supply
stoichiometric quantities of reducing equivalents. For nearly 700
known classes of redox enzymes (Ullah et al., 2015), 1,4-NAD(P)H2 is
the required cofactor. Since the total intracellular pool of both the
oxidized and reduced forms of the cofactor is fairly low, the cofactor
undergoes reduction and oxidation turnovers on the order of 103 to 105
in typical reactions (Angelastro et al., 2017; Ströhle et al., 2016). The
current technology for regeneration of 1,4-NAD(P)H2, exemplified in
Fig. 2, includes 1) sacrificial substrates, e.g., a non-valuable alcohol is
present in the reaction mixture and is simultaneously oxidized to an
aldehyde (or acid). This was used by Whitesides and Wong (1982),
Wong and Whitesides (1983) to utilize all of the reducing power

Abbreviations: NAD, oxidized nicotinamide adenine dinucleotide; NADP, oxidized nicotinamide adenine dinucleotide phosphate; NADH2, reduced nicotinamide adenine dinucleotide
(any isomer); NADPH2, reduced nicotinamide adenine dinucleotide phosphate (any isomer); [NAD]2, nicotinamide adenine dinucleotide dimer; [NADP]2, nicotinamide adenine dinu-
cleotide phosphate dimer; DPN, diphosphopyridine nucleotide; DPNH, reduced diphosphopyridine nucleotide; TPN, triphosphopyridine nucleotide; TPNH, reduced triphosphopyridine
nucleotide; 1,n-NADH2, reduced 1,n-nicotinamide adenine dinucleotide; FAD, flavin adenine dinucleotide; HPLC, high performance liquid chromatography; MSE, saturated mercury
sulfate electrode; mt, metric tonne; SCE, saturated calomel electrode; Poly-His, polyhistidine; HRP, horseradish peroxidase
⁎
Corresponding author at: Department of Chemical and Biological Engineering, Rensselaer Polytechnic Institute, 110 8th Street, Troy, NY 12180, United States.
E-mail address: koffam@rpi.edu (M.A.G. Koffas).

http://dx.doi.org/10.1016/j.biotechadv.2017.10.003
Received 23 August 2017; Received in revised form 2 October 2017; Accepted 6 October 2017
0734-9750/ © 2017 Elsevier Inc. All rights reserved.

Please cite this article as: Morrison, C.S., Biotechnology Advances (2017), http://dx.doi.org/10.1016/j.biotechadv.2017.10.003
C.S. Morrison et al. Biotechnology Advances xxx (xxxx) xxx–xxx

Fig. 1. The numbered chemical structures for NAD(P) co-

O NH2
factor material. The isomers of NAD(P)H and a re-
presentative example of an NAD dimer are also shown
H2 N N H
HO O O O- here.
N O P P O N
+ NAD (zwitterion)
N O O O
N

HO OH HO OH

O NH2

H
H2N N -
HO O O O Na+ H
O O
1,4-NADH2
N N P P N
O O O (sodium salt)
N

HO OH HO OH

O NH2

H2 N N H
HO O O O-
O P O + NADP (disodium salt)
N N P N
O O O pKa1 = 3.9; pKa2 = 6.1
N

O OH HO OH

O P O- Na+
O- Na+

O
O NH2 O NH2
NH2
H 3
3 H
H 2
2 4 4 R N 4 4' N R
H
N N 6 5
5
R 1 R 1
6 H2N
H H
O

1,2-NADH2 1,6-NADH2 4,4'-dimer; [NAD]2

H2 N N
HO O O O- Na+
N O P P O
R= N O O O
N
HO OH HO OH

available by employing three enzymes and oxidizing methanol all the
way to CO2; 2) whole cell methods, using resting cells or active fer-
mentations, in which a nutrient such as glucose is metabolized to
provide reducing equivalents (Zaks and Dodds, 1997); and 3) combined
chemoenzymatic methods (Eikeren, 1989). While all of these methods
can be used, they require handling whole cells, multiple enzymes, and/
or other molecules, such as the sacrificial substrate. Direct regeneration
of the 1,4-NAD(P)H2 would be simpler and less expensive if electrons
and protons could be provided directly to the oxidized form of NAD(P).
Electrochemical methods, under specific reaction conditions, are able to
achieve exactly that.
Formally, the reduced form of the cofactor, NAD(P)H2, carries one
molecule of H2, which is provided to the reaction being catalyzed by the
enzyme requiring reducing power. Reduced cofactor is thus the che-
mical equivalent of hydrogen, and competition with other sources of
hydrogen must be considered when planning a chemical process that
could use NAD(P)H2.
Historically, the regeneration of 1,4-NAD(P)H2 by electrochemical
means has been well reported, but has never attained wide-spread
practical application due to the barrier caused by the formation of
biologically inactive isomers of NAD(P)H2 and dimers of NAD(P). The
recognition of these well-known chemistries has held back the appli-
cation of electrochemical 1,4-NAD(P)H2 regeneration to industrial
processes. As discussed in Section 4.2, strategies for dimer mitigation
have been recognized for decades, but no similar approaches for isomer
mitigation were known. However, naturally occurring enzymes (re-
nalases) were recently discovered, as discussed in Section 4.2.1, in both
eukaryotic and prokaryotic cells whose primary function is the recovery
of the biologically inactive forms of 1,2- and 1,6-NAD(P)H2 by re-oxi-
dizing them back to NAD(P). This recent discovery of nature's system
for conserving the biologically active NAD(P) molecule within the cell
can be incorporated into electrochemical systems for removing this
barrier.
An electrochemical bioreactor must achieve a number of cofactor

2
C.S. Morrison et al. Biotechnology Advances xxx (xxxx) xxx–xxx

Fig. 2. Examples of current industrial strategies used for

SACRIFICIAL
SUBSTRATE the regeneration of 1,4-NAD(P)H2. 1,4-NAD(P)H2 provides
reducing power to a variety of biocatalytic processes.
reduced cofactor
oxidized 1,4-NAD(P)H2 substrate
by-product
1st 2nd
redox redox
enzyme enzyme

sacrificial
NAD(P) product
substrate
oxidized cofactor

WHOLE CELL
nutrient substrate
e.g. glucose

cell

oxidized metabolite(s) product

e.g. gluconolactone, CO2

CHEMO-ENZYMATIC

chemical reagent reduced cofactor

oxidized by-
1,4-NAD(P)H2 substrate
e.g. mBH4 product
e.g. CH3CHO 1st 2nd
redox redox
enzyme enzyme
spent reagent sacrificial (reduced)
e.g. substrate NAD(P) product
mB(OH)4 e.g. CH3CH2OH oxidized cofactor

turnovers to maximize process efficiency and economic value. To
maximize the number of cofactor turnovers, strategies for mitigating
the formation of the adventitious but undesired reduced cofactor side
products, 1,2- and 1,6-NAD(P)H2 as well as the dimer [NAD(P)]2, must
be considered. These mitigation strategies can take the form of re-
covering NAD(P) or 1,4-NAD(P)H2 from the undesired reduction pro-
ducts or preventing their formation in the first place.
This review surveys the literature of NAD(P)H2 reduction and oxi-
dation over the last 70 years and documents the important electro-
chemical products resulting from NAD(P) reduction along with the
observations that have confounded and confused the field for many
decades. This literature review, along with recent developments in the
last 18 months, provides insights into the issues that have held back the
use of electrochemical regeneration of 1,4-NAD(P)H2.
2. Redox enzymes and pyridine nucleotide cofactors
2.1. Properties of NAD(P)H2
As of this review, thousands of articles in the scientific literature
refer to a charged species NAD(P)+ as part of a redox pair with neutral
NAD(P)H2. Indicating a positive charge on the nicotinamide moiety of
NAD(P) without the context of the rest of the molecule leads one to
conclude that the entire molecule has a formal charge of + 1 due to the
nitrogen at the 1-position of the nicotinamide ring. This is commonly
represented in the literature by either of the following representations:
NAD(P)+ + H+ + 2e− → NAD(P)H (a)
NAD(P)+ + 2H+ + 2e− → NAD(P)H + H+ (b)
While commonly encountered, these descriptions lead to confusion
when writing balanced electrochemical reactions showing the formal
transfer of a hydrogen molecule, as two electrons and two protons, to
the oxidized species NAD(P). These molecules are best regarded as a
neutral zwitterion, in the same way that amino acids are regarded, in
order to account for a holistic perspective on the overall charge balance
within their native biological environments. To avoid further confusion,
this review suggests to the chemical community the nomenclature of
Reaction 2 below, despite the lengthy historical use of the nomen-
clature of Reactions 1a and b.
In its oxidized state, non-phosphorylated NAD is an electron ac-
ceptor with a formal positive charge at the quaternary nitrogen atom of
the nicotinamide ring and two ionizable hydroxyl groups on the di-
phosphate linkage. The pKa of these hydroxyl groups are approximately
2.4 and 6.6, respectively. The more acidic hydroxyl provides a formal
negative charge on the molecule to satisfy the positive charge on the
nitrogen atom in the nicotinamide ring as a zwitterion. Depending on
the exact pH, the second hydroxyl group will be deprotonated to
varying degrees giving NAD a net negative charge, contrary to the
commonly used NAD + notation. This net negative charge is balanced
by a suitable cation in aqueous solution.
A similar situation exists with NADP, albeit with two additional
ionizable hydroxyl groups on the extra phosphate moiety. Thus NADP
should be properly represented as a di- or even tri-anionic species, not
as NADP +, which suggests it is a mono-cation.
The aforementioned discussion on net charge is important when
considering electrochemical redox reactions in solution. Critically, two
electrons and two protons are required for the reduction of NAD(P) to
NAD(P)H2, and this is clear experimentally using mass spectrometry
Reaction (1) with NAD and NADH2 having a molecular weight difference of 2 amu.
Therefore, it is more accurate to report the overall redox reaction with

3
C.S. Morrison et al.
e-
O NH2 O NH2
H H
e-
N+ Step 1 N Step 3
R R
NAD NAD NAD
H2N N
HO O O O- Na+
N O P P O
R= N O O O
N
HO OH HO
the following representation:
NAD(P) + 2H+ + 2e− → NAD(P)H2
In some reports, NAD(P)H2 is referred to in its enzymatically active
form as 1,4-NAD(P)H2 in order to specifically reflect that the hydrogen
on the dihydropyridyl moiety is at the C-4 position. Other reported
names for the non-phosphorylated 1,4-NADH2 include β-NADH, DPNH,
and reduced coenzyme I. Other reported names for the phosphorylated
form, NADPH2, include β-NADPH, TPNH, and reduced coenzyme II.
1,4-NAD(P)H2 is significant in a variety of industrial processes. For
example, the enzymatic production of butanol (Kang et al., 2007) and
ethanol (Shin et al., 2004); the hydroxylation of alkanes and aromatic
compounds (Urlacher and Eiben, 2006); the epoxidation of alkenes
(McClay et al., 2000); heteroatom oxygenations and Baeyer–Villiger
reactions (Zambianchi et al., 2002); as well as the hydrolysis of tri-
glycerides (Hollmann et al., 2006) are a few of the relevant industrial
biocatalytic processes that utilize 1,4-NAD(P)H2 as a redox cofactor.
1,4-NAD(P)H2 must be used in stoichiometric quantities for these
biocatalytic reactions, but the high cost and lability (Mädje et al., 2012)
of these cofactors presents an imposing economic obstacle for industrial
processes relying on 1,4-NAD(P)H2 without the ability to reduce NAD
back to 1,4-NAD(P)H2. Furthermore, it is disadvantageous to add high
concentrations of cofactor to an in vitro enzyme reaction since excess
amounts can lead to enzyme inhibition (Williams, 1952). Therefore,
there has been motivation to achieve continuous in situ regeneration of
1,4-NAD(P)H2 from a limited pool of NAD(P).
Several approaches to 1,4-NAD(P)H2 regeneration have been in-
vestigated, including electrochemical, photochemical, and chemical
reduction methods (Lee and Whitesides, 1985) in addition to whole cell
and coupled enzyme/substrate methods (Zhao and van der Donk,
2003). Due to their low cost and simplicity, electrochemical techniques
have received increased attention, and therefore this review focuses on
1,4-NAD(P)H2 regeneration in electrochemical systems.
2.2. The issue of nonspecific NAD(P) reduction
As opposed to enzymatic regeneration methods, which are highly
selective for the catalytically active form 1,4-NAD(P)H2, non-biological
regeneration strategies are driven by nonspecific electrochemical,
photochemical or chemical reduction (Lee and Whitesides, 1985). Such
nonspecific approaches result in a mixture of reduced nicotinamide ring
isomers, including the desired 1,4-species and the non-biologically ac-
tive 1,2- and 1,6- reduction isomers (Chenault and Whitesides, 1987).
The formation of the inactive isomers leads to inefficiency in co-
factor regeneration systems and ultimately limits the recycling turnover
due to dead-end products. In addition, a dimer of NAD(P), noted in this
Biotechnology Advances xxx (xxxx) xxx–xxx
Fig. 3. The mechanism for the electrochemical production
O NH2 O NH2
of 1,4-NADH2 and [NAD]2 involving one-electron transfer
H reactions with a radical intermediate. The radical species is
- H H+
H the branching point of the reaction.
N Step 2a N
R R
1,4-NADH2
Step 2
O
NH2
R N N R 4,4'-dimer; [NAD]2
H2N
O
OH
paper as [NAD(P)]2, is also reported to form during the electrochemical
reduction of NAD(P). This reaction also locks some of the cofactor
Reaction (2)
material into a biologically inactive form with each electrochemically-
driven reduction cycle. This problem, referred to as “the nonspecific
reduction problem,” was pointed out in the 1980s by Whitesides et al.
as a major reason for the inability of electrochemical 1,4-NAD(P)H2
regeneration methods to be competitive with biological methods
(Chenault and Whitesides, 1987).
1,2-NAD(P)H2 has a half-life of ~30 min (Beaupre et al., 2015).
However, this was probably not known to earlier researchers in the
field who reported unusual situations where their experiments did not
give expected results. For example, Jaegfeldt (1981) observed dis-
crepancies in HPLC analyses of NADH2 isomer preparations. In another
study, 1,6-NAD(P)H2 was positively identified by Godtfredsen and Ot-
tesen as an inhibitor of lactate dehydrogenase (Godtfredsen and
Ottesen, 1978), but the authors were probably unaware that 1,2-NAD
(P)H2 also behaves as an inhibitor of 1,4-NAD(P)H2-dependent enzymes
as the isomer's instability likely prevented its detection or isolation
(Beaupre et al., 2015). Similarly, Biellman et al. reported difficulties in
isolating and studying certain inhibitors of lactate dehydrogenase
(Biellmann et al., 1979).
The mechanism for the electrochemical formation of the 4,4′-dimer
[NAD]2 is shown in Fig. 3. A one-electron reduction occurs at a re-
ducible position of the pyridine ring to form an NAD% radical, which
then undergoes either a second electron transfer to form an anion
(which is then quenched by a proton) or a radical dimerization to form
an [NAD]2 dimer. If the second electron transfer is slower than di-
merization, then dimer will be the predominant product (Ali et al.,
2012b). Migration of the radical leads to the 1,2-NADH2 or the 1,6-
NADH2 isomers.
Given the three reducible positions on the pyridine ring, as well as
additional stereoisomers of the dimer due to the asymmetric carbons at
the point of dimerization, one might expect a combinatorial number of
possible dimers, e.g. 2,2′-, 2,4′-, 2,6′-, 4,4′-, 4,6′-, and 6,6′-dimers, plus
the stereoisomers for each dimer, thus totaling 21 possible dimers. The
dimers are typically named by the combination of monomeric species
that constitute the dimer, e.g. n,m′-[NAD(P)]2.
Experimental evidence suggests that only seven of the 21 possible
dimers are sufficiently stable to persist in solution. These are the three
possible 4,4′-dimers and the four possible 4,6′-dimers. Indeed, three
stereoisomers of 4,4′-dimers comprise 90% of electrochemically pre-
pared dimer solutions, while the other 10% consists of three stereo-
isomers of 4,6′-dimers. No 6,6′-dimers nor any dimers involving the C-2
position were found (Jaegfeldt, 1981). Generally similar results were
obtained by Carelli et al., who observed only 4,4′-dimers after
4
C.S. Morrison et al. Biotechnology Advances xxx (xxxx) xxx–xxx

externally driven current flow
oxidized
cofactor
Anode Cathode
6 H2O
4 e- 4 e-
O2
+
4 H3O+ 4 H3O+
Anode Chamber Cathode Chamber
proton-
permeable
membrane
Fig. 4. The overall electrochemistry and general arrangement of the electrochemical cell.
Shown here is the direct electrochemical reduction of the oxidized form of the cofactor to
the reduced form of the cofactor.
electrochemical reduction (Carelli et al., 1981), and Kovář et al. who
found that 4,4′-dimers constituted the majority of electrochemical
dimer preparations, with only small amounts of 4,6′-dimers present in
the mixtures (Kovář et al., 1985). Thus, dimers of the 4,4′ type appear
to form the most stable structures out of the 21 possible dimers, as
predicted by Burnett and Underwood (Burnett and Underwood, 1968).
3. Electrochemical bioreactors
3.1. Function of electrochemical bioreactors
Electrochemical bioreactors operate on the principle of either pro-
ducing or using electrical current. Examples of electrochemical bior-
eactors include bio-batteries, microbial fuel cells, microbial electro-
synthesis cells, enzymatic electrolysis cells, and electrolyzers (Rabaey
et al., 2010). Electrolyzers are used most frequently in cofactor re-
generation and are thus the focus of electrochemical bioreactors de-
scribed in this review.
As shown in Fig. 4, the electrolyzer consists of an anode chamber,
where four electrons are obtained from O2 reduction to water, and a
cathode chamber, where these electrons are donated to the oxidized
cofactor to generate the reduced cofactor. The electrode chambers are
typically separated by an ion-selective membrane to allow for the
transfer of protons in the form of hydronium ions to cross from the
anode chamber to the cathode chamber while the electrons are trans-
ferred from the anode to the cathode chamber over the wire connecting
the chambers.
Reduced cofactor can be regenerated electrochemically with as
much as 98% of 1,4-NAD(P)H2 regenerated from NAD(P) in solution.
This highly efficient regeneration of 1,4-NAD(P)H2 was reportedly ob-
tained with an electrochemical bioreactor utilizing a glassy carbon
electrode as the cathode (Ali et al., 2012b; Ali et al., 2011). While this is
promising, most systems reported to date are too small to be
commercially feasible. Moreover, the low surface area of these elec-
trodes results in a slow NAD(P) conversion rate, and the poorly-defined
fluid flow and mass transport characteristics of laboratory-scale elec-
trochemical systems present scale-up challenges (Walsh and Reade,
1994). In addition, high overpotentials or use of electron mediators are
often required for the direct reduction of NAD(P) to 1,4-NAD(P)H2 (Lau
reduced et al., 2005). Electrochemical bioreactor technology development for
cofactor
research use or for industrial production should, therefore, focus on
systems capable of achieving a practical number of cofactor turnovers
in a reasonable amount of time.
The turnover efficiency for cofactors regenerated by any strategy
4 H2O
must be high enough to render the adventitious formation of useless
reduced cofactor products, i.e. the loss of usable cofactor per each
turnover, negligible. Regenerating the reduced form of cofactor in situ
allows for large-scale reactions to proceed economically. Though to be
truly economical, the regeneration method should be capable of re-
cycling the cofactor 102–106 times, depending on the cost of the co-
factor and the value of the products made by the process. The turnover
efficiency required to have at least 50% remaining cofactor activity
after 102 turnovers is 99.3% and for 106 turnovers the efficiency needed
is 99.99993%. It then follows that of the various regeneration strate-
gies, the only strategy historically capable of delivering such turnover
efficiencies are those mediated by coupled redox enzyme reactions
using sacrificial substrates (Chenault et al., 1988).
Coupled redox enzyme reactions, whether through the use of sa-
crificial substrates or resting whole cell methods, utilize the simulta-
neous reduction of NAD(P) in one direction and the oxidation of 1,4-
NAD(P)H2 in the other direction. This can occur by using an enzyme
with a reversible reaction in both reaction directions but acting on
different substrates, or by using two enzymes with different activities
for different substrates that, when combined, result in the cycling of
NAD(P)/1,4-NAD(P)H2 (see Fig. 2). Indeed, the production of many
modern industrially-relevant biocatalysis products requiring the re-
generation of 1,4-NAD(P)H2 are only made possible by the use of
coupled enzyme systems. Examples of such products include the statin
drug Lipitor® (atorvastatin calcium) (see Fig. 5) (Schroer and Lütz,
2009; Schroer et al., 2007; Ma et al., 2010), the antilipemic agent
ZETIA® (ezetimibe) (Homann and Previte, 1996), the HIV protease in-
hibitor REYATAZ® (atazanavir) (Patel, 2006), the anti-malarial drug
precursor amorphadiene (Ma et al., 2011), and the pharmaceutical
building block L-tert-leucine (Leuchtenberger et al., 2005) among other
drugs (Zaks and Dodds, 1997).
Electrochemical cofactor regeneration falls broadly into two stra-
tegies: 1) direct reduction of NAD(P) in the cathode chamber, and 2)
indirect reduction of NAD(P) using a mobile electron transport med-
iator in the cathode chamber that is recycled between an enzyme re-
action chamber where NAD(P) reduction takes place with the oxidized
electron mediator returning to the cathode chamber. In general, elec-
tron transport mediators can be scaled to maximize the availability of
electron-donating species throughout the system. This could help to
minimize electrolyzer size relative to reactor volume.
An electrochemical bioreactor system capable of simultaneously
performing an 1,4-NAD(P)H2-dependent reaction of interest while ef-
ficiently regenerating 1,4-NAD(P)H2 in situ offers an advantage over
currently-practiced technology by eliminating the requirements of

substrate Fig. 5. An example of a coupled enzyme system for the

regeneration of 1,4-NADPH2. 1,4-NAD(P)H2 provides re-
reduced cofactor O O
ducing power to a variety of industrially-relevant products
oxidized by-product 1,4-NADPH2 Cl
O such as pharmaceutical intermediates, adapted from (Ma
gluconate, C6H10O6
et al., 2010).
GDH KRED

OH O
sacrificial (reduced)
NADP Atorvastatin
substrate Cl
glucose, C6H12O6 oxidized cofactor O
reduced product

5
C.S. Morrison et al.
Table 1
Five guiding principles for process development for a successful electrochemical bior-
eactor.
Principle Function
Reduction Reduction of oxidized form of the cofactor
Reaction Useful reduction of substrate to desired product with concomitant
oxidation of the cofactor
Recovery Recovery of unused forms of reduced cofactor (i.e. non-
biologically useful isomers, dimers, etc.)
Regeneration Reduction of recovered oxidized cofactor
Recycle Practical engineering to allow continuous closed system cofactor
recycle
using a sacrificial substrate, which requires separation of its associated
products from the desired product.
3.2. Criteria for successful electrochemical bioreactors
The co-production of 1,2- and 1,6-NAD(P)H2 and [NAD(P)]2 re-
mains a challenge for electrochemical bioreactor systems. Therefore,
successful electrochemical bioreactor systems need to follow five
guiding principles for process development, as shown in Table 1.
Recovery processes, which have yet to be given proper treatment in
the literature in the context of electrochemical bioreactor systems, are
vital to the success of a process by which 1,4-NAD(P)H2 is regenerated
electrochemically. Therefore, it is imperative for the success of an
electrochemical bioreactor process to follow the above-mentioned
guidelines and to include “recovery” as a fifth necessary principle. It is
essential that NAD(P) is recovered from the adventitious reduced co-
factor products, namely 1,2- and 1,6-NAD(P)H2 and [NAD(P)]2, so that
recycling the cofactor results in a minimal loss of cofactor per turnover
event, and the overall process can be performed as economically as
possible.
3.3. Electrochemical 1,4-NAD(P)H2 regeneration
Various embodiments of electrochemical 1,4-NAD(P)H2-re-
generating systems have been reported in the literature to varying de-
grees of success. Elving et al. reported in 1982 that the direct reduction
of NAD(P) on an unmodified metallic electrode generally results in the
formation of [NAD(P)]2 via NAD(P)% radicals which can be reduced to
1,4-NAD(P)H2 and its isomers at high overpotentials (Elving et al.,
1982). Since operating at higher overpotentials is not economically
efficient, much work in the literature has focused since then on creating
modified electrodes to promote 1,4-NAD(P)H2 formation at lower
overpotentials.
Siu et al. reported the conversion of α-ketoglutarate to glutamate by
glutamate dehydrogenases following the direct electrochemical reduc-
tion of NAD in the presence of vanadia-silica xerogels using unmodified
platinum electrodes. The authors reported that 3300 turnovers of NAD
were achieved using their system, but they do not discuss formation of
adventitious side products that may have limited the total number of
turnovers (Siu et al., 2007).
Hildebrand et al. reported the conversion of acetophenone to (R)-
phenylethanol by alcohol dehydrogenase from Lactobacillus brevis fol-
lowing the indirect reduction of NAD on a carbon felt electrode using
Cp*Rh(bpy) as an electron mediator. The authors focused on enzyme
productivity rather than cofactor turnover numbers, so their system
with 35 turnovers of NAD does not immediately lend itself to industrial
application (Hildebrand and Lütz, 2007).
Hollman et al. reported the conversion of several α-substituted
phenol derivatives to their ortho-hydroxylated counterparts by 2-hy-
droxybiphenyl-3-monooxygenase following the indirect electro-
chemical reduction of NAD on a glassy carbon electrode using [Cp*Rh
(bpy)Cl]Cl as an electron mediator. No information was provided on the
6
Biotechnology Advances xxx (xxxx) xxx–xxx
number of NAD turnovers achieved (Hollmann et al., 2001).
Some of the more successful embodiments of 1,4-NAD(P)H2 re-
generating systems utilize glassy carbon electrodes and their modified
counterparts to achieve efficient regeneration. Azem et al. reported in
2004 that a glassy carbon electrode modified with a submonolayer of
ruthenium is capable of the direct reduction of NAD to NADH2 with
little to no concomitant formation of [NAD]2 or 1,6-NADH2. The au-
thors made no mention of 1,2-NADH2 in their system. Overall, the ru-
thenium modified glassy carbon electrode allowed a 96% yield of en-
zymatically-active 1,4-NADH2. They reported that the reaction was
highly irreversible, independent of pH, and occurs at high cathodic
overpotentials where the reaction is diffusion controlled and is first
order with respect to NAD. The authors postulated that ruthenium
serves simultaneously to provide sites for adsorbed hydrogen on the
electrode, thereby selectively protonating NAD(P)% radicals as they
form, as well as acting as a physical barrier against the formation of
dimers at the electrode surface (Man and Omanovic, 2004; Azem et al.,
2004).
Ali et al. observed that even unmodified glassy carbon electrodes
are capable of selective 1,4-NADH2 regeneration. At −1.40 V vs. MSE
on an unmodified glassy carbon electrode, the yield of 1,4-NADH2 was
only ~32%. However, at a high overpotential of − 2.30 V vs. MSE on
the same electrode, the yield of 1,4-NADH2 was ~98% (Ali et al.,
2011). Various modifications to glassy carbon electrodes have been
explored to lower high overpotentials. For example, Ali et al. reported
that their glassy carbon electrode nano-patterned with platinum and
nickel gave 100% yield of 1,4-NADH2 from NAD. They posited that
platinum and nickel on the electrode surface provided sites for ad-
sorbed hydrogen close to the sites of dimer formation in order to in-
crease radical protonation kinetics. The authors state that this electrode
has good potential for industrial applications since they were able to
operate at relatively lower overpotentials than on an unmodified glassy
carbon electrode (Ali et al., 2012a).
These authors also indicated other electrode configurations could be
used. Specifically, a carbon nanofiber cathode grown on a stainless steel
mesh was able to yield ~99.3% 1,4-NADH2 at −2.3 V vs MSE. The
authors posited that this electrode could be more commercially viable if
the carbon nanofiber cathode is nano-patterned with platinum or nickel
and further suggested that the high overpotentials could be reduced to
− 1.5 V vs. MSE with the nano-patterned platinum or nickel (Ali et al.,
2012b). Ullah et al. reported that a titanium substrate coated with an
iridium/ruthenium-oxide coating gave a yield of ~88% 1,4-NADH2 at
− 1.70 V vs. MSE. The authors attribute this result to the relatively
weak hydrogen-metal bonds formed on the electrode surface which
provides surface-adsorbed hydrogen to NAD(P)% radicals to favor 1,4-
NAD(P)H2 formation rather than dimer formation while requiring less
energy to break the bond from the electrode surface, resulting in a
lower overpotential (Ullah et al., 2015). This result is particularly in-
teresting in that the authors achieved a relatively high yield of 1,4-
NADH2 at a relatively low overpotential without using a glassy carbon
electrode.
Although many of these reports state there is little to no formation
of 1,6-NAD(P)H2, 1,2-NAD(P)H2, or [NAD(P)]2, these experiments were
done at a relatively small scale compared to what would be expected for
industrial reactors. It may be reasonable, therefore, to expect greater
variability in 1,4-NAD(P)H2 regeneration efficiencies during scale-up.
To that end, Wang and Yiu demonstrated for the first time the use of
a heterogeneous catalyst with H2 as a reductant to regenerate 1,4-
NADH2 in situ. The yield of 1,4-NADH2 was found to increase with
temperature and pH. The 1,4-NADH2 regeneration was coupled with
the conversion of propanal to propanol by alcohol dehydrogenase. The
turnover number for the cofactor was found to be 7. Interestingly, the
catalyst appears to provide surface reactive hydrogen to reduce NAD to
1,4-NADH2 (Wang and Yiu, 2016). Although the mechanistic details are
unclear, such surface reactive hydrogen is analogous to the findings of
Azem, Man and Omanovic, Ullah, and Ali. This is indicative of a more
C.S. Morrison et al.
fundamental, underlying mechanism involving chemisorbed hydrogen
for selective 1,4-NAD(P)H2 regeneration irrespective of the regenera-
tion driving force.
4. Mitigation strategies
4.1. Prevention
One of the most significant factors in achieving efficient electro-
chemical regeneration of 1,4-NAD(P)H2 is the choice of electrode ma-
terials. In general, the electrode should be engineered to improve the
kinetics of the second electron transfer (Step 2) versus the dimerization
of two NAD(P)% radicals (Step 3) in Fig. 3 or to prevent dimerization
altogether. This can be achieved by physically preventing the coupling
of NAD(P)% radicals on the electrode surface, or by enhancing hydrogen
and electron transfer kinetics so that the rate of formation of all isomers
of NADH2 outcompetes the rate of dimer formation.
Long and Chen reported that a silver electrode modified with L-
histidine resulted in the selective formation of 1,4-NADH2 without
concomitant formation of [NAD]2. They used an alcohol dehydrogenase
to catalyze the conversion of acetaldehyde to ethanol with an 82% yield
of active 1,4-NADH2 and a cofactor turnover number of 9 (Long and
Chen, 1997). Although this turnover number is commercially im-
practical, it is nonetheless interesting that relatively high selectivity
was achieved in the formation of 1,4-NADH2.
Along similar lines, Baik et al. reported that a yield of ~10% 1,4-
NADH2 was achieved on a bare gold amalgam electrode, but a yield of
~ 75% 1,4-NADH2 was achieved on a gold amalgam electrode modified
with cholesterol. The authors speculated that cholesterol likely served
as a physical barrier to the formation of [NAD]2 on the electrode sur-
face. Using this system, they were able to catalyze the conversion of
pyruvate to D-lactate by D-lactate dehydrogenase. The NAD turnover
number was reported to be 1400. The authors also suggested that
keeping the concentration of NAD low discourages [NAD]2 formation
and increases total NAD turnovers since there would be a smaller
probability of NAD% radicals undergoing dimerization if there are less
radicals available in the first place (Baik et al., 1999), and this will be
clear from inspection of Fig. 3 above.
As mentioned in Section 3.3, it has been demonstrated that bare and
modified glassy carbon electrodes are also capable of the selective
formation of enzymatically active 1,4-NAD(P)H2 via electrolysis. This
selectivity is likely due to surface adsorbed hydrogen, which is avail-
able to hydrogenate NAD% radicals in apparently the correct orientation
to minimize the formation of isomers and dimers. Various catalysts can
also be employed to encourage the favored reactions.
Established electrochemical preparations of [NAD(P)]2 use alkaline
buffers to stabilize the dimer for further use or analysis (Jaegfeldt,
1981; Schmakel et al., 1975). The dimer is known to be acid-labile
(Schmakel et al., 1975; Bresnahan and Elving, 1981), so 1,4-NAD(P)H2
regeneration systems seeking to avoid the formation of dimer may
benefit from the use of a slightly acidic buffer if the prepared NAD(P)H2
is to be used quickly after preparation.
Indeed, it has been reported that NAD(P)H2 can be electro-
chemically prepared in higher concentrations relative to other mixture
components by increasing the concentration of protons in the buffer
(Jaegfeldt, 1981).
Azem et al. reported an apparent voltage selectivity for the amount
of enzymatically active NADH2 prepared electrochemically at various
electrolysis potentials (Azem et al., 2004). This was corroborated in the
reports by Ullah et al. and Ali et al. where the amount of 1,4-NADH2
produced via electrolysis was maximized at a specific electrolysis po-
tential relative to the choice of electrode materials and running buffers
(Ullah et al., 2015; Ali et al., 2011; Ali et al., 2012a). This clearly in-
dicates that the electrochemical properties of 1,4-NAD(P)H2 allow for a
voltage selectivity effect that an engineered electrochemical bioreactor
system could take advantage of in order to maximize the production of
7
Biotechnology Advances xxx (xxxx) xxx–xxx
1,4-NAD(P)H2.
Clearly, operational controls and conditions can be implemented
into engineered electrochemical bioreactor systems to favor the pro-
duction of the catalytically active 1,4-NAD(P)H2 isomer while si-
multaneously discouraging the formation or persistence of unfavorable
isomers and dimers. Optimization of any particular electrochemical
bioreactor process should begin by investigating the optimum buffer
pH, buffer composition, electrolysis potential, and various other op-
erational controls for the process of interest. In the likely case that there
is no optimal set of conditions under which to selectively form 1,4-NAD
(P)H2, recovery processes should be implemented in tandem with pre-
vention strategies in order to avoid wastage of NAD(P).
4.2. Recovery
4.2.1. Enzymatic recovery of NAD(P) cofactor material by renalase
In 2005, Xu et al. reported the discovery of a novel enzyme, called
renalase, secreted into the blood by the human kidney that modulates
systemic blood pressure by metabolizing circulating catecholamines.
While the highest renalase gene expression was found in the kidney, the
enzyme was also expressed in cardiac tissue, skeletal muscle tissue, and
the small intestine. The enzyme was characterized as an FAD-dependent
amine oxidase important to human health (Xu et al., 2005). In 2010,
Pandini et al. reported that by expressing recombinant human renalase
in E. coli with an N-terminal poly-His tag, enough enzyme could be
obtained to perform biochemical studies (Pandini et al., 2010).
Beaupre et al. reported that there had been no convincing demon-
stration of the claim that renalase does in fact catalyze the oxidation of
circulating catecholamines since earlier work did not sufficiently con-
trol for spontaneous oxidation of catecholamines in buffers containing
dissolved oxygen. Hence, the physiologically-relevant metabolic func-
tion of the enzyme remained unclear (Beaupre et al., 2013). In later
work, this group reported that renalase catalyzes the oxidation of the
undesired isomeric forms of NAD(P)H2 that may arise in vivo via events
such as tautomerization and nonspecific redox events. They postulated
that the biological significance of such an enzyme could be to act as a
housekeeping enzyme that prevents 1,2- or 1,6-NAD(P)H2 from in-
hibiting 1,4-NAD(P)H2-dependent reactions, as they had found that 1,2-
and 1,6-NAD(P)H2 is highly inhibitory to dehydrogenase enzymes. The
biochemical reaction catalyzed by renalase is given in Fig. 6. Renalase
requires an oxidizing environment to catalyze the reaction in which
1,2- and 1,6-NAD(P)H2 are oxidized and the two electrons and two
protons are eventually passed on to O2 as an electron acceptor to form
H2O2 (Beaupre et al., 2015).
Interestingly, a bacterial analog of renalase has been identified in
Pseudomonas phaseolicola which features the same substrate specificity
and catalytic function as human renalase. Hoag et al. reported that this
bacterial renalase is characterized by a marked substrate preference for
NAD isomers over of NADP isomers. The existence of a renalase analog
in an organism without a circulatory system lends further credibility to
the hypothesis that renalase is a cellular housekeeping enzyme (Hoag
et al., 2015) and not a modulator of blood pressure as was originally
proposed.
Recent fundamental work by Moran et al. has been convincing in
elucidating the catalytic activity of renalase, and by extension, the
potential biological implications of the various NAD(P)H2 isomers with
respect to primary metabolism (Moran, 2016). The potential implica-
tions renalase has for applied biochemistry and industrial applications
are the subject of patent applications (Morrison et al., 2017). Since the
catalytic activity of renalase is fundamentally a recovery reaction, an
engineered electrochemical bioreactor system taking advantage of im-
mobilized renalase could theoretically ameliorate the adventitious
electrochemical production of the undesired isomers of NAD(P)H2 that
would otherwise inhibit 1,4-NAD(P)H2-dependent enzymes that gen-
erate the product of interest, and potentially lock all NAD(P) cofactor
material into an unusable form. Either of these events severely limit the
C.S. Morrison et al. Biotechnology Advances xxx (xxxx) xxx–xxx

H2O2 H2O2 Fig. 6. Renalase catalyzes an NAD(P) recovery reaction.

O NH2 O2 O NH2 O2 O NH2
The biochemical reaction catalyzed by renalase oxidizes
H
H 1,2- and 1,6-NAD(P)H2 to NAD(P) with concomitant pro-
H H
H
duction of H2O2.
N+
renalase renalase
N N
R R R
H H

1,2-NAD(P)H2 NAD(P) 1,6-NAD(P)H2

H2N N -
HO O O O
N O P P O
N O O O
R= N
HO OH HO OH

economic potential of any industrial process relying on the electro-
chemical regeneration of 1,4-NAD(P)H2.
4.2.2. Photochemical recovery of NAD(P) from [NAD(P)]2
Bresnahan and Elving report that solutions containing the dimer
[NAD]2 are characteristically yellow-green in color. They observed that
upon exposure to daylight, the dimer solutions lost their yellow-green
color in a few minutes while the solution in their dark electrolytic
chamber remained yellow-green. They posited that a photochemical
reaction was responsible for the loss of dimer. They report that they
illuminated their dimer solutions with UV light at 254 nm or at wave-
lengths exceeding 320 nm and found that the illuminated dimer solu-
tion produced NAD in stoichiometric amounts. They also suggested that
the reaction would produce some isomer NADH2, but their results did
not corroborate this hypothesis (Bresnahan and Elving, 1981).
Avigliano et al. described the light-induced homolytic cleavage of
[NAD]2 dimer to NAD% free radicals which are further oxidized by
oxygen to NAD. They reported significant concomitant oxygen con-
sumption and H2O2 production when [NAD]2 was exposed to UV light
above 300 nm. These observations led them to propose the following
overall reaction:
[NAD]2 + O2 + 2H+ → 2NAD + H2 O2
acceptor in this reaction (Avigliano et al., 1983). 1,4-NADH2 is also
known to participate in photooxidation reactions, presumably through
similar mechanisms as those proposed by Noguchi et al. (Noguchi et al.,
1997).
Czochralska et al. reported that [NADP]2 oxidizes to NADP and
NADPH2 when irradiated at 365 nm with concomitant production of
H2O2. They found that NADPH2 is also subject to oxidation by light, and
that its photooxidation is strongly dependent on the presence of oxygen
whereas the rate of photooxidation of [NADP]2 is similar under aerobic
and anaerobic conditions. The fact that the rate of anaerobic [NADP]2
photooxidation is competitive with aerobic [NADP]2 photooxidation
challenges the aerobic mechanism proposed by Liberatore et al. for
[NAD]2 photooxidation. Based on their experimental work, Czochralska
et al. proposed the following reaction scheme for the anaerobic pho-
toconversion of [NADP]2 at 365 nm as shown in Fig. 8a and b.
While the authors state that it is unclear what intermediates would
be responsible for NADP and H2O2 production, they maintain that it
must be due to some interaction with water (Czochralska et al., 1990).
Photochemical recovery of NAD(P) from [NAD(P)]2 could be simi-
larly engineered into an electrochemical bioreactor system to achieve
the same advantages as an enzymatic or electrochemical recovery re-
action as mentioned in Sections 4.2.3 and 4.2.4.
Reaction (3)

and the reaction products were confirmed with enzymatic assays. The
proposed mechanism is as shown in Fig. 7.
[NAD]2 is put into an excited state by photon absorption. The ex-
cited dimer then breaks up into NAD% free radicals, which can either
interact with oxygen to form the oxidized cofactor NAD and H2O2, or
they can interact with themselves to reform the dimer. Indeed, anae-
robic pathways for [NAD]2 photooxidation involving the formation of
dihydopyridyl radicals via interaction with solvent have been proposed,
but it is not obvious what intermediate would behave as an electron
h zymatically active (Jaegfeldt, 1981). Similar results for the oxidation of
[NAD]2 [NAD]2*
4.2.3. Electrochemical recovery of NAD(P) from [NAD(P)]2
Dimeric species can be electrochemically oxidized to NAD(P).
Specifically, Carelli et al. reported that the dimers undergo oxidation at
− 0.2 V vs. SCE on a hanging mercury drop electrode, 0.5 V vs. SCE on
a platinum electrode, and 0.6 V vs. SCE on a glassy carbon electrode
(Carelli et al., 1981). Jaegfeldt reported that the dimers can be elec-
trochemically oxidized to NAD(P) at − 0.1 V vs. SCE on a mercury
electrode. The author also adds that following electrochemical oxida-
tion, the product was tested for activity against alcohol dehydrogenase
and ethanol. With this assay, 100% of the resulting NAD was en-
[NAD]2 and [NADP]2 dimer were reported by Schmakel et al. (1975).

4.2.4. Enzymatic recovery of NAD(P) from [NAD(P)]2

[NAD]2* 2 NAD Kôno and Suekane reported in 1958 that crude extracts from etio-
lated mung bean seedlings seemingly (but incorrectly) catalyzed the
complete oxidation of all NADH2 isomers to NAD. This was significant
2 NAD + O2 NAD + O2 - given that prior reports had determined that alcohol dehydrogenase,
malic acid dehydrogenase, and lactic acid dehydrogenase could only
partially oxidize the isomers. The discovery of a new enzyme that
NAD + NAD [NAD]2 completed the reaction was promising, but the enzyme they identified
required an unknown cofactor and was stimulated by methylene blue.
Based on their observations, they categorized it as a diaphorase despite
2 O2 - + 2 H+ H2O2 + O2 their experiments involving yeast or porcine diaphorases that showed
that only the mung bean enzyme can completely oxidize the substrate
Fig. 7. The proposed mechanism for the light-induced homolytic cleavage of [NAD]2, (Kôno and Suekane, 1958).
adapted from (Avigliano et al., 1983). In this scheme, [NAD]2 is oxidized to NAD in the Burnett and Underwood (1968) expanded on the work by Kôno and
presence of oxygen.
Suekane, and corrected Kôno's assumption that the mung bean enzyme

8
C.S. Morrison et al.
a duced anaerobic cleavage of [NADP]2, adapted from
[NADP]2 + H2O NADPH2 + NADP
[NADP]2 + H2O NADP + NADPH2 + OH-
b
h
[NADP]2 [NADP]2*
[NADP]2* 2 NADP
NADP + NADP [NADP]2
NADP + H2O NADPH2 + OH
h
2 [NADP]2 2 [NADP]2*
2 [NADP]2* + H+ NADP + further photochemical reactions
used NADH2 isomers as its substrate. They demonstrated experimen-
tally that the mung bean enzyme actually catalyzed the oxidation of
[NAD]2 to NAD. Fricks et al. reported that the catalytic activity of the
enzyme was greatly accelerated upon addition of phenol as Kôno's
unknown cofactor to the reaction mixture, and thus they reclassified the
enzyme as a phenol oxidase. Further, they found analogs of the enzyme
in various other plants such as corn, cotton, wheat, and other beans.
However, the activity of the enzyme from crude extracts was lost upon
purification. They reported that commercially-available mushroom
polyphenol oxidase also catalyzed the oxidation of [NAD]2 to NAD
(Fricks et al., 1973).
Carelli et al. reported that cytoplasmic fractions from rat liver cells
stimulated the uptake of oxygen by [NAD]2, and they verified the
production of NAD and H2O2 by an enzyme assay (Carelli et al., 1980).
Avigliano et al. showed that horseradish peroxidase (HRP) catalyzes
the oxidation of [NAD]2 to NAD in a pH-dependent manner, which
suggests a free radical mechanism similar to that known for the oxi-
dation and peroxidation of NADH2 to NAD under the same conditions
(Avigliano et al., 1985). Two reaction schemes were proposed – one
consuming molecular oxygen as the oxidant, the other using hydrogen
peroxide. In either case, the result is conversion of the [NAD]2 dimer to
two molecules of NAD plus a water molecule. Interestingly, bacterial
NADH2 peroxidase was unable to catalyze the oxidation of [NAD]2 to
NAD. The authors suggest it is because bacterial NADH2 peroxidase
involves a two-electron transfer, but [NAD]2 oxidation is only catalyzed
by one-electron transfers which further supports the hypothesis that the
responsible mechanism involves free radicals. Kirkor and Scheeline also
reported on the enzymatic activity of HRP with [NAD]2 to form NAD. In
addition, they reported that [NAD]2 participates in other biological
redox reactions that allows for the reintroduction of HRP into the os-
cillatory peroxidase-oxidase cycle (Kirkor and Scheeline, 2000).
While these reports have hinted at the potential biological sig-
nificance of [NAD(P)]2 in primary metabolism, none to date have
identified the potential implications of enzymatic [NAD(P)]2 recovery
methods to industrial applications. Similar to the case for renalase,
immobilized enzymes that are able to catalyze the recovery of NAD(P)
Biotechnology Advances xxx (xxxx) xxx–xxx
Fig. 8. a. The proposed overall reactions for the light-in-
+ OH (Czochralska et al., 1990). In this scheme, [NADP]2 un-
dergoes homolytic cleavage and rearrangement to form
NADP and NADPH2. b. The proposed mechanism for the
light-induced anaerobic cleavage of [NADP]2, adapted
from (Czochralska et al., 1990). In this scheme, a photo-
excited [NADP]2 can react to form NADP and NADPH2.
from [NAD(P)]2 could be engineered into an electrochemical bioreactor
system to deal with the adventitious formation of dimer that would
otherwise lock all input NAD(P) cofactor material into an unusable
form and severely limit the number of turnovers theoretically achiev-
able by the 1,4-NAD(P)H2 regeneration system (Morrison et al., 2017).
4.2.5. Other recovery reactions
Molecular oxygen is known to oxidize [NAD(P)]2 to NAD(P) with
increasing oxidation rates at higher temperatures and lower pH. In an
anoxic environment, [NAD(P)]2 has been reportedly stable for several
days (Kovář et al., 1985). This knowledge was gained in the context of
[NAD(P)]2 stability in various storage conditions, so it is unlikely that
molecular oxygen will be a good choice of method for intentional NAD
(P) recovery from [NAD(P)]2 since the rate of [NAD(P)]2 oxidation is on
the time scale of several days.
Doxorubicin, an anthracycline chemotherapeutic drug, has been
reported to oxidize [NAD]2 to NAD in anaerobic conditions when ex-
posed to light. It is thought to occur through a series of one-electron
transfers involving NAD% radicals. The same reaction is not observed
when 1,4-NADH2 is used in place of [NAD]2 (Carelli et al., 1988). Given
its clinical use and deleterious health effects in humans, it is unlikely
that doxorubicin will be a good choice for NAD(P) recovery from [NAD
(P)]2 in larger-scale applications.
One-electron-accepting proteins, such as cytochrome c, azurin and
methemoglobin, have also reportedly been able to oxidize [NAD(P)]2 to
NAD(P) via a reaction involving NAD(P)% radicals. Avigliano et al. re-
port that while [NAD(P)]2 reacts generally favorably with one-electron-
acceptors, the dimers are unable to react with two-electron-acceptors.
While these proteins may not be the best choices for engineering NAD
(P) recovery reactions in larger-scale applications, the dimers may in-
stead be suitable species for discriminating between one-electron-ac-
ceptors and two-electron-acceptors in biological redox reactions
(Avigliano et al., 1986).
Czochralska et al. reported that [NADP]2 is also oxidized to NADP
by iron chelates. Fe(CN)63 − oxidized [NADP]2 in the dark due simply
to the more positive redox potential of the Fe(CN)63 −/Fe(CN)64 −
9
C.S. Morrison et al.
couple than that of the NAD(P)/NAD(P)% couple. They also found that
Fe(EDTA)2 + and Fe(CN)63 − were able to increase the rate of [NADP]2
oxidation to NADP upon irradiation at 365 nm (Czochralska et al.,
1990). It follows, then, that other redox pairs with redox potentials
more positive than that of the NAD(P)/NAD(P)% couple should be able
to chemically oxidize [NAD(P)]2 to NAD(P). While this may be inter-
esting to some researchers hoping to recover NAD(P) from NAD(P) in
small-scale applications, it is unlikely that this strategy could be em-
ployed for larger scale cofactor regeneration since the addition of extra
chemicals to the system constitutes an added material cost in addition
to adding to the cost and complexity of downstream purifications.
In summary, it is now well established in the literature that various
enzymatic, chemical, and photochemical dimer oxidation mechanisms
involve the formation of NAD(P)% radicals in aerobic and anaerobic
reactions. This may point to some fundamental biochemical property of
the dimer that is important to redox processes involving NAD(P)H2
(Avigliano et al., 1985; Avigliano et al., 1983; Czochralska et al., 1990;
Carelli et al., 1988; Avigliano et al., 1986).
5. Conclusions
5.1. Opportunities for in vitro and in vivo biocatalysis
5.1.1. P450 monooxygenase reactions
There are many other potential applications beyond ketoreductases.
The P450 monooxygenases, which require a reducing equivalent to take
up one oxygen atom from an oxygen molecule transforming it to water,
are very important reactions in living systems that are responsible for
inserting an oxygen atom between a carbon hydrogen bond (Urlacher
and Eiben, 2006; Vilker et al., 1999; Leonard and Koffas, 2007; Leonard
et al., 2006; Leonard et al., 2005). There are no known chemical cat-
alysts that are capable of accomplishing this transformation that is es-
sential to steroid hydroxylation, metabolite formation, and general
biological degradation of organic materials in the environment (see
Fig. 9).
5.1.2. Microbial electrosynthesis
The application of electrochemical bioreactors to the field of mi-
crobial whole-cell electrosynthesis is not comprehensively reviewed in
this paper. However, it deserves to be mentioned in order to make the
reader aware of the applications of electrochemical bioreactors in the
context of metabolic engineering, synthetic biology, and industrial
microbiology. The reader who is interested in microbial electrosynth-
esis is encouraged to read some recent review articles that have given
an extensive treatment of the topic (Kato, 2015; Mohan et al., 2014;
Choi and Sang, 2016; TerAvest and Ajo-Franklin, 2016; Tremblay and
Zhang, 2015; Harnisch et al., 2015).
In general, the concept of microbial electrosynthesis is driven by the
R1
R1
H O
H H O2 H2O
R2
R2
R1
R1
P450
O
enzyme
R2
R2
R1
R1 O
O 1,4-NAD(P)H2 NAD(P)
O
R2
R2
Fig. 9. The utility of 1,4-NAD(P)H2-requiring P450 enzymes for biocatalysis is high-
lighted through various chemical transformations. The cost-efficient regeneration of 1,4-
NADPH2 is essential to give access to these reactions on a meaningful scale.
Biotechnology Advances xxx (xxxx) xxx–xxx
notion that the redox states of cells in an actively fermenting culture
can be fundamentally modified in order to drive up the production of a
reduced fermentation product. This is thought to occur by reducing the
amount of carbonaceous feedstock that is sacrificially oxidized by the
cells to provide electrons for the reduction of NAD(P) to 1,4-NAD(P)H2,
thereby increasing the efficiency of carbon flux through metabolic
pathways toward the target product. For example, the yield of 1,3-
propanediol by a mixed culture growing on glycerol was enhanced from
24.8% to 50.1% when an external current was applied to the culture.
The authors found through a metabolic flux analysis that glycerol me-
tabolism was redirected from propionate fermentation to 1,3-propane-
diol production (Zhou et al., 2013). An example of a process catalyzed
by a recombinant culture is the production of biofuels by Ralstonia
eutropha. The authors reported the electrically-driven generation of
140 mg L− 1 of biofuels from CO2 alone (Li et al., 2012).
However, the mechanisms by which electrically-driven cultures in-
crease the carbon efficiencies and yields of reduced fermentation pro-
ducts is not well understood. Harrington et al. looked at the metabolite
profiles of Escherichia coli, Klebsiella pneumoniae, and Zymomonas mobilis
following neutral red-mediated electrosynthesis to observe changes in
the organisms' metabolisms during electrically-driven fermentations.
They found through stoichiometric analysis that there were significant
metabolite changes observed during electrosynthesis for E. coli and K.
pneumoniae. No changes in Z. mobilis metabolism were observed. While
the authors claim that NAD reduction alone could not account for all
the metabolite changes observed, it is clear from their work that a
variety of electrically-driven effects on microbial metabolism culminate
in an increase of carbon efficiency and yield of fermentation products
(Harrington et al., 2015). There is a significant need for more research
to be done to understand the mechanisms by which some organisms are
able to interact with their external environment to alter their own in-
tracellular redox environment. A clearer fundamental understanding of
these phenomena could lead to the development of industrial microbial
strains that are engineered to optimize target production without in-
creasing the amount of input carbonaceous material in electrically-
driven fermentations.
Although more work remains to be done to understand microbial
electrosynthesis before extensive engineering can take place, the reader
is encouraged to read the review by White et al. for a current, com-
prehensive understanding of how bacterial extracellular electron ex-
change is thought to occur (White et al., 2016).
5.1.3. In vitro biocatalysis
For the first time, strategies for dealing with the adventitious for-
mation of useless and inhibitory reduced cofactor side products are
presented in the context of applied biochemistry and industrial bioca-
talysis. Such reactions are henceforth designated as “recovery reac-
tions” and are crucial to the success of engineered electrochemical NAD
(P)H2 regeneration systems. Recovery reactions should be used in
tandem with operational controls to minimize or eliminate the forma-
H
tion of reduced cofactor side products while simultaneously maximizing
the selective formation of 1,4-NAD(P)H2.
5.2. Current state-of-the-art for electrochemical reactor technology
To date, engineered electrochemical bioreactor systems have in-
evitably suffered from the nonspecific reduction problem described in
Section 2.2. It has not been possible to engineer an efficient or an
economic system for the concomitant in situ regeneration of 1,4-NAD
(P)H2 with the formation of a product of interest via the catalytic ac-
tivity of a redox enzyme. This review therefore serves as a path toward
engineered solutions by which to achieve efficient and economic large-
scale 1,4-NAD(P)H2-dependent biocatalytic processes using electro-
chemical methods.
10
C.S. Morrison et al.
5.3. Future outlook
Many practical challenges remain for the electrochemical re-
generation of 1,4-NAD(P)H2, and this paper concludes by calculating
some benchmarks which should be kept in mind for practical systems.
Generating hydrogen gas by conventional electrolysis using power
from the electrical grid at US$60/KW gives a cost approximately US
$3200–$3500/mt. If hydrogen is supplied within a biological process
by the metabolism of glucose having a cost of US$450/mt, and running
a “costless” process by which glucose is metabolized completely to CO2
and 1,4-NAD(P)H2, the equivalent hydrogen cost is US$6750/mt.
If 1,4-NAD(P)H2 is generated by oxidizing glucose to glucono-
lactone in a coupled enzyme system using glucose dehydrogenase, the
equivalent hydrogen cost is US$40,500/mt. Replacing glucose with
sodium formate at a cost of US$250/mt and using formate dehy-
drogenase to generate 1,4-NAD(P)H2 and CO2, the equivalent hydrogen
cost is US$8500/mt. Using ethanol at the current price of US$1.50/
USgal, or US$500/mt, in a process in which acetaldehyde is generated,
the equivalent hydrogen cost is US$16,000/mt. This could be reduced
to US$8000/mt hydrogen if the ethanol were oxidized to acetic acid.
Using an electrochemical bioreactor with the same efficiency as a
current electrolyzer would give the same equivalent hydrogen cost; US
$3200–3500/mt, but this would be delivered as 1,4-NAD(P)H2 to the
enzyme of interest. Like a standard electrolyzer, this allows the current
electrical grid to be used as a fungible hydrogen source, with any
proportion of power from renewable sources that may be available.
Of course, the cofactor material itself has a cost, and at present this
is approximately US$1200/kg for NAD. One mole of NAD can be loaded
with one mole of hydrogen as 1,4-NADH2, and with a molecular weight
of 665 for 1,4-NADH2, that would give a cofactor cost of just under US
$500 per mol of reduced cofactor – completely swamping any con-
sideration of the cost of the hydrogen that mol of cofactor is carrying!
Hence the importance of recycling the cofactor material, and ensuring it
is all recovered as NAD ready to be put through the reduction cycle
again. This will be true for any in vitro system that is used to perform
cofactor recycling, enzymatic or non-enzymatic. With 1000 recycles of
the cofactor, the cost contribution of the cofactor material drops to US
$0.50 per mol of delivered hydrogen. But with 500,000 mols of hy-
drogen per mt, this is still US$250,000 cofactor cost per mt of delivered
hydrogen.
Products in which one mole of hydrogen is consumed in a reaction
giving a product with a molecular weight of 400, and with a cofactor
recycle of 1000 times, gives a cofactor cost of contribution of just over
US$100 per kilogram of product. This is well within the acceptable cost
range of a process step in the synthesis of a pharmaceutical, and at
10,000 turnovers of the cofactor, would give a step cost acceptable for
use in high value specialty products.
It is anticipated that laboratory-scale in vitro electrochemical bior-
eactor systems capable of routinely supplying g/hour stoichiometric
quantities of 1,4-NAD(P)H2 will become practical in the near future.
These systems will significantly lower the cost of 1,4-NAD(P)H2 while
facilitating easy use of 1,4-NAD(P)H2 dependent enzymes in potentially
useful enzyme reactions.
The larger-scale commercial production of 1,4-NAD(P)H2-depen-
dent products will be on the horizon with the first processes most like
involving P450 monooxygenase and other high value-added pharma-
ceutical intermediates.
Conflict of interest statement
The authors declare that CSM, WBA, DRD, and MAGK are co-in-
ventors on a patent application related to the application of renalase to
electrochemical bioreactor systems.
11
Biotechnology Advances xxx (xxxx) xxx–xxx
Acknowledgments
This work was supported by BioChemInsights, Inc., Malvern, PA
through a Sponsored Research Agreement with Rensselaer Polytechnic
Institute; and the National Science Foundation Graduate Research
Fellowship Program [grant number DGE-1247271].
References
Ali, I., Gill, A., Omanovic, S., 2012a. Direct electrochemical regeneration of the enzymatic
cofactor 1,4-NADH employing nano-patterned glassy carbon/Pt and glassy carbon/Ni
electrodes. Chem. Eng. J. 188, 173–180. http://dx.doi.org/10.1016/j.cej.2012.02.
005.
Ali, I., McArthur, M., Hordy, N., Coulombe, S., Omanovic, S., 2012b. Electrochemical
regeneration of the cofactor NADH employing a carbon nanofibers cathode. Int. J.
Electrochem. Sci. 7, 7675–7683.
Ali, I., Soomro, B., Omanovic, S., 2011. Electrochemical regeneration of NADH on a glassy
carbon electrode surface: the influence of electrolysis potential. Electrochem.
Commun. 13, 562–565. http://dx.doi.org/10.1016/j.elecom.2011.03.010.
Angelastro, A., Dawson, W.M., Luk, L.Y.P., Allemann, R.K., 2017. A versatile disulfide-
driven recycling system for NADP+ with high cofactor turnover number. ACS Catal.
7, 1025–1029. http://dx.doi.org/10.1021/acscatal.6b03061.
Avigliano, L., Carelli, V., Casini, A., Finazzi-Agrò, A., Liberatore, F., 1983. Photocatalyzed
oxidation of NAD dimers. Biochim. Biophys. Acta Bioenerg. 723, 372–375. http://dx.
doi.org/10.1016/0005-2728(83)90043-9.
Avigliano, L., Carelli, V., Casini, A., Finazzi-Agrò, A., Liberatore, F., 1985. Oxidation of
NAD dimers by horseradish peroxidase. Biochem. J. 226, 391–395.
Avigliano, L., Carelli, V., Casini, A., Finazzi-Agrò, A., Liberatore, F., Rossi, A., 1986.
Oxidation of nicotinamide coenzyme dimers by one-electron-accepting proteins.
Biochem. J. 237.
Azem, A., Man, F., Omanovic, S., 2004. Direct regeneration of NADH on a ruthenium
modified glassy carbon electrode. J. Mol. Catal. A Chem. 219, 283–299. http://dx.
doi.org/10.1016/j.molcata.2004.04.041.
Baik, S.H., Kang, C., Jeon, I.C., Yun, S.E., 1999. Direct electrochemical regeneration of
NADH from NAD + using cholesterol-modified gold amalgam electrode. Biotechnol.
Tech. 13, 1–5. http://dx.doi.org/10.1023/A:1008865212773.
Beaupre, B.A., Carmichael, B.R., Hoag, M.R., Shah, D.D., Moran, G.R., 2013. Renalase is
an α-NAD(P)H oxidase/anomerase. J. Am. Chem. Soc. 135, 13980–13987. http://dx.
doi.org/10.1021/ja407384h.
Beaupre, B.A., Hoag, M.R., Roman, J., Försterling, F.H., Moran, G.R., 2015. Metabolic
function for human renalase: oxidation of isomeric forms of β-NAD(P)H that are
inhibitory to primary metabolism. Biochemistry 54, 795–806. http://dx.doi.org/10.
1021/bi5013436.
Biellmann, J.F., Lapinte, C., Haid, E., Weimann, G., 1979. Structure of lactate dehy-
drogenase inhibitor generated from coenzyme. Biochemistry 18, 1212–1217. http://
dx.doi.org/10.1021/bi00574a015.
Bresnahan, W., Elving, P., 1981. Spectrophotometric investigation of products formed
following the initial one-electron electrochemical reduction of nicotinamide adenine
dinucleotide (NAD+). Biochim. Biophys. Acta Gen. Subj. 678, 151–156. http://dx.
doi.org/10.1016/0304-4165(81)90200-2.
Burnett, R.W., Underwood, A.L., 1968. A dimer of diphosphopyridine nucleotide.
Biochemistry 7, 3328–3333. http://dx.doi.org/10.1021/bi00850a003.
Carelli, V., Casini, A., Finazzi-Agro, A., Liberatore, F., Tortorella, S., 1988. Photocatalyzed
anaerobic oxidation of nicotinamide coenzyme dimers to NAD+ by adriamycin. Free
Radic. Res. Commun. 4, 397–402.
Carelli, V., Liberatore, F., Casini, A., Mondelli, R., Arnone, A., Carelli, I., Rotilio, G.,
Mavelli, I., 1980. Dimers of nicotinamide adenine dinucleotide: new evidence for the
structure and the involvement in an enzymatic redox process. Bioorg. Chem. 9,
342–351. http://dx.doi.org/10.1016/0045-2068(80)90044-9.
Carelli, I., Rosati, R., Casini, A., 1981. Electrochemical oxidation of dimers of nicotina-
mide adenine dinucleotide in aqueous medium. Electrochim. Acta Pergamon. http://
dx.doi.org/10.1016/0013-4686(81)85149-3.
Chenault, H.K., Simon, E.S., Whitesides, G.M., 1988. Cofactor regeneration for enzyme-
catalysed synthesis. Biotechnol. Genet. Eng. Rev. 6, 221–270. http://dx.doi.org/10.
1080/02648725.1988.10647849.
Chenault, H.K., Whitesides, G.M., 1987. Regeneration of nicotinamide cofactors for use in
organic synthesis. Appl. Biochem. Biotechnol. 14, 147–197. http://dx.doi.org/10.
1007/BF02798431.
Choi, O., Sang, B.-I., 2016. Extracellular electron transfer from cathode to microbes:
application for biofuel production. Biotechnol. Biofuels 9, 11. http://dx.doi.org/10.
1186/s13068-016-0426-0.
Czochralska, B., Bojarska, E., Pawlicki, K., Shugar, D., 1990. Photochemical and enzy-
matic redox transformations of reduced forms of coenzyme NADP +. Photochem.
Photobiol. 51, 401–410. http://dx.doi.org/10.1111/j.1751-1097.1990.tb01731.x.
Eikeren, P. Van, 1989. Use and regeneration of reagents using coupled reactions and
permselective barriers. EP0349204 A3.
Elving, P.J., Bresnahan, W.T., Moiroux, J., Samec, Z., 1982. NAD/NADH as a model redox
system: mechanism, mediation, modification by the environment. Bioelectrochem.
Bioenerg. 9, 365–378. http://dx.doi.org/10.1016/0302-4598(82)80026-3.
Fricks, D.H., Bechtel, J.T., Underwood, A.L., 1973. Reactivity of nicotinamide-adenine
dinucleotide dimer in plant phenol oxidase systems. Arch. Biochem. Biophys. 159,
837–841. http://dx.doi.org/10.1016/0003-9861(73)90525-0.
Godtfredsen, S.E., Ottesen, M., 1978. 1,6-dihydro-NAD as an humidity-induced lactate
dehydrogenase inhibitor in NADH preparations. Carlsb. Res. Commun. 43, 171–175.
http://dx.doi.org/10.1007/BF02914239.
Harnisch, F., Rosa, L.F.M., Kracke, F., Virdis, B., Krömer, J.O., 2015. Electrifying white
biotechnology: engineering and economic potential of electricity-driven bio-
C.S. Morrison et al.
production. ChemSusChem 8, 758–766. http://dx.doi.org/10.1002/cssc.201402736.
Harrington, T.D., Mohamed, A., Tran, V.N., Biria, S., Gargouri, M., Park, J.-J., Gang, D.R.,
Beyenal, H., 2015. Neutral red-mediated microbial electrosynthesis by Escherichia
coli, Klebsiella pneumoniae, and Zymomonas mobilis. Bioresour. Technol. 195,
57–65. http://dx.doi.org/10.1016/j.biortech.2015.06.005.
Hildebrand, F., Lütz, S., 2007. Electroenzymatic synthesis of chiral alcohols in an
aqueous–organic two-phase system. Tetrahedron Asymmetry 18, 1187–1193. http://
dx.doi.org/10.1016/j.tetasy.2007.05.002.
Hoag, M.R., Roman, J., Beaupre, B.A., Silvaggi, N.R., Moran, G.R., 2015. Bacterial re-
nalase: structure and kinetics of an enzyme with 2- and 6-dihydro-β-NAD(P) oxidase
activity from Pseudomonas phaseolicola. Biochemistry 54, 3791–3802. http://dx.doi.
org/10.1021/acs.biochem.5b00451.
Hollmann, F., Hofstetter, K., Schmid, A., 2006. Non-enzymatic regeneration of nicotina-
mide and flavin cofactors for monooxygenase catalysis. Trends Biotechnol. 24,
163–171. http://dx.doi.org/10.1016/j.tibtech.2006.02.003.
Hollmann, F., Schmid, A., Steckhan, E., 2001. The first synthetic application of a
monooxygenase employing indirect electrochemical NADH regeneration. Angew.
Chem. Int. Ed. 40, 169–171. http://dx.doi.org/10.1002/1521-3773(20010105)
40:1<169::AID-ANIE169>3.0.CO;2-T.
Homann, M.J., Previte, E., 1996. Stereoselective microbial reduction of 5-fluorophenyl-5-
oxo-pentanoic acid and a phenyloxazolidinone condensation product thereof.
US5618707 A.
Jaegfeldt, H., 1981. A study of the products formed in the electrochemical reduction of
nicotinamide-adenine-dinucleotide. Bioelectrochem. Bioenerg. 8, 355–370. http://
dx.doi.org/10.1016/0302-4598(81)80018-9.
Kang, H.S., Na, B.K., Park, D.H., 2007. Oxidation of butane to butanol coupled to elec-
trochemical redox reaction of NAD +/NADH. Biotechnol. Lett. 29, 1277–1280.
http://dx.doi.org/10.1007/s10529-007-9385-7.
Kato, S., 2015. Biotechnological aspects of microbial extracellular electron transfer.
Microbes Environ. 30, 133–139. http://dx.doi.org/10.1264/jsme2.ME15028.
Kirkor, E.S., Scheeline, A., 2000. Nicotinamide adenine dinucleotide species in the
horseradish peroxidase-oxidase oscillator. Eur. J. Biochem. 267, 5014–5022. http://
dx.doi.org/10.1046/j.1432-1327.2000.01554.x.
Kôno, T., Suekane, M., 1958. A diphosphopyridine nucleotide oxidase in mung bean
seedlings. Bull. Agric. Chem. Soc. Jpn. 22 (6), 404–410.
Kovář, J., Turánek, J., Hlaváč, C., Vala, V., Kahle, V., 1985. Liquid chromatographic se-
parations of dimers of nicotinamide adenine dinucleotide and nicotinamide adenine
dinucleotide phosphate. J. Chromatogr. A 319, 341–349. http://dx.doi.org/10.1016/
S0021-9673(01)90570-9.
Lau, C., Flechsig, G.-U., Gründler, P., Wang, J., 2005. Electrochemistry of nicotinamide
adenine dinucleotide (reduced) at heated platinum electrodes. Anal. Chim. Acta 554,
74–78. http://dx.doi.org/10.1016/j.aca.2005.08.044.
Lee, L.G., Whitesides, G.M., 1985. Enzyme-catalyzed organic synthesis: a comparison of
strategies for in situ regeneration of NAD from NADH. J. Am. Chem. Soc. 107,
6999–7008. http://dx.doi.org/10.1021/ja00310a043.
Leonard, E., Koffas, M.A.G., 2007. Engineering of artifical plant cytochrome P450 en-
zymes for synthesis of isoflavones by Escherichia coli. Appl. Environ. Microbiol. 73,
7246–7251.
Leonard, E., Yan, Y., Koffas, M.A., 2006. Functional expression of a P450 flavonoid hy-
droxylase for the biosynthesis of plant-specific hydroxylated flavonols in Escherichia
coli. Metab. Eng. 8, 172–181.
Leonard, E., Yan, Y., Lim, K.H., Koffas, M.A., 2005. Investigation of two distinct flavone
synthases for plant-specific flavone biosynthesis in Saccharomyces cerevisiae. Appl.
Environ. Microbiol. 71, 8241–8248.
Leuchtenberger, W., Huthmacher, K., Drauz, K., 2005. Biotechnological production of
amino acids and derivatives: current status and prospects. Appl. Microbiol.
Biotechnol. 69, 1–8. http://dx.doi.org/10.1007/s00253-005-0155-y.
Li, H., Opgenorth, P.H., Wernick, D.G., Rogers, S., Wu, T.-Y., Higashide, W., Malati, P.,
Huo, Y.-X., Cho, K.M., Liao, J.C., 2012. Integrated electromicrobial conversion of CO2
to higher alcohols. Science 80 (335), 1596. http://dx.doi.org/10.1126/science.
1217643.
Long, Y.-T., Chen, H.-Y., 1997. Electrochemical regeneration of coenzyme NADH on a
histidine modified silver electrode. J. Electroanal. Chem. 440, 239–242. http://dx.
doi.org/10.1016/S0022-0728(97)80061-7.
Ma, S.M., Garcia, D.E., Redding-Johanson, A.M., Friedland, G.D., Chan, R., Batth, T.S.,
Haliburton, J.R., Chivian, D., Keasling, J.D., Petzold, C.J., Soon Lee, T., Chhabra,
S.R., 2011. Optimization of a heterologous mevalonate pathway through the use of
variant HMG-CoA reductases. Metab. Eng. 13, 588–597. http://dx.doi.org/10.1016/
j.ymben.2011.07.001.
Ma, S.K., Gruber, J., Davis, C., Newman, L., Gray, D., Wang, A., Grate, J., Huisman, G.W.,
Sheldon, R.A., 2010. A green-by-design biocatalytic process for atorvastatin inter-
mediate. Green Chem. 12, 81–86. http://dx.doi.org/10.1039/B919115C.
Mädje, K., Schmölzer, K., Nidetzky, B., Kratzer, R., 2012. Host cell and expression en-
gineering for development of an E. coli ketoreductase catalyst: enhancement of for-
mate dehydrogenase activity for regeneration of NADH. Microb. Cell Fact 11, 7.
http://dx.doi.org/10.1021/bp0200954.
Man, F., Omanovic, S., 2004. A kinetic study of NAD + reduction on a ruthenium mod-
ified glassy carbon electrode. J. Electroanal. Chem. 568, 301–313. http://dx.doi.org/
10.1016/j.jelechem.2004.02.006.
McClay, K., Fox, B.G., Steffan, R.J., 2000. Toluene monooxygenase-catalyzed epoxidation
of alkenes. Appl. Environ. Microbiol. 66, 1877–1882. http://dx.doi.org/10.1128/
AEM.66.5.1877-1882.2000.
Mohan, S.V., Velvizhi, G., Vamshi Krishna, K., Lenin Babu, M., 2014. Microbial catalyzed
electrochemical systems: a bio-factory with multi-facet applications. Bioresour.
Technol. 165, 355–364. http://dx.doi.org/10.1016/j.biortech.2014.03.048.
Moran, G.R., 2016. The catalytic function of renalase: a decade of phantoms. Biochim.
Biophys. Acta 1864, 177–186. http://dx.doi.org/10.1016/j.bbapap.2015.04.010.
12
Biotechnology Advances xxx (xxxx) xxx–xxx
Morrison, C.S., Armiger, W.B., Dodds, D.R., Koffas, M.A.G., 2017. Improved Method for
Using Electrochemical Bioreactor Module With Recovery of Cofactor.
(WO2017160793).
Noguchi, N., Tachikawa, M., Takahashi, H., 1997. One-electron photooxidation of Nadh:
laser photolysis, time-resolved raman and Ab initio Mo calculation study. In:
Spectroscopy of Biological Molecules: Modern Trends. Springer, Netherlands,
Dordrecht, pp. 161–162. http://dx.doi.org/10.1007/978-94-011-5622-6_71.
Pandini, V., Ciriello, F., Tedeschi, G., Rossoni, G., Zanetti, G., Aliverti, A., 2010. Synthesis
of human renalase1 in Escherichia coli and its purification as a FAD-containing ho-
loprotein. Protein Expr. Purif. 72, 244–253. http://dx.doi.org/10.1016/j.pep.2010.
03.008.
Patel, R., 2006. Biocatalysis: synthesis of chiral intermediates for pharmaceuticals. Curr.
Org. Chem. 10, 1289–1321. http://dx.doi.org/10.2174/138527206777698011.
Rabaey, K., Angenent, L., Schroder, U., Keller, J., 2010. Bioelectrochemical Systems:
From Extracellular Electron Transfer to Biotechnological Application, 1st ed. IWA
Publishing, London.
Schmakel, C.O., Santhanam, K.S.V., Elving, P.J., 1975. Nicotinamide adenine dinucleo-
tide (NAD +) and related compounds. Electrochemical redox pattern and allied
chemical behavior. J. Am. Chem. Soc. 97, 5083–5092. http://dx.doi.org/10.1021/
ja00851a010.
Schroer, K., Lütz, S., 2009. A continuously operated bimembrane reactor process for the
biocatalytic production of (2R,5R)-hexanediol. Org. Process. Res. Dev. 13,
1202–1205. http://dx.doi.org/10.1021/op9001643.
Schroer, K., Mackfeld, U., Tan, I.A.W., Wandrey, C., Heuser, F., Bringer-Meyer, S.,
Weckbecker, A., Hummel, W., Daußmann, T., Pfaller, R., Liese, A., Lütz, S., 2007.
Continuous asymmetric ketone reduction processes with recombinant Escherichia
Coli. J. Biotechnol. 132, 438–444. http://dx.doi.org/10.1016/j.jbiotec.2007.08.003.
Shin, I.H., Jeon, S.J., Park, H.S., Park, D.H., 2004. Catalytic oxidoreduction of pyruvate/
lactate and acetaldehyde/ethanol coupled to electrochemical oxidoreduction of NAD
+/NADH. J. Microbiol. Biotechnol. 14, 540–546.
Siu, E., Won, K., Park, C.B., 2007. Electrochemical regeneration of NADH using con-
ductive vanadia-silica xerogels. Biotechnol. Prog. 23, 293–296. http://dx.doi.org/10.
1021/bp060247l.
Ströhle, F.W., Kranen, E., Schrader, J., Maas, R., Holtmann, D., 2016. A simplified process
design for P450 driven hydroxylation based on surface displayed enzymes.
Biotechnol. Bioeng. 113, 1225–1233. http://dx.doi.org/10.1002/bit.25885.
TerAvest, M.A., Ajo-Franklin, C.M., 2016. Transforming exoelectrogens for biotechnology
using synthetic biology. Biotechnol. Bioeng. 113, 687–697. http://dx.doi.org/10.
1002/bit.25723.
Tremblay, P.-L., Zhang, T., 2015. Electrifying microbes for the production of chemicals.
Front. Microbiol. 6, 201. http://dx.doi.org/10.3389/fmicb.2015.00201.
Ullah, N., Ali, I., Omanovic, S., 2015. Direct electrocatalytic reduction of coenzyme NAD
+ to enzymatically-active 1,4-NADH employing an iridium/ruthenium-oxide elec-
trode. Mater. Chem. Phys. 149–150, 413–417. http://dx.doi.org/10.1016/j.
matchemphys.2014.10.038.
Urlacher, V.B., Eiben, S., 2006. Cytochrome P450 monooxygenases: perspectives for
synthetic application. Trends Biotechnol. 24, 324–330. http://dx.doi.org/10.1016/j.
tibtech.2006.05.002.
Vilker, V.L., Reipa, V., Mayhew, M., Holden, M.J., 1999. Challenges in capturing oxy-
genase activity in vitro. J. Am. Oil Chem. Soc. 76, 1283–1289. http://dx.doi.org/10.
1007/s11746-999-0140-1.
Walsh, F., Reade, G., 1994. Design and performance of electrochemical reactors for ef-
ficient synthesis and environmental treatment. Part 2. Typical reactors and their
performance. Analyst 119, 797. http://dx.doi.org/10.1039/an9941900797.
Wang, X., Yiu, H.H.P., 2016. Heterogeneous catalysis mediated cofactor NADH re-
generation for enzymatic reduction. ACS Catal. 6, 1880–1886. http://dx.doi.org/10.
1021/acscatal.5b02820.
White, G.F., Edwards, M.J., Gomez-Perez, L., Richardson, D.J., Butt, J.N., Clarke, T.A.,
2016. Mechanisms of bacterial extracellular electron exchange. Adv. Microb. Physiol.
68, 87–138. http://dx.doi.org/10.1016/bs.ampbs.2016.02.002.
Williams, J.N., 1952. Inhibition of coenzyme I-requiring enzyme by adenine and Adenyl
metabolites in vitro. J. Biol. Chem. 195, 629.
Wong, C.H., Whitesides, G.M., 1982. Enzyme-catalyzed organic synthesis: NAD(P)H co-
factor regeneration using ethanol/alcohol dehydrogenase/aldehyde dehydrogenase
and methanol/alcohol dehydrogenase/aldehyde dehydrogenase/formate dehy-
drogenase. J. Org. Chem. 47, 2816–2818. http://dx.doi.org/10.1021/jo00135a037.
Wong, C.H., Whitesides, G.M., 1983. Enzyme-catalyzed organic synthesis: regeneration of
deuterated nicotinamide cofactors for use in large-scale enzymatic synthesis of
deuterated substances. J. Am. Chem. Soc. 105, 5012–5014. http://dx.doi.org/10.
1021/ja00353a026.
Xu, J., Li, G., Wang, P., Velazquez, H., Yao, X., Li, Y., Wu, Y., Peixoto, A., Crowley, S.,
Desir, G.V., 2005. Renalase is a novel, soluble monoamine oxidase that regulates
cardiac function and blood pressure. J. Clin. Investig. 115, 1275–1280.
Zaks, A., Dodds, D.R., 1997. Application of biocatalysis and biotransformations to the
synthesis of pharmaceuticals. Drug Discov. Today 2, 513–531. http://dx.doi.org/10.
1016/S1359-6446(97)01078-7.
Zambianchi, F., Pasta, P., Carrea, G., Colonna, S., Gaggero, N., Woodley, J.M., 2002. Use
of isolated cyclohexanone monooxygenase from recombinantEscherichia coli as a
biocatalyst for Baeyer-Villiger and sulfide oxidations. Biotechnol. Bioeng. 78,
489–496. http://dx.doi.org/10.1002/bit.10207.
Zhao, H., van der Donk, W.A., 2003. Regeneration of cofactors for use in biocatalysis.
Curr. Opin. Biotechnol. 14, 583–589. http://dx.doi.org/10.1016/j.copbio.2003.09.
007.
Zhou, M., Chen, J., Freguia, S., Rabaey, K., Keller, J., 2013. Carbon and electron fluxes
during the electricity driven 1,3-propanediol biosynthesis from glycerol. Environ. Sci.
Technol. 11199–11205.