Improving Marine Algae for Bioenergy 287
synthesis of [FeFe]-hydrogenase. During this period, a low activity of PSII keeps electron flow from
water through the photosynthetic chain, and ultimately into ferredoxin and the hydrogenase. In addition,
limited oxygen availability reduces respiratory rates at the mitochondria, with glycolysis and anaerobic
fermentation becoming important pathways for ATP formation (Timmins et al. 2009b). On the other hand,
the mobilization of starch reserves generates reducing power in the form of NAD(P)H, feeding electrons
to plastoquinone through the enzyme Nda2 (Fig. 1A). Moreover, pyruvate derived from glycolysis
may also contribute to reduce ferredoxin through pyruvate:ferredoxin oxidoreductase. Both sources
of reducing power (water and starch) are important for hydrogen production under these conditions
(Hemschemeier et al. 2008; Chochois et al. 2009). Eventually, hydrogen production comes to an end,
as a result of the generalized cell damage induced by the extended sulfur shortage. Therefore, hydrogen
production through sulfur deprivation requires not only a previous lag-phase for anaerobic conditions to
be established, but also a regeneration phase in a non-deficient medium for cell recovery after nutrient
stress. Anyhow, this induction system may compromise cell viability for extended periods, and maximum
light-conversion efficiencies (less than 0.5%) are anyway below the values recommended for technical
and economic feasibility (i.e., above 10%) (Posewitz et al. 2008). Subsequent studies have evaluated
hydrogen production by C. reinhardtii subjected to different nutrient deprivation regimes, including
N-, P-, and Mg-deficient media (reviewed in González-Ballester et al. 2015). Despite hydrogen production
rates were comparable (in the same order of magnitude) to those achieved with S-deprived cultures, the
underlying physiological mechanisms differed, specially, in the relative contribution of PSII or in the
routes to feed electrons to ferredoxin. Alternative electron sources arise from starch mobilization (in
P- and Mg-deficient cultures), as previously indicated, or protein degradation (in N-deficient cultures).
Furthermore, the finding that pyruvate:ferredoxin oxidoreductase from C. reinhardtii was able to
transfer electrons from oxaloacetate to hydrogenase through ferredoxin (Fig. 1A) indicated that acetate
fermentation via the glyoxylate pathway may also be coupled to hydrogen production (Noth et al. 2013).
This route has been shown to operate in vivo with mixotrophic, non-stressed cultures of C. reinhardtii,
grown with acetate under low light intensity, which were able to produce hydrogen to similar yields
than S-deprived cultures (Jurado-Oller et al. 2015). In this case, acetate consumption rather than starch
mobilization was used as the source of reducing power by hydrogenase. Low light intensities were
required to restrict oxygen evolution through PSII. Interestingly, slight aeration that kept low oxygen
levels improved hydrogen production, by enhancing acetate uptake and releasing inhibitory hydrogen
partial pressure. Altogether, these results indicate that manipulation of growth conditions is a key aspect
to optimize hydrogen yields. On the other hand, genetic engineering approaches have also aimed at
increasing hydrogen production, mainly directed to alleviate the problem of the oxygen-sensitivity of
the [FeFe]-hydrogenase or to enhance the electron flow from ferredoxin into hydrogenase, as will be
discussed later.
In cyanobacteria, a two-stage strategy has been also addressed to sustain hydrogen production
(reviewed by Srirangan et al. 2011). In the case of heterocyst-forming cyanobacteria (Anabaena cylindrica,
Nostoc muscorum, Anabaena variabilis), nitrogen starvation is required to induce the differentiation of
vegetative cells into heterocysts. Therefore, cell growth under aerobic conditions in light is promoted
at a first stage; and hydrogen production is induced at a second stage, by imposing a metabolic stress
(nitrogen deficiency) under an argon atmosphere supplemented in some cases with CO 2 . The slower
turnover rate of nitrogenases (6.4 s
–1
) compared to bidirectional [NiFe]-hydrogenases (98 s
–1
), together
with the high energetic cost of the nitrogenase catalyzed reaction have encouraged the study of hydrogen
production in non-diazotrophic cyanobacteria (Synechoccoccus, Synechocystis, Gloebacter violaceus).
In this case, a temporal separation is established between oxygen-evolving photosynthesis (generating
energy and reducing power) and anoxic hydrogen evolution—as happens in green algae and in contrast
to heterocyst-forming cyanobacteria, where a physical separation exists between the two processes. The
two-stage operational mode consists of an initial phase in a nutrient-enriched medium in the light to
accumulate cell biomass, which is subsequently transferred to a nutrient-limited medium (deficient in
nitrogen or sulfur). In this second stage, anaerobic conditions are kept via argon bubbling, and glycogen
is further accumulated. As previously mentioned, hydrogen production by the bidirectional [NiFe]hydrogenase is promoted in the dark (which is opposed to the heterocysts system, requiring light for
synthesis of [FeFe]-hydrogenase. During this period, a low activity of PSII keeps electron flow from
water through the photosynthetic chain, and ultimately into ferredoxin and the hydrogenase. In addition,
limited oxygen availability reduces respiratory rates at the mitochondria, with glycolysis and anaerobic
fermentation becoming important pathways for ATP formation (Timmins et al. 2009b). On the other hand,
the mobilization of starch reserves generates reducing power in the form of NAD(P)H, feeding electrons
to plastoquinone through the enzyme Nda2 (Fig. 1A). Moreover, pyruvate derived from glycolysis
may also contribute to reduce ferredoxin through pyruvate:ferredoxin oxidoreductase. Both sources
of reducing power (water and starch) are important for hydrogen production under these conditions
(Hemschemeier et al. 2008; Chochois et al. 2009). Eventually, hydrogen production comes to an end,
as a result of the generalized cell damage induced by the extended sulfur shortage. Therefore, hydrogen
production through sulfur deprivation requires not only a previous lag-phase for anaerobic conditions to
be established, but also a regeneration phase in a non-deficient medium for cell recovery after nutrient
stress. Anyhow, this induction system may compromise cell viability for extended periods, and maximum
light-conversion efficiencies (less than 0.5%) are anyway below the values recommended for technical
and economic feasibility (i.e., above 10%) (Posewitz et al. 2008). Subsequent studies have evaluated
hydrogen production by C. reinhardtii subjected to different nutrient deprivation regimes, including
N-, P-, and Mg-deficient media (reviewed in González-Ballester et al. 2015). Despite hydrogen production
rates were comparable (in the same order of magnitude) to those achieved with S-deprived cultures, the
underlying physiological mechanisms differed, specially, in the relative contribution of PSII or in the
routes to feed electrons to ferredoxin. Alternative electron sources arise from starch mobilization (in
P- and Mg-deficient cultures), as previously indicated, or protein degradation (in N-deficient cultures).
Furthermore, the finding that pyruvate:ferredoxin oxidoreductase from C. reinhardtii was able to
transfer electrons from oxaloacetate to hydrogenase through ferredoxin (Fig. 1A) indicated that acetate
fermentation via the glyoxylate pathway may also be coupled to hydrogen production (Noth et al. 2013).
This route has been shown to operate in vivo with mixotrophic, non-stressed cultures of C. reinhardtii,
grown with acetate under low light intensity, which were able to produce hydrogen to similar yields
than S-deprived cultures (Jurado-Oller et al. 2015). In this case, acetate consumption rather than starch
mobilization was used as the source of reducing power by hydrogenase. Low light intensities were
required to restrict oxygen evolution through PSII. Interestingly, slight aeration that kept low oxygen
levels improved hydrogen production, by enhancing acetate uptake and releasing inhibitory hydrogen
partial pressure. Altogether, these results indicate that manipulation of growth conditions is a key aspect
to optimize hydrogen yields. On the other hand, genetic engineering approaches have also aimed at
increasing hydrogen production, mainly directed to alleviate the problem of the oxygen-sensitivity of
the [FeFe]-hydrogenase or to enhance the electron flow from ferredoxin into hydrogenase, as will be
discussed later.
In cyanobacteria, a two-stage strategy has been also addressed to sustain hydrogen production
(reviewed by Srirangan et al. 2011). In the case of heterocyst-forming cyanobacteria (Anabaena cylindrica,
Nostoc muscorum, Anabaena variabilis), nitrogen starvation is required to induce the differentiation of
vegetative cells into heterocysts. Therefore, cell growth under aerobic conditions in light is promoted
at a first stage; and hydrogen production is induced at a second stage, by imposing a metabolic stress
(nitrogen deficiency) under an argon atmosphere supplemented in some cases with CO 2 . The slower
turnover rate of nitrogenases (6.4 s
–1
) compared to bidirectional [NiFe]-hydrogenases (98 s
–1
), together
with the high energetic cost of the nitrogenase catalyzed reaction have encouraged the study of hydrogen
production in non-diazotrophic cyanobacteria (Synechoccoccus, Synechocystis, Gloebacter violaceus).
In this case, a temporal separation is established between oxygen-evolving photosynthesis (generating
energy and reducing power) and anoxic hydrogen evolution—as happens in green algae and in contrast
to heterocyst-forming cyanobacteria, where a physical separation exists between the two processes. The
two-stage operational mode consists of an initial phase in a nutrient-enriched medium in the light to
accumulate cell biomass, which is subsequently transferred to a nutrient-limited medium (deficient in
nitrogen or sulfur). In this second stage, anaerobic conditions are kept via argon bubbling, and glycogen
is further accumulated. As previously mentioned, hydrogen production by the bidirectional [NiFe]hydrogenase is promoted in the dark (which is opposed to the heterocysts system, requiring light for