Bioresource Technology 121 (2012) 454457

Contents lists available at SciVerse ScienceDirect

Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech

Short Communication

Methane production from glycolate excreting algae as a new concept in the production of biofuels
Anja Gnther a, Torsten Jakob a, Reimund Goss a, Swetlana Knig b, Daniel Spindler b, Norbert Rbiger c, Saskia John c, Susanne Heithoff c, Mark Fresewinkel d, Clemens Posten d, Christian Wilhelm a,
a

University of Leipzig, Institute of Biology, Department of Plant Physiology, Johannisallee 23, D-04103 Leipzig, Germany Saxon Institute of Applied Biotechnology, Permoserstr. 15, D-04318 Leipzig, Germany University of Bremen, Institute for Environmental Process Engineering, Leobener Str. 33, D-28359 Bremen, Germany d Karlsruhe Institute of Technology, Institute of Life Science Engineering, Bioprocess Engineering, Fritz-Haber-Weg 2, D-76131 Karlsruhe, Germany
b c

h i g h l i g h t s
" We show the feasibility of a photo-bioreactor concept to generate bio-methane. " The concept is based on the direct fermentation of algal exudates. " We show the metabolic engineering strategy to improve productivity. " We show a concept for the technical realization of the bio-reactor.

a r t i c l e

i n f o

a b s t r a c t
It is the aim of the present work to introduce a new concept for methane production by the interaction of a glycolate-excreting alga (Chlamydomonas reinhardtii) and methanogenic microbes operating in separate compartments within one photobioreactor. This approach requires a minimum number of metabolic steps to convert light energy to methane thereby reducing the energetic and nancial costs of biomass formation, harvest and renement. In this feasibility study it is shown that the physiological limitations for sustained glycolate production can be circumvented by the use of C. reinhardtii mutants whose carbon concentrating mechanisms or glycolate dehydrogenase are suppressed. The results also demonstrate that methanogenic microbes are able to thrive on glycolate as single carbon source for a long time period, delivering biogas composed of CO2/methane with only very minor contamination. 2012 Elsevier Ltd. All rights reserved.

Article history: Received 3 April 2012 Received in revised form 14 June 2012 Accepted 15 June 2012 Available online 9 July 2012 Keywords: Biofuel Biomethane Bioreactor Chlamydomonas

1. Introduction Microalgal biofuel technologies have been reported as a potential approach to displace fossil fuels in signicant amounts in the post-fossil carbon era (Stephens et al., 2010). Recent progress in reactor design (Posten, 2009) and genetically improved algal strains (Beer et al., 2009) raise the hope that the costs of algal production and product processing can be greatly improved. However, the major constraints of microalgal biomass-based biofuel production include the energetic and nancial costs of mixing, harvesting and renement to the nal product oil, ethanol or biogas (Whitaker et al., 2010). Therefore, strategies that bypass these production costs will have the greatest potential to increase the efciency from the absorbed photon to the nal product. One such strategy could be algal biohydrogen production. However, at present the
Corresponding author. Tel.: +49 341 9736874.
E-mail address: cwilhelm@rz.uni-leipzig.de (C. Wilhelm). 0960-8524/$ - see front matter 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.biortech.2012.06.120

major drawback of this technology is the discontinuity of its sub-processes and the low photosynthetic conversion efciency (Hallenbeck et al., 2012). In the present study, a new concept of algal based biogas production is introduced which aims to minimize the metabolic costs of the algal cells and the energetic costs of the entire process. This is achieved by the conversion of photosynthetic energy not into cellular biomass, but into glycolate. This metabolite is produced via the process of photorespiration and is easily excreted by algal cells. In a second step, glycolate is rened into methane by methanogenic bacteria under anaerobic conditions. To reduce the energetic costs of the process, a new reactor design is required and is described in detail here. Since in this new concept it is not intended to produce and harvest cellular biomass but to convert the cells into glycolate producing units, nutrients are needed only in minimal amounts for cellular maintenance. An elimination of nitrogen, sulfur and salts during renement is not required. In addition, metabolic costs

A. Gnther et al. / Bioresource Technology 121 (2012) 454457

455

(e.g. respiratory losses) which are associated with biomass production will be largely prevented. The feasibility of the concept is demonstrated by a set of experimental data. Furthermore, it is demonstrated that physiological limitations of short-term and long-term excretion of glycolate can be circumvented by the use of genetically modied algal cells, i.e. specic mutants of Chlamydomonas reinhardtii. 2. Materials and methods 2.1. Culture conditions of algae The green alga C. reinhardtii (SAG 11-32b, Culture Collection of Algae, Gttingen, Germany) was grown semi-continuously as airlift cultures in Kuhl-medium (Kuhl and Lorenzen, 1964) at a temperature of 20 C and an irradiance of 150 lmol photons m2 s1 (14:10 light/dark). The chlorophyll content of the cultures was adjusted to a concentration of 5 mg Chl a L1. Unless otherwise stated, the semi-continuously grown cultures were aerated with CO2-enriched air (3% CO2). The following strains of mutants of C. reinhardtii were obtained from the Chlamydomonas Research Center (Minnesota): CC-1860 (CCM-mutant) and CC-4160 (GDHmutant). 2.2. Experimental setup for glycolate production For the measurements of glycolate excretion, 50 mL of algal cell suspension were harvested. The cell suspension was aerated with ambient air or oxygen-enriched air and illuminated as described above. The following O2/CO2 mixtures were used throughout the experiments: 30/0.036, 40/0.03, 47/0.027, and 70/0.015 (%O2/ %CO2). At different time points, samples (4 mL) were collected from the reaction vessel. Isoniazid (10 mM) was added to inhibit the internal metabolization of glycolate. Algal cells were pelleted by centrifugation and 3 1 mL of the supernatant were sterileltered (0.2 lm; Roth, Karlsruhe, Germany) and freeze-dried. These samples were used for the quantitative determination of glycolate based on the colorimetric method with 2,7-dihydroxynaphthalene (Takahashi, 1972). 2.3. Methanogenesis and gas measurement The selection of glycolate-utilizing microorganisms was started from an anaerobic culture obtained from a biogas plant. The singlestage cultivation was run at anaerobic conditions at 37 C in the 5 L vessel MD 5 (B. Braun Biotech International GmbH, Melsungen, Germany) with 3.5 L working volume. Glycolic acid was added to the medium as the sole carbon source. Within the cultivation time of 600 days, the amount of this carbon source was increased from 1.12 to 2.35 mM. Gas volume production was measured by Milligascounter (Dr.-Ing. Ritter Apparatebau GmbH, Bochum, Germany). Oxygen, hydrogen, nitrogen, methane, carbon dioxide and hydrogen sulde concentrations were estimated by gas chromatography (Chrompack CP 9001). The ammonia content was analyzed with the ammonium cuvette test (LCK 303, Hach Lange GmbH, Duesseldorf, Germany). 2.4. Estimation of the expected area-based photo-production rate of methane By taking into account the maximum production rate of glycolate (P C2 H3 O ; lmol (mg Chl a)1 h1) obtained from the 3 experiments of this study, the annual area-based photoproduction rate of methane (rCH4 ; g ha1 a1) was estimated by:

rCH4

PC2 H3 O 3 Chl aMC2 H3 O 3 yCH4

Sept

t d2h
Apr

with [Chl a] being the maximum Chl a content (mg m2), the molar mass of glycolate (M C2 H3 O ; 75 g mol1), td 2 h (h) being 3 the daily astronomic duration of solar insolation (minus 2 h for sunrise and sunset), and the glycolate to methane fermentation 1 yield (yCH4 ; g CH4 g C2 H3 O ).
3

3. Results and discussion 3.1. The new concept of algal based biogas production The basic design of the photo-bioreactor places the photosynthesizing algal cells and the methanogenic bacteria in neighboring compartments (Supplemental Fig. 1). This design requires a separation by an interface providing a strong oxygen barrier between the algal and bacterial compartment (see below). In the photosynthetic compartment, algae are embedded in a solid matrix covered by a minimum volume of liquid culture medium to ensure the supply of nutrients and a specic CO2/O2-gas mixture. The produced glycolate is transported through the separation unit (see below) into the anaerobic compartment, where methanogenic bacteria convert it to carbon dioxide and methane. In comparison to conventional bioreactors, the design of this new reactor is expected to reduce the energetic costs signicantly. There is no demand for auxiliary energy input for (i) a complete mixing of the algal suspension or (ii) harvesting the algal biomass. It is also expected that (iii) only low pressure gradients are needed for the transport of glycolate from the photosynthetic to the methanogenic compartment. 3.2. Proof of concept 3.2.1. The photosynthetic compartment The green alga C. reinhardtii was used for the feasibility study because it provides the best algal experimental system for further genetic engineering. The results of the present study demonstrate that the glycolate production of C. reinhardtii cells can be controlled by the gas composition and that the excretion potential of the cells can be improved signicantly. Cultivation of C. reinhardtii under optimal growth conditions with a low O2/CO2-ratio (20/3%) did not lead to excretion of glycolate. A transfer of the cells back to ambient air conditions (21% O2, 0.04% CO2), however, resulted in signicant glycolate production at a rate of 9 lmol (mg Chl a)1 h1 (data not shown). This glycolate excretion was further enhanced to a rate of 13 lmol (mg Chl a)1 h1 by an increase of the O2/CO2-ratio to 30/0.03% and 40/0.024%, respectively. Two cellular processes have a strong inuence on the rate of glycolate excretion and provide important targets for the metabolic engineering of the algal cells: (i) the carbon concentrating mechanisms (CCMs), and (ii) the metabolization of glycolate. Numerous algal species possess CCMs to increase the internal CO2 concentration and to favor carbohydrate production via the carboxylating reaction of RubisCO (Kaplan and Reinhold, 1999). It is known that CCMs can be suppressed by a high CO2 concentration during the cultivation of algae. In line with these data, the results of the present study show that the increase of the CO2 concentration from 0.04% (ambient air) to 1% during cell cultivation induced stimulation of the oxygenating reaction of RubisCO and thus, a strong increase of the glycolate production (Fig. 1a). However, long-term experiments indicated a stagnation of glycolate excretion in WT cells of C. reinhardtii after 4 h under photorespiratory conditions (Fig. 1b). It is likely that the suppression of glycolate excretion was caused by the activation of CCMs. This means that sustained glycolate excretion is only possible when cells with

456

A. Gnther et al. / Bioresource Technology 121 (2012) 454457

a
Glycolate [mol (mg Chl a)-1 h-1]

18 15 12 9 6 3 0 control 10 mM NIA

in the presence of NIA (Fig. 1b). However, like in the WT cells, glycolate excretion of the GDH mutant started to stagnate after 45 h under photorespiratory conditions, again due to the induction of CCMs. The data from the present study clearly demonstrate that the metabolic limitations for a high and long-term stable excretion of glycolate can be circumvented by the use of respective mutants.

0.04 %

0.5 %

1%

2%

3%

CO2 concentration

b
Glycolate [mol (mg Chl a)-1]

90 75 60

+ NIA

3.2.2. The methanogenic compartment In the present study, methanogenic bacteria were isolated from a commercial biogas fermentation facility and fed with glycolic acid as the single source of carbon. After an adaptation phase, the gas composition changed from values of a conventional biogas plant to the optimum values of 40% methane and 60% CO2 (Supplemental Table 1), which are close to the CH4/CO2-ratio given by Friedrich et al. (1991). Hydrogen sulde and ammonia as contaminants could only be found in concentrations below 0.1%, and oxygen and nitrogen were detected as minor impurities. The fermentation process was observed to be stable for 600 days at a rate of 77 mL biogas per day and liter working volume and a methane yield of 0.173 m3 CH4 kg1 glycolate was calculated.

GDH mutant
45 30 15 0

CCM mutant WT
0 1 2 3 4 time [h] 5 6 7 8

Fig. 1. Rates of glycolate excretion of illuminated cells of C. reinhardtii under different experimental conditions. (a) Comparison of the rate of glycolate excretion during a 4 h incubation period at an O2/CO2 ratio of 40/0.02% in cells which were previously acclimated to different concentration levels of CO2. Data are shown as mean values (n = 3, SD) for control samples and samples in the presence of the inhibitor of glycolate metabolism (Isoniazid, NIA). (b) Representative examples of time-dependent changes in the amount of excreted glycolate in the comparison WT cells, the CCM mutant, and the GDH mutant (NIA) during incubation with a O2/CO2 ratio of 40/0.02%.

constitutively suppressed CCM activity are used. This assumption was proven by using the CCM mutant CC-1860 of C. reinhardtii. Fig. 1b illustrates that, in contrast to WT cells of C. reinhardtii, the excretion rate of glycolate was stable for at least 7 h. However, the suppression of CCMs in CC-1860 mutants did not change the maximum rate of glycolate excretion, which was dependent on the rate of glycolate metabolization. To evaluate the maximum potential of glycolate excretion, the inhibition of glycolate metabolization is required. In the present experiments, Isoniazid (NIA) was used to inhibit the C2-pathway (Nelson and Tolbert, 1970) and glycolate oxidation by the glycolate dehydrogenase (GDH; Colman et al., 1974). In the presence of NIA the glycolate excretion rates in WT cells of C. reinhardtii were enhanced by a factor of at least three (Fig. 1a). Prolonged incubation with NIA, however, exerted toxic effects on the algal cells, preventing the long-term application of this inhibitor. Thus, the use of transgenic cells with a suppressed expression of the GDH is the only alternative to achieve a high glycolate excretion rate. In the present study, the GDH mutant CC-4160 was tested for its glycolate excretion rates. Within 4.5 h of incubation under photorespiratory conditions, a comparable rate of glycolate excretion was observed for both the GDH mutant and the C. reinhardtii WT cells

3.2.3. Realization of technical aspects The conversion of the presented concept into a technical application requires the establishment of an algal biolm in the photosynthetic compartment. The aeration of this reactor compartment can be realized by a membrane integrated into the reactor-inside. This will reduce the energy consumption for aeration and provide greater contact areas for mass transport in comparison to bubble or surface aeration. The immobilization of algal cells in a biolm can be realized by different techniques, such as inclusion or adhesion/ adsorption of the cells on appropriate materials (Mallick, 2002). Vlchez et al. (1991) showed that long-term excretion of glycolate can be achieved with immobilized cells of C. reinhardtii, which is in line with our rst results with C. reinhardtii cells embedded in alginate (data not shown). With respect to the illumination of the algal cells, it could be necessary to nd a technical solution for an optimal vertical distribution of light in the biolm. To realize a dense package of algal cells in a biolm while still maintaining a high light absorption of the individual cells, the use of transgenic cells with a decreased absorptivity of light may be promising (Melis, 2009). Since the results from laboratory experiments may not be directly applicable to outdoor conditions (Scoma et al., 2012), natural aerophytic algae have also been investigated in the present study (data not shown). The fermentation of glycolate to methane requires anaerobic conditions in the methanogenic compartment. Thus, the transport of oxygen from the photosynthetic biolm into the bacterial compartment has to be restricted. Consequently, a separation unit (Supplemental Fig. 1) must fulll two functions: (i) the provision of a high transport rate of glycolate and (ii) the retention of oxygen. These two requirements can be realized by two different technical approaches. The rst approach combines a membrane, which allows the transfer of glycolate, with a gas stripping facility that removes oxygen from the glycolate containing medium. A selective diffusion of glycolate can be obtained by e.g. micro-ltration membranes (RM MV 020, Microdyn-Nadir GmbH, Wiesbaden, Germany). The rst experiments with such a membrane demonstrate the applicability of microltration membranes to allow the diffusion of glycolate between two separated compartments and a gas stripping unit (data not shown). A second approach could employ a specic semi-permeable membrane that allows a maximum passage of glycolate molecules while simultaneously providing the retention of oxygen.

A. Gnther et al. / Bioresource Technology 121 (2012) 454457

457

3.3. Estimation of the current and future potential of methane production The present study demonstrates the suitability of the principle processes and selected organisms to produce methane on the basis of glycolate excretion. Fig. 2 summarizes the process components and the key parameters in order to estimate a glycolate production rate per unit area. The initial process of photosynthesis requires light, water and CO2. The glycolate production rate (k1) depends on the presence of high concentrations of oxygen, the daily insolation hours, the amount of biomass per area (based on Chlorophyll or cell number), and the oxygenation rate of RubisCO. A reasonable assumption for the biomass concentration is a value of 300 mg Chlorophyll per m2, which corresponds to an absorption efciency of the photosynthetically available radiation (PAR) of 90%. Assuming a light saturation point of 100 lmol photons m2 s1 for the photorespiratory glycolate production, the daily light period, which is able to promote maximum glycolate production, can be derived from the duration of daily solar insolation (minus 2 h for sunrise and sunset). With the additional assumption that algal cells can be physiologically active at a minimum temperature of 10 C, this results in a vegetation period from April to September with a total of 2240 h of solar insolation. Based on the maximum glycolate excretion rates (Fig. 1), an annual, area-based glycolate production rate of 664 g m2 a1 is estimated. The rate of methane production (k2) by methanogenic bacteria basically depends on their fermentation yield (y). Using the experimentally measured glycolate production and fermentation yield, a maximum annual production rate of approximately 1100 m3 methane per hectare is estimated. With the presumption of very low energy costs during the operation of the proposed bioreactor (10% of gross energy production), an annual net energy gain of 40 GJ per hectare is calculated. It should be emphasized that the currently measured glycolate excretion rates are well below the expected physiological maximum value. Due to the fact that for

every molecule of excreted glycolate, two molecules of CO2 must be rexed, the maximum glycolate excretion rate can be as high as one-third of the total CO2 xation rate (appr. 230 lmol CO2 mg Chl a1 h1 according to Wagner et al., 2006). Thus, based upon optimized conditions for photorespiration and glycolate excretion, a maximum annual net energy gain of 240 GJ per hectare is estimated. This is about 2.6 times higher than the net energy of methane production based on maize. Work is under progess to increase the glycolate production under high light conditions which would strongly improve the annual net gain. 4. Conclusions The concept for biofuel production on the basis of glycolate fermentation has several synergistic advantages: (i) the bio-methane production via glycolate excretion can be much higher compared to biofermentation of biomass, because metabolic losses can be strongly reduced. (ii) The energetic costs for harvesting, extraction and renement can be decreased to a minimum. (iii) Nutrients are needed only in minimal amounts, because cells consume them only for cellular maintenance but not for growth. (iv) The biolm technology does not require heavy hardware. The ideal concept of a photo-bioreactor incorporates a thin biolm enclosed in a transparent robust foil. Acknowledgements The nancial support from the German Ministry of Science and Education (VIP16V0001) is greatly acknowledged. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.biortech.2012. 06.120. References
Beer, L.L., Boyd, E.S., Peters, J.W., Posewitz, M.C., 2009. Engineering algae for biohydrogen and biofuel production. Curr. Opin. Biotechnol. 20, 264271. Colman, B., Miller, A.G., Grodzinski, B., 1974. A study of the control of glycolate excretion in Chlorella. Plant Physiol. 53, 395397. Friedrich, M., Laderer, U., Schink, B., 1991. Fermentative degradation of glycolic acid by dened syntrophic cocultures. Arch. Microbiol. 156, 398404. Hallenbeck, P.C., Abo-Hashesh, M., Ghosh, D., 2012. Strategies for improving biological hydrogen production. Bioresour. Technol. 110, 19. Kaplan, A., Reinhold, L., 1999. CO2 concentrating mechanisms in photosynthetic microorganisms. Annu. Rev. Plant Physiol. 50, 539570. Kuhl, A., Lorenzen, H., 1964. Handling and culturing of Chlorella. In: Methods in Cell Physiology. Academic Press, New York and London, pp. 152187. Mallick, N., 2002. Biotechnological potential of immobilized algae for wastewater N, P and metal removal: a review. BioMetals 15, 377390. Melis, A., 2009. Solar energy conversion efciencies in photosynthesis: minimizing the chlorophyll antennae to maximize efciency. Plant Sci. 177, 272280. Nelson, E., Tolbert, N.E., 1970. Glycolate dehydrogenase in green algae. Arch. Biochem. Biophys. 141, 102110. Posten, C., 2009. Design principles of photo-bioreactors for cultivation of microalgae. Eng. Life Sci. 9, 165177. Scoma, A., Gianelli, L., Faraloni, C., Torzillo, G., 2012. Outdoor H2 production in a 50L tubular photobioreactor by means of a sulfur-deprived culture of the microalga Chlamydomonas reinhardtii. J. Biotechnol. 157, 620627. Stephens, E., Ross, I.L., King, Z., Mussgnug, J.H., Kruse, O., Posten, C., Borowitzka, M.A., Hankamer, B., 2010. An economic and technical evaluation of microalgal biofuels. Nat. Biotechnol. 28, 126128. Takahashi, K., 1972. A colorimetric method for quantitative determination of glycolic acid with 2,7-dihydroxynaphthalene. J. Biochem. 71, 563565. Vlchez, C., Galvn, F., Vega, J.M., 1991. Glycolate photoproduction by free and alginate-entrapped cells of Chlamydomonas reinhardtii. Appl. Microbiol. Biotechnol. 35, 716719. Wagner, H., Jakob, T., Wilhelm, C., 2006. Balancing the energy ow from captured light to biomass under uctuating light conditions. New Phytol. 169, 95108. Whitaker, J., Ludley, K.E., Rowe, R., Taylor, G., Howard, D.C., 2010. Sources of variability in greenhouse gas and energy balances for biofuel production: a systematic review. Bioenergy 2, 99112.

Fig. 2. Process diagram of the photosynthetic conversion of solar energy into glycolate and its subsequent fermentation to methane. The glycolate production rate (k1) results from the duration of daily solar insolation, the oxygenation rate of RubisCO (POx), and the biomass concentration (BM). The oxygenation rate will depend on the recycling of the primary acceptor of the Calvin cycle (ribulose 1,5bisphosphate) by the carboxylation reaction of RubisCO. The kinetics of methane production will result from k1, the fermentation yield of methane formation from glycolate (yCH4 ), and the volume (V; or respective amount of biomass concentration). It is assumed that the diffusion of glycolate from the photosynthetic into the methanogenic compartment is not a rate-limiting step. It is further assumed that CO2 produced by the fermentation of glycolate can be re-used in the photosynthetic reaction.