View Online / Journal Homepage / Table of Contents for this issue Energy & Dynamic Article LinksC< Environmental Science

Cite this: Energy Environ. Sci., 2012, 5, 5531 www.rsc.org/ees REVIEW -to-fuels: from sunlight to , isoprene, and botryococcene production

Anastasios Melis*

Received 31st August 2011, Accepted 21st November 2011 DOI: 10.1039/c1ee02514g

The concept of ‘‘Photosynthetic Biofuels’’ entails the direct application of photosynthesis for the generation of fuels and chemicals, in a process where a single organism acts both as catalyst and processor, synthesizing and secreting ready to use product. Examples of photosynthetic fuel and chemicals generation are offered in this perspective.1 Physiological and genetic manipulation of green microalgae enabled diverting the natural flow of photosynthetic electron transport toward sustained generation of hydrogen gas, instead of the normally produced oxygen.2 Heterologous expression of a plant isoprene synthase gene in cyanobacteria and microalgae enabled the renewable generation of

volatile isoprene (C5H8) hydrocarbons, derived from sunlight, (CO2) and water (H2O). Photobioreactor design concepts and conditions for capturing the isoprene are presented. The process of

generating isoprene (C5H8) hydrocarbons serves as a case study in the development of technologies for the renewable generation of a multitude of other fuels and useful chemicals.3 Green microalgae of the genus Botryococcus afford naturally occurring examples on how to divert the flow of photosynthetic metabolites toward high-value long-chain hydrocarbon products instead of the normally produced sugars. Members of the genus Botryococcus direct photosynthate toward tri-terpenoid botryococcene hydrocarbons or diene and triene alkanes, at the expense of investing photosynthate toward biomass accumulation.4 Lastly, methods are offered on how to improve the solar-to-biomass energy conversion Downloaded on 20 October 2012 efficiency of photosynthesis in high cell-density cultures, or canopies, under bright sunlight conditions. Goal of this effort is to improve efficiency, from the current-best of 2–3% up to a theoretical maximum of about 10%. Advances in each of the above-mentioned fields suggest blueprints of organism genetic and

Published on 12 December 2011 http://pubs.rsc.org | doi:10.1039/C1EE02514G metabolic engineering, by which to improve the performance of photosynthesis and to divert cellular metabolic flux toward alternative high-value bio-products. Successful implementation of this R&D would lead to the generation of commercially viable fuels, chemicals, and other useful bio-products.

University of California, Department of Plant & Microbial Biology, Berkeley, CA, 94720-3102, USA. E-mail: [email protected]; Fax: +1 510 642 4995

Broader context The quest for fuels and chemicals to serve human industrial consumption can be met by the emerging science of synthetic biology. Processing of biomass to fuels and chemicals has attracted considerable attention in the recent past. An alternative approach, and a shortcut in this endeavor, is the direct application of photosynthesis, where absorption and utilization of sunlight, in combination with existing or endowed biosynthetic pathways in the cell, result in the generation of designer fuels and chemicals. Such a direct Photosynthetic Biofuels approach bypasses intermediate steps associated with the multistep processes of biomass generation, harvesting, deconstruction and fermentation, and promises greater efficiencies and lower costs. The premise of a ‘‘Photosynthesis-to- Fuels’’ approach, therefore, is the ability to transform the primary products of photosynthesis directly, in a single organism, into

useful fuels and chemicals for human consumption. Examples of the current state of the art include the generation of hydrogen (H2), isoprene (C5H8), and longer chain fatty acids and terpenoid hydrocarbons. Such fuels and chemicals, generated renewably and without environmental degradation, offer the promise of closing the gradually widening ‘‘energy gap’’ between the global supply and demand, mitigate the release of greenhouse gases in the , alleviate pollution, and improve the economies of all nations on earth.

This journal is ª The Royal Society of Chemistry 2012 Energy Environ. Sci., 2012, 5, 5531–5539 | 5531 View Online

Prologue The premise of photosynthesis for the direct generation of fuels (Photosynthetic Biofuels) is that a single organism can serve both as a photo-catalyst and a producer of ready-made fuel. This

concept is exemplified in the schematic of Fig. 1, where H2O, sunlight and CO2 are inputs and O2, biomass, fuels (H2, hydro- carbons) and chemicals are outputs. In this model, conversion of

solar-to-chemical energy, and the biosynthesis of H2, hydrocar- bons, or other chemicals take place within a single cell, involving the photosynthetic apparatus and possibly the adjacent cellular metabolism. Commercially viable generation of photosynthetic biofuels Fig. 2 Microscopic observation of a live Chlamydomonas reinhardtii cell. requires employment of organisms that possess a robust photo- (Left panel) Differential interference contrast microscopy revealing synthetic apparatus and otherwise place limited demands on the peripherally dark grey areas of higher matter density originating from the photosynthetic output. Photosynthetic microorganisms, such as thylakoid membranes of photosynthesis. (Right panel) Viewing the same unicellular microalgae and cyanobacteria, fit these criteria. cell by chlorophyll fluorescence, revealing the presence and localization Desirable features of candidate microalgae and cyanobacteria of photosynthetic pigments and, by extension, the shape of the chloro- include high rates of photosynthesis, easiness of cultivation, plast. The chloroplast occupied 65% of the cellular volume. Scale of the m acclimation to the prevailing environmental and climatic condi- horizontal bar is 10 m. tions, and amenability to genetic transformation so as to direct photosynthetic and metabolic activity toward the desirable fluorescence microscopy (Fig. 2, right panel). This fully auton- product generation. Some of the critical features of these omous microorganism is about 10 mm in length. Viewed against microorganisms are illustrated in Fig. 2, where a single cell of the the shape and size of the cell (DIC) the red chlorophyll fluores- green microalga Chlamydomonas reinhardtii is viewed by differ- cence denotes the presence and distribution of chlorophyll, i.e., ential interference contrast (Fig. 2, left panel) and chlorophyll defines the boundaries of the photosynthetic apparatus, revealing a three-dimensional U-shaped chloroplast that occupies 65% of the cell volume. These cells can rightfully be viewed as factories of photosynthesis, especially so as the chloroplast to cytosol volume ratio is 2 : 1 or greater. In a photosynthetic biofuels effort, the microalgal chloroplast can be directed to generate product for commercial exploitation. An advantage of the single- Downloaded on 20 October 2012 Fig. 1 Schematic depicting the concept of ‘‘Photosynthetic Biofuels’’, cell configuration of such microalgae and cyanobacteria is that via where a single organism converts, the process of oxygenic photo- products of photosynthesis do not need to be shared for the synthesis, H O and CO into biomass and O . Alternatively, photosyn- 2 2 2 purpose of plant tissue formation, as the case would be for the thate can be directed toward the generation of fuels and chemicals.

Published on 12 December 2011 http://pubs.rsc.org | doi:10.1039/C1EE02514G Oxygen is a by-product of photosynthesis. The issue of photosynthetic formation of roots, stems, branches and leaves in plants. Thus, carbon partitioning between biomass and fuels/chemicals is further dis- a greater proportion of the photosynthetic products in micro- cussed in this work. algae and cyanobacteria can be immediately directed toward fuel or chemical generation.

H production Anastasios Melis received his 2 BSc in Biology at the University Uniquely among organisms of oxygenic photosynthesis, many of Athens, Greece, and PhD in but not all green microalgae encode for genes of hydrogen molecular biophysics from the metabolism. Included are two HydA [Fe–Fe] hydrogenases,1,2 as Florida State University, Talla- well as genes encoding proteins that are required for the [Fe–Fe] hassee. He received postdoctoral hydrogenase assembly.3 Hydrogen metabolism-related proteins training under the auspices of in green microalgae are localized and function in the chloroplast, the European Molecular such that the [Fe–Fe] hydrogenase can receive high potential- Biology Organization at the energy electrons directly from reduced ferredoxin (FD) at the end Rijksuniversiteit Leiden, The of the photosynthetic electron transport chain (Fig. 3). Since the

Netherlands, and under the discovery of the green microalgal H2 metabolism by Hans Gaf- auspices of the Carnegie Insti- fron and co-workers in the early 1940s,4,5 it escaped no-one’s tution for Science at Stanford attention that green microalgae can serve as the photosynthetic

Anastasios Melis University. He is currently a full producers of H2, essentially derived from sunlight and H2O. A Professor at the University of measure of the promise of green microalgal hydrogen production California, Berkeley, and asso- is offered upon consideration of the presence of approximately ciated Faculty Biologist at the Lawrence Berkeley National 3 000 000 photosynthetic electron transport chains per green Laboratory. algal cell,6–8 each capable of transporting 100 electrons per

5532 | Energy Environ. Sci., 2012, 5, 5531–5539 This journal is ª The Royal Society of Chemistry 2012 View Online

Fig. 4 Coordinated photosynthetic and respiratory electron transport Fig. 3 Linked H2O oxidation and H2 production in the photosynthetic apparatus of green microalgae. Electrons originate from PS II upon the that leads to anoxia (absence of oxygen) and H2 production in green microalgae. In a sealed culture, oxygen generated by the process of photooxidation of H2O molecules. They are transported via PS II (P680), the intermediate electron transport chain (PQ, Cyt b–f complex, PC), and photosynthesis is efficiently utilized by the cell’s own mitochondria, enabling cellular respiration. Electrons for the latter are derived upon PS I (P700) to ferredoxin (FD). In normal photosynthesis, high potential- endogenous substrate catabolism (starch), which yields reductant energy electrons from reduced ferredoxin are transferred to NADP+ to generate NADPH. The latter serves as the main reductant in support of (NADH) and CO2. Release of molecular H2 by the chloroplast sustains cellular metabolic reactions. ATP is also generated during the process of the function of this coordinated photosynthesis–respiration process. Internally generated anoxia is necessary and sufficient to permit induc- electron transport in the photosynthetic apparatus. In H2 production, the HydA-[Fe–Fe] hydrogenase receives high potential-energy electrons from tion of the H2 production metabolism in green microalgae.

reduced ferredoxin to generate molecular H2. Schematic adapted from Melis and Happe (2001). platform of green algal H2 production in the field, and is currently employed by many labs in several countries, as a vehicle second. Operating at this capacity, a 1 L culture containing by which to further explore the properties and premise of green 10 106 cells per mL could produce 200 mL hydrogen per hour! microalgal H production.14–16 In practice, anoxic conditions are a strict requirement for the 2 The process of H production, depicted in the mechanism of expression and activity of the H production machinery in the 2 2 Fig. 4, depends on the availability of starch or endogenous green microalgal cell. Oxygen, produced at the H O-oxidation 2 substrate to help sustain cellular respiration for the consumption site of the photosynthetic apparatus (Fig. 3), is a potent inhibitor of photosynthetic oxygen. In wild type microalgae, starch of the [Fe–Fe] hydrogenase and a positive suppressor of reserves can suffice to sustain hydrogen production for about 4–5 H -related gene expression, blocking the transcription of all 2 days. When starch reserves are consumed, cells need to go back

Downloaded on 20 October 2012 genes associated with hydrogen metabolism. This may be seen as to normal photosynthesis, where biomass accumulation and O nature’s provision of a powerful and effective mechanism that 2 evolution would take place. The latter is necessary and sufficient prevents the co-production of hydrogen and oxygen from the to replenish endogenous substrate and otherwise to rejuvenate photosynthetic apparatus. Thus, upon turning ‘‘on’’ illumination

Published on 12 December 2011 http://pubs.rsc.org | doi:10.1039/C1EE02514G the microalgae, so that the stage of H production can be of a dark-adapted anoxic green microalgal culture, hydrogen 2 repeated. Experimental results from such cycling of the ‘‘stages’’ production has been observed to last for as long as 90 seconds, are shown in Fig. 5, where alternating O2 and H2 production before the oxygen produced by the photosynthetic apparatus 12 9 could be sustained ad infinitum. Furthermore, the critical role of fully inhibits hydrogen production. endogenous substrate in maintaining anoxia in the cells was The conundrum of the O2 problem in H2 production could not be solved in 70 years of related research.10 In 2000, however, an experimental approach was successfully designed and applied to

bypass the O2 problem. Continuous photosynthetic hydrogen gas production was sustained for several days, achieved upon a regulated slow down of oxygen evolution in the green algae Chlamydomonas reinhardtii.11,12 This breakthrough successfully employed the cell’s own respiration to consume the photosyn- thetically generated oxygen,2,12 in a process where internal starch reserves were used to sustain the cells’ respiration.13 A schematic depicting the coupling of the cellular chloroplast photosynthesis with mitochondrial respiration (Fig. 4) explains how anoxic conditions can be maintained in the cell permitting expression of the HydA hydrogenase, and enabling sustained hydrogen metabolism in the chloroplast. Initially, balancing of photosyn- thesis and respiration was achieved upon sulfur-deprivation of the algae,11 a condition that lowered the level of photosynthesis Fig. 5 Cycling of a green microalgal culture between the stages of H2 to just below that of respiration, resulting in an anoxic envi- production and normal photosynthesis (Normal P). Cells deplete starch

ronment that supported hydrogen production. Maintenance of reserves during H2 production, but recover and replenish endogenous anoxia by the cell’s own respiration has since become the substrate during normal photosynthesis.

This journal is ª The Royal Society of Chemistry 2012 Energy Environ. Sci., 2012, 5, 5531–5539 | 5533 View Online

demonstrated with mutants of Chlamydomonas reinhardtii that Isoprene production over-accumulated starch. These were able to sustain H2 production for about twice as long, and reach yields about twice Microalgal H2 is the direct product of the light reactions of as high, compared to those measured with wild type strains.17 photosynthesis (Fig. 3). To bypass the H2 storage problems, an In the laboratory, sequestration and quantification of alternative approach would be to enable and harvest biofuel products from the carbon reactions of photosynthesis. Of hydrogen can be achieved upon collection of H2 gas in upside- down graduated cylinders or burettes by the method of water particular interest is the process of generating and accumulating displacement (Fig. 6). The method of hydrogen storage by the hydrocarbons via the fatty acid or terpenoid biosynthetic path- 32–34 displacement of water in glass containers satisfies the require- ways. Hydrocarbons can be viewed as a biological way of storing hydrogen (Fig. 7). Of the great variety of such molecules, ment of easy H2 sequestration and a subsequent easy retrieval for use in the laboratory. However, the method is not practical for isoprene (C5H8) is the smallest hydrocarbon that can be gener- large-scale commercial exploitation, where substantial amounts ated by photosynthetic systems (Fig. 8). In fact, isoprene is 35 of hydrogen must be reversibly stored-and-retrieved without naturally produced by many herbaceous and deciduous plants, a significant energetic expenditure.18 To date, there are no simple derived from the early steps of the DXP-MEP terpenoid biosynthetic pathway of chloroplasts.36,37 However, microalgae storage alternatives, especially when considering H2 as a fuel for the transportation sector. A main barrier is the requirement of and aquatic plants are not endowed with the isoprene synthase high capacity storage and on-demand retrieval, in a reversible gene and enzyme. process where the energetic requirements of storing-and- Table 1 presents a comparison of volatile hydrocarbons that retrieving are low. Different approaches have been investigated, are currently in use in industrial, automotive, and synthetic including hydrogen liquefaction,19,20 compression up to 5000 chemistry industries, but also used for residential and commer- psi,21,22 storage in metal hydrides,23–27 and adsorption (phys- cial heating. Of those listed (methane, propane, butane, isoprene isorption) in porous materials, notably carbon nanotubes.28,29 and pentane), only isoprene can be presently generated via the Current problems associated with these approaches include process of oxygenic photosynthesis. Isoprene (C5H8) has a combination of low capacity, high cost, high energetic a hydrogen storage capacity of 13.3% (H/C w/w), far surpassing the so-far untenable target of 6% w/w set for H2 storage in requirement, and safety. Alternative methods of storing H2 in N2 carbon nanotubes and similar devices. Moreover, the boiling (conversion to NH3), CO2 (conversion to CH4 or CH3OH) have also been proposed. Energetic and economic feasibility of the point of isoprene (34 C) makes the product easy to volatilize or latter has not been established as yet. Difficulties in hydrogen condense to a liquid form, as needed. These physicochemical storage are an impediment in transportation, distribution, and properties would greatly facilitate storage, transportation and on-board storage, all of which raise questions as to the present- distribution of a product that can be used by itself as a fuel and day practicality of renewable hydrogen in industrial and auto- feedstock in the synthetic chemistry industry (rubber) or, upon transformation, converted into hydrogen for stationary and on-

Downloaded on 20 October 2012 motive applications. board fuel cell applications. Regardless of difficulties in H2 storage, techno-economic 30 Isoprene naturally abounds in certain ecotypes, such as the boundary analysis of biological H2 production has shown promise for the generation of this biofuel product at costs Blue Mountain, Australia, and the Blue Ridge Mountain in the Published on 12 December 2011 http://pubs.rsc.org | doi:10.1039/C1EE02514G approaching US$ 2 per kg, i.e., US$ 2 per gallon gasoline equivalent (the energy content equivalence between these prod-

ucts is as follows: 1 kg H2 ¼1 gallon gasoline ¼40 kWh). Future R&D improvements hold promise in terms of further 31 lowering the cost of H2 production to less than US$ 2 per kg.

Fig. 7 Photosynthetic product generation from the light and carbon reactions of photosynthesis. Shown are the pathways of hydrogen production as well as those of carbon partitioning leading to the gener- ation of sugar (1), terpenoid (2) and fatty acid (3). In the vast majority of photosynthetic systems, carbon partitioning is primarily directed toward sugars (80–85%), with fatty acid (10%) and terpenoid biosynthesis

Fig. 6 Light-driven green microalgal H2 production, sequestration, and (5%) lagging far behind. Note that products such as isoprene (C5H8) quantification measurements conducted in the laboratory. Hydrogen gas and botryococcene (Btc-oil; C30H48) emanate, respectively, from the early was removed from the sealed cultures by suitable tubing and collected in and late steps of the isoprenoid biosynthetic pathway. (Adapted from upside-down graduated cylinders by the method of water displacement. Lindberg et al., 2010.39)

5534 | Energy Environ. Sci., 2012, 5, 5531–5539 This journal is ª The Royal Society of Chemistry 2012 View Online

Sequestration and trapping of a volatile product require growth of the photosynthetic microorganism in a closed biore- actor system, so as to prevent escape of the volatile isoprene. A tubular horizontal photobioreactor system is offered as an illustrative example (Fig. 9 and 10) for this purpose. Tubular photobioreactors are examples of a structure suitable for microalgae or cyanobacteria cultivation, designed to support cell growth and isoprene production. The tube structure can be made from a variety of materials ranging from polyethylene to plex- iglass and glass. Tube diameter and length can vary, e.g., depending on the size of the bioreactor and the optical properties of the cells that define sunlight penetration (see below). For example, the diameter may vary from about 5–30 cm and the length can vary from a few to tens or hundreds of metres. On the basis of the modular bioreactor schematic shown in Fig. 9, a tube Fig. 8 Chemical structure and molecular configuration of the isoprene length of 50 m and diameter of 15 cm would translate into 884 L unit. total capacity reactor volume. If the tube diameter of the bioreactor is 30 cm, then the total internal volume of the reactor would be 3534 L. It is understood that far greater capacity Table 1 Physicochemical properties of selected volatile hydrocarbons. reactor volumes can be attained with 5–30 cm diameter reactors The hydrogen storing capacity of the molecule was calculated as % H/C upon increasing the overall length of the tubes. (w/w) Specifically for isoprene production, an experimental approach developed was to fill the photobioreactor with the Chemical H/C ratio Molecular Boiling Hydrocarbon formula (% w/w) mass/g mol1 point/C microalgae or cyanobacteria-containing growth media to 50% volume (Fig. 10, aqueous phase). The remainder of the reactor Methane CH4 33.3 16 161 space (headspace) was then filled with 100% CO to support the 2 Propane C3H8 22.2 44 42 growth of the photosynthetic microorganisms (Fig. 10, gaseous Butane C4H10 20.8 58 0.5 Isoprene C5H8 13.3 68 34 Pentane C5H12 20.0 72 36

Downloaded on 20 October 2012 eastern US, where the photosynthesis of the trees naturally produces it.38 Isoprene is a small hydrophobic and volatile molecule, passes easily through the chloroplast and cellular membranes and is released into the atmosphere through the Published on 12 December 2011 http://pubs.rsc.org | doi:10.1039/C1EE02514G stomata of leaves. The undesirable effects of isoprene as an Fig. 9 Top view of the tubular modular photo-bioreactor layout. atmospheric pollutant are well discussed in the literature.38 Given Tubular modular photobioreactors for microalgae or cyanobacteria are the volatile nature of isoprene (boiling point 34 C) and the designed to support cell growth and hydrogen or isoprene production. extended canopy of herbaceous, deciduous, and conifer plants, it Tube diameter can be e.g. 5–30 cm; length can vary e.g. from a few to tens is impractical to attempt to harvest this hydrocarbon from plant or hundreds of metres. The control box enables culture manipulations ecosystems. However, microorganisms endowed with the including addition/removal of growth media and nutrient supply. isoprene synthesis property are amenable to cultivation in fully enclosed bioreactors that offer advantages in product contain- ment and sequestration. Heterologous over-expression of a plant isoprene synthase (IspS) gene in a photosynthetic microorganism can effectively yield isoprene production from the cell’s own metabolism. This was reported for the first time with the cyanobacterium Synechocystis sp. PCC 6803 upon the heterol- ogous transformation with the kudzu (Pueraria montana) IspS gene.39,40 The use of a photosynthetic microorganism, such as cyanobacteria or microalgae, to synthesize isoprene from sunlight, carbon dioxide, and water is an example of a shortcut in synthetic biology, as it eliminates other steps including growth of Fig. 10 Cross-sectional view of a photo-bioreactor tube, showing the gaseous/aqueous two-phase partition configuration. The schematic the biomass, harvesting, dewatering, deconstruction and shown depicts a 50 : 50 partition between the gaseous and aqueous fermentation of the sugars therein in order to arrive at the phases. However, the gaseous/aqueous partition ratio could vary desirable product. Isoprene affords other advantages over larger- depending on organism and growth conditions. In a sealed reactor, this

size molecules, such as fatty acids, since it is volatile, and is configuration permits spontaneous diffusion-driven CO2 uptake from the naturally emitted by the cells, eliminating the need for costly gaseous phase, and its replacement by the O2 and isoprene (Isp) products product extraction processes. of photosynthesis from the aqueous phase.

This journal is ª The Royal Society of Chemistry 2012 Energy Environ. Sci., 2012, 5, 5531–5539 | 5535 View Online

phase). This concept and approach was recently developed and systems, such as E. coli, a common approach to addressing this applied by Bentley and Melis.40 Fig. 10 depicts a cross-sectional problem entailed precursor flux amplification, coupled with view of such a photobioreactor tube, showing the gaseous/ enhancement in the accumulation of the terminal step.41–43 Such aqueous two-phase partition of the reactor interior space. The improvements, however, have not yet been attempted with schematic depicts a 50 : 50 (v/v) partition between the gaseous photosynthetic systems. and aqueous phases. However, the partition ratio can vary The fuel-to-biomass carbon-partitioning ratio impacts the depending on the growth conditions and activity of the photo- product yield and process economics. For example, commercial synthetic microorganisms. Also depicted in Fig. 10 is the diffu- viability of the photosynthetic isoprene production process was

sion-driven uptake of CO2 from the gaseous headspace and its assessed in techno-economic boundary analysis on the basis of exchange with O2 and isoprene from the aqueous phase. As currently achieved microalgal/cyanobacterial productivity (25 g discussed by Bentley and Melis,40 the gaseous/aqueous method dry weight m2 d1). From the business model point-of-view, it avoids continuous bubbling of the culture (expensive) and relies was determined that a break-even point would be reached when only on periodic product removal from the headspace and re- a minimum of 20% of photosynthate can be directed toward 2 1 filling of the gaseous phase with CO2. The headspace of the isoprene, e.g. generating about 5 g isoprene m d , while photobioreactor is otherwise maintained in a sealed and resting limiting biomass to about 20 g dry weight m2 d1. Of course,

state, and the process relies on a diffusion-driven uptake of CO2 a greater than 20% isoprene-to-biomass carbon partitioning from the gaseous phase and its exchange for O2 and isoprene. ratio, when achieved, would favorably impact the overall Continuous operation of the gaseous/aqueous two-phase biore- economics of the process. actor was achieved upon periodic removal (harvesting) of isoprene from the headspace and replacement with fresh CO2 for Hydrocarbons synthesis, accumulation and secretion further photosynthesis and production. The photosynthetic in the colonial green microalgae Botryococcus braunii microorganisms in the aqueous phase were also maintained in a state of continuous growth by periodically removing a portion In photosynthetic systems thus far, there are no well-established of the culture and replacing with an equal volume of fresh growth experimental or genetic engineering approaches by which to medium.40 constitutively direct large proportions of endogenous carbon to

Passing the gaseous phase that contains the isoprene and O2 terpenoid biofuel synthesis. Similarly, constitutive alkane through a cooled condenser and/or passing the condensed production in microalgae and cyanobacteria has not occurred at

isoprene through a solvent efficiently separates the O2 by- high yields, relative to biomass accumulation, in spite of having product and enables harvesting of the volatile isoprene hydro- been pursued vigorously by the private sector. carbon. Alcohols such as methanol (boiling point 64.7 C), However, there are naturally occurring microalgal systems in ethanol (boiling point 78.0 C), butanol (boiling point 117.7 C), which photosynthetic carbon partitioning is substantially and

Downloaded on 20 October 2012 and pure hydrocarbons such as hexane (boiling point 69.0 C), constitutively altered toward either terpenoid or fatty acid heptane (boiling point 98.4 C), octane (boiling point 125.5 C), (alkane) biosynthesis. The colonial green microalgae Botryo- and dodecane (boiling point 216.2 C) readily form stable coccus braunii B-race constitutively produce, accumulate and ‘‘blends’’ with isoprene, thereby facilitating its retention and secrete about 35% of photosynthetic carbon in the form of tri- Published on 12 December 2011 http://pubs.rsc.org | doi:10.1039/C1EE02514G stabilization in a liquid solution. Concentrations of isoprene (or terpenoid oils (botryococcenes), a process that occurs naturally other volatile hydrocarbon) collected in this manner can vary in these microalgae.44,45 Thus, in comparison to other photo- between 1%, when a solvent is used to trap the isoprene, and synthetic systems, where carbon partitioning among the sugar/ 100% isoprene when no solvent is used, but a condenser- fatty acid/terpenoid pathways is about 85 : 10 : 5 w/w/w (Fig. 7), temperature low enough is employed to enable retention of Botryococcus braunii B-race showed sugar/fatty acid/terpenoid isoprene in the liquid phase. 50 : 10 : 40 w/w/w C-partitioning ratios.46,47 Currently on-going research aims to elucidate the regulation of this unique photo- The fuel-to-biomass carbon-partitioning ratio synthetic carbon-partitioning phenomenon in B. braunii B-race, so as to enable application of this property to other photosyn- Whereas the photosynthesis of microalgae and cyanobacteria thetic biofuel production systems. can be directed toward terpenoid or fatty acid accumulation Similarly, alkane production relative to biomass has not (Fig. 7), yields are limited due to a preference by the organism to occurred at high yields in current photosynthetic production direct photosynthetic carbon toward sugar synthesis. A flux systems. However, the colonial green microalgae Botryococcus analysis conducted in this laboratory39 suggested that more than braunii A-race constitutively produce, accumulate and secrete 80% of photosynthetic carbon is directed toward sugars, which >20% of the overall biomass equivalent in the form of diene and eventually are converted into protein, cell wall, nucleic acid and triene alkanes.47–49 In these systems, sugar/fatty acid/terpenoid polysaccharide (components of biomass). Under normal and 65 : 30 : 5 w/w/w C-partitioning ratios were reported.47 These physiological growth conditions, the fatty acid and terpenoid naturally altered C-partitioning ratios afford concept feasibility biosynthetic pathways consume only about 10% and 5% of in efforts to alter photosynthetic carbon flux and to direct it more photosynthetic carbon, respectively. It stands to reason that toward hydrocarbon rather than sugar biosynthesis. a commercial biofuels and chemicals production process would An interesting example of terpenoid oil accumulation is require a substantially different carbon partitioning in the chlo- afforded by Botryococcus braunii var Showa (the Berkeley roplast and cell, such that bio-product accumulation is enhanced strain). These colonial green microalgae constitutively synthesize at the expense of sugar and/or biomass yield. In model bacterial and secrete up to 35% of photosynthetic carbon in the form of

5536 | Energy Environ. Sci., 2012, 5, 5531–5539 This journal is ª The Royal Society of Chemistry 2012 View Online

saturation curve of photosynthesis,55–59 showing high efficiency of photosynthesis at low irradiances but over-absorption of sunlight, saturation of photosynthesis, and wasteful dissipation of excess energy at high irradiances. Photosynthesis in most species saturates at 20% incident solar irradiance, or less, meaning that 80%, or more, of the absorbed bright sunlight would be wastefully dissipated. The problem of the early light-saturation of photosynthesis impacts all photosynthetic organisms. It is most severe in green and purple photosynthetic bacteria, as these tend to have sizable, possibly in the thousands, light-harvesting antenna pigments.60 Among oxygenic photosynthetic organisms, cyanobacteria and red algae are especially subject to this early light-saturation effect Fig. 11 Microscopic observation of a pressed (flattened) Botryococcus as, in addition to their chlorophyll pigment beds, they possess braunii var. Showa (the Berkeley strain) micro-colony. Pressing of the phycobilisomes, containing hundreds of light-harvesting antenna 61 micro-colony permitted visualization of individual green cells delimited pigments. Green and brown microalgae, and C3 plants have 62,63 by the circle of the entity and by droplets of botryococcene (C30H48) comparatively smaller chlorophyll antenna sizes, averaging hydrocarbons exuding from the micro-colony into the growth medium. 600 Chl molecules per PSII and PSI, i.e., per linear electron- Also shown is the chemical structure of the C30H48 botryococcene transport chain in the photosynthetic apparatus. However, deep- 46 molecule. (Adapted from Eroglu and Melis, 2010. ) water microalgae and terrestrial obligate shade species possess a larger than 600 chlorophyll antenna size.63 botryococcene, a tri-terpenoid hydrocarbon, which is a squalene analog (Fig. 11). Unfortunately, microalgae of the genus Botryococcus braunii B-race cannot be directly applied for the commercial generation of botryococcene oil due to their slow growth. Recent effort in this field50 succeeded in cloning three novel genes, termed ‘‘squalene synthase like’’ SSL1, SSL2, and SSL3, so as to distinguish them from the known squalene syn- thase51 that was cloned in this species. From the newly identified genes, two (SSL1 and SSL3) were necessary and sufficient to catalyze robust botryococcene biosynthesis from farnesyl

Downloaded on 20 October 2012 diphosphate (FPP), via the squalene precursor presqualene diphosphate (PSPP). Findings by Niehaus et al.50 suggest that over-expression of the SSL1 and SSL3 genes in fast-growth production-type photosynthetic and fermentative microorgan- Published on 12 December 2011 http://pubs.rsc.org | doi:10.1039/C1EE02514G isms might open the door to commercial exploitation of botryococcene (melting point ¼75 C; boiling point ¼ +285 C) as a fuel precursor and feedstock in the synthetic chemistry industry.

The efficiency and productivity of photosynthesis Fig. 12 Schematic illustrations of the pathway of incident sunlight Organisms of oxygenic photosynthesis can operate with a theo- absorption and utilization/dissipation by Chlamydomonas reinhardtii in retical maximum of 8–10% solar-to-biomass energy conversion liquid culture. (A) Fully pigmented (dark green) microalgae in a high- 52 efficiency. Experimental evidence showed that, under low-light density mass culture. Cells at the surface of the culture over-absorb conditions, i.e., when the rate of light absorption limits the rate incoming sunlight (more than can be utilized by photosynthesis), and of photosynthesis, this theoretical maximum is attained.53,54 dissipate most of it via non-photochemical quenching, thus limiting Under ambient sunlight conditions, however, solar-to-biomass productivity (P). Note that a high probability of absorption by the first energy conversion efficiencies are substantially lower than under layer of cells would cause shading, i.e., prevent cells deeper in the culture low-light intensities, with a best-case scenario of 2–3% for green from being exposed to sunlight. Measured microalgal solar-to-biomass microalgae, with even lower efficiencies for cyanobacteria, red energy conversion efficiencies under these conditions were at best 2–3%. algae, and C3 plants. A primary reason for such inefficiency is (B) Truncated chlorophyll antenna size (light green) microalgae in a high- density mass culture. Individual cells with a truncated Chl antenna for the that, at high photon flux densities, the rate of photon absorption photosystems have a diminished probability of absorbing sunlight, by the photosystem antenna pigments far exceeds the rate at thereby permitting greater penetration and a more uniform distribution which photosynthesis can utilize them, resulting in dissipation of and utilization of irradiance through the culture. This alleviates heat the excess absorbed photons as fluorescence or heat (Fig. 12A). dissipation of the absorbed sunlight and enhances photosynthetic These properties of the photosynthetic apparatus are readily productivity (P) by the culture as a whole. Extrapolated solar-to-biomass manifested in field or laboratory measurements of the light- energy conversion efficiencies under these conditions were 8–10%.

This journal is ª The Royal Society of Chemistry 2012 Energy Environ. Sci., 2012, 5, 5531–5539 | 5537 View Online

Assembly of large arrays of light-absorbing antenna pigment saturation of photosynthesis,68 improving solar-to-biomass molecules in photosynthetic organisms is an evolutionary trait energy conversion efficiencies and culture productivities.8,69 that confers a survival advantage in the wild, especially so as In laboratory or small-scale microalgal cultures under bright most species occupy an ecotype where light is often limiting. This sunlight and otherwise optimal growth conditions, many studies selective pressure, however, becomes counterproductive when have shown maximal sustained productivities of 20–40 g dry strains are cultivated in high-density cultures or canopies under weight m2 d1.70–74 These productivities are, however, substan- direct and bright sunlight. The resulting over-absorption and tially lower than the calculated 77 5 g dry weight m2 d1 wasteful dissipation of sunlight from the top layers of cells lowers theoretical maximum,52 suggesting the possibility of an overall solar energy conversion efficiencies and plant productivities to 3-fold improvement that can be achieved in mass culture.52,75 The levels that are substantially less than the theoretical maxima. To latter will be achieved upon optimization of the parameters dis- make the situation worse, chloroplasts at the surface of a high- cussed by Long et al.66 and also upon optimization of sunlight density cultivation are subject to photoinhibition of photosyn- absorption and utilization in microalgal cultures and plant thesis,64,65 whereas cells deeper in the canopy or culture are canopies. Thus, in view of efforts to increase the yield of food, deprived of much needed sunlight, as this is strongly attenuated fuel and other renewable bio-products, approaches by which to due to filtering (Fig. 12A). improve the solar energy conversion efficiency of photosynthesis In general terms, improvements in the efficiency and produc- merit serious consideration. tivity of photosynthesis can be achieved upon optimization of the photochemical or carbon reactions. An authoritative analysis of Fill’er up with photosynthetic biofuels? strategies by which to achieve improvement in photosynthesis and crop yield was presented66,67 suggesting potential additive The answer to this pivotal question, arguably, is ‘‘Yes, may be’’. gains upon several improvements. Long et al.66 showed that The above perspective on biofuels would suggest that we are increases in leaf photosynthesis are closely associated with beginning to see the proverbial ‘‘light at the end of the tunnel’’. similar increases in plant yield, thereby suggesting that Hydrogen, isoprene, and longer hydrocarbons such as botry- improvements in photosynthesis would benefit productivity. coccene, bio-oils and other biofuel products can be generated They identified a number of strategies to improve photosynthesis directly by photosynthesis in a single microorganism. Questions as a way by which to improve yield. Included were concepts of of concept validation, including product separation from the altered canopy architecture for better deployment of leaves and biomass, improvements in the solar energy conversion efficiency avoidance of shading, faster relaxation of the photoprotected of photosynthesis, and reduction to practice of the above are all state in the photochemical apparatus to improve exciton utili- in the realm of the feasible. Why then reluctancy with a ‘‘may zation, minimizing the impact of photorespiration on biomass be’’? The common ‘‘barrier’’ in the commercial exploitation of all yield, increasing the catalytic efficiency of Rubisco, and faster of the above mentioned photosynthetic biofuel processes and products is unfavorable at present process economics, due to the

Downloaded on 20 October 2012 regeneration of the CO2 accepting molecule ribulose- 1,5-bisphosphate (RuBP). Collectively, a 50% improvement in low relative yield of the fuel generating reactions. Thus, with the photosynthesis and productivity could be achieved by these current state of the art, ‘‘process reduction to practice’’ does not strategies. immediately translate into ‘‘commercial viability’’. Lacking is the Published on 12 December 2011 http://pubs.rsc.org | doi:10.1039/C1EE02514G In addition to the above-described biochemical consider- ability to alter flux and to increase the efficiency and yield of ations, there are agronomic features that would impact yield. For product generation beyond the threshold that constitutes example, productivity on a surface area basis would be impacted commercial viability. Enhancements in the product-to-biomass by the so-called ‘‘carpeting effect’’, meaning the totality of foliage ratio, and the metabolic engineering required to achieve it, coverage, or lack thereof, of a given land surface area. In this constitute a fundamental challenge in the biofuels field, regard- case, a low plant density would compromise yield. Likewise, less of the product to be generated. The reason for this difficulty a limited growth season would tend to lower the annual per is that altering cellular biochemistry to generate biofuel product surface area productivity of a fuel crop. (s) entails no incentive for the organism, whose priorities are In all photosynthetic systems, however, over-absorption of growth and competitive survival, rather than helping to solve bright sunlight and wasteful dissipation of most of it via non- energy-related problems in the human economy. The above photochemical quenching are the primary and most important analysis and perspective may also constitute recognition of source of the lower-than-theoretical efficiency and productivity a bottleneck in current progress, and a call to action by of photosynthesis. Rectifying this pitfall could improve produc- researchers, whose focal effort should be to develop protocols tivity by up to 300%. Such endeavor requires minimizing the size enabling stable alteration in metabolic flux, so as to achieve of the light-harvesting pigment arrays in the photosynthetic enhanced efficiencies of product generation for human industrial apparatus, so as to prevent the absorption of excess irradiance at consumption. the surface of a microalgal culture of plant canopy. A smaller Chl antenna size for the photosystems will compromise the ability of References individual cells to compete for sunlight in the wild. However, it 1 L. Florin, A. Tsokoglou and T. Happe, J. Biol. Chem., 2001, 276, will permit greater irradiance penetration through a high cell- 6125–6132. density culture, enabling more cells to participate in photosyn- 2 A. Melis and T. Happe, Plant Physiol., 2001, 127, 740–748. thesis and productivity (Fig. 12B). It has been demonstrated that 3 M. L. Ghirardi, M. C. Posewitz, P. C. Maness, A. Dubini, J. Yu and M. Seibert, Annu. Rev. Plant Biol., 2007, 58, 71–91. a ‘‘truncated light-harvesting chlorophyll antenna’’ can alleviate 4 H. Gaffron, Nature, 1939, 143, 204–205. many of the optical pitfalls associated with the early light- 5 H. Gaffron and J. Rubin, J. Gen. Physiol., 1942, 26, 219–240.

5538 | Energy Environ. Sci., 2012, 5, 5531–5539 This journal is ª The Royal Society of Chemistry 2012 View Online

6 H. Cao, L. Zhang and A. Melis, J. Appl. Phycol., 2001, 13, 25–34. 41 V. J. Martin, D. J. Pitera, S. T. Withers, J. D. Newman and 7 J. E. W. Polle, J. R. Benemann, A. Tanaka and A. Melis, Planta, 2000, J. D. Keasling, Nat. Biotechnol., 2003, 21, 796–802. 211, 335–344. 42 S. Wu, M. Schalk, A. Clark, R. B. Miles, R. Coates and J. Chappell, 8 J. E. W. Polle, S. Kanakagiri and A. Melis, Planta, 2003, 217, 49–59. Nat. Biotechnol., 2006, 24, 1441–1447. 9 M. L. Ghirardi, R. K. Togasaki and M. Seibert, Appl. Biochem. 43 E. Leonard, P. K. Ajikumara, K. Thayer, W.-H. Xiaoa, J. D. Moa, Biotechnol., 1997, 63, 141–151. B. Tidor, G. Stephanopoulos and K. L. J. Prather, Proc. Natl. 10 V. A. Boichenko, E. Greenbaum and M. Seibert, in Photoconversion Acad. Sci. U. S. A., 2010, 107, 13654–13659. of Solar Energy: Molecular to Global Photosynthesis, ed. M. D. 44 A. Banerjee, R. Sharma, Y. Chisti and U. C. Banerjee, Crit. Rev. Archer and J. Barber, Imperial College Press, London, 2004, pp. Biotechnol., 2002, 22, 245–279. 397–452. 45 P. Metzger and C. Largeau, Appl. Microbiol. Biotechnol., 2005, 66, 11 A. Melis, L. Zhang, M. Forestier, M. L. Ghirardi and M. Seibert, 486–496. Plant Physiol., 2000, 122, 127–136. 46 E. Eroglu and A. Melis, Bioresour. Technol., 2010, 101, 2359–2366. 12 M. L. Ghirardi, L. Zhang, J. W. Lee, T. Flynn, M. Seibert, 47 E. Eroglu, S. Okada and A. Melis, J. Appl. Phycol., 2011, 23, 763–775. E. Greenbaum and A. Melis, Trends Biotechnol., 2000, 18, 506–511. 48 P. Metzger, E. Casadevall, M. J. Pouet and Y. Pouet, Phytochemistry, 13 L. Zhang, T. Happe and A. Melis, Planta, 2002, 214, 552–561. 1985, 24, 2995–3002. 14 A. Melis, Planta, 2007, 226, 1075–1086. 49 M. Audino, K. Grice, R. Alexander, C. J. Boreham and R. I. Kagi, 15 A. Hemschemeir, A. Melis and T. Happe, Photosynth. Res., 2009, 102, Geochim. Cosmochim. Acta, 2001, 65, 1995–2006. 523–540. 50 T. D. Niehaus, S. Okada, T. P. Devarenne, D. S. Watt, V. Sviripad 16 O. Kruse and B. Hankamer, Curr. Opin. Biotechnol., 2010, 21, 238– and J. Chappell, Proc. Natl. Acad. Sci. U. S. A., 2011, 108, 12260– 243. 12265. 17 O. Kruse, J. Rupprecht, K. P. Bader, S. Thomas-Hall, P. M. Schenk, 51 S. Okada, T. P. Devarenne and J. Chappell, Arch. Biochem. Biophys., G. Finazzi and B. Hankamer, J. Biol. Chem., 2005, 280, 34170–34177. 2000, 373, 307–317. 18 U. Bossel, B. Eliason, G. Taylor, http://www.evworld.com/library/ 52 A. Melis, Plant Sci., 2009, 177, 272–280. h2economyFinalReport.pdf, 2003. 53 A. C. Ley and D. Mauzerall, Biochim. Biophys. Acta, Bioenerg., 1982, 19 C. R. Baker and R. L. Shaner, Int. J. Hydrogen Energy, 1978, 3, 321– 680, 95–106. 334. 54 O. Bjorkman and B. Demmig, Planta, 1987, 170, 489–504. 20 M. T. Syed, S. A. Sherif, T. N. Veziroblu and J. W. Sheffield, Int. J. 55 P. H. Homann and G. H. Schmid, Plant Physiol., 1967, 42, 1619– Hydrogen Energy, 1998, 23, 565–576. 1632. 21 M. Temporal, J. J. Lopez cell, A. R. Piriz, N. Grandjouan, 56 O. Bjorkman, N. K. Boardman, J. M. Anderson, S. W. Thorne, N. A. Tahir and D. H. H. Hoffmann, Laser Part. Beams, 2005, 23, D. J. Goodchild and N. A. Pyliotis, Carnegie Inst. Yearbook, 1972, 137–142. vol. 71, pp. 115–135. 22 Z. Dehouche, M. Savard, F. Laurencelle and J. Goyette, J. Alloys 57 N. K. Boardman, J. M. Anderson, O. Bjorkman, D. J. Goodchild, Compd., 2005, 400, 276–280. L. H. Grimme and S. W. Thorne, Port. Acta Biol., Ser. A, 1974, 23 J. K. Kim, I. S. Park, K. J. Kim and K. Gawlik, Int. J. Hydrogen XIV, 213–236. Energy, 2008, 33, 870–877. 58 K. Okabe, G. H. Schmid and J. Straub, Plant Physiol., 1977, 60, 150– 24 A. Zaluska, L. Zaluski and J. O. Strom-Olsen,€ Appl. Phys. A: Mater. 156. Sci. Process., 2001, 72, 157–165. 59 M. Mitra and A. Melis, Opt. Express, 2008, 16, 21807–21820. 25 B. Sakintuna, F. Lamari-Darkrim and M. Hirscher, Int. J. Hydrogen 60 M. F. Hohmann-Marriott, R. E. Blankenship and R. W. Roberson, Energy, 2007, 32, 1121–1140. Photosynth. Res., 2005, 86, 145–154. 26 D. J. Collins and H. C. Zhou, J. Mater. Chem., 2007, 17, 3154–3160. 61 A. N. Glazer and A. Melis, Annu. Rev. Plant Physiol., 1987, 38, 11–45.

Downloaded on 20 October 2012 27 S. I. Orimo, Y. Nakamori, J. R. Eliseo, A. Andreas Zuttel€ and 62 A. Melis, Biochim. Biophys. Acta, Bioenerg., 1991, 1058, 87–106. C. M. Jensen, Chem. Rev., 2007, 107, 4111–4132. 63 A. Melis, in Oxygenic Photosynthesis: the Light Reactions, ed. D. R. 28 N. L. Rosi, J. Eckert, M. Eddaoudi, D. T. Vodak, J. H. Kim, Ort and C. F. Yocum, Kluwer Academic Publishers, Dordrecht, M. O’Keeffe and O. M. Yaghi, Science, 2003, 300, 1127–1129. Netherlands, 1996, pp. 523–538. 29 H. S. Kim, H. Lee, K. S. Han, J. H. Kim, M. S. Song, M. S. Park, 64 S. B. Powles, Annual Review of Plant Physiology and Plant Molecular Published on 12 December 2011 http://pubs.rsc.org | doi:10.1039/C1EE02514G J. Y. Lee and J. K. Kang, J. Phys. Chem. B, 2005, 109, 8983–8986. Biology, 1984, 35, 15–44. 30 B. D. James, G. N. Baum, J. Perez, K. N. Baum, 2009, http://www. 65 A. Melis, Trends Plant Sci., 1999, 4, 130–135. osti.gov/bridge. 66 S. P. Long, X.-W. Zhu, S. L. Naidu and D. R. Ort, Plant, Cell 31 A. Melis and M. Melnicki, Int. J. Hydrogen Energy, 2006, 31, 1563– Environ., 2006, 29, 315–330. 1573. 67 X. G. Zhu, S. P. Long and D. R. Ort, Curr. Opin. Biotechnol., 2008, 32 J. Chappell, Annu. Rev. Plant Physiol. Plant Mol. Biol., 1995, 46, 521– 19, 153–159. 547. 68 A. Melis, J. Neidhardt and J. R. Benemann, J. Appl. Phycol., 1999, 10, 33 D. McGarvey and R. Croteau, Plant Cell, 1995, 7, 1015–1026. 515–525. 34 F. X. Cunningham and E. Gantt, Annu. Rev. Plant Physiol. Plant 69 Y. Nakajima and R. Ueda, J. Appl. Phycol., 1999, 11, 195–201. Mol. Biol., 1998, 49, 557–583. 70 Z. Dubinski, T. Berner and S. Aaronso, Biotechnol. Bioeng. Symp., 35 T. D. Sharkey and S. Yeh, Annu. Rev. Plant Physiol. Plant Mol. Biol., 1979, 8, 51–68. 2001, 52, 407–436. 71 A. Ben-Amotz, in Biochemical and Photosynthetic Aspects of Energy 36 H. K. Lichtenthaler, Annu. Rev. Plant Physiol. Plant Mol. Biol., 1999, Production, ed. A. San Pietro, Academic Press, New York, 1980, 50, 47–65. pp. 191–208. 37 H. K. Lichtenthaler, Biochem. Soc. Trans., 2000, 28, 785–789. 72 A. Ben-Amotz and M. Avron, Trends Biotechnol., 1990, 8, 121–126. 38 T. D. Sharkey, A. E. Wiberley and A. R. Donohue, Ann. Bot., 2008, 73 Y. K. Lee, J. Appl. Phycol., 1997, 9, 403–411. 101, 5–18. 74 G. C. Dismukes, D. Carrieri and N. Bennette, et al., Curr. Opin. 39 P. Lindberg, S. Park and A. Melis, Metab. Eng., 2010, 12, 70–79. Biotechnol., 2008, 19, 235–240. 40 F. K. Bentley and A. Melis, Biotechnol. Bioeng., 2012, 109, 100–109. 75 D. R. Ort, X.-G. Zhu and A. Melis, Plant Physiol., 2011, 155, 79–85.

This journal is ª The Royal Society of Chemistry 2012 Energy Environ. Sci., 2012, 5, 5531–5539 | 5539