Subscribe rss-microbe
Home Home Harnessing Bacteria That Use Light To Produce Hydrogen
Harnessing Bacteria That Use Light To Produce Hydrogen Print E-mail

 

The challenge is to increase efficiencies among bacteria that depend on light and simple compounds to make hydrogen

James B. McKinlay and Caroline S. Harwood

James B. McKinlay is Assistant Professor in the Department of Biology, Indiana University, Bloomington, and Caroline S. Harwood is Gerald and Lyn Grinstein Endowed Professor of Microbiology in the Department of Microbiology, University of Washington, Seattle. 

Author profile--Harwood: Pond Biology and Personal Experiences Shaped Her Research Interests

Summary

 

 

 

● Rhodopseudomonas palustris is one of several types of microorganisms that produce hydrogen gas (H2)-in this case, anaerobically and using simple compounds as electron donors and light as a source of energy.
● R. palustris nitrogenase produces pureH2 when N2 is unavailable; this microbe and others like it make large amounts of nitrogenase.
● To produce H2 more efficiently, we seek nitrogenase- producing R. palustris strains that overcome repressors such as ammonium, which is plentiful in carbon and electron-donating feedstocks.
● Single-nucleotide changes in nifA, encoding the master transcriptional regulator of nitrogenase, overcome nitrogenase regulation and increase H2 productivity.
● Genomic sequence analysis of R. palustris strains and novel ways for capturing light energy could lead to additional means for improving overall H2 productivity.

Rhodopseudomonas palustris is one of several types of microorganisms that produce hydrogen gas (H2), a deceptively simple molecule consisting of two protons and two electrons. We study how R. palustris produces H2 in part because this organism so aptly illustrates the different types of data and skill sets that are required to bring a microbial metabolic property into practical use. Focused studies on specific mechanisms, broader approaches to decipher extensive biological networks, and nonbiological skills such as those needed to design and operate hydrogen-producing reactors all come into play as part of an effort to harness microbes to augment and replace unsustainable petrochemical processes.

R. palustris is but one among diverse microbes that release H2 during growth, ridding themselves of excess electrons that they generate. These electrons combine with dissolved protons that are freely available within the cells. Notably, the energy content of H2, a mere waste product for R. palustris, is three times that of gasoline. Indeed, H2 is being used to power automobile prototypes whose engines are reconfigured to burn this gas, yielding energy and water as a nonpolluting combustion byproduct. Although widespread use of H2 as a transportation fuel is not yet a practical reality, H2 is being used industrially in petroleum refining operations and to produce ammonia. Even though commercially produced H2 derives mainly from fossil fuels, many microbiologists are asking whether biologically produced H2 can contribute significantly to our economy.

Several Types of Microorganisms Produce H2

Several major categories of microbes produce H2, including anaerobic fermentative bacteria, cyanobacteria/green algae, and purple nonsulfur photosynthetic bacteria (PNSB). Each of these microbial catalysts has its pros and cons. Fermentative bacteria produce H2 at high rates from glucose and other sugars, but yields are low because this type of microbe simultaneously excretes large amounts of electron-rich organic compounds. Meanwhile, cyanobacteria and green algae rely on solar energy to generate energy-rich electrons to produce H2 from water, the most abundant source of electrons. This process would be ideal if these microbes did not also produce oxygen, to which the enzymes that produce H2 are extremely sensitive.

PNSB have advantages over cyanobacteria and algae. For example, because PNSB produce H2 under anaerobic conditions, their H2-generating enzymes are not exposed to oxygen molecules. Moreover, because they obtain energy from light, H2 yields from electron-donating compounds are very high. However, the electron donors that PNSB use to produce H2 are not as abundant as water. On the plus side, PNSB metabolize many different kinds of organic compounds, including agricultural and food wastes.

harwoodf1It might prove possible to grow all three types of H2-generating microbes together to maximize resources and overall efficiencies (Fig. 1). One key challenge is to understand the inner workings of H2-producing microbes from these different groups to learn how to maximize their individual capabilities and how to intervene when things go wrong.

Among PNSB, R. palustris Is Ideal for Studying H2 Production

R. palustris, which is widespread and readily isolated from waterlogged soils and stagnant bodies of water, is a nearly ideal model organism for studying biological production of H2.

When oxygen in its environment is depleted, R. palustris cells turn a characteristic deep purple color, synthesizing light-absorbing pigments that they need to carry out photosynthesis. R. palustris generates ATP anaerobically by a photosynthetic process that differs from that of green plants, algae, and cyanobacteria in that it does not produce oxygen. The genome of R. palustris encodes all the components of the Calvin cycle for fixing CO2, enabling it to grow photoautotrophically with CO2 as its sole carbon source. However, it prefers to grow photoheterotrophically on organic compounds, especially acetate and lignin monomers, while generating ATP photosynthetically.

When growing anaerobically in light, R. palustris and other PNSB can fix nitrogen gas from the atmosphere and convert it to ammonium. Although the main function of nitrogenase, the central enzyme of nitrogen fixation, is considered to be ammonium production, it also harwoodf2produces H2, as an obligate product of its catalytic cycle (Fig. 2). Producing H2 can be important for R. palustris, allowing it to dispose of electrons generated during metabolism. Nitrogenase devotes about 25% of its electrons to producing H2 even in the presence of 50 atmospheres of N2, according to Frank Simpson and Robert Burris at the University of Wisconsin, whose findings date back several decades.

A lesser-known fact is that nitrogenase produces pure H2 when N2 is unavailable. In 1949, as the first graduate student of Martin Kamen at Washington University, Howard Gest learned that, in the PNSB Rhodospirillum rubrum, H2 production is lower when N2 gas is present. This observation led Gest to hypothesize, test, and prove, that the enzyme responsible for producing H2 also fixes N2. Remarkably, his observing H2 production was possible because he used a growth medium containing glutamate instead of ammonium as the nitrogen source. Ammonium completely represses nitrogenase activity, and thereby H2 production, whereas glutamate does not.

PNSB bacteria make large amounts of nitrogenase (up to 2% of cellular protein), an enzyme that proves to be a slow catalyst. Moreover, making this enzyme is complicated in itself. Specifically, the enzyme contains an iron-molybdenum cofactor that must be synthesized separately and then inserted into the apo-enzyme by a series of auxiliary proteins. Finally, the nitogenase reaction uses 16 ATP molecules to fix a single molecule ofN2 to make 2NH4+. For these reasons, microbes tightly regulate their synthesis of nitrogenase and its activity.

harwoodf3For example, R. palustris regulates nitrogenase by several mechanisms in a multi-tiered fashion (Fig. 3). These mechanisms are found in other bacterial metabolic pathways. However, the ways in which nitrogenase regulatory networks are put together vary from microbe to microbe. The R. palustris nitrogenase regulatory hierarchy, which we inferred from its gene content and from gene expression experiments, depends on cellular levels of the metabolite α-ketoglutarate and the amino acid glutamine. Levels of α-ketoglutarate are high in cells that are starving for nitrogen and therefore must resort to nitrogen fixation, whereas high levels of glutamine are found in cells with access to a rich nitrogen source such as ammonium.

The uridylyltransferase GlnD senses differing intracellular ratios of α-ketoglutarate to glutamine. GlnD then acts via small trimeric signal transduction proteins called PII proteins, adding uridylyl groups to them when α-ketoglutarate is abundant and removing those groups when glutamine is abundant. The uridylylation-state of the PII proteins determines the degree to which they physically interact with several other cellular proteins. For example, they can modulate (i) the activities of the NtrBC two-component regulatory system, which controls expression of genes involved in nitrogen metabolism, including nifA; (ii) NifA, the master transcriptional activator protein of nitrogenase, and (iii) DraT and DraG, enzymes that control nitrogenase activity by adding or removing ADPribosyl groups to it.

Harnessing R. palustris To Produce More H2

For R. palustris to produce H2 more efficiently, we will need strains that synthesize and operate nitrogenase even when repressors such as ammonium are present. This goal is particularly important if we use agricultural and industrial wastes as carbon and electron-donating feedstocks; both contain ammonium.

Understanding how nitrogenase fits into the overall physiology of R. palustris already is leading to improvements in H2 production efficiencies. Importantly, nitrogenase helps in producing H2, providing a means for cells to dispose of electrons generated during metabolism. When R. palustris is growing photoheterotrophically and generating energy from light, it typically incorporates most of the carbon it is provided into new cell material. Through 13C-labeling experiments, we found that about half of the NADH that cells generate during this process is reoxidized to NAD+ in cellular biosynthetic reactions. The other half of the NADH must be converted to NAD+ by other means or metabolism would stop. One way that R. palustris and other PNSB can do this is to transfer electrons from NADH to nitrogenase, which makes H2 gas, thereby releasing electrons from cells and regenerating NAD+.

Several research groups identified mutants of PNSB that produce H2 under all conditions. Our approach involved inoculating R. palustris into anaerobic tubes of mineral growth medium, in which it could not grow unless it mutated to produce H2. The growth medium contained ammonium as a nitrogen source and cyclohexanecarboxylate, an electron-rich organic compound that cells can use as carbon for growth only if they have a way of disposing of excess electrons generated from its oxidation. After several months without growth, the anaerobic tubes became turbid with mutants of purple R. palustris cells that produced H2. They carried mutations allowing R. palustris nitrogenase to escape repression by ammonium.

Remarkably, only a single nucleotide change in nifA, encoding the master transcriptional regulator of nitrogenase, can overcome regulation of nitrogenase in R. palustris. Such a mutation led to an amino acid substitution that locked the NifA protein into a constitutively active state. We were fortunate to work with R. palustris because single mutations are enough to allow nitrogenase to escape ammonium repression. Other PNSB typically require mutations in at least two genes to become constitutive for nitrogenase.

Because producing H2 rids R. palustris of excess electrons, perhaps blocking other electron-disposing pathways could increase H2 yields. For instance, even when CO2 is not provided, R. palustris scavenges considerable amounts of CO2 from organic compounds that it metabolizes-disposing of excess electrons while recycling NADH to NAD_. Furthermore, some CO2 is fixed even when cells produce H2 via nitrogenase. Therefore, we blocked the CO2-fixing step within the Calvin cycle by deleting genes encoding the enzyme ribulose 1,5-bisphosphate carboxylase. Sure enough, this change increased H2 yields by an amount corresponding to the amount of CO2 that is fixed via the Calvin cycle in the parent strain. These two examples-one involving a regulatory mutant, and the other, a mutant that cannot fix CO2-illustrate how understanding the physiology of a bacterium can be important for increasing yields of desirable products such as H2.

Other Approaches to Improving R. palustris H2 Productivity

Analyses of the genome sequences of several R. palustris strains suggest additional ways for improving H2 productivity. For instance, one strain encodes three nitrogenase isozymes-the most commonly found molybdenum-containing nitrogenase and the alternative vanadium- and iron-containing nitrogenases. Although the molybdenum nitrogenase is the most efficient at converting N2 to NH3, the alternative vanadium- and iron-containing nitrogenases are better at producing H2. Thus, they could be used to improve H2 yields, especially when it is impractical to remove the competing substrate, N2, from a bioreactor. Although little is known about the underlying mechanisms, extreme nitrogen starvation appears to activate the alternative nitrogenases of R. palustris.

Other strain-specific features could increase H2 productivity in R. palustris. For instance, one strain encodes fermentative enzymes that enable it to produce H2 in the dark, opening the possibility of producing H2 continuously by producing H2 photosynthetically during the day and fermentatively at night. Yet another strain absorbs light at wavelengths that most R. palustris strains do not, thereby improving the photosynthetic efficiency of its H2 production.

Implementing a commercially viable, biological process for producing H2 will depend on appropriate electron-rich feed stocks. For instance, R. palustris can produce H2 from partially fermented vegetable waste, according to Roberto de Philippis at the University of Florence in Italy and his collaborators. We recently found that inorganic thiosulfate, a byproduct of paper manufacturing, can serve as an electron donor for H2 production. In principle, by using thiosulfate, R. palustris could be grown photoautotrophically, consuming the greenhouse gas CO2 while producing H2-a process now claimed only by cyanobacteria and algae.

Another practical challenge when biologically producing H2 on a large scale is how to capture adequate light energy. Light intensity decreases exponentially with depth in a turbid culture. Thus, traditional vat-shaped bioreactors are impractical. Nontraditional glass bioreactors that are thin rectangles or long cylinders provide one way to maximize the ratio of surface area to volume.

Another strategy, developed by Michael Flickinger at North Carolina State University, is to immobilize bacterial cells in thin nanoporous and transparent latex films. At least on a small scale, nongrowing cells of R. palustris can be embedded in latex at very high concentrations of 1011 cells/ml, forming very thin, 50-μm films. When illuminated and supplied with buffer containing acetate as an electron donor, such cells produce H2 continuously for several months. One can further imagine layering films of different H2-producing photosynthetic bacteria to maximize photosynthetic efficiencies. H2-producing biological films might also be painted onto light-conducting plastics, and arrayed as cylindars in enclosed bioreactors. While these approaches are now too expensive to be practical, someday with further improvements, they might reduce the space needed to convert light energy into large quantities of H2.

SUGGESTED READING

Dixon, R., and D. Kahn. 2004. Genetic regulation of biological nitrogen fixation. Nature Rev. Microbiol. 2:621-631. Gest, H. 1999. Memoir of a 1949 railway journey with photosynthetic bacteria. Photosynthesis Res. 61:91-96.

Gosse, J. L., B. J. Engel, J. C. Hui, C. S. Harwood, and M. C. Flickinger. 2010. Progress toward a biomimetic leaf: 4,000 h of hydrogen production by coating-stabilized nongrowing photosynthetic Rhodopseudomonas palustris. Biotechnol. Prog. 26:907-918.

Huang, J. J., E. K. Heiniger, J. B. McKinlay, and C. S. Harwood. 2010. Production of hydrogen gas from light and the inorganic electron donor thiosulfate by Rhodopseudomonas palustris. Appl. Environ. Microbiol. 76:7717-7722.

McKinlay, J. B., and C. S. Harwood. 2010. Carbon dioxide fixation as a central redox cofactor recycling mechanism in bacteria. Proc. Natl. Acad. Sci. USA 107:11669-11675.

McKinlay, J. B., and C. S. Harwood. 2010. Photobiological production of hydrogen gas as a biofuel. Curr. Opin. Biotechnol. 21:244-251.

Melis, A., and M. R. Melnicki. 2006. Integrated biological hydrogen production. Int. J. Hydrogen Energy 31:1563-1573.

Oda, Y., F. W. Larimer, P. S. Chain, S. Malfatti, M. V. Shin, L. M. Vergez, L. Hauser, M. L. Land, S. Braatsch, J. T. Beatty, D. A. Pelletier, A. L. Schaefer, and C. S. Harwood. 2008. Multiple genome sequences reveal adaptations of a phototrophic bacterium to sediment microenvironments. Proc. Natl. Acad. Sci. USA 105:18543-18548.

Rubio, L. M., and P. W. Ludden. 2008. Biosynthesis of the iron-molybdenum cofactor of nitrogenase. Annu. Rev. Microbiol. 62:93-111.