OUP user menu

Inactivation of the Escherichia coli K-12 twin-arginine translocation system promotes increased hydrogen production

David W. Penfold, Frank Sargent, Lynne E. Macaskie
DOI: http://dx.doi.org/10.1111/j.1574-6968.2006.00333.x 135-137 First published online: 1 September 2006


The effect of deleting the genes encoding the twin-arginine translocation (Tat) system on H2 production by Escherichia coli strain MC4100 and its formate hydrogenlyase upregulated mutant (ΔhycA) was investigated. H2 evolution tests using two mutant strains defective in Tat transport (ΔtatC and ΔtatA-E) showed that the rate doubled from 0.88±0.28mL H2mg dry weight−1L culture−1 in the parental strain, to 1.70±0.15 and 1.75±0.18mL H2mg dry weight−1L culture−1, respectively, in the ΔtatC and ΔtatA-E strains. This increase was comparable to that of a previously characterized hydrogen over-producing E. coli strain carrying a ΔhycA allele. Construction of a tatC, ΔhycA double deletion strain did not increase hydrogen production further. Inactivation of the Tat system prevents correct assembly of the uptake hydrogenases and formate dehydrogenases in the cytoplasmic membrane and it is postulated that the subsequent loss of basal levels of respiratory-linked hydrogen and formate oxidation accounts for the observed increases in formate-dependent hydrogen evolution.

  • Escherichia coli
  • hydrogenase
  • hydrogen production
  • Tat mutants


The growing concern about the greenhouse effect and depletion in fossil fuel reserves has led to the search for a clean fuel. One such possibility is hydrogen as the only product of its combustion is water. Conventional methods of producing hydrogen, such as coal gasification and the electrolysis of water, are energy intensive. Biological hydrogen production uses ambient temperatures and lower pressures and is therefore potentially economic (Das & Veziroğlu, 2001), especially as biohydrogen production can be coupled to the utilization of organic industrial wastes and this is, therefore, a potentially useful alternative method of producing hydrogen. The majority of studies have used Rhodobacter sp. (Zhuet al,1999; Eroğluet al,2004), Clostridium sp. (Yokoiet al,1998, 2001), Enterobacter sp. (Tanisho & Ishiwate, 1995; Kumar & Das, 2000) and anaerobic digestion methods (Mizunoet al,2000; Lay, 2000). A previous study by Penfold (2003) investigated hydrogen evolution by Escherichia coli K-12 strain HD701, a derivative of the well-characterized parental strain MC4100. Strain HD701 does not possess the gene encoding the HycA regulator of the formate hydrogenlyase (FHL) system (Sauteret al,1992). This FHL enzyme system is responsible for the generation of hydrogen from formate as a homeostatic response to the drop in pH during mixed acid fermentation (Sauteret al,1992). The absence of the HycA repressor and consequently increased activity of the FHL system significantly increased hydrogen production by HD701 compared with strain MC4100 (Penfoldet al,2003). It is possible however, that net hydrogen production by E. coli is reduced by the competing activities of the uptake hydrogenases, which would recycle a portion of the hydrogen produced, and the respiratory formate dehydrogenases which would oxidize a portion of the formate produced, from which all hydrogen evolved by E. coli is derived.

The E. coli respiratory hydrogenases and formate dehydrogenases are located on the periplasmic side of the cytoplasmic membrane and are transported by the twin-arginine translocation (Tat) protein translocation system. Escherichia coli uses the Tat system in parallel to the Sec protein transport system to transport a group of exported proteins which are synthesized with a characteristic N-terminal signal peptide containing a SRRxFLK ‘twin-arginine’ amino acid motif (Berkset al,2003). Proteins possessing this twin-arginine signal peptides are posttranslationally directed to the membrane-embedded Tat system which allows transport of fully folded, matured proteins without rendering the membrane permeable to protons and other ions (Berkset al,2003). The Tat translocase of E. coli comprises the TatA, TatB, TatC and TatE membrane proteins. The core Tat transporter is a large oligomeric complex of the TatA, TatB and TatC proteins (Berkset al,2003). Known proteins transported by the Tat system include the two uptake hydrogenases, hydrogenase-1 and -2, and the two formate dehydrogenases, FdHN and FdHO. The FHL system also contains a hydrogenase component (hydrogenase-3) and a formate dehydrogenase component (the isoenzyme FdHH), however both of these isoenzymes are cytoplasmically located and hence are not to be transported by the Tat system.

Here, the effect of deleting the genes encoding the Tat system on hydrogen production was investigated.

Materials and methods

Strains of E. coli and construction of Tat mutants

Strains used in this study were derivatives of E. coli K12 strain MC4100 (F−, ΔlacU169, araD139, rpsL150, relA1, ptsF, rbs, flbB5301) and were HD701 (ΔhycA) (Sauteret al,1992), FTD701 (ΔhycA, ΔtatC::SpecR), DADE (ΔtatABCD, ΔtatE) (Wexleret al,2000) and B1LK0 (ΔtatC) (Bogschet al,1998). Strain FTD701 was constructed by P1 transduction of the tatC mutant allele from NRS-1 (ΔtatC::SpecR) (Stanleyet al,2000) into HD701 as described (Wexleret al,2000).

Media and culture growth

All strains were maintained on nutrient agar (NA) plates (Oxoid, UK) as described previously (Penfoldet al,2003). Cells were grown for hydrogen production tests in nutrient both (NB) no. 2 (Oxoid, UK). The composition of phosphate-buffered saline (PBS) was 0.8gL−1 NaCl, 0.2gL−1 KCl, 1.43gL−1 Na2HPO4 and 0.2gL−1 KH2PO4. The cell density of samples was measured spectrophotometrically at OD600nm (Ultraspec III, Pharmacia) and related to bacterial dry weight using a predetermined calibration curve. An optical density of 1.0 was calculated to equate to c. 0.5g dry weight L culture−1.

Hydrogen evolution experiments

Cells were grown aerobically (200r.p.m.) overnight in nutrient broth (Oxoid) at 30°C. Reactors (working volume of 310mL) were inoculated with 10% (v/v) overnight culture, 248mL PBS (pH 7.3) and glucose (100mM final concentration). The reactors were sparged with argon for 1h to produce anaerobic conditions and incubated at 30°C. The carbon dioxide coproduced with the hydrogen was removed using a 1M NaOH trap and the volume of evolved hydrogen was measured as described previously (Penfoldet al,2003).

Results and discussion

The Tat system was inactivated in both E. coli MC4100 and HD701 (ΔhycA). The MC4100 tat strains were B1LK0 (ΔtatC) and DADE (ΔtatA-E) and have been described previously (Bogschet al,1998; Wexleret al,2000). In this study, a ΔtatC, ΔhycA double mutant was constructed and designated FTD701.

Growth studies on the five strains under aerobic conditions (used to grow the cell inocula) showed the three Tat mutants grew at a slower rate than MC4100 and HD701 (not shown). This slow growth may be attributed to the cell separation and defective outer membrane phenotype associated with Tat mutants (Stanleyet al,2001). As hydrogen evolution experiments utilized resting cells suspended in PBS the pregrowth conditions for each strain were set to ensure that the cultures reached approximately the same OD600nm simultaneously.

Both B1LK0 (ΔtatC) and DADE (ΔtatA-E) strains showed a significant increase in total hydrogen production to 1.70±0.15 and 1.75±0.18mL H2mg dry weightL culture−1, respectively, compared with the 0.88±0.28mL H2mg dry weightL culture−1 evolved by the parental strain MC4100 (Fig. 1). This is comparable to values obtained for strain HD701 (ΔhycA) −2.20±0.20mL H2mg dry weightL culture−1. This increased level of hydrogen production in the tat mutants can be explained by the inactivity of the Tat-dependent respiratory formate dehydrogenases, FdHN and FdHO, and uptake hydrogenases −1 and −2. Although respiratory chains are largely inactive under fermentative conditions, there are large amounts of respiratory electron donors produced endogenously by E. coli under such conditions (hydrogen and formate), and a small amount of respiratory electron acceptors are also produced under fermentative conditions (fumarate). In the tat mutants, however, even these low levels of respiratory formate and hydrogen oxidation are lost and as a result more formate will be available to be channelled into the FHL system, and more hydrogen so-produced will avoid recycling and escape to the atmosphere.

Figure 1

Comparison of hydrogen evolution by Escherichia coli MC4100, HD 701, MC4100 ΔtatC (B1LK0) Δtat A–E (DADE), and FTD701. Tat mutants were made as described in the text. ▪, MC4100 (wild type); ◆, HD701 (MC4100ΔhycA); ○, B1LK0 (MC4100ΔtatC); ▴, DADE (MC4100ΔtatA-E); □, FTD701 (HD701ΔtatC).

The volume of hydrogen evolved by B1LK0 (ΔtatC) and DADE (ΔtatA-E) was not significantly different (Fig. 1) indicating that the effect of deleting tatC or tatA-E has the same effect on hydrogen production. In an attempt to further increase hydrogen production, tatC was therefore deleted in strain HD701 (ΔhycA). This gave the strain FTD701 (ΔhycA, ΔtatC) which would have increased hydrogen production due to the loss of HycA and increased hydrogen production due to the loss of Tat-dependent respiratory enzymes. However, deleting the genes encoding the Tat system in HD701 (ΔhycA) had no significant accumulative effect on hydrogen production (Fig. 1). One reason for this could be that the FHL system is already evolving hydrogen at a maximum rate at the concentrations of glucose being used.

It can be concluded that inactivation of the Tat system provides an alternative route to the construction of hydrogen-overproducing bacterial strains, which may be combined with other phenotypes e.g. the ability to utilize sucrose as a fermentable carbon source (Penfold & Macaskie, 2004) which is a major constituent of confectionery waste, for use in biotechnological applications.


This work was funded by the BBSRC (studentship no. 00/B1/E/0698 to D.W.P.) and grant No BB/C516128/1. The authors wish to thank Professor A. Böck (Lehrstühl für Mikrobiologie der Universität, Germany) for permission to use the E. coli strains, Professor J.A. Cole (School of Biosciences, University of Birmingham, UK) for useful discussions and the Royal Society for Fellowships to L.E.M. and F.S.


View Abstract