Conversion of a Peroxiredoxin into a Disulfide Reductase by a Triplet Repeat Expansion

See allHide authors and affiliations

Science  05 Oct 2001:
Vol. 294, Issue 5540, pp. 158-160
DOI: 10.1126/science.1063143


Pathways for the reduction of protein disulfide bonds are found in all organisms and are required for the reductive recycling of certain enzymes including the essential protein ribonucleotide reductase. An Escherichia coli strain that lacks both thioredoxin reductase and glutathione reductase grows extremely poorly. Here, we show that a mutation occurring at high frequencies in the gene ahpC, encoding a peroxiredoxin, restores normal growth to this strain. This mutation is the result of a reversible expansion of a triplet nucleotide repeat sequence, leading to the addition of one amino acid that converts the AhpC protein from a peroxidase to a disulfide reductase. The ready mutational interconversion between the two activities could provide an evolutionary advantage to E. coli.

The cytoplasms of all living cells are highly reducing environments compared with their external milieu. A host of cytosolic proteins are responsible for maintaining a redox balance. In E. coli, two pathways in the cytoplasm function in overlapping ways to maintain cysteine residues in the reduced state (1). The thioredoxin system includes thioredoxin reductase (trxB) and two thioredoxins; the glutaredoxin system includes glutathione reductase (gor), glutathione, and three glutaredoxins (2, 3). These two systems recycle certain reductive enzymes by reducing disulfide bonds formed at their active sites (4–7).

A trxB,gor double mutant, lacking both cytosolic disulfide reductive pathways, grows poorly under aerobic conditions unless a reducing agent such as dithiothreitol (DTT) is present. The poor growth is likely due to the inability of these cells to synthesize deoxyribonucleotides, because ribonucleotide reductase is not maintained in the reduced state.

We have previously described a derivative of thetrxB,gor mutant strain, FÅ113, in which a suppressor mutation has restored a normal growth rate (8). Despite the likelihood that a source of disulfide-reducing power has been restored by the suppressor mutation, FÅ113 is able to efficiently catalyze disulfide bond formation in cytoplasmically retained secreted proteins. FÅ113 and its derivatives also enhance the expression in the cytoplasm of active tissue plasminogen activator (vtPA), a complex eukaryotic protein with multiple disulfide bonds (8).

We wished to determine the nature of the suppressor mutation in FÅ113. However, even though the cells were grown in the presence of DTT, suppressors arose in this strain at such a high frequency that we could not map them using transposon insertions that were cotransduced by bacteriophage P1 with the suppressor mutation. To overcome this problem, we introduced into the trxB,gor mutant a plasmid that contained the trxB gene under the control of the pBADara promoter (9). The frequency of suppressor mutations that occurred under restrictive growth conditions was still 0.5 × 10−3. Transposon insertions linked to the suppressor mutation clustered to minute 14 of the E. colichromosome, and the linkage data suggested that a mutation in theahpCF locus might be responsible for the suppression (Fig. 1A).

Figure 1

The ahpC locus is responsible for the restoration of disulfide bond reduction in a trxB,gorstrain. (A) Identification of the FÅ113 suppressor locus by analysis of randomly inserted λNK1324 transposons. The open reading frames that were disrupted by the transposon insertions are shown in gray and the ahpCF locus is marked in black. (B) Sequence comparison of the wild-type and mutant loci. The mutation is a triplet insertion (TCT) leading to the expansion of a triplet repeat sequence in the ahpC open reading frame. The insertion leads to the addition of a single amino acid (Phe) near Cys46, which is essential for the peroxidase activity of the peroxiredoxin AhpC. (C) Peroxiredoxins are present in all organisms, and the region containing the active site cysteine is well conserved. Ec, Escherichia coli; St, Salmonella typhi; Kp, Klebsiella pneumoniae; Sf, Shigella flexneri; Pp,Pseudomonas putida; Ta, Thermus aquaticus; Ll,Lactococcus lactis; Bs, Bacillus subtilis; Sa,Staphylococcus aureus; Tp, Treponema pallidum; Tv, Thermoplasma volcanium; TSA, thiol-specific antioxidant protein from Rattus norvegicus. Sequences were retrieved from GenBank and aligned using ClustalW.

The ahpCF locus of E. coli encodes the alkyl hydroperoxidase system, which reduces both hydrogen peroxide and alkyl hydroperoxides, thus protecting the cells from oxidative damage. Peroxides oxidize the active site Cys46 of the AhpC dimer to a sulfenic acid intermediate that is converted to a disulfide bond when it is attached by Cys165 in the other subunit of the homodimer (10). The mechanism of reductive electron transfer in the AhpCF system involves a cascade of disulfide bond reduction steps. First, electrons are transferred from NADH via a flavin moiety in the COOH-terminus of AhpF to reduce a disulfide bond in that portion of the protein. The electrons from these reduced cysteines are then donated to a disulfide bond in the NH2-terminal domain of AhpF (11), from which they are transferred to the disulfide bond in AhpC, regenerating the reduced active form of the protein (12, 13).

Sequence analysis of the entire ahpCF locus from the wild-type, FÅ113, and nine independently isolated suppressor strains revealed that all of the mutants contained the same mutation, an insertion of three bases (TCT) after base 103 of theahpC coding region (Fig. 1B). The wild-type gene contains four direct repeats of TCT, whereas the mutants contain five. This change results in the addition of a single amino acid (Phe38), located only nine amino acids away from the active site Cys46 (Fig. 1B). We refer to the mutant protein as AhpC*.

Expansion of a repeated sequence, as is found in the ahpC* mutation, is usually a very frequent event (14) and would explain the finding that all the suppressors represent the same mutational event. The conserved triplet repeat sequence in which theahpC* mutation occurs is present in the ahpC gene of several facultative anaerobic enteric bacteria [Salmonella, Shigella, and Klebsiella(Fig. 1C) (15)].

To examine whether the ahpC* triplet repeat expansion is responsible for the restoration of growth in the trxB,gordouble mutant, we generated plasmid constructs containing the genes from the ahpCF or ahpC*F operon (16). These were transformed into either thetrxB,gor strain (JL10), or into a derivative that has a deletion of both ahpC and ahpF (DR456). Only the ahpC* gene restored normal growth to thetrxB,gor strain, whereas expression of both ahpC* and ahpF was necessary for suppression in theahpCF null background (Table 1). These results show that the effect of AhpC* is dominant over the wild-type allele and that AhpF is necessary for the phenotype of this mutation. Further, they suggest that the mutation in ahpC* is a gain-of-function mutation, altering this peroxiredoxin so that it has a new function.

Table 1

Ability of plasmid-borne ahpCF to restore the growth defect of trxB,gor and trxB,gor,ahpCFmutant strains. Strains were grown on plates containing 0.2% arabinose (to allow the expression of the araBAD controlledtrxB gene), and the antibiotics to maintain the plasmids were restreaked onto plates containing only 0.2% glucose. Growth was scored after 24 hours at 37°C as the ability to form single colonies. The strain backgrounds used are JL10 (trxB,gor) and DR456, which was obtained by introducingahpCF::Km (deletion of codons 62 to 187 of ahpC and codons 1 to 406 of ahpF) into JL10. NZ, rich medium containing NZ amine.

View this table:

If the AhpC protein has a new function in the ahpC* mutant, does it still retain its function as a peroxidase? To address this question, we determined whether the mutation had altered the resistance of E. coli to alkyl hydroperoxides. Comparison of strains expressing the plasmid-borne wild-type ahpCF and mutant ahpC*F operons revealed that the mutant operon conferred little or no increased resistance to cumene hydroperoxide (17). We also looked at the oxidative stress response mediated by the transcription factor OxyR in vivo. OxyR is activated by high intracellular concentrations of hydrogen peroxide to turn on genes involved in peroxide defense (ahpCF, katG) and disulfide bond reduction (gor, grxA, trxC) (18, 19). If the AhpC* mutant eliminates peroxidase activity, the concentration of cytoplasmic hydrogen peroxide should increase, inducing the OxyR response (20). Indeed, strains containingahpC* showed the same elevated OxyR response as a strain deleted for the ahpC locus (21). Combination ofahpC* with a mutation inactivating the katG gene should then result in a maximal OxyR response. This is true of akatG strain that is deleted for the ahpCF locus (20). However, the double mutants were highly unstable; 75% of the mutants that had initially the desired ahpC*,katG genotype segregated to give colonies with an ahpC +, katG genotype (22). The instability of individual colonies suggests that there is a strong selective advantage to restore peroxidase activity in the ahpC*, katG strain background.

We also purified the wild-type and mutant AhpC and assayed their peroxidase activity in vitro in the presence of AhpF, using stopped flow spectrophotometry with hydrogen peroxide and cumene hydroperoxide as substrates (23). A sustained high rate of loss of absorbance at 340 nm (A 340) due to NADH consumption was observed only when AhpC and peroxide were included. In contrast, with AhpC*, a sustained A 340 loss was not observed with either cumene hydroperoxide or hydrogen peroxide. In a separate assay for disulfide reductase activity, AhpC* was able to reduce the substrate DTNB [5,5′-dithio-bis(2-nitrobenzoic acid)]; AhpC showed a similar activity (24). Thus, AhpC* has disulfide reductase activity in the presence of AhpF, but no peroxidase activity. Although the wild-type AhpC is capable of reducing the generic disulfide bond–containing substrate DTNB in vitro, unlike AhpC* it does not exhibit any activity in vivo contributing to the reduction of ribonucleotide reductase, even when overexpressed.

These results suggest that the mutant AhpC* protein, although it has acquired a new reductive function, has effectively lost its original ability to reduce peroxides. Because the mutant protein still functions as an electron acceptor from AhpF, the electrons necessary to restore growth to the trxB,gor strain must be derived from NADH.

The ahpC* mutation must have restored essential disulfide-reducing capacities to the cytoplasm. To determine whether AhpC* acts by reducing any of the known components of the cytoplasmic thiol-redox pathways, we introduced additional mutations into thetrxB,gor suppressor strain and determined their effect on the suppression. Both glutathione and glutaredoxin 1 were required forahpC* to restore growth to the trxB,gor strain; the thioredoxins are not required (Fig. 2A).

Figure 2

Thiol redox pathways in a strain lacking glutathione reductase and thioredoxin reductase. (A) Additional cellular factors are required for AhpC* to function as a disulfide reductase in vivo. Strains were constructed by generalized transduction of the mutant alleles into atrxB ,gor ,ahpC* strain expressing trxB from plasmid under the control of thearaBAD promoter. Growth on rich medium with 0.2% glucose was scored after 1 day as the ability to form single colonies. (B) Model of the electron flow in a trxB,gordouble mutant. The mutant form of the peroxiredoxin AhpC* is required to restore growth, because it can donate electrons to oxidized cytoplasmic substrates, such as ribonucleotide reductase. The effect depends on the presence of glutathione and glutaredoxin 1 (Grx1) as well as AhpF. The channeling of electrons into the glutathione pathway has not been shown biochemically (gray arrow). Oxidized thioredoxins are presumably involved in disulfide bond formation in the cytoplasm.

So far, we have been unable to detect significant glutathione or glutaredoxin reductase activities using the AhpC*F enzyme system in vitro. Nevertheless, the in vivo data suggest that electrons are being channeled into the glutathione pathway from the AhpC*F complex, thus replacing NADPH with NADH as the source of the electrons. The most likely explanation for these findings is that the AhpC* protein has been altered so that it now uses its disulfide reductase activity to partially replace glutathione reductase in the double mutant (Fig. 2B).

Even though the AhpC*F system sustains sufficient disulfide-reducing capacity to allow cells to grow, the cytoplasm remains oxidizing and favors the formation of disulfide bonds in proteins. The reducing ability of the thioredoxin branch is not restored in the suppressor strains, and it is therefore likely that the oxidized forms of the thioredoxins actively catalyze disulfide bond formation in the cytoplasm, as they do in a trxB mutant strain (25).

The conservation of the triplet repeat sequence across species and the high frequency of genetic interconversion between the peroxidase and the apparent glutathione- glutaredoxin reductase forms of alkyl hydroperoxidase raise the possibility that this change represents an evolutionarily significant advantage for these bacteria. The ability to switch back and forth between the peroxidase activity and the activity that channels electrons into the glutathione pathway may allow a fraction of the bacterial population to survive either oxidative stress or disulfide stress, respectively. Disulfide stress would occur when bacteria encounter oxidative agents, other than peroxides, that promote disulfide bond formation (1,26). Alternatively, the switch may be important for growth conditions in which bacteria become depleted of NADPH. The switch would allow the cells to use NADH via the AhpC*F system as the normal source of electrons for glutathione. The frequency of the ahpC* mutation is comparable to other cases of evolutionarily significant genetic switches, including the activation of cryptic β-glucosidases (27) and the alternation between flagella types in phase variation (28–30). The AhpC example differs from these in that the genetic event is a reversible interconversion of a single enzyme between two forms that have substantially different activities.

In humans, triplet repeat expansions are deleterious and have been causally linked to certain neurodegenerative disorders (31,32). In contrast, the repeat expansion in AhpC represents a genetic switch leading to a protein with a new and advantageous enzymatic function.

  • * To whom correspondence should be addressed. E-mail: jon_beckwith{at}


View Abstract

Navigate This Article