A glutathione-based system for defense against carbonyl stress in Haemophilus influenzae
- First Online:
- Cite this article as:
- Kidd, S.P., Jiang, D., Tikhomirova, A. et al. BMC Microbiol (2012) 12: 159. doi:10.1186/1471-2180-12-159
adhC from Haemophilus influenzae encodes a glutathione-dependent alcohol dehydrogenase that has previously been shown to be required for protection against killing by S-nitrosoglutathione (GSNO). This group of enzymes is known in other systems to be able to utilize substrates that form adducts with glutathione, such as aldehydes.
Here, we show that expression of adhC is maximally induced under conditions of high oxygen tension as well as specifically with glucose as a carbon source. adhC could also be induced in response to formaldehyde but not GSNO. An adhC mutant was more susceptible than wild-type Haemophilus influenzae Rd KW20 to killing by various short chain aliphatic aldehydes, all of which can be generated endogenously during cell metabolism but are also produced by the host as part of the innate immune response.
These results indicate that AdhC plays a role in defense against endogenously generated reactive carbonyl electrophiles in Haemophilus influenzae and may also play a role in defense against the host innate immune system.
KeywordsStress responseH. influenzaeReactive aldehydes
Haemophilus influenzae is a γ-Proteobacterium adapted to the human host. It exists as a commensal in up to 80% of the healthy population. It survives in the nasopharnyx, and can spread to other sites within the body and cause disease . H. influenzae requires a number of exogenous cofactors for growth including cysteine for the production of glutathione (GSH) . In addition to its role in defence against oxidative stress [2, 3] GSH forms adducts with toxic electrophilic molecules. Glutathione-dependent alcohol dehydrogenase (AdhC) catalyses the NAD+-dependent oxidation of a GSH-formaldehyde adduct [4, 5]. Expression of adhC in a variety of bacteria is associated with defense against formaldehyde stress and is correspondingly regulated in the response to the presence of formaldehyde . It is also established that AdhC catalyses the NADH-dependent reduction of S-nitrosoglutathione (GSNO), a molecule generated during the conditions of nitrosative stress that occurs in human cells in response to invading pathogens such as H. influenzae. Unlike other aldehyde dehydrogenase enzymes AdhC cannot use ethanol or formaldehyde directly, but uses the adducts which spontaneously form with GSH (hence the nomenclature, GSH-dependent formaldehyde dehydrogenase) . AdhC from different sources is known to catalyse the concurrent oxidation of formaldehyde and reduction of GSNO [8, 9]. We have previously observed that AdhC of H. influenzae does function in GSNO metabolism .
H. influenzae does not use methanol as a carbon source (the by-product of which is formaldehyde) and cannot assimilate formaldehyde. Therefore, a source of formaldehyde substrate for AdhC from the host environment is not obvious; however, bacteria do encounter a variety of aldehydes. Neutrophils use myeloperoxidase to produce glycoaldehyde and acrolein at sites of inflammation while short chain sugar aldehydes such as glyceraldehyde 3-P and erythrose 4-P are produced endogenously as intermediates of the bacterial metabolism of sugars . All these short chain aldose sugars mentioned can undergo auto-oxidation to more toxic dicarbonyl species . In this paper we report the effect of reactive carbonyl species on growth of H. influenzae. This provides a new insight into the physiological role of AdhC in non-methylotrophic bacteria.
Bacterial strains and growth conditions
H. influenzae strains were cultured on Brain heart infusion (BHI) medium or chemically defined media (CDM). BHI was prepared with 3.7% (wt/vol) BHI Powder (Oxoid). For solid medium, 1.5% (wt/vol) agar powder was added. Medium was sterilized by autoclaving at 121°C for 20 min. Levinthal blood (10% [wt/vol]) was added for solid medium. BHI broth required NAD (2 μg/ml) and 10 μg/ml hemin solution (0.1% [wt/vol] hemin, 0.1% [wt/vol] L-histidine, 4% [vol/vol] triethanolamine). Solutions for media were sterilized individually, either by filter sterilizing or by autoclaving. The solutions were mixed under sterile conditions. CDM was prepared mostly as described by Coleman et al. . The exception to this protocol is the use of RPMI 1640 without glucose (Invitrogen) and the addition of 0.4% of the appropriate sugar or carbon source. In standard procedures the final pH of CDM was adjusted to 7.56 by NaHCO3. CDM was sterilized by filter sterilization through a 0.22-μm filter.
Reverse transcriptase PCR
RNA was extracted from H. influenzae Rd KW20 at the time points 3 h, 5.5 h and 8 h during growth cycle by using a QIAGEN RNeasy minikit (QIAGEN). RNA was quantified using an A260 reading and then checked for DNA contamination by PCR; no product was detected. RNA was further treated to remove any residual DNA by using Promega DNase (Promega). The reverse transcriptase (RT) reaction was performed using a QIAGEN Omniscript reverse transcriptase kit. The products of this reaction were used in a multiplex PCR with primers for the 16 S rRNA gene:
16SFOR: 5’-AGTCCACGCCCTAAACGATGT-3’ and
16SREV: 5’-TACTCCCCAGGCGGTCAAT-3’; and primers from estD to adhC:
Est1: 5’-CCCAAGGCTGCTCGGTC-3’ and
PCR was carried out with New England Biolabs Taq polymerase using an initial 96°C for 10 min followed by 30 cycles of 96°C for 45 s, 54°C for 45 s, and 72°C for 30 s and a final elongation step of 72°C for 10 min.
Cells were cultured in rich media (BHI, Oxoid UK) or chemically defined media (CDM). Unless otherwise stated, analysis of the growth of H. influenzae strains was carried out using CDM. For rich media cells were grown on BHI medium supplemented with NAD (2 μg/ml) and 10 μg/ml hemin solution. Overnight growth cultures were inoculated into 5 ml of media and grown until log phase prior to the assay. The OD600 of the cultures were measured (Hitachi U-3000 spectrophotometer) to standardize the amount of cells inoculated. Routinely 1 ml was inoculated into 50 ml of CDM in a 250-ml conical flask. For analysis of the effects of oxygen supply to the cells, cultures were grown in 250 ml conical flasks with 25 ml, 75 ml and 150 ml medium. This has been previously used and shown to provide the oxygen transfer coefficents (kLa) values of 87.4 h-1 (high), 27.8 h-1 (medium) and 11.5 h-1 (low) respectively [14, 15].
Different specific concentrations of stress agent were added to the medium. Cultures were incubated aerobically at 37°C with shaking at 190 rpm. OD600 measurements were taken at different time points for 10 h. The assays were done in triplicate. Assay results were represented as growth curves over this period or, for clarity for the large set of clinical isolates, as percentages of survival at this time point.
GSNO reductase enzyme assays
NADH-dependent GSNO reductase activity was measured as previously described . Fresh overnight cultures of H. influenzae were inoculated into 100 ml of CDM in 500 ml conical flasks and grown aerobically at 37°C with shaking at 190 rpm until an OD600 measurement between 0.4 and 0.6 was obtained. The cells were harvested (5,000 × g at 4°C for 10 min) and washed twice with 0.1 M phosphate buffer (pH 7.0) before resuspending in 2 ml of phosphate buffer. The suspension was frozen at −80°C, thawed at room temperature, given a brief vortexing, and frozen again at −80°C. This freeze-thaw process was performed four more times before the cells were centrifuged at 13,000 × g at 4°C for 15 min. The final supernatant (cell extract) was used for assays. The total protein concentration of the supernatant was determined spectrophotometrically using the formula protein (mg/ml) (1.55 × A280) – (0.76 × A260) 19. GSNO reductase activity was expressed as μmol of NADH oxidized per minute per mg of total protein. The assays were done in triplicate.
AdhC is expressed under aerobic conditions and required for aerobic growth in H. influenzae
AdhC is required for defense against reactive aldehydes
The growth rates of Rd KW20 andadhC; with 2 mM glycolaldehyde and 1 mM 1,2-diaminobenzene (DAB)
Growth rate (doubling per hour)
1.10 ± 0.14
Rd KW20 + glycolaldehyde
0.80 ± 0.37
Rd KW20 + glycol. + DAB
1.47 ± 0.35
0.79 ± 0.34
adhC + glycolaldehyde
0.20 ± 0.10
adhC + glycol. + DAB
0.51 ± 0.27
AdhC is induced by formaldehyde but not by GSNO
The expression of adhC is regulated by the MerR family transcription factor NmlRHI . Regulators of this family generally function as both weak repressors, and as activators when in the presence of their cognate stress effector. We have previously reported that expression of GSNO reductase activity in H. influenzae requires both adhC, the structural gene encoding the enzyme activity, as well as its regulator nmlRHI under growth conditions with no exogenous stress. Mutant strains of H. influenzae in which the adhC or nmlRHI genes have been inactivated do not express detectable GSNO reductase activity . A reasonable conclusion was that under these conditions NmlRHI is in its activator conformation and therefore endogenously generated molecules are the cognate “stress” for which it responds. Attempts to identify the cognate ligand or the environmental stimuli, which acts to switch NmlRHI, to an activator form have been unsuccessful.
In mammalian systems AdhC functions in detoxification of a range of reactive aldehyde species as well as in defense against GSNO. Our results suggest that there may be a similar role for AdhC in H. influenzae. Glycoaldehyde is produced from serine by the action of myeloperoxidase . This is one of several types of reactive aldehydes that are produced by activated neutrophils at sites of inflammation. The toxicity of glycoaldehyde arises from the oxidation of its ene-diol tautomer to form a highly reactive α, β-dicarbonyl species. This reaction requires oxygen or superoxide, consistent with AdhC activity being highest with increased oxygen levels and during the highest periods of metabolic reactions. Our observations are also consistent with previous in silico analyses analysis of gene expression in H. influenzae which showed that adhC expression was highest under aerobic conditions during which glucose would be metabolised mainly via the pentose phosphate pathway [18, 19]. Our data is consistent with these results as AdhC was required for growth with glucose as the carbon source under high oxygen culture conditions (Figures 1 and 2). Glyceraldehyde 3-phosphate and erythrose 4-phosphate are both intermediates in this pathway. It has been noted that the equilibrium constant for the aldolase reaction means that in glycolysis the concentration of glyceraldehyde 3-phosphate is kept very low. This may not be the case when the pentose phosphate pathway is the dominant glucose oxidation pathway that occurs under conditions of high oxygen tension [18, 19].
Recently, it is has been observed that an NmlR homologue in Bacillus subtilis (AdhR) activates gene expression in response to methylglyoxyl and formaldehyde . One cysteine (C54) was shown to be required for activation of gene expression and this led Antelmann and co-workers  to propose that Bacillus AdhR is activated by S-alkylation of this cysteine residue. AdhR contains a single conserved cysteine, as in the NmlRsp transcription factor from Streptococcus pneumoniae . In H. influenzae we only observed induction of adhC by NmlRHI upon addition of formaldehyde. In contrast to the situation in B. subtilis and S. pneumoniae, NmlRHI possesses three conserved cysteine residues and is closely related to the NmlR regulators from Neisseria species . Thus, there may be significant differences in the mechanism of the sensing of reactive carbonyl compounds by transcription factors of the NmlR family.
Uniquely, H. influenzae utilizes an AdhC enzyme for the concurrent roles of protection against an exogenous stress (GSNO) as well as the endogenously generated and harmful reactive aldehydes. AdhC is essential for H. influenzae growth under conditions of high oxygen and with glucose as the carbon source. This role is through the detoxification of different reactive carbonyl compounds.
We acknowledge support from program grants 284214 from the National Health and Medical Research Council of Australia to M. P. J. and A. G. M. and DP0986578 from the Australian Research Council to A. G. M.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.