Functional genomics of pH homeostasis in Corynebacterium glutamicum revealed novel links between pH response, oxidative stress, iron homeostasis and methionine synthesis.

BACKGROUND: The maintenance of internal pH in bacterial cells is challenged by natural stress conditions, during host infection or in biotechnological production processes. Comprehensive transcriptomic and proteomic analyses has been conducted in several bacterial model systems, yet questions remain as to the mechanisms of pH homeostasis.
RESULTS: Here we present the comprehensive analysis of pH homeostasis in C. glutamicum, a bacterium of industrial importance. At pH values between 6 and 9 effective maintenance of the internal pH at 7.5 +/- 0.5 pH units was found. By DNA microarray analyses differential mRNA patterns were identified. The expression profiles were validated and extended by 1D-LC-ESI-MS/MS based quantification of soluble and membrane proteins. Regulators involved were identified and thereby participation of numerous signaling modules in pH response was found. The functional analysis revealed for the first time the occurrence of oxidative stress in C. glutamicum cells at neutral and low pH conditions accompanied by activation of the iron starvation response. Intracellular metabolite pool analysis unraveled inhibition of the TCA and other pathways at low pH. Methionine and cysteine synthesis were found to be activated via the McbR regulator, cysteine accumulation was observed and addition of cysteine was shown to be toxic under acidic conditions.
CONCLUSIONS: Novel limitations for C. glutamicum at non-optimal pH values were identified by a comprehensive analysis on the level of the transcriptome, proteome, and metabolome indicating a functional link between pH acclimatization, oxidative stress, iron homeostasis, and metabolic alterations. The results offer new insights into bacterial stress physiology and new starting points for bacterial strain design or pathogen defense.

Background

Bacteria have to cope with changing environmental conditions in order to survive in different habitats. A key determinant is the pH value because it has an impact on the solubility of nutrients and trace elements, like iron, and on the cellular metabolism in general. Most bacteria maintain a neutral or slightly alkaline internal pH when subjected to acidic or alkaline conditions [1]. This pH homeostasis is important for the function of all cellular enzymes as well as their stability. The pH gradient across the membrane (ΔpH) can be very high at low pH values or can even be reversed at high external pH values. Beside the electrical membrane potential ΔΨ, ΔpH represents the chemical constituent of the proton motive force (pmf) which is essential for generation of ATP by the F1F0ATPase.

Corynebacterium glutamicum is a work horse in biotechnology for the production of glutamate and lysine and a model strain for the investigation of its pathogenic relatives C. diphtheriae, C. jeikeium or mycobacteria [2-4]. Its sensitivity towards acidic pH was noticed, but regarding the mechanism of pH homeostasis and the components participating in the acclimatization process, little is known.

Several general mechanisms are known to be important during pH acclimatization in bacteria. Under alkaline conditions, sodium proton antiporters like MdfA and NhaA mediate resistance in E. coli [5,6]. However, in C. glutamicum an MdfA homologue is missing and the participation of further sodium proton antiporters in the pH response is unknown. Arginine, lysine, and glutamate decarboxylases are predominant for acid tolerance in many bacteria. During decarboxylation of amino acids CO2 is liberated and affects the internal pH by formation of bicarbonate. The decarboxylated product is excreted in exchange for the corresponding amino acid [7]. In C. glutamicum genes encoding homologous proteins of the AdiCA (arginine:agmatine antiporter and arginine decarboxylases), GadABC (glutamate decarboxylase AB and glutamate:gamma-aminobutyric acid antiporter), and CadAB (lysine decarboxylases and lysine:cadaverine antiporter) systems are absent [8]. In Gram-positive bacteria like Bacillus subtilis, or lactic acid bacteria the arginine deiminase pathway is important for acid stress response [7]. By arginine utilization ammonium is liberated which induces the alkalization of the cytoplasm as well as the periplasm. In C. glutamicum, however, a homolog of the arcA gene is missing (Kalinowski et al., 2003). The F1F0ATPase was found to function as a proton exporter under acidic conditions in Enterococcus hirae and its role in pH homeostasis in other bacteria was discussed [7,9]. In C. glutamicum the atp gene cluster encoding the F1F0ATPase was found to be transcriptionally induced at alkaline pH under the control of the sigma factor SigH and subsequent studies indicated that the expression is correlated with growth rate rather than the pH value of the medium [10,11]. Furthermore a putative cobalt transporter encoded by the gene cg1447 was found to be important under alkaline conditions [12].

Further studies on acidic pH response revealed the participation of multiple cellular processes in acclimatization of various bacteria. Among them are the activation of the protein folding and stabilization machinery [13], the induction of iron uptake systems [11], or metabolic adaptations including the induction of the methionine pathways [14]. Furthermore, observations were made indicating the occurrence of oxidative stress at low pH values [15]. In conclusion, a shift of the external pH seems to act on various levels and affects a multiplicity of cellular processes finally limiting growth at non optimal pH conditions.

In the present study we identified limitations of pH homeostasis that restrict growth at non-optimal pH conditions in C. glutamicum. We excluded short term effects and focused on the steady state regulation in exponentially growing cells under neutral, acidic, and alkaline conditions. Applying transcriptome studies, soluble as well as membrane proteome analyses we found that C. glutamicum cells are exposed to oxidative stress at low pH and concomitantly iron starvation response is induced leading to the alteration of a variety of metabolic pathways which was reflected by differential metabolite pattern as well. We present comprehensive data showing that a decrease of the external pH affects particular cellular processes at various levels which finally limit growth of C. glutamicum under acidic conditions.

Results

Effective pH homeostasis is correlated with optimal growth in C. glutamicum

We first quantified the efficiency of pH homeostasis in C. glutamicum. We performed growth assays in shaken micro titer plates (MTP) in minimal medium in presence of optimized buffers at a pH of 4 to 11 with subsequent determination of growth rates. As seen in Fig. 1, optimal growth rates were observed at a pH of 7 to 8.5. At an external pH below 6 and above 9 growth rates decreased drastically and at a pH of 4 as well as 10.5 and 11 no significant growth was observed.

Figure 1

Comparison of growth rate and internal pH of . Growth experiments were performed in selected buffer systems in shaken microtiter plates (pH 4-5.5 black circles: Homopipes, pH 5.5-6.5 white triangles down: Mes, pH 6.5-7.75 black squares: Mops, pH 7.5-8.75 white diamonds: Hepps, pH 8.75-9.75 black triangles: Ches, pH 9.75-11 white hexagons: Caps). Determination of the cytoplasmic pH (white triangles top) was performed in a separate experiment by measuring the distribution of the radioactive probes benzoic acid (pH 4 to 7.5) and methylammonium (pH 7.5-11, see Methods section).

Subsequently, we determined the internal pH of C. glutamicum cells grown at pH 7.5 after exposure to different external pH values. At a pH of 7.5 the internal pH value was found to be 7.5. This value was kept constant (± 0.5 pH units) after lowering the external pH down to 6 or increasing the pH up to 9. Below or above these external pH values the internal pH decreased respectively increased much faster in response to an external pH shift. We concluded that C. glutamicum can perform effective pH homeostasis in a range of external pH values from 6 to 9.

The failure of effective pH homeostasis at low or high external pH values could result from an impaired energy metabolism. The pH gradient across the cytoplasmic membrane (ΔpH) is important for generation of the proton motive force (pmf = ΔΨ - 2.3RT/F × ΔpH) which is essential for ATP synthesis by the F1F0ATPase. In order to prove whether the pmf is affected in C. glutamicum we determined ΔpH as well as the membrane potential ΔΨ in cells exposed to different pH values and calculated the pmf. The results are shown in Fig. 2. As expected, the pH gradient is zero at pH 7.5. At lower external pH values ΔpH increased up to 60 mV, whereas at higher external pH values, ΔpH was found to be reverted and decreased to -80 mV. The values for ΔΨ were found to be 110 mV at an external pH of 4.5, increased to 200 mV at pH 8, and at very high pH values (pH 10.5) 245 mV were measured. As a consequence of the decreasing ΔpH and increasing ΔΨ values the resulting pmf was kept relatively constant at a surprisingly broad pH range of 4.5 to pH 11, varying between 150 to 200 mV. At the most acidic pH of 4 the membrane potential ΔΨ collapsed and the resulting pmf value was 40 mV only.

Figure 2

The pH dependent bioenergetic homeostasis in . Membrane potential (triangles) and pH gradient (circles) across the cytoplasmic membrane of C. glutamicum exposed to different external pH values and values for the resulting proton motive force (squares). All values are given in mV.

Transcriptome and proteome analyses of pH acclimatization in C. glutamicum

In order to unravel components and processes involved in pH homeostasis we performed transcriptome analyses by DNA microarrays and proteome studies by 1D-nLC-ESI-MS/MS. For this purpose, two independent batch fermentations were carried out at a pH of 6, 7.5, and 9 in stirred bioreactors under continuous pH control and samples were drawn during the exponential phase in order to focus on the steady state pH homeostasis and to prevent additional perturbations by short term responses. We observed growth rates of 0.14 ± 0.01 at pH 6, 0.32 ± 0.02 at pH 7.5, and 0.14 ± 0.01 at pH 9. Cells were harvested and immediately frozen for metabolic inactivation. Transcriptome patterns were analyzed by co-hybridization of cDNA derived from cells grown at pH 6 vs. pH 7.5 and from cells grown at pH 9 vs. pH 7.5. The data analysis was performed as described previously [16], using a m-value (log2 of the relative change in the respective mRNA ratio) cut-off of ± 1 which corresponds to transcription changes equal or greater than twofold. For the comparative proteome analysis we performed identification and quantification of peptides in the enriched soluble and membrane fractions as well as in the cell envelope fraction. Relative quantification of protein abundance and its change was performed by the spectral counting technique and the results are also given as log2 values [17]. The complete set of data is available as supplementary material.

The transcriptome analyses revealed 42 genes with increased expression at pH 9 in comparison to pH 7.5 (Table 1). In the respective proteome studies, 19 corresponding proteins were found with an increased peptide number in at least one protein fraction whereby four of them were present at significantly higher levels at pH 9 in comparison to pH 7.5 (Table 1). For 39 genes we found a decreased mRNA level at pH 9 in comparison to pH 7.5 whereby for 26 corresponding proteins a lower content was indicated by lower peptide numbers in at least one fraction and for four of them a significant lower abundance at pH 9 was observed (Table 1). For 10 genes differentially expressed at pH 9 we did not find a corresponding change of the peptide number and for 18 proteins we did not find any corresponding peptides at all indicating a low abundance of these proteins in C. glutamicum cells. The comparison of cells grown at pH 6 and pH 7.5 revealed higher mRNA pools for 88 genes whereby for 49 corresponding proteins (for 10 of them significantly) increased peptide numbers were found (Table 2). A lower mRNA content at pH 6 was found for 91 genes whereby for 52 corresponding proteins (for 16 of them significantly) decreased peptide numbers were found at least in one protein fraction at acidic pH (Table 2). In case of 17 proteins alterations of mRNA and protein content do not match at acidic pH. For 35 genes differentially expressed at pH 6 in comparison to pH 7.5 no peptide was found at all. In summary, we found many overlaps of transcriptome and proteome data for C. glutamicum grown at different pH values.

Table 1

Differential expression pattern at pH 9 in comparison to pH 7.5

Genes induced at pH 9						Transcriptome4		Proteome5									Regulators6
								cytoplasma			envelope			membrane
No	gene ID1	op2	gene	function	TMH3	pH 6	pH 9	6	7.5	9	6	7.5	9	6	7.5	9
1	cg0077			Hypothetical protein	0	n.d.	1.04	-0.6	-	-	-	-	-	-0.9	-	-
2	cg0105			Hypothetical protein Cgl0077	0	-0.83	1.65	-	-	-	-0.4	-	-	-	-	-
3	cg0310		katA	Catalase	0	-0.62	1.73	4.2	5.1	6.4	-	-	0.0	-	-	-0.6	RipA
4	cg0444		ramB	Regulator of acetate metabolism	0	0.76	1.11	-	-	-0.8	-	0.7	3.5	-	-	-	RamA, RamB
5	cg0445	c	sdhCD	Succinate dehydrogenase component CD	5	-1.31	1.31	-	-	-	1.8	1.4	2.4	3.2	4.0	5.0	RipA, DtxR, RamA*, RamB*
6	cg0446	c	sdhA	Succinate dehydrogenase A	0	-1.08	1.47	1.0	2.7	1.7	6.7	7.0	9.0	8.5	9.0	9.9	RipA, DtxR, RamA*, RamB*
7	cg0447	c	sdhB	Succinate dehydrogenase B	0	-1.43	1.14	-	-0.4	-	4.6	5.0	6.7	3.5	3.4	5.3	RipA, DtxR, RamA*, RamB*
8	cg0448	c		Conserved hypothetical membrane protein	2	-0.98	1.18	-	-	-	-0.4	0.8	1.7	-	-	-	RipA*, DtxR, RamB*
9	cg0778		fecC	putative iron or siderophore ABC-type transporter, permease component	8	-0.05	1.07	-	-	-	-	-	-	-	-	-
10	cg0825			Dehydrogenases with different specificities	0	-0.19	1.07	4.1	4.6	5.0	-0.4	-0.3	1.8	-0.7	-	-0.4
11	cg0858			putative gamma subunit of the nitrate reductase	1	-0.28	1.22	-	-	-	-	-	-	-	-	-
12	cg1095			Hypothetical protein	0	0.32	1.28	-	-	-	-	-	-	-	-	-
13	cg1136			Hypothetical protein	0	0.15	1.02	1.6	0.5	2.2	-	-	-	-	-	-
14	cg1206			PEP phosphonomutase or related enzyme	0	0.38	1.05	2.6	2.8	2.7	-	-	-	-	-	-
15	cg1292	r		Flavin-containing monoxygenase 3	0	n.d.	3.87	-	-	2.4	-	-	4.4	-	-	-0.6
16	cg1293	r		Hypothetical membrane protein	3	n.d.	1.71	-	-	-	-	-	-	-	-	-
17	cg1312			Hypothetical membrane protein	4	n.d.	1.32	-	-	-0.2	1.9	2.1	4.7	5.3	4.4	6.3
18	cg1344	s	narG	Nitrate dehydrogenases 2 (Fe4S4 containing)	0	1.36	1.94	-	-	-0.6	-	-	-	-	-	-0.4	RipA, GlxR*
19	cg1345	s	narK	putative nitrate/nitrite transporter	12	n.d.	1.66	-	-	-	-0.4	-	3.0	1.5	-0.6	3.0	RipA, GlxR*
20	cg1695			SAM dependent methyltransferase	0	n.d.	1.72	-	-	-	-	-	-	-	-	-
21	cg1737		acn	Aconitase A	0	-0.58	1.66	4.4	5.3	6.2	4.4	5.0	4.9	2.2	2.6	2.5	RipA, RamA*, RamB*
22	cg1759	x	sufX	Predicted metal-sulfur cluster biosynthetic enzyme	0	0.94	1.34	-	-	-0.8	-0.4	-	-	-	-	-	SufR, SigM
23	cg1761	x	nifS2	Cysteine desulfhydrase, Selenocysteine lyase	0	0.44	1.47	2.3	2.0	3.2	-0.4	-	-	-	-	-	SufR, SigM
24	cg1762	x	sufC	Suf related ABC-type transporter, ATPase component	0	0.81	1.51	3.7	3.4	4.6	3.4	4.4	4.1	0.1	2.9	2.1	SufR, SigM
25	cg1764	x	sufB	Suf related ABC-type transporter SufB, permease component	0	0.93	1.81	2.8	3.4	5.0	3.4	3.4	3.8	0.1	0.4	1.6	SufR, SigM
26	cg1765	x	sufR	transcriptional regulator SufR	0	0.76	1.53	-	-	-	-	-	-	-	-	-	SufR, SigM
27	cg1790		pgk	Phosphoglycerate kinase	0	0.15	1.07	4.8	4.2	5.0	5.0	4.8	5.0	2.5	2.7	3.3	SigB*
28	cg1884			putative copper export protein CopC	2	-0.39	1.12	-	-	-	-	-	-	-	-	-	LexA
29	cg1904			ABC-type transporter, permease component	6	0.09	1.12	-	-	-	-	-	-	-	-0.6	1.2
30	cg2191			putative 3-demethylubiquinone-9 3-methyltransferase	0	n.d.	1.02	-	-	-	-	-	-	-	-	-
31	cg2274	ab		Hypothetical protein	0	-0.17	1.15	-0.6	0.4	1.8	-	-	-	-	-	-
32	cg2275	ab		Hypothetical protein	0	-0.28	1.03	1.0	2.4	1.3	-	-	-	-0.9	-	-
33	cg2320			ArsR type transcriptional regulator	0	-0.26	1.25	-	-	-	-	-	-	-	-	-
34	cg2572			Hypothetical protein	0	0.45	1.44	-	-	-	-	-	-	-	-	-
35	cg2636		catA1	Protocatechuate 3,4-dioxygenase beta subunit	0	n.d.	3.09	-	-	1.6	-	-	-	-	-	-	RipA, GlxR*
36	cg2736		bcp	putative 3-demethylubiquinone-9 3-methyltransferase	0	0.12	1.12	0.5	-0.5	0.4	-	-	-	-	-	-
37	cg2782		ftn	Ferritin-like protein	0	0.56	1.29	-	-	0.8	-	-	-	-	-	-	DtxR
38	cg2853			Hypothetical protein	0	0.41	1.4	0.6	-0.5	1.8	-	-	-	-	-	-
39	cg3117	ai	cysX	Hypothetical protein	0	3.5	1.08	-	-	-	-	-	-	-	-	-	DtxR, McbR*, CysR*
40	cg3118	ai	cysI	Sulfite reductase hemoprotein beta-component	0	3.5	1.06	0.6	1.5	2.4	3.5	2.1	-0.3	-	-	-	DtxR, McbR, CysR*
41	cg3236			MFS-type transporter	0	1.27	1.01	-	-	-0.8	-	-	-	-	-	-
42	cg3331		ogt	Methylated DNA-protein cysteine methyltransferase	0	0.2	1.34	-	-0.8	-	-0.1	-	1.3	-	-	-
Genes repressed at pH 9
1	cg0071			Metallo-beta-lactamase superfamily	0	n.d.	-1.01	-	-	-	-	-	-	-	-	-
2	cg0133		abgT	P-aminobenzoyl-glutamate transporter	13	-0.86	-1.02	-	-	-	-	-	-	0.3	3.7	2.1
3	cg0202		iolD	Putative acetolactate synthase	0	n.d.	-1.13	-	-	-	-	-	-	-	-	-
4	cg0265			putative ABC-type molybdate transporter, ATPase component	0	-0.22	-1.08	-	-	-	-	-	-	-	-	-
5	cg0303		leuA	2-isopropylmalate synthase	0	-0.97	-1.06	4.3	4.7	4.5	-	-0.3	-	-	-	-
6	cg0404	b		putative nitroreductase	0	1.77	-1.50	3.4	2.5	0.9	0.9	-	-	-0.9	-	-
7	cg0467			ABC-type cobalamin/Fe3+-siderophores transporter, periplasmic component	0	-0.50	-1.10	-	0.2	-	-	0.7	-	0.2	1.0	-0.4	DtxR
8	cg0527			ArsR type transcriptional regulator	0	-0.32	-1.13	-	-	-	-	-	-	-	-	-	DtxR
9	cg0589			ABC-type cobalamin/Fe3+-siderophores transporter, ATPase component	0	-0.15	-1.01	-	-	-	1.2	-	-	0.1	-	-	DtxR
10	cg0623			putative ABC-type cobalt transporter, permease components	8	-1.05	-1.35	-	-	-	-	-	-	-	-	-
11	cg0624	h		Hypothetical membrane protein	2	-1.09	-1.09	-	-	-	-	-	-	1.3	1.6	-0.4
12	cg0625	h		putative terpenoide cylase	1	-1.28	-1.04	-	2.2	0.0	2.8	3.6	2.2	-	-0.6	-0.4
13	cg0723		crtE	Geranylgeranyl-pyrophosphate sythase	0	0.66	-1.01	-	-	-	-	-0.2	-	-	-	-
14	cg0748			ABC-type Fe3+-siderophores transporter, periplasmic components	0	-1.07	-2.14	-0.3	1.3	-	1.2	0.7	-	0.1	-0.6	-	DtxR
15	cg0756		cstA	putative carbon starvation protein A	16	-0.83	-1.56	-	-	-	1.9	2.1	1.0	2.1	2.8	-0.6
16	cg0767	m		Siderophore-interacting protein	0	-0.34	-2.99	0.5	-0.8	-	-	-	-	-	-	-	DtxR, RamB*
17	cg0768	m		ABC-type cobalamin/Fe3+-siderophores transporter, ATPase component	0	-0.17	-2.03	-	-	-	1.7	1.7	-	-	-0.6	-	DtxR, RamB*
18	cg0924	n		ABC-type cobalamin/Fe3+-siderophores transporter, periplasmic component	0	-0.60	-2.91	4.8	5.2	2.7	3.6	4.1	0.7	5.1	4.5	3.4	DtxR
19	cg0927	n		ABC-type cobalamin/Fe3+-siderophores transporter, permease component	9	-0.10	-2.08	-	-	-	1.2	-0.2	0.0	3.0	3.3	-	DtxR
20	cg0928	n		ABC-type cobalamin/Fe3+-siderophores transporter, ATPase component	0	-0.14	-2.35	-	-0.8	-	3.3	4.0	-	1.8	1.0	-	DtxR
21	cg0952	o	mctB	putative monocarboxylate transporter subunit	2	-2.55	-1.32	-	0.5	-	2.4	4.7	4.1	-	2.1	-0.6	RamA, RamB
22	cg0953	o	mctC	monocarboxylate transporter	13	-2.44	-1.01	-	3.0	-0.2	1.9	4.7	4.7	3.2	4.5	3.2	RamA, RamB
23	cg1091			Hypothetical protein	0	0.95	-1.18	-	-	-	-	-	-	-	-	-	SigM*
24	cg1167		metS	putative methionine transporter subunit	0	-1.04	-1.43	-	-	-	-	-	-	-0.9	-	-0.6
25	cg1290		metE	5-methyltetrahydropteroyltri-glutamate-homocystein-emethyltransferase	0	0.65	-1.06	8.2	6.8	6.4	8.0	8.0	6.5	6.2	6.4	4.2	McbR*
26	cg1365	t	atpH	F0F1 ATP synthase delta subunit	0	-1.26	-1.03	1.9	2.3	1.3	3.9	4.8	4.7	3.2	4.1	3.4	SigH*
27	cg1367	t	atpG	F0F1 ATP synthase gamma subunit	0	-1.71	-1.11	1.1	4.2	3.4	3.3	4.9	4.7	4.0	4.0	2.5	SigH*
28	cg1451		serA	Phosphoglycerate dehydrogenase	0	-0.22	-1.10	6.5	5.4	6.1	7.7	7.5	7.2	5.3	5.3	4.2
29	cg1537		ptsM	PTS system mannose-specific EIIBCA component	10	-0.39	-1.06	2.6	3.5	1.8	5.9	5.6	6.0	6.6	6.4	6.1	RamB*, GlxR*
30	cg1859			Putative secreted protein	0	0.04	-1.34	-	1.7	-	-	3.6	1.5	0.1	5.0	-0.4
31	cg1930	z		Putative secreted hydrolase	0	-1.18	-3.26	-	-	-	-	-	-	-	-	-	DtxR
32	cg1931	z		Hypothetical protein	0	-0.88	-2.55	-	-	-	-	-	-	-0.7	-0.6	-	DtxR
33	cg2283			Hypothetical protein	0	0.00	-1.51	4.2	4.3	3.4	-	-0.2	-	-	-	-
34	cg2336			Putative secreted protein	0	-0.09	-1.31	-	-	-	-	2.1	-	3.6	3.9	1.9
35	cg2445		hmuO	putative heme oxygenase	0	-0.93	-1.19	-0.2	2.5	-	-	-0.2	-	-	-0.6	-	DtxR
36	cg2560	ad	aceA	Isocitrate lyase	0	-2.73	-2.19	-	4.9	2.4	-	-0.3	-	-	-	-	RamA, RamB
37	cg2962			Uncharacterized enzyme involved in biosynthesis of extracellular polysaccharides	0	0.45	-1.18	-0.3	0.2	-	-	-	-	-	-	-
38	cg3156		htaA	secreted protein implicated in iron acquisition and transport	0	-2.50	-2.76	-	-	-	-	-	-0.3	2.7	5.7	-	DtxR
39	cg3254			Hypothetical membrane protein	3	0.07	-1.29	-	-	-	-	-	-	-0.5	-	-

Genes for which an increased or decreased mRNA level was found at pH 9 in comparison to pH 7.5. The gene locus tag, organisation in operons, the gene name, the (proposed) function of the protein as well as the predicted number of transmembrane helices are given. The results of the transcriptome analysis are given as induction factor at pH 6 and pH 9 in comparison to pH 7.5. Results of the proteome analysis are indicated for the soluble, membrane and envelope fraction. Regulators of particular genes are given based on the CoryneRegNet data base.

1 The geneID according to the accession number BX927147 was used.

2 Genes known to form an operon and closely adjacent, equally oriented genes that likely form an operon were indicated by equal Latin letters.

3 Prediction of transmembrane helices were performed by using the TMHMM 2.0 sever at http://www.cbs.dtu.dk/services/TMHMM/.

4 The induction factors are given as log2 values of the ration of mRNA levels at pH 6 and pH 9 in comparison to pH 7.5, respectively.

5 The determined relative peptide numbers are given as log2 values in order to allow calculation of ratios by simple subtraction of values. Peptide numbers found to be significantly altered at pH 6 and pH 9 in comparison to pH 7.5 are shown in bold and peptide numbers found to be significantly altered at pH 6 in comparison to pH 9 are shown in italic (see M&M section for the details of calculation).

6 Data whether a particular gene was experimentally proven or predicted (*) to be regulated by a transcription factor was obtained by using the data base CoryneRegNet http://coryneregnet.cebitec.uni-bielefeld.de/v4/.

Table 2

Differential expression pattern at pH 6 in comparison to pH 7.5

Genes induced at pH 6						Transcriptome4		Proteome5									Regulators6
								cytoplasm			envelope			membrane
No	gene ID1	op2	gene	function	TMH3	pH 6	pH 9	6	7.5	9	6	7.5	9	6	7.5	9
1	cg0012			Hypothetical protein	0	1.45	n.d.	-	-	-	2.5	-	-	-	-	-	McbR*
2	cg0325			Multisubunit Na+/H+ antiporter	2	1.04	0.17	-	-	-	0.6	1.3	0.8	-	-	-
3	cg0360			Putative phosphatase	1	1.35	n.d.	-	-	-	-	-	-	-	-	-
4	cg0403	b	rmlB1	dTDP-glucose 4,6-dehydratase	0	1.04	n.d.	-	-	-	-	-	-	-	-	-
5	cg0404	b		Nitroreductase family	0	1.77	-1.50	3.4	2.5	0.9	0.9	-	-	-0.9	-	-
6	cg0550			Putative peptidase E	0	1.10	-0.46	-	-	-	-	-	-	-	-	-
7	cg0736	k	metN	ATPase component	0	1.45	n.d.	2.2	-	-	5.1	2.8	2.6	4.0	1.7	-0.6	McbR*, RamB*
8	cg0737	k	metQ	periplasmic component	0	1.85	-0.69	5.4	3.4	3.5	7.2	4.1	4.7	9.1	6.8	6.7	McbR*, RamB*
9	cg0754	l	metX	Homoserine O-acetyltransferase	0	1.19	n.d.	2.1	-	-0.8	2.5	-	-0.3	-	-	-	McbR*
10	cg0755	l	metY	O-acetylhomoserine sulfhydrylase	0	2.43	-0.08	6.6	4.8	5.4	3.3	0.7	1.4	1.7	-	-	McbR*
11	cg0874			Uncharacterized ACR, COG2135	0	1.03	0.46	-	-	-	-0.4	-	-	-	-	-
12	cg0878		whiB1	Stress response transcription factor WhiB1	0	1.26	0.49	-	-	-	-	-	-	-	-	-	SigH, GlxR*
13	cg1081	p		ABC-type multidrug transport system, ATPase component	0	1.18	0.86	-0.3	-0.8	0.0	2.0	2.6	3.3	3.0	3.3	4.4
14	cg1082	p		ABC-type multidrug transporter, permease components	6	1.33	0.68	-	-	-	-	-	1.4	-	-	-
15	cg1083	p	cgtS10	Two-component system, sensory transduction histidine kinases	5	1.38	0.64	-	-	-	-	-	1.0	-	-	-	SigB*
16	cg1129		aroG	Phospho-2-dehydro-3-deoxyheptonate aldolase	0	1.44	n.d.	1.1	1.5	-	2.4	2.6	-	-0.7	1.4	-
17	cg1150			NADPH dependent FMN reductase	0	1.02	n.d.	1.2	-	-	-	-	-	-	-	-
18	cg1202			Hypothetical protein	0	1.23	n.d.	-	-	-	-	-	-	-	-	-
19	cg1214	q	sufS	Cysteine desulfurase involved in maturation of Fe-S clusters	0	1.30	-0.37	-0.6	-	-	1.4	-0.3	-	-	-	-	NrtR*
20	cg1215	q	nadC	Nicotinate-nucleotide pyrophosphorylase	0	1.30	-0.06	5.1	4.2	3.6	1.2	-	-	2.1	-	1.6	NrtR*
21	cg1216	q	nadA	Quinolinate synthetase A	0	1.77	-0.20	-	0.2	0.7	2.2	2.6	-	-	0.7	-	NrtR*
22	cg1218	q	nrtR	Regulator NrtR, ADP-ribose pyrophosphatase	0	1.58	-0.24	-	-	-	-	-	-	-	-	-	NrtR*
23	cg1256		dapD	Tetrahydrodipicolinate N-succinyltransferase	0	1.25	n.d.	2.5	0.4	-	5.0	3.6	2.4	-0.9	-	-
24	cg1291			Hypothetical membrane protein	2	2.43	n.d.	-	-	-	0.9	-	-	-	-	-
25	cg1322			Uncharacterized beta barrel protein	0	1.24	-0.62	6.3	5.1	2.7	-	-	-	2.6	-0.6	-
26	cg1337		hom	Homoserine dehydrogenase	0	1.68	0.22	4.4	2.4	3.0	5.4	4.5	4.0	3.1	3.2	1.4	McbR
27	cg1344	s	narG	Nitrate reductase	0	1.36	1.94	-	-	-0.6	-	-	-	-	-	-0.4	RipA, GlxR*
28	cg1476		thiC	Thiamine biosynthesis protein	0	1.00	0.62	3.9	3.8	4.1	2.9	2.3	3.6	1.8	1.4	1.0
29	cg1478			Hypothetical protein	0	4.89	n.d.	-0.6	-	-	-	-	-	-	-	-	LexA
30	cg1580	v	argC	N-acetyl-gamma-glutamyl-phosphate reductase	0	1.65	-0.17	2.9	2.2	2.2	1.3	-0.2	-	-0.9	-	-	ArgR
31	cg1581	v	argJ	Glutamate N-acetyltransferase	0	2.03	-0.02	5.8	5.0	5.2	2.8	2.1	2.1	-	-	-	ArgR
32	cg1582	v	argB	Acetylglutamate kinase	0	1.84	n.d.	1.1	-	-0.8	3.1	-	0.0	-	-	-	ArgR
33	cg1583	v	argD	Acetylornithine aminotransferase	0	1.66	n.d.	1.9	0.9	0.2	-	-	-	-	-	-	ArgR
34	cg1584	v	argF	Ornithine carbamoyltransferase	0	1.75	-0.12	1.2	0.4	-0.2	-	-	-	-	-	-	ArgR
35	cg1586	v	argG	Argininosuccinate synthase	0	1.16	0.05	4.1	4.3	4.3	5.4	5.2	5.1	3.1	3.5	2.8
36	cg1626	w		Hypothetical secreted protein	0	2.12	n.d.	-	-	-	-	-	-	-	-	-
37	cg1628	w		Putative hydrolase	0	2.25	n.d.	-	-	-	1.9	-	-	-	-	-
38	cg1647			ABC-type multidrug transporter, permease components	5	1.06	-0.01	-	-	-	-0.4	-	-0.3	-	-	-	AmtR*
39	cg1701		metH	Homocysteine Methyltransferase	0	2.22	0.12	1.1	-	-	3.7	3.1	0.7	-	-	-0.6	McbR*
40	cg1739			GMP synthase-Glutamine amidotransferase domain	0	1.36	n.d.	-	-	-	-	-	-	-	-	-	McbR*
41	cg1806		metK	S-adenosylmethionine synthetase	0	1.29	-0.17	2.9	2.3	1.1	3.6	2.9	-	-	-	-	McbR
42	cg1940			Hypothetical protein	1	1.18	-0.08	-	-	-	-	-	-	-	-	-
43	cg2051			Hypothetical protein	0	1.25	n.d.	-	-	-	-	-	-	-	-	-
44	cg2157		terC	putative tellurium exporter TerC	9	1.05	-0.13	-	-	-	2.8	2.4	1.5	1.5	1.0	2.3
45	cg2250			Hypothetical protein	0	1.16	n.d.	-	-	-	0.6	-	-0.3	1.6	-	1.1
46	cg2260		glnK	Nitrogen regulatory protein PII	0	1.13	n.d.	1.2	2.2	1.4	-	-	-	-	-	-	AmtR
47	cg2338		dnaE1	DNA polymerase III, alpha chain	0	1.12	0.29	-	-	-	0.0	-	-0.3	-	-	-
48	cg2380			Hypothetical membrane protein	2	1.00	0.12	-	-	-	1.7	1.3	0.8	-	-	-0.6
49	cg2462			TetR-type transcription factor	0	1.12	n.d.	-	-	-	-	-	-	-	-	-
50	cg2472			Predicted hydrolase or acyltransferase	0	1.14	0.48	-	-	-	-	-	-	-	-	-
51	cg2576			DNA polymerase III delta subunit	0	1.17	n.d.	-	-	-	-	-	-	-	-	-
52	cg2590			Putative xanthine/uracil permease	12	1.87	n.d.	-	-	-	-	-	-	0.2	-	-0.6
53	cg2591		dkgA	2,5-diketo-D-gluconic acid reductase	0	1.72	n.d.	3.2	0.9	2.0	-	-	-	-	-	-
54	cg2674	ae		Putative Carboxymuconolactone decarboxylase	0	1.08	-0.25	3.0	2.8	2.5	1.5	-	-	-0.9	-	-	McbR*
55	cg2675	ae		ABC-type putative peptide transporter, duplicated ATPase component	0	1.70	n.d.	0.8	-0.4	-	3.9	0.8	-	0.2	-	-	McbR*
56	cg2677	ae		ABC-type putative peptide transporter, permease subunit	6	2.00	n.d.	-	-	-	3.9	0.8	1.9	2.9	1.6	1.5	McbR*
57	cg2678	ae		ABC-type putative peptide transporter, periplasmic subunit	0	1.90	n.d.	1.2	-	-0.2	4.9	1.3	3.4	6.0	3.3	2.7	McbR*
58	cg2687		metB	Cystathionine beta-lyases/cysta-thionine gamma-synthases	0	1.34	0.37	5.3	4.2	5.1	-	-	-	-	-	-0.4	McbR*
59	cg2748			Hypothetical membrane protein	2	1.21	0.86	-	-	-	-0.4	-	-0.3	2.7	-	1.8
60	cg2761			Metal-dependent hydrolases of the beta-lactamase superfamily III	0	1.05	-0.30	0.4	-0.8	-	-	-	-	-	-	-
61	cg2766			MarA-type transcription factor	0	1.20	0.69	0.7	-	1.5	-	-0.3	-	-	-	-	RamB*
62	cg2796	af	prpD	Putative 2-methylcitrate dehydratase	0	2.79	n.d.	4.7	-	-	2.0	-	-	-	-	-	DtxR
63	cg2797	af		Putative 2-phospho-3-sulpholactate synthase	0	2.92	n.d.	2.4	-	-	-	-0.2	-	-	-	-	DtxR
64	cg2834		cysE	Serine O-Acetyltransferase	0	1.48	n.d.	-	-0.5	-	3.4	1.3	2.1	0.4	-	-
65	cg2835			Predicted acetyltransferase	0	1.52	0.56	2.5	-	1.9	0.8	-	-	-	-	-
66	cg2890			Putative amino acid processing enzyme	0	2.18	n.d.	-	-	-	-	-	-	-	-	-
67	cg2891		poxB	Pyruvate:quinone oxidoreductase	0	2.02	-0.21	3.6	0.4	0.3	0.6	-	0.9	-0.9	-	-	SigB*
68	cg3049		fprA	putative ferredoxin NADP oxidoreductases	0	1.26	0.24	3.3	3.1	3.6	3.3	3.8	3.0	0.2	2.4	-0.6
69	cg3082	ag		ArsR-type transcription factor	0	3.34	n.d.	-	-	-	-	-	-	-	-	-	DtxR
70	cg3083	ag		Predicted Co/Zn/Cd cation transporter	6	3.42	n.d.	-	-	-	1.2	-	-	-	-	-	DtxR
71	cg3084	ag		putative FAD dependent NAD(P)H disulphide oxidoreductase	0	3.14	n.d.	-	-	-	2.6	-	-	-	-	-	DtxR
72	cg3085	ag		putative FAD dependent oxidoreductase	0	1.47	n.d.	-0.6	-	-	-	-	-	-	-	-	DtxR
73	cg3112	ah	cysZ	Sulfate permease	7	4.03	0.32	-	-	-	3.8	0.8	2.6	2.6	0.7	1.6	DtxR, McbR*, CysR*
74	cg3113	ah		Putative metal chelatase	0	3.31	n.d.	-	-	-	-	-	-	-	-	-	DtxR, McbR*, CysR*
75	cg3114	ah	cysN	GTPases-Sulfate adenylate transferase subunit 1	0	3.42	0.48	4.4	1.4	3.2	6.2	4.1	4.0	4.6	3.1	1.0	DtxR, McbR*, CysR*
76	cg3115	ah	cysD	3-phosphoadenosine 5-phosphosulfate sulfotransferase (PAPS reductase)	0	3.27	0.78	4.4	2.8	3.0	5.5	4.0	3.7	4.2	2.4	1.0	DtxR, McbR*, CysR*
77	cg3116	ah	cysH	Phosphoadenosine-Phosphosulfate Reductase	0	2.98	n.d.	-	-0.5	-	2.7	2.3	2.4	-	-	-	DtxR, McbR*, CysR*
78	cg3117	ai	cysX	Hypothetical protein	0	3.50	1.08	-	-	-	-	-	-	-	-	-	DtxR, McbR*, CysR*
79	cg3118	ai	cysI	Sulfite reductase hemoprotein beta-component	0	3.50	1.06	0.6	1.5	2.4	3.5	2.1	-0.3	-	-	-	DtxR, McbR, CysR*
80	cg3119		cysJ	Probable NADPH-dependent Sulfite Reductase	0	2.99	0.29	4.9	4.2	3.5	2.8	1.3	1.4	2.0	1.4	-	DtxR, McbR*, CysR*
81	cg3157			Uncharacterized vancomycin resistance protein	1	1.54	0.23	-	-	-	2.7	2.1	3.1	3.7	-0.6	-0.4
82	cg3215		glpQ1	Putative glycerophosphoryl diester phosphodiesterase	0	1.02	n.d.	-	-	-	-	-	-	-	-	-
83	cg3219		ldh	Anaerobic L-lactate DH	0	1.83	0.12	5.1	3.1	4.8	4.3	1.3	2.3	-	-	-0.4	GlxR*
84	cg3227		lldD	Aerobic FMN-L-lactate DH	0	1.14	0.45	1.2	2.1	0.7	5.2	4.1	4.0	1.5	-0.6	-	GlxR*
85	cg3236		msrA	Methionine sulfoxide reductase	0	1.27	1.01	-	-	-0.8	-	-	-	-	-	-
86	cg3372			Hypothetical membrane protein	0	2.80	-0.16	-	-	-	1.8	-	-	-0.7	-	-	McbR*, CysR*
87	cg3374	ak		Putative NADH-dependent flavin oxidoreductase	0	2.48	n.d.	-	-	-	-	-	-	-	-	-	McbR*, CysR*
88	cg3375	ak		Putative NAD dependent dehydratase	0	1.61	n.d.	1.5	-0.8	0.8	-	-	-	-	-	-	McbR*, CysR*
Genes repressed at pH 6
1	cg0148		panC	Pantoate--beta-alanine ligase	0	-1.29	-0.56	-0.6	1.7	-0.8	-	-	-	-	-	-
2	cg0244	a		Hypothetical membrane protein	4	-1.23	-0.49	-	-	-	-	-	-	-	-	-
3	cg0245	a		Putative Moco sulfurase involved in sulphur metal clusters formation	0	-1.02	-0.64	-	-	-	-	-	-	-	-	-
4	cg0252			Hypothetical membrane protein	5	-1.16	-0.58	-	-	-	-	-	-	-	-	-
5	cg0308			Putative membrane protein	4	-1.34	n.d.	-	-	-	-	-	-	-	-	-
6	cg0337		whiB4	Transcriptional regulator	0	-1.04	0.39	-	-	-	-	-	-	-	-	-
7	cg0350		glxR	Transcriptional regulator	0	-1.18	-0.07	4.0	4.7	4.6	3.3	2.8	3.1	-0.7	-0.6	1.5
8	cg0445	c	sdhCD	Succinate dehydrogenase CD	5	-1.31	1.31	-	-	-	1.8	1.4	2.4	3.2	4.0	5.0	RipA, DtxR, RamA*, RamB*
9	cg0446	c	sdhA	Succinate dehydrogenase A	0	-1.08	1.47	1.0	2.7	1.7	6.7	7.0	9.0	8.5	9.0	9.9	RipA, DtxR, RamA*, RamB*
10	cg0447	c	sdhB	Succinate dehydrogenase B	0	-1.43	1.14	-	-0.4	-	4.6	5.0	6.7	3.5	3.4	5.3	RipA, DtxR, RamA*, RamB*
11	cg0465			Conserved hypothetical membrane protein	3	-1.14	n.d.	-	-	-	-	-0.2	-	-	-	-	DtxR
12	cg0466			Heme transport system, substrate binding subunit	0	-1.33	n.d.	-	-	-	-	-	-	-	-	-	DtxR
13	cg0470	d		Heme transport associated protein	2	-1.66	n.d.	-0.6	1.9	-	3.5	3.8	-	3.3	3.6	-	DtxR, LexA
14	cg0471	d		Heme transport associated protein	1	-1.30	n.d.	-	-	-	-	-0.3	-	0.1	1.0	-	DtxR, LexA
15	cg0493			Hypothetical protein	0	-1.05	-0.41	-	-	-	-	-	-	-	-	-
16	cg0563	e	rplK	50S ribosomal protein L11	0	-1.28	-0.59	-	-	-	0.9	0.7	0.8	-0.9	-	-0.6
17	cg0564	e	rplA	50S ribosomal protein L1	0	-1.10	-0.57	4.5	5.0	4.8	3.9	4.2	4.0	4.1	4.4	4.5
18	cg0572	f	rplJ	50S ribosomal protein L10	0	-1.60	-0.69	3.4	4.3	3.6	2.1	2.3	2.1	4.2	4.5	3.9
19	cg0573	f	rplL	50S ribosomal protein L7/L12	0	-1.53	-0.59	5.3	5.5	4.5	-	1.3	0.0	0.2	0.4	1.6
20	cg0582		rpsG	30S ribosomal protein S7	0	-1.07	-0.35	3.1	3.4	2.1	2.2	2.6	3.3	1.3	1.6	1.5
21	cg0599		rpsS	30S ribosomal protein S19	0	-1.09	-0.51	1.7	2.2	1.1	-	-	-0.3	-	-	-
22	cg0601	g	rpsC	30S ribosomal protein S3	0	-1.05	-0.35	3.3	3.2	3.0	4.7	4.0	4.5	0.3	0.4	0.4
23	cg0602	g	rplP	Ribosomal protein L16/L10E	0	-1.02	-0.27	3.1	3.3	3.1	4.5	4.8	5.0	3.0	3.4	3.5
24	cg0603	g	rpmC	50S ribosomal protein L29	0	-1.01	-0.20	-	-0.8	-	-	-	-	-	-	-
25	cg0604	g	rpsQ	30S ribosomal protein S17	0	-1.18	-0.21	3.3	3.2	2.8	-	-	0.0	-	-	-
26	cg0623	h		ABC-type cobalt exporter unit	8	-1.05	-1.35	-	-	-	-	-	-	-	-	-
27	cg0624	h		Hypothetical membrane protein	2	-1.09	-1.09	-	-	-	-	-	-	1.3	1.6	-0.4
28	cg0625	h		Putative terpene cylase or prenyltransferase subunit	1	-1.28	-1.04	-	2.2	0.0	2.8	3.6	2.2	-	-0.6	-0.4
29	cg0690	i	groS	10 kDa chaperonin	0	-1.77	-0.02	4.1	4.3	4.7	-	-	-	-	-	-	SigM*
30	cg0691	i	groEL	60 kDa chaperonin HSP60	0	-2.19	-0.11	4.0	3.4	3.8	-	-	-	1.2	-0.6	1.1	SigM*
31	cg0693	i	groL1	60 kDa chaperonin1 Hsp60	0	-1.47	0.02	5.6	5.8	5.8	1.8	0.7	1.8	2.5	1.7	1.6
32	cg0748			ABC-type Fe3+-siderophores transport systems, periplasmic components	0	-1.07	-2.14	-0.3	1.3	-	1.2	0.7	-	0.1	-0.6	-	DtxR
33	cg0752			Putative flotillin like protein	1	-1.55	-0.47	0.6	2.7	1.6	3.4	2.3	1.8	4.4	4.7	3.4
34	cg0760		prpB2	Methylisocitrate lyase 2	0	-1.54	-0.80	-	3.2	0.9	-0.4	0.8	-	-	-	-
35	cg0762		prpC2	2-methylcitrate synthase 2	0	-1.39	n.d.	-	1.7	0.8	-	-	-	-	-	-
36	cg0832			ABC-type transporter, permease components	5	-1.33	-0.40	-	-	-	-0.4	-	-	-	1.0	1.1
37	cg0834			ABC-type transporter, periplasmic component	0	-2.46	-0.03	1.6	3.4	2.7	2.5	3.6	3.7	2.9	4.2	4.1	LexA
38	cg0842			Putative DNA helicase	0	-1.03	n.d.	-	-	-	-	-	-	-	-	-
39	cg0898		pdxS	Pyridoxal biosynthesis lyase pdxS	0	-1.09	-0.61	4.6	5.0	4.3	-	-	-	-0.7	-	-	LexA, PdxR
40	cg0952	o	mctB	putative monocarboxylate transporter subunit	2	-2.55	-1.32	-	0.5	-	2.4	4.7	4.1	-	2.1	-0.6
41	cg0953	o	mctC	monocarboxylate transporter	13	-2.44	-1.01	-	3.0	-0.2	1.9	4.7	4.7	3.2	4.5	3.2
42	cg0961			Homoserine acetyltransferase	0	-1.70	n.d.	-	-	-	-	-	-	-	-	-
43	cg1072		rplY	50S ribosomal protein L25	0	-1.26	-0.49	0.7	1.1	0.6	-	-	-	-	-	-0.4
44	cg1108		porC	Putative secreted protein	0	-1.11	0.06	-	-	-	0.8	2.3	-	-	1.0	-
45	cg1122			Hypothetical protein	0	-1.31	0.23	-	-0.8	-0.8	3.3	4.7	6.1	5.8	6.3	6.6
46	cg1123		greA	Transcription elongation factor	0	-1.21	-0.14	2.0	0.4	2.4	-	-	-	-	-	-
47	cg1167		metS	putative methionine transporter subunit	0	-1.04	-1.43	-	-	-	-	-	-	-0.9	-	-0.6
48	cg1362	t	atpB	F0F1-type ATP synthase a subunit	6	-1.35	-0.85	-	-	-	2.7	3.2	3.8	5.8	6.0	5.5	SigH*
49	cg1363	t	atpE	F0F1-type ATP synthase c subunit	2	-1.60	-0.88	-	-	-	4.4	5.0	3.6	5.4	6.6	6.4	SigH*
50	cg1364	t	atpF	F0F1-type ATP synthase b subunit	1	-1.52	-0.96	-0.6	0.5	-0.8	4.5	4.8	4.9	4.9	5.4	4.3	SigH*
51	cg1365	t	atpH	F0F1-type ATP synthase delta subunit	0	-1.26	-1.03	1.9	2.3	1.3	3.9	4.8	4.7	3.2	4.1	3.4	SigH*
52	cg1366	t	atpA	F0F1-type ATP synthase alpha subunit	0	-1.65	-0.89	4.9	5.1	5.1	5.6	6.2	6.1	6.8	6.9	6.4	SigH*
53	cg1367	t	atpG	F0F1-type ATP synthase gamma chain	0	-1.71	-1.11	1.1	4.2	3.4	3.3	4.9	4.7	4.0	4.0	2.5	SigH*
54	cg1368	t	atpD	F0F1-type ATP synthase beta chain	0	-1.72	-0.94	5.3	5.5	5.4	5.7	5.8	5.6	6.3	6.0	5.7	SigH*
55	cg1369	t	atpC	F0F1-type ATP synthase epsilon chain	0	-1.09	n.d.	1.6	0.4	0.7	-	-	-0.1	3.1	2.5	1.9	SigH*
56	cg1436		ilvN	Acetolactate synthase, subunit	0	-1.41	-0.69	2.0	2.2	2.4	1.5	3.2	1.5	-0.9	-0.6	-0.6
57	cg1437		ilvC	Ketol-acid reductoisomerase	0	-1.44	-0.91	4.0	3.8	3.9	2.9	3.9	3.3	-0.9	1.7	-0.6
58	cg1564	u	rpmI	50S ribosomal protein L35	0	-1.04	-0.31	-	-0.8	-	1.6	1.4	2.4	-	-	-
59	cg1565	u	rplT	50S ribosomal protein L20	0	-1.26	-0.29	1.2	1.5	-0.2	-	-0.2	-	-	-	-
60	cg1579			Hypothetical protein	0	-1.36	-0.66	-	-	-	-	-	-	-	-	-
61	cg1612			Putative acetyltransferases	0	-1.32	n.d.	-	-	-	-	-	-	-	-	-
62	cg1905	y		Putative protein kinase	0	-1.33	0.72	-	-	-	-	0.7	-	-	-	-
63	cg1906	y		Putative protein phosphatase	0	-1.17	0.63	-	-	-	-	-0.2	-	-	-	-
64	cg1913			Hypothetical protein	0	-1.19	n.d.	-	-	-	-	-	-	-	-	-
65	cg1930	z		Putative secreted serine protease	0	-1.18	-3.26	-	-	-	-	-	-	-	-	-	DtxR
66	cg2136	aa	gluA	ABC-type glutamate transporter, ATPase component	0	-1.36	-0.07	-0.6	-0.5	0.8	1.9	4.3	3.8	2.2	3.4	2.7	GlxR*, AmtR
67	cg2137	aa	gluB	ABC-type glutamate transporter, substrate binding component	0	-1.38	-0.39	-0.6	2.7	1.3	2.4	3.3	3.5	4.9	5.5	5.1	GlxR*, AmtR
68	cg2138	aa	gluC	ABC-type glutamate transporter, permease component	6	-1.19	-0.33	-	-	-	1.8	1.3	3.0	-	-	-0.6	GlxR*, AmtR
69	cg2167		rpsO	30S ribosomal protein S15	0	-1.01	-0.14	1.1	1.8	1.9	-0.4	-	-	-	-0.6	0.6
70	cg2181			ABC-type peptide transporter, periplasmic component	0	-2.20	-0.53	-0.2	3.5	1.5	2.8	4.0	4.1	3.9	5.7	5.6	AmtR
71	cg2234			ABC-type cobalamin/Fe3+-siderophores transporter, secreted component	0	-1.32	n.d.	-0.6	-	-	-0.4	0.7	-	-0.9	-0.6	-	DtxR, RamB*
72	cg2235		rplS	50S ribosomal protein L19	0	-1.40	-0.46	2.5	2.2	1.9	6.6	6.9	7.5	3.0	3.9	3.7
73	cg2253		rpsP	30S ribosomal protein S16	0	-1.14	-0.34	4.1	4.7	4.0	1.2	1.3	2.5	1.6	-0.6	-0.4
74	cg2467	ac		ABC-type peptide transporter, substrate binding component	0	-1.21	-0.17	-	-	-	-0.4	2.8	2.8	1.4	-0.6	2.1
75	cg2470	ac		ABC-type peptide transporter, substrate binding component	0	-1.66	-0.36	0.7	2.2	1.2	1.2	1.3	0.0	2.8	3.4	3.4
76	cg2559	ad	aceB	Malate synthase G	0	-1.21	n.d.	-0.6	4.7	2.0	-	-	-	-	-	-	RamA, RamB
77	cg2560	ad	aceA	Isocitrate lyase	0	-2.73	-2.19	-	4.9	2.4	-	-0.3	-	-	-	-	RamA, RamB
78	cg2573		rpsT	30S ribosomal protein S20	0	-1.33	-0.56	2.8	3.3	2.1	2.7	2.7	3.5	1.6	1.4	1.2
79	cg2603		ndk	Nucleoside diphosphate kinase	0	-1.19	-0.20	2.8	3.3	3.2	-	-	-	-	-	-
80	cg2647		tig	Trigger factor	0	-1.47	-0.27	4.8	5.4	5.0	-	-	-	-	-	-
81	cg2703			ABC-type transporter, permease component	6	-1.90	-0.32	-	-	-	-	-	-	-	-	-
82	cg2705			ABC-type transporter, periplasmic component	0	-1.68	-0.28	4.9	7.0	5.8	6.9	7.7	8.0	6.8	7.6	7.4
83	cg2840		actA	Butyryl-CaA-acetate coenzyme A transferase	0	-1.67	-0.01	6.4	7.1	7.6	3.4	4.9	3.2	-	2.0	0.4	RamB*
84	cg2953		xylC	Benzaldehyde dehydrogenase	0	-1.39	0.61	1.6	3.3	4.1	1.3	4.3	3.7	-	-	1.5	GlxR*
85	cg3011		groL2	60 kDa chaperonin 2 (HSP60)	0	-2.20	-0.44	5.9	6.7	6.8	-0.4	1.7	3.5	2.2	2.4	3.0
86	cg3048		pta	Phosphate acetyltransferase	0	-1.07	0.16	-	1.7	2.3	-	2.9	2.8	-	-	-	RipA, RamA, RamB
87	cg3096			NAD-dependent aldehyde dehydrogenases	0	-3.24	0.10	3.2	6.6	5.9	3.0	6.1	5.7	-	1.0	1.4
88	cg3107		adhA	Zn-dependent alcohol dehydrogenases	0	-1.97	0.06	-	2.9	1.8	-	5.6	4.9	-	3.7	-0.4	RamA, RamB, GlxR*
89	cg3156		htaA	secreted protein implicated in iron acquisition and transport	0	-2.50	-2.76	-	-	-	-	-	-0.3	2.7	5.7	-	DtxR
90	cg3195			Flavin-containing monooxygenase	0	-2.23	0.32	-	2.2	-0.8	-	4.9	3.9	-	3.7	2.1
91	cg3212			Hypothetical membrane protein	0	-1.92	-0.50	-	-	-	-	-	-	-	-	-

Genes for which an increased or decreased mRNA level was found at pH 6 in comparison to pH 7.5. The gene locus tag, organisation in operons, the gene name, the (proposed) function of the protein as well as the predicted number of transmembrane helices are given. The results of the transcriptome analysis are given as induction factor at pH 6 and pH 9 in comparison to pH 7.5. Results of the proteome analysis are indicated for the soluble, membrane and envelope fraction. Regulators of particular genes are given based on the CoryneRegNet data base. For further details see legend of Table 1.

In addition, a number of proteins with changed abundance was detected for which no change in gene transcription was observed. We identified for 43 proteins increased and for 30 proteins decreased peptide numbers at pH 6 (Additional file 1). The same held for 32 proteins with increased and for 20 proteins with decreased peptide numbers under alkaline conditions, (Additional file 2). An example is the gene cg1111 encoding enolase. The mRNA content was neither significantly changed at pH 6 (m-value 0.24) nor at pH 9 (m-value 0.09) but 229 peptides were found in the cytoplasmic fraction at pH 7.5, 334 at pH 6, and 104 at pH 9 (Additional file 2). Other examples with stable mRNA level and varying peptide numbers include the porines of the outer membrane PorA and PorH (decreased amounts of peptides found at pH 9 and 6 in comparison to pH 7.5 in the membrane fraction) as well as MetE (increased peptide numbers at pH 6 in the cytoplasmic fraction, Additional file 2). This indicates that posttranscriptional or posttranslational control might be involved and that the regulation of protein stability is important during pH acclimatization. Furthermore, because of the (putative) function of many genes that are differentially expressed in a pH dependent manner the rearrangement of the cell wall might take place and influence the gene expression response.

Subsequently, we checked whether transcriptional regulators are known to be involved in expression control of genes that were found to be regulated. This was done using the CoryneRegNet data base which provides information on 72 regulators in C. glutamicum [18]. For 21 of the 39 genes found to be repressed at pH 9, predictions were made or experimental evidence was obtained, for regulation by particular transcription factors (Table 1). Accordingly, for approx. 50% of the genes found to be induced at pH 9 or differentially expressed at pH 6 the transcriptional regulator was proposed or identified (Table 1, 2).

Iron homeostasis of C. glutamicum is affected by the external pH

The iron availability is monitored in C. glutamicum by the binding of ferrous iron to the transcription factor DtxR [19]. At high internal concentrations of ferrous iron, the regulator binds to operator sites in the promoter regions of target genes, including RipA, the second regulator of iron homeostasis. Whereas DtxR can act both as repressor and activator, RipA acts as repressor only [20,21]. The combined transcriptome and proteome data suggest that the external pH value influences the availability of iron. At alkaline pH, DtxR-repressed genes like cg0925-28, encoding a siderophore ABC transporter, or cg0767, encoding a siderophore interacting protein, are found to be repressed, while the mRNA levels respectively peptide numbers of DtxR-activated genes like ftn (encoding a ferritin-like protein involved in iron storage) and dps (cg3327) are increased. We found RipA-regulated genes encoding iron containing enzymes like succinate dehydrogenase (cg0446-0447), aconitase (cg1737), or catalase (cg0310) to be (slightly) repressed at pH 6, whereas the same genes were found to be induced at alkaline pH (Table 1). Additionally, higher peptide numbers were found for SdhA, SdhB, Acn, and KatA under alkaline conditions (Table 1). Furthermore, genes of the SufR regulon, cg1759-65, including the genes nifS2, sufC, and sufB which encode components of the FeS cluster assembly machinery, as well as the regulator SufR itself are induced at pH 9 (Table 1). In contrast, the ABC type transporter for ferric iron uptake encoded by cg0508-0506 is not under the control of DtxR and no change of the transcript or protein level was detected (data not shown). In summary, we found a pH-dependent regulation of genes of the RipA and DtxR regulon indicating the activation of the iron starvation response at pH 6 and iron excess conditions at pH 9.

At neutral and acidic pH H2O2 can be detected in C. glutamicum cultures

The induction of iron starvation response at pH 6 was surprising because the solubility of iron is increased at low pH values and the availability should be increased at pH 6. Therefore, we speculated that activation of iron starvation could be caused by an impaired function of the cytoplasmic regulators. By oxidation of the cytoplasmic ferrous iron to ferric iron, the co-activator of DtxR, DtxR-mediated regulation might be triggered. Such a process could be induced by the endogenous formation of reactive oxygen species as described for the Fur protein in E. coli [22]. In order to test for the pH dependent occurrence of oxidative stress in C. glutamicum cells, we performed again batch fermentations in bioreactors in minimal medium under continuous pH control. During the exponential phase we detected significantly higher levels of H2O2 at pH 6 (6.5 μM, OD600 4) than under neutral (2.2 μM, OD600 12) or alkaline pH conditions (0.9 μM, OD600 6). Additionally, in cultures grown in buffered minimal medium in Erlenmeyer flasks we could detect H2O2 during exponential growth in C. glutamicum cultures. We measured 3 μM H2O2 in cultures grown at pH 9 but in cultures grown at pH 7.5 and pH 6 we measured unexpected high concentrations of H2O2, namely 20 μM after eight hours of incubation. The results indicate the increased occurrence of oxidative stress in C. glutamicum and/or suggest that the defense against oxidative stress is impaired in a pH dependent manner. In order to assess an effect of H2O2 production at low pH we applied a well established method for the measurement of protein carbonylation by using the OxyBlot assay. Total proteins of cells grown at pH 6, 7.5 and 9 were extracted and subjected to 1D SDS PAGE before and after the OxyBlot treatment (Additional file 3). Interestingly, a high number of proteins can be detected to harbour carbonyl groups in C. glutamicum protein extracts of cells grown at every pH. We could not find a significant increase in carbonylation at low pH.

Furthermore we performed growth experiments in Erlenmeyer flasks at pH 7.5 and pH 6 in presence of external catalase enzyme (Fig. 3). Interestingly, for C. glutamicum cells grown at pH 7.5 in presence of catalase (16 KU/ml) a higher growth rate was observed (μ = 0.393 ± 0.005) in comparison to the absence of external catalase (μ = 0.343 ± 0.006). At pH 6 addition of catalase had no significant beneficial effect because the growth rates in presence or absence of catalase were comparable (Fig. 3). Catalase was also added after every hour of incubation in order to prevent loss of enzymatic activity and to provide continuous catalase activity but comparable results were obtained (data not shown). In conclusion, elimination of H2O2 by addition of external catalase enzyme facilitates growth of C. glutamicum at neutral pH but not at acidic pH conditions.

Figure 3

Impact of externally added catalase enzyme on growth of . Wild type cells were exposed to pH 6.0 (white symbols) and 7.5 (black symbols) in buffered medium in Erlenmeyer flasks and growth was determined in absence (circles) or presence (squares) of purified catalase protein of C. glutamicum. The enzyme (16 KU/ml) was added at the beginning.

Metabolic alterations during response to acidic pH

The amounts of several enzymes were found to be affected by the changed external pH including succinate dehydrogenase and aconitase. In order to unravel metabolic alterations caused by the differing protein content, we performed GC-MS based metabolic profiling of cells grown at pH 6 and pH 7.5 under continuous pH control. Thereby, we identified numerous amino acids, intermediates of TCA, glycolysis, pentose phosphate pathway, and methionine pathway to be present at significantly different levels (Table 3). For example, pyruvate was found at pH 6 at an eleven fold higher concentration than at pH 7.5. Within the TCA, citrate, which is the substrate of aconitase, was found to accumulate like malate and fumarate. In contrast the metabolites 2-oxoglutarate and succinate were found in significantly lower concentrations at pH 6 (Table 3). Among the amino acids accumulation of phenylalanine, valine, glutamine, and alanine was observed, and proline and β-alanine were found in lower concentrations at pH 6. The pool size of methionine was slightly decreased, but we identified intermediates of the methionine pathway to be present in high concentrations like cystathionine and cysteine (Table 3). On the other hand nearly all enzymes of the methionine pathway were found to be induced at the mRNA and/or protein level.

Table 3

Differential metabolite pattern at pH 6 in comparison to pH 7.5

Metabolite	differential content	t-test
Methionine synthesis
homolanthionine	-15.28	0.0392
Serine	-1.84	0.0307
Glycine	-1.67	0.0010
homocysteine	-1.41	0.2019
methionine	-1.09	0.9449
threonine	1.26	0.0302
Aspartate	1.33	0.1295
homoserine	1.58	0.3969
O-acetyl-serine	2.34	0.0165
S-adenosyl-homocysteine	2.45	0.0441
cysteine	2.46	0.0030
O-acetyl-homoserine	3.71	0.0002
cystathionine	6.96	0.0044
Glycolysis and Pentosephosphate pathway
Pep	-3.58	0.0025
DHAP	-2.89	0.0004
DHAP	-2.79	0.0010
glycerate-3-P	-1.99	0.0043
gluconate-6-P	-1.77	0.1733
ribose-5-P	-1.44	0.0373
fructose-1-6-P	2.07	0.0360
glycerate-2-P	2.86	n.d.
fructose-6-P	3.00	0.0009
glucose-6-P	3.18	0.0000
glucose-6-P	3.85	0.0006
pyruvate	11.10	0.0341
TCA cycle
alpha-ketoglutarate	-6.07	0.0131
succinate	-5.37	0.0039
fumarate	1.25	0.1755
citrate	1.84	0.0775
malate	2.06	0.0065
Amino acids
proline	-8.67	0.0006
beta-alanine	-7.30	0.0000
ornithine, citrulline, arginine	-2.38	0.2871
serine	-1.84	0.0307
glycine	-1.67	0.0010
lysine	-1.58	0.3466
asparagine	-1.45	0.3712
methionine	-1.09	0.9449
leucine	1.01	0.9353
glutamate	1.02	0.7068
tyrosine	1.17	0.1464
threonine	1.26	0.0302
L-aspartate	1.33	0.1295
tryptophan	1.40	0.0606
isoleucine	1.73	0.0007
histidine	2.28	0.2898
cysteine	2.46	0.0030
alanine	3.28	0.0117
glutamine	4.85	0.0002
valine	5.52	0.0013
phenylalanine	9.08	0.0020

Pool size ratios determined by GC-MS in extracts of C. glutamicum cells grown at pH 7.5 and pH 6. The ratios have been calculated using the mean values from two independently grown cultures and three technical replicates each. A t-test was applied for determining if observations were significantly different. Numbers in italics indicate values above the chosen significance cut off (P < 0.05), n.d. means not determined.

The McbR regulon is induced at acidic pH

At pH 6 we observed induction of genes encoding proteins of the methionine and cysteine pathway (Table 2, Fig. 4). Intermediates of these pathways are involved in essential cellular functions including the assembly of iron sulfur clusters (cysteine), the de novo synthesis of proteins (cysteine, methionine) or the metabolism of C1 units (S-adenosyl-methionine, methyltetrahydrofolate; Fig. 4). Many of the genes are under control of McbR and the ancillary regulators CysR and SsuR [23-25]. Among them are, e.g., the fpr2-cysIXHDNYZ cluster and cysK, encoding the sulfate permease CysZ, the complete set of enzymes involved in sulfate reduction to sulfide (CysDN, CysH, CysIX, and Fpr2) as well as the serine-O-acetylserine sulfhydrylase CysK, involved in cysteine synthesis (Fig. 4). Furthermore, the genes hom, metB, metH, metK, metXY, and metQN, encoding enzymes of the methionine pathway and subunits of the primary methionine uptake system MetQNI, were found to be induced [8,26]. The genes encoding the cysteine synthase (cysK), the homocysteine methyltransferase (metE), the β-C-S lyase (aecD), and the S-adenosyl-homocysteine hydrolase (sahH) were not found to be induced at the mRNA level (Table 2, Fig. 4). Corresponding to the unaffected mRNA level of aecD and sahH no differential peptide numbers were found (Fig. 4). For the AecD enzyme we determined unchanged cystathionine lyase activities in cells grown at pH 7.5 and at pH 6 using an enzymatic assay (data not shown). However, a higher protein level was found for MetE and a lower amount for CysK in spite of the unaffected mRNA levels (Additional file 1 and 2, Fig. 4). This might be an indication for increased protein stability of MetE and CysK at low pH. In contrast to pH 6 the McbR and CysR regulon were not found to be differentially expressed at pH 9 (Table 2, Fig. 4). It should be noted that we are not able to report on genes under the control of the transcription factor SsuR, because no transcription data were obtained for these genes and no peptides were found representing the corresponding proteins.

Figure 4

The pH dependent regulation of the methionine and cysteine metabolism in . The metabolite pool sizes at pH 6 in comparison to pH 7.5 are indicated below the intermediates. The involved proteins as well as the encoding genes are given in circles and the regulation by McbR (M) and/or CysR (C) is indicated. Below the proteins the relative expression levels at pH 6 and pH 9 in comparison to pH 7.5 are given and the peptide numbers detected in the soluble protein fraction at pH 6/7.5/9 are given. (n.d. means not detected)

At the metabolite level we observed the accumulation of intermediates of the methionine pathway upstream of the AecD enzyme including L-homoserine, O-acetyl-L-homoserine, L-cysteine, and L,L-cystathionine (Table 3, Fig. 4). Furthermore, the content of the McbR effector S-adenosyl-homocysteine was increased at low pH. In contrast, the pool sizes of homocysteine and methionine, representing metabolites downstream of AecD, were found to be slightly reduced.

From the observed metabolic imbalance we inferred that accumulation of intermediates of the methionine pathway upstream of AecD or the lower pool size of the downstream intermediates could contribute to the growth defect of C. glutamicum cells at acidic pH. In order to test this hypothesis we performed growth experiments at pH 7.5 and 6 in absence or presence of 10 mM cystathionine, cysteine, homocysteine, or methionine. Based on these assumptions, the addition of cystathionine or cysteine should increase pH dependent growth inhibition whereas homocysteine and methionine should supplement a putative demand for these compounds at pH 6. The addition of cystathionine, homocysteine and methionine had no significant effect on C. glutamicum growth at pH 6 (data not shown). However, addition of cysteine significantly decreased growth rates of cells exposed to acidic pH values. Further experiments revealed that the extent of growth inhibition by cysteine was indeed pH dependent. Whereas at pH 9 and 7.5 cysteine addition had no effect on the growth rate, at pH 7 growth was retarded and at pH 6.5 and 6 cells were hardly able to grow (Fig. 5).

Figure 5

The pH dependent impact of cysteine on growth of . Wild type cells were exposed to different pH values in MTP and growth rates were determined in absence (black bars) or presence 10 mM cysteine (white bars).

Differential expression of further regulatory modules

Beside the induction of methionine and cysteine synthesis, the complete arg cluster was found to be induced at pH 6. The expression of the arg genes, encoding all enzymes for synthesis of arginine from glutamate via the urea cycle, was proven to be under the control of the two repressors ArgR and FarR (Table 2). The investigation of the metabolite pattern revealed, however, a lower pool size for ornithine, citrulline, and/or arginine, represented by only one signal in the GC-MS analysis (Table 3).

The transcription factors RamA and RamB as well as GlxR are major regulators of the carbon flux in C. glutamicum [19,27,28]. At pH 9 and pH 6 we observed the repression of several genes indicating alterations of the carbon metabolism. Among them are aceA, encoding isocitrate lyase, aceB, encoding malate synthase, and mctC, encoding an uptake system for pyruvate, acetate and propionate [8,29]. All of these genes are under the control of RamA and RamB. Additionally, alternative oxidases were found to be induced, like the FMN containing lactate dehydrogenase LldD and the pyruvate quinone oxidase Pqo. Whereas lldD expression was proposed to be under the control of GlxR, pqo expression was found to be part of the sigma factor SigB regulon [28,30].

Another regulatory module found to be induced at pH 6 comprises the genes cg1214-18. These genes encode the NadAC proteins, involved in NAD synthesis, a putative cysteine desulfurase, possibly involved in maturation of FeS clusters necessary for function of the NadAC complex, and the regulator NrtR [8,31]. All genes of the operon were induced at pH 6 and for NadC (cg1215), an increased number of peptides (33) was found at acidic conditions in comparison to pH 9 (12 peptides, Table 1). Under alkaline conditions genes of the NrtR regulon were not induced. Subsequently, we determined the cellular concentration of all NAD derivatives in C. glutamicum cells grown at pH 7 and pH 6. Whereas the NADP and NADPH concentrations at pH 6 were only half of those observed at pH 7 (NADP pH 6: 0.11 ± 0.01 mM, pH 7: 0.18 ± 0.03 mM; NADPH pH 6: 0.25 ± 0.03 mM, pH 7 0.48 ± 0.07 mM) the NAD and NADH concentrations were only one third at pH 6 in comparison to pH 7 (NAD pH 6: 0.57 ± 0.05 mM, pH 7: 1.53 ± 0.14 mM; NADH pH 6: 0.46 ± 0.05 mM, pH 7 1.49 ± 0.39 mM). Calculation of the ratios of the oxidized and reduced forms revealed that the reduction state of the cell was not affected by acidic pH. However, significantly lower levels of NAD derivatives were found under acidic conditions accompanied by the induction of genes encoding enzymes involved in the first steps of their synthesis.

Discussion

The purpose of this study was to achieve a general conception of the acclimatization of the Gram-positive soil bacterium C. glutamicum towards acidic as well as alkaline external pH on the transcriptome, proteome, as well as on the physiological level. Of particular interest was the question which metabolic processes are impaired under conditions of non-optimal pH and thereby represent limitations for growth. As a prerequisite the growth optimum was determined and found to be in the range between pH 7 and 8.5. Consequently, C. glutamicum can be regarded as a moderately alkali-tolerant strain in comparison to E. coli with a pH optimum at 6-7 [1]. In E. coli the capacity of pH homeostasis is higher than in C. glutamicum, because E. coli can maintain pHi values of 7.6 ± 0.2 at external pH values ranging from pH 5 to 9 [32]. Consequently, mechanisms of pH homeostasis are less effective in C. glutamicum in comparison to E. coli. In agreement with the observed growth optimum, pH homeostasis was effective in a range between 6 and 9 in C. glutamicum. At lower or higher external pH values maintenance of the internal pH at a level of 7.5 was not achieved leading to reduced growth rates. In order to identify targets which are affected by pH values exceeding the boundaries of effective pH homeostasis in C. glutamicum, we checked at first whether the proton motive force (pmf) was affected in a pH-dependent manner. The pmf is the major driving force for the generation of ATP by oxidative phosphorylation. However, the pmf was kept constant over a surprisingly broad pH range.

Subsequently, the dissection of pH acclimatization by transcriptome and proteome studies uncovered many physiological processes that are affected in a pH-dependent manner. The gene expression pattern observed in this work overlaps with the results obtained for gene expression analysis during growth on lactate at pH 5.7 [11]. Beside a high number of genes which seem to be expressed in dependence of the carbon source, 15 out of 88 genes which were found in our studies to be transcriptionally induced at pH 6 were also identified during growth on lactate as induced at low pH [11]. On the other hand, 31 out of 91 genes that were found to be repressed under acidic conditions were also found to be repressed by Jakob et al. (2007). Among them are genes encoding subunits of the succinate dehydrogenase, the F1F0-ATPase, and rRNA genes. For several of them the expression was shown to be dependent on the growth rate which is in agreement with a lower expression at pH 6 and pH 9 at which we observed lower growth rates than at neutral pH [10]. The evaluation of differential transcript and protein patterns by the comparison with targets of transcriptional regulators in C. glutamicum unraveled numerous regulatory modules that are activated during pH acclimatization. Novel findings are represented by the induction of the iron starvation response as well as the induction of expression of the methionine and cysteine pathway under acidic conditions.

At neutral and low pH C. glutamicum is impaired by oxidative stress

The permanent formation of H2O2 in living cells was already previously described for E. coli. However, the detection of hydrogen peroxide in the medium was only possible after inactivation of the primary H2O2 scavenging enzyme alkyl hydroperoxide reductase (AhpCF) [33]. Homologues of the Ahp proteins are missing in C. glutamicum and we found that WT cells produce significant amounts of H2O2 at neutral and especially at acidic pH. The formation of H2O2 can cause cellular damage by oxidation of sulfur atoms in cysteine or methionine residues at the protein surface, by protein carbonylation, or by oxidation of iron sulfur clusters [34]. The analysis of protein carbonylation was performed for the first time for C. glutamicum and in contrast to other bacteria a high number of carbonylations were detected under all conditions [35]. In consequence, C. glutamicum might be exposed towards a certain level of oxidative stress under all our experimental conditions and the pH dependent differences might be overlooked.

The most likely source for the formation of hydrogen peroxide is not the respiratory chain but alternative oxidases like the lactate oxidase and the pyruvate oxidase. Both enzymes are utilized by lactic acid bacteria in order to excrete large amounts of H2O2 [36,37]. In C. glutamicum lactate oxidase as well as pyruvate oxidase were found to be significantly induced at low pH. The decrease of the internal pH in C. glutamicum cells at an external pH of 6 might also cause formation of reactive oxygen species by soluble oxidoreductases, especially those using FADH2 as cofactor, upon malfunctions at non optimal pH conditions [34,38]. Furthermore, the catalase content of cells is reduced at pH 6 in comparison to pH 7.5 and even more to pH 9. Thereby, an additional decrease of the H2O2 scavenging capacity at neutral and low pH is expected. Accordingly, addition of catalase enzyme at pH 7.5 caused a significant increase in growth rate pointing to the limitation of growth by H2O2 formation under neutral conditions and the insufficient activity of the cellular catalase enzyme. At pH 6 production of H2O2 is accompanied by other limitations for C. glutamicum cells and consequently, addition of catalase did not increase growth rate significantly.

Additional limitations at pH 6 could be cause by the disruption of iron sulfur clusters by ROS or by oxidation of the amino acids cysteine and methionine. In agreement with this, expression of the methionine sulfoxide reductase which is involved in repair functions is induced at low pH (Table 1). Thioredoxins are in general thought to be involved in this process but were also found to be not induced by H2O2 in other bacteria like in Bacillus subtilis [39]. Additionally, NAD synthesis in C. glutamicum potentially depends on iron sulfur clusters [40] and these seem to be affected by H2O2. The total NAD concentration was found to be reduced significantly. In agreement to this, in E. coli the NadA enzyme was identified as a target of oxidative stress [41]. The formation of H2O2 in C. glutamicum cells might impair the function of the enzymes NadC and/or NadA directly or indirectly because significantly reduced levels of NAD derivatives were found. Consequently, induction of the NrtR regulon was found (Fig. 6).

Figure 6

Regulatory modules activated under acidic conditions in . Putative stimuli and signals are indicated as stars, regulators involved are shown in circles. The affected processes and genes are indicated in boxes and the physiological consequences are given at the bottom.

The occurrence of oxidative stress interferes with iron availability and control of metabolic fluxes in C. glutamicum

We propose that the link between acidic pH and iron starvation response may be caused be the H2O2 dependent conversion of ferrous into ferric iron, resulting in the inactivation of DtxR and the consecutive activation of RipA, the two regulators of iron homeostasis in C. glutamicum (Fig. 6). In agreement with our assumption is the observation that the addition of higher amounts of iron sulfate at low pH values did not diminish the expression of DtxR controlled genes [11]. We observed a reduced transcription of the acn and sdhABC genes which was correlated with lower contents of the corresponding proteins (Table 2), due to the lack of DtxR-mediated activation and repression by RipA. In turn, this could cause a reduced activity of the TCA cycle and indeed we could determine lower levels of α-ketoglutarate and succinate and drastically increased levels of pyruvate and higher levels of citrate and malate (Fig. 6). As a consequence of the increased pyruvate pool, alternative metabolic routes are activated. Among them is the oxidation of pyruvate by the pyruvate oxidase Pqo or the synthesis of the branched chain amino acids valine and isoleucine at pH 6 (Table 3).

In conclusion, at acidic pH endogenous formation of hydrogen peroxide occurs in C. glutamicum cells at an extent that can obviously not be compensated by ROS defense mechanisms. As a consequence the induction of the iron starvation response was observed including lower levels of iron containing enzymes of the TCA cycle, and oxidative damage of iron sulfur cluster containing enzymes as proposed for the NAD synthesis pathway. All effects represent metabolic limitations and contribute to the impaired growth of C. glutamicum cells under acidic conditions (Fig. 6).

Cysteine accumulation inhibits the cysthationine-β-lyase AecD and causes additional limitations for growth of C. glutamicum under acidic conditions

High internal concentrations of H2O2 can cause DNA damage and require DNA repair as well as de novo synthesis [42]. Newly synthesized DNA is methylated whereby S-adenosylmethionine (SAM) serves as the methyl donor [43]. We could not measure SAM directly but found that the corresponding pool size of the resulting intermediate S-adenosylhomocysteine (SAH) was increased at acidic pH. This increase causes inactivation of the McbR repressor and led to the transcriptional induction of genes of the methionine and cysteine pathway (Fig. 4) [23].

An exception is the aecD gene encoding the cystathionine-β-lyase AecD which is not under McbR control. In agreement with this, the AecD protein content was unaffected by changed pH values and comparable enzyme activities were determined under neutral and acidic pH conditions. The increased pool sizes of metabolites upstream of AecD including cystathionine and cysteine and slightly reduced pool sizes of the downstream metabolites homocysteine and methionine could be caused by the missing induction of the aecD gene at acidic pH and/or by inhibition of the AecD enzyme activity. As a consequence, accumulation of cysteine was found, indicating an imbalance of thiol homeostasis in C. glutamicum under acidic stress conditions. High levels of cysteine can cause oxidative stress by formation of H2O2 and hydroxyl radicals via the Fenton reaction which, concomitantly, would increase damage of proteins and DNA [44]. Consequently, cysteine would be converted into cystine. Because cystine is preferred over cystathionine by AecD, this could cause inhibition of methionine and SAM biosynthesis as well as further cysteine accumulation [45]. Internal accumulation of cysteine was found to be toxic for E. coli cells [44]. In conclusion, cysteine addition would result in its accumulation in C. glutamicum and would thus amplify oxidative stress at acidic pH and thereby cause the severe growth inhibition.

Interestingly, methionine synthesis was also affected at acidic pH in E. coli and, in this case, accumulation of homocysteine was observed [14]. This indicates that the homocysteine methyltransferase MetE is affected at low pH. Because the MetE protein represents the major target for oxidative stress in E. coli we assume that oxidative stress may occur in this strain at low pH as well [46,47]. In contrast to E. coli, in C. glutamicum inhibition of the AecD enzyme was found. This suggests that a significant flux from homoserine to homocysteine via trans-sulfuration occurs at least during growth at pH 6. In addition, AecD inhibition prevents the accumulation of homocysteine which is more toxic for bacteria than cystathionine [14,48].

Conclusions

At non-alkaline pH values oxidative stress was found to occur in C. glutamicum and the reactive oxygen species defense was found to be impaired. As a consequence maintenance of cellular NAD levels is impaired and iron starvation response is activated. This leads to reduced protein levels of iron-containing enzymes affecting the TCA cycle and other metabolic pathways at low pH, among them methionine synthesis. McbR-dependent activation leads to cysteine accumulation which is toxic under acidic conditions. We thus have unraveled regulatory modules activated during acidic pH response in C. glutamicum (Fig. 6) and have identified targets as well as physiological consequences for the cellular stress response. Beside insights into bacterial physiology conclusions can be drawn for acclimatization of pathogenic Actinomyces and for the optimization of biotechnological production processes.

Methods

Strains and culture conditions

Strain ATCC 13032 served as Corynebacterium glutamicum wild type. C. glutamicum cells were grown either in Brain Heart Infusion (BHI) medium (Becton-Dickenson, Heidelberg, Germany) or in minimal medium MM1 [49] at 30°C. For all experiments C. glutamicum cells were precultured in 5 ml BHI medium for approx. 8 h and subsequently used for inoculation of 20 ml MM1. After approx. 20 h the culture was used to inoculate fresh MM1 medium of a desired pH to an OD600 of 1-2. Batch cultivations at different pH were performed in 2 l stirred bioreactors (Biostat B, Sartorius, BBI Systems, Melsungen, Germany) under continuous control of pH (6, 7.5, or 9) temperature (30°C) and pO2 (>30%) at a flow rate of 1 vvm air. Growth was followed by measuring the optical density at 600 nm (OD600). Each cultivation at pH 6, 7.5, and 9 was performed twice and parameters were determined always in triplicate. In order to screen for the impact of amino acid addition on growth at different pH values cultivations were performed in 96 well micro titer plates in a volume of 200 μl MM1 medium and the OD600 was followed by using a plate reader. Growth was also investigated in Erlenmeyer flasks (20 ml) MM1 medium and purified C. glutamicum catalase enzyme was a kind gift of Roche Diagnostics, Manheim, Germany.

Determination of bioenergetic parameters

During the exponential phase of growth cells were harvested, washed twice and resuspended in 100 mM MES buffer of the respective pH. Cell volumes were determined by the distribution of 3H-labelled H2O (0.55 mCi/l) and 14C-labelled inulin (0.14 mCi/l) between the cell pellet and the supernatant. The membrane potential was determined by measuring the distribution of 14C-labelled TPP (5 μM final concentration, sp. radioactivity 0.995 Ci/mol). Processing of samples for rapid separation of extra- and intracellular fluids was performed by using silicone oil centrifugation with perchloric acid in the bottom layer (Rottenberg, 1979). Internal pH was determined by measuring the distribution of 14C-labelled benzoic acid (15 μM final concentration, sp. radioactivity 3.12 Ci/mol). All measurements were performed at least in triplicate and standard deviations were calculated.

Detection and elimination of H2O2, protein carbonylation and cysthationine lyase activity

Concentrations of H2O2 in the medium were detected by use of the Amplex Red Hydrogen Peroxide/Peroxidase Assay Kit (A22188) from Molecular Probes (Karlsruhe, Germany) according to the supplier information. Fluorescence of the Amplex Red reagent was measured at 590 ± 4 nm after excitation at 530 ± 4 nm. In medium of the desired pH (6, 7.5, and 9) specific calibrations were done and H2O2 values were calculated accordingly. Because of the membrane permeability of H2O2 the external concentrations were regarded to be in equilibrium with the internal concentrations. In order to exclude unspecific formation of H2O2 in the medium control experiments were performed by incubation of medium without cells under the same conditions. In order to eliminate H2O2 produced in cell cultures purified catalase enzyme of C. glutamicum (kind gift of Roche Applied Science, Mannheim, Germany) was added to cultures.

The oxidative damage of proteins was analyzed using the OxyBlot™ Protein Oxidation Detection Kit (S7150) provided by Millipore. Basically, carbonylations of protein side chains are regarded as marker for oxidative stress and the occurrence of reactive oxygen species. Total proteins were extracted from cells grown at pH 6, 7.5 and 9 and carbonyl groups were derivatized to 2,4-dinitrophenylhydrazone which can be detected by a specific antibody. The cystathionine lyase activity was measured in cell extracts of C. glutamicum cells grown at different pH values as described previously [50]. All measures were performed at least in triplicate and standard deviations were calculated or representative results are shown.

Transcriptome analysis

Total RNA isolation (including cell harvest and lysis), cDNA synthesis, and array hybridisation were performed as described previously [16], using 70 mer oligo microarrays instead of dsDNA microarrays. Spot finding, signal background segmentation and intensity quantification were carried out with the ImaGene 6.0 software (BioDiscovery). Normalization using the lowess function, which computes the logarithmic intensity ratio (m) and the logarithmic mean signal intensity (a) for each spot was performed and t-test statistics was accomplished with the EMMA microarray data analysis software [51]. Evaluation of the hybridization experiment was done as described in [23], using a m-value cut-off of ± 1, which corresponds to expression changes equal or greater than twofold. Since Hüser et al. (2003) found that an m-value cutoff of ± 0.6 equals a false-positive rate of 1%, at ± 1 this rate is 0.04% (roughly one false-positive among 3000 genes). The microarray data are available at the public repository ArrayExpress http://www.ebi.ac.uk/arrayexpress by the accession number E-MTAB-151.

Proteome analysis

C. glutamicum ATCC13032 cells were harvested by centrifugation for 15 min at 4500 × g; cells were washed (PBS, 137 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4, 1.8 mM KH2PO4; pH 7.4) and disintegrated (PBS containing additional 20 mM MgCl2, 10 mM MnCl2, 200 U/ml DNaseI, protease inhibitor mix for bacterial cells (Sigma, St. Louis, MO, USA)) by four French press treatments (20000 psi, Thermo Spectronic, Rochester, USA). After centrifugation (5000 × g, 4°C) fractionation of the supernatant was performed. The soluble protein fraction was obtained as the supernatant after ultracentrifugation (100,000 × g; 4°C; 35 min). The membrane fraction was obtained after repeated ultracentrifugation and resuspension of the pellet in PBS with 10% glycerol and protease inhibitors. All samples were stored at -80°C.

For all approaches two technical replicates were performed. After inactivation of the protease inhibitor (60°C; 1 h) 100 μg of soluble proteins were incubated over night at 60°C with 4 μg trypsin (Promega, Madison, USA) and samples were desalted by Spec PT C18 AR solid phase extraction pipette tips (Varian, Lake Forest, CA, USA). The membrane fraction was treated according to two different protocols: (a) the enriched membrane fraction whereas a predigest removes membrane-associated proteins was achieved by the SIMPLE (Specific Integral Membrane Peptide Level Enrichment) protocol [52]; (b) the cell envelope fraction was achieved by a direct tryptic digest [53]. After removal of membranes by centrifugation (100,000 × g; 4°C; 35 min) samples were desalted with Spec PT C18 AR tips.

All desalted samples were resuspended in buffer A (2% acetonitrile, 0.1% formic acid) and subjected to 1D-nLC-ESI-MS using an autosampler. A self-packed capillary column was used for LC (Eclipse C18-RP XDB, Hewlett Packard) in combination with the Accela gradient HPLC pump system (Thermo Electron) coupled to an LTQ Orbitrap mass spectrometer (Thermo Electron). For elution of the peptides a multiple step gradient of buffer A to buffer B (80% acetonitril, 0.1% formic acid) was applied (0-5 min: 0% buffer B; 5-10 min: 10% buffer B; 10-175 min: 40% buffer B; 175-200 min: 100% buffer B; 200-210 min: 0% buffer B) at a flow rate of ~250 nl/min and a spray voltage of 1.5-1.8 kV. The LTQ Orbitrap was operated via instrument method files of Xcalibur (Rev. 2.0.7). The linear ion trap and orbitrap were operated in parallel, i.e. during a full MS scan on the orbitrap in the range of 300-2000 m/z at a resolution of 60,000, MS/MS spectra of the four most intense precursors were detected in the ion trap. Singly charged and more than triply charged ions were rejected from MS/MS and dynamic exclusion was enabled.

All database searches were performed using SEQUEST algorithm, embedded in Bioworks™ (Rev. 3.3, Thermo Electron), according the Corynebacterium glutamicum ATCC 13032 Bielefeld database [8]. The mass tolerance for precursor ions was set to 10 ppm; the mass tolerance for fragment ions was set to 1 amu. For protein identification a threshold for both protein and peptide probability was set to 0.001 in Bioworks™. MS2 spectra per protein were counted using an in-house Perl script from Bioworks™ result tables. Spectral counts [17], i.e. number of identified peptide MS spectra per protein of different samples were normalized according the sum of all spectra in the sample. The significance of protein abundance changes was calculated in relation to the total peptide counts for each protein [54]. The proteome data are available at the public repository PRIDE at http://www.ebi.ac.uk/pride/ by the accession numbers 9355-9390.

Metabolome analysis

Cell disruption and metabolite extraction was performed as described previously [55,56]. Derivatisation of samples as well as GC-MS analysis using a TraceGC gas chromatograph equipped with an AS2000 auto sampler and coupled to a PolarisQ ion trap mass spectrometer (Thermo Finnigan, Dreieich, Germany) was performed according to previous analyses [56]. The t-test algorithm of Excel 2000 (Microsoft, Seattle, WA) was used for determining whether observations were significantly different (P < 0.05).

Authors' contributions

MF performed the analysis of the internal pH and bioenergetic parameters, IO performed the enzyme assays and the determination of H2O2, DS performed the OxyBlot assay, CT and AP performed the proteome analysis, AH and CR performed the DNA microarray analyses, MP performed the metabolome analysis, JK supervised the transcriptome and metabolome analysis. KM and RK designed the research and wrote the manuscript with assistance by CR, CT and JK.

Additional file 1

Exclusive alterations at the protein level at pH 6. Table of proteins for which a differential peptide number was found at pH 6 in comparison to pH 7.5 but no alteration of the mRNA level was observed. footnotes for Table. 1 The geneID according to the accession number BX927147 was used. 2 Prediction of transmembrane helices were performed by using the TMHMM 2.0 sever at http://www.cbs.dtu.dk/services/TMHMM/. 3 The induction factors are given as log2 values of the ration of mRNA levels at pH 6 and pH 9 in comparison to pH 7.5, respectively. 4 The determined relative peptide numbers are given as log2 values in order to allow calculation of ratios by simple subtraction of values. Peptide numbers found to be significantly altered at pH 6 and pH 9 in comparison to pH 7.5 are shown in bold and peptide numbers found to be significantly altered at pH 6 in comparison to pH 9 are shown in italic (see M&M section for the details of calculation).

Click here for file

Additional file 2

Exclusive alterations at the protein level at pH 9. Table of proteins for which a differential peptide number was found at pH 9 in comparison to pH 7.5 but no alteration of the mRNA level was observed. footnotes for Table. 1 The geneID according to the accession number BX927147 was used. 2 Prediction of transmembrane helices were performed by using the TMHMM 2.0 sever at http://www.cbs.dtu.dk/services/TMHMM/. 3 The induction factors are given as log2 values of the ration of mRNA levels at pH 6 and pH 9 in comparison to pH 7.5, respectively. 4 The determined relative peptide numbers are given as log2 values in order to allow calculation of ratios by simple subtraction of values. Peptide numbers found to be significantly altered at pH 6 and pH 9 in comparison to pH 7.5 are shown in bold and peptide numbers found to be significantly altered at pH 6 in comparison to pH 9 are shown in italic (see M&M section for the details of calculation).

Click here for file

Additional file 3

Analysis of protein modifications by oxidative stress using the detection of carbonyl groups in protein side chains. Total protein extracts of cells grown at pH 6, 7.5 and 9 were obtained and subjected to an 1D SDS-PAGE before (A) and after the derivatization by 2,4-dinitrophenylhydrazine (DNP, B). The DNP mojety was detected using a specific antibody of the OxyBlot detection kit.

Click here for file