trans-Acting Factors and cis Elements Involved in Glucose Repression of Arabinan Degradation in Bacillus subtilis

Abstract

In Bacillus subtilis, the synthesis of enzymes involved in the degradation of arabinose-containing polysaccharides is subject to carbon catabolite repression (CCR). Here we show that CcpA is the major regulator of repression of the arabinases genes in the presence of glucose. CcpA acts via binding to one cre each in the promoter regions of the abnA and xsa genes and to two cres in the araABDLMNPQ-abfA operon. The contributions of the coeffectors HPr and Crh to CCR differ according to growth phase. HPr dependency occurs during both exponential growth and the transitional phase, while Crh dependency is detected mainly at the transitional phase. Our results suggest that Crh synthesis may increase at the end of exponential growth and consequently contribute to this effect, together with other factors.

Bacillus subtilis secretes a vast number of polysaccharide-degrading enzymes that are able to hydrolyze plant cell wall material (30). Arabinose and xylose are the two most abundant pentoses in nature and often occur associated with hemicellulosic substrates, such as arabinans, xylans, and arabinoxylans (25, 26, 29). B. subtilis synthesizes enzymes, namely, endo-1,5-α-arabinanases (ABN; EC 3.2.1.99) and α-l-arabinofuranosidases (AF; EC 3.2.1.55), capable of releasing arabinosyl oligomers and arabinose from the plant tissue homoglycan arabinan (1, 13, 14, 35). In previous studies, we characterized the transcriptional regulation of three arabinan-degrading enzyme genes, abnA (ABN), xsa (AF), and abfA (AF), which are clustered with genes encoding enzymes that further catabolize l-arabinose (23). Expression of these genes is (i) induced by arabinose and arabinan, (ii) negatively controlled by AraR, the key regulator of the arabinose regulon, and (iii) repressed in the presence of glucose (6, 12, 23). The last phenomenon, carbon catabolite repression (CCR), is a global regulatory mechanism that ensures the selection of rapidly metabolized carbon and energy sources, such as glucose, for optimal bacterial growth (3, 30). CcpA is the master regulator of CCR in Bacilli and other gram-positive bacteria with low GC contents and regulates approximately 10% of the B. subtilis genome (2, 18). This global regulator binds to cres (catabolite-responsive elements) either upstream of or in the promoter region or within coding regions of target genes (17, 34). This interaction is modulated by phosphorylation of HPr, a phosphocarrier protein of the phosphoenolpyruvate phosphotransferase system (PTS), or of Crh (an HPr-like protein) (4, 30, 33). In this work, we identified trans-acting factors and cis elements involved in glucose repression of arabinan-degrading enzyme genes. Moreover, we address the question of temporal regulation of these genes (23) by determination of the contributions of CcpA and the coeffectors HPr and Crh to glucose repression at different growth stages (exponential and postexponential [transitional] growth).

cis-Regulatory elements in the promoter regions of abnA and xsa.

The involvement of two cres present in the araABDLMNPQ-abfA operon (cre araA and cre araB) (Fig. 1) in CCR was shown in previous studies (12, 17). Both cre araA, located between the promoter region and the araA gene, and cre araB, placed 2 kb downstream within the araB gene, are independently functional, and both contribute to glucose repression (12). Moreover, the level of glucose repression, 21.8-fold, measured in a strain bearing an abfA′-lacZ reporter fusion (IQB450) (23; see below) was only 1.6-fold higher than that observed with a strain bearing an araAB′-lacZ fusion, i.e., 13.4-fold (12), suggesting the absence of other cre active sites in glucose-mediated CCR of abfA expression. Expression of abnA and xsa is also subject to glucose repression (23). Potential cre sequences, namely, cre abnA (+83-TGTAAGCGCTTTCT [SubtiList position 2949135]) (Fig. 1) (17) and cre xsa (+1-TAAAAGCGCTTACA [SubtiList position 2914316]) (Fig. 1) (18), are present in the promoter regions of the genes. To assess the functionality of these putative regulatory sites, we introduced a single-base-pair substitution which destroyed the central symmetry in both cre abnA (+89 C→A) and cre xsa (+7 C→A) (Fig. 1). The two promoter regions, together with the respective mutated cres, were fused to the lacZ reporter gene and introduced into the amyE locus. In the resulting B. subtilis strains, IQB472 and IQB473 (Table 1), respectively, the levels of accumulated β-galactosidase activity in the absence or presence of the inducer arabinose and under repressing conditions (arabinose plus glucose) were determined. Glucose repression was almost abolished in the mutants compared to that in strains bearing the wild-type promoters (IQB410 and IQB405) (23) (Table 2). These results indicate that both cre abnA and cre xsa are functional and that the designed mutation prevented binding of CcpA (see below) in the presence of glucose. Thus, cre abnA and cre xsa are the cis-acting elements involved in glucose-mediated CCR of abnA and xsa expression at the transcriptional level. In addition, the results suggest that only one cre in each promoter is responsible for this phenomenon.

FIG. 1.

Organization of the B. subtilis chromosome region comprising the arabinan-degrading enzyme genes abnA, abfA, and xsa. The three genes are represented by dotted arrows pointing in the direction of transcription. abfA belongs to the araABDLMNPQ-abfA metabolic operon, abnA is located immediately upstream, and xsa is positioned 23 kb downstream of the metabolic operon. Hairpin structures denote terminators. The dotted boxes below the physical map represent extensions of the inserts fused to the lacZ reporter gene in the indicated plasmids. Plasmid pSA1 was integrated into the host chromosome by means of a single-crossover (Campbell-type) recombinational event that occurred in the region of homology of the resulting strain (Table 1). Linearized DNAs from plasmids pSN40, pZI26, pZI27, and pRIT1 were used to transform B. subtilis strains (Table 1), and the fusions were integrated into the chromosome via double recombination at the amyE locus. Above the map, the DNA sequences of cre abnA and cre xsa are depicted, and the numbers indicate the positions of the abnA and xsa genes relative to the transcriptional start site. The single nucleotide change (C→A) introduced into each cre is shown.

TABLE 1.

B. subtilis strains, plasmids, and oligonucleotides used in this study

Plasmid, strain, or oligonucleotide	Relevant construction, genotype, or sequence (5′→3′)	Source or referencea
Plasmids
pJM783	Integrative plasmid; promoterless lacZ preceded by rbsspoVG and multicloning site (MCS); cat bla	22
pSN32	Promoterless lacZ preceded by rbsspoVG and MCS; cat bla; flanked by amyE-5′ and amyE-3′	20
pMutin4	Promoterless lacZ preceded by rbsspoVG and MCS; lacI bla ery Pspac	31
pRIT1	pSN32 containing a 281-bp fragment of the xsa promoter region in the MCS	23
pSN40	pSN32 carrying a 292-bp fragment of the abnA promoter region in the MCS	23
pSA1	pJM783 containing a 687-bp fragment comprising the abfA 5′ region in the MCS	23
pZI26	Same as pSN40, but with a single-base-pair substitution at position +89 (C→A)	This work
pZI27	Same as pRIT1, but bearing a single-base-pair substitution at position +7 (C→A)	This work
pGP211	pBSK+ derivative that contains the spc gene inserted into the hprK gene	10
pZI45	Identical to pBGM6 (9), pMutin4 bearing a crh′-lacZ fusion	This work
pZI48	Identical to pLF2 (9), pMutin4 bearing a ptsH′-lacZ fusion	This work
Strains
168T+	Prototroph	12
QB5223	trpC2 ptsH1	15
QB7097	trpC2 crh::spc	15
WLN29	trpC2 aroG932 ccpA::Tn917	11
IQB405	amyE::[xsa′-lacZ cat]	23
IQB406	amyE::[xsa′-lacZ cat] araR::km	23
IQB410	amyE::[abnA′-lacZ cat]	23
IQB411	amyE::[abnA′-lacZ cat] araR::km	23
IQB422	amyE::[abnA′-lacZ cat] ccpA::Tn917	WLN29→IQB410
IQB423	amyE::[xsa′-lacZ cat] ccpA::Tn917	WLN29→IBQ405
IQB450	abfA::pSA1[abfA′-lacZ cat]	23
IQB453	abfA::pSA1[abfA′-lacZ cat] araR::km	23
IQB466	amyE::[abnA′-lacZ cat] ptsH1	pRIT1→QB5223b
IQB467	amyE::[xsa′-lacZ cat] ptsH1	pSN40→QB5223b
IQB468	amyE::[abnA′-lacZ cat] crh::spc	QB7097→IQB410
IQB469	amyE::[xsa′-lacZ cat] crh::spc	QB7097→IQB405
IQB470	amyE::[abnA′-lacZ cat] ptsH1 crh::spc	QB7097→IQB466
IQB471	amyE::[xsa′-lacZ cat] ptsH1 crh::spc	QB7097→IQB467
IQB472	amyE::[abnA′(+89 C→A)-lacZ cat]	pZI26→168T+b
IQB473	amyE::[xsa′(+7 C→A)-lacZ cat]	pZI27→168T+b
IQB474	abfA::pSA1[abfA′-lacZ cat] ccpA::Tn917	WLN29→IBQ450
IQB475	abfA::pSA1[abfA′-lacZ cat] ptsH1	pSA1→QB5223
IQB476	abfA::pSA1[abfA′-lacZ cat] crh::spc	QB7097→IQB450
IQB477	abfA::pSA1[abfA′-lacZ cat] ptsH1 crh::spc	QB7097→IQB476
IQB478	abfA::pSA1[abfA′-lacZ cat] hprK::spc	pGP211→IQB450b
IQB479	amyE::[abnA′-lacZ cat] hprK::spc	pGP211→IQB410b
IQB480	amyE::[xsa′-lacZ cat] hprK::spc	pGP211→IQB405b
IQB495	crh::pZI45[crh′-lacZ erm (::pMutin4)]	pZI45→168T+
IQB496	ptsH::pZI48[ptsH′-lacZ erm (::pMutin4)]	pZI48→168T+
Oligonucleotidesc
ARA204	−8-GGTTACTTTAAAAGAGCTTACATTCATGC-+21 (xsa)
ARA205	GCATGAATGTAAGCTCTTTTAAAGTAACC
ARA206	+66-GGATAAATAATCTAATTTGTAAGAGCTTTCTAAAATAAAGG-+106 (abnA)
ARA207	CCTTTATTTTAGAAAGCTCTTACAAATTAGATTATTTATCC

^aArrows indicate transformation, performed as previously described (27), and point from donor DNA to the recipient strain.

^bTransformation was carried out with linearized plasmid DNA.

^cSubstituted residues in oligonucleotides are underlined. The position relative to the transcriptional start site of the xsa and abnA genes is indicated in the forward primer. Site-directed mutagenesis was performed using a QuikChange kit (Stratagene).

TABLE 2.

Single-base-pair substitutions in the cres located in the promoter regions of the abnA and xsa genes^a

Promoter fusion and strain (relevant genotype)	β-Galactosidase activity (Miller units)b			Glucose repression indexc
	−Ara	+Ara	+Ara +Glc
abnA′-lacZ
IQB410d (WT)	3.2 ± 0.7	10.0 ± 0.8	1.6 ± 0.3	6.5
IQB472 (cre abnAmut)	9.1 ± 1.3	80.7 ± 11.4	41.3 ± 3.3	1.9
xsa′-lacZ
IQB405d (WT)	7.5 ± 0.2	117.3 ± 3.9	9.2 ± 0.5	19.3
IQB473 (cre xsamut)	4.0 ± 0.5	163.1 ± 12.6	76.8 ± 9.5	2.1

^aStrains containing different promoter-lacZ fusions were grown on C minimal medium (21) supplemented with 1% (wt/vol) casein hydrolysate in the absence of sugar (−Ara), in the presence of 0.4% (wt/vol) arabinose (+Ara), and in the presence of 0.4% (wt/vol) arabinose plus 0.4% (wt/vol) glucose (+Ara +Glc). Antibiotics used as selective markers were added as appropriate. Samples were analyzed 2 h after the addition of sugar, as previously described (27).

^bThe levels of accumulated β-galactosidase activity, determined as previously described (27), represent the averages ± standard deviations of two measurements each from three independent experiments.

^cThe glucose repression index was calculated as the ratio of the level of expression (in Miller units) obtained in the presence of arabinose to the value determined in the presence of glucose.

^dData are from the work of Raposo et al. (23).

A major role of CcpA and different contributions of effectors HPr and Crh to CCR.

The arabinan-degrading enzyme genes are subject to temporal regulation, since the levels of expression of abfA, xsa, and abnA increase at early postexponential phase in response to arabinose and arabinan (23). Here we analyzed this effect in the presence of glucose. The expression of abfA′-lacZ, xsa′-lacZ, and abnA′-lacZ transcriptional fusions, integrated into the chromosomes of strains IQB450, IQB405, and IQB410, respectively (Table 1) (23), was analyzed as described above, and samples were collected 2 and 4 h after induction (t₂ and t₄, respectively), corresponding to the middle exponential and early postexponential (transitional) growth phases, respectively (Fig. 2A). Although expression of abfA′-lacZ, xsa′-lacZ, and abnA′-lacZ increased at t₄, the glucose repression indexes, measured as ratios between expression in the presence of arabinose and the values obtained in the presence of arabinose plus glucose, were similar during exponential growth and transitional phase (Fig. 2B, C, and D). These results suggest that the effect of glucose repression under these conditions is not affected over time.

FIG. 2.

Effects of ccpA, ptsH1, crh, hprK, and araR mutations on catabolite repression of the xsa, abnA, and abfA genes. (A) Growth curves for wild-type (squares) and ccpA-null mutant (triangles) B. subtilis strains in C minimal medium (21) supplemented with 1% (wt/vol) casein hydrolysate in the presence of 0.4% (wt/vol) arabinose plus 0.4% (wt/vol) glucose. Exponential growth phase (t₂) and transitional phase (t₄) are indicated. The strains bearing the xsa′-lacZ (B), abnA′-lacZ (C), and abfA′-lacZ (D) transcriptional fusions were grown in the absence of sugar, in the presence of 0.4% (wt/vol) arabinose, and in the presence of 0.4% (wt/vol) arabinose plus 0.4% (wt/vol) glucose. The levels of accumulated β-galactosidase activity, expressed as Miller units, were measured at hours 2 (t₂) and 4 (t₄) after induction (addition of arabinose). The results are represented by the glucose repression index, calculated as the ratio between the level of accumulated Miller units obtained in the presence of arabinose and the value determined in the presence of arabinose plus glucose. Each value is the average of two measurements each from three independent experiments.

To identify the trans-acting factors involved in CCR of the arabinan-degrading enzyme genes, we examined the expression of abfA′-lacZ, xsa′-lacZ, and abnA′-lacZ transcriptional fusions, as described above, in B. subtilis ccpA, pstH1, crh, hprK, and araR mutant backgrounds (Table 1). In a wild-type background, the addition of glucose caused 21.8-, 19.3-, and 6.5-fold repression of abfA′-lacZ, xsa′-lacZ, and abnA′-lacZ expression, respectively (strains IQB450, IQB405, and IQB410, respectively) during exponential growth phase (23). Disruption of the ccpA gene led to a complete loss of glucose-mediated CCR of expression from the abfA′-lacZ (IQB474), xsa′-lacZ (IQB423), and abnA′-lacZ (IQB422) fusions (Fig. 2), revealing its major role in glucose-mediated CCR of the arabinan-degrading enzyme genes. The ptsH1 mutation (HPr Ser46 to Ala), which impaired HPr Ser46 phosphorylation (5), partially abolished glucose repression of abfA′-lacZ (IQB475), xsa′-lacZ (IQB467), and abnA′-lacZ (IQB466) fusion expression (Fig. 2). In contrast, disruption of the crh gene caused almost no effect on glucose repression of expression from the abfA′-lacZ, xsa′-lacZ, and abnA′-lacZ fusions (IQB476, IQB469, and IQB468, respectively) during exponential growth phase (t₂). Accordingly, in the ptsH1 crh double mutant background, the levels of glucose repression, at t₂, of abfA′-lacZ (IQB477), xsa′-lacZ (IQB471), and abnA′-lacZ (IQB470) fusions were only slightly lower than those observed in the single ptsH1 mutant (Fig. 2), suggesting no or a small contribution of Crh to CCR mediated by glucose. These results are in agreement with previous observations that both CcpA and HPr, but not Crh, participate in glucose repression of the araE and araAB genes (12). However, in the crh-null mutant background, a significant relief of glucose repression of abfA′-lacZ and xsa′-lacZ expression, of 6.4-fold and 2.1-fold, respectively, was observed at transitional phase (t₄), suggesting a contribution of Crh to glucose-mediated CCR at late growth phases (discussed below). The disruption of the hprK gene, encoding the bifunctional HPr kinase/phosphorylase, which reversible phosphorylates HPr and Crh (8, 16, 24), also caused a decrease of glucose repression (in IQB478, IQB479, and IQB480) (Fig. 2).

During exponential growth of cells with an araR-null mutant background, we observed a twofold derepression of expression of xsa′-lacZ (IQB406) and abfA′-lacZ (IQB453) in the presence of glucose compared to that of the wild type (Fig. 2B and C). AraR is a negative regulator of the arabinose-inducible genes (6, 7, 20), and this effect was previously observed with araE′-lacZ and araAB′-lacZ fusions (12). The AraR binding sites and cre sequences are located close to each other in the promoter regions of the araABDLMNPQ-abfA, araE, and xsa genes (12, 19, 23). Since under particular physiological conditions both CcpA and AraR might be bound to the DNA in close proximity, a possible interaction (cooperation) between the two proteins in the negative control of these genes was proposed (12, 19). Interestingly, this effect was not observed for the abnA′-lacZ fusion in an araR-null mutant background (strain IQB411) (Fig. 2D; Table 1). This observation is in agreement with a different localization of cre abnA relative to the AraR binding site in the abnA promoter and also to a weak control of abnA expression by AraR as a result of a distinct mechanism of action (23).

Crh expression increases at transitional growth phase.

The contribution of both effectors HPr and Crh to CCR by glucose can depend on the growth conditions, as shown for the hut operon, which responds to HPr in rich LB medium and to both HPr and Crh in minimal medium (36). Crh is the major effector for CCR of the citM gene in a medium containing succinate and glutamate (32). In connection with this, it was recently shown that the nature of the carbon source influences ptsH and crh expression (9). Using ptsH′-lacZ and crh′-lacZ transcription and translational fusion analysis, it was suggested that the different contributions of Crh and HPr to CCR might be due to the drastic differences in their synthesis rates under the conditions that cause CCR. In the presence of PTS carbohydrates, such as glucose, the synthesis level of HPr during exponential growth is 100-fold higher than that of Crh, which may explain the minor role of the latter in certain circumstances (9). Moreover, CcpA binds Crh(Ser-P) more weakly than HPr(Ser-P) (28). We have previously shown the absence of a Crh contribution to glucose repression of the araABDLMNPQ-abfA and araE genes in both C minimal medium supplemented with casein hydrolysate and CSK minimal medium (12). To further investigate the lack of Crh dependency during exponential growth and its possible contribution to glucose repression of the arabinan-degrading enzyme genes during the transitional growth phase, we analyzed the expression of both ptsH′-lacZ and crh′-lacZ transcriptional fusions at different growth stages. The constructed transcriptional fusions are identical to those described by Görke et al. (9). In B. subtilis strains bearing the fusions, the complete ptsH and crh coding regions are restored downstream from the lacZ insertion cassette. Their expression is directed by the IPTG (isopropyl-β-d-thiogalactopyranoside)-inducible Pspac promoter in order to prevent polar effects caused by insertions into their natural chromosomal loci (Table 1). Expression from the crh′-lacZ (IQB495) (Table 1) and ptsH′-lacZ (IQB496) (Table 1) transcriptional fusions during exponential growth phase (t₂) was comparable in both C minimal medium supplemented with casein hydrolysate and CSK minimal medium (Table 3). The presence of arabinose, a non-PTS sugar, did not influence crh and ptsH expression (Table 3). In addition, the β-galactosidase activities obtained for crh′-lacZ and ptsH′-lacZ in CSK minimal medium plus glucose (29 and 229, respectively) (Table 3) were very similar to those obtained by Görke et al. (9). Interestingly, at transitional growth phase (t₄), the expression of crh increased about twofold in the presence of glucose and in both media. In contrast, ptsH expression was unaffected by the growth phase. These results suggest that increased expression of crh at the transcriptional level during transitional phase in the presence of glucose may lead to a higher cellular concentration of Crh than of HPr. Thus, this difference in synthesis level may contribute to the glucose repression relief effect on abfA′-lacZ and xsa′-lacZ expression in the crh-null mutant background (Fig. 2). However, due to a still expected higher concentration of HPr than Crh under these conditions, other factors, such as intracellular variation of certain metabolites which may influence Crh(Ser-P) phosphorylation and/or CcpA-Crh(Ser-P) complex formation, are likely to contribute to Crh(Ser-P)-mediated glucose repression at transitional phase.

TABLE 3.

Expression from crh′-lacZ and ptsH′-lacZ fusions in minimal media at different growth stages^a

Medium	Strain (promoter fusion)	Timeb	β-Galactosidase activity (Miller units)c
			−Ara	+Ara	+Ara +Glc	+Glc
C	IQB495 (crh′-lacZ)	t2	36.0 ± 3.8	31.5 ± 2.4	25.6 ± 1.2	23.9 ± 2.4
		t4	37.0 ± 0.8	37.8 ± 1.2	56.8 ± 3.2	56.9 ± 2.6
	IQB496 (ptsH′-lacZ)	t2	97.4 ± 14.9	94.9 ± 10.1	315.4 ± 18.4	322.7 ± 16.2
		t4	87.5 ± 7.1	94.3 ± 13.3	323.2 ± 16.3	322.7 ± 47.2
CSK	IQB495 (crh′-lacZ)	t2	31.4 ± 3.4	28.4 ± 3.5	24.5 ± 3.7	29.1 ± 2.3
		t4	28.3 ± 1.3	27.1 ± 3.2	43.7 ± 6.8	45.5 ± 2.7
	IQB496 (ptsH′-lacZ)	t2	59.7 ± 5.4	56.7 ± 6.0	204.0 ± 31.1	229.1 ± 16.2
		t4	55.0 ± 5.9	46.5 ± 8.3	195.9 ± 10.3	228.5 ± 38.8

^aStrains containing different promoter-lacZ fusions were grown on C minimal medium (21) supplemented with casein hydrolysate or in CSK minimal medium (9) with IPTG (1 mM) in the absence of sugar (−Ara), in the presence of arabinose (+Ara), in the presence of arabinose plus glucose (+Ara +Glc), and in the presence of glucose (+Glc). Antibiotics used as selective markers were added as appropriate.

^bSamples were analyzed 2 h (t₂) and 4 h (t₄) after the addition of sugars, which correspond to exponential growth phase and transitional phase, respectively.

^cThe levels of accumulated β-galactosidase activity represent the averages of two measurements each from three independent experiments.

Concluding remarks.

CCR of the arabinan-degrading enzyme genes is mediated by binding of CcpA to one cre located downstream near the promoter regions of abnA and xsa and to two cres present in the araABDLMNPQ-abfA operon, with cre araA located near the promoter and cre araB positioned within the araB gene. Our data suggest that during exponential growth phase, glucose-mediated CCR is mediated by CcpA associated with the coeffector HPr(Ser-P), and that towards the end of exponential growth, at transitional phase, glucose repression is achieved by both CcpA-HPr(Ser-P) and CcpA-Crh(Ser-P) complexes.