Control of asparagine homeostasis in Bacillus subtilis: identification of promiscuous amino acid importers and exporters

ABSTRACT The Gram-positive model bacterium B. subtilis is able to import all proteinogenic amino acids from the environment as well as to synthesize them. However, the players involved in the acquisition of asparagine have not yet been identified for this bacterium. In this work, we used d-asparagine as a toxic analog of l-asparagine to identify asparagine transporters. This revealed that d- but not l-asparagine is taken up by the malate/lactate antiporter MleN. Specific strains that are sensitive to the presence of l-asparagine due to the lack of the second messenger cyclic di-AMP or due to the intracellular accumulation of this amino acid were used to isolate and characterize suppressor mutants that were resistant to the presence of otherwise growth-inhibiting concentrations of l-asparagine. These screens identified the broad-spectrum amino acid importers AimA and BcaP as responsible for the acquisition of l-asparagine. The amino acid exporter AzlCD allows detoxification of l-asparagine in addition to 4-azaleucine and histidine. This work supports the idea that amino acids are often transported by promiscuous importers and exporters. However, our work also shows that even stereo-enantiomeric amino acids do not necessarily use the same transport systems. IMPORTANCE Transport of amino acid is a poorly studied function in many bacteria, including the model organism Bacillus subtilis. The identification of transporters is hampered by the redundancy of transport systems for most amino acids as well as by the poor specificity of the transporters. Here, we apply several strategies to use the growth-inhibitive effect of many amino acids under defined conditions to isolate suppressor mutants that exhibit either reduced uptake or enhanced export of asparagine, resulting in the identification of uptake and export systems for l-asparagine. The approaches used here may be useful for the identification of transporters for other amino acids both in B. subtilis and in other bacteria.


D-Asn is toxic for B. subtilis, and mutations in the mleN gene overcome this toxicity
Several D-amino acids are harmful to B. subtilis, as they exert toxic effects during protein synthesis (21).We tested the effect of D-Leu, D-Arg, D-Gln, D-Asp, and D-Asn on the growth of B. subtilis.Of these D-amino acids only D-Asn and D-Gln showed toxic effects (the results with D-Asn are shown in Fig. 1).The wild-type strain B. subtilis 168 was unable to grow at D-Asn concentrations above 5 mM.We isolated stable suppressor mutants that were able to grow in the presence of D-Asn.We isolated four suppressor mutants and subjected two of them to whole-genome sequencing.The data revealed an identical single mutation in the malate/lactate antiporter MleN (32) in both mutants, causing the insertion of two base pairs, resulting in a premature stop codon, and thus likely in the inactivation of the transporter (see Table 1).We then sequenced the mleN region of the two remaining suppressor mutants and found that the mleN gene was mutated there as well.Interestingly, all mutants had insertions of one or two base pairs in the mleN coding region that caused frame shifts in the N-terminal half of the protein.
We then examined the growth of an mleN deletion strain (GP1460), compared to the wild type strain (168) and one of the suppressors (GP4158), in the presence of increasing amounts of D-Asn (see Fig. 1).While low amounts of D-Asn (up to 5 mM) could be tolerated by the cells, higher amounts led to cell death and suppressor formation in the wild type strain.Both the suppressor mutant GP4158 and the ΔmleN deletion mutant GP1460 were able to grow at up to 30 mM D-Asn.This result indicates that the truncation of the MleN protein in the suppressor mutant indeed results in an inactivation of the protein, and shows that this mutation is responsible for the suppression.The fact that the ΔmleN deletion led to full D-Asn resistance suggests that MleN can transport D-Asn in B. subtilis.

Expression of the mleN gene
The putative D-Asn transporter MleN and the NAD + -dependent malate dehydrogenase MleA seem to be expressed in an L-asparagine-inducible operon with the asparaginase AnsA and the aspartase AnsB (31,35).According to a large-scale transcriptome analysis with B. subtilis and the SubtiWiki database, there might be one promoter in front of ansA and a second promoter in front of mleN (35,36).We wanted to verify this observation, as well as examine a possible induction of the operon by amino acids, particularly by the Dand L-enantiomers of asparagine and aspartate.For this purpose, we made use of strains that carry fusions of the regions upstream of the ansA and mleN genes to a promoterless lacZ reporter gene.The strains were grown in C glucose minimal medium in the absence or presence of casein hydrolysate, and the asparagine and aspartate enantiomers (see Table 2).In agreement with previous reports (31,37), the ansA promoter was highly active in the presence of L-Asn, whereas only weak expression was observed under all other tested conditions.In contrast, only weak activities were observed for the mleN promoter.The observed readthrough in the ansAB-mleNA region (35) as well as the absence of a promoter in front of mleN suggest, that the four genes are co-expressed in the ansAB-mleNA operon, via the asparagine-inducible promoter in front of ansA.Due to the toxicity of D-Asn, we were unable to assay promoter activities in the wild type background.To circumvent this problem, we transferred the lacZ fusion constructs into the strain lacking the D-asparagine importer MleN.In this case (GP4191), we observed a very weak increase of ansA promoter actvity which is, however, negligible as compared to the observed induction of the promoter in the presence of L-Asn (Table 2) and in comparison to promoters of basic carbon metabolism (38).However, this activity seems to be sufficient to express the mleN gene to a level that allows uptake of D-Asn and concomitant intoxication of the B. subtilis cells.

The toxicity of L-Asn for a Δdac mutant allows the identification of AimA as an L-Asn transporter
In some cases, the D-and L-enantiomers of an amino acids are taken up by the same transporter, as it is the case for D-and L-alanine that are both taken up by AlaP (12).This raised the question whether MleN might also be involved in the transport of L-Asn.This hypothesis is highly attractive since mleN is part of an operon involved in L-Asn utilization that is induced by this amino acid.The B. subtilis wild-type strain 168 tolerates L-Asn in minimal medium (see Fig. 2).We made, therefore, use of the observation that a strain that is unable to produce c-di-AMP due to the deletion of the three genes encoding the diadenylate cyclases (Δdac) is sensitive to glutamate and several other amino acids (4,17).As shown in Fig. 2, the Δdac mutant GP2222 is highly sensitive to growth inhibition a For the complete genotypes of all strains, see Table 3.All gene designations are used in SubtiWiki (36), and the genes can be retrieved there.
by both D-Asn and L-Asn.To test the role of MleN in the uptake of L-Asn, we constructed the Δdac ΔmleN mutant GP4177.Previous studies have shown that AimA is a major low affinity glutamate and serine importer in B. subtilis (3,4).Thus, we also used the Δdac ΔaimA mutant strain GP3054 (Fig. 2).The wild type as well as the single and double mutants were tested for growth in the absence and presence of D-and L-Asn.All strains grew well in the absence of any added amino acid.As expected, growth of the wild-type strain 168 and the Δdac mutant was inhibited In the presence of D-Asn, whereas the deletion of mleN restored growth in both genetic backgrounds.In addition, the Δdac ΔaimA double mutant was unable to grow in the presence of D-Asn, confirming that AimA does not contribute to the uptake of D-Asn.L-Asn was well tolerated by both the wild type and the ΔmleN mutant.In contrast, the Δdac mutant GP2222 was unable to grow if L-Asn was present.The analysis of the double mutants revealed that only the deletion of aimA but not of mleN confers resistance to L-Asn.The drop dilution assay (see Fig. 2) shows that the deletion of aimA allows growth in the presence of L-Asn; however, growth was not restored to the full level as compared to the wild-type strain, indicating that B. subtilis encodes additional L-Asn transporter(s).Taken together, our results indicate that MleN does not contribute to the uptake of L-asparagine, whereas AimA is able to transport this amino acid.This experiment demonstrates that the L-and D-forms of Asn, though structurally very similar, are not taken up by the same transporter.
To get more insights into the sensitivity of the Δdac mutant GP2222 to L-asparagine, we isolated suppressor mutants that were able to grow in the presence of 15 mM L-asparagine.Of four isolated mutants, two were subjected to whole-genome sequenc ing.Both mutants had frameshift mutations in the ktrD gene that result in the formation of truncated proteins (see Table 1).The ktrD alleles of the two remaining mutants were sequenced, and both had nucleotide substitutions that resulted in the replacement of Gly-314 by a Trp residue.The KtrD protein is the membrane subunit of the low affinity KtrCD potassium channel (39).It has been shown that the presence of glutamate results in high-affinity potassium uptake by KtrCD, resulting in intoxication of the Δdac mutant (19).It is, therefore, possible that either L-asparagine itself can also activate KtrCD or KtrCD is activated by the aspartate formed upon asparagine degradation.The possible link between KtrCD and asparagine transport will be addressed elsewhere.The fact that we did not find suppressor mutations with inactivated amino acid transporter genes supports the idea that multiple transporters for L-Asn are present in B. subtilis.

Identification of the AzlCD system as an potential exporter for L-Asn
Amino acids can be growth-inhibiting by themselves or due to the generation of harmful intermediates (16).To distinguish between these possibilities, we investigated the toxicity of L-asparagine in a strain that lacks the two asparaginases AnsA and AnsZ.This strain (GP4197) is sensitive to L-asparagine on minimal medium suggesting that the accumulation of L-asparagine itself is harmful for the cells (see Fig. 3).We observed suppressor mutant formation of the ΔansAB ΔansZ strain GP4197 after incubation at 42°C for 48 h with 5, 10, 15, and 30 mM L-asparagine.We isolated four suppressor mutants, one for each of the tested concentrations of L-asparagine.
To identify the mutations responsible for the resistance to L-Asn, we performed whole genome sequencing for the four suppressor mutants.The sequencing results are summarized in Table 1.In each mutant, we found single mutations that all affected the transcriptional repressor AzlB, which controls the expression of the azl operon.Two of the mutations (in GP4231 and GP4232) result in the formation of truncated proteins (due to a frame shift or a point mutation that converts a codon for glutamine to a stop codon), and two mutations (in GP4230 and GP4233) result in AzlB proteins with amino acid substitutions.Two of the mutations, the 8 base pair insertion of CATTAATG after nucleotide 37 (in GP4231) and the Asn24-Ser substitution (in GP4230) were already found in a previous study as a result of selection for resistance to histidine of the Δdac strain (11).It is, therefore, likely that the azlB gene is inactivated in all suppressor mutants.The inactivation of azlB causes overexpression of the genes of the azl operon that are expressed downstream of azlB (10,11).This involves the bipartite amino acid exporter AzlCD.To test the role of AzlB and AzlCD in the resistance to and export of L-Asn, we constructed strains carrying deletions of azlB and azlBCD and performed drop dilution assays with the resulting ΔansAB ΔansZ ΔazlB and ΔansAB ΔansZ ΔazlBCD triple mutants (GP4236 and GP4237, respectively) (Fig. 3).We observed improved growth when azlB was mutated or deleted, demonstrating that the inactivation of the azlB gene is responsible for the resistance to L-Asn.The additional deletion of the azlCD genes encoding the amino acid exporter resulted in the loss of growth.This finding suggests that AzlCD exports L-Asn in addition to 4-azaleucine and histidine (10, 11) (see Fig. 7).

AimA is the major transporter for L-Asn on complex medium
The B. subtilis strain SP1 is a prototrophic derivative of the laboratory strain 168 that has been used throughout this study.Due to its prototrophy, this strain is of interest for biotechnological applications (40).SP1 lacking the ansAB operon (strain BP269) is also sensitive to L-Asn, even on complex medium.Again, we were able to isolate suppres sor mutants that could tolerate the addition of 19 mM L-Asn to LB medium.Genome sequencing of two suppressor mutants (see Table 1) revealed two distinct mutations that affect the amino acid importer AimA.In the mutant BP391, there was an insertion of three bases in the aimA gene resulting in the insertion of a Thr residue after Ser-64 in AimA.In the second mutant, BP392, there was a single-nucleotide insertion in aimA, resulting in the expression of a truncated protein.In addition, BP391 had a second mutation that resulted in the substitution of Ala-96 by a Glu residue in the YetL protein, which acts as a transcriptional repressor to control the expression of the yetM (encodes a FAD-dependent monooxygenase) and yetL genes (41).Since these functions seem to be unrelated and the inactivation of AimA was already shown to be sufficient to cause resistance to L-Asn, the yetL mutation was not further investigated.In conclusion, the suppressor screen using the SP1-derived strain BP269 confirms the important role of AimA in the uptake of L-Asn.
As we have identified AimA as an importer for L-asparagine, we also deleted the aimA gene in the ΔansAB ΔansZ mutant in order to test whether the strain would still be sensitive to L-Asn stress on minimal medium.Interestingly, the deletion of the aimA gene only conferred limited resistance to L-asparagine (see Fig. 3, GP4239) if the bacteria grew on minimal medium.Again, this finding indicates that there are multiple transporters for L-Asn in B. subtilis.

Further adaptation to L-Asn identifies BcaP as additional L-Asn importer
In order to increase the selective pressure, we employed a disc diffusion assay of a strain that lacks the main known options to adapt to the presence of L-Asn (the ΔansAB ΔansZ ΔazlBCD ΔaimA quadruple mutant GP4245) with high concentrations of L-Asn (500 mM) (see Fig. 4).We observed suppressor mutant formation after 48 h of incubation at 42°C.We isolated a total of eight suppressors and sequenced the whole genomes of two of them.The results are summarized in Table 1.Both mutants carried mutations affecting the amino acid transporter BcaP.In strain GP4251, BcaP had a substitution of Ala-68 in Glu, and in GP4252, there was a base substitution that resulted in the direct formation of a stop codon and, thus, to a truncated BcaP protein.In addition, strain GP4251 had a second mutation in the ydeC gene that resulted in a substitution of His-43 to Gln in the AraC-type transcription factor YdeC of unknown function.We then sequenced the bcaP and ydeC alleles of the remaining six mutants.All had mutations in bcaP, while none of them had the mutation in ydeC.Therefore, we focussed the further analysis on bcaP.
All eight suppressors carried mutations in bcaP, the gene coding for the branched chain amino acid permease BcaP, which is involved in the uptake of isoleucine, valine, threonine, and serine (3,5,9).The mutations appear to make BcaP nonfunctional, which suggests a role of BcaP in the uptake of L-asparagine.When we compared our suppres sors as well as the effect of a deletion of the bcaP gene (in strain GP4267) with the original strain (ΔansABΔansZ ΔazlBCD ΔaimA) in a drop dilution assay, we found that both the suppressor mutants and the bcaP deletion mutant are fully resistant to L-Asn stress (see Fig. 5).Thus, the lack of both AimA and BcaP confers full resistance to L-Asn, indicating that these two transporters are responsible for the uptake of this amino acid.
The AimA amino acid importer and the AzlCD amino acid exporter have been shown to be involved in the uptake and export, respectively, of multiple amino acids.In order to characterize the suppressor mutants further, we also carried out an experiment with L-serine.L-Ser is toxic to B. subtilis in minimal medium (3,15).To test, whether AimA, BcaP, and AzlCD are also involved in the homeostasis of L-Ser, we examined the growth of the wild type strain 168, the ΔazlB mutant GP3600, the ΔansAB ΔansZ ΔazlBCD ΔaimA mutant GP4245 and of the two suppressor mutants GP4251 and GP4252 at different concentrations of L-Ser (see Fig. 6).As observed before (3), the wild-type strain was FIG 5 BcaP is also an Importer for L-Asn.Sensitivity of both suppressors (GP4251, GP4252) isolated via disc diffusion assay, as well as the ΔansAB ΔansZ ΔazlBCD ΔaimA ΔbcaP mutant (GP4267) to L-Asn is shown.The cells were grown in C-Glc minimal medium to an OD 600 of 1.0 and were then diluted 10-fold to create dilutions ranging from 10 −1 to 10 −6 .The dilution series was dropped onto C-Glc plates with and without L-asparagine (0 and 15 mM), respectively.The plates were incubated at 42°C for 48 h.

FIG 6
The isolated suppressors are also resistant against L-serine.Sensitivity of wild-type B. subtilis (168), the ΔazlB mutant (GP3600), the ΔansAB ΔansZ ΔazlBCD ΔaimA mutant (GP4245), and the two suppressors isolated from L-Asn stress (GP4251 and GP4252) to L-serine is shown.The cells were grown in C-Glc minimal medium to an OD 600 of 1.0 and were then diluted 10-fold to create dilutions ranging from 10 −1 to 10 −6 .The dilution series was dropped onto C-Glc plates with and without L-serine (0, 5, 15, and 30 mM, respectively) The plates were incubated at 42°C for 48 h.
unable to grow if L-Ser was present.The same result was obtained for the azlB mutant GP3600 indicating that the AzlCD amino acid exporter does not play a significant role in serine export.At L-Ser of 5 mM, the loss of the amino acid importer AimA in GP4245 provided a very faint protection against growth inhibition by L-Ser.However, both suppressor mutants that are otherwise isogenic to the progenitor GP4245, exhibited a strong resistance to the presence of serine.This oberservation is in good agreement with previous reports that BcaP and AimA make the major contributions to L-Ser uptake.Interestingly, the two suppressor mutants differ in their resistance to L-Ser at higher concentrations.The aimA bcaP double-mutant GP4252 barely grew at 15 mM serine or above.In contrast, GP4251 which has an additional mutation affecting the AraC family regulator YdeC grew well at 15 mM of L-Ser indicating that the mutation in ydeC contributes to amino acid resistance (see Discussion).

DISCUSSION
This work aimed to utilize the toxic effects of D-asparagine to identify uptake mecha nisms for both D-and L-asparagine, following the hypothesis that both enantiomers enter the cell via the same transporter, as it is the case for alanine in B. subtilis (12).The mutations in MleN revealed it to be a D-Asn importer.Even in the presence of MleN, L-Asn was well tolerated if the two transporters AimA and BcaP were missing indicating that MleN is not involved in the uptake of L-Asn.This demonstrates that, although structurally similar, D-and L-asparagine are taken up by different uptake systems.Our suppressor screen did not give any indication of the possiblity of D-Asn export in B. subtilis.This is interesting since we were easily able to find mutations in the azlB repressor gene that result in consitutive expression of the amino acid exporter AzlCD in the presence of L-Asn.This observation again supports the idea, that even though amino acid transport ers are often quite promiscous (2), it can not be taken for granted that transporters that import or export any given amino acid do so for the L-and D-forms of the amino acid.It should be noted that the high concentration of D-Asn used in the suppressor screen does not reflect physiological conditions.Therefore, it cannot be ruled out that B. subtilis encodes additional high-affinity transporters for D-Asn that might be active at lower concentrations.
A critical point for the identification of bacterial amino acid transporters is the use of strains and/or conditions that make the target amino acid toxic for the cells.In this study, we used strains lacking the second messenger c-di-AMP (Δdac) and strains that are unable to degrade asparagine (ΔansAB mutant) resulting in the intracellular accumula tion of L-Asn.Interestingly, the different strains seem to have experienced different selective pressures resulting in the identification of highly specific suppressor mutations according to the selection scheme.For the Δdac mutant, the well-established sensitivity to potassium was the major bottleneck, resulting in the isolation of ktrD mutants in all four cases (see below).For the strain lacking both asparaginases, the intracellular accumulation of L-Asn was the main issue which could be efficiently solved by activating the AzlCD amino acid exporter in all four independently isolated strains.Interestingly, the prototrophic strain SP1 lacking the asparaginase AnsA inactivated the AimA amino acid importer rather than activating AzlCD.This might result from the use of complex medium in which several amino acids share transporters.Moreover, the expression of the second transporter BcaP is repressed during growth on complex medium by the transcription factor CodY (42).Thus, under these conditions, AimA might be the only relevant transporter for L-Asn, resulting in its preferred inactivation to cope with L-Asn stress.Finally, when degradation or export of L-Asn were blocked, and one of the uptake systems (AimA) was also missing, mutations in a second transporter BcaP were selected in all eight cases.Thus, our results highlight the specificity of the outcomes that result from even subtle differences in the selective scenario.
In the case of the Δdac mutant GP2222, all suppressor mutants had the KtrD potassium channel inactivated.This protein imports potassium with low affinity, but it has a very high affinity for the ion in the presence of glutamate (4).It can, therefore, be concluded that the well-established potassium sensitivity of the Δdac mutant (4,18,43) is the major factor that causes growth inhibition in the presence of L-Asn, resulting in mutations that inactivate the strongly expressed KtrD potassium channel.Similarly, a mutation in ktrD was also selected if the Δdac mutant grew in the presence of high concentrations of histidine (11).While glutamate is one of the final degradation products of histidine, no glutamate is formed during the utilization of L-Asn.This suggests that either L-Asn or its degradation product L-Asp can also cause an increased affinity of KtrD for potassium, as shown for glutamate (4).
Our previous work has shown that the AimA protein is a major non-specific player in amino acid uptake (3,4).Based on our results with both the ΔdacΔaimA mutant GP3054 (see Fig. 2) and the SP1 ΔansAB mutant BP269 (see Table 1), AimA also plays the major role in the uptake of L-Asn.Thus, AimA is the major importer for glutamate, serine, and also asparagine (3,4).It is tempting to speculate that AimA might also be involved in the transport of other amino acids.Since AimA was discovered only recently, its presence may have masked previous genetic attempts to identify amino acid transporters in B. subtilis.AimA is a member of the large amino acid-polyamine-organocation superfamily (44).These proteins typically transport amino acids, but they may also be involved in the uptake of methyl-thioribose (MtrA) or potassium ions (KimA) (18,45).Interestingly, AimA seems to be limited to B. subtilis, as close orthologs of the protein are missing in most species, even in other Bacillus species.In B. subtilis, AimA has a paralog, YveA.This protein is expressed during sporulation in the forespore (46) and may, thus, be important for amino acid uptake once the spores start to germinate.However, based on the expression profile, YveA is unlikely to contribute to amino acid transport in growing cell.
Based on the suppressor mutants, the selective pressure caused by L-Asn was different in strains that are unable to degrade this amino acid from those that lack c-di-AMP.In the case of the 168-derived ansAB ansZ mutant all mutations affected the AzlB transcription repressor.Such mutations were previously shown to cause expression of the amino acid exporter complex AzlCD and to allow export of toxic histidine (11).Moreover, an amino acid export activity of AzlCD was also suggested for 4-azaleucine (10).Indeed, the loss of the exporter subunits in addition to the repressor results in an increased sensitivity toward L-Asn as compared to the azlB repressor mutant (see Fig. 3).This suggests that the AzlCD amino acid exporter is responsible for the resistance to L-Asn by exporting this amino acid.However, we cannot exclude the possibility that the accumulation of L-Asn in the ansAB ansZ mutant results in the formation of toxic products that are the actual substrate for export by AzlCD.Taking into account that AzlCD is a member of a family of amino acid exporters (47), and all the observations of this study it seems most likely that L-Asn is, indeed, the substrate for AzlCD.It is interesting to note that L-Asn is already the third amino acid that is likely to be exported by the AzlCD exporter.Moreover, expression of AzlCD also provides resistance to otherwise toxic diaminopropionic acid (Warneke et al., unpublished results), thus suggesting that this complex is a broad-spectrum amino acid exporter.So far, no conditions have been identified that would allow a substantial expression of the AzlCD amino acid exporter (35), and the acquisition of resistance to growth-inhibiting amino acids always depends on the inactivation of the azlB repressor gene.It is tempting to speculate that the untimely expression of the amino acid exporter might cause a loss of amino acids from the cell.Since the bacteria invest a lot of energy to either synthesize or import amino acids, their loss would be disadvantage for the bacteria.Indeed, the fitness of the azlB mutant is reduced in minimal media with ammonium as the single source of nitrogen, when the cells depend on de novo biosynthesis of amino acids (48).Thus, it might be a good strategy to encode the exporter in the genome but to keep the gene silent unless expression is required because a toxic amino acid causes the corresponding selective pressure.
As stated above, the identification of conditions that render an amino acid toxic is crucial to find the corresponding transport systems.Another step in this strategy is to use mutants that are unable to activate e.g., amino acid exporters or that are already defective in main transporters.Indeed, this approach allowed us to identify BcaP as a second importer for L-Asn.As for AimA, several substrates have been identified for BcaP.BcaP is the major transporter for the branched-chain amino acids isoleucine and valine as well as for threonine (3,5,9).Moreover, in addition to AimA, BsaP plays a minor role in serine uptake (3).The observation that BcaP is also involved in the acquisition of L-Asn suggests that it is also a more non-specific amino acid importer.
In one of the suppressor mutants that had acquired the mutations in bcaP, we found a second mutation in the ydeC gene encoding an unknown transcription regulator of the AraC family.It is tempting to speculate that YdeC controls the expression of an amino acid transporter.The observation that the presence of the ydeC mutation causes an increased resistance toward serine suggests again a general role for the correspond ing transporter.Indeed, this transporter, AexA (previously YdeD), can export β-alanine (Warneke et al., unpublished results) and probably also L-Asn and L-Ser.
The work described here has identified systems for the uptake and export of D-and L-Asn in B. subtilis (see Fig. 7).As for many other proteinogenic amino acids, L-Asn can be imported and exported by multiple broad-range transport proteins.experiments, B. subtilis was cultivated in LB, SM, MSSM, or C-Glc minimal medium (18).SM is a minimal medium that uses KH 2 PO 4 as buffer.MSSM is a modified SM medium in which KH 2 PO 4 was replaced by NaH 2 PO 4 and KCl was added as indicated.C-Glc is a chemically defined medium that contains glucose (1 g/L) as carbon source (52).The media were supplemented with ampicillin (100 µg/mL), kanamycin (10 µg/mL), chloramphenicol (5 µg/mL), spectinomycin (150 µg/mL), tetracycline (12.5 µg/mL), or erythromycin plus lincomycin (2 and 25 µg/mL, respectively) if required.

DNA manipulation and transformation
Transformation of E. coli and plasmid DNA extraction were performed using standard procedures (50).All commercially available plasmids, restriction enzymes, T4 DNA ligase, and DNA polymerases were used as recommended by the manufacturers.B. subtilis was transformed with plasmids, genomic DNA, or PCR products according to the two-step protocol (51).DNA fragments were purified using the QIAquick PCR Purification Kit (Qiagen, Hilden, Germany).DNA sequences were determined by the dideoxy chain termination method (50).

Construction of mutant strains by allelic replacement
Deletion of the ansAB, aimA, and bcaP genes was achieved by transformation of B. subtilis 168 with a PCR product constructed using oligonucleotides to amplify DNA fragments flanking the target genes and an appropriate intervening resistance cassette as described previously (53).The integrity of the regions flanking the integrated resistance cassette was verified by sequencing PCR products of about 1,100 bp amplified from chromosomal DNA of the resulting mutant strains.In the case of ansAB, the cassette carrying the resistance gene lacked a transcription terminator to ensure the expression of the downstream genes.

Phenotypic analysis
In B. subtilis, amylase activity was detected after growth on plates containing nutrient broth (7.5 g/L), 17 g Bacto agar/L (Difco), and 5 g hydrolyzed starch/L (Connaught).Starch degradation was detected by sublimating iodine onto the plates.
Quantitative studies of lacZ expression in B. subtilis were performed as follows: cells were grown in C-Glc medium.Cells were harvested at OD 600 of 0.5-0.8.β-Galactosidase specific activities were determined with cell extracts obtained by lysozyme treatment as described previously (51).One unit of β-galactosidase is defined as the amount of enzyme which produces 1 nmol of o-nitrophenol per min at 28°C.All β-galactosidase assays were performed in triplicate.
To assay the growth of B. subtilis mutants at different asparagine concentrations, multiple drop dilution assays was performed.Briefly, precultures in MSSM, C-Glc, or SM minimal medium at the indicated L-and D-asparagine concentration were washed three times and resuspended to an OD 600 of 1.0 in a 1× MSSM buffer, C-salts, or SM-buffer solution.A dilution series was then pipetted onto the respective minimal medium plates containing the desired asparagine concentration.All drop dilution assays were performed in triplicate.

Disc diffusion assay
For the disc diffusion assay, a preculture of GP4245 in C-Glc medium was prepared.The cells were harvested during exponential growth phase, washed three times and then resuspended to an OD 600 of 1.0 in a 1× C-salts solution.Then, 150 µL of the cells was spread onto C-Glc agar plates and left for 2 min to dry.A sterile filter disc was placed in the middle of the agar plates and 15 µL of highly concentrated (500 mM) L-asparagine solution was then pipetted onto the filter disc.A plate with 150 mM L-glutamate solution was used as positive control.The plates were incubated at 42°C for 48 h and then photographed and the suppressors isolated.
Frequently occurring hitchhiker mutations (56) and silent mutations were omitted from the screen.The resulting genome sequences were compared to that of our in-house wild type strain.Single-nucleotide polymorphisms were considered as significant when the total coverage depth exceeded 25 reads with a variant frequency of ≥90%.All identified mutations were verified by PCR amplification and Sanger sequencing.

Plasmid construction
To construct translational fusions of the potential mleN and ansA promoter regions to the promoterless lacZ gene, we used the plasmids pAC7 (58) and pAC5 (59), respectively.Briefly, the promoter regions were amplified using oligonucleotides that attached EcoRI and BamHI restriction to the ends of the products, and the fragments were cloned between the EcoRI and BamHI sites of pAC5 or pAC7.The resulting plasmids were pGP388 for mleN and pGP872 for ansA, respectively.

FIG 2
FIG 2 MleN is specific to D-asparagine import, while AimA is an L-asparagine importer.Sensitivity of wild-type B. subtilis (168), the mleN mutant GP1460 and Δdac mutants to D-Asn and L-Asn is compared on MSSM minimal medium.The cells were grown in MSSM minimal medium to an OD 600 of 1.0 and were then diluted 10-fold to create dilutions ranging from 10 −1 to 10 −6 .The dilution series was dropped onto MSSM plates with and without L-asparagine.The plates were incubated at 42°C for 48 h.

FIG 3 FIG 4
FIG3 AzlCD provides resistance to L-asparagine when AzlB is mutated.Sensitivity of the ΔansAB ΔansZ (GP4197) double mutant is shown compared to the suppressor GP4233, isolated from L-Asn stress, as well as the knockout strains for the exporter AzlCD and the importer AimA.The cells were grown in C-Glc minimal medium to an OD 600 of 1.0 and were then diluted 10-fold to create dilutions ranging from 10 −1 to 10 −6 .The dilution series was dropped onto C-Glc plates with and without L-asparagine (0, 5, 15, and 30 mM, respectively).The plates were incubated at 42°C for 48 h.

FIG 7
FIG 7 Asparagine metabolism in B. subtilis.L-asparagine enters the cell via the importers AimA and BcaP.Within the cell it is degraded to L-aspartate and then subsequently feeds into the tricarboxylic acid cycle (TCC) and production of L-glutamate via a newly discovered AspB-dependent bypass (49).L-asparagine is exported from the cell due to mutations in the regulators AzlB and YdeC.Export happens by the broad range amino acid exporters AexA (YdeD) and AzlCD.D-asparagine enters the cell via MleN and is toxic to B. subtilis.No form of degradation or export was detected.

TABLE 1
Identification of suppressor mutations a

TABLE 2
Activity of the putative mleN and ansA promoters

Units of β-galactosidase per µg of protein Addition to C-Glc minimal medium Strain Relevant genotype None CAA D-Asn
a L-Asn a D-Asp a L-Asp a The amino acids were added to a concentration of 5 mM.b NG, no growth.