The filamentous fungus Aspergillus niger contains two "differentially regulated" trehalose-6-phosphate synthase-encoding genes, tpsA and tpsB.

Two genes encoding trehalose-6-phosphate synthase were cloned from Aspergillus niger. tpsA was cloned using the Saccharomyces cerevisiae GGS1/TPS1 gene as a probe. It encodes a 517-amino acid polypeptide with 64-70% similarity to trehalose-6-phosphate synthase of S. cerevisiae, Kluyveromyces lactis, and Schizosaccharomyces pombe. Its transcription occurs constitutively and is enhanced on carbon-derepressing carbon sources, coinciding with the presence of a CreA-binding nucleotide motif in the 5′-noncoding region of tpsA. Disruption of tpsA only weakly reduces growth on glucose, and neither influences the glucose induction of a low affinity glucose permease nor interferes with the catabolite repression of a pectinase; it causes reduced the heat tolerance of conidia. tpsB was cloned by a polymerase chain reaction-based strategy. Its 480 amino acid sequence showed 76.5% identity to tpsA. Its transcription was hardly detectable at ambient temperatures but was enhanced strongly upon heat shock, which agrees with the presence of several copies of a C4T stress-responsive element in its 5′-upstream sequences. Hence the function of yeast GGS1/TPS1 has been split into two differentially regulated genes in A. niger, of which none appears to be involved in glucose sensing.

Two genes encoding trehalose-6-phosphate synthase were cloned from Aspergillus niger. tpsA was cloned using the Saccharomyces cerevisiae GGS1/TPS1 gene as a probe. It encodes a 517-amino acid polypeptide with 64 -70% similarity to trehalose-6-phosphate synthase of S. cerevisiae, Kluyveromyces lactis, and Schizosaccharomyces pombe. Its transcription occurs constitutively and is enhanced on carbon-derepressing carbon sources, coinciding with the presence of a CreA-binding nucleotide motif in the 5-noncoding region of tpsA. Disruption of tpsA only weakly reduces growth on glucose, and neither influences the glucose induction of a low affinity glucose permease nor interferes with the catabolite repression of a pectinase; it causes reduced the heat tolerance of conidia. tpsB was cloned by a polymerase chain reaction-based strategy. Its 480 amino acid sequence showed 76.5% identity to tpsA. Its transcription was hardly detectable at ambient temperatures but was enhanced strongly upon heat shock, which agrees with the presence of several copies of a C 4 T stress-responsive element in its 5-upstream sequences. Hence the function of yeast GGS1/TPS1 has been split into two differentially regulated genes in A. niger, of which none appears to be involved in glucose sensing.
Trehalose (␣-glucosido-1,1-glucose) is a nonreducing disaccharide found in such diverse organisms as bacteria, fungi, algae, plants, invertebrates, and insects (1). In fungi, trehalose is accumulated by mycelia during the stationary phase and by conidia, and it is metabolized rapidly once growth is resumed or germination initiated. For a long time, therefore, trehalose has been considered a reserve carbohydrate (2). However, an additional role of trehalose as a protectant against various conditions of physical stress such as heat, dehydration, and hyperosmotic shock has emerged more recently (3)(4)(5).
The biosynthesis of trehalose from UDP-glucose and glucose 6-phosphate involves two enzymatic steps catalyzed by a multienzyme complex (6). In the yeast Saccharomyces cerevisiae, at least three proteins participate in this complex: trehalose-6phosphate synthase (EC 2.4.1.15), trehalose-6-phosphate phosphatase (EC 3.1.3.12), and a regulatory protein (7). Interestingly, S. cerevisiae and Kluyveromyces lactis mutants in trehalose-6-phosphate synthase exhibit a pleiotropic phenotype; they are unable to grow on glucose or other rapidly fermentable sugars, exhibit strongly reduced sporulation, and display improper control of glucose-induced signaling pathways (8). Thevelein and co-workers (8) proposed that the trehalose-6-phosphate synthase protein may act as a component of a "glucose sensor" that transduces the glucose signal to the various signaling pathways. However, trehalose-6-phosphate synthase does not have such a role in the yeast Schizosaccharomyces pombe (9), and these authors suggest that the effects observed in the GGS1/TPS1/tps1 mutants are the result of a different susceptibility of hexokinase to inhibition by trehalose-6-phosphate in S. pombe and S. cerevisiae (9), which in S. cerevisiae GGS1/TPS1 mutants would lead to an unrestricted flux through hexokinase and consequently inhibition of growth.
Aspergillus niger is a filamentous fungus that has been used for decades for the industrial production of enzymes and organic acids (10,11). Both product accumulations are strongly, yet divergently, influenced by the presence of glucose (12)(13)(14). Several enzymes have been shown to take part in the control of glycolysis in A. niger (15), but under citric acid-producing conditions a major control point occurs at the level of hexokinase (16). Interestingly, the hexokinase of A. niger is inhibited only weakly by trehalose-6-phosphate (17). We have therefore investigated the role of trehalose-6-phosphate synthase in the growth of and glucose metabolism by A. niger.

EXPERIMENTAL PROCEDURES
Strains and Media-A. niger ATCC 11414 was used throughout this study. Conditions for strain maintenance and preparation of conidia were described previously (18). Growth tests were performed on minimal medium (19), containing 2.0% (w/v) agar without biotin. Liquid cultures of the fungus were grown at 30°C in 1-liter flasks containing 400 ml of complete medium, consisting of minimal medium plus 0.5% (w/v) yeast extract and 0.5% (w/v) peptone, and a carbon source (3% (w/v)). Flasks were inoculated with 2 ϫ 10 8 conidia and incubated on a rotary shaker at 250 rpm for 14 -18 h.
To induce heat stress response, A. niger cultures were pregrown in complete medium for 18 h, and aliquots of 20 ml of culture broth in 100-ml flasks were then transferred to a shaking water bath and incubated at 40°C for 60 min and 120 rpm.
Escherichia coli strain DH5a was used for propagation of plasmids and grown under standard conditions (20).
Cloning of the tpsA Gene-A BamHI/XbaI fragment of GGS1/TPS1 * This study was supported in part by Grant P10482-BIO from the Austrian Science Foundation (to C. P. K) and by Ö AD (Austrian Academic Exchange Fellowship program; acciones integradas) Contract 20.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) U07184 (Fig. 1) and U63416 (Fig. 6).
‡ Recipient of a research stipendium from the Technical University of Vienna.
§ To whom correspondence should be addressed. Tel.: 43-1-58801-4707; Fax: 43-1-581-6266; E-mail: ckubicek@fbch.tuwien.ac.at. was labeled with [ 32 P]dCTP by random priming and used to screen 40,000 plaques of the genomic library. Hybridization was carried out at 60°C. The final washing was done in 2 ϫSSC, 0.1% SDS at the same temperature. Inserts of hybridizing clones were analyzed by restriction mapping and subcloned.
Cloning of the tpsB Gene-Using sequence similarities of the GGS1/ TPS1/tps1 genes from S. cerevisiae (8), K. lactis (22), S. pombe (9), and A. niger tpsA (see "Results") two degenerated primers were designed for amplification: ggsdegI (5Ј-GGGTNCAYGAYTAYCAYYTNATG-3Ј) and ggsdegII (5Ј-GNGGNACNCCYTTDATRTARTC-3Ј). PCRs were performed in a total volume of 50 l containing 10 mM Tris-HCl, pH 8.8, 50 mM KCl, 4 mM MgCl 2 , 1% Triton X-100, 50 ng of genomic DNA, 25 pmol of each primer, 200 M dNTPs, and 0.5 units of Taq polymerase (Biomedica). The program for amplification consisted of a 1-min incubation at 95°C followed by 35 cycles of 1 min at 95°C, 1.5 min at 55°C and 1 min at 72°C and was terminated by a final extension for 7 min at 72°C. Protruding 3Ј-termini of PCR products were removed with Vent polymerase (New England Biolabs) according to the manufacturer's instructions and were analyzed by restriction cleavage with TaqI. A 450-bp fragment, verified as a fragment of tpsB by sequencing, was consequently used as a probe to screen 40,000 plaques of the genomic library. Hybridization was done at 64°C. The final washing was performed in 0.2 ϫ SSC, 0.1% SDS at the same temperature. Appropriate fragments of hybridizing clones were subloned, subjected to restriction mapping, and sequenced.
DNA Sequencing and Analysis-Sequencing was performed with the Sequenase version 2.0 system (U. S. Biochemical Corp.) and by means of an automatic sequencer (Applied Biosystems), using both universal primers as well as a series of primers specific for both strands of the tpsA locus. Sequence similarities were investigated using the BLAST server (23). Alignments were performed using the program MACAW (24) and improved by visual inspection.
Gene Disruption-A 4.5-kb NdeI/NotI fragment of the tpsA locus containing the whole tpsA structural gene and 2.3 kb and 0.38 kb of its 5Ј-and 3Ј-noncoding sequences, respectively, was cloned into pBluescript SKϩ (previously cut with SalI and NotI) to yield pGARP10. To disrupt the tpsA locus of A. niger, a 2.6-kb fragment of the E. coli hph gene flanked by the Aspergillus nidulans gpdA promoter and the trpC terminator was released from pAN7-1 (21) by restriction cleavage with SacI and HindIII. Protruding 5Ј-and 3Ј-termini were filled or removed, respectively, with Klenow enzyme, and the fragment was ligated into the unique EcoRV site of pGARP10, located within the tpsA coding sequences (395 bp) to yield pGARP11 (see Fig. 4A). The NotI/XhoI insert of pGARP11 was introduced into A. niger protoplasts by transformation as described by Yelton et al. (25). They were regenerated in liquid medium for 3 h at 30°C containing 2% (w/v) glycerol, 0.5% (w/v) yeast extract, and 0.5% (w/v) peptone for 3 h at 30°C followed by plating on solid minimal medium containing 300 g/ml hygromycin B (Calbiochem) and glycerol as a carbon source.
Standard Molecular Biological Techniques-Plasmid constructions, E. coli transformation, plasmid isolations from E. coli, restriction enzyme cleavage, isolation of chromosomal DNA, Southern and Northern blotting, preparation of 32 P-labeled DNA probes, hybridization of blots, and construction of the gene library in EMBL3 were performed according to standard protocols (20). Isolation of DNA fragments from agarose gels was performed with a Quiaex II kit (Quiagen) according to the manufacturer's protocols. RNA was isolated as described by Chomczynski and Sacchi (26) Poly(A) ϩ RNA was isolated from total RNA using the FastTrack 2.0 system (Invitrogen, San Diego, CA) according to the manufacturer's instructions.
Gel Mobility Shift Assays and Footprinting Techniques-A ScaI/ HpaII fragment of the 5Ј-noncoding sequences of tpsA was prepared and radioactively labeled with [ 32 P]dCTP and appropriate dNTPs by filling in with the Klenow fragment of DNA polymerase. A CreA::GST fusion protein was obtained by overexpression of a NcoI fragment (from nucleotides 103-722) of the A. nidulans creA gene (27) containing the two C 2 H 2 zinc fingers and an alanine-rich region as a GST fusion in E. coli, contained in vector pGEX-CreA, following the expression and purification protocol of Kulmburg et al. (28). Polyacrylamide gel shifts and methylation protection footprinting were performed as described previously (29).
Determination of Trehalose-Mycelia were thereafter harvested by filtration, ground under liquid nitrogen, and trehalose extracted by their suspension in distilled water at 95°C for 30 min. After centrifugation (15 min, 12,000 ϫ g, 18°C), the extracts were passed through a 0.45-m filter. Trehalose was determined using the method in Ref. 33 except that commercial porcine kidney trehalase (Sigma) was used.
Electrophoretic Techniques-For demonstration of individual pectinases in the supernatant, samples from the culture broth were subjected to SDS-polyacrylamide gel electrophoresis followed by Western blotting to nitrocellulose and immunological detection as described previously (34). A polyclonal antiserum was used to detect pectinase PL II (35).

RESULTS
Cloning of the A. niger tpsA Gene-To clone the A. niger equivalent of GGS1/TPS1 we screened a genomic EMBL3 library of A. niger by heterologous hybridization with the S. cerevisiae GGS1/TPS1 gene and obtained seven positive clones. Restriction mapping and Southern hybridization showed that all clones were part of the same gene locus (data not shown). Hybridizing DNA fragments were subcloned into pGEM-5Zf(ϩ) and sequenced. The clones revealed the presence of an ORF of 1,825 bp ( Fig. 1), which was interrupted by four introns (identified by the presence of consensus sequences of splicing signals for filamentous fungi; (36)). High stringency Southern hybridization of A. niger genomic DNA cleaved with XhoI, which does not cut within the gene, with a tpsA-specific probe yielded a single band, suggesting that tpsA occurs as a single copy in the A. niger genome. The predicted 517-amino acid sequence shows high identity to the GGS1/TPS1 gene products from S. cerevisiae (64.0%), K. lactis (64.3%), and S. pombe (69.5%).
The 5Ј-noncoding region of tpsA contains two putative CreA binding sites (consensus 5Ј-SYGGRG-3Ј; (28,37)) at positions Ϫ169 and Ϫ177, organized as a tandem repeat. Binding sites for the yeast carbon catabolite repressor protein MIG1, whose DNA binding domain shows high similarity to that of A. nidulans and A. niger CreA (27,38), are also found in the promoter of GGS1/TPS1 from S. cerevisiae (8) and K. lactis (22). Using a CreA::GST fusion protein, binding to a tpsA restriction fragment resembling bp Ϫ58 to Ϫ265 (ScaI-HpaII) was shown by gel retardation assays and by methylation protection footprinting in vitro (Fig. 2). To investigate whether transcription of tpsA was indeed regulated by carbon catabolite repression, Northern analysis of mRNA, harvested from A. niger grown on various carbon sources, was carried out (Fig. 3A). tpsA expression was observed on all carbon sources tested, but the ratio between the tpsA and the actA (control) transcripts was higher during growth on citrate and arabinose than on glucose and lactose, which would be consistent with partial regulation by carbon catabolite.
Transcription of the S. cerevisiae and the S. pombe GGS1/ TPS1 is known to be inducible by heat stress (8,9). We did not find any of the consensus sequences for binding of heat shockregulating proteins in the tpsA promoter. However, a single copy of a C 4 T motif (consensus core sequence CCCCT), which mediates transcription in response to various conditions of stress in S. cerevisiae (39), was present at Ϫ386. Northern analysis of A. niger cultures subjected to 40°C heat shock showed that tpsA transcription declined immediately upon transfer to 40°C (Fig. 3B). We conclude that tpsA transcription is not triggered by heat shock. Sequences resembling the consensus for binding of yeast GCR1 (CCTTC (40)) were found at Ϫ249 and Ϫ457, but those for binding of other activators of glycolytic genes were not found.
Disruption of the tpsA Gene-To prove that tpsA encodes a trehalose-6-phosphate synthase of A. niger and to study its function in this fungus, we replaced it with a tpsA derivative into which the E. coli hph gene from pAN7-1 had been inserted by transformation (Fig. 4A). Transformants, in which the tpsA gene had been replaced by the disrupted gene, were identified by the presence of a 9.6-kb DNA fragment, which replaced the wild type 7.0-kb band (Fig. 4B). One of these, strain A. niger tpsA⌬1-3, was selected for further analysis. When grown at 30°C on glucose or glycerol its trehalose-6-phosphate synthase activity was below the detection limit (Fig. 5), which proves that the gene encodes trehalose-6-phosphate synthase. Interestingly, the mycelial trehalose content at 30°C was reduced strongly compared with the parental strain, but it was not depleted completely. In view of the clear reduction of trehalose-6-phosphate synthase activity, this suggests the presence of alternative enzymes/pathways for trehalose biosynthesis in A. niger. This assumption was supported strongly by the findings that very high trehalose-6-phosphate synthase activity was observed in the disruptant strain upon growth of A. niger at 40°C, which was virtually similar to that of the parent strain where it represented a 6-and 15-fold increase (on glucose and glycerol as carbon source, respectively). This was also reflected in drastically increased mycelial trehalose concentrations. Interestingly, although this trehalose-6-phosphate synthase activity was twice as high on glycerol as the carbon source, the trehalose content on this carbon source was 30 -40% lower, which may be due to a lower glucose-6-phosphate supply and/or a different regulation of trehalose-6-phosphatase by glycerol and glucose.
Disruption of GGS1/TPS1 in S. cerevisiae and K. lactis results in an impairment of growth on glucose, fructose, and other rapidly fermentable sugars (8,22). To see whether there was a similar phenotype in A. niger tpsA⌬1-3, the wild type as well as the disruptant were pregrown for 24 h at 30°C on solid minimal medium under derepressing conditions (0.1% glycerol) and then transferred to fresh medium containing different concentrations of glucose; the increase in colony diameter was measured over a period of 5 days. Virtually the same growth rate on glucose was observed with the wild type strain and strain tpsA⌬1-3 at 30°C, but the disruptant grew slightly lower at 37 and at 40°C. Similar results were also obtained with glycerol as a carbon source. No significant differences in the morphology of the growing hyphae were observed between the parent strain and the disruptant at either temperature.
In contrast to S. cerevisiae and K. lactis, the only effect of disruption of tsp1 in S. pombe was a failure to germinate in a medium with glucose as a carbon source (9). A. niger tpsA⌬1-3 showed no retardation or even deficiency in germination on glucose, however (data not shown).
The data described above suggest a role of trehalose-6-phosphate synthase A for mycelial growth at higher temperature. To study the possibility that the reduced trehalose content in A. niger tpsA⌬1-3 also affects its viability at temperatures where growth no longer takes place, we incubated conidiospores of the wild type and the disruptant strain at 50 and 55°C, respectively, and thereafter analyzed their ability to germinate at 30°C. At 50°C germination of the disruptant exhibited a 27% reduction compared with the parent strain (94 versus 70% of the control), and this effect was even more dramatic at 55°C (7.4 versus 2.6% of the control; i.e. 32%). Considering the fact that the trehalose content in A. niger tpsA⌬1-3 is reduced by 56%, these data support the assumption that trehalose contributes to the heat stability of A. niger conidiospores.
To learn whether the functional impairment of tpsA leads to glucose derepression (as has been shown for yeast invertase and ␣-glucosidase (8)), A. niger tpsA⌬1-3 and its parent strain were grown on glucose as a carbon source, and the formation of pectinase A (which is subject to glucose repression (35)) was determined by Western blotting and immunostaining. Under these conditions, pectinase A remained below the limit of detection in both strains, whereas it was demonstrated clearly during growth on pectin (data not shown). We conclude that a disruption of tpsA in A. niger does not lead to carbon catabolite derepression.
A. niger has been shown to respond to the presence of elevated concentrations of glucose or sucrose (Ͼ5%, w/v), by increasing the glycolytic flux and citric acid accumulation (14,18). This effect is, among others, reflected in the induction of a low affinity glucose permease by elevated glucose concentrations (32). However, in the presence of 10% glucose the low affinity transporter was formed in the wild type as well as the disruptant strain, and both its K m (4.1 mM) as well as V max (0.14 mol/min/mg dry weight) were virtually the same. We therefore conclude that tpsA is not involved in the signaling of high glucose concentrations to A. niger.
Cloning and Transcriptional Analysis of the A. niger tpsB Gene-The fact that A. niger tpsA was only expressed constitutively, yet the disruptant strain still contained high trehalose-6-phosphate synthase activity at 40°C, suggested the presence of a second, heat-inducible trehalose-6-phosphate synthase-encoding gene in A. niger. Using the predicted protein sequences of the three GGS1/TPS1 genes from S. cerevisiae, K. lactis, and S. pombe as well as of tpsA, a pair of degenerated oligonucleotide primers was designed which corresponded to the nucleotide sequences encoding amino acids 149 -156 and 190 -198 in the tpsA polypeptide. PCR amplified a single nucleotide fragment of 0.5 kb. Restriction fragment length polymorphism analysis with HinfI and TaqI showed that this fragment corresponded to two different amplicons and that the restriction fragments of one of them were in accordance with those predicted from the tpsA sequence. The other amplicon was unique. Using a 450-bp fragment of this amplicon as a probe, the A. niger library was screened under stringent con-  (panel B). Panel A, A. niger wild type was grown for 18 h on complete medium, containing the carbon sources as indicated. 100 g of total RNA was applied to each lane. A PCR fragment covering the entire tpsA ORF was used as a probe. The blot was rehybridized with restriction fragment containing the whole ORF of the A. nidulans actA gene as control. Hybridization was carried out at 42°C. The final washing was generally done in 0.3 ϫ SSC, 0.1% SDS at 60°C for 40 min, except for the actA control where the temperature was lowered to 55°C. Panel B, A. niger wild type was grown at 30°C on complete medium containing either glucose or glycerol as carbon source and the temperature then shifted to 40°C for 1 h. RNA was isolated at the times indicated, and poly(A) ϩ RNA corresponding to 500 g of total RNA was applied to each lane. Hybridization conditions were as described in panel A. ditions and three positive clones were obtained. Subcloning and sequencing showed that the tpsB gene is contained in an ORF of 1,732 bp (Fig. 6). Four putative introns occur at the same relative position as in tpsA. The predicted 480-amino acid sequence shows 76.5% identity to tpsA, and 64.0, 64.0, and 65.1% identity to the GGS1/TPS1 gene products from S. cerevisiae, K. lactis, and S. pombe, respectively (Fig. 7).
In contrast to tpsA, the 5Ј-nontranscribed sequences of tspB did not contain nucleotide sequences homologous to CreA binding sites. In analogy with tpsA, sequences resembling the consensus for binding of yeast GCR1 (CCTTC (40)) were also present at Ϫ17, Ϫ152, Ϫ300, Ϫ375, and Ϫ412. Interestingly, the C 4 T motif identified in tpsA occurred in five copies in tpsB, i.e. at Ϫ14, Ϫ103, Ϫ224, Ϫ230, and Ϫ343, respectively. Probing the Northern blots of Fig. 3 with the tpsB gene revealed only a low level of transcript at 30°C, whose level was even lower on arabinose and citrate, but the transcript increased strongly upon transfer to 40°C (Fig. 8). These findings are consistent with the assumption that tpsB encodes the heat-induced trehalose-6-phosphate synthase activity of A. niger. DISCUSSION We have isolated from A. niger two different genes encoding trehalose-6-phosphate synthase (tpsA and tpsB), as indicated by the fact that the genes have roughly 65% identity at the amino acid level with the GGS1/TPS1 genes of S. cerevisiae (8) and K. lactis (22) and the TPS1 gene of S. pombe (9). Furthermore, disruption of tpsA virtually reduces the trehalose-6-phosphate synthase activity under conditions where this gene is selectively expressed (e.g. growth on glucose as a carbon source) to zero.
Since tpsA is expressed more strongly during vegetative growth on a variety of carbon sources at ambient temperatures than tpsB we considered it a more likely candidate for exerting a regulatory effect on glucose metabolism similar to that of GGS1/TPS1 in yeast. The phenotype of tpsA gene disruption in A. niger, however, was quite different from that produced by the disruption of the GGS1/TPS1 genes in S. cerevisiae and K. lactis and resembled that observed in S. pombe in some aspects. In agreement with the latter but in contrast to S. cerevisiae and K. lactis, a disruption of tpsA did not influence the capacity of A. niger to grow on glucose. Blá zquez et al. (9) attributed this difference between S. cerevisiae (and K. lactis) and S. pombe to the different behavior of the hexokinases of these three yeasts toward trehalose 6-phosphate, a strong inhibitor of S. cerevisiae and K. lactis hexokinases, but not that of S. pombe (9). Such an explanation also may be valid for A. niger, whose hexokinase has a rather low K i for trehalose 6-phosphate (17) and is virtually not inhibited by glucose 6-phosphate (41). The low K i for trehalose 6-phosphate is reflected by the findings that disruption of tpsA leads to an increase in glycolytic flux at very high sugar concentrations (Ͼ5%, w/v) only, apparently because an intracellular concentration of trehalose 6-phosphate, sufficient for inhibiting hexokinase, can only be accumulated under these conditions (17). All of these findings argue against a regulatory role of tspA in A. niger glycolysis during vegetative growth on conventional media containing low carbon source concentrations.
Trehalose has been attributed a dual role as a reserve carbohydrate (2) and as a protectant of proteins against denaturation by dehydration (5). Consistent with the first role, disruption of tps1 in S. pombe prevented germination of spores (9). In contrast, disruption of tpsA in A. niger had no effect on the germination of conidia. These differences may be because disruption of tpsA does not produce a null phenotype of trehalose-6-phosphate synthase activity because of the presence of tpsB. Interestingly, tpsA disruption and the accompanying reduction of the conidial trehalose concentration lead to a reduction in the heat stability of the conidia. This suggests that tpsA contributes to trehalose formation during conidiation and that its lack cannot be fully compensated by tpsB.
In contrast to tpsA, the tpsB transcript was hardly detectable during vegetative growth of A. niger but accumulated strongly during heat shock. A similar behavior has been described for S. cerevisiae GGS1/TPS1 (8) and S. pombe tps1 (9) gene expression. This heat shock-triggered gene expression is consistent with the presence of an element with the sequence AAGGG-GAT in the 5Ј-noncoding sequences of GGS1/TPS1 (8), which confers regulation of certain S. cerevisiae genes by heat shock, nitrogen starvation, and the RAS-cAMP pathway (39,42). Several copies of this motif are also present in the 5Ј-upstream sequences of tpsB, and it is tempting to speculate that this motif confers regulation by heat shock also in A. niger. However, as the functionality of this motif in Aspergillus and other filamentous fungi has not yet been demonstrated, this assumption must still be treated cautiously. The fact that a single AAGGGGAT motif also occurs in the 5Ј-upstream regions of tpsA, whose expression is not triggered by heat shock, indicates that its mere presence in A. niger does not imply functionality and suggests at least that they may be subject to position effects. Kobayashi and McEntee (42) reported that a single copy of the C 4 T motif is active in S. cerevisiae, yet the presence of two such motifs evokes a more than additive response.
We have shown here for the first time that the biological function of a gene that is present in a single copy in several yeasts (S. cerevisiae, K. lactis, S. pombe) is fulfilled by two genes in the filamentous fungus A. niger, which encode highly similar polypeptides but are regulated differently. A comparison of the amino acid sequences of the tpsA and tpsB gene products does not provide a clue as to the reason for this. Interestingly, while this paper was prepared for publication, Borgia et al. (43) reported on the presence of two trehalose-6phosphate phosphatases in A. nidulans but did not investigate whether the two gene products may have different functions. It will be interesting to see if this is a general feature of multicellular fungi or a feature unique to aspergilli. FIG. 8. Northern blot analysis of tpsB transcription during growth on different carbon sources (panel A) and upon exposure to heat stress (panel B). Panel A, A. niger wild type was grown for 18 h on complete medium containing the carbon sources as indicated. 100 g of total RNA was applied to each lane. A 1.56-kb HindIII/ SalI fragment (corresponding to position Ϫ295 and ϩ1283 of the tpsB sequence) was used as a probe. The blot was rehybridized with a restriction fragment containing the whole ORF of the A. nidulans actA gene as control. Hybridization was carried out at 42°C. The final washing was generally done in 0.3 ϫ SSC, 0.1% SDS at 60°C for 40 min, except for the actA control where the temperature was lowered to 55°C. Panel B, A. niger wild type was grown at 30°C on complete medium containing either glucose or glycerol as carbon source and the temperature then rose to 40°C for 1 h. RNA was isolated at the times indicated, and poly(A) ϩ RNA corresponding to 500 g of total RNA was applied to each lane. Hybridization conditions were the same as in panel A.