Fructose utilization and pathogenicity of Spiroplasma citri: characterization of the fructose operon

Gene 252 (2000) 61–69 www.elsevier.com/locate/gene

Fructose utilization and pathogenicity of Spiroplasma citri: characterization of the fructose operon Patrice Gaurivaud, Fre´de´ric Laigret *, Monique Garnier, Joseph M. Bove Laboratoire de Biologie Cellulaire et Mole´culaire, Institut de Biologie Ve´ge´tale Mole´culaire, Institut National de la Recherche Agronomique, Universite´ Victor Segalen Bordeaux 2, Domaine de la Grande Ferrade, 71 avenue Edouard Bourleaux, PB 81, 33883 Villenave d’Ornon cedex, France Received 8 March 2000; received in revised form 10 May 2000; accepted 24 May 2000 Received by D.L. Court

Abstract Transposon Tn4001 mutagenesis of Spiroplasma citri wild-type (wt) strain GII-3 led to the isolation and characterization of non-phytopathogenic mutant GMT 553. In this mutant, transposon Tn4001 is inserted within the first gene of the fructose operon. This operon comprises three genes. The first gene ( fruR) codes for a putative transcriptional regulator protein belonging to the deoxyribonucleoside repressor (DeoR) family. Sequence similarities and functional complementation of mutant GMT 553 with different combinations of the wt genes of the fructose operon showed that the second gene ( fruA) codes for the permease of the phosphoenolpyruvate:fructose phosphotransferase system (fructose PTS ), and the third, fruK, for the 1-phosphofructokinase (1-PFK ). Transcription of the fructose operon in wt strain GII-3 resulted in two messenger RNAs, one of 2.8 kb and one of 3.8 kb. Insertion of Tn4001 in the genome of mutant GMT 553 abolished transcription of the fructose operon, and resulted in the inability of this mutant to use fructose. Functional complementation experiments demonstrated that fructose utilization was restored with fruR–fruA–fruK, fruA–fruK or fruA only, but not with fruR or fruR–fruA. This is the first time that an operon for sugar utilization has been functionally characterized in the mollicutes. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Complementation; DeoR protein; Phosphofructokinase; PTS

1. Introduction Mollicutes are wall-less prokaryotes, phylogenetically related to Gram-positive eubacteria with low guanine and cytosine content in their DNA ( Weisburg et al., 1989). Spiroplasmas are mollicutes characterized by helical morphology and motility. Certain spiroplasmas

Abbreviations: aa, amino acid(s); DeoR, deoxyribonucleoside repressor; IS, insertion sequence; Km, kanamycin; LB, Luria–Bertani; ORF, open reading frame; oriC, origin of DNA replication; p, plasmid; P, promoter; PFK, phosphofructokinase; 1-PFK, 1-phosphofructokinase; 6-PFK, 6-phosphofructokinase; PCR, polymerase chain reaction; PPi, inorganic pyrophosphate; PTS, phosphoenolpuryvate:sugar phosphotransferase system; SDS, sodium dodecyl sulfate; Tc, tetracycline; Tn, transposon; T, terminator; wt, wild-type. * Corresponding author. Present address: Laboratoire de Biologie Cellulaire et Mole´culaire, Institut de Biologie Ve´ge´tale Mole´culaire, Institut National de la Recherche Agronomique, Domaine de la Grande Ferrade, 71 avenue Edouard Bourleaux, PB 81, 33883 Villenave d’Ornon cedex, France. Tel.: 33-5-56-84-31-50; Fax: 33-5-56-84-31-59. E-mail address: [email protected] (F. Laigret)

are plant pathogens. Spiroplasma citri, the causal agent of citrus stubborn disease (Saglio et al., 1973), is transmitted from plant to plant by phloem-feeding leafhoppers such as Circulifer haematoceps and Circulifer tenellus (Calavan and Bove´, 1989), and is restricted to the sieve tube elements of the phloem tissue. Little is known about the mechanisms of pathogenicity of S. citri, even though toxins, shortage of plant auxin, and lactic acid production have been suggested as possible virulence factors (Daniels, 1983; Chang, 1998). In spite of these studies, the molecular mechanisms of the interactions between the spiroplasma and the host plant remain obscure. In an attempt to identify genetic determinants involved in S. citri phytopathogenicity, mutagenesis by random insertion of transposon Tn4001 into the genome of wild-type (wt) strain GII-3 ( Foissac et al., 1997b), as well as insect-transmission and pathogenicity assays were developed ( Foissac et al., 1997a). As many as 257 transpositional mutants were screened for their ability to (i) multiply in the leafhopper vector C. haematoceps, (ii) be transmitted to the periwinkle plants and (iii) multiply and induce symptoms in

0378-1119/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S0 3 7 8 -1 1 1 9 ( 0 0 ) 0 0 23 0 - 4

62

P. Gaurivaud et al. / Gene 252 (2000) 61–69

these plants ( Foissac et al., 1997a). One mutant, GMT 553, was found to be non-phytopathogenic for periwinkle plants (Foissac et al., 1997a). While the wt strain GII-3 induced symptoms during the first week following the insect-transmission period, it took 4 weeks for symptoms to appear with the mutant, and the symptom was associated with reversion to wt by spontaneous excision of transposon Tn4001 from the genome of mutant GMT 553. In addition, the rate of multiplication of mutant GMT 553 in the plants was approx. twofold lower than that of the wt strain. Here, we show that Tn4001 is inserted in the fructose operon, and the mutant is unable to use fructose. This result led us to characterize the fructose operon genes by sequence comparisons, complementation experiments, and determination of phosphoenolpyruvate:fructose phosphotransferase system (fructose PTS ) and 1-phosphofructokinase (1-PFK ) activities. We show elsewhere that fructose utilization by the spiroplasma in the phloem is linked to phytopathogenicity (Gaurivaud et al., 2000).

2. Materials and methods 2.1. Bacterial strains and plasmids Escherichia coli XL1 Blue MRF ∞ Kan (Stratagene, LAJolla, CA, USA) was used as host strain for cloning experiments and plasmid propagation. It was grown in Luria–Bertani (LB) medium (Sambrook et al., 1989) or on LB plates containing 10 mg/ml of km and 50 mg/ml of ampicillin. Electroporation was used for transformation of E. coli (Dower et al., 1988). S. citri GII-3 wt strain was originally isolated from C. haematoceps leafhoppers captured in Morocco ( Vignault et al., 1980). Mutant GMT 553 was obtained by mutagenesis of the wt strain GII-3 using the transposon Tn4001 ( Foissac et al., 1997a, b). S. citri was grown in SP-4 medium ( Whitcomb, 1983). Mutant GMT 553 was grown in the presence of 100 mg/ml of gentamycin. HSI medium ( Whitcomb, 1983) [mycoplasma broth base 15 g/l (Gibco/BRL, Gaithersburg, MD, USA), PPLO serum fraction 1% (v/v) (Difco, Detroit, MI, USA), sorbitol 7% (w/v), phenol red 30 mg/l, penicillin 2.105 U/ml ] was used for carbohydrate fermentation experiments. Carbohydrates were added to medium HSI at the final concentration of 0.5% (w/v). Plasmid pBS was purchased from Stratagene. PGEM-T vector (Promega Corporation, Madison, WI, USA) was used for cloning amplified DNA. Replicative plasmid pBOT carrying the S. citri oriC region and the tetracycline resistance determinant tetM was described previously (Renaudin et al., 1995). Plasmid pBOTP1 was obtained by subcloning the spiroplasmal fibril gene promoter ( Williamson et al., 1991) into plasmid pBOT (Renaudin, unpublished

results). These plasmids were used for gene expression experiments in S. citri. Transformation of S. citri by electroporation with pBOT and pBOTP1 was as previously described (Renaudin et al., 1995). S. citri transformants were selected by plating on SP-4 solid medium supplemented with 2 mg/ml of tetracycline. Transformants were propagated in SP-4 liquid medium for at least 15 passages until integration of plasmid into the chromosomal DNA had occurred as verified by Southern blot hybridizations. 2.2. DNA analyses Total DNAs from spiroplasmas were isolated as follows. 30 ml of spiroplasmal culture were collected by centrifugation and resuspended in 270 ml of PBS (phosphate buffered saline) buffer. Cells were lysed by adding 30 ml of 10% SDS and incubated for 30 min at 37°C. DNA was further purified by phenol–chloroform deproteinization and ethanol–acetate precipitation. Restricted DNA was fractionated by agarose gel electrophoresis, blotted onto charged nylon membranes by the alkali transfer procedure and hybridized with appropriate [a-32P]dATP-labelled probes using standard stringency conditions (Sambrook et al., 1989). DNA sequencing was performed with the T7 DNA sequencing kit (Pharmacia, Uppsala, Sweden), [35S ]dATP and appropriate primers. Search for similarity in GenBank involved the Blast program (Altschul et al., 1990). The ProDom database was used to study protein-domain arrangements (Corpet et al., 1999). The GOR method was used to study secondary structure (Garnier et al., 1996). The sequence has been deposited in GenBank/ EMBL under Accession No. AF202665. 2.3. Determination of Tn4001 insertion site and relevant plasmids Tn4001 is bordered at both ends by insertion sequence IS256 (Byrne et al., 1989). A hybridization probe containing IS256 was obtained from plasmid pMUT ( Foissac et al., 1997b) and used to screen a HindIII DNA library of mutant GMT 553 obtained in plasmid pBS. Two inserts hybridizing with the probe were selected: H12.3 (2.5 kbp) was from recombinant plasmid pH12.3 and H13.2 (3.3 kbp) came from pH13.2 (Fig. 1). The sequences of inserts H12.3 and H13.2 were determined and revealed the presence of IS256 sequences. In order to clone and analyze the corresponding region of the wt strain, a HindIII DNA library from S. citri wt strain GII-3 was produced in plasmid pBS, and probed with the 0.5 kbp HindIII–Sau3A fragment (P0.5) obtained from insert H13.2 ( Fig. 1). In this way, a 3.3 kbp insert, named GH3 ( Fig. 1), was detected and the recombinant plasmid was designated pGH3. The nucleotide sequence of insert GH3 was determined and

P. Gaurivaud et al. / Gene 252 (2000) 61–69

63

Fig. 1. Physical map of S. citri fructose operon. The locations of recognition sites for restriction endonucleases deduced from sequence analysis are shown. The infB, fruR, fruA and fruK encode the IF2 translation factor, hypothetical regulatory protein (FruR), fructose permease (FruA) and 1-PFK (FruK ), respectively. PF1 and PF2 are primers used for amplification of the infB–fruR intergenic region carrying the putative fructose operon promoter (PFRU ). Inserts used in cloning of the insertion region and complementation experiment are shown. The names of the strains are given in parentheses.

was found to contain the same open reading frames (ORFs) as those present on insert H13.2 and H12.3, i.e. the 3∞ end of infB, the complete fruR and the 5∞ half of fruA. The 3∞ half of fruA was cloned using the P0.5 probe ( Fig. 1) on a Sau3A DNA library of the wt strain; plasmid pS32 containing an insert of 1.0 kbp (S32, Fig. 1) was obtained. However, because this fragment did not contain the 3∞ end of fruA, a new screening procedure using as a probe the 0.2 kbp HindIII–EcoRI fragment (P0.2, Fig. 1) of insert S32, was carried out on a TaqI DNA library. Insert GT6 (2.8 kbp, Fig. 1) from plasmid pGT6 was detected and sequenced. Sequence analysis showed that pGT6 carried the 3∞ end of fruA and a fourth ORF, fruK. 2.4. Plasmids for functional complementation For functional complementation of mutant GMT 553 with fructose operon genes from wt S. citri, several plasmids were constructed containing various combinations of the fructose operon genes. Plasmid pRAK containing all three fructose operon genes ( fruRAK, Fig. 1) was obtained by cloning the BamHI–EcoNI fragment from plasmid pGH3 into plasmid pGT6. Plasmid pR was obtained by partial digestion of pRAK with HpaI ( Fig. 1). Plasmid pRA resulted from total digestion of pRAK with BstXI and partial digestion with PstI; single-stranded protruding ends were removed

with the Klenow fragment of DNA polymerase I, and the resulting blunt ends were ligated with T4 DNA ligase. For plasmid pAK, pRAK was submitted to total digestion with ThaI and partial digestion with SpeI ( Fig. 1), followed by re-insertion into SmaI–HincIIlinearized pBS. Plasmid pA was derived from pAK by total digestion with BstXI and partial digestion with PstI, treated by the Klenow enzyme, and blunt ends ligated with T4 DNA ligase. Plasmid pK came from pGT6 submitted to total digestion with XbaI and BstXI, Klenow enzyme treatment and T4 DNA ligase. For transformation of S. citri, inserts RAK and RA were excised from their respective plasmids by PstI digestion and inserted into PstI-linearized plasmid pBOT, yielding plasmids pBOT-RAK, and pBOT-RA. Inserts AK, A and K do not possess the promoter sequence of the fructose operon (PFRU on Fig. 1). Therefore, this sequence was PCR-amplified with primers PF1 (5∞-GGAGCTGCAGGTGGCC-3∞) and PF2 (5∞-TTGTTTCTGCAGGATCCCACCAAC-3∞) corresponding, respectively, to nucleotides 1227–1242 and 1384– 1407 in the nucleotide sequence of the fructose operon. These sequences were used as mutagenic primers to create a unique BamHI site (boldface in PF2) and two PstI sites (underlined in PF1 and PF2). The resulting 180 bp sequence was cloned into plasmid pGEM-T, reisolated by PstI or SphI treatment and inserted into PstI or SphI linearized plasmid pBOTP1. In this way,

64

P. Gaurivaud et al. / Gene 252 (2000) 61–69

the fibril P1 promoter of pBOTP1 was replaced by the fructose operon promoter PFRU, yielding pBOTPFRU. Plasmids pBOT-AK, pBOT-A and pBOT-K were obtained by replacing the SphI–BamHI fragment of pBOT-PFRU with fragment AK, A and K excised, respectively, from pAK, pA and pK by SphI and BamHI. 2.5. Northern blot hybridization Total RNA from S. citri cells was extracted by the guanidinium thiocyanate–cesium chloride method (Chirgwin et al., 1979) or using the RNeasy kit from Qiagen ( Valencia, CA, USA). RNAs were separated by electrophoresis on 1% agarose formaldehyde 16.6% (v/v), MOPS 1X gels. After blotting onto C-extra membranes (Amersham) as recommended by the manufacturer, hybridizations were performed as previously described (Sambrook et al., 1989). Expression of fruR, fruA and fruK genes in S. citri mutant GMT 553 and GII-3 wt strain was monitored using specific probes from fruR, fruA and fruK. The specificity of all probes used was verified by Southern blots. FruR and fruA specific probes were obtained from insert H13.2: the 240 bp HincII (PfruR, Fig. 1) and 0.8 kbp DraI (PfruA, Fig. 1) fragments were used for transcription analysis of the fruR and fruA genes, respectively. The fruK specific probe was the 0.4 kbp HincII (PfruK, Fig. 1) fragment obtained from insert GT6.

2.7. Biochemical analyses Protein concentration was determined by the procedure of Bradford (1976), using the Biorad protein assay. The activity of PTS was measured as described by Navas-Castillo et al. (1993). The 6-PFK and 1-PFK activities were determined as described by Baumann and Baumann (1975) and Pollack (1995). Identification of fructose phosphates was as follows: 100 ml of exponentially growing cells of wt GII-3 or mutant GMT 553 were harvested by centrifugation at 12 000×g for 45 min and washed with HEPES 8 mM (pH 7.4), sorbitol 10% (w/v). The cells were resuspended in 1 ml of the same buffer and incubated with 0.1 mM -[ U-14C ]fructose (specific activity: 235 mCi/mmol ) for 15 min at 32°C. Preparation of sugar extracts was as described by Ferenci and Kornberg (1973). Briefly, the cells were harvested by centrifugation for 5 min at 20 000×g and were suspended in 1 ml of 7% (v/v) HClO . The undissolved 4 material was removed by centrifugation. The supernatant was neutralized with 0.3 M of 4 M-KOH, the precipitate was removed by centrifugation and the supernatant solution was used for chromatography on Whatman 3MM paper, as described by Wawszkiewicz (1961).

3. Results 3.1. Insertion site of Tn4001 in S. citri mutant GMT 553

2.6. Carbohydrate fermentation Fermentative catabolism of carbohydrates in S. citri results in lactic acid and acetic acid production (Miles, 1992). Acidification of the growth medium was used as an indicator of carbohydrate utilization. S. citri growth requires complex media supplemented with calf fetal or horse serum, the latter bringing carbohydrates and enzymes such as invertase which may interfere with fermentation of sugar added to the medium. For these reasons we used HSI medium for monitoring carbohydrate fermentation. In HSI, a sugar-free serum fraction replaces the normal serum, thus, without sugar addition no fermentation is observed. S. citri cells grown in SP-4 medium were harvested by centrifugation (20 min, 12 000×g, 20°C ), washed twice, and resuspended in HEPES 8 mM (pH 7.4), sorbitol 10% (w/v). HSI medium supplemented with sugar (0.5%, w/v) was inoculated with the washed cells. Decrease of pH was measured daily during 1 week on an aliquot of the culture medium. Concentrations of glucose or fructose in HSI medium inoculated with 5.108 colony forming unit/ml of wt strain GII-3 or mutant GMT 553, after 5 days at 32°C, were determined by the Test Combination sucrose/-glucose/-fructose (Boehringer, Mannheim, Germany).

Tn4001 is bordered at both ends by insertion sequence IS256. Southern blot hybridization with the IS256specific probe showed that Tn4001 was present as a single copy on the chromosomal DNA of mutant GMT 553 (data not shown). DNA fragments encompassing the insertion site of the transposon were cloned from mutant GMT 553. Sequence analyses revealed the presence of two direct repeat sequences of eight nucleotides ( TGGTATTA, Fig. 2) bordering Tn4001 as a result of integration. 3.2. Putative proteins encoded by the insertion region The region corresponding to the insertion site of Tn4001 was cloned from the wt strain GII-3, as indicated in Materials and methods. Four ORFs in the same orientation were identified. The first ORF ( Fig. 1) was truncated at its 5∞ extremity by the HindIII cloning site. The potential protein product showed 57% identity with the carboxy terminal amino acid sequence of translation factor IF2 from Bacillus stearothermophilus, encoded by infB (Brombach et al., 1986). The insertion site of Tn4001 is located 52 bp downstream of the ATG start codon of the second ORF ( Figs. 1 and 2). The 702 bp sequence encodes a putative

P. Gaurivaud et al. / Gene 252 (2000) 61–69

65

Fig. 2. Genetic map of S. citri wt GII-3 fructose operon. Nucleotide sequences of the intergenic regions are shown, as well as insertion of transposon Tn4001 (Tn) at the beginning of fruR. Direct and inverted repeat sequences are indicated by arrows. The putative ribosome binding site preceding fruR, fruA and fruK is underlined. The putative −35 and −10 box of the fructose operon promoter is indicated.

protein of 233 amino acids (aa) (26 kDa) which shows significant similarity with several transcriptional regulator proteins of bacterial carbohydrate catabolic operons belonging to the DeoR family ( Van Rooijen and de Vos, 1990). For instance, the identities with a hypothetical repressor of Mycoplasma capricolum (Accession No.: S48604), and FruR, a repressor in the fructose operon of Bacillus subtilis ( Kunst et al., 1997) were 43% and 39%, respectively. For these reasons, the second ORF was named fruR. Sequence analyses revealed that protein FruR of S. citri had the same structural characteristics as the other DeoR proteins: a helix-turn-helix motif involved in DNA binding ( Van Rooijen and de Vos, 1990), four conserved aa involved in binding the phosphorylated carbohydrate operon inducer ( Van Rooijen et al., 1993), and a highly conserved region supposed to be also involved in inducer binding ( Van Rooijen and de Vos, 1990). The sequence signature of the DeoR family (Reizer et al., 1996b) was also identified at the N-terminal end of FruR. The 2064 bp sequence of the third ORF encodes a protein of 687 aa (73 kDa) with a significant identity (29%) to the permease of the PEP:fructose phospho-

transferase system (fructose PTS ) from Escherichia coli (Prior and Kornberg, 1988). The identities with the putative fructose permease from Mycoplasma pneumoniae (Himmelreich et al., 1996) and Mycoplasma genitalium (Fraser et al., 1995) are 34% and 33%, respectively. The third ORF is thus likely to be the fruA gene of S. citri and to encode the fructose permease of the fructose phosphotransferase system. Sequence analysis showed that the putative fructose permease of S. citri is made of three domains linked to each other in the order IIA, IIB and IIC, similar to the situation in the fructose permeases of M. genitalium ( Fraser et al., 1995; Reizer et al., 1996a), and M. pneumoniae ( Himmelreich et al., 1996), but also in the HrsA protein of E. coli ( Utsumi et al., 1996). The fourth ORF (933 bp) was named fruK because it encodes a putative protein of 310 aa (34 kDa) with identities of 37%, 28%, 27% and 27% to the 1-PFK ( EC 2.7.1.56) of B. subtilis ( Kunst et al., 1997), Rhodobacter capsulatus ( Wu et al., 1991), M. genitalium (Fraser et al., 1995), and M. pneumoniae (Himmelreich et al., 1996), respectively. All these proteins belong to the phosphofructokinase PfkB family ( Wu et al., 1991).

66

P. Gaurivaud et al. / Gene 252 (2000) 61–69

The sequences downstream of fruK were identified as ORF14, ORF12 and ORF7 of virus SpV1 DNA sequences inserted in the S. citri genome (Renaudin et al., 1990). As shown in Fig. 2, all intergenic regions revealed repeat sequences and particularly terminatorlike structures at the end of infB, between fruA and fruK, and in the SpV1 sequence. 3.3. Transcription of the fructose operon Northern blot experiments, with probes specific for fruR or fruA, revealed the occurrence of two transcripts in the wt strain GII-3, one of 2.8 kb and one of 3.8 kb. The fruK specific probe revealed only a 3.8 kb messenger RNA for the wt strain GII-3. The 2.8 RNA ( fruR+fruA) and 3.8 kb RNA ( fruR+fruA+fruK ) were not detected in mutant GMT 553 (data not shown, Table 1). These results indicate that in strain GII-3, the three genes are organized as a single transcription unit: the putative fructose operon of S. citri. 3.4. Phenotype of mutant GMT 553 for sugar utilization S. citri is able to catabolize three sugars: trehalose, glucose and fructose. In order to demonstrate that the fructose operon is involved only in fructose utilization, trehalose, glucose and fructose utilizations were moni-

tored by acidification of the culture medium. Medium with sorbitol, a non-metabolizable carbohydrate, was used as negative control. As expected for the wt strain GII-3, acidification was observed with trehalose, glucose and fructose but not with sorbitol, whereas for mutant GMT 553, fermentation was observed only with glucose and trehalose (data not shown; Table 1). In addition, glucose and fructose titers were determined at the beginning and at the end of a 5 day culture of the wt strain and mutant GMT 553. From these data it was calculated that the wt strain used 22 mmol of glucose and 28 mmol of fructose, while mutant GMT 553 used 19 mmol of glucose but no fructose, as no variation in the fructose titer was observed. 3.5. Enzymes for sugar uptake by mutant GMT 553 Since transcription of fruA and fruK is abolished in mutant GMT 553, no fructose PTS activity and no 1-PFK activity should be detected in the mutant. Glucose and fructose PTS activities were determined in S. citri wt strain GII-3 and found to be, respectively, 1.2 and 4.9 nmol of sugar-phosphate formed/30 min/mg ( Table 1). As expected, mutant GMT 553 showed a phophoenolpyruvate-dependent phosphorylation of glucose (1.1 nmol of sugar-phosphate formed/30 min/mg) with the same specific activity as the wt strain GII-3,

Table 1 Properties of strains selected for functional complementation experiments

a + and −: presence and absence of gene transcription, respectively. b nmol of fructose-1,6-bisphosphate formed/min/mg protein. c nmol of sugar phosphate formed/30 min/mg protein. d +,−: sugar (Glu=glucose, Fru=fructose) is fermented (+) or not fermented (−). For each strain able to use both glucose and fructose, maximum rate of acidification was calculated for fructose (Vfru) and glucose (Vglu) and the ratio Vfru/Vglu (Flu/Glu) was deduced. e e: not detectable. f nd: not done. g 553pBOT and GII3pBOT: respectively mutant GMT 553 and wt strain GII-3 transformed with pBOT.

P. Gaurivaud et al. / Gene 252 (2000) 61–69

but fructose PTS activity was undetectable ( Table 1). The specific activity of 1-PFK was found to be 155 nmol of fructose-1,6-bisphosphate formed/min/mg of protein in the wt strain, whereas no activity could be detected in mutant GMT 553 ( Table 1). However, activity of 6-PFK was found to be present in both wt strain GII-3 and mutant GMT 553 and was measured to be 255 and 200 nmol of fructose-1,6-bisphosphate formed/min/mg, respectively ( Table 1). For both PFK activities, PPi could not replace ATP as phosphate donor (data not shown). Paper chromatography demonstrated that only fructose-1-phosphate, not fructose-6-phosphate, could be detected in the wt strain GII-3 incubated with [ U-14C ]fructose. In mutant GMT 553, no fructose phosphate was detected. 3.6. Functional complementation of mutant GMT 553 Complementation experiments, i.e. introduction of various combinations ( Table 1) of the wt genes of the fructose operon into mutant GMT 553 were performed to confirm that fruR and/or fruA and fruK were involved in fructose utilization and to clearly associate enzyme activities with the genes of the fructose operon. During the insect-transmission period and, thereafter, during the time required for symptom development, tetracycline was obviously absent from the insects or plants, and the spiroplasmas were thus not maintained under tetracycline selection pressure. For this reason, it was important to use complemented mutants in which the complementing plasmid had integrated into the chromosome. Integration sites of complementing plasmids are shown in Table 1. Transcription of the fructose operon genes, PTS fructose activity, 1-PFK and 6-PFK activities as well as glucose and fructose fermentation were determined for each complemented strain ( Table 1). In each case, it was verified that the gene carried by the plasmid was expressed in the relevant mutant, and the transcript lengths were in agreement with the plasmid construction. However, with plasmids pBOT-A and pBOT-RA transcripts were longer than those predicted from the fruA and fruR+fruA sequences, respectively. This may be explained by the lack of a transcription terminator at the end of the fruA gene, resulting in transcription of part of the pBOT plasmid. This was indeed confirmed by hybridization of total RNA from strain Fru21 ( Table 1) with a pBOT derived probe (data not shown). The infB+fruR intergenic region was amplified by PCR into a 180 bp fragment, and used as a promoter for expression of fruA+fruK in clone Fru31 and Fru36, fruA in clone Fru21, and fruK in clone FruK7 (Fig. 1 and Table 1). The 180 bp fragment carried the promoter of the fructose operon as shown by transcription analysis of clones Fru36, Fru21 and FruK7. In clone Fru36, expression of fruA+fruK resulted in two transcripts of

67

2.3 kb and 3.3 kb, showing that the intercistronic terminator between fruA and fruK is not affected in the absence of fruR expression. Since transcription of fruA+fruK is correctly achieved in clone Fru36, the inverted repeat sequence in the SpV1 sequence following fruK is sufficient to stop transcription of the operon. Activities of 1-PFK, 6-PFK and fructose PTS were measured in crude extracts of the complemented strains ( Table 1). The 6-PFK activity was detected in all complemented strains and was the same than in wt strain GII-3 ( Table 1). Only mutant GMT 553 complemented with fruK (clone FruK7), fruA–fruK (clones Fru31 and Fru36) or with the entire fructose operon (clone 13.1.1) showed restoration of 1-PFK activity. This result is in agreement with transcription analysis and confirms that fruK codes for the 1-PFK of S. citri. The fructose PTS activity was restored when the entire fructose operon, fruA–fruK, or fruA alone, was introduced into mutant GMT 553. This result confirms that fruA codes for the fructose PTS permease of S. citri. As shown in Table 1, as expected, all complemented mutants used glucose. Regarding fructose fermentation, all mutant strains complemented with fruA–fruK (strain 13.1.1, Fru31 and Fru36) were able to use fructose. Surprisingly, strain Fru21 complemented with fruA alone did use fructose, even though less efficiently. Fructose fermentation did not occur with the complemented strains lacking fruA (Table 1). One exception is strain TXA4 which possesses fruA and, yet, was unable to use fructose. In this clone, fruA was transcribed (together with fruR), but fructose PTS activity was not detected. For each strain able to use fructose, the maximum rate of acidification was determined in HSI medium from pH measurements for glucose (Vglu) and fructose (Vfru). The Vfru/Vglu ratio indicates that the strains can be divided into two groups. Strain GII-3 and strain 13.1.1 ( fruR–fruA–fruK ) had a ratio slightly greater than 1, indicating that fructose fermentation was at least as efficient as glucose fermentation. Strains Fru31 and Fru36 ( fruA–fruK ), and Fru21 ( fruA), which all lack fruR, had a ratio of 0.5, corresponding to a twofold decrease in fructose fermentation when compared to glucose fermentation. This may be correlated with the absence of the FruR protein.

4. Discussion In previous work, it was found that mutant GMT 553 was unable to induce symptoms in periwinkle plants as long as no revertants occurred, and the rate of multiplication of the mutant in planta was lower than that of the wt strain GII-3 (Foissac et al., 1997a). Transposon Tn4001 was found to be inserted in ORF1 of a transcription unit identified as the fructose operon

68

P. Gaurivaud et al. / Gene 252 (2000) 61–69

for the following reasons: (i) The products of the three ORFs of the unit, i.e. ORF1, ORF2 and ORF3, have significant similarities to, respectively, DeoR regulatory proteins, the fructose permease FruA and the 1-PFK FruK; (ii) Transcription of the three wt ORFs leads to a 3.8 kb RNA. A 2.8 kb RNA also occurs and correspond to ORF1 and ORF2. An inverted repeat sequence has been identified between ORF2 and ORF3 and could be a conditional terminator. In mutant GMT 553, none of the three ORFs is transcribed. (iii) The wt spiroplasma is able to use glucose and trehalose as well as fructose. Mutant GMT 553 is also able to use glucose and trehalose but cannot catabolize fructose. (iv) Fructose PTS activity and 1-PFK activity are detected in the wt spiroplasma but not in mutant GMT 553. (v) Complementation of mutant GMT 553 with ORF1+ORF2+ORF3, or ORF2+ORF3 or ORF2 restores fructose utilization. Complementation with ORF2 restores fructose PTS activity; complementation with ORF3 restores 1-PFK activity. These results indicate that ORF2 and ORF3 represent, respectively, the fruA permease gene and the fruK 1-PFK gene of S. citri. ORF1 codes for a protein belonging to the DeoR family. Because of its similarity with fruR of B. subtilis, we have named it fruR. The role of S. citri or B. subtilis FruR proteins is not clearly established. Complementation experiments show that mutant GMT 553 strains complemented with fruA+fruK use fructose less efficiently than the wt strain or the strain complemented with fruR+fruA+fruK. Knowing that DeoR proteins are transcriptional regulators of carbohydrate catabolic operons, it is possible that FruR plays a similar role in the fructose operon. However, fruR is not directly involved in fructose catabolism, since complementation of mutant GMT 553 with fruA+fruK in the absence of fruR, restores fructose catabolism. Different organizations of fructose operons have been found among mollicutes. M. pneumoniae and M. genitalium possess putative fruA and fruK genes in the same reciprocal organization as in S. citri, but no genes homologous to fruR have been found. No inverted repeat sequence was observed in the fruA–fruK intergenic region of M. genitalium and M. pneumoniae. The first gene of the putative M. capricolum fructose operon is homologous to S. citri fruR but, in contrast to S. citri, fruK is upstream of fruA. This organization is the same as that of B. subtilis. The modular organization of the fructose operon genes in mollicutes suggest that the genes have been acquired independently and/or have been reshuffled after acquisition. This hypothesis is supported by the presence, in S. citri, of repeat sequences within the intergenic regions between fruR and fruA, and between fruA and fruK, and by the fact that the terminator region of the fructose operon is located within the SpV1 viral sequences located down-

stream of fruK. Mollicutes are phylogenetically related to Gram-positive bacteria such as B. subtilis. If we consider that the organization of the B. subtilis fructose operon reflects that of the ancestral operon, i.e. promoter–fruR–fruK–fruA, as suggested by Itoh et al. (1999), then in S. citri, the order promoter–fruR–fruA– fruK must be the result of translocation events. The inverted repeat sequence downstream of fruA in S. citri, might represent the vestige of the ancestral terminator of the fructose operon in Gram-positive bacteria. Transposon Tn4001 is inserted in the first gene of the fructose operon of mutant GMT 553. As a result, the fructose operon is not transcribed, no fructose PTS activity and no 1-PFK activity could be detected. The product of fructose uptake, fructose-1-phosphate, was not detected, and the mutant was unable to use fructose as carbon and energy source. These results seem to indicate that the only pathway for fructose utilization in S. citri is via the fructose operon. This is the first time that an operon for sugar utilization has been functionally characterized in the mollicutes. In addition, we have shown previously that late development of severe symptoms in plants infected with mutant GMT 553 was concomitant with the genotypic reversion of the mutant to the wt strain (Foissac et al., 1997a). These results strongly support the role of the fructose operon in the phytopathogenicity of S. citri.

Acknowledgements We are grateful to J. Renaudin and S. Duret for making available of plasmids pBOT and pBOTP1, and for helpful discussions. This work was financially supported in part by grants from the Conseil Re´gional d’Aquitaine (number 96 03 07 003) and by the AIP Microbiologie Fondamentale grant from INRA (number P00188). Support for P. Gaurivaud was provided by the Ministe`re de l’Enseignement Supe´rieur et de la Recherche.

References Altschul, S.F., Gish, W., Miller, W., Myers, E.W., Lipman, D.J., 1990. Basic local alignment search tool. J. Mol. Biol. 215, 403–410. Baumann, L., Baumann, P., 1975. Catabolism of -fructose and -ribose by Pseudomonas doudoroffii. II. Properties of 1-phosphofructokinase and 6-phosphofructokinase. Arch. Microbiol. 105, 241–248. Bradford, M.M., 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein–dye binding. Anal. Biochem 72, 248–254. Brombach, M., Gualerzi, C.O., Nakamura, Y., Pon, C.L., 1986. Molecular cloning and sequence of the Bacillus stearothermophilus translational initiation factor IF2 gene. Mol. Gen. Genet. 205, 97–102. Byrne, M.E., Rouch, D.A., Skurray, R.A., 1989. Nucleotide sequence

P. Gaurivaud et al. / Gene 252 (2000) 61–69 analysis of IS256 from the Staphylococcus aureus gentamicin–tobramycin–kanamycin-resistance transposon Tn4001. Gene 81, 361–367. Calavan, E.C., Bove´, J.M., 1989. Ecology of Spiroplasma citri. In: Whitcomb, R.F., Tully, J.G. ( Eds.), The Mycoplasmas vol. 5. Academic Press, New York, pp. 425–485. Chang, C.-J., 1998. Pathogenicity of Aster Yellows phytoplasma and Spiroplasma citri on periwinkle. Phytopathology 88, 1347–1350. Chirgwin, J.M., Przybyla, A.E., MacDonald, R.J., Rutter, W.J., 1979. Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry 18, 5294–5299. Corpet, F., Gouzy, J., Kahn, D., 1999. Recent improvements of the ProDom database of protein domain families. Nucleic Acids Res. 27, 263–267. Daniels, M.J, 1983. Mechanisms of spiroplasma pathogenicity. Annu. Rev. Phytopathol. 21, 29–43. Dower, W.J., Miller, J.F., Ragsdale, C.W., 1988. High efficiency transformation of E. coli by high voltage electroporation. Nucleic Acids Res. 16, 6127–6145. Ferenci, T., Kornberg, H.L., 1973. The utilization of fructose by Escherichia coli. Properties of a mutant defective in fructose 1-phosphate kinase activity. Biochem. J. 132, 341–347. Foissac, X., Danet, J.L., Saillard, C., Gaurivaud, P., Laigret, F., Bove´, J.M, 1997a. Mutagenesis by insertion of Tn4001 into the genome of Spiroplasma citri: characterization of mutants affected in plant pathogenicity and transmission to the plant by leafhopper vector Circulifer haematoceps. Mol. Plant-Microbe Interact. 10, 454–461. Foissac, X., Saillard, C., Bove´, J.M., 1997b. Random insertion of transposon Tn4001 in the genome of Spiroplasma citri strain GII-3. Plasmid 37, 80–86. Garnier, J., Gibrat, J.F., Robson, B., 1996. GOR method for predicting protein secondary structure from amino acid sequence. Methods Enzymol. 266, 540–553. Gaurivaud, P., Danet, J.L., Laigret, F., Garnier, M., Bove´, J.M., 2000. Utilization of fructose and pathogenicity of Spiroplasma citri. Mol. Plant-Microbe Interact. in press. Fraser, C.M., et al., 1995. The minimal gene complement of Mycoplasma genitalium. Science 270, 397–403. Himmelreich, R., Hilbert, H., Plagens, H., Pirkl, E., Li, B.C., Herrmann, R., 1996. Complete sequence analysis of the genome of the bacterium Mycoplasma pneumoniae. Nucleic Acids Res. 24, 4420–4449. Itoh, T., Takemoto, K., Mori, H., Gojobori, T., 1999. Evolutionary instability of operon structures disclosed by sequence comparisons of complete microbial genomes. Mol. Biol. Evol. 16, 332–346. Kunst, F., Ogasawara, N., Moszer, I., Albertini, A.M., Alloni, G., Azevedeo, V., et al., 1997. The complete genome sequence of the gram-positive bacterium Bacillus subtilis. Nature 20, 249–256. Miles, R.J., 1992. Catabolism in mollicutes. J. Gen. Microbiol. 138, 1773–1783. Navas-Castillo, J., Laigret, F., Hocquellet, A., Chang, C.J., Bove´, J.M., 1993. Evidence for a phosphoenolpyruvate dependent sugarphosphotransferase system in the mollicute Acholeplasma florum. Biochimie 75, 675–679. Pollack, J.D., 1995. Methods for testing metabolic activities in mollicutes. In: Razin, S., Tully, J.G. ( Eds.), In: Molecular and Diagnostic Procedures in Mycoplasmology vol. I. Academic Press, New York, pp. 277–286. Prior, T.I., Kornberg, H.L., 1988. Nucleotide sequence of fruA, the

69

gene specifying enzyme IIfru of the phosphoenolpyruvate-dependent sugar phosphotransferase system in Escherichia coli K12. J. Gen. Microbiol. 134, 2757–2768. Reizer, J., Paulsen, I.T., Reizer, A., Titgemeyer, F., Saier Jr., M.H., 1996a. Novel phosphotransferase system genes revealed by bacterial genome analysis: the complete complement of pts genes in Mycoplasma genitalium. Microb. Comp. Genom. 1, 151–164. Reizer, J., Ramseier, T.M., Reizer, A., Charbit, A., Saier Jr., M.H., 1996b. Novel phosphotransferase genes revealed by bacterial genome sequencing: a gene cluster encoding a putative N-acetylgalactosamine metabolic pathway in Escherichia coli. Microbiology 142, 231–250. Renaudin, J., Aullo, P., Vignault, J.C., Bove´, J.M., 1990. Complete nucleotide sequence of the genome of Spiroplasma citri virus SpV1-R8A2 B. Nucleic Acids Res. 18, 1293 Renaudin, J., Marais, A., Verdin, E., Duret, S., Foissac, X., Laigret, F., Bove´, J.M., 1995. Integrative and free Spiroplasma citri oriC plasmids: expression of the Spiroplasma phoeniceum spiralin in Spiroplasma citri. J. Bacteriol. 177, 2870–2877. Saglio, P., Lhospital, M., Lafleche, D., Dupont, G., Bove´, J.M., Tully, J.G., Freundt, E.A., 1973. Spiroplasma citri gen. and sp.: a mycoplasma like organism associated with ‘‘stubborn’’ disease of citrus. Int. J. Syst. Bacteriol. 23, 191–204. Sambrook, J., Fritsch, E.F., Maniatis, T.E., 1989. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Utsumi, R., Horie, T., Katoh, A., Kaino, Y., Tanabe, H., Noda, M., 1996. Isolation and characterization of the heat-responsive genes in Escherichia coli. Biosci. Biotechnol. Biochem. 60, 309–315. Van Rooijen, R.J., de Vos, W.M., 1990. Molecular cloning, transcriptional analysis, and nucleotide sequence of lacR, a gene encoding the repressor of the lactose phosphotransferase system of Lactococcus lactis. J. Biol. Chem. 265, 18499–18503. Van Rooijen, R.J., Dechering, K.J., Niek, C., Wilmink, J., de Vos, W.M., 1993. Lysines 72, 80 and 213 and aspartic acid 210 of the Lactococcus lactis LacR repressor are involved in the response to the inducer tagatose-6-phosphate leading to induction of lac operon expression. Protein. Eng. 6, 201–206. Vignault, J.C., et al., 1980. Mise en culture de spiroplasmes a` partir de mate´riel ve´ge´tal et d’insectes provenant des pays circum-mediterrane´ns et du proche-orient. C.R. Acad. Sci. Paris 290, 775–778. Wawszkiewicz, E.J., 1961. A two dimensional system of paper chromatography of some sugar phosphates. Anal. Chem. 33, 252–254. Weisburg, W.G., Tully, J.G., Rose, D.L., Petzel, J.P., Oyaizu, H., Yang, D., Mandelco, M., Schrest, J., Lawrence, T.G.van, Etten, J., Maniloff, J., Woese, C.R., 1989. A phylogenetic analysis of the mycoplasmas: basis for their classification. J. Bacteriol. 171, 6455–6467. Whitcomb, R.F., 1983. Culture media for spiroplasma. In: Razin, S., Tully, J.G. ( Eds.), Methods in Mycoplasmology vol. I. Academic Press, New York, pp. 147–158. Williamson, D.L., Renaudin, J., Bove´, J.M., 1991. Nucleotide sequence of the Spiroplasma citri fibril protein gene. J. Bacteriol. 173, 4353–4362. Wu, L.F., Reizer, A., Reizer, J., Cai, B., Tomich, J.M., Saier Jr., M.H., 1991. Nucleotide sequence of the Rhodobacter capsulatus fruK gene, which encodes fructose-1-phosphate kinase: evidence for a kinase superfamily including both phosphofructokinases of Escherichia coli. J. Bacteriol. 173, 3117–3127.