Biochimie (1993) 75,6X-679 0 Societe francaise de biochimie et biologie moleculaire / Elsevier, Paris

675

Evidence for a phosphoenolpyruvate dependent sugar- phosphotransferase system in the mollicute Acholeplasmaflorum

J Navas-Castillo, F Laigret *, A Hocquellet, CJ Chang, JM Bove

Laboratoire de biologic cellulaire et molkulaire, Institut National de la Recherche Agronomique et Universite’ de Bordeaux lI, 33883 Villenave d’Ornon Cedex, France

(Received 4 February 1993; accepted 8 February 1993)

Summary - In order to confirm the presence of a phosphoenolpyruvate (PEP)-dependent sugar-phosphotransferase system (PTS) in the mollicute Acholeplasmaflorum we studied the ability of cell free extracts of this organism to phosphorylate glucose and/or fruc- tose in the presence of PEP We also cloned and sequenced a DNA fragment coding for a putative polypeptide showing significant similarity with the enzyme II of the P-glucoside PTS of Escherichia coli. Taken together, these results show that AJiorum possesses a fructose-PTS, but not a glucose-PTS, and that the amino acid sequence deduced from the DNA fragment is related to P-glucoside and sucrose enzymes II of PTS from various bacteria.

Acholeplasmajlorum / mollicute / metabolism / phosphotransferase system / Spiroplasma

Introduction

Mollicutes are wall-less eubacteria with a small genome size and a low guanine-plus-cytosine (G + C) content in their DNA. According to several metabolic and morphological characteristics, they have been divided into six genera. Phylogenetic analyses have shown that they are closely related to the low G + C Gram-positive bacteria [ 1,2] and represent four phylo- genetic groups 133. One of these groups, the spiro- plasma group, contains not only all the Spiroplasma species (helical, motile mollicutes) but also organisms classified as Mycoplasma and Acholeplusma species, t,he unifying characteristic of all the organisms in this group being their association with arthropods, mainly insects. However, this phylogenetic grouping contra- dicts the current phenotypic classification in which the genus Acholeplasma is clearly distinct from the genera Mycoplasma and Spiroplasma, while, in the phylogenetic classification, Acholeplasma species are split between two phylogenetic groups: the spiro- plasma group and the anaeroplasma group. A number of metabolic and molecular properties support this subdivision into two groups. For example, Achole- plasma florum, which belongs to the spiroplasma group, has metabolic pathways closer to Spiroplasma species than to those of the typical acholeplasma,

*Correspondence and reprints

Acholeplasma laidlawii [4-6]. As a matter of fact, A jbrum has been placed in the genus Acholeplasma mainly because it does not require sterol for its growth, a property which defines the genus Achole- plasma, however, it requires the presence of 0.04% Tween 80 in its culture medium 171. Two closely related species, Acholeplasma en tomophilum and Acholeplasma seifertii share these properties [8, 91. Furthermore, for several organisms within the spiro- plasma group it has been shown that they have a genetic code in which UGA is not a stop codon but specifies tryptophan [ 10-l 21. In contrast, A kaidkawii uses the universal genetic code [ 131. It was therefore important to determine the codon usage of A florrrm. We have now found that, like all &he organisms within the spiroplasma group, but not the anaeroplasma group, it uses UGA for tryptophan. Also, in the course of this study, we came across a DNA fragment, the deduced amino acid sequence of which showed simi- larity with the sequence of the p-glucoside specific enzyme II of the phosphoenolpyruvate (PEP)-depen- dent sugar-phosphotransferase system (PTS) of Escherichiu coli. Because acholeplasmas do not pos- sess a PTS [ 141, this finding was an additional argu- ment to exclude A Jlorum and related species from the genus Acholeplasma. A new genus name, Meso- plasma, has been proposed to accomodate these spe- cies (Tully et al, in press).

Little is known about sugar metabolism in molli- cutes, especially because until recently they could only be grown in complex media. The isolation of a

676

Table 1. PTS activity of crude extracts”.

[W]su,par

OkMG

Species None” ATP PEP

Fructose

None ATP PEP

A jIorum 0.76 0.69 0.75 0.65 0.53 8.30

S citri 0.8 1 0.83 3.52 0.68 0.70 7.78

A laidlawii 1.06 1.23 0.86 1.04 0.89 0.91

“Expressed a s n mol of phosphate-sugar formed by mg of proteins per assay. bPhosphate donor in the assay.

DNA fragment coding for a PTS component of A &rum is insufficient to demonstrate the presence of a functional PTS and, therefore, we have tested the eventual phosphorylation of glucose and fructose by crude extracts from A j7orunz. Indeed, the phosphoryl- ation of a sugar, in the presence of PEP, but not ATP, is a good indication of the occurrence of a PTS for this sugar [ 151. We have also used the cloned DNA fragment of the PTS of A Jlorum to isolate a larger fragment and, thus, to further extend the previous nucleotide sequence. We have obtained a deduced amino acid sequence covering the hydrophobic domain IIC of the enzyme II and this has made it pos- sible to make comparisons between sequences from various enzymes II of different bacteria. This paper confirms and extends preliminary results /16].

Materials and methods

A jkww~, strain L l’r (ATCC 33453), Sir~j?kuuntcr citri, strain R8AY’ (ATCC 27556) and A kdlawii, strain PGST :ATCC 23206) were routinely grown in BSR medium [ 171 supple- mented with 10% foetal calf serum.

E= coli, strain XLI-blue (Stratagene, La Jolla, CA), was used to propagate the recombinant plasmids.

PTS activity measrrr~ment

Preparation of crude extracts and measurement of sugar phos- phorylation were carried out as previously described [ 181. Briefly, cells were harvested by low speed centrifugation and washed twice with 250 mM NaCl and 10 mM MgC&. The pellet was resuspended in 150 mM NaCl, 50 mM Tris-HCI (pH 7.4), 1 mM EDTA, 10 mM B-mercaptoethanol and 0.1 mM dithiotreitol. The cells were sonicated and centrifugated at 15 000 R for 20 min to remove cell debris,

Cell free extracts (about 0.4 mg of protein in 250 pi) were incubated at 37OC for 30 min in complete assay medium containing 50 mM sodium phosphate buffer (pH 7.4), 25 mM NaE 5 mM PEP or ATP, 5 mM MgQ, and about 2 mM @t4C]methyI-o-glucopyranoside (a-MG) or [ Klfructose (304 mCi/mmol and 287 mCi/mmol, Amersham, Burkingham- shire, UK) in a final volume of 500 p.1. The reaction was

stopped at 100°C for 4 min. The reaction mixture was applied on a econo-column (Bio-Rad, Richmond, CA) containing 2 ml of AG 1 -x8 ( 100 mesh, Cl- form) anion-exchange resin (Bio- Rad). Non-phosphorylated sugar was etuted with 10 ml de- ionized water and phosphorylated sugar was recovered with 10 ml 1 M LiCl. The radioactivity of the samples was recorded in a scintillation counter.

Molecular biology techniques

Restriction and modification enzymes (Gibco-BRL, Gaithers- burg, MD; Biolabs, Beverly, MA), random priming labelling kit (Gibco-BRL), and T7 sequencing kit (Pharmacia, Mil- waukee, WI), were used as recommended by the manufactu- rers.

Plasmid PBS+ (Stratagene) was used for cloning procedures. E coli cells were transformed by high voltage electroporation as previously described [ 191. The small DNA fragment of a PTS gene from A f’iorum, previously cloned by us [ 161, was used as a probe to screen a DNA library of A Jorum genomic DNA cleaved with DrwI. A recombinant plasmid containing a 0,8-kbp fragment was selected and this fragment was sequenced. The nucleotide sequence has been deposited in GenBank (Los Alamos, NM) under the accession number 219057.

Sequence analyses

Nucleotide and amino acids sequences used for comparisons were imported from GenBank using the retrieve e-mail server of the National Center for Biotechnology Information at the National Library of Medecine, NIH, Bethesda, MD, Sequence analyses were performed on a Compaq Deskpro 386s micro- computer using SEQAID II (obtained from EMBL, Heidelberg, Germany) and CLUSTAL V [20] software.

Results and discussion

Glucose and fructose phosphorylation

Because phosphorylation of sugar by cell free extracts is a good indication of the occurrence of a PTS, we tested the eventual presence of a glucose and/or a fructose PTS in A jiorum, by monitoring the phospho- rylation of [W]labelled a-MG and of fructose by cell free extracts from this organism in the presence or

677

absence of PEP or ATP. We also used, in our test, extracts from S citri and A Zaidlawii as positive and negative controls respectively [ 14, 211. As shown in table I, cell free extracts from A Jiorum and S citri were able to efficiently phosphorylate fructose in the presence of PEP but not in the presence of ATP. This is a strong argument for the occurrence of a fructose- PTS in both organisms. In contrast, only extracts from S citri were able to phosphorylate a-MG in presence of PEP Thus A Jlorum seems to possess no glucose- PTS. As expected, with extracts from A laidlawii no phosphorylation of the two sugars was obtained.

Amino acid sequence comparisons

The previously cloned DNA fragment of A jlorum contains an open reading frame (ORF), the deduced amino acid sequence of which shows similarity with the enzyme II of the P-glucoside-PTS of E coli (IIBsi)

Ec-bgl Af-ptsX

Ec-bgl Af-ptsX

Ec-bgl Af-ptsX

Ec-bgl Af-ptsX

Ec-bgl Af-ptsX

Ec-bgl Af-ptsX

229

encoded by the bglC gene [22]. Because the A jlor-um DNA fragment was small (204 bp), the amino acid sequence was rather limited to allow significant comparisons with other known enzyme II sequences. Therefore, we used this fragment as a probe to clone a larger one. In this way a 0%kb DraI fragment was obtained and sequenced. An amino acid sequence of 265 residues was deduced, designated Af-ptsX, and compared to that of enzyme IIu@ of E co/i (EC-bgl). The alignment shown in figure 1 indicates that Af- ptsX encompasses more than three-fourth of the C domain of PTS enzyme II (IIC). Af-ptsX contains six tryptophan residues, all encoded by UGA codons and four of them are conserved in the enzyme IIBsr sequence from E coli (fig 1) and Erwinia chrysan- themi (data not shown).

According to the newly proposed nomenclature [23], enzymes II are composed of three to four domains

LERRLNAWLPSAIKNFFTPLLCLMVITPVTFLLVGPLS LN~IK~IPISLELMFRPFLLVMVIVPTSFFIL--LP IETLAGTL *. . ...*.* . . * *.* .*** * .*.*. *. * *..

1 YLWLYQAVPAFAGAVMGG IFVMFGLRWGLVPLCINNFTVLGYDTMIP ~IGQAPLGIGVGFYIG VAVIFGIHMGLIIVGILDSIQRGGAGIFM

** ** * ..o . *.**. ****** **. . * . . .

LLMPAIMAQVGAALGVFLCERDAQKKWAGSAALTSLFGITEPAVYGVNL IMGIS~QVGALVGVILVTQNSKLKKDAIHMLPAGC~~~~~~I;~~IS~

* * mm . . ***** l **.* ~0~~ . . . .

PRKYPF~IACISGALGATI--- IGYAQTKVYSFGLPSIFTFMQTIPSTGI PKKRPLIAGCPAAFFAGAYCNAVGVTARAGTGFG---IFEFIGFFSSPTM ’

*.* *.. .**.. . . . . .* 0 .** ** *o ..* .

end of domain IIC r-

>

start oti domain IIA

DFTV-WASVIGGVIAIGCAFV---- GTV-MLHFITAKRQPAQGAPQEKT GGTAELSNISNGLLYISGMIALALGTVFSLLIYIE---RPNEKSSISKS . * * . . . . . *. * . *** ,* l * .* . . . *.

491 PEVITPPEQGGICSPMTGEIVPLI ANALLKFIK--VKDNLTDEEIQIL . . . . . .*.* . . .

Fig 1. Alignment of the amino acid sequence of enzyme IIBsl from E coli (Ec-bgl) and AfiptsX from A jkrrm. Numbering is according to the E coli sequence, Stars indicate identity, while points indicate similarity. Tryptophan residues (W) encoded by UGA are in white. Conserved histidine residue (H) is indicated by a vertical arrow.

678

T,ble II. Percent identities (and simiiarities) between the homologous segments of domain IIC of several PTS proteins”.

Percentage of identin, (and percentage of similarity)h Gth

PTS proteins” E&g! E&V-bgl Bs-sacX Bs-scr St-scr Bs-glc Ec-mtl EC-nag

Af-ptsX Ec-bgl Er+bgl Bs-sacX Bs-scr St-scr Bs-glc Ec-mtl

28 (59) 29 (58) 27 (59) 24 (53) 27 (56) 18 (40) 17 (47) 19 (43)

56 (82) 29 (61) 27 (60) 26 (61) 20 (44) 21 (50) 20 (48)

30 (61) 27 (59) 31 (61) 18 (43) 22 (47) 17 (44) 55 (77) 48 (78) 18 (44) 16 (48) 18 (42)

55 (75) 18 (40) 19 (48) 20 (44) 18 (45) 18 (47) 20 (42)

17 (47) 42 (71) 20 (44)

aThe amino sequences compared correspond to position 229 to 457 of bglC from E coli and are entirely located in domain IIC. bPercentage similarities represent the percentage of identical residues plus the percentage of conservative changes. CThe two first letters indicate the bacterial species; Af, A Jlorwm; EC, E coli; Bs, Bacillus subtilis; Ew, Elwinia chrysanthemi; St, Salmo- nella t~phimrrrium. The following letters indicate the specificity of the enzyme II: bgl, P-glucoside; scr, sucrose; glc, glucose; mtl, mannitol; nay. N-acetyl glucosamine.

(HA to IID) which may be harboured by one or several polypeptides, depending on the sugar and the bacterial species involved. Domain IIC is the hydro- phobic region presumed to bear the transmembrane channel and sugar binding site [23, 241. The equiva- lent regions from several enzymes II, from various bacteria and differing in their sugar specificity, have been compared and the results are shown in table II. The sequences showing more than 50% similarity with that of A fior~mz are those of enzymes IIngl of E coli and Erw>du chrysanthemi, protein sacX from Bacilfl~s suhtilis and enzymes IIscr from Saimntr~IIa

typhimltritdm and B sdmYis. SacX is a negative regula- tory protein of the levansucrase operon having homo- logy with enzymes IIM and enzymes IIscr [25, 261. A number of features are highly conserved such as the histidine residue, indicated by a vertical arrow in figure 1, which is known to undergo phosphorylation [27]. The hydrophobic nature of the protein is also conserved (data not shown). In contrast, our sequence has limited similarity with enzymes IIorc, IIMtl and IIM (table II) and almost none with fructose per- meases (data not shown).

We have demonstrated that A jforum is able to phosphorylate fructose but not glucose in the presence of PEP but not ATP. This result is in agreement with the fact that in a chemically defined medium, fructose, but not glucose, is able to support the growth of AJo- rum (Chang et al, manuscript in preparation). We did not test for the labelling of sucrose, because A Jorum does not grow with this sugar as sole source of carbon (Chang et al, manuscript in preparation). Hence the enzyme II of A florum (Af-ptsX) is not related to the glucose- or sucrose-PTS, but nothing more can be said concerning the sugar specificity of this enzyme II. We

have only shown that it shares significant similarity with domain C of enzymes IIn@ from E coli and E chrysantemi, with the negative regulatory protein SacX of the sucrose PTS of B subtilis and with various enzymes 11s cr. It is probably not the fructose permease because there is no significant similarity with any of these proteins. An interesting feature is that the reading frame of Af-ptsX remains open after the end of domain IIC (fig 1). In enzymes II@‘, domain IIC is immediately followed by IIA (fig 1) while, in enzymes IIscr, IIA is harboured by a distinct polypeptide 123, 241. In Af-ptsX, IIA may well follow IIC, as several conserved amino acids between &bgl and AfiptsX have already been revealed (fig 1). It should also be possible, with our chemically defined medium, to identify the sugar requirements of A Jo- rum and the potential PTS present in this organism.

Acknowledgmer(lts

We specially thank .i Reitzer for fruitful advice. Jesus Naves- Castillo is a recipient of a fellowship from the Ministerio de Education y Ciencia (Spain).

References

Woese CR, Maniloff J, Zabien LB (1980) Phylogenetic analysis of the mycoplasma. Proc Nat1 Acad Sci USA 77, 494-498 Woese CR (1987) Bacterial evolution. Microbial Rev 51, 221-271 Weisburg WG, Tully JG, Rose DL, Petzel JP, Oyaizu H, Yang D, Mandelco L, Sechrest J, Lawrence TG, Van Etten J, Maniloff J, Woese CR (1989) A phylogenetic analysis of the Mycoplasmas: basis for their classification. J Bacte- riol 17 1,6455-6467

679

4

5

6

7

8

9

10

11

12

13

14

15

16

Pollack JD, Beaman KD, Robertson JA (1984) Synthesis of lipids from acetate is not characteristic of Achole- plasma or Ureaplasma species. Int J Syst Bacterial 34, 124-126 Williams MV, Pollack JD ( 1985) Pyrimidine deoxyribonu- cleotide metabolism in members of the class mollicutes. In J Syst Bacterial 35227-230 Tryon VV, Pollack JD (1985) Distinction in mollicutes purine metabolism: pyrophosphate-dependent nucleoside kinase and dependence on guanylate salvage. Znt J Syst Bacterial 35,497-501 MC Coy RE, Basham HG, Tully JG, Rose DL, Carle P, BovC JM (1984) Acholeplasma jlorum: a new species iso- lated from plants. Int J Syst Bacterial 34, 1 l-l 5 Tully JG, Rose DL, Carle P, BovC JM, Hackett KJ, Whit- comb RF (1988) Acholeplasma entomophilum: sp nov from gut content of a wide range of host insects. Int J Syst Bacterial 38, 164-167 Bonnet F, Saillard C, Vignault JC, Gamier M, Carle P, Bovd JM, Rose DL, Tully JG, Whitcomb RF (1911) Acho- leplasma seisfertii sp nov, a mollicute from plant surfaces. Int J Syst Bacterial 41,45+9 Yamao F, Muto A, Kawauchi Y, Iwami M, Iwagami S, Azumi Y, Osawa S (1985) UGA is read as tryptophan in Mycoplasma capricolum. Proc Nat1 Acad Sci USA 82, 2306-2309 Renaudin J, Pascarel MC, Saillard C, Chevalier C, Bove JM (1986) Chez les spiroplasmes le codon UGA n’est pas non sens et semble coder pour le tryptophane. CR Acad Sci Paris 303,539-540 Guindy YS, Samuelsson T, Johansen TI (1989) Uncon- ventional codon reading by Mycoplasma mycoides tRNAs as revealed by partial sequence analysis. Biochem J 258, 869-873 Tanaka R, Muto A, Osawa S (1989) Nucleotide sequence of tryptophan tRNA gene in Acholeplasma laidlawii. Nucleic Acids Res 17.5842 Tarshis MA, Migoushina VL, Panchenko LF (1973) On the phosphorylation of sugars in Acholeplusma luidlawii. FEBS Lett 31,111-113 Roman0 AH, Eberhard SJ, Dingle SL, McDowell TD ( 1970) Distribution of the ,phosphoenolpyruvate:,glucose @&;transferase system m bacterra. J Bactertol 104,

Nav&Castillo J, Laigret F, Tully JG, Bov6 JM (1992) Le mollicute Acholeplasma florum possbde un gene du sys- t&me phospho6nolpyruvate sucre-phosphotransferase et il

17

18

19

20

21

22

23

24

25

26

27

utilise UGA comme codon tryptophane. CR Acad Sci Paris 3 15,43-48 Whitcomb RF (1983) Culture media for spiroplasmas. In: Methods in mycoplasmology (Razin S, Tully JG, eds) Aca- demic Press, New York, 147-I 59 Mugharbil U, Cirillo VP (1978) Mycoplasma phospho- enolpyruvate-dependent sugar phosphotransferase system: glucose-negative mutant and regulation of intracellu!ar cyclic AMP J Bacterial 133,203-209 Dower WJ, Miller JF, Ragsdale CW (1988) High effi- ciency transformation of E coli by high voltage electropo- ration. Nucleic Acids Res 16,6127-6145 Higgins DG, Sharp PM (1989) Fast and sensitive multiple sequence alignments on a microcomputer. Cabios 5, 15 l- 153 Tarshis MA (1991) Distribution of a phosphoenolpyru- vate-dependent sugar phosphotransferase system in myco- plasmas. Can J Microbial 37,477-479 Schnetz K, Toloczyki C, Rak B (1987) P-Glucoside (bgl) operon of Escherichia coli K-12: nucleotide sequence, genetic organisation, and possible evolutionary relation- ship to regulatory components of two Bacillus subtilis genes. J Bacterial 169,2579-2590 Saier MH, Reizer J (1992) Proposed uniform nomencla- ture for the proteins and protein domains of the bacterial phosphoenolpyruvate:sugar phosphotransferase system. J Bacterial 174, 1433- 1438 Ebner R, Lengeler JW (1988) DNA sequence of the gene scrA encoding the sucrose transport protein enzyme IIsrc of the phosphotransferase system from enteric bacteria: homology of the enzyme 11s~ and enzyme IIt@ proteins. Mol Microbial 2,9-l 7 Fouet A, Amaud M, Klier A, Rapoport G (1987) Bacillus subtilis sucrose-specific enzyme II of the phosphotransfe- rase system: expression in Escherichia coli and homology to enzyme II from enteric bacteria. Proc Nat1 Acad Sci USA 84,8773-8777 Amaud M, Vary P, Zagorec M, Klier A, Debarbouille M, Postma P, Rapoport G (1992) Regulation of the sacPA operon of Bacillus subtilis: identification of phosphotrans- ferase system components involved in SacT activity. J Bacterial 174,3161-3170 Bramley HF, Komberg HL (1987) Sequence homology between proteins of bacterial phosphoenolpyruvate-depen- dent sugar phosphotransferase systems: identification of possible phosphate-carrying histidine residues. Proc Nat1 Acad Sci USA 84,4777-4780