By continuing to browse this site you agree to us using cookies as described in About Cookies
Wiley Online Library is migrating to a new platform powered by Atypon, the leading provider of scholarly publishing platforms. The new Wiley Online Library will be migrated over the weekend of March 17 & 18. You should not experience any issues or loss of access during this time. For more information, please visit our migration page: http://www.wileyactual.com/WOLMigration/
Haloarchaeal flagella are composed of a number of distinct flagellin proteins, specified by genes in two separate operons (A and B). The roles of these flagellins were assessed by studying mutants of H. salinarum with insertions in either the A or the B operon. Cells of the flgA− mutant produced abnormally short, curved flagella that were distributed all over the cell surface. The flgA2− strain produced straight flagella, mainly found at the poles. The flgB− mutant had flagella of the same size and spiral shape as wild-type cells, but these cells also showed unusual outgrowths, which appeared to be sacs filled with basal body-like structures. In broth cultures of this mutant, the medium accumulated flagella with basal body-like structures at their ends.
The Archaea is a domain of microorganisms distinct from Eukarya and Bacteria. Archaea often live in extreme environments (e.g. high temperatures, extreme pH and salt concentration, extreme anaerobiosis), and many are motile, possessing flagella that, at least superficially, resemble those of Bacteria. However, biochemical and genetic studies of the flagella and their component flagellins have demonstrated that archaeal flagella have features distinct from bacterial flagella. The main differences are as follows: (i) the archaeal flagella are generally thinner (10–13 nm) than their bacterial counterparts (Jarrell and Koval, 1989; Pyatibratov et al., 1993); (ii) they are composed of a mixture of different flagellins (Alam and Oesterhelt, 1984; Kalmokoff et al., 1992; Pyatibratov et al., 1993) in contrast to a single flagellin in the flagella of bacteria, such as Escherichia coli and Salmonella typhimurium; (iii) frequently, archaeal flagellins are glycosylated (Wieland et al., 1985; Southam et al., 1990; Bayley et al., 1993), and their glycosylation occurs outside the cytoplasmic membrane (Sumper, 1987); (iv) flagellins from all archaea have highly conserved hydrophobic N-terminal portions and leader peptides (Kalmokoff et al., 1990; Jarrell et al., 1996a); (v) archaeal flagellins have homology with bacterial type IV pilins but not with bacterial flagellins (Faguy et al., 1994).
The requirement for the presence of multiple flagellin genes in archaea has remained an unsolved question. Are all gene products necessary for the formation of a spiral filament, or do some flagellins in a minor quantity have a specific role in flagellar filament assembly (for example, as initiators or terminators, or as anchor or hook-associated proteins of bacterial flagella)? Flagellin genes have been cloned from three archaea, Methanococcus voltae (Kalmokoff and Jarrell, 1991), Methanococcus vannielii (Bayley et al., 1998) and Halobacterium salinarum (formerly called Halobacterium halobium) (Gerl and Sumper, 1988), and identified from the completely sequenced genomes of M. jannaschii (Bult et al., 1996), Archaeoglobus fulgidus (Klenk et al., 1997), Pyrococcus horikoshii (Kawarabayasi et al., 1998) and Aeropyrum pernix (Kawarabayasi et al., 1999). According to SDS–PAGE, purified flagella of M. voltae are composed of two flagellin bands corresponding to molecular masses of 33 kDa and 31 kDa (Kalmokoff et al., 1988). The flagellins are encoded by a multigene family of four related genes (flaA1, flaB1, flaB2 and flaB3) organized into two transcriptional units. Northern blot and primer extension analysis indicated that all four genes are transcribed (Kalmokoff and Jarrell, 1991). Mutant studies of M. voltae indicated that the four flagellins are not simply interchangeable entities and that A flagellin may play a special role in the assembly or function of flagella (Jarrell et al., 1996b). Three closely spaced and highly similar flagellin genes (flaB1, flaB2 and flaB3) have been found in M. vannielii (Bayley et al., 1998). Three, two, five and two flagellin genes have been identified for M. jannaschii (Bult et al., 1996), A. fulgidus (Klenk et al., 1997), P. horikoshii (Kawarabayasi et al., 1998) and A. pernix (Kawarabayasi et al., 1999).
Flagellins of H. salinarum have been shown to be encoded by five different but homologous genes. Two flagellin genes are arranged tandemly at one locus (flgA1 and flgA2), and the other three are in a tandem arrangement at different loci (flgB1, flgB2 and flgB3) (Gerl and Sumper, 1988). The estimated lengths of both mRNAs indicated that they contained only flagellin genes, and the flagellin genes are not co-transcribed with other possible downstream open reading frames (ORFs), in contrast to M. voltae (Bayley and Jarrell, 1998). According to genomic mapping of Halobacterium sp. GRB, the two flagellin gene loci are spaced at least 42 kb apart (St Jean et al., 1994). All the five flagellin gene products are expressed and integrated into a flagellar bundle (Gerl et al., 1989). These proteins were characterized as sulphated glycoproteins with N-glycosidically linked oligosaccharides, and each polypeptide possesses three N-glycosylation sites (Wieland et al., 1985).
In the present paper, we describe insertional mutant strains of H. salinarum with inactivated flagellin A and B operons and A2 flagellin gene.
Construction of flagellin A− and B− strains of H. salinarum
Two insertion vectors, pAN and pBN, were constructed as described in Experimental procedures (Fig. 1A and B). The insertion plasmids were incapable of self-replication and could only be maintained in cells by incorporation (via a single cross-over event) into the chromosome, resulting in insertional inactivation of the flgA or flgB genes of cells. Southern blots of the DNA from these gene knock-out strains were hybridized with flagellin gene probes to determine the presence of the insertion plasmids in the A or B loci of the chromosome. As shown in 1Fig. 1C (lane 2), BamHI digestion of DNA from wild-type cells produced two hybridizing bands, i.e. a 15 kb band containing the A operon and a 5.5 kb band containing the B operon. The sizes of these bands are similar to those reported previously by Gerl and Sumper (1988). In the DNA of mutant strains, three bands were seen (lanes 1 and 3), consistent with the insertion vector adding a BamHI site to the middle of an operon, so producing an extra BamHI fragment. The flgA recombinant showed bands of 1.1 kb and 19.9 kb (as well as the 5.5 kb B operon-containing band). The flgB recombinant had bands of 4.6 kb and 6.9 kb (as well as the 15 kb A operon-containing band). While this figure shows only one transformant for each plasmid, a total of five separate transformants for each plasmid (pAN or pBN) were examined, and all produced hybridizing bands of identical size and number to those seen in 1Fig. 1C.
Transcriptional activity of the A and B operons was assessed by Northern blot hybridization of total cellular RNA extracted from the wild-type and both mutant strains (Fig. 1D). Wild-type cells displayed two mRNAs, one from the A operon (≈1300 bp) and the other from the B operon (≈1900 bp) (Gerl and Sumper, 1988). Insertions in the A or B operons might be expected to cause the production of extended or truncated transcripts but, in the case of the flgA mutant, no flgA gene-specific transcripts could be detected at all (Fig. 1D, lane 3). In addition, flgB mRNA in this mutant was reduced dramatically to around background levels (not seen in Fig. 1D). In the case of the flgB mutant, flgB mRNA was undetectable, whereas the flgA transcript was relatively unaffected. No additional transcripts were observed. As the promoters for both operons are unchanged after plasmid insertion, the most likely explanation for the lack of detectable transcripts from inactivated operons is that the transcripts were rapidly degraded, perhaps because of a loss of transcription termination signals. This would not explain the reduction in mRNA from the B operon in the flgA mutant, and it is proposed that A flagellins positively regulate the B operon.
Construction of a flgA2− strain of H. salinarum
The insertion vector pAN1 (see Fig. 2A) was used specifically to inactivate the flgA2 gene within the A operon. As shown in the figure, the plasmid contained the 3′ part of the flgA1 gene [nucleotide 524 to the stop codon of the A1 ORF according to Gerl and Sumper (1988)] fused to the terminator region of the A operon. The method of construction inadvertently substituted the A1 sequence for the A2 sequence at nucleotide 985, producing a Thr to Lys change at the penultimate residue of the flgA1 protein. This change was not expected to be significant.
Transformants carrying insertions of the pAN1 plasmid were analysed by Southern blot hybridization (Fig. 2B), which confirmed that the plasmid had inserted into the flgA1 gene in the expected manner. Three separate transformants gave identical patterns after BamHI digestion (i.e. flgA-containing bands of 1.1 kb and 19.9 kb; see Fig. 2B). Northern blots hybridized with a flgA DNA probe showed that mutant strains did not produce the flgB operon transcript (about 1.9 kb in wild-type cells) and produced a smaller flgA specific transcript of about 650 nucleotides (Fig. 2C). These results were consistent with the addition of a transcriptional terminator to the flgA1 gene, with transcription of the B operon being dependent upon the A operon.
Characterization of flagellin A−, A2− and B− strains of H. salinarum
Electron microscopy indicated that the cells of mutant flgA−, flgA2− and flgB− strains had flagella distinct from those of the wild type. All flagellated strains (wild type and mutants) showed a proportion of cells without flagella, but flagellated cells within a pure culture were found to have flagella that were characteristic in size, shape and cellular location. Approximately 200 flagellated cells were examined for each strain. Flagella of flgA− strain, compared with the wild-type ones, were curved, significantly shorter and located at both the poles and the sides of cells (Fig. 3A). The cells of the flgA2− strain possessed straight flagella (Fig. 3B), which were mainly located at the poles of the cells. The flagella of the flgB− strain were located only at the cell poles, had a spiral shape and were as long as the wild-type flagella (Fig. 3C). Apart from the flagella, some of the flgB− cells had outgrowths, sometimes with a flagellum at their ends, while some had only outgrowths. The yield of flagella, isolated using the method described in Experimental procedures, per gram of flgB− strain biomass was similar to that of the wild-type strain. When separated on CsCl gradients, flagella of wild-type cells showed a clear density difference from cell membranes, but the flagella from this mutant was found to band very close to the red-coloured cell membrane band. Apart from flagellins, other protein bands were detected on SDS–PAGE of this sample (Fig. 4). Electron microscopy of this sample showed filaments resembling normal flagella as well as flagella with a basal body-like structure at the filament ends (Fig. 5A). In addition to flagella with basal body-like structures at the ends, outgrowths attached to the ends of flagella were also revealed (Fig. 5B), some extending up to 0.8 μm. Taking into consideration the presence of the membrane component in the flagella sample, this suggests that the outgrowth is a membrane sac filled with basal body-like structures.
Motility of the wild-type and mutant strain cells was determined by plating on semi-solid medium. The flgA−, flgA2− and the flgB− cells migrated threefold less than the wild-type cells, indicating that their motility was significantly impaired (Fig. 6). Reduction of motility for the flgA−, flgA2− and flgB− cells was also observed by light microscopy.
The content of flagellins in the flagella and cell lysates of the flgA−, flgA2− and flgB− strains was tested by Western blot analysis with polyclonal antiserum raised to purified flagella of wild-type H. salinarum.
In the case of the flgA− strain, an immunocross-reacting band for flagella sample was observed mainly for B2 flagellin (Fig. 7B, lane 1, top band). For the cell lysate, no immunocross-reacting band was observed (Fig. 7A, lane 1). The absence of B flagellins in the cell lysate was probably a result of the low level of flgB mRNA observed in this mutant compared with the wild-type strain (Fig. 1D). The absence of immunocross-reacting bands of A1 (the middle band) and A2 (the bottom band) flagellins in both flagella and cell lysate samples corresponded with the absence of flgA mRNA in this mutant (Fig. 1D). These results confirmed that the plasmid insertion into the flgA2 gene had inactivated the synthesis of both A1 (the middle band) and A2 (the bottom band) flagellins.
In the case of the flgB− strain, the immunocross- reacting band corresponding to B2 flagellin was not observed in immunoblots of either the flagella (Fig. 7B, lane 2) or the cell lysate samples (Fig. 7A, lane 2). B1 and B3 flagellins migrate in SDS–PAGE as a single large band with A2; thus, Western blotting cannot reveal the presence or the lack of B1 and B3 flagellins. The band corresponding to A1 flagellin was clearly observed in both samples.
In the case of the flgA2− strain, one narrow immunocross-reacting band running immediately above the A2, B1 and B3 flagellin bands was observed in immunoblots of both the flagella (Fig. 7B, lane 4) and the cell lysate samples (Fig. 7A, lane 4). The shifting of the immunocross-reacting band of A1 flagellin towards the lower value may be caused by the inhibition of flagellin glycosylation. The inhibition of glycosylation in vivo was shown previously to result in shifting the 30 kDa A1 flagellin band to 23 kDa (Sumper, 1987). As the shape of flagella was changed from spiral to straight, it can lead to steric inaccessibility of glycosylation sites on the surface of straight flagella in contrast to spiral flagella.
The assembly process for archaeal flagella has not been completely determined and, at present, there is little information showing how it may occur (Jarrell et al., 1996a,b). The assembly process would be clearer if the multiflagellin nature of archaeal flagella was explained. In the present study, the phenotypes of three flagella mutants (flgA−, flgB− and flgA2−) were likely to be the result of flagellin gene inactivation only, as indicated by the Northern blot data. In contrast to wild-type flagella, which consist of both A and B flagellins, mutants unable to make either A or B flagellins were unable to make functional flagella. Despite a high level of sequence similarity, these results indicate that flagellins encoded by the A and B operons are not interchangeable and have different specialized roles in flagella formation. Cells with an inactivated flg A operon were less motile than wild-type cells and produced short, curved filaments, composed of B flagellins, that were distributed all over the cell surface. When the B operon was inactivated, cells produced flagella (consisting only of A flagellins) that appeared to be similar to wild-type flagella and had normal polar distribution. However, cells were also significantly impaired in their motility. The conclusion that can be drawn is that A flagellins compose the basic filament part of the flagellum. This is supported by the previous studies showing that A flagellins represent the major component of flagellin species from wild-type flagella (e.g. V. Y. Tarasov, unpublished). Despite the formation of long, spiral filaments from A flagellins, flgB− cells do not possess functional flagella. This may be the result of a disruption occurring in the basal parts of the flagella, as the flgB− cells have outgrowths some of which were shown to be attached to the end of flagella. The outgrowths contain basal body-like structures connected to flagellum. A similar structure was found previously at the ends of flagella of Methanococcus voltae (Kalmokoff et al., 1988) (as knob), Methanococcus thermolithotrophicus and Methanospirillum hungatei (Cruden et al., 1989). If some of the B flagellins are included in the proximal part of the filament adjacent to the basal body (for example, as terminators, or as anchor or hook-associated proteins of bacterial flagella), the lack of B flagellins could cause this disruption. Consequently, A flagellins, as basic components of the flagellar filament, must be included in flagella filaments at the initial stage of assembly. This is also more likely because the normal polar location of flagella was revealed for the flgB− strain, i.e. when the flagellar filaments consisted of A flagellins. When flagellar filaments were located at both the poles and the sides of the cell, it was probably a result of the incorrect inclusion of B flagellins (instead of A flagellins) at the initial stage of flagella assembly.
Northern blot results indicated that inactivation of the whole A operon or only the A2 flagellin gene led to inhibition of transcription of B flagellin genes in the B operon, but not vice versa. This indicates that the synthesis of B flagellins is dependent upon the synthesis of A flagellins and cannot occur before the synthesis of A flagellins. This also supports a model in which A flagellins are produced first and are included in the initial stage of assembly. The regulation of transcription in the B operon probably depends on transcription in the A operon but, as the two operons are spaced over 40 kb apart in Halobacterium sp. GRB (St Jean et al., 1994), this regulation probably involves soluble factors.
Our results indicate that only two A flagellins of H. salinarum are sufficient to form a longitudinal and spiral filament. In this case, there are two possibilities for the formation of flagella helicity: (i) two flagellins (for H. salinarum, they are A1 and A2) of differing conformation are necessary for the assembly of spiral flagella; (2) a single flagellin can be in two conformations, thus providing flagellar helicity as in the case of bacteria (Calladine, 1978). If the first possibility is correct, the cells of mutant strains with one A flagellin should have straight flagella. The data obtained for the flagellin A2− strain of H. salinarum confirmed this. In this case, when cells contained only A1 flagellin, flagella were straight. The first hypothesis is confirmed also by the fact that two flagellins are arranged into a spiral flagellar filament both for the constructed flgB− strain of H. salinarum and for the wild type of M. vannielii (Bayley et al., 1998). The flagella isolated by cell shearing (i.e. mainly the distal part of flagella) were shown to consist of only two of the three flagellins determined in M. vannielii. Moreover, three B flagellins of the insertional mutant of M. voltae are sufficient for the formation of long spiral filaments and, in this case, it is known that one of these (flaB3) was synthesized in a small quantity (Jarrell et al., 1996b).
Thus, the multicomponent nature of flagella from H. salinarum may be explained first by the requirement of two A flagellins for the formation of a flagellar filament with a spiral shape and, secondly, by the possible participation of some B flagellins in the final stage of flagellar assembly.
To isolate total RNA, a 1 ml cell culture sample (A550 = 0.5–0.8) was removed to a 1.5 ml microcentrifuge tube and placed on ice. The pelleted cells (13 000 r.p.m., 1 min, 4°C) were resuspended in 80 μl of cold lysis buffer (25 mM NaOH, 0.5% SDS, 5 mM EDTA, 8% sucrose, 5 μM CDTA). After incubation at 37°C for 15 min, the tubes were placed on ice for 2 min, and then 30 μl of precooled sodium acetate (3 M, pH 5.6) was added, and the solution was vortexed for a few seconds. The lysate was centrifuged (13 000 r.p.m., 30 min, 4°C), and the supernatant was removed to a fresh tube. The RNA was precipitated from the supernatant by adding two volumes of ethanol, and the subsequent RNA pellet was washed twice with 70% ethanol, before being air dried (30 min at room temperature under slight vacuum) and finally dissolved in DEPC-treated water.
SDS–PAGE was performed according to the method of Laemmli (1970) using 12% acrylamide gel. The proteins were stained with Coomassie brilliant blue G 250.
Polymerase chain reaction (PCR) amplification of A and B operons and A1–A2 DNA fragment
DNA fragments containing H. salinarum A (PCR products of 1330 bp) and B (PCR products of 1986 bp) operons were PCR amplified from genomic DNA using the following oligonucleotides: 5′-TTCCCGGGTACCGCCGTTTT-3′, forward primer (Afp) corresponding to position 276–295 of A operon (Gerl and Sumper, 1988), and 5′-AGAAGGATCCGCGCGTCTTACAGCTTG-3′, reverse primer (Arp) (position 1579–1605 of A operon); 5′-AAGGTTCGGTACCGCGCAAAT-3′, forward primer (Bfp) (position 290–310 of B operon), and 5′-TGGCACGTCTAGACAGACGA-3′, reverse primer (Brp) (position 2257–2276 of B operon). The primers included restriction sites for cloning: Afp, KpnI; Arp, BamHI; Bfp, KpnI; Brp, XbaI. The flgA1 fragment was PCR amplified from a KpnI–SalI fragment of cloned A operon using the following oligonucleotides: forward primer (Hfp), 5′-ACACTGCAAGCTTCCTCCAGTC-3′ (position 514–535 of A operon); reverse primer (Hrp), 5′-GGACGTTCGCCCAGTAGGTGG-3′ (position 934–954 of A operon). The flgA2 fragment was PCR amplified from genomic DNA using the following oligonucleotides: (BHfp), 5′-GTGACCACGCAGTACGGCTCG-3′ (position 1505–1525 of A operon); reverse primer (BHrp), 5′-TGCCACAGGGATCCACGCACCT-3′ (position 1663–1684 of A operon). The A1–A2 fragment was PCR amplified from mixed flgA1 and flgA2 fragments using Hfp and BHrp. The primers included restriction sites for cloning: Hfp, HindIII; BHrp, BamHI. The PCR reaction contained 100 pmol of each primer and 10 ng of genomic DNA or KpnI–SalI fragment of cloned A operon. The reaction mixtures were heated for 5 min at the temperature of boiling water and cooled on ice before the addition of 2 U of Taq or Deep-Vent polymerase. The PCR reaction consisted of 30 cycles of (i) 1 min at 95°C; (ii) 1.5 min at 54°C in the case of the A locus, at 60°C for the flg B locus and 58°C in the case of the flg A1–A2 fragment; (iii) 1.5 min for the A locus, 2.5 min for the B locus and 1 min for A1, A2, A1–A2 fragments at 72°C. The final cycle was performed with an extension of 8 min at 72°C.
Construction of the mutagenesis vectors pAN and pBN
For construction of pAN and pBN insertion plasmids the NruI fragment (600 bp) of the A operon and the PvuII fragment (600 bp) of the B operon were used accordingly. To construct the pAN plasmid, the NruI fragment of the A operon was ligated to the HincII-cut pUC19, and then the SmaI–XbaI fragment, containing a novobiocin resistance (NovR) marker from pMDS20 (Holmes et al., 1994), was ligated to the resulting plasmid (also cut with SmaI and XbaI). The pBN plasmid was made using the same procedures; in this case, the PvuII fragment of the B operon was used instead of the NruI fragment of the A operon. Plasmids pAN and pBN, containing the flgA and flgB inserts in the opposite direction compared with the novobiocin gene, were used for transformation.
Construction of the mutagenesis vector pAN1
To construct the pAN1 insertion plasmid, the flg A1–A2 PCR fragment was exchanged for the HindIII–BamHI fragment, containing the NruI fragment of the A operon, of the pAN plasmid.
Transformation of H. salinarum
Transformations were carried out by the PEG method described by Cline et al. (1989) with the following modifications. Spheroplast formation was performed separately for each aliquot of the cell suspension; 1.5 ml of culture was harvested in a 1.5 ml Eppendorf tube by centrifugation at 3300 × g for 10 min, and then the cells were resuspended in 150 μl of spheroplasting buffer. Spheroplasts were formed by the addition of 15 μl of 0.5 M EDTA for 10 min at room temperature, after which DNA (about 5 μg) was added in a 10 μl volume of spheroplasting solution, and the mixture was incubated for 5 min. An equal volume of a solution containing 60% PEG 600 and 40% spheroplasting solution (w/v) was then added. The cells were incubated for 20 min at room temperature, after which they were added to 15 ml of usual growth medium containing 15% sucrose. The cells were allowed to recover for 12–24 h at 37°C with shaking. Selection of transformants was achieved by plating the cells into agar growth medium containing 0.1–0.2 μg ml−1 novobiocin (Sigma) and incubating at 37°C for 10–14 days.
Labelling of probes
The DNA fragment containing the A operon was labelled with [α-32P]-dATP, using a Multiprimer labelling kit (Amersham) to prepare probes for Southern and Northern hybridizations.
After electrophoresis in 0.8% agarose gels and denaturation, the DNA fragments were transferred to Biotrans nylon membranes (ISN) and hybridized. Hybridization was performed in 33% formamide, 0.05% SDS, 6 × SSC, 5 × Denhardt's solution and 100 μg ml−1 denatured fish sperm DNA at 42°C (Maniatis et al., 1982). The membranes were washed twice with 2 × SSC at room temperature and then in 2 × SSC, 0.1% SDS at 60°C for 15 min. Membranes were exposed to X-ray film for 12–24 h.
The electrophoresis was performed using a modified version of the method described by Goda and Minton (1995). RNA was denatured by adding the appropriate volume of loading buffer (6 × 0.25% bromophenol blue, 0.25% xylene cyanol, 30% glycerol, 1.2% SDS, 60 mM sodium phosphate buffer, pH 6.8) and heating at 75°C for 5 min. Samples were immediately loaded into the wells of 1% agarose gel containing 1 × TBE, 20 mM guanidine thiocyanate, ethidium bromide and electrophoresed in 1 × TBE. RNA was transferred to nylon membrane (GeneScreen Plus; NEN Research Products) using 1 × TBE (Kevil et al., 1997). RNA was cross-linked to the membrane by UV cross-linking (Spectrolinker XL-1000; Spectronics), and the membrane was air dried. The membrane was prehybridized (30 min, 65°C) in hybridization solution (7% SDS, 0.25 M Na2HPO4, pH 7.2) before the radiolabelled DNA probe was added, and hybridization was allowed to proceed at 65°C overnight. The membrane was rinsed in 0.1 × SSC/0.1% SDS at room temperature and then washed at 65°C in 0.1 × SSC/0.1% SDS. Membranes were exposed to X-ray film for 12–24 h.
The proteins separated by SDS–PAGE were transferred from the polyacrylamide slab gel to the nitrocellulose membrane (Sigma) according to the method described by Towbin et al. (1979). Electroblotting was carried out in a transblot apparatus (Hoffer Scientific Instruments) for 2 h at 250 V. Rabbit antiserum was raised by injecting purified flagella, 0.1 mg of protein per injection, using the method described by Southam et al. (1990). Antibody reactivity was detected by incubation with alkaline phosphatase conjugated with goat anti-rabbit immunoglobulin G and development with nitroblue tetrazolium and 5-bromo-4-chloro-3-indolyl phosphate (Promega).
The cell and flagella specimens were prepared by negative staining with 2% uranyl acetate on Formvar-coated copper grids. A sample of the culture (OD 1–1.2) was diluted from five- to ninefold in basal salt solution. The grid was placed on top of a drop of diluted culture medium and left for 1.5–2 min. Then, it was placed on top of a drop of 2% uranyl acetate and left for 1–1.5 min. Excess stain was removed by touching the grid to filter paper, and the grid was air dried. The grids were viewed on a JEM-100c electron microscope (JEOL) operating at 80 kV.
We thank Dr I. S. Serganova for help with the genetic techniques, and M. A. Stoylov and Jocelyn Carpenter for providing technical support in electron microscopic studies. This work was supported by a grant from the Russian Foundation for Basic Research (No. 98-04-48318). V.Y.T was supported by a UNESCO-MCBN Fellowship.