Flagella and pili are both necessary for efficient attachment of Methanococcus maripaludis to surfaces


  • Editor: Robert Gunsalus

Correspondence: James P.J. Chong, Department of Biology, University of York, Heslington, York YO10 5DD, UK. Tel.: +44 1904 328 628; fax: +44 1904 328 505; e-mail: james.chong@york.ac.uk


Methanococcus maripaludis has two surface appendages, namely flagella and pili. Flagella have been shown to be required for swimming, but no specific role has been assigned as yet to pili. In this report, wild-type M. maripaludis cells are compared with mutants lacking either pili or flagella or both surface appendages in their ability to attach to a variety of surfaces including nickel, gold and molybdenum grids as well as glass, silicon and mica. Wild-type cells attached to varying degrees to all surfaces tested, except mica, via their flagella as observed by scanning electron microscopy. Large cables of flagella were found to leave the cell and to be unwound on the surface. In addition, such cables were often found to connect cells. In contrast, cells lacking either flagella or pili or both surface appendages were unable to attach efficiently to any surfaces. This indicates a second role for flagella in addition to swimming in M. maripaludis, as well as a first role for pili in this organism, namely in surface attachment.


Methanococcus maripaludis has been shown to produce two surface appendages: flagella and pili (Wang et al., 2008; VanDyke et al., 2009; Ng et al., 2011). The flagella of archaea are a unique prokaryotic motility structure and the best studied of several different unusual appendages observed in various archaea (Ng et al., 2008; Albers & Pohlschroder, 2009; Jarrell et al., 2009). Archaeal flagella have many similarities to bacterial type IV pili (Peabody et al., 2003; Ng et al., 2006), an organelle that is involved in a type of surface motility called twitching (Bradley, 1980; Merz et al., 2000; Mattick, 2002). Both archaeal flagella and type IV pili are composed of proteins made with class III signal peptides cleaved by a specific signal peptidase (Pohlschroder et al., 2005) and both contain homologous genes for an ATPase and conserved membrane protein required for appendage assembly (Bayley & Jarrell, 1998; Peabody et al., 2003). There are significant structural similarities as well (Trachtenberg & Cohen-Krausz, 2006).

The flagella of M. maripaludis, shown to be essential for swimming, are composed of three flagellin glycoproteins modified with a tetrasaccharide N-linked at multiple positions in each flagellin (Kelly et al., 2009; VanDyke et al., 2009). Interference in glycan assembly or attachment leads to either nonflagellated cells or cells that can make flagella, but that are impaired in swimming, depending on the severity of the glycan defect (VanDyke et al., 2008, 2009). A number of accessory genes located downstream of, and transcribed with, the flagellins have been shown, by inframe deletion analysis, to also be essential for flagella formation (Thomas & Jarrell, 2001; VanDyke et al., 2009).

In M. maripaludis, the pili, like the archaeal flagella, are assembled from type IV pilin-like proteins (Szabo et al., 2007; Ng et al., 2011). The main structural protein is a very short glycoprotein (MMP1685), although at least three other type IV pilin-like proteins are all necessary for normal pili formation (Ng et al., 2011). The glycan attached to the pilins is a modified version of that found on flagellins, with a fifth sugar found attached as a branch to the N-acetylgalactosamine (Ng et al., 2011). No function has been assigned as yet to pili in this organism.

Methanococcus maripaludis is a model organism for study in archaea (Leigh et al., 2011). We have taken advantage of numerous genetic tools that allow for efficient transformation, inframe deletion and complementation studies (Tumbula et al., 1994; Hendrickson et al., 2004; Moore & Leigh, 2005) to generate mutants in M. maripaludis that lack one or other, or both, surface appendages. Examination of these strains by scanning electron microscopy demonstrated that strains lacking either or both of the surface structures were severely compromised in their ability to attach to various surfaces, demonstrating a second role for flagella and the first function for pili in this organism.

Materials and methods

Strains and growth conditions

Methanococcus maripaludis strain Mm900 (Moore & Leigh, 2005), obtained as a gift from John Leigh, was used to generate mutants having deletions in either flaK or eppA and the double mutant flaK eppA. Cells were grown in Balch medium III at 35 °C with shaking (Balch et al., 1979; Kalmokoff et al., 1988).

Generation of mutants in flaK and eppA

An inframe deletion mutant in flaK derived from M. maripaludis Mm900 was described previously (Ng et al., 2009). These cells are nonflagellated, but piliated and complementation of flaK in trans restores flagellation. Using the flaK mutant as a starting host, the subsequent deletion of the prepilin peptidase eppA was accomplished using the technique of Moore & Leigh (2005). An approximately 1-kb region upstream of eppA was amplified using the primers P1: 5′-CGCGGATCCCATTTCTATCAATTTTCCAC and P2: 5′-TTGGCGCGCCGGGGAATTATTCGCTCTTTGATAT. Primers P3: 5′-TTGGCGCGCCGGCGTTATAAATTATCTGGTGGGA and P4: 5′-CGCGGATCCCGTTTGACTGTTTGAACAGC were used to amplify approximately 1 kb downstream of the gene. Both P2 and P3 primers had AscI sites incorporated into them (underlined in primer), allowing for AscI cleavage of the two PCR products, followed by ligation and PCR amplification with primers P1 and P4 to generate an approximately 2-kb fragment that contained an inframe deletion version of eppA. Using the BamHI sites incorporated into primers P1 and P4 (underlined), this piece was cloned into pCRPrtNeo and transformed into the existing M. maripaludis flaK deletion strain with transformants screened for the eppA deletion by PCR using primers P5: 5′-CTGGAGCTGTATGAAATGCAACTGG and P6: 5′-CCTGCATTATCCCAGGTCATCC, which amplify across the deleted region. Similarly, the same plasmid was transformed into the Mm900 strain and transformants screened for the eppA deletion leading to mutants that were wild type for flaK, but deleted only for eppA.

Attachment studies

Wild-type and mutants cells were grown for 18 h at 35 °C before ethanol-sterilized substrates to be tested for attachment were added. Incubation continued at 35 °C with gentle agitation for a further 24 h. Tested substrates included various grids [200 or 400 mesh uncoated gold, nickel (Agar Scientific, Essex, UK) and molybdenum grids (Gilder Grids, Grantham, UK)], mica, silicon wafer chips (Agar Scientific) and glass.

To examine whether surface contact influenced the production of pili, the flaK deletion mutant was grown on Balch medium III plates (with 1.5% w/v Noble agar) for 4 days. Colonies were removed, resuspended in medium and briefly centrifuged. The pellet was gently resuspended in 2% glutaraldehyde in 100 mM sodium phosphate buffer, pH 7.2–7.4, containing 2% w/v NaCl for 30 min and examined by transmission electron microscopy (TEM), as described below.

Electron microscopy

Wild-type and mutant cells were examined by TEM to identify the presence of surface appendages. Cells were fixed with 2% glutaraldehyde in 100 mM sodium phosphate buffer, pH 7.2–7.4, containing 2% NaCl, applied to 200 mesh copper grids (Agar Scientific) coated with a formvar/carbon support film before staining with 2% phosphotungstic acid, pH 7.4. Samples were examined in a Tecnai 12 BioTWIN (FEI, Eindhoven, the Netherlands) operated at 120 kV.

For scanning electron microscopy, substrates were removed from the serum bottles and washed twice for 15 min in medium. Samples were then fixed with 2% glutaraldehyde in 100 mM sodium phosphate buffer containing 2% NaCl for 30 min at room temperature. The samples were then taken through a series of ethanol dehydration steps (25%, 50%, 70%, 90% and 100% ethanol) for 15 min each, followed by hexamethyl-disilazane. Dried specimens were mounted on aluminum stubs, sputter coated with approximately 2 nm of gold/palladium and examined in a JEOL JSM-7500F scanning electron microscope.


Methanococcus maripaludis possesses two surface appendages, flagella and pili, which could both potentially be involved in attachment of cells in the environment. To investigate the role of these appendages in attachment, mutants that lacked one or the other, or both, appendages were generated. The nonflagellated, but piliated mutant in the preflagellin peptidase flaK has been described previously (Ng et al., 2009). To create mutant strains that lacked pili, the eppA gene, the prepilin peptidase necessary for the removal of the signal peptide from pilins, was targeted. This would be predicted to prevent the incorporation of the nonprocessed pilins into pili fibers, leading to nonpiliated cells (Strom & Lory, 1993). If this gene is knocked out in the wild-type background, then cells should be flagellated, but nonpiliated. Mutants deleted for eppA were readily isolated and identified by a PCR screen (Fig. 1). Examination by TEM demonstrated that these mutants were, as predicted, flagellated (approximately 12 nm diameter fibers), but nonpiliated (Fig. 2). If eppA is deleted in the flaK mutant background, then such double-deletion mutants should lack both flagella due to the loss of flaK and also pili due to the deletion of eppA. Such mutants were readily isolated and identified by PCR screening (Fig. 1). Examination of the double deletion strains indicated that the cells did lack both surface appendages (Fig. 2). Complementation of the eppA deletion strain with a plasmid copy of the gene restored the piliated state (data not shown). Wild-type cells synthesized both appendages while the previously reported flaK mutant was nonflagellated, but piliated (approximately 6 nm diameter fibers) (Fig. 2).

Figure 1.

 PCR screening of the appendage mutants. (a) PCR screening using primers that amplify the region across the eppA gene. (b) PCR screening using primers that amplify the region across the flaK gene. Lane 1, wild type; lane 2, flaK mutant; lane 3, flaK eppA double mutant; lane 4, eppA mutant. M, 1 kb ladder.

Figure 2.

 Electron micrographs of wild type cells as well as the flaK, eppA and flaK eppA double mutant. Flagella and pili are apparent on the wild-type cells, whereas only flagella are seen on the eppA mutant. The flaK mutant possesses only pili, whereas the double mutant lacks both surface appendages. Cells stained with 2% phosphotungstic acid. Arrows indicate pili. Scale bar=200 nm.

The four strains were examined for their ability to attach to a variety of available substrates. Substrates tested included numerous uncoated electron microscopy grid types, as well as glass, mica and silicon wafer chips. After 24 h, wild-type cells were shown to attach to varying degrees to all surfaces tested, except mica (Fig. 3 for molybdenum grids and silicon chips; others not shown), although the number of cells attached to glass were few. Cells often preferred the edges of grids, where the rough surface seemed favorable for attachment (Fig. 3a). Fewer wild-type cells appeared to attach to the smooth surfaces of the grids. In stark contrast to the observation of wild-type cells, examination of the various mutants indicated that attachment of any of the mutants to any tested surface was almost nonexistent (Fig. 4b shows the result for the flaK mutant on gold grids; others are not shown). In the case of the flaK mutant (piliated, nonflagellated), a few attached cells were observed compared with the wild type, but only in the case of the nickel grids. In these cases, no cable-like appendages were seen arising from the cells, as expected if these cables are flagella (data not shown). Even after a 48-h incubation, where a large number of wild-type cells had accumulated on silicon, there was still no attachment of any of the mutant cells (Fig. 4c and d for eppA mutant; others not shown). Attachment of wild-type cells appeared to require metabolizing cells, because when the extremely oxygen-sensitive cells were exposed to air for 6 h and then allowed an opportunity to attach to silicon pieces over the course of a further 40-h incubation under aerobic conditions, they did not attach, although both appendages were still observed on the cell surface (data not shown). In addition, a mixture of the flaK mutants with the eppA mutants was also unable to attach to silicon pieces after a 48-h incubation (data not shown).

Figure 3.

 Scanning electron micrographs of wild-type cells attached to a molybdenum grid (a) and silicon (b). Scale bar=1 μm.

Figure 4.

 Scanning electron micrographs of wild type (a) and flaK mutant (b) cells attached to gold grids after 24 h and wild type (c) and eppA mutant (d) cells attached to silicon after 48-h incubation. Scale bar=1 μm.

Closer examination of the attached cells demonstrated that they were often tethered to the surfaces by a thick cable of flagella, which often was observed to unwind to strands of thinner diameter and ultimately to apparently single flagella (Fig. 5). The unwound flagella were most clearly observed when cells were attached to substrates with smooth backgrounds, such as glass and silicon (Fig. 5a and b). Here, one could follow bundles of flagella leaving the cell and then unwinding into thinner bundles and finally to apparently single flagella filaments attached to the substrate. Examination of grids with rougher surfaces, such as nickel, often led to the observation of individual cells attached to the surface in a more three-dimensional setting by multiple flagella cables, while other cables attached to neighboring cells (Fig. 5c). Again, the thicker cables could be seen to be unwound to thinner filaments, although this was harder to follow on the rougher surfaces. In some cases, it could be observed that the individual flagella were joining together into the thick bundle as they left the cell (Fig. 6).

Figure 5.

 Scanning electron micrographs of wild-type cells attached to various surfaces: (a) glass, (b) silicon, (c) nickel and (d) molybdenum. Thick arrows point out bundles of flagella leaving the cells, while the thin arrows indicate some of the individual flagella after unwinding from the thicker cable. Cables of flagella connecting neighboring cells (c) were common. Scale bar=100 nm.

Figure 6.

 Scanning electron micrograph of a wild-type cell attached to a nickel grid. Note the individual flagella (thin arrows) leaving the cell surface and joining into a thicker cable (thick arrow) as it leaves the cell. Scale bar=1 μm.

We attempted to see whether pili production was increased when cells were grown on a surface. As mutants were unable to grow attached to any surface tested, we examined the M. maripaludis flaK mutant after 4-day growth on plates. Cells were scraped off the plates and examined by negative staining. No evidence of increased pili number on the surface of these cells was observed; cells examined typically had only one or two pili and often no pili were observed on cells (data not shown).


Studies on the functions of various archaeal appendages are rare. In this report, deletions of key genes required for the biogenesis of flagella and pili led to the generation of M. maripaludis strains lacking pili or flagella or both appendages. Mutants missing either or both flagella and pili were shown to be extremely compromised in their ability to attach to any of the many potential substrates tested compared with wild-type cells. These studies show that besides their previously documented role in swimming (Chaban et al., 2007), flagella of M. maripaludis are also critical for attachment and are involved in cell-to-cell contacts. Similarly, a role in attachment is demonstrated for pili, the first role assigned to these unusual organelles, in this organism.

Very few studies on any archaea have been devoted to determining the functions of the several different types of archaeal appendages. Some organisms studied have only pili or flagella and some lack appropriate genetic systems in which to further characterize the roles of the various appendages. In Methanothermobacter thermautotrophicus, pili are the sole known surface appendages. Cells grown planktonically are poorly piliated; the expression of surface pili is much enhanced, however, under conditions where the cells adhere (Thoma et al., 2008). These pili were shown to be essential for the adherence of cells to a variety of surfaces, as antibodies to the major pilus structural protein lead to detachment of the cells. A genetic system that would allow the generation of nonpiliated mutants in this species is not available. This study was the first to demonstrate a role for pili in any archaeon. In the hyperthermophile, Pyrococcus furiosus, on the other hand, only flagella have been reported on the cell surface and these organelles were shown to be responsible for the adherence of cells to many types of surfaces, including ones found in the organism's natural environment (Nather et al., 2006), although adhesion to glass and mica was limited, as observed here with M. maripaludis. Again, lacking a genetic system in which to generate nonflagellated mutants, it was shown that adherent cells could be detached by antibodies directed against flagella. This was the first report of an adhesion role for archaeal flagella. In some of the electron micrographs, large cables of flagella can be observed to leave the cell before unwinding to the single flagella that are involved in adherence, as observed for M. maripaludis. Large cables of flagella were also seen to connect cells, an additional novel role for archaea flagella and an observation again made in this study for the flagella of M. maripaludis. Pyrococcus furiosus can also attach to Methanopyrus kandleri cells via its flagella, forming a unique archaeal bispecies biofilm (Schopf et al., 2008). Cable-like groups of flagella were shown to mediate cell-to-cell contact and attachment to gold grids in Methanocaldococcus villosus (Bellack et al., 2010). It was reported that, in some cases, flagella tufts unwound to single filaments about 500 nm from the cell surface. Interestingly, the cells were more heavily flagellated when attached to a surface than when free swimming.

It is only in two very recent studies that the roles for surface organelles in archaea that have both pili and flagella have been presented. In Haloferax volcanii, a member of the euryarchaeota kingdom, it was reported that flagella did not play a role in surface attachment and that type IV pili-like structures were responsible (Tripepi et al., 2010). In Sulfolobus solfataricus, a member of the crenarchaeota kingdom, it was shown that both pili and flagella were involved in attachment to surfaces (Zolghadr et al., 2010). Both of these studies utilized organisms with genetic systems that allowed the generation of mutants defective in each organelle. In S. solfataricus, it was shown that expression of the major flagella structural protein, FlaB, was dramatically reduced in adherent cells, leading the authors to conclude that flagella may be most important in the initial attachment to surfaces, but not for persistence after the initial attachment has occurred (Zolghadr et al., 2010). Although M. maripaludis is a euryarchaeon like H. volcanii, the findings with regard to the role of pili and flagella in attachment were unlike that of the halophile, but instead identical to those reported in the more distantly related organism, S. solfataricus.

As the M. maripaludis cells are clearly attached firmly by flagella, it begs the question of why the flagellated, nonpiliated strains could not attach well to surfaces. It may be that the pili render the initial attachment to a surface and only after this is formed can the flagella make the more permanent attachment. Even though abundant, flagella on their own cannot usually bind strongly enough to merit attachment, perhaps because they are involved in swimming until the cells are bound to a surface by pili. Unlike the current belief for S. solfataricus, it appears for M. maripaludis [as well as P. furiosus (Nather et al., 2006) and M. villosus (Bellack et al., 2010)] that the flagella are critical for continued attachment to surfaces, although further research will be necessary to ultimately discriminate between the roles played by pili and flagella in adherence for M. maripaludis. What is emerging in the few studies reported thus far is that the flagella of archaea play multiple roles in addition to their presumed primary role in motility, and that these roles are not consistent across different species.


This work was supported by grants from the Natural and Engineering Research Council of Canada (to K.F.J.) and Cancer Research UK (to J.P.J.C.). K.F.J. was the recipient of a Leverhulme Trust Visiting Professorship.