Salt‐dependent regulation of archaellins in Haloarcula marismortui

Abstract Microorganisms require a motility structure to move towards optimal growth conditions. The motility structure from archaea, the archaellum, is fundamentally different from its bacterial counterpart, the flagellum, and is assembled in a similar fashion as type IV pili. The archaellum filament consists of thousands of copies of N‐terminally processed archaellin proteins. Several archaea, such as the euryarchaeon Haloarcula marismortui, encode multiple archaellins. Two archaellins of H. marismortui display differential stability under various ionic strengths. This suggests that these proteins behave as ecoparalogs and perform the same function under different environmental conditions. Here, we explored this intriguing system to study the differential regulation of these ecoparalogous archaellins by monitoring their transcription, translation, and assembly into filaments. The salt concentration of the growth medium induced differential expression of the two archaellins. In addition, this analysis indicated that archaellation in H. marismortui is majorly regulated on the level of secretion, by a still unknown mechanism. These findings indicate that in archaea, multiple encoded archaellins are not completely redundant, but in fact can display subtle functional differences, which enable cells to cope with varying environmental conditions.

. These regulatory proteins are only conserved in organisms of the phylum Sulfolobales.
Studies of methanogenic euryarchaea have shown that archaellum biosynthesis is induced at low hydrogen pressure, which is an electron donor in methanogenesis, the main source of energy for these organisms (Hendrickson et al., 2008;Mukhopadhyay, Johnson, & Wolfe, 2000). Changes in temperatue influence archaellum expression in the archaeon Methanococcus maripaludis (Ding et al., 2016). In addition, the first transcriptional regulator of archaellation in Euryarchaea was identified in this organism (Ding et al., 2016). As for another large group of euryarchaeota, the halophiles, the regulation of archaellum biosynthesis, is not fully understood.
The only information about regulation of archaellation in halophiles comes from the model Haloferax volcanii. It was shown that there is post-translational regulation of archaellation, because the presence of the conserved H-domain of pilins is required for correct archaellum assembly (Esquivel & Pohlschröder, 2014).
In contrast to the regulation of archaellation, the structure and biochemical composition of the archaellum have lately been mapped in increasingly greater detail (Banerjee, Neiner, Tripp, & Albers, 2013;Banerjee et al., 2015;Briegel et al., 2017;Chaudhury et al., 2016;Daum et al., 2017;Poweleit et al., 2016;Reindl et al., 2013). The protein components of the archaellum are quite conserved amongst motile archaea. However, there are some distinct differences between the archaellum operons of archaea belonging to the major phyla Euryarchaea and Crenarchaea. Euryarchaea generally encode multiple archaellins, the constituents of the archaellum filament (Albers & Jarrell, 2015;Pohlschröder, Ghosh, Tripepi, & Albers, 2011). These are made as preproteins and cleaved by a type III peptidase (Albers, Szabó, & Driessen, 2003;Bardy & Jarrell, 2003;Tripepi, Imam, & Pohlschröder, 2010). This raises the question why euryarchaea need several archaellins, while crenarchaea can generally achieve motility with one filament protein. The archaella of the euryarchaeon Haloferax volcanii were shown to be made up of one major and one minor archaellin (Tripepi, Esquivel, Wirth, & Pohlschröder, 2013;Tripepi et al., 2010). A possibility is that some of the additional archaellins in euryarchaea are required as minor structural component, for example, to construct a hook complex at the base of the archaellum filament.
The existence of this hook complex has been reported in one case, but no detailed studies on its composition or structure have been performed (Bardy, Mori, Komoriya, Aizawa, & Jarrell, 2002). Interestingly, a completely different explanation for the need of several archaellins came when studying the halophilic euryarchaeon Haloarcula marismortui Syutkin, Pyatibratov, Galzitskaya, Rodríguez-Valera, & Fedorov, 2014).
Haloarcula marismortui harbors four genes encoding archaellins on its genome. Two archaellins were found to constitute the archaellum filament; (a) flaB (WP_011224081.1) encoded on the major circular chromosome I, in the neighborhood of the fla accessory genes encoding components of the archaellum motor, and (b) flaA2 (WP_011222137.1) on the pNG100 plasmid (~33 kbp) (see Figure 1) (Syutkin, Pyatibratov, Galzitskaya, et al., 2014). The function of the other two archaellin genes is unknown: flaB2 (WP_049938970.1), located next to flaB on the main chromosome I, but transcribed from the opposite DNA strand, and flaA1 (YP_137938.1), encoding a truncated archaellin without a leader peptide, located on chromosome II. The isolation of an H. marismortui strain that had lost the pNG100 plasmid (named ΔpNG100) enabled detailed characterizsation of the different archaellins (Pyatibratov et al., 2008). The wild type archaella are constituted mainly of the FlaA2 archaellin, encoded on the pNG100 plasmid, with minor amounts of the genomic encoded FlaB. In contrast, the archaella of the ΔpNG100 strain are built exclusively from the FlaB archaellin ( Figure 1). It was demonstrated that both archaellins are able to form functional archaella alone (Syutkin, Pyatibratov, Beznosov, & Fedorov, 2012;Syutkin, Pyatibratov, Galzitskaya, et al., 2014). However, the FlaB protein proved very stable at high ionic strengths, while the FlaA2 protein can better resist media with low ionic strengths. Hence, it was proposed that these genes are ecoparalogs, that is, their products perform the same function under various environmental conditions (Syutkin, Pyatibratov, Galzitskaya, et al., 2014).
We have employed this interesting system to study regulation of motility. After varying the nutrient composition and salinity of the growth medium, we studied regulation of motility by studying archaellin transcription, translation, and correct assembly in to archaella. This study enabled us to identify the first environmental factors influencing archaellum expression in haloarchaea.
The culture was grown at 40°C on a shaking platform. The 30% salt medium consisted of 4 M NaCl, 0.6 M MgCl 2 *6H 2 O, 160 mM KCl, 10 mM CaCl 2 *2H 2 O, and 30 mM Bis-Tris. For experiments with different salinities, the salt solutions with 20% and 25% salinities were prepared by dilution the 30% solution with pure water. For experiments with different carbon sources 1% yeast extract (rich media), 1% glucose or 1% tryptone were added to 25% salt solution. The pH of the media was adjusted to 7.0.
To study the influence of different carbon sources, cells of both strains were grown in rich medium (25% salinity, 1% yeast extract, 40°C) to mid-log phase. Next the cells were pelleted, washed 3 times in the prewarmed final media and resuspended in media of 25% salinity containing either 1% yeast extract, 1% of glucose, or 1% of tryptone. Afterward, cells were incubated again at 40°C for 4 hr.
Samples were taken at 4 hr after starvation.

| qRT-PCR
Total RNA was isolated from 2 ml culture taken at the indicated time-points and conditions using the RNeasy Mini Kit (Qiagen, the Netherlands

| Isolation of archaella
Archaella filaments were prepared by precipitation with polyethylene glycol 6,000 as described by Gerl, Deutzmann, and Sumper F I G U R E 1 Schematic representation of archaella of the Haloarcula marismortui wild type and the ΔpNG100 strain. In the lower panel, the genomic locus of the archaellin genes is depicted. FlaB is encoded on the main chromosome next to the chemotaxis and assecory archaellin genes. FlaA2 is encoded on the plasmid pNG100. In wild-type H. marismortui, the archaellum filament is majorly built of plasmid encoded FlaA2 archaellin, with minor amounts of FlaB. As flaA2 is not present in the ΔpNG100 strain, the archaella filaments in this strain are exclusively made of FlaB. Archaellins are made as preproteins and N-terminally cleaved by a PibD signal peptidase after which they are added at the base of the growing filament. Besides the archaellins, several archaellum assecory genes (fla genes) encode constituents of the archaellum motor (flaHIJ), the archaellum stator (flaFG) or have still an unknown function (flaCDE)

| Electron microscopy
Archaella samples were prepared for EM by negative staining with 1% uranyl acetate on Formvar-coated Copper grids. A grid was floated on a 20µl drop of archaella solution (about 0.01-0.05 mg/ ml, in 3.4 M NaCl, 50 mM Tris-HCl, pH 8.0) for 2 min, blotted with filter paper, placed on top of a drop of 1% uranyl acetate and left for 1 min. Excess stain was removed by touching the grid to filter paper, and the grid was air dried. Samples were examined on a Jeol JEM-1400 transmission electron microscope (JEOL, Japan) operated at 120 kV. Images were recorded digitally using a highresolution water-cooled bottom-mounted CCD camera.

| Western blot analysis
Whole cell sample aliquots were harvested at exponential growth phase and equal amounts were loaded on 7% SDS-PAGE and blotted on a polyvinylidene fluoride membrane (Thermo Fisher). Western blots with independent biological replicates were repeated at least three times. To detect archaellum components, polyclonal antibodies were raised in rabbit against FlaB archaellin protein. These antibodies were diluted 1:2,000 and the membranes incubated at 4°C overnight. After incubating the membranes with the secondary goat anti-rabbit-horseradish peroxidase antibody, Clarity Western ECL blotting substrate (Bio-Rad) was added and chemiluminescent signals were recorded using an INTAS ECL Chemocam Imager (Intas, India).

| Chromatography mass spectrometry analysis
Proteins bands were excised and treated with Proteinase K (Promega) and trypsin (Sigma) at 37°C in a Thermo Mixer thermo shaker (Eppendorf, Germany). To stabilize proteinase K, CaCl 2 was added to the solution to a final concentration of 5 mM. The molar ratio of enzyme-to-protein was 1/50. The reaction was stopped by adding trifluoroacetic acid to the solution. Prior to mass spectrometric analysis, the peptides were separated by reversed-phase highperformance liquid chromatography using an Easy-nLC 1000 Nano liquid chromatography (ThermoFisher Scientific). The separation was carried out in a homemade column 25 cm in length and 100 μm in diameter packed with a C18 adsorbent; with an adsorbent particle size of 3 μm, and a pore size of 300 Å. The column was packed under laboratory conditions at a pressure of 500 atm. The peptides were eluted in a gradient of acetonitrile from 3% to 40% for 180 min; the mobile phase flow rate was 0.3 μl/min. Mass spectra of the samples were obtained using an OrbiTrap Elite mass spectrometer (Thermo Scientific, Germany). The peptides were ionized by electrospray at nano-liter flow rates with 2 kV ion spray voltage; ion fragmentation was induced by collisions with an inert gas (collision induced dissociation in a high-energy cell). The mass spectra were processed and peptides were identified using Thermo Xcalibur Qual Browser and PEAKS Studio (ver. 7.5) programs based on the sequences of UniRef-100. Parent Mass Error Tolerance was 2.0 ppm and fragment Mass Error Tolerance was 0.1 Da. Only peptides were taken into account with a "10 l gP." threshold value higher than 15.

| RE SULTS
We aimed at identifying differential regulation of archaellation in First we compared the flaB level of the ΔpNG100 strain with that of the wild type in the same medium containing 25% salinity (standard growth conditions). flaB transcription could be detected in both strains, but the flaB levels were upregulated with a log 2 fold of ~8 in the ΔpNG100 strain compared to the wild type, indicating that flaB expression is strongly repressed in the presence of pNG100 ( Figure 3a).
Next, we compared transcription of flaB, flaA2 and flaI in medium with 20% and 30% salinity, to that of the standard medium with 25% salinity. Previous work has shown that the archaella of the wild-type strain grown under optimal conditions are mainly formed by FlaA2 . In the wild type, expression of flaA2 was strongly decreased at 30% salinity compared with those at 25%. In addition expression of flaI was also found to be downregulated (Figure 3b), whereas flaB transcription was not affected. In medium with 20% salinity, the expression of archaellum genes was similar to that in medium with 25% salinity (Figure 3b). The ΔpNG100 strain lacks the flaA2 gene and therefore only expression of flaB and flaI was studied. In contrast to the wild-type strain, mRNA levels of both genes were slightly lower in the medium with 20% salinity compared to that of 25% (Figure 3c). Increasing the ionic strength to 30% salinity did not alter the expression of archaellum genes. Thus, in the wild-type archaellin flaA2 expression is repressed at high ionic strength. Comparison between the ΔpNG100 strain and the wild type shows that expression of flaB and flaI increases at higher ioning strength specifically in the ΔpNG100 strain ( Figure S1b).
As clear differences in archaellin transcription between the two strains were observed under growth in medium with different salinities, in a next step we aimed to study the translation of archaellins.  Pyatibratov et al., 2008;Syutkin, Pyatibratov, Galzitskaya, et al., 2014). In wild-type cultures,

SDS-PAGE and
we detected minor amounts of FlaB and higher levels of FlaA2, which is in correspondence with the qRT-PCR results (Figures 3a and 4).
While the low levels of FlaB seemed constant under different salinities, FlaA2 levels varied significantly. At 20% salinity, the FlaA2 signal was much increased compared to 25%, while at 30% salinity, the FlaA2 signal was reduced (Figure 4). This decrease in FlaA2 protein After the wild type, we also analyzed the ΔpNG100 strain that lacks flaA2 with Western blot. The FlaB signal was in all tested conditions much more intense as the FlaB signal of the wild type, corresponding with the qRT-PCR levels (Figures 2a and 3). At 20% salinity, a small band was detected just below the FlaB band, which might represent a not completely glycosylated form or proteolysis products of intracellular FlaB (Figure 4). The intensity of the FlaB signal was approximately similar at 20%, 25%, and 30% salinity, corresponding with the qRT-PCR results, which indicated no strong regulation at different salinities in the ΔpNG100 strain.
After measuring transcription and translation of archaellin genes in conditions with different ionic strengths, we proceeded to analyze their correct assembly of the proteins in archaella represented on the cell surface. To this end, extracellular archaella were isolated from both H. marismortui strains grown to stationary phase at 40°C in liquid media with different salinities. The samples were analyzed by SDS-PAGE on which again FlaA2 and FlaB could be distinguished based on their electrophoretic mobility ( Figure 5). The identity of the proteins was verified with Mass Spectrometry analysis (see Materials and Methods). This confirmed (>90% coverage, 340 unique peptides) that the major protein bands in the wildtype samples at 20% and 25% salinity represented archaellin A2: "H. marismortui Flagellin protein A2 (Q5V881_HALMA)" ( Figure S2). The integrity of archaella was checked by Electron Microscopy (Figure 6). The wild-type strain produced most archaella in its regular growth medium with 25% salinity, which is demonstrated by the intense FlaA2 band on the SDS gel. In medium with 20% salinity, a low amount of archaella was detected and no archaella could be isolated from cells growing at 30% salinity. The archaella of the ΔpNG100 strain were formed exclusively of FlaB. The archaella produced by ΔpNG100 were only detected in medium with 30% salinity. In addition, using optical microscopy we confirmed that only in the media where archaella could be isolated, cells were motile (data not shown). The strong differences between the amount of successfully assembled and secreted archaellins is not reflected in the intracellular archaellin levels as measured by qRT-PCR and Western blot. This suggests that an additional step of regulation might occur at the level of assembly of the filament. It was shown that transcription and translation of the flaA2 gene in the wild type are optimal at low salinity and are reduced at higher salinities. In contrast, the ΔpNG100 strain showed under the same conditions only subtle or no differences of flaB mRNA and intracellular FlaB archaellin concentration. The highest number of archaella in the wild type could be isolated at 25% salinity and in the ΔpNG100 strain at 30% salinity. Interestingly, FlaB (pre)-archaellins could be detected by immunoblotting in total cell lysates of cultures grown at all three salinities. This indicates that the archaellins are produced at different ionic strengths, but that their assembly in correct archaella occurs mainly in medium of a specific salinity. Thus, even though archaellins seem to be produced intracellularly, their secretion and assembly in archaella is subject to an additional level of regulation.

| D ISCUSS I ON
The differential regulation of archaellation in the wild type and ΔpNG100 strain fit very well to the suggested role of their ecoparalogous archaellins. FlaB was previously shown to be more stable at high ionic strength (Pyatibratov et al., 2008 and what is the evolutionary advantage of maintaining FlaA2. It was shown previously that the archaellin FlaA2 is more stable at low ionic strengths (Syutkin, Pyatibratov, Galzitskaya, et al., 2014) and indeed the above described analysis shows that its expression is induced in media with 20% salinity. As especially low ionic strength causes stress to halophilic organisms, the presence of the FlaA2 archaellin, and specifically the possibility to adapt the occurrence of FlaB and FlaA2 in archaella, represents a distinct advantage.
The observation of the two ecoparalogous archaellins in H. marismortui resembles that of Haloferax volcanii, which encodes six different pilins. It was shown that all six pilins could individually build pili, but that these pili had slightly different functionalities (Esquivel, Xu, & Pohlschröder, 2013). While some pili attached even better to surfaces as the wild type pili, others could not adhere anymore Pohlschröder & Esquivel, 2015). These findings, in combination with the present study, indicate that, at least in haloarchaea, multiple encoded pilins and archaellins are not redundant, but in fact have subtle functional differences that allow cells to cope with varying environmental conditions.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

AUTH O R CO NTR I B UTI O N
AS, MP, TEFQ, S-VA designed research. ASS, AKS, and MP performed experiments. MW analyzed qRT-PCR data. All authors interpreted data and contributed to writing the article.

DATA ACCE SS I B I LIT Y
The datasets generated and analyzed during this study are available from the corresponding authors T.E.F.Q and A.S.S. on reasonable request.