Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Hopanoid-free Methylobacterium extorquens DM4 overproduces carotenoids and has widespread growth impairment

  • Alexander S. Bradley ,

    Contributed equally to this work with: Alexander S. Bradley, Paige K. Swanson

    abradley@eps.wustl.edu (AB); cmarx@uidaho.edu (CM)

    Affiliations Department of Organismic and Evolutionary Biology, Harvard University, Cambridge, MA, United States of America, Department of Earth and Planetary Sciences, Washington University in St Louis, St Louis, MO, United States of America

    ORCID http://orcid.org/0000-0002-4044-2802

  • Paige K. Swanson ,

    Contributed equally to this work with: Alexander S. Bradley, Paige K. Swanson

    Current address: Finch Therapeutics, Somerville, MA, United States of America

    Affiliation Department of Organismic and Evolutionary Biology, Harvard University, Cambridge, MA, United States of America

  • Emilie E. L. Muller,

    Affiliation Equipe Adaptations et interactions microbiennes, Université de Strasbourg, UMR 7156 UNISTRA–CNRS Génétique Moléculaire, Génomique, Microbiologie, Strasbourg, France

  • Françoise Bringel,

    Affiliation Equipe Adaptations et interactions microbiennes, Université de Strasbourg, UMR 7156 UNISTRA–CNRS Génétique Moléculaire, Génomique, Microbiologie, Strasbourg, France

  • Sean M. Caroll,

    Current address: Pronutria Biosciences, Cambridge, MA, United States of America

    Affiliation Department of Organismic and Evolutionary Biology, Harvard University, Cambridge, MA, United States of America

  • Ann Pearson,

    Affiliation Department of Earth and Planetary Sciences, Harvard University, Cambridge, MA, United States of America

  • Stéphane Vuilleumier,

    Affiliation Equipe Adaptations et interactions microbiennes, Université de Strasbourg, UMR 7156 UNISTRA–CNRS Génétique Moléculaire, Génomique, Microbiologie, Strasbourg, France

  • Christopher J. Marx

    abradley@eps.wustl.edu (AB); cmarx@uidaho.edu (CM)

    Affiliations Department of Organismic and Evolutionary Biology, Harvard University, Cambridge, MA, United States of America, Department of Biological Sciences, University of Idaho, Moscow, ID, United States of America, Institute for Bioinformatics and Evolutionary Studies, University of Idaho, Moscow, ID, United States of America, Center for Modeling Complex Interactions, University of Idaho, Moscow, ID, United States of America

Hopanoid-free Methylobacterium extorquens DM4 overproduces carotenoids and has widespread growth impairment

  • Alexander S. Bradley, 
  • Paige K. Swanson, 
  • Emilie E. L. Muller, 
  • Françoise Bringel, 
  • Sean M. Caroll, 
  • Ann Pearson, 
  • Stéphane Vuilleumier, 
  • Christopher J. Marx
PLOS
x

Abstract

Hopanoids are sterol-like membrane lipids widely used as geochemical proxies for bacteria. Currently, the physiological role of hopanoids is not well understood, and this represents one of the major limitations in interpreting the significance of their presence in ancient or contemporary sediments. Previous analyses of mutants lacking hopanoids in a range of bacteria have revealed a range of phenotypes under normal growth conditions, but with most having at least an increased sensitivity to toxins and osmotic stress. We employed hopanoid-free strains of Methylobacterium extorquens DM4, uncovering severe growth defects relative to the wild-type under many tested conditions, including normal growth conditions without additional stressors. Mutants overproduce carotenoids–the other major isoprenoid product of this strain–and show an altered fatty acid profile, pronounced flocculation in liquid media, and lower growth yields than for the wild-type strain. The flocculation phenotype can be mitigated by addition of cellulase to the medium, suggesting a link between the function of hopanoids and the secretion of cellulose in M. extorquens DM4. On solid media, colonies of the hopanoid-free mutant strain were smaller than wild-type, and were more sensitive to osmotic or pH stress, as well as to a variety of toxins. The results for M. extorquens DM4 are consistent with the hypothesis that hopanoids are important for membrane fluidity and lipid packing, but also indicate that the specific physiological processes that require hopanoids vary across bacterial lineages. Our work provides further support to emerging observations that the role of hopanoids in membrane robustness and barrier function may be important across lineages, possibly mediated through an interaction with lipid A in the outer membrane.

Introduction

Hopanoids are pentacyclic triterpenoids present in some bacteria that are ubiquitous in sediments and sedimentary rocks. These types of molecules are among the best documented organic geochemical biomarkers in the rock record, and have been detected in rocks at least 1.7 billion years old [1]. Hopanoids are widely used by geochemists as general indicators of bacterial presence in ancient environments [2].

Previous studies of the physiological role and regulation of hopanoid production have revealed diverse results across taxa. Streptomyces coelicolor synthesizes hopanoids only during the formation of aerial hyphae, possibly to decrease the permeability of water across the membrane [3], whereas Streptomyces scabies produced hopanoids during submerged growth [4]. In the root nodule-colonizing bacterium Frankia alni, hopanoids were over-expressed in nodules [5], and enriched in nitrogen-fixing vesicles in Frankia spp., where they may function as a permeability barrier to oxygen [6]. A permeability role for hopanoids was also inferred in the alphaproteobacterium Zymomonas mobilis, which displayed an increased sensitivity to ethanol concentrations when hopanoid production was diminished by the addition of azasqualene, which inhibits the key enzyme squalene-hopene cyclase (SHC) that catalyzes the synthesis of hopanoids [7]. In the gammaproteobacterium Frateuria aurantia, hopanoid production increased with growth temperature [8]. Similarly, in the Firmicute Alicyclobacillus (formerly Bacillus) acidocaldarius, the abundance of side-chain containing hopanoids is positively correlated with growth at increased temperature or decreased pH [9]. The authors of the latter study hypothesized that hopanoids help overcome the expected increased permeability and decreased stability of biological membranes under these conditions.

Hopanoid-free mutants have yielded further insight into hopanoid function. A hopanoid-free mutant of S. scabies showed no growth phenotype under normal conditions, formed aerial hyphae, and did not differ from wild-type (WT) in its response to oxidative stress, osmotic stress, or tolerance of ethanol, high temperature, or pH stress [4]. A hopanoid-free mutant of the betaproteobacterium Burkholderia cenocepacia displayed no growth defect at neutral pH, but was sensitive to pH stress, antibiotics, and detergents [10]. In Burkholderia multivorans, a role for hopanoids in multiple antimicrobial resistance was suggested, based on increased membrane permeability in the mutant [11], and susceptibility to polymixin antibiotics in the presence of the isoprenoid synthesis inhibitor fosmidomycin [12]. Growth of a hopanoid-free mutant of the alphaproteobacterium Rhodopseudomonas palustris was indistinguishable from WT under standard conditions, but exhibited sensitivity to pH shock, bile salts, and antibiotics [13]. A Bradyrhizobium strain, another alphaproteobacterium, was found to covalently link hopanoids to lipid A in the outer leaflet of the outer membrane, and the hopanoid-free mutant shows increased sensitivity to stress [14]. Furthermore, in B. diazoefficiens hopanoids are apparently essential for growth [15]. In the cyanobacterium Nostoc punctiforme, a hopanoid-free mutant grew more poorly than WT under temperature stress, but better at low temperatures [16]. These results across organisms suggest that hopanoids play a role in membrane rigidity and integrity, with a pronounced link with the outer leaflet of the outer membrane. This is supported by biophysical data suggesting that hopanoids are associated with lipid A in the outer leaflet of the outer membrane, where they interact with glycolipids [17].

One type of microbial metabolism that has been little explored with regard to hopanoid function is methylotrophy. During growth on single-carbon compounds like methanol, methylotrophs generally oxidize these substrates to formaldehyde in the periplasm, and the formaldehyde is then utilized in the cytoplasm. This partitioning of the production and use of the toxic intermediate formaldehyde may make it necessary to maintain a high level of inner and outer membrane integrity. Members of the alphaproteobacterial genus Methylobacterium have long been known to produce hopanoids. These hopanoids include C31 hopanoids containing an additional methyl group at C2 [18,19], as well as those with side chains added through the successive action of enzymes encoded by hpnG and hpnH [20]. The first indication that the loss of hopanoids causes growth defects in Methylobacterium arose from having identified a M. extorquens DM4 isolate with a minitransposon insertion into shc that caused it to lose the ability to grow on dichloromethane (DCM) [21]. The mutant strain was determined to be hopanoid-free [20] [21]. Recently, a hopanoid-free M. extorquens PA1 mutant was used in a biophysical study that reported this strain to have lowered membrane order [17]. That work established that in Methylobacterium hopanoids are preferentially localized to the outer membrane, where they interact with lipid A. The only phenotypic analysis of this mutant, however, was to look at sensitivity to Triton X-100, which was increased by 1000-fold [17]. Here we characterize the changes in lipid content for the previously-isolated M. extorquens DM4 shc::miniTn5 strain (hereafter shc mutant) [21] and characterize a wide variety of growth defects.

Materials and methods

Media and growth conditions

M. extorquens DM4 WT strain (DSMZ 6343) and its hopanoid-free, shc mutant 41C5 [21] were grown in minimal medium at 30°C as previously described [22,23]. For liquid growth, carbon sources were added alone or in combination at the following concentrations (unless stated otherwise): succinate (3.5 mM), methanol (15 mM), DCM (10 mM) or betaine (i.e., trimethylglycine, 10 mM). On 1.8% agar plates, substrate concentrations were increased to 15 mM succinate, or 125 mM methanol. Where the pH of the medium was adjusted, it was via addition of HCl or KOH. In some experiments, which we specifically note below, we utilized an optimized version of Hypho medium described in [24], that we refer to as MPIPES. Where utilized, bile salts were added to a concentration of 1.5%. Cell culture was grown aerobically to mid-exponential phase in up to 50 mL batches for harvest for lipid analysis. Growth rate analyses were performed in 48-well plates (Costar) containing 640 μL of medium per well in an automated system situated in a warm room (30°C, 80% humidity) that consisted of a shaking plate tower (Liconic), a TwisterII microplate handler (Caliper), and a Wallac Victor2 plate reader (Perkin Elmer) operated by the software program Clarity [25]. Plates were shaken at a rate of 650 rpm and optical density was measured hourly at 600 nm. Growth rates for DCM were determined in 250 mM Erlenmeyer flasks containing 50 mL medium supplied with 32 μL of pure liquid DCM (Fluka) with gas-tight screwcaps equipped with miniert valves (Sulpelco). Tolerance to various stresses was evaluated by triplicate measures of the sizes of growth inhibition halos on plates as described previously [21]. Escherichia coli strains were grown at 37°C in Luria-Bertani broth. Antibiotics were added for selection at the following final concentrations: ampicillin 50 μg/mL, chloramphenicol 20 μg/mL, rifamycin 50 μg/mL, streptomycin 35 μg/mL, and tetracycline 10 μg/mL.

Liquid cultures were supplemented with an exogenous solution of cellulase enzyme to reduce cell clumping and permit accurate calculations of growth rates from a time series of optical density measurements. The cellulase solution was prepared by diluting 2 g of purified Aspergillus niger cellulase (Sigma-Aldrich, St. Louis, MO) into 10 mL of molecular grade water. To remove chemical impurities, the cellulase was loaded into a pre-wetted 10K MWCO Slide-A-Lyzer cassette (Thermo Fisher Scientific, Rockford, IL) and dialyzed twice for 1.5 hours in 4 L of deionized water at 4°C; longer dialysis times were found to impair the integrity of the dialysis cassette membrane. To ensure sterility, the dialyzed cellulase was passed through a 0.2 μm syringe filter (VWR) and diluted with sterile molecular grade water to a final concentration of 10 mg/mL or 20 mg/mL, as quantified on a ND-1000 spectrophotometer (Thermo). The final solution was found to be stable at 4°C for at least several months. Cellulase added to approximately 0.1 mg/mL in growing cultures eliminated most of the noise in optical density measurements caused by clumping, yet had no effect on cellular growth rate or yield (see below).

Lipid extraction and analysis

Lipids were extracted from M. extorquens DM4 cells following the method of Bligh and Dyer [26], with phosphate-buffered saline (PBS) substituted for the aqueous portion of the mixture to improve extraction of intact polar lipids (IPLs). Total lipid extracts were dried under a stream of N2 and stored at -20°C in the dark until analysis.

Hopanoids were analyzed by subjecting the total lipid extract to oxidation by periodic acid. Aldehydes produced by this process were subsequently reduced to alcohols with sodium borohydride (NaBH4) after the method of Rohmer et al. [27]. Alcohols were derivatized with N,O-bis(trimethylsilyl)trifluoro-acetamide (BSTFA) and 1% trimethylchlorosilane (TMCS) in the presence of pyridine, and analyzed via GC-MS on an Agilent 6890 gas chromatograph coupled to an Agilent 5973 quadrupole MS operating in full scan mode between m/z 50 and 750. In both LC-MS and GC-MS analyses, hopanoids were identified by comparison of retention times and mass spectra to previously published information [2830], or to samples with established hopanoid compositions. Hopanoid quantification was via GC-MS relative to an internal standard (epiandrosterone, Sigma). Squalene was also present in some samples and was quantified relative to an internal standard (squalane, Aldrich).

The fatty acid content of M. extorquens DM4 strains was determined by converting fatty acids to fatty acid methyl esters (FAMEs) in methanolic boron trifluoride (BF3/MeOH kit, Sigma) for 15 minutes at 70°C, followed by quenching with 2 mL of DCM-extracted deionized water. Transesterified lipids were extracted from this mixture with hexane, and dried over Na2SO4. Analysis occurred via GC-MS on the Agilent system described above, and FAMEs were quantified relative to an internal standard (lignoceric acid methyl ester; Sigma-Aldrich).

Total carotenoid content was quantified by suspending Bligh-Dyer lipid extractions in 250 μL of chloroform and determining the absorbance of the extract at 512 nm utilizing a Nanodrop 5000c spectrophotometer. Absorbance was converted to carotenoid concentration by comparison to a carotenoid standard calibration curve and normalization to cell number.

Construction of a plasmid to complement the shc mutant

A plasmid for squalene-hopene cyclase expression was obtained by first amplifying the shc gene from M. extorquens DM4 with primers that incorporated the ribosomal binding site from fae [31], a highly-expressed protein that catalyzes the first step of formaldehyde oxidation [32]. The resulting product was cloned into pCR-BluntII TOPO (Invitrogen), yielding the plasmid pAB169. The insert was then excised via XbaI/KpnI and ligated into the corresponding sites of the plasmid pCM62 [33], resulting in pAB170. pAB170 was mated from E. coli (10-beta, New England Biolabs) into the hopanoid-free mutant via tri-parental mating with M. extorquens DM4 and E. coli strain pRK2073 [34], resulting in strain CM3926. After analysis of the plasmid-complemented strain, a second plasmid, pPS10, was designed to complement the mutant by reinsertion of the native shc gene onto the chromosome in place of the miniTn5 disruption. pPS10 was constructed by amplifying the shc gene from M. extorquens DM4 and cloned into pCM433 [35]. pPS10 was integrated into the shc mutant via tri-parental mating with E. coli strain pRK2073, resulting in the double-crossover chromosomal shc restoration strain CM4201.

Toxicity disk assays

Toxicity assays were performed as described previously [21]. Briefly, on exactly 20 mL of solidified mineral medium containing 1.5% agar containing 10 mM methanol, 2.5 mM succinate, and 10 μg/mL tetracycline base, an overlay was added of precisely 7 mL of the same medium containing 0.7% agar to which a suspension of the strain to be investigated, grown in the same medium, was added to a final OD600 of 0.2. One 6 mm filter disk was added to each plate and spotted with 5 μL of a solution of the toxicant to be investigated, and plates were incubated face-up for 4 days at 30°C before measuring the diameter of growth inhibition.

Illumina resequencing of mutant genome

Cells of the shc mutant were grown to an optical density of 0.2 and harvested by centrifugation. Genomic DNA was extracted using a Wizard Genomic DNA purification kit (Promega) and checked for purity using a Nanodrop 5000c spectrophotometer. Genomic libraries were prepared at the Genome Technology Access Center at Washington University in St. Louis, followed by paired end (2x101) sequencing on an Illumina HiSeq 2000. Sequence files were compared to the M. extorquens DM4 reference genome using breseq [36].

Results

The shc mutant strain of M. extorquens DM4 lacks hopanoids

As a first stage in characterization of the M. extorquens DM4 shc mutant, its lipid content was compared to WT. While the GC-MS trace of the lipid extracts analysis for WT showed the presence of several hopanoid structures (Fig 1; Table 1), the shc mutant did not accumulate hopanoids (Fig 1). Although hopanoid structures were absent in shc mutant, the hopanoid precursor compound squalene was observed. The concentration of squalene in the mutant was lower than that of hopanoids in the WT cells (Table 1). Complementing the shc mutant with a plasmid expressing shc from the E. coli lac promoter (which is constitutively expressed at a low level in M. extorquens [31] recovered the hopanoid structures observed in the WT. The level of hopanoid accumulation in the complemented mutant was not to the same concentration as in the WT, and the level of squalene remained elevated (Table 1).

thumbnail
Fig 1. GC-MS m/z = 191 trace of total lipid extracts of Methylobacterium extorquens DM4.

A) wild-type: hopanoid peaks shown are diploptene (I) and diplopterol (II) + methyldiplopterol (III). B) hopanoid-free shc mutant showing absence of hopanoids C) shc mutant complemented with an shc-expressing plasmid showed a product accumulation similar to WT. Roman numerals refer to structures given in S1 Fig.

https://doi.org/10.1371/journal.pone.0173323.g001

thumbnail
Table 1. Lipid content of M. extorquens DM4 strains grown on succinate (μg/mg dry wt ± 1σ).

Homohopanoids are defined as hopanoids containing a side chain.

https://doi.org/10.1371/journal.pone.0173323.t001

The shc mutant strain of M. extorquens DM4 has an altered fatty acid profile

Both WT and shc mutant strains contained C16 and C18 saturated and mono-unsaturated fatty acids, with C18 mono-unsaturated fatty acid as the overwhelmingly most abundant fatty acid (> 85% of total fatty acids in all strains). The IPL distributions suggested that the hopanoid-free mutant strain produced relatively more C16 fatty acid than C18 fatty acid, as compared to the WT. This was confirmed by analysis of the fatty acid content. The mutant strain had a slightly higher proportion of C16 fatty acids overall (Fig 2A), and about twice the proportion of saturated fatty acids as the WT (Fig 2B).

thumbnail
Fig 2. Fatty acid production of wild-type and shc mutant strain grown on succinate.

A) Ratio of total C16 fatty acids to the sum of C16 + C18 fatty acids. B) Ratio of saturated/unsaturated fatty acids. Complementation plasmid carries the shc gene.

https://doi.org/10.1371/journal.pone.0173323.g002

The shc mutant strain of M. extorquens DM4 overproduces carotenoids

Given that hopanoids and carotenoid pigments are both derived from isoprenoid precursors, we examined the carotenoid content of the WT and mutant strains. Visually, the shc mutant was substantially darker pink than WT (Fig 3). When quantified spectrophotometrically following Bligh-Dyer extraction, this difference was revealed to result from a greater than 10x increase in carotenoid concentration compared to WT strain (Table 1).

thumbnail
Fig 3. Growth of wild type and mutant.

A) Growth on succinate in liquid media of WT DM4, shc mutant (showing clumping), and complemented shc mutant. B) Growth on succinate plates showing over-accumulation of pigmentation in shc and complemented shc relative to WT strain. C) Growth on plate containing succinate as growth substrate and 1 mM formaldehyde. WT strain and complemented mutant grow, but shc mutant is unable to grow. Complementation plasmid carries the shc gene.

https://doi.org/10.1371/journal.pone.0173323.g003

Hopanoid-free mutant strain is hypersensitive to a wide range of stressors

In order to determine whether the shc mutant renders M. extorquens DM4 sensitive to external stresses, we examined the inhibitory effect of a variety of agents relative to WT. It should be noted that our complementation plasmid contains the tetA tetracycline resistance pump, which is well know for affecting sensitivity to other factors [37]. However, we needed to use this plasmid because the transposon already introduced kanamycin resistance into the chromosome. Growth of the hopanoid-free mutant was inhibited on agar plates containing 1 mM formaldehyde, while WT strain and complemented strains grew normally (Fig 3). Similar results were obtained during growth on bile salts (1.5%), although in this case the complemented strain did not recover the ability to grow. We also examined the zone of growth inhibition imparted by a number of potential toxins and antibiotics (Table 2). Many such compounds affected the shc mutant more than the WT, consistent with the phenotype predicted by Saenz et al. [17]. In some instances, these effects were only partially relieved by complementation–for example rhodamine toxicity is only slightly alleviated through recovery of hopanoids by the shc plasmid.

thumbnail
Table 2. Sensitivity of strains to toxic compounds in disk diffusion assays.

https://doi.org/10.1371/journal.pone.0173323.t002

Hopanoid-free mutant strain grows poorly in liquid media and flocculates due to cellulose production

The hopanoid-free shc mutant strain was able to grow on both single-carbon and multi-carbon substrates such as methanol, formate, succinate, acetate, ethanol, glycerol, pyruvate, and betaine on agar plates, but was unable to grow on DCM in liquid culture, as reported previously [21]. Regardless of the carbon substrate, however, growth of the mutant on liquid media showed a defect in comparison to the WT, with low yields (as measured by total dry weight) and pronounced flocculation (Figs 3 and 4).

thumbnail
Fig 4. Growth curves of M. extorquens DM4 strains with succinate as the growth substrate.

The colors represent triplicate experiments for A) wild type (pink circle) and mutant (blue diamond) strains without cellulase addition, and B) wild type (pink circle) and mutant (blue diamond) strains with cellulase addition.

https://doi.org/10.1371/journal.pone.0173323.g004

We utilized an automated, robotic growth analysis system to determine growth rates for the mutant strain on C1 and multi-C substrates. Due to the flocculation in the shc mutant, however, it was difficult to determine a growth rate for this strain. Other reports had demonstrated that minor flocculation in M. extorquens can be ameliorated through removal of genes required for cellulose synthesis, suggesting a role of cellulose in biofilm formation [24]. This suggested that the severe flocculation in the shc mutant might be similarly ameliorated, and when the media was supplemented with purified cellulase, a large decrease in flocculation occurred. This allowed accurate determination of growth rates for wild-type and mutant strains. During growth on succinate in media amended with cellulase, the WT and complemented strains doubled in 2.6 hours, while the shc mutant had a doubling time of 4.9 hours. Similarly, on methanol plus cellulase the WT and complemented strains doubled in 2.9 hours, while the shc mutant had a doubling time 5.6 hours (Table 3), with final optical densities of the mutants indistinguishable from that of the wild type. The growth defect was also severe on formate, where the WT doubling time was 4.5 hours, while the mutant doubling time was 7.2 hours. This analysis confirms a substantial, generic growth defect for the mutant strain, in addition to the observed clumping (Fig 5).

thumbnail
Fig 5. Growth curves of M. extorquens DM4 strains in optimized media.

Succinate is the growth substrate. The colors represent 6x replicate experiments for wild-type (dark pink), shc mutant (dark blue), shc mutant with shc plasmid (green square), and shc mutant that has been rescued with chromosomal shc insertion (purple hexagon).

https://doi.org/10.1371/journal.pone.0173323.g005

thumbnail
Table 3. Doubling times (hours ± 1σ) of M. extorquens DM4 strains grown on various substratesa.

Mean of three replicates per strain/substrate combination.

https://doi.org/10.1371/journal.pone.0173323.t003

Hopanoid-free mutant strain grows poorly under pH stress

We grew the WT and mutant strains of M. extorquens DM4 in liquid media amended with cellulase, with succinate as a growth substrate under a range of pH conditions. We found that at near-neutral pH, the growth rate of the mutant strain lagged behind both the WT and the complemented strains. When grown in media where the pH had been adjusted to more extreme alkaline or acidic values, the growth rate of the mutant strain became markedly worse, with growth rates dropping to levels that were approximately 10% that of the WT and complemented strains (Fig 6).

thumbnail
Fig 6. Doubling times of M. extorquens DM4 wild-type and mutant strains as a function of medium pH.

The colors represent triplicate experiments for wild-type (pink circles), wild-type with empty vector (orange triangles), shc mutant (blue diamonds), shc mutant with shc plasmid (green square). Growth rates measured in the presence of cellulase.

https://doi.org/10.1371/journal.pone.0173323.g006

Partial complementation reveals that the hopanoid-free mutant strain accumulated another mutation

A plasmid expressing shc was generated to complement the mutant phenotypes observed. The growth defects observed for the shc mutant were partially reversed when complemented (Table 3). This included the previously reported [21] growth defect of the shc mutant with DCM as the sole carbon source. The intermediate growth phenotypes of the complemented strain may either be due to an incomplete return to the lipid profile of wild-type (see above) and/or toxicity from expressing shc from a multi-copy plasmid, polar effects of the transposon insertion, or to the accumulation of secondary mutations that were acquired during the mutagenesis of the WT.

To distinguish between these three possibilities, we compared the growth of the WT and mutant strain to a second complemented mutant strain in which the shc gene was inserted into the chromosome of M. extorquens, thereby restoring the shc locus to its wild-type state (Table 4). These experiments were performed in media optimized for Methylobacterium [24]. Results demonstrated that even with insertion of shc into the site of Tn5 disruption, full recovery of the WT phenotype was not achieved.

thumbnail
Table 4. Doubling times (hours ± 1σ) of M. extorquens DM4 strains on succinate in optimized media.

Mean of three replicates per strain/substrate combination.

https://doi.org/10.1371/journal.pone.0173323.t004

After observation that neither complementation of shc on a plasmid nor insertion of shc back onto the chromosome resulted in full recovery of the wild-type phenotype, we hypothesized that mutations had occurred during the mutagenesis of the WT to recover the shc mutant. We sequenced the genome of the shc mutant to detect these mutations. This analysis revealed only a single confirmed mutation, a single nucleotide polymorphism (CCT → CCC) at site 216230, which results in a synonymous mutation in a proline of a gene encoding a sulfonate transport system binding protein (METDI0211).

Discussion

Although hopanoids are not required for the growth of M. extorquens DM4, their absence imposes myriad constraints on growth. This result is similar to that observed in two other alphaproteobacteria, Bradyrhizobium [14], and R. palustris [13], which had shc mutants that display phenotypes under osmotic and pH stress. In M. extorquens a growth rate defect is observed even under optimal conditions, but final cellular densities of the mutant are as great as the WT. This suggests that the general importance of hopanoids is most significant under particular environmental conditions, which may relate to their physiological function. While this function is unknown, it likely relates to membrane permeability, and perhaps to cell division [38]. A recent suggestion, based on data from M. extorquens PA1, is that hopanoids order the outer membrane allowing for critical functions such as multidrug efflux [17]. Some hopanoid structures have been suggested to have a more particular function: for example, methylation at the C2 position has been related to membrane rigidification [39] and correlated with plant-microbe associations [38] [40]. Accumulation of high amounts of methyldiplopterol in Methylobacterium strains–a plant-associated microbe [41]–is consistent with this hypothesis. Deletion of the gene responsible for hopanoid methylation results in accumulation of enhanced amounts of tetrahymanol in R. palustris TIE-1 [42]. Methylobacterium does not accumulate tetrahymanol, a hopanoid-derived lipid that is produced in some hopanoid-containing bacteria by an additional enzymatic step [43].

Despite the fact that the shc mutant was highly sensitive to formaldehyde, and that growth with methanol leads to formaldehyde production, the relative growth defect of the shc mutant was similar for growth on methanol as on succinate, or almost any of the other compounds tested. Except for DCM, where it is essential to have shc, and ethanol, where there was no significant growth difference, the shc mutant grew 50–100% slower on all other substrates tested. We had expected to see a defect on methanol or betaine in particular, as these compounds lead to formaldehyde production, but this result suggests that either the formaldehyde level during methanol (or betaine) growth was sufficiently low to have less effect than 1 mM exogenous formaldehyde, or that the growth substrate may alter other aspects of sensitivity to formaldehyde. One possible explanation is that the documented induction in formaldehyde oxidation capacity (e.g., the tetrahydromethanopterin pathway) during growth on methanol is sufficient to buffer the effects of formaldehyde production [44,45].

Although the particulars of the selective environments of hopanoid-containing microbes clearly differ, there may still be some generalities as to their immediate biochemical role. Altered susceptibility to bile salts was interpreted to indicate that membrane damage occurs in the hopanoid-free versions of R. palustris [13], B. cenocepacia [10], B. multivorans [11,12], and Bradyrhizobium [14]. This prompted speculation that hopanoids play a role in membrane permeability to cations and/or protons [13]. The hypothesis that hopanoids play a role in membrane permeability has a long history [6,9], and our results are consistent with this idea. Other work has demonstrated that hopanoids are localized to the outer membrane, where they interact with peptidoglycan and play a role in cell division [38]. The increased toxicity of a broad range of compounds in the absence of hopanoids suggests that hopanoids play a role in modulating toxicity, and perhaps hopanoid-free mutants render cells prone to leakiness or other inability to deal with toxins. Nevertheless, the peptidoglycan biosynthesis inhibitor ampicillin does not produce significantly different zones of inhibition in these two strains, suggesting that the potential barrier role of hopanoids is not involved in resistance to ampicillin. Recent data imply that diplopterol may function in modulating the membrane ordering of saturated fatty acids [46], a suggestion supported by observations that hopanoids localize to particular cellular locations in some strains [47]. Saturated fatty acids comprise less than 10% of the membrane lipids in Methylobacterium, but saturated moieties are also present in the lipid A present in the outer membrane, which may also be modulated by diplopterol [46]. This supports previously suggested [46] interactions between hopanoids and lipid A.

Although we have thus far treated hopanoids as a single category, the diverse structures of hopanoids may play different physiological roles in bacteria. In M. extorquens AM1, approximately half the hopanoid product is diplopterol, which is the simplest amphiphilic hopanoid [4]. Adenosylhopane is not accumulated, but is an intermediate in hopanoid synthesis which is converted to bacteriohopanetetrol cyclitol ether (BHT-CE; structure VI), and guanidine-substituted BHT-CE (structure VII) [20]. Yet other bacteria, such as Rhodopseudomonas and Nitrosomonas, accumulate adenosylhopane under some conditions [30]. It remains unresolved whether these are biosynthetic by-products or whether this biosynthetic pathway is tightly regulated by the bacterium to produce a specific suite of products, each with different function. Recent work on expressing hopanoid genes in a cell-free system derived from Methylobacterium organophilum has begun to shed light on the function of these genes [48]. The wide range of hopanoid structures suggests the possibility that they serve a wide range of functions; this is consistent with biophysical evidence suggesting a diversity of functional roles [49].

Complementation by providing shc via a plasmid or as a chromosomal replacement resulted in a full return to wild-type for several phenotypes, such as lipid composition, but only partial restoration for growth rate. These results suggested that the original 41C5 isolate with the shc::miniTn5 allele contained an additional compensatory mutation. Accordingly, we identified an additional mutation relative to WT DM4. This was in a sulfonate transporter, which is consistent with hopanoids functioning in membranes. It is currently unclear whether the synonymous mutation identified affects expression of this gene, but there is precedent for single synonymous mutations to result in large changes in expression and fitness in other metabolic pathways in M. extorquens [50]. Future work will be required to determine whether this mutations underlies the lack of complementation, as well as to determine it arose as a suppressor mutations that was selectively advantageous in strains lacking hopanoids. These results may be quite interesting in terms of further uncovering the physiological function of hopanoids.

Our results indicate that hopanoids contribute to the fitness of M. extorquens DM4 under nearly all tested conditions. We observed a series of growth defects for the shc mutant beyond what had been previously tested, including extreme flocculation that appears to be related to cellulose biosynthesis. Consistent with other microbes, the shc mutant displayed increased sensitivity to a wide variety of toxic agents. These results confirm an emerging picture of the physiological role of hopanoids as important agents of membrane barrier function, while the cellulose interaction suggests that the specific functions that are dependent on hopanoids may differ among organisms.

In many other hopanoid-containing bacteria in which hopanoid-free mutants have been examined, growth defects have been seen under extremes of pH, temperature, and when subjected to stresses such as detergents and antibiotics [4,1012]. These phenotypes are also present in Methylobacterium, which additionally shows growth defects under usual growth temperature and pH. However, hopanoids are not required for growth as in B. diazoefficiens [15]. The phenotype of hopanoid-free M. extorquens DM4 offers one more example to the growing set of model organisms showing membrane stress in the absence of hopanoids. Understanding the important role of hopanoids may inspire new hypotheses to understand how most bacteria have adapted to life without hopanoids [51].

Supporting information

S1 Fig. Structures of hopanoids and hopanoid precursor in M. extorquens DM4.

I) diploptene, II) diplopterol, III) methyldiplopterol, IV) C32 hopanol V) bacteriohopanetetrol, VI) bacteriohopanetetrol cyclitol ether, VII) guanidine-substituted bacteriohopanetetrol cyclitol ether, VIII) tetrahymanol, IX) squalene.

https://doi.org/10.1371/journal.pone.0173323.s001

(PDF)

Acknowledgments

This work was supported by an Agouron Institute postdoctoral fellowship to A.S.B. (AI-F-GB9.08.2) and a grant from the National Science Foundation to C.J.M. and A.P. (EAR 1024723). Support of Région Alsace (France) for a PhD thesis grant to Emilie E.L. Muller is also gratefully acknowledged. We thank the Genome Technology Access Center in the Department of Genetics at Washington University School of Medicine for help with genomic analysis. The Center is partially supported by NCI Cancer Center Support Grant #P30 CA91842 to the Siteman Cancer Center and by ICTS/CTSA Grant# UL1 TR000448 from the National Center for Research Resources (NCRR), a component of the National Institutes of Health (NIH), and NIH Roadmap for Medical Research. This publication is solely the responsibility of the authors and does not necessarily represent the official view of NCRR or NIH. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Author Contributions

  1. Conceptualization: ASB EELM FB AP SV CJM.
  2. Data curation: ASB SV CJM.
  3. Formal analysis: ASB PKS EELM FB SMC AP SV CJM.
  4. Funding acquisition: ASB EELM AP SV CJM.
  5. Investigation: ASB PKS EELM FB SMC AP SV CJM.
  6. Methodology: ASB PKS EELM FB SMC AP SV CJM.
  7. Project administration: ASB SV CJM.
  8. Writing – original draft: ASB PKS EELM FB SMC AP SV CJM.
  9. Writing – review & editing: ASB CJM.

References

  1. 1. Summons RE, Jahnke LL, Hope JM, Logan GA. 2-Methylhopanoids as biomarkers for cyanobacterial oxygenic photosynthesis. 1999;400: 554–557. pmid:10448856
  2. 2. Brocks JJ, Summons RE. Sedimentary Hydrocarbons, Biomarkers for Early Life. In: Holland HD, Turekian KK, editors. Treatise on Geochemistry: Second Edition. Oxford, United Kingdom: Elsevier; 2004. pp. 63–115.
  3. 3. Poralla K, Munth G, Hartner T. Hopanoids are formed during transition from substrate to aerial hyphae in Streptomyces coelicolor A3 (2). FEMS Microbiol Lett. 2000;189: 93–95. pmid:10913872
  4. 4. Seipke RF, Loria R. Hopanoids are not essential for growth of Streptomyces scabies 87–22. J Bacteriol. 2009;191: 5216–5223. pmid:19502399
  5. 5. Alloisio N, Queiroux C, Fournier P, Pujic P, Normand P, Vallenet D, et al. The Frankia alni symbiotic transcriptome. Mol Plant Microbe Interact. 2010;23: 593–607. pmid:20367468
  6. 6. Berry AM, Harriott OT, Moreau RA, Osman SF, Benson DR, Jones AD. Hopanoid lipids compose the Frankia vesicle envelope, presumptive barrier of oxygen diffusion to nitrogenase. Proc Natl Acad Sci U S A. 1993;90: 6091–4. pmid:11607408
  7. 7. Horbach S, Neuss B. Effect of azasqualene on hopanoid biosynthesis and ethanol tolerance of Zymomonas mobilis. 1991;79: 347–350.
  8. 8. Joyeux C, Fouchard S, Llopiz P, Neunlist S. Influence of the temperature and the growth phase on the hopanoids and fatty acids content of Frateuria aurantia (DSMZ 6220). FEMS Microbiol Ecol. 2004;47: 371–379. pmid:19712325
  9. 9. Poralla K. Effect of temperature and pH on the hopanoid content of Bacillus acidocaldarius. FEMS Microbiol Lett. 1984;23: 253–256.
  10. 10. Schmerk CL, Bernards MA, Miguel A. Hopanoid Production Is Required for Low-pH Tolerance, Antimicrobial Resistance, and Motility in Burkholderia cenocepacia Hopanoid Production Is Required for Low-pH Tolerance, Antimicrobial Resistance, and Motility in Burkholderia cenocepacia. J Bacteriol. 2011;193: 6712–6723. pmid:21965564
  11. 11. Malott RJ, Steen-Kinnaird BR, Lee TD, Speert DP. Identification of hopanoid biosynthesis genes involved in polymyxin resistance in Burkholderia multivorans. Antimicrob Agents Chemother. 2012;56: 464–471. pmid:22006009
  12. 12. Malott RJ, Wu CH, Lee TD, Hird TJ, Dalleska NF, Zlosnik JE a, et al. Fosmidomycin decreases membrane hopanoids and potentiates the effects of colistin on Burkholderia multivorans clinical isolates. Antimicrob Agents Chemother. 2014;58: 5211–5219. pmid:24957830
  13. 13. Welander P V, Hunter RC, Zhang L, Sessions AL, Summons RE, Newman DK. Hopanoids Play a Role in Membrane Integrity and pH Homeostasis in Rhodospeudomonas TIE-1. J Bacteriol. 2009;191: 6145–6156. pmid:19592593
  14. 14. Silipo A, Vitiello G, Gully D, Sturiale L, Chaintreuil C, Fardoux J, et al. Covalently linked hopanoid-lipid A improves outer-membrane resistance of a Bradyrhizobium symbiont of legumes. Nat Commun. 2014;5: 5106. pmid:25355435
  15. 15. Kulkarni G, Busset N, Molinaro A, Gargani D, Chaintreuil C, Silipo A, et al. Specific hopanoid classes differentially affect free-living and symbiotic states of Bradyrhizobium diazoefficiens. MBio. 2015;6: 1–9.
  16. 16. Ricci JN, Morton R, Kulkarni G, Summers ML, Newman DK. Hopanoids play a role in stress tolerance and nutrient storage in the cyanobacterium Nostoc punctiforme. Geobiology. 2016; 1–11.
  17. 17. Sáenz JP, Grosser D, Bradley AS, Lagny TJ, Lavrynenko O, Broda M, et al. Hopanoids as functional analogues of cholesterol in bacterial membranes. Proc Natl Acad Sci. 2015;112: 11971–11976. pmid:26351677
  18. 18. Zundel M, Rohmer M. Prokaryotic triterpenoids 1. 3β-methylhopanoids from Acetobacter species and Methylococcus capsulatus. Eur J Biochem. 1985;150: 23–27. pmid:3926494
  19. 19. Zundel M. Hopanoids of the methylotrophic bacteria Methylocoocus capsulatus and Methylomonas sp. as possible precursors of C29 and C30 hopanoid chemical fossils. FEMS Microbiol Lett. 1985;28: 61–64.
  20. 20. Bradley AS, Pearson A, Sáenz JP, Marx CJ. Adenosylhopane: The first intermediate in hopanoid side chain biosynthesis. Org Geochem. Elsevier Ltd; 2010;41: 1075–1081.
  21. 21. Muller EEL, Hourcade E, Louhichi-jelail Y, Hammann P. Functional genomics of dichloromethane utilization in Methylobacterium extorquens DM4. Environ Microbiol. 2011;13: 2518–2535. pmid:21854516
  22. 22. Lee M, Chou H, Marx CJ. Assymetric, bimodal trade-offs during adaptation of Methylobacterium to distinct growth substrates. Evolution (N Y). 2009;63: 2816–2830.
  23. 23. Studer A, McAnulla C, Büchele R, Leisinger T, Vuilleumier S. Chloromethane-induced genes define a third C1 utilization pathway in Methylobacterium chloromethanicum CM4. J Bacteriol. 2002;184: 3476–3484. pmid:12057941
  24. 24. Delaney NF, Kaczmarek ME, Ward LM, Swanson PK, Lee MC, Marx CJ. Development of an Optimized Medium, Strain and High-Throughput Culturing Methods for Methylobacterium extorquens. PLoS One. 2013;8.
  25. 25. Delaney NF, Rojas Echenique JI, Marx CJ. Clarity: an open-source manager for laboratory automation. J Lab Autom. 2013;18: 171–177. pmid:23032169
  26. 26. Bligh EG, Dyer WJ. A rapid method of total lipid extraction and purification. Can J Biochem Physiol. 1959;37: 911–917. pmid:13671378
  27. 27. Rohmer M, Bouvier-Nave P, Ourisson G. Distribution of hopanoid triterpenes in prokaryotes. Microbiology. 1984;130: 1137–1150.
  28. 28. Talbot HM, Summons R, Jahnke L, Farrimond P. Characteristic fragmentation of bacteriohopanepolyols during atmospheric pressure chemical ionisation liquid chromatography/ion trap mass spectrometry. Rapid Commun Mass Spectrom. 2003;17: 2788–96. pmid:14673828
  29. 29. Talbot HM, Squier AH, Keely BJ, Farrimond P. Atmospheric pressure chemical ionisation reversed-phase liquid chromatography/ion trap mass spectrometry of intact bacteriohopanepolyols. Rapid Commun Mass Spectrom. 2003;17: 728–37. pmid:12661028
  30. 30. Talbot HM, Rohmer M, Farrimond P. Rapid structural elucidation of composite bacterial hopanoids by atmospheric pressure chemical ionisation liquid chromatography/ion trap mass spectrometry. Rapid Commun Mass Spectrom. 2007;21: 880–892. pmid:17294511
  31. 31. Chou HH, Berthet J, Marx CJ. Fast growth increases the selective advantage of a mutation arising recurrently during evolution under metal limitation. PLoS Genet. 2009;5.
  32. 32. Vorholt JA, Marx CJ, Lidstrom ME, Thauer RK. Novel Formaldehyde-Activating Enzyme in Methylobacterium extorquens AM1 Required for Growth on Methanol. J Bacteriol. 2000;182: 6645–6650. pmid:11073907
  33. 33. Marx CJ, Lidstrom ME. Development of improved versatile broad-host-range vectors for use in methylotrophs and other gram-negative bacteria. Microbiology. 2001;147: 2065–2075. pmid:11495985
  34. 34. Figurski DH, Helinski DR. Replication of an origin-containing derivative of plasmid RK2 dependent on a plasmid function provided in trans. Proc Natl Acad Sci U S A. 1979;76: 1648–52. pmid:377280
  35. 35. Marx CJ. BMC Research Notes. 2008;8: 1–8.
  36. 36. Deatherage DE, Barrick JE. Identification of mutations in laboratory-evolved microbes from next-generation sequencing data using breseq. In: Sun L, Shou W, editors. Engineering and Analyzing Multicellular Systems. Springer New York; 2014. pp. 165–188.
  37. 37. Bochner BR, Huang H, Schieven GL, Ames BN. Positive selection for loss of tetracycline resistance. TL—143. J Bacteriol. 1980;143: 926–933. pmid:6259126
  38. 38. Doughty DM, Coleman ML, Hunter RC, Sessions a. L, Summons RE, Newman DK. PNAS Plus: The RND-family transporter, HpnN, is required for hopanoid localization to the outer membrane of Rhodopseudomonas palustris TIE-1. Proc Natl Acad Sci. 2011;108: E1045–E1051. pmid:21873238
  39. 39. Wu C-H, Bialecka-Fornal M, Newman DK. Methylation at the C-2 position of hopanoids increases rigidity in native bacterial membranes. Elife. 2015;
  40. 40. Ricci JN, Coleman ML, Welander P V, Sessions AL, Summons RE, Spear JR, et al. Diverse capacity for 2-methylhopanoid production correlates with a specific ecological niche. ISME J. Nature Publishing Group; 2014;8: 675–84. pmid:24152713
  41. 41. Vorholt JA. Microbial life in the phyllosphere. Nat Rev Microbiol. Nature Publishing Group; 2012;10: 828–40. pmid:23154261
  42. 42. Neubauer C, Dalleska NF, Cowley ES, Shikuma NJ, Wu C-H, Sessions a. L, et al. Lipid remodeling in Rhodopseudomonas palustris TIE-1 upon loss of hopanoids and hopanoid methylation. Geobiology. 2015; n/a-n/a.
  43. 43. Banta AB, Wei JH, Welander P V. A distinct pathway for tetrahymanol synthesis in bacteria. Proc Natl Acad Sci. 2015;2015: 201511482.
  44. 44. Chistoserdova L. Methylotrophic Bacteria and Methanogenic Archaea Transfer Enzymes and Coenzymes Linking 1 C. Eur J Immunol. 1998;99: 1083–1830.
  45. 45. Marx CJ, Dien SJ Van, Lidstrom ME. Flux Analysis Uncovers Key Role of Functional Redundancy in Formaldehyde Metabolism. 2005;3. pmid:15660163
  46. 46. Saenz JP, Sezgin E, Schwille P, Simons K. Functional convergence of hopanoids and sterols in membrane ordering. Proc Natl Acad Sci. 2012;109: 14236–14240. pmid:22893685
  47. 47. Doughty DM, Dieterle M, Sessions AL, Fischer WW, Newman DK. Probing the subcellular localization of hopanoid lipids in bacteria using NanoSIMS. PLoS One. 2014;9: e84455. pmid:24409299
  48. 48. Bodlenner A, Liu W, Hirsch G, Schaeffer P, Blumenberg M, Lendt R, et al. C 35 Hopanoid Side Chain Biosynthesis: Reduction of Ribosylhopane into Bacteriohopanetetrol by a Cell-Free System Derived from Methylobacterium organophilum. ChemBioChem. 2015;16: 1764–1770. pmid:26032177
  49. 49. Poger D, Mark AE. The relative effect of sterols and hopanoids on lipid bilayers: When comparable is not identical. J Phys Chem B. 2013;117: 16129–16140. pmid:24299489
  50. 50. Agashe D, Sane M, Phalnikar K, Diwan GD, Habibullah A, Martinez-Gomez NC, et al. Large-Effect Beneicial Synonymous Mutation Meidate Rapid and Parallel Adaptation in a Bacterium. Mol Biol Evol. 2016;33: 1542–1543. pmid:26908584
  51. 51. Pearson A, Page SRF, Jorgenson TL, Fischer WW, Higgins MB. Novel hopanoid cyclases from the environment. Environ Microbiol. 2007;9: 2175–2188. pmid:17686016