Characterization of methanotrophic bacteria on the basis of intact phospholipid profiles

Abstract

The intact phospholipid profiles (IPPs) of seven species of methanotrophs from all three physiological groups, type I, II and X, were determined using liquid chromatography/electrospray ionization/mass spectrometry. In these methanotrophs, two major classes of phospholipids were found, phosphatidylglycerol (PG) and phosphatidylethanolamine (PE) as well as its derivatives phosphatidylmethylethanolamine (PME) and phosphatidyldimethylethanolamine (PDME). Specifically, the type I methanotrophs, Methylomonas methanica, Methylomonas rubra and Methylomicrobium album BG8 were characterized by PE and PG phospholipids with predominantly C16:1 fatty acids. The type II methanotrophs, Methylosinus trichosporium OB3b and CSC1 were characterized by phospholipids of PG, PME and PDME with predominantly C18:1 fatty acids. Methylococcus capsulatus Bath, a representative of type X methanotrophs, contained mostly PE (89% of the total phospholipids). Finally, the IPPs of a recently isolated acidophilic methanotroph, Methylocella palustris, showed it had a preponderance of PME phospholipids with 18:1 fatty acids (94% of total). Principal component analysis showed these methanotrophs could be clearly distinguished based on phospholipid profiles. Results from this study suggest that IPP can be very useful in bacterial chemotaxonomy.

Phospholipid profile, LC/ESI/MS, Methanotroph

1 Introduction

Methane-oxidizing bacteria or methanotrophs are a vital link in the global carbon cycle, particularly through the oxidation of biogenic methane to carbon dioxide [1]. Methanotrophic bacteria are also recognized for their ability to metabolize or co-metabolize chlorinated solvents, such as trichloroethylene (TCE) [2–4]. Due to their importance in both natural and engineered processes, these cells have been examined extensively and can be divided into three general categories, type I, II and X, based on a wide variety of properties, including but not limited to GC content, optimal growth temperature, ability to fix nitrogen, mechanism of carbon assimilation, as well as predominant phospholipid fatty acids. Currently, categorization of specific methanotrophs requires lengthy experimentation. It would be desirable to construct rapid and accurate means of characterization to better measure in situ methanotrophic diversity.

Current approaches used for chemical characterization of microbial populations in natural environments include two techniques that analyze the cell membrane phospholipids. These are (1) phospholipid ester-linked fatty acid analysis by gas chromatography/mass spectrometry [5], and (2) intact phospholipid profiling (IPP) using liquid chromatography/electrospray ionization/mass spectrometry (LC/ESI/MS) analysis of bacterial membrane phospholipids [6]. Both techniques rely on the fact that phospholipids are found in the membranes of all living cells, but not in storage lipids, and are turned over rapidly in dead cells. Thus, their quantification provides an estimation of viable biomass [7]. Both techniques can give valuable insight into microbial community structure, based on the premise that there are a great number of dissimilar fatty acids in bacterial phospholipids and some bacteria contain unique fatty acids. The overall goal of the present study was to determine the IPPs of seven methanotrophic bacteria including a recently isolated acidophilic methanotroph and to demonstrate the potential of utilizing IPP for microbial identification.

2 Materials and methods

Phospholipid standards were prepared in methanol. Ammonium acetate was purchased from Aldrich (Milwaukee, WI, USA). All solvents were of optima grade (Aldrich). Deionized water was obtained from a Milli-Q water system (Millipore, Milford, MA, USA).

Seven strains of methanotrophs were selected for application of phospholipid profiling in microbial characterization, representing type I (Methylomonas methanica, Methylomonas rubra and Methylomicrobium album BG8), type II (Methylosinus trichosporium OB3b and CSC1, isolated from Moffet NAS, Moutain View, CA, USA) and type X (Methylococcus capsulatus Bath) methanotrophs [8,9] and a recently isolated acidophilic methanotroph, Methylocella palustris[10]. With the exception of M. palustris, the cells were grown in nitrate mineral salt medium [8] at 30°C with 5 μM copper (added as Cu(NO₃)₂) in 250-ml batch flasks. For M. palustris, the cells were grown in M2 medium with no added copper [10]. In all cases, the culture medium was no more than 15% of the total flask volume to prevent methane mass transfer limitations on growth. The cells were grown to the mid-exponential phase at an optical density (OD₆₀₀) of 0.75–0.8, and then collected for lipid extraction.

Total lipids were extracted with a modified Bligh and Dyer extraction method [5,11]. Approximately 6 ml of liquid bacteria culture was added to a test tube filled with 22.5 ml of methanol, dichloromethane (DCM) and phosphate buffer (2:1:0.8) extraction solution. The extraction mixture was allowed to stand overnight in darkness at 4°C. The lipids were partitioned by adding DCM and water such that the final ratio of DCM–methanol–water was 1:1:0.9. The upper aqueous phase was discarded and the lower organic phase was then decanted through a cellulose #4 filter into a test tube. The solid residue retained on the filter was washed three times with 1 ml DCM. The total lipid extract was dried under a gentle stream of nitrogen and was once again dissolved in methanol.

The LC/ESI/MS analysis was performed on a HP 1090 liquid chromatography/HP 5989B single quadrupole mass spectrometer with an electrospray interface. The LC was equipped with a 250 μl sample loop. A Zorbax (Hewlett Packard) C8 (150 mm×4.6 mm, 5 μm) or C18 (150 mm×3.2 mm, 5 μm) reverse phase high performance liquid chromatography (HPLC) column was used for the chromatographic separation of phospholipids. A gradient solvent system composed of solvent A (10 mM ammonium acetate) and solvent B (methanol) was used with a flow rate of 0.3 ml min⁻¹. At the beginning of the gradient, the mobile phase was 50% of A and 50% of B for 2 min. Solvent B was increased to 80% at 20 min and 100% at 60 min. The mobile phase was then held isocratically for 5 min.

The mass spectrometer was operated in the negative ionization mode. Nitrogen drying gas flow was approximately 12 l min⁻¹ at a temperature of 300°C. The electrospray needle was held at ground potential, the capillary, end plate and cylinder voltages were set at 3500, 2500 and 1500 V, respectively, and maintained at high positive potential for negative ionization. An Iris hexapole ion guide (Analytica of Bradford, Bradford, CT, USA) in the MS source enhanced the efficiency of ion transfer and the sensitivity of the mass spectrometer. The capillary exit voltage was set at 200 V at all times. The mass spectrometer was tuned using a solution provided by the manufacturer and was scanned from 70 to 1000 amu at approximately 0.4 scan s⁻¹. The concentrations of phospholipids were calculated based on the chromatographic area response of individual phospholipids relative to that of internal standard (18:1-lyso-phosphatidylglycerol (PG)) and reported as μg ml⁻¹ of liquid culture. The reproducibility of the analysis was better than 87% (n=5).

Phospholipids were designated as follows: C1:d1/C2:d2-PL (e.g. C16:0/C18:1-PE), where C1 and C2 are the numbers of carbon atoms in the fatty acyl chains on the sn-1 and sn-2 positions, respectively; d1 and d2 are the numbers of double bonds of the sn-1 and sn-2 fatty acyl chains, respectively; and PL is the abbreviation for phospholipids.

3 Results and discussion

The IPPs of the methanotrophic bacteria are shown in Fig. 1 and the identification of the phospholipids is presented in Table 1. Phospholipids were identified based on their mass spectra under negative ionization mode. Position of fatty acids on the sn positions was determined based on the ratio of intensity of fragment ions representing each fatty acid [11]. Fatty acids with the same degree of unsaturation but different double bond positions were not differentiated. Methanotrophic bacteria analyzed in this study contained two major classes of phospholipids, i.e. PG and phosphatidylethanolamine (PE) and its derivatives phosphatidylmethylethanolamine (PME) and phosphatidyldimethylethanolamine (PDME). Fatty acids on the sn-1 and sn-2 positions of the phospholipids were either saturated or monounsaturated, with chain lengths from 14 to 18. As can be seen in Table 1, methanotrophs had significantly different phospholipid profiles that can be used to categorize these cells as either type I, II or X. For example, the intact phospholipids of the known type I methanotrophs, M. rubra, M. album BG8 and M. methanica, predominantly had PG and PEs with hexadecenoic acids (C16:1) as the most common fatty acid chain. There were some differences, however, in these strains. All three strains had measurable amounts of PG and PE phospholipids, but M. rubra and M. methanica also had PME. Furthermore, M. album BG8 was not found to contain C18 fatty acids while M. methanica and M. rubra had low amounts of 18:1/18:0-PME. The type II methanotrophs, M. trichosporium OB3b and CSC1, exhibited a different IPP. M. trichosporium OB3b had approximately equal amounts of PG and PME with the major fatty acids being octodecanoate (C18:1) in both the sn-1 and sn-2 positions. CSC1, like M. trichosporium OB3b, had primarily octadecenoic fatty acids but the predominant phospholipid was PME. Furthermore, all but one phospholipid had unsaturated fatty acids in both sn-1 and sn-2 positions. The type II methanotrophs are the only bacteria that contained PDME (Table 1). For M. capsulatus Bath, PE accounted for more than 89% of the total phospholipids with the predominant fatty acids being C16:1. Most of the phospholipids (84%) were symmetrical phospholipids. The final organism in this study, an acidophilic methanotroph, M. palustris, represents a new genus recently isolated from an acidic Sphagnum peat bog [10]. The phospholipid composition was relatively simple for this organism, with three phospholipids detected, and the primary phospholipid and fatty acid were PME and octadecenoate.

1

Total ion chromatogram of intact phospholipids isolated from methanotrophic bacteria. Phospholipids of M. rubra (1), M. methanica (2), M. album BG8 (3) and M. trichosporium OB3b (4) were analyzed using a Zorbax C8 HPLC column (150 mm×4.6 mm, 5 μm), and those of CSC1 (5), M. capsulatus Bath (6) and M. palustris (7) were determined using a C18 HPLC column (150 mm×3.2 mm, 5 μm). See Table 1 for compound identifications.

1

Intact phospholipids isolated from methanotrophic bacteria (% of total phospholipids)

Peak #	Compound	Concentration (%)
		1a	2	3	4	5	6	7
1	16:1/14:0-PG		9.6				1.5
2	14:0/16:1-PG	11.8
3	16:1/16:1-PG	19.3	7.3	19.8			3.2
4	16:1/16:1-PG		6.3
5	16:0/16:1-PG	4.4	2.7	17.1			6.0
6	18:1/18:0-PME	5.9	6.7
7	18:1/16:1-PG					2.3
8	18:1/18:1-PG				53.1	15.9
9	18:0/18:0-PDME				3.9
10	16:1/14:0-PE		1.2				8.4
11	16:1/16:1-PE	11.8	5.0				38.8
12	16:1/16:1-PME	2.2	18.2
13	16:1/16:1-PE	23.1	11.6
14	16:1/16:1-PE			12.4
15	16:1/16:1-PE	19.1		19.0
16	16:1/16:1-PE		18.7	12.1
17	16:1/16:1-PE			12.5
18	16:1/16:1-PE		7.7
19	16:0/16:1-PE	1.9	4.1				42.0
20	16:1/16:1-PE	0.5	0.9
21	16:0/16:1-PE			7.0
22	18:1/16:0-PME				8.0
23	18:1/16:1-PDME					16.1
24	18:1/16:1-PME					14.7
25	18:1/16:1-PE							6.0
26	18:0/18:1-PME				35.1	25.1
27	18:1/18:1-PME					25.9		29.9
28	18:1/18:1-PME							64.1
	PE/PG ratio	1.8	2.9	1.7	0.9	4.5	18.8	–

^a1=M. rubra; 2=M. Methanica; 3=M. album BG8; 4=M. trichosporium OB3b; 5=strain CSC1; 6=M. capsulatus Bath; 7=M. palustris.

One interesting conclusion that can be drawn from these data is that all examined methanotrophs contained only phospholipids PG and either PE or its methylated forms, PME and PDME. These characteristics were observed to be typical for Gram-negative bacteria with extensive intracytoplasmic membranes [12], particularly methanotrophs [13]. The precursor of phospholipid synthesis, phosphatidic acid, was not detected in these organisms. The dominance of PE and the absence of phosphatidylserine can be explained by the fact that PEs are synthesized from the decarboxylation of phosphatidylserine. The methylated derivatives of PE (PME and PDME) are formed subsequently by stepwise methylation of PE [14].

Methanotrophic bacteria have been previously differentiated based on their fatty acid compositions. Type I methanotrophs contain predominantly C16 monounsaturated fatty acids, whereas type II contains mostly C18 monoenoics [15–17]. However, these characteristics are not unique to methanotrophs. For example, the ammonia-oxidizing bacteria showed a predominance of C16:1 fatty acids and nitrite-oxidizing bacteria have mostly C18:1 fatty acids [18], while the photosynthetic bacteria Rhodopseudomonas spheroides and Rhodomicrobium vannielii were found to contain 76.8% and 90%, respectively, of C18:1 fatty acids in total lipids [19]. As such, fatty acid analysis alone cannot be used to determine in situ methanotrophic diversity. Furthermore, as can be seen in Table 1, if one only considers the PE/PG ratio from IPP, it is also not possible to make accurate conclusions as to methanotrophic community composition. This ratio was greater than 1 (between 1.7 and 2.9) for the tested type I methanotrophs, and varied between 0.9 and 4.5 for the tested type II methanotrophs. Interestingly, a previous study found that M. trichosporium OB3b had a PE/PG ratio of less than 1.0 [13]. If the entire data set generated by IPP is considered using principal component analysis (PCA), however, one can see clear differences between methanotrophic types. In PCA, the original variables (phospholipids) were orthogonally transformed and a new set of uncorrelated variables, or principal components (PC) were extracted consecutively with the first PC accounting for most of original variability, the second PC the second largest variability, etc., as indicated by the eigenvalues. The number of PCs retained was determined by the Kaiser criterion (eigenvalues=1) [20]. Based on the Kaiser criterion, three PCs were selected. The percentage of variance expressed by the first three factors was 29.7%, 17.5% and 16.0%, respectively, with a cumulative total of 63.2% of the variance. As shown in Fig. 2, PCA of IPPs showed that type II methanotrophs (CSC1 and M. trichosporium OB3b) were closely clustered within one group and clearly separated from type I methanotrophs (M. rubra, M. methanica and M. album BG8), suggesting variations within groups were generally smaller than variations between groups. The type X methanotroph, M. capsulatus Bath, was placed on the loading plot close to the type I strains but distant from type II methanotrophic bacteria. Such a result confirms the opinion of some researchers that M. capsulatus Bath can be considered a type I strain [21]. Finally, M. palustris, a moderately acidophilic methane oxidizer, had an IPP separate from the tested type I, II and X strains. This finding is in agreement with the more extensive analyses of the characteristics of this strain which suggest it is a novel subtype of methanotrophs [10].

2

Factor loading plot showing variations in IPPs among type I (M. rubra (1), M. methanica (2) and M. album BG8 (3)), type II (M. trichosporium OB3b (4) and CSC1 (5)) and type X (M. capsulatus Bath (6)) methanotrophs and a recently isolated acidophilic methanotroph (M. palustris (7)).

In conclusion, we have presented a basis for the comparison and differentiation between functionally similar but different bacteria based on IPP. Although direct determination of the double bond position of unsaturated fatty acids in intact phospholipids remains to be done, results in this study suggest that IPP can be very useful in microbial chemotaxonomy [22]. For example, the IPP technique was successfully applied for differentiating five archetypes of pseudomonads that harbor different toluene degradation pathways [22]. However, before this technique can be extensively applied to study in situ microbial communities, further work is needed to: (1) establish a more comprehensive database of bacterial IPP profiles; and (2) determine if the IPP profile of a bacterial species changes under different environmental conditions, including in response to contaminants (e.g. TCE).