Novel lipids in Myxococcus xanthus and their role in ...davidcwhite.org/fulltext/570.pdfM. xanthus contains signiﬁcant amounts of unsatur-ated fatty acids at the sn-1 position. Mixtures

Environmental Microbiology (2006)

8

(11) 1935ndash1949 doi101111j1462-2920200601073x

copy 2006 The AuthorsJournal compilation copy 2006 Society for Applied Microbiology and Blackwell Publishing Ltd

Blackwell Publishing LtdOxford UKEMIEnvironmental Microbiology1462-2912copy 2006 The Authors Journal compilation copy 2006 Society for Applied Microbiology and Blackwell Publishing Ltd

2006

8

1119351949

Original Article

Unusual biosynthetic enzymes lead to diverse PE compositionP D Curtis R Geyer D C White and L J Shimkets

Received 5 May 2006 accepted 8 May 2006 For correspondenceE-mail shimketsugaedu Tel (

+

1) 706 542 2681 Fax (

+

1)706 542 2674

Novel lipids in

Myxococcus xanthus

and their role in chemotaxis

Patrick D Curtis

1

Roland Geyer

23

David C White

2

and Lawrence J Shimkets

1

1

Department of Microbiology University of Georgia Athens GA 30602 USA

2

Center for Biomarker Analysis University of Tennessee Knoxville TN 37932 USA

3

Department of Environmental Microbiology UFZ Leipzig-Halle D-04318 Leipzig Germany

Summary

Organisms that colonize solid surfaces like

Myxo-coccus xanthus

use novel signalling systems toorganize multicellular behaviour Phosphatidyletha-nolamine (PE) containing the fatty acid 161

w

5 (

D

11

)elicits a chemotactic response The phenomenon wasexamined by observing the effects of PE species withvarying fatty acid pairings Wild-type

M xanthus

con-tains 17 different PE species under vegetative condi-tions and 19 at the midpoint of development 13 of the17 have an unsaturated fatty acid at the

sn

-1 positiona novelty among Proteobacteria

Myxococcus xan-thus

has two glycerol-3-phosphate acyltransferase(PlsB) homologues which add the

sn

-1 fatty acidEach produces PE with 161 at the

sn

-1 position andsupports growth and fruiting body developmentDeletion of

plsB1

(MXAN3288) results in more dra-matic changes in PE species distribution than dele-tion of

plsB2

(MXAN1675) PlsB2 has a putative N-terminal eukaryotic fatty acid reductase domain andmay support both ether lipid synthesis and PE syn-thesis Disruption of a single

sn

-2 acyltransferasehomologue (PlsC of which

M xanthus

contains five)results in minor changes in membrane PE Derivati-zation of purified PE extracts with dimethyldisulfidewas used to determine the position of the doublebonds in unsaturated fatty acids The results suggestthat

D

5

and

D

11

desaturases may create the doublebonds after synthesis of the fatty acid Phosphatidyle-thanolamine enriched for 161 at the

sn

-1 positionstimulates chemotaxis more strongly than PE with161 enriched at the

sn

-2 position It appears that the

deployment of a rare fatty acid (161

w

5) at an unusualposition (

sn

-1) has facilitated the evolution of a novelcell signal

Introduction

In nature many bacteria are attached to surfaces wherethey are subject to different physical stresses from plank-tonic cells As a consequence surface dwelling bacteriaoften employ distinctive mechanisms including differentmotility motors chemical signals and sensory transduc-tion mechanisms A model surficial organism is

Myxococ-cus xanthus

a soil-dwelling

δ

-proteobacterium with acomplex life cycle Vegetative cells travel in swarms feed-ing on other bacteria When starved cells aggregate intolarge fruiting bodies and then differentiate into sporesDuring development swarms of

M xanthus

maintain sig-nificant species purity (Fiegna and Velicer 2005) even ina complex soil microbial community through mecha-nisms that are as yet unknown Lipid signals may beimportant

Myxococcus xanthus

cells respond chemotac-tically to purified

M xanthus

membrane phosphatidyleth-anolamine (PE) (Kearns and Shimkets 1998) containingthe fatty acid 161

ω

5 (Kearns

et al

2001a) Thisresponse is dependent on the presence of the extracellu-lar matrix and has only been observed under starvationconditions suggesting developmental relevance Chemo-taxis is also promoted by PE containing 181

ω

9 (Kearnsand Shimkets 1998) a fatty acid not found in

M xanthus

(Kearns

et al

2001a) but common in prey bacteria suchas

Escherichia coli

(Cronan and Rock 1996) Theresponse to PE containing 181

ω

9 uses a different sen-sory pathway from the response 161

ω

5 and may beinvolved in sensing prey (Kearns

et al

2000 Bonner

et al

2005)Mutations that alter the fatty acid composition of the

membrane can have dramatic effects on developmentFor example mutation of the genes encoding the E1

α

and E1

β

subunits of the branched-chain keto-acid dehy-drogenase complex (BCKAD) (involved in the synthesisof branched-chain fatty acids) decreases the abundanceof branched-chain fatty acids with a subsequent rise inunsaturated fatty acids (Toal

et al

1995) Thesemutants are deficient in fruiting body formation andsporulation Partial restoration of branched-chain fattyacid synthesis by feeding cells isovalerate improves

1936

P D Curtis R Geyer D C White and L J Shimkets


Environmental Microbiology

8

1935ndash1949

fruiting body formation and sporulation (Toal

et al

1995) These results argue that a balance of fatty acidsis necessary to maintain the complex life cycle of theorganism

Phosphatidylethanolamine is usually the most abun-dant phospholipid in Proteobacteria membranes Its bio-synthesis starts with the addition of an acyl chain (fattyacid) at the

sn

-1 position of glycerol-3-phosphate by glyc-erol-3-phosphate acyltransferase PlsB (Cronan and Rock1996) The second acyl chain is added at the

sn

-2 positionby PlsC (1-acyl-

sn

-glycerol-3-phosphate acyltransferase)forming phosphatidic acid A serine molecule is added tothe phosphate group forming phosphatidylserine whichis then decarboxylated to create the PE The fatty aciddiversity of PE molecules is due in part to the activities ofthe PlsB and PlsC acyltransferases and in part to the poolof available fatty acids

In this work the molecular diversity of membrane PEwas examined in wild-type and mutant strains showingthat

M xanthus

contains significant amounts of unsatur-ated fatty acids at the

sn

-1 position Mixtures of PE mol-ecules enriched in

sn

-1 161 are stronger attractants thanmolecules with

sn

-2 161 This effect is associated with161

ω

5 as it is the major monounsaturated fatty acid in

M xanthus

Results

Analysis of double bond position in unsaturated fatty acids

Fatty acid methyl esters (FAMEs) were prepared fromwild-type cells (Haumlrtig

et al

2005) separated by gaschromatography (GC) and analysed by mass spectrom-etry (MS) (Fig 1A) The most abundant fatty acid wasiso150 while the second most abundant was 161

ω

5 thisresult agrees with previous analyses (Kearns

et al

2001a Ware and Dworkin 1973) In order to identify thepositions of the double bond unsaturated fatty acids werederivatized with dimethyldisulfide (DMDS) which deriva-tizes the carbons flanking the double bond and facilitatesfragmentation between the derivatized carbons (Fig 1Binset) The resulting mass spectrum verifies the positionof unsaturation In the case of the major 161 GC peakthe double bond was as the

ω

5 position (Fig 1B) Theminor 161 species was the

ω

11 isomer 151 exists asboth

ω

10 and

ω

4 isomers (data not shown) although the

ω

4 isomer is more abundant (Fig 1A) Finally an iso171was found Because it eluted from the GC column earlierthan the iso170 it appeared to be a branched monoun-saturated fatty acid The fragmentation of the DMDS-derivitized product indicated unsaturation at the

ω

5 posi-tion (

∆

11

data not shown)

Fig 1

Analysis of fatty acids from

Myxococcus xanthus

A Fatty acid methyl esters were prepared from wild-type

M xanthus

DK1622 cells separated by gas chromatography and analysed by mass spectrometryB Unsaturated fatty acid methyl esters were derivatized with dimethyldisulfide (DMDS) to determine the point of unsaturation by detecting the size of resultant fragments after derivatiza-tion and fragmentation The mass spectrum of the 161

ω

5 DMDS derivative is depicted Frag-mentation between the

ω

5 and

ω

6 carbons would result in

mz

245 and 117 fragments respectively (inset) The parent ion (

mz

362) and fragment ions are detected in addition to the ion

mz

213 resulting from the loss of meth-anol from the fragment

mz

245

Unusual biosynthetic enzymes lead to diverse PE composition

1937



8

1935ndash1949

Analysis of

M xanthus

vegetative and developmental PE

Phosphatidylethanolamine from vegetative and develop-mental cells was purified by thin-layer chromatography(TLC) and liquid chromatography (LC) Finally high-performance liquid chromatography (HPLC)-coupledelectrospray ionization tandem mass spectrometry(HPLC-ESI-MS-MS) was used to identify the locationand abundance of specific acyl chains in PE Themolecular ions of PE molecules were isolated by theirmass (

mz

amu) in the negative mode (MndashH)

middot

frag-mented with N

2

causing collision-induced dissociation(CID) and the fragments were analysed The fatty-ester bonds at

sn

-1 and

sn

-2 exhibit different stabilitiesduring CID enabling the positions of fatty acids in PEmolecules to be determined by the relative abundanceof the fragment ions (Ekroos

et al

2002 Taguchi

et al

2005)

Phosphatidylethanolamine from vegetative

M xanthus

cells contains at least 17 different molecular species thisnumber rises to at least 19 during development (Table 1)The most abundant PE species during vegetative growthis PE-150150 (PE with the fatty acid 150 at the

sn

-1and

sn

-2 positions respectively 203) followed by PE-161150 (120) The lipids PE-161161 and PE-172150 (combined 120) have the same

mz

ratio and couldnot be separated Because CID is a partial hydrolysis itis difficult to separately quantify the abundances of differ-ent PE species with the same molecular weight In thisexample however relative abundance of the two speciesis easy to infer from the CID The abundance of 150 afterCID is 30-fold lower than 161 indicating that there ismuch more PE-161161 than PE-172150 This conclu-sion is supported by GC-MS which shows that 172 is alow abundance fatty acid (data not shown) (Kearns

et al

2001a)

Table 1

Mass spectrometry ndash collision-induced dissociation (CID) analysis of phosphatidylethanolamine (PE) in

Myxococcus xanthus

strains

Progenitorion (

mz

amu)

PE speciesFragmention

Per cent of total PE (wild type and mutants)

sn

-1

sn

-2WT Veg(DK1622)

WT Dev(DK1622)

plsB1

(LS2300)

plsB2

(LS2301)

esg

a

(JD300)

plsC2

(LS2305)

plsC4

(LS2306)

plsC5

(LS2307)

6466 141161

150131

25 28 30 33 3981

6486 140 150 56 53 80 44 63 68 616586 161

151141151

3272

6606 151161

150140

102 110 29 58 96

b

142 107197

6624 150 150 203 122 299 212 103 215 2796723 161

162151150

71 43 61 46 50 62123

6744 161 150 120 61 131 79 113 109 1036766 150 160 47 506846 161 162 71 65 30 47 97 44 46 406863 161

172161150

120 155 73 82 278 95 122 11663

6885 171160

150161

41 70 93 36 78

c

60 6543

6905 170 150 45 33 144 616984 172 161 34 45 57 53 397006 171 161 60 94 35 87 61 807023 170

181161150

75 86 40 46 53 47

7047 170 160 267122 nd nd 337144 nd nd 497167 nd nd 257187 nd nd 35

a

Strain JD300 (

esg

) was created previously (Toal

et al

1995)

b

PE-161140 is more prevalent than PE-151150 in this strain

c

PE-171150 is undetectable and both PE-161160 and PE-160161 are detected in this strainPhospholipids were extracted from whole cells and purified by thin-layer chromatography The PE fraction was extracted from the silica andseparated by liquid chromatography followed by electrospray ionization tandem mass spectrometry wherein collision with nitrogen gas causes apreferential fragmentation of the

sn

-2 fatty-ester linkage The most abundant fatty acid after CID is the

sn

-2 fatty acid The abundance of the PEspecies was determined by signal intensity (cps) In the case that two PE species have the same mass (

mz amu) the more abundant speciesis listed above the less abundant as determined by the ratio of sn-2 fatty acid abundances after CID Phosphatidylethanolamine species below2 of total were not found consistently and have been omitted

1938 P D Curtis R Geyer D C White and L J Shimkets

copy 2006 The AuthorsJournal compilation copy 2006 Society for Applied Microbiology and Blackwell Publishing Ltd Environmental Microbiology 8 1935ndash1949

The sn-1 position in vegetative PE is occupied by fivesaturated fatty acids and 12 unsaturated fatty acids whilethe sn-2 position is occupied by nine saturated fatty acidsand eight unsaturated fatty acids The 161 fatty acid thatis associated with chemotaxis is found in the sn-1 positionof five PE species In sum they account for approximately382 of all PE molecules 161 is also found at the sn-2position where it accounts for about 289 of total

The fatty acid pairings in vegetative PE species did notchange during development (Table 1) although somespecies changed in abundance In vegetative cells PE-150150 was most abundant (203) whereas underdevelopmental conditions PE-161161 was most abun-dant (155 including a small amount of PE-172150)Four new PE species appear at the midpoint of develop-ment and two disappear Development induces a notableincrease in larger PE species many of which contain 161in the sn-2 position Concordantly the summed percent-age of PE containing 161 at the sn-2 position rises from289 to 380

Phosphatidylethanolamine from M xanthus plsB1 and plsB2 mutants

PlsB initiates PE biosynthesis by transferring a fatty acidfrom the acyl carrier protein to the sn-1 position of glyc-erol-3-phosphate (Cronan and Rock 1996) The plsBgene is considered to be one of the minimal set of 256that are essential in bacteria (Gil et al 2004) Synthesisof PEs with unsaturated fatty acids at the sn-1 positioncould involve a PlsB that uses unsaturated fatty acyl sub-strates While there is typically one plsB gene per organ-ism a TBLASTN search with the E coli PlsB proteinsequence identified two significant M xanthus matchesPlsB1 (MXAN3288) showed 29 identity and 49 simi-larity to E coli PlsB over the full length of the protein (5e-56 expect) As seen in Fig 2 PlsB1 contains the con-served amino acids shown to be essential for cell viability

in E coli (Heath and Rock 1998 Lewin et al 1999) TheTIGR annotation of MXAN3288 places 22 additionalamino acids at the N-terminus of the product these maybe incorrectly assigned Our proposed plsB1 start sitehas a near-consensus ribosome binding site (RBS)AGGACG 12 bp upstream of the start site whereasMXAN3288 has no recognizable RBS Using our startsite plsB1 was deleted from the chromosome leavingonly the start codon separated from the stop codon by a6 bp XbaI site The resulting mutant LS2300 exhibitedno growth or motility defects and formed fruiting bodiescontaining 99 viable spores compared to the wild type

PlsB2 (MXAN1675) showed 33 identity and 48 sim-ilarity over the 452 C-terminal amino acids of the E coliPlsB sequence (807 amino acids total) PlsB2 also con-tains the conserved amino acids shown to be necessaryfor E coli viability (Fig 2) The N-terminal half of PlsB2showed some homology to eukaryotic fatty acid reduc-tases (FARs) An alignment of PlsB2 amino acids 1ndash309with eukaryotic FARs is shown in Fig 3 The NAD(P)H-binding motif [I V F]-X-[I L V]-T-G-X-T-G-F-L-[G A] con-served among eukaryotic FARs is indicated by the blackbox (Aarts et al 1997 Metz et al 2000 Moto et al2003 Cheng and Russell 2004) The M xanthussequence has a near-consensus NAD(P)H binding site aswell as other conserved blocks of amino acids suggestingthat the N-terminal portion of PlsB2 may function in fattyacid reduction MXAN1675 has a consensus RBS(AGGAGG) 8 bp upstream of the predicted start Theportion of the gene containing the PlsB2 active sitecorresponding to amino acids 416ndash868 was deletedto eliminate putative acyltransferase activity The plsB2mutant had no obvious growth motility or developmentaldefects (102 viable spores compared with the wild type)

Phosphatidylethanolamine was purified from vegetativeplsB1 and plsB2 cells and analysed by LC-ESI-MS-MS(Table 1) The most significant finding is that each mutantcontains unsaturated fatty acids at the sn-1 position While

Fig 2 Alignment of Escherichia coli PlsB and Myxococcus xanthus homologues The E coli PlsB protein (Accession No P0A7A7 amino acids 273ndash446) was aligned with the M xanthus PlsB1 (MXAN3288 313ndash487) and PlsB2 (MXAN1675 469ndash635) proteins Stars indicate conserved amino acids known to be essential for viability in E coli (Heath and Rock 1998 Lewin et al 1999)

EcoliPlsB

MxanthusPlsB1

MxanthusPlsB2

EcoliPlsB

MxanthusPlsB1

MxanthusPlsB2

EcoliPlsB

MxanthusPlsB1

MxanthusPlsB2

Unusual biosynthetic enzymes lead to diverse PE composition 1939


the majority of the fatty acid pairings for individual PEspecies remain the same there are some changes in PEcomposition for these mutants The plsB1 mutant lacksmany of the high-molecular-weight PE species present inwild-type notably the species with mz 6984 7006 and7023 Each of these species has 161 at the sn-2 posi-tion Furthermore mz 6863 the principal species with161 at the sn-2 position declines by nearly 40 For thisreason total unsaturated fatty acids at the sn-2 positiondecline from about 44 to roughly 10 in the plsB1mutant The plsB2 mutant contains the high-molecular-weight PE species missing in the plsB1 mutant in additionto other larger PE species (mz 7047 7167 and 7187)Despite these changes the three most abundant PE spe-cies are present in quantities similar to those of the wildtype These results argue that PlsB1 and PlsB2 haveoverlapping substrate specificities

The fatty acid composition at the sn-2 position appearsto be dependent on the sn-1 fatty acid given that a muta-tion that alters sn-1 acyltransferase activity also alters thesn-2 fatty acid content If the sn-2 acyltransferase addedfatty acids irrespective of which fatty acids occupy the sn-

1 position then the sn-2 fatty acid content would remainunchanged in this mutant

Analysis of PE from an esg mutant

The esg mutant (JD300) is deficient in the BCKAD nec-essary for branched-chain fatty acid synthesis resultingin PE that is enriched for 161 (Toal et al 1995) Aspreviously reported this mutant has reduced amounts ofbranched-chain fatty acids 150 and 170 (Table 1) (Toalet al 1995) PE-150150 in the esg mutant decreasednearly twofold compared with the wild type Branched-chain fatty acid synthesis is not abolished in this mutantas a bypass mechanism has been recently elucidated(Mahmud et al 2002) The mass-spectral signal corre-sponding to PE-161161 increased 23-fold constituting278 of the total PE In the sn-1 position the amount of161 rises from roughly 382 in wild type to approxi-mately 716 in the esg mutant While this mutant lacksmany of the high-molecular-weight PE species found inthe wild type due to the fact that the 17-carbon branched-chain fatty acids are missing the sn-2 161 content is

Fig 3 Alignment of several eukaryotic fatty acid reductases (FARs) and Myxococcus xanthus PlsB2 (amino acids 1ndash309) Mouse Mus musculus FAR 1 BC007178 amino acids 1ndash319 mouse FAR 2 BC055759 1ndash319 silkworm Bombyx mori FAR BAC79426 1ndash328 jojoba plant Simmondsia chinensis FAR AD38040 1ndash353 The boxed portion indicates the NAD(P)H-binding motif [I V F]-X-[I L V]-T-G-X-T-G-F-L-[G A] (Aarts et al 1997 Metz et al 2000 Moto et al 2003 Cheng and Russell 2004)

mouseFAR1mouseFAR2insectFARjojobaFARMxanthusPlsB2








similar to that in the wild type largely due to the increasein PE-161161 Fatty acid pairings are similar in the esgmutant and in the wild type but the total fatty acid contentis changed This again suggests that the fatty acids addedto the sn-2 position are dependent on the sn-1 acyl chainand not the overall fatty acyl-ACP pool Again the 161acyl chains can be attributed mainly to the ω5 specieswhich is by far the most abundant 161 isomer inM xanthus

Analysis of PE from three M xanthus plsC mutants

PlsC is an essential enzyme that adds a fatty acid to thesn-2 position of 1-acyl-sn-glycerol-3-phosphate (Cronanand Rock 1996) Typically there are one to two plsCgenes per organism (Cronan and Rock 1996 Shih et al1999) Searching the M xanthus genome with the E coliPlsC sequence (TBLASTN) revealed five homologuesplsC1 (MXAN3330) plsC2 (MXAN3969) plsC3(MXAN5578) plsC4 (MXAN0955) and plsC5(MXAN0147) The related protein products had 28ndash36identity and 49ndash56 similarity to the E coli protein overthe entire length of the protein PlsC enzymes have con-served amino acid blocks similar to PlsB enzymes (Liet al 2003 Weier et al 2005) and the M xanthus PlsChomologues contain many of the conserved residues(Fig 4) All the homologues had the characteristic smallsize of PlsC enzymes The plsC2 plsC4 and plsC5 geneswere each mutated by homologous integration of a plas-mid bearing an internal fragment of each gene This

method results in a merodiploid in which both copies ofthe gene are truncated Phosphatidylethanolamine fromeach mutant was purified and analysed by LC-ESI-MS-MS (Table 1) Although some changes were observedsuch as an increase in PE-150150 in the plsC4 mutantmost values stay within 15-fold of each other for compa-rable PE species These results suggest that the homo-logues are either functionally redundant or biologicallyinactive

Chemotactic excitation by PE from different strains

Previous work demonstrated that M xanthus alters itsmotility in response to PE with specific fatty acids (Kearnsand Shimkets 1998) Solitary M xanthus cells travel oversolid surfaces in one direction for approximately 7 minbefore reversing direction (reversal period) (Blackhart andZusman 1985) In the presence of certain PEsM xanthus cells are stimulated to travel longer distancesbefore reversing directions Not all PEs produce thisresponse indicating a molecular specificity related to aparticular fatty acid PE-120120 and PE-181ω9181ω9elicit a response whereas PE-140140 PE-160160PE-170170 and PE-180180 do not (Kearns andShimkets 1998) Cells adapt over time and the stimulatedreversal period falls back to the unstimulated level Stim-ulation and adaptation are hallmarks of chemotaxis(Kearns and Shimkets 1998)

Neither of the chemoattractants PE containing 120 orPE containing 181ω9 is found in M xanthus Phosphati-

Fig 4 Alignment of Escherichia coli PlsC and Myxococcus xanthus PlsC homologues The E coli PlsC protein (P26647 amino acids 22ndash245) was aligned with the five M xanthus PlsC homologues PlsC1 (MXAN3330 22ndash254) PlsC2 (MXAN3969 30ndash263) PlsC3 (MXAN5578 26ndash245) PlsC4 (MXAN0955 45ndash264) and PlsC5 (MXAN0147 61ndash282)

MxanthusPlsC4MxanthusPlsC5MxanthusPlsC2MxanthusPlsC3MxanthusPlsC1EcoliPlsC






dylethanolamine containing 181 is found in many bacteriaM xanthus preys on and may be involved in prey recog-nition The nature of a native attractant found in theM xanthus membrane was revealed through metabolicengineering and linked to the fatty acid 161ω5 (Kearnset al 2001a) Phosphatidylethanolamine 161ω5161ω5was synthesized and found to be about 1000-fold moreactive than the other two attractants and active at physi-ological concentrations The chemical synthesis was per-formed with identical fatty acids at both positions simplybecause it was easier and was completed without priorknowledge of the PE composition of the M xanthus mem-brane Table 1 illustrates the unexpected finding that PE161161 is both present and abundant in M xanthus andthere are a wide variety of other species containing 161at the sn-1 position This raises the question of whetherthe chemotactic response is more sensitive to 161ω5 atthe sn-1 position or the sn-2 position

The per cent PE with 161 at a given sn position wasdetermined from the data in Table 1 by summing the per-centages for each PE containing 161 at that position Incases where the PE contained multiple fatty acid pairingsonly the most abundant species was considered The PEwith mz 6885 peak in the esg mutant did not contain PE-171150 (like the wild type) but instead had a mixture ofPE-160161 and PE-161160 as represented by equalamounts of each fatty acid after CID For the purpose ofthis analysis 39 was assigned to sn-1 and 39 to sn-2 161 The 161 composition for vegetative wild type 24 hdeveloping wild type plsB1 and esg cells is reported inTable 2 and represent the most varied examples of 161

distribution among the strains reported in Table 1 esg PEhad by far the largest amount of sn-1 161 at 716 Thenext highest is vegetative wild-type PE at 382 thendevelopmental wild-type PE at 281 and finally plsB1 PEat 234 However this order is not preserved for the sn-2 position The highest sn-2 161 composition is develop-mental wild-type PE at 380 followed by esg PE at316 vegetative wild-type PE at 289 and plsB1 PE at73

The ability of the collective pool of PE from each strainto stimulate chemotaxis was assayed by tracking themovement of wild-type cells by time-lapse video micros-copy The reversal period of cells in the absence of PEwas 70 plusmn 10 min and provides the baseline for unstimu-lated cells (Table 2) In trial experiments the optimalresponse was found with 2 microg of PE distributed over anarea of about 016 mm2 and overlayed with about 5 times 104

cells Phosphatidylethanolamine from the esg mutantwas most active and produced a reversal period of473 plusmn 170 min (Table 2) esg PE has the largest abun-dance of 161 at the sn-1 position and more average levelsof sn-2 161 Vegetative and developmental wild-type PEand plsB1 PE produced comparable reversal periodsPhosphatidylethanolamine from the plsB1 mutant has asimilar amount of sn-1 161 to wild-type cells but muchlower sn-2 161 Together these results argue that the sn-1 position is more important than the sn-2 position forstimulating chemotaxis A linear regression analysis wasperformed and an excellent relationship was observedwith sn-1 161 (R2 = 095) but not sn-2 161 (R2 = 009)

Although the sn-1 position appears to be more impor-tant for chemotactic stimulation than sn-2 this relationshipmay be driven by a particular PE species The most dra-matic correlation between reversal period and specificspecies containing 161 was PE-161161 (R2 = 088Table 2) PE-161ω5161ω5 has been synthesized and isthe strongest attractant described for M xanthus (Kearnset al 2001a Bonner et al 2005) Whether other PE spe-cies with 161 at the sn-1 position are chemoattractantsand contribute to the observed stimulation is unknown butit is quite evident that the abundance of 161 at the sn-1position is strongly correlated with chemotaxis

Abundance of 161ω5 in analysed strains and communities

The fatty acid 161ω5 is rarely found in cultured bacteriaand generally represents a small proportion of the totalfatty acids (Table 3) There are only a few exceptions Forexample 161ω5 has been suggested as a biomarker forarbuscular mycorrhizal fungi in soil (Olsson et al 1995)Elevated proportions of 161ω5 and a high ratio of 161ω5to 161ω7 (including 161ω6) were found in the phospho-lipid fatty acids (PLFA) of a strain of Flexibacter flexilis

Table 2 Phosphatidylethanolamine (PE) enriched in sn-1 161 stim-ulates chemotaxis

PE type Reversal period sn-1 161 sn-2 161 PE-161161

None 70 plusmn 10 ndash ndash ndashplsB1 160 plusmn 19 234 73 73Dev WT 174 plusmn 52 281 380 155Veg WT 184 plusmn 11 382 289 120esg 473 plusmn 170 716 317 278

R 2 095 009 088

Phosphatidylethanolamine was purified from vegetative DK1622 (VegWT) 24 h developing DK1622 (Dev WT) LS2300 (plsB1) and JD300(esg) cells Relative amounts of 161 at the sn-1 and sn-2 positionswere calculated from the data in Table 1 by summing the percentagesof each molecular PE species containing 161 Where two molecularspecies pairings have the same mz only the most abundant pairingwas considered Phosphatidylethanolamine from these same strainswas used to assess chemotaxis of wild-type DK1622 cells using thestimulation assay In this assay attractants increase the reversalperiod from an unstimulated level of 7 min depending on the abun-dance of attractant molecules in the preparation Phosphatidyletha-nolamine preparations are listed in order of increasing stimulatedreversal period Linear regression was performed between the rever-sal period stimulated by each sample and the amount of 161 at thesn-1 position sn-2 position or the species PE-161161 Regressionanalysis was performed using Microsoft Excel



which contained 51 of its PLFA as 161ω5 with aratio = [161ω5]([161ω7] + [161ω6]) of 96 (Nichols et al1986) This is unusual for the cultured FlavobacteriumCytophaga Another unusual strain from this groupCytophaga hutchinsonii also has a high level of 161ω5Cultured type I methanotrophs like Methylomonas sp 761contain high levels of 161ω5 (16 of its PLFA as 161ω5and a ratio of 094) (Nichols et al 1985) whereas othercultured methanotrophs do not In the environmental data-base compiled by MIDI (Newark DE) 161ω5 was foundin only 46 of 906 bacterial species Of these 46 only threecontained 161ω5 above 4 of the total fatty acidsreleased by alkaline saponification Bradyrhizobiumjaponicum 64 Mycobacterium marinum 80 andEmpedobacter brevis 86 (data kindly supplied by GaryJackoway MIDI Newark DE) In these cases howeverthe 161 fatty acids are mixtures of isomers with differentpoints of unsaturation The ratio of 161ω5 to161ω7 + 161ω6 is 30 for B japonicum 13 forM marinum and 04 for E brevis In a previous study161ω5c in M xanthus constitutes an even larger percent-age of the total fatty acids (24) than in the previousexamples and is in much higher proportion to other 161isomers (a ratio of 800 Table 3) (Ware and Dworkin1973) While in this study the relative abundance of161ω5 was smaller (163 Table 3) the ratio of isomersstill strongly favours the ω5 isomer at 1254 A high ratioof 161ω5 to 161ω7 + 161ω6 in PLFA can also be foundin Syntrophomonas wolfeiDesulfovibrio sp co-culturesgrowing with valerate instead of butyrate (285 Henson

et al 1988) however the total 161ω5 was comparativelylow at 57 161 found in Stigmatella aurantiaca anothermyxobacterium is predominantly in the ω5 form and isrelatively high in abundance (Dickschat et al 2005)

The conclusion that M xanthus and S aurantiaca areunusual in having both a high level of 161ω5 and a highratio of 161ω5 relative to other 161 isomers is strength-ened by examining microbial communities In surface soiland in an oak rhizosphere 161ω5 can be detected but isnot the most abundant 161 isomer (Table 3) A similarresult is obtained with soils enriched with different carbonsources such as methane and propane

Discussion

Myxococcus xanthus vegetative PE contains at least 17different PE species (Table 1) rising to 19 during devel-opment Among all the strains examined 27 different PEspecies were detected The number is surprising as mostProteobacteria have on average five to seven different PEspecies (Cronan and Rock 1996 Rahman et al 2000)Even more unusual is the fact that 12 of the 17 fatty acidsfound at the sn-1 position in M xanthus PE are unsatur-ated (Table 1) which violates the well-established para-digm for Proteobacteria (Cronan and Rock 1996)

In addition to the diversity observed in PE species thereis a large amount of diversity in the fatty acids themselvesas exemplified by the multiple points of unsaturation foundin unsaturated fatty acids 151 fatty acids were repre-sented by ω4 (∆11) and ω10 (∆5) isomers 161 by ω5 (∆11)

Table 3 161ω5 and 161ω7161ω6 content in cultured bacteria and environment samples

Organism 161ω5 161ω7ω6 Ratioa Reference

Bacillus firmus 38 09 42 Jackowayb

Bradyrhizobium japonicum GC subgroup B 64 21 30 JackowayBradyrhizobium japonicum GC subgroup A 18 07 26 JackowayMycobacterium marinum 80 63 13 JackowayRoseomonas cervicalis 35 29 12 JackowayHyphomonas neptunium 13 18 07 JackowayRoseomonas genomospecies 5 12 18 06 JackowayEmpedobacter brevis 85 214 04 JackowayStreptococcus mitis 22 60 04 JackowayExiguobacterium acetylicum 26 73 04 JackowayGeobacter metallireducens 15 407 004 Lovley et al (1993)Caulobacter-like aerobic dye degrader 14 257 005 Govindaswami et al (1993)Desulfomonile tiedjei 21 237 01 Ringelberg et al (1994)Subsurface Sphingomonas 03 21 01 Balkwill et al (1997)Pseudoalteromonis tunicata 04 475 0008 Holmstrom et al (1998)Syntrophomonas wolfei (valerate) 57 02 285 Henson et al (1988)Soil methanotrophs 53 103 05 Nichols et al (1987)Surface soil 17 35 05 Ringelberg et al (1989)Propanotrophs 32 220 01 Ringelberg et al (1989)Oak rhizosphere 93 112 08 Ringelberg et al (1997)Myxococcus xanthus 240 03 800 Ware and Dworkin (1973)Myxococcus xanthus 163 013 1254 This study

a Ratio is determined by dividing the per cent 161ω5 by the per cent 161ω7 (including 161ω6)b Gary Jackoway MIDI Newark DE (pers comm)



and ω11 (∆5) isomers and iso171 by ω5 (∆11) Howeverthe mechanism for generating this diversity is unclear Inthe well-characterized E coli fatty acid biosynthesis path-way unsaturated fatty acids are created by the action ofFabA (Cronan and Rock 1996) FabA isomerizes thetrans-∆2 double bond intermediate (which is usuallyreduced to create a fully saturated acyl chain) to the cis-∆3 double bond which is preserved in further acyl-chainelongation Given that only one point of unsaturation isobserved in E coli FabA must have molecular specificityfor the intermediate to be isomerized If M xanthus pro-duces unsaturated fatty acids using a similar pathway itwould require multiple FabA homologues to introduceunsaturations at different positions on different fatty acids

At least four FabA homologues are found in theM xanthus genome Additionally in this model fatty acidswith even numbers of carbons must have unsaturationswith odd-numbered ω positions (and the opposite case forfatty acids with odd numbers of carbons) In fact thispattern is observed There is however a far simpler expla-nation for the pattern of unsaturations In the case of151ω4 161ω5 and iso171ω5 the point of unsaturationin each fatty acid is conspicuously 11 carbons from the ∆terminus Similarly for 151ω10 and 161ω11 the point ofunsaturation is five carbons from the ∆ terminus There-fore the pattern of unsaturations observed could easilybe explained by the action of a ∆11 desaturase and a ∆5

desaturase and in fact desaturases are found in thegenome This model for desaturation is similar to themethod used by Bacillus (Aguilar et al 1998) and cyano-bacteria (Hongsthong et al 2004) In Bacillus and incyanobacteria the acyl-lipid desaturases introduce dou-ble bonds at specific positions (usually relative to the ∆terminus) in fatty acids that have already been esterifiedto glycerolipids While the E coli model for biosynthesisof unsaturated fatty acids is viable for M xanthus thedesaturase model is the simplest explanation for the pat-tern of unsaturations observed Further testing will beneeded to determine exactly what role each model has incontributing to unsaturated fatty acid diversity in thisorganism

The chemotactic activity of M xanthus PE is associatedwith the fatty acid 161ω5 This compound is rare in bac-teria and in natural samples It may be a chemoattractantfor the myxobacteria as M xanthus and S aurantiacaboth contain high levels of the ω5 isomer High levels of161 are also present in other myxobacteria Nannocystisexedens and Sorangium cellulosum although the iso-mer(s) is unknown (Iizuka et al 2003) Synthetic PE-161ω5161ω5 is a potent chemoattractant for M xanthus(Kearns et al 2001a) and here we show that PE-161161 is one of the major PE constituents However thereare several other PE species containing 161 at either snposition Myxococcus xanthus may respond to PE with

161 at the sn-1 position the sn-2 position or a particular161-containing species In addition PE may be hydroly-sed by a phospholipase and the free fatty acid detectedRegression analysis with chemotactic stimulation by PEof varying composition shows strong linear relationshipswith sn-1 161 and with the species PE-161161 How-ever the alternate hypothesis of position-specific cleav-age by phospholipase A1 cannot be ruled out Taken asa whole the results argue that the wide diversity in PEspecies by M xanthus is due to the fact that some speciesplay a role in cellndashcell signalling It appears thatM xanthus has deployed a rare unsaturated fatty acid(161ω5) at an unusual position in PE (sn-1) to obtain aspecies or group-specific signalling molecule

A key to understanding the evolution of this signallingsystem is the enzymology leading to the deployment ofunsaturated fatty acids at the sn-1 position One case ofunsaturated fatty acid localization at the sn-1 position hasbeen examined previously phospholipids of Clostridiumbutyricum have both saturated and unsaturated fatty acidsat the sn-1 position while the sn-2 position has only sat-urated fatty acids (Heath et al 1997) The sn-1 fatty acidbias in C butyricum involves a unique PlsC-like sn-1 acyl-transferase PlsD However there is no evidence thatM xanthus uses a similar strategy as PlsB1 and all fivePlsC homologues found in the M xanthus genome closelyresemble their E coli counterparts (Figs 2 and 4) In addi-tion M xanthus PE does not have a reversed fatty acidbias like C butyricum (variable saturation at the sn-1 posi-tion and predominantly saturated at the sn-2) but ratheran aberrant one (predominantly unsaturated at the sn-1and variable at the sn-2) Therefore it appears thatM xanthus PE has properties that differ from not only theestablished lipid paradigm but also other unusual cases

Myxococcus xanthus is unusual in the large number ofputative acyltransferase genes present in the genome Wesuspected that one of the two plsB genes might be thesource of the unsaturated fatty acids at the sn-1 positionHowever deletion of plsB1 or plsB2 did not eliminateunsaturated fatty acids at the sn-1 position suggestingthat both acyltransferases contribute unsaturated fattyacids The N-terminal half of PlsB2 is similar to severaleukaryotic FARs These enzymes reduce fatty acyl-CoAsubstrates to free fatty alcohols (Kolattukudy 1970) eitherfor synthesis of waxes a reaction where a fatty alcohol isesterified to another fatty acid or for ether-linked phos-pholipids synthesis (for review see Nagan and Zoeller2001) Prokaryotic fatty acid reduction utilizes twoenzymes one to reduce the fatty acyl-CoA to the fattyaldehyde and a second to reduce the fatty aldehyde tothe fatty alcohol (Reiser and Somerville 1997) EukaryoticFARs use a single enzyme with a NAD(P)H-binding motifto reduce fatty acyl-CoA to the fatty alcohol in a singleenzyme (Kolattukudy 1970) In both eukaryotic and



prokaryotic systems the reductases are separateenzymes from the acyltransferases Therefore PlsB2 isunique in that the FAR is more similar to eukaryotic reduc-tases and is uniquely coupled to an acyltransferasedomain It appears unlikely that PlsB2 functions solely inether lipid production although ether lipids are detectedin myxobacteria (Kleinig 1972 Caillon et al 1983)because the plsB1 mutation would be lethal The plsB1gene was deleted without complication and while there isa loss of some of the higher-molecular-weight species themajority of the species and their abundances remain sim-ilar to the wild type (Table 1) Therefore it seems likelythat the PlsB2 acyltransferase domain functions in PEbiosynthesis

The plsB2 gene is located next to a putative alkyl-dihydroxyacetone phosphate synthase gene the stopcodon of the synthase overlaps the start codon of plsB2by 1 bp suggesting translational coupling This enzymecatalyses the exchange of a fatty alcohol for a fatty acidat the sn-1 position of 1-acyl-dihydroxyacetone phos-phate creating an intermediate in eukaryotic ether lipidbiosynthesis The proximity of these genes combined withthe nature of the FAR domain would suggest that PlsB2functions in ether lipid biosynthesis Therefore it is possi-ble that PlsB2 has a dual role in both PE biosynthesis andether lipid biosynthesis

Lipid signalling may be involved in maintaining culturespecificity during fruiting body formation in the midst of acomplex microbial community The output from lipid sig-nalling assayed in these experiments is an alteration ofmotility The chemotaxis signal passes through the difchemosensory pathway (Kearns et al 2000 Bonneret al 2005) which also regulates extracellular matrix pro-duction another component of surface-dwelling organ-isms that is essential for fruiting body formation (Yanget al 2000 Bellenger et al 2002 Black and Yang 2004)The coupling of matrix production and lipid sensing isunusual and may reflect the fact that extracellular matrixis essential for a response to some lipid attractants

Myxococcus xanthus responds chemotactically toanother lipid PE-181ω9181ω9 The fatty acid 181ω9was not observed in the cells of M xanthus but is foundin many other proteobacteria such as E coli (Cronan andRock 1996) which M xanthus preys on in the soil(Yamanaka et al 1987 Shimkets et al 2006) It is spec-ulated that 181 chemotaxis is used for finding prey Cor-relatively neither the conditions nor part of the machineryfor the 181 response has much in common with the 161response Whereas 161 stimulation requires starvationconditions and production of the extracellular matrix (fac-tors for the developmental cycle) 181 can stimulatechemotaxis under nutrient-rich conditions and indepen-dently of the presence of the extracellular matrix (Kearnset al 2000) It was also recently shown that the 181

signalling pathway is partially independent of the Difchemosensory pathway used for 161 signalling 181requires the histidine kinase (DifE) and response regula-tor (DifD) but not the methyl-accepting chemotaxis protein(DifA) or the coupling protein (DifC) to stimulate chemot-axis (Bonner et al 2005) Therefore it appears thatM xanthus has two largely independent lipid sensorysystems

Pseudomonas aeruginosa also travels up gradients ofPE-181ω9181ω9 and PE-120120 (Kearns et al2001b Barker et al 2004) Interestingly the response toPE-181ω9181ω9 is dependent on the activity of a phos-pholipase C (PlcB) while the response to PE-120120 isnot (Barker et al 2004) indicating that P aeruginosa likeM xanthus has two lipid-signalling pathways IndeedP aeruginosa shares several characteristics in commonwith M xanthus They are both soil-dwelling microbesutilize surface motility produce an extracellular matrix anddisplay forms of multicellular behaviour (biofilm formationin the case of P aeruginosa) Lipids may be ideal signalsfor surface-translocating organisms surface motility isprohibitively slow compared with flagellar motility of plank-tonic cells Whereas soluble chemical signals could freelydiffuse and therefore collapse gradients before surface-motile organisms have a chance to respond lipids willestablish more stable gradients due to the insoluble natureof the molecules Other microorganisms may provide abroad variety of unique fatty acids and lipid species (sig-nature lipid biomarkers) and may also utilize lipid-basedsignalling

Experimental procedures

Strains and growth conditions

Myxococcus xanthus wild-type strain DK1622 (Kaiser 1979)and mutant strains LS2300 (plsB1 MXAN3288) LS2301(plsB2 MXAN1675) JD300 (Downard and Toal 1995)LS2305 (plsC2 MXAN3969) LS2306 (plsC4 MXAN0955)and LS2307 (plsC5 MXAN0147) were grown at 32degC in CYE[10 Bacto Casitone 05 Difco yeast extract 10 mM 3-[N-morpholino]propanesulfonic acid (MOPS) pH 76 and 01MgSO4] broth with vigorous shaking (Campos et al 1978)Cultures were grown on plates containing CYE with 15Difco agar To induce development 75 ml of 5 times 109 cellsmlminus1 were plated on TPM agar [10 mM Tris (hydroxymethyl)aminomethane HCl 8 mM MgSO4 1 mM KHPO4-KH2PO415 Difco agar pH 76] in a 33 times 22 cm tray and incubatedat 32degC for 24 h

Construction of plsB mutants

Preliminary sequence data were obtained from The Institutefor Genomic Research through the website at httpwwwtigrorg Two M xanthus plsB homologues were identi-fied using the E coli PlsB protein sequence (P0A7A7) tosearch the M xanthus genome (available at http



cmrtigrorgtigr-scriptsEMRCmrHomePagecgi) and analy-sed using the CLUSTALW (httpaligngenomejp) and BOX-

SHADE (httpbiowebpasteurfrseqanalinterfacesboxshadehtml) programs In-frame deletion mutants of plsB1 and plsB2were constructed as described previously (Julien et al2000) Wild-type M xanthus chromosomal DNA was purified(Easy-DNA kit Invitrogen) and approximately 750 bp ofsequence upstream of plsB1 was amplified using primersINFRB1P1F and INFRB1P1R which add BamHI and XbaIrestriction sites respectively (Table 4) Approximately 750 bpof downstream sequence was amplified using primersINFRB1P2F and INFRB1P2R which add XbaI and HindIIIsites respectively Polymerase chain reaction (PCR) productswere separated on 10 agarose and the desired fragmentswere extracted using the UltraClean 15 DNA PurificationKit (Mo Bio Laboratories) Extracted fragments were usedin a third PCR reaction with primers INFRB1P1F andINFRB1P2R and the products were separated on 10 aga-rose A band of approximately 15 kb was extracted andcloned into pCR21-TOPO (TOPO TA Cloning Kit Invitrogen)to create plasmid pINFR6 The integrity of the construct wasverified by DNA sequencing Plasmids pBJ113 (Julien et al2000) and pPDC4 were digested with BamHI and HindIII andgel purified The sim15 kb insert from pPDC4 was ligated withpBJ113 The resultant plasmid (pPDC5) was electroporatedin M xanthus DK1622 and transformants were selected onCYE agar plates containing 50 microg mlminus1 kanamycin Transfor-mants were then grown in CYE broth in the absence ofselection and plated on CYE agar plates supplemented with10 galactose Galactose-resistant colonies were screenedfor kanamycin sensitivity and analysed by PCR reactions withprimers INFRB1P1F and INFRB1P2R Polymerase chainreaction fragments of sim15 kb were verified as deletions bydigestion of the PCR fragments with XbaI into approximatelyequal size fragments A colony with appropriate amplificationsize and digestion pattern was kept as strain LS2300

Similarly the plsB2 mutant (LS2301) was generated usingprimers INFRB2P1F and INFRB2P1R in the first PCR reac-

tion INFRB2P2F and INFRB2P2R in the second reactionINFRB2P1F and INFRB2P2R in the third reaction andINB2DIAGF and INB2DIAGR in the verification reactionPolymerase chain reaction fragments were cloned digestedand ligated exactly as described above

Construction of plsC mutants

Five M xanthus plsC homologues were identified by genomesearch using the E coli PlsC protein sequence (P26647)They were named plsC1 plsC2 plsC3 plsC4 and plsC5Campbell insertion mutants were created by amplifying inter-nal fragments of approximately 450 bp These were amplifiedusing primers MYXplsC2F and MYXplsC2R for the plsC2gene MYXplsC4F and MYXplsC4R for plsC4 andMYXplsC5F and MYXplsC5R for plsC5 The PCR productswere separated on a 10 agarose gel purified and clonedinto pCR21-TOPO to create plasmids pPDC1 pPDC2 andpPDC3 respectively Each plasmid was electroporated inDK1622 and grown on CYE + 50 microg mlminus1 kanamycin Trans-formants were examined by Southern hybridization andprobed with labelled PCR fragments generated using thesame primers

Characterization of mutants

Mutants were examined for motility fruiting body formationand sporulation For motility the edges of colonies on CYEagar were viewed under 400times magnification on a Leitz Labor-lux D phase-contrast microscope To analyse fruiting bodyformation and sporulation 15 times 109 cells were plated onTPM agar plates and incubated at 32degC with observation ona Wild Heerbrugg dissecting microscope over 5 days Fruitingbodies were removed with a sterile razor blade and resus-pended in 10 ml of TPM buffer The fruiting bodies were thenincubated at 55degC for 2 h and sonicated at a 60 duty cyclefor 2 times 15 s on an Ultrasonic Processor Sonicator (Heat Sys-temsndashUltrasonics) Refractile myxospores were quantifiedusing a PetroffndashHauser counting chamber Spores werediluted and plated on CYE plates to enumerate viable spores

Fatty acid methyl ester analysis

Lyophilized cells of M xanthus were extracted and deriva-tized according to the whole-cell hydrolysate procedure of theMicrobial Identification System (MIDI Newark DE) (Haumlrtiget al 2005) Samples were resolved in hexane containingthe FAME 210 as an internal standard (59 pmol microlminus1) foranalysis by GC-MS (Geyer et al 2005)

Dimethyldisulfide derivatization

Fatty acid methyl esters were derivatized using a modifiedprocedure of Leonhardt and deVilbiss (1985) Samples forderivatization with dimethyldisulfide were dissolved in 10 mlof diethyletherhexane (5050) Approximately 40 mg ofiodine and 300 microl of DMDS were added and the mix wasincubated at 37degC for 30 min The reaction was completed byadding sodium thiosulfate (10 solution) dropwise until the

Table 4 Primers used in construction and examination of plsB andplsC mutants

Primername Sequence (5primerarr3prime)

INFRB1P1F GGA TCC GCA CCG TCC ACG TCG CGT TCGINFRB1P1R CAT CTA GAC ATG GGG CCG AAT TCG TCC TTC

AGCINFRB1P2F ATG TCT AGA TGA GCA GCC CTC CCT GGC CCCINFRB1P2R AAG CTT GGC GCT GAA CAC CAC GGC GGA GINFRB2P1F GGA TCC GGT GGT GGG CAT GGT CGA CGT GINFRB2P1R TCA TCT AGA CAT GCC CTC GTT CAC CAC GCG

CINFRB2P2F GTC TAG ATG AAG ACA CTC GTG ACG GGA GCCINFRB2P2R AAG CTT CGA GCA CCA CGG GGT CCG GCA GCINB2DIAGF CAT GTT CGT CGG CCG CAA GGA CINB2DIAGR CGA GCT GTC GAT GTT GAA GCG CMYXplsC2F AAG CAA CCA CGA GTC CAAMYXplsC2R TTG GTG GAG ATG GGC GTGMYXplsC4F TCA TTG GTC TGT CGT TGGMYXplsC4R GTC TGG ATG CAG CAG CCCMYXplsC5F CCG TGC TCG TGT CCA ACCMYXplsC5R GAG GAC CAC GGG GAT GAC



mixtures remained colourless The upper organic phase wastransferred dried over sodium sulfate evaporated to drynessand resuspended in 500 microl of hexane An aliquot of 10 microlwas analysed with GC-MS using the same temperature pro-gramme as for FAME analysis Results showed a nearlycomplete derivatization Details to interpretation of massspectra from DMDS derivatives are outlined elsewhere(Christie 2006)

Phosphatidylethanolamine purification

Phosphatidylethanolamine was purified using a modifiedmethod of Bligh and Dyer (1959) Cells were grown in CYEto a density of 75 times 108 cells mlminus1 (approximately mid-logphase) and pelleted by centrifugation at 12 100 g for 10 minTo 08 g (wet cell weight) of M xanthus cells was added75 ml of methanolchloroform (21) The mixture was vigor-ously shaken for 1 h Insoluble material was pelleted by cen-trifugation at 12 100 g for 5 min and the supernatant wascollected The pellet was re-extracted with 95 ml of metha-nolchloroformwater (2108) centrifuged again the super-natant collected and combined with the supernatant from thefirst extraction Chloroform (50 ml) and 50 ml of water wereadded to the combined supernatants and mixed by vortexingThe phases were separated by centrifugation at 3020 g for30 s The chloroform layer was air-dried resuspended in075 ml of methanolchloroform (21) and spotted onto silicagel 60 TLC plates (EM Science) The plates were developedfor 3 h in 100 ml of chloroformacetonemethanolacetic acidwater (104221) and dried Two-centimetre horizontalswaths of silica were scraped from the plate leaving only asmall vertical section of silica for staining Lipid was extractedfrom the scraped silica with methanolchloroform (21) fol-lowed by centrifugation at 15 400 g to pellet the silica andthe extract was dried The extract was re-extracted 3 times 05 mlmethanolchloroform (21) centrifuged twice at 15 400 g for5 min to remove residual silica and the final extract was driedfor analysis The remainder of the plate was then stained with05 ninhydrin (3 acetic acid in 1-butanol) and incubatedat 100degC for 5 min Phosphatidylethanolamine resolved withan Rf value of approximately 055

Phosphatidylethanolamine analysis

The polar lipid extracts were analysed by LC-ESI-MS-MSperformed on an Applied BiosystemsMDS SCIEX 365 tan-dem MS system with an electrospray interface (LC-ESI-MS-MS) The LC system (Agilent 1100 LC) used for introductionof the sample into the mass spectrometer consisted of an in-line vacuum degasser a quaternary solvent pump anautosampler a column oven and a diode-array detector (Agi-lent 1100 Agilent Palo Alto) The autosampler was equippedwith a 100 microl sample loop Details of phospholipid analysesat this system were described elsewhere (Lytle et al 2000White et al 2002) A Thermo Dash-8 20 times 21 mm HPLCcolumn was used for the reversed phase chromatographicseparation of phospholipids A gradient solvent system com-posed of solvent A (water) and solvent B (methanolacetoni-trile 9010 + 0002 piperidine) was used with a flow rate of

100 microl minminus1 At the beginning of the gradient the mobilephase was 80 of B for 05 min Solvent B was increased to100 at 15 min The mobile phase was then held isocraticallyfor 10 min Solvent B was decreased within 05 min to thestarting value and the column equilibrated for 5 min Thecolumn oven was held at 40degC

The mass spectrometer was operated in the positive ion-ization mode to uniquely detect the PE molecules and deter-mine their mass (mz) by a neutral-loss scan for the PE headgroup (mz 141) The acyl-chain composition within thedetected PEs was determined by a product ion scan (MS2)in the negative ionization mode Overviews of methods forESI-MS-MS of phospholipids were recently given (Sturt et al2004 Taguchi et al 2005)

With a palmitoyl-oleoyl PE standard (PE-160181 AvantiPolar Lipids) infused into the ESI source the instrumentsource parameters (eg curtain gas nebulizer gas sourceheater) were optimized as follows The source temperaturewas 450degC For positive ionization the ion transfer voltage(IS) between electrospray needle and vacuum interface wasset to 5000 V or in the negative mode to minus4400 V with theskimmer held at ground potential The declustering potential(DP voltage between orifice plate and ground) the focusingpotential (FP voltage between skimmer and ring potential)and the collision energy (CE) for CID were optimized at 30 V290 V and 30 V in the positive mode and at minus40 V minus280 Vand minus45 V in the negative mode respectively Nitrogen wasused as collision gas The nitrogen curtain gas and the ion-source gases 1 (nebulizer gas) and 2 (turbo gas at theheater) were set to pressures of approximately 20 30 and60 psi (14 20 41 bar) respectively The mass spectrometerwas tuned from mz 5 to 2000 amu according to the protocolprovided by the manufacturer

The fragmentation of the [M+H]+ ions of PEs is dominatedby the mz [Mminus141]+ ion which unequivocally characterizesmolecules of this lipid class This decomposition releasingthe PE head group as a neutral molecule[H2PO4CH2CH2NH2] can be assessed by tandem mass spec-trometer as loss of mz 141 The obtained masses uniquelycharacterized the mz [M+H]+ of PEs

The relative concentration of PE within a sample wasassessed based on the ion currents of the individual phos-pholipids set in relation to the sum of all detectable PE spe-cies in the neutral-loss scan To exclude molecules with lowabundances and with background signals a cut-off of 15signal intensity was applied Absolute concentrations of PEspecies were calculated based on a calibration curveobtained with a PE 160180 standard (Avanti Polar Lipids)

In the negative mode subsequent collision-induced frag-mentation of the proposed PE ion mz [MndashH]ndash revealed theacyl-chain composition within the individual PE The positionon the glycerol backbone was assigned based on the signalintensities of the abundant carboxylate anion as the sn-2-derived anion is usually preferred at a ratio greater than 2 to1 (Pulfer and Murphy 2003) The assignment as PEs couldbe additionally verified based on the diagnostic ion of the PEheadgroup at mz 140 [HPO4CH2CH2NH2]macr which was visibleat a low abundance

Fragmentation scans in the negative mode indicated thatthe samples still contained phosphatidylglycerol (PG)characterized by the diagnostic ions mz 153



[CH2C(OH)CH2HPO4]macr and mz 171[HPO4CH2CH(OH)CH2OH]macr However the PG did not com-promise either the quantification of PE or analysis of fattyacid composition

Phospholipids were designated as follows PL-C1d1C2d2 (eg PE-160181) where C1 and C2 are the numbersof carbon atoms in the fatty acyl chains on the sn-1 and sn-2 positions respectively d1 and d2 are the numbers of dou-ble bonds of the sn-1 and sn-2 fatty acyl chains respectivelyand PL is the abbreviation for phospholipids class

Stimulation assay

Chemotaxis was quantified using the stimulation assay(Kearns and Shimkets 1998) This assay measures theperiod of time between cell reversals (reversal period whichis approximately 7 min for unstimulated cells) (Blackhart andZusman 1985) Attractants increase the reversal period(Kearns and Shimkets 1998) The stimulation assay wasperformed as previously described with minor modifications(Kearns and Shimkets 1998) Phosphatidylethanolaminepurified from DK1622 LS2300 or JD300 was solubilized inchloroform to 05 mg mlminus1 A TPM agar plate was dried for10 min at 37degC Then 4 microl of PE solution was applied to thesurface of the agar plate and the solvent was evaporated byincubation at 37degC for 15 min Five microlitres of DK1622cells diluted to 5 times 107 cells mlminus1 in MOPS buffer (10 mMMOPS 8 mM MgSO4 pH 76) were applied to the PE spotand incubated at 32degC for 15 min Cells were then observedat 25degC with a Leitz Laborlux D microscope for 45 min at640times magnification Stop-motion digital movies were pro-duced with a microscope-mounted Sony Power HAD 3CCDcolour video camera and a Macintosh 9500 with Adobe PRE-MIERE software (Adobe Systems Mountain View CA framecapture rate 12 frames per minute) The reversal period wasmanually enumerated for 15ndash20 isolated cells per movie Themean and standard deviation were calculated from threemovies

Acknowledgements

The authors wish to thank Pamela Bonner for her insightfuldiscussions and critical analysis Roy Welch for supplyingannotated versions of the genome and Gary Jackoway forsupplying fatty acid data Sequencing of M xanthus DK1622was accomplished with support from the National ScienceFoundation This article is based on work supported by theNational Science Foundation under Grant No 0343874 toLJS and Grant DE-FG02-04ER63939 Department ofEnergy Office of Science to DCW as well as support ofthe UFZ Leipzig-Halle to RG

References

Aarts MG Hodge R Kalantidis K Florack D WilsonZA Mulligan BJ et al (1997) The Arabidopsis MALESTERILITY 2 protein shares similarity with reductases inelongationcondensation complexes Plant J 12 615ndash623

Aguilar PS Cronan JE Jr and de Mendoza D (1998)A Bacillus subtilis gene induced by cold shock encodes amembrane phospholipid desaturase J Bacteriol 1802194ndash2200

Balkwill DL Drake GR Reeves RH Fredrickson JKWhite DC Ringelberg DB et al (1997) Taxonomicstudy of aromatic-degrading bacteria from deep-terrestrial-subsurface sediments and description of Sphingomonasaromaticivorans sp nov Sphingomonas subterranea spnov and Sphingomonas stygia sp nov Int J Syst Bacteriol47 191ndash201

Barker AP Vasil AI Filloux A Ball G Wilderman PJand Vasil ML (2004) A novel extracellular phospholipaseC of Pseudomonas aeruginosa is required for phospholipidchemotaxis Mol Microbiol 53 1089ndash1098

Bellenger K Ma X Shi W and Yang Z (2002) A CheWhomologue is required for Myxococcus xanthus fruitingbody development social gliding motility and fibril biogen-esis J Bacteriol 184 5654ndash5660

Black WP and Yang Z (2004) Myxococcus xanthuschemotaxis homologs DifD and DifG negatively regulatefibril polysaccharide production J Bacteriol 186 1001ndash1008

Blackhart BD and Zusman DR (1985) lsquoFrizzyrsquo genes ofMyxococcus xanthus are involved in control of frequencyof reversal of gliding motility Proc Natl Acad Sci USA 828767ndash8770

Bligh EG and Dyer WJ (1959) A rapid method of totallipid extraction and purification Can J Biochem Physiol 37911ndash917

Bonner PJ Xu Q Black WP Li Z Yang Z andShimkets LJ (2005) The Dif chemosensory pathway isdirectly involved in phosphatidylethanolamine sensorytransduction in Myxococcus xanthus Mol Microbiol 571499ndash1508

Caillon E Lubochinsky B and Rigomier D (1983) Occur-rence of dialkyl ether phospholipids in Stigmatella auranti-aca DW4 J Bacteriol 153 1348ndash1351

Campos JM Geisselsoder J and Zusman DR (1978)Isolation of bacteriophage MX4 a generalized transducingphage for Myxococcus xanthus J Mol Biol 119 167ndash178

Cheng JB and Russell DW (2004) Mammalian wax bio-synthesis I Identification of two fatty acyl-Coenzyme Areductases with different substrate specificities and tissuedistributions J Biol Chem 279 37789ndash37797

Christie W (2006) The lipid library mass spectra of methy-lesters of fatty acids In Part 7 Further Derivatizations[WWW document] URL httpwwwlipidlibrarycoukms04filepdf last updated 02 February 2002

Cronan JE and Rock CO (1996) Biosynthesis of mem-brane lipids In Escherichia coli and Salmonella NeidhardtFC Curtiss R Ingram JL Lin ECC Low KB andMagasanik B (eds) Washington DC USA AmericanSociety for Microbiology Press pp 612ndash636

Dickschat JS Bode HB Kroppenstedt RM Muller Rand Schulz S (2005) Biosynthesis of iso-fatty acids inmyxobacteria Org Biomol Chem 3 2824ndash2831

Downard J and Toal D (1995) Branched-chain fatty acidsthe case for a novel form of cellndashcell signalling duringMyxococcus xanthus development Mol Microbiol 16 171ndash175



Ekroos K Chernushevich IV Simons K andShevchenko A (2002) Quantitative profiling of phospho-lipids by multiple precursor ion scanning on a hybrid qua-drupole time-of-flight mass spectrometer Anal Chem 74941ndash949

Fiegna F and Velicer GJ (2005) Exploitative and hierar-chical antagonism in a cooperative bacterium PLoS Biol3 e370

Geyer R Peacock AD Miltner A Richnow HH WhiteDC Sublette KL and Kastner M (2005) In situ assess-ment of biodegradation potential using biotraps amendedwith 13C-labeled benzene or toluene Environ Sci Technol39 4983ndash4989

Gil R Silva FJ Pereto J and Moya A (2004) Determi-nation of the core of a minimal bacterial gene set MicrobiolMol Biol Rev 68 518ndash537

Govindaswami M Schmidt TM White DC and LoperJC (1993) Phylogenetic analysis of a bacterial aerobicdegrader of azo dyes J Bacteriol 175 6062ndash6066

Haumlrtig C Loffhagen N and Harms H (2005) Formationof trans fatty acids is not involved in growth-linked mem-brane adaptation of Pseudomonas putida Appl EnvironMicrobiol 71 1915ndash1922

Heath RJ and Rock CO (1998) A conserved histidine isessential for glycerolipid acyltransferase catalysis J Bac-teriol 180 1425ndash1430

Heath RJ Goldfine H and Rock CO (1997) A gene(plsD) from Clostridium butyricum that functionally substi-tutes for the sn-glycerol-3-phosphate acyltransferase gene(plsB) of Escherichia coli J Bacteriol 179 7257ndash7263

Henson JM McInerney MJ Beaty PS Nickels J andWhite DC (1988) Phospholipid fatty acid composition ofthe syntrophic anaerobic bacterium Syntrophomonaswolfei Appl Environ Microbiol 54 1570ndash1574

Holmstrom C James S Neilan BA White DC andKjelleberg S (1998) Pseudoalteromonas tunicata sp nova bacterium that produces antifouling agents Int J SystBacteriol 48 1205ndash1212

Hongsthong A Subudhi S Sirijuntarat M and Chee-vadhanarak S (2004) Mutation study of conserved aminoacid residues of Spirulina ∆6-acyl-lipid desaturase showinginvolvement of histidine 313 in the regioselectivity of theenzyme Appl Microbiol Biotechnol 66 74ndash84

Iizuka T Jojima Y Fudou R Hiraishi A Ahn JW andYamanaka S (2003) Plesiocystis pacifica gen nov spnov a marine myxobacterium that contains dihydroge-nated menaquinone isolated from the Pacific coasts ofJapan Int J Syst Evol Microbiol 53 189ndash195

Julien B Kaiser AD and Garza A (2000) Spatial controlof cell differentiation in Myxococcus xanthus Proc NatlAcad Sci USA 97 9098ndash9103

Kaiser D (1979) Social gliding is correlated with the pres-ence of pili in Myxococcus xanthus Proc Natl Acad SciUSA 76 5952ndash5956

Kearns DB and Shimkets LJ (1998) Chemotaxis in agliding bacterium Proc Natl Acad Sci USA 95 11957ndash11962

Kearns DB Campbell BD and Shimkets LJ (2000)Myxococcus xanthus fibril appendages are essential forexcitation by a phospholipid attractant Proc Natl Acad SciUSA 97 11505ndash11510

Kearns DB Venot A Bonner PJ Stevens B BoonsG-J and Shimkets LJ (2001a) Identification of a devel-opmental chemoattractant in Myxococcus xanthus throughmetabolic engineering Proc Natl Acad Sci USA 9813990ndash13994

Kearns DB Robinson J and Shimkets LJ (2001b)Pseudomonas aeruginosa exhibits directed twitching motil-ity up phosphatidylethanolamine gradients J Bacteriol183 763ndash767

Kleinig H (1972) Membranes from Myxococcus fulvus (Myx-obacterales) containing carotenoid glucosides I Isolationand composition Biochim Biophys Acta 274 489ndash498

Kolattukudy PE (1970) Reduction of fatty acids to alcoholsby cell-free preparations of Euglena gracilis Biochemistry9 1095ndash1102

Leonhardt BA and deVilbiss ED (1985) Separation anddouble-bond determination on nanogram quantities of ali-phatic monounsaturated alcohols aldehydes and carbox-ylic acid methyl esters J Chromatogr 322 484ndash490

Lewin TM Wang P and Coleman RA (1999) Analysisof amino acid motifs diagnostic for the sn-glycerol-3-phos-phate acyltransferase reaction Biochemistry 38 5764ndash5771

Li D Yu L Wu H Shan Y Guo J Dang Y et al (2003)Cloning and identification of the human LPAAT-zeta genea novel member of the lysophosphatidic acid acyltrans-ferase family J Hum Genet 48 438ndash442

Lovley DR Giovannoni SJ White DC Champine JEPhillips EJP Gorby YA and Goodwin S (1993) Geo-bacter metallireducens gen nov sp nov a microorganismcapable of coupling the complete oxidation of organic com-pounds to the reduction of iron and other metals ArchMicrobiol 159 336ndash344

Lytle CA Gan YD and White DC (2000) Electrosprayionizationmass spectrometry compatible reversed-phaseseparation of phospholipids piperidine as a post columnmodifier for negative ion detection J Microbiol Methods 41227ndash234

Mahmud T Bode HB Silakowski B Kroppenstedt RMXu M Nordhoff S et al (2002) A novel biosyntheticpathway providing precursors for fatty acid biosynthesisand secondary metabolite formation in myxobacteria JBiol Chem 277 32768ndash32774

Metz JG Pollard MR Anderson L Hayes TR andLassner MW (2000) Purification of a jojoba embryofatty acyl-coenzyme A reductase and expression of itscDNA in high erucic acid rapeseed Plant Physiol 122635ndash644

Moto K Yoshiga T Yamamoto M Takahashi S OkanoK Ando T et al (2003) Pheromone gland-specific fatty-acyl reductase of the silkmoth Bombyx mori Proc NatlAcad Sci USA 100 9156ndash9161

Nagan N and Zoeller RA (2001) Plasmalogens biosyn-thesis and functions Prog Lipid Res 40 199ndash229

Nichols PD Smith GA Antworth CP Hanson RS andWhite DC (1985) Phospholipid and lipopolysaccharidenormal and hydroxy fatty acids as potential signatures formethane-oxidizing bacteria FEMS Microbiol Ecol 31 327ndash335

Nichols P Stulp BK Jones JG and White DC (1986)Comparison of fatty acid content and DNA homology of the



filamentous gliding bacteria Vitreoscilla Flexibacter Fili-bacter Arch Microbiol 146 1ndash6

Nichols PD Henson JM Antworth CP Parsons JWilson JT and White DC (1987) Detection of a micro-bial consortium including type-II methanotrophs by use ofphospholipid fatty-acids in an aerobic halogenated hydro-carbon-degrading soil column enriched with natural-gasEnviron Toxicol Chem 6 89ndash97

Olsson PA Baath E Jakobsen I and Soderstrom B(1995) The use of phospholipid and neutral lipid fatty acidsto estimate biomass of arbuscular mycorrhizal fungi in soilMycol Res 99 623ndash629

Pulfer M and Murphy RC (2003) Electrospray mass spec-trometry of phospholipids Mass Spectrom Rev 22 332ndash364

Rahman MM Kolli VS Kahler CM Shih G StephensDS and Carlson RW (2000) The membrane phospho-lipids of Neisseria meningitidis and Neisseria gonorrhoeaeas characterized by fast atom bombardment mass spec-trometry Microbiology 146 1901ndash1911

Reiser S and Somerville C (1997) Isolation of mutants ofAcinetobacter calcoaceticus deficient in wax ester synthe-sis and complementation of one mutation with a geneencoding a fatty acyl coenzyme A reductase J Bacteriol179 2969ndash2975

Ringelberg DB Davis JD Smith GA Pfiffner SMNichols PD Nickels JS et al (1989) Validation of sig-nature polarlipid fatty acid biomarkers for alkane-utilizingbacteria in soils and subsurface aquifer materials FEMSMicrobiol Ecol 62 39ndash50

Ringelberg DB Townsend GT Deweerd KA SuflitaJM and White DC (1994) Detection of the anaerobicdechlorinating microorganism Desulfomonile tiedjei in envi-ronmental matrices by its signature lipopolysaccharidebranched-long-chain hydroxy fatty acids FEMS MicrobiolEcol 14 9ndash18

Ringelberg DB Stair JO Almeida J Norby RJO NeillEG and White DC (1997) Consequences of risingatmospheric carbon dioxide levels for the below-groundmicrobiota associated with white oak J Environ Qual 26495ndash503

Shih GC Kahler CM Swartley JS Rahman MM

Coleman J Carlson RW and Stephens DS (1999)Multiple lysophosphatidic acid acyltransferases in Neis-seria meningitidis Mol Microbiol 32 942ndash952

Shimkets LJ Dworkin M and Reichenbach H (2006) Themyxobacteria In The Prokaryotes Dworkin M (ed) NewYork USA Springer-Verlag

Sturt HF Summons RE Smith K Elvert M and Hin-richs KU (2004) Intact polar membrane lipids in prokary-otes and sediments deciphered by high-performance liquidchromatographyelectrospray ionization multistage massspectrometry ndash new biomarkers for biogeochemistry andmicrobial ecology Rapid Commun Mass Spectrom 18617ndash628

Taguchi R Houjou T Nakanishi H Yamazaki T IshidaM Imagawa M and Shimizu T (2005) Focused lipidom-ics by tandem mass spectrometry J Chromatogr B AnalTechnol Biomed Life Sci 823 26ndash36

Toal DR Clifton SW Roe BA and Downard J (1995)The esg locus of Myxococcus xanthus encodes the E1αand E1β subunits of a branched-chain keto acid dehydro-genase Mol Microbiol 16 177ndash189

Ware JC and Dworkin M (1973) Fatty acids of Myxococ-cus xanthus J Bacteriol 115 253ndash261

Weier D Muller C Gaspers C and Frentzen M (2005)Characterisation of acyltransferases from Synechocystissp PCC6803 Biochem Biophys Res Commun 334 1127ndash1134

White DC Lytle CA Gan YD Piceno YM WimpeeMH Peacock AD and Smith CA (2002) Flash detec-tionidentification of pathogens bacterial spores and biot-errorism agent biomarkers from clinical and environmentalmatrices J Microbiol Methods 48 139ndash147

Yamanaka S Kawaguchi A and Komagata K (1987) Iso-lation and identification of myxobacteria from soils andplant materials with special reference to DNA base com-position quinone system and cellular fatty acid composi-tion and with a description of a new species Myxococcusflavescens J Gen Appl Microbiol 33 247ndash265

Yang Z Ma X Tong L Kaplan HB Shimkets LJ andShi W (2000) Myxococcus xanthus dif genes are requiredfor biogenesis of cell surface fibrils essential for socialgliding motility J Bacteriol 182 5793ndash5798

1936

P D Curtis R Geyer D C White and L J Shimkets



8

1935ndash1949

fruiting body formation and sporulation (Toal

et al

1995) These results argue that a balance of fatty acidsis necessary to maintain the complex life cycle of theorganism

Phosphatidylethanolamine is usually the most abun-dant phospholipid in Proteobacteria membranes Its bio-synthesis starts with the addition of an acyl chain (fattyacid) at the

sn

-1 position of glycerol-3-phosphate by glyc-erol-3-phosphate acyltransferase PlsB (Cronan and Rock1996) The second acyl chain is added at the

sn

-2 positionby PlsC (1-acyl-

sn

-glycerol-3-phosphate acyltransferase)forming phosphatidic acid A serine molecule is added tothe phosphate group forming phosphatidylserine whichis then decarboxylated to create the PE The fatty aciddiversity of PE molecules is due in part to the activities ofthe PlsB and PlsC acyltransferases and in part to the poolof available fatty acids

In this work the molecular diversity of membrane PEwas examined in wild-type and mutant strains showingthat

M xanthus

contains significant amounts of unsatur-ated fatty acids at the

sn

-1 position Mixtures of PE mol-ecules enriched in

sn

-1 161 are stronger attractants thanmolecules with

sn

-2 161 This effect is associated with161

ω

5 as it is the major monounsaturated fatty acid in

M xanthus

Results

Analysis of double bond position in unsaturated fatty acids

Fatty acid methyl esters (FAMEs) were prepared fromwild-type cells (Haumlrtig

et al

2005) separated by gaschromatography (GC) and analysed by mass spectrom-etry (MS) (Fig 1A) The most abundant fatty acid wasiso150 while the second most abundant was 161

ω

5 thisresult agrees with previous analyses (Kearns

et al

2001a Ware and Dworkin 1973) In order to identify thepositions of the double bond unsaturated fatty acids werederivatized with dimethyldisulfide (DMDS) which deriva-tizes the carbons flanking the double bond and facilitatesfragmentation between the derivatized carbons (Fig 1Binset) The resulting mass spectrum verifies the positionof unsaturation In the case of the major 161 GC peakthe double bond was as the

ω

5 position (Fig 1B) Theminor 161 species was the

ω

11 isomer 151 exists asboth

ω

10 and

ω

4 isomers (data not shown) although the

ω

4 isomer is more abundant (Fig 1A) Finally an iso171was found Because it eluted from the GC column earlierthan the iso170 it appeared to be a branched monoun-saturated fatty acid The fragmentation of the DMDS-derivitized product indicated unsaturation at the

ω

5 posi-tion (

∆

11

data not shown)

Fig 1

Analysis of fatty acids from

Myxococcus xanthus

A Fatty acid methyl esters were prepared from wild-type

M xanthus

DK1622 cells separated by gas chromatography and analysed by mass spectrometryB Unsaturated fatty acid methyl esters were derivatized with dimethyldisulfide (DMDS) to determine the point of unsaturation by detecting the size of resultant fragments after derivatiza-tion and fragmentation The mass spectrum of the 161

ω

5 DMDS derivative is depicted Frag-mentation between the

ω

5 and

ω

6 carbons would result in

mz

245 and 117 fragments respectively (inset) The parent ion (

mz

362) and fragment ions are detected in addition to the ion

mz

213 resulting from the loss of meth-anol from the fragment

mz

245


1937



8

1935ndash1949

Analysis of

M xanthus



mz


middot

frag-mented with N

2


sn

-1 and

sn


et al

2002 Taguchi

et al

2005)


M xanthus


sn

-1and

sn


mz


et al

2001a)

Table 1


Myxococcus xanthus

strains

Progenitorion (

mz

amu)



sn

-1

sn

-2WT Veg(DK1622)

WT Dev(DK1622)

plsB1

(LS2300)

plsB2

(LS2301)

esg

a

(JD300)

plsC2

(LS2305)

plsC4

(LS2306)

plsC5

(LS2307)

6466 141161

150131

25 28 30 33 3981

6486 140 150 56 53 80 44 63 68 616586 161

151141151

3272

6606 151161

150140

102 110 29 58 96

b

142 107197

6624 150 150 203 122 299 212 103 215 2796723 161

162151150

71 43 61 46 50 62123

6744 161 150 120 61 131 79 113 109 1036766 150 160 47 506846 161 162 71 65 30 47 97 44 46 406863 161

172161150

120 155 73 82 278 95 122 11663

6885 171160

150161

41 70 93 36 78

c

60 6543

6905 170 150 45 33 144 616984 172 161 34 45 57 53 397006 171 161 60 94 35 87 61 807023 170

181161150

75 86 40 46 53 47


a

Strain JD300 (

esg


et al

1995)

b


c


sn


sn













EcoliPlsB

MxanthusPlsB1

MxanthusPlsB2

EcoliPlsB

MxanthusPlsB1

MxanthusPlsB2

EcoliPlsB

MxanthusPlsB1

MxanthusPlsB2










































R 2 095 009 088







Discussion



































































Stimulation assay


Acknowledgements


References









































































1937



8

1935ndash1949

Analysis of

M xanthus



mz


middot

frag-mented with N

2


sn

-1 and

sn


et al

2002 Taguchi

et al

2005)


M xanthus


sn

-1and

sn


mz


et al

2001a)

Table 1


Myxococcus xanthus

strains

Progenitorion (

mz

amu)



sn

-1

sn

-2WT Veg(DK1622)

WT Dev(DK1622)

plsB1

(LS2300)

plsB2

(LS2301)

esg

a

(JD300)

plsC2

(LS2305)

plsC4

(LS2306)

plsC5

(LS2307)

6466 141161

150131

25 28 30 33 3981

6486 140 150 56 53 80 44 63 68 616586 161

151141151

3272

6606 151161

150140

102 110 29 58 96

b

142 107197

6624 150 150 203 122 299 212 103 215 2796723 161

162151150

71 43 61 46 50 62123

6744 161 150 120 61 131 79 113 109 1036766 150 160 47 506846 161 162 71 65 30 47 97 44 46 406863 161

172161150

120 155 73 82 278 95 122 11663

6885 171160

150161

41 70 93 36 78

c

60 6543

6905 170 150 45 33 144 616984 172 161 34 45 57 53 397006 171 161 60 94 35 87 61 807023 170

181161150

75 86 40 46 53 47


a

Strain JD300 (

esg


et al

1995)

b


c


sn


sn













EcoliPlsB

MxanthusPlsB1

MxanthusPlsB2

EcoliPlsB

MxanthusPlsB1

MxanthusPlsB2

EcoliPlsB

MxanthusPlsB1

MxanthusPlsB2










































R 2 095 009 088







Discussion



































































Stimulation assay


Acknowledgements


References


















































































EcoliPlsB

MxanthusPlsB1

MxanthusPlsB2

EcoliPlsB

MxanthusPlsB1

MxanthusPlsB2

EcoliPlsB

MxanthusPlsB1

MxanthusPlsB2










































R 2 095 009 088







Discussion



































































Stimulation assay


Acknowledgements


References

















































































































R 2 095 009 088







Discussion



































































Stimulation assay


Acknowledgements


References



































































































R 2 095 009 088







Discussion



































































Stimulation assay


Acknowledgements


References





















































































R 2 095 009 088







Discussion



































































Stimulation assay


Acknowledgements


References













































































Discussion



































































Stimulation assay


Acknowledgements


References



































































































































Stimulation assay


Acknowledgements


References


























































































































Stimulation assay


Acknowledgements


References













































































































Stimulation assay


Acknowledgements


References


























































































Stimulation assay


Acknowledgements


References












































































Stimulation assay


Acknowledgements


References





















































































































































Documents

Novel lipids in Myxococcus xanthus and their role in ...davidcwhite.org/fulltext/570.pdfM. xanthus contains signiﬁcant amounts of unsatur-ated fatty acids at the sn-1 position. Mixtures