The modular xylanase Xyn10A from Rhodothermus marinus cell

88 Pages·2016·4.97 MB·English

Bench-Scale Production of Heterologous Proteins from Extremophiles- Escherichia
coli and Pichia pastoris based expression systems
Ramchuran, Santosh
2005
Link to publication
Citation for published version (APA):
Ramchuran, S. (2005). Bench-Scale Production of Heterologous Proteins from Extremophiles- Escherichia coli
and Pichia pastoris based expression systems. [Doctoral Thesis (compilation), Biotechnology]. Santosh O.
Ramchuran, Biotechnology (LTH), Lund University.
Total number of authors:
1
General rights
Unless other specific re-use rights are stated the following general rights apply:
Copyright and moral rights for the publications made accessible in the public portal are retained by the authors
and/or other copyright owners and it is a condition of accessing publications that users recognise and abide by the
legal requirements associated with these rights.
• Users may download and print one copy of any publication from the public portal for the purpose of private study
or research.
• You may not further distribute the material or use it for any profit-making activity or commercial gain
• You may freely distribute the URL identifying the publication in the public portal
Read more about Creative commons licenses: https://creativecommons.org/licenses/
Take down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove
access to the work immediately and investigate your claim.
LUND UNIVERSITY
PO Box 117
221 00 Lund
[phone]
FEMSMicrobiologyLetters241(2004)233–242
www.fems-microbiology.org
The modular xylanase Xyn10A from Rhodothermus marinus is
cell-attached, and its C-terminal domain has several
putative homologues among cell-attached proteins within the phylum
Bacteroidetes
Eva Nordberg Karlsson a,*, Maher Abou Hachem a,1, Santosh Ramchuran a,
Hugo Costa a, Olle Holst a, A˚sa Fex Svenningsen b, Gudmundur O. Hreggvidsson c
aDepartmentBiotechnology,CenterforChemistryandChemicalengineering,LundUniversity,P.O.Box124,SE-22100Lund,Sweden
bDepartmentPhysiologicalSciences,DivisionofNeuroendocrineCellBiology,BMCF10,LundUniversity,SE-22184Lund,Sweden
cProkariaLtd,Gylfaflo¨t5,IS-112,Iceland
Received27August2004;receivedinrevisedform13October2004;accepted14October2004
Firstpublishedonline28October2004
EditedbyE.Ricca
Abstract
Untilrecently,thefunctionofthefifthdomainofthethermostablemodularxylanaseXyn10AfromRhodothermusmarinuswas
unresolved.Aputativehomologuetothisdomainwashoweveridentifiedinamannanase(Man26A)fromthesamemicroorganism
whichraisedquestionsregardingacommonfunction.Anextensivesearchofallaccessibledata-basesaswellasthepartially
sequencedgenomesofR.marinusandCytophagahutchinsoniishowedthathomologuesofthisdomainwereencodedbymultiple
genesinmicroorganismsinthephylumBacteroidetes.Moreover,thedomainoccurredinvariablyattheC-terminiofproteinsthat
werepredominantlyextra-cellular/cellattached.Aprimarystructuremotifofthreeconservedregionsincludingstructurallyimpor-
tantglycinesandaprolinewasalsoidentifiedsuggestingaconserved3Dfold.Thisbioinformaticevidencesuggestedapossibleroleof
thisdomaininmediatingcellattachment.Toconfirmthistheory,R.marinuswasgrown,andactivityassaysshowedthatthemajor
partofthexylanaseactivitywasconnectedtowholecells.Moreover,immunocytochemicaldetectionusingaXyn10A-specificanti-
bodyprovedpresenceofXyn10AontheR.marinuscellsurface.Inthelightofthis,arevisionofexperimentaldatapresentonboth
Xyn10AandMan26Awasperformed,andtheresultsallindicateacell-anchoringroleofthedomain,suggestingthatthisdomain
representsanoveltypeofmodulethatmediatescellattachmentinproteinsoriginatingfrommembersofthephylumBacteroidetes.
(cid:1)2004FederationofEuropeanMicrobiologicalSocieties.PublishedbyElsevierB.V.Allrightsreserved.
Keywords:Cellattachment;Xylanase;Bacteroidetes
1.Introduction
Glycosidehydrolase(GH)multiplicityisacommon
*Correspondingauthor.Tel.:[phone];fax:[phone] theme among microorganisms capable of degrading
4713. thecomplexandrecalcitrantpolysaccharidecomposites
E-mail address: [email protected] (E.N.
foundinplantandalgalcellwalls.Theseenzymestypi-
Karlsson).
1Presentaddress:BiochemistryandNutritiongroup,Biocentrum callyhaveamodularorganisationconsistingofcatalytic
DTU,DK-2800KgsLyngby,Denmark modules (CMs) usually, but not always, joined to
0378-1097/$22.00(cid:1)2004FederationofEuropeanMicrobiologicalSocieties.PublishedbyElsevierB.V.Allrightsreserved.
doi:10.1016/j.femsle.2004.10.026
234 E.N.Karlssonetal./FEMSMicrobiologyLetters241(2004)233–242
non-catalytic modules (NCMs) by flexible linker se- moduleclassifiedasGH10andafinallya5thdomain
quences [1,2]. The most common types of NCMs are (D5) at its C-terminus [14,16]. Alignments of the Rm
carbohydrate-binding modules (CBMs), but a number Xyn10A-catalyticmodulewithfamily10xylanases,un-
of other domains or modules, some of yet unknown masked D5 as an extended C-terminal sequence [14],
function, have been reported and include NCMs in- preceded by a short stretch of repeated glutamic acid
volvedincelladhesion,orproteinanchoring[3]. and proline residues, typical for the linker sequences
Two different types of domains have earlier been often found joining modules in glycoside hydrolases.
suggested to play a role in cell adhesion/anchoring of Tocastmorelightonthepossiblefunctionofthisdo-
glycosidehydrolases,thesearetheso-calledfibronecti- main a search for similar sequences was accomplished
nIII-like domains (Fn3-like domains), and the S-layer using accessible databases as well as available partial
homology domains (SLH-domains). The Fn3-like do- genome sequences from R. marinus and from the re-
main, which has a length of approximately 100 resi- lated organism Cytophaga hutchinsonii. Based on the
dues, is phylogenetically spread and presented in a findingsinthissearchandcombinedwithexperimental
superfamilyofsequencesrepresentingreceptorproteins evidence the possible role in cell-adhesion of the
or proteins involved in cell-surface binding mainly in Xyn10A C-terminal domain and its homologues is
eukaryotes.Itisalsofoundinsomeextra-cellularbac- discussed.
terialglycosidehydrolases[4].Thesedomainsarecom-
monly distributed in multiple copies in modular
glycosidehydrolases,andareoftenfoundbetweencat- 2.Materialsandmethods
alytic modules and CBMs. Recently it has however
been demonstrated that the Fn3-like domains have a 2.1.Sequenceanalysisandsimilaritysearches
role in hydrolysis of insoluble substrates [5]. The
SLH-domain is a domain of about 50–60 residues, Bioinformatictoolswereusedtoexploretheprimary
found at the N- or C-termini of mature proteins [6] structureofD5inXyn10AfromR.marinus.Similarity
and is believed to be anchored to the peptidoglycan searchesbyBLAST,usingD5ofR.marinusastemplate,
[7], or some other structure in the bacterial cell wall were performed on the NCBI server (http://
[8]. This domain is almost exclusively bacterial with www.ncbi.nlm.nih.gov) or locally using BioEdit v.
63 of the 64 sequences reported to pfam-database 5.0.6.onavailablegenomesequencesofC.hutchinsonii
(http://pfam.wustl.edu) in the bacterial branch, and (from the KEGG database), or partial genome se-
most commonly occurring within the Bacillus/Clostri- quencesofR.marinus(availableviaProkariaLtd,Rey-
dium group andinrelated Gram-positive bacteria[6]. kjavik, Iceland). Location of a putative signal peptide
ThethermophilicmarineaerobicbacteriumRhodoth- waspredictedbySignalPv.1.1.(http://www.cbs.dtu.dk/
ermusmarinusstainsGram-negativeandisphylogeneti- services/SignalP).Matcheswithopenreadingframesof
callyaffiliated totheBacteroidetes(alsoknownasthe unknown function were subjected to an additional
Cythophaga/Flexibacter/Bacteroides-group) [9], a phy- searchbyBLASTafterdeletionofthepartshowinghigh
lumwithmanyknowndegradersoforganicmatter.This similaritytoD5,topredicttheputativefunctionofthe
groupofbacteriaisknowntoproduceanumberofcel- remainingpartoftheORFs.
lulose degrading enzymes. Moreover, members of the The ClustalW tool on the EBI server (http://www.
Cytophaga,oneofthebetterstudiedgenerawithinthis ebi.ac.uk/clustalw)wasusedtocreatemultiplesequence
phylum,donotproducesolubleextra-cellularcellulose alignmentsandphylogenetictrees,displayedusingGene
hydrolases,butinsteadkeeptheirenzymesattachedto doc2.6.02[17],andTreeView1.5[18],respectively.The-
the cell-envelope [10]. Despite established cell-attach- oreticalisoelectricpoints,andaminoacidcomposition
ment, only one gene within the phylum Bacteroidetes ofdeducedaminoacidsequenceswereanalysedbyProt-
hasbeenreportedtoencodeahomologuetotheSLH- Param (http://www.expasy.org/tools/protparam.html),
domain(anS-layerproteinprecursorfrom Cytophaga andMicrosoftExcel.
sp.Jeang1995).
Rhodothermus marinus resembles other microorgan- 2.2.CultivationofR.marinus
isms within this phylum, in its ability to produce a
number of glycoside hydrolase activities [11–13], as Rhodothermusmarinuswasgrownwithandwithout
well as in displaying enzyme activities suggested to be xylan(5g/L,Birch7500.1fromCarlRoth,Karlsruhe,
cell-attached. Primary structures of some of the R. Germany)at65(cid:2)C,pH7.1,withaerationon5L/min,
marinus glycoside hydrolases are known, including in2.5LmodifiedM162medium[11,19]ina3Lbiore-
one family 10 xylanase [14], and one family 26 man- actor inoculated with a 100 ml shake-flask culture
nanase [15]. The xylanase (Xyn10A) is a modular en- (OD =0.7). Optical density measurements
620nm
zyme that consists of two N-terminal family 4 CBMs (OD )monitoredcellgrowth.Mannanaseproduc-
620nm
followedbyadomainofunknownfunction,acatalytic ingR.marinusweregrownin100mlshake-flaskcultures
E.N.Karlssonetal./FEMSMicrobiologyLetters241(2004)233–242 235
usingtheM162mediumincludingLocustBeanGum6 Theslideswerethenrinsed2·10mininblockingsolu-
g/L(Sigma–Aldrich,St.Louis,Mo). tionandonceinPBS.
Samplesforactivityanalysisandelectrophoresiswere AfterthisprocedureflourescentSytoxgreen(Molec-
withdrawnduringtheearlylog,mid-log,latelogandin ularProbes,WA,USA)wasaddedina1:3000solution
the stationary phases and kept at 4 (cid:2)C until analysis. ofPBSfor10min,andthenrinsedforanother15min
Theculturesupernatantandcell-fractionwereseparated inPBSbeforemounting.
by centrifugation at 25,000g for 30 min at 4 (cid:2)C. The The immuno-labelled cells were visualized using an
whole cell-fraction was washed with 20 mM sodium Olympus BX-60 microscope connectedto anOlympus
phosphate buffer at pH 7.0, recentrifuged and resus- DP-50digitalcamera.Photomicrograpsweretakenwith
pended to the original sample volume in the above theviewfinderLitesoftware.
buffer.
3.Resultsanddiscussion
2.3.Activityanalysisandelectrophoresis
3.1.SimilaritybetweenC-terminalpartsofR.marinus
Xylanaseactivitywasmeasuredintheculturesupern-
xylanaseandmannanase
atant,andonwholecellsusingtheDNSmethodasde-
scribed elsewhere [20] with birch xylan (Carl Roth) as
Initially,theonlydomainamongthepubliclyacces-
substrateandusingindividualenzymeblanks.Xylanase
sible sequences found to share primary structure simi-
production in R. marinus was also analysed using so-
larity with the Xyn10A D5 domain of R. marinus,
diumdodecylsulfatepolyacrylamidegelelectrophoresis
was from another hemicellulose degrading enzyme
(SDS–PAGE) with 10% separation gels [21], activity
originating from the same organism (Rm Man26A)
stainedwithCongo-redaspreviouslydescribed[22],ex-
[15]. The similarity was restricted to the C-terminal
ceptforthefollowingmodifications:theover-layeragar-
partof the two enzymes (33% identity) (Fig. 1). Eval-
ose gel contained 0.05% oat spelt xylan (Sigma), the
uation of a multiple sequence alignment including the
bufferusedwas50mMTris–HCl,pH7.5,andincuba-
tiontimeat65(cid:2)Cwas60min. R. marinus mannanase and a number of known cata-
lytic modules (CMs) of GH 26 suggested also the C-
Mannanaseactivitywasdeterminedbyahaloplate
terminal part (residues 939–1021) of this enzyme to
assaycontaining3.5%(w/v)agarand0.1%(w/v)Azo-
beaseparatedomain,asitflankstheCMdownstream
Carob Galactomannan (Megazyme, Bray, Ireland).
Samples(80lL)wereloadedintowellsandplateswere the consensus region, rather than lying within it even
incubatedat65(cid:2)Covernight. thoughnolinkersequenceseparatingitfromthecata-
lyticmoduleisdistinguishableinthisenzyme(datanot
shown).
2.4.Immunocytochemistry It was also noted that D5 of RmXyn10A (residues
913–997,intheXyn10A-sequence)hasatheoreticalpI
Drops of cell suspension were dried on SuperFrost value of 11.05, which is strikingly higher than either
microscope slides (Menzel-Gla¨ser, Germany). When thefull-lengthxylanaseoranyisolatedmodulethereof
completelydry,thecellswerefixedfor20mininStefa- (allwithpI:sof4–4.5)butinbetteragreementwiththe
ninifixative(2%paraformaldehydeand15%saturated C-terminaldomainofRmMan26A(pI11.65).Thehigh
aqueouspicricacidsolutionin0.1Mphosphatebuffer, similarity betweenthese two domains andtheir occur-
pH 7.2), followed by repeated rinsing in sucrose-en- renceattheC-terminusoftwomodularhemicellulases
riched10%Tyrodes(cid:1)solutionandfinallyinphosphate-
bufferedsaline(PBS).Thecellswerethenpermeabilized
andblocked(inthesamestep)usingablockingsolution
containing;0.25%TritonX-100and0.25%BSA(both
fromSigma)inPBS.Thiswasfollowedbyincubation
inamoistchamberwitharabbitanti-CBM4-2primary
antibody[immunoglobulinfraction(10mg/ml)fromser-
umdrawnfromarabbitimmunisedwithrecombinant
produced purified carbohydrate binding module Fig.1.DomainstructureofthexylanaseXyn10Aandmannanase
(CBM4-2) of Xyn10A] in a 1:100 dilution in blocking Man26Aof Rhodothermusmarinus. The identifieddomains (D)or
solution,overnight.Thenextmorningexcessantibody unknown regions in the primary sequence are shown as blocks.
wasrinsedoffandthecellswerefurtherincubatedfor Regions/domainsofunknownfunctionareshowninwhite,catalytic
modules(CM)ingreyandcarbohydratebindingmodules(CBM)ina
2 h, with flourescent Rhodamine Red-X-conjugated
squaredpattern.Identifiedlinkersequencesareshowninblack,and
donkeyanti-rabbitsecondaryantibody(JacksonLabo- signal peptides striped. The C-terminal domains of the respective
ratories, PA,USA)diluted 1:400 in blocking solution. proteinaremarkedbydashes.
236 E.N.Karlssonetal./FEMSMicrobiologyLetters241(2004)233–242
Rm1992
Rm4942
Rm3381
Rm2220
Rm4825
Rm5009
RmMan26A
Ch18
RmXyn10A
Rm2102
RmPL6
Ch3
Rm3906
Rm4503
Rm3935
Rm5177
Ch37
Ch32
Ch42
Ch4
Ch6
Ch15
Ch38
Ch5
Ch22
Ch23
Ch12
Ch17
Ch13
Ch20
Ch11
Ch8
Ch39
MMS116
Ch36
Zg
Ch33
Ch26
Ch14
Ch9
Ch40
Ch24
Ch10
MMS130
Pg PG102
Pg PG91
Pg pOMB
Pg PG97
Ch31
Ch21
Ch28
Ch35
Ch7
Ch30
Ch34
Ch16
Pg PG99
Ch29
Ch27
Ch19
Ch41
0.1
Fig.2.PhylogramovertheputativedomainsfoundbyBLAST-similaritysearchusingD5ofRmXyn10Aastemplate.Openreadingframesfrom
the respective organism are numbered, and labelled: Ch (Cythophaga hutchinsonii), M (Microscilla sp.), Pg (Porphyromonas gingivalis), Rm
(Rhodothermusmarinus),Zg(Zobelliagalactanivorans).ThephylogramisdisplayedusingTreeView,andcreatedusingtheEBI-ClustalW-toolwith
theoutputinthephylipformat,andusingdefaultparameters,withcorrectdist.‘‘on’’,andignoregaps‘‘off’’.
fromthesameorganismssuggestedthepossibilityofa can-bindingcapacityofthexylanasedomain,rulingout
commonfunction.Affinityelectrophoresisanalysisina acarbohydratebindingfunctiontosubstratesrelatedto
previousworkfailedtoshowanymannan,xylanorglu- eitherofthecatalyticmodules[23].
E.N.Karlssonetal./FEMSMicrobiologyLetters241(2004)233–242 237
Fig.3.MultiplesequencealignmentofC-terminaldomainsencodedintheproteinsfoundbyBLAST-similaritysearchusingD5ofRmXyn10Aas
template.Thefirstconsensusregionstartsatposition7inthealignment(residue915,Xyn10Anumbering)andspanssevenresidueswiththemotif
[(I/L/M/V),X,(I/L/M/V),(F/W/Y),P,N,P].Thesecondconsensusregionspanspositions37–44[(I/L/V),X,(I/L/M/V),(I/L/M/T/V/F/W/Y),(D,N),
(I/L/M/V),X,G],andmostlyinvolvesconservedhydrophobicresidues.Thisisalsotrueforthethirdregion,whichislocatedatposition74–85inthe
alignment,andhas1–2insertedresiduesinafewofthesequences[(I/L/M/V),X,X,G,(I/L/M/T/V),Y,-,-,(F/I/L/M/V),(I/L/M/V),X,(I/L/M/V)].
ThedomainsoriginatefromfivespeciesallaffiliatedtotheBacteroidetesphylum.Openreadingframesfromtherespectiveorganismarenumbered,
Ch(Cythophagahutchinsonii),M(Microscillasp.),Pg(Porphyromonasgingivalis),Rm(Rhodothermusmarinus),Zg(Zobelliagalactanivorans).The
alignmentiscreatedusingtheEBI-ClustalW-tool,anddefaultparameters.TheresultingalignmentwasanalysedinGeneDoc.Conservedresiduesare
identified [The following residues are grouped, and considered conserved within the group: 1.(D,N); 2.(E,Q); 3.(S,T); 4.(K,R); 5.(F,Y,W);
6.(L,I,V,M).]andshadedifpresentinmorethan60%ofthesequences.
3.2.HomologousC-terminaldomainsareencodedin ofthisdomain.Usingthisapproach,anumberofhits
multiplegenesinR.marinusandrelatedorganisms in mostly putative open reading frames were found,
invariablylocatedattheC-terminiinthededucedami-
The observations presented above prompted an ex- no acid sequence of the respective gene. The highest
tendedsearchforhomologuestoD5ofXyn10Ausing scorehitsoriginatedfromfivedifferentmicroorganisms
partialgenomicsequencedatafromR.marinus(avail- allclassifiedwithinthephylumBacteroidetes.Mostof
ableviaProkariaLtd,Iceland),andpubliclyavailable the hits originated from genome sequences of two
sequencedatabasesattemptingtounravelthefunction microorganisms (R. marinus and C. hutchinsonii), but
238 E.N.Karlssonetal./FEMSMicrobiologyLetters241(2004)233–242
Table1
AsummaryofaveragepropertiesfortheC-terminaldomainsfoundinthefivemicrobialspecies,inwhichgenesencodingtheputativehomologueto
theC-terminaldomainofRmXyn10Awerefound
Ch M Pg Rm Zga
Numberofgenes 39 2 5 13c 1
Averagelength(residues) 79±0.5 78±0 75±1 85±0.5 73
Theoreticalmolecularweight(kDa) 8.64±0.06 8.93±0.04 8.43±0.13 9.54±0.09 7.97
PI 5.6±0.2 5.4±0.3 7.5±0.7 9.4±0.5 6.0
Contentofselectedresiduesb(%oftotalnumberofresidues)
Ala 6.1±0.5 4.5±1.9 5.9±1.4 8.7±1.2 4.1
Arg 1.6±0.2 3.8±1.3 4.5±0.9 11.2±0.9 0
Asn 7.4±0.4 3.2±0.6 4.8±0.3 2.4±0.4 8.2
His 0.9±0.1 1.3±0 1.3±0.02 2.4±0.4 0
Ile 9.5±0.6 7.0±2.0 6.1±1.1 2.0±0.6 9.6
Lys 6.2±0.3 7.7±0 7.7±1.3 1.0±0.3 8.2
Pro 4.1±0.2 3.2±0.6 3.7±0.7 6.6±0.6 4.1
Ser 7.9±0.5 7.0±2.0 6.1±0.8 3.2±0.6 11
Themicrobialspeciesare:Cytophagahutchinsonii(Ch),Microscillasp.(M),andPorphyromonasgingivalis(Pg),Rhodothermusmarinus(Rm),and
Zobelliagalactanivorans(Zg).Valuesaregivenastheaverage±SEM.
aOnlyonegenefound,nostatisticspossible.
bValuesaregivenforthoseresidueswherethedomainsfromenzymesofthermophilicR.marinusdiffersignificantlyfromthedomainsfromthe
mesophilicmicroorganisms.
cTheORFencodingaputativefamily6pectatelyasewasexcluded,duetoashortC-terminaldomain(58residues)indicatingaframeshift
possiblycausedbyasequencingerror.
additionalhitswerealsofoundwithinthreeotherspe- the thermophilicity of this organism. There is for in-
cies,fromthesamephylum.Inaddition,threeofthese stance a dramatic increase in the Arg/Lys-ratio, as
fivemicroorganisms(R.marinus,C.hutchinsonii,anda previously reported for many thermophilic proteins
Microscillasp.),wereclassifiedintothesameclassand [24], a decrease of residues prone to deamidation (in
order (Sphingobacteria; Sphingobacteriales), and both this case seen as a decrease in Asn content), and an
R. marinus and C. hutchinsonii to the same family increase in Pro (which could increase rigidity in the
(Chrenotrichaceae) within the bacterial lineage. The structure).
remaining two microorganisms (Porphyromonas gingi- Most of the genes (except Xyn10A and Man26A)
valisand Zobellia galactanivorans),belongto separate found during the search were uncharacterised ORFs,
classes of Bacteroidetes: Bacteroides and Flavobacte- which motivated analysis of the deduced polypeptide
ria, respectively. In the phylogram analysis, most of regions upstream the D5 homologues. Analysis of
the sequences originating from R. marinus (n=14) the full-length genes revealed that all had a potential
clustered together, and had the shortest distance to N-terminal signal peptide. Furthermore, the best
some of the sequences originating from its closest rel- matches encoded proteins predominantly implicated
ative, C. hutchinsonii (n=39). Sequences from the lat- in extra-cellular functions including glycoside hydro-
ter were, however, distributed throughout the lases, and various other cell attached proteins (Table
phylogram (Fig. 2). Presence of only two genes from 2), while regions encoding polypeptides of clear intra-
Microscilla sp., and one from Z. galactanivorans, ren- cellularfunctionwereabsent.Therelativelyubiquitous
dered discussion of evolutionary relationships prema- occurrence of this domain within taxonomic subdivi-
ture. In contrast, four of the five sequences from P. sions of a single phylum, its conserved location at
gingivalis (n=5) clustered together in the phylogram, the C-termini, and the identification of a conserved
in a position relatively distant to the R. marinus motif within its primary structure (which is likely to
sequences. reflectastructuralcommonfoldormotif),alongwith
A multiple sequence alignment including all the the common extra-cellular nature of the proteins or
putative domains revealed three consensus regions of activities the domain is associated to, strongly suggest
which the first was most conserved (Fig. 3). Some that this is a novel type of module conserved among
physico-chemical properties of respective domain were members of this particular phylum. This along with
also analysed, revealing that most of the R. marinus arevisionoftheexperimentalevidence onthecellular
sequences,whichareonaverageafewresidueslonger, location ofthe R. marinus enzymes (see below)makes
have a relatively high pI (Table 1). It was also noted it extremely tempting topostulate that the cell-attach-
that the content of a number of amino acid residues ment of glycoside hydrolases and of other enzymes
in the R. marinus domains differed significantly from could be mediated by this domain within the
the domains of the other species, probably reflecting addressed taxonomic group.
E.N.Karlssonetal./FEMSMicrobiologyLetters241(2004)233–242 239
dothermus (Q97P71)
Rhom TJS8); 32)
C-terminaldomainofXyn10Afro Q97P71);(Q8TJE3);(Q8TI59);(Q8 Q97FL4);(Q97FL4);(Q97FL4)Q8TTC5)Q8XQF7)Q8XQP2);(Q8Y366);(Q8XPU1)Q934I7)Q9RBJ4);(P71140)Q8YWJ8)Q8TR26) AAM21605);(Q9F1V3);(BAC163Q8PKM0)Q8X4H5)Q8P3F9)Q97TN3)O30678);(Q9C105)Q97F11)Q97TN3andQ9KQP6) P37701andQ9C105)O30654)Q9AQH0);(Q95YN1) Q934I6) Q8YDM7) Q8P9F2)Q9RKE2)
utativehomologuetothe Identification Bysequencesimilarity( Bysequencesimilarity(Bysequencesimilarity(Bysequencesimilarity(Bysequencesimilarity(Bysequencesimilarity(Bysequencesimilarity(Bysequencesimilarity(Bysequencesimilarity( Bysequencesimilarity(Bysequencesimilarity(Bysequencesimilarity(Bysequencesimilarity(Bysequencesimilarity(Bysequencesimilarity(Bysequencesimilarity(Bysequencesimilarity( Bysequencesimilarity(Bysequencesimilarity(Bysequencesimilarity( Annotated(Q93PA1);( Bysequencesimilarity( Bysequencesimilarity(Bysequencesimilarity( Annotated(Q9RGX9)
p
a
d)encodingmodularproteinsthatcontain N-terminalfunction Cellwallsurfaceanchorprotein/cellsurfaceprotein RCC1repeatsproteinSurfaceantigenPutativesignalpeptideproteinHemagglutinin/hemolysinrelatedproteinbPutative-agaraseGH9Subtilasefam.proteinSecretedmetal-bindingprot.(plastocyanin/azurinfam.)GH10PutativeautotransporterproteinPutativeRTX-familyexoproteinKelch-likeproteinSecretedmetalloproteaseGH18IntegrinlikerepeatspredictedsecretedmetalloproteaseandGH18GH8(and18)GH26GH5 bMS130andMS116,putative-agarase Extracellularserineprotease GGDEFfamilyproteinPolysaccharidelyasefamily6 b-AgaraseAprecursor
fie
denti 42
bei 27,
similaritycould GeneorORF ORF4,10,19, ORF5,23,31ORF6ORF7ORF8,9,15ORF11ORF13,14ORF16ORF17 ORF18,33,38ORF20ORF22ORF24ORF26ORF28,41ORF29ORF30 ORF32ORF35ORF36,40 ORF1,2 ORF4825 ORF3935ORFPL6 ORF1
here
w
Table2PutativefunctionofORFs(inthecasesmarinus Bacteriallineage Bacteroidetes;Sphingobacteria;Sphingobacteriales;Crenotrichacae;Cytophagahutchinsonii Bacteroidetes;Sphingobacteria;Sphingobacteriales;Flexibacteraceae;Microscillasp. Bacteroidetes;Sphingobacteria;Sphingobacteriales;Crenotrichaceae;Rhodothermusmarinus Bacteroidetes;Flavobacteria;Flavobacteriales;Flavobacteriaceae:Zobelliagalactanivorans
240 E.N.Karlssonetal./FEMSMicrobiologyLetters241(2004)233–242
3.3.Experimentaldatasupportcell-attachment
Aseriesofbatch-cultivationsofR.marinuswereper-
formedinordertoinduceexpressionofthenativexylan-
ases and monitor their cellular location as to provide
experimental evidence for the hypothesis presented
above. The gene encoding Rm Xyn10A encodes an
N-terminal putative signal peptide [14], ruling out an
intracellular location. This finds support in aprevious
investigation in which it was demonstrated that xyla-
naseactivitywasnotlocatedintheintra-cellularprotein
fraction[11].
In the presence of xylan, R. marinus had a specific
growthrate of0.35 h(cid:1)1, and grew to afinal OD
620nm
of approximately 5. Without xylan, the final OD
620nm
was lower, and xylanase activity not detectable. Xyla-
naseactivitywashencemeasuredinbothR.marinuscell
slurries (cell attached activity), and in the cultivation
medium,oftheculturegrowninpresenceofxylan.In
both fractions, the activity was increasing throughout
the cultivation and the cell associated activity was al-
wayshigherthan thatofthe cultivation medium (Fig.
4(a)), while intracellularly, it was below the detection
limitoftheassayasdeterminedbyproteinfractionation
[11]. Evidence that the cell-bound activity originated
from Xyn10A emerged from two techniques: activity
staining and immunocytochemistry. Activity staining
afterelectrophoresisofsamples(identicaltothoseana-
lysedforactivity)showedasinglebandofxylanohydro-
lyticactivityoftheexpectedmolecularmass(M (cid:2)110
r
kDa) in the cell fraction (showing intracellular and
cell-attachedproteins,Fig.4(b)),excludingthepossibil-
ity that the cell-bound activity arose from additional
cell-attachedxylanasesproducedbytheorganism.Def-
initeproofoftheattachmentofXyn10AtoR.marinus
cells was collected after immunocytochemical analysis,
in which primary antibodies against the carbohydrate
bindingmoduleofXyn10A(producedasarecombinant
proteininEscherichiacoli)wereeffectivelyboundtothe
cells(Fig.5).
Two activity bands were observed in samples from
Fig.4.XylanaseactivityintheculturesupernatantandonR.marinus
the cultivation medium, one with high ((cid:2)100 kDa) wholecells. Activitywas analysedbythe DNS-assayforreducing
andonewithlow(<25kDa)apparentmolecularmass sugars(a)usingbirchxylanasthesubstrate,andbyactivitystaining
(Fig. 4(b)). The low molecular mass activity band afterelectrophoreticseparation(bySDS–PAGE)andin-gelrefolding
proved presence of an additional secreted xylanase in (b), using Congo-red staining of a xylan-containing overlayer-gel.
Samplesforactivityanalysis(a),weretakenattimescorrespondingto
R. marinus. Interestingly the M differences between
r differentphasesofthegrowth-curve,after3h(earlylogphase),9h
the higher molecular mass (100 kDa) activity-band (mid-logphase),13h(latelog-phase)and16h(stationaryphase)and
and that of Xyn10A conformed well to the M of theactivityofthesupernatant(grey)andthewholecells(black)was
r
D5, suggesting the possibility of release after (proteo- analysed.Thepositionsofthestandardsintheprestainedmolecular
lytic) cleavage at the linker preceding D5. This would massladder(Biorad)afterseparationisindicatedbyascalpelinthe
overlayergel,andactivitybandsfromthemid-logsample(9h)ofthe
explain the previously observed slow release of xyla-
supernatant(left)andcellpellet(right)ofasizecorrespondingto
nase to the growth medium during the stationary Xyn10Aareindicatedbyarrows(b).
phase [11]. An alternative explanation for the release
would be cell lysis, but the apparent decrease in lected from production patterns of recombinant Rm
molecularmassinourcurrentresultssupportsthefor- Xyn10A and truncated variants in E. coli, which
mer.Furthersupportforproteolyticcleavagewascol- showed two linker positions in the heterologous full-
E.N.Karlssonetal./FEMSMicrobiologyLetters241(2004)233–242 241
Fig.5.FluorescencemicroscopyimageafterimmunocytochemicalstainingofR.marinuscellsfromthelategrowthphasegrowninthepresenceof
xylan.Bindingoftheprimaryxylanase-antibodytothecellsisvisualisedafterincubationwithsecondaryantibodyconjugatedtothefluorophore
RhodamineRed-X(a).SubsequentstainingofallcellswithfluorescentSytoxGreennucleicacidstain(b)demonstratedthatthexylanasewas
expressedonthesurfaceofallcells.Barrepresents10lm.
length enzyme to be susceptible to proteolytic cleav- ties with SLH-domains (proposed to anchor GH to
age, one of them being the linker preceding D5 [23]. many G-positive microorganisms by noncovalent
Since the functions of the other modules in this en- interactionswithsecondary cellwallpolymers,inturn
zymeareknownexceptforthethirddomain(D3),cell covalentlylinkedtopeptidoglucan[7,8])sonoconclu-
attachment could theoretically be mediated by either sion can be made based on their mechanism. Another
domain. The third domain, however, has no homo- differenceisthatR.marinusisaG-negativebacterium
logues in this microorganism or in related ones (an [25]. Based on the observation of the three conserved
does not so far show significant sequence similarities regions (with a number of hydrophobic residues) in
to any deposited sequence), which practically dis- the domain, attachment could be mediated by hydro-
counts it from any common function such as cell phobic stretches (specific or unspecific), but a protein-
attachment, and leaves D5 as the most credible site carbohydrate interaction (e.g., via lipopolysaccharides
of cell anchoring. (LPS),foundtooftenbemediatedbyCa2+)[6]isalso
Cultivations in presence of galactomannan (locust possible. At this point further discussion will have to
beangum)inourlaboratoryshowedalsothemajorpart await isolation and/or characterisation of R. marinus
ofthemannanaseactivityassociatedwiththecellslurry cell wall polymers, in order to study possible
(data not shown). This finding together with the facts alternatives.
thattheonlysharedsimilaritybetweenthetwoR.mar-
inusenzymesisintheC-terminaldomain,andthedata
fromPolitzandcoworkers[15]promotethecell-attach- 4.Conclusion
ment hypothesis. Politz and co-workers found cell-
bound (and no intracellular) mannanase activity after Bioinformaticandexperimentallinesofevidenceare
proteinfractionationofR.marinuscells.NoN-terminal analysed in this study to elucidate the function of the
signalpeptidewasrecognisedinMan26AbyPolitzand fifthdomain(D5)inXyn10A.Ourresultsclearlydem-
coworkers,butasthecollecteddataonthelocationof onstrate that this domain is relatively wide-spread
theenzymeprovedexportoftheenzymeatleasttothe within subdivisions of the Bacteroidetes taxonomic
periplasmicspace[15]wehave(usingtheprogramSign- groupandthatitdisplaysacommonprimarystructure
alP)recognisedalikelylocationofasignalpeptideon motif and location at the C-termini which reflects its
theN-terminalsideofthecatalyticmodule(comprising commonfunctionandevolutionaryhistory.Moreover,
the sequence MTLLLVWLIFTGVA). In accordance experimental evidence from R. marinus cells in combi-
with the xylanase experiments and with the activity nation with evidence gathered from previous studies
graphspublishedbyGomesandSteiner[13],someman- leadstothesuggestionthatthisdomaintypemediates
nanaseactivitywasalsoreleasedtotheculturemedium cell-attachment in proteins produced by members of
duringthelatergrowthphaseandthestationaryphase. the Bacteroidetes. Additional studies are motivated to
highlight the mechanism of cell-attachment and the
3.4.Howarethedomainsattachedtothecells? structural basis of this function. Finally, identification
of a potential cell receptor for this module (if any)
CurrentlywecanonlyspeculateonhowXyn10Ais opensthedoorforuseofthereceptor/moduleinterac-
attachedtotheR.marinuscell.Thedomainsfoundin tion in biochemical studies, and for biotechnological
this search did not share significant sequence similari- applications.