6487_Van Der Meij

12 Pages·2017·1.24 MB·English

AntonievanLeeuwenhoek(2018)111:679–690
https://doi.org/10.1007/s10482-018-1014-z
ORIGINAL PAPER
Inter- and intracellular colonization of Arabidopsis roots
by endophytic actinobacteria and the impact of plant
hormones on their antimicrobial activity
. . .
Anne van der Meij Joost Willemse Martinus A. Schneijderberg
Rene´ Geurts . Jos M. Raaijmakers . Gilles P. van Wezel
Received:20November2017/Accepted:3January2018/Publishedonline:15January2018
(cid:2)TheAuthor(s)2018.Thisarticleisanopenaccesspublication
Abstract Many actinobacteria live in close associ- clavifer as well represented species. When seeds of
ation with eukaryotes such as fungi, insects, animals Arabidopsis were inoculated with spores of Strepto-
and plants. Plant-associated actinobacteria display myces strain coa1, which shows high similarity to S.
(endo)symbiotic,saprophyticorpathogeniclifestyles, olivochromogenes, roots were colonised intercellu-
and can make up a substantial part of the endophytic larly and, unexpectedly, also intracellularly. Subse-
community. Here, we characterised endophytic acti- quent exposure of endophytic isolates to plant
nobacteria isolated from root tissue of Arabidopsis hormonestypicallyfoundinrootandshoottissuesof
thaliana (Arabidopsis) plants grown in soil from a Arabidopsis led to altered antibiotic production
natural ecosystem. Many of these actinobacteria against Escherichia coli and Bacillus subtilis. Taken
belong to the family of Streptomycetaceae with together, our work reveals remarkable colonization
Streptomyces olivochromogenes and Streptomyces patterns of endophytic streptomycetes with specific
traits that may allow a competitive advantage inside
roottissue.
Electronicsupplementarymaterial Theonlineversionof
thisarticle(https://doi.org/10.1007/s10482-018-1014-z)con- Keywords Streptomyces(cid:2)Plant–microbe
tainssupplementarymaterial,whichisavailabletoauthorized
interactions(cid:2)Planthormone(cid:2)Crypticantibiotics(cid:2)
users.
Electronmicroscopy
AnnevanderMeijandJoostWillemsehavecontributed
equallytothiswork.
A.vanderMeij(cid:2)J.Willemse(cid:2)G.P.vanWezel(&)
MolecularBiotechnology,InstituteofBiology,Leiden
Introduction
University,Sylviusweg72,2333BELeiden,The
Netherlands
e-mail:[email protected] Actinobacteria represent a diverse phylum composed
of both rod-shaped and filamentous bacteria that can
M.A.Schneijderberg(cid:2)R.Geurts
be found in soil, marine and fresh water ecosystems
DepartmentofPlantSciences,WageningenUniversity,
Wageningen,TheNetherlands (Goodfellow 2012). The filamentous actinobacteria
are versatile producers of bioactive natural products,
J.M.Raaijmakers
includingtwo-thirdsofallknownantibioticsaswellas
DepartmentofMicrobialEcology,NetherlandsInstitute
many anticancer, antifungal and immunosuppressive
ofEcology(NIOO-KNAW),Wageningen,The
Netherlands agents (Barka et al. 2016; Be´rdy 2005; Hopwood
123
680 AntonievanLeeuwenhoek(2018)111:679–690
2007).Theactinobacteriaarealsomajorproducersof actinobacteria by the plant hormone salicylic acid
industrially relevant enzymes (Anne et al. 1990; van (SA)(Lebeisetal.2015).SAplaysaroleinavariety
Dissel et al. 2014). Hence, the actinobacteria are of of physiological and biochemical processes and is
utmost importance for human health, agriculture and importantasanendogenoussignalmediatinglocaland
biotechnology. systemic plant defense responses against pathogens
Theconceptofactinobacteriaasfree-livingbacte- and abiotic stress factors (Rivas-San Vicente and
ria has more recently been challenged by studies Plasencia 2011). SA is detectable in Arabidopsis
pointing to their intimate relationships with diverse leaves and roots at concentrations up to 1 lg/g fresh
eukaryotes (Seipke et al. 2012; van der Meij et al. weight(vandeMorteletal.2012).Hence,endophytic
2017). Indeed, they have been found in association actinobacteriaaremostlikelyexposedandresponsive
with vertebrates, invertebrates, fungi and plants. toSAandotherplanthormonessuchasjasmonicacid
Actinobacteriaareoftenwelcomegueststotheirhosts andauxin(Halimetal.2006;Zhao2010).
due to their ability to produce chemically diverse Inthisstudyweisolatedendophyticactinobacteria
natural products. Much of the chemical diversity of from Arabidopsis roots, grown under controlled
secondarymetabolitesproducedbyactinobacteriahas conditions as well as from plants grown in an
likelyevolvedbecauseoftheirinteractionswithother ecological setting. Colonization patterns were moni-
(micro)organisms in highly diverse environments tored and visualised by confocal as well as electron
(Seipke et al. 2012; van der Meij et al. 2017). It is microscopy for Streptomyces strain coa1, which
becomingincreasinglyclearthatactinobacteriaplaya adapted from a soil dwelling to an endophytic
keyroleinmaintainingplanthealthbycontributingto lifestyle, on axencially-grown A. thaliana. Finally,
bioticandabioticstresstolerance(Viaeneetal.2016). weinvestigatedhowspecificplanthormonesaffectthe
Forexample,actinobacteriaproducesiderophoresfor antimicrobial activity of endophytic Streptomyces
iron acquisition as well as antibacterials and antifun- isolated from wild Arabidopsis providing a first step
gals to protect their host against pathogens (Viaene towards evaluating the concept of plant-mediated
etal.2016). ‘antibioticproductionondemand’.
Actinobacteriacanmakeupasubstantialpartofthe
rootendophyticcommunityacrosstheplantkingdom,
which is largely determined by an increased relative Resultsanddiscussion
abundance of the family of Streptomycetaceae (Bul-
garellietal.2012;Edwardsetal.2015;Lundbergetal. Isolationofendophyticactinobacteria
2012). The composition of rhizospheric and root
endophytic bacterial communities is strongly influ- To confirm the presence of actinobacteria in the
encedbysoiltype(Bulgarellietal.2012)aswellasby endosphere, sterile A. thaliana Col-0 plants were
the plant genotype (Perez-Jaramillo et al. 2017). growninapottingsoil:sandmixturefor2 weeksunder
Therefore, part of the microbiome composition of controlled conditions, followed by harvesting roots
the rhizosphere and endosphere of Arabidopsis andshoot,surface-sterilizationandhomogenizationof
thaliana grown under controlled conditions in differ- theroottissueandplatingontovariousmediaselective
ent natural soils is conserved with specific bacterial foractinobacteria(Zhuetal.2014b).Tenmorpholog-
taxa including members of the actinobacteria (Bul- ically distinct actinobacterial isolates were obtained
garelli et al. 2012). Despite their occurrence in the fortheCol-0plants.Asimilarisolationapproachwas
endosphere, very little is known on how most adoptedforA.thalianaecotypemossel(Msl)obtained
actinobacteria colonise the endosphere and where from a natural ecosystem (Mossel, Veluwe, the
theyreside.AnexceptionisStreptomycesscabies,the Netherlands). Next to plating onto selective media
causalagentofpotatoscab,whichhasbeenstudiedin for the isolation of actinobacterial endophytes from
detail (Loria et al. 2006; Jourdan et al. 2016; Bignell Mslplants,totalDNAwasextractedineightreplicates
etal.2010;Bukhalidetal.1998). fromsamplesobtainedfromthesoil,therhizosphere,
Colonizationofrootandshoottissuebyactinobac- therootendosphereandtoothpicksinsertedinthesoil
teria is dependent, at least in part, on chemical cues were utilised as wood compartment (Bulgarelli et al.
from the plant as was shown for enrichment of 2012). These DNA samples were then analyzed by
123
AntonievanLeeuwenhoek(2018)111:679–690 681
16S rDNA-amplicon sequencing. Based on total analysed both belonged to the most abundant OTUs
sequencereadstheresultsofthelatteranalysisshowed StreptomycesclaviferandStreptomycesolivochromo-
thatactinobacteriarepresentedonaverage22%ofthe genes (Fig. 2a). Out of the 35 isolates, five were
total endophytic population (Fig. 1a). Among the closelyrelatedtoS.claviferandfourtoS.olivochro-
actinobacterial operational taxonomic units (OTUs), mogenes. Supportive evidence comes from several
the Streptomycetaceae and Micromonosporaceae independent studies, reporting S. olivochromogenes
were overrepresented. Additionally, 10% the reads fromtheendosphereofChinesecabbageroots,potato
oftheendophyticpopulationwasrepresentedbyonly tubers,medicinalplantsandpurplehenbit(Singhand
two OTUs (137 and 48), belonging to the family of Gaur2016;Leeetal.2008;Doumbouetal.1998;Kim
Streptomycetaceae(Fig. 2b).Thisenrichmentwasnot etal.2012).Additionally,strainssimilartoS.clavifer
as much observed for the soil, the rhizosphere and were previously isolated from sugarcane, lucerne
bacteria associated with a toothpick, suggesting a plants and wheat (Kruasuwan et al. 2017; Franco
species-specific selection among the root endophytic etal.2017;MiskandFranco2011).
actinobacteria of A. thaliana. Numbers on total Subsequent phenotyping of these endophytic acti-
sequencereadsandOTUsarelistedinTableS1. nobacterial isolates by high resolution imaging by
Isolation of actinobacteria from the endosphere of scanning electron microscopy (SEM) revealed spiny
A. thaliana ecotype Msl resulted in 35 morphologi- spores that are typical of S. clavifer (Goodfellow
callydifferentisolates.The34-465regionofthe16S 2012), supporting the 16S rDNA-based taxonomic
RNA gene of all isolates was sequenced and the classificationoftheisolatesasS.clavifer(Fig. 2b).In
closest hitsontheEzBioCloudwereconfirmedusing contrast,S.olivochromogenespoorlydevelopssporu-
the MLSA-based phylogeny published recently latingmyceliumaccordingtoBergey’smanual,which
(Labeda et al. 2017). Linking the OTUs detected by is shown as well for an isolate classified as S.
16S rDNA amplicon sequencing to a single Strepto- olivochromogenes (Fig. 2c). However, impaired
myces species was not always possible due to a sporulation was encountered frequently, a phenotype
relatively low resolution (Girard et al. 2013; Labeda thatwasdependentonthegrowthmedia,whichmakes
et al. 2017). Nevertheless, the two isolates that were this feature non-specific (see next paragraph). The
Fig.1 Relative abundance of the actinobacteria in the endo- community. b The enrichment of actinobacteria is predomi-
phyticcompartmentofA.thalianaMsl.aAmpliconsequencing natelydrivenbyStreptomycetaceaeOTU48and137.OTU48
data show that actinobacteria represent 22% of the total and137makeup10%ofthetotalECandalmost50%ofallthe
endophytic compartment (EC), while in soil and rhizosphere actinobacteriapresentintheEC.TheseOTUsarenotasmuch
(RS)communitiesandthoseinassociationwithatoothpick(TP) enrichedinthesamplesderivedfromnon-endophyticorigins
insertedinthesoil,theytakeuproughly2%ofthetotalbacterial
123
682 AntonievanLeeuwenhoek(2018)111:679–690
Fig.2 CharacterizationofStreptomycesendophytesandtheir species level based on the 16S rDNA sequence. b Scanning
taxonomic distribution. a Several isolates show highest simi- electron micrograph of Streptomyces sp. MOS18, which
laritywithStreptomycesclaviferorStreptomycesolivochromo- produces spiny spores. Scale bar 3lM. c Scanning electron
genesonthebasisof16SrDNAanalysis.Themajorityofthe micrograph of Streptomyces sp. MOS38 showing poor sporu-
isolates showed a wide variation in closest species assigned, lation.Scalebar5lM.StrainsweregrownonSFMmediafor
indicating a diverse endophytic compartment. ‘Unclassified’ 6days
means that these actinobacteria could not be classified at the
endophyticisolatesofA.thalianaecotypeMslshowed EndophyticcolonizationofArabidopsis
diverse colony morphologies (for examples see byStreptomyces
Fig. 3). Remarkably, 20 out of 35 isolates failed to
developonR5agarplates,suggestingtheyhadeither Wethenwantedtoknowifandwheretheisolatesenter
lost key sporulation genes or lacked the ability to therootendosphereofA.thalianatogetmoreinsight
develop in the absence of specific nutrients or trace into the yet elusive endophytic biology of Strepto-
elements. For example, R5 agar is known to lack myces. Therefore we inoculated spores of Strepto-
sufficient iron and copper, which is a major reason myces strain coa1, which was recruited by sterile A.
why bld mutants cannot sporulate on this media thaliana Col-0 plants grown in a potting soil:sand
(Keijser et al. 2000; Lambert et al. 2014). Addition- mixture, onto sterilised seeds of A. thaliana Col-0.
ally, several endophytes showed the propensity to Noteworthy,16SrDNAanalysisclusterscoa1intheS.
openuptheircolony,withtheirvegetativemycelium olivochromogenes branch based on phylogenetic
facing upward. This feature is well exemplified by relationships within the family of Streptomycetaceae
Streptomyces sp. MOS31 and MOS18 (Fig. 3h, c). (Labedaet al. 2017).Attachment of the sporesto the
MOS18 is taxonomically related to S. clavifer and seedswasconfirmedbySEM(Fig.S1).Seven-day-old
MOS31is the only endophyticisolate closely related seedlings grown from the treated seeds were stained
toStreptomycesbobili,a‘neighbour’ofS.olivochro- with propidium iodide and subjected to confocal
mogenes (Labeda et al. 2017). To study the morpho- fluorescencemicroscopy.Theresultsshowedthatthe
logical characteristicsofMOS31 at higher resolution endophyte had colonised both leaves and roots
SEM was applied, which revealed a thick sheet of (Fig. 5). Strain coa1 attached to the lateral roots in
hyphaethatturnedawayfromtheinsideofthecolony dense pellets, whereas colonization of the leaves
(Fig. 4a). Additionally, we observed hyphae extend- involved hyphal growth over the leaf surface. The
ing from these sheets, away from the vegetative colonised Arabidopsis roots were then fixed for
mycelium(Fig. 4b). sectioningandhighresolutionimagingwithtransmis-
sion electron microscopy (TEM). Regions of interest
were identified by obtaining 1-lm sections that were
123
AntonievanLeeuwenhoek(2018)111:679–690 683
Fig.3 Streptomyces endophytes display a wide range of DSM40236 and the endophytic streptomycetes MOS18 (c),
morphologies. Strains shown above are grown on SFM agar MOS38 (d), MOS14 (e), MOS32 (f), MOS25 (g), MOS31
platesfor6days.S.coelicolorM145andS.griseusDSM40236 (h)andMOS35(i).Scalebar2mm
areshownasreferencestrains.aS.coelicolorM145,bS.griseus
Fig.4 Scanning electron micrograph of Streptomyces sp. Scale bar 100lM. b Hyphae extending from the mycelial
MOS31. Images of mycelial sheets as is seen in Fig.3H. sheets, thereby growing away from the vegetative mycelium.
aThicklayersofvegetativemyceliummadeupofhyphaeand Scalebar30lM.Streptomycessp.MOS31wasgrownonSFM
extracellularmatrixturnawayfromtheinnerpartofthecolony. agarplatesfor6days
stained with toluidine to visualise the bacteria tissue and, remarkably, the intracellular space
(Fig. 6a). The results showed that coa1 not only (Fig. 6b, c). No plant cell defects were observed in
colonised the root surface but also the internal root the imaged samples. Strikingly, there is no plant
123
684 AntonievanLeeuwenhoek(2018)111:679–690
Fig.5 ColonizationofArabidopsisbyStreptomycesendophyte the root as a dense pellet. Scale bar 50lM. b Confocal
coa1. a Confocal micrograph of a colonised lateral root. The micrographoftheborderoftheleaf.Singlehyphaearegrowing
sample is stained with propidium iodide, resulting in red overtheleafsurface(arrowheads).Scalebar15lM
fluorescenceofbothbacterialandplantcells.Coa1attachesto
Fig.6 SectionsofArabidopsisrootsinvadedbyStreptomyces micrographs of Arabidopsis roots invaded by cao1. Coa1
coa1.aToluidinestainedsectionofanArabidopsisrootinvaded colonisestherootintracellularly.Thebacteriumcanbefound
byStreptomyces coa1.Coa1 enters theroot viathe epidermis inbetweentheCO-andEPcells,andinbetweentheendodermis
(EP) cells and colonises the root in between the cortex cells (ED) and CO cells (image B). In addition, in EP cells
(CO)andEPcells.Scalebar10lM.Boxedpartoftheimageis intracellulargrowthofStreptomycescoa1wasobserved(image
shownasmagnificationontheright.b,cTransmissionelectron C).Scalebar2lM
cellular membrane separating Streptomyces from the turnip rape and carrot by Streptomyces (Bulgarelli
intracellularspace. etal.2012;Bonaldietal.2015;Kortemaaetal.1994).
Previous studies showed colonization of the root In addition, endophytic colonization of germinating
surfaceofArabidopsisbyactinobacteria,andlettuce, wheat seed has been reported by reintroduction of a
123
AntonievanLeeuwenhoek(2018)111:679–690 685
GFP expressing Streptomyces strain (Franco et al. Genome sequencing revealed that actinobacteria
2017).Tothebestofourknowledge,ourresultsshow have a lot of biosynthetic gene clusters for natural
forthefirsttimethepresenceofastreptomycetewithin products that are poorly expressed (Kolter and van
an Arabidopsis root cell. Furthermore, the hyphae of Wezel 2016; Nett et al. 2009). This offers a vast
coa1 appeared less electron dense in the endosphere reservoirofpotentiallyimportantbioactivemolecules,
than on the plant surface, which may reflect physio- including antibiotics. To allow screening of these
logicaldifferencesbetweenlifeinsideandoutsidethe compounds, strategies are required to elicit their
plant. Still there are major gaps in our understanding expression (Rutledge and Challis 2015; van Wezel
of the endophytic biology of streptomycetes. For et al. 2009; Wu et al. 2015; Zhu et al. 2014a). We
example, we did not find reproductive structures or assessed the potential of phytohormones as elicitors,
spores within the endosphere. It therefore remains to thereby mimicking the chemical environment of the
beseenwhetherendophyteshaveacompletelifecycle rhizosphereandendosphere.Forthis,weexposedthe
inside plants oronly remain in the vegetative growth endophytestotheplanthormonessalicylicacid(SA),
phase, although we have to take into account that a indoleaceticacid(IAA,knownasauxin)orjasmonic
7 daytimeframemightnotbesufficienttoformspores acid(JA).Thestrainsweregrownonminimalmedia,
within a plant. It has also been suggested that the withorwithout0.01,0.001or0.0001%(w/v)ofeither
endophyticlifestyleofactinobacteriamayincludethe planthormoneasconcentrationsofSAcanbeashigh
formationofsmallprotoplast-likecellsthatlackacell as0.01%(w/w).Square12 9 12 cmagarplateswere
wall (Ramijan et al. 2016). These so-called L-forms inoculated with spots from spore stocks of the
may explain how the relatively large mycelial acti- endophytes. Interestingly, the percentage of strains
nobacteriacanstillmigrateandproliferateinsideplant exhibiting antibiotic activity roughly doubled, with
tissue. To further unravel the endophytic biology of 20% of the strains inhibiting E. coli cells, and 29%
Streptomyceswewillfocusfutureexperimentsonthe inhibitinggrowthofB.subtilis(TableS1).Thiseffect
identificationofmajorgeneticormorphologicaltraits couldmostlybeattributedtoIAA,whichhadamore
associatedwiththeendophyticlifestyle. significantelicitingeffectontheendophytesthanSA
orJA(TableS2).Anexampleofelicitedantimicrobial
Responseofendophytestophytohormones production byIAAis presentedin Fig.S2. Addition-
ally, we observed increased activity against the
Actinobacteriaplayanimportantroleinantibiosisand indicator strains as well, indicating either more
as probiotics to the plant due to the production of a production of the same antibiotic or production of a
diverse array of bioactive molecules (Viaene et al. different (set of) antibiotic(s) (See Fig. S3). After
2016). We assessed the antimicrobial activity of the elicitation, comparative metabolomics and/or geno-
actinobacterial endophytes in overlay assays, using mics on samples obtained from producing and non-
Bacillussubtilis(Gram-positive)andEscherichiacoli producing conditions can be used to identify the
(Gram-negative) as the target. By applying agar compound and gene cluster of interest, respectively
diffusion assays with the streptomycetes grown on (Gubbensetal.2014;Nguyenetal.2013).
minimalmediumagarwithmannitolandglycerolwe Although the concept of eliciting antimicrobial
showedthat11%oftheisolatesproducedoneormore production by actinobacteria is already well-estab-
compoundsthatinhibitedgrowthofE.colicells,while lished, this experimental setup was connected to the
14%inhibitedgrowthofB.subtilis.Generally,thevast biotic interactions between plant and streptomycete,
majority of actinomycetes isolated from soil samples aimedtomimicthechemicalenvironmentoftheplant.
are able to inhibit B. subtilis growth under routine Plant-endophyteinteractionshavelikelyplayedakey
laboratoryconditions,whichdiffersfromwhatwefind role in the evolution of the chemical diversity of
for our collection of endophytes (Zhu et al. 2014b). actinomycete-derived natural products and signals
While the small numbers make it difficult to apply thatcontroltheproductionoftheseantimicrobialsare
statistics,wecannotruleoutthatthereducedantibiotic likelytiedtothebioticinteractions.Thisideaisfurther
activity against Gram-positive bacteria and in partic- explored by the ‘‘cry for help’’ hypothesis, which
ular Firmicutes, may be typical of endophytic additionallystatesthatactinobacteriaencountertrade-
actinobacteria. offs between the costs of producing complex natural
123
686 AntonievanLeeuwenhoek(2018)111:679–690
productsandtheirbenefits,andmaythereforeproduce from the DSMZ culture collection (Braunschweig,
thesemoleculesspecificallyinresponsetoecological Germany). S. coelicolor M145 and S. griseus
demands (van der Meij et al. 2017). To broaden our DSM40236weregrownonSFMfor6 daysat30 (cid:3)C,
understandingofthisconceptitwouldbeofinterestto unless indicated differently. A sterile A. thaliana
know which type(s) of compounds are produced in ecotype Columbia (Col-0) was used for the recruit-
responsetophytohormoneexposure.Identificationof ment of endophytic actinobacteria under lab condi-
theantimicrobialcompoundsproducedinresponseto tions.Plantsweregrownonamixtureof9:1substrate
exposure to phytohormones and the corresponding soil and sand (Holland Potgrond) at 21 (cid:3)C, a 16 h
metabolic networks will give better insights into the photoperiod, and 70% relative humidity. After
bacterial responses to the plant’s ‘‘cry for help’’. In 2 weeks of growth, the plants were harvested. Soil
addition, the concentration-dependent effect of phy- for the field experiment with A. thaliana ecotype
tohormones may be of great importance as local Mossel(Msl)wascollectedinMay2016attheMossel
phytohormoneconcentrationsinaplantvary.Inorder area at the ‘Hoge Veluwe’ in the Netherlands (coor-
toharnessplant-actinomyceteinteractionsaselicitors dinates: N52(cid:3)03035.500E5(cid:3)45006.400), from four differ-
for antimicrobial production, further studies should ent spots within a radius of 100 m. If there was any
focus on these concentration-dependent effects, as vegetationpresent,thetop5–10 cmsoilwasremoved.
wellassynergismorantagonismofdifferentphythor- Thesoilwashomogenisedandalllargepartssuchas
moneswhenusedaselicitor. deadrootsandstoneswereremoved.Thesoilwaskept
inacoldroomat4 (cid:3)Cuntiluse.Seedsweresterilised
in 49 diluted household bleach for 10 min, washed
Conclusion seventimeswithsterileMQwater,ashortrinsewith
70%ethanolandtransferredtoplateswithawetfilter
Inthisstudy,weshowtherecruitmentofStreptomyces paper, placed at 4 (cid:3)C for 48 h and then moved to a
endophytesbyA.thalianaCol-0andA.thalianaMsl. 21 (cid:3)Cincubatorinthedark.Mosselsoilwasplacedin
Our pilot study shows that isolates falling within the atraywith3 9 3 cmpotsandwatered.Toremovethe
groups of S. olivochromogenes and S. clavifer are endogenousseedpopulation,thetraywasplacedinthe
overrepresented in the endosphere, suggesting that greenhouse for 2 days. After weeding, the sterile
these endophytic streptomycetes may have specific seedlings on the plates were transplanted to the tray
characteristics that allow them to adapt to life inside withMosselsoilandafter7 daystheplantsincluding
theplant.Thisneedstobestudiedinmoredetail,and the soil were planted into to the Mossel field. After
mayalsobeextendedtothestudyofotheractinobac- 6 weeks ofgrowthinthefield, plantswere harvested
terial genera found in the endosphere. Additionally, using a small shovel 3–4 cm around the base of the
weprovideafirststeptowardstheproofofconceptfor plant.
the‘cryforhelp’hypothesis,wherebyplanthormones,
in particular IAA, have a stimulating effect on Isolationofendophytes
antibiotic production by endophytic actinobacteria.
These baseline experiments highlight the importance A.thalianaCol-0wassurfacesterilisedbywashingthe
of exploring and exploiting plant-actinomycete inter- plant three timesin70%EtOH,after which theplant
actions as elicitors for ‘antibiotic production on material was crushed in MQ. Sterilization was
demand’. confirmed by adding a sterilised plant onto LB agar,
after which no bacterial growth was observed. Root
tissue of A. thaliana Msl was cleaned, sonicated and
Materialsandmethods ground with mortar and pestle in 1 mL phosphate
buffer(perlitre:6.33 gofNaH PO (cid:2)H O,10.96 gof
2 4 2
Bacteria,plantsandgrowthconditions Na HPO (cid:2)2H O and 200 lL Silwet L-77). Both the
2 4 2
crushed plant material (Col-0) and sonicated plant
Streptomyces coelicolor A3(2) M145 was obtained material(Msl)werespreadontothesurfaceofarange
from the John Innes Centre strain collection in ofselectiveisolationmediaincludinghumicacidagar
Norwich, UK, and Streptomyces griseus DSM40236 (HA) (Hayakawa and Nonomura 1987), glucose
123
AntonievanLeeuwenhoek(2018)111:679–690 687
casein agar (GCA) (Zhang 1985), soy flour mannitol mitochondial and chloroplast sequences were
medium (SFM) or minimal medium (MM) (Kieser removed, as were the OTUs that did not have 25
et al. 2000). Initial selection was done on media reads in at least 5 samples (‘‘rare taxa’’). To obtain
supplemented with the antifungal agent nystatin relativeabundanceoftheOTUs,thenumberofreads
(50 lg/mL) and the antibacterial agent nalidixic acid fromasingleOTUpersamplewasdividedbythetotal
(10 lg/mL). Plates were incubated at 30 (cid:3)C for numberofreadsofthatsampleafterfiltering forrare
4–25 days. taxa.
PlantharvestingandDNAisolation Analysisofactinobacteriabasedon16SrRNA
ofthemicrobialcommunity sequences
ForA.thalianaMslweappliedtheharvestingprotocol The 16S rRNA genes of the actinobacteria were
as described before (Lundberg et al. 2012). In short, amplified by PCR from liquid-grown mycelia using
roots including rhizospheric soil were collected in a primers F1 (50-GCGTGCTTAACACATGCAAG-30)
50 mL tube containing 25 mL of phosphate buffer, and R1 (50-CGTATTACCGCGGCTGC T G-30),
and vortexed for 15 s. Replacing the buffer and which correspond to nt positions 15-34 and 465-483
vortexing was repeated until the buffer stayed clear. of the 16S rRNA locus of S. coelicolor (van Wezel
Roots were transferred toa 15 mL tube, sonicated(5 et al. 1991), respectively. PCRs were conducted as
bursts of 30 s with 30 s breaks), vortexed, washed described (Colson et al. 2007) and sequenced using
withphosphatebufferanddriedonfilterpaper.Then, oligonucleotide F1. Sequencing was done at Base-
therootswereeithergroundforbacterialisolation,or Clear in Leiden, the Netherlands. 16S rRNA gene
flash frozen and stored at - 80 (cid:3)C for later DNA analysiswasperformedusingwebbasedidentifytool
isolation.Fourindividualplantswerepooledintoone on EzBioCloud (https://www.ezbiocloud.net/
sample. Untreated toothpicks were inserted into the identify). The identify service provides proven simi-
Mslsoilataminimaldepthof4 cmandwereutilised larity-based searches against quality-controlled data-
as wood compartment. DNA was isolated from soil basesof16SrRNAsequences.Thetop-hitinformation
usingtheMoBioPowerSoilkitandECandtoothpick for each Identify Job was checked against Strepto-
sampleswith theMPBioFast DNAspinkit.Quality myces focused MLSA based phylogenetic tree pub-
and quantity of the DNA was checked by nanodrop lishedelsewhere(Labedaetal.2017).
and gel electrophoresis. Around 400 ng was sent for
16 s rDNA sequencing at Beijing Genomics Institute Microscopy
(BGI).
Lightmicroscopy
Ampliconsequencing
StereomicroscopywasdoneusingaZeissLumarV12
Using primers 515F (50-GTGYCAGCMGCCGCGG- microscope equipped with a AxioCam MRc, and
TAA-30) and 806R (50-GGACTACNVGGGTWTC- confocal microscopy using a Zeiss Observer Micro-
TAAT-30),theV4regionwassequencedatBGIonthe scope. For confocal microscopy, spores of the endo-
HiSeq 2500sequencingplatform(Illumina).Rawdata phyte were added to the seeds. Seeds were put on (cid:2)
from BGI was processed using a previously reported Murashige and Skoog medium with 1% sucrose and
custom implementation (Perez-Jaramillo et al. 2017) 0.8%agarandkeptinthedarkat4(cid:3)for48 h.Afterthe
of QIIME (Caporaso et al. 2010) with minor modifi- coldshock,seedswereincubatedintheclimateroom
cations(describedbySchneijderbergetal.,inprep).In for 1 week (see bacteria, plants and growth condi-
short, reads were quality filtered and filtered for tions) after which they were imaged. Samples were
chimerasusingChimeraSlayer.Using a97%identity stained with propidium iodide 1:1000 (1 lg/mL).
threshold, de novo OTUs were determined, which Sampleswereexcitedwithlaserlightatawavelength
weretaxonomicallyassignedusingtheRDPclassifier 535 nmtodetectthepropidiumiodide.
2.10(Coleetal.2014)withtheGreenGenesdatabase
28 (DeSantis et al. 2006). OTUs related to
123
688 AntonievanLeeuwenhoek(2018)111:679–690
Electronmicroscopy Streptomyces grow equally well on either mannitol
or glycerol. The agar plates were supplemented with
Morphological studies on single colonies of endo- either(±)-jasmonicacid(Caymanchemicalcompany,
phytes by SEM were performed using a JEOL cas: 88-30-0), 3-indoleacetic acid (Sigma-Aldrich,
JSM6700F scanning electron microscope. For Strep- cas: 87-51-4) or salicyclic acid (Alfa Aesar, cas:
tomycete only samples, pieces of agar with biomass 69-72-7).Theendophytesweretypicallyincubatedfor
from6-day-oldcoloniesgrownonSFMwerecutand 5 days at 30 (cid:3)C, following which they were overlaid
fixed with 1.5% glutaraldehyde (1 h). Subsequently, withLBsoftagar(0.6%w/vagar)containing300 lL
samplesweredehydrated(70%acetone15 min,80% ofoneoftheindicatorstrains(OD0.4–0.6),andthen
acetone 15 min, 90% acetone 15 min, 100% acetone incubated overnight at 37 (cid:3)C. The following day,
15 min and critical point dried (Baltec CPD-030). antibacterialactivitywasdeterminedbymeasuringthe
Hereafter the samples were coated with gold using a inhibition zones (mm) of the indicator strain sur-
goldsputtercoater,anddirectlyimagedusingaJEOL roundingthecolonies.
JSM6700F.ForSEMofArabidopissamples,imaging
timing was increased four-fold to optimise fixation Acknowledgements This work was supported by a Grants
14218 and 14221 from the Netherlands Organization for
and dehydration, and the 70%acetone step was done
Scientificresearch(NWO)toJRandGPvW,respectively.
overnight.
Transmission electron microscopy (TEM) for the Conflict of interest The authors declare no conflict of
analysis of cross-sections of Arabidopsis roots and interests.
StreptomyceswasperformedwithaJEOL1010trans-
OpenAccess Thisarticleisdistributedunderthetermsofthe
mission electron microscope as described previously
CreativeCommonsAttribution4.0InternationalLicense(http://
(Pietteetal.2005). creativecommons.org/licenses/by/4.0/), which permits unre-
Samples were grown in the same way as for the stricted use, distribution, and reproduction in any medium,
lightmicroscopy,andfixedwith1.5%glutaraldehyde provided you give appropriate credit to the original
author(s)andthesource, providealinktotheCreativeCom-
for4 h.Post-fixationwasperformedwith1%Osmium
monslicense,andindicateifchangesweremade.
tetroxidefor4 h.Initialdehydrationwith70%ethanol
was done overnight. Hereafter using a high magnifi-
cation (9150) stereo microscope (MZ16AF) root
sections containing Streptomyces growth were
References
selected, followed by dehydration in 1 h steps (80,
90, 100% ethanol, 100% propylene oxide, 50/50
Anne´J,VanMellaertL,EyssenH(1990)Optimumconditions
propylene oxide/EPON, 100% EPON). Subsequently
for efficient transformation of Streptomyces venezuelae
sampleswereembeddedinEPONandpolymerisedfor protoplasts.ApplMicrobiolBiotechnol32:431–435
2 days at 60 (cid:3)C. Ultrathin sections were cut on an BarkaEA,VatsaP,SanchezL,Gavaut-VaillantN,JacquardC,
KlenkHP,Cle´mentC,OudouchY,vanWezelGP(2016)
ultramicrotome (Reichert Ultracut E), collected on
Taxonomy,physiology,andnaturalproductsoftheActi-
copper grids and examined using a Jeol1010 trans-
nobacteria.MicrobiolMolBiolRev80:1–43
missionelectronmicroscopeat70 kV. Be´rdy J (2005) Bioactive microbial metabolites. J Antibiot
(Tokyo)58:1–26
Bignell DRD, Seipke RF, Huguet-Tapia JC, Chambers AH,
Antimicrobialactivityassays
Parry RJ, Loria R (2010) Streptomyces scabies 87-22
contains a coronafacic acid-like biosynthetic cluster that
Antimicrobial activity of the endophytes was tested contributes to plant-microbe interactions. Mol Plant-Mi-
againstB.subtilis168oraderivativeofE.coliAS19- crobeInteract23:161–175
BonaldiM,ChenX,KunovaA,PizzattiC,SaracchiM,Cortesi
RrmA (Liu and Douthwaite 2002). Indicator bacteria
P(2015)Colonizationoflettucerhizosphereandrootsby
were cultured in LB broth and incubated at 37 (cid:3)C
taggedStreptomyces.FrontMicrobiol6:25
overnight.Antimicrobialassayswereconductedusing BukhalidRA,ChungSY,LoriaR(1998)nec1,ageneconfer-
the double-layer agar method. Briefly, actinobacteria ringanecrogenicphenotype,isconservedinplant-patho-
genic Streptomyces spp. and linked to a transposase
were inoculated on minimal medium agar plates
pseudogene.MolPlant-MicrobeInteract11:960–967
containing both mannitol and glycerol (1% w/v) as
BulgarelliD,RottM,SchlaeppiK,LorenVer,vanThemaatE,
non-repressing carbon sources, since not all AhmadinejadN,AssenzaF,RaufP,HuettelB,Reinhardt
123