Gutell 110.ant.v.leeuwenhoek.2010.98.195


Published on

Mueller U.G., Ishak H., Lee J.C., Sen R., and Gutell R.R. (2010).
Placement of attine ant-associated Pseudonocardia in a global phylogeny (Pseudonocardiaceae, Actinomycetales): a test of two symbiont-association models.
Antonie van Leeuwenhoek International Journal of General and Molecular Microbiology, 98(2):195-212.

Published in: Technology
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Gutell 110.ant.v.leeuwenhoek.2010.98.195

  1. 1. ORIGINAL PAPERPlacement of attine ant-associated Pseudonocardiain a global Pseudonocardia phylogeny (Pseudonocardiaceae,Actinomycetales): a test of two symbiont-association modelsUlrich G. Mueller • Heather Ishak • Jung C. Lee •Ruchira Sen • Robin R. GutellReceived: 23 December 2009 / Accepted: 3 March 2010 / Published online: 24 March 2010Ó Springer Science+Business Media B.V. 2010Abstract We reconstruct the phylogenetic relation-ships within the bacterial genus Pseudonocardia toevaluate two models explaining how and why Pseud-onocardia bacteria colonize the microbial communi-ties on the integument of fungus-gardening ant species(Attini, Formicidae). The traditional Coevolution-Codivergence model views the integument-colonizingPseudonocardia as mutualistic microbes that arelargely vertically transmitted between ant generationsand that supply antibiotics that specifically suppressthe garden pathogen Escovopsis. The more recentAcquisition model views Pseudonocardia as part of alarger integumental microbe community that fre-quently colonizes the ant integument from environ-mental sources (e.g., soil, plant material). Under thislatter model, ant-associated Pseudonocardia mayhave diverse ecological roles on the ant integument(possibly ranging from pathogenic, to commensal, tomutualistic) and are not necessarily related toEscovopsis suppression. We test distinct predictionsof these two models regarding the phylogeneticproximity of ant-associated and environmentalPseudonocardia. We amassed 16S-rRNA genesequence information for 87 attine-associated and238 environmental Pseudonocardia, aligned thesequences with the help of RNA secondary structuremodeling, and reconstructed phylogenetic relation-ships using a maximum-likelihood approach. Wepresent 16S-rRNA secondary structure models ofrepresentative Pseudonocardia species to improvesequence alignments and identify sequencing errors.Our phylogenetic analyses reveal close affinities andeven identical sequence matches between environ-mental Pseudonocardia and ant-associated Pseud-onocardia, as well as nesting of environmentalPseudonocardia in subgroups that were previouslythought to be specialized to associate only with attineants. The great majority of ant-associated Pseudono-cardia are closely related to autotrophic Pseudono-cardia and are placed in a large subgroup ofPseudonocardia that is known essentially only fromcultured isolates (rather than cloned 16S sequences).The preponderance of the known ant-associatedPseudonocardia in this latter clade of culturablelineages may not necessarily reflect abundance ofthese Pseudonocardia types on the ants, but isolationElectronic supplementary material The online version ofthis article (doi:10.1007/s10482-010-9427-3) containssupplementary material, which is available to authorized users.U. G. Mueller (&) Á H. Ishak Á J. C. Lee ÁR. Sen Á R. R. GutellSection of Integrative Biology, University of Texas atAustin, Austin, TX 78712, USAe-mail: umueller@mail.utexas.eduJ. C. Lee Á R. R. GutellCenter for Computational Biology and Bioinformatics,University of Texas at Austin, Austin, TX 78712, USAU. G. Mueller Á J. C. Lee Á R. R. GutellInstitute for Cellular and Molecular Biology, Universityof Texas at Austin, Austin, TX 78712, USA123Antonie van Leeuwenhoek (2010) 98:195–212DOI 10.1007/s10482-010-9427-3
  2. 2. biases when screening for Pseudonocardia (e.g.,preferential isolation of autotrophic Pseudonocardiawith minimum-nutrient media). The accumulatedphylogenetic patterns and the possibility of isolationbiases in previous work further erode support for thetraditional Coevolution-Codivergence model andcalls for continued revision of our understandinghow and why Pseudonocardia colonize the microbialcommunities on the integument of fungus-gardeningant species.Keywords Attine ant-microbe symbiosis ÁMutualism Á Antibiotic Á Secondary rRNA structureIntroductionFungus-growing ants (Attini, Formicidae) cultivate abasidiomycete fungus in gardens that the ant nurturewith dead or fresh plant substrate. The main foodsource for the ants is the cultivated fungus, which isthought to comprise the main living biomass in attinegardens (Mueller et al. 2005). The gardens areactually mini-ecosystems of competing, commensal,and mutualistic microbes, including a diverse assem-bly of filamentous fungi, yeasts, and bacteria (Bacciet al. 1995; Carreiro et al. 1997; Currie et al. 1999a;Santos et al. 2004; Mueller et al. 2005; Rodrigueset al. 2008; Rodrigues et al. 2009). The interactionnetwork of these microbes is engineered in part bythe physical manipulations and secretions of the ants(Santos et al. 2004; Mueller et al. 2005). Significantpopulations of microbes, including actinomycetes,also colonize the integument and the alimentary canalof the farming ants (the ant microbiome), as isprobably true for most arthropods (Goodfellow andWilliams 1983; Dattagupta et al. 2009 and referencestherein). A diversity of integumental actinomycetebacteria can be readily isolated from most attine antswith minimum-nutrient media favoring autotrophicbacterial growth, including bacteria in the genusPseudonocardia (Cafaro and Currie 2005; Currieet al. 2006; Mueller et al. 2008; Ferna´ndez-Marı´net al. 2009; Sen et al. 2009). As an extreme case,workers of the attine ant Mycocepurus smithii can becolonized by significant populations of several spe-cies of Pseudonocardia, one species of the closelyrelated genus Amycolatopsis, and many otheractinomycete and non-actinomycete bacteria (Senet al. 2009).At the time of the discovery of actinomycetegrowth on the attine integument 10 years ago, it wasthought that the ants promote the growth of antibi-otic-secreting actinomycetes on their integument as aspecific defense to suppress the garden parasiteEscovopsis (Hypocreales, Ascomycota) (Currieet al. 1999a, b; Currie 2001). The almost simulta-neous discovery of the parasite Escovopsis and theintegumental actinomycete growth led to the imme-diate speculation that Escovopsis and the growth mustsomehow be ecologically linked. Specifically, Currieet al. (1999b) and Currie (2001) argued that attineants sustain growth of the actinomycete Pseudono-cardia as mutualists on their integument to specifi-cally combat Escovopsis parasites; that themutualistic Pseudonocardia have been engaged in aco-evolutionary arms race with Escovopsis since theorigin of the attine ant-fungus mutualism 50 millionyears ago; and that Escovopsis largely failed toevolve effective resistance against Pseudonocardiabecause of some unknown disadvantage in the co-evolutionary arms race (see also recent reviewsdiscussing the Escovopsis-Pseudonocardia arms race;Suen and Currie 2008; Poulsen et al. 2009).The original report and a string of follow-up studiesseemed to support this ant-Escovopsis-Pseudonocar-dia co-evolution model, including (a) derived cuticularstructures that seem designed to facilitate actinomy-cete colonization of the worker integument (Currieet al. 2006); (b) glandular structures underlying areasof significant actinomycete accumulation on theintegument, suggesting that the ants may manipulatemicrobial growth through secretions (Currie et al.2006); (c) greater susceptibility of gardens to Escov-opsis infection if the integumental accretions arescraped off from workers (a procedure also likely tostress or injure ants) (Currie et al. 2003); (d) apparentclade-to-clade correspondences between ant, Escov-opsis, and Pseudonocardia phylogenies, supportingthe possibility of tight co-diversification in an ancientant-Pseudonocardia association (Cafaro and Currie2005; Poulsen et al. 2009); (e) marginal (but notstrong) inhibition of Escovopsis by a peptide secretedin high concentrations by a Pseudonocardia isolate invitro (Oh et al. 2009); (f) a growth-enhancing effect onthe fungal cultivar by an unidentified actinomyceteisolated from an ant integument (Currie et al. 1999b);196 Antonie van Leeuwenhoek (2010) 98:195–212123
  3. 3. and (g) specific antibiosis of a single Pseudonocardiaisolate against Escovopsis but not against 17 non-Escovopsis test fungi (Currie et al. 1999b). Takentogether, the accumulating evidence seemed to supportant-Escovopsis-Pseudonocardia co-evolution and thepossibility of a 50-million-year-old Pseudonocardia-Escovopsis arms race.The image of co-evolved, pesticide-secreting, in-tegumental microbes of attine ants seemed so con-vincing that the National Science Teacher Association(NSTA) and several science museums in the USAbegan promoting the attine-Pseudonocardia-Escov-opsis association as an indisputable example of co-evolution (Diamond 2006). To visualize the evidenceof co-evolved, specific antibiosis against Escovopsis, ahands-on exercise developed by NSTA asks biologystudents to measure growth of two garden pathogens,Escovopsis and Trichoderma, growing under theinfluence of Pseudonocardia secretions. Because thegrowth tracks on paper show that Pseudonocardiastrongly inhibits Escovopsis but not Trichoderma (afinding originally reported in Currie et al. 1999b),Pseudonocardia apparently exhibits specialized anti-biosis against Escovopsis, because, if Pseudonocardia‘‘developed a substance that would kill only [Escov-opsis], then that would support the idea that the twoorganisms had coevolved’’ (Diamond 2006, page 95;see series of recent studies have eroded much of thekey evidence that was thought to support the originalformulation of specific antibiosis and ant-Pseudono-cardia-Escovopsis co-evolution. For example, Kostet al. (2007) and Sen et al. (2009) showed thatPseudonocardia from the integument of attine antshave generalized, broad-spectrum antibiotic activi-ties, comparable to the activities of environmental(free-living) actinomycete bacteria, and that thereforethe attine-associated Pseudonocardia are not antibi-oticially specialized and evolutionarily derived aspostulated by Currie et al. (1999b). Moreover,antibiotics secreted in vitro by ant-associated Pseud-onocardia strongly suppress or kill the cultivatedfungi from the corresponding nests (Sen et al. 2009),contrary to the original report of a growth-enhancingeffect of Pseudonocardia on the cultivated fungi(Currie et al. 1999b). Even more unfortunate was agreat delay before other research groups couldreplicate some of the original experiments. Becauseof a regrettable oversight, the original report did notspecify the microbial methods for isolating Pseud-onocardia from attine ants (Currie et al. 1999b), andit was not until 2005, with the first publications of theexact methods to isolate Pseudonocardia from ants(Cafaro and Currie 2005), that other research groupsbegan to evaluate the earlier findings. Reviewing thehistory of ant-actinomycete research in a recentcommentary, Boomsma and Aanen (2009) concludedthat there exists a need for more careful hypothesistesting to prevent oversimplified evolutionary con-clusions, and that a key message from ten years ofant-Pseudonocardia research may be the reminder toscientists at large ‘‘to continuously re-evaluate whatwe know for a fact and what we merely infer’’.Following Boomsma and Aanen’s (2009) call formore careful hypothesis testing to avoid prematureconclusions of adaptive and co-evolved antibioticdesign (sensu Gould and Lewontin 1979), we expandhere on a test of phylogenetic predictions that was firstexplored by Mueller et al. (2008), testing for closephylogenetic affinities between ant-associated andenvironmental Pseudonocardia lineages. This sametest can now be applied to a much larger dataset,because four times the sequence information hasbecome available at GenBank for Pseudonocardia.Our test specifically juxtaposes critical predictionsof the traditional ant-Pseudonocardia-EscovopsisCoevolution-Codivergence model (Currie et al.1999b; Poulsen et al. 2009) and predictions of thealternate Acquisition model (Kost et al. 2007; Muelleret al. 2008). These two models represent extremeviewpoints, but they make different predictionsregarding the phylogenetic proximity of ant-associ-ated and environmental Pseudonocardia.Predictions of the traditional Coevolution-Codivergence modelThe Coevolution-Codivergence model assumes thatthe diversity of attine ants associate only with alimited number of Pseudonocardia lineages, and thatthese few lineages have co-diverged with correspond-ing ant lineages during a long association spanningmillions of years (Cafaro and Currie 2005; Suen andCurrie 2008; Poulsen et al. 2007; Poulsen et al. 2009).These predictions rest on two main assumptions; firstthat each attine colony associates with only a singlePseudonocardia strain; and second, that Pseudono-cardia lineages are largely vertically transmittedAntonie van Leeuwenhoek (2010) 98:195–212 197123
  4. 4. across many ant generations, punctuated by occa-sional horizontal transfer between ant lineages, andpunctuated very rarely by de novo acquisition ofPseudonocardia from environmental sources (Cafaroand Currie 2005; Poulsen et al. 2007; Poulsenet al. 2009; Caldera et al. 2009). The traditionalCoevolution-Codivergence model emphasizes rareacquisition from environmental sources, long-termvertical transmission of Pseudonocardia associatesacross many ant generations, and strong co-evolu-tionary interactions between Pseudonocardia andEscovopsis.Prediction of the alternate Acquisition modelThe Acquisition model assumes continual coloniza-tion of the attine integument by diverse microbes(including Pseudonocardia), microbial turnover onthe integument, and regular loss (purging) of someintegumental microbes. In essence, the Acquisitionmodel assumes a microbiome on the ant integumentthat may parallel the complex surface microbiome ofother arthropods (e.g., Dattagupta et al. 2009 andreferences therein). Specifically, the Acquisitionmodel assumes that (i) a great diversity of microbesinhabits the ant integument; (ii) specific microbescolonize the ants predictably because the antsencounter these microbes inevitably in their particu-lar environment or in their interactions with thegardening substrate; (iii) some of these integumentalmicrobes may be transmitted through foundressqueens between generations, but long-term verticaltransmission across many ant generations is limitedbecause of continual microbial turnover on the antintegument; (iv) not all Pseudonocardia strains on theant integument are mutualists (some may be com-mensal, some may even be pathogenic); (v) antibioticeffects against Escovopsis may be incidental byprod-ucts—but not necessarily co-evolved adaptations—ofmicrobial evolution occurring before colonization ofthe ant integument or occurring in situ on theintegument under microbe-microbe competition forlimited resources (Kost et al. 2007; Mueller et al.2008; Sen et al. 2009). The Acquisition modelemphasizes frequent acquisition of Pseudonocardiafrom environmental sources, limited vertical trans-mission of Pseudonocardia associates across manyant generations, and low potential or absence ofPseudonocardia-Escovopsis co-evolution.Using 16S rRNA gene sequence information from45 ant-associated and 41 environmental Pseudono-cardia, Mueller et al. (2008) recently began toreconstruct the phylogenetic relationships betweenthese two types of Pseudonocardia within the genus.This previous study identified a number of ant-associated Pseudonocardia that are sequence-identi-cal at the 16S gene to environmental Pseudonocardia,and documented that well-supported clades includeclosely related environmental and ant-associatedPseudonocardia that differ in only a few base pairs(Fig. 1). However, phylogenetic analyses alsorevealed some Pseudonocardia clades that appearedderived and that consisted exclusively of Pseudono-cardia isolated from ants (Fig. 1). These derivedPseudonocardia clades could be ant-specific eitherbecause (a) these Pseudonocardia were ancientlyacquired by the ants from environmental sources andsince then had been exclusively associated with attineants; or (b) the clades appear only ant-specificbecause free-living Pseudonocardia are undersam-pled. This later explanation predicts that morecomprehensive sampling of free-living Pseudonocar-dia will eventually show that unknown environmentalstrains exist that are nested within the putative ant-specific clades or that are very closely related tothem. Because over 500 Pseudonocardia sequenceshave accumulated to date at GenBank (compared to85 sequences in the analysis of Mueller et al. 2008),the above predictions of the Acquisition model cannow be tested with a more comprehensive samplingof Pseudonocardia diversity. Additionally, we showFig. 1 Phylogeny of Pseudonocardia reconstructed by Muel-ler et al. (2008) from 16S rRNA gene sequence information.This reconstruction was based on a limited sample of 85Pseudonocardia sequences available at GenBank in September2007 and excluded sequence information in regions ofunambiguous alignment (i.e., the reconstruction was not basedon the 16S secondary-structure modeling employed in thepresent study). The clade labels on the far right correspond tothe labels in Fig. 2a and b of the present study (Clade 1&2 wascalled ‘‘compacta-group’’ and Clade 3 was called ‘‘alni-group’’in Mueller et al. 2008). This previous analysis revealed closeaffinities between many ant-associated (shaded) and environ-mental Pseudonocardia (unshaded), and predicted that futuresurveys will discover environmental Pseudonocardia nestingin groups appearing ‘‘ant-specific’’ at that time. This predictionis tested in the phylogenetic reconstruction shown in Fig. 2aand b, which is based on four times the sequence informationthat has become available for Pseudonocardia since the studyby Mueller et al. (2008)c198 Antonie van Leeuwenhoek (2010) 98:195–212123
  5. 5. Pseudon.sp.AY944253.(N16.marine.sponge.China.Sea)Pseudon.sp.DQ144225.(HPA192.marine.sponge.China.Sea)Pseudon.carboxydovorans.sp.nov.EF114314.(SWP-2006)Pseudon.sp.AY974793.(AB632.marine.sample.Mariana.Trench)Pseudon.sp.AY177725.(T4.marine.sample.Wadden.Sea)Pseudon.sp.DQ856329.(Dw)Pseudon.sp.EF588207.(Atta.colombica.CC031208-10.worker)Pseudon.sp.AY996838.(80821)Pseudon.sp.AB354140.(1a-G6.soil.Japan)Pseudon.sp.EF588229.(Trachy.zeteki.AL030107-17.worker)Pseudon.sp.EF588219.(Acro.sp.SES030406-01.worker)Pseudon.sp.EF588221.(Atta.sp.SP030328-02.worker)Pseudon.sp.EF588216.(Trachy.sp.NMG030609-02.worker)Pseudon.sp.AY376890.(Acro.sp.AcroEC.worker)Pseudon.sp.EF588212.(Acro.sp.CC030402-02.worker)Pseudon.sp.EU135642.(YIM-C680.soil.China)Pseudon.sp.EF588222.(Acro.sp.SP030405-01.worker)Pseudon.sp.AY376888.(Acro.octospinosus.AcroPA.worker)Pseudon.sp.EF588225.(Acro.octospinosus.CC031210-22.worker)Pseudon.sp.EF588210.(Acro.octospinosus.CC011128-01A.worker)EU718349.(Trachy.arizonensis.UGM010817-01.D2.worker)EU718348.(Trachy.arizonensis.UGM010817-01.D10.larva)Pseudon.sp.AY376889.(Acro.sp.A30.worker)AF491899.(Acro.octospinosus.symbiont3.worker)Pseudon.sp.EF588209.(Acro.echinatior.AL040118-01.worker)Pseudon.sp.EF588206.(Acro.echinatior.CC031212-01.worker)Pseudon.antarctica.AJ576010.(DVS5a1.soil.Antarctica)Pseudon.sp.AM403499.(BHF008.industrial.sludge.France)Pseudon.sp.EF588214.(Acro.sp.CC030405-05.worker)Pseudon.alni.AJ252823.(IMSNU20049T.root.nodule.Russia)Pseudon.alni.Y08535.(DSM44104.root.nodule)EU718350.(Trachy.desertorum.UGM061207-01.G3.DD.worker)Pseudon.fici.EU200678.(YIM56250.endophyte.China)EU718351.( 0.001 substitutions/siteClade3Clades1&2Clade4Clades6-986608158996583989994Antonie van Leeuwenhoek (2010) 98:195–212 199123
  6. 6. here how 16S rRNA secondary structure modeling ofPseudonocardia can identify sequencing errors andhelp resolve ambiguities in alignments for phyloge-netic analyses.MethodsUtility of the 16S rRNA geneOur phylogenetic analysis takes advantage of theexponentially increasing number of 16S rRNA genesequences that have become available at the NCBIGenBank ( forreconstruction of phylogenetic relationships. Some ofthese 16S sequences were generated for taxonomicdescriptions of Pseudonocardia species (summarizedin Huang and Goodfellow 2011), but the majority ofthe Pseudonocardia sequences derived from environ-mental surveys characterizing general bacterialdiversity with either culture-dependent methods (e.g.,specialized isolation media) or culture-independentmethods (e.g., cloning and sequencing of 16S ampli-cons; direct pyrosequencing of 16S amplicons). Reli-ance on only the 16S gene in our phylogenetic analyseshas disadvantages because this single gene does notfully resolve all relationships between closely relatedtaxa, because bacterial genomes can contain several16S genes that differ minimally in sequence (althoughthis is currently unknown for Pseudonocardia),because of potential artifacts in multitemplate PCR onwhole microbial communities in environmental sur-veys (Acinas et al. 2005), and because of the possi-bility of lateral gene transfer among bacteria (althoughlateral gene transfer is thought to be very unlikely foran isolated 16S gene because of the integrated, centralrole of this gene within a cell’s genomic network;Woese 2000). Despite these potential disadvantages,our 16S-based approach is justified for three reasons.First, the amount of 16S information available fromenvironmental surveys exceeds by several orders ofmagnitude that of any other gene, and the most com-prehensive sample on Pseudonocardia diversity fromdiverse sources is critical to address close phyloge-netic affinities between environmental and ant-asso-ciated Pseudonocardia lineages. Second, the mainsubgroups of Pseudonocardia identified in our phy-logenetic reconstruction using only 16S informationessentially matches the groupings inferred also by fourprotein-coding housekeeping genes also available atGenBank (UG Mueller, unpublished), showing that16S and these four housekeeping genes have recordedparallel evolutionary histories that have not been ero-ded by lateral gene transfer. Third, our analysis doesnot aim to resolve relationship at all levels between alllineages of Pseudonocardia, but aims to identify clo-sely-related lineages (i.e., we are less interested inresolving basal relationships between major subgroupswithin the genus) and, most importantly, aims to testfor nesting of previously unknown environmentallineages within known clades of Pseudonocardia thatcurrently are believed to be specialized to associateonly with attine ants.Selection of sequence informationTo compile a comprehensive sample of 16S sequencesfor Pseudonocardia, we obtained all 456 sequenceslisted under Pseudonocardia by 1. November 2009 atthe Ribosomal Database Project (RDP; Center forMicrobial Ecology, Michigan State University; Thislist included all taxa classified under Pseudonocardiaat the NCBI Taxonomy Browser (sequences obtainedfrom isolated cultures;, as wellas additional Pseudonocardia sequences from envi-ronmental surveys deposited at the NCBI CoreNucleotide Database (CND; To find additional sequen-ces not listed at the RDP (e.g., very recently depositedsequences at CND), we also searched the CNDby BLASTing full sequence information ([1400 bp)of Pseudonocardia carboxydivorans, callicarpae,ammonioxydans, halophobica, spinosispora, petr-oleophila, thermophila, and zijingensis. These Pseud-onocardia species were chosen because they span thephylogenetic diversity of the genus, and because theyare closely related to Pseudonocardia lineages previ-ously isolated from attine ants (Mueller et al. 2008).Because BLASTing of longer sequences can missshorter sequences deposited at the CND, we alsoBLASTed, for each of these eight species, 16S-sequence segments of about 300–500 bp length thatspanned one or two of the V2, V3, V4, or V9 regions ofthe 16S gene. These regions were chosen because theyrepresent the most variable regions of 16S within thegenus Pseudonocardia. We added to our dataset two200 Antonie van Leeuwenhoek (2010) 98:195–212123
  7. 7. sequences (GU318369, GU318370) of Pseudonocardiaisolated in an ongoing survey of actinomycete diversityin nests of the attine ant Trachymyrmex septen-trionalis (Miller, Sen, Ishak, Mueller unpublished), and16S sequences for P. compacta (NRRL B-16170;GU318372) and P. nitrificans (NRRL B-1664;GU318371; strains courtesy of David Labeda at theARS Culture and Patent Culture Collections, UnitedStates Department of Agriculture, Peoria, Illinois). Thissearch strategy accumulated a total of 519 sequences.We excluded from this dataset those sequencesactually belonging to other pseudonocardiaceousgenera (e.g., some sequences currently listed at theRDP under Pseudonocardia actually belong intoCrossiella, such as NCBI accessions FN297974,EU299989, FN297967, EF447067, EF447073,EF447078, EF188338, EF447079; sequenceAJ871440 belongs into the genus Actinostreptospora;etc.), and we excluded poor-quality sequencesflagged as such at the RDP. Second, to ensuresufficient overlap across phylogenetically informativeregions of the 16S gene, we excluded all sequencesshorter than 500 bp in length. This criterion of500 bp minimum length also excluded the shortsequences generated with next-generation sequencingplatforms (e.g., 454 pyrosequencing) that are cur-rently deposited at the NCBI Short Read Archive( Third,we excluded all sequences that did not span both theV3 and the V4 regions. These two variable regionswere targeted for three reasons: first, the greatmajority of partial 16S sequences available forPseudonocardia included these two regions; second,these two regions are among the most variable 16Sregions differentiating species in the genus Pseud-onocardia; third, we believed it is important for treestability in the phylogenetic analysis that all taxaoverlapped in at least two contiguous variable regions(i.e., V3&V4). This criterion of V3&V4 sharing of alltaxa means that these regions may have greaterinfluence on the inferred phylogenetic relationshipsthan other, similarly variable regions (e.g., V9). Weexcluded identical sequences that had been obtained inthe same environmental survey, such as duplicatesequences from a large marine sediment survey fromthe Mariana Trench (AY974775-AY974796; retainingonly AY974776, AY974783, and AY974793 in ouranalysis), and duplicate sequences from a large soilsurvey in Czechoslovakia (EF506948-EF507159;retaining from these only 13 non-identical Pseudono-cardia sequences). Exclusion of these duplicate taxadoes not affect any of our conclusions, but reducedcomputation time in our analyses. After exclusion ofredundant taxa, our dataset included 325 Pseudono-cardia taxa.In the final dataset of 325 sequences, 67.4% of thePseudonocardia sequences had been obtained byculture-dependent methods, 25.8% from cloning of16S sequences amplified from microbial communities,2.2% from direct PCR, and 4.6% from 454-pyrose-quencing of 16S amplicons (Fig. 2a, b). Also fromthese 325 sequences, 26.8% of the sequences had beenobtained from attine ants, 9.2% from plant material(e.g., endophytes, root bacteria), 3.4% from marineorganisms (e.g., sponge symbionts), 5.2% from humanor mouse skin, 31.7% from soil, 1.5% from rocksurfaces, 5.2% from marine sediments, 5.5% fromindustrial sources (e.g., sludge, biofilters), and 11.4%from other sources (e.g., art paintings, house dust, air,etc.). Two Kibdelosporangium, two Crossiella, twoActinokineospora, and three Amycolatopsis wereincluded as outgroups (Fig. 2b), covering pseudono-cardiaceous genera most closely related to the genusPseudonocardia (Huang and Goodfellow 2011; UGMunpublished).On 30. November 2009, 16S sequences from anadditional 24 ant-associated Pseudonocardia werereleased at GenBank (accessions EU283917-EU283940; Poulsen Cafaro, Erhardt, Little, Gerardo,Currie, unpublished, Host-pathogen dynamics in anancient antibiotic system). While it was too late toFig. 2 Phylogeny of Pseudonocardia reconstructed for 325taxa under the maximum-likelihood criterion from 16S rRNAgene sequence information. Ant-associated Pseudonocardiaare boxed (shaded background); environmental Pseudonocar-dia are unboxed. Attine-associated lineages are closely relatedto soil-dwelling or endophytic Pseudonocardia lineages,indicating frequent recruitment of free-living Pseudonocardiainto association with ants. Support values are inferred from 100ML bootstrap pseudoreplicates. Each taxon label gives themethod of characterization (sequencing of 16S amplicons fromcultured isolates; sequencing of cloned 16S amplicons fromenvironmental surveys; pyrosequencing of 16S amplicons;direct PCR from ants) and the GenBank accession, followed inparentheses by the ant species name from which a bacteriumwas sampled, the ant collection ID (e.g., UGM010817-01), thesource from which the sequence was obtained (worker, garden,soil, plant, etc.), and the collection location. The attine antgenera are abbreviated as: Ac. = Acromyrmex; Ap. = Aptero-stigma; At. = Atta; Cy. = Cyphomyrmex; My. = Mycocepu-rus; Tr. = TrachymyrmexcAntonie van Leeuwenhoek (2010) 98:195–212 201123
  8. 8. 202 Antonie van Leeuwenhoek (2010) 98:195–212123
  9. 9. Fig. 2 continuedAntonie van Leeuwenhoek (2010) 98:195–212 203123
  10. 10. incorporate these sequences in our phylogeneticanalyses, comparison of these new sequences withthose in our phylogenetic reconstruction permittedprecise placement of the new sequences into ourphylogenetic framework (UG Mueller unpublished).Placement of these new sequences into the phyloge-netic context (Fig. 2) does not change any of theconclusions emphasized below.Alignment and character selectionSequences were aligned in MacClade version 4.06(Maddison and Maddison 2000). The initial align-ment of 1521 characters was adjusted manually tocorrect obvious alignment errors. To validate andadjust our initial alignment, we used the sequencealignment editor AE2 (Olsen et al. 1992) to manuallyalign characters according to similar positioning inthe RNA secondary structure. Regions of the 16SrRNA sequence that have significant sequence sim-ilarity can be aligned solely based on this identity.However, regions of the alignment that do not containsufficient sequence similarity require additional con-straints to accurately place nucleotides in columnsthat represent a similar part of the overall rRNAhigher-order structure. The secondary structure dia-grams were drawn with the interactive graphicsprogram XRNA (B Weiser, H Noller unpublished).Inspection of our alignment revealed that sequencesfrom ant-associated Pseudonocardia and P. halophob-ica (accessions EU139566.1-EU139584.1; thesesequences were later replaced after completion of ouranalyses and once we had pointed out sequencingerrors; see Supplemental Information) submitted byPoulsen et al. (2007) contained highly unusual char-acter scorings, some of them at positions within the 16Sgene that are thought to be invariable or highlyconserved across all Eubacteria (see SupplementalInformation). The secondary structure analysis of thesesuspect 16S sequences revealed that these unusualcharacters are likely sequencing artifacts. Theseunusual characters were furthermore rendered suspectbecause they also appeared in a sequence deposited byPoulsen et al. (2007) for P. halophobica (EU139576.1,from strain NRRL B-16514), whereas all othersequences deposited at GenBank for P. halophobica(AJ252827 = strain IMSNU21327; Z14111 = strainDSM43089T; Y08534 = strain DSM43089T) do notshow the unusual character scorings appearing inPoulsen et al. 2007 (see Figures S3 and S7 in theSupplemental Information for secondary-structuremodels of the P. halophobica 16S rRNA sequenceswith and without the sequencing errors). We thereforeexcluded from our phylogenetic analyses 11 suspectcharacters introduced by the Poulsen et al. (2007)sequences (these characters did not vary among all theother taxa, and exclusion therefore did not affectphylogenetic inferences about the remaining taxa).However, we were unsure about the error of fouradditional suspect characters in the Poulsen et al.(2007) sequences, which therefore were retained in ourphylogenetic analyses and were expected to renderthe corresponding taxa as phylogenetically derived.The derived phylogenetic positions of some of thesequences from Poulsen et al. (2007) (EU139566-EU139584) in Fig. 2 therefore may be an artifact ofuncorrected sequence-scoring errors.16S RNA secondary structure modelingTo facilitate future 16S RNA alignments for phyloge-netic reconstruction, we generated secondary structuremodels of 16S rRNA for five Pseudonocardia speciesspanning the diversity of the genus: P. carboxydivo-rans (GenBank accession EF114314; type strain Y8;Fig. 3), P. kongjuensis (AJ252833; type strain LM157), P. halophobica (AJ252827; type strain IMSNU21327), P. spinosispora (AJ249206; type strain LM141), and P. yunnanensis (AJ252822; type strainIMSNU 22019) (see Supplemental Information). The16S rRNA secondary structures were predicted withcomparative methods for a secondary structure that iscommon to all of the known 16S rRNA sequences,using Covariation analysis (Woese et al. 1980, 1983;Gutell et al. 1985, 1992, 1994; Cannone et al. 2002).We are confident that the 16S rRNA secondarystructure predictions are accurate, because, in aprevious analysis, approximately 98% of the predictedmodel base pairs were present in the high-resolutioncrystal structure of the ribosomal subunit (Gutell et al.Fig. 3 16S rRNA secondary structure model for P. carboxydivo-rans (GenBank accession EF114314; type strain Y8). TheSupplemental Information presents additional secondary struc-ture models for four additional species within the genusPseudonocardia (P. kongjuensis, P. halophobica, P. spinosis-pora, P. yunnanensis). Canonical base pairs (G:C, A:U) areshown with tick marks, wobble (G:U) base pairs with smallclosed circles, A:G base pairs with large open circles, and allother non-canonical base pairs with large closed circlesc204 Antonie van Leeuwenhoek (2010) 98:195–212123
  11. 11. Antonie van Leeuwenhoek (2010) 98:195–212 205123
  12. 12. 2002). Bacterial 16S rRNA secondary structurediagrams for the modeled Pseudonocardia species areavailable at the Comparative RNA Web Site, analysesOur final alignment utilized all characters of the 16Sgene, but excluded the 11 suspect characters fromPoulsen et al. (2007; see above) and excluded 72characters from the conserved region at the beginningand end of the 16S gene because the great majority ofthe sequences were missing character information inthese end regions. Our final alignment included 1436characters (289 informative, 77 autapomorphic, and1070 invariable characters). Phylogenetic relation-ships were inferred under the maximum likelihood(ML) criterion using the General Time Reversible(GTR ? I ? G) model of nucleotide substitution(Garli version 0.951; Zwickl 2006; The appropri-ate model of nucleotide substitution was selectedusing the Akaike Information Criterion (AIC)implemented in Modeltest 3.7 (Posada and Crandall1998). We performed twenty separate likelihoodsearches to explore likelihood space more thor-oughly. Branch support was calculated by performingbootstrap analyses with 100 repetitions. Figure 2reports the best phylogenetic topology under the MLcriterion (lowest negative log-likelihood score)among the twenty ML searches.Results and discussionSub-groupings within the genus PseudonocardiaThe maximum-likelihood (ML) reconstruction shownin Fig. 2a and b places 87 known ant-associatedPseudonocardia within the larger phylogenetic con-text of 238 environmental Pseudonocardia obtainedfrom diverse sources (e.g., soil, marine sediment,plants, marine invertebrates, mammal skin, industrialmaterial, etc.). We recover the main sub-groupings ofPseudonocardia that already emerged in the 16Sanalyses of Lee et al. (2000) and Mueller et al.(2008), but our increased taxon sampling tentativelyidentifies at least 10 subgroups within the genus(labeled as Clades 1–10 in Fig. 2a, b). The genusPseudonocardia appears to be split basally into twomain groups, one clade comprised of five subgroupsthat contain the great majority of the known ant-associated Pseudonocardia (Fig. 2a); and a second,less satisfactorily resolved clade comprised of at least5 subgroups (Fig. 2b). However, this basal split intothese two main groups is only weakly supported inour ML analyses, as Clades 4 and 5 (shown inFig. 2a) exhibited also affinities with the other maingroup (shown in Fig. 2b) in some of the less-likelyML reconstructions. Only few subgroups within thegenus received adequate support in the ML bootstrapanalyses, but two subgroups with acceptable supportvalues (Clade 1 and a larger clade comprised ofClades 1–3; Fig. 2a) contain the great majority of theant-associated Pseudonocardia. Clearly, informationfrom additional genes will be necessary to resolve allsubgroupings and affiliations with confidence. Fortu-nately, Wen-Jun Li’s Lab at Yunnan University,China, is currently compiling sequence informationof four protein-coding genes (rpoB, fusA, gyrB, andlepA) for a multilocus phylogenetic analysis of keyspecies in the genus Pseudonocardia. This frame-work should be useful in future studies to place theant-associated Pseudonocardia into the subgroups ofthe genus. Preliminary analyses (UG Mueller unpub-lished) of the sequence information available atGenBank for these four protein-coding genes alsorevealed the monophyly of Clade 1 and the mono-phyly of the larger clade comprised of Clades 1–3, asshown in Fig. 2a.Nesting of environmental Pseudonocardialineages within clades previously thoughtto be ant-specificOur analysis confirms the prediction of the Acquisitionmodel that there exist previously unknown environ-mental Pseudonocardia that insert themselves intoclades that originally were thought to be ‘‘ant-spe-cific’’. For example, Cafaro and Currie (2005) hadidentified a three-taxon clade of Pseudonocardia(isolated from Trachymyrmex ants from Panama) thathad emerged also as phylogenetically unique and ‘‘ant-specific’’ in the phylogenetic analysis by Mueller et al.(2008; Fig. 1); however, we found two environmentalPseudonocardia nested within that group of allegedlyTrachymyrmex-specific Pseudonocardia (Clade 4 in206 Antonie van Leeuwenhoek (2010) 98:195–212123
  13. 13. Fig. 2a). One of these is an endophytic Pseudonocar-dia that was isolated from the surface-sterilized root ofan Eucalyptus tree in Australia (GenBank accessionFJ805426; 99.8% sequence-identical to AY376893isolated from a Trachymyrmex worker from Panama),the second Pseudonocardia was isolated in a survey ofmicrobial growth that degraded prehistoric cavepaintings in Spain (Stomeo et al. 2008, 2009;EU810881; 99.3% sequence-identical to AY376893isolated from a Trachymyrmex worker). The nesting ofthese two environmental Pseudonocardia inside aclade of ant-associated Pseudonocardia is less com-patible with the phylogenetically derived position ofant-associated Pseudonocardia predicted by theCoevolution-Codivergence Hypothesis. Moreover,representatives of Clade 4 have so far been isolatedonly six times (four times from Trachymyrmex ants,twice from other sources; details in SupplementalInformation) and they have so far not been detected inany culture-independent screen. This suggests that itmay be sometimes inherently difficult to find environ-mental Pseudonocardia that are nested within ant-specific clades, either because these environmentalPseudonocardia are difficult to isolate, or they are rare,or a combination of both factors. Because of thisinherent difficulty of detection, the absence of evi-dence for environmental Pseudonocardia that arenested within ant-associated Pseudonocardia cladesshould not be interpreted as evidence for absence, andtherefore also not as evidence in support for theCoevolution-Codivergence hypothesis (e.g., Cafaroand Currie 2005; Poulsen et al. 2009).Figure 2a and 2b reveal additional cases whereenvironmental Pseudonocardia nest within clades thatwere previously thought to be largely ant-specific. Forexample, for the main group of ant-associated Pseud-onocardia (Clade 3 in Fig. 2a), only one environmen-tal counterpart was known previously (Fig. 1),whereas now a number of environmental affiliates areknown (sequence identical or near sequence identical at[99% similarity), such as the endophytic P. tropica sp.nov. from China, EU931094 from Antarctic soil,P. carboxydivorans from soils in China and Korea,P. nitrificans from soil in the USA, etc. (Fig. 2a). Forthe second main group of ant-associated Pseudono-cardia (Clade 1 in Fig. 2a), one environmental affiliatewas known previously (Fig. 1) whereas now threeenvironmental affiliates are known (Fig. 2a), includingthe endophytic P. endophytica from China, DQ490989from an endophyte in China, and EF682939 from ahotspring in the USA. Another interesting case is aconglomerate of ant-associated Pseudonocardia fromthree species of lab-reared attine ants, whose Pseud-onocardia have close affinities with P. khuvsgulensissp. nov. from soil in Mongolia, P. thermophila fromcompost in Germany, and the cloned EU132555 fromprairie soil in the USA (Fig. 2b). Overall, the phylo-genetic patterns conform with the prediction of theAcquisition model that, because the attine integumentis frequently colonized by bacteria from the environ-ment, ant-associated Pseudonocardia have closephylogenetic links to diverse environmental Pseud-onocardia. As we predicted before (Mueller et al.2008), future studies are likely to uncover manyadditional close links between environmental counter-parts and ant-associated Pseudonocardia, especially ifmore environmental Pseudonocardia are surveyedacross the Neotropics (the realm of attine ants) and ifmore of the neglected basal attine ant lineages aresurveyed for their integumental microbes.Single Pseudonocardia lineages colonizethe integument of a wide diversity of attine antsIn previous analyses (Cafaro and Currie 2005; Zhanget al. 2007; Mueller et al. 2008), Clade 3 appeared tocontain Pseudonocardia that had been isolatedmostly from Atta and Acromyrmex leafcutter ants.This near exclusive association between leafcutterants and Clade 3 Pseudonocardia could be inter-preted as evidence for co-diversification between theleafcutter ant clade and Clade 3 Pseudonocardia(sensu Poulsen et al. 2009). However, the accumu-lated evidence now documents also the presence ofClade 3 Pseudonocardia in the microbial communi-ties of the integument of many Trachymyrmex ants(Fig. 2a), and it is possible that a more thoroughsurvey of basal attine lineages could reveal associ-ations of Clade 3 Pseudonocardia with ant lineagesacross the entire diversity of attine ants. As a parallelcase, Sen et al. (2009) recently isolated from Clade 1Pseudonocardia from laboratory nests of the lower-attine species Mycocepurus smithii (Fig. 2a), show-ing that more comprehensive sampling uncovered thefirst exceptions to the original ‘‘specificity’’ of Clade1 Pseudonocardia on leafcutter and Apterostigmaants. Likewise, a group of closely related Clade 7Pseudonocardia isolated from three species ofAntonie van Leeuwenhoek (2010) 98:195–212 207123
  14. 14. Cyphomyrmex (costatus, muelleri, longiscapus) wasinterpreted by Poulsen et al. (2007, 2009) as possiblyCyphomyrmex-specific, but our phylogenetic analysesnow reveal Pseudonocardia from Clade 7 that hadbeen isolated from diverse Trachymyrmex ant species(Zhang et al. 2007; Sen et al. 2009; Fig. 2b). Theisolation of Clade 7 Pseudonocardia from Trachy-myrmex laboratory colonies by Sen et al. (2009)could potentially mean that the screened Trachymyr-mex workers were colonized by Pseudonocardiaderived from Cyphomyrmex species that had beenkept in the same lab room for several years. If so, thisthen would document how readily Pseudonocardiacan colonize the integument of Trachymyrmex spe-cies from other sources. We conclude, therefore, thattests for ant-Pseudonocardia specificities and co-diversification should avoid conclusions based onsmall sample sizes (which too readily can underes-timate the true diversity of Pseudonocardia associ-ates). Rather, tests for specificities should firstcarefully document the full diversity of Pseudono-cardia that colonize the integument of a single antand different workers from the same colony, evaluatepotential isolation biases (e.g., preferential isolationof autotrophic Pseudonocardia with minimum-nutri-ent medium; see below), and rule out cross-contam-ination from environmental sources.Culturable versus unculturable PseudonocardiaPerhaps the most striking difference between the twomain Pseudonocardia clades in Fig. 2a and b is that99% of the sequences in Fig. 2a (Clades 1–5) derivedfrom Pseudonocardia that had been cultured,whereas the sequences in Fig. 2b derived from amix of cultured and uncultured Pseudonocardia(cloned, pyrosequenced). In fact, the sizable Clade9 in Fig. 2b consists entirely of uncultured Pseud-onocardia, with the exception of the culturedP. asaccharolytica. Absence or underrepresentationof cultured Pseudonocardia in a clade suggests thatthese types of bacteria may be inherently difficult toisolate and were therefore not captured with theculture-dependent methods used so far. In contrast,the near absence of cloned Pseudonocardia in Clades1–5 (Fig. 2a) suggests potential cloning biases, forexample due to disruption of E. coli growth by the16S sequences inserted into the E. coli genomeduring cloning (e.g., Hamady and Knight 2009, andreferences therein). Interestingly, the great majorityof ant-associated Pseudonocardia characterized todate belong to two clades (Clade 1 and Clade 3) thatare heavily biased towards culturable Pseudonocar-dia (this bias is also true for all the environmentalPseudonocardia in these clades; Fig. 2a), suggestingthat these kind of Pseudonocardia are readilycaptured with the isolation methods used in thesurveys so far, whereas the less culturable Pseud-onocardia (i.e., from Clades 6–8) may have beenmissed by the same surveys.The only study that compared culture-dependentand culture-independent methods for surveying ant-associated Pseudonocardia was Sen et al. (2009), whodocumented that, whereas the standard culture-depen-dent method of Cafaro and Currie (2005) yielded only asingle Pseudonocardia type from workers of the antTrachymyrmex zeteki, a culture-independent pyrose-quencing survey of T. zeteki workers revealed thepresence of several species from different Pseudono-cardia subgroups. Taken together, the phylogeneticdistribution of the cultured and uncultured Pseudono-cardia (Fig. 2a versus 2b) and the preliminary findingsof Sen et al. (2009) suggest that the standard isolationmethod of attine-associated Pseudonocardia withminimum-carbon, chitin medium (Cafaro and Currie2005) may underestimate the true Pseudonocardiadiversity present on attine ants. If so, the isolation ofonly one kind of Pseudonocardia from the integumentof a particular ant (e.g., a Pseudonocardia from Clades1 and 3) should therefore not be taken as evidence thatthis kind of Pseudonocardia is the only Pseudonocar-dia species in the microbial community on the antintegument. We therefore caution that previous studiesthat relied on minimum-nutrient isolation methodsmight have underestimated the true diversity ofPseudonocardia in the microbial communities grow-ing on the integument of attine ants.Colonization of the ant integumentby Pseudonocardia from plant and soil sourcesAs already noted by Mueller et al. (2008), many ant-associated Pseudonocardia are closely relatedto Pseudonocardia from soil and plant material(e.g., endophytic Pseudonocardia), but several addi-tional close affinities now emerge from our largerphylogenetic analysis. For example, ant-associatedPseudonocardia that have sequences identical to208 Antonie van Leeuwenhoek (2010) 98:195–212123
  15. 15. P. carboxydivorans, P. tropica sp. nov., and P. endo-phytica are now known (Fig. 2a), all of which have beencultured as endophytes from plant material. Likewise,environmental Pseudonocardia from soil can besequence-identical in 16S to attine-associated Pseud-onocardia (Fig. 2a). Both soil-dwelling and endophyticPseudonocardia therefore remain possible sources ofPseudonocardia that colonize the attine integument.This calls for experimental evaluation of the coloniza-tion rate of the attine integument by microbes from suchenvironmental sources, which may cause fast or slowmicrobial turnover on the integument depending on thenature of the resident microbial community, the natureof the colonizing microbes, and the general disturbancein the nest. Such studies will require in-depth screeningof the entire microbial community on the integument,as well as understanding of any biases of culture-dependent and culture-independent Pseudonocardiascreens (see above Culturable versus unculturablePseudonocardia).Pseudonocardia symbiosis beyond fungus-growing antsPseudonocardia has been found not only in associationwith fungus-growing ants, but also in association witha wide diversity of other organisms, including plants,marine sponges, an anemone, a Clorella alga, an aphid,and mammals (human gut; human and mouse skin)(Fig. 2). Excluding the 87 ant-associated Pseudono-cardia in Fig. 2, 12.6% of the surveyed Pseudonocar-dia derived from plants (endophytes), 4.6% frommarine invertebrates, 0.4% from terrestrial arthropods(27% when including ants), and 7% from mammal skin(human, mouse). The absence or rarity of Pseudono-cardia in other well-characterized microbial commu-nities (e.g., human gut, termite gut, ungulate rumen)indicates that Pseudonocardia are not omnipresent andcapable of life in all symbiotic niches. However, thefrequent isolation from marine sponges and plants isintriguing and suggests that Pseudonocardia evolutionmay have been shaped by symbiosis with a wide arrayof organisms.Autotrophic versus heterotrophic Pseudonocardiaon the ant integumentMany Pseudonocardia isolated from attine ants areclosely related to autotrophic Pseudonocardia, suchas P. carboxydivorans, P. autotrophica, P. nitrificans(Fig. 2a), suggesting the possibility that the integu-mental Pseudonocardia of attine ants may alsoinclude autotrophs that are not limited in growth (orare less limited) by nutrients supplied by the ants, asassumed by the Coevolution-Codivergence hypothe-sis. The traditional formulation of the Coevolution-Codivergence hypothesis assumes that the ants haveprecise control over growth (upregulation of Pseud-onocardia growth to combat an acute Escovopsisinfection of the garden; Currie et al. 2003); suchcontrol may be less possible if Pseudonocardia areautotrophs that can thrive in nutrient-poor niches. Thepossibility of autotrophic Pseudonocardia on the antintegument could explain why it is relatively easy toisolate slow-growing Pseudonocardia with mini-mum-nutrient chitin medium (e.g., preferential selec-tion of autotrophs), perhaps with similar biochemicalproperties as the closely related P. carboxydivoransthat can fix carbon from carbon-monoxide (Park et al.2008). While one could speculate that removal oftoxic carbon-monoxide could be a potential functionof Pseudonocardia in the interior of nests of fungus-growing ants, it is equally possible that the anoxicconditions in the bottom of fungus gardens promotesnon-adaptive proliferation of autotrophic Pseudono-cardia (e.g., on the immobilized Acromyrmex work-ers at the bottom of fungus gardens, as documentedby Currie et al. 2003).Phylogenetic position of Pseudonocardiacompacta and P. nitrificansOur new 16S sequence of P. nitrificans (strain NRRLB-1664) places this species into a clade with P. alni,P. carboxydivorans, P. antarctica, and P. tropica sp.nov. (Clade 3, Fig. 2a). Our P. nitrificans sequence(GU318371) is identical to AJ252831 (= strain IMS-NU 22071 = P. nitrificans; Lee et al. 2000), whichcurrently is reported at GenBank as an undescribedPseudonocardia. We believe that these two sequencesare of much better quality than sequence X55609(=strain IFAM 379 = P. nitrificans), which has beenused in previous phylogenetic analyses (Warwick et al.1994; McVeigh et al. 1994; Huang and Goodfellow2011) but whose phylogenetic placement in theseanalyses is now dubious because of the poor sequencequality of X55609. More surprising was the placementof our 16S sequence of P. compacta (NRRL B-16170;Antonie van Leeuwenhoek (2010) 98:195–212 209123
  16. 16. GU318372) close to P. saturnea (Fig. 2b), that is, in adifferent subgroup of Pseudonocardia than accessionAJ252825 (=IMSNU 20111; Fig. 2a) of P. compactathat is generally used in phylogenetic analyses of thegenus. Interestingly, the placement of P. compacta inthe vicinity of P. saturnea conforms approximatelywith the placement of P. compacta (DSM 43592) whenusing information from four protein-coding genesavailable at GenBank (EU722624, EU722594,EU722569, EU722535; UG Mueller unpublishedobservation), suggesting the possibility of strain mix-up between culture collections. [Note inserted at proofstage: We recently obtained a strain labeledP. saturnea (NRRL B-16172) from the ARS Cultureand Patent Culture Collections (USDA, Peoria, Illi-nois) and the 16S sequence of this strain was identicalto the sequence of P. compacta (AJ252825; IMSNU20111T) (H Ishak and UG Mueller unpublished).Because we had received the strain labeled NRRLB-16170 (believed to be P. compacta, but sequencingto P. saturnea) and the strain labeled NRRL B-16172(believed to be P. saturnea, but sequencing toP. compacta) in two separate shipments from theARS-USDA culture collection, and because we com-pleted sequencing of NRRL B-16170 before wereceived NRRL B-16172, it seems that these twostrains were mixed up prior to shipment from the ARS-USDA culture collection to our lab.] To resolve theseissues, it will be necessary to re-sequence the typestrains of P. compacta, P. nitrificans, and P. saturneafor 16S and other informative genes.ConclusionThe two models juxtaposed in our analyses – theCoevolution-Codivergence model and the Acquisi-tion model—represent two extreme viewpoints. Thetraditional Coevolution-Codivergence model stressesmutualism and long-term ant-Pseudonocardia-Escov-opsis co-evolution and co-divergence, the morerecent Acquisition model stresses diversity in eco-logical association and close links between ant-associated and environmental Pseudonocardia. Thesetwo models make distinct predictions regarding thephylogenetic proximity of ant-associated and envi-ronmental Pseudonocardia. Our expanded phyloge-netic analyses conform more with the predictions ofthe Acquisition model, adding to recent antibioticassays and phylogenetic analyses (Kost et al. 2007;Mueller et al. 2008; Haeder et al. 2009; Sen et al.2009) that had started to question the plausibility ofant-Pseudonocardia-Escovopsis co-evolution and co-divergence. Specifically, our phylogenetic analysesreveal previously unknown cases of 16S-sequence-identical ant-associated and environmental Pseud-onocardia (from soil and plants), and environmentalPseudonocardia that are nested within clades previ-ously thought to contain only specialized, ant-asso-ciated Pseudonocardia. These phylogenetic patternsindicate frequent and continuous colonization of theattine integument by Pseudonocardia from environ-mental sources, microbial turnover on the ant integ-ument over evolutionary time, and consequentlydisruption of long-term ant-Pseudonocardia associa-tions. Such processes greatly reduce the potential forant-Pseudonocardia co-evolution and for adaptivemodification of Pseudonocardia to serve the fitnessinterests of the ants. Inadequate inclusion of envi-ronmental Pseudonocardia in previous analysesappears to have led to the premature identificationof ‘‘ant-specific’’ Pseudonocardia clades. More sam-pling of environmental Pseudonocardia is needed tofully address the ecological links between ant-asso-ciated and environmental Pseudonocardia. Likewiseneeded are genomic and biochemical studies thatevaluate genetic and phenotypic differences (orabsence of differences) between ant-associated andenvironmental Pseudonocardia, particularly in theNeotropics where only few environmental bacterialsurveys have been carried out to date. Becauseprevious culture-dependent methods characterizingPseudonocardia diversity on attine ants may havebeen biased (see Culturable and unculturable Pseud-onocardia), future studies should also attempt todevelop unbiased methods to characterize the truemicrobial diversity on attine ants. We encouragestudies that test for both non-adaptive and adaptiveroles of integumental microbes in carefully designedexperiments, and we concur with the recent assess-ment by Boomsma and Aanen (2009) that thetraditional Coevolution-Codivergence model greatlyoversimplifies the nature and ecological diversity ofant-Pseudonocardia associations.Acknowledgements We thank Linquan Bai and AmandaJones for the invitation to contribute to this special issuehonoring Mike Goodfellow and his lifetime contribution to210 Antonie van Leeuwenhoek (2010) 98:195–212123
  17. 17. actinomycete ecology, systematics, and evolution. We thankLinda Kinkel, Liz Wellington, and Linquan Bai for theinvitation to give a plenary talk at the 2009 ISBA Congressin Shanghai, for which we developed many ideas presentedhere. We thank two anonymous reviewers for constructivecomments on the manuscript. The research was supported byNSF awards DEB-0919519, DEB-0639879, and DEB-0110073to UGM; and NIH award R01GM067317 to RRG.ReferencesAcinas SG, Sarma-Rupavtarm R, Klepac-Ceraj V, Polz MF(2005) PCR-induced sequence artifacts and bias: insightsfrom comparison of two 16S rRNA clone libraries con-structed from the same sample. Appl Environ Microbiol71:8966–8969Bacci M, Ribeiro SB, Casarotto MEF, Pagnocca FC (1995)Biopolymer-degrading bacteria from nests of the leaf-cutting ant Atta sexdens rubropilosa. Braz J Med Biol Res28:79–82Boomsma JJ, Aanen DK (2009) Rethinking crop-diseasemanagement in fungus-growing ants. Proc Natl Acad SciUSA 106:17611–17612Cafaro MJ, Currie CR (2005) Phylogenetic analysis of mutu-alistic filamentous bacteria associated with fungus-grow-ing ants. Can J Microbiol 51:441–446Caldera EJ, Poulsen M, Suen G, Currie CR (2009) Insectsymbioses: a case study of past, present, and future fun-gus-growing ant research. Environ Entomol 38:78–92Cannone JJ, Subramanian S, Schnare MN, Collett JR, D’SouzaLM, Du Y, Feng B, Lin N, Madabusi LV, Mu¨ller KM,Pande N, Shang Z, Yu N, Gutell RR (2002) The com-parative RNA Web (CRW) site: an online database ofcomparative sequence and structure information forribosomal, intron, and other RNAs. BioMed Cen Bioin-formatics 3:2Carreiro SC, Pagnocca FC, Bueno OC, Bacci M, Hebling MJA,da Silva OA (1997) Yeasts associated with the nests of theleaf-cutting ant Atta sexdens rubropilosa Forel, 1908.Antonie van Leeuwenhoek 71:243–248Currie CR (2001) A community of ants, fungi, and bacteria: amultilateral approach to studying symbiosis. Annu RevMicrobiol 55:357–380Currie CR, Mueller UG, Malloch D (1999a) The agriculturalpathology of ant fungus gardens. Proc Natl Acad Sci USA96:7998–8002Currie CR, Scott JA, Summerbell RC, Malloch D (1999b)Fungus-growing ants use antibiotic-producing bacteria tocontrol garden parasites. Nature 398:701–704Currie CR, Bot ANM, Boomsma JJ (2003) Experimental evi-dence of a tripartite mutualism: bacteria protect ant fungusgardens from specialized parasites. Oikos 101:91–102Currie CR, Poulsen M, Mendenhall J, Boomsma JJ, Billen J(2006) Coevolved crypts and exocrine glands supportmutualistic bacteria in fungus-growing ants. Science311:81–83Dattagupta S, Schaperdoth I, Montanari A, Mariani S, Kita N,Valley JW, Macalady JL (2009) A novel symbiosisbetween chemoautotrophic bacteria and a freshwater caveamphipod. Int Soc Microb Ecol 3:935–943Diamond J (2006) Virus and the whale: exploring evolution increatures small and large. National Science TeachersAssociation Press, Arlington, VAFerna´ndez-Marı´n H, Zimmerman JK, Nash DR, Boomsma JJ,Wcislo WT (2009) Reduced biological control andenhanced chemical pest management in the evolution offungus farming in ants. Proc R Soc Lond B 276:2236–2269Goodfellow M, Williams ST (1983) Ecology of actinomycetes.Annu Rev Microbiol 37:189–216Gould SJ, Lewontin RC (1979) The spandrels of San Marcoand the Panglossian Paradigm: a critique of the Adapta-tionist Programme. Proc R Soc Lond B 205:581–598Gutell RR, Weiser B, Woese CR, Noller HF (1985) Compar-ative anatomy of 16S-like ribosomal RNA. Prog NuclAcid Res Mol Biol 32:155–216Gutell RR, Power A, Hertz G, Putz E, Stormo G (1992)Identifying constraints on the higher-order structure ofRNA: continued development and application of com-parative sequence analysis methods. Nucleic Acids Res20:5785–5795Gutell RR, Larsen N, Woese CR (1994) Lessons from anevolving ribosomal RNA: 16S and 23S rRNA structurefrom a comparative perspective. Microbiol Rev 58:10–26Gutell RR, Lee JC, Cannone JJ (2002) The accuracy of ribo-somal RNA comparative structure models. Curr OpinStruct Biol 12:301–310Haeder S, Wirth R, Herz H, Spiteller D (2009) Candicidin-producing Streptomyces support leaf-cutting ants to pro-tect their fungus garden against the pathogenic fungusEscovopsis. Proc Natl Acad Sci USA 106:4742–4746Hamady M, Knight R (2009) Microbial community profilingfor human microbiome projects: tools, techniques, andchallenges. Genome Res 19:1141–1152Huang Y, Goodfellow M (2011) Genus I. Pseudonocardia(Henssen 1957, 408AL) emend. Park, Park, Lee andKim 2008, 2477VP. In: Goodfellow M, Ka¨mpfer P,Busse H-J, Trujillo M, Suzuki K, Ludwig W, WhitmanWB (eds) Bergey’s manual of systematic bacteriology,2nd ed., vol 5: the actinobacteria. Springer, New York(in press)Kost C, Lakatos T, Bo¨ttcher I, Arendholz W-R, Redenbach M,Wirth R (2007) Non-specific association between fila-mentous bacteria and fungus-growing ants. Naturwis-senschaften 94:821–828Lee SD, Kim ES, Hah YC (2000) Phylogenetic analysis of thegenera Pseudonocardia and Actinobispora based on 16Sribosomal DNA sequences. FEMS Microbiol Lett182:125–129Maddison DR, Maddison WP (2000) MacClade 4: analysis ofphylogeny and character evolution. Sinauer Associates,Sunderland, MAMcVeigh HP, Munro J, Embley TM (1994) The phylogeneticposition of Pseudoamycolata halophobica (Akimov et al.1989) and a proposal to reclassify it as Pseudonocardiahalophobica. Int J Syst Bacteriol 44:300–302Mueller UG, Gerardo NM, Aanen DK, Six DL, Schultz TR(2005) The evolution of agriculture in insects. Ann RevEcol Evol Sys 36:563–595Mueller UG, Dash D, Rabeling C, Rodrigues A (2008)Coevolution between attine ants and actinomycete bacte-ria: a reevaluation. Evolution 62:2894–2912Antonie van Leeuwenhoek (2010) 98:195–212 211123
  18. 18. Oh D-C, Poulsen M, Currie CR, Clardy J (2009) Dentigeru-mycin: a bacterial mediator of an ant-fungus symbiosis.Nat Chem Biol 5:391–393Olsen GJ, Overbeek R, Larsen N, Marsh TL, McCaughey MJ,Maciukenas MA, Kuan WM, Macke TJ, Xing Y, WoeseCR (1992) The ribosomal database project. Nucleic AcidsRes 20 suppl:2199–2200Park SW, Park ST, Lee JE, Kim YM (2008) Pseudonocardiacarboxydivorans sp. nov., a carbon monoxide-oxidizingactinomycete, and an emended description of the genusPseudonocardia. Int J Syst Evol Microbiol 58:2475–2478Posada DK, Crandall A (1998) Model selection and modelaveraging in phylogenetics: advantages of the AIC andBayesian approaches over likelihood ratio tests. Bioinfor-matics 14:817–818Poulsen M, Erhardt DP, Molinaro DJ, Ting-Li L, Currie CR(2007) Antagonistic bacterial interactions help shape host-symbiont dynamics within the fungus-growing ant-microbe mutualism. PLoS ONE 2:e960Poulsen M, Little AEF, Currie CR (2009) Fungus-growing ant-microbe symbiosis: using microbes to defend beneficialassociations with symbiotic communities. In: White JF,Torres MS (eds) Defensive mutualism in microbial sym-biosis. CRC Press, Boca Raton, pp 149–162Rodrigues A, Bacci M, Mueller UG, Ortiz A, Pagnocca FC(2008) Microfungal ‘‘weeds’’ in the leafcutter ant sym-biosis. Microb Ecol 56:604–614Rodrigues A, Cable RN, Mueller UG, Bacci M, Pagnocca FC(2009) Antagonistic interactions between garden yeastsand microfungal garden pathogens of leaf-cutting ants.Antonie Leeuwenhook 96:331–342Santos AV, Dillon RJ, Dillon VM, Reynolds SE, Samuels RI(2004) Occurrence of the antibiotic producing bacteriumBurkholderia sp. in colonies of the leaf-cutting ant Attasexdens rubropilosa. FEMS Microbiol Lett 239:319–323Sen R, Ishak HD, Estrada D, Dowd SE, Hong E, Mueller UG(2009) Generalized antifungal activity and 454-screeningof Pseudonocardia and Amycolatopsis bacteria in nests offungus-growing ants. Proc Natl Acad Sci USA 106:17805–17810Stomeo F, Portillo MC, Gonzalez JM, Laiz L, Saiz-Jimenez C(2008) Pseudonocardia in white colonizations in twocaves with Paleolithic paintings. Int Biodeterior Biodeg-radation 62:483–486Suen G, Currie CR (2008) Ancient fungal farmers of the insectworld. Microbiol Today 35:172–175Warwick S, Bowen T, McVeigh H, Embley TM (1994) Aphylogenetic analysis of the family Pseudonocardiaceaeand the genera Actinokineospora and Saccharothrix with16S rRNA sequences and a proposal to combine thegenera Amycolata and Pseudonocardia in an emendedgenus Pseudonocardia. Int J Syst Bacteriol 44:293–299Woese CR (2000) Interpreting the universal phylogenetic tree.Proc Natl Acad Sci USA 97:8392–8396Woese CR, Magrum LJ, Gupta R, Siegel RB, Stahl DA, Kop J,Crawford N, Brosius R, Gutell R, Hogan JJ, Noller HF(1980) Secondary structure model for bacterial 16S ribo-somal RNA: phylogenetic, enzymatic and chemical evi-dence. Nucleic Acids Res 8:2275–2294Woese CR, Gutell R, Gupta R, Noller HF (1983) Detailedanalysis of the higher-order structure of 16S-like ribo-somal ribonucleic acids. Microbiol Rev 47:621–669Zhang M, Poulsen M, Currie CR (2007) Symbiont recognitionof mutualistic bacteria in Acromyrmex leaf-cutting ants.Int Soc Microb Ecol J 1:313–320Zwickl DJ (2006) Genetic algorithm approaches for the phy-logenetic analysis of large biological sequence datasetsunder the maximum likelihood criterion. Ph.D. disserta-tion, The University of Texas at Austin212 Antonie van Leeuwenhoek (2010) 98:195–212123