fmicb-08-02016 October 16, 2017 Time: 17:44 # 1
ORIGINAL RESEARCH
published: 17 October 2017
doi: 10.3389/fmicb.2017.02016
Evidence for Widespread
Associations between Neotropical
Hymenopteran Insects and
Actinobacteria
Bernal Matarrita-Carranza1,2, Rolando D. Moreira-Soto2,3, Catalina Murillo-Cruz2,
Marielos Mora4, Cameron R. Currie5 and Adrián A. Pinto-Tomas2,4,6*
1 La Selva Biological Station, Organization for Tropical Studies, Heredia, Costa Rica, 2 Centro de Investigación en Estructuras
Microscópicas, Universidad de Costa Rica, San José, Costa Rica, 3 Centro de Investigación en Enfermedades Tropicales,
Universidad de Costa Rica, San José, Costa Rica, 4 Centro de Investigación en Biología Celular y Molecular, Universidad de
Costa Rica, San José, Costa Rica, 5 Department of Bacteriology, University of Wisconsin-Madison, Madison, WI,
United States, 6 Departamento de Bioquímica, Facultad de Medicina, Universidad de Costa Rica, San José, Costa Rica
The evolutionary success of hymenopteran insects has been associated with complex
physiological and behavioral defense mechanisms against pathogens and parasites.
Edited by:
Robert Brucker, Among these strategies are symbiotic associations between Hymenoptera and
Rowland Institute at Harvard, antibiotic-producing Actinobacteria, which provide protection to insect hosts. Herein,
United States
we examine associations between culturable Actinobacteria and 29 species of tropical
Reviewed by:
Amparo Latorre, hymenopteran insects that span five families, including Apidae (bees), Vespidae (wasps),
Universitat de València, Spain and Formicidae (ants). In total, 197 Actinobacteria isolates were obtained from 22
Regina Lamendella, of the 29 different insect species sampled. Through 16S rRNA gene sequences of
Juniata College, United States
161 isolates, we show that 91% of the symbionts correspond to members of the
*Correspondence:
Adrián A. Pinto-Tomas genus Streptomyces with less common isolates belonging to Pseudonocardia and
adrian.pinto@ucr.ac.cr Amycolatopsis. Electron microscopy revealed the presence of filamentous bacteria
with Streptomyces morphology in brood chambers of two different species of the
Specialty section:
This article was submitted to eusocial wasps. Four fungal strains in the family Ophiocordycipitacea (Hypocreales)
Microbial Symbioses, known to be specialized insect parasites were also isolated. Bioassay challenges
a section of the journal
Frontiers in Microbiology between the Actinobacteria and their possible targeted pathogenic antagonist (both
Received: 05 July 2017 obtained from the same insect at the genus or species level) provide evidence that
Accepted: 29 September 2017 different Actinobacteria isolates produced antifungal activity, supporting the hypothesis
Published: 17 October 2017 of a defensive association between the insects and these microbe species. Finally,
Citation: phylogenetic analysis of 16S rRNA and gyrB demonstrate the presence of five
Matarrita-Carranza B,
Moreira-Soto RD, Murillo-Cruz C, Streptomyces lineages associated with a broad range of insect species. Particularly
Mora M, Currie CR and our Clade I is of much interest as it is composed of one 16S rRNA phylotype repeatedly
Pinto-Tomas AA (2017) Evidence
for Widespread Associations between isolated from different insect groups in our sample. This phylotype corresponds to a
Neotropical Hymenopteran Insects previously described lineage of host-associated Streptomyces. These results suggest
and Actinobacteria. Streptomyces Clade I is a Hymenoptera host-associated lineage spanning several new
Front. Microbiol. 8:2016.
doi: 10.3389/fmicb.2017.02016 insect taxa and ranging from the American temperate to the Neotropical region. Our
Frontiers in Microbiology | www.frontiersin.org 1 October 2017 | Volume 8 | Article 2016
fmicb-08-02016 October 16, 2017 Time: 17:44 # 2
Matarrita-Carranza et al. Widespread Neotropical Insect-Actinobacteria Associations
work thus provides important insights into the widespread distribution of Actinobacteria
and hymenopteran insects associations, while also pointing at novel resources that
could be targeted for the discovery of active natural products with great potential in
medical and biotechnological applications.
Keywords: symbiosis, ants, bees, wasps, Streptomyces, Cordyceps, Ophiocordyceps, Hirsutella
INTRODUCTION The first is comprised of leaf-cutter ants and bacteria in the
genus Pseudonocardia. These bacteria are found on the cuticle
Hymenoptera is the third largest species-rich insect order, of the ants and can occur in crypts in their exoskeleton (Currie
containing more than 150000 extant species described to et al., 2006), and they produce secondary metabolites that can
date (Aguiar et al., 2013), only surpassed by Coleoptera and inhibit the growth of parasitic fungi (Escovopsis) that can colonize
Lepidoptera. All ants, bees, and wasps belong to this order the ant’s fungal cultivar (Currie et al., 1999; Oh et al., 2009;
and, together with termites (Blattodea), represent the best- Marsh et al., 2013; Sit et al., 2015). The second group consists of
known groups of insects where eusociality, the highest-level solitary wasps belonging to the genera Philanthus, Trachyphus,
of social organization, is present. Eusociality is characterized and Philanthinus (Kaltenpoth et al., 2005, 2010, 2012) in
by the presence of reproductive division of labor, brood which females harbor Streptomyces bacteria in specialized gland
care and overlapping of generation of individuals of the reservoirs inside the antenna. Before laying their eggs, female
same colony (Wilson, 1971). These complex social systems wasps secrete their Streptomyces symbiont inside the brood
and their associated adaptations are what have allowed some chambers. During larval development, the Streptomyces produce
hymenopteran groups to achieve ecological dominance playing a mixture of at least nine different antimicrobial compounds that
prominent roles in ecosystem function as pollinators (stingless will cover immature wasps and protect them from fungal attack
bees and the honey-bee), herbivores (leaf-cutter ants) and during hibernation periods (Kaltenpoth et al., 2005; Kroiss et al.,
predators (e.g., army ants and paper wasps; Hanson and Nishida, 2010). Actinobacteria have been characterized in several other
2016). The success of eusocial insects is reflected for example in social Hymenoptera, including within three ant-plant systems
tropical forests where they can represent more than 50% of all (Hanshew et al., 2015) and in bees (Promnuan et al., 2009),
animal biomass (Hölldobler and Wilson, 2009). suggesting these associations may be more widespread.
However, social lifestyles also represent an opportunity Neotropical Hymenoptera are an extremely diverse group of
for pathogens and parasites to exploit means of dispersion insects (Hanson and Gauld, 1995) with largely unexplored
between colony members in order to infect healthy individuals. Actinobacteria associations. Here we examine possible
Moreover, environmental conditions such as humidity and stable associations between Actinobacteria and tropical eusocial
temperatures inside the nest are factors that predispose these and subsocial Hymenoptera, including species of bees, wasps,
insects to pathogenic infestations (Oi and Pereira, 1993). and ants, in a tropical rain forest at La Selva Biological Station
As defensive responses against pathogenic microorganisms, in Costa Rica (Bawa et al., 1994). Culture-dependent targeted
many social Hymenoptera have developed collective hygienic isolations for Actinobacteria from 29 hymenopteran species were
strategies and altruistic behaviors in order to evade, control conducted. Patterns of associations between insect species and
or eliminate parasitic infections (Wilson-Rich et al., 2009). Actinobacteria isolates were characterized through 16S rRNA
Beneficial microbial associations may augment integral immune and gyrB DNA sequencing and Bayesian phylogenetic analyses.
defenses and help provide protection against pathogens via Moreover, different strains of entomopathogenic fungi known
microbial competition, by stimulating immune responses, or to infect some of the insects under study, including strains
through secretion of anti-microbial compounds (Berg, 1996; Bot in the genus Ophiocordyceps (Hirsutella) were isolated. These
et al., 2002; Dillon and Dillon, 2004; Evans and Lopez, 2004; de fungi are known as highly specialized pathogens capable of
Souza et al., 2009; Kaltenpoth and Engl, 2013). manipulating the behavior of their host (de Bekker et al., 2014).
As suggested by Kaltenpoth (2009), the factors that may Finally, the results from in vitro bioassays show that some strains
have predisposed Actinobacteria to engaging in defensive of Actinobacteria associated with different species of insects were
symbiotic associations with insects are their ubiquity in terrestrial capable of inhibiting the growth of these entomopathogenic
environments (Mayfield et al., 1972; van der Meij et al., 2017), fungi.
the capability to degrade recalcitrant carbon and nitrogen sources
(e.g., lignin, chitin and cellulose; Bhattacharya et al., 2007;
Větrovský et al., 2014), and the potential to produce many MATERIALS AND METHODS
bioactive secondary metabolites (Barka et al., 2016). The genus
Streptomyces alone produces more than 80% of all antibiotics with Insect Sampling for Actinobacteria
pharmaceutical applications that are used today, either directly as Screening
natural products or their semisynthetic derivatives (Bérdy, 2005). Insect colonies were collected at La Selva Biological Station
There are two well-studied hymenopteran groups that have (Organization for Tropical Studies) from 2010 to 2014 (Table 1).
established defensive symbiotic associations with Actinobacteria. This station is located in the province of Heredia, Costa Rica (10◦
Frontiers in Microbiology | www.frontiersin.org 2 October 2017 | Volume 8 | Article 2016
fmicb-08-02016 October 16, 2017 Time: 17:44 # 3
Matarrita-Carranza et al. Widespread Neotropical Insect-Actinobacteria Associations
25′ 53.14′′ N, 84◦ 0′ 10.51′′ W) and is composed of 1500 hectares leaves of different plants, and Pheidole bicornis (Myrmicinae)
of lowland tropical rain forest. A few other samples were collected that obligatorily lives in cavities of the petioles and stems shrubs
in Las Brisas Nature Reserve, Limón Province, Costa Rica (10◦ from the genus Piper. Four different ant species in the genus
04′ 09.7′′ N, 83◦37′ 57.5′′ W). Samples from each colony were Odontomachus (Ponerinae) which live under rocks or trunks,
aseptically collected using flame sterilized forceps and placed were also sampled. Finally, Paraponera clavata (Paraponerinae),
directly into a capped sterile container for transport to the an omnivorous large ant that constructs nests in the soil next
laboratory. Each colony was divided into different sub-samples to the base of trees and is a member of a recent paraphyletic
for further Actinobacteria isolation. Colony components when ancestor group to the formicoid ants lineage was also studied
present included larvae, adults, nest material and honey or pollen (Ward, 2014).
in the case of bees. Our sampling approach encompassed eusocial In the case of eusocial wasps, Actinobacteria associations
insects (all ants, paper wasps, and stingless bees) and non-eusocial with colonies of Agelaia cajennensis, Polybia plebeja, Polybia
insects (bees from the tribe Euglossini, and wasps from the family occidentalis, and Metapolybia docilis were studied. Subsocial
Pompilidae and Crabronidae). wasps in the families Pompilidae and Crabronidae were also
Ten different species of ants representing four sub-families sampled (Table 1). The Crabronidae family is related to the
and spanning different lifestyles and habitats were collected. Apidae lineage (Debevec et al., 2012) and known as mud daubers
The ant species sampled comprised, Tapinoma ramulorum together with Sphecidae.
(Dolichoderinae), Paratrechina caeciliae (Formicinae), and From the eusocial stingless bees, Tetragonisca angustula
Pheidole fiorii (Myrmicinae), which build carton nests on the and Trigona sp. 1 were investigated. Adults, larvae, comb,
TABLE 1 | Insect species and colony components sampled and Actinobacteria isolates obtained.
Number of isolates obtained from:
Insect family Species Colonies sampled Adult Larvae Nest Others Total
Apidae Euglossa allosticta‡ 1 0 — — — 0
Euglossa hansoni‡ 1 0 — — — 0
Euglossa heterosticta‡ 6 2 — — — 2
Euglossa ignita‡ 1 0 — — — 0
Euglossa imperilais‡ 2 2 — — — 2
Euglossa cybelia‡ 1 0 0 0 — 0
Eulaema speciosa‡ 2 0 0 0 0 0
Tetragonisca angustula† 16 7 2 0 10∗ 19
Trigona sp. 1 † 4 0 — 0 — 0
Formicidae † Crematogaster longispina 10 5 4 1 — 10
Odontomachus bauri 3 0 1 0 — 1
Odontomachus erythrocephalus 18 11 7 3 — 21
Odontomachus hastatus 1 0 1 0 — 1
Odontomachus opaciventris 5 0 2 0 — 2
Paraponera clavata 16 18 — — — 18
Paratrechina caeciliae 10 7 4 1 — 12
Pheidole bicornis 13 0 2 5 — 7
Pheidole fiorii 8 4 1 0 — 5
Tapinoma ramulorum inrectum 8 4 1 3 — 8
Pompilidae ‡ Pompilidae sp. A 3 1 2 0 — 3
Crabronidae ‡ Trypoxylon sp. A 4 3 8 4 — 15
Trypoxylon sp. B 1 1 1 0 — 2
Vespidae † Agelaia cajennensis 11 3 16 3 — 22
Metapolybia docilis 10 3 6 15 — 24
Polybia plebeja 13 1 6 5 — 12
Polybia occidentalis bohemani 7 2 2 4 — 8
Vespidae sp. A 1 0 0 0 — 0
Vespidae sp. B 1 0 1 0 — 1
Vespidae sp. C 1 0 1 1 — 2
Total 178 74 68 45 10 197
†Eusocial insects studied (all ants, paper wasps, and stingless bees). ‡Non-eusocial insects (mud dauber wasps and Euglossini bees). ∗ Isolates obtained as follows: three
from honey, one comb, one pollen, and five propolis.
Frontiers in Microbiology | www.frontiersin.org 3 October 2017 | Volume 8 | Article 2016
fmicb-08-02016 October 16, 2017 Time: 17:44 # 4
Matarrita-Carranza et al. Widespread Neotropical Insect-Actinobacteria Associations
honey, and pollen from colonies of T. angustula were collected; containers and brought to the laboratory for processing and
antimicrobial activity has been reported for this species’ honey identification.
(DeMera and Angert, 2004). Adults and beehive entrance Insect cadavers with fruiting bodies developed or with active
material were the only colony components of the Trigona sp. 1 conidiophores on stromata were attached with double-sided tape
collected. to the inner side of a Petri plate lid. Then the lid was placed on a
Five species of the weakly social orchid bees (Cardinal et al., plate containing PDA media for collecting discharged ascospores
2011) were studied using essential oils as bait to collect males or conidia (Lacey, 2012). Pieces of stromata or synnemata were
(Table 1). Finally, adults and larvae specimens from two colonies also cut with sterile scissors and directly stroked against PDA
of the solitary-nesting orchid bee Eulaema speciosa were included media with flame-sterilized forceps, in a laminar flow hood. Agar
in the study. plates were examined under a microscope and pure cultures
were obtained after cutting pieces of agar with single ascospores
Actinobacteria Isolation or hyphal tips and transferred into new PDA plates that were
A culture-dependent approach for isolating Actinobacteria was later incubated at 27
◦C for 4 weeks. Agar slants with sterile
performed, using chitin agar as a selective media (Cafaro and mineral oil and 20% glycerol stocks were prepared to preserve the
Currie, 2005). For small insects (e.g., C. longispina, T. ramulorum, isolates.
P. bicornis, P. fiorii, and P. caeciliae) five adults, five immature
forms (larvae and pupae) and five nest fragments from each DNA Extractions and PCR
colony were placed separately into 0.5 mL of 0.1% sterile tween- From each colony component sampled, unique isolate
20 PBS buffer solution in an autoclaved 1.5 mL Eppendorf morphotypes were chosen for DNA extraction and 16S
tube. For medium sized insects (T. angustula, Odontomachus rRNA sequencing. Standard cetyltrimethylammonium bromide
ants, and all wasps), pools of three adult insects or larvae were (CTAB) DNA extractions were performed for Actinobacteria
used. For P. clavata ants, only adult insects were collected for isolates (Cafaro et al., 2011). PCR amplifications were employed
Actinobacteria isolation. Each sample was vortexed for 15–20 s. on the genomic DNA using the eubacterial universal primers,
Tween-20 PBS solution was pipetted off, discarded and replaced 27F and 1492R (Lane, 1991) for 16S rRNA. As 16S rRNA gene
with 0.5 mL of PBS solution. The sample was macerated with sequence analysis cannot distinguish between closely related taxa
a sterile pestle and 0.5 mL of PBS was added. Then the (Han et al., 2012) the gyrB coding gene for 71 representative
microcentrifuge tube was vortexed for 15 s. An aliquot of 100 L Streptomyces isolates from our collection were sequenced, usingµ
was plated onto chitin agar plates and incubated at 27◦C for the primers gyrB-F1 (GAGGTCGTGCTGACCGTGCTGCA)
three to 4 weeks. Isolated Actinobacteria were subcultured onto and gyrB-R1 (GTTGATGTGCTGGCCGTCGACGT) (Hatano
yeast-malt extract agar (YMEA) with antifungals (nystatin and et al., 2003). The PCR amplification reactions were performed
cycloheximide) for macro and micromorphology examination. in a Veriti (Applied Biosystems) Thermal cycler. The reaction
A Gram stain screening of the isolates was performed prior to conditions for 16S rRNA gene amplification consisted of
◦ ◦ ◦
DNA extractions of the most promising isolates. A 20% glycerol 35 cycles at 94 C for 1 min, 52 C for 1 min, 72 C for
stock of Actinobacteria spores and cells were prepared by freezing 1 min. The amplification of gyrB consisted of 30 cycles of
a homogenized mixture in liquid nitrogen and storing them in denaturation at 95
◦C for 1 min, annealing for 0.5 min at 65◦C,
and extension at 72◦cryovials at 80◦C. C for 1.5 min. Bacterial 16S rRNA and gyrB−
sequences obtained in this study were deposited in the GenBank
database under accession numbers KY067229-KY067322 and
Entomopathogenic Fungi Sampling and KY082974-KY083044, respectively.
Isolation In the case of fugal genomic DNA extractions, a simple
A total of 14 insect cadavers with conspicuous signs of thermolysis method was applied (Zhang et al., 2010). Mycelium
Cordyceps-like fungal infection were collected and brought to from 3-day cultures from fast growing isolates was used to
the lab for processing. The majority (11) of specimens were extract DNA. In the case of Ophiocordyceps isolates PCFB and
collected in the same geographic area where insect colonies BA18 fresh stromata tip from old cultures were used. PCR
for Actinobacteria screening were sampled. Two wasps (Agelaia amplification of ITS region was performed with primers ITS4
areata and Polybia sp.) and one Camponotus ant cadavers were and ITS5 (White et al., 1990). SSU was amplified using the
collected in a secondary forest in San Carlos, Costa Rica (10◦ 34′ primers NS1 and NS4 (White et al., 1990). Amplification of the
26.2′′ N, 84◦30′ 08.2′′ W). Insect cadavers corresponding to the elongation factor 1-α (EF1-α) was performed with the primers
same species or genus as one of the Actinobacteria hosts under 983F and 2218R and cycling conditions used by Rehner and
study were chosen for fungal isolation, further characterization Buckley (2005). ITS amplification cycling conditions were of
and bioassay analysis. 35 cycles of 0.5 min at 95◦C, 0.5 min at 52◦C and 1.5 min
Ophiocordyceps infected specimens were of much interest as at 72◦C. SSU amplification was performed with 35 cycles of
these fungi have been described as being a hymenopteran host 94◦C 1 min, 52◦C 0.5 min and 72◦C for 1 min. PCR products
specialist having a narrow host range (Boomsma et al., 2014). cleaning and sequencing was performed either at Macrogen
Specimens were carefully collected, including a portion of the or at the Biotechnology Center DNA Sequencing Facility at
substrate (leaves or twigs) at which they were attached to avoid the University of Wisconsin-Madison. The GenBank accession
damaging the specimen. Samples were placed in sterile plastic numbers for the entomopathogenic fungi isolates sequences
Frontiers in Microbiology | www.frontiersin.org 4 October 2017 | Volume 8 | Article 2016
fmicb-08-02016 October 16, 2017 Time: 17:44 # 5
Matarrita-Carranza et al. Widespread Neotropical Insect-Actinobacteria Associations
are KY053448-KY053455 (ITS), KY055528-KY055532 (EF1-α) y by inoculation of 100 µl spore suspension (∼1 × 106 spore
KX579044-KX579053 (18S rRNA). mL−1) on a PDA plate and incubation in the dark at 27◦C and
counting germinated spores within 24 h (Lacey, 2012).
Analysis of Sequences PDA media was prepared and inoculated with the spore
Sequences from this study were edited using Geneious version suspension to reach a final approximate concentration of 1× 104
9.1.2 and high-quality sequences were clustered into operational viable spore mL−1. As we determined that temperatures above
taxonomic units (OTUs) using mothur (Schloss et al., 2009) 40◦C compromised spore viability of at least one of our fungal
with the standard divergence cut-off of 98% similarity (Schloss strains, the suspension spores were carefully preheated to 38◦C
and Handelsman, 2004; Hong et al., 2006; Hulcr et al., 2011). and inoculated in PDA media at around 38–40◦C. This avoid
Interactions bipartite network analysis were developed using the the formation of lumps when inoculating agar media with lower
R package-bipartite software (Dormann et al., 2008). temperature spore suspension, keeping a homogenous mixture
Sequence alignments were performed using GUIDANCE with viable spores. 24 mL PDA inoculated media was poured
(Penn et al., 2010) with MAFFT algorithm and 100 bootstrap in Petri plate (8.5 cm in diameter). For each fungal strain, the
repeats. Final editing of alignments was done in MEGA v6.0 panel of Actinobacteria isolates that corresponded to the same
(Tamura et al., 2013). Analyses for the best model of nucleotide insect host were cultured in YMEA at 27◦C for 2 weeks to reach
substitution were performed using jModelTest2 version 2.1.6 the stationary phase growth. Then YMEA with Actinobacteria
(Darriba et al., 2012). The chosen model for 16S rRNA and gyrB lawn growth was cut into 5 mm diameter disks and carefully
sequences was GTR (General Time Reverse) and K80 (Kimura transferred to the PDA media inoculated with entomopathogenic
two-parameter) for ITS sequences. Phylogenetic analyses were fungal spore suspension. Bioassays plates were incubated at 27◦C
performed through Bayesian inference, using MrBayes 3.2 for 2–4 weeks and halos of inhibition were measured. Negative
(Ronquist and Huelsenbeck, 2003). All analyses employed one controls were run with YMEA agar disks media only.
cold chain and three incrementally heated chains. A temperature
parameter set to 0.2 for ribosomal RNA genes and 0.1 for gyrB. Scanning Electron Microscopy
Four separate Markov Chain Monte Carlo runs were performed, The ultrastructure of adults and nest material for two vespid
with 3 million generations each (5 million generations for gyrB), wasps (M. docilis and P. plebeja) was examined. Two nest
discarding the initial 25% generations from each run as burn- samples and adults per colony were fixed with glutaraldehyde
in and sampling one in every 100 generations to calculate and paraformaldehyde and were left at 4◦C overnight. Samples
posterior probabilities for each branch. jModelTest and MrBayes were post-fixed in 1% osmium tetroxide for at least 1 h and
analysis were carried out using the Cipress Science Gateway dehydrated with ethanol (SEM; 30–100%) SEM and then dried
bioinformatics service platform (Miller et al., 2010). Final editing by sublimation in a freeze dryer (VFD-20, VD Inc.). Due to
of each phylogenetic tree was done in FigTree v1.4 and Adobe the fragile nature of these samples, the processing was repeated
Illustrator CS5.1. with a modified Karnovsky and Osmium vapors fixation protocol
Bioassays (Kim, 2008; Alves et al., 2013). In this case, samples were putin a Petri plate containing a moistened cotton ball. In a fume
Duplicate bioassay challenges between Actinobacteria and hood, 1 mL of Karnovsky’s solution (Karnovsky, 1965) was placed
entomopathogenic fungi were conducted. In each bioassay, both in a small opened container that was then transferred into the
the bacteria and the fungi were isolated from the same insect host Petri plate containing the sample. The Petri plate was kept closed
at the genus or species level. To obtain spore suspensions, fungal and the sample was fixed at room temperature for 2 h. For
fermentations were done with a modified methodology (Suay a second fixation step, the container with Karnovsky solution
et al., 2000). In the case of Ophiocordyceps/Hirsutella isolates, was replaced with another container that contained 1 mL of 2%
pieces of agar from pure cultures were transferred to 100 mL osmium tetraoxide. The Petri plate was covered with aluminum
sterile Erlenmeyer’s flasks containing 20 mL of Grace Insect cell foil inside the fume hood and the sample was fixed for 12–24 h
medium (Gibco) supplemented with 10% fetal bovine serum and then dried with silica gel in a hermetic container for 24 h.
(Gibco). The inoculated flasks were shaken at 250 rpm at 25◦C The samples were then mounted on aluminum stubs, with a
for 2–3 weeks. Spore suspensions were prepared by filtering the double-stick carbon, coated with gold in a sputter Eiko ID 2 and
liquid culture through sterile cheesecloth placed in the tip of examined under a scanning electron microscope Hitachi S-3700.
a sterile pipette while applying pressure with a pipette pump.
Isolates classified in the genera Metarhizium, Chlorocillium, and
Engyodontium were cultured directly in PDA plates for 3 weeks
at 27◦C. Spore suspensions were obtained by scrapping the RESULTS
mycelium directly from the agar media and transferring to 100 ml
sterile Erlenmeyer flasks containing 20 mL of 0.1% tween-20 Actinobacteria Isolation and
solution. Eight sterile crystal balls (3 mm diameter) were added to Phylogenetic Analysis
the flasks and then shaken at 250 rpm for 10 min. The mycelium In our sampling of 178 colonies of tropical hymenopteran insects,
was filtered through sterile cheesecloth and spore suspension 197 isolates of Actinobacteria were obtained. Of the 29 different
collected in a sterile beaker. Suspension spore concentration was species of insects sampled, Actinobacteria was obtained from 22
determined with a hemocytometer. Spore viability was assessed (Table 1). Isolation of Actinobacteria from a particular species
Frontiers in Microbiology | www.frontiersin.org 5 October 2017 | Volume 8 | Article 2016
fmicb-08-02016 October 16, 2017 Time: 17:44 # 6
Matarrita-Carranza et al. Widespread Neotropical Insect-Actinobacteria Associations
FIGURE 1 | Insect-Actinobacteria interaction network using OTUs clustering with 2% cut-off sequence similarity. Upper level boxes represent insect species and
lower level boxes depict associated Actinobacteria OTUs. Boxes and link size are proportional to the total number of isolates obtained and to the frequency of this
particular interaction, respectively. Genera are abbreviated as follow: Streptomyces (Str), Pseudonocardia (Ps) Amycolatopsis (Amy) Nocardia (Noc) Saccharothrix
(Sac).
was only unsuccessful for species where sampling involved only OTUs accounted for 75% of all sequences included in the
a single colony, except in the case of the bee Trigona sp. 1 analysis. Two Pseudonocardia strains associated with bees and
In comparing bees (34 colonies), ants (92 colonies), wasps (44 three Amycolatopsis strains associated with ants were also found.
colonies), 23, 82, and 70 strains of Actinobacteria, were isolated Within the Actinobacteria genus Nocardia, one strain was
respectively. By grouping the isolates according to the colony isolated from a bullet ant (P. clavata) and wasp larvae (A.
component from which they were obtained, 74 corresponded cajennensis) and a different strain obtained from nest material
to adult insects (176 samples), 68 with immature insects (131 of a trap-jaw ant colony (O. erythrocephalus). Additionally, one
samples), and 45 with nest material (149 samples). In addition, Sacchrothrix strain was isolated from O. erythrocephalus larvae
from the stingless bee T. angustula other bee hive components (Figure 1).
were sampled and isolates obtained as follow: three from honey, Four main clades of Actinobacteria, each one composed
one from brood comb, one from pollen and five from propolis of sequences with 99% 16S rRNA identity were found to
(Table 1). be associated with different groups of insects under study
From the 197 Actinobacteria-like isolates obtained, we choose (Figure 2). In general, phylogenetic analysis of gyrB, grouped
only unique morphotypes per colony component sampled (174), representative sequences from the 16S rRNA tree in the same
for DNA extraction and 16S rRNA sequence analysis. From clades (Figures 2, 3). In the 16S rRNA phylogenetic tree, the
those isolates 161 successfully amplified, 23 were replicates, and paraphyletic group I represents a single 16S rRNA phylotype
138 Actinobacteria sequences were chosen for further analysis. isolated repeatedly from 29 different samples, from nine different
A sequence alignment was generated and then 44 short sequences hymenopteran insect species (Figure 2). This group contained
were removed and the remaining 94 were clustered into 22 16 Streptomyces isolates associated with vespid wasp colonies, 11
different OTUs with a 2% cutoff used as an indicator of bacterial from ants and 2 from the stingless bee T. angustula. The closest
species (Figure 1, see Supplementary Table S1 for details about relatives to this group were S. fulvissimus DSM 43, S. flavogriseus
the distribution of isolates differentiating their origin) (Hong ATCC 33331, S. globisporus KTC 9026 and S. cavourensis NRRL
et al., 2006; Hulcr et al., 2011). Interaction network analysis using 2740, based on >99% sequence similarity. Moreover, sequences
OTUs clustering did not show evidence of specific associations from other studies corresponding to known host associated
between Actinobacteria isolates and the insect species from Streptomyces strains with antimicrobial (Poulsen et al., 2011) or
which they were obtained (Figure 1). Streptomyces isolates cellulolytic activity (Book et al., 2016) were clustered in clade
were grouped into 22 different OTUs with ten of them being I (Figure 2). In the gyrB phylogeny, representative sequences
represented by singletons (Figure 1). Six dominant Streptomyces within clade I were resolved into different taxonomic groups
Frontiers in Microbiology | www.frontiersin.org 6 October 2017 | Volume 8 | Article 2016
fmicb-08-02016 October 16, 2017 Time: 17:44 # 7
Matarrita-Carranza et al. Widespread Neotropical Insect-Actinobacteria Associations
FIGURE 2 | Bayesian, phylogenetic analysis of 16S rRNA sequences of Actinobacteria associated with different hymenopteran insect species. Insect taxonomic
identity is indicated next to sample code. Posterior probabilities (PP) are indicated at the corresponding nodes. NCBI reference sequences included in the analysis
are denoted by the corresponding type culture collection acronym. Clades within 99% sequence similarity cutoff are highlighted except for clade II that it is
composed of different phylotypes and was highlighted to show a group of sequences closely related to the well-studied solitary wasp symbiont Candidatus
Streptomyces philanthi. Alignment 950 pb.
Frontiers in Microbiology | www.frontiersin.org 7 October 2017 | Volume 8 | Article 2016
fmicb-08-02016 October 16, 2017 Time: 17:44 # 8
Matarrita-Carranza et al. Widespread Neotropical Insect-Actinobacteria Associations
FIGURE 3 | Bayesian, phylogenetic analysis of partial gyrB sequences of Streptomyces associated with different hymenopteran insect species. Insect taxonomic
identity is indicated next to sample code. Posterior probabilities (PP) are indicated at the corresponding nodes. NCBI reference sequences included in the analysis
are denoted by the corresponding type culture collection acronym. Representative isolates with 99% sequence similarity cutoff in the 16S rRNA phylogeny are
highlighted as well as clade II that it is composed of different phylotypes. Clade II was highlighted to show a group of sequences closely related to the well-studied
solitary wasp symbiont Candidatus Streptomyces philanthi. Alignment 780 pb.
Frontiers in Microbiology | www.frontiersin.org 8 October 2017 | Volume 8 | Article 2016
fmicb-08-02016 October 16, 2017 Time: 17:44 # 9
Matarrita-Carranza et al. Widespread Neotropical Insect-Actinobacteria Associations
FIGURE 4 | SEM image of the interior of wasp brood chambers. (A) Two brood chambers of M. docilis, after larvae removal. Scale, 1 mm. (B) Filamentous bacteria
forming a dense cluster at the base of the chamber, hyphae diameters of 1.3 – 2.4 µm. Scale, 20 µm. (C) P. plebeja larvae in brood chambers. Scale, 1 mm.
(D) Streptomyces like hyphae growing on P. plebeja larvae. Hyphae diameter of 1.2 – 1.7 µm. Scale, 60 µm.
FIGURE 5 | Bayesian phylogenetic relationships among fungal isolates in this study as inferred from ITS-5.8S sequences. Posterior probability (PP) at the
corresponding nodes. The tree was rooted on Candida albicans.
Frontiers in Microbiology | www.frontiersin.org 9 October 2017 | Volume 8 | Article 2016
fmicb-08-02016 October 16, 2017 Time: 17:44 # 10
Matarrita-Carranza et al. Widespread Neotropical Insect-Actinobacteria Associations
FIGURE 6 | Ophiocordyceps fungi infecting P. clavata ants. (A) Ophiocordyceps kniphofioides (H. stilbelliformis) PCFB in the sexual state. Scale, 1 cm (B)
H. stilbelliformis PCFB isolate pure culture on PDA, showing dense growth of synnemata. (C) H. stilbelliformis PCFB ovoid conidia borne at compacted terminal
phialides. Scale, 20 µm (D) SEM micrograph of PCFB verrucose apical phialide with conidias inside mucous droplet. Scale, 4 µm. (E) a- and b- phialides and
conidia of Hirsutella stilbelliformis PCFB. Scale, 20 µm. (F) P. clavata ant, infected by Ophiocordyceps sp. Scale, 1 cm. (G) Isolate BA18 in PDA pure culture.
(H) BA18, verrucose compact and apical phialides with ovoid conidia. Scale, 20 µm.
but within a single clade with the Streptomyces strains with high the vespid wasp P. plebeja, six different colonies were studied
cellulolytic activity (Figure 3). from which were obtained 11 Streptomyces isolates associated
Clades III, IV, and V also represent different 16S rRNA with brood or nest samples. Four of these isolates were isolated
phylotypes (99% cut off) with their closest relatives being from different colonies and grouped together as one 16S rRNA
S. prasinopilosus, S. sampsonii, and S. misionensis, respectively. Streptomyces phylotype in Clade I (Figure 2), suggesting being a
Sequences corresponding to Streptomyces isolates associated with common strain associated with this wasp. The SEM ultrastructure
subsocial Crabronidae or Pompilidae insects grouped together of P. plebeja brood chambers was also studied and bacteria with
in the monophyletic Clade II (Figure 3) with Candidatus Streptomyces morphology were found growing on the larvae
Streptomyces philanthi, the solitary wasp Philanthus triangulum (Figures 4C,D). These hyphae have diameters between 1.2 and
symbiont (Kaltenpoth et al., 2005, 2014). 1.7 µm. The ultrastructure of adult M. docilis wasp cuticle was
also studied; however, evidence of microorganism growth was
not found. Similarly, no hyphae with morphology similar to
Electron Microscopy for Actinobacteria Streptomyces were seen when samples from adults, immature
Localization in Brood Chambers and stages, and nest material from the vespid A. cajennensis and adult
Insect’s Cuticle ants within the species O. erythrocephalus and P. clavata were
From the 24 Actinobacteria isolates associated with the wasps seen under SEM.
M. docilis, 62% were obtained from the nest substrate.
Scanning electronic micrographs from brood chambers from two Isolation and Identification of
different M. docilis nests revealed the presence of filamentous Entomopathogenic Fungi
microorganisms (Figure 4B). Vegetative hyphae with diameters The ITS region from seven fungal isolates that were
of 1.3 – 2.4 µm was abundant at the base of the brood chamber obtained from insect cadavers was successfully amplified
but was not observed externally (Figures 4A,B). In the case of and a phylogenetic tree was constructed (Figure 5). These
Frontiers in Microbiology | www.frontiersin.org 10 October 2017 | Volume 8 | Article 2016
fmicb-08-02016 October 16, 2017 Time: 17:44 # 11
Matarrita-Carranza et al. Widespread Neotropical Insect-Actinobacteria Associations
FIGURE 7 | Paraponera clavata ant parasitized by fungi. (A) Healthy P. clavata ant. (B) Infected ant, arrow pointing at the stroma that grew on the back of the ant’s
head with no perithecia, note the dense yellowish aerial hyphae with chains of conidia growing on the ant’s body. Scale, 5 mm (C) Isolate BAI obtained from the
P. clavata cadaver. (D) SEM micrograph, showing the conidia chains growing directly from the ant’s infected body. Scale, 2 µm.
isolates grouped with the families Ophiocordycipitaceae, (Figure 6H) was examined. A Metarhizium strain isolated from
Metacordycipitacea, and other closely related groups. an O. erythrocephalus ant, was included in our entomopathogen
PCFB and BA18, two fungal strains isolated from P. clavata collection. This asymptomatic specimen was collected in the field
ants were classified by macro and micromorphology as well as and brought to the lab and later died developing Metarhizium like
ITS, SSU and E-1α sequences, within the Ophiocodycipitaceae fungal growth. The identity of this fungus was confirmed by ITS
family where all Cordyceps fungi infecting ants are grouped. sequence.
Isolate PCFB was obtained from a specimen with fungal growth BAI was an isolate obtained from an apparently
in its teleomorph stage, with a characteristic single brown stroma hyperparasitized P. clavata specimen presenting a brown
with orange fertile tip emerging from the back of the ant’s dried stroma stalk extending from the back of the ant’s head
head (Figure 6A). This isolate was identified as O. kniphofioides and a dense yellowish aerial mycelium with spores growing
(H. stilbelliformis) with a 98% Ef-1α sequence identity. Moreover, out from the joints of the cadaver (Figure 7). Isolate BAI was
microscopic study of coordinated hyphal structures (synnemata) grouped together with Chlorocillium spp. by its ITS and 18S
of isolate PCFB, showed the presence of distinctive verrucose gene sequencing. Isolates CdyAgAr and CdyAvB, obtained
hyphae, phialides, and conidia (Figures 6B–E) (Evans and from wasp specimens collected at San Carlos were identified as
Samson, 1984). Isolate BA18 (Figure 6G) was obtained from Hirsutella sp. (Figure 8). For these isolates, ITS and 18S rRNA
a P. clavata ant attached by its mandibles to the underside genes were sequenced, and both genes shared 99% identity with
of a palm leaf (Figure 6F) and the fungus was growing out H. citriformis when compared against sequences from NCBI
of the ant’s body in its anamorphic state. Isolate BA18 was databases. Sequence similarity, as well as macro and microscopic
recognized as a close relative to O. kniphofioides sharing 95% morphology examination, suggest that both wasp specimens
Ef-1α and 98% 18S partial sequence similarity when compared were infected by the same fungal species. Finally, a fungal
against the closest relatives sequence available at NCBI database. isolate classified by ITS and SSU sequences as Engyodontium sp.
Conidiophores and conidia structures similar to isolate PCFB was included in the study. This isolate was obtained from an
were found when the micromorphology of BA18 synnemata Agelaia wasp cadaver infected by an Ophiocordyceps fungus that
Frontiers in Microbiology | www.frontiersin.org 11 October 2017 | Volume 8 | Article 2016
fmicb-08-02016 October 16, 2017 Time: 17:44 # 12
Matarrita-Carranza et al. Widespread Neotropical Insect-Actinobacteria Associations
FIGURE 8 | Wasps infected with Hirsutella fungi. (A) A. areata cadaver. Note solitary stromata, filiform, arising from thorax to the abdomen region of the host, erect
or curved and of brown color. Scale, 2 mm. (B) Ultrastructure of stromata from isolate CdyAgAr showing phialides and citriform conidia. Scale, 20 µm. (C) Polybia
sp. wasp attached to a leaf by their mandibles. Scale, 1 mm. (D) Light micrograph of stromata from CdyAVB isolate in PDA pure culture. Scale, 20 µm.
was collected at La Selva in November 2014. Although, to our All Actinobacteria isolates with antifungal activity that had
knowledge there is no evidence of Engyodontium fungus being their 16S rRNA sequence included in the phylogenetic study were
pathogens of hymenopteran insects, there have been reports mapped in the tree (Figure 2). From the four Actinobacteria
of this fungal genus being a pathogen of another insect group isolates not belonging to the genus Streptomyces that were
(Samson et al., 2013). screened for antifungal properties, only B86 (identified in the
genus Saccharothrix) was effective in inhibiting the growth of the
Bioassays: Actinobacteria Strains with fungal strain that was tested against (Metarhizium).
Antifungal Properties against Insects’
Natural Enemies of Fungal Origin DISCUSSION
In total 125 bioassays were carried out in duplicates between
Actinobacteria and entomopathogenic fungi. Only some of the Our results indicate that Actinobacteria associations with
Actinobacteria isolates were used in bioassays and were chosen hymenopteran insects are more common than previously
so that both the bacteria and the fungi were obtained from described (Currie et al., 1999; Kaltenpoth et al., 2005; Promnuan
the same insect host, at the species or genus level. These et al., 2009; Poulsen et al., 2011; Madden et al., 2013; Reyes and
bioassays were categorized into seven different groups each Cafaro, 2014). With SEM analysis, the presence of filamentous
one corresponding to a different fungus. For every group of microorganisms with cell morphology similar to Streptomyces
bioassays, at least three different Actinobacteria isolates that inside M. docilis and P. plebeja brood chambers (Figure 4)
could inhibit the growth of the pathogenic fungi in vitro were confirmed. Moreover, Actinobacteria from most insect
(Figure 9) were found. The highest proportion of antifungal species sampled were successfully isolated, including insects
activity was seen for bioassays between P. clavata Actinobacteria from five different Hymenopteran families, which represent
and the fungal isolate BA18. In this group of bioassays, 10 eusocial (paper wasps, ants, and the stingless bee, T. angustula)
out 12 Actinobacteria isolates inhibited the growth of the and non-eusocial lifestyles (orchid bees, mud dauber wasps
Ophiocordyceps fungus. In the case of P. clavata Actinobacteria in the family Crabronidae and a spider wasp in the family
and the PCFB (O. kniphofioides) fungal isolate, 8 out 15 bacterial Pompilidae). The highest proportion of isolates per colony was
isolates produced halos of inhibition. Similarly, in A. cajennensis obtained in Trypoxylon sp. wasps, followed by paper wasps
and Hirsutella bioassays, 12 out of 23 isolates presented antifungal in the tribe Epiponini (M. docilis, P. plebeja, P. occidentalis,
activity. and A. cajennensis). In these cases, Actinobacteria were isolated
Frontiers in Microbiology | www.frontiersin.org 12 October 2017 | Volume 8 | Article 2016
fmicb-08-02016 October 16, 2017 Time: 17:44 # 13
Matarrita-Carranza et al. Widespread Neotropical Insect-Actinobacteria Associations
from all colony compartments sampled (brood, adults and
nest material). This was also the case for ant species such as
O. erythrocephalus, C. longispina, P. caeciliae, and T. ramulorum
inrectum. On the other hand, with the exception of the meliponid
bee Trigona sp. 1 (four colonies sampled), all insect species that
did not yield Actinobacteria were sampled only once. In the case
of Trigona bees, only adults and material from nest entrances
were sampled, as the colonies studied were located inside tree
trunk cavities. It is possible that no Actinobacteria were isolated
from Trigona bees because of the inability to sample other colony
components. It has been suggested that the role of defensive
Actinobacteria symbionts could be linked to specific vulnerable
life stages of their host (e.g., larvae and pupae; Kaltenpoth and
Engl, 2013). Likewise, in a different study, Promnuan et al.
(2009) reported the isolation of Actinobacteria from brood cells
of Trigona bee hives. Consequently, we obtained 12 actinomycete
isolates from the internal colony components of a related stingless
bee, T. angustula.
Streptomyces was the predominant genus of culturable
Actinobacteria associated with insect colonies in the present FIGURE 9 | Actinobacteria isolates with antifungal properties againstentomopathogenic fungi. In gray, the number of isolates with antifungal
study. These results are consistent with similar studies employing properties are shown and in white the ones without activity. OEMC3
culture-dependent isolation approaches in Apis and Trigona (Metarhizium isolate against Actinobacteria obtained from Odontomachus
bees (Promnuan et al., 2009), Dendroctonus beetles (Hulcr ants), CdyAgAR (H. citriformis isolate against Actinobacteria obtained from
et al., 2011), Polistes dominulus wasps (Madden et al., 2013) A. cajennensis), CdyAvB (H. citriformis against Actinobacteria isolated form
P. occidentalis y P. plebeja wasps), AC2 (Engyodontium sp. against
and Pseudomyrmex penetrator, Petalomyrmex phylax, and Actinobacteria associated with A. cajennensis), BAI (Chlorocillium sp.
Crematogaster margaritae ants as well as culture-independent challenged against P. clavata Actinobacteria isolates), PCFB (O. kniphofioides
techniques, as the case for P. longicornis (Reyes and Cafaro, against P. clavata Actinobacteria) y BA18 (Ophiocordyceps sp. against
2014). Moreover, our interaction network analysis suggests P. clavata Actinobacteria).
that the dominant Streptomyces phylotypes had diverse
associations with hymenopteran hosts at both the genus
and family levels (Figure 1). The binning of isolates from potential nutritional role of these Actinobacteria in their host
different insect groups into a few OTUs imply that possible niche.
horizontal transmission and selective processes may have To test the resolution of our 16S rRNA tree, the gyrB gene from
occurred. These results agree with other well-studied insect- representative isolates included in our 16S rRNA phylogeny were
Actinobacteria defensive symbiosis in which the symbiont has amplified and sequenced. Our 16S rRNA and gyrB phylogenies
undergone dynamic substitution over the evolutionary history had corresponding sequences grouped in the same clades
of their host (Kaltenpoth et al., 2010, 2012; Cafaro et al., 2011; with minor changes related to more variability in sequence
Kaltenpoth and Engl, 2013). Although some Actinobacteria composition in clade I from the gyrB tree (Figure 3). As pointed
strains classified into genera different than Streptomyces were out elsewhere, among important housekeeping genes markers,
isolated, our culture-dependent screening approach may gyrB produced higher correlations with genome relatedness
have underestimated the insect host Actinobacteria diversity. in phylogenetic analysis with Actinobacteria belonging to the
Independent cultivation techniques are necessary to address the Streptomycetaceae family (Han et al., 2012). When comparing
real diversity, abundance and stability of Actinobacteria in these the phylogenetic trees, there is evidence of apparently unique
systems. phylotypes represented in clades III, IV, and V (Figures 2, 3)
Our results indicate that 1/3 of all the isolates obtained in in both phylogenies. These monophyletic clades consist of one
our sampling occur within clade I of our 16S rRNA phylogenetic taxonomic unit that was isolated repeatedly from different insect
tree. This clade corresponds to a lineage of highly cellulolytic hosts. Moreover, isolates within these clades that were included
Streptomyces previously identified in temperate zones, which in bioassays showed antifungal properties against different
is associated with a variety of insect species that feed on pathogenic fungi, suggesting that these isolates shared similarities
plant biomass (Book et al., 2016). The high cellulolytic activity in terms of production of antifungal compounds active against
present in these bacteria appears to be the result of evolutionary natural enemies of their host.
adaptations acquired through horizontal transfer and selective Our phylogenetic analysis results also suggest that subsocial
retention of genes as well as the expansion of their regulatory wasps included in our study and the Streptomyces isolates
elements (Book et al., 2016). Further studies evaluating the obtained from them are taxonomically related to P. triangulum
cellulolytic activity and genomic content enrichment for key and its mutualistic symbiont Ca. Streptomyces philanthi
enzyme families related to plant biomass degradation in our respectively, showing congruency in both Actinobacteria-
isolates from clade I would provide important insight into this insect association systems. These findings require further
Frontiers in Microbiology | www.frontiersin.org 13 October 2017 | Volume 8 | Article 2016
fmicb-08-02016 October 16, 2017 Time: 17:44 # 14
Matarrita-Carranza et al. Widespread Neotropical Insect-Actinobacteria Associations
FIGURE 10 | Bioassay, antifungal activity screening from Actinobacteria isolates associated with P. clavata against O. kniphofioides. The Actinobacteria isolate codes
are shown. Negative controls are marked with the letter “B.” For illustrative purposes the photographs were taken after 4 weeks of incubation because of the slow
growing of this fungus.
study, as the P. triangulum-Streptomyces defensive symbiosis from P. clavata ants could inhibit the ant’s parasitic fungus
is thought to be confined into the subfamily Philanthinae O. kniphofioides (Figure 10).
(Kaltenpoth et al., 2012), whereas Trypoxylon is a genus Two H. citriformis (Ophiocordycipitaceae: Hypocreales)
belonging to the subfamily Crabroninae. Both of these strains were isolated from two different vespid wasps (Figure 8).
subfamilies are classified in different monophyletic clades Hirsutella is a genus of anamorphic fungi whose teleomorphs
in the diverse paraphyletic family Crabronidae (Debevec et al., has commonly been linked to the genus Ophiocordyceps (Sung
2012). et al., 2007). H. citriformis is a known pathogen of insects in
One important goal of this study was to test if the the orders Hemiptera and Psocoptera (Simmons et al., 2015).
Actinobacteria associated with insects can protect them against To our knowledge, this is the first report on H. citriformis
natural enemies. Different strains of Ophiocordyceps (Hirsutella) infecting Hymenopteran insects. In this study, H. citriformis
fungi were isolated and blastospores were obtained in high was also inhibited by several wasp associated Actinobacteria
concentrations to screen for Actinobacteria antifungal activity strains. Our results suggest that the antifungal substances
in vitro. All of the insect species examined (those from which produced by symbiotic Actinobacteria may occur in the natural
entomopathogenic fungi were isolated) had more than three environment, but additional studies are necessary to demonstrate
different associated Actinobacteria strains that were capable of their presence.
producing zones of inhibition in co-culture in vitro bioassays This study shows associations of a few Streptomyces lineages
(Figure 9). with diverse neotropical hymenopteran insects, one of them
Cordyceps fungi are characterized by being host specialized being previously described in temperate zones, suggesting a
parasites. For example, the zombie ant fungus Ophiocordyceps consistent association that spans from temperate to tropical
unilateralis is an antagonist of ants in the tribe Camponotini regions. Moreover, the results of the present study provide
(Formicinae, Formicidae), with less common infections reported evidence of antagonism between some Actinobacteria strains
on Polyrhachis ants (Pontoppidan et al., 2009; Andersen et al., and the fungal entomopathogens tested and supports the idea
2012). Similarly, Ophiocordyceps kniphofioides infects Cephalotes that certain insects may use some of the active secondary
atratus (Formicidae: Myrmicinae), Paraponera clavata, and metabolites from associated Actinobacteria as an additional
Dinoponera longipes (Formicidae: Ponerinae) ants (Evans and protection mechanism against natural enemies. Collectively,
Samson, 1982; Sanjuan et al., 2015). In this study, it has these findings provide insight into the widespread and dynamic
been shown that different strains of Streptomyces isolated host-niche selection of Actinobacteria and points to promising
Frontiers in Microbiology | www.frontiersin.org 14 October 2017 | Volume 8 | Article 2016
fmicb-08-02016 October 16, 2017 Time: 17:44 # 15
Matarrita-Carranza et al. Widespread Neotropical Insect-Actinobacteria Associations
models of mutualistic symbiosis, which harbor great potential University of Costa Rica; Resolution Number 012); and
in finding new biologically active compounds with medical and authorized by La Selva Biological Station and Las Brisas Nature
biotechnological applications. Reserve. The authors thank Dr. Shanna Gofredi, Dr. Sean
O’Donnell, and Dr. Ralph Saporito for critical comments on the
manuscript and Dr. Fernando García and Dr. César Rodríguez
AUTHOR CONTRIBUTIONS for advice and support. Terry McGlynn for facilitating the use of
personal laboratory equipment and Dr. Carlos García Robledo
Conceived and designed the experiments: CC, AP-T, MM, for help on interaction network analysis. Ludovic Le Renart and
and BM-C. Performed the experiments: BM-C. Processed SEM David Hughes for advice on fungal taxonomy and physiology.
samples: RM-S. Analyzed the data: BM-C, CC, and AP-T. Ronald Vargas for field work and insect identification. Natalie
Contributed reagents/materials/analysis tools: CM-C and MM. Whitehead, Jane Li, and Mauricio García for field assistance.
Wrote the paper: BM-C, CC, and AP-T. Danilo Brenes, Jessica Smith, Marcela Fernandez, Mariel Víquez
and many other members of the “Grupo de Investigación
FUNDING Simbiosis Hospedero-Microorganismo” for their help withlab work. Dr. Deedra McClearn, Dr. Carlos de la Rosa and
Authors gratefully acknowledge financial support from the staff members of La Selva Biological Station (Organization for
National Science Foundation, United States (MCB-0702025), Tropical Studies) for logistic support on field and lab work on
Sistema de Estudios de Posgrado y Vicerrectoría de Investigación site; MSc. Ethel Sánchez for advice on electron microscopy and
de la Universidad de Costa Rica (research projects 801-B0-538 to the scientific personnel at CIEMic and CIBCM, Universidad
and 810-B3-185), and the National Institute of Health, de Costa Rica, for providing the use of research facilities.
United States (U19 Al109673).
SUPPLEMENTARY MATERIAL
ACKNOWLEDGMENTS
The Supplementary Material for this article can be found
Collecting permits were granted by the “Comisión Institucional online at: https://www.frontiersin.org/articles/10.3389/fmicb.
de Biodiversidad” (Institutional Biodiversity Committee, 2017.02016/full#supplementary-material
REFERENCES metapleural glands of leaf-cutting ants. Insectes Soc. 49, 363–370. doi: 10.1007/
PL00012660
Aguiar, A. P., Deans, A. R., Engel, M. S., Forshage, M., Huber, J. T., Jennings, J. T., Cafaro, M. J., and Currie, C. R. (2005). Phylogenetic analysis of mutualistic
et al. (2013). Order Hymenoptera. Zootaxa 3703, 51–62. doi: 10.1603/EN09221 filamentous bacteria associated with fungus-growing ants. Can. J. Microbiol. 51,
Alves, E., Gilvaine Ciavareli, L., Ampélio Pozza, E., and de Carvalho, M. (2013). 441–446. doi: 10.1139/w05-023
“Scanning electron microscopy for fungal sample examination,” in Laboratory Cafaro, M. J., Poulsen, M., Little, A. E. F., Price, S. L., Gerardo, N. M., Wong, B.,
Protocols in Fungal Biology: Current Methods in Fungal Biology, eds G. V. et al. (2011). Specificity in the symbiotic association between fungus-growing
Kumar, M. G. Tuohy, M. Ayyachamy, K. M. Turner, and A. O’Donovan ants and protective Pseudonocardia bacteria. Proc. Biol. Sci. 278, 1814–1822.
(New York, NY: Springer), 133–150. doi: 10.1007/978-1-4614-2356-0_8 doi: 10.1098/rspb.2010.2118
Andersen, S. B., Ferrari, M., Evans, H. C., Elliot, S. L., Boomsma, J. J., and Hughes, Cardinal, S., Danforth, B. N., Pitts, J., Gillespie, J., and Cameron, S. (2011).
D. P. (2012). Disease dynamics in a specialized parasite of ant societies. PLOS The antiquity and evolutionary history of social behavior in bees. PLOS ONE
ONE 7:e36352. doi: 10.1371/journal.pone.0036352 6:e21086. doi: 10.1371/journal.pone.0021086
Barka, E. A., Vatsa, P., Sanchez, L., Gaveau-Vaillant, N., Jacquard, C., Klenk, H.-P., Currie, C. R., Poulsen, M., Mendenhall, J., Boomsma, J. J., and Billen, J. (2006).
et al. (2016). Taxonomy, physiology, and natural products of Actinobacteria. Coevolved crypts and exocrine glands support mutualistic bacteria in fungus-
Microbiol. Mol. Biol. Rev. 80, 1–43. doi: 10.1128/MMBR.00019-15 growing ants. Science 311, 81–83. doi: 10.1126/science.1119744
Bawa, K. S., McDade, L. A., and Hespenheide, H. (1994). La Selva: Ecology and Currie, C. R., Scott, J. A., Summerbell, R. C., and Malloch, D. (1999). Fungus-
Natural History of a Neotropical Rain Forest. Chicago, IL: University of Chicago growing ants use antibiotic-producing bacteria to control garden parasites.
Press. Nature 398, 701–704. doi: 10.1038/19519
Bérdy, J. (2005). Bioactive microbial metabolites. J. Antibiot. 58, 1–26. doi: 10.1038/ Darriba, D., Taboada, G. L., Doallo, R., and Posada, D. (2012). jModelTest 2:
ja.2005.1 more models, new heuristics and parallel computing. Nat. Methods 9, 772.
Berg, R. (1996). The indigenous gastrointestinal microflora. Trends Microbiol. 4, doi: 10.1038/nmeth.2109
430–435. doi: 10.1016/0966-842X(96)10057-3 de Bekker, C., Quevillon, L. E., Smith, P. B., Fleming, K. R., Ghosh, D., Patterson,
Bhattacharya, D., Nagpure, A., and Gupta, R. K. (2007). Bacterial chitinases: A. D., et al. (2014). Species-specific ant brain manipulation by a specialized
properties and potential. Crit. Rev. Biotechnol. 27, 21–28. doi: 10.1080/ fungal parasite. BMC Evol. Biol. 14:166. doi: 10.1186/s12862-014-0166-3
07388550601168223 de Souza, D. J., Bézier, A., Depoix, D., Drezen, J.-M., and Lenoir, A. (2009).
Book, A. J., Lewin, G. R., McDonald, B. R., Takasuka, T. E., Wendt-Pienkowski, E., Blochmannia endosymbionts improve colony growth and immune defence
Doering, D. T., et al. (2016). Evolution of high cellulolytic activity in symbiotic in the ant Camponotus fellah. BMC Microbiol. 9:29. doi: 10.1186/1471-21
Streptomyces through selection of expanded gene content and coordinated gene 80-9-29
expression. PLOS Biol. 14:e1002475. doi: 10.1371/journal.pbio.1002475 Debevec, A. H., Cardinal, S., and Danforth, B. N. (2012). Identifying the sister
Boomsma, J. J., Jensen, A. B., Meyling, N. V., and Eilenberg, J. (2014). Evolutionary group to the bees: a molecular phylogeny of Aculeata with an emphasis on
interaction networks of insect pathogenic fungi. Annu. Rev. Entomol. 59, the superfamily Apoidea. Zool. Scr. 41, 527–535. doi: 10.1111/j.1463-6409.2012.
467–485. doi: 10.1146/annurev-ento-01163-162054 00549.x
Bot, A., Ortius-Lechner, D., Finster, K., Maile, R., and Boomsma, J. J. (2002). DeMera, J. H., and Angert, E. R. (2004). Comparison of the antimicrobial activity
Variable sensitivity of fungi and bacteria to compounds produced by the of honey produced by Tetragonisca angustula (Meliponinae) and Apis mellifera
Frontiers in Microbiology | www.frontiersin.org 15 October 2017 | Volume 8 | Article 2016
fmicb-08-02016 October 16, 2017 Time: 17:44 # 16
Matarrita-Carranza et al. Widespread Neotropical Insect-Actinobacteria Associations
from different phytogeographic regions of Costa Rica. Apidologie 35, 411–417. Kroiss, J., Kaltenpoth, M., Schneider, B., Schwinger, M.-G., Hertweck, C., Maddula,
doi: 10.1051/apido:2004033 R. K., et al. (2010). Symbiotic streptomycetes provide antibiotic combination
Dillon, R. J., and Dillon, V. M. (2004). The gut bacteria of insects: nonpathogenic prophylaxis for wasp offspring. Nat. Chem. Biol. 6, 261–263. doi: 10.1038/
interactions. Annu. Rev. Entomol. 49, 71–92. doi: 10.1146/annurev.ento.49. nchembio.331
061802.123416 Lacey, L. A. (ed.). (2012). Manual of Techniques in Invertebrate Pathology, 2nd Edn.
Dormann, C., Gruber, B., and Fründ, J. (2008). Introducing the bipartite package: Oxford: Academic Press.
analysing ecological networks. R News 8, 8–11. Lane, D. (1991). “16S/23S rRNA sequencing,” in Nucleic Acid Techniques in
Evans, H. C., and Samson, R. A. (1982). Cordyceps species and their anamorphs Bacterial Systematics, eds E. Stackebrandt and M. Goodfellow (Chichester: John
pathogenic on ants (Formicidae) in tropical forest ecosystems I. The Cephalotes Wiley & Sons), 115–147.
(Myrmicinae) complex. Trans. Br. Mycol. Soc. 79, 431–453. doi: 10.1016/S0007- Madden, A. A., Grassetti, A., Soriano, J. N., and Starks, P. T. (2013). Actinomycetes
1536(82)80037-5 with antimicrobial activity isolated from paper wasp (Hymenoptera: Vespidae:
Evans, H. C., and Samson, R. A. (1984). Cordyceps species and their anamorphs Polistinae) Nests. Environ. Entomol. 42, 703–710. doi: 10.1603/EN12159
pathogenic on ants (Formicidae) in tropical forest ecosystems II. The Marsh, S. E., Poulsen, M., Gorosito, N. B., Pinto-Tomás, A., Masiulionis, V. E., and
Camponotus (Formicinae) complex. Trans. Br. Mycol. Soc. 82, 127–150. Currie, C. R. (2013). Association between Pseudonocardia symbionts and Atta
doi: 10.1016/S0007-1536(84)80219-3 leaf-cutting ants suggested by improved isolation methods. Int. Microbiol. 16,
Evans, J. D., and Lopez, D. L. (2004). Bacterial probiotics induce an immune 17–25. doi: 10.2436/20.1501.01.176
response in the honey bee (Hymenoptera: Apidae). J. Econ. Entomol. 97, Mayfield, C. I., Williams, S. T., Ruddick, S. M., and Hatfield, H. L. (1972). Studies on
752–756. doi: 10.1093/jee/97.3.752 the ecology of actinomycetes in soil IV. Observations on the form and growth of
Han, J. H., Cho, M. H., and Kim, S. B. (2012). Ribosomal and protein coding streptomycetes in soil. Soil Biol. Biochem. 4, 79–91. doi: 10.1016/0038-0717(72)
gene based multigene phylogeny on the family Streptomycetaceae. Syst. Appl. 90045-4
Microbiol. 35, 1–6. doi: 10.1016/j.syapm.2011.08.007 Miller, M. A., Pfeiffer, W., and Schwartz, T. (2010). “Creating the CIPRES Science
Hanshew, A. S., McDonald, B. R., Díaz Díaz, C., Djiéto-Lordon, C., Blatrix, R., Gateway for inference of large phylogenetic trees,” in Proceedings of the 2010
and Currie, C. R. (2015). Characterization of Actinobacteria associated with Gateway Computing Environments Workshop (GCE) (IEEE), New Orleans, LA,
three ant–plant mutualisms. Microb. Ecol. 69, 192–203. doi: 10.1007/s00248- 1–8. doi: 10.1109/GCE.2010.5676129
014-0469-3 Oh, D.-C., Poulsen, M., Currie, C. R., and Clardy, J. (2009). Dentigerumycin: a
Hanson, P. E., and Gauld, I. D. (1995). The Hymenoptera of Costa Rica. Oxford: bacterial mediator of an ant-fungus symbiosis. Nat. Chem. Biol. 5, 391–393.
Oxford University Press. doi: 10.1002/mmnd.19970440203 doi: 10.1038/nchembio.159
Hanson, P. E., and Nishida, K. (2016). Insects and Other Arthropods of Tropical Oi, D. H., and Pereira, R. M. (1993). Ant behavior and microbial pathogens
America. Ithaca, NY: Cornell University Press. (Hymenoptera: Formicidae). Florida Entomol. 76, 63–74. doi: 10.2307/3496014
Hatano, K., Nishii, T., and Kasai, H. (2003). Taxonomic re-evaluation of Penn, O., Privman, E., Ashkenazy, H., Landan, G., Graur, D., and Pupko, T. (2010).
whorl-forming Streptomyces (formerly Streptoverticillium) species by using GUIDANCE: a web server for assessing alignment confidence scores. Nucleic
phenotypes, DNA-DNA hybridization and sequences of gyrB, and proposal of Acids Res. 38, W23–W28. doi: 10.1093/nar/gkq443
Streptomyces luteireticuli (ex Katoh and Arai 1957) corrig., sp. nov., nom. rev. Pontoppidan, M.-B., Himaman, W., Hywel-Jones, N. L., Boomsma, J. J., and
Int. J. Syst. Evol. Microbiol. 53, 1519–1529. doi: 10.1099/ijs.0.02238-0 Hughes, D. P. (2009). Graveyards on the move: the spatio-temporal distribution
Hölldobler, B., and Wilson, E. O. (2009). The Superorganism: The Beauty, Elegance, of dead Ophiocordyceps-infected ants. PLOS ONE 4:e4835. doi: 10.1371/
and Strangeness of Insect Societies. New York City, NY: W.W. Norton & journal.pone.0004835
Company. Poulsen, M., Oh, D.-C., Clardy, J., and Currie, C. R. (2011). Chemical analyses
Hong, S.-H., Bunge, J., Jeon, S.-O., and Epstein, S. S. (2006). Predicting microbial of wasp-associated Streptomyces bacteria reveal a prolific potential for
species richness. Proc. Natl. Acad. Sci. U.S.A. 103, 117–122. doi: 10.1073/pnas. natural products discovery. PLOS ONE 6:e16763. doi: 10.1371/journal.pone.
0507245102 0016763
Hulcr, J., Adams, A. S., Raffa, K., Hofstetter, R. W., Klepzig, K. D., and Currie, Promnuan, Y., Kudo, T., and Chantawannakul, P. (2009). Actinomycetes isolated
C. R. (2011). Presence and diversity of Streptomyces in Dendroctonus and from beehives in Thailand. World J. Microbiol. Biotechnol. 25, 1685–1689.
sympatric bark beetle galleries across North America. Microb. Ecol. 61, 759–768. doi: 10.1007/s11274-009-0051-1
doi: 10.1007/s00248-010-9797-0 Rehner, S. A., and Buckley, E. (2005). A Beauveria phylogeny inferred from
Kaltenpoth, M. (2009). Actinobacteria as mutualists: general healthcare for insects? nuclear ITS and EF1-α sequences: evidence for cryptic diversification and
Trends Microbiol. 17, 529–535. doi: 10.1016/j.tim.2009.09.006 links to Cordyceps teleomorphs. Mycologia 97, 84–98. doi: 10.3852/mycologia.
Kaltenpoth, M., and Engl, T. (2013). Defensive microbial symbionts in 97.1.84
Hymenoptera. Funct. Ecol. 28, 315–327. doi: 10.1111/1365-2435.12089 Reyes, R. D. H., and Cafaro, M. J. (2014). Paratrechina longicornis ants in a tropical
Kaltenpoth, M., Göttler, W., Herzner, G., and Strohm, E. (2005). Symbiotic dry forest harbor specific Actinobacteria diversity. J. Basic Microbiol. 55, 11–21.
bacteria protect wasp larvae from fungal infestation. Curr. Biol. 15, 475–479. doi: 10.1002/jobm.201300785
doi: 10.1016/j.cub.2004.12.084 Ronquist, F., and Huelsenbeck, J. P. (2003). MrBayes 3: Bayesian phylogenetic
Kaltenpoth, M., Roeser-Mueller, K., Koehler, S., Peterson, A., Nechitaylo, T. Y., inference under mixed models. Bioinformatics 19, 1572–1574. doi: 10.1093/
Stubblefield, J. W., et al. (2014). Partner choice and fidelity stabilize coevolution bioinformatics/btg180
in a Cretaceous-age defensive symbiosis. Proc. Natl. Acad. Sci. U.S.A. 111, Samson, R. A., Evans, H. C., and Latge, J.-P. (2013). Atlas of Entomopathogenic
6359–6364. doi: 10.1073/pnas.1400457111 Fungi. Berlin: Springer Science & Business Media.
Kaltenpoth, M., Schmitt, T., Polidori, C., Koedam, D., and Strohm, E. (2010). Sanjuan, T. I., Franco-Molano, A. E., Kepler, R. M., Spatafora, J. W., Tabima, J.,
Symbiotic streptomycetes in antennal glands of the South American digger Vasco-Palacios, A. M., et al. (2015). Five new species of entomopathogenic fungi
wasp genus Trachypus (Hymenoptera, Crabronidae). Physiol. Entomol. 35, from the Amazon and evolution of neotropical Ophiocordyceps. Fungal Biol.
196–200. doi: 10.1111/j.1365-3032.2010.00729.x 119, 901–916. doi: 10.1016/j.funbio.2015.06.010
Kaltenpoth, M., Yildirim, E., Herzner, G., and Strohm, E. (2012). Refining the Schloss, P. D., and Handelsman, J. (2004). Status of the microbial census. Microbiol.
roots of the beewolf-Streptomyces symbiosis: antennal symbionts in the rare Mol. Biol. 68, 689–691. doi: 10.1128/MMBR.68.4.686-691.2004
genus Philanthinus. Appl. Environ. Microbiol. 78, 822–827. doi: 10.1128/AEM. Schloss, P. D., Westcott, S. L., Ryabin, T., Hall, J. R., Hartmann, M., Hollister,
06809-11 E. B., et al. (2009). Introducing mothur: open-source, platform-independent,
Karnovsky, M. J. (1965). A formaldehyde-glutaraldehyde fixative of high osmolality community-supported software for describing and comparing microbial
for use in electron-microscopy. J. Cell Biol. 27, 137–138. communities. Appl. Environ. Microbiol. 75, 7537–7541. doi: 10.1128/AEM.
Kim, K. W. (2008). Vapor fixation of intractable fungal cells for simple and versatile 01541-09
scanning electron microscopy. J. Phytopathol. 156, 125–128. doi: 10.1111/j. Simmons, D. R., Kepler, R. M., Rehner, S. A., and Groden, E. (2015). Phylogeny of
1439-0434.2007.01331.x Hirsutella species (Ophiocordycipitaceae) from the UsA: remedying the paucity
Frontiers in Microbiology | www.frontiersin.org 16 October 2017 | Volume 8 | Article 2016
fmicb-08-02016 October 16, 2017 Time: 17:44 # 17
Matarrita-Carranza et al. Widespread Neotropical Insect-Actinobacteria Associations
of Hirsutella sequence data. IMA Fungus 6, 345–356. doi: 10.5598/imafungus. White, T., Bruns, T., Lee, S., and Taylor, J. (1990). “Amplification and direct
2015.06.02.06 sequencing of fungal ribosomal RNA genes for phylogenetics,” in PCR Protocols:
Sit, C. S., Ruzzini, A. C., Van Arnam, E. B., Ramadhar, T. R., Currie, C. R., A Guide to Methods and Applications, eds N. Innis, D. Gelfand, J. Sninsky, and
and Clardy, J. (2015). Variable genetic architectures produce virtually identical T. White (New York, NY: Academic Press), 315–322.
molecules in bacterial symbionts of fungus-growing ants. Proc. Natl. Acad. Sci. Wilson, E. O. (1971). The Insect Societies. Cambridge, MA: Belknap Press of
U.S.A. 112, 13150–13154. doi: 10.1073/pnas.1515348112 Harvard University Press.
Suay, I., Arenal, F., Asensio, F. J., Basilio, A., Cabello, M. A., Díez, M. T., et al. Wilson-Rich, N., Spivak, M., Fefferman, N. H., and Starks, P. T. (2009). Genetic,
(2000). Screening of basidiomycetes for antimicrobial activities. Antonie Van individual, and group facilitation of disease resistance in insect societies. Annu.
Leeuwenhoek 78, 129–140. doi: 10.1023/A:1026552024021 Rev. Entomol. 54, 405–423. doi: 10.1146/annurev.ento.53.103106.093301
Sung, G.-H., Hywel-Jones, N. L., Sung, J.-M., Luangsa-Ard, J. J., Shrestha, B., Zhang, Y. J., Zhang, S., Liu, X. Z., Wen, H. A., and Wang, M. (2010). A simple
and Spatafora, J. W. (2007). Phylogenetic classification of Cordyceps and the method of genomic DNA extraction suitable for analysis of bulk fungal strains.
clavicipitaceous fungi. Stud. Mycol. 57, 5–59. doi: 10.3114/sim.2007.57.01 Lett. Appl. Microbiol. 51, 114–118. doi: 10.1111/j.1472-765X.2010.02867.x
Tamura, K., Stecher, G., Peterson, D., Filipski, A., and Kumar, S. (2013). MEGA6:
molecular evolutionary genetics analysis version 6.0. Mol. Biol. Evol. 30, Conflict of Interest Statement: The authors declare that the research was
2725–2729. doi: 10.1093/molbev/mst197 conducted in the absence of any commercial or financial relationships that could
van der Meij, A., Worsley, S. F., Hutchings, M. I., and van Wezel, G. P. (2017). be construed as a potential conflict of interest.
Chemical ecology of antibiotic production by actinomycetes. FEMS Microbiol.
Rev. 41, 392–416. doi: 10.1093/femsre/fux005 Copyright © 2017 Matarrita-Carranza, Moreira-Soto, Murillo-Cruz, Mora, Currie
Větrovský, T., Steffen, K. T., Baldrian, P., Rouard, N., and Martin-Laurent, F. and Pinto-Tomas. This is an open-access article distributed under the terms of
(2014). Potential of cometabolic transformation of polysaccharides and lignin in the Creative Commons Attribution License (CC BY). The use, distribution or
lignocellulose by soil Actinobacteria. PLOS ONE 9:e89108. doi: 10.1371/journal. reproduction in other forums is permitted, provided the original author(s) or licensor
pone.0089108 are credited and that the original publication in this journal is cited, in accordance
Ward, P. S. (2014). The phylogeny and evolution of ants. Annu. Rev. Ecol. Evol. with accepted academic practice. No use, distribution or reproduction is permitted
Syst. 45, 23–43. doi: 10.1146/annurev-ecolsys-120213-091824 which does not comply with these terms.
Frontiers in Microbiology | www.frontiersin.org 17 October 2017 | Volume 8 | Article 2016