Figs ( Ficus spp.) harbor diverse communities of animals, fungi, and microbes that interact within the enclosed environment of the fig syconium itself. Although the fig–fig wasp mutualism has served as a model for studying coevolution for decades (Cook and Rasplus 2003), comparatively little is known about the microbial members of these communities and their ecological roles. Recent work has demonstrated that the fig community harbors a diverse assemblage of bacterial associates (Woodruff et al. 2026), yet our understanding of bacterial functional ecologies within the fig environment remains muddied. Interestingly, other recent research suggests that bacterial associates may mitigate fungal pathogens within figs (Van Goor et al. 2026), highlighting the need for cultured isolates and genomic resources to investigate these interactions. This raises an important question: which bacteria are implicated in fungal mitigation, and why?

To support future studies of fig-associated microbial communities, bacterial isolates were recovered from recently pollinated syconia of two fig species native to southern Florida, USA ( Ficus aurea and Ficus citrifolia ). The isolates served as the basis for a course-based undergraduate research experience (CURE) in the University of Missouri Microbiology Laboratory course (BIO_SC 3760). Course-based undergraduate research experiences are high-impact educational practices that improve student learning and retention while simultaneously generating valuable biological datasets (Martinez-Vaz and Bell 2025; Ksiazek Mikenas 2026).

The interior of fig syconia were aseptically sampled and plated on lysogeny broth (LB) agar. Individual colonies were purified by repeated streaking, and each strain was maintained as a frozen stock with 25% glycerol at −80 °C. Each isolate is available by request from the Brown lab. During the semester, each student characterized one of the 22 bacterial isolates through phenotypic analyses, including Gram staining, swimming motility, β-lactamase screening, and antagonistic activity against Aspergillus niger and Saccharomyces cerevisiae .

For whole-genome sequencing, isolates were cultured on LB agar for 2 days at 30 °C before a single colony was inoculated into 6 mL LB medium and grown with shaking for 2 days at 30 °C. Cells were pelleted (13,000 × g ), washed with phosphate-buffered saline (PBS), resuspended in 0.5 mL DNA/RNA Shield (Zymo Research), and submitted to Plasmidsaurus (Eugene, OR, USA) for Oxford Nanopore sequencing (Purushothaman et al. 2026). Genome assembly and annotation were performed using the Plasmidsaurus bacterial assembly pipeline, which incorporates Filtlong (Wick 2026), Flye (Kolmogorov et al. 2019), Autocycler (Wick et al. 2025), Dnaapler (Bouras et al. 2024), and Plassembler (Bouras et al. 2023) for read filtering, genome assembly, consensus generation, genome reorientation, and plasmid assembly, respectively. Whole genome sequencing statistics for fig-associated bacterial isolates are shown in Table 1 and illustrate excellent coverage, completeness, and rNG50 values. Genome assemblies and raw sequence data have been deposited under BioProject PRJNA1419296 with associated BioSample accessions SAMN55952996–SAMN55953017 and Sequence Read Archive accessions SRX33714672–SRX33714693.

Genome assemblies were used for taxonomic classification, phylogenetic analysis, and functional annotation. As part of the Plasmidsaurus pipeline Mash (Ondov et al 2016) was used to identify the closest named relative available in RefSeq (O’Leary et al 2016) for the largest contig in each genome assembly (Table 1). The estimated percent identity (Estimated %ID, Table 1) is estimated from the number of k -mers that matched the reference genome for the closed named relative. To validate and refine the taxonomic classification of the isolates, a multiple sequence concatenated alignment of 16S rRNA, 23S rRNA, RNA polymerase beta subunit ( rpoB ), DNA gyrase subunit B ( gyrB ), and recombinationA ( recA ) gene sequences were generated using MUSCLE (Edgar 2004) implemented in Geneious Prime V2025.2.2 (http://www.geneious.com). Bayesian phylogenetic inference was performed with MrBayes (Huelsenbeck and Ronquist 2001) using the GTR+G+I substitution model and 5,000,000 generations. Trees were edited in FigTree (Rambaut 2018) and visualized with Thermotoga neapolitana as the outgroup using Interactive Tree of Life (iTOL; Letunic and Bork 2024). Functional annotation was performed using BlastKOALA (Kanehisa et al. 2016), and candidate antimicrobial and secondary metabolite biosynthetic gene clusters were identified using antiSMASH (Blin et al. 2025).

Whole-genome sequencing and taxonomic analyses identified 22 isolates representing six bacterial species spanning four genera: Klebsiella-Raoultella complex, Kosakonia , Neobacillus , and Pantoea ( Figure 1A ). Representative phenotypes are shown and summarized in Figure 1B and strain-specific phenotypic profiles are indicated in Table 1 (ND = not determined). The Klebsiella-Raoultella complex, Kosakonia , and Pantoea isolates were motile, Gram-negative rods, although substantial variation was observed in antagonistic activity against fungi ( Figure 1B, Table 1). In contrast, the sole Gram-positive bacterium in our dataset (presumed Neobacillus ) displayed no apparent motility, no detectable antifungal inhibition behaviors, and no apparent microbial competition abilities ( Figure 1B, Table 1).

As expected in all communities, this analysis of the microbial communities of fig species identifies associates with varying ecologies. When viewing through the scope of potential fungal mitigation, our phenotypic analyses (Table 1, Figure 1B ) strongly suggest that isolated members of the Klebsiella-Raoultella complex and Neobacillus function as commensals within the internal fig environment (Mathis and Bronstein 2020). However, members of the Klebsiella–Raoultella complex and the genus Neobacillus have been described as plant growth promoters in other agricultural systems (Singh et al. 2015, Kumar et al. 2021, Tsotetsi et al. 2022), particularly under environmental/abiotic stressors. It is possible that they are commensals, or that they interact with figs in ways that our current experiments have not explored.

Intriguingly, Kosakonia and Pantoea isolates consistently showed antifungal activities and evidence of antimicrobial molecular defenses, both of which suggest a protective mutualism between these microbes and the figs they inhabit (Table 1, Figure 1B ). Indeed, isolates of Kosakonia consistently inhibited the growth of S. cerevisiae and most isolates also inhibited the growth of both A. niger . Some Kosakonia isolates exhibited β-lactamase activity, and genes encoding predicted β-lactamases were identified in their genomes. Similarly, most of the Pantoea isolates inhibited the growth of S. cerevisiae and A. niger and their genomes contain predicted siderophore biosynthetic gene clusters (Table 1, Figure 1B ). These data suggest that Kosakonia and Pantoea isolates may have some beneficial properties in fungal mitigation or interbacterial competition within the fig microenvironment. While Kosakonia and Pantoea have both been previously implicated in fungal mitigation for other plants (Jiang et al. 2019, Thanwisai et al. 2024), this is the first experimental evidence of such activities in fig associates. These findings may inform future management of fungal diseases for the cultivated fig, such as Fig Endosepsis (Michailides and Morgan 1998, Van Goor et al. 2026). Further, while microbially produced molecules such as siderophores are known to benefit associates through nutrient acquisition and competitive abilities (Kramer et al. 2019), production of siderophores by Kosakonia has also been linked to fungal mitigation abilities (Lambrese et al. 2018), warranting future examination within the context of the fig community. Together, these genome assemblies and accompanying phenotypic analyses provide a valuable resource for future investigations into the ecological roles of bacteria within fig-associated microbial communities.

While the data presented here suggest a promising connection between the presence of some fig microbes and the mitigation/limitation of fungal growth, it also presents a curious conundrum: if anti-fungal bacteria are present within the fig environment, why do we see figs infested with fungus? Local infestation rate of Fusarium fungus within the Floridian fig species examined here sometimes exceeded 95% of the figs sampled; how could this be? Future research efforts will prioritize the examination of multiple microbial taxa working in concert, and the potential role of alternative eukaryotic fig associates (notably nematodes) in the mitigation of pathogenic fungus. Finally, it appears as if our previous culturing efforts largely selected for Gram-negative bacteria, but more recent culture-independent metabarcode efforts suggest the presence of a large diversity of Gram-positive bacteria that were present within the fig environment as well. Future iterations of the BIO_SC 3760 CURE course seek to utilize alternative media in an effort to culture a wider diversity of microbial associates to further understand their functional ecological roles within the fig environment.

Bacterial strains and growth conditions

All bacterial strains used in this study were grown in LB Miller broth (10 g tryptone, 5 g yeast extract, and 10 g NaCl L⁻¹) or agar at 30°C with shaking unless otherwise indicated.

Phase contrast microscopy

To prepare for microscopy, strains were grown to exponential phase in LB medium and 0.5 mL of cells were spotted onto 1.75% LB low-melt agarose pads. Cells were imaged using an inverted Nikon Eclipse TiE equipped with a 63X 1.4 NA Plan Apochromat oil-immersion phase-contrast objective using a Rolera em-c ² 1K EMCCD and Nikon Elements Imaging Software.

Gram staining

A thin smear was prepared by suspending a fresh bacterial colony in 5 µL sterile water on a glass microscope slide. Smears were air dried for 10–15 min, heat fixed, and Gram stained using a Remel™ Gram Stain Kit (Thermo Fisher Scientific) according to the manufacturer's instructions. Gram-stained cells were observed and imaged using a Leica ICC50 W microscope equipped with a 40× brightfield objective.

Wet mount microscopy

Bacterial isolates were grown overnight in 1 mL LB medium with shaking at 30 °C. A 5 µL aliquot of culture was placed on a glass microscope slide, covered with a coverslip, and examined using a Leica ICC50 W microscope equipped with a 100× oil immersion phase-contrast objective. Phase-contrast videos were recorded to document the presence or absence of swimming motility.

β-Lactamase screening

Bacterial isolates were screened for β-lactamase production using a disk diffusion assay on Mueller-Hinton agar. Bacterial lawns were prepared from overnight cultures, and commercial ampicillin and ampicillin-sulbactam disks were placed onto the agar surface. Plates were incubated at 30 °C for 2 days, after which zones of inhibition were measured in millimeters. Increased susceptibility to ampicillin in the presence of sulbactam was interpreted as evidence of β-lactamase production.

Antagonistic activity against fungi

Yeast growth inhibition assay (Saccharomyces cerevisiae). Commercial active dry baker's yeast ( Saccharomyces cerevisiae ) was prepared by hydrating 1 g of yeast in 10 mL sterile water for 10 min and vortexing to obtain a homogeneous suspension. The suspension was serially diluted from 10 ^-2 to 10 ^-6 in sterile water. Bacterial isolates were spread onto yeast peptone dextrose (YPD) agar using sterile cotton swabs to produce confluent lawns and allowed to dry for approximately 15 min. Five-microliter aliquots of each yeast dilution were spotted in technical duplicate onto bacterial lawns and uninoculated control plates. After the inoculum had dried, plates were sealed with parafilm and incubated at 30 °C. After 2 days of incubation, yeast growth on bacterial lawns was compared with growth on control plates lacking bacteria. Growth inhibition was determined by comparing the maximum dilution supporting visible yeast growth on bacterial lawns with the maximum dilution supporting visible growth on control plates and was reported as the corresponding log-fold reduction in yeast growth.

Mycelial growth inhibition assay (Aspergillus niger). Aspergillus niger was obtained as a live culture from Carolina Biological Supply (Burlington, NC, USA) and maintained on potato dextrose agar (PDA). Bacterial isolates were spread onto PDA using sterile cotton swabs to produce confluent lawns and allowed to dry for approximately 15 min. A 1-cm² plug of actively growing A. niger mycelium was excised from the stock culture and transferred to the center of each bacterial lawn. Control plates consisted of fungal plugs transferred to PDA without bacterial lawns. Plates were sealed with parafilm, incubated in an inverted orientation at 30 °C, and monitored for fungal growth. Mycelial diameter was measured on days 2, 5, and 7. Fungal mat area was calculated from colony diameter, assuming circular colony morphology, and used to quantify fungal growth.