Catalyzing rapid discovery of gold-precipitating bacterial lineages with university students

Introduction

The potential benefits from the study of the unique abilities of bacteria to everyday human life is ever more obvious. Bacteria are used industrially in food preparation, drug production, waste treatment and many other roles. Advances in biotechnology techniques have facilitated the use of known bacterial species and their enzymes, proteins and pathways (Berini et al., 2017). For example, it is now possible, and indeed not very difficult, to identify genes of interest in a bacterial species, clip those genes out of that species and insert them into another work horse species of bacteria to allow the products of those genes to be produced industrially. Ironically, as our ability to harness the power of bacteria becomes ever more sophisticated, one of the key challenges is still finding the useful bacteria in the first place. In a world with as many as a trillion bacterial species (Locey & Lennon, 2016; Pike, Viciani & Kumar, 2018), how does one speed the discovery of bacterial species with a particular use or even simply strains of a particular bacterial taxon with sequences of interest?

One approach is to engage citizen scientists. In as much as the first step in the discovery of novel, useful microbes is often collection from nature, collections made by the public have the potential to speed up this key and often rate-limiting first step. What is more, in a rapidly interconnected digital era, the potential for truly global projects that rely on hundreds, thousands, or even hundreds of thousands of individuals is ever greater (Cooper, 2016). Citizen scientists contribute data to many publicly-accessible projects, from birdwatchers helping conservation efforts with the e-Bird project (https://ebird.org/home; Sullivan et al., 2014), game enthusiasts folding proteins for the FoldIt project (https://fold.it/portal/; Cooper et al., 2010), to homeowners exploring the microbial diversity in their houses (http://robdunnlab.com/projects/wild-life-of-our-homes/; Dunn et al., 2013). Additionally, projects like the Science Education Alliance—Phage Hunters Advancing Genomics and Evolutionary Science (SEA—PHAGES) and Tiny Earth engage students in large research projects as part of course-based undergraduate research experiences (CUREs) (Hanauer et al., 2017; Handelsman et al., 2018). Citizen scientists, we argue, can also help discover bacteria with novel, useful traits.

Delftia is a genus first discovered in the city Delft (Den Dooren de Jong, 1927; Wen et al., 1999), where bacteria themselves were discovered by Leeuwenhoek (Gest, 2004). Delftia has genes capable of precipitating gold by excreting a metabolite called delftibactin (Johnston et al., 2013). Gold in solution as gold chloride is toxic to bacteria, so Delftia has evolved this novel mechanism for precipitating aqueous gold out of solution to nontoxic solid gold nanoparticles. This mechanism has obvious potential uses in gold recycling in used electronics, gold mining, and urban waste (Gold recycling, 2013; Reith et al., 2007; Subhabrata, Natarajan & Ting, 2017), but to date, the existing genetic diversity of Delftia in strain collections is modest. There are only six known species of Delftia. Full genome assemblies exist for four of these species within the National Center for Biotechnology Information (NCBI) database (Wen et al., 1999). Discovery of novel Delftia species and their relatives has the potential to better elucidate variations in Delftia genetic sequences, especially within the gold precipitation gene cluster and other industrially and human health related sequences. The more information about these gold precipitation genes, for example, the greater potential for using Delftia or its genetic potential to recycle our electronics and make mining more sustainable.

Here we leverage a citizen science approach to detect new Delftia species on a university campus. We simultaneously test whether students are able to aid the speed of discovery of novel lineages and consider the biology of the lineages we have discovered. The Wolfpack Citizen Science Challenge for spring 2018 (go.ncsu.edu/wpc18) was a collaborative project to document the presence and genetic diversity of Delftia spp. across the North Carolina State University campus and create a scalable and interdisciplinary model to continue learning about this and other organisms. In addition to involving students in two introductory courses in the initial data collection, we also involved students in two upper-level courses in the downstream study of the microbes detected during the Challenge.

Materials and Methods

Results

Discussion

Here, we sought to simultaneously test whether we could engage students campus-wide in a citizen science style microbial research project and in doing so, understand the distribution and diversity of strains of one particular bacterial genus, Delftia. In short, we were indeed able to engage students from diverse majors across campus. In doing so, we discovered that some sampling sites had many more Delftia counts than did others, that Delftia was relatively ubiquitous and that some of the strains we identified had gold genes that appeared relatively divergent from those known from the literature. Although we were unable to accurately determine the diversity of Delftia strains present, this unanswered question presents a new challenge and opportunity for our citizen science and Delftia research efforts.

Collectively, the qPCR, Sanger DNA sequencing and BLAST comparison results showed that strains of Delftia are diverse, abundant and frequent (found at many sites) in environments in and around the college campus. Based on available genomic sequences deposited in the NCBI database and partial sequencing of the highly conserved gold gene, the strains students discovered best matched the reference strains D. tsuruhatensis CM13 and D. acidovorans ANG1 and SPH-1. However, 14 of 17 samples contained strains that were a 97% or lower match to strains in the NCBI database. Our suspicion is that these strains represent uncharacterized genetic diversity among strains in Delftia’s gold gene. However, because we sequenced from complex environmental samples we can’t preclude the possibility that some of this variation is due to cases in which the forward and reverse sequences obtained were from different Delftia species or strains in the sample.

The sequenced Delftia gold gene from many of the participant samples matched well to known Delftia species, but some samples matched two different existing strains equally well. For example, samples from 7-1 to 24-1 were equally similar to the strains D. acidovorans ANG1 and SPH-1 (Table 2). Clearly further work can be done to sequence additional portions or the entire genomes of these samples to identify what known strain is present or discover a new lineage of Delftia. More extensive community analyses of the samples using both targeted (16S rRNA gene) and whole genome shotgun sequencing would aid in the identification of which microbes associate with the presence of Delftia and the identity of the gold sequences in the environment, respectively. Additionally, high-throughput sequencing approaches such as Hi-C from Phase Genomics (“Hi-C Proximity-Guided Assembly,” 2018) (Sieber et al., 2018) or Nanopore single-molecule long-read sequencing can be employed to attempt to sequence and assemble the entire Delftia genome in metagenomic samples positive for Delftia by qPCR. Ultimately, selective media capable of isolating and identifying Delftia would allow us to increase our collection of Delftia strains for basic functional studies and genome sequencing.

Our sequencing results best matched the species D. acidovorans and D. tsuruhatensis, both of which have been found in environments similar to those we studied. D. acidovorans was originally discovered in soil and has been found in drains, waterspouts and showerheads in the built environment (Wen et al., 1999). D. tsuruhatensis was first discovered in a wastewater treatment plant and has been found in similar locations along with D. acidovorans (Hou et al., 2015). The Delftia species we did not encounter in our study are species that have so far been associated with more restricted habitats. Delftia deserti has been found to inhabit desert environments (Li et al., 2015), D. lacustris in lake water (Jørgensen et al., 2009), D. litopenaei in pond water (Chen et al., 2012), and D. rhizosphaerae in the rhizosphere of Cistus ladanifer, a plant native to the Mediterranean region (Carro et al., 2017). The apparent ubiquity of the genus Delftia hides the reality that individual species appear to show considerable habitat restriction. In the future, it would be interesting to understand which traits and genes of individual Delftia species confer the ability to survive in particular habitats.

It is unclear the extent to which the life history of Delftia in the above habitats is the same as that of Delftia in the built environment of a college campus. Nor is it well understood whether the presence of Delftia in water systems is problematic or potentially beneficial. Like many bacterial taxa, Delftia species are recorded as opportunistic pathogens that can infect hospitalized or immunocompromised patients (Patel et al., 2019) (Bilgin et al., 2015). However, there is no indication that human bodies are a common habitat for this genus. Instead, in buildings such as those we sampled it appears to be much more common in water systems—in drains, showerheads and downspouts. In as much as the ecological conditions of water systems differ greatly, it is possible that a comparative study of water systems, such as those that are or are not chlorinated, might reveal more about the built environment natural history of this organism.

Our approach kindled campus-wide student interest in microbial diversity and molecular biology techniques through the excitement of discovering this unique microbe in places that students frequent on campus. Groups of students from various academic disciplines and courses produced and analyzed samples that contributed to a large public dataset. The findings helped teach the student community about Delftia and also reinforced the importance of the collaborative nature of scientific discovery. The success of this project, in terms of the documentation of Delftia’s distribution helps to validate our general approach. In addition, our approach has the potential to encourage future students to participate. We aim to continue the challenge of accurately identifying new Delftia lineages and engage others by expanding the sampling opportunity to a multi-section first-year English class that is required for all undergraduate students on our campus. Using a similar approach and incorporating the expertise of faculty in the English department, we will engage students in writing tasks related to the project. Additionally, an upper-level metagenomics course will tie into this endeavor by processing, sequencing and analyzing the microbial communities in samples with high numbers of Delftia sequences. With relatively minor changes to the course schedules and curricula, 100 more students per semester can participate, learn and contribute to the project. We are creating resources that are accessible for other faculty and campuses to implement this project and share findings. For this, students participating in the project are writing the Delftia book (go.ncsu.edu/delftiabook), and we have created a group for instructor resources on the QUBES web portal (https://qubeshub.org/community/groups/delftia/projects). Liquid handlers can be cost-prohibitive, but less expensive models such as the Opentrons OT-2 are available, and we are developing scripts for this instrument. Student groups in lab-based courses can always set up qPCRs manually to participate in this project.

Conclusions

As the future plans for integrating this project into courses indicate, enthusiasm for the project was high among our colleagues and grew as the project proceeded. However, if we are to continue the project it is key that it continues to yield new scientific insights. Fortunately, this seems very likely to be the case. For example, although Delftia abundance was very patchy on campus, we have yet to explain what factors account for such patchiness. Additional samples will help us to have sufficient coverage across sample types to allow spatial models of Delftia diversity and abundance. In addition, our results suggest that new variants of the Delftia gold gene and even new Delftia strains remain to be discovered. Conversely, there is a lack of genomic diversity represented in the NCBI database. By leveraging the enthusiasm of university students and staff, interconnecting courses and researchers, and using our model pipeline, new lineages of Delftia can be rapidly identified and studied (e.g., groups of students cloning novel gold gene cluster into a host such as Escherichia coli or yeast for functional characterization). This will yield a better understanding of the ecological and environmental significance of these organisms and simultaneously help to connect students and faculty across campus in a common scientific project. Finally, it is of note that Delftia species, while little known, are of potentially great applied importance. In addition, they contain genes that allow many strains to precipitate gold. Given the many waste streams in which gold is present but hard to concentrate, this ability has the potential to be very useful moving forward.

Supplemental Information