Bridging Place-Based Astrobiology Education with Genomics, Including Descriptions of Three Novel Bacterial Species Isolated from Mars Analog Sites of Cultural Relevance

Abstract

Democratizing genomic data science, including bioinformatics, can diversify the STEM workforce and may, in turn, bring new perspectives into the space sciences. In this respect, the development of education and research programs that bridge genome science with "place" and world-views specific to a given region are valuable for Indigenous students and educators. Through a multi-institutional collaboration, we developed an ongoing education program and model that includes Illumina and Oxford Nanopore sequencing, free bioinformatic platforms, and teacher training workshops to address our research and education goals through a place-based science education lens. High school students and researchers cultivated, sequenced, assembled, and annotated the genomes of 13 bacteria from Mars analog sites with cultural relevance, 10 of which were novel species. Students, teachers, and community members assisted with the discovery of new, potentially chemolithotrophic bacteria relevant to astrobiology. This joint education-research program also led to the discovery of species from Mars analog sites capable of producing N-acyl homoserine lactones, which are quorum-sensing molecules used in bacterial communication. Whole genome sequencing was completed in high school classrooms, and connected students to funded space research, increased research output, and provided culturally relevant, place-based science education, with participants naming three novel species described here. Students at St. Andrew's School (Honolulu, Hawai'i) proposed the name Bradyrhizobium prioritasuperba for the type strain, BL16A^T, of the new species (DSM 112479^T = NCTC 14602^T). The nonprofit organization Kauluakalana proposed the name Brenneria ulupoensis for the type strain, K61^T, of the new species (DSM 116657^T = LMG = 33184^T), and Hawai'i Baptist Academy students proposed the name Paraflavitalea speifideiaquila for the type strain, BL16E^T, of the new species (DSM 112478^T = NCTC 14603^T).

Keywords: Bradyrhizobium; Brenneria; Lava caves; MinION; Paraflavitalea.; Place-based education.

FIG. 1.

ʻĀina-Informatics Network (AIN) lab: genome sequencing in the classroom, Hawaiʻi. Top left: Students at St. Andrew's School sequencing BL16A^T, a novel Bradyrhizobium bacteria isolated from Kaūmana cave, on the Island of Hawaiʻi. Bottom left: Eric ʻIwakeliʻi Tong with students and teachers sequencing. Right: Students at ʻIolani school load a MinION flow cell.

FIG. 2.

Genes and pathways of interest from whole genome sequencing and annotation of the 13 bacterial genomes described here. Annotations were completed using PROKKA within the EDGE Bioinformatics pipeline, as well as tools FeGenie, HMM-S-S, and an in-house Hidden Markov Model (HMM) for the detection of putative luxI and luxR genes involved in quorum sensing, a type of bacterial chemical signaling process that uses the molecule N-acyl homoserine lactone (AHLs). The strain ID is listed at the top of each column, with the various genes of interest in each row. Each pathway of interest is represented by a different color, and larger colored circles represent a higher gene count in that species. For example, luxR-like gene homologs, which make the receptor proteins for various AHL signaling molecules, and up or down regulating genes controlled by a particular AHL signaling molecule, ranged in gene count from 10 luxRs in Y38-1Y to 43 in C9-3.

FIG. 3.

Images of three novel species described here, along with their transmission electron microscopy (TEM) images: BL16E^T, BL16A^T, and K61^T. Left, top and bottom panels: BL16E^T with TEM image (direct magnification 2500 × ). Center, top and bottom panels: BL16A^T with TEM image. Right, top and bottom panels: K61^T with TEM image (scale bar = 600 nm). TEM images: cells were negatively stained with uranyl acetate and viewed on a 120 kV Hitachi HT7700 transmission electron microscope in the Biological Electron Microscope Facility at the University of Hawai‘i at Mānoa.

FIG. 4.

Dot-plot from FeGenie bioinformatics tool, which helps identify iron-related pathways and genes in the genomes of the 13 bacteria discussed here. Bacterial strain ID is given on the x-axis and is grouped by the location they were cultivated from. Iron-related gene categories are on the y-axis and represented by different colors in the key on the right of the dot-plot. Larger circles represent a higher count of genes in a given category, with a size key on the right side of the plot.

FIG. 5.

Maximum-likelihood tree based on 16S rRNA gene sequences extracted from the genomes of BL16A^T and related type strains, inferred under the GTR+GAMMA model and rooted by midpoint-rooting. Branches are scaled in terms of the expected number of substitutions per site. Scale bar represents 0.01 nucleotide substitutions per site. Numbers above branches are support values for the maximum likelihood tree (left) and maximum parsimony tree (right). Species and strain are followed in parentheses by the GenBank accession number of the 16S rRNA sequence for that strain. The ML bootstrapping did not converge, so 1000 replicates were conducted: average support was 49.83%.

FIG. 6.

Maximum-likelihood tree based on 16S rRNA gene sequences extracted from the genomes of BL16E^T and related type strains, inferred under the GTR+GAMMA model and rooted by midpoint-rooting. Branches are scaled in terms of the expected number of substitutions per site. Scale bar represents 0.01 nucleotide substitutions per site. Numbers above branches are support values for the maximum-likelihood tree (left) and maximum parsimony tree (right). Species and strain are followed in parentheses by the GenBank accession number of the 16S rRNA sequence for that strain. The maximum likelihood bootstrapping converged after 900 replicates; average support was 56.70%.

FIG. 7.

Maximum-likelihood tree based on 16S rRNA gene sequences extracted from the genomes of K61^T and related type strains, inferred under the GTR+GAMMA model and rooted by midpoint-rooting. Branches are scaled in terms of the expected number of substitutions per site. Scale bar represents 0.01 nucleotide substitutions per site. Numbers above branches are support values for the maximum-likelihood tree (left) and maximum parsimony tree (right). Species and strain are followed in parentheses by the GenBank accession number of the 16S rRNA sequence for that strain. Maximum-likelihood bootstrapping did not converge, so 1000 replicates were conducted; average support was 42.18%.