This study is the first to report the assembly, genome and functional annotation, as well as biosynthetic gene cluster prediction of Nodosilinea sp. isolated from a copper mine tailing storage facility in Tuba, Benguet, Philippines. The data and information generated would enable further understanding of this species’ phylogeny and biology.

The cyanobacterium Nodosilinea sp. used in this study was from the existing cyanobacterial cultures of the Plant Molecular Biology and Genetics Laboratory (PMBGL) of the Institute of Biology, University of the Philippines Diliman, Quezon City, Philippines. Sediment samples were previously collected from an active Tailing Storage Facility 3 (TSF3) of a large-scale mining site in Tuba, Benguet, Philippines. Upon collection, a portion of the sediment samples was subjected to serial dilution. The sediment solutions were filtered, and the filter used was inoculated in BG-11 (HiMedia, India) liquid medium to allow enrichment of cyanobacteria. The cultures were then placed under white fluorescent lamps for a 12-hour light-dark cycle. This served as the mother culture for cyanobacterial isolation. To obtain unialgal cultures, classical microbiological techniques were employed using both solid and liquid BG-11 media. Isolated cyanobacteria were transferred into sterile 250mL Erlenmeyer flasks with BG-11 medium for cultivation and maintenance.

Morphological examination of the isolated cyanobacteria was performed using a light compound microscope (Labomed LB-221, USA). Photomicrographs were taken with the aid of ScopeImage 9.0. One particular isolate was morphologically characterized by examining the filament and trichome width, cell color and sheath, apical cells, and the presence of nodules from 10 randomly selected replicate specimens. The examined features were compared to the descriptions from the literature (Anagnostidis and Komarek, 2005; Heidari et al., 2018; Vázquez-Martínez et al., 2018).

After two months of growth, planktonic colonies of the cultured cyanobacterium were collected for the extraction of the total genomic DNA using the Quick-DNA Fungal/Bacterial Miniprep Kit (Zymo Research, USA). Agarose gel electrophoresis was performed to detect the presence of the extracted DNA, while a spectrophotometer (Epoch™) was used to determine its purity. Furthermore, the 16S rRNA gene region of the extracted genomic DNA was amplified in a T100 Thermal Cycler (Bio-Rad, Singapore) using forward and reverse primers specific to cyanobacteria: CYA106F (5′ CGG ACG GGT GAG TAA CGC GTG A 3′) and CYA781R (mixed in equimolar quantities: (a) (5′- GAC TAC TGG GGT ATC TAA TCC CAT T 3′) and (b) (5′- GAC TAC TGG GGT ATC TAA TCC CAT T 3′) (Nübel et al., 1997). The PCR cycle conditions were set with an initial denaturation step at 94°C for 5 minutes, followed by 35 cycles of denaturation at 94°C for 1 minute, an annealing step at 60°C for 1 minute, an extension step at 72°C for 1 minute, and a final elongation step at 72°C for 5 minutes. The amplicons were then visualized in 1% agarose gel at 100V for 30 minutes using 1kb universal DNA ladder (HyperLadder™). The gel was stained with gel red, and the image was recorded from a Bio-Rad Gel Doc EZ Gel Documentation System (Bio-Rad, USA). In addition, the quality of the extracted DNA was assessed using a microplate spectrophotometer (Epoch™). The PCR products were sent to Macrogen, Inc. (South Korea) for bidirectional capillary sequencing. The resulting sequences were assembled using Sequencher v5.4.6 (Gene Codes Corporation, 2017) and a nucleotide BLAST search was conducted on the NCBI website using the generated consensus sequences to identify the cyanobacterium. Upon confirmation, the genomic DNA of the cyanobacterial isolate was also sent to Macrogen, Inc. (South Korea) for whole genome shotgun sequencing (2x150 PE) on the Illumina NovaSeq platform.

Paired-end reads were filtered and trimmed with BBtools v.38.92 (Bushnell, 2014) and PRINSEQ (Preprocessing and Information and SEQuence 0.20.4) (Schmieder and Edwards, 2011). De novo assembly was performed using SPAdes 3.15.4 with k-mer sizes 21, 33, 55, 77, 99, and 127 (Bankevich et al., 2012). CheckM v1.1.6 (Parks et al., 2015) was used to assess the quality of assembled contigs. Since the culture was not axenic but unialgal, taxonomic binning was performed to remove contaminants (Teikari et al., 2022) using MetaBAT2 v1.7 (Kang et al., 2019), MaxBIN2 v2.2.4 (Wu et al., 2015), and CONCOCT v1.1 (Alneberg et al., 2014). The resulting bins were compared using DAS Tool v1.1.2 (Sieber et al., 2018). The highest quality bins were selected and subjected to GTDB-Tk v2.2.1 (Chaumeil et al., 2022) for taxonomic classification. The genome assembly was submitted to the Prokaryotic Genome Annotation Pipeline (PGAP) (Tatusova et al., 2016) at the National Center for Biotechnology Information (NCBI) for rapid annotation. Then, the cyanobacterial genome was functionally annotated using RASTtk server v2.0 (Brettin et al., 2015). Meanwhile, AntiSMASH 6.0 (Blin et al., 2019), DeepBGC 0.1.23 (Hannigan et al., 2019) and NaPDoS2 (Klau et al., 2022) were used to predict gene clusters related to secondary metabolite production. Default parameters were used unless specified.

The phylogenetic tree construction was performed using the 16S rRNA partial sequence of Nodosilinea sp. PGN35, including the representative sequences from the genera of filamentous cyanobacteria: Leptolyngbya (16), Nodosilinea (13), Arthronema (3), Lyngbya (5), and Phormidesmis (5). Gloeobacter violaceus was used as the outgroup. Gene alignment was performed using ClustalW. The nucleotide substitution model utilized was Kimura two parameter with gamma (K2+G) distribution, as determined by MEGA11 (Tamura et al., 2021). The phylogenetic tree was also generated using MEGA11; running bootstrap method with 1,000 replications. For further analysis, a phylogenomic tree of Nodosilinea sp. PGN35 was constructed using the Species Tree App present in KBase (Arkin et al., 2018). All parameters were set at default.

The cyanobacterial isolate was unialgal and non-axenized. It has uniseriate, unbranched filaments, which during nodule formation becomes multiseriate. It also has a blue-green trichome enclosed in a thin, colorless sheath. Each cell is longer than wide. The cell length measured 4.01 μm and the cell width was 3.45 μm. Their apical cells are rounded and the filaments form nodules ( Figures 1A–C ).

The whole genome sequencing of Nodosilinea sp. PGN35 generated a total of 22,265,312 reads. After quality control and assembly, 25 contigs were generated. The cyanobacterial genome has an estimated size of 6,390,537 bp and a GC content of 59.3%. Furthermore, the assembled genome showed a 98.37% completeness and 1.54% contamination.

The annotated genome using the PGAP revealed a total of 5,688 genes, 5617 protein coding genes (CDS), 45 rRNAs, 41 tRNAs, 4 ncRNAs, 26 pseudo genes, and 1 CRISPR arrays ( Table 1A ).

A circular graphical display from the RASTtk annotation is shown in Figure 1F . There were 17% proteins annotated by the RASTtk server. Annotations were grouped according to their likelihood to be involved in a cellular process or subsystem. Following the pie chart clockwise, subsystem categories are listed in the legend from top to bottom. Among all the cellular subsystem components, (1^st) the Amino Acids and Derivatives, (2^nd) Protein Metabolism, (3^rd) Cofactors, Vitamins, Prosthetics Groups, Pigments and (4^th) the Carbohydrates were the biggest groups, with 227, 165, 158, and 142 annotated protein-coding genes, respectively.

Moreover, three automated genome mining pipelines were used to annotate the major biosynthetic gene clusters (BGCs) of Nodosilinea sp. PGN35 ( Table 1B ). Analysis using AntiSMASH resulted in the identification of four BGCs which correspond to one cluster of arylpolyene, one cluster of resorcinol and two clusters of terpenes. It can be noted that these BGCs have not been characterized, and are currently referred to as orphan/cryptic, thus, warrants further exploration to determine whether these gene clusters are novel for this cyanobacterial strain. Aryl polyene is expected to be a prominent secondary metabolite for the prokaryotic Nodosilinea sp. since it is a widespread metabolite found in gram-negative bacteria (Lee et al., 2021) that functions as antioxidative carotenoids (Schöner et al., 2016). Meanwhile, resorcinol and its derivative is a class pf polyketide with antioxidant and antibacterial properties (Hiasa et al., 2013), hence commonly used as antiseptic and disinfectant. On the other hand, terpene synthases and its precursor pathways were expressed in cyanobacteria efficiently converting light and carbon dioxide to terpene HC and terpenoids (Lin and Pakrasi, 2019), which in turn, has vast applications in the pharmaceutical industry, as antiviral among others. Comparatively, DeepBGC was also able to predict two terpenes, the product class corresponding to saccharide and polyketide, as well as BGCs with predicted antibacterial and cytotoxic activity. The ribosomally synthesized and post-translationally modified peptides (RiPPs) are the dominant class of cyclic peptides reported to possess high antimicrobial activity (Carpine and Sieber, 2021). Meanwhile, NaPDoS2 was able to detect two ketosynthase domains identified as type II fatty acid synthase.

Phylogenetic tree construction of PGN35 using maximum likelihood of partial 16S rRNA gene sequences revealed that it belongs to the Nodosilinea genus. Likewise, it shows that it is related to the species of nodulosa and chupicuarensis with a similarity of 70% ( Figure 1D ). Meanwhile, phylogenomic analysis grouped PGN35 with one representative genome of the genus Leptolyngbya with 95.2% bootstrap support ( Figure 1E ). Although phylogenetic and phylogenomic trees of Nodosilinea sp. PGN35 showed high similarity to other Nodosilinea species, a close relationship with species of Leptolyngbya was also observed. This may suggest that additional genomic data is needed to further delineate the genus of Nodosilinea from Leptolyngbya.