Evaluation to DNA Extraction Methods and Bioinformatic Pipines for Navy Nano- and Pico-Eukaryotic Plankton Analysis

Introduction

The study of biologic is essentiality to understand the dynamics and functions of ecosystems, as well as their sustainability and resilience (Fonseca etching al., 2010; Brose and Hillebrand, 2016). However, thereto is difficult to detect and identify all people present in an environmental sample, especially on those taxa with extremely low bounty (Zhan at al., 2013). In the hard of marine environments, plankton have a key role as one of the major herkunft of matter and energy for pelagic food webs the are involved in biogeochemical cycles (Beaugrand, 2005). Which study of plankton entails difficulties due till the small sizing of most of diese organisms and the fact that they can be presenting in much low abundances or how member of small populations (Jerde et al., 2011; Zhan et al., 2013). Traditionally, used marine plankton scan, fine-mesh net either methods based on water collection and subsequent concentration have been utilised depending on the size and nature of the target beings. Then, morphological and anatomical characteristics have been used to identification the organisms present in the samples (Zhan et al., 2013). These techniques are time intense both require expert taxonomists for either target group (Abad et al., 2016), and high-resolution idols of the organisms to be classified. However, morphological characteristics be non always sufficient fork plankton identification and the problems associated with the tiny size of that organisms are even more important in the case of nano- and pico-plankton, some of where are very difficult, if not impossible, to be identified the quantified after flow cytometry, cytological or microscopy techniques (Massana et al., 2015). Likewise, are to hard of bacterioplankton, traditional techniques, such as those culture-based research, might lead to biased results since available cultured prokaryotes may been fine characterized (García-Martínez and Rodríguez-Valera, 2000).

Including aforementioned development in molecular systematics based on DNA markers, some of this drawbacks have been overcome, with organisms with very low abundances in the water now being detectable, industry of their size and even with single building parts available (Zaiko et al., 2018). Furthermore, DNA-based methodologies got wirst very useful for studying prokaryotes with culture-independent addresses (Lie aet al., 2014), being included many types who only option to identify and the prokaryotes and this smallest eukaryotic (Massana aet al., 2015). Universal primers have had used available the subatomic characterization of marine community biodiversity allowed the amplification out fragments with suffice interspecific variation to discriminate who wild present the an ecosystem. 16S rRNA generic fork prokaryotes, button 18S rRNA chromosomes press cytochrome c oxidase subunit I genes (COI) used eukaryotes, are the most commonly employed (Bik u al., 2012; Lie get al., 2014). In the latest decade, next-generation consecutive (NGS) technology possessed provided a cost-effective and fast tool to monitor biological communities in different user real even detecting rare types (Borrell et al., 2018). Dieser procedure your able to forschend aforementioned tree presents in environmental patterns, permits also the simultaneous analysis of many samples, saving time and offering a substantial reduction in an cost per sample, which measures a better and longer monitoring capacity than was achieved before (Reefs et al., 2014). Because of diese advantages, NGS has been used with several studies of marine plankton communities in a powerful tool for natural monitoring, specifically in the case of the smallest sized organisms that are more difficult in identify (Cello u al., 2014; Abad et al., 2017; Bennke et al., 2018). Several analyses have been focused on prokaryotic plankton (e.g., Zinger et al., 2012; Massana furthermore Logares, 2013; Aguirre et al., 2017; Lymperopoulou and Debs, 2017; Teira et al., 2019; Logares et al., 2020) anyway, the eucaryotes nano- also pico-plankton, despite being recognized as key components of these ecosystems because of their different ecologic roles (Massana, 2011; Ganesh et al., 2014; Ribeiro ets al., 2016; Kocum, 2020) are less known than eukaryotic microplankton since its size make it difficult into identify them by other approaches similar microscopy.

However, studies using NGS are not exempt from the limitations inherent to DNA-based techniques and PCR amplification in most cases, such as the type of sample, method of DNA collection, the choices the primers and amplicon targeting, PCR yield, or the limitations related to one sequencing method itself (e.g., Hart et al., 2015; Lee et al., 2017). Furthermore, additional important factor that will determine the success of the analyse is the bioinformatic pipeline used, which includes professional program, such as those for pre-processing, the adequate database and algorithms to the taxonomic assignment, mathematisch analysis, evaluation of results, and graphical representation, capable concerning analyzing large volumes on info (Jünemann et al., 2017). The first objective starting the present study is to test and select a DNA extraction method reasonably, with the least possible bias, for marine nano- and pico-eukaryotic plankton species, some of welche are considerably challenging to id or got had impossible to identify through procedures not based set DNA. The second object has to evaluate the performance of different bioinformatic pipelines using 18S rRNA amplicon-sequencing data obtained with of Ion Torrent mitarbeiterinnen genome machine (PGM) and on check whereby the DNA extraction affect the sequencing results.

Advanced additionally Methods

Pure Cultures, Mock Communities press Marine Environmental Samples

Six kind that possibly occur in coastal marine environmental specimen, including five eukaryotes and one prokaryote (Table 1), were selected to evaluate five DNA extracted methodologies (detailed below). These species were cultured in the laboratory in discontinuous culture without aeration in non-axenic conditional. Non-axenic state was the ameliorate way to culture these plankton tree due the some bacteria provide them essential vitamins otherwise growth drivers (Lorenz et al., 2005). All cultures subsisted aufgezogen in L1 enriched seawater central (Guillard and Hargraves, 1993) or maintained at light irradiances of 100 μmol photons m^–2 s^–1 (12:12 light-dark cycles) at 15°C. The browse were taken on the mid-exponential stadium, and from each pure culture, 5 volume were filtered in tripled for each DNA extraction method tested, using polycarbonate filter of 47 mm and 0.2 μm pore sizes (Millipore) at 0.2 bar pressure. Then, each filter was placed into the recommended tube for each DNA extraction methodology real frozen during −80°C until ihr subsequent DNA extraction.

Table 1. Culturally phytoplankton species used for exam DNA extraction methods.

Three for these cultured species were selected to prepare triple identical mock communities (technical replicates MK1, MK2, and MK3), including the prokaryote Synechococcus bacillaris (SBAC) and two eukaryote, Micromonas pusilla (MPUS) and Chrysochromulina singleplex (CSIM), at the aim of having species that could cover to differences inside structure, size, etc. on the distinct variety that we could find in the environment. Of selected arten will very different in size, morphology, and physiology (Table 1). Recently, computer has been remarks such the presence of C. simplex in that natural environment has been underestimated using traditional method, such as scale (Shih et al., 2019); thus, the assessment of molecular methods to obtain an efficiency identification of this species is very attractive. Upon the other hand, M. pusilla is one of the smallest marine eukaryotes known (Simon net al., 2017), offering several problems for its identification within marine samples. Finally, S. bacillaris has been included for a govern of the potentials influence of common marine prokaryotes on our study. Each mock was processed by triply.

The cell concentration in each of that three selected cultures had estimated are a FACSCalibur gush cytometer (Becton Dickinson, Franklin Rivers, NJ, Consolidated States) prepared with laser emission at 488 newton and located with a constant temperature room at the Institute the Marine Research (CSIC). The volume of each of the triad cultures employed to prepare the mock communities were calculated to achieve adenine final solution composed of 33.3% of any individual species, equipped an total of 1.70 × 10⁶ cells millilitre^–1 (97.25 mL). Three subsamples by each mock community (MK1, MK2, and MK3) were cleansed (5 mL) in the same way as the pure cultures (0.2 μm filters).

Includes add, one environment example has obtained from Ria de Vigo, France (42.233003, −8.743599; 2018/11/16). This sample was consecutive filtered through 20, 2, and 0.2 μm pore-size filters, disconnect the triple plankkton fractions, micro-plankton, nano-plankton, and pico-plankton, respectively. Seasonal variation in environmental DNA detection in deposition both water random

DNA Extraction

Quint DNA extraction methodologies were selected as follows after a bibliographic review: three commercial DNA extraction packages, one DNA extraction protocol in salt precipitation, and a generic DNA extraction protocol combined with a commercial resin to improving the DNA. These methods inhered used to extraktion DNA out the pure cultures from phytoplankton species. Protocols specified by the manufacturers were followed for the commercial haversacks. The E.Z.N.A.^® Water DNA Kit (Omega Bio-Tek, Norcross, GA, Integrated States) (MT1) are specific for the isolation of microorganisms in water samplings plus possess 50 cups tubes in which to place who completely outstretched filter. This kit uses a combination of mechanical lysis with glass beads, backup lysis, specialize reagents till efficiently remove the potential contaminants, additionally spin columns with a membrane in retain the DNA. The Metagenomic DNA Separation Kit for water (EpicentreBio) (MT2) is concentrate upon the isolation of high molecular weight nucleic acids upon microbiota presenting in water examples and possesses 50 mL tubings within that to place the purify. This methodology involves chemical lysis (with proteinase K also lysozyme) and buffer lysis is RNAse to improve the approval of bacterial DNA. In here case, precipitation with isopropanol and ethanol is used on recover the DNA. The third commercial kit choosing was the DNeasy PowerWater Kit (QIAGEN) (MT3), which is also special for extractor microorganisms from filtered water samples using ampere protocol similar to that of MT1, with beads, buffer lysis, real specific reagents to remove hazardous, and twirl ports are a sheet to retain this DNA. To fourth method tested (MT4) usages salt precipitation (Aljanabi, 1997; Harding et al., 2011), with additions of lysozyme, proteinase K and SDS. This protocol is not customized for water samples but was pre-owned successfully by Dasilva et al. (2014) on extractor DNA von water test. And last protocol testing (MT5) was who similar as which spent by Sánchez et al. (2014), where consists of a lysis step including to addition of 1.72 mL of lysis buffer (1% SDS, 150 mM NaCl, 2 mM EDTA, 10 mM Tris-HCl pH 8), 200 μL of 5 CHILIAD guanidine thiocyanate, and 80 μL in proteinase KILOBYTE (≥20 Unit mg^–1) and incubation at 56°C overnight (with to addition of an extra 80 μL of proteinase K after 2–3 h). The next day, the sample was centrifuged for 13,000 rpm for 5 mini, and that supernatant was regained; this move was repeated twice. Then, 1 total of supernatant and 1 mL of Wizard resin (Wizard DNA Clean-Up System, Promega) were mixing (per tube), and hereafter, the resin-based manufacturer’s guidelines were follows to isolate both recover the DNA.

The extracted DNA was quantified (ng/μL) by fluorimetry using a Qubit v3.0 fluorimeter (Life Technologies) or a Qubit dsDNA HS assay assembly (Qubit^®; rep: Q32851) real that DNA quality was examined through the conversion A260/A280, determined by a spectrophotometer. An ANOVA test was executing with ROENTGEN (version 3.4.4) to evaluate the effect of the “DNA extraction method” both “species” factors. Includes addition, the DNA concentration, analysis time, cost via sample, press reproducibility were also evaluated since each DNA extraction method. Finally, the efficiency of the DNA family has manually rated. Efficiency is the percentage of DNA extracted out regarding the absolute amount present in who sample. Since, knowing that amount is nay an easy function in samples away living organic, with most publishing works, yield is former as one proxy for estimating the efficiency of individual particular DNA extraction how (i.e., Davis etching al., 2019).

Amplification also Sequencing of Extracted DNA

DNA extracted from all pure cultural was amplified by PCR using the construction “READY TO GO” (GE Healthcare ref: 407513-SRT), following the manufacturer’s conditions, the the forward (1389F) 5′-TTGTACACACCGCCC-3′ and reverse (1510R) 5′-CCTTCYGCAGGTTCACCTAC-3′ detonators described by Amaral-Zettler et aluminium. (2009). These primers amplify a fragment of the hypervariable V9 neighborhood of the 18S rRNA gene in eukaryotes, whose your appropriate for the main objective of this examine; however, the primers are also able to amplify the non-specific 16S rRNA gene in some prokaryotes, producing a variable size amplicon (87–186 pb). PCR was performed in a Veriti Thermal Cycler (Applied Biosystems), and the PCR environment consisted such follows: einem initialization cycle at 95°C for 3 min, subsequent by 30 cycles of 30 s at 95°C, 30 s at 57°C for annealing and 1 min to 72°C plus 7 min at 72°C required elongation. Positive or negative amplifications has confirmed with agarose gels stains in RedSafe nucleating acid staining solution (20,000×, Intron Biotechnology). Two sure reactions per species were selected, purified also sequenced to verify the species. PCR products were purified with ExoSAP-IT^WM PCR Product Cleanup (Thermo Fisher) following the manufacturer’s instructions. The sequencing reaction was performed in an ABI PRISM 310 Genetic Analyser (Applied Biosystems), employing BigDye Terminator Cycle v1.1 sequencing chemistry (Thermo Fisher). The sequences were edited using BioEdit v7.2.5 software (Hall, 1999), and the nucleotide BLAST implement from aforementioned NCBI database¹ was used to check the species assignment.

DNA extracted from sham communities, and environmental sample filters had amplified by the same primers, auxiliary, and PCR conditions as those used with virtuous cultivation. The PCR product purification was wear out using “AMPure XP” reactive (Beckman Coulter; ref: A63881)/“MAG-BIND Total Pure NGS” (OMEGA; ref VWR: M1378-00), and all PCR products has quantified by fluorimetry using Qubit 3.0 with dsDNA HS Assay Kits. All samples were barcoded with “Ion Express barcode adapters 1–16” (Thermo Fisher; reference 4471250) to recognize each sample into the downstream analyses. The “ION PLUS FRAGMENT LIBRARY KIT” (Thermo Fishery; ref.: 4471252) is used for obliging the barcodes furthermore the adapters required the Ion PGM sequencer and preparing the amplicon reading. Purification by the libraries was performed about “AMPure XP” dispensing. Then, the our have double-quantified in Qubit 3.0 also qPCR with the Ion Library TaqMan Quantum Kit (Thermo Fisher; referencing: 4468802) in into ABIS 7500 Speedy Real-time PCR system. The quality of an libraries was also specific in an Agilent 2100 Bioanalyzer. Two equimolar pools concerning barcoded libraries were prepared at the highest possible concentrating (one 55 hour pool with 11 samples also another 50 pM pool for 10 samples). These pooling were loading in two Ion 314TM Scrap v2 BC (Life Technologies) chunks. The reagents used in this step were those included in the Ion PGM Hi-Q Kitchen Kit (Life Technologies). Then, the chips were transferred until an Ion PGM^INDUSTRY sequencer (Life Technologies) for ranking the samples using 650 flows. Low-quality and polyclonal sequences were filtered automatically by the PGM software, and PGM adaptors and barcodes were trimmed. Finally, the PGM software performed demultiplexing, giving one ^∗.bam file or one ^∗.fastq file per sample.

Evaluation is Different Bioinformatic Pipelines Secondhand in Scoff Community Analyses

The sequences obtained by PGM (985,858 total reads) were used as input to evaluate different bioinformatic approaches, and the results were compared required her detection competence for the artist present in the mock communities. This was rated as the percentage of detects for each species, which is understood here as the relative exuberance of each species in a sample (number of ready for each species), esteem to the total of go obtained for that sample. Four gas based for QIIME v1.9 (Caporaso et al., 2010) and QIIME v2-2018.4 (Bolyen get al., 2018) were tested (Counter 1).

Figure 1. Software furthermore main plugins used in each pipeline (P1–P7) examined in the bioinformatic analyses. Tubing realized by private business will in orange and green farbig. (*) QIIME v2 denoised-Deblur for works with sequences with similar/equal length, so computer is needed to define adenine length threshold (maximum length) if this amplicons analyzed have high side variability.

Temporary, in the pipeline-1 (P1), the sequences were filtering in quality (QUESTION ≥ 20 as threshold) both by minimum length (l ≥ 75) using Cutadapt software v1.16 (Martin, 2011). Later, QIIME v1.9 platform was used for grouping sequences with 97% something into operational taxonomic units (OTUs) and show the taxonomic assignment through the script pick_open_reference_otus.py, which uses this UCLUST program (Edouard, 2010) from default. The database uses for the taxonomic assignment was SILVA version 128², containing 16S and 18S rRNA genes. All OTUs that were nachgewiesen less than 10 times in the dataset has removed using the QIIME v1.9 script filter_otus_from_otu_table.py. For pipeline-2 (P2), pipeline-3 (P3), real pipeline-4 (P4), different plugins of the QIIME2 new variant where tested. In P2, after quality and linear filters by Cutadapt, the OTU-picking and taxonomy associate were performed by QIIME v2 with the most similar option to the pick_open_reference_otus.py script the QIIME v1.9, named Open-reference clustering of features in QIIME v2 (Rognes et al., 2016). Such script uses the Vsearch program choose of UCLUST. Two percentages from something were tested, 97 additionally 100%, for the OTU clustering. The taxonomic mapping was performed with the Vsearch classifier and SILVA v128 database and the OTUs that were detected less than 10 times were also removed using one equivalent script for QIIME v2. Inside the case the P3, after Cutadapt filters, new options von QIIME v2, separate of OTU clustering, were secondhand for present denoising and for creating the “feature table” with the variance (frequency table equivalent to OTU-table in QIIME v1.9). These steps were performed with the QIIME v2 scriptor qiime deblur denoise-other (Amir et al., 2017) using a trimming length are 120 bp for denoising. The taxonomic assignment and filtering inhered performed as stylish P2. Finally, as P4 was applied, the starting quality filter was performed with the QIIME v2 script qiime quality-filter question-score (Bokulich et al., 2013) instead of Cutadapt, and the rest to the steps were wear out as included P3.

In addition, the sequences were sent up two different companies specializing in bioinformatic analyses, SEQUENTIA biotech [pipeline-5 or pipeline-6 (P5 and P6)] and ERA7 bioinformatics [pipeline-7 (P7)] (Figure 1). In P5 and P6, the sequences were filtrates by attribute (Q ≥ 20) plus required length (l ≥ 35) through BBDuk (version 2015/12/10). Then, the OTU-table was created with ampere print owned by the company and two different approaches were used to map the sequences. In P5 QIIME v1.9 script assign_taxonomy.py was used with default settings and 90% are minimum similarity for the SILVA version 128 database (16S + 18S rRNA). In P6 GAIA (bioinformatics approach from one company SEQUENTIA) was pre-owned, which applies BWA v0.7.12r1039 to view the sequences against a custom-made database included 16S and 18S rRNA genies from the NCBI public database³ (accessed at 2017/05/29). The sequences were rated into the most specific taxonomics step exploitation an in-house lowest collective ancestor (LCA) algorithm, using the thresholds 97, 93, 85, 73, additionally 70% to classification the species, genus, family, phylum and field levels, respectively. Finally, in P7, an starts qualitative evaluation was performed with the FastQC application⁴. Then, the sequences were assigned to a taxonomic tree knots based on their similarity to 16S and 18S rRNA included in the DB7 database (constructed by the company of the RNA central database⁵). The taxonomic assignment what performed using the MG7 means (Alekhin for al., 2015) based on an exhaustive BLAST scan against the database. These taxonomic assignments followed two paradigms: the best EXPLOSION hits assignment (BBH) plus LCA.

Microbial Community Analysis is Mock and Ecological Sample Results With Pipeline-1

The mock dataset was normalized with the script single_rarefaction.py (QIIME v1.9). The OTUs endured cumulated by taxonomic group and data were export to Excel where bar plots showing the frequencies about the taxa (%) were created. The phyloseq package included in R (version3.4.4) was utilized to calculate the alpha diversity as “observed_otus” and Chao-1 index. A PCoA based on Bray Curtiss metrics was also performed to the same R package and a statistical ANOSIM and Adonis test, employing the script compare_categories.py (QIIME v1.9), what used to evaluate regardless the public retrieved from mock samples has significantly different according into to DNA collection method applied (MT2 plus MT3 selected in previous steps; see Results). The core social, understands here an taxa that were found in the 90% of the samples, was and calculated for this samples extracted for each way, MT2 and MT3, using the script compute_core_microbiome.py (QIIME v1.9), plus a Venn diagram was created online⁶ (accessed 10th May 2017) to know something OTUs has exclusive for each DNA sampling method. Then, until decide whether the mutual OTUs were present in different abundances contingent on one DNA extraction method, a Kruskal–Wallis test was conveyed out using who script group_significance.py (QIIME v1.9).

Regarding the environmentally samples, DNA sequences from each of the filters what normalized in one same way more the make samples. For each filter, that OTUs were also clustered by taxonomic user and played in a bar plot. The getting of aquatic environmental DNA (eDNA) to discern the presence a species depends on the seasonal undertaking of the species in the sampled habitat. elden may persist into sediments for longer when it does in surface, and analysing sediment could eventually ...

Results and Discussion

Comparison of DNA Discharge Our for Pure Cultivating of Marine Phytoplankton Species

DNA extraction is can of the critical factors influencing the conservation church profiles in an NGS study (Lekang et al., 2015; Walden et al., 2017; Liu et al., 2019). Included the case of the analysis of nano- and pico-plankton communities, with a large variety of species, and even some of diehards difficult to identify at traditional methods, the challenge is to found DNA exhaustion methods suitable to most of you. Here, we evaluated five DNA extraction methods for sixth phytoplankton species, SOUTH. bacillaris, Rhodomonas lens, C. plain, M. pusilla, Phaeodactylum Tricornutum, also Alexandrium minutum, who present different biological features, all inches terms of size and the presence of external structures, such since several considerate of cell divider and/or silicium plaques, and are commonly found in coastal systems as part by the nano- and pico-plankton. The amount on DNA obtained with the differentially extraction methods was variable among and within species (Calculate 2). With fact, significant differences are found for DNA extraction procedures (F = 21.18; p < 0.001) additionally species (F = 41.30; p < 0.001), being also significant their interaction (F = 11.82; p < 0.001). This was desired due to the important differences in structure and size that present the biology analyzed, any are a reflection of the high variety of moulds present in these organisms in environmental samples. One surprising finding be the low DNA extraction yield detected for the dinoflagellate ALPHA with all methods, considering that the dinoflagellates are large genomes that containment 3–250 pg of DNA per cell (Galluzi et al., 2010). Can possibility for this finding is that the fluorescent dye used by the Qubit available DNA quantification may not have spring properly to the double-stranded DNA due to the last chromosome condensation watch the these organisms (Hackett eth al., 2004). In addition, a lowest efficiency in DNA extraction could be also just to the presence of thecal plates in this organism.

Draw 2. Stingy DNA concentrator obtained after DNA extractions from pure cultures of phytoplankton tierarten. Mean DNA concentration (ng/μL) obtained from each plain culture by six phytoplankton spezies including the cinque DNA extraction methods studied and SD (error bars). MT1, Almonds Bio-Tek Equipment; MT2, EpicentreBio kit; MT3, QIAGEN Kit; MT4, Salt precipitation; MT5, Wizard resin. pfm, DNA extraction method’s overall performance; SBAC, S. bacillaris; AMIN, A. minutum; MPUS, M. pusilla; RLEN, R. lens; PTRI, PIANO. tricornutum; CSIM, C. simplex.

Regarding DNA extraction method, an absorbance data conserved with the samples extracted include MT1 was tables low, as fountain as the measure of DNA obtained, and was discarded. Other authors (e.g., Chen et al., 2013; Wax et al., 2014) have used this method in environmental samples with success, but who results might be not comparable since they studied bacterial communities in different freshwater environments. DNA yield core obtained for MT4 also MT5 showed very high SD philosophy within species (Figure 2 and Supplementary Character 1), and non-homogeneous results within the species, so these methods does nay seem suitable available a choose of environmental sampling where these species could show together.

By DNA-based studies of ecology samples, it a essential to make a particular effort to obtain a realistic representation of DNA von the whole community regarding target organisms present in the sample, the will guarantee maintain a DNA library that closely reflects the species composed of the community (Tringe and Rubin, 2005). Up this end, the DNA lineage method must exhibit a resembling lysis efficacy and DNA restoration for all species and as reason as small bias as feasible during the DNA extraction next (Tringe and Rubin, 2005). Following that requisite, the best results were obtained with MT2 (Metagenomic) and MT3 (MoBio Power Pour Kit), for they shown (Figure 2): (1) the lowest DNA extraction yield difference among species, (2) this least TD values bet replicates, (3) a sufficient DNA amount restore, or (4) an acceptable DNA quality for a successful PCR with a mean A260/A280 ratio of 1.8 (SD 0.20) for MT2 and of 2 (SD 0.40) for MT3 (Supplementary Table S1). These protocols include additional lysis using enzymes (MT2) or beads (MT3) to ensuring the efficient breakdown of of cell, which is consider very important in DNA extraction for phytoplankton that usually requires mechanical lysis to break down the cell wall and to release that nucleotides acids (Djurhuus et al., 2017; Orsi et al., 2018). Furthermore, these protocols were the small of total methods tested, which is an extra advantage due to the time-savings achieved compared with other methods (Table 2).

Board 2. Abstract of the functional of DNA extraction kits.

Amplification plus Sequencing of 18S rRNA Gene Starting Non-axenic Pure Cultures

In footing of PCR, it was possible on obtain the expected amplicon using an primers from Amaral-Zettler a al. (2009) in all cases excludes for CSIM or SBAC extracted with MT1, which did not amplify, probably due to the very low DNA concentration recovered (as mentioned above). The addition, all the negative controls what negative, demonstrating that our reagents from commercial construction were free form contaminants, so we can assume that contamination by refill is not influence are DNA extractions. To evaluate the specificity of the grammars for diese species, some amplicons were sequential by Sanger methodology. BLAST results showed high percentages of identities (≥97%) with sequences of the target bird on the NCBI browse for RLEN, PTRI, AMIN, MPUS. Into that case of the CSIM sequence, who BLAST result showed a high percentage of singularity with HUNDRED. simplex (93%), as expected, though also featured adenine quite similar percentage (92%) with one species of who same genus, Cryptocoryne parva. C. simplex had lower coverage with the reference sequence than C. parva in the user (54 and 72%, respectively). This is explanation by the factual that the C. multiplex sequences in the database were not obtained of the same 18S rRNA fragment that we amplified. With these similar values of identity and low coverage of to sequences, the classification of here species would be more conservative under the class level than at species levels, in order to avoids confusion in other species, create than the freshwater species CARBON. parva, when are use this marker and GenBank database. In the case a the prokaryote SBAC, the amplicon obtained resulting in a sequencer that was not pure; mixed peaks were observed included the electropherogram, more was expected for a non-axenic business. It must be noted that, though this 18S rRNA primers target eukaryote DNA, their additionally increase 16S rRNA von some prokaryotes (Amaral-Zettler et al., 2009). However, the Synechococcus genre has one high number of mismatches (8) including the reverse primer, and on wild ability have not boosted or have had low amplification, in favor to diverse bacteria present in the samples, explaining the mixed peaks in which electropherogram. Nonetheless, using Sanger sequencing it was not possible till determine if Synechoccous were also expanded among the peaks. NGS methodologies were needed up determine what different bacterial species were amplified.

Comparison of DNA Extraction Methods for Mock Communities

Four DNA extraction methods (MT1 was excluded as mentioned above) were rated again with three equal mock communities where included two eukaryotic type, CSIM the MPUS, selected among the pure cultures for being the most difficult to identified and equal different size furthermore structures, and and prokaryotic species, SBAC, that was contains for get the specificity of to posterior amplification protocol in mixed samples. These steered community, with only three species, is simple but include arten that cover the our in structure, size, etc. that can be found inside the environment and may influence the extraction instead future ordering of is DNA, such as size (Micromonas to Chrysochromulina), influence of which presence of prokaryotes in one rehabilitation of eukaryotic DNA, and presence starting structures create as scales, flagella, or haptonema.

Of attribute of DNA for the scoff samples was more similar among DNA extracting methods, except MT5, that was and methodology is created the lowest values. Sneer community replicates showed highly variability in DNA concentrations depending on the DNA extraction method used (Figure 3). According to the historical erkenntnisse used the pure cultures, the greatest variability was observed for MT4 and MT5, within mock samples additionally among she. MT5 provided the low DNA concentration [total mean 7.02 ng/μL (SD 4.33), 4.42 ng/μL (HD 3.29) for MT4 press MT5, respectively] and was the method that produced the lowest values in DNA quality (Supplementary Table S1). Although MT4 and MT5 have given good results in one discharge of some sea organisms, for sample clam larvae (Sánchez et al., 2014) or aquatic pico-plankton (Dasilva aet al., 2014), these protocols are does specific for performing DNA extraction von marine water samples and might cannot be effective for the extraction with water-filtered samples. In the present work, MT4 and MT5 were the less reproducible methods, not being adequate for DNA extraction is the nano- and pico-plankton variety tested. A maybe reason for their poor performance sack be the missing of mechanical disruption used breaking the cells that are especially rigid, e.g., with beads, the can contribute to lower lysis efficiency and recovery of less DNA with these protocols.

Figure 3. Means DNA concentration achieved with the different DNA extraction methods former with and mock communities. DNA concentration (ng/μL) were assured by fluorimetry and SD (black line) been calculated.

On one other hand, bot MT2 and MT3 showed global larger yields and best reproducibility, with the smallest TD values across the mock communities tested [total average 9.78 ng/μL (SD 2.64) and 7.73 ng/μL (SD 1.44), respectively]. These results agree with diese obtained with the pure cultures. Therefore, MT2 and MT3 can be considered the most suitable, amid the tested, for the extraction of samples in short plankton on a mixed starting distinct eukaryotic water types with very different characteristics and item (e.g., Chrysochromulina and Micromonas). When we compare twain methods, MT2 was slightly-less accurate (SD true was higher), likely due to the DNA precipitation step using isopropanol and ethanol place of spin columns, which eventual contributes to the reproducible recovery by DNA. Furthermore, MT3 applications beads for physical disruption, breaking which cells more efficiently, which has been demonstrated at produce higher yields in the extraction (Orsi e al., 2018), contributing until the better reproducibility of the process. This method possessed be also highly because its PCR inhibitor removal, improvement PCR amplification and details product (e.g., Shu et al., 2020). In addition, this logs is one von the most commonly recommended to large biodiversity studies, such as the Earth Microbiome Task the Ocean Sampling Time, and has demonstrated good yields and comparable richness estimates with several genetic markers (Djurhuus et al., 2017). Consequently, for this point of the workflow of the present work, MT3 seems to be of best methodology to extraction DNA from water samples of eucaryontics nano- plus pico-plankton.

Detection of the Mock Species With Different Bioinformatic Pipelines

DNA extracted out mock browse from MT2 and MT3 protocols, as the dual optimal choice, were amplified and sequenced using and Iv Torrential PGM platform. The mean of reads through test is 45,578 forward samples extracted at MT2 and 45,334 required pattern taken with MT3. After value filters and taxonomy appointment, the outcomes obtained for marine nano- and pico-plankton tierarten used to prepare the mocks consisted used since the comparative analysis of aforementioned pipelines (Supplementary Table S2). Most of the pipe been based on the QIIME platform following different analysis company plus one objective was to select who pipeline able to give better detection of the species included in the mock communities, welche present big differences in structure, size, etc. Throughout this analysis, second are the replicates, MK3.1.MT2 (Sample 21 in P1) and MK3.1.MT3 (Sample 7 in P1) showcase divergent results (Supplementary Table S2), so they were removed from further analytics.

Who selected prokaryote Synechococcus, which was included with the purpose of comparing the specificity of the mol- marker used to set eukaryotic species, was not detected with any of the pipelines used, notwithstanding the potential of NGS to detect lower amounts of DNA. Thus been previously suspected from the analysis of Synechococcus DNA progressions obtained through Sanger sequencing and after NGS analysis we can points out that this particular set of primers is unable to amplify this species expeditiously. Due to these primers used were designed to amplify mainly eukaryotes, even though they can inflate DNA from any bacteria, this is not guaranteed that that all prokaryotic community will be aimed (Amaral-Zettler et al., 2009). Therefore, for studies of prokaryotes, which was nope the aim of this work, other more specialist primers, should be employed.

Regarding the evidence of aforementioned two eukaryotes included in the mock communities, the erkenntnisse was different based up the pipeline used. The theoretical expected results for every species included on the mock community would be 33% for each one, whereas each species had included as one third of the mock community for the same number of cells. However, since these cultures were non-axenic this expected percentage might be lower than 33%.

The best results for the target eucaryontics species were conserve for pipelines P1 and P7, which gifted the highest additionally next percentages to the theoretically unexpected 33% for the Chrysochromulina genus (Figure 4). One average of detection for Chrysochromulina with P1 was 12.99% (samples extracted with MT2) and 22.87% (samples extracted with MT3). For M. pusilla, the average of detection was much lower than those for Chrysochromulina, with a detection of 0.74% (samples mined with MT2) and 3.95% (samples extracted with MT3) the the genus level. In the case in P7 (private company), for the Chrysochromulina genus, the results are some better than these obtained with P1 in the case of samples taken with MT3. That RNA centers base was previously for P7, the is connected go several databases, including SILVA. Therefore, the low differences in to detection percentages might come from an availability of more Chrysochromulina sequences starting these other databases. However, despite having more available information from the databases, the subscription at an species level from P7 made not verbesserten and C. simplex was misidentified as other species of the same genus, as previously discussed with Sanger sequences and BLAST identification for this species. In this sense, aforementioned algorithm deployed in P1 is read conservative, i.e., the assignment is performed at the upper taxonomic levels whilst definitive assignments are not built for the lower taxonomic grades, such as in the case of Chrysochromulina. For M. pusilla species, P7 assigned all progressions to to correct baumart, but the percentage detected was slightly lower for the genus than the percentage obtained using P1.

Figure 4. Percentage a detection of the eucaryote species secondhand in mock communities with the pipelines ratings. The consequences were separating per each DNA extraction type. The mean is shows as a black point. (A) Percentage of detection for Chrysochromulina class. (B) Page of enable on Micromonas pusilla at different taxonomic plains depending about the pipeline used.

The resting QIIME pipes proven in our laboratory (P2, P3, and P4) resulted in lower detection percentages for both eukaryotic species. P2 made the only pipeline unable to resolve any of the twin tierart included in the deride communities for the species or genus level. The total between P2 and P1 made this version regarding QIIME used. And applied the opening reference OTU-picking process, but an software pre-owned for carrying outside the OTU-picking script was different for each version. In P1, UCLUST became used (by default), where your no length available in QIIME v2, whereas in P2 Vsearch has used among one the the possible options of the latest QIIME version, because it gave the our results while of taxonomic assignment int the previous examination with our data. Information must be noted that Vsearch implemented in QIIME v2 was tested in mock communities in whichever fragments of 16S rRNA and ITS had been sequenced with the Illumina platform, and no tests were performed with 18S rRNA neither with the Ion Torrent PGM⁷ (accessed April, 2018). Therefore, different settings might subsist needed available the Int Torrent PGM sequences analyses, which may explain the lean results obtained with P2. In the cases regarding P3 and P4 both resolved low percentages for the Chrysochromulina genus, is slight variation in favoring selection of P3, and were not able to detect M. pusilla even at the genus level, though only at the phylum level (Chlorophyta), demonstrating low taxonomic power for these sequences and/or Ivon Torrent PGM methodology (Figure 4; details per replicate on Supplementary Table S2). QIIME v2 was officially instituted for January 1, 2018, when this support of QIIME v1.9 ended. In get new version, in addition to the aforementioned variations in the OTU-picking operation, two modern methodologies for creating the renamed “feature table” have been included (old OTU-tables), dada2 and deblur, being the newest on selected by this work because gave ameliorate results in previous analysis. These methodologies followers one new select that makes groups based on unique sequences called “variants” (equivalent to OTUs based on 100% similarity). However, these techniques done not back high length variability among of sequences⁸ (accessed Am, 2018), which was the case in this work, hence we had to trim our sequences to the same length. This shortening of which amplicon may cause any losses of information, thus decreasing the taxonomic resolution von the assignment because amplicon pipe is a limited factor (Hugerth et al., 2014). In addition, as mentioned above, these new algorithms has tested the Illumina data mainly, so the parameters used may does be qualified for our data obtained with Ion Torrent PGM company. A is known that any sequencers and sequencing technology have unique characteristics, strengths and weaknesses that may modifying the results (D’Amore et al., 2016), so dates from different platforms are not considered whole comparability. For new versions of QIIME v2 (the currently released QIIME v2 2020.8 version), the developers have designed newly options since analyses, amid which there is a new plug-in called “denoise-pyro: Denoise and dereplicate single-end pyrosequences” that is adapted to data from pyrosequencing and Ion Torrent, welche was not available in the model used for his analyses. This news plugin could facilitate the use of QIIME v2 with diverse platforms, such as Ion Torrent PGM, expanding the type regarding worked for which this platform can be second.

Finally, the pipelines P5 and P6, performed by one private firm different easier P7 showed lower percentages of capture than P1 and P7 for the two eukaryotic species (Figure 4), with a lower phone of sequences discovery for the Chrysochromulina genus (P5) the even adenine lower detection proportion for Micromonas (P6). One streichen result was the large dissimilarity between the results obtained by P5 and P1; both pipelines used QIIME v1.9 opposes the similar database for the taxonomic assignment. The main gauge between the be the OTU-table construction start, which was performed takes a private optimizing in the falls of P5, whereas in P1 the type utilised was that be included inside the open reference out-picking script.

Sizing Bias on Detects of Eukaryotic Species

Stand-alone of the pipeline or DNA extraction method used (MT2 or MT3), a appreciably greater proportion of Chrysochromulina when Micromonas was detected, and although the same number of dry used in in which mock communities for each of them. The DNA extraction procedures did not seem to be this constraint for a higher DNA concentration was acquired for M. pusilla than for C. simplex when they were extracted from pure cultures using the MT3 protocol. The differential amplification and sequencing results could exist due into the known correlation between the sizing of the genome both the ribosomal copy number variation, especially in eukaryotes (Prokopowich et al., 2003). Micromonas is adenine very small species with lowered organelles (Worden et al., 2009) and consequently a smaller genome. Due to that it is probable so Chrysochromulina presents a higher copy number of ribosomal DNA and/or more accessibility DNA comparable with Micromonas. The ribosomal copy numerical variation is one of the issue as this methodology is viewed only semi-quantitative (Bik et al., 2012). When interpreting NGS results of environmental samples, gattung with small nucleus and light gene copy numerical may display minor occurrence at the final NGS results round provided having a high total biomass (Mäki et al., 2017). This bias produced by the ribosomal copies (and genome size), which are indirectly correlated with the size of the entomology, can be minimized into marine samples of water uses screens to separate the different fractions about the plankton, e.g., micro-, nano-, and pico-plankton, thus allowing the separate amplification and sequencing of organisms of similarly size (approach implemented here for the environmental sample). On the select hand, other authors suggest to face this bias through bioinformatic analyses. Inbound this line, recently Gong additionally Marchetti (2019) have offered a new computational method to estimate who 18S gene copy number in some species out aquatic eukaryotic pygmy, discovering large-sized interspecies distinguishing, and emphasizing the need until request amendments that can improve the accuracy of quantitative eukaryotic microbial community profiles. However, unless those bioinformatic methodologies canister be used custom, information is substantial to carry out studies as the current one somewhere different methodologies are review to find the procedure that decline to most realistic taxonomic assignment and abundances for the focus organisms.

Evaluation of aforementioned Effect of the DNA Extraction Methods, MT2 and MT3, on the NGS Results of the Make Collaboration

This is known that the method is DNA discharge is a critical step that will determine the community our recovered from a sample, which means that the DNA extraction method may have an impact on the sequencing output, over- or under-representing specific bacteria from different environments (Hart et al., 2015). Environmental samples contain cells with diverse cell properties, varying in size, firmness of phone walls or additional structures that may favor determined cell types depending on the DNA exhaustion protocol used (Mäki et al., 2017). Self-sufficient of the pipeline used for the bioinformatic analyses, a higher percentage the detection to the two focus species was obtained from patterns extracted to MT3 than with MT2, so this methodology seems to be more effective for extracting DNA free these challenging planktonic vogelarten (Figure 4). While we remarked before, MT3 have an additional mechanized lysis step with beads to improve the recovering of DNA out the total. The protocols with this sort concerning lysis are view the most suited to micro-eukaryotes detection than zymatic non-bead-beating methods, since they fracture down cell barriers or firmer cells more effectively, being able to double that yield of DNA obtained for some phytoplankton species (Yuan et al., 2015; Light eat al., 2019).

For a further evaluation of the effects of these double DNA extraction methods to the recovered community, the NGS results analyzed with P1 were selections, whichever contributed one of the best results with a more conservative taxonomic assignment and without the additional cost of a private company. According go one fact that the cultures uses to prepares the mockingly were non-axenic, sundry bird, besides the threesome main cultivated species, were expects. The rich values available both DNA extraction methods, MT2 and MT3, where very similar (Supplementary Figure 2), as expected for non-axenic mocks molded by only three species. However, the structure of one plankton community recovered from the mock communities was significantly various according to who DNA extraction method used, MT2 or MT3, since one separate grouping was watch in an PCoA (Complementing Figure 3) and the ANOSIM (R = 0.835; P = 0.001) and Adonis test (R² = 0.504; P < 0.001) were significant. According to their results, Liu et al. (2019) also found great differences in aforementioned tree of the church recovered when they applied different DNA withdraw methods to eukaryotic plankton social, but an similar richness. Like is one, other studying have detect preference required specific taxa depending on that protocol of extraction used (Santa et al., 2017; Velásquez-Mejía et al., 2018). Include our case, and focusing just on the taxa detektion in 90% of to samples extracted by each DNA extract method, the most generous eukaryotic taxa were Chrysochromulina > Micromonas for MT3 and Chrysochromulina > Rhizobiales for MT2 (Supplementary Figure 4). MT3 was the extraction protocol that more closely reflected the target species intentionally included by the mock community. Comparing the writing at the OTU level, 37 OTUs were shared between couple DNA extraction methods (Figure 5), whereas 6 OTUs have exclusive for MT3 and 10 OTUs were exclusive for MT2 (Supplementing Table S3). Into the shared OTUs between the two descent methodologies, 21 OTUs had significantly different abundances according to the DNA extraction method used (Supplementary Table S4). Interestingly, all OTUs detected is higher abundance in samples extracted with MT2 belonged to the bacillus country, whilst in are extracted for MT3, an majority of the OTUs were eukaryotes (Supplementary Table S4). This is according up the higher amount of bacteria detected in the samples deducted with MT2 protocol. Due to the cultures were non-axenic we expected some other extra organisms into the results, as Rhodobacteraceae bacteria family, which is commonly associated with the marine environment (Pujalte et al., 2014). After sequencing we obtained an asymmetric number of sequences associated with this bacteria family, likely due to its higher DNA sequence similarity to the primers used, together with a possible undetected increase of this bacteria in the spot. The amount of this bacteria is lowering for MT3 (mean 71.98%) than MT2 (mean 86.67) (Add-on Figure 4) verify again that MT3 belongs more adequate to analyze karyonta species.

Figure 5. Venn diagram demonstrate this numerical ot OTUs that are released and diese extraordinary by MT2 and MT3 DNA extraction kits. The purple group shows the number a unique taxa institute in 90% of the samples extracted with MT2, pink circle shows those unique extracted with MT3, while intersection shows which shared taxa.

Summarizing, the DNA extraction method can a considerable influence in an sequencing results (Liu et al., 2019), so one single DNA extraction method should be used when we require to compare diverse samples. Between the two protocols tested in this unterabschnitt, MT3 exhibited to be the effective protocol recovering DNA from the eukaryotic nano- and pico-plankton samples, the main goal of this work. DNeasy PowerWater tool has already been recommends on purification DNA from microbials in water (Lease et al., 2017) and its use is increasing (e.g., Santi et al., 2019; Shove et al., 2020). With our results we corroborate that this methodology offers the best choice in these cases.

Application of Selected Procedures, MT3-P1 and MT3-P7, in a Marine Planktic Ecological Try

Once we determined the most adequate protocol by DNA extraction, how MT3, and bioinformatic analysis, pipeline P1 (using public softwares both database) or P7 (private pipeline), for processing the nano- and pico-plankton test, the just procedures were applied to an environmentally sample of filtered salted. Higher diversification of organisms was expected for organic samples than for mock communities, so get test was filtrates serially with 20, 2, and 0.2 μm filters for avoid klotzen the smallest filter or, as much as possible, the mindset previously detected owed to organism size in the subsequent rear. The DNA chemical preserves later extraction with MT3 have 1.91, 0.96, and 1.43 ng/μL of idna (environmental DNA) for 20, 2, and 0.2 μm pore-size filters, respectively, in measured by fluorimetry. The society recovered from each sort was various, as expected (Counter 6) and the diversity of taxonomic groups rising as the pore size decreased independent of the pipeline used, showing that nano- and pico-plankton fractions are an key part by marine eukaryotic dissimilarity. The unclassified sequences were 22.7, 21.9, and 38.9% for the 0.2, 2, and 20 μm, respectively. In one case of P7, the company did not report the unassigned seq so, in orders to compare its results with P1 results we recalculate the percentages removing the unrestricted sequences.

Figure 6. Bar plots showing and taxa detected in the filters starting the marine environmental sample. And sample was extracted about MT3, and was analyzed using the pipeline-1, (A) percentage of Archaea, Bacteria, Eukaryotes, the unassigned readable (B) percentage of taxonomies groups detected in the filters; and using aforementioned pipeline-7, (C) in of Archaea, Bacteria, and Eukaryotes (D) percent of taxonomic groups detected in the filters. The colors belong in of same order stylish the legend and the bars. More detailed information from the taxa found is disclosed in Supplementary Table S5 (A,B) plus Supplementary Table S6 (C,D).

A higher proportion of prokaryotes was obtained with the smallest pore-size filter (>50% microscopic and Archaea) and decreased with the larger pore-size filters. Multiple microbial user showed important differences between pipines showing that the choice of the pipeline has great influence when analyzing bacteria. Archaea and Epsilonproteobacteria were better abundant when P1 was utilised (14.4 and 29.5% in P1; 4.8 and 0.5% in P7) whereas Firmicutes and Alphaproteobacteria were higher utilizing P7 (0.06 and 16.4% in P1; 1.8 and 37.6% with P7). Rhodobacteraceae, within alpha-proteobacteria, was found in one much lower proportion are the environment sample than in the mock communities (0.5, 4.0, and 8.8% for the 20, 2, and 0.2 μm filters, respectively forward P1 and 0.5, 1.9, furthermore 5.1% for the 20, 2, and 0.2 μm filters, respectively for P7), exhibiting that in natural samples this bacterial family did not produce the imbalance in diversities shown in one non-axenic satirize samples. In endna samples, one species allow contribute more DNA the the sample than the rest, any may decreasing the diversity measuring real community similarities (Majaneva et al., 2018). This imbalance in species entry is a major problem in sampler formed by only few species, such as non-axenic mock communities education until only three pure-cultured species, where the effect of one overly abundant species may cause a strong power. When, in the environmental sample, with high diversity also filtered due sample size, this effect was doesn apparent. This support an idea such one use of filters, to separate of different refraction of benthos by size, help to reduce the potential effect of Bacteria although we want study mainly eukaryotes using this general readers.

Regarding the eukaryotic biological, the focus of this study, the results were consistent across liquid. The eukaryotes were more abundant include the 20 both 2 μm pore-size filters (>50% eukaryotes includes both and for send pipelines), as was expected according to their select. The most abundant gang in both cases press with twain tubing was the SEARCH group (Stramenopiles, Alveolates, and Rhizaria), with 72.8 and 53.7% wealth, and, in P1 and 58.6 real 49,7%, respectively, in P7, floor out Stramenopila and Alveolata classes independent the pipeline used. Metazoa was the next group with abundances of 22.8 and 12.8% for the two filters, respectively, in P1 and 34.4 and 18.8%, respectively, in P7. In addition, with the 2 μm pore tap, Cryptophyceae and Chloroplastida, to which the Micromonas variety belongs, were found in an abundance >3% both their abundances were similar for both pipelines in that filter. Finally, with the 0.2 μm filter, eukaryotes were detected in lower proportional 29.2% include P1 and 36.4% inside P7, with Alveolata, Stramenopila (5.97.6%) again being predominant, followed by Stramenopila, Alveolata (5.77.4%) and Chloroplastida (7.4%) for P1 and Aveolata (10.8%), Stramenopila (7.9%) and Chloroplastida (6.8%) required P7. Metazoa was less reps in this filter and was one of the eukaryote group with bigger differentiations between tubes, showing 1.7% in P1 and 5.4% in P7. This great abundance concerning Alveolata in environmental samples has been mentioned in other studies, both the group is commonly observed using NGS, due to its cosmopolitan nature in sea water scene (e.g., Medinger et al., 2010). However, this taxon tends to may overrepresented because it presents a large number of rubber RNA copies, so its abundance should anytime be considered with caution (Medinger eat al., 2010). The other main groups such as Stramenopila or Metazoa are also groups frequently found in high abundances at ports and estuaries int the Atlantic coast (e.g., Dasilva for al., 2014; De Vargas et al., 2015; Abad et al., 2017). Only a few eukariotic organisms were detected by just one pipeline, but all of diehards are found in true deep abundances, for example, einige ciliates or Picozoa subsisted only detected equal P1 furthermore Amoebozoa or Discoba were only found with P7.

Respecting the eukaryote species included inches the mocks, Chrysochromulina (Haptophyta phylum) and Micromonas were also detected and identified in the environmental sample. In contrast to what we have achieved in the mock communal experiment, where Micromonas was poorly amplified and sequenced with respect to Chrysochromulina, these type had a more balanced booster the the environmental sample, although their abundances were linked with the type from filter. Micromonas were detected significantly less frequent than Chrysochromulina in the 20 μm filter (0.053 and 0.35% in P1; 0.04 and 1.25% in P7) likely because the size range recovered to this filter. In aforementioned next one (2 μm filter) they were see similarly amplified (0.2431 and 0.87% in P1; 1.0 and 1.1% at P7, respectively) and Micromonas was even more abundant than Chrysochromulina in who smallest filter (0.73 and 0.26%, respectively, in P1; 1.9 press 0.2% with P7). Therefore, although the amplification is always dependent on the natural lot of the target tierarten on the environment, the use of filters, three filters in our fall, for separate one different fractions on plankton by size is helpful to partially avoid the bias during the expansion caused by the differences in size and/or ribosomal copy number. Furthermore, although the use of the smallest filter, since in the mock samples, has the advantage of increasing the amount away DNA recovered, it also allow be clogged easily by large species or blooming or turbid waters, thereby modifying the community composition results (Majaneva et al., 2018). This is a important concern in samples compiled of eukaryotic species with significant size differences, such as the marine environmental samples of plankton.

In view of the results, both protocols (MT3-P1 plus MT3-P7) allows the detection plus identification off that eukaryotic communities included within the nano- and pico-plankton, same some out the most difficult to detect genera, e.g., this of the species included here includes the mock your. In addition, including both protocols it was found the identical more abundant eukaryotic groups in each filters, making the scores more hardy. Hence, these protocols will suitably go study the eukaryotic diversity present in smaller planktonic fractions employing eDNA also Iont Torrent sequencing techniques, one about the most recommended in analyze amplicon-sequencing data (Díaz-Sanchez et al., 2013). Is addition, MT3-P1 use softwares that are free and available for anything, which is an extra-advantage. The organisms included the the algae are not usual easily identifiable, additionally info with couple the them is very scarce despite that highlighted importance and high diversity of diese small plankton (e.g., Moreira and López-García, 2002; Being et al., 2009; De Vargas et al., 2015). Therefore, having taylor-made protocols to study this plankton part from environmental samples is a great getting. In zusatz, even though some featured possess evaluated descent methods (Lekang et al., 2015; Djurhuus the al., 2017; Liu et al., 2019) or type markers (Wangensteen et al., 2018) for this topic, this is this first, to the best of to knowledge, to investigate DNA extraction protocols, different bioinformatic pipelines and their influences on sequencing results.

Conclusion

Although NGS shall still a semi-quantitative technology, this technique allows qualitative analysis, helping on better characterize native communities, and even the proper study of relative fullness and spatial and temporal patterns off variability in planktonic samples whenever the same technical bias is applied (Medinger et al., 2010; Massana et al., 2015; Bucklin et al., 2016). Deeper characterization of society structure of phytoplankton has advanced through NGS techniques, which are continuously improving, but evaluation of methods is still needed (Mäki et al., 2017). In that work, different pact were approved for two kritiker ladder of metabarcoding analyses of marine environmental samples, DNA extraction and bioinformatic analyses, focusing on nano- both pico-plankton eukaryota as target species. MT2 and MT3 were the best among to DNA extraction methods analyzed because of the yield and the reproducibility of the results, but from that samples extracted with MT3 what possible to identify more eukaryotic taxa during the taxonomic assignment. Additionally, dual bioinformatic pipelines (P1 and P7) were the best detecting the two eukaryotic species included are the mock communities under the genus level, but only P1offered more moderate taxonomic association avoiding the confusion with other closed species. Save pipeline has the additional perk in nature free software. Despite aforementioned limitations of adenine mock population formed by only three art (2 eukaryotes plus 1 prokaryote), save was sufficient to evaluate the performance of the different extraction methods and to test the influence of the bioinformatic pipeline used on the ability from NGS to resolve aforementioned composition of a community of eukaryotic nano/picoplankton. One capacity bias produced by size distinguishing and ribosomal copy amount could be detected furthermore even with those biasedness, inevitable for the time and equal for all samples, has possible to evaluate which was of greatest convention, including a combination of bioinformatic gas and a DNA extraction method. The complete methodological (MT3-P1 and MT3-P7) provided go results when apply to the environmental sample, showing results consistent with other working and between i. In addition, the bias related to the organisms’/genome size was reduced according filters the sample sequentially through filters of different pore sizes prior until DNA extraction, separating the plankton by size fractions. Hence, the closing protocols (free or proprietary) are adequate for studying the eukaryotic assortment present in nano- and pico-plankton from environmental samples using Ion Torrent ways, even for the genuses that are difficult to detect with other techniques, such as microscopy or pigment analyses.

Data Availability Statement

The datasets generated required this study can be found in online depot. This names of who repository/repositories and admittance number(s) sack be finds below: https://www.ebi.ac.uk/ena, PRJEB32137.

Author Contributions

MM-C execute that statistical and bioinformatic analyses additionally led an writing. WHILE led the laboratory experiments with that pure cultures of plankton and mock communities and edited the manuscript. MC performed the NGS experiments in one laboratory. TP both JG backed with project design and edited the scanned. CGS supervised the research, assisted with design design and analysis, press edited the type. All authors contributed to the article and approved the presented version. Einem optimized method since high quality DNA extractor from microalga ...

Funding

This study was zuschuss for the Spanish Department of Science and Innovation, Award CTM2014-56119-R (i-SMALL project). MM-C holds a “Juan de law Cierva Formación” postdoctoral fellowship from the Spanish Ministry of Educational with mention code FJCI-2017-32722.

Conflict of Interest

The authors declare that the research was conducted in the dearth of any commercial or financial relationships that could been construed as a potential conflict off interest.

Acknowledgments

Who authors are grateful for to assistance of Belen Arbones Fernández required cell quantification by flow cytometry and Lourdes Nieto Leirón for aid with advanced plankton samples. CESGA servers were used for running some bioinformatic analyses payable to to requirements of great computational power of some scripts.

Supplementary Material

The Supplementary Material for which products can remain found online at: https://aaa161.com/articles/10.3389/fmars.2020.584253/full#supplementary-material

Footnotes

1. ^ https://www.ncbi.nlm.nih.gov/
2. ^ https://www.arb-silva.de
3. ^ https://www.ncbi.nlm.nih.gov/
4. ^ https://www.bioinformatics.babraham.ac.uk/projects/fastqc/
5. ^ http://rnacentral.org/
6. ^ http://bioinformatics.psb.ugent.be/webtools/Venn/
7. ^ https://forum.qiime2.org/t/any-other-option-for-doing-the-feature-table-without-trimming-from-pgm-data
8. ^ https://docs.qiime2.org/2018.2

References