Direct PCR matches standard metabarcoding for chironomid communities

Direct PCR can recover chironomid community composition without the conventional DNA-extraction step, according to a comparison of artificial and natural insect communities. The study found that homogenized tissue and purified DNA produced broadly comparable results in high-throughput DNA metabarcoding, with no significant effect of preparation method on the overall composition of natural chironomid communities. The findings suggest that a mechanical tissue-breakdown step can replace extraction for presence–absence monitoring of these abundant, species-rich aquatic insects, although direct PCR was less reliable at very low DNA inputs and generated greater variation in sequence-read abundance among specimens.

Chironomidae are well suited to testing alternative molecular workflows because they are superabundant and species-rich members of freshwater communities, while their morphological identification is often difficult. Researchers collected emerging adults from artificial ponds at the Eußerthal Ecosystem Research Station near Landau, Germany, in 2019 and 2020. Specimens were sorted into four size classes because individual chironomids in the study differed in mass by more than a factor of 20. For tissue preparation, the insects were dried at 60 °C for 16 to 24 hours, finely ground with a TissueLyser II at 30 hertz for three one-minute cycles, suspended in PCR-grade water, vortexed and frozen at −20 °C. This procedure served as the starting material for direct PCR, whereas the conventional workflow began with extraction and purification of DNA from an equivalent portion of each homogenate.

Both workflows used a two-step PCR process targeting a 421-base-pair fragment of the mitochondrial cytochrome c oxidase I gene. The first PCR either received 1 microliter of purified DNA or 5 microliters of tissue-water suspension. Products were diluted 1:20 before the second PCR, in which fusion primers added Illumina adapters and sample-specific inline barcodes. Libraries were purified, size-selected with magnetic beads and sequenced as 250-base-pair paired-end reads on a MiSeq system. Demultiplexed reads were merged, quality-filtered and clustered into operational taxonomic units, or OTUs, at a 3% similarity threshold. For one artificial community whose members had very similar COI sequences, the researchers instead used amplicon sequence variants and assigned them to individual specimens through their Sanger-sequenced barcodes. Negative controls and duplicate technical PCRs were used to identify low-level contamination and inconsistent detections.

The artificial communities provided a controlled test of whether tissue mass and preparation method altered species detection. Mock A contained 16 chironomid specimens selected to differ by at least 1.5% in the targeted COI region, and five replicates were prepared under each principal workflow. One set of replicates used the same tissue-water volume for every specimen, simulating the variable masses expected without size-sorting; another used specimen-specific volumes adjusted to provide approximately 0.006 milligrams of tissue per individual. A second mock community, Mock B, contained 13 specimens with at least 13% sequence dissimilarity and provided an additional comparison. Across methods, 13 of the 16 Mock A specimens and all Mock B specimens were reliably detected in every replicate. Three Mock A taxa were detected in 95% to 97.5% of samples because one technical replicate in a small number of samples contained no reads for them.

Equalizing the amount of material supplied to the PCR reaction improved detection of the smallest specimens, but it did not make sequence-read abundance a reliable measure of individual biomass. Read counts varied more among taxa within a mock community than they did between the extracted-DNA and direct-PCR approaches, with Hedges’ g values greater than 0.8 in the comparisons. The observed heterogeneity also persisted after specimen mass was adjusted. In a separate experiment, researchers created four mock communities from 13 specimens using either equal amounts of purified total DNA or equal amounts of amplified target fragments. Equal purified-DNA inputs produced significantly different read abundances among OTUs, whereas equalizing the target fragments yielded no significant differences between specimens. Deviations from mean read abundance were significantly greater with equal total-DNA inputs, and varying the total-DNA input had essentially no linear relationship with read counts. By contrast, increasing the amount of target fragment in the second PCR explained 74% to 94% of read-abundance variation in the four mock communities.

These results indicate that much of the among-specimen variation arose during amplification rather than directly from specimen mass or the quantity of total DNA introduced into the first PCR. The authors identified primer bias as a probable contributor, although the study did not directly measure it, and noted that variation in mitochondrial copy number, heteroplasmy and nuclear pseudogenes could also influence read abundance. Their results therefore support interpreting chironomid metabarcoding primarily through presence–absence data rather than assuming that sequencing reads directly represent tissue mass or biomass proportions. A separate test of sensitivity also exposed a clear limit of the extraction-free workflow. Detection declined in every approach as the starting material was diluted, but the decline was substantially faster for direct PCR, whose amplification bands weakened more rapidly on agarose gels. OTUs that began with fewer than 50 reads disappeared in at least one dilution, whereas those with more than 700 reads were detected almost consistently across the tested range.

Natural communities were tested using 40 samples collected across five dates within 17 days from eight artificial ponds, four treated with the mosquito-control agent Bacillus thuringiensis israelensis and four left as controls. The samples came from an ongoing ecotoxicological study, and every community was processed by both direct PCR and the conventional extraction-based method. Sequencing identified 114 chironomid OTUs representing 36 genera. The methods differed for an average of 8.3% of the OTUs in pairwise sample comparisons, although six OTUs deviated between methods in more than four samples. Those frequently discrepant OTUs had relatively low mean abundance—no more than 10 reads—compared with an overall mean of 447 reads. A correspondence analysis showed substantial overlap between the workflows, and a permutational multivariate analysis of variance found significant effects of Bti treatment and sampling date, but no significant effect of DNA-preparation method on presence–absence community composition. In other words, both workflows supported the same conclusions about the ecological response to the treatment.

The practical implication is that DNA extraction can be omitted in at least some chironomid community metabarcoding without sacrificing the presence–absence information needed to characterize biodiversity or compare environmental treatments. Direct PCR is attractive for monitoring programs because extraction can be costly, labour-intensive and difficult to automate. The approach used here required drying, mechanical homogenization and suspension in water before amplification, rather than placing intact tissue directly into a PCR reaction. Modern polymerases and buffers also tolerate inhibitors more effectively than earlier formulations, which may further support the method. Eliminating extraction could reduce sample-processing time and costs and may fit workflows that aim to miniaturize molecular analysis for field or on-site use. However, the study did not directly measure a formal cost comparison, so the expected economic benefit remains a proposed advantage rather than a quantified result.

Important limits constrain how broadly the findings can be applied. The study focused on chironomids and used a specific homogenization, dilution and PCR protocol, so performance may differ among arthropod groups with tougher tissues, pigmentation, secretions or other PCR inhibitors. Earlier direct-PCR work had reported lower success for some beetles, copepods, ants and dragonflies. Although the chironomid protocol produced successful COI amplification for specimens spanning several arthropod orders in supporting tests, the authors emphasized that mixed arthropod samples would require representative mock communities to reveal taxon-specific bias. Direct PCR was also less robust than purified DNA when very small amounts of material were pipetted, and its read counts were more variable. The evidence therefore supports direct PCR as a valid option for chironomid presence–absence monitoring, particularly when speed and simpler sample preparation matter, rather than as a universal replacement for extraction or as a direct measure of organism abundance.

Overall, the experiments show that a carefully prepared tissue suspension can substitute for purified DNA in chironomid community metabarcoding. Artificial communities demonstrated similar detection rates, while natural pond samples showed high agreement in species composition and identical treatment-related conclusions under the two workflows. The approach cannot yet match extraction in sensitivity to diluted material, nor can its sequence reads be treated as exact proxies for biomass, but it preserves the principal compositional signal needed for ecological monitoring. For researchers working with chironomid communities, the study offers a practical alternative that removes one of the most demanding stages of metabarcoding and could make large-scale biodiversity surveys faster and less expensive, while still requiring validation when transferred to other insect assemblages or lower-input samples.

Subject of Research: Direct PCR for high-throughput metabarcoding of chironomid insect communities

Article Title: Direct PCR meets high-throughput sequencing – metabarcoding of chironomid communities without DNA extraction

Article References: Röder, N., & Schwenk, K. (2023). Direct PCR meets high-throughput sequencing – metabarcoding of chironomid communities without DNA extraction. Metabarcoding and Metagenomics, 7, Article e102455. https://doi.org/10.3897/mbmg.7.102455

Image Credits: AI Generated

DOI: 10.3897/mbmg.7.102455

Keywords: direct PCR, DNA metabarcoding, chironomids, biodiversity monitoring, high-throughput sequencing, environmental DNA, COI barcoding, ecotoxicology, arthropod communities, Direct, meets, high-throughput