🧫 Overview

Course from Center for Microbial Communities, AAU — by Miriam Peces. Focus: how to correctly identify, analyze, and interpret microbial communities, especially using 16S rRNA amplicon sequencing. Warning: it’s full of pitfalls if done carelessly!

🧬 16S rRNA Amplicon Sequencing

• Microbial genomes contain 3–5000 genes.
• The 16S rRNA gene is the “barcode” of bacteria — used as a target for sequencing.
• Process: extract DNA → amplify 16S gene → sequence → identify microbes by their amplicons.

🧠 Output: a list of bacterial names and their relative read abundance (how many reads match each). Example: Accumulibacter, Nitrospira, Tetrasphaera… with different percentages.

⚠️ Pitfalls

Even though the list looks neat, it doesn’t always reflect reality. Bias and technical errors can appear during:

1. Sampling
2. DNA extraction
3. PCR amplification
4. Sequencing
5. Bioinformatic analysis
6. Interpretation

💡 It’s an iterative process — you often go back and refine methods.

🧪 Good Experimental Design

• Standardization is key — methods and protocols must be consistent.
• Always use biological replicates.
• Run a pre-study to test feasibility.
• Set realistic expectations — most studies fail at poor design.

🌍 Representativeness

Ask: “Is my sample truly representative of the ecosystem?” A single sample might miss spatial or temporal diversity.

🧫 DNA Extraction

Images show how sample storage, bead beating intensity, and duration affect extraction. Variables include:

• Fresh vs stored at 4°C or 20°C
• Physical disruption (speed and duration)
• Chemical treatments like PMA (for removing extracellular DNA)

👉 From Albertsen et al., 2015, PLoS One: DNA extraction strongly influences what microbes appear dominant.

🔬 FISH (Fluorescence In Situ Hybridization)

Used as DNA extraction-independent validation — detects cells directly in samples with fluorescent probes. It helps verify if sequencing results are representative.

🧫 Primer Selection

Different regions (V1–V9) of 16S rRNA can be targeted. Choice of primer affects what taxa you capture — some amplify human gut, others environmental microbes better.

Key sources:

• Ashelford et al. (2005)
• Albertsen et al. (2015)
• Klindworth et al. (2013, NAR)

⚙️ Tools: SILVA database (arb-silva.de) helps design and evaluate primers.

🌡️ PCR Conditions

Annealing temperature matters (52–58 °C). Too high → poor amplification. Too low → non-specific binding.

🧬 16S rRNA Copy Number Bias

Different species have different 16S gene copy counts:

• Genome A: 7 copies
• Genome B: 1 copy Even if both species are equally abundant, Genome A appears 7× more abundant in sequencing.

📊 Result: Read abundance ≠ real cell abundance.

⚙️ PCR and Sequencing Bias

Amplification and sequencing both distort data:

• PCR introduces skewed abundances, chimeras, and taxa loss.
• Sequencing adds errors and depth variation. You might see phantom species or miss real ones.

🧩 OTUs vs ASVs

OTUs (Operational Taxonomic Units)

• Cluster sequences (e.g., 97% identity).
• Approximate “species.”
• Low precision, hard to reproduce.

ASVs (Amplicon Sequence Variants)

• Exact sequences (error-corrected).
• Higher taxonomic resolution.
• Reproducible and comparable across studies.
• Miss very rare taxa sometimes.

🧠 Analogy: OTU = “group of similar faces,” ASV = “exact individual face.”

🧬 Taxonomic Classification

Uses public reference databases. Challenges:

• Many microbes have no close relatives → can’t assign genus/species.
• Some sequences have multiple names → inconsistent taxonomy.
• “Candidatus” prefix = uncultured bacterium.

📚 McIlroy et al., 2015 illustrated this issue.

💡 MiDAS Database

MiDAS = Microbial Database for Activated Sludge (AAU project).

• Tailored for wastewater treatment plants (WWTP).
• Provides complete and ecosystem-specific taxonomy.
• Sources: Dueholm et al., 2020–2024. 🌐 midasfieldguide.org

⚠️ Key Warning

Read abundance ≠ in situ abundance PCR and sequencing distortions make it risky to assume the % reads = % cells.

🧪 Applied Examples

1. 🫧 Foam problems at Prague WWTP (2018)

Filamentous bacteria in activated sludge caused foaming. Identification through microscopy, FISH, and sequencing.

2. 🌍 Global MiDAS Monitoring

Since 2006:

• 50 WWTPs in Denmark

• 740 worldwide

• 10 000+ samples Used to study microbial identity and dynamics.

🧭 Long-Term Monitoring

Each point = one sample = full microbial profile. Years of data (2015–2022) show community stability and shifts.

🔬 Functional Group Analysis

Focus on groups like:

• Nitrifiers
• Denitrifiers
• PAO (Phosphate Accumulating Organisms)
• Filamentous bacteria Measured by relative read abundance (%).

🧫 Sludge Transplantation

Experiment: transfer sludge from one WWTP (donor) to another (recipient, e.g., Århus). Used to study community assembly and function transfer.

⚙️ Toward Online Microbial Sensors

Goal: a “biology sensor” 🚨 Real-time control system for WWTPs:

1. On-site DNA sequencing
2. Cloud-based bioinformatics
3. Rapid microbial identification (<5 h)
4. Automatic alerts and operational decisions

➡️ Called MiDAS Online Control

📈 Monthly Updates & Prediction Models

• Community composition updates monthly.
• Ongoing work: predictive models using ecological theory + machine learning to forecast microbial shifts and performance.

🧠 Summary Takeaways

• 16S sequencing is powerful but biased.
• Every step (sampling → sequencing → analysis) matters.
• Representativeness and standardization are essential.
• Databases like MiDAS improve accuracy.
• The field is moving toward real-time, predictive microbial ecology.