Comparison of Three DNA Extraction Kits for Assessment of Bacterial Diversity in Activated Sludge, Biofilm, and Anaerobic Digestate

Abstract

Direct DNA analysis is the most widely used approach for microorganism characterization in natural and built environments; therefore, reliable and effective methods of nucleic acid extraction for samples from particular types of environments are needed. In this study, we compared three commercial kits for metagenomic DNA extraction from three types of biomass: activated sludge, biofilm, and anaerobic digestate. The yield, purity, and quality of DNA were measured, and the effect that the DNA kit had on the subsequent microbial community analysis was assessed with amplified ribosomal intergenic spacer analysis (ARISA). Amplicons were analyzed automatically utilizing capillary electrophoresis. For the activated sludge and digestate, the suggested kit is FastDNA™ Spin Kit for Soil (MP Biomedicals). This kit allowed the highest DNA yield to be obtained and provided the highest biodiversity. For biofilm with a high content of extracellular polymeric substances, the FavorPrep™ Soil DNA Isolation Mini Kit (FAVORGEN) is recommended. This kit allowed to obtain the highest biodiversity and provided the most reliable results of genetic distance assessment in this type of biomass.

1. Introduction

Biomass, such as activated sludge, biofilm, or digestate from anaerobic digesters, is considered to be a potential source of many valuable chemicals. However, ultimate success in obtaining valuable substances depends not only on the type of biomass and its content but also on the microorganisms engaged in its conversion. Therefore, gaining knowledge about microorganisms in biomass would be helpful in steering the metabolic processes towards efficient synthesis of valuable chemicals. Usually, the processes leading to the synthesis of valuable chemicals are not conducted by single strains of microorganisms, but by specialized microbial communities. Identification and characterization of the members of these communities is important, not only to understand the processes, but also to make it possible to construct synthetic microbial communities in the future [1]. Unfortunately, most microorganisms cannot be cultivated in laboratory conditions, but can only be identified by direct DNA analysis. Usually, the first stage of this approach is the extraction of metagenomic DNA, the efficiency of which affects the success of the entire analysis [2]. To ensure that the metagenomic DNA that is obtained will be representative of most of the microorganisms that inhabit the analyzed sample, the method of DNA extraction must be properly chosen and executed.

The available methods should be examined for each habitat prior to beginning full-scale studies [3,4]. Most investigators use commercial kits for extracting and purifying nucleic acids for metagenomic analysis of samples from wastewater treatment plants. These DNA isolation kits have been successfully applied to study environmental samples, as well as samples from activated sludge and anaerobic digesters [5,6]. Importantly, however, the microbial communities in soils, activated sludge, and anaerobic digesters differ in terms of their bacterial community compositions [7]. Moreover, wastewater habitats are not only heterogeneous in nature, but they tend to change over time, which could affect the performance of DNA isolation [3]. Despite these facts, extraction kits which proved suitable for one habitat are oftentimes used for other habitats without further validation [4]. However, to determine the diversity of wastewater communities, extraction methods must be appropriately selected and optimized for each individual habitat.

DNA yield, purity, and bacterial diversity assessments vary depending on the DNA extraction methodology that is used [8,9,10,11]. The method of DNA extraction chosen for an experiment should yield the highest possible number of OTUs (operational taxonomic units), while not excluding the detection of any taxonomic groups [4]. Such a method should provide high yield and purity and generate the most complete description possible of a community’s diversity [4]. However, a difficulty lies in the fact that wastewater plants are designed to maximize nutrient recovery, which is accomplished by forcing communities to live closely together, for example, embedded in a self-produced matrix of extracellular polymeric substances (EPS). Because EPS contain considerable amounts of proteins, polysaccharides, lipids, and humic substances, which can strongly affect the purity and yield of extracted DNA, it can be difficult to obtain quality results with wastewater microorganisms. This is unfortunate because the purity of DNA is especially important for subsequent metagenomic procedures. Downstream procedures such as DNA amplification using polymerase chain reaction or preparation of DNA libraries are susceptible to contamination, which should be avoided or minimized [4].

ARISA has been widely used in metagenomic analysis and gives reproducible estimates of a bacterial community’s diversity [12]. ARISA provides estimates of microbial richness and diversity based on the length heterogeneity of the 16S–23S internal transcribed spacer region (RIS) of bacterial rRNA [13,14]. This region displays substantial variability in the length and nucleotide sequences between different microbial genotypes. Moreover, ARISA, which was originally designed to be performed with a genetic analyzer (GA-ARISA), could also give accurate and reproducible results faster and without trained personnel while using less specialized equipment, such as a microfluidic electrophoresis device (MF-ARISA) [15].

Here, we compared three commercial DNA extraction kits for DNA extraction from activated sludge, EPS-rich biofilm, and anaerobic digestate from an anaerobic digester. These three types of biomass have different properties, which is useful for highlighting the strengths and weaknesses of these DNA extraction kits. Each method was tested in triplicate to assess its extraction efficiency, the yield, purity, and integrity of the obtained DNA, and how accurately the method reflects the bacterial community composition and diversity. The results showed that most useful kit was the FastDNA Spin kit for Soil (MP Biomedicals, Solon, OH, USA), which was especially efficient for DNA extraction and purification from activated sludge and anaerobic digestate. For the biofilm with a high content of exopolysaccharides, the FavorPrep Soil DNA Isolation Mini Kit (Favorgen Biotech, Ping Tung, Taiwan) is recommended because it allowed us to obtain the highest biodiversity index and provided the most reliable results of genetic distance analysis.

2. Materials and Methods

2.1. Samples

For each kind of biomass, eight samples were analyzed. The samples of activated sludge were obtained from wastewater treatment plants located in Gdynia Dębogórze, Poland (samples S1–S5, S7 and S8) and Słupsk, Poland (S6). Samples from Gdynia Dębogórze were taken in different seasons of the year. Samples of biofilm (S11–S18) were taken from a sequencing batch biofilm reactor (Department of Ecological Engineering, University of Warmia and Mazury in Olsztyn, Olsztyn, Poland) and contained significant exopolysaccharides at an average concentration of 32% [16]. Samples S21–S28 were obtained from the anaerobic digestion chamber of a wastewater treatment plant located in Gliwice, Poland.

2.2. DNA Extraction Procedures

The following three DNA extraction kits were compared: a FastDNA Spin kit for Soil (MP Biomedicals), referred to as FASTD; a PureLink MicrobiomeTM Purification Kit (ThermoFisher Scientific, Waltham, MA, USA), referred to as PUREL; and a FavorPrep Soil DNA Isolation Mini Kit (Favorgen Biotech), referred to as FAVOR. All three extraction kits use bead beating, after which DNA is separated and purified on spin columns. For all methods, DNA extraction was conducted according to the manufacturer’s instructions. Bead beating was performed at maximum speed in a Uniequip device (Uniequip, Planegg, Germany) until the biomass was completely homogenous; for FASTD, it was 5 min., whereas for PUREL and FAVOR, it was 10 min. The wet weight of the biomass used for extraction was dependent on the extraction kit used: for FASTD and FAVOR, 0.25 g of each sample was used, but for PUREL, 0.2 g of each sample was used, according to the manufacturers’ instructions. To elute DNA, 100 µL of elution buffer or ddH₂O was used, depending on the manufacturers’ instructions. Extracted DNA was stored at −20 °C.

2.3. Determination of DNA Concentration, Purity, and Integrity

The concentration and purity of extracted DNA were determined spectrophotometrically with a NanoDrop 8000 (Thermo Fisher Scientific). Additionally, it was also quantified fluorometrically using a Qubit fluorometer with a dsDNA BR Assay kit (Invitrogen, Carlsbad, CA, USA). Because different amounts of source materials were used for extraction procedures, the DNA extraction efficiency was calculated as the total amount of DNA extracted per gram of wet biomass used for extraction (μg g _wet wt⁻¹). The purity of the extracted DNA was estimated by calculating the 260/280 ratio with a NanoDrop 8000 (ThermoFisher Scientific). Low absorption ratios at 260/280 nm (<1.7) were considered an indicator of protein impurities, and low absorption ratios at 260/230 nm (<2) were considered an indicator of contamination from humic acids and polysaccharides [4]. The integrity of DNA extracts was estimated based on electrophoresis in 1.2% agarose gels with the addition of Midori Green Advance DNA Stain (Nippon Genetics, Tokyo, Japan). Electrophoresis was conducted at 90 V in TBE buffer for 60 min.

2.4. Polymerase Chain Reaction

DNA fingerprinting of bacterial communities was undertaken by amplifying the 16S and 23S rRNA intergenic spacer region using previously validated primers (S-D-Bact-1522-β-S-20 and L-D-Bact-132-α-A-18) [17]. The PCR mixtures contained 15 ng of extracted DNA, 0.5 μM of each primer, 150 μM of deoxynucleoside triphosphate (Promega, Madison, WI, USA), 1.5 U of GoTaq Flexi DNA polymerase (Promega), 5 μL of reaction buffer (500 mM KCl, pH 8.5; Triton X-100), 1.5 mM MgCl₂, and water up to 30 μL of total volume. The PCR temperature program was as follows: 94 °C for 5 min, 35 cycles of denaturation at 94 °C for 45 s, annealing at 50 °C for 1 min, extension at 72 °C for 1 min, and a single final elongation at 72 °C for 10 min. The PCR result was checked by separation in 1.5% agarose gels stained with Midori Green Advance DNA (Nippon Genetics).

2.5. Microfluidic-Based Automated Ribosomal Intergenic Spacer Analysis (MF-ARISA)

The PCR mixture was loaded into chip wells that were prepared according to the manufacturer’s recommendations (DNA 1000 LabChip kit; Agilent Technologies, Santa Clara, CA, USA). Samples were analyzed with the use of Agilent 2100 Bioanalyzer (Agilent Technologies) and Agilent 2100 Expert software (Version B.02.10.SI76). The peak sizes and areas based on the data for internal size standards in each lane (15 and 1500 bp) were determined by Agilent 2100 Expert software (Version B.02.10.SI76). Peak sizes were compared for all samples; values within ±5% were assumed to be the same, according to the kit instructions, and binned together.

2.6. Bacterial Abundance Determination and Community Composition Analysis

The differences between different methods of DNA extraction and biomass were determined by calculating the genetic distance between samples. To create a binary matrix, presence–absence data were used. The pairwise similarity of the samples was estimated by calculating the Dice coefficient, Dc = 2j/(a + b), where j is the number of bands common to both samples, a is the number of bands in sample A, and b is the number of bands in sample B. The obtained data were analyzed and clustered by an unweighted pair group method using the arithmetic average (UPGMA) algorithm.

2.7. Statistical Analyses

UPGMA cluster analysis was conducted with DGGEstat software (Version 1.0 beta; Erik van Hannen, The Netherlands Institute of Ecology). The biodiversity of the microbial community was estimated using the Margalef index of diversity [18]. This index was calculated based on the presence of peaks. A Mann–Whitney U test was used to find significant differences between groups. A significance level of p ≤ 0.05 was used in all analyses. PCA was performed with PAST, version 4.03 [19].

3. Results

3.1. DNA Quantity, Concentration, and Quality Assessment

The molecular weight of the DNA varied dramatically between kits (Figure 1). The PUREL kit generated the highest weight of DNA. Mostly degraded DNA was observed when DNA was extracted using the FAVOR method, especially when samples of activated sludge and biofilm were processed. In the case of activated sludge, when the DNA concentration was investigated using the fluorescent approach, the highest mean value of DNA yield was noticed when the FASTD method of DNA extraction was used (0.84 ± 0.17 ng of DNA/1 mg of biomass) (Tables S1–S3). For comparison, the respective values were 0.41 (±0.02) and 0.11 (±0.02) when the PUREL and FAVOR methods were used. When the spectrophotometric approach was used, the mean values were 1.42 (±0.17), 0.54 (±0.03), and 2.34 (±0.66) for the FASTD, PUREL, and FAVOR methods, respectively. All differences between all methods of DNA extraction were statistically significant when using both the fluorescent and spectrophotometric approaches. The quality of purified DNA was the worst when FAVOR was used (260/280 = 2.09 ± 0.04); when FASTD and PUREL were used, this value was 1.91(±0.03, and ±0.04, respectively). This difference was statistically significant.

In the case of biofilm, when the fluorescent approach was used for DNA concentration analysis, the highest mean value of DNA yield was noticed when the FASTD method of DNA extraction was used (0.38 ± 0.15 ng of DNA/1 mg of biomass). With the PUREL and FAVOR methods, these values were 0.14 (±0.06) and 0.13 (±0.05), respectively. When the spectrophotometric approach was used, the mean values were 0.57 (±0.16), 0.16 (±0.06), and 5.37 (±4.49) for the FASTD, PUREL, and FAVOR methods, respectively. With the fluorescence measurements, significant differences were found between FASTD and PUREL, as well as between FASTD and FAVOR. With the spectrophotometric approach, all methods of DNA extraction differed significantly. The quality of DNA was best when the PUREL method was used for DNA purification (260/280 = 1.83 ± 0.07). With FASTD and FAVOR, these values were 1.98 (±0.05) and 2.12 (±0.08), respectively. All these differences were statistically significant.

When the fluorescent approach was used to measure the concentration of DNA extracted from digestate, the highest yield was noticed with FASTD (0.20 ± 0.07 ng of DNA/1 mg of biomass). With PUREL and FAVOR, these values were 0.07 (±0.01) and 0.05 (±0.01), respectively. The calculated DNA yield values when the spectrophotometric approach was used were as follows: 0.82 (±0.28), 0,16 (±0.04), and 0.43 (±0.17) for FASTD, PUREL, and FAVOR, respectively. All differences between all methods of DNA extraction were statistically significant with both the fluorescent and spectrophotometric approaches. The quality of DNA was the best when the PUREL method was used for DNA purification (260/280 = 1.81(±0.11). These values were 1.93 (±0.07) and 2.07 (±0.07) when FASTD and FAVOR were used, respectively. The differences between FASTD and FAVOR as well as between PUREL and FAVOR were statistically significant.

3.2. Automated Ribosomal Intergenic Spacer Analysis (ARISA) Profiles

The automated ribosomal intergenic spacer analysis showed that the FASTD kit generated the fluorograms with the highest number of peaks. Fewer numbers of peaks were provided by FAVOR and PUREL (Figures S1–S3). In the case of activated sludge samples, the number of peaks ranged from 4 to 23. The mean number of peaks was highest for FASTD (20.0 ± 2.45 peaks), followed by FAVOR (17.6 ± 1.99 peaks) and PUREL (12.6 ± 4.79 peaks). For the biofilm, the number of peaks ranged from 6 to 26. The mean number of peaks was highest with FAVOR (16.5 ± 3.42) peaks), followed by PUREL (13.0 ± 5.75 peaks) and FASTD (12.1 ± 5.11 peaks). In contrast, for the digestate, the number of peaks ranged from 17 to 30. The mean number of peaks was highest with FASTD (25.3 ± 4.06 peaks), followed by FAVOR (20.6 ± 1.50 peaks) and finally PUREL (19.5 ± 4.81 peaks). In general, looking at all samples together, the number of peaks was highest for FASTD (19.2 ± 6.74 peaks), then FAVOR (18.4 ± 2.96 peaks), and the lowest for PUREL (15.0 ± 5.95 peaks). The Mann–Whitney test found statistically significant differences in the number of peaks between FASTD and PUREL and between PUREL and FAVOR.

The genetic distance between samples shown in the form of dendrograms varied between the tested kits (Figures S4–S6). There were no statistically significant differences in genetic distance values obtained by particular pairs of DNA purification method and overall genetic distance for each type of sample. However, some pairs of DNA purification kits gave more similar results than others. Activated sludge samples analyzed by FASTD and PUREL were clustered together four times (samples S1, S5, S7, and S8), PUREL and FAVOR paired three times (samples S2–S4), whereas FASTD and FAVOR only once (sample S6). In the biofilm, PUREL and FAVOR were clustered together four times (samples S14 and S16–S18). FASTD and FAVOR were grouped three times (samples S12, S13 and S15), whereas FASTD and PUREL were only grouped once (sample S11). In digestate, the FASTD and PUREL methods were usually clustered together (samples S22–S26 and S28). Two times, PUREL and FAVOR were grouped together (samples S21 and S27). It was assumed that kits that are outgrouped more often are less reliable. For activated sludge, FAVOR was outgrouped 4 times; for biofilm, FASTD was out-grouped 4 times; and for digestate, FAVOR was outgrouped six times.

In the case of activated sludge, the highest mean value of the Margalef index was noticed for FASTD (3.37 ± 0.40), then for FAVOR (3.32 ± 0.43), then PUREL (2.30 ± 0.92) (Figure 2). The Mann–Whitney pairwise test showed a statistically significant difference between FASTD and PUREL (p = 0.01), as well as between PUREL and FAVOR (p = 0.013). In the biofilm, the highest mean value of the Margalef index was noticed in FAVOR (2.59 ± 0.60), then in PUREL (2.03 ± 1.04), and lowest in FASTD (1.91 ± 0.83). The observed differences were not statistically significant. In digestate, FASTD showed the highest mean value of the Margalef index (3.96 ± 0.61). The intermediate value was noticed in FAVOR (3.45 ± 0.28), whereas the lowest was in PUREL (3.18 ± 0.77). The observed differences were not statistically significant.

A principal component analysis (PCA) was performed to construct a global picture of the relations between DNA yield, DNA purity, and biodiversity expressed in values of the Margalef index (Figure 3). DNA yield and DNA purity explained 67.32% of total variation. The obtained results confirm a stronger effect of the kit used for DNA extraction and purification. There was a strong positive correlation between DNA yield and the FASTD kit. A positive correlation was evident between DNA purity and the PUREL kit; in contrary, the FAVOR kit did not provide pure DNA.

4. Discussion

The extraction and purification of DNA is a crucial stage in metagenomic analysis; however, kits dedicated for the types of biomass analyzed in this study are still not available. Because the number of kits for metagenomic DNA extraction is still increasing, making the right decision is a hard task. In this paper, we evaluated three commercially available kits to identify which is suitable for sampling activated sludge, biofilm, and anaerobic digestate. The obtained results showed that for the activated sludge and digestate, the most promising is the FASTD method. This method allowed the highest DNA yield to be obtained and provided the highest biodiversity. For the biofilm with a high content of EPS, the FAVOR is recommended. The DNA yields measured with fluorometry were always highest when the FASTD kit was used, but other parameters should also be taken under consideration when metagenomic DNA is purified. Additionally, the DNA integrity analysis performed by gel electrophoresis showed that FASTD provided more disintegrated DNA than PUREL. It is not a disadvantage if short DNA fragments are going to be analyzed in further steps. However, if longer DNA fragments are needed (e.g., for third-generation sequencing techniques such as those developed by Oxford Nanopore Technologies and Pacific Biosciences), alternative kits should be considered.

The most commonly used method for nucleic acid assessment is spectrophotometry, which relies on the absorbance of light by nucleic acids at different wavelengths. Alternatively, nucleic acids can be measured using fluorometry, which is shown to be a much more accurate and reproducible method than spectrophotometry, which is known to overestimate DNA or RNA quantity by manifolds [20,21]. In our study, the values of DNA concentration were significantly higher when the spectrophotometric method was used for DNA concentration analysis. The differences were more significant when the values of the A₂₆₀/A₂₈₀ ratio were higher than 2.0. The differences between the results obtained by spectrophotometry and fluorometry were not significant when DNA purity was higher, as in the case of the PUREL method. It confirms previous suggestions that the fluorometric method should be used for DNA concentration assessment when DNA extract is not ideally purified.

The best kit should show the highest alpha diversity. Thus, such a kit should generate the highest biodiversity estimates and minimize the bias of not extracting DNA from the specific taxonomic group [4]. Our results showed that the best kits for describing microbial communities are the FASTD and FAVOR kits. In activated sludge and digestate, the highest values were obtained using FASTD, whereas in biofilm with a high content of EPS, the FAVOR kit provided higher values of biodiversity. The advantage of the FASTD kit, in light of its high efficiency in DNA extraction, is not surprising, but the superiority of the FAVOR kit over the PUREL kit could surprise. The index of biodiversity in the biofilm with a high content of EPS was lowest for the FASTD kit, although the DNA yield was almost three times higher than in the PUREL and FAVOR methods. This could suggest that presence of EPS may be an obstacle in obtaining DNA from some taxa of bacteria when the FASTD kit is used.

In our work, we used microfluidic chip electrophoresis to obtain a picture of the biodiversity of analyzed samples. The study based on the analysis of the RIS showed the differences in biodiversity. The peak number, although limited by PCR bias, provided reliable information about biodiversity differences caused by DNA extraction using various kits. The main advantage of this method is its simplicity and fast execution time [15]. However, it is not enough to provide detailed information about the taxonomic characterization of the analyzed samples. Therefore, to obtain knowledge of which groups of bacteria are not represented in DNA extracted by a particular method, it would be better to analyze 16S rRNA amplicons sequences generated by the next-generation sequencing approach.

In summary, this study showed that the choice of the proper DNA extraction method is not trivial and should be preceded by an in-depth analysis. Generally, the method of choice that will provide enough DNA for the most environmental samples is FastDNA™ Spin Kit for Soil (MP Biomedicals). Additionally, further steps should also determine which method should be used. If longer DNA fragments are needed, the vigorous method that destroys long DNA fragments should be omitted. Finally, the method of DNA concentration analysis should also be performed with caution; if DNA extracts could have some impurities, the researchers should know that the results of DNA concentration analysis using the spectrophotometric method could provide false results.