Weak toxicity of chemical additives extracted from car tire rubber on the microbiome of Atlantic cod and Norway pout
Sverre Hjelset
Thesis submitted for the degree of Master of Science in Toxicology and Environmental Science
60 Credits
Department of Biosciences
Faculty of Mathematics and Natural Sciences UNIVERSITY OF OSLO
SUPERVISORS:
Alexander Eiler (UIO) Ketil Hylland (UIO) Tânia Gomes(NIVA)
Abstract
Plastics are now ubiquitous in the marine environment releasing diverse chemical additives that are absorbed to them. When marine animals are exposed to these chemicals this may lead to potential adverse responses in the animals, including alterations to their associated microbiome. It is well documented that the microbiome influences host health and that toxin-induced changes in the microbiome can be associated with decreased lifespan for the affected individual. We exposed cultured microbiomes from two marine fish species, Atlantic cod (Gadus morhua) and Norway pout (Trisopterus esmarkii) sampled from two locations: inner and outer Oslo fjord, to extracts derived from car tire rubber (MPS). An in vitro model of microbiome
enrichments was chosen to explore the effects of extracts as it permitted the study of microbiome-specific responses in a simple, convenient and high-throughput fashion.
Functional responses of the microbiome were assessed by determining growth
characteristics such as lag phase and maximum growth rate in batch incubations, while diversity responses were determined by metabarcoding of the 16s rRNA gene. The experiments revealed weak effects of the extracts on functional responses of the enriched gut microbiome, as expressed by an increase in lag-phase and decrease in maximum growth rate in few biological replicates. Similarly, metabarcoding revealed minor responses in alpha and beta diversity of the microbiome following exposure to the extracts when compared to control treatments and other factors such as differences in association with fish species and location. Still, three bacterial genera were affected already at extract concentrations around 0.01 mg L-1. Overall, our results indicate toxicity and hazardous effects from car tire extracts on in vitro models of fish microbiomes from two gadid species.
Acknowledgment
This master thesis was a collaboration between the Department of Biosciences at the University of Oslo (UiO) and Norwegian institute for water research (NIVA). Main
supervisor was Alexander Eiler (UIO) with co supervisor Ketil Hyland (UIO) and external supervisor Tânia Gomes (NIVA). I feel very well taken care of by these supportive people and this project would be impossible without the valuable instructions from them.
Especially thanks to Alex with all the help and instructions I have received when working in the lab and crunching statistics, your experience has been invaluable to me. I would also like to thank Dirk Linke and his lab team, for being so helpful when I had questions and letting me use their lab and equipment for our research. It has been inspiring to be part of the lab team on Aqua and socially it has been a blast working with some of the PHD students in our department, providing both humour and intellectual discussions along the way. I would also like to thank the MicroLEACH project for providing the car tire extract, and in particular Amaia Igartua for preparing it. Lastly, I would like to give a shoutout to my beloved girlfriend, friends and family for supporting me.
Table of Contents
1 Introduction 5
2 Aims of the study 8
3 Methods and Materials 9
3.1 Sampling 9
3.2 Plastic extraction 10
3.3 Microbiome enrichment 11
3.4 Exposure experiments 13
3.5 Growth rate analysis 14
3.6 DNA extraction, amplicon library preparation and sequencing 14
3.7 Processing of 16s rRNA gene amplicon sequences 15
3.8 Statistical analysis 16
4 Results 17
4.1 Growth Results 17
4.2 Sequencing results 20
5 Discussion 30
6 Conclusion 33
References 34
APPENDIX 39
1 Introduction
Since the early 1950s, plastic production has increased nearly exponentially. This is due to plastic replacing earlier traditional materials because of being cheap and easy to manufacture. It is estimated that globally over 350 million tons of plastic are produced each year, with an expected annual increase in production by about 10%(Andrady et al., 2011). The increased production and improper waste management leads to plastic getting discharged into every known ecological compartment. It is estimated that 8 million pieces of plastic find their way into the ocean every day(Koelmans et al., 2015).
We can now find plastics on every beach in the world. Multiple studies have even found plastics frozen in the arctic ice, where these particles can stay for decades or centuries since they biodegrade slowly.
Plastics and in turn microplastics (MPs) are either produced or are the byproduct of large plastic pieces degrading in the environment. Plastic litter breaks down into smaller pieces slowly, while suspended in the water column. This happens when plastics are exposed to sunlight (UV) and higher temperatures (Geyer et al.,2017,Zalasiewicz et al., 2016). While there are challenges to making a clear identifiable list of plastics, many agree on labeling them by measuring their size: Macroplastics = (<1mm), Meso-plastic = (2.5-5.5mm), Microplastic = (5mm-1µm) and nanoplastics = (<1µm). Apart from that, it is important to know that there is no clear linear way how plastics degrade, because of the dynamic nature of breaking down into un uniform smaller pieces. Plastics usually contain additives incorporated during production, which define the property of the plastics such as: Flame retardants, plasticizers, antioxidants, UV stabilizers and
pigments. MPs can also act as surfaces absorbing chemicals already present in aquatic and terrestrial environments.
There is a lot of documentation on the uptake of MPs in a large variety of marine
organisms (Desforges et al., 2015,Lusher et al., 2015; Phamet al., 2015,), including wild
and agricultural marine animals. Ingested plastics accumulate in the digestive tract (Wright et al., 2013)leading to blockades and abrasions where they further compete with food particles(Kühn et al,. 2015).In addition, MPs can alter gut tissue(Peda et al.
2016)even leading to physical and mechanical damages to the tissue (Jovanovićet al., 2017). There is also a growing number of studies supporting the hypothesis that MPs and their additives (i.e., plasticizers like bisphenol and phthalates) induce
gastrointestinal dysbiosis in aquatic animals, including fish. However, the role of additives leaching from the MPs in exerting physical irritation and inflammation to the gut, as well as inducing changes in the gut microbiome is not well documented. But there is evidence that chemicals adsorbed to plastics may be released when in contact with digestive fluids and then possibly interact with the host tissue and its microbiome (Teuten et al., 2009).
In the digestive system, plastic particles will come in contact with the host microbiome.
These microbiomes play an important role in all animals’ health through their function in digestion, lipid metabolism, vitamin synthesis, inflammation, and immune function (Turnbaugh et al., 2009:Lu et al., 2014). However, this host-microbiome system, also referred to as holobiont, is very sensitive to external hazardous stressors, likely including MPs and their additives, which may result in compositional and functional changes, leading to poor health and increased susceptibility to diseases(Hu et al., 2021).
A recent study demonstrated that DEHP (Di(2-ethylhexyl)phthalate) targets both the microbiome and intestinal transcriptome of male and female zebrafish (Adamovsky et al., 2020, Buerger et al., 2020). Similar to the phthalate example, the microbiome of fish was shown to be sensitive to perturbations by BPA (Bisphenol A,
2,2-(4,4′-dihydroxydiphenyl) propane) exposure (Javurek et al., 2016).
Whileexposure studies using single additives have provided an initial understanding of microbiota disturbances in fishes by plastics and plasticizers, the effect of
multiple additives as found in most MPs has not been explored. For example, the chemical makeup of car tire rubber has been previously assessed and some of the chemicals that are commonly found are: polypropylene, polyethylene terephthalate, polystyrene, polyvinyl chloride, benzothiazole, phthalide, acetophenone, cobalt, zinc, lead and antimony (Capolupoet al, 2020).Another study done by Akvaplan-niva found
evidence for additives such as benzothiazole, N-1,3-dimethylbutyl-N0
-phenyl-p-phenylenediamine, phenolic compounds (e.g., bisphenols) and a range of polycyclic aromatic hydrocarbons (PAHs) such as pyrene being most abundant
(Halsband et al., 2020).Thus, we decided to use phenanthrene as a positive control for our experiment, and since it is a PAH we expect to have an effect on in vitro microbiome models (Loibner et al., 2004).
Given the difficulty to access the microbiome of the gut for in vivo studies, in vitro models have been proposed as timely and cost-efficient solutions to study microbiome
responses to drugs and toxic compounds. For this purpose, in vitro models that maintain the functional and compositional profiles of in vivo gut microbiomes are extremely valuable (Venema et al., 2013). Examples range from simple batch fermentations to complex multicompartmental continuous systems including the incorporation of host cells (i.e. Bahrami et al., 2011). In this study we will use the simplest and most frequently used in vitro model, enrichment cultures of the gut microbiome, in static anaerobic batch cultures. These systems operate far from in situ conditions, since they suffer for example from accumulation of microbial metabolites potentially inhibiting microbial activity. In fact, due to changes in pH, redox potential and community structure during incubations, these systems have been shown to be
suboptimal in providing accurate results (Gibson and Fuller 2000). However, inter-individual variations in microbiota composition (Weirdt et al 2010) and the high-throughput of such batch incubations make them valuable systems for initial screening assays.
2 Aims of the study
This study aims to increase the knowledge of how chemical additives from car tire rubber affect the microbiome of two species of gadidae. We hypothesize that the chemical additives found in car tire rubber will affect the growth characteristics and biodiversity of the microbiomes. Chopped up car tires were chosen since they are highly abundant in Norway due to usage in artificial grass football fields and it is estimated that these plastics are contaminating many ecological compartments (Herzke et al 2019.
Bråte et al 2018)
The main aim can be divided as follows:
● Clarify whether exposure to rubber extracts affect growth characteristics such as length of lag phase and maximum growth rate of microbiomes.
● Clarify whether exposure to rubber extracts can potentially affect microbiome biodiversity in Atlantic cod (Gadus morhua) and Norway pout (Trisopterus esmarkii)
● Clarify differences in sensitivity to the exposure between microbiomes from two fish species sampled at two different locations with different pollution levels.
3 Methods and Materials
3.1 Sampling
Sampling was performed in the inner (59.86 N 10.62 E) and outer (59.41 N-10.51 E) Oslofjord (Figure 1) in early August 2020 using the research vessel Trygve Braarud. Fish were captured by bottom trawl, and then quickly killed by a blow to the head. Next, gut samples were collected onboard the boat by opening the fish with a scalpel. A section of approximately 3 cm of the intestine was extracted, and its content was squeezed out into a cryotube. For the smaller fish, the entire intestine was extracted, frozen and ground into a paste, using a pestle and mortar.Samples were labelled and immediately frozen in liquid nitrogen. Back at the lab, samples were stored at -80ºC until the initialization of the enrichment experiment.
A total of 12 samples were taken for Atlantic cod (Gadus morhua) and Norway pout (Trisopterus esmarkii) at each location.
Figure 1: Map of the Oslofjord with the two sampling locations (marked by red rings) Midtmeie (inner) and Hortenskrakken (outer).
3.2 Plastic extraction
The car tire rubber granulates used were obtained commercially from RagnSells and processed under the Project MicroLEACH (Grant 295174) funded by the Norwegian Research Council. These granulates consist of "medium Infill" of recycled tire rubber in the size range 1.0 – 2.8 mm, commonly used in soccer pitches (Capolupo et al., 2020;
Halsband et al., 2020).
Extract preparation was performed according to Zimmerman et al. (2019, 2020). Briefly, 3 grams of car tire granulates were placed in amber glass vials and 20 mL of methanol were added (LC-MS grade, Sigma- Aldrich). Samples were then extracted by sonication in an ultrasound bath for 1 h at room temperature. The methanol was transferred into clean glass vials, and 500 μL of the methanol extracts were retained for chemical analysis. Subsequently, 200 μL of dimethyl sulfoxide (DMSO, Merck) was added and
samples were evaporated under a gentle stream of nitrogen to a final volume of 200 μL.
Extracted samples were stored at −20 °C prior to screening analysis. Procedural blanks consisting of only methanol and DMSO blanks were treated identically as the samples to control potential contamination. To uniform the expression of the toxicity data, the concentration of car tire granulates present in the extracts was translated into “plastic equivalents”, in which “1 mg plastic” implies the toxicity extracted from 1 mg of plastic material (Zimmerman et al., 2019). Accordingly, the plastic concentration present in the extracted sample corresponds to 151 mg/mL of plastic.
3.3 Microbiome enrichment
A custom-made Yeast Casitone Fatty Acid (YCFA) media was chosen as growth media as it supposedly enriches the most diverse microbiome from gut samples. The media was prepared as given in the original manual(See appendix)with the exception that we did not gas the media withN2gas while cooling in step 4. Since inoculation was performed in a glovebox under a nitrogen atmosphere, we presumed that filter sterilization and
purging of the media in the glovebox would suffice in draining most accumulated O2
present in the media.
Gut samples from three individuals from the same species and same location were pooled to yield higher diversity and less variability between individual enrichments than when using individual fish guts. We also attempted to pool gut samples from fish of similar size and weight. Thus, we ended up with four samples per fish species per location for the experiment, resulting in a total of 16 samples. Samples were stored at -20ºC until further processing.
After thawing the gut samples at room temperature, they were quickly inoculated into the YCFA media inside an anaerobic chamber (888-SIMPLICITY PLAS LABS, INC). The incubation was conducted in 96-well microliter plates (Microtest plate 96 well, F non-pyrogenic, non-cytotoxic) with 1 µL of sample as inoculate and 199 µL of media.
Three replicates per gut sample were cultivated. Once inoculated inside the glove-box, microplates were sealed with clear adhesive film to prevent entry of oxygen when continuing the experiment in a spectrophotometer outside the glove-box. The samples
were kept at all times in an anaerobic state and growth rates were monitored on a
spectrometer with wavelengths 630 nm every 20-30 minutes. This first enrichment grew for 75 hours before reaching a stationary phase. Triplicates were pooled, and new
triplicates were created, as previously described, by inoculating 199 µL YCFA media with 1 µL of the original enrichment. This second batch ran for 24 hours before reaching a stationary phase. Again, triplicates were pooled and transferred to a third adaptation round. After the third batch incubation, the enrichments were frozen at -20ºC until the exposure experiments were performed. For a visual summary see figure 2.
Figure 2: Diagram of experimental design. Illustrating fish gut extraction, pooling of samples, preparation of enrichments, DNA extraction, primer attachment, DNA metabarcoding, Growth curves and characterisation of the microbial
communities.
3.4 Exposure experiments
All bioassays were conducted in 96- well microliter plates (see figure 3 for an
illustration, the plates have the same specs as in the enrichment process), and 8 different chemical concentrations were selected ranging from 0.472mg/ml – 0.004mg/ml halving the concentration of the extracts each step. All samples except blanks contained 0.1%
DMSO. We also included negative controls, DMSO and procedural blanks and positive controls with phenanthrene (1% final concentration) as the toxicant. After inoculation, the enrichments were incubated in the spectrophotometer as previously mentioned and cell density was measured at >600Nm in 20-30 minute intervals. It should be mentioned that absorbance measures turbidity (or optical density - OD) and thus only approximates cell density. Measurements are influenced by a variety of factors including species and media composition. The experiment lasted for at least 24 hours.
Figure 3, Experimental setup illustrating how one of the two plates ran in this experiment was set up. Note that each fish represents a pool of 3 unique individual fish. The second plate contained the rest of the concentrations as well as replicates of 3 doses from the first plate (0.236 mg/ml, 0.059mg/ml and 0.0015mg/ml).
3.5 Growth rate analysis
Resulting OD curves were analysed with R package ‘growthrates’ in R version 3.6.2.
The packages allows for analyses of data using three methods:
● Nonlinear fitting of parametric growth models
● Fitting of linear models to the period of exponential growth
● Nonparametric growth rate estimation by using smoothers
The package contains methods to fit single data sets or complete series of data sets organized in a data frame. More information about the R growth rates packages can be found here byThomas Petzoldt 2020.
3.6 DNA extraction, amplicon library preparation and sequencing
Metabarcoding of the 16s rRNA genes was used, spanning the variable regions V3 and V4. This is a powerful method to investigate biodiversity in vast biomes such as marine and terrestrial ecosystems or small host associated systems within a gut of an animal (Taberlet et al., 2018;Zinger et al., 2020).When conducting an experiment involving metabarcoding there is usually a three-step process. (1) Sample collection and
preservation; (2) DNA extraction and amplification; and (3) DNA sequencing (Tsuji et al., 2019).At the end of the experiment, the 96-well plates were centrifuged at 12000 RPM for 5 minutes to create a “cell pellet”. The supernatant was removed and a lysis buffer was added to each well before the plates were frozen at -20C. DNA extraction was carried out as described in the Protocol:Pre-treatment for Gram Negative Bacteria and”
Purification of Total DNA from Animal Tissues Spin- Column Protocol” of the DNeasyR Blood & Tissue kit. DNA Samples were amplified using the Bacteria primers 341F (Herlemann et al., 2011)and 805RN(Apprill et al., 2015).
Primer sequences were as follows:
Illumina adapter-N4-341F:
5’-ACACTCTTTCCCTACACGACGCTCTTCCGATCTNNNNCCTACGGGNGGCWGCAG-3’
Illumina adapter-805NR:
5’-AGACGTGTGCTCTTCCGATCTGACTACNVGGGTATCTAATCC-3’
PCR reactions were conducted in a Mastercycler epgradient S(Eppendorf AG 22331 Hamburg) set to the conditions described in the protocol. The PCR products were
checked by gel electrophoresis to determine if the PCR was successful. Then the samples were purified usingAgencourt AMPure XP PCR purification beads. Standard illumina sequencing primers and unique index tags were attached to each sample through a second PCR (each sample receiving a unique combination). Then we performed another cleanup withAgencourt AMPure XP PCRcr purification beads.DNA was quantified using Quant-iT™ PicoGreen® dsDNA Reagent (Thermo Fisher Scientific) and a Microplate reader (SYNERGYMx BioTek). Samples were pooled and sent off to the Norwegian Sequencing Center (Ullevåll sykehus, Oslo, Norway) for sequencing on a MiSeq in PE300 v3 mode and with 20% phiX.(For more information about library prep, see the protocol here or see “PCR reactions” in the appendix).
3.7 Processing of 16s rRNA gene amplicon sequences
After primers were removed from the raw sequences using cutadapt (version 1.6), the sequence data was processed using the R package dada2 (Callahan et al., 2016, version 1.8) for de-replication, denoising and sequence pair merging. After manual inspection of quality score plots, the forward and reverse reads of the bacterial 16S rRNA gene
amplicons were trimmed at 255 and 215 bp length, respectively. Additional quality filtering removed any reads with a single paired score below 20. After de-replication of reads, forward and reverse error models were created in dada2 with a subset of the sequences (~ 109nucleotides). Chimeras were removed using the
‘removeBiomeraDenova’-function provided in the R package ‘dada2’. Taxonomy was assigned using the Bayesian classifier and SILVA non-redundant database 138 (Quast et al., 2013; Yilmaz et al., 2014). This resulted in 180 unique amplicon sequence variants (ASVs) comprising a total of 9.9 million paired end reads. The average number of reads per sample was 51486 with a range from 11 to 703257. Non-bacterial ASVs (including chloroplasts) were removed resulting in 180ASVs. Next alpha diversity was estimated
using the abundance-based coverage estimator (ACE) and Simpson’s index, after
samples were rarefied to 3000 reads. Raw sequence reads were deposited to NCBI-SRA under accession number PRJNA749010.
3.8 Statistical analysis
Statistical analyses were performed in R version 3.6.2. Variables, i.e. growth
characteristics, diversity indices and concentrations were tested for normality using the Shapiro-Wilks test. Due to severe positive skewness, data was transformed: ACE was inverted while growth characteristics were square root and concentrations log +1 transformed. Multiple comparisons were performed using multi-factor analysis of variance (ANOVA) and generalized linear mixed models (GLMMs). The Holm correction was used to account for multiple testing. Beta-diversity was analysed using an
“altGower” dissimilarity matrix produced by standardizing the rarefied ASV with Hellinger transformation. The distance matrix was visualized using
non-metric-multi-dimensional-scaling (NMDS) and permutational multivariate analysis of variance (PERMANOVA, implemented as adonis in R) on the altGower matrix was used to test if centroids of microbial community structure differed between groups of the multifactorial variables (location, fish species, concentration, enrichment).
Both ANOVA and PERMANOVA were used for pair-wise comparison of treatments with various controls (Procedural blank, phenanthrene amendment, media blank without DMSO-solvent). Cohen’s D test as implemented in the R package “lsr” was used to estimate effect size (partial eta-squared). Significant fold-changes of individual ASVs were estimated using Deseq2 implemented in R. Deseq2 estimates variance-mean dependence in count data from high-throughput sequencing assays without the need of standardization and rarefying. Deseq2 tests for differential abundance between
treatments using the negative binomial distribution. We used “Wald’s test” and
“poscounts” for the type of size factor estimation. The “pheatmap” package was used to visualize the Deseq2 results.
4 Results
4.1 Growth Results
Resulting growth curves were obtained from 192 individual batch cultures and are shown in the supplementary data. Replicates were highly similar as indicated by the overlapping growth curves, still few exceptions were observed suggesting variability across the technical replicates. There was a significant shift in lag phase between replicates derived from individual plates. Since 78 of the samples did not achieve stationary phase in this experiment, we did not assess growth yields. A potential for cross contamination among biological replicates is indicated by the observation that out of all of the six blanks, exponential growth occurred in one.
Maximum growth rate and lag phase were assessed using the growthrates package in R.
Using the fit easy linear method, we estimated the maximum growth rates and lag phases in individual wells. We also used the fit splines function to estimate maximum growth rates providing nonparametric splines. Maximum growth rates as estimated by linear model fitting varied between 0.01 and 0.29 (mean of 0.14) across the two species and two locations while fitted splines estimated maximum growth rates ranging from 0.01 to 0.744 (mean 0.16). The lag phase as estimated by the linear model fitting was on average 12.6 hours and ranged from 0 to 22.1 hours. Considering only wells with
extracts and DMSO additions, maximum growth rates were on average 0.15 and 0.17 as estimated by the linear and splines fitting, respectively.
After correcting for multiple testing, we found no significant responses in either maximum growth rate and lag phase using linear model fitting in response to the concentration gradient in rubber extracts. However, there was a significant response in the maximum growth rate of the entire community in association with the extract gradient using the spline fitting results (p value 0.017 and F value 5.831). This
significant overall decrease in maximum growth rates of the community with increasing leachate concentrations was highly noisy, likely due to the high variation among
replicated enrichments (figure 4). Cohen’s D further revealed an effect size of 0.024 and thus extracts having a minor effect on maximum growth rates. This can also be
compared with the Cohen’s D of replicates’ effect size (d = 0.065) and interactive term location:replicates (d = 0.35). A generalized mixed model (glmm) corroborated that there is a minor significant effect of leachates on the maximum growth rates from splines fitting (z value = -2.478, p = 0.0132) .
To remove some of the variability in the data we also analysed the data separately for each fish species. The results confirmed the significant responses in growth
characteristics due to plastic leachate additions as the cultured microbiomes from the Norwegian cod exhibited a significant response in maximum growth rates from splines fitting (glmms: z-value = -2.322, p < 0.02, p < ; ancova: p < 0.01, Cohen’s D = ) while a significant response in the lag phase was observed in pout (glmms: z-value = -2.099, p <
0.036, p < ; ancova: p < 0.01, Cohen’s d = 0.061). As such, similar to the entire dataset the statistical evaluation estimates only minor responses in overall community growth characteristics of the in vitro microbiomes.
Figure 4 shows a weak response when looking at the mean ranges. (1 = highest concentration, while 8 = lowest
concentration, F= Positive control using Phenanthrene, P = Procedural blank, M = media sample blank with no DMSO, D=
DMSO blank)
We did find a significant difference in the length of the lag phase between the fish species (F values = 49.8, p < 0.001) and locations (F value= 20.3, p < 0.001). as well as between the biological replicates (individual microbiome enrichments; F value = 6.43, p < 0.002)-.
When analyzing the fish species separately the difference among the two locations in the in vitro microbiome composition could only be confirmed for the pout (F value = 39.7, p
< 0.001). Across controls, there were no significant responses in the growth
characteristics of the microbial communities such as lag phase and maximum growth rates (including the positive controls and negative controls) such as procedural blank and DMSO control when compared with microbiomes grown in unmodified YCFA media.
This was further confirmed by effect size as estimated by Cohen’s D showing that
location and fish species had much larger effects on growth characteristics than the leachate or other control additives.
4.2 Sequencing results
In total 192 samples were submitted to the sequencing facility, 13 samples had low yield, most likely due to low DNA template concentration. The sequencing resulted in almost 9.9 million paired reads across all samples. Filtering and denoising by dada2 removed 38% of the total reads. Merging forward and reversed reads resulted in 6,143,055 sequences, of which 589,357 were deemed to be chimeric sequences. The remaining 5.5 million reads were distributed among the 192 samples, with a mean of 28926 and a range from 2 to 436137 sequences per sample. Rarefaction curves indicated that at 3000 reads most of the diversity in the samples seemed to be covered (see figure 5). Thus samples with less than 3000 sequences were removed prior to further analysis This resulted in 170 total samples (of the 192) with 180 unique ASVs.
Figure 5 Rarefaction curves of the sequencing results showing that if we cut the number of sequences per sample at 3000 sequences, we will capture most of the ASV diversity.
Most taxa belonged to: Bacilli (50833 sequences) and Gammaproteobacteria (1579).
Other taxa observed were Actinobacteria, Cyanobacteria, Clostridia, Oligophrenia, Thermleophillia, Negativicutes, Alphaproteobactera, Planctomycetes, Bacteroidia and Saccharimonadia. At the genus level, the main taxa were Bacillus, Lactococcus,
Paenibacillus, Lysinibacillus and several more at lower proportions (figure 6). Bacillus was the most dominant, comprising multiple ASVs, in 69 samples while Lactococcus, the second most abundant taxon, dominated 50 samples. Visual inspections and Deseq2 analysis of the taxonomic patterns at the genus level did not reveal any responses of individual genera to the extracts at any concentration while three genera (Paenibacillus, Lactococcus and Psychrobacter) exhibited significant (padj < 0.05) differential
abundances in the enrichments derived from the two cod species and the genus Lactococcus in the enrichment from the two locations.
Figure 6 Proportion of the 15 most abundant genera including Bacillus, Lactococcus and
Paenibacillus across 170 experimental treatments. Y-axis indicates the various treatments in replicates where the first number indicates the concentrations of the extract (1 highest and 8 lowest), letters I (inner Oslofjord) and Y (outer
Oslofjord). A letter at the first position indicates the various controls: F = negative control containing phenantrene, P = procedural blank, M= Media blank (only media and sample, no DMSO) D = DMSO blank, containing media, DMSO and sample).
There were no responses in ACE richness and other alpha diversity measures such as Simpson’s index to the extract gradient as revealed by ANOVA, while alpha diversity showed significant differences between enrichments derived from the two fish species.
Similarly, generalised linear models revealed no significant response in the three diversity estimates to extract additions. Furthermore, the absence of significant responses in alpha diversity measures was also observed among the various controls (positive control, negative control, DMSO control and procedural blank).
Beta diversity was estimated for 170 samples and visualized using NMDS (Figure 7).
Visual inspection of this plot seems to show a weak effect of the treatments. Thus, we performed a clustering analysis and visualized the outputs in a heatmap (figure 8). This revealed clustering among fish species and location while extract concentrations had no trend. Cluster analysis corroborated the differentiation in beta diversity across locations and fish species, as can be observed in the heat map (Figure 8). Again, there seems to be no association between extract concentrations and beta diversity. This lack of an extract response was also confirmed by PERMANOVA. There were not any significant
concentration responses while communities were significantly different across location (p<0.002) and fish species (p<0.001), as well as replicated enrichments (p<0.001). Effect size as given by R2 from the PERMANOVA model was highest (0.21) among individual biological replicates (enrichments), followed by the extracts exposure (0.06), fish species (0.05) and location (0.04).
Performing the PERMANOVAs separately for each fish species revealed a significant response in pout (R2 = 0.11, p < 0.01) while there was an equally strong response in cod (R2 = 0.13) it was only marginally significant (p = 0.091). A significant difference in microbiome composition could be identified between the sampling locations in both fish species (cod: R2 = 0.057, p < 0.001; pout: R2 = 0.060, p < 0.001). Like in the overall PERMANOVA, biological replicates had a large effect in both fish species (cod: R2 = 0.12, p < 0.001; pout R2 = 0.29, p < 0.001).
Figure 7 Non-multidimensional scaling plot visualizing Bray Curtis distances among treatments. As evident from the plot no clear responses to the extract could be observed. PERMANOVAs confirmed these results.
Figure 8 Heatmap of Euclidean distances across the leachate treatments. Treatments are sorted by complete
hierarchical clustering using ‘pheatmap’ package in R. Some clusters are revealed in correspondence with fish species and location.
To further disentangle the effects on bacterial composition, we performed a differential abundance analysis using Deseq2. This identified five ASVs with significant responses to leachate exposure while the rest did not show any or only a single significant response at intermediate extract concentrations. ASVs with a negative response belonged to the genera Lysinibacillus (Figure 9), Unidibacterium (Figure 10) and Lautropia, with the latter already disappearing at the second lowest concentration (Figure 11).
Figure 09. ASV responses to the extract amendments as determined by Deseq2. Figure shows significant changes to Lysinibacillus at different concentrations, Y-axis represents the fold change in relation to the DMSO control (0 mg l-1 extract concentration). ‘ns’ indicates non significant fold changes while * indicates significant fold changes (p<0.05), **
(p<0.01), *** (p<0.001). On the top of the plots, the taxonomic assignments of the ASVs at the genus level are shown.
Figure 10. ASV responses to the extract amendments as determined by Deseq2. Figure shows significant changes to Unidibacterium at different concentrations. Y-axis represents the fold change in relation to the DMSO control (0 mg l-1 extract concentration). ‘ns’ indicates non significant fold changes while * indicates significant fold changes (p<0.05), **
(p<0.01), *** (p<0.001). On the top of the plots, the taxonomic assignments of the ASVs at the genus level are shown.
In these cases, significant responses could already be observed at concentrations of 0.007 and 0.015 mg/mL of extract.. ASVs assigned to the genus Bacillus exhibited both increasing (positive fold change) and negative (negative fold change) responses to the extract (figure 11 and 12). More bacteria with significant responses at a single
concentration were ASVs assigned to the genera Sporosarcina, Psychrobacter,
Paenibacillus, Lactococcus and Lautropia, with results shown in the supplementary data.
Figure 11. ASV responses to the extract amendments as determined by Deseq2. Only ASVs with significant responses to the extract are shown. Y-axis represents the fold change in relation to the DMSO control (0 mg l-1 extract concentration). ‘ns’
indicates non significant fold changes while * indicates significant fold changes (p<0.05), ** (p<0.01), *** (p<0.001). On the top of the plots, the taxonomic assignments of the ASVs at the genus level are shown.
Figure 12. ASV responses to the extract amendments as determined by Deseq2. Only ASVs with significant responses to the extract are shown. Y-axis represents the fold change in relation to the DMSO control (0 mg l-1 extract concentration). ‘ns’
indicates non significant fold changes while * indicates significant fold changes (p<0.05), ** (p<0.01), *** (p<0.001). On the top of the plots, the taxonomic assignments of the ASVs at the genus level are shown.
5 Discussion
The results from this study show that extracts derived from car tire granulates can affect the microbiome of two fish species. Our results point towards altered microbial diversity as shown by overall community responses and individual ASVs showing either an
increase or decrease in proportion due to the additions of extracts. This corroborates previous studies, where altered structures of fish microbiomes were observed in response to BAP and Phthalates additions (Gaulke et al 2016, Buerger et al 2020,
DeBofsky et al 2020). Still growth and diversity responses were often weaker than when compared to differences observed between locations, fish species and biological
replicates. This was also reflected in the effect size of the experimental treatments, thus suggesting a minor influence of MP extracts on fish microbiomes.
While there were only minor effects on overall microbiome characteristics, several genera were negatively affected by the extract additions as indicated by their decrease in differential abundance under leachate exposure. The mechanism explaining the
individual taxa responses remains elusive, in particular that certain genera seem to respond in non-linear fashion to the concentration gradient. Examples are ASVs annotated as Sporosacarina and Psychrobacter. Negatively affected genera were Lysinibacillus and Undibacterium showing up in lower quantities at higher
concentrations of extracts. Most notably is the ASV annotated to the genus Lautropia disappearing entirely at concentrations higher than 0.0037mg/ml. This indicates that certain genera might be highly sensitive to leachates from MPs and that this sensitivity might not be uniformly distributed across bacterial taxa.
The effect size between the extracts was generally smaller when compared with effect size resulting when comparing microbiome characteristics between fish species and location. The significant results when comparing the microbiome of the two fish species are likely due to different diets and habitats of the fish species. Cod favors depths
ranging from 6 to 60 meters while pout tends to prefer being a bottom dweller, preferring depths at 50-250 meters which has been shown to affect the gut flora
composition (Obadia et al 2018). The responses with regards to location may be linked to fish being exposed to higher levels of pollution in the inner fjord due to urbanization (Ruus., et al 2015), or be the result of different habitat and food choice between fish from the inner and outer fjord. As such ecological circumstances seem to be more important for the development of the in vitro microbiomes than extract additions up to 0.473 mg/l.
When we designed our study we decided to use different approaches from the norm, having two cod species from two different locations and using a complex mixture of compounds. Instead of the model fish organism Zebrafish, we used two species which allowed comparative analysis and drawing more general conclusions on observed effects. In addition, we obtained fish from two different locations with differences in microbiome composition as suggested by our in vitro results. Third, we used a car tire extract representing a complex mixture of compounds including PAHs (e.g. pyrene), additives (e.g. benzothiazole), phenols (e.g. bisphenol A) and metals (e.g. zinc) (Capolupo et al., 2020; Halsband et al., 2020). Previous screening detected a complex mixture of chemicals in plastic products (Groh et al 2019). While we see effects in our study, we cannot pinpoint to the actual chemicals that are responsible for the observed responses.
There is also evidence that many plastic products contain unknown chemicals, Zimmermann et al. (2019) demonstrated that most (82%) of the leachates screened from different plastic products, could not be identified using the NIST database. At least, as previously mentioned, benzo[a] pyrene is a well documented chemical that, among other things, induces a variety of toxic effects on the microbiome of organisms.(Defois et al 2017, Ribière et al 2016, Zhao et al 2019 Xie et al 2020). Still our results cannot be solely attributed to BAP, as there are other potential toxic compounds such as …..
The problem with forming knowledge on complex chemical cocktails, like MP leachate, pose a huge problem for manufacturers, health authorities and researchers alike. When
many chemicals interact with a host at the same time, the effects become hard to measure. In a situation where many chemicals are introduced, the single chemical may react independently as in a single exposure, or many chemicals may interact to modify the effects of the whole exposure (Koppe et al., 2006). In addition, being exposed to repeated low dose stressors may sometimes be more toxic than one single high dose exposure.(2008 VKM report "Combined toxic effects of multiple chemical exposures") This is why we decided to use an in vitro model, since we expected it would be easier to measure effects from an extract, rather then if we would conduct the experiment using an in vivo system were we would have a magnitude more factors to account for ie (Host health, diet etc) than using an in vitro model. However it is important to note that transferability to real life scenarios is difficult, since in vitro models are simple systems compared to in vivo models who can potentially experience way more external influence including host health and diet. However in vitro models are practical to work with, since we can work under controlled and thus isolate effector variables when compared to in vivo models.
There are multiple improvements that could be made to more closely resemble natural conditions of the gut. Trace elements and macronutrient concentrations of the media could more closely simulate in vivo conditions. Also the multiple freezing and thawing of the cultures resulted in the only remaining bacteria being spore-forming and highly freezing resistent taxa highly resistant against toxicants (Dimitraki et al 2007). The variable temperature in our incubation device could also be an important factor, as temperature can increase metabolic rate (Dehler et al., 2017, Fontaine et al., 2018). We could account for gender which has been shown to have substantial effects on the gut microbiome instead of pooling fish with potentially highly variable microbiomes. In vivo experiments could also allow longer exposure times and in parallel monitor host health in combination with alterations to the microbiome.
6 Conclusion
This study has focused on the toxic effects from chemical additives added to car tires on in vitro microbiomes extracted from two common fish species. Besides ecological differences, we observed significant responses due to the car tire plastic extracts although of minor importance for the overall observed variability in the in vitro
microbiomes. Overall, our results agree with previous research, pointing towards effects from extracts derived from plastics on marine organisms (Hermabessiere et al 2017, Gallo et al 2018, Rødland et al 2019). Thus, our results contribute to the increasing evidence that plastics affect the microbiome of fish.
The study concludes with these following points:
● We find at most a minor responses in growth characteristics of the microbiome and its diversity when exposed to different concentrations of the plastic extracts,
● We observe significant responses in the in vitro microbiomes when comparing fish species and location of the fish
● Although we got minor effects of leachates on overall community composition, we observed 4-10 fold decreases in the abundance of 4 taxa as well as 2 taxa
disappearing entirely already at very low leachate concentrations (0.0037 mg/l).
The conclusions from this study can be extrapolated into how fish holobionts and their associated microbiome react when exposed to leachates derived from microplastics. Still the insights gained from using an in vitro model to measure toxic effects from plastic extractions need to be validated in vivo.
References
Adamovsky O, Bisesi. J.H, Christopher J. Martyniuk,Plastics in our water: Fish
microbiomes at risk?,Comparative Biochemistry and Physiology Part D: Genomics and Proteomics,Volume 39,2021,100834,ISSN 1744-117X,https://doi.org/10.1016/j .cbd.2021.100834.
Andrady A (2011) Microplastics in the marine environment. Mar Pollut Bull 62(8):1596–1605
https://www.plasticseurope.org/application/files/5715/1717/4180/Plastics_the_facts_
2017_FINAL_for_website_one_page.pdf
Apprill A, McNally S, Parsons R, Weber L (2015) Minor revision to V4 region SSU rRNA 806R gene primer greatly increases detection of SAR11 bacterioplankton. Aquat Microb Ecol 75:129-137. https://doi.org/10.3354/ame01753
Bahrami, B., Macfarlane, S. and Macfarlane, G. (2011), Induction of cytokine formation by human intestinal bacteria in gut epithelial cell lines. Journal of Applied Microbiology, 110: 353-363. https://doi.org/10.1111/j.1365-2672.2010.04889.x
Bråte I.L.N., R.R. Hurley, K. Iversen, J. Beyer, K.V. Thomas, C.C. Steindal, N. Green, M. Olsen, A.L. Lusher Mytilus spp. as sentinels for monitoring microplastic pollution in Norwegian coastal waters: a qualitative and quantitative study Environ. Pollut., 243 (2018), pp.
383-393, 10.1016/j.envpol.2018.08.077
Buerger A, Dillon D, Schmidt J, Yang T, Zubcevic J, Martyniuk C, Bisesi J,Gastrointestinal dysbiosis following diethylhexyl phthalate exposure in zebrafish (Danio rerio): Altered microbial diversity, functionality, and network connectivity,Environmental
Pollution,Volume 265, Part B, https://doi.org/10.1016/j.envpol.2020.114496.
Capolupo M, Sørensen L, Jayasena K, Booth. A.M, Fabbri E,Chemical composition and ecotoxicity of plastic and car tire rubber leachates to aquatic organisms,Water Research,Volume 169,2020,115270,ISSN
0043-1354,https://doi.org/10.1016/j.watres.2019.115270.
DeBofsky A, Y. Xie, C. Grimard, A.J. Alcaraz, M. Brinkmann, M. Hecker, J.P. Giesy Differential responses of gut microbiota of male and female fathead minnow
(Pimephales promelas) to a short-term environmentally-relevant, aqueous exposure to benzo[a]pyrene Chemosphere, 252 (2020), Article 126461
Defois C, Ratel J, Denis S, Batut B, Beugnot R, Peyretaillade E, Engel E, Peyret P.
Environmental Pollutant Benzo[a]Pyrene Impacts the Volatile Metabolome and
Transcriptome of the Human Gut Microbiota Frontiers in Microbiology VOLUME 8 2017 PAGES 1562 DOI 10.3389/fmicb.2017.01562
Dehler, C.E., Secombes, C.J., Martin, S.A.M., 2017. Environmental and physiological
factors shape the gut microbiota of Atlantic salmon parr (Salmo salar L.).
Aquaculture 467, 149 e157
Desforges, JP.W., Galbraith, M. & Ross, P.S. Ingestion of Microplastics by Zooplankton in the Northeast Pacific Ocean. Arch Environ Contam Toxicol 69, 320–330 (2015).
https://doi.org/10.1007/s00244-015-0172-5
Dimitraki, P. & Velonakis, E.. (2007). The survival of pathogens in frozen food as a health risk. Archives of Hellenic Medicine. 24. 432-439.
Fontaine, S.S., Novarro, A.J., Kohl, K.D., 2018. Environmental temperature alters the digestive performance and gut microbiota of a terrestrial amphibian. J. Exp. Biol.221, jeb187559
Gallo, F., Fossi, C., Weber, R. et al. Marine litter plastics and microplastics and their toxic chemicals components: the need for urgent preventive measures. Environ Sci Eur 30, 13 (2018). https://doi.org/10.1186/s12302-018-0139-z
Gaulke CA, Barton CL, Proffitt S, Tanguay RL, Sharpton TJ (2016) Triclosan Exposure Is Associated with Rapid Restructuring of the Microbiome in Adult Zebrafish. PLoS ONE 11(5): e0154632. https://doi.org/10.1371/journal.pone.0154632
Geyer R, Jambeck JR, Law KL2017. Production, use, and fate of all plastics ever made. Sci Adv 3:e1700782. doi:10.1126/sciadv.1700782
Gibson and Fuller, Aspects of In Vitro and In Vivo Research Approaches Directed Toward Identifying Probiotics and Prebiotics for Human Use, The Journal of Nutrition, Volume 130, Issue 2, February 2000, Pages 391S–395S, https://doi.org/10.1093/jn/130.2.391S Groh K, Backhaus T, Carney-Almroth B, Geueke B, Inostroza P, Lennquist A, Muncke J Sci.
Total Environ., 651 (2) (2019), pp. 3253-3268, 10.1016/j.scitotenv.2018.10.015
Halsband C, Sørensen L, Booth A, Herzke Dorte. Car Tire Crumb Rubber: Does Leaching Produce a Toxic Chemical Cocktail in Coastal Marine Systems?. Frontiers in
Environmental Science 2020, 8 https://doi.org/10.3389/fenvs.2020.00125
Halsband C, Claudia Herzke, Dorte 2019/01/01 308- 318 Plastic litter in the European Arctic: What do we know? 5 10.1016/j.emcon.2019.11.001 Emerging Contaminants Herlemann, D., Labrenz, M., Jürgens, K. et al. Transitions in bacterial communities along the 2000 km salinity gradient of the Baltic Sea. ISME J 5, 1571–1579 (2011).
https://doi.org/10.1038/ismej.2011.41
Hermabessiere L, Dehaut A, Paul-Pont I, Lacroix C, Jezequel R, Soudant P, Duflos G.
Occurrence and effects of plastic additives on marine environments and organisms: A review. Chemosphere. 2017 Sep;182:781-793. doi: 10.1016/j.chemosphere.2017.05.096.
Epub 2017 May 16. PMID: 28545000.
Hernandez A G, O'Brien J,Berrow S, O'Connor I, Officer R, Microplastic and macroplastic ingestion by a deep diving, oceanic cetacean: The True's beaked whale Mesoplodon mirus, Environmental Pollution,Volume 199,2015,Pages 185-191,ISSN 0269-7491 , https://doi.org/10.1016/ j.envpol.2015.01.023.
Hu C, Liu M, Wan T, Tang L, Sun B, Zhou B, James C. W. Lam, Paul K. S. Lam, and Chen L Environmental Science & Technology 2021 55 (1), 529-537
DOI: 10.1021/acs.est.0c06884
Javurek A, Spollen W, Johnson S, Bivens J, Bromert K, Givan S & Rosenfeld C (2016) Effects of exposure to bisphenol A and ethinyl estradiol on the gut microbiota of parents and their offspring in a rodent model, Gut Microbes, 7:6, 471-485, DOI:
10.1080/19490976.2016.1234657
Jovanović, B. (2017), Ingestion of microplastics by fish and its potential consequences from a physical perspective. Integr Environ Assess Manag, 13: 510-515.
https://doi.org/10.1002/ieam.1913
Koelmans A.A., Besseling E., Shim W.J. (2015) Nanoplastics in the Aquatic Environment.
Critical Review. In: Bergmann M., Gutow L., Klages M. (eds) Marine Anthropogenic Litter.
Springer, Cham. https://doi.org/10.1007/978-3-319-16510-3_12
Koppe, j.g., bartonova, a., bolte, g., bistrup, m.l., busby, c., butter, m., dorfman, p., fucic, a., gee, d., van den hazel, p., howard, v., kohlhuber, m., leijs, m., lundqvist, c., moshammer, h., naginiene, r., nicolopoulou-stamati, p., ronchetti, r., salines, g., schoeters, g., tusscher, g.t., wallis, m.k. and zuurbier, m. (2006), exposure to multiple environmental agents and their effect. acta pædiatrica, 95: 106-113.
https://doi.org/10.1080/08035320600886646
Kubo k, Arai O, Omura M, Watanabe R, Ogata R, Aou S, Low dose effects of bisphenol A on sexual differentiation of the brain and behavior in rats, Neuroscience Research, Volume 45, Issue 3, 2003, Pages 345-356, ISSN 0168-0102,
https://doi.org/10.1016/S0168-0102(02)00251-1.
Kühn S., Bravo Rebolledo E.L., van Franeker J.A. (2015) Deleterious Effects of Litter on Marine Life. In: Bergmann M., Gutow L., Klages M. (eds) Marine Anthropogenic Litter.
Springer, Cham. https://doi.org/10.1007/978-3-319-16510-3_4
Lear, G., Kingsbury, J.M., Franchini, S. et al. Plastics and the microbiome: impacts and solutions. Environmental Microbiome 16, 2 (2021).
https://doi.org/10.1186/s40793-020-00371-w
Loibner, A.P., Szolar, O.H.J., Braun, R. and Hirmann, D. (2004), Toxicity testing of 16 priority polycyclic aromatic hydrocarbons using Lumistox®. Environmental Toxicology and Chemistry, 23: 557-564. https://doi.org/10.1897/03-59
Lu K, Phillip R.A, Schlieper K.A, Graffam M.E, Levine S, Wishnok J.S, Swenberg J.A, Tannenbaum S.R, and Fox J.G 2014 Arsenic Exposure Perturbs the Gut Microbiome and Its Metabolic Profile in Mice: An Integrated Metagenomics and Metabolomics Analysis Environmental Health Perspectives 122:3 CID: https://doi.org/10.1289/ehp.1307429
Maćczak, M. Cyrkler, B. Bukowska, J. Michałowicz Bisphenol A, bisphenol S, bisphenol F and bisphenol AF induce different oxidative stress and damage in human red blood cells (in vitro study) Toxicol. In Vitro, 41 (2017), pp. 143-149, 10.1016/j.tiv.2017.02.018 Obadia B, Keebaugh ES, Yamada R, Ludington WB, Ja WW (2018) Diet influences host-microbiota associations in Drosophila. Proc Natl Acad Sci USA 115:E4547–E4548.
Pedà, C., Caccamo, L., Fossi, M. C., Gai, F., Andaloro, F., Genovese, L., et al. (2016). Intestinal alterations in European sea bass Dicentrarchus labrax (Linnaeus, 1758) exposed to microplastics: preliminary results. Environ. Pollut. 212, 251–256. doi:
10.1016/j.envpol.2016.01.083
Pham C, Rodríguez Y, Dauphin A, Carriço R, P.G.L J. Frias, Vandeperre F, Otero V, R. Santos, R. Martins, B. Bolten, Karen A., Plastic ingestion in oceanic-stage loggerhead sea turtles (Caretta caretta) off the North Atlantic subtropical gyre, Marine Pollution Bulletin, Volume 121, Issues 1–2, 2017, Pages 222-229, ISSN 0025-326X,
https://doi.org/10.1016/j.marpolbul.2017.06.008.
Ribière, C., Peyret, P., Parisot, N. et al. Oral exposure to environmental pollutant benzo[a]pyrene impacts the intestinal epithelium and induces gut microbial shifts in murine model. Sci Rep 6, 31027 (2016).https://doi.org/10.1038/srep3102
Rosemarie De Weirdt, Sam Possemiers, Griet Vermeulen, Tanja C.W. Moerdijk-Poortvliet, Henricus T.S. Boschker, Willy Verstraete, Tom Van de Wiele, Human faecal microbiota display variable patterns of glycerol metabolism, FEMS Microbiology Ecology, Volume 74, Issue 3, December 2010, Pages 601–611,
https://doi.org/10.1111/j.1574-6941.2010.00974.x
Ruus A, Allan I, Beylich B, Bæk K, Schlabach M, Helberg M. 2015. Environmental Contaminants in an Urban Fjord, 2014. Report M375-2015 from the Norwegian Environment Agency. 110 pp + Appendix.
Teuten E L., Saquing J M., Knappe D R. U., Barlaz M A., Jonsson S, Björn A, Rowland S J., Thompson R C., Galloway T S., Yamashita R, Ochi D, Watanuki Y, Moore C, Viet P H, Tana T Se, Prudente M, Boonyatumanond R, Zakaria M P., Akkhavong K, Ogata Y, H H, Iwasa S, Mizukawa K, Hagino Yi, Imamura A, Saha M and Takada H 2009Transport and release of chemicals from plastics to the environment and to wildlifePhil. Trans. R. Soc.
B3642027–2045
Turnbaugh, P., Hamady, M., Yatsunenko, T. et al. A core gut microbiome in obese and lean twins. Nature 457, 480–484 (2009). https://doi.org/10.1038/nature07540
Venema K, P. Van Den Abbeele Experimental models of the gut microbiome Best Practice
& Research Clinical Gastroenterology, 27 (1) (2013), pp. 115-126, 10.1016/j.bpg.2013.03.002
Wang D, Ren J, Tan Z, and You J. Microbial Profiles in Nereis succinea and Their Contribution to the Degradation of Organic Pollutants Environmental Science &
Technology 2020 54 (10), 6235-6243 DOI: 10.1021/acs.est.9b07854
Wright S.L, Thompson R.C, Galloway T.S, The physical impacts of microplastics on marine organisms: A review, Environmental Pollution,Volume 178,2013,Pages 483-492,ISSN 0269-7491,
https://doi.org/10.1016/j.envpol.2013.02.031.(https://www.sciencedirect.com/science /article/pii/S0269749113001140)
Xie S, Zhou A, Xu N, Feng Y, Pan Z, Junaid M, Wang J, Zou J, Benzo[a]pyrene induces microbiome dysbiosis and inflammation in the intestinal tracts of western mosquitofish (Gambusia affinis) and zebrafish (Danio rerio),Fish & Shellfish Immunology, Volume 105, 2020, Pages 24-34, ISSN 1050-4648,https://doi.org/10.1016/j.fsi.2020.06.041.
Zhao Y, Liu H, Wang Q, Li B, Zhang H, Pi Y, The effects of benzo[a]pyrene on the composition of gut microbiota and the gut health of the juvenile sea cucumber Apostichopus japonicus Selenka,Fish & Shellfish Immunology,Volume 93,2019,Pages 369-379,ISSN 1050-4648,https://doi.org/10.1016/j.fsi.2019.07.073.
Zalasiewicz J, Colin N., Juliana A. Sul I, Patricia L. Corcoran, D. Barnosky A, Cearreta A, Edgeworth M, Gałuszka A, Jeandel C, Leinfelder R, J.R. McNeill, Steffen W, Summerhayes C, Wagreich M, Williams M, P. Wolfe A, Yonan Y,The geological cycle of plastics and their use as a stratigraphic indicator of the Anthropocene, Anthropocene, Volume 13, 2016, Pages 4-17, ISSN 2213-3054, https://doi.org/10.1016/j.ancene.2016.01.002.
Zimmermann L, Dierkes G, Ternes T, Völker C, and Wagner M Environmental Science &
Technology 2019 53 (19), 11467-11477DOI: 10.1021/acs.est.9b02293
APPENDIX
YCFA Medium
1611. YCFA MEDIUM (modified) Casitone 10.00 g
Yeast extract 2.50 g Glucose 5.00 g
MgSO4 x 7 H2O 45.00 mg CaCl2 x 2 H2O 90.00 mg
K2HPO4 0.45 g KH2PO4 0.45 g NaCl 0.9 g
Resazurin 1.0 mg
Distilled water 1000.00 ml
NaHCO3 4.00 g L-Cysteine-HCl 1.00 g Hemin 10.00 mg
Volatile fatty acids:
Acetic acid 1.90 ml Propionic acid 0.70 ml
iso-Butyric acid 90.00 µl n-Valeric acid 100.00 µl iso-Valeric acid 100.00 µl
Vitamin solution:
Biotin 2.00 mg Folic acid 2.00 mg
Pyridoxine-HCl 10.00 mg Thiamine-HCl x 2 H2O 5.00 mg
Riboflavin 5.00 mg Nicotinic acid 5.00 mg D-Ca-pantothenate 5.00 mg Vitamin B12 0.10 mg
p-Aminobenzoic acid 5.00 mg Lipoic acid 5.00 mg
Distilled water 1000.00 ml
Dissolve ingredients except NaHCO3, hemin, cysteine in water and boil for 10 min, then cool to room temperature while gassing with 100% CO2.Add the NaHCO3, hemin, cysteine and adjust pH to 6.7 -6.8. Distribute under N2 into Hungate tubes (10 ml) and autoclave. Filter sterilize vitamin solution and aseptically add 100 µl per Hungate tube.
DNA Extraction Protocol: Pretreatment for Gram-Positive Bacteria
Procedure
1. Harvest cells (maximum 2 x 109 cells) in a microcentrifuge tube by centrifuging for 10 min at 5000 x g (7500 rpm). Discard supernatant.
2. Resuspend bacterial pellet in 180 µl enzymatic lysis buffer.
3. Incubate for at least 30 min at 37°C. After incubation, heat the heating block or water bath to 56°C if it is to be used for the incubation in step 5. DNeasy Blood & Tissue Handbook 07/2020 51
4. Add 25 µl Proteinase K and 200 µl Buffer AL (without ethanol). Mix by vortexing. Note:
Do not add Proteinase K directly to Buffer AL. Ensure that ethanol has not been added to Buffer AL (see “Buffer AL”, page 19). Buffer AL can be purchased separately (see ordering information starting on page 59).
5. Incubate at 56°C for 30 min. Optional: If required, incubate at 95°C for 15 min to inactivate pathogens. Note that incubation at 95°C can lead to some DNA degradation.
6. Add 200 µl ethanol (96–100%) to the sample, and mix thoroughly by vortexing. It is important that the sample and the ethanol are mixed thoroughly to yield a homogeneous solution. A white precipitate may form on addition of ethanol. It is essential to apply all of the precipitate to the DNeasy Mini spin column. This precipitate does not interfere with the DNeasy procedure.
7. Continue with step 4 of the protocol “Purification of Total DNA from Animal Tissues SpinColumn Protocol)”, page 30.
8. Pipet the mixture from step 3 (including any precipitate) into the DNeasy Mini spin column placed in a 2 ml collection tube (provided). Centrifuge at ≥6000 x g (8000 rpm) for 1 min. Discard flow-through and collection tube.*
9. Place the DNeasy Mini spin column in a new 2 ml collection tube (provided), add 500 µl Buffer AW1, and centrifuge for 1 min at ≥6000 x g (8000 rpm). Discard flow-through and collection tube.* * Flow-through contains Buffer AL or Buffer AW1 and is therefore not compatible with bleach. See page 6 for safety information. DNeasy Blood & Tissue Handbook 07/2020 33
10. Place the DNeasy Mini spin column in a new 2 ml collection tube (provided), add 500 µl Buffer AW2, and centrifuge for 3 min at 20,000 x g (14,000 rpm) to dry the DNeasy membrane. Discard flow-through and collection tube. It is important to dry the
membrane of the DNeasy Mini spin column, since residual ethanol may interfere with subsequent reactions. This centrifugation step ensures that no residual ethanol will be carried over during the following elution. Following the centrifugation step, remove the DNeasy Mini spin column carefully so that the column does not come into contact with the flow-through, since this will result in carryover of ethanol. If carryover of ethanol occurs, empty the collection tube, then reuse it in another centrifugation for 1 min at 20,000 x g (14,000rpm).
11. Place the DNeasy Mini spin column in a clean 1.5 ml or 2 ml microcentrifuge tube (not provided), and pipet 200 µl Buffer AE directly onto the DNeasy membrane. Incubate at room temperature for 1 min, and then centrifuge for 1 min at ≥ 6000 x g (8000 rpm) to elute. Elution with 100 µl (instead of 200 µl) increases the final DNA concentration in the eluate, but also decreases the overall DNA yield (see Figure 2, page 23).
12. Recommended: For maximum DNA yield, repeat elution once as described in step 7.
This step leads to increased overall DNA yield. A new microcentrifuge tube can be used for the second elution step to prevent dilution of the first eluate. Alternatively, to
combine the eluates, the microcentrifuge tube from step 7 can be reused for the second elution step. Note: Do not elute more than 200 µl into a 1.5 ml microcentrifuge tube because the DNeasy Mini spin column will come into contact with the eluate.
PCR reactions
PCR purification
96 Well Format Procedure
1. Determine whether or not a plate transfer is necessary. If the PCR reaction volume multiplied by 2.8 exceeds the volume of the PCR plate, a transfer to a 300 μL round bottom plate or a 1.2 mL deep-well plate is required.
2. Shake the Agencourt AMPure XP bottle to resuspend any magnetic particles that may have settled. Then add Agencourt AMPure XP according to the sample reaction volume shown in Table 4. The volume of Agencourt AMPure XP for a given reaction can be derived from the following equation: (Volume of Agencourt AMPure XP per reaction) =
1.8 x (Reaction Volume)
3. This step binds DNA fragments 100 bp and larger to the magnetic beads. Pipette mixing is preferable to vortexing as it tends to be more reproducible. The color of the mixture should appear homogenous after mixing: -- Mix reagent and sample thoroughly by pipette mixing 10 times. Let the mixed samples incubate for 5 minutes at room temperature for maximum recovery.
4. Place the reaction plate onto an Agencourt SPRIPlate 96 Super Magnet Plate for 2 minutes to separate beads from the solution. IMPORTANT Wait for the solution to clear before proceeding to the next step.
5. This step must be performed while the reaction plate is situated on the Agencourt SPRIPlate 96 Super Magnet Plate: -- Aspirate the cleared solution from the reaction plate and discard. Leave 5 μL of supernatant behind, otherwise beads are drawn out with the supernatant. IMPORTANT Do not disturb the ring of separated magnetic
beads.IMPORTANT Perform the next step with the reaction plate situated on an Agencourt SPRIPlate 96 Super Magnet Plate. Do not disturb the separated magnetic beads. Also, be sure to remove all of the ethanol from the bottom of the well.
6. Dispense 200 μL of 70% ethanol to each well of the reaction plate and incubate for 30 seconds at room temperature. Aspirate out the ethanol and discard. NOTE If the total
volume of sample plus reagent exceeds 200 μl, then use a wash volume of at least the volume of sample plus reagent. Repeat for a total of two washes. The beads are not drawn out easily when in alcohol, so it is not necessary to leave any supernatant behind.
NOTE A dry time is optional to ensure all traces of Ethanol are removed. For fragments 10 kb and larger, do not over dry the bead ring (bead ring appears cracked if over dried) as this will significantly decrease elution efficiency.
7. Remove the reaction plate from the magnet plate, and then add 40 μL of elution buffer to each well of the reaction plate and pipette mix 10 times. Incubate for 2 minutes. The liquid level will be high enough to contact the magnetic beads at a 40 μL elution volume.
A greater volume of elution buffer can be used, but using less than 40 μL will require extra mixing (to ensure the liquid comes into contact with the beads), and may not be sufficient to elute the entire PCR product. 8. Place the reaction plate onto an Agencourt SPRIPlate 96 Super Magnet Plate for 1 minute to separate beads from the solution.
9. Transfer the eluate to a new plate. NOTE Bead carryover into the final plate is usually not a cause for concern. The samples can be stored in the freezer with beads and the beads are inert in downstream enzymatic reactions. If bead carryover must be limited for any reason, 2 μL – 5 μL of eluate can be left behind in the original plate. In addition, a second transfer away from the beads is optional. To do so, place the final plate containing beads and eluate onto the magnet for 1 minute to separate the beads. Transfer the eluate into another clean plate.
