Environmental DNA Metabarcoding Supporting Community

Subscriber access provided by Kaohsiung Medical University

Ecotoxicology and Human Environmental Health

eDNA metabarcoding supporting community assessment of environmental stressor in a field-based sediment microcosm study Jianghua Yang, Katherine Joanna Jeppe, Vincent J. Pettigrove, and Xiaowei Zhang Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.8b04903 • Publication Date (Web): 12 Nov 2018 Downloaded from http://pubs.acs.org on November 12, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 38

Environmental Science & Technology

1

TOC (Table of Contents Art)

2 3

ACS Paragon Plus Environment


metabarcoding

supporting

community

Page 2 of 38

4

eDNA

assessment

5

environmental stressor in a field-based sediment microcosm study

of

6 7

Jianghua Yang†, Katherine Jeppe§, Vincent Pettigrove§ and Xiaowei Zhang†*

8 9

†State Key Laboratory of Pollution Control & Resource Reuse, School of the

10

Environment, Nanjing University, Nanjing 210023, China

11

§Aquatic Pollution Prevention Partnership, School of Science, RMIT University, PO

12

Box 71, Bundoora, 3083, Victoria, Australia.

13 14 15 16 17 18 19 20

Corresponding author: Xiaowei Zhang School of Environment, Nanjing University 163 Xianlin Avenue, Nanjing, 210023, China Tel: (86)-25-89680623 E-mail address: [email protected] [email protected]

21 22


Page 3 of 38


23

Abstract

24

Conventional ecological risk assessment on toxic stressors in sediment is

25

limited to a small and select fraction of benthic communities. Ecogenomic

26

approaches provide unprecedented capacity to monitor the changes of biodiversity and

27

community composition in the field, but how to utilize it to assess ecological impact by

28

contaminates remains largely unexplored. Here, an environmental DNA (eDNA)

29

metabarcoding approach was used to assess the effect of copper on changes in

30

biodiversity and community composition across the Tree of Life (including bacteria,

31

protist, algae, fungi and metazoa) in a field-based microcosm. Many micro-

32

organisms across a broad range of taxa groups changed their relative abundance in

33

response to increased copper concentrations in sediments. Changes in community

34

structure of microbiota appeared to be more sensitive to copper than survival of

35

laboratory-bred organisms and indigenous macroinvertebrates. Copper caused a

36

significant shift in prokaryotic community composition via substitution of

37

dominant species. Network heterogeneity and Shannon diversity of the bacterial

38

community decreased in the high copper treatments. eDNA metabarcoding assessed the

39

effects of copper contaminated sediment with less effort than manually

40

processing samples. Our study highlighted the value of community profiling by eDNA

41

based approach in prospective and retrospective risk assessment of environmental

42

stressors.

43



44

Introduction

45

Historically, assessing the impacts of environmental stressors on aquatic

46

ecosystem relies on the laboratory-based animal bioassays1-4, which have been

47

criticized for lacking environmental relevance and for being less sensitive than field

48

responses5. Sediment contaminants are widespread environmental stressors and can

49

severely impact aquatic ecosystems and human health6. Field-based microcosm

50

approaches establish a semi-controlled lentic habitat for field exposure, which

51

allow us to isolate the effects of specific contaminants, and determine the

52

responses of a range of indigenous species7. In addition, these field-based

53

experiments can describe the response of indigenous species that cannot be

54

reared or tested in laboratory. Therefore, these tests provide a more

55

environmentally relevant assessment of sediment toxicity than laboratory-based tests 8,

56

9.

57

A major objective of ecological risk assessment of toxic substance is to protect the

58

local biodiversity and ecosystem function or service10. However, it has been

59

increasingly recognized that the traditional focus of ecotoxicology on individuals is

60

insufficient to guide environmental management decision making11. A stressor may

61

cause complex indirect responses in the field, such as sensitive taxa being replaced by

62

more tolerant organisms

63

be altered by shifts of community composition14. Furthermore, meio- and microbiota

64

play a critical role in many biogeochemical processes. Yet only the responses of a

12, 13.

It has been recognized that ecosystem function can


Page 4 of 38

Page 5 of 38


65

biased fraction of biodiversity, such as macroinvertebrates are assessed whereas

66

the responses of other ecological components, such as bacteria and eukaryotic

67

microbiota which are difficult to identify, are often overlooked.

68

Omics advances in ecotoxicological studies provide a wholly new paradigm

69

for ecotoxicology by linking ecological models to mechanism-based, systems

70

biology approaches. With the increasing development of DNA high-throughput

71

sequencing (HTS), environmental DNA (eDNA) metabarcoding has greatly improved

72

our ability to identify diverse taxonomic groups

73

potential to not only assess phylogenetic biodiversity more reliably and rapidly than

74

ever before, but also provides a cost-effective tool to evaluate community structures

75

that are relevant to the function of biota residing in the ecosystem17. This technical

76

advance also provides an opportunity to overcome some limitations of

77

traditional ecotoxicology research that focused on the adverse effects of

78

chemicals on individual organisms, and allow assessing the responses from

79

individual level to the community and ecosystem level18.

15, 16.

eDNA metabarcoding has the

80

Here a field-based microcosm was used to study the effects of copper on

81

the biodiversity across the Tree of Life with the eDNA metabarcoding. Copper is one

82

of the most common heavy metals and is toxic, in trace amounts, to both

83

humans and aquatic organisms. Sediment copper contamination is also

84

widespread across the world. In a previous laboratory-based toxicity tests, we

85

used metabarcoding to describe the impact of copper on microbial



Page 6 of 38

86

communities after exposure with two macroinvertebrates, (Chironomus tepperi

87

and Austrochiltonia subtenuis). The microbial communities were found to be

88

more sensitive to copper than that of laboratory-bred macroinvertebrates19.

89

However, the natural variation of community composition and the indirect

90

effects of interspecies interaction were not considered in these tests. The aims

91

of this study are to: 1) determine the impacts of copper at community level (biodiversity

92

and community structure) of the meio- and microbiota in field-based microcosms; 2)

93

describe the effects of copper on the bio-interactions among bacteria, protist, algae and

94

fungi communities; 3) evaluate the sensitivity of novel community endpoints by

95

metabarcoding technology to derive the sediment criteria in both the field-based

96

microcosms and laboratory-based tests.

97

Methods and Materials

98

Microcosm experiment

99

The field-based microcosm experiment was set up at the reference Glynns wetland

100

(Warrandyte, Melbourne AUS) which supports a variety of aerial invertebrates and has

101

been used for previous microcosm experiments 20. Sediment was spiked following the

102

methods of Simpson et al. 21, 22. A high-spike sediment stock (6000 mg/kg) was diluted

103

to five exposure concentrations of copper (62.5, 125, 250, 500 and 750 mg/kg) which

104

span the range of environmentally relevant concentrations

105

treatment was analyzed at the start and end of the experiment for total copper through


23.

Sediment from each

Page 7 of 38


24.

106

ICP-AES method

Sediments were fully mixed in the laboratory for four weeks

107

before the experiment began. Each treatment had five replicate microcosms that

108

consisted of a 20 L polypropylene plastic tub containing 500 g sediment and 15 L

109

reference site water. Microcosms were covered with a polypropylene net (2 cm2 mesh)

110

to prevent large predators entering the microcosm and allowing small invertebrates to

111

colonize. Three laboratory-bred test organisms (two snail species, Potamopyrgus

112

antipodarum (Gray, 1843) and Physella acuta (Draparnaud, 1805), and

113

Chironomus tepperi (Skuse, 1889)) were deployed in the microcosm experiments. Two

114

snail species were added in week one and exposed for six weeks. Each replicate

115

microcosm contains 40 individuals of P. antipodarum and five individuals of P. acuta.

116

Five days before the experiment concluded, 40 five-day old Chironomus tepperi larvae

117

were added to each replicate to simulate a standard five-day toxicity test 25. At the end

118

of the experiment, the macroinvertebrates were collected and identified in laboratory.

119

The toxicological response of these macroinvertebrates and laboratory deployed

120

species was reported elsewhere 24.

121

Environmental DNA isolation

122

One week before the experiment concluded, three microcosms of each treatment

123

were randomly selected to collect about 10 g of sediment using the polypropylene

124

pipette. Each treatment had three replicates. Sediments were freeze-dried at -40 ℃ and

125

homogenized before DNA isolation. About 0.3 g of homogenized sediment was used

126

to extract the eDNA with the MoBio Power Soil DNA Kit (MoBio Laboratories Inc.,



127

CA, USA) following the manual. Genomic DNA was checked for purity in a Take3

128

microplate in the Synergy H4 Hybrid Multi-Mode Microplate reader (BioTek, VT,

129

USA) and quantified by Qubit dsDNA HS assay kits (Invitrogen, USA).

130

PCR amplification and sequencing

131

Bacterial 16S rRNA genes and eukaryotic 18S rRNA genes were amplified using

132

V3 primers (modified primers 341F and 518R) 26 and V9 primers (1380F and 1510R)

133

27,

134

potential PCR bias. The PCR amplicons were sequenced by the Ion Torrent Proton

135

(Life Technologies, CA, USA) according to the manufacturer’s protocols.

136

Bioinformatics

137

respectively. Triplicate PCR reactions were performed for each sample to minimize

QIIME v.1.8.0 (Quantitative Insights Into Microbial Ecology) pipeline was used 28.

138

to process the raw sequences

Low quality sequences were trimmed via the

139

“split_libraries.py” script with “-w 50 -q 20”. PCR chimera filtering was performed via

140

“parallel_identify_chimeric_seqs.py” script with the default parameter. Operational

141

taxonomic units (OTUs) were selected with a sequence similarity cutoff of 97%

142

following the UPARSE pipeline in both 16S and 18S sequencing29. The homopolymer

143

errors were removed before OTUs clustering in QIIME by the split_libraries.py

144

function with the -H 6 parameter. For each OTU, a representative sequence was chosen

145

and taxonomy was assigned using the RDP classifier30 against the Greengenes


Page 8 of 38

Page 9 of 38


146

database31 and SILVA database32 for prokaryote and eukaryote community,

147

respectively.

148

Statistical analysis

149

All statistical analysis was conducted in R (version 3.3.0, released on 2016-05-

150

03). Samples were rarefied at the lowest sequencing depth to reduce biases resulting

151

from differences in sequencing depth (8134 and 10410 for eukaryote and prokaryote

152

community, respectively). Beta diversity was characterized by weighted UniFrac

153

distances calculated between samples

154

scaling (CSS) normalization. The response of relative abundance of microbiota taxa to

155

copper was modeled with a 3-parameter log-logistic model and the 50% effects

156

concentration (EC50) was calculated. A correlation network of bacteria community was

157

generated by SparCC with 100 bootstraps to assign two-sided pseudo p values34.

158

Samples were arbitrarily divided into three subgroups (Low copper: control and 62.5

159

mg/kg, Medium copper: 125 and 250 mg/kg, High copper: 500 and 750 mg/kg) to

160

increase the reliability of the correlation analysis. The network was filtered by

161

correlation magnitudes > 0.6 and < -0.6 which indicating strong co-abundant and co-

162

exclusion relationships. Principal component analysis (PCA) was employed to observe

163

differences in microbiota between copper treatments and the significance was assessed

164

by the ANOVA test with Dunnett’s post hoc test. Bio-interaction among taxonomic

165

groups were measured by the Mantel test based on the OTUs metrics. Differences of

33.

OTU tables underwent cumulative sum



166

unique taxon number between copper treatments were assessed using the t tests.

167

Sediment copper benchmark was determined using the following: (i) the 3 parameters

168

log-logistic model was used to screen the “responsive OTU” from the spectrum of

169

biodiversity profiled by DNA metabarcoding based on the changes of relative reads

170

abundance. The responsive OTUs were determined by statistical significance with p

171

value < 0.01. (ii) the concentration causing 50% of the maximum effects (changes of

172

relative abundance) (EC50) was calculated. (iii) the EC50 of all responsive OTUs were

173

fitted by the species sensitivity distribution (SSD) model and the hazard concentration

174

for 5% of OTUs (HC5) was calculated by the bootstrap method.

175

Results

176

Composition of benthic communities in the field-based microcosm

177

939 prokaryotic taxa and 878 eukaryotic taxa were recovered by eDNA

178

metabarcoding across the Tree of Life from the sediments in the field-based

179

microcosms (Fig.1, Tables S1& S2). The three replicates from each treatment showed

180

high consistency in OTU and taxon composition (mean R2 > 0.88, p < 0.0001, see SI,

181

Fig.S1-S2). The differences of UniFrac distance between the treatments are

182

significantly greater than that within treatments (see SI, Fig.S3-S4). The prokaryotic

183

communities were dominated by Proteobacteria, Cyanobacteria, Bacteroidetes,

184

Chloroflexi, Planctomycetes and Firmicutes. Metazoans were dominated by

185

Arthropoda, Mollusca, Rotifer and Nematoda. Protists were dominated by


Page 10 of 38

Page 11 of 38


186

Dinophyceae, Spirotrichea and Endomyxa. Algal communities were dominated by

187

Chlorophyceae, Trebouxiophyceae and Bacillariophyta. Fungi were dominated by

188

Ascomycota and Blastocladiomycota (see SI, Fig.S5 & Table S2).

189

Micro-prokaryotes responses to copper exposure

190

Prokaryotes showed diverse responses to copper in the microcosms (Fig.2).

191

Although the total abundance of cyanobacteria was stable in different copper treatments,

192

the relative abundance of different taxa varied significantly and the dominant OTUs

193

changed (Fig.S6). Nineteen of the 38 cyanobacteria taxa were obviously correlated with

194

copper, where 7 taxa had higher abundance levels and 12 taxa were less abundant in

195

higher copper concentrations than the control (Fig.S7 & Table S3). Pseudanabaenales

196

replaced Stramenopiles (formally, Heterokonta) and Synechococcales, as the new

197

dominant cyanobacteria at high copper treatments. Substitution of dominant species

198

also occurred in the communities of Verrucomicrobia, Proteobacteria, Spirochaetes,

199

Planochaetes and Armatimonadetes (Fig.S6 & Fig.S7). Some prokaryotes showed only

200

monotonic response pattern to copper exposure. For example, Chloroflexi (LL.3

201

models, AIC = -26.2, p < 0.001) and Actinobacteria (LL.3 models, AIC = -9.0, p
10 was considered as present. The singletons were

658

removed before OTU clustering. (C) changes in bio-interaction that measured by

659

mantel test. Width of lines shows the Rho in mantel test between two communities.

660

Low: 0 and 62.5 mg/kg; medium: 125 and 250 mg/kg; high: 500 and 750 mg/kg.



661

Fig.5 Responses of eukaryote and prokaryote community structure in copper

662

contaminated sediment. Principal components analysis was used to differentiate the

663

changes of community structure. Boxplot shows the differences of PC1 among different

664

copper treatments with the t test. * p < 0.05, ** p < 0.01 and *** p < 0.001.

665

Fig.6 Comparison between field microcosm and laboratory test. (A) shared responsive

666

species; (B) response pattern at the class level; (C) species sensitivity distribution of

667

macroinvertebrates and microbiotas. Sediment copper criteria (HC5) was derived by

668

the 3-parameter log−logistic model.

669


Page 32 of 38

Page 33 of 38

670


Fig.1

671 672



673

Fig.2

674 675


Page 34 of 38

Page 35 of 38

676


Fig.3

677



678

Fig.4

679


Page 36 of 38

Page 37 of 38

Fig.5

PC1 ( 49.3 % variance explained)


0.0

Cryptomycota Basidiomycota Ascomycota

-0.4

-0.2

Blastocladiomycota

***


681


0.6

-0.2

-0.4

***

***

*** 0.4

**

0.2

Cu750 Cu500 Cu250 Cu125 Cu62.5 Cu0

0.4 0.2 0.0 -0.2 -0.4

***

0.4

0.2

0.0

-0.2

**

*** ***

0.6

***

-0.4

0.6

0.4

0.0

0.4 0.2

Chytridiomycota

-0.6

0.2

0.0

-0.2

-0.4

-0.6

***

0.4

*** *** ***


0.2 0.4 0.6 -0.2

Rotifera Platyhelminthes Nematoda

-0.6


Mollusca

Trebouxiophyceae Bacillariophyta


Fungi

***

Arthropoda

Cu750 Cu500 Cu250 Cu125 Cu62.5 Cu0

-0.2


Metazoa


0.4 0.2 0.0

***

Glaucocystophyceae Hyphochytriomyceta Cryptophyceae

Chlorophyceae

Cu750 Cu500 Cu250 Cu125 Cu62.5 Cu0

*** ***

***

-0.4

0.2

0.0

-0.2

-0.4

***

Cu750 Cu500 Cu250 Cu125 Cu62.5 Cu0

*** ***

***

Endomyxa

0.4

Cu750 Cu500 Cu250 Cu125 Cu62.5 Cu0

Filosa-Granofilosea

Spirotrichea

0.2

Planctomycetes

-0.2

Bacteroidetes Acidobacteria

Dinophyceae

-0.4

Actinobacteria

-0.2

Chloroflexi


0.0

Firmicutes

Algae

-0.6

0.2

0.4

Protist

-0.4


Bacteria

0.0

680



682

Fig.6

683


Page 38 of 38

Environmental DNA Metabarcoding Supporting Community

Recommend Documents