Content uploaded by David J Beale
Author content
All content in this area was uploaded by David J Beale on Nov 01, 2021
Content may be subject to copyright.
Short Communication
Bioaccumulation and metabolic response of PFAS mixtures in
wild-caught freshwater turtles (Emydura macquarii macquarii)using
omics-based ecosurveillance techniques
David J. Beale
a,
⁎,Katie Hillyer
a
,Sandra Nilsson
b
, Duncan Limpus
c
, Utpal Bose
d
,
James A. Broadbent
d
,Suzanne Vardy
e
a
Land and Water, Commonwealth Scientific and Industrial Research Organisation, Ecosciences Precinct, Dutton Park, QLD 4102, Australia
b
Queensland Alliance for Environmental Health Sciences, The University of Queensland, Woolloongabba, QLD 4102, Australia
c
Aquatic Threatened Species, Wildlife and Threatened Species Operations, Department of Environment and Science, Queensland Government, Australia
d
Agriculture and Food, Commonwealth Scientific and Industrial Research Organisation, Queensland Bioscience Precinct, St Lucia, QLD 4067, Australia
e
Water Quality and Investigation, Science and Technology Division, Department of Environment and Science, Queensland Government, Australia
HIGHLIGHTS
•First ecosurveillance assessmentof PFAS
mixtures on freshwater turtles.
•First report of FBSA and FHxSA in
Australian freshwater turtles.
•Bioaccumulation of PFOSin turtle serum
235 times higher than water concentra-
tions.
•Positive PFAS correlation with purine/
lipid metabolism related to immune re-
sponses.
•Negative PFAS correlation with lipid
transport and binding activity.
GRAPHICAL ABSTRACT
abstractarticle info
Article history:
Received 17 August 2021
Received in revised form 14 October 2021
Accepted 22 October 2021
Available online xxxx
Editor: Adrian Covaci
PFAS mixtures in the environment are common and identifying PFAS constituents, bioaccumulation, and biolog-
ical impacts of mixtures remains a challenge. Here, an omics-based ecosurveillance approach was taken to
investigate the impacts of PFAS pollution in freshwater turtles (Emydura macquarii macquarii). Four turtles
were collected from an impacted waterway downstream from an industrial source of PFAS contamination in
Queensland, Australia and analysed for 49 different PFAS. One turtle was collected from a suitable control site.
PFAS concentrations were quantified in turtle serum using an established targeted methodology. The serum
PFAS concentration was ten-fold greaterat the impacted site (Σ49 PFAS 1933 ± 481 ng/mL) relative to thecon-
trol sample (Σ49 PFAS 140 ng/mL). Perfluorooctane sulfonate (PFOS; 889 ± 56 ng/mL) was 235 times higher in
turtle serum than in the water that they were collected from (ΣPFAS 32.0 μg/L). Perfluorobutane sulfonamide
(FBSA; 403 ± 83 ng/mL) and perfluorohexane sulfonamide (FHxSA; 550 ± 330 ng/mL) were also reported at
substantial concentrations in the serum of impacted turtles. Biochemical profiles were analysed using a mixture
of liquid chromatography triple quadrupole (QqQ) and quadrupole time-of-flight (QToF) mass spectrometry
methodologies. These profiles demonstrated a positive correlation in the impacted turtles exposed to elevated
PFAS with an enhanced purine metabolism, glycerophosphocholines and an innate immune response, which
suggest an inflammation response, metabolic preservation and re-routing of central carbon metabolites.
Conversely, lipid transport and binding activity were negatively correlated. Using these preliminary data, we
were able to demonstrate the negative metabolic impact from PFAS mixtures on turtle metabolic health. With
Keywords:
Environmental exposure
Ecosurveillance
PFOS
PFAS
Multi-omics
Metabolomics
Proteomics
Lipidomics
Metabolic health
Science of the Total Environment xxx (xxxx) xxx
⁎Corresponding author at: Environmental Systems Biology, Land and Water, CSIRO, Australia.
E-mail address: david.beale@csiro.au (D.J. Beale).
STOTEN-151264; No of Pages 10
https://doi.org/10.1016/j.scitotenv.2021.151264
0048-9697/Crown Copyright © 2021 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Contents lists available at ScienceDirect
Science of the Total Environment
journal homepage: www.elsevier.com/locate/scitotenv
Please cite this article as: D.J. Beale, K. Hillyer, S. Nilsson, et al., Bioaccumulation and metabolic response of PFAS mixtures in wild-caught
freshwater turtles (Emydura ..., Science of the Total Environment, https://doi.org/10.1016/j.scitotenv.2021.151264
further research on a larger turtle cohort, omics-based data will contribute towards linking adverse outcome
pathways for turtle populations exposed to PFAS mixtures. Moreover, expanding the use of ecosurveillance
tools will inform mechanistic toxicological data for risk assessment and regulatory applications.
Crown Copyright © 2021 Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
1. Introduction
Per-andpolyfluoroalkyl substances (PFAS) encompass a large, het-
erogeneous group of chemicals of potential concern to human health
and the environment (Ankley et al., 2021). Data suitable for determin-
ing risks of adverse effects are lacking for the majority of PFAS com-
pounds, for both prospective and retrospective assessments. To date,
most of the biological impacts of PFAShave been delineated in the lab-
oratory, using simple PFAS mixtures, exposure concentrations beyond
environmental relevance and in model organisms (Lee et al., 2017;
Mishra et al., 2017). Compounding this further is the uncertainty in eco-
logical risk assessments of PFAS chemical mixtures, in terms of PFAS
identification and biological impact (Naidu et al., 2020), with evidence
emerging of synergistic effects of PFOS together with other PFAS (Ojo
et al., 2020). PFAS mixtures are often introduced to the environment
as end-use products, industrial waste/by-products and discharged
wastewater treatment effluents. In Australia, aqueous film-forming
foam (AFFF) formulations (i.e., firefighting foam), have been identified
as a significant source of PFAS in the environment, and these products
typically contain several PFAS such as perfluorooctane sulfonate
(PFOS), perfluorooctanoic acid (PFOA), perfluorohexane sulfonate
(PFHxS), and 6:2 fluorotelomer sulfonate (6:2 FTSA) (Barzen-Hanson
et al., 2017). It is then no surprise that multiple discrete PFAS are rou-
tinely detected in aquatic environments, groundwater systems, and
wastewater treatment effluents. Assessing and predicting the relative
ecological risk of PFAS contaminants and any transformation products
within the environment poses a significant challenge (Naidu et al.,
2020;Phillips et al., 2007).
The effect of pollutants, such as PFAS, on metabolic systems for
aquatic species remains unclear. For example, Morris et al. (2019)
assessed the potential influences of contaminant exposure on the he-
patic metabolome of male polar bears in Canada. While this was a
multi-stressor exposure study in wild mammals that included a variety
of contaminants such as PFAS, pesticides, pharmaceuticals etc., it was
unable to conclusively tie observations to specific contaminant classes.
However, correlations between fatty acid uptake, processing and me-
tabolism with increased bioaccumulated contaminant levels were ob-
served (Morris et al., 2019). Similarly, PFAS effected the endocrine
systems in prawns (Metapenaeus macleayi) by disrupting fatty acid
and lipid metabolism homeostasis resulting in fatty- and amino acid
level alterations (Taylor et al., 2019). Zhang et al. (2020a) also studied
the impacts of a 60-day PFOA exposure (0.05–5.0 mg/kg) experiment
on lizards (Eremias argus). Therein the female lizards exhibited an in-
creased body weight, stronger immune response, and reduced egg
mass that suggested a metabolic trade-off towards self-maintenance.
Whereas, for the male lizards, reduced body weight and a decreased
survival rate were observed, indicative of oxidative damage (Zhang
et al., 2020a). Surprisingly, there is currently an international paucity
of information relating to the impacts of PFAS on freshwater turtles,
and reptiles in general, with fish and mammals being the primary area
of biological focus. In this respect, turtles and other reptiles have a
very important role in the food chain, as predators, scavengers
and prey, making them an important candidate to assess the
biomagnification and impact within an ecosystem structure
(Kannan et al., 2005).
Although the impacts from PFAS on reptiles are poorly studied, re-
cent investigations have focussed on marine turtles. Bioaccumulation
of PFAS in marine turtle blood has been reported in previous studies
(de Solla et al., 2012;Guerranti et al., 2013;Kannan et al., 2005;Keller
et al., 2012;Morikawa et al., 2006), but the correlation between PFAS
concentrations and metabolic alterations remains elusive. Impacts that
have been associated with elevated PFAS levels in these turtles include
maternal offloading of PFAS into eggs along with reduced emergence
success of hatchlings (Wood et al., 2021) and negative correlations be-
tween body size and PFAS exposure (Bangma et al., 2019). Although
not described in turtles, alterations in metabolic processes associated
with PFAS have been reported broadly in humans (Chen et al., 2020;
Fan et al., 2020;Kobayashi et al., 2021) and other animals (Dale et al.,
2020;Ortiz-Villanueva et al., 2018;Pfohl et al., 2021;Seyoum et al.,
2020;Taylor et al., 2019;Zhang et al., 2020b). These alterations can be
assessed precisely using metabolomics, which is an approach that can
monitor and rapidly screen biochemical perturbations in an organism
and help to elucidate the biological impacts of toxicants (Ankley et al.,
2021).
The aim of this study was to demonstrate to environmental regula-
tors and stakeholders the extent of PFAS bioaccumulation and qualita-
tive metabolic disturbances within impacted turtles. To achieve this
aim, we used an omics-based ecosurveillance approach (which couples
measuredbiological metabolism data with quantitative bioaccumulated
contaminant measurements) to assess the metabolic health impact of
PFAS on freshwater turtles. Omics within this context is focused on
the functional biological outputs (i.e., metabolomics, lipidomics and
proteomics), which is closest to the impacted phenotype related to
PFAS exposure (Pinu et al., 2019). Such an approach is critical for iden-
tifying key biological activities associated with less well-characterized
PFAS and PFAS mixtures in the environment (Ankley et al., 2021). In
this study, common freshwaterturtles were collected from a PFAS con-
taminated water body that was identified downstream of an industrial
site within Queensland in 2021 (site location is confidential and thus
unspecified). This short communication presents the results of a prelim-
inary study of the metabolic impact on freshwater turtles and elevated
levels of PFAS mixtures in the waterway. To the best of our knowledge,
no studies have been undertaken on freshwater turtles in Australia, and
the actual measured biological effects and accumulation of PFAS on
wild-caught turtle metabolism and functional health. Furthermore,
here we present preliminary data on how PFAS exposure can impact
the turtle population and their metabolic health, which necessitates
larger-scale ‘omics’-based studies to understand the contaminant expo-
sure and impact on various stages of the turtle lifecycle.
2. Material and methods
2.1. Animal ethics
Freshwater turtles (n= 5) were captured, and an aliquot of blood
was collected for a preliminary investigation into the impacts of ele-
vated PFAS exposure and its bioaccumulation in exposed turtles. This
was conducted as per the Queensland Department of Environment
and Science (DES) Wildlife and Threatened Species animal population
dynamics survey (AEC permit SA 2018/11/663).
2.2. Study site and turtle sample collection
The study site in question was an undisclosed waterway down-
stream of an industrial source of PFAS contamination in Queensland,
Australia (the impact site). A risk-based monitoring approach was un-
dertaken that considered known contaminant sources and associated
industrial activities (present and in the past). In doing so, PFAS was
D.J. Beale, K. Hillyer, S. Nilsson et al. Science of the Total Environment xxx (xxxx) xxx
2
identified as the chief contaminant of concern. The site was also
screened for metals, which were found to be negligible. A previous un-
published study of fish, macroinvertebrates, emergentplants, algae, and
duckweedfrom the impact site confirmed perfluorooctanesulfonic acid
(PFOS) at concentrations that exceeded the current dietary intake
guidelines for the protection of mammalian and avian wildlife, which
is set at 4.6 μg/kg (sum of PFOS and PFHxS, wet weight) for a mamma-
lian diet and 8.2 μg/kg for an avian diet (Heads of EPA Australia and New
Zealand, 2020) (Supplementary Tables S1 and S2 provides supporting
biota PFOS concentration data). The omnivorous freshwater turtle
(Emydura macquarii macquarii), which are a common short-necked tur-
tle species to Australia, were observed at the site. The biota data sug-
gests a food source that exceeded the PFAS dietary threshold and was
likely a key mechanism for high levels of PFAS to enter the turtle popu-
lation. As such, a preliminary investigation into the bioaccumulation of
PFAS and the utility of omics based ecosurveillance tools to assess the
metabolic impact of elevated PFAS was undertaken. As per the limita-
tions of the population dynamics permit, only four turtles were
captured from the impact site and one turtle from a secondary back-
ground (control) site. The control site location was selected as per the
definition stated in ANZECC and ARMCANZ (2018). That being, the
site was similar in all relevant respects to the impact site except for
the activities being undertaken.
In July 2021, three cathedral nets were set overnight at each site
as described in Hamann et al. (2007). All turtles were transported
to the Ecosciences Precinct Turtle Laboratory in an enclosed
blacked-out tub. In the laboratory, 2 mL of whole blood was collected
according to methods described in Rogers and Booth (2004) and
Todd et al. (2013). The blood samples were collected and split into
two 1 mL aliquots, transferred directly from the needle to serum sep-
arator tubes (SST), then centrifuged at 3000 rcf for15minat4°C.
One sample was transported immediately to Queensland Alliance
for Environmental Health Sciences (QAEHS) research laboratory
where samples were stored in a freezer (−20 °C) for seven days be-
fore PFAS analysis. The second sample was taken immediately to the
CSIRO Environmental Systems Biology Laboratory at the Ecosciences
Precinct for functional omics analysis and analysed within 24 h of
collection.
2.3. PFAS analysis in turtle serum
Serum (200 μL) was extracted for PFAS analysis using a method de-
scribed by Toms et al. (2019). Targeted PFAS analysis was performed
using high-performance liquid chromatography (HPLC, Nexera, Shimadzu
Corp., Kyoto Japan), coupled to a tandem mass spectrometer (SCIEX Triple
Quad 6500+, Concord, Ontario, Canada)equippedwithanelectrospray
ionisation source and using scheduled multiple reaction monitoring
mode (sMRM). Quantification was performed using the isotope dilution
method. Both linear and total (linear + branched isomers) concentrations
were quantified for PFOS and PFHxS. Linear isomers were quantified for
all other compounds. Two samples were extracted and analysed as dupli-
cates, a standard reference material (SRM NIST 1957), three pooled serum
samples (one non-spiked, one spiked before extraction and one spiked
after extraction before analysis) were extracted and analysed alongside
the serum samples to control for precision and accuracy. Two procedural
blanks (MilliQ® and Acetonitrile) ensured no contamination. The method
detection limit (MDL) for all analysed PFASs ranged from 0.18 to
1.5 ng/mL. The Supplementary Information presents further details on
the extraction, analysis, and quality control/quality assurance results.
2.4. Functional omics analysis (proteomics, lipidomics, and metabolomics)
Metabolites, lipids and proteins were extracted from collected
serum using a one-pot extraction method (Fig. 1). This process offers
an opportunity to maximise the biological information from the one
sample, eliminating the need to collect multiple samples per workflow.
Briefly, 100 μL of serum was combined with 450 μLofice-cold(−20 °C)
methanol:ethanol (50% v/v; LiChrosolv®, Merck, Darmstadt, Germany),
and vortexed for 2 min. The samples were centrifuged (Centrifuge
5430R, Eppendorf, Hamburg, Germany) at 14,000 rcf at 4 °C for 5 min.
The supernatant was transferred and filtered using a positive pressure
manifold (Agilent PPM48 Processor, Agilent Technologies, Santa Clara,
Fig. 1. Overview of the functional omics workflow for the analysis of PFAS exposed turtle serum samples.
D.J. Beale, K. Hillyer, S. Nilsson et al. Science of the Total Environment xxx (xxxx) xxx
3
California, USA) with Captiva EMR cartridges (40 mg, 1 mL; Agilent,
Mulgrave, VIC, Australia) to remove the lipid fraction. The cartridges
were then washed with two 200 μL aliquots of MilliQ:methanol:ethanol
(50/25/25% v/v/v). The combined filtered supernatant and cartridge
washes, representing the serum polar metabolite fraction, were
combined in a 1.5 mL high recovery vial (30 μL reservoir, silanized
glass vials, Agilent Technologies, Mulgrave, Australia) and dried in a
SpeedVAC (10 mBar). The retained lipid fraction on the Captiva EMR
cartridges was eluted with two 500 μL aliquots of dichloromethane:
methanol (33:67% v/v; Sigma-Aldrich, Mulgrave, Victoria, Australia).
The lipid fractions were combined in high recovery vials and dried
under a stream of nitrogen (RATEK-Evaporation manifold, AdelLab
Scientific, Thebarton, South Australia). The centrifuged protein pellet
was collected and then prepared for proteome measurement. The
polar metabolite fraction was reconstituted with 100 μL MilliQ:metha-
nol (80/20% v/v) and the lipid fraction was reconstituted in 100 μL
butanol:Methanol (50/50% v/v). Internal standard set #1 comprised
100 ppb of L-Phenylalanine (1-
13
C) and L-Glutamine (amide-
15
N);
Internal standard set #2 comprised 200 ppb of Succinic Acid
(1,4-
13
C
2
) and DL-Homocysteine-D4. Internal standards were sourced
from Cambridge Isotope Laboratories (Andover, MA, USA). The residual
relative standard deviation (RDS%) of the internal standards were 3.16%
(L-Phenylalanine, 1-
13
C), 2.63% (L-Glutamine, amide-
15
N), 1.98%
(Succinic Acid, 1,4-
13
C
2
), and 3.46% (DL-Homocysteine-D4).
Central carbon metabolism metabolites were measured on an
Agilent Infinity Flex II UHPLC coupled to an Agilent 6470 Triple
Quadrupole Mass Spectrometer (QqQ-MS) following Sartain (2016)
and Gyawali et al. (2021); untargeted lipids and metabolites were
analysed on an Agilent Infinity Flex II UHPLC coupled to an Agilent
6546 Quadrupole Time-of-Flight Mass Spectrometer (QToF-MS) follow-
ing Shah et al. (2021) and Beale et al.(2021); turtle serumproteins were
extracted and analysed via SWATH-MS-based proteomics on an
SCIEX 6600 TripleToF-MS modified from Bose et al. (2020) and
Bose et al. (2021). The Supplementary Information presents
further details on the extraction, analysis and reported QAQC of
metabolites, lipids, and proteins. All samples were extracted and
analysed in triplicate, with exception to Sample 2 and 4 which
were analysed in duplicate.
2.5. Data analysis
The ‘omics’datasets (metabolites, lipids and proteins) were log-
transformed and multivariate data analysis conducted using SIMCA
(v17.0.01, Sartorius Stedim Biotech, Umeå, Sweden) and MetaboAnalyst
5.0 (Pang et al., 2021). Functional omics outputs were enriched using
MetaboAnalyst 5.0 to further explore the contribution of measured bio-
molecules to corresponding metabolic pathways, which then facilitated
a pathway impact assessment (i.e., its criticality in ensuring pathway
expression). Chemical clusters based on structuralsimilarities were cre-
ated for metabolic examination using the ChemRICH analysis (Barupal
and Fiehn, 2017). Due to the sampling constraints, the data lacked ade-
quate biological replication of the control samples. Therefore, qualita-
tive statistical analysis of metabolome, lipidome and proteome was
undertaken using a fold change of ≥1.5–2.0 (De Livera et al., 2013)and
a Benjamini–Hochberg adjusted p-value of ≤0.05 (Pang et al., 2021)of
the enriched outputs for discussion. The fold change in this instance de-
scribes the difference (increased or decreased) between the impacted
and control turtle groups.
3. Results and discussion
3.1. PFAS concentrations at the sites
Water samples from the two sites were measured for PFAS by an
Australian National Association of Testing Authorities (NATA) accredited
commercial laboratory using a different methodology reported herein
(method not provided). The water samples were collected as part of
a site investigation and the total PFAS (n= 28) concentration at
the impacted site was reported to be 32.0 μg/L; n= 6. For the control
site, PFAS was significantly lower (0.011 μg/L; n=3).Thiswasabove
the limit of reporting stated by the commercial laboratory (Supple-
mentary Table S4). The dominant PFAS in the impacted waters
were (PHHxA > PFHxS > PFOS), where the control site it was
PFOS > PFOA > PFHxA. Water samples were also screened for metals,
which were found to be negligible. Supplementary Table S3 provides
additional water quality data collected.
3.2. The physical condition of collected turtles
The collected turtles across both sites were all females and within
2109 ± 395 g total body weight, with a straight carapace length (SCL)
of 26.00 ± 1.84 cm, and tail-to-carapace length of 2.53 ± 0.57 cm.
This is consistent with the size and weight of adult freshwater turtles
for this species. All turtles wereconsidered ‘healthy’on visual inspection
and examination (no abnormal growths or irregularities). Table 1 pro-
vides a summary of the turtle physical measurements at the time of
sampling, in addition to measuring the serum total lipid concentration
(Cheng et al., 2011).
3.3. Turtle PFAS serum concentrations
ThePFASserumprofile and concentrations of detected PFAS in the im-
pacted and control cohorts is presented in Fig. 2 and Supplementary
TableS5.Theserumofthecontrolturtle(n= 1) had a total PFAS concen-
tration (Σ49 PFAS) of 167 ng/mL. Serum from the impacted turtles (n=
4) was observed at levels over ten-fold greater than the control, with a Σ
49 PFAS level of 1933 ± 481 ng/mL (ranging from 1499 to 2461 ng/mL).
Dominant PFAS in the serum from the impacted turtles were PFOS (total
linear + branched; 89 ± 56 ng/mL), PFHxS (total linear + branched;
48 ± 6 ng/mL), perfluoroheptane sulphonate (PFHpS; 12 ± 3 ng/m),
perfluorodecanoic acid (PFDA; 5 ± 1 ng/mL), perfluorobutane sulfon-
amide (FBSA; 403 ± 83 ng/mL), perfluorohexane sulfonamide (FHxSA;
550 ± 330 ng/mL) and perfluorooctane sulfonamide (FOSA; 18 ±
9 ng/mL). Interestingly, PFOS was the dominant PFAS in the control sam-
ple with a concentration of 158 ng/mL It should be noted that the water
analysis did not include FHxSA or FBSA. Amongst the other PFAS analysed,
only PFDA (6.4 ng/mL) and perfluoroundecanoic acid (PFUnDA;
1.5 ng/mL) were detected at concentrations >1 ng/mL in the control sam-
ple (PFHxS, PFHpS and Perfluorotetradecanoic acid (PFTreDA) were de-
tected in concentrations ranging from 0.32–0.87 ng/mL). Furthermore,
of the other dominant PFAS in the analysed water samples, only PFHxS
was in abundance in the impacted (48.0 ± 6.5 ng/mL) and control
(0.9 ng/mL) turtle serum samples.
PFOS being the dominant PFAS in the turtle serum is consistent with
previousstudies that have reported blood PFAS concentrations in turtles
(Bangma et al., 2019;Kannan et al., 2005;Morikawa et al., 2006). The
concentrations of PFOS detected in the freshwater turtles herein are rel-
atively high, both in the impacted and control site collected turtles,
Table 1
Physical measurements of the sampled turtles.
Characteristic Control site Impacted site
Sample ID 1 2345
Sex Female Female Female Female Female
Age class Adult Adult Adult Adult Adult
Weight (g) 1985 1940 2715 1925 1980
Carapace (SCL; cm) 26.25 25.77 28.43 24.75 24.78
Tail-to-carapace (cm) 2.2 3.1 2.0 2.9 2.4
Total serum lipids
(mg/mL; n = 4)
10.2 ± 2.1 6.4 ± 0.5 3.4 ± 0.3 4.6 ± 0.2 4.0 ± 0.5
Note: SCL is defined as ‘straight carapace length’.
D.J. Beale, K. Hillyer, S. Nilsson et al. Science of the Total Environment xxx (xxxx) xxx
4
compared to other reports on the freshwater turtle. However, higher
blood PFAS concentrations have been reported in snapping turtles
(Chelydra serpentina), downstream from an international airport
(2223 ng/mL plasma with standard error 247 ng/mL) (de Solla et al.,
2012). In the same study, turtles from a control site had plasma PFOS
levels ranging from 9 to 171 ng/mL, which is comparable to the PFOS
levels in the control turtle in the current study. Further, to our knowl-
edge, this is the first study to report the perfluoroalkane sulfonamide
(FASA) constituents FBSA and FHxSA in the blood serum of turtles,
with only a few studies having reported these compounds in biota
(Chu et al., 2016). FBSA has previously been detected in fish tissue
(Baygi et al., 2021;Chu et al., 2016) and was reported in Australian sur-
face water for the first time earlier this year (Marchiandi et al., 2021).
FHxSA has previously been reported in surface water, soil and ground-
water at AFFF impacted sites (Backe et al., 2013;Martin et al., 2019;
McGuire et al., 2014;Nickerson et al., 2021), but to our best knowledge,
no studieshave reported the presenceof FHxSA in biota. However, it is
possible that the limited reporting of these compounds in biota is a re-
sult of the analytical approach used and the recent expansion of PFAS
constituents in targeted analysis. In the impacted turtles investigated
in this study, FBSA and FHxSA together make up to nearly 50% of the
total PFAS concentration. This suggests that these FASAs are present in
the environment and have the potential to bioaccumulate, and thus
should be considered in future environmental monitoring of this site
and others. This supports Munoz et al., who observed C8 sulfonamide
bioaccumulated at a greater rate than its metabolite PFOS in fish and a
broad range of invertebrates (Munoz et al., 2017). To link the analysed
PFAS from the environment to measured constituents in the turtles, a
consistent suite of target analytes needs to be applied across the full
complement of sample types to ensure PFAS measurements are compa-
rable (in terms of the total number of PFAS constituents and reportable
limits). Otherwise, bioaccumulation and transformation assessments
will only be achieved on the common PFAS constituents across each
method.
3.4. Biochemical profiling of PFAS exposed turtles
The functional biochemical profile of the analysed turtle serum
yielded 133 central carbon metabolism metabolites, 210 annotated
polar metabolites, 332 annotated lipids and 102 annotated proteins.
Central carbon metabolism metabolites are a major energy source of
all living organisms (Eylem et al., 2021), while the polarmetabolites col-
lected viathe QToF analyses representing additional associated aqueous
metabolites. The lipid fraction represents the non-polar complement
Fig. 2. PFAS serum concentrations inthe blood serum of collected freshwater turtles from animpacted site downstream of a PFAS contamination source (n= 4) and a control site(n=1)
within Queensland, Australia. PFAS constituents were analysed via targeted LC-MS/MSusing scheduled multiple reaction monitoring mode (sMRM) and authentic labelled standards.
D.J. Beale, K. Hillyer, S. Nilsson et al. Science of the Total Environment xxx (xxxx) xxx
5
(Ebshiana et al., 2015). the Of these, 32 central carbon metabolism metab-
olites and 64 polar metabolites were significantly elevated or depleted
(i.e., FC ≥2.0) in the impact group with reference to the control sample,
representing a broad range of metabolite classes such as purine metabo-
lism (e.g., allantoin, uric acid, and adenosine), leucine-isoleucine-valine
metabolism (e.g., 3-hydroxyisovaleric acid, ketoleucine, isoleucine, and va-
line), guanine nucleotides (e.g., dGTP, dGDP, guanosine diphosphate, and
adenylsuccinic acid), pyrimidine nucleotides (e.g., dTDP, CDP, and dCTP),
uracil nucleotides (e.g., uridine diphosphate glucose, uridine triphosphate,
deoxyuridine triphosphate), amongst others. For the lipids, 115 com-
pounds were found to be significantly elevated or depleted in the impact
group with reference to the control sample, representing such lipid classes
as glycerophosphocholines (n= 48), sphingomyelins (n= 20),
triacylglycerolceramides (n= 10), and glycerophosphoethanolamines
(n = 4). Supplementary Figs. S1, S2 and S3 provide a summary of all sig-
nificant central carbon metabolism metabolites and annotated lipids. Vol-
cano plots (Fig. 3) summarise significant fold changes in the measured
biochemical data. These significantly differentiated metabolites and lipids
were further subjected to pathway impact analysis (Fig. 4), in which py-
rimidine metabolism (p-value < 0.001), purine metabolism (p-
value < 0.001), glycerophospholipid metabolism (p-value = 0.021),
linoleic acid metabolism (p-value = 0.031), and valine-leucine-
isoleucine biosynthesis (p-value = 0.031) were deemed significantly
enriched and impacted. Other important pathways that were enriched
but not identified as significantly perturbed include glutamate metabolism
(p-value = 0.0561), starch and sucrose metabolism (p-value = 0.064), ga-
lactose metabolism (p-value = 0.104), and glutathione metabolism (p-
value = 0.130). While these pathways did not meet the p-value threshold
assigned (p-value ≤0.05), they do represent pathways with sufficient
effect size differences (based on fold change data) that warrants mention-
ing for further investigations.
Interestingly, prenatal exposure of elevated PFAS in humans resulted
in a higher risk of liver injury and increased serum levels of branched-
chain amino acids (valine, leucine, and isoleucine), aromatic amino acids
(tryptophan and phenylalanine), and glycerophospholipids citrate cycle,
and purine metabolism (Stratakis et al., 2020). These pathways were like-
wise significantly enriched and impacted in the PFAS exposed turtle sam-
ples (impact site). Exposure to PFOA/PFOS has previously been shown to
increase fatty acid metabolism, glutathione, ascorbate, purine, and TCA
cycle metabolism (Kingsley et al., 2019;Li et al., 2020). These data demon-
strated an altered metabolic activity in key pathways relating to metabolic
health, which if coupled with the ‘healthy’physical appearance of the tur-
tles, suggests an inflammation response, metabolic preservation and re-
routing of central carbon metabolites (Jiang et al., 2015). This re-routing
could result in metabolic deficiencies or perturbations in other parts of
the body (such as liver, kidney, heart, brain, and reproductive organs)
that, if left uncorrected, could lead to a decline in physical condition and
necessitates further research to establish such linkages.
Further, the proteomics data revealed 24 proteins that were signifi-
cantly elevated or depleted (FC ≥1.5) in the impact group with reference
to the control sample, which related to antioxidant activity (−3.5 FC), en-
dopeptidase inhibitor activity (−2.7 FC), lipid transporter activity (−1.9
FC), phospholipid binding (−1.6 FC), calcium ion binding (1.7 FC), ATP
binding (1.7 FC), and innate immune response activity (1.9 FC), amongst
others. Fig. 3(G) provides an overview of these important proteins (nor-
malised fold change data), while Supplementary Fig. S4 provides an over-
view of the annotated and significantly altered proteins. Interestingly,
protein M7BHM5 (UniProt ID; biological function annotation of
Fig. 3. Volcano plot of significantly depleted (red circles) and elevated (green circles) biomolecules based on the functional omics analyses performed. (A) central carbon metabolism
metabolites analysed by LC-QqQ-MS; (B) polar metabolites analysed by LC-QToF-MS; (C) lipids analysed by LC-QToF-MS; and (D) proteins analysed by LC-QToF-MS. (E) Illustrates the
most enriched pathways from the metabolite datasets, while (F) illustrates the most enriched pathways from the lipid dataset and (G) illustrates the impacted proteins based on
normalised fold change values. Note, for thecentral carbon metabolism metabolites, polar metabolites, and lipids, the fold change threshold for significant features was set at FC ≥2;
for proteins the fold change threshold was set at FC ≥1.5. The red coloured protein bars indicate a negative fold change in the impacted turtles with reference to the control, while
yellow protein bars indicate a positive fold change in the impacted turtles with reference to the control. (For interpretation of the references to colour in this figure legend, the reader
is referred to the web version of this article.)
D.J. Beale, K. Hillyer, S. Nilsson et al. Science of the Total Environment xxx (xxxx) xxx
6
superoxide dismutase) was observed 3.06 times higher in the impacted
turtles compared to the control. This protein has been observed to be el-
evated post PFOA and PFOS exposure in mouse cell lines, leading to al-
tered antioxidant enzyme activities and oxidative stress response (Xu
et al., 2019). A positive correlation between PFAS and measured proteins
related to innate immune response (UniProt ID A0A674I4V8) was also
observed in the turtle serum samples. A similar trend has been found in
shellfish (Bernardini et al., 2021) and mice (Wang et al., 2021). This
may also account for an increase in the protein M7BPB9 (hyaluronan
metabolic process), which is tied to immune response modulation in
invertebrates (Erickson and Stern, 2012). Hydrolase activity (on carbon-
nitrogen; M7B8P9 in Supplementary Fig. S4) was also elevated in the im-
pacted turtle serum and has been linked to an increase in activity within
liver tissues due to enzyme induction when exposed to PFOA
(Kawashima and Kozuka, 1992). PFOS has also been linked to disturbance
in calcium homeostasis (M7AWD1, Supplementary Fig. S4) by inducing
extracellular calcium influx and intracellular calcium release, as observed
here in the turtle cohort and previously in PFAS clinical observation stud-
ies (Zeng et al., 2019). Conversely, antioxidant, lipid transport (M7ANL9
and A0A4D9DRN0) and phospholipid-binding activity (K7FHQ5) were
negatively correlated in the impacted turtle samples; a similar trend
was previously observed in frogs (Foguth et al., 2019) and mice (Wang
et al., 2014). Here it was hypothesised that PFAS may block the export
of lipids from the liver to peripheral tissues (Wang et al., 2014), which
may account for the observed depletion of binding and transportation-
related proteins in the impacted turtle serum. Foguth et al. (2019) also
measured a significant elevation of phospholipids in frog brain tissues
post-PFAS exposure. These phenomena warrant further study in these
PFAS exposed turtle brain and liver tissues.
4. Concluding remarks and limitations
This preliminary study has provided substantial evidence from an
omics-based ecosurveillance approach perspective for the investigation
of PFAS bioaccumulation and its impacts on turtle metabolic health.
PFAS serum profiles and metabolome composition differed between tur-
tles from the PFAS impacted site and the control site, noting the control
turtle had a significantly reduced bioaccumulated PFAS profile compared
to the impacted turtles. Despite the limited data and PFAS bioaccumulated
in the control turtle, the patterns in the functional ecosurveillance (multi-
omics) dataset concorded strongly with the findings from previous stud-
ies, across wild-caught observational studies, laboratory-controlled chal-
lenge experiments and human clinical studies. Here, we found a positive
PFAS correlation with purine/lipid metabolism related to immune re-
sponses and a negative PFAS correlation with lipid transport and binding
activity in the impacted turtles. This demonstrates the usefulness of
omics-based ecosurveillance approaches in assessing the metabolic health
of PFAS impacted turtles, where physical appearance was unchanged. The
approach here can be further extended andappliedtoassessotherecosys-
tem taxa, exposed to multiple contaminant classes. The main aim of func-
tional omics-based approaches is to measure the biological response
(phenotype), not necessarily all the compounds and chemicals that im-
pact the response.
4.1. Study limitations and benefits
It is acknowledged that the number of collected turtles is a significant
limitation of this study (with only four turtles collected from the impact
site and one from the control site). It is noted that the Metabolomics
Standard Initiative guideline documents for minimum reporting
standards suitable for biological samples set a minimum number of bi-
ological replicates to three (n= 3) per group (Sumner et al., 2007).
While we meet this requirement for the turtles collected from the im-
pact site, the number of control turtles collected did not. To overcome
this limitation and facilitate a qualitative omics assessment of measur-
able biomolecules (i.e., fold changes), all sampled turtles were analysed
in triplicate, where possible. The site was screened for metals, however,
the risk-based monitoring approach followed didnot identify any other
organic contaminants of concern. While it appears PFAS is the only
point of concern, along with elevated potassium (amongst others), the
measured metabolic response of the turtles did confirm findings from
the literature related toelevated PFAS exposure (Zeng et al., 2019). Fur-
thermore, although the serum concentrations of PFAS were greater in
the turtles from the impacted sites, high levels of PFAS (compared to
other reported studies) were also found in the control site turtle. There-
fore, more samples are needed to establish the baseline of PFAS in wild-
caught freshwater turtles within suitable control sites. It is widely spec-
ulated the PFAS has far reached into many aquatic ecosystems and accu-
mulated within living organisms where it has bioaccumulated into
higher order taxa that are transient (Teunen et al., 2021).
Associations with environmental physicochemical data were not
attempted. However, there was some association between the dominant
PFAS in the water through to the dominant PFAS in the turtle serum, not-
ing that FBSA and FHxSA were not part of the water PFAS assay targets.
Further work is needed to strengthen these associations, and exploring
additional environmental factors and industrial contaminants not
Fig. 4. Significant metabolic pathways identified via enrichment and impact analysisof (A) metabolites and (B) lipids analysed from turtleserum using MetaboAnalyst 5.0 (Enrichment
Analysis and Pathway Impact Toolbox). Metabolites and lipids were selected for enrichment based on elevated or depleted fold change (FC) ≤2.0. M-1 represents “Valine, leucine and
isoleucine biosynthesis”; M-2 represents “Purine metabolism”; M-3 represents “Taurine and hypotaurine metabolism”; M-4 represents “Aminoacyl-tRNA biosynthesis”;M-5
represents “Pyrimidine metabolism”; L-1 represents “Glycerophospholipid metabolism”; L-2 represents “Linoleic acid metabolism”; and, L-3 represents “Glycosylphosphatidylinositol
(GPI)-anchor biosynthesis”impacted pathways.
D.J. Beale, K. Hillyer, S. Nilsson et al. Science of the Total Environment xxx (xxxx) xxx
7
captured in the risk-based monitoring program thus far, which are known
to affect metabolic composition (Morrison et al., 2007). These associations
will be considered in future investigations. Further, the turtles collected
for the present study are from the same geological region, yet there
were differences in the water chemistry between the two sites, in partic-
ular total water hardness, calcium, magnesium, sodium, and potassium,
which were an order(s) of magnitude higher at the impacte d site; Supple-
mentary Table S3. While these water chemistries may impact the turtle
metabolic measurements, the majority of impacted pathways were relat-
able to PFAS exposure studies (Jiang et al., 2015). Future studies will aim
to identify suitable controls and multiplePFASimpactsitestoaccountfor
such non-PFAS related variations. It should also be noted that turtles are
transient animals. They are known to be broadly connected through
inter-pond movements in rural and agricultural settings (Bowne et al.,
2006) and therefore, they will move from waterway-to-waterway if the
local conditions are not favourable, in terms of biota and water chemistry
(i.e., saline water).
5. Future research
Due to the limitations of the current study outlined above, such as the
limited sample size, no definitive associations can be established. How-
ever, biochemical metabolic correlations as result of elevated PFAS expo-
sure were observed that support findings reported in model and non-
Fig. 5. Proposed strategy to comprehensively analyse PFAS bioaccumulation and its impact on freshwater turtle populations.
D.J. Beale, K. Hillyer, S. Nilsson et al. Science of the Total Environment xxx (xxxx) xxx
8
model PFAS exposure studies. This demonstrates the utility of omics-
based ecosurveillance tools in assessing the metabolic impacts of contam-
inants in concert with conventional monitoring approaches; coupling
datasets on exposure thresholds with metabolic health impacts.However,
notwithstanding these benefits, this study highlights the need for further
research on the overall metabolic function in turtles and the PFAS contri-
bution/perturbation to individual organ function (such as liver, kidney,
brain, and reproductive organs). Thus, unraveling the mechanism of in-
teractions and impacts of PFAS on turtle organs and tissues, and endocrine
components throughout their lifecycle. Through this preliminary data, we
have demonstrated to the regulator and other stakeholders the power of
omics-based ecosurveillance approaches in assessing the metabolic
health impact of elevated PFAS exposure that facilitated an extension of
our ethics permits to now encompass three interlinked and detailed ex-
periments. These experiments are conceptually presented in Fig. 5.
As illustrated in Fig. 5, the proposed ecosurveillance strategy for
assessing the elevated PFAS exposure impacts on freshwater turtles in-
cludes the following project elements: (1) an assessment of PFAS bioaccu-
mulation and its metabolic impact on adult male and female freshwater
turtles. This metabolic assessment will include analyses of turtle serum
metabolites, lipids and proteins coupled with PFAS screening. Additional
analyses of dissected liver, kidney, brain tissues, reproductive organs,
and fatty deposits coupled with conventional histology and veterinary
biochemistry assays for liver and kidney function will also be carried
out. Furthermore, these data will be correlated with associated re-
sampled biota (such as planktonic and sediment microorganisms, sedi-
ment invertebrates and aquatic plants) to establish PFAS linkages through
the turtle's food web; (2) An assessment of PFAS maternal transfer from
adult female turtles to their egg clutch. This will include PFAS serum
screening of the adult female turtles before they lay their clutch. In addi-
tion, half of the clutch will be collected and analysed for PFAS and profiled
for metabolites and lipids to assess the impact on embryo development;
and, (3) An assessment of PFAS maternaltransfertohatchlings.There-
maining 50% of eggs in Stage 2 will be taken through to hatchlings, from
the impacted and control clutches, with a physical assessment of turtle
health and condition, and a determination of clutch survival concerning
PFAS exposure. One hatchling from each clutch will be sacrificed for
deeper metabolomics, lipidomic and proteomic analysis in addition to
screening for PFAS residues that have carried through from the mother
to the eggs to the hatchlings. PFAS screening, tissue histology and conven-
tional veterinary biochemistry assays will be performed on all samples
collected to tie functional omics analysis to d efined phenotypical PFAS im-
pacts. While the proposed new samples are still relatively low, turtles
being a higher order organism within the ecosystem/food chain, we
would expect these impacts to be amplifiedasthePFASbioaccumulates.
CRediT authorship contribution statement
David J. Beale: Conceptualization, Methodology, Formal analysis, In-
vestigation, Data curation, Writing –original draft, Writing –review &
editing, Visualization, Supervision. Katie Hillyer: Investigation, Formal
analysis, Data curation, Writing –review & editing. Sandra Nilsson:
Methodology, Investigation, Formal analysis, Data curation, Writing –
original draft, Writing –review & editing, Visualization. Duncan
Limpus: Investigation, Data curation, Writing –review & editing.
Utpal Bose: Investigation, Methodology, Formal analysis, Data curation,
Writing –review & editing. James A. Broadbent: Methodology, Data
curation, Writing –review & editing. Suzanne Vardy: Investigation,
Data curation, Writing –original draft, Writing –review & editing, Su-
pervision, Project administration.
Declaration of competing interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
Acknowledgements
The authors would like to acknowledge Brenda Baddiley and
Christoph Braun for organising and undertaking the field work associ-
ated with water, biota, and turtle sampling. We would also like to thank
Mrs. Sally Stockwell (CSIRO) for carrying out the Bradford assay on the
turtle serumsamples. QAEHS gratefully acknowledgesthe financial sup-
port of the Queensland Department of Health. The Bravo Metabolomics
Workbench and the 1290 Infinity II Flex pump coupled to a 6470 LC-
QqQ-MS instrument used in this study were provided by Agilent
Technologies to the CSIRO for preparation and extraction of metabolites
and lipids,and for the determination of centralcarbon metabolism me-
tabolites. Appreciation is alsoextended to theinternal CSIRO reviewers
who provided input and comments on earliermanuscript drafts.All im-
ages and figures were created or modified using BioRender.com,with
instrumentation images supplied by Agilent Technologies.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.scitotenv.2021.151264.
References
Ankley, G.T., Cureton, P., Hoke, R.A., Houde, M., Kumar, A., Kurias, J., Lanno, R., McCarthy,
C., Newsted, J., Salice, C.J., Sample, B.E., Sepúlveda, M.S., Steevens, J., Valsecchi, S.,
2021. Assessing the ecological risks of per- and polyfluoroalkyl substances: current
state-of-the science and a proposed path forward. Environ. Toxicol. Chem. 40,
564–605.
ANZECC, ARMCANZ, 2018. Australian and New Zealand Guidelines for Fresh and Marine
Water Quality. 2021 Canberra, Australia.
Backe, W.J., Day, T.C., Zwitterionic, Field J.A., 2013. Cationic, and anionic fluorinated
Chemicals in Aqueous Film Forming Foam Formulations and Groundwater from
U.S. Military bases by nonaqueous large-volume injection HPLC-MS/MS. Environ.
Sci. Technol. 47, 5226–5234.
Bangma, J.T., Ragland, J.M., Rainwater, T.R., Bowden, J.A., Gibbons, J.W., Reiner, J.L., 2019.
Perfluoroalkyl substances in diamondback terrapins (Malaclemys terrapin) in coastal
South Carolina. Chemosphere 215, 305–312.
Barupal, D.K., Fiehn, O., 2017. Chemical similarity enrichment analysis (ChemRICH) as al-
ternative to biochemical pathway mapping for metabolomic datasets. Sci. Rep. 7,
14567.
Barzen-Hanson, K.A., Roberts, S.C., Choyke, S., Oetjen, K., McAlees, A., Riddell, N.,
McCrindle, R., Ferguson, P.L., Higgins, C.P., Field, J.A., 2017. Discovery of 40 classes of
per- and Polyfluoroalkyl substances in historical aqueous film-forming foams
(AFFFs) and AFFF-impacted groundwater. Environ. Sci. Technol. 51, 2047–2057.
Baygi, S.F., Fernando, S., Hopke, P.K., Holsen, T.M., Crimmins, B.S., 2021. Nontargeted dis-
covery of novel contaminants in the Great Lakes region: a comparison of fish fillets
and fish consumers. Environ. Sci. Technol. 55, 3765–3774.
Beale, D.J., Shah, R., Karpe, A.V., Hillyer, K.E., McAuley, A.J., Au, G.G., Marsh, G.A., Vasan, S.S.,
2021. Metabolic profiling from an asymptomatic ferret model of SARS-CoV-2 infec-
tion. Metabolites 11, 327.
Bernardini, I., Matozzo, V., Valsecchi, S., Peruzza, L., Rovere, G.D., Polesello, S., Iori, S.,
Marin, M.G., Fabrello, J., Ciscato, M., Masiero, L., Bonato, M., Santovito, G., Boffo, L.,
Bargelloni, L., Milan, M., Patarnello, T., 2021. The new PFAS C6O4 and its effects on
marine invertebrates: first evidence of transcriptional and microbiota changes in
the Manila clam ruditapes philippinarum. Environ. Int. 152, 106484.
Bose, U., Broadbent, J.A., Byrne, K., Blundell, M.J., Howitt, C.A., Colgrave, M.L., 2020. Prote-
ome analysisof hordein-null barley lines reveals storage protein synthesis and com-
pensation mechanisms. J. Agric. Food Chem. 68, 5763–5775.
Bose, U., Broadbent, J.A., Juhász, A., Karnaneedi, S., Johnston, E.B., Stockwell, S., Byrne, K.,
Limviphuvadh, V., Maurer-Stroh, S., Lopata, A.L., Colgrave, M.L., 2021. Protein extrac-
tion protocols for optimal proteome measurement and arginine kinase quantitation
from cricketAcheta domesticus for food safety assessment. Food Chem. 348, 129110.
Bowne, D.R., Bowers, M.A., Hines, J.E., 2006. Connectivity in an agricultural landscape as
reflected by interpond movements of a freshwater turtle. Conserv. Biol. 20, 780–791.
Chen, Z., Yang, T., Walker, D.I., Thomas, D.C., Qiu,C., Chatzi, L., Alderete, T.L., Kim, J.S., Conti,
D.V., Breton, C.V., Liang, D., Hauser, E.R., Jones, D.P., Gilliland, F.D., 2020. Dysregulated
lipid and fattyacid metabolism link perfluoroalkyl substancesexposure and impaired
glucose metabolism in young adults. Environ. Int. 145, 106091.
Cheng, Y.S., Zheng, Y., VanderGheynst, J.S., 2011. Rapid quantitative analysis of lipids
using a colorimetric method in a microplate format. Lipids 46, 95–103.
Chu, S., Letcher, R.J., McGoldrick, D.J., Backus, S.M., 2016. A new fluorinated surfactant
contaminant in biota: perfluorobutane sulfonamide in several fish species. Environ
Sci Technol 50, 669–675.
Dale, K., Yadetie, F., Müller, M.B., Pampanin, D.M., Gilabert, A., Zhang, X., Tairova, Z., Haarr,
A., Lille-Langøy, R., Lyche, J.L., Porte, C., Karlsen, O.A., Goksøyr, A., 2020. Proteomics
and lipidomics analyses reveal modulation of lipid metabolism by perfluoroalkylsub-
stances in liver of Atlantic cod (Gadus morhua). Aquat. Toxicol. 227, 105590.
D.J. Beale, K. Hillyer, S. Nilsson et al. Science of the Total Environment xxx (xxxx) xxx
9
De Livera, A.M., Olshansky, M., Speed,T.P., 2013. Statistical analysis of metabolomics data.
Metabolomics Tools for Natural Product Discovery. Springer, pp. 291–307.
Ebshiana, A.A., Snowden, S.G., Thambisetty, M., Parsons, R., Hye, A., Legido-Quigley, C.,
2015. Metabolomic method: UPLC-q-ToF polar and non-polar metabolites in the
healthy rat cerebellum using an in-vial dual extraction. PLoS One 10, e0122883.
Erickson, M., Stern, R., 2012. Chain gangs: new aspects of hyaluronan metabolism.
Biochem. Res. Int. 2012, 893947.
Eylem, C.C., Reçber, T., Waris, M., Kır, S., Nemutlu, E., 2021. State-of-the-art GC-MS ap-
proaches for probing central carbon metabolism. Microchem. J. 106892.
Fan, Y., Li, X., Xu, Q., Zhang, Y., Yang, X., Han, X., Du, G., Xia, Y., Wang, X., Lu, C., 2020.
Serum albumin mediates the effect of multiple per- and polyfluoroalkyl substances
on serum lipid levels. Environ. Pollut. 266, 115138.
Foguth, R.M., Flynn, R.W., de Perre, C., Iacchetta, M., Lee, L.S., Sepúlveda,M.S., Cannon, J.R.,
2019. Developmental exposure to perfluorooctane sulfonate (PFOS) and
perfluorooctanoic acid (PFOA) selectively decreases brain dopamine levels in north-
ern leopard frogs. Toxicol. Appl. Pharmacol. 377, 114623.
Guerranti, C., Ancora,S., Bianchi, N., Perra,G., Fanello, E.L., Corsolini, S., Fossi, M.C., Focardi,
S.E., 2013. Perfluorinated compounds in blood of Caretta caretta from the Mediterra-
nean Sea. Mar. Pollut. Bull. 73, 98–101.
Gyawali, P., Karpe, A.V., Hillyer, K.E., Nguyen, T.V., Hewitt, J., Beale, D.J., 2021. A multi-
platform metabolomics approach to identify possible biomarkers for human faecal
contamination in Greenshell™mussels (Perna canaliculus). Sci. Total Environ. 771.
Hamann, M., Schauble, C., Limpus, D., Emerick, S., Limpus, C., 2007. Management Plan for
the Conservation of Elseya sp. (Burnett River) in the Burnett River Catchment.
Queenslnd Environmental Protection Agency, Brisbane, QLD.
Heads of EPA Australia and New Zealand, 2020. PFAS National Environmental Manage-
ment Plan Version 2.0. Department of Agriculture Water and the Environment,
Australia.
Jiang, Q.,Gao, H., Zhang, L., 2015.Metabolic effects PFAS.In: JC, DeWitt (Ed.), Toxicological
Effects of Perfluoroalkyl and Polyfluoroalkyl Substances. Springer International Pub-
lishing, Cham, pp. 177–201.
Kannan, K., Tao, L., Sinclair, E., Pastva, S.D., Jude, D.J., Giesy, J.P., 2005. Perfluorinated com-
pounds in aquatic organisms at various trophic levels in a Great Lakes food chain.
Arch. Environ. Contam. Toxicol. 48, 559–566.
Kawashima, Y., Kozuka, H., 1992. Cytosolic long-chain acyl-CoA hydrolase, a suitable pa-
rameter to measure hepatic response to peroxisome proliferators. Toxicology 71,
151–160.
Keller, J.M., Ngai, L., Braun McNeill, J., Wood, L.D., Stewart, K.R., O'Connell, S.G., Kucklick,
J.R., 2012. Perfluoroalkyl contaminants in plasma of five sea turtle species: compari-
sons in concentration and potential health risks. Environ. Toxicol. Chem. 31,
1223–1230.
Kingsley, S.L., Walker, D.I., Calafat, A.M., Chen, A., Papandonatos, G.D., Xu, Y., Jones, D.P.,
Lanphear, B.P., Pennell, K.D., Braun, J.M., 2019. Metabolomics of childhood exposure
to perfluoroalkyl substances: a cross-sectional study. Metabolomics 15, 95.
Kobayashi, S., Sata, F., Goudarzi, H.,Araki, A., Miyashita, C., Sasaki, S., Okada, E., Iwasaki, Y.,
Nakajima, T., Kishi, R., 2021. Associations among perfluorooctanesulfonic/
perfluorooctanoic acidlevels, nuclear receptor gene polymorphisms, and lipid levels
in pregnant women in the Hokkaido study. Sci. Rep. 11, 9994.
Lee, J.W.,Lee, J.-W., Kim, K.,Shin, Y.-J., Kim, J.,Kim, S., Kim, H., Kim,P., Park, K., 2017. PFOA-
induced metabolism disturbance and multi-generational reproductive toxicity in
Oryzias latipes. J. Hazard. Mater. 340, 231–240.
Li, P., Xiao, Z., Sun, J., Oyang, X., Xie, X., Li, Z., Tian, X., Li, J., 2020. Metabolic regulations in
lettuce root under combined exposure to perfluorooctanoic acid and perfluorooctane
sulfonate in hydroponic media. Sci. Total Environ. 726, 138382.
Marchiandi, J., Szabo, D., Dagnino, S., Green, M.P., Clarke, B.O., 2021. Occurrence and fate
of legacy and novel per- and polyfluoroalkyl substances (PFASs) in freshwater after
an industrial fire of unknown chemical stockpiles. Environ. Pollut. 278, 116839.
Martin, D., Munoz, G.,Mejia-Avendaño,S., Duy, S.V., Yao, Y.,Volchek, K., Brown,C.E., Liu, J.,
Sauvé, S., 2019. Zwitterionic, cationic, and anionic perfluoroalkyl and polyfluoroalkyl
substances integrated into total oxidizable precursor assay of contaminated ground-
water. Talanta 195, 533–542.
McGuire, M.E., Schaefer, C., Richards, T., Backe, W.J., Field, J.A., Houtz, E., Sedlak, D.L.,
Guelfo, J.L., Wunsch, A., Higgins, C.P., 2014. Evidence of remediation-induced alter-
ation of subsurface poly- and perfluoroalkyl substance distribution at a former fire-
fighter training area. Environ. Sci. Technol. 48, 6644–6652.
Mishra, P., Gong, Z., Kelly, B.C., 2017. Assessing biological effects of fluoxetine in develop-
ing zebrafish embryos using gas chromatography-mass spectrometry based metabo-
lomics. Chemosphere 188, 157–167.
Morikawa, A., Kamei, N., Harada, K., Inoue, K., Yoshinaga, T., Saito, N., Koizumi, A., 2006.
The bioconcentration factor of perfluorooctane sulfonate is significantly larger than
that of perfluorooctanoate in wild turtles (Trachemys scripta elegans and Chinemys
reevesii): an ai river ecological study in Japan. Ecotoxicol. Environ. Saf. 65, 14–21.
Morris, A.D., Letcher, R.J., Dyck, M., Chandramouli, B., Cosgrove, J., 2019. Concentrations of
legacy and new contaminants are related to metabolite profiles in Hudson Bay polar
bears. Environ. Res. 168, 364–374.
Morrison, N., Bearden, D., Bundy, J.G., Collette, T., Currie, F., Davey, M.P., Haigh, N.S.,
Hancock, D., Jones, O.A.H., Rochfort, S., Sansone, S.-A., Štys, D., Teng, Q., Field, D.,
Viant, M.R., 2007. Standard reporting requirements for biological samples in metabo-
lomics experiments: environmental context. Metabolomics 3, 203–210.
Munoz, G., Budzinski, H., Babut, M., Drouineau, H., Lauzent, M., Menach, K.L., Lobry, J.,
Selleslagh, J., Simonnet-Laprade, C., Labadie, P., 2017.Evidence for the trophictransfer
of perfluoroalkylated substances in a temperate macrotidal estuary. Environ. Sci.
Technol. 51, 8450–8459.
Naidu, R.,Nadebaum, P., Fang,C., Cousins, I., Pennell, K., Conder, J., Newell, C.J., Longpré, D.,
Warner, S., Crosbie, N.D., Surapaneni, A., Bekele, D., Spiese, R., Bradshaw, T., Slee, D.,
Liu, Y., Qi, F., Mallavarapu, M., Duan, L., McLeod, L., Bowman, M., Richmond, B.,
Srivastava, P., Chadalavada, S., Umeh, A., Biswas, B., Barclay, A., Simon, J., Nathanail,
P., 2020. Per- and poly-fluoroalkyl substances (PFAS): current status and research
needs. Environ. Technol. Innov. 19, 100915.
Nickerson, A., Rodowa, A.E., Adamson, D.T., Field, J.A., Kulkarni, P.R., Kornuc, J.J., Higgins,
C.P., 2021. Spatial trends of anionic, zwitterionic, and cationic PFASs at an AFFF-
impacted site. Environ. Sci. Technol. 55, 313–323.
Ojo, A.F., Peng, C., Ng, J.C., 2020. Combined effects and toxicological interactions of
perfluoroalkyl and polyfluoroalkyl substances mixtures in human liver cells
(HepG2). Environ. Pollut. 263, 114182.
Ortiz-Villanueva, E., Jaumot, J., Martínez, R., Navarro-Martín, L., Piña, B., Tauler, R., 2018.
Assessment of endocrine disruptors effects on zebrafish (Danio rerio) embryos by
untargetedLC-HRMS metabolomic analysis. Sci. Total Environ. 635, 156–166.
Pang, Z., Chong, J., Zhou, G., de Lima Morais, D.A., Chang, L., Barrette, M., Gauthier, C.,
Jacques, P.-É., Li, S., Xia, J., 2021. MetaboAnalyst 5.0: narrowing the gap between
raw spectra and functional insights. Nucleic Acids Res. 49, W388–W396.
Pfohl, M., Marques, E., Auclair, A., Barlock, B., Jamwal, R., Goedken, M., Akhlaghi, F., Slitt,
A.L., 2021. An 'Omics approach to unraveling the paradoxical effect of diet on
perfluorooctanesulfonic acid (PFOS) and perfluorononanoicacid (PFNA)-induced he-
patic steatosis. Toxicol. Sci. 180, 277–294.
Phillips, M.M., Dinglasan-Panlilio, M.J., Mabury, S.A., Solomon, K.R., Sibley, P.K., 2007.
Fluorotelomer acids are more toxic than perfluorinated acids. Environ Sci Technol
41, 7159–7163.
Pinu, F.R., Beale, D.J., Paten, A.M., Kouremenos, K., Swarup, S., Schirra, H.J., Wishart, D.,
2019. Systems biology and multi-omics integration: viewpoints from the metabolo-
mics research community. Metabolites 9, 76.
Rogers, K.D., Booth, D.T., 2004. A method of sampling blood from Australian freshwater
turtles. Wildl. Res. 31, 93–95.
Sartain, M., 2016. The agilent metabolomics dynamic MRM database and method.Agilent
Technologies technical overview, publication number. 5991–6482EN.
Seyoum, A., Pradhan, A., Jass, J., Olsson, P.E.,2020. Perfluorinated alkyl substances impede
growth, reproduction, lipid metabolism and lifespan in Daphnia magna. Sci. Total En-
viron. 737, 139682.
Shah, R.M., Hillyer, K.E., Stephenson, S., Crosswell, J., Karpe, A.V., Palombo, E.A., Jones,
O.A.H., Gorman, D., Bodrossy, L., van de Kamp, J., Bissett, A., Whiteley, A.S., Steven,
A.D.L., Beale, D.J., 2021. Functional analysis of pristine estuarine marine sediments.
Sci. Total Environ. 781.
de Solla, S.R., De Silva, A.O., Letcher, R.J., 2012. Highly elevated levels of perfluorooctane
sulfonate and other perfluorinated acids found in biota and surface water down-
stream of an international airport, Hamilton, OntarioCanada. 39, 19–26.
Stratakis, N., Conti D, V., Jin, R., Margetaki, K., Valvi, D., Siskos, A.P., Maitre, L., Garcia, E.,
Varo, N., Zhao, Y., Roumeliotaki, T., Vafeiadi, M., Urquiza, J., Fernández-Barrés, S.,
Heude, B., Basagana, X., Casas, M., Fossati, S., Gražulevičienė, R., Andrušaitytė,S.,
Uppal, K., RRC, McEachan, Papadopoulou, E., Robinson, O., Haug, L.S., Wright, J., Vos,
M.B., Keun, H.C., Vrijheid, M., Berhane, K.T., McConnell, R., Chatzi, L., 2020. Prenatal
exposure to perfluoroalkyl substances associated with increased susceptibility to
liver injury in children. Hepatology 72, 1758–1770.
Sumner, L.W., Amberg, A., Barrett, D., Beale, M.H., Beger, R., Daykin, C.A., Fan, T.W.M.,
Fiehn, O., Goodacre, R., Griffin, J.L., Hankemeier, T., Hardy, N., Harnly, J., Higashi, R.,
Kopka, J., Lane, A.N., Lindon, J.C., Marriott, P., Nicholls, A.W., Reily, M.D., Thaden, J.J.,
Viant, M.R., 2007. Proposed minimum reporting standards for chemical analysis.
Metabolomics 3, 211–221.
Taylor, M.D., Bräunig, J., Mueller, J.F., Crompton, M., Dunstan, R.H., Nilsson, S., 2019.
Metabolomic profiles associated with exposure to per- and polyfluoroalkyl sub-
stances (PFASs) in aquatic environments. 21, 1980–1990.
Teunen, L.,Bervoets, L., Belpaire, C., De Jonge, M., Groffen, T., 2021. PFAS accumulation in
indigenous and translocated aquatic organisms from Belgium, with translation to
human and ecological health risk. Environ. Sci. Eur. 33, 39.
Todd, E.V., Blair, D., Limpus, C.J., Limpus, D.J., Jerry, D.R., 2013. High incidence of multiple
paternity in an australian snapping turtle Elseya albagula. Aust. J. Zool. 60, 412–418.
Toms, L.M.L., Bräunig, J., Vijayasarathy, S., Phillips, S., Hobson, P.,Aylward, L.L., Kirk, M.D.,
Mueller, J.F., 2019. Per- and polyfluoroalkyl substances (PFAS) in Australia: current
levels and estimated population reference values for selected compounds. Int.
J. Hyg. Environ. Health 222, 387–394.
Wang, L., Wang, Y., Liang, Y., Li, J., Liu, Y., Zhang, J., Zhang, A., Fu, J., Jiang, G., 2014. PFOS
induced lipid metabolismdisturbances in BALB/cmice through inhibition of low den-
sity lipoproteins excretion. Sci. Rep. 4, 4582.
Wang, L.Q., Liu,T., Yang, S., Sun, L., Zhao, Z.Y.,Li, L.Y., She, Y.C., Zheng, Y.Y.,Ye, X.Y., Bao, Q.,
Dong, G.H., Li, C.W., Cui, J., 2021. Perfluoroalkyl substance pollutants activate the in-
nate immune system through the AIM2 inflammasome. Nat. Commun. 12, 2915.
Wood, C., Balazs, G.H., Rice, M., Work, T.M., Jones,T.T., Sterling, E.,Summers, T.M., Brooker,
J., Kurpita, L., King, C.S., Lynch, J.M., 2021. Sea turtles across the North Pacific are ex-
posed to perfluoroalkyl substances. Environ. Pollut. 279, 116875.
Xu, M., Wan, J., Niu, Q., Liu, R., 2019. PFOA and PFOS interact with superoxide dismutase
and induce cytotoxicity in mouse primary hepatocytes: a combined cellular and mo-
lecular methods. Environ. Res. 175, 63–70.
Zeng, Z., Song, B., Xiao, R., Zeng, G., Gong, J., Chen, M., Xu, P., Zhang, P., Shen, M., Yi, H.,
2019. Assessing the human health risks of perfluorooctane sulfonate by in vivo and
invitrostudies.Environ.Int.126,598–610.
Zhang, L., Meng, Z., Chen, L.,Zhang, G., Zhang, W., Tian,Z., Wang, Z., Yu, S., Zhou, Z.,Diao, J.,
2020a. Perfluorooctanoic acid exposure impact a trade-off between self-maintenance
and reproduction in lizards (Eremias argus) in a gender-dependent manner. Environ.
Pollut. 262, 114341.
Zhang, L., Sun, W., Chen, H., Tian, F., Cai,W., 2020b. Transcriptome analysisof acute expo-
sure of the Manila clam, ruditapes philippinarum to perfluorooctane sulfonate
(PFOS). Comp. Biochem. Physiol. C Toxicol. Pharmacol. 231, 108736.
D.J. Beale, K. Hillyer, S. Nilsson et al. Science of the Total Environment xxx (xxxx) xxx
10
