Article

Matrix Solid-Phase Dispersion Procedure for Determination of

Antibiotics and Metabolites in Mussels: Application in
Exposure Bioassays
Carmen Mejías 1 , Tainá G. Fonseca 2 , Noelia García-Criado 1 , Julia Martín 1 , Juan Luis Santos 1 ,
Irene Aparicio 1 and Esteban Alonso 1, *

1 Departamento de Química Analítica, Escuela Politécnica Superior, Universidad de Sevilla, C/Virgen de

África, 7, E–41011 Seville, Spain; cmpadilla@us.es (C.M.); ngarcia5@us.es (N.G.-C.); jbueno@us.es (J.M.);
jlsantos@us.es (J.L.S.); iaparicio@us.es (I.A.)
2 Centre for Marine and Environmental Research—CIMA/ARNET—Infrastructure Network in Aquatic
Research, University of Algarve, Campus de Gambelas, 8000-139 Faro, Portugal; tgfonseca@ualg.pt
* Correspondence: ealonso@us.es; Tel.: +34-954-55-28-58

Abstract: The presence of antibiotics in seafood for human consumption may pose a risk for con-
sumers. Furthermore, some marine organisms, such as mussels, can result in appropriate bioindica-
tors of marine contamination. In this work, a multiresidue analytical methodology suitable for the
determination of antibiotics and metabolites in mussels is proposed. The target compounds include
three sulphonamides and trimethoprim (TMP) and six of their main metabolites. Sample treatment
involves extraction and clean-up in a single step using matrix solid-phase dispersion with acetonitrile.
Analytical determination was carried out by liquid chromatography–tandem mass spectrometry.
Good linearity (R2 > 0.99), accuracy (from 80.8 to 118%), and limits of quantification (lower than
5 ng g−1 (dry matter, dm)) were obtained for all selected compounds. The method was applied
to the determination of antibiotics in mussel samples from an exposure assay with contaminated
seawater with TMP and sulfamethoxazole (SMX). Both antibiotics were detected in the analysed
samples with concentrations up to 77.5 ng g−1 dm. TMP was bioconcentrated to a higher extent than
SMX, attributable to its higher hydrophobicity. None of the metabolites were detected. These results
Citation: Mejías, C.; Fonseca, T.G.;
demonstrate that Mytilus galloprovincialis is a suitable bioindicator to assess marine pollution.
García-Criado, N.; Martín, J.; Santos,
J.L.; Aparicio, I.; Alonso, E. Matrix
Keywords: mussels; antibiotics; metabolites; MSPD; LC-MS/MS; pharmaceuticals
Solid-Phase Dispersion Procedure for
Determination of Antibiotics and
Metabolites in Mussels: Application
in Exposure Bioassays. Molecules 2024,
29, 5478. https://doi.org/10.3390/ 1. Introduction
molecules29225478 In 2021, antibiotic consumption in Europe was measured at 125 and 92.6 mg per kg
Academic Editor: Irene Panderi
of biomass for humans and food-producing animals, respectively [1]. From 2014 to 2021,
antibiotic use in animals raised for food dropped by 44%, whereas, in humans, it stayed
Received: 18 October 2024 relatively constant [1] despite the efforts carried out to reduce the consumption of antibiotics
Revised: 11 November 2024 as growing concerns are being raised about their overuse and misuse. After consumption,
Accepted: 14 November 2024 antibiotics are released into the environment primarily through wastewater treatment
Published: 20 November 2024
plants, including domestic effluents and effluents from industries and hospitals, due to their
incomplete degradation. The world oceans are regarded as the final repository for sewage
and other human by-products [2]. The most significant antibiotic in the marine environment
Copyright: © 2024 by the authors.
in terms of its detection frequency and concentration levels is sulfamethoxazole (SMX) [3].
Licensee MDPI, Basel, Switzerland. It was detected at various locations, with concentrations of up to 6.3 ng L−1 in the Saronic
This article is an open access article Gulf and the Eleusis Bay [4] and up to 70.1 ng L−1 in the Baltic Sea [5]. Trimethoprim
distributed under the terms and (TMP), commonly administered jointly with SMX, was reported at lower concentration
conditions of the Creative Commons levels (concentrations up to 21.8 ng L−1 in the Hong Kong Bay [6]) than SMX but with
Attribution (CC BY) license (https:// remarkably high detection frequency [3].
creativecommons.org/licenses/by/ The continuous exposure of aquatic ecosystems to antibiotics is an issue of environ-
4.0/). mental and human health concern, as antibiotics are conceived to exert specific biological

Molecules 2024, 29, 5478. https://doi.org/10.3390/molecules29225478 https://www.mdpi.com/journal/molecules

Molecules 2024, 29, 5478 2 of 15

effects, even at low concentrations. The increasing environmental occurrence of antibiotics

may promote the proliferation and dispersion of antibiotic-resistant bacteria strains that
threaten the treatment of infectious diseases [7]. An extensive list of antibiotics has already
been found in the tissues of marine organisms, revealing their bioaccumulative poten-
tial [8,9]. Residues of TMP and SMX were quantified in mussel tissues at concentrations
of up to 9.22 and 13.9 ng g−1 dry matter (dm), respectively [8,9]. Therefore, long-term
exposure to antibiotics and their transference to higher trophic levels may pose potential
risks to marine biota, besides posing negative consequences on human health through the
consumption of chemically contaminated seafood [10]. The main public health concern
associated with this practice is the human acquisition of antimicrobial-resistant bacteria
or genes through consuming contaminated food products, which has been demonstrated
previously in mussels [11].
Marine mussels are suitable sentinel organisms to assess the impacts of anthropogenic
stressors in coastal waters [12]. They are sessile filter-feeding organisms that can efficiently
accumulate chemical pollutants from the surrounding water, providing an integrative
measure of the concentration and bioavailability of seawater pollutants [13,14]. Moreover,
mussels are globally distributed, easily accessible, and have a high tolerance to a wide
range of environmental parameters, including temperature, oxygen levels, salinity, and
food availability.
Antibiotics and their metabolites have been extracted from marine bivalves using the
QuEChERS method [9], pressurised liquid extraction [15,16], ultrasound-assisted extrac-
tion [17], and microwave-assisted extraction [18]. Extract clean-up has been performed
by solid-phase extraction [15,16] and liquid–liquid extraction [17]. The determination has
been made mainly by liquid chromatography–tandem mass spectrometry (LC-MS/MS) [9].
However, the preparation of samples can sometimes be complex and expensive, leading
to time-consuming procedures that require large volumes of solvent and result in poor
reproducibility. Therefore, there is a growing need to develop new, reliable analytical
methods for determining antibiotics in mussels. Matrix solid-phase dispersion (MSPD) is
a technique that integrates extraction and purification into a single step, facilitating the
analysis of contaminants in environmental, food, and other complex matrices. The soft
tissues of marine animals are indeed complex matrices, containing various compounds
like lipids and proteins, which can interfere with detection and quantification processes,
impacting both selectivity and sensitivity. In this technique, the solid sample is mixed
with a dispersant to achieve complete sample disruption and isolate the analytes. The
analytes are then eluted using an appropriate solvent [19]. This technique reduces solvent
consumption, extraction time, and sample amount, making the extraction method highly
suitable, fast, economical, and simple in terms of sample preparation. MSPD has been
applied with success for the determination of bisphenols [20–22], phthalates [22,23], pes-
ticides [24], parabens [25], polychlorinated biphenyls [26], herbicides [27], UV filters [28],
perfluorinated compounds [29], and flame retardants [30,31] in mussels, demonstrating the
suitability of the technique to extract chemical compounds from such biological matrix.
The aim of this work was to develop and validate a new method for the determination
of TMP, sulphonamides, and their main metabolites in marine mussels, Mytilus galloprovin-
cialis, using MSPD as an extraction technique. This method offers several advantages over
other conventional extraction techniques, including reduced time consumption, lower costs,
and decreased solvent usage. To the best of our knowledge, this is the first methodology
of this approach being applied for antibiotic determination in mussel samples. Further-
more, the methodology was applied to whole soft tissues of M. galloprovincialis, exposed to
environmentally relevant concentrations of SMX and TMP, in an in vivo 28-day bioassay.

2. Results and Discussion

2.1. Method Optimisation
The key parameters influencing MSPD, namely the type of extraction solvent, the
type and amount of clean-up sorbents, and the number of extraction cycles, were assessed.
Molecules 2024, 29, 5478 3 of 15

Optimisation was conducted using lyophilised mussels (0.2 g dm), spiked to achieve a
final concentration of 125 ng g−1 dm for each target compound. The mussel samples were
spiked by adding a methanol solution containing the selected antibiotics at the required
spiking concentration in each case. The volume of the solution added was 200 µL, which
was enough volume to impregnate the entire solid sample. Once the solution was added, it
was vortexed for 1 min for homogenisation. The spiked mussel samples were incubated in
the dark for 12 h to reach equilibrium and for the total evaporation of the methanol. All
experiments were performed in triplicate.

2.1.1. Optimisation of the Extraction Solvent

For the optimisation of the extraction solvent, an aprotic solvent (acetonitrile) and a
protic solvent (methanol) were assessed. The use of water as extraction solvent required more
than 12 h under a nitrogen stream for evaporation to dryness, which makes the method
neither environmentally friendly, easy to perform, nor cheap without obtaining better
recovery percentages of the analytes. For these reasons, the use of water was discarded.
These two solvents were tested as pure solvents and acidified at 0.1% v/v with formic
acid. The initial conditions did not include sorbents for clean-up in the extraction procedure.
For that, 0.2 g dm of spiked sample was mixed with 1 g of silica. The empty cartridge
was fitted with a frit at the bottom. The mixture of the sample dispersed in the dispersant
was placed on top of the frit and compacted in the cartridge, and another frit was added
on top. The cartridge was eluted by gravity using three consecutive aliquots of 4 mL of
each extraction solvent. Experiments were performed in triplicates. Mixtures of solvents
were not optimised because previous works about the determination of sulphonamide
antibiotics in similar samples revealed that better recoveries were obtained using a single
solvent. For example, recently, Sun et al. (2023) [32] determined five sulphonamides in
shrimp samples using acetonitrile as the extraction solvent. Similarly, Mu et al. (2022) [33]
also employed only acetonitrile for the extraction of TMP and 17 sulphonamides from
fish samples. Extraction efficiencies (peak area of spiked sample compared with the peak
area of a standard at the same concentration in pure solvent) were calculated. Results can
be seen in Figure 1. In most cases, acetonitrile showed better extraction efficiencies than
methanol. It has been reported that acetonitrile has the ability to precipitate the proteins
present in the matrix [33,34], obtaining better results in food samples. The addition of
0.1% v/v of formic acid to acetonitrile increased the extraction efficiencies but only in
a few compounds, specifically in those that obtained better extraction efficiencies (such
as N4 -acetylsulfadiazine (AcSDZ) or N4 -acetylsulfamethazine (AcSMZ)), while in the
compounds that extracted worse the efficiency obtained was higher when acetonitrile alone
was used. Given that the method proposed is a multiresidue approach, where compounds
with different physical and chemical properties are analysed, it is necessary to achieve a
balance that ensures optimal recovery for all analytes. For this reason, an extraction solvent
was selected that, having good extraction for most of the compounds, best extracted the
compounds with the worst extraction efficiencies. From these results, acetonitrile was
selected as the extraction solvent for further experiments, as this solvent provided the best
average results.

2.1.2. Optimisation of d-SPE Sorbents and Their Amount

Three sorbents were evaluated for extract clean-up. One reversed-phase sorbent
(C18), one normal-phase sorbent (Florisil), and a weak anion exchanger sorbent (PSA). A
Box–Behnken design (BBD) was applied to the clean-up optimisation, enabling a proper
assessment of the influence and interactions of each variable. The number of experiments
(N) necessary for BBD optimisation is calculated using Equation (1):

N = 2k(k − 1) + C0 (1)

where k represents the number of variables, and C0 is the number of central points. Three
variables were considered, corresponding to the types of clean-up sorbents assessed, with
 Molecules 2024, 29, 5478 4 of 15

three central points. The experiments were conducted randomly. Each sorbent was tested
at three levels (0, 0.4, and 0.8 g). Thus, 15 experiments were required for the simultaneous
optimisation of the type and amount of sorbents. Table S1 presents the values assigned
to each variable for each experiment. For that, 0.2 g dm of spiked sample was mixed
with 1 g of silica. A mixture of the clean-up sorbents, according to the 15 experiments
(Table S1), was placed on top of the frit. The sorbents were compacted in the cartridge,
and another frit was added on top of them. The sample dispersed in the dispersant was
then added on top of the second frit. It was compacted, and another frit was added on
top of the sample. The cartridge was eluted by gravity using three consecutive aliquots
of 2 mL of acetonitrile. Extraction efficiencies (peak area of spiked sample compared with
the peak area of a standard at the same concentration in pure solvent) were calculated.
BBD was calculated using the optimisation of multiple responses mode in Statgraphics
18-X64 version 18.1.16. Figure 2 illustrates the response surface plots of the geometric mean
relative signals. As shown, a larger amount of Florisil (0.8 g) resulted in higher desirability.
Regarding PSA, lower amounts yielded better desirability, with higher amounts producing
2024, 29, x FOR PEER REVIEW 4 of 15
worse results; the optimal amount was 0.3 g. Similarly, for C18, both very high and very low
amounts resulted in lower desirability, with an intermediate amount (0.5 g) being optimal.

s 2024, 29, x FOR PEER REVIEW

Figure 1. Extraction efficiency
Figure percentage
1. Extraction of compounds
efficiency percentageobtained using different
of compounds obtainedextraction 5 extraction
of 15
solvents
using different solvents
(n = 3). (n = 3).

2.1.2. Optimisation of d-SPE Sorbents and Their Amount

Three sorbents were evaluated for extract clean-up. One reversed-phase sorbent
(C18), one normal-phase sorbent (Florisil), and a weak anion exchanger sorbent (PSA). A
Box–Behnken design (BBD) was applied to the clean-up optimisation, enabling a proper
assessment of the influence and interactions of each variable. The number of experiments
(N) necessary for BBD optimisation is calculated using Equation (1):
N = 2k(k − 1) + C0 (1)
Figure 2. Response surface plots corresponding
Response surface to desirability versus
plots corresponding (A) C18versus
and PSA amounts (g); amounts (g);
Figure
where k represents 2. number
the of variables, and C0 is the to desirability
number (A) C18
of central points. and PSA
Three
(B) C18 and Florisil amounts (g); (C) Florisil and PSA amounts (g).
(B) C18 and Florisil
variables were considered, amounts (g);
corresponding to (C)
theFlorisil
types and PSA amounts
of clean-up (g). assessed, with
sorbents
three central points. The experiments were conducted randomly. Each sorbent was tested
2.1.3. Optimisation of Extraction Cycles
at three levels (0, 0.4, and 0.8 g). Thus, 15 experiments were required for the simultaneous
In the optimisation
optimisation of the type andof amount
the technique, the remaining
of sorbents. volume the
Table S1 presents thatvalues
couldassigned
be addedtoof
extraction
each variablesolvent, based
for each on the capacity
experiment. For that,of0.2
thegcartridge used and
dm of spiked the was
sample amounts ofwith
mixed sample,
1
sorbent, and dispersants added, was 3.5 mL. The number of cycles (between 1 and
g of silica. A mixture of the clean-up sorbents, according to the 15 experiments (Table S1),3) of
 Figure 2. Response surface plots corresponding to desirability versus (A) C18 and PSA amounts (g);
Molecules 2024, 29, 5478 (B) C18 and Florisil amounts (g); (C) Florisil and PSA amounts (g). 5 of 15

2.1.3. Optimisation of Extraction Cycles

2.1.3.InOptimisation
the optimisationof Extraction Cycles
of the technique, the remaining volume that could be added of
In thesolvent,
extraction optimisation
based of onthe
thetechnique,
capacity ofthe theremaining
cartridge usedvolume andthat could be added
the amounts of
of sample,
extraction
sorbent, solvent,
and based on
dispersants the capacity
added, was 3.5ofmL. theThecartridge
number used of and the(between
cycles amounts of sample,
1 and 3) of
sorbent, and
extraction wasdispersants added, was
therefore optimised, 3.5the
i.e., mL. The number
number of times ofthat
cycles3.5(between 1 andsolvent
mL of elution 3) of
extraction was therefore optimised, i.e., the number of times that
was added to the cartridge. For that, 0.2 g dm of spiked sample was mixed with 1 g of 3.5 mL of elution solvent
was added
silica. to thecartridge
An empty cartridge. wasFor that,with
fitted 0.2 a
g filter
dm offrit spiked
at thesample
bottom.was mixed with
A mixture of the1clean-
g of
up sorbents was placed on top of the frit, consisting of 0.8 g of Florisil, 0.5 g of C18,the
silica. An empty cartridge was fitted with a filter frit at the bottom. A mixture of and
clean-up sorbents was placed on top of the frit, consisting of 0.8
0.3 g of PSA. The sorbents were compacted in the cartridge, and another frit was added g of Florisil, 0.5 g of C18,
and 0.3 g of PSA. The sorbents were compacted in the cartridge, and another frit was
on top of them. The sample dispersed in the dispersant was then added on top of the
added on top of them. The sample dispersed in the dispersant was then added on top
second frit. It was compacted, and another frit was added on top of the sample. The car-
of the second frit. It was compacted, and another frit was added on top of the sample.
tridge was eluted by gravity using 1, 2, or 3 consecutive aliquots of 3.5 mL of acetonitrile.
The cartridge was eluted by gravity using 1, 2, or 3 consecutive aliquots of 3.5 mL of
Experiments were performed in triplicates. Extraction efficiencies (peak area of spiked
acetonitrile. Experiments were performed in triplicates. Extraction efficiencies (peak area
sample compared with the peak area of a standard at the same concentration in pure sol-
of spiked sample compared with the peak area of a standard at the same concentration
vent)
in pure were calculated.
solvent) Results canResults
were calculated. be seen can inbeFigure
seen in3. Figure
The highest
3. Theextraction efficiency
highest extraction
was obtained for two cycles. As expected, the higher the number
efficiency was obtained for two cycles. As expected, the higher the number of cycles, the of cycles, the higher the
extraction
higher the efficiency
extraction when comparing
efficiency one cycleone
when comparing andcycle
two andcycles
twodue to the
cycles duehigher
to the volume
higher
of
volume of extraction solvent. The reduction in efficiency of the extraction from two to
extraction solvent. The reduction in efficiency of the extraction from two cycles three
cycles
cycles can be attributable to the extraction of a major number of interferences
to three cycles can be attributable to the extraction of a major number of interferences with with a high
matrix
a high effect
matrixthat results
effect that in a lowinextraction
results efficiency.
a low extraction Therefore,
efficiency. two cycles
Therefore, twowerecycles selected
were
as the optimum.
selected as the optimum.

Figure 3. Extraction efficiency percentage of compounds obtained using different extraction cycles
Figure 3. Extraction efficiency percentage of compounds obtained using different extraction cycles
(n ==3).
(n 3).

2.2. Method Validation

The optimised methodology was validated for the determination of the antibiotics and
their metabolites in mussel samples. The validation was conducted in terms of linearity,
method detection limits (MDL), method quantification limits (MQL), precision (expressed
as relative standard deviation, RSD), absolute recovery (R), and accuracy (A), expressed as
relative recovery.
First, two different calibration curves were prepared: the first one in pure solvent
(external calibration) and the other one by matrix-matched standards, both at six different
concentration points and in triplicate. The slopes of both calibration curves were com-
Molecules 2024, 29, 5478 6 of 15

pared, and significant differences were obtained at 95% confidence using Student’s t-test.
Therefore, the presence of matrix effects was confirmed, and a matrix-matched calibration
curve was needed to be applied for the quantification of antibiotics. For this purpose,
mussel samples were spiked in triplicate at six different concentration levels, and the final
optimised methodology was applied to each of them. The ratio of the peak area after
subtracting the blank peak area from the peak area of the internal standard was plotted
versus the concentration of the analyte to obtain the matrix-matched calibration curve for
the determination of antibiotics in mussel samples.
The determination coefficients (R2 ) for matrix-matched calibration curves with six
different concentrations (from MQL to 125 ng g−1 dm) in triplicates were higher than 0.99
for all compounds (Table 1). MDL and MQL were obtained as the sample concentrations
provided a signal-to-noise ratio of 3 and 10, respectively. For their calculation, spiked
samples were employed. For all compounds, MQL values were in the range of 0.10
to 5.00 ng g−1 dm, and MDL values were from 0.03 to 1.50 ng g−1 dm. The obtained
MDL, MQL, and matrix-matched calibration curve correlation coefficients (R2 ) values are
presented in Table 1.

Table 1. Method detection limits (MDL), method quantification limits (MQL), and matrix-matched
calibration curve correlation coefficients (R2 ).

Mussels
Compound MDL MQL
R2
(ng g−1 dm) (ng g−1 dm)
TMP 0.30 1.00 0.997
4-OH-TMP 1.50 5.00 0.990
DM-TMP 0.30 1.00 0.997
SMX 0.03 0.10 0.998
AcSMX 0.30 1.00 0.994
SMX-GL 1.50 5.00 0.990
SDZ 0.15 0.50 0.996
AcSDZ 1.50 5.00 0.993
SMZ 0.03 0.10 0.999
AcSMZ 0.03 0.10 0.995
Parent compounds are marked in bold.

After that, matrix effects, recovery, precision, and accuracy values were calculated at three
different concentration levels: 1.25, 12.5, and 125 ng g−1 dm, except for 4-hydroxytrimethoprim
(4-OH-TMP), AcSDZ, and sulfamethoxazole N4 -glucoside (SMX-GL), concentrations levels
of which were 6.25, 12.5, and 125 ng g−1 dm.
The results obtained for recovery, accuracy, matrix effect, and precision at three spiking
levels are shown in Table 2. The matrix effect was in the range from −56.4 to −2.14%.
Matrix suppression was obtained in all compounds in mussel samples. Matrix suppression
values were low considering the large amount of interferents that can be extracted from
the mussel matrix, such as lipids or proteins. Recovery values ranged from 27.0 to 71.6%
(Table 2). Accuracy ranged from 80.8 to 118% (Table 2). Analytical guidelines, such as those
from the AOAC Peer-Verified Methods program [35], recommend recoveries ranging from
60 to 120% at part per billion levels, typically referring to absolute recovery or accuracy
for individual analyses. Given that the method proposed in this work is a multiresidue
approach, where compounds with highly varied properties are analysed, it is necessary
to achieve a balance that ensures optimal recovery for all analytes. Furthermore, despite
recoveries achieved for some of the target compounds being low (lower than 30%), as
shown in Table 2, the accuracy reaches values between 80.8 and 120% for all analytes. This
parameter and the low limits of quantification reflect the suitability of the method.
Molecules 2024, 29, 5478 7 of 15

Table 2. Recovery (R%), accuracy (A%), matrix effect (ME%), and precision, expressed as relative
standard deviation (RSD%), for mussels’ matrix at three spiking levels.

Mussels
Compound
1.25 (ng g−1 , dm) 12.5 (ng g−1 , dm) 125 (ng g−1 , dm)
R (%) A (%) ME (%) RSD (%) R (%) A (%) ME (%) RSD (%) R (%) A (%) ME (%) RSD (%)
TMP 28.1 102 −14.6 9.34 27.2 92.6 −26.8 17.7 33.1 118 −28.0 6.04
4-OH-TMP * 55.8 80.8 −12.2 11.6 45.0 98.4 −12.1 15.2 49.8 94.6 −20.1 5.14
DM-TMP 27.0 120 −54.2 7.41 27.5 97.5 −50.0 14.8 30.3 101 −44.6 12.3
SMX 57.5 98.0 −31.8 2.13 58.6 95.6 −26.0 2.52 54.7 96.6 −32.3 5.46
AcSMX 63.3 118 −20.7 13.8 63.3 89.4 −21.1 13.9 67.3 98.9 −3.42 7.20
SMX-GL * 53.2 101 −10.7 14.0 64.2 111 −56.4 17.4 53.7 93.5 −24.1 9.18
SDZ 28.5 109 −45.3 9.42 33.0 95.2 −45.3 3.80 29.4 100 −50.7 3.65
AcSDZ * 54.4 93.1 −2.14 13.2 54.3 102 −4.50 17.7 53.1 84.1 −6.51 17.3
SMZ 37.8 117 −12.3 6.53 36.3 88.0 −16.8 12.5 36.4 97.5 −17.6 6.32
AcSMZ 71.6 98.5 −4.12 14.9 71.4 86.4 −4.59 2.29 70.0 103 −14.8 7.80
Parent compounds are marked in bold; *: spiking levels: 6.25, 12.5, and 125 ng g−1 dm.

Precision was expressed as relative standard deviation (RSD, %). RSD values were
below 17.7% for all compounds at the three spike concentrations (mean value: 10.0%)
(Table 2).
Finally, the method’s selectivity was assessed by visualising potential interferences
in the obtained chromatograms. No interference was observed at the retention times of
the target compounds. Figure S1 shows the LC-MS/MS chromatogram of a 10 ng g−1 dm
spiked mussel sample.

2.3. Method Comparison

Table 3 presents data on various analytical methodologies reported for determining
antibiotics in bivalve mollusc samples.

Table 3. Comparison of proposed methodology with other methods for determination of antibiotics
in bivalve molluscs.

Sample Solvent
Clean MQL
Compounds Sample Amount Extraction Volume Determination Recovery Reference
Up (ng g−1 , dm)
(g) (min)
6 macrolides,
7 sulphonamides, Mussels and
0.5 QuEChERS 10 - LC-MS/MS 28–60 0.05–1.03 [9]
metronidazole, TMP, clams
3 metabolites
3 nitroimidazoles,
Mussels,
1 sulphonamide,
oysters and 0.5 PLE 200 SPE LC-MS/MS 30.2–115.7 0.02–2.66 [15]
2 macrolides,
clams
1 metabolite
2 b-lactams,
2 tetracyclines,
2 amphenicols, HPLC-DAD-
Mussels 2.0 UAE 15 LLE 60.1–83.3 50–580 [17]
5 sulphonamides, FLD
TMP and
5 metabolites
2 b-lactams,
2 tetracyclines,
2 amphenicols,
Mussels 2.0 MAE 10 - LC-MS/MS 63–97 5–55 [18]
5 sulphonamides,
TMP and
5 metabolites
TMP Mussels 1.0 PLE - SPE LC-MS/MS 91 4 [16]
3 sulphonamides,
Proposed
TMP and Mussels 0.2 MSPD 7 - LC-MS/MS 27.0–71.6 0.1–5
methodology
6 metabolites
-: no data; HPLC-DAD-FL: high-pressure liquid chromatography coupled to diode array detection and fluo-
rescence; LC-MS/MS: liquid chromatography–tandem mass spectrometry; LLE: liquid-liquid extraction; MAE:
microwave-assisted extraction; MQL: method quantification limits; MSPD: matrix solid-phase dispersion; PLE:
pressurised-liquid extraction; QuEChERS: quick, easy, cheap, effective, rugged and safe method; SPE: solid-phase
extraction; UAE: ultrasound-assisted extraction.
Molecules 2024, 29, 5478 8 of 15

Among the studied methods, mussels were the most frequently analysed molluscs,
with clams and oysters also included. The proposed method used the smallest sample
amount (0.2 g), while other methods required up to 2.0 g of sample. Various extraction
techniques were employed (including MAE, PLE, UAE, and QuEChERS), but this work
introduces the first method based on MSPD. Additionally, the proposed method used
the lowest solvent volume (7 mL) compared with other methodologies, which used up
to 200 mL [15]. For extract clean-up, Álvarez-Muñoz et al. (2015) [15] and McEneff et al.
(2013) [16] employed an additional SPE step, while Fernández-Torres et al. (2010) [17] used
LLE. The proposed methodology performed extraction and purification simultaneously,
similar to the QuEChERS method reported by Serra-Compte et al. (2017) [9]. The main
advancement of the developed MSPD method is the minimal amount of mussel sample and
solvent required for the extraction of the target analytes and the simultaneous extraction
and purification. Therefore, in terms of green sample preparation, the proposed method
can be considered one of the most sustainable when compared with similar reported
methodologies for analysing antibiotics residues in bivalve molluscs. Most methodologies
used LC-MS/MS for antibiotic determination, whereas Fernández-Torres et al. (2010) [17]
applied high-pressure liquid chromatography coupled with diode array detection and
fluorescence detection (HPLC-DAD-FLD), which results in a loss of sensitivity, as it is the
method with the highest MQLs. The cost of performing the technique is low as in the
QuEChERS method in comparison with the other methodologies since it does not require
the acquisition of high-cost equipment, as the PLE and MAE methods might require.
Although at a lower cost, the method using UAE also requires expensive equipment. In
addition, this technique does not require any additional steps such as sample centrifugation
(which is necessary for MAE, UAE, and QuEChERS) since, thanks to the frits, the extraction
solvent is filtered as it elutes from the cartridge. The extraction recoveries obtained with the
proposed method were comparable to those of previously reported methods, even when
using lower extraction solvent volumes. The dispersion of the matrix allows for a better
extraction of compounds attributable to a better interaction between individual sample
particles and the extraction solvent. The mechanical forces involved in mixing fragment
the material into smaller parts with higher specific surfaces. Obtained MQL values were
similar or lower in comparison with the previously reported methods, even when using
smaller amounts of sample and extraction solvents.

2.4. Method Application

The proposed analytical method was applied to the monitorisation of sulphonamides
and TMP in mussels exposed to target compounds. The results from the exposure assay are
presented in Table S2. Field samples (mussels directly taken from the natural environment)
showed concentrations below the MDL for all compounds. Consequently, samples at time
0 of the experiment (after a 7-day acclimation period but before antibiotic exposure) were
also below the MDL for all compounds. After 14 days of exposure to 1 µg L−1 of SMX
and TMP, the concentrations were 1.17 ng g−1 dm and 62.3 ng g−1 dm, respectively. These
results indicate that TMP bioconcentrates more efficiently in mussels than SMX. This could
be explained by the higher hydrophobicity of TMP, expressed by its higher octanol–water
partitioning coefficient (log Kow = 1.26), compared with SMX (0.79). An increase in SMX
and TMP concentrations in mussels’ tissues was observed on the 28th day of exposure,
reaching up to 1.26 ng g−1 dm and 77.5 ng g−1 dm, respectively, although no differences
were observed compared with the 14th day. These findings underscore the significant
health risks posed by the presence of antibiotics in the marine environment.
Serra-Compte et al. (2019) [36] revealed an increase in SMX concentration in mussels
M. galloprovincialis up to 13.2 ± 0.7 ng g−1 dm after 96 h of exposure to the 10 µg L−1
of the antibiotic. The concentrations found in the whole soft tissues of mussels were
10 times higher, while the concentrations in seawater used were also 10 times higher, thus
finding a 10-fold correlation. In contrast to the parent compounds detected in the whole
tissues of mussels, no TMP- or SMX-related metabolites nor other sulphonamides and their
Molecules 2024, 29, 5478 9 of 15

metabolites were detected in the biological samples analysed herein. Similarly, none of the
analytical approaches performed by Serra-Compte et al. (2019) [36] allowed the detection of
SMX metabolites (i.e., SMX-GL and N4 -acetylsulfamethoxazole (AcSMX)), suggesting that
the parent compounds SMX and TMP were not metabolised by the selected marine species.
McEneff et al. (2014) [8] found residues of TMP at concentrations of 9.22 ng g−1 and
7.28 ng g−1 dm in the Mytilus spp. from organisms exposed to marine water naturally
contaminated at 0.16 and 0.29 µg L−1 , respectively. These concentrations were lower than
those obtained in this study.
Bioconcentration is the process by which a pollutant is absorbed by an organism from
the environment via the dermal and/or respiratory routes, with dietary intake not included.
Bioconcentration factor (BCF, mL g−1 ) was calculated using measured concentration levels
in seawater and biota using the following Equation (2):

BCF = Cb /Cw (2)

where Cb is the concentration of a chemical in the biota (ng g−1 dm), and Cw is the concen-
tration of a chemical in the water a (ng mL−1 ) at the same time at the end of the experiment
(28th day). Calculated BCFs were 1.26 and 77.5 mL g−1 for SMX and TMP, respectively.
Similar BCFs (up to 1.5 mL g−1 ) have been previously reported for SMX in mussels [36].
These results demonstrate that Mytilus galloprovincialis is a suitable bioindicator to as-
sess marine pollution. In addition, it also highlights the high bioconcentration rates of
antibiotics, especially TMP, in mussels, which can have negative effects on humans through
the consumption of contaminated mussels. This scenario could worsen as the presence of
antibiotics in mussels may promote the development of antimicrobial-resistant bacteria or
genes and affect humans through the consumption of contaminated food products.

3. Materials and Methods

3.1. Chemicals and Reagents
Sigma-Aldrich (Steinheim, Germany) supplied high-purity standards of AcSDZ (≥99.0%),
AcSMX (≥98.5%), sulfadiazine (SDZ, ≥99.0%), and sulfamethazine (SMZ, ≥99.0%). AcSMZ
(≥98.0%), SMX-GL (>99.0%), 3-desmethyltrimethoprim (DM-TMP, ≥98.0%), and 4-OH-
TMP (≥97.0%) were obtained from Toronto Research Chemicals (Toronto, ON, Canada).
TMP (≥99.5%) and SMX (≥99.0%) were supplied by Dr. Ehrenstorfer GmbH (Augsburg,
Germany). Physical–chemical properties of selected compounds can be seen in Table 4.
The isotopically labelled compound, sulfamethoxazole-(phenyl-13 C6 ) (SMX-13 C, ≥99.0%),
used as an internal standard, was supplied by Sigma-Aldrich (Steinheim, Germany).
Florisil® , silica, and ammonium formate were provided by Sigma-Aldrich (Steinheim,
Molecules 2024, 29, x FOR PEER REVIEW 10 of 15
Germany). Primary–secondary amine (PSA) and C18 were supplied by Scharlab (Barcelona,
Molecules 2024, 29, x FOR PEER REVIEW 10 of 15
Spain). Formic acid (≥98.0%) was obtained from Panreac (Barcelona, Spain). All reagents
were of high
water puritychromatography–tandem
of liquid and analytical grade. Acetonitrile, methanol,
mass spectrometry and water
(LC-MS/MS) of were
grade liquid
chromatography–tandem
supplied
water of by Merck
liquid mass spectrometry (LC-MS/MS) grade were supplied by Merck
(Darmstadt, Germany). mass spectrometry (LC-MS/MS) grade were
chromatography–tandem
(Darmstadt,
supplied byGermany).
Merck (Darmstadt, Germany).
Table 4. Physical–chemical properties of the target compounds.
Table 4. Physical–chemical
Table properties
4. Physical–chemical ofof
properties the target
the targetcompounds.
compounds.
Compound Molecular Weight (g mol ) −1 pKa Log Kow Chemical Structure
Compound Compound
Molecular Weight Molecular
(g mol −1 ) Weight
pKa (g mol )
−1 pK
Loga KowLog Kow Chemical
Chemical Structure
Structure
7.16, 17.3 1.26
TMP 290.3
[37]
7.16, 17.3 [37]
1.26
TMP TMP
290.3 290.3
7.16, 17.3 [37] 1.26 [37]
[37] [37]

4-OH-TMP 4-OH-TMP
306.3 306.3
8.18 [38] 8.18 [38]
- -
4-OH-TMP 306.3 8.18 [38] -

DM-TMP 276.3 9.40 [38] -

DM-TMP 276.3 9.40 [38] -

1.97, 6.16 0.79

SMX 253.3
[39] [37]
 TMP
Compound 290.3 (g mol−1)
Molecular Weight [37]a
pK [37]
7.16,
[37] 17.3 Log1.26
[37]
Kow Chemical Structure
TMP 290.3 7.16, 17.3 1.26
TMP 290.3 [37]
7.16, 17.3 [37]
1.26
TMP 290.3 [37]
7.16, 17.3 [37]
1.26
TMP 290.3 7.16,
[37] 17.3 1.26
[37]
TMP
4-OH-TMP 290.3
306.3 [37]
8.18 [38] [37]
-
[37] [37]
4-OH-TMP 306.3 8.18 [38] -
Molecules 2024, 29, 5478 4-OH-TMP 306.3 8.18 [38] - 10 of 15
4-OH-TMP 306.3 8.18 [38] -
4-OH-TMP 306.3 8.18 [38] -
4-OH-TMP 306.3 8.18 [38] -
4-OH-TMP 306.3 8.18 [38] -
Table4-OH-TMP
4. Cont.
DM-TMP 306.3
276.3 8.18
9.40 [38]
[38] --
DM-TMP 276.3 9.40 [38] -
Compound DM-TMP
Molecular Weight (g mol−1 ) 276.3
pKa 9.40Log [38]
Kow - Chemical Structure
DM-TMP 276.3 9.40 [38] -
DM-TMP 276.3 9.40 [38] -
DM-TMP 276.3 1.97, 6.16
9.40 [38] 0.79
-
SMX
DM-TMP 253.3
276.3 9.40 [38] -
DM-TMP DM-TMP
276.3 276.3
9.40 [38] 1.97,
[39]
9.40 6.16
[38]
- 0.79
[37]
-
SMX 253.3 1.97, 6.16 0.79
SMX 253.3 [39] [37]
1.97,
[39] 6.16 0.79
[37]
SMX 253.3 1.97, 6.16 0.79
SMX 253.3 [39]
1.97, 6.16 [37]
0.79
SMX 253.3 [39]
1.97, 6.16 [37]
0.79
1.18
SMX SMX
253.3
AcSMX 253.3
1.97, 6.16 [39]
295.3 1.97,
[39]
5.540.796.16
[37] 0.79
[40] [37]
SMX 253.3 [39] [37]
1.18
[40]
AcSMX 295.3 [39]
5.54 [40] [37]
1.18
[40]
AcSMX 295.3 5.54 [40] 1.18
AcSMX 295.3 5.54 [40] [40]
1.18
AcSMX 295.3 5.54 [40] [40]
1.18
AcSMX AcSMX
295.3 295.3
5.54 [40] 5.541.18[40] [40]
[40] 1.18
AcSMX 295.3 5.54 [40] 1.18
[40]
AcSMX
SMX-GL 295.3
415.4 5.54-[40] [40]
-
[40]
SMX-GL 415.4 - -
SMX-GL 415.4 - -
SMX-GL 415.4 - -
SMX-GL SMX-GL
415.4 415.4- - - -
SMX-GL 415.4 2.01,-- 6.99 -
0.25
SMX-GL
SDZ 415.4
250.3 -
SMX-GL 415.4 2.01,- 6.99
[37] -
0.25
[37]
SDZ 250.3 2.01, 6.99 0.25
SDZ 250.3 [37] [37]
2.01,
[37] 6.99 0.25
[37]
SDZ SDZ
250.3 250.3
2.01, 6.99 [37] 2.01,
0.256.99
[37] 0.25
SDZ 250.3 [37]
2.01, 6.99 [37]
0.25
SDZ 250.3 [37]
2.01, 6.99 [37]
0.25
0.39
SDZ
AcSDZ 250.3
292.3 2.01,
[37]
6.10 6.99
[41] 0.25
[37]
SDZ 250.3 [37] [37]
0.39
[39]
AcSDZ 292.3 [37]
6.10 [41] [37]
0.39
[39]
AcSDZ AcSDZ
292.3 292.3
6.10 [41] 6.10 [41]
0.39 [39] 0.39
AcSDZ 292.3 6.10 [41] [39]
0.39
AcSDZ 292.3 6.10 [41] [39]
0.39
AcSDZ 292.3 6.10 [41] [39]
0.39
AcSDZ 292.3 6.10 [41] 0.39
[39]
AcSDZ 292.3 6.10 [41]
2.04, 6.99 0.43
[39]
SMZ 278.3 [39]
2.04,
[37] 6.99 0.43
[37]
SMZ SMZ
278.3 278.3
2.04, 6.99 [37] 2.04,0.43 [37] 0.43
6.99
SMZ 278.3 [37] [37]
2.04,
[37] 6.99 0.43
[37]
SMZ 278.3 2.04, 6.99 0.43
SMZ 278.3 [37]
2.04, 6.99 [37]
0.43
SMZ 278.3 [37]
2.04, 6.99 [37]
0.43
SMZ 278.3 2.04,
[37] 6.99 0.43
[37]
SMZ 278.3 [37] [37]
[37] [37]
1.48
AcSMZ AcSMZ
320.4 320.4
7.16 [42] 7.161.48[42]
[39] 1.48
[39]
AcSMZ 320.4 7.16 [42] 1.48
AcSMZ 320.4 7.16 [42] [39]
1.48
[39]
AcSMZ 320.4 7.16 [42] 1.48
-: no -: AcSMZ
no
data; data;
parentparent 320.4
compounds
compounds are marked
are 320.4
marked 7.16 [42]
in bold.in bold.
[39]
1.48
AcSMZ 7.16 [42] [39]
1.48
AcSMZ
-: no 320.4
data; parent compounds are marked in bold.7.16 [42] 1.48
[39]
AcSMZ 320.4 7.16 [42] [39]
3.2. -:Sample
no data;Collection
parent compounds are marked
and Exposure Assayin bold. [39]
-: no data; parent compounds are marked in bold.
-: no data; parent compounds are markedlength:
in bold.
-:Mussels M. galloprovincialis
no data; parent marked in bold. 6.82 ± 0.53 cm; mean width: 3.85 ± 0.30 cm)
compounds are(mean
-:
were no data;
harvestedparent compounds
from are
a coastal marked in bold. (Western Coast of Portugal), during win-
-: no data; parent compounds arefarm ininSagres
marked bold.
ter (February 2024), and immediately transported alive to the laboratory. Upon arrival,
specimens were cleaned by scraping off any fouling. The mussels were then placed in
glass aquaria filled with 12.5 L of seawater (2 mussels per liter) with a salinity of 35 g L−1
and a temperature of 17 ◦ C. They were kept under constant aeration and a light cycle of
12:12 h for a one-week acclimation period. The seawater was renewed every 48 h, and the
mussels were fed with marine microalgae Tetraselmis chuii, cultured at a concentration of
150,000 cells per mussel, once per renewal day. After the acclimation period, the mussels
were exposed for 28 days, in triplicate, to 1 µg L−1 of SMX and TMP. The light cycle was
maintained at 12 h of light and 12 h of dark. The water was renewed every 48 h, with
the antibiotic concentration re-established each time. During the exposure period, the
mussels were also fed with Tetraselmis chuii, cultured at 150,000 cells per mussel, once per
renewal day. Mussels were collected at 0, 14, and 28 days of exposure. They were sacrificed,
and their tissues were freeze-dried in a Cryodos-50 lyophiliser (Telstar, Barcelona, Spain),
homogenised in a mortar, and sieved (particle size < 100 µm). Lyophilised samples were
frozen until analysis.
Molecules 2024, 29, 5478 11 of 15

3.3. Sample Treatment

An MSPD technique was employed to extract the target analytes from the mussel
samples. For that, 0.2 g dm of homogenised and lyophilised mussel sample was mixed
with 1 g of silica (used as a dispersant/solid support). To achieve a thorough dispersion,
the mixture was ground using a laboratory mini ball mill Pulverisette 23 (Fritsch, Idar-
Oberstein, Germany) for 2 min at 35 oscillations per second. An empty SPE polypropylene
cartridge of 6 mL was fitted with a polyethylene filter frit at the bottom. A mixture of
the clean-up sorbents was placed on top of the frit, consisting of 0.8 g of Florisil, 0.5 g of
C18, and 0.3 g of PSA. The sorbents were compacted in the cartridge, and another frit was
added on top of them. The sample dispersed in the dispersant was then added on top of
the second frit. It was compacted, and another frit was added on top of the sample. The
cartridge was eluted by gravity using two consecutive aliquots of 3.5 mL of acetonitrile.
To recover the maximum volume of extraction solvent, at the end of the elution process,
a vacuum manifold system (Waters, Milford, AL, USA) connected to a vacuum pump
was used. The eluate was collected and evaporated to dryness under a gentle nitrogen
stream. The dried extract was then resuspended in 100 µL of methanol and 150 µL of water
containing the internal standard and achieving a final concentration of 50 µg L−1 , filtered
with a syringe filter (0.22 µm) and collected in an automatic injector vial for LC–MS/MS
determination.

3.4. Liquid Chromatography–Tandem Mass Spectrometry

Chromatographic determination was performed using an Agilent 1290 Infinity II
chromatograph (Agilent, Palo Alto, CA, USA) equipped with a vacuum degasser, a binary
pump, and an automatic injector. Chromatographic separation was conducted on a Zorbax
RRHD Eclipse Plus C18 column (150 mm × 3.0 mm inner diameter, 1.8 µm particle size)
(Agilent, Palo Alto, CA, USA), thermostated at 35 ◦ C and protected with a Zorbax RRHD
Eclipse Plus C18 guard column (3.0 mm inner diameter, 1.8 µm particle size) (Agilent, Palo
Alto, CA, USA). The injection volume was 10 µL. The chromatographic conditions were
those previously optimised [43]. The optimised LC-MS/MS conditions were as follows:
gradient elution with a flow rate of 0.4 mL min−1 using a mobile phase composed of 10
mM ammonium formate (0.05% v/v formic acid) and methanol. Elution started with 5%
methanol, maintained for 1 min, increased to 30% in 3 min, then to 60% in 8 min, and
finally to 100% in 2 min, holding at 100% for 2 min. The return to initial conditions was
achieved in 2 min and maintained for 2 min for re-equilibration. The LC system was
coupled with a 6495 triple quadrupole (QQQ) mass spectrometer (Agilent, Palo Alto, CA,
USA) equipped with an electrospray ionisation source. The mass spectrometry parameters
were as follows: capillary voltage, 4000 V; fragmentor, 166 V; nebuliser pressure, 40 psi;
sheath gas temperature, 250 ◦ C; sheath gas flow rate, 12 L min−1 ; gas temperature, 350
◦ C; and gas flow rate, 11 L min−1 . Analysis was performed in dynamic multiple reaction

monitoring mode in positive ionisation mode. Therefore, the precursor ions corresponded
to the molecular ions after protonation. The two most abundant transitions for each analyte
were monitored, with the most abundant transition used for quantification and the other for
confirmation. LC-MS/MS parameters for compounds can be seen in Table 5. LC–MS/MS
parameters were optimised by injection of individual and mixture standard solutions of
the selected compounds at 1 mg L−1 . The type and composition of mobile phase solvents
were optimised to achieve the highest compound ionisation to improve analytical signals
and lower limits of detection.
Molecules 2024, 29, 5478 12 of 15

Table 5. LC-MS/MS conditions and retention times for the selected compounds.

Product Ions
Compound Precursor Ion (m/z) CE (eV) Retention Time (min) Ratio
(Quantifier/Qualifier) (m/z)
TMP 291.2 261.1/229.8 28/24 7.79 98.2
4-OH-TMP 279.2 93.0/121.1 40/40 8.27 1.10
DM-TMP 277.3 261.4/123.0 28/44 6.80 63.1
SMX 254.3 92.1/108.0 28/28 8.96 76.1
AcSMX 296.3 134.0/108.1 24/28 10.74 49.8
SMX-GL 416.4 254.0/108.0 8/44 7.59 9.50
SDZ 251.3 92.1/156.0 28/12 6.45 98.0
AcSDZ 293.3 134.1/198.0 24/16 7.71 74.9
SMZ 279.3 186.0/92.0 16/36 8.28 76.4
AcSMZ 321.4 186.0/134.0 20/28 9.08 81.3
SMX-13 CIS 260.2 98.1/162.0 32/16 8.95 94.5
CE: collision energy; IS: internal standard. Parents compounds are marked in bold.

Resolution of analytes was higher than 1.0 for all antibiotics, with the exception of
the separation of SMZ and 4-OH-TMP. Even so, the differences in their precursor ion and
product ions employed for their monitoring make it possible to perform the determination
of all of them in just one injection.

3.5. Method Validation

The matrix effect was evaluated by comparison of the peak area of the target com-
pounds in matrix extract (Aextract ), after subtracting the peak area obtained from non-spiked
extracts (Ablank ), and in pure solvent standard solutions (Astandard ) applying Equation (3):

ME (%) = (Aextract − Ablank − Astandard )/Astandard × 100) (3)

Extraction recoveries were assessed by comparison of the peak areas obtained from the
spiked samples (Asample ) with those from spiked extracts (Aextract ) after blank correction
(Ablank ) following the Equation (4):

R (%) = (Asample − Ablank )/(Aextract − Ablank ) × 100 (4)

Accuracy (A), expressed as relative recovery, was determined by comparison of the

concentration obtained from spiked samples using matrix-matched calibration curves
(Cspiked sample ), after blank correction (Cblank ), with the spike concentration (Cspike concentration )
applying Equation (5):

A (%) = (Cspiked sample − Cblank ) × 100/Cspike concentration (5)

4. Conclusions
An MSPD-based method has been developed and validated for the first time for the
determination of four antibiotics (TMP and three sulphonamides, namely SDZ, SMZ, and
SMX) and six of their metabolites in mussel samples. Sample extraction and extract clean-
up were performed in a single step and based on MSPD, using low volumes of solvent
(7 mL) and sample amount (0.2 g). MQL values were in the range of 0.10–5.00 ng g−1 dm
for all compounds. Accuracy, expressed as relative recovery, was in the range of 80.8–118%.
Precision, expressed as relative standard deviation, obtained a mean value of 10%. The
application of the method revealed that mussels can be suitable bioindicators of marine pol-
lution since, after the exposure of the assay to TMP and SMX, concentrations in the mussels
were found to be up to 77.5 ng g−1 dm. The bioconcentration of TMP in M. galloprovincialis
was shown to be higher than that of SMX, attributable to its higher hydrophobicity.
The method was proven to be suitable for routine control of the presence of tar-
get antibiotics in mussel samples. Therefore, the proposed method offers a valuable
Molecules 2024, 29, 5478 13 of 15

tool for (i) obtaining information on the occurrence and fate of high-concern antibiotic
classes and their metabolites in the marine environment, as mussels comprise suitable
sentinel organisms applied for coastal pollution monitoring, (ii) supporting environmen-
tal risk assessments for antibiotic residues, (iii) evaluating the presence of antibiotics in
edible tissues of mussels before human consumption, (iv) assessing their bioconcentra-
tion/bioaccumulation/biomagnification in mussels, and (v) revealing antibiotic overuse
and environmental burden. Obtaining this information falls within the objectives of “One
Health”, which is an approach to optimise the health of humans, animals, and ecosystems
by integrating these fields rather than treating them separately.

Supplementary Materials: The following supporting information can be downloaded at https:

//www.mdpi.com/article/10.3390/molecules29225478/s1, Table S1: Box–Behnken design matrix for
the optimisation of clean-up sorbent amount; Table S2: Method application to mussels (expressed as
ng g−1 dm). Figure S1: LC-MS/MS chromatogram of a 10 ng g−1 dm spiked mussel sample.
Author Contributions: Conceptualisation, T.G.F. and E.A.; methodology, I.A. and J.M.; validation,
T.G.F.; formal analysis, C.M.; investigation, C.M. and N.G.-C.; resources, C.M. and N.G.-C.; data
curation, J.M. and I.A.; writing—original draft preparation, C.M.; writing—review and editing, T.G.F.,
J.M., J.L.S. and I.A.; visualisation, C.M.; supervision, E.A.; project administration, J.L.S. and E.A.;
funding acquisition, J.L.S. and E.A. All authors have read and agreed to the published version of
the manuscript.
Funding: This research was funded by the Ministerio de Ciencia e Innovación-Agencia Estatal de
Investigación MICIU/AEI/10.13039/501100011033, grant number PID2020-117641RB-I00. C. Mejías
acknowledges University of Seville for her predoctoral contract (grant number VI PPIT-US 2021 II.2A)
and for a research stay grant (VIIPPIT-2024-EBRV).
Institutional Review Board Statement: Ministerio de Ciencia, Innovación y Universidades (Spain)
PID2020-498 117641RB-I.
Informed Consent Statement: Not applicable.
Data Availability Statement: Data will be made available on request.
Conflicts of Interest: The authors declare no conflicts of interest.

References
1. European Centre for Disease Prevention and Control (ECDC); European Food Safety Authority (EFSA); European Medicines
Agency (EMA). Antimicrobial consumption and resistance in bacteria from humans and food-producing animals. Fourth joint
inter-agency report on integrated analysis of antimicrobial agent consumption and occurrence of antimicrobial resistance in
bacteria from humans and food-producing animals in the EU/EEA. JIACRA IV 2019–2021. EFSA J. 2024, 22, e8589. [CrossRef]
2. Jiang, J.J.; Lee, C.L.; Fand, M.D. Emerging organic contaminants in coastal waters: Anthropogenic impact, environmental release
and ecological risk. Mar. Pollut. Bull. 2014, 85, 391–399. [CrossRef] [PubMed]
3. Alygizakis, N.A.; Slobodnik, J.; Thomaidis, N.S. Sources and occurrence of pharmaceutical residues in offshore seawater. In
Pharmaceuticals in Marine and Coastal Environments. Occurrence, Effects and Challenges in a Changing World; Durán-Álvarez, J.C.,
Jiménez-Cisneros, B., Eds.; Estuarine and Coastal Sciences Series; Elsevier: Amsterdam, The Netherlands, 2021; Volume 1,
pp. 329–350. [CrossRef]
4. Alygizakis, N.A.; Gago-Ferrero, P.; Borova, V.L.; Pavlidou, A.; Hatzianestis, I.; Thomaidis, N.S. Occurrence and spatial distribution
of 158 pharmaceuticals, drugs of abuse and related metabolites in offshore seawater. Sci. Total Environ. 2016, 541, 1097–1105.
[CrossRef] [PubMed]
5. Borecka, M.; Siedlewicz, G.; Halinski, L.P.; Sikora, K.; Pazdro, K.; Stepnowski, P.; Bialk-Bielinska, A. Contamination of the
southern Baltic Sea waters by the residues of selected pharmaceuticals: Method development and field studies. Mar. Pollut. Bull.
2015, 94, 62–71. [CrossRef]
6. Benotti, M.J.; Brownawell, B.J. Distributions of pharmaceuticals in an urban estuary during both dry and wet-weather conditions.
Environ. Sci. Technol. 2007, 41, 5795–5802. [CrossRef]
7. World Health Organization. Fact Sheet: Antibiotic Resistance. 2020. Available online: https://www.who.int/news-room/fact-
sheets/detail/antibiotic-resistance (accessed on 5 October 2024).
8. McEneff, G.; Barron, L.; Kelleher, B.; Paull, B.; Quinn, B. A year-long study of the spatial occurrence and relative distribution
of pharmaceutical residues in sewage effluent, receiving marine waters and marine bivalves. Sci. Total Environ. 2014, 476–477,
317–326. [CrossRef]
Molecules 2024, 29, 5478 14 of 15

9. Serra-Compte, A.; Álvarez-Muñoz, D.; Rodríguez-Mozaz, S.; Barceló, D. Multi-residue for the determination of antibiotics and
some of their metabolites in seafood. Food Chem. Toxicol. 2017, 104, 3–13. [CrossRef]
10. Omar, T.F.T.; Aris, A.Z.; Yusoff, F.M.; Mustafa, S. Occurrence and level of emerging organic contaminant in fish and mollusk from
Klang River estuary, Malaysia and assessment on human health risk. Environ. Pollut. 2019, 248, 763–773. [CrossRef]
11. Van Hamelsveld, S.; Kurenbach, B.; Paull, D.J.; Godsoe, W.A.; Ferguson, G.C.; Heinemann, J.A. Indigenous food sources as
vectors of Escherichia coli and antibiotic resistance. Environ. Pollut. 2023, 334, 122155. [CrossRef]
12. Digka, N.; Patsiou, D.; Hatzonikolakis, Y.; Raitsos, D.E.; Skia, G.; Koutsoubas, D.; Dimitriadis, C.; Tsangaris, C. Microplastic
ingestion in mussels from the East Mediterranean Sea: Exploring its impacts in nature and controlled conditions. Sci. Total
Environ. 2024, 946, 174268. [CrossRef]
13. Gonçalves, C.; Martins, M.; Sobral, P.; Costa, P.M.; Costa, M.H. An assessment of the ability to ingest and excrete microplastics by
filter-feeders: A case study with the Mediterranean mussel. Environ. Pollut. 2019, 245, 600–606. [CrossRef] [PubMed]
14. Cruz, P.; Cuccaro, A.; Pretti, C.; He, Y.; Soares, A.M.V.M.; Freitas, R. Comparative subcellular responses to pharmaceutical
exposures in the mussel Mytilus galloprovincialis: An in vitro study. Environ. Toxicol. Pharmacol. 2023, 104, 104314. [CrossRef]
[PubMed]
15. Álvarez-Muñoz, D.; Huerta, B.; Fernández-Tejedor, M.; Rodríguez-Mozaz, S.; Barceló, D. Multi-residue method for the analysis of
pharmaceuticals and some of their metabolites in bivalves. Talanta 2015, 136, 174–182. [CrossRef] [PubMed]
16. McEneff, G.; Barron, L.; Kelleher, B.; Paull, B.; Quinn, B. The determination of pharmaceutical residues in cooked and uncooked
marine bivalves using pressurised liquid extraction, solid-phase extraction and liquid chromatography-tandem mass spectrometry.
Anal. Bioanal. Chem. 2013, 405, 9509–9521. [CrossRef]
17. Fernández-Torres, R.; Bello-López, M.A.; Olias-Consentino, M.; Callejón-Mochón, M.; Pérez-Bernal, J.L. Application of enzymatic
probe sonication extraction for the determination of selected veterinary antibiotics and their main metabolites in fish and mussel
samples. Anal. Chim. Acta 2010, 675, 156–164. [CrossRef]
18. Fernández-Torres, R.; Bello-López, M.A.; Olias-Consentino, M.; Callejón-Monchón, M.; Ramos-Payán, M. Enzymatic-microwave
assisted extraction and high-performance liquid chromatography-mass spectrometry for the determination of selected veterinary
antibiotics in fish and mussel samples. J. Pharm. Biomed. Anal. 2011, 54, 1146–1156. [CrossRef]
19. Barker, S.A. Matrix solid phase dispersion (MSPD). J. Biochem. Biophys. Methods 2007, 70, 151–162. [CrossRef]
20. Carro, N.; Fernández, R.; Cobas, J.; García, I.; Ignacio, M.; Mouteira, A. Optimization of a modified Captiva EMR-lipid method
based on micro-matrix solid-phase dispersion coupled with gas chromatography-mass spectrometry for the determination of
nine bisphenols in mussel samples. Anal. Methods 2024, 16, 3957–3967. [CrossRef]
21. Carro, N.; Fernández, R.; García, I.; Ignacio, M.; Mouteira, A. Optimization of an analytical method based on miniaturized matrix
solid-phase dispersion combined with gas chromatography-mass spectrometry for the determination of bisphenols in mussel
samples. Food Chem. 2024, 4, 100714. [CrossRef]
22. Castellani, F.; Vitali, M.; Antonucci, A.; Del Morrone, G.; Cofone, L.; D’Ancona, G.; Pindinello, I.; Mattiucci, S.; Protano, C.
Optimization of a fast and sensitive method based on matrix solid-phase dispersion-LC-MS/MS for simultaneous determination
of phthalates and bisphenols in mussel simples. Food Addit. Contam. Part A 2024, 41, 537–549. [CrossRef]
23. Carro, N.; Mouteira, A.; García, I.; Ignacio, M.; Cobas, J. Fast determination of phthalates in mussel samples by micro-matrix
solid-phase dispersion (micro-MSPD) coupled with GC–MS/MS. J. Anal. Sci. Technol. 2021, 12, 51. [CrossRef]
24. Carro, N.; López, A.; Cobas, J.; García, I.; Ignacio, M.; Mouteira, A. Development and Optimization of a Method for Organochlorine
Pesticides Determination in Mussels Based on Miniaturized Matrix Solid-Phase Dispersion Combined with Gas Chromatography–
Tandem Mass Spectrometry. J. Anal. Chem. 2021, 76, 603–612. [CrossRef]
25. Villaverde-de-Sáa, E.; Rodil, R.; Quintana, J.B.; Cela, R. Matrix solid-phase dispersion combined to liquid chromatography–
tandem mass spectrometry for the determination of paraben preservatives in mollusks. J. Chromatogr. A 2016, 1459, 57–66.
[CrossRef] [PubMed]
26. Carro, N.; Vilas, L.; García, I.; Ignacio, M.; Mouteira, A.M. Optimization of a method based on micro-matrix solid-phase dispersion
(micro-MSPD) for the determination of PCBs in mussel samples. Anal. Chem. Res. 2017, 11, 1–8. [CrossRef]
27. Rodríguez-González, N.; Gónzalez-Castro, M.J.; Beceiro-González, E.; Muniategui-Lorenzo, S. Development of a Matrix Solid
Phase Dispersion methodology for the determination of triazine herbicides in mussels. Food Chem. 2015, 173, 391–396. [CrossRef]
28. Negreira, N.; Rodríguez, I.; Rodil, R.; Rubí, E.; Cela, R. Optimization of matrix solid-phase dispersion conditions for UV filters
determination in biota samples. Int. J. Environ. Anal. Chem. 2013, 93, 11. [CrossRef]
29. Villaverde-de-Sáa, E.; Quintana, J.B.; Rodil, R.; Ferrero-Refojos, R.; Rubí, E.; Cela, R. Determination of perfluorinated compounds
in mollusks by matrix solid-phase dispersion and liquid chromatography–tandem mass spectrometry. Anal. Bioanal. Chem. 2012,
402, 509–518. [CrossRef]
30. Castro, V.; Montes, R.; Quintana, J.B.; Rodil, R.; Cela, R. Determination of 18 organophosphorus flame retardants/plasticizers in
mussel samples by matrix solid-phase dispersion combined to liquid chromatography-tandem mass spectrometry. Talanta 2020,
208, 120470. [CrossRef]
31. Villaverde-de-Sáa, E.; Valls-Cantenys, C.; Quintana, J.B.; Rodil, R.; Cela, R. Matrix solid-phase dispersion combined with gas
chromatography–mass spectrometry for the determination of fifteen halogenated flame retardants in mollusks. J. Chromatogr. A
2013, 1300, 85–94. [CrossRef]
Molecules 2024, 29, 5478 15 of 15

32. Sun, W.; Hu, X.; Xiang, Y.; Zhang, L.; Ye, N. Rapid fabrication of magnetic covalent organic frameworks for the effective extraction
and determination of sulfonamides in shrimp samples. Talanta Open 2023, 8, 100255. [CrossRef]
33. Mu, S.; Teng, T.; Zhou, R.; Liu, H.; Sun, H.; Li, J. Development and evaluation of a semi-automatic single-step clean-up apparatus
for rapid analysis of 18 antibiotics in fish samples. J. Food Comp. Anal. 2022, 113, 104718. [CrossRef]
34. Bitas, D.; Kabir, A.; Locatelli, M.; Samanidou, V. Food sample preparation for the determination of sulfonamides by high-
performance liquid chromatography: State-of-the-art. Separations 2018, 5, 31. [CrossRef]
35. AOAC International. Method Validation Programs (OMA/PVM Department), Including Appendix D: Guidelines for Collabo-
rative Study Procedures to Validate Characteristics of a Method of Analysis. 2000. Available online: https://www.aoac.org/
(accessed on 8 November 2024).
36. Serra-Compte, A.; Álvarez-Muñoz, D.; Solé, M.; Cáceres, N.; Barceló, D.; Rodríguez-Mozaz, S. Comprehensive study of
sulfamethoxazole effects in marine mussels: Bioconcentration, enzymatic activities and metabolomics. Environ. Res. 2019, 173,
12–22. [CrossRef] [PubMed]
37. DrugBank Online. Database for Drug and Drug Target Info. Available online: https://go.drugbank.com/ (accessed on 5 October 2024).
38. Chemical BooK. Sourcing and Integrating Center of Chemicals Materials in China. Available online: https://www.chemicalbook.com/
(accessed on 5 October 2024).
39. Li, S.; Shi, W.; Liu, W.; Li, H.; Zhang, W.; Hu, J.; Ke, Y.; Sun, W.; Ni, J. A duodecennial national synthesis of antibiotics in China’s
major rivers and seas (2005–2016). Sci. Total Environ. 2018, 615, 906–917. [CrossRef] [PubMed]
40. Malvar, J.L.; Santos, J.L.; Martín, J.; Aparicio, I.; Alonso, E. Routine analytical method for monitoring the main metabolites for
a recurrent group of parabens and pharmaceuticals in wastewater and tap water. Anal. Bioanal. Chem. 2019, 411, 6625–6635.
[CrossRef]
41. Scudi, J.V.; Plekss, O.J. Chemotherapeutic Activity of Some Sulfapyridine-1-Oxides. Proc. Soc. Exp. Biol. Med. 1958, 97, 639–641.
[CrossRef]
42. Yuan, S.; Liu, Z.; Yin, H.; Dang, Z.; Wu, P.; Zhu, N.; Lin, Z. Trace determination of sulfonamide antibiotics and their acetylated
metabolites via SPE-LC-MS/MS in wastewater and insights from their occurrence in a municipal wastewater treatment plant. Sci.
Total Environ. 2019, 653, 815–821. [CrossRef]
43. Mejías, C.; Martín, J.; Santos, J.L.; Aparicio, I.; Sánchez, M.I.; Alonso, E. Development and validation of a highly effective analytical
method for the evaluation of exposure of migratory birds to antibiotics and their metabolites by faeces analysis. Anal. Bioanal.
Chem. 2022, 414, 3373–3386. [CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.