Page 1 of 25
Environmental Science & Technology
A physiologically based pharmacokinetic model for inorganic and methylmercury in a marine fish Xun Wang† and Wen-Xiong Wang*,†
†Division of Life Science, The Hong Kong University of Science and Technology (HKUST), Clear Water Bay, Kowloon, Hong Kong
Corresponding author:
[email protected]
1 ACS Paragon Plus Environment
Environmental Science & Technology
1 2
ABSTRACT A physiologically based pharmacokinetic (PBPK) model was developed to simulate
3
the uptake, distribution and elimination of inorganic mercury [Hg(II)] and methylmercury
4
(MeHg) in a marine fish, Terapon jarbua. In this model, fish were schematized as a six-
5
compartment model by assuming that blood was the medium linking the exchange between
6
the different compartments. The transfer rates between blood and other compartments were
7
determined during a period of 10-day dietary Hg(II) or MeHg exposure, followed by a 30-day
8
depuration. For both Hg species, the exchange rates between liver and blood were high,
9
indicating that liver served as a ‘transferring station’ in the distribution. Their accumulation
10
in the kidney was relatively constant and low. The carcass (mainly muscle) represented a
11
large sink for both Hg(II) and MeHg with the highest input rate constants and relatively lower
12
output rate constants. Significant differences were observed in the rate constants between the
13
two Hg species, suggesting great variations in their exchange and transportation routes.
14
Modeling simulation for the first time demonstrated that the gill was the most important route
15
in Hg(II) elimination in marine fish, with a rate constant of 0.90 d-1. A long timeframe is
16
needed to study the exact rate of MeHg elimination in marine fish. This study showed that
17
the PBPK modeling provided critical information for the uptake, distribution and elimination
18
of Hg(II) and MeHg in the fish body, especially in elucidating the role of each compartment.
19 20 21
Keywords: Physiologically based pharmacokinetic model; Mercury; Methylmercury;
22
Transfer rate; Marine fish
2 ACS Paragon Plus Environment
Page 2 of 25
Page 3 of 25
23
Environmental Science & Technology
TOC art 24
3 ACS Paragon Plus Environment
Environmental Science & Technology
25
INTRODUCTION
26
Mercury (Hg) contamination in marine fish has aroused a worldwide concern for public
27
health. Among the different species of Hg, methylmercury (MeHg) is extremely toxic for
28
humans and can be readily biomagnified along food chains.1,2 Hg can be absorbed by marine
29
fish from aqueous and dietary phases, with the latter route generally considered to be the
30
primary exposure pathway.3,4 Several studies have determined the kinetics of Hg
31
accumulation in fish feeding on contaminated diets.5-7 However, most of these earlier studies
32
only focused on the kinetics of Hg from the perspectives of whole-body5,8 or isolated organs
33
(e.g. gill, intestine).9,10 Little information is available on the transfer of Hg between organs
34
(or tissue groups). It is therefore important to quantify the distribution kinetics of dietary Hg
35
between (among) organs.
36
Previous studies have shown that inorganic Hg [Hg(II)] and MeHg differed greatly in the
37
uptake and accumulation by fish, with a greater trophic transfer of MeHg over Hg(II), which
38
resulted in >70% of Hg in whole wild fish presenting as MeHg.11,12 Besides that, Hg(II) was
39
inclined to be accumulated in liver and kidney,13 whereas MeHg was primarily accumulated
40
in the muscle.14 Compared to Hg(II), the greater uptake and accumulation of MeHg can be
41
attributed to its higher assimilation efficiencies (AEs) and lower excretion rates in whole
42
fish.5,8,15 However, the exact roles of different organs (or tissue groups) in the uptake and
43
distribution of Hg, and their contributions to the different behaviors of Hg(II) and MeHg in
44
fish are still not well understood.
45
Physiologically based pharmacokinetic (PBPK) modeling has been increasingly used to
46
describe the absorption, distribution and excretion of compounds among multiple tissues or
47
organs of aquatic organisms.16-18 Combining experimental measurements and PBPK
48
modeling allows researchers to analyze and explain the temporal changes in the
49
concentrations of compounds in biological compartments of the exposed organism, taking
50
into account the effects of exposure scenarios, routes, doses and species.19,20 Therefore,
51
PBPK modeling can not only help develop a deeper understanding of the roles, kinetics, and
52
relative importance of the organs involved in the uptake, distribution and excretion of Hg, but
53
can also elucidate the underlying reasons for the different biokinetics of Hg(II) and MeHg in
54
fish.
55
Applications of PBPK modeling in describing the kinetics of Hg uptake, distribution and
56
elimination in fish are very limited.21,22 Ribeiro et al.22 quantified the distribution kinetics of 4 ACS Paragon Plus Environment
Page 4 of 25
Page 5 of 25
Environmental Science & Technology
57
MeHg in arctic charr (Salvelinus alpinus) with a simple three-compartment catenary model,
58
which divided the fish into well-perfused viscera and blood (VB), gut (G) and the rest of
59
body (R), and found that the translocation rates between compartments were surprisingly low.
60
Based on a multi-compartmental model, Leaner and Mason23 determined the transfer rates
61
between organs of sheepshead minnows (Cyprinodon variegatus) after a single dose of
62
MeHg-spiked food, and observed that the exchange between blood and visceral organs was
63
relatively low, but the calculated transfer rate from viscera to the rest of body was higher than
64
those measured for larger fish. However, these studies only described the uptake and
65
distribution of MeHg within fish body without addressing the elimination process. The
66
related routes and transfer rates for Hg(II) and MeHg elimination thus remain elusive. In the
67
present study, we for the first time developed a PBPK model of Hg(II) and MeHg in a marine
68
fish Terapon jarbua, with the aims to simulate the uptake, distribution and elimination of
69
Hg(II) and MeHg and determine their kinetics between the organs (or tissue groups). We
70
then explored the roles of each organ in these processes using the PBPK modeling approach.
71 72
MATERIALS AND METHODS
73
Fish and Diet. Juvenile thornfish T. jarbua (7-8 cm in length, ~4 g) were collected from a
74
fish farm located in Sai Kung, Hong Kong, and were acclimated in natural sand-filtered Clear
75
Water Bay seawater (pH = 8.0, salinity = 33 psu, DOC = 4.2 mg L-1, dissolved Hg = 0.2 ng L-
76
1
77
(Spectrum small fish formula, New Life International, Inc.) for 2 weeks at a daily rate of 4%
78
body weight. The clean fish diet contained 0.037±0.002 µg total Hg g-1, 0.022±0.004 µg
79
MeHg g-1 (n = 4), with the macronutrient being 38% crude protein, 7% fat, 5% fiber and 8%
80
ash. The concentrations of Hg(II) and MeHg in the spiked food were set as 50 µg g-1 and 1.5
81
µg g-1, respectively. Since the duration of exposure was within weeks, we employed a
82
relatively high Hg(II) spiked concentration to ensure an observable Hg(II) trend. Briefly, 100
83
grams of clean food pellets were incubated with 125 mL of freshly prepared solution of
84
HgCl2 or MeHgCl in MiliQ water for 4 h, and then dried at room temperature for two days.
85
The measured concentrations of Hg(II) and MeHg in spiked fish diet were 47.6±1.04 µg g-1
86
and 1.56±0.04 µg g-1, respectively (n = 4).
87
Fish Exposure, Depuration and Sampling. Fish were randomly divided into 10 tanks (size
88
of 60 × 30 × 45 cm3, 32 fish per tank), with 4 tanks for each Hg treatment and 2 tanks for
) at 25 oC under a 14 h light:10 h dark regime. Fish were fed clean commercial fish diet
5 ACS Paragon Plus Environment
Environmental Science & Technology
89
control, and then acclimated under the same conditions as described above for another week
90
before exposure. The fish were not fed for 2 d before Hg exposure in order to empty the gut
91
and maximize their digestion and assimilation of Hg-spiked food. The dietary-exposed fish
92
were fed the Hg-spiked fish diet for 10 d (exposure) and then were fed clean fish diet for 30 d
93
(depuration), whereas the control fish were fed clean fish diet during the entire period. All the
94
fish were fed twice a day at a daily rate of 4% body weight. The diet consumption time by
95
fish was limited to 1 h. Fish feeding behavior was monitored and almost all food pellets
96
(>95%) were found to be eaten; the remaining food pellets and feces were then siphoned off.
97
For all the treatments, the seawater was cycled at a flow rate of 1.5 L/min to ensure that the
98
water was kept clean.
99
Sampling of the Hg-exposed fish took place at 0, 2, 4, 6, 8, 10, 11, 12, 14, 16, 20, 25, 31
100
and 40 d, whereas that of the control fish took place every 5 d during the experiment
101
(primarily because the change of Hg concentration in the fish body was small for the control
102
fish). For each treatment at each time point, eight fish (two from each of the four tanks) were
103
randomly collected. Four of the collected fish were rinsed by deionized water and narcotized
104
in cold ice water, then were dissected into gill, liver, intestine, kidney and carcass (mostly
105
muscle) immediately after blood was collected (100 to 150 µL) from caudal fin by capillary
106
pipet. The remaining four fish were sampled as whole body. All samples were weighed for
107
fresh weight, freeze-dried at -80 oC (except blood), and stored at -20 oC for further analysis.
108
Chemical Analysis. THg concentrations in the tissues were analyzed followed the method
109
of EPA 7474 with a few modifications. Briefly, the homogenized tissues were digested in
110
aqua regia in a heating block at 80 oC overnight. In the hydrochloride/bromate/bromide
111
mixture (Sigma–Aldrich), mercury was oxidized by stannous chloride (Sigma–Aldrich) and
112
analyzed by cold vapor atomic fluorescence spectrometry (CVAFS, Brooks Rand Model III).
113
MeHg analysis followed the method 1630.24 Tissues were digested in 25% KOH/methanol
114
solution in an oven at 80 oC for 4 h, and were converted to volatile MeHg after addition of
115
citrate buffer and freshly thawed 1% NaBEt4 solution. MeHg was determined with a MERX
116
Automatic Methylmercury System (Brooks Rand Laboratories, Seattle, WA). Standard
117
reference materials (Fish protein DORM-4, National Research Council of Canada) were
118
concurrently digested and measured for total mercury and MeHg. Recoveries of the
119
standards were 93-101% for THg and 96-106% for MeHg.
6 ACS Paragon Plus Environment
Page 6 of 25
Page 7 of 25
120
Environmental Science & Technology
Model Development. Hg(II) and MeHg taken up with food pellets were firstly received in
121
the intestine (I) of the fish, and were then absorbed across the intestinal wall and transferred
122
into blood (B). Since blood was directly in contact with the other organs (or tissue groups),
123
metals were then distributed to the gill (G), liver (L), kidney (K) and carcass (C) via blood
124
stream. The Hg previously transferred and stored in these organs was transferred back into
125
blood and then redistributed around the fish body. This process was simulated by a
126
pharmacokinetic model as shown in Fig. 1. Each of the heavily perfused organs (gill, liver
127
and kidney) was treated as a separate compartment, since liver had been shown to play an
128
important role in Hg redistribution in fish,25 while gill and kidney might serve as the
129
excretion route of Hg by fish.26 The rate coefficient, k(i, j) (d-1) represents the transfer rate
130
constant between the compartments, where the first subscript (i) refers to the receptor organ
131
and the second (j) refers to the source. The coefficient k(0, 2) represents the elimination rate
132
constant of Hg through feces (fast phase) from intestine to outside.
133
The SAAM II modeling software version 2.3.1 (SAAM Institute, University of
134
Washington, Seattle, WA, USA)27 was used for model development and parameter
135
calculation for each Hg species. This software has been widely used to simulate the kinetic
136
processes in multiple compartments,16,28,29 due to its convenience in the production of
137
differential equations of the model structures and association of these equations to the
138
experimental data. Assuming that the transfer of Hg between the compartments follows first-
139
order kinetics, the flux between compartments can be expressed by the following equations:
140
F(j, i) = k(i, j)•Qj
141
where F(j, i) is the mass flux (µg d-1) of Hg from the source compartment (j) to the receptor
142
compartment (i) and Qj refers to the mass (µg) of Hg located in the source compartment (j) at
143
the time point t. The equations for the different compartments can be expressed by the
144
following equations:
145
Blood
dQB/dt = k(1,2)•QI + k(1,3)•QL + k(1,4)•QG + k(1,5)•QK + k(1,6)•Qc - [k(2,1) + k(3,1) + k(4,1) + k(5,1) + k(6,1)]•QB
146 147
Intestine dQI/dt = D + k(2,1)•QB - k(1,2)•QI - k(0,2)•QI
148
Liver
149
Gill
dQL/dt = k(3,1)•QB - k(1,3)•QL dQG/dt = k(4,1)•QB - k(1,4)•QG - k(0,4)•QG 7 ACS Paragon Plus Environment
Environmental Science & Technology
150
Kidney
151
Carcass dQC/dt = k(6,1)•QB - k(1,6)•QC
152
where parameter D refers to the amount of Hg assimilated in the fish body daily (µg d-1). The
153
coefficients k(0,2) and k(0,4) refer to the elimination rates of Hg through feces and gills to
154
outside environments, respectively. Qj equals to the concentration of Hg in the compartment
155
(Cj) multiplied by the fresh weight (wj). The total weight of blood in fish body is calculated
156
by assuming 40 ml blood/kg tissue in teleosts and 1 mL blood ≈ 1 g.30 The fish weights
157
parameters used in the calibration are listed in Table S1.
158
dQK/dt = k(5,1)•QB - k(1,5)•QK
When the Hg was taken up along with food pellets into the intestine lumen, some would be
159
absorbed by intestine cells and transported into the blood, whereas the residuals would be
160
eliminated through feces. In this study, the Heaviside function was introduced to simulate the
161
elimination process of Hg(II) and MeHg through feces and to calculate their residence time in
162
the lumen. The Heaviside function is a discontinuous step function whose value is zero for
163
negative argument and one for positive argument. It can be used in operational calculus for
164
the solution of differential equations, and represents a signal that switches on at a specified
165
time and stays switched on indefinitely. The Heaviside function is expressed by the
166
following equation:
167
heaviside = 0.5•(1.0 + atan[lamba•(t - tlag)]•2/π)
168
where atan is the arctangent function which is used to obtain a smooth approximation of the
169
step function, t is the time, and tlag is the value at which the Heaviside function changes
170
between 0 and 1. For values of t less than tlag, the value of the Heaviside function equals to
171
0; for values of t greater than tlag, the value equals to 1. The parameter, lambda, controls the
172
sharpness of the function, i.e. how fast the values change between zero and 1. In this study,
173
tlag refers to the residence time of Hg(II) and MeHg in the intestine, and the elimination rate
174
is expressed by the following equation:
175
k(0,2) = (1-heaviside)*loss1
176
Thus, Hg was eliminated through feces at a constant rate of loss1 during the residence period
177
(t < tlag, k(0,2) = loss1). This process would be terminated after that period (t ≥ tlag, k(0,2) = 0).
8 ACS Paragon Plus Environment
Page 8 of 25
Page 9 of 25
Environmental Science & Technology
The set of different model equations was solved using the Runge-Kutta integrator. The
178 179
SAAM II software is very functional since it can evaluate the standard deviations of the
180
parameter estimates, the parameter correlation matrixes, the weighted sums of squares of the
181
residuals, and the appearance of the data-model plots. To evaluate the adequacy of the
182
parameter estimates, we chose a correlation coefficient threshold of 0.9, as advocated by
183
Miller et al.31
184 185
RESULTS AND DISCUSSION Uptake and Distribution Kinetics of Hg(II) and MeHg. In the present study, fish (T.
186 187
jarbua) were schematized as a six-compartment pharmacokinetic model based on the
188
biological properties, containing blood, intestine, gill, liver, kidney and carcass. The
189
concentrations of Hg(II) and MeHg within the body of control fish did not show significant
190
change during the experiment period (Fig. S1), indicating that the clean fish diet had no
191
impact on the accumulation and distribution of Hg in fish. By contrast, both concentrations
192
of Hg(II) and MeHg in different compartments increased dramatically during the exposure
193
(0-10 d) (Fig. 2). During depuration (10-40 d), Hg(II) decreased continuously and resulted in
194
this trend at the end of experiment: kidney > liver ~ intestine > gill ~ carcass > blood. MeHg
195
in liver kept declining during depuration, whereas that in carcass kept relatively stable. At
196
the end, the levels of MeHg followed carcass ~ kidney > liver ~ intestine > gill > blood.
197
Similarly, previous studies have shown that kidney and liver were the target organs for Hg(II)
198
accumulation,13 whereas MeHg was primarily accumulated in the muscle.32 For whole-body
199
concentrations (Fig. 3), Hg(II) showed a similar trend to the compartments, whereas MeHg
200
increased up to day 12, decreased to day 16, and did not show obvious change towards the
201
end of experiment. The estimated rate constants of Hg(II) and MeHg are listed in Table 1,
202
and the simulation curves obtained by PBPK models are shown in Fig. 4 and Fig. 5. The
203
correlation coefficients were highly correlated (correlation coefficient > 0.9) in each of the
204
Hg treatment. Dietary metals were taken up from and eliminated into the environment through the
205 206
intestine. The coefficient from intestine to blood (k(1,2)) reflects the transfer of metals across
207
the inner membrane of the intestinal epithelium. The estimated k(1,2) of MeHg (1.31±0.156 d-
208
1
209
much more easily absorbed by the intestine. Previous study has reported that the assimilation
) was significantly higher than that of Hg(II) (0.04±0.004 d-1), indicating that MeHg was
9 ACS Paragon Plus Environment
Environmental Science & Technology
210
efficiency (AE) of MeHg (90.0±1.8%) was significantly higher than that of Hg(II)
211
(41.4±9.7%) in the same fish species.8 The different efficiencies in the absorption also
212
suggest that they were taken up through different routes. For MeHg, the uptake by gut tissues
213
was suggested to be comprised of about 40% active and 60% passive mechanisms, and the
214
active uptake was predominantly mediated by neutral amino-acid carriers, which targeted
215
MeHg-L-cysteine complexes, thus the uptake was dependent on the reaction between MeHg
216
and thiolates in the gastric and intestinal fluids.33 On the other hand, the major routes for
217
inorganic Hg entering into intestine cells were diffusion of electroneutral complexes10 and
218
active uptake through mucosal Na+ channels or divalent cation transporter.34,35 However, the
219
absorption can be blocked by the gut mucosal membrane due to the high affinity of mucus for
220
Hg(II) ions,36 and this may contribute to the relatively lower transfer rate of Hg(II). It is
221
worth noting that the simulated curve of Hg(II) concentrations in the intestine was lower than
222
the determined values during the uptake phase (Fig. 4). When the fish were sampled, the
223
residual feces in the intestine were inevitable. During the exposure period, the concentration
224
of Hg(II) in the intestine would be elevated by the residues containing some unabsorbed Hg.
225
Therefore, the actual concentrations of Hg(II) in the intestine might be lower than the
226
determined values. This also suggested the accuracy of the PBPK models.
227
Since the intestine was highly vascularized and kept in contact with blood, the exchange of
228
mercury was expected to be bi-directional. The rate constants of Hg(II) and MeHg from
229
blood to intestine were estimated to be 0.07 ± 0.402 d-1 and 7.93 ± 1.51 d-1, respectively.
230
Similarly, it has been reported that the rate constants of MeHg from blood to intestine in
231
sheepshead minnow were 7.34 ± 2.84 and 7.11 ± 2.68 d-1, respectively, after algae or flake-
232
food dosage.23 It has been reported that approximately 90% of MeHg was quickly taken up
233
into erythrocytes and bound to hemoglobin thiol groups within the cells after being
234
transferred into the blood.22,37 This binding was reversible and the loss of MeHg from
235
erythrocytes to the surrounding medium was determined by the relative proportion of thiol
236
groups inside and outside the cells.37 Thus, MeHg is more likely drawn out from the red
237
blood cells when the bloodstream is passing by the intestine, since there is large amount of
238
available amino acids with thiol groups existed in digestive fluids or derived from tissue
239
digestion.33 MeHg can then be transferred into intestinal cells through the b0,+ and B0,+ amino
240
acid transporters, which are involved in the transport of MeHg-Cys in the intestinal cells.38
241
However, little is known about how the Hg(II) in blood is exchanged with other organs,
242
although it has been suggested that Hg(II) was initially associated almost exclusively (91%) 10 ACS Paragon Plus Environment
Page 10 of 25
Page 11 of 25
Environmental Science & Technology
243
with plasma following exposure,39 and was associated dominantly with erythrocytes over
244
long time (>150 d).40 In this study, the relatively lower rate constants of Hg(II) from blood to
245
intestine suggest that the Hg(II) transfer across membranes is rather difficult.
246
Following the exchange across the intestinal epithelium to the blood, metals are distributed
247
to other visceral organs where they are either stored, detoxified, or further redistributed to
248
other parts of the body. The concentrations of Hg(II) and MeHg in the kidney were relatively
249
higher and stable, compared to other viscera organs (Fig. 2), whereas the proportions
250
transferred to kidney were lower and could only reach up to approximately 3% and 1% of the
251
body burden, respectively (Fig. S2). It appeared that kidney played a role as ‘binding site’,
252
where the binding and release of Hg were simultaneously processed. When the sites for Hg-
253
binding were saturated, the rates of binding and release would be balanced and the levels of
254
Hg in kidney were stable. Previous studies have shown that most Hg(II) was restricted to
255
lysosomes in the kidney cells,41 whereas MeHg was associated with numerous organelles as
256
well as dispersed throughout the cytoplasm.42 Thus, MeHg in the cytoplasm of kidney cells
257
was probably more easily released and transferred back to the blood, which might contribute
258
to a higher transfer rate from kidney to blood (1.42 ± 0.697 d-1) than that of Hg(II) (0.26 ±
259
0.053 d-1). Comparatively, larger proportions of both Hg(II) and MeHg were transported
260
from the blood to liver and reached a maximum of approximately 10% and 8% of the body
261
burden, respectively (Fig. S2). Besides, the exchange rates between liver and blood were
262
high (Table 1). All of these evidences indicated that the liver served as a ‘transfer site’ in the
263
distribution of Hg. This was similar to the previous reports for freshwater fish,22 in which the
264
liver was considered to be an important site for storage and possibly detoxification during the
265
first few days of exposure. Sequestration of Hg(II) and MeHg in the lysosomes of the liver
266
functioned as a storage mechanism to remove them from cellular process.41,42 GSH was
267
involved in Hg(II) and MeHg detoxification,43,44 but the related mechanisms remained
268
unclear. After the initial handling, Hg in the liver was transferred back to the blood and
269
redistributed into other parts of the body (especially to the muscle), thus the concentrations of
270
Hg kept decreasing during the depuration (Fig. 2). During the experiment, little Hg(II) or
271
MeHg was observed to be methylated or demethylated (Fig. S3), possibly due to the
272
extremely low rate of methylation25 and lack of selenium to promote the demethylation.45
273 274
The transfer rates of Hg(II) and MeHg from blood to carcass (muscle) were much higher compared to the other compartments (Table 1), indicating that metals were distributed 11 ACS Paragon Plus Environment
Environmental Science & Technology
275
effectively to this compartment, where >60% of Hg(II) and >90% of MeHg in the body were
276
located at the end of experiment (Fig. S2). The carcass was dominantly comprised of muscle
277
tissue, which contributed to the bulk of the body mass of fish, thus this compartment
278
represented a large sink for Hg. The storage of Hg in the muscle helped to remove it from
279
circulation and cellular function, thus protecting other tissues from the threat. The estimated
280
rate constants of Hg(II) (11.9 ± 1.98 d-1) and MeHg (10.8 ± 0.753 d-1) from blood to carcass
281
in our study were even higher than those of MeHg determined in a freshwater fish (sheeshead
282
minnow),23 with values of 8.39 ± 1.57 and 7.26 ± 1.37 d-1, respectively, after algae or flake-
283
food dosage. This suggested that the transfer of Hg into muscle tissue by marine fish was
284
more efficient than freshwater fish. It is worthy to note that the transfer rate of MeHg from
285
carcass to blood (0.04 ± 0.004 d-1) was much lower than that of Hg(II) (0.23 ± 0.040 d-1).
286
Previous studies have also shown that the elimination of MeHg from muscle tissue was a
287
slow process,14,15 whereas Hg(II) did not present a similarly high affinity to muscle tissue.46
288
Inside the muscle cells, MeHg was bound tightly to cysteine-rich proteins47 or peptide
289
glutathione in heat-stable protein, whereas Hg(II) was bound preferentially to heat-denatured
290
protein.8 Thus, the variations in the binding sites of Hg(II) and MeHg in muscle cells may be
291
responsible for the differences in transfer rates.
292
Hg transport across membranes can occur through channels or mediated by Hg binding to
293
proteins and low molecular weight thiols such as GSH and cysteine.48,49 Carrier-mediated
294
transport and protein binding are commonly described by Michaelis-Menten kinetics due to
295
the saturation of carriers- and binding sites at high concentrations. However, the
296
concentrations of Hg(II) and MeHg in our experiments were not high enough to induce the
297
saturation of carriers- and binding sites. Taking the blood as an example, the concentration
298
of Hg(II) and MeHg increased continuously following a linear-like trend during the exposure
299
period (0-10 d) (Fig. 2), suggesting that the transport process of Hg into blood was not
300
saturated. Similar trends were also observed in other organs such as gill, liver, kidney and
301
carcass. Based on these observations, the first-order kinetics was appropriate for describing
302
the uptake and distribution of Hg in our experiments. The application of Michaelis-Menten
303
kinetics may be needed under higher Hg concentrations or longer exposure period.
304
Elimination Kinetics of Hg(II) and MeHg. Generally, the elimination of Hg in fish body
305
occurred in two phases, with a fast phase associated with feces from intestine and a slow
306
phase associated with other organs.5 The results showed that the elimination of Hg(II) 12 ACS Paragon Plus Environment
Page 12 of 25
Page 13 of 25
Environmental Science & Technology
307
through feces (4.32 ± 0.324 d-1) was faster than that of MeHg (2.84 ± 0.141 d-1), consistent
308
with the finding that the whole-body elimination rate of Hg(II) (0.0957 ± 0.0179 d-1) in
309
Plectorhinchus gibbosus was much higher than that of MeHg (0.0116 ± 0.0059 d-1) during
310
the first 7 d in depuration.5 Based on the observations of whole-body levels (Fig. 2), the
311
elimination rate constant (ke, d-1) for slow phase Hg(II) could be estimated as the slope of the
312
least-squares regression equation relating Ln Ct/C0 to the number of the days,8 where Ct
313
referred to the concentrations of Hg(II) at the time points after day 14, and C0 referred to the
314
initial concentrations of Hg(II) at day 10. The estimated ke value for Hg(II) was 0.0258 ±
315
0.0017 d-1, which was lower than those measured in P. gibbosus (0.0547 ± 0.0093 d-1)5 and
316
Acanthopagrus schlegeli (0.042 ± 0.023 d-1),50 but comparable to those in Gambusia affinis
317
(0.021-0.042 d-1) and Lepomis microlophus (0.003-0.035 d-1).6 The present study also
318
simulated the elimination process of slow phase Hg(II) through gills, and the value of k(0,4)
319
was estimated to be 0.90 ± 0.148 d-1. Although high loss rates were observed in the gills
320
following aqueous and dietary exposures to Hg(II),26,51 the contributions of gills to the whole-
321
body elimination of Hg(II) have not been thoroughly investigated previously. Our modeling
322
strongly suggested that gills played the most important role in the excretion of Hg(II) in
323
marine fish. We further calculated the efflux rates (µg d-1) of the gills during the last 25 days
324
in the experiment (Fig. 6). Based on these data, the simulated elimination rate constant of
325
Hg(II) was 0.0303 ± 0.0025 d-1, which was very close to the actual measured ke (0.0258 ±
326
0.0017 d-1). Such comparison indicated that our simulation of the Hg(II) excretion was
327
satisfactory and confirmed the role of gills in Hg elimination. On the other hand, the
328
excretion of Hg(II) through urine or bile has been suggested by previous studies,40,52 although
329
the relative importance of these routes is still unknown. In this study, these two pathways
330
were also considered and the related rate constants were added in the model. However, the
331
simulation results showed that the related rate constants were infinitely close to zero,
332
suggesting that the excretion through urine or bile was minimal, thus these two routes could
333
be excluded in the present study. As mentioned above, Hg(II) entered into cells through
334
mucosal Na+ channels, as well as possibly by passive diffusion. It can be speculated that
335
Hg(II) was transported to the gills and excreted into the environment through similar
336
pathways, since the gills were the important sites for ion exchange in fish.53 Further studies
337
are needed to test this notion.
338 339
In this study, the value of k(0,4) for MeHg was set to be zero since the elimination of slow phase MeHg was limited and undetectable. This might be caused by the following two 13 ACS Paragon Plus Environment
Environmental Science & Technology
340
reasons. First, the whole-body ke of MeHg in fish was very low, as shown by many previous
341
studies, such as in Gadus morhua (0.0018 ± 0.0008 d-1),14 Perca flavescens (approximately
342
0.0013 d-1)15 and A. schlegeli (0.0058 ± 0.0034 d-1).50 The elimination process may exceed
343
well beyond the time-frame of this study (30 d depuration). Second, the ke was usually
344
measured based on the radioisotope method, which is sufficiently enough to detect the minor
345
changes of MeHg radioactivity even within a relatively short time scale,54 whereas the
346
determination of MeHg concentrations in this study could not achieve such a high sensitivity.
347
This study demonstrated that PBPK models are helpful to describe the uptake, distribution
348
and elimination kinetics of Hg(II) and MeHg within the body of T. jarbua as well as to
349
elucidate the exact role of each compartment. For both Hg species, the intestine was the
350
major site for the uptake and fast phase elimination, and the liver served as a ‘transfer site’,
351
whereas the kidney was the ‘binding site’ in the distribution. The carcass represented a large
352
sink where most Hg was stored. Compared to Hg(II), MeHg exerted higher exchange rates
353
between blood and viscera organs, but much lower transfer rates from carcass to blood,
354
indicating the great differences in the exchange and transportation routes between these two
355
species. Our study using modeling approach showed that gills played the most important role
356
in the excretion of Hg(II) during the slow phase in marine fish. Since the excretion of MeHg
357
in marine fish is usually a slow process, a longer timeframe study may be required to
358
investigate the exact rate of MeHg elimination.
359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374
REFERENCES
(1) Harris, H. H.; Pickering, I. J.; George, G. N. The chemical form of mercury in fish. Science. 2003, 301 (5637), 1203-1203. (2) Chan, H. M.; Egeland, G. M. Fish consumption, mercury exposure, and heart diseases. Nutr Rev. 2004, 62 (2), 68-72. (3) Hall, B. D.; Bodaly, R. A.; Fudge, R. J. P.; Rudd, J. W. M.; Rosenberg, D. M. Food as the dominant pathway of methylmercury uptake by fish. Water Air Soil Poll. 1997, 100 (1-2), 13-24. (4) Wang, W.-X. Interactions of trace metals and different marine food chains. Mar Ecol Prog Ser. 2002, 243, 295-309. (5) Wang, W.-X.; Wong, R. S. K. Bioaccumulation kinetics and exposure pathways of inorganic mercury and methylmercury in a marine fish, the sweetlips Plectorhinchus gibbosus. Mar Ecol Prog Ser. 2003, 261, 257-268. (6) Pickhardt, P. C.; Stepanova, M.; Fisher, N. S. Contrasting uptake routes and tissue distributions of inorganic and methylmercury in mosquitofish (Gambusia affinis) and redear sunfish (Lepomis microlophus). Environ Toxicol Chem. 2006, 25 (8), 2132-2142. 14 ACS Paragon Plus Environment
Page 14 of 25
Page 15 of 25
375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416
Environmental Science & Technology
(7) Dang, F.; Wang, W.-X. Antagonistic interaction of mercury and selenium in a marine fish is dependent on their chemical species. Environ Sci Technol. 2011, 45 (7), 3116-3122. (8) Dang, F.; Wang, W.-X. Subcellular controls of mercury trophic transfer to a marine fish. Aquat Toxicol. 2010, 99 (4), 500-506. (9) Pedersen, T. V.; Block, M.; Part, P. Effect of selenium on the uptake of methyl mercury across perfused gills of rainbow trout Oncorhynchus mykiss. Aquat Toxicol. 1998, 40 (4), 361-373. (10) Hoyle, I.; Handy, R. D. Dose-dependent inorganic mercury absorption by isolated perfused intestine of rainbow trout, Oncorhynchus mykiss, involves both amiloride-sensitive and energydependent pathways. Aquat Toxicol. 2005, 72 (1-2), 147-159. (11) Bank, M. S.; Chesney, E.; Shine, J. P.; Maage, A.; Senn, D. B. Mercury bioaccumulation and trophic transfer in sympatric snapper species from the Gulf of Mexico. Ecol Appl. 2007, 17 (7), 21002110. (12) Wyn, B.; Kidd, K. A.; Burgess, N. M.; Curry, R. A. Mercury biomagnification in the food webs of acidic lakes in Kejimkujik National Park and National Historic Site, Nova Scotia. Can J Fish Aquat Sci. 2009, 66 (9), 1532-1545. (13) Elia, A. C.; Galarini, R.; Taticchi, M. I.; Dorr, A. J. M.; Mantilacci, L. Antioxidant responses and bioaccumulation in Ictalurus melas under mercury exposure. Ecotox Environ Safe. 2003, 55 (2), 162-167. (14) Amlund, H.; Lundebye, A. K.; Berntssen, M. H. G. Accumulation and elimination of methylmercury in Atlantic cod (Gadus morhua L.) following dietary exposure. Aquat Toxicol. 2007, 83 (4), 323-330. (15) Van Walleghem, J. L. A.; Blanchfield, P. J.; Hintelmann, H. Elimination of mercury by yellow perch in the wild. Environ Sci Technol. 2007, 41 (16), 5895-5901. (16) Van Campenhout, K.; Bervoets, L.; Redeker, E. S.; Blust, R. A kinetic model for the relative contribution of waterborne and dietary cadmium and zinc in the common carp (Cyprinus Carpio). Environ Toxicol Chem. 2009, 28 (1), 209-219. (17) Weijs, L.; Yang, R. S. H.; Covaci, A.; Das, K.; Blust, R. Physiologically based pharmacokinetic (PBPK) models for lifetime exposure to PCB 153 in male and female harbor porpoises (Phocoena phocoena): model development and evaluation. Environ Sci Technol. 2010, 44 (18), 7023-7030. (18) Parhizgari, Z.; Li, J. A physiologically-based pharmacokinetic model for disposition of 2,3,7,8-Tcdd in fathead minnow and medaka. Environ Toxicol Chem. 2014, 33 (5), 1064-1071. (19) Thomann, R. V.; Shkreli, F.; Harrison, S. A pharmacokinetic model of cadmium in rainbow trout. Environ Toxicol Chem. 1997, 16 (11), 2268-2274. (20) Krishnan, K.; Peyret, T., Physiologically based toxicokinetic (PBTK) modeling in ecotoxicology. In Ecotoxicology modeling, Springer: 2009; pp 145-175. (21) Rouleau, C.; Gobeil, C.; Tjalve, H. Pharmacokinetics and distribution of dietary tributyltin compared to those of methylmercury in the American plaice Hippoglossoides platessoides. Mar Ecol Prog Ser. 1998, 171, 275-284. (22) Ribeiro, C. A. O.; Rouleau, C.; Pelletier, E.; Audet, C.; Tjalve, H. Distribution kinetics of dietary methylmercury in the arctic charr (Salvelinus alpinus). Environ Sci Technol. 1999, 33 (6), 902-907.
15 ACS Paragon Plus Environment
Environmental Science & Technology
417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459
(23) Leaner, J. J.; Mason, R. P. Methylmercury uptake and distribution kinetics in sheepshead minnows, Cyprinodon variegatus, after exposure to CH3Hg-spiked food. Environ Toxicol Chem. 2004, 23 (9), 2138-2146. (24) USEPA, Method 1630: methylmercury in water by distillation, aqueous ethylation, purge and trap, and CVAFS; EPA-821-R-01-020; Washington, DC. 2001. (25) Wang, R.; Feng, X. B.; Wang, W.-X. In vivo mercury methylation and demethylation in freshwater Tilapia quantified by mercury stable isotopes. Environ Sci Technol. 2013, 47 (14), 79497957. (26) Handy, R. D.; Penrice, W. S. The influence of high oral doses of mercuric chloride on organ toxicant concentrations and histopathology in rainbow trout, Oncorhynchus mykiss. Comp Biochem Phys C. 1993, 106 (3), 717-724. (27) Barrett, P. H. R.; Bell, B. M.; Cobelli, C.; Golde, H.; Schumitzky, A.; Vicini, P.; Foster, D. M. SAAM II: simulation, analysis, and modeling software for tracer and pharmacokinetic studies. Metabolism. 1998, 47 (4), 484-492. (28) Carneiro, M. F. H.; Souza, J. M. O.; Grotto, D.; Batista, B. L.; Souza, V. C. D.; Barbosa, F. A systematic study of the disposition and metabolism of mercury species in mice after exposure to low levels of thimerosal (ethylmercury). Environ Res. 2014, 134, 218-227. (29) Consoer, D. M.; Hoffman, A. D.; Fitzsimmons, P. N.; Kosian, P. A.; Nichols, J. W. Toxicokinetics of perfluorooctanoate (PFOA) in rainbow trout (Oncorhynchus mykiss). Aquat Toxicol. 2014, 156, 65-73. (30) Horn, M. H., Feeding and digestion. In The physiology of fishes, Evans, D. H., Ed. CRC, Boca Raton, FL: USA, 1998; pp 65-97. (31) Miller, L. V.; Krebs, N. F.; Hambidge, K. M. Development of a compartmental model of human zinc metabolism: identifiability and multiple studies analyses. Am J Physiol-Reg I. 2000, 279 (5), R1671-R1684. (32) Cizdziel, J.; Hinners, T.; Cross, C.; Pollard, J. Distribution of mercury in the tissues of five species of freshwater fish from Lake Mead, USA. J Environ Monitor. 2003, 5 (5), 802-807. (33) Leaner, J. J.; Mason, R. P. Factors controlling the bioavailability of ingested methylmercury to channel catfish and Atlantic sturgeon. Environ Sci Technol. 2002, 36 (23), 5124-5129. (34) Vazquez, M.; Velez, D.; Devesa, V.; Puig, S. Participation of divalent cation transporter DMT1 in the uptake of inorganic mercury. Toxicology. 2015, 331, 119-124. (35) Vazquez, M.; Devesa, V.; Velez, D. Characterization of the intestinal absorption of inorganic mercury in Caco-2 cells. Toxicol in Vitro. 2015, 29 (1), 93-102. (36) Vazquez, M.; Calatayud, M.; Velez, D.; Devesa, V. Intestinal transport of methylmercury and inorganic mercury in various models of Caco-2 and HT29-MTX cells. Toxicology. 2013, 311 (3), 147-153. (37) Giblin, F.; Massaro, E. J. The erythrocyte transport and transfer of methylmercury to the tissues of the rainbow trout (Salmo gairdneri). Toxicology. 1975, 5 (2), 243-254. (38) Vazquez, M.; Velez, D.; Devesa, V. Participation of b0,+ and B0,+ systems in the transport of mercury bound to cysteine in intestinal cells. Toxicol Res-Uk. 2015, 4 (4), 895-900. (39) Olson, K. R.; Fromm, P. O. Mercury uptake and ion distribution in gills of rainbow trout (Salmo gairdneri): tissue scans with an electron microprobe. J Fish Res Board Can. 1973, 30 (10), 1575-1578. 16 ACS Paragon Plus Environment
Page 16 of 25
Page 17 of 25
460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495
Environmental Science & Technology
(40) Schultz, I.; Peters, E.; Newman, M. Toxicokinetics and disposition of inorganic mercury and cadmium in channel catfish after intravascular administration. Toxicol Appl Pharm. 1996, 140 (1), 3950. (41) Baatrup, E.; Nielsen, M. G.; Danscher, G. Histochemical demonstration of two mercury pools in trout tissues: mercury in kidney and liver after mercuric chloride exposure. Ecotox Environ Safe. 1986, 12 (3), 267-282. (42) Baatrup, E.; Danscher, G. Cytochemical demonstration of mercury deposits in trout liver and kidney following methyl mercury intoxication: differentiation of two mercury pools by selenium. Ecotox Environ Safe. 1987, 14 (2), 129-141. (43) Monteiro, D. A.; Rantin, F. T.; Kalinin, A. L. Inorganic mercury exposure: toxicological effects, oxidative stress biomarkers and bioaccumulation in the tropical freshwater fish matrinx, Brycon amazonicus (Spix and Agassiz, 1829). Ecotoxicology. 2010, 19 (1), 105-123. (44) Larose, C.; Canuel, R.; Lucotte, M.; Di Giulio, R. T. Toxicological effects of methylmercury on walleye (Sander vitreus) and perch (Perca flavescens) from lakes of the boreal forest. Comp Biochem Phys C. 2008, 147 (2), 139-149. (45) Khan, M. A. K.; Wang, F. Y. Chemical demethylation of methylmercury by selenoamino acids. Chem Res Toxicol. 2010, 23 (7), 1202-1206. (46) Ribeiro, C. A. O.; Guimaraes, J. R. D.; Pfeiffer, W. C. Accumulation and distribution of inorganic mercury in a tropical fish (Trichomycterus zonatus). Ecotox Environ Safe. 1996, 34 (2), 190-195. (47) Lemes, M.; Wang, F. Y. Methylmercury speciation in fish muscle by HPLC-ICP-MS following enzymatic hydrolysis. J Anal Atom Spectrom. 2009, 24 (5), 663-668. (48) Bridges, C. C.; Zalups, R. K. Molecular and ionic mimicry and the transport of toxic metals. Toxicology and applied pharmacology. 2005, 204 (3), 274-308. (49) Onsanit, S.; Wang, W.-X. Sequestration of total and methyl mercury in different subcellular pools in marine caged fish. J Hazard Mater. 2011, 198, 113-122. (50) Dang, F.; Wang, W.-X. Why mercury concentration increases with fish size? Biokinetic explanation. Environ Pollut. 2012, 163, 192-198. (51) Weisbart, M. The distribution and tissue retention of mercury-203 in the goldfish (Carassius auratus). Can J Zool. 1973, 51 (2), 143-150. (52) Greif, R. L.; Du Vigneaud, M. Distribution of a radiomercury-labelled diuretic (Chlormerodrin) in tissues of marine fish. Biol Bull-US. 1959, 117 (2), 251-257. (53) Kidd, K.; Batchelar, K., 5-Mercury. In Fish Physiology: Homeostasis and Toxicology of NonEssential Metals, Academic Press: 2011; Vol. 31, pp 237-295. (54) Wang, W.-X.; Fisher, N. S. Assimilation efficiencies of chemical contaminants in aquatic invertebrates: a synthesis. Environ Toxicol Chem. 1999, 18 (9), 2034-2045.
496 497 498
17 ACS Paragon Plus Environment
Environmental Science & Technology
499
AUTHOR INFORMATION
500
Corresponding Author
501
*Phone: (852) 23587346. E-mail:
[email protected].
502 503
ACKNOWLEDGEMENTS
504
We are grateful to the anonymous reviewers for their comments. This study was supported
505
by the National Key Basic Research Program of China (2013CB430004) and the General
506
Research Fund from the Hong Kong Research Grants Council (663112).
507 508
ASSOCIATED CONTENT
509
Supporting Information Available
510
Figure. S1. Concentrations of Hg(II) (a) and MeHg (b) in the different compartments (blood,
511
gill, carcass, liver, intestine and kidney) in the controlled fish during the experiment. Data are
512
mean + SD; Figure. S2. Proportions of Hg(II) (a) and MeHg (b) in different compartments
513
(blood, gill, carcass, liver, intestine and kidney) of Terapon jarbua during dietary exposure
514
(10 d) and depuration (30 d); Figure. S3. Ratio of Hg(II) and MeHg in the whole-body of
515
Terapon jarbua during dietary exposure (10 d) and depuration (30 d); Table S1. Fish weights
516
parameters used for calibration. This information is available free of charge via the Internet
517
at http://pubs.acs.org.
18 ACS Paragon Plus Environment
Page 18 of 25
Page 19 of 25
Environmental Science & Technology
Figure. 1 Schematic representation of Hg(II) and MeHg uptake and distribution in the different compartments of Terapon jarbua used in the PBPK model. Values of k are the model’s intercompartmental rate constant (d-1)
19 ACS Paragon Plus Environment
Environmental Science & Technology
Figure. 2 Concentrations of Hg(II) (a, b) and MeHg (c, d) in different compartments (blood, gill, carcass, liver, intestine and kidney) of Terapon jarbua during dietary exposure (10 d) and depuration (30 d). Data are means ± SD (n = 4).
20 ACS Paragon Plus Environment
Page 20 of 25
Page 21 of 25
Environmental Science & Technology
Figure. 3 Concentrations of Hg(II) and MeHg in the whole-body of Terapon jarbua during dietary exposure (10 d) and depuration (30 d). Data are means ± SD (n = 4).
21 ACS Paragon Plus Environment
Environmental Science & Technology
Figure. 4 The distribution kinetics of Hg(II) in different compartments of Terapon jarbua during dietary exposure (10 d) and depuration (30 d). Data points represent means ± SD (n = 4). Fitted curves were obtained by physiologically based pharmacokinetic (PBPK) models using the SAAM II modeling software (SAAM Institute, University of Washington, Seattle, WA, USA). Compartments: (a) blood, (b) intestine, (c) liver, (d) gill, (e) kidney, and (f) carcass.
22 ACS Paragon Plus Environment
Page 22 of 25
Page 23 of 25
Environmental Science & Technology
Figure. 5 The distribution kinetics of MeHg in different compartments of Terapon jarbua during dietary exposure (10 d) and depuration (30 d). Data points represent means ± SD (n = 4). Fitted curves were obtained by physiologically based pharmacokinetic (PBPK) models using the SAAM II modeling software (SAAM Institute, University of Washington, Seattle, WA, USA). Compartments: (a) blood, (b) intestine, (c) liver, (d) gill, (e) kidney, and (f) carcass.
23 ACS Paragon Plus Environment
Environmental Science & Technology
Figure. 6 The efflux rates (µg d-1) of whole-body (solid circles) and the gills (open circles) in Terapon jarbua during the last 25 days. The solid curve (whole-body) was calculated to be Y = 0.056e-0.0258t, and the dashed curve (the gills) was calculated to be Y = 0.062e-0.0303t.
24 ACS Paragon Plus Environment
Page 24 of 25
Page 25 of 25
Environmental Science & Technology
Table 1. Estimated rate constants for Hg(II) and MeHg in Terapon jarbua under dietary exposure (n = 4). Rate constant
Definition
(d-1)
Value ± SD
MeHg a
Value ± SD
k(2,1)
Blood to intestine
0.07 ± 0.402
7.93 ± 1.51
k(3,1)
Blood to liver
3.06 ± 0.830
6.10 ± 2.67
k(4,1)
Blood to gill
2.64 ± 1.10
1.16 ± 0.219
k(5,1)
Blood to kidney
0.58 ± 0.097
0.95 ± 0.466
k(6,1)
Blood to carcass
11.9 ± 1.98
10.8 ± 0.753
k(1,2)
Intestine to blood
0.04 ± 0.004
1.31 ± 0.156
k(1,3)
Liver to blood
0.41 ± 0.124
1.27 ± 0.587
k(1,4)
Gill to blood
0.11 ± 0.453
0.43 ± 0.096
k(1,5)
Kidney to blood
0.26 ± 0.053
1.42 ± 0.697
k(1,6)
Carcass to blood
0.23 ± 0.040
0.04 ± 0.004
k(0,2)
Intestine to feces
4.32 ± 0.324
2.84 ± 0.141
k(0,4)
Gill to water
0.90 ± 0.148
/
14.4 ± 0.191
11.4 ± 0.088
tlag (d)
a
Hg(II)
Loss time by intestine
SD = Standard deviation.
25 ACS Paragon Plus Environment
