ayllon 2000

Prévia do material em texto

Ž .Mutation Research 467 2000 177–186
www.elsevier.comrlocatergentox
Community address: www.elsevier.comrlocatermutres
Induction of micronuclei and other nuclear abnormalities in
European minnow Phoxinus phoxinus and mollie Poecilia
latipinna: an assessment of the fish micronucleus test
F. Ayllon), E. Garcia-Vazquez
UniÕersidad de OÕiedo, Departamento de Biologia Funcional, Area de Genetica, Facultad de Medicina, crJulian ClaÕeria srn,
33006 OÕiedo, Spain
Received 22 December 1999; received in revised form 7 March 2000; accepted 8 March 2000
Abstract
In this work, we have measured both micronuclei and other nuclear abnormalities in renal erythrocytes from European
minnow Phoxinus phoxinus and mollie Poecilia latipinna, with the aim to contribute to the standardisation of the
Ž . Ž .micronucleus test for fish species. Intraperitoneal injections of colchicine 10 mgrkg , cyclophosphamide 40 mgrkg , or
Ž .mitomycin C 20 mgrkg for 24 h induced diverse nuclear abnormalities in minnow erythrocytes, therefore nuclear
abnormalities should be added to micronuclei as genotoxicity indicators in fish micronucleus tests. The adequacy of
administration protocols based on intraperitoneal injections has been evaluated by injecting saline solution to both species:
single or double injections have not induced neither micronuclei nor other nuclear abnormalities in any case. Finally, the
differential sensitivity of both species to toxic heavy metals was evaluated by exposing individuals of both species to
Ž .different doses 0.17, 1.7, 2=1.7, and 3.4 mgrkg of cadmium and mercury for 24 h; we concluded that the mollie is
sensitive to both metals whereas the minnow is not sensitive to mercury. q 2000 Elsevier Science B.V. All rights reserved.
Keywords: Genotoxicity; Heavy metals; Micronucleus; Nuclear abnormalities; Phoxinus phoxinus; Poecilia latipinna
1. Introduction
The count of micronuclei has served as an index
of chromosome breaks and mitotic spindle apparatus
dysfunction for over 20 years. Some of the advan-
tages of the micronucleus test are its simplicity,
reliability, and sensitivity. Although originally devel-
) Corresponding author. Fax: q34-985103534.
Ž .E-mail address: ayllon@sauron.quimica.uniovi.es F. Ayllon .
oped for its application in mouse, it was subse-
w xquently modified by Hoofman and de Raat 1 for the
application in the laboratory to fish. It is widely
employed to assess the biological impact of water
w xpollution 2–4 and to test the genotoxicity of chemi-
cal compounds after direct or indirect exposure in
w xvivo 5–7 . For fish species, the micronucleus test
presents several advantages over other cytogenetic
studies such as sister chromatid exchanges or chro-
mosome aberrations, which are time-consuming and
not very effective due to the relatively large number
w xof small chromosomes of many fish species 8 .
1383-5718r00r$ - see front matter q 2000 Elsevier Science B.V. All rights reserved.
Ž .PII: S1383-5718 00 00033-4
Nevertheless, some problems have been reported
for the application of the micronucleus test in fish
species. One of them is the lack of consensus about
the type of abnormalities of the nuclear morphology
that should be considered micronuclei analogous. In
fish, there are several types of nuclear lesions whose
origin is not still understood. These nuclear abnor-
malities have been described and photographed by
w xCarrasco et al. 9 . Basically, they were classified as
notched, lobed, blebbed, and vacuolated, in addition
to the typical micronuclei. These abnormalities have
been used by some authors as a signal of cytogenetic
w x w xdamage in fish species 10,11 . Carrasco et al. 9
have not found a significant association between the
Žvariations of nuclear morphology including mi-
.cronuclei observed and the levels of chemical pollu-
tion in sediments or fish tissues, pointing out another
weakness of the micronucleus test in fish species: its
possible lack of sensitivity is due to the low and
variable frequency of micronuclei existing in wild
fish. This problem could be solved by analysing a
cell type with a high mitotic index, an essential
w xcondition for the micronuclei formation 12 . Most of
the piscine micronuclei surveys have been carried
w xout in peripheral blood erythrocytes 2,3 , but an
alternative possibility could be to try micronucleus
test in other tissues, like cephalic kidney, with a high
w xmitotic index 13 .
To test genotoxicity of different substances in fish
species, intraperitoneal injection is the most common
exposure method in laboratory conditions, because
there are many factors which are difficult to control
that could influence the metabolic process of product
absorption if other administration protocols are em-
w xployed 14 . However, this method causes stress for
the fish. The stress can induce metabolic changes in
fish; for example, it induces cortisol in the fish
w xyellow perch Perca flaÕescens 15 , as well as
changes in the number and affinity of cortisol recep-
w xtors in stripped bass Morone saxatilis 16 . In mam-
mals, the stress induces DNA abducts, chromosome
w xaberrations, and micronuclei 17 ; it could also influ-
ence the results of genotoxicity assays in fish, for
example by altering the normal cell cycle. Therefore,
the possible role of the stress associated to intraperi-
toneal injection should be evaluated to know its
possible interaction with the micronuclei induction
due to the compounds tested.
When assessing genotoxicity in fish, heavy metals
are very interesting group of elements due to their
strong impact on the stability of aquatic ecosystems
w xand their bioaccumulation in living organisms 18 .
Current available data indicate that many routine
tests such as bacterial mutation gene assays and
cytogenetic assays with mammalian cells are not
adequate for the detection of the genotoxic effects of
heavy metals in complex environmental mixtures
w x19 . Cadmium and mercury are considered two of
the most toxic metals by national and international
legislations. In addition to their toxicity, they have a
genotoxic effect that has been demonstrated in many
studies. However, in vivo genotoxicity tests for mer-
cury and cadmium compounds have been poorly
Ždeveloped in fish. For killifish Fundulus heterocli-
.tus , inorganic mercury resulted to be teratogenic
w x20 and methylmercury chloride was found to in-
w xduce chromosomal aberrations and micronuclei 21 .
w xManna and Sadhukhan 22 showed also the induc-
tion of micronuclei in Oreochromis mossambicus by
cadmium chloride.
The aim of this work is twofold. First, to con-
tribute to the standardisation of the micronucleus test
in fish species, we have investigated the type of
nuclear abnormalities that are induced by compounds
of well-known genotoxic activity to determine what
nuclear forms may be included in the micronucleus
test; in addition, we determined if the administration
Ž .protocol intraperitoneal injection induces micronu-
clei to validate such protocol for laboratory genotox-
icity tests. Second, we studied the genotoxic effects
of cadmium and mercury in two fish species, Phoxi-
nus phoxinus and Poecilia latipinna, and the suit-
ability of these two species as targets for laboratory
genotoxicity tests.
2. Material and methods
2.1. Fish employed
European minnow Pho. phoxinus individuals were
collected by electrofishing in a wild area of the Sella
Ž .River Asturias, North Spain , with low human influ-
ence. Mollie Poe. latipinna individuals were pur-
chased in a local pet shop. Only well-differentiated
males were used in this work. Before any treatment,
fish were acclimatised in tanks with well-oxygenated
Ž .dechlorinated water at 108C minnow , and 248C
Ž .mollie for at least 24 h. Fish were not allowed to
feed during the experimental period. For intraperi-
toneal injections, fish were anaesthetised with ethy-
Ž .lene-glicol-monophenylether Merck to avoid ani-
mal pain.
2.2. Induction of nuclear abnormalities by genotoxic
compoundsIn order to determine what type of nuclear abnor-
malities are really induced by genotoxic chemicals,
three well-known genotoxic compounds were em-
Ž . Žployed: colchicine 10 mgrkg , mitomycin C 20
. Ž .mgrkg , and cyclophosphamide 40 mgrkg ; the
doses were selected according to previous works
w x23,24 . Each compound was intraperitoneally in-
jected to 15 minnows in order to evaluate the induc-
tion of micronuclei and other nuclear abnormalities
Žin cephalic kidney erythrocytes. The control group 8
. Žindividuals was injected with saline solution NaCl
.0.9% . Chemicals were diluted for their administra-
tion in volumes of 50 ml. 24 h after the injection,
animals were sacrificed and the slides were prepared.
w xFollowing Carrasco et al. 9 , the nuclear abnor-
malities observed were classified into four cate-
gories. Briefly, micronuclei were only non-refractory
particles, with the same colour as the nucleus cell
and with a round or ovoid shape; blebbed nuclei
present a relatively small evagination of the nuclear
envelope, which seemed to contain euchromatin.
Lobed nuclei are those presenting evaginations larger
that the blebbed nuclei. Finally, a notched nucleus
presents an appreciable depth into a nucleus that
does not contain nuclear material.
2.3. Effect of administration protocol
To assess the possible induction of micronuclei by
the stress produced by intraperitoneal injections in
minnow and mollie, fish were injected with saline
Ž . Žsolution NaCl 0.9% . One group C1, 8 individuals
.of each species received a single intraperitoneal
injection with saline solution; the animals were sacri-
Žficed 24 h later. Another group of individuals C2, 8
.individuals of each species were twice injected with
saline solution. The animals were sacrificed 24 h
Žafter the last injection. One more group of fish B,
.10 minnows and 8 mollies was not injected and they
were kept together with the injected groups in the
same aquarium for 48 h, then sacrificed for micronu-
clei counting.
2.4. Induction of micronuclei by heaÕy metals
Ž .Cadmium chloride and mercury nitrate Sigma
Ž .were dissolved in saline solution NaCl 0.9 % and
diluted in order to be injected to fish in final vol-
umes of 50 mlrg body weight. Both metals were
Žintraperitoneally injected at three doses 0.17, 1.7
.and 3.4 mg of metal per kg of body weight to
different groups of fish, each group size being 12
individuals for both minnow and mollie; control
Ž .animals 12 minnows and 16 mollies were injected
with the same volume of saline solution. Double-dose
experiments with cadmium and mercury were per-
formed by injecting two consecutive doses of 1.7
mgrkg separated 24 h; then animals were sacrificed
24 h after the last dose. Twelve individuals per dose
and species were used. The doses were selected on
the basis of the results found by Sanchez-Galan et al.
w x25 .
2.5. Fish sacrifice and slide preparation
Ž .Fish were killed by overdosing 50 mlrl the
Ž .ethylene-glycol-monophenylether Merck anaesthet-
ics, then length and weight were recorded and a
Ž .portion around 20% of cephalic kidney was re-
moved by scraping and dragged along a slide in a
single layer of well-spread cells, then allowed to dry
for a few minutes. Two slides per animal were
prepared. Staining procedure followed Sanchez-
w xGalan et al. 4 . Briefly, slides were sequentially
stained with May-Grunwald for 2 min, May-¨
Grunwaldrdistilled water 1:1 for 3 min, and¨
Giemsardistilled water 1:6 for 10 min. Then slides
were rinsed with distilled water, allowed to dry, and
mounted with Eukitt.
ŽFrom each animal, 1000 renal erythrocytes 500
.per slide whenever possible were scored under 1000
= magnification, to determine the frequency of mi-
cronuclei and other nuclear lesions. Coded and ran-
domised slides were scored using a blind review by a
single observer.
Table 1
Biological characteristics of fish samples
Pho. phoxinus Poe. latipinna
N Weight Length N Weight Length
Ž . Ž .Control Group 8 3.308 "2.028 6.787 "1.175 n.a. n.a. n.a.
Ž . Ž . Ž .Colchicine 10 mgrkg 13 3.243 "1.863 6.030 "1.538 n.a. n.a. n.a.
Ž . Ž . Ž .Mitomycin C 20 mgrkg 10 5.273 "2.275 6.136 "3.363 n.a. n.a. n.a.
Ž . Ž . Ž .Cyclophosphamide 40 mgrkg 11 4.231 "1.670 5.840 "1.140 n.a. n.a. n.a.
Ž . Ž . Ž . Ž .B group 10 2.879 "1.281 5.350 "0.726 8 1.145 "0.353 4.287 "0.372
Ž . Ž . Ž . Ž .C1 group 6 3.143 "1.293 5.916 "0.835 8 1.463 "0.258 4.500 "0.991
Ž . Ž . Ž . Ž .C2 group 6 3.301 "0450 5.160 "0.288 8 1.428 "0.292 4.637 "0.362
Ž . Ž . Ž . Ž .Control group 12 3.225 "0.927 5.538 "0.714 16 1.446 "0.267 4.568 "0.724
Ž . Ž . Ž . Ž .Cd 0.17 mgrkg 11 3.548 "0.914 5.999 "0.989 6 1.136 "0.316 4.166 "0.314
Ž . Ž . Ž . Ž .Cd 1.7 mgrkg 8 3.883 "1.190 5.838 "0.566 8 1.581 "0.383 4.625 "0.337
Ž . Ž .Cd 3.4 mgrkg 8 3.785 "1.130 6.025 "0.636 n.a. n.a. n.a.
Ž . Ž . Ž . Ž .Cd 2=1.7 mgrkg 6 3.926 "0.955 6.416 "0.614 6 1.498 "0.262 4.550 "0.339
Ž . Ž . Ž . Ž .Hg 0.17 mgrkg 7 3.425 "0.664 5.838 "0.566 6 1.290 "0.286 3.983 "0.292
Ž . Ž . Ž . Ž .Hg 1.7 mgrkg 11 4.022 "2.014 6.854 "1.227 6 1.353 "0.376 4.083 "0.231
Ž . Ž .Hg 3.4 mgrkg 9 2.840 "0.667 5.644 "0.367 n.a. n.a. n.a.
Ž . Ž .Hg 2=1.7 mgrkg 6 3.440 "2.141 6.100 "1.540 n.a. n.a. n.a.
Ž .Weight and length averages and standard deviations "SD of the fish used for each experimental group. N: number of animals analysed;
n.a.: product not assayed; B group: fish receiving zero injection of saline solution; C1 group: fish receiving one injection of saline solution;
C2 group: fish receiving two injections of saline solutions.
2.6. Statistical analysis
Homogeneity of the experimental groups for
length and weight was tested with one way ANOVA.
The frequencies of micronuclei and other nuclear
Ž .lesions were expressed per 1000 cells ‰ . The
nuclear abnormalities other than micronuclei were
considered together for statistical analysis in absence
of evidences or suggestions of differences in their
origin. Micronuclei, the standard indicators of geno-
toxicity, were always considered separately from the
other nuclear abnormalities.
Ž .Kruskal–Wallis K–W tests were used to com-
pare nuclear abnormalities frequencies between con-
trol and treatment groups. Mann–Whitney tests were
employed a posteriori. We used the SPSS 8.0 pro-
Ž .gram SPSS for PC computers.
3. Results
Table 1 presents averages and standard deviations
of lengths and weights for the groups of fish used in
this work. For each species, significant differences
Table 2
Micronuclei and nuclear abnormalities in minnow renal erythrocytes treated with different compunds
Micronuclei Blebbed Lobed Notched Other abnormalities
Ž . Ž . Ž . Ž . Ž .Control 0.625 "0.517 0.5 "0.755 0.875 "0.991 0.25 "0.462 1.625 "0.916
) ) )Ž . Ž . Ž . Ž . Ž .Colchicine 2.076 "1.441 1.692 "1.250 1.769 "1.423 0.538 "0.660 4 "2.380
) )Ž . Ž . Ž . Ž . Ž .Mitomycin C 2 "1.460 1.25 "1.183 1.5 "1.751 0.812 "0.834 3.562 "2.279
) ) )Ž . Ž . Ž . Ž . Ž .Cyclophosphamide 0.4 "0.516 n.s. 2.1 "0.994 3.5 "3.171 1 "1.154 6.6 "4.550
Ž . Ž .Averages and standard deviations "SD of micronuclei and nuclear abnormalities in minnows injected with saline solution control ,
colchicine, cyclophosphamide, and mitomycin C. Nuclear abnormalities: the sum of blebbed, lobed, and notched nuclei.
Ž .n.s. not significant .
)P-0.05.
))P-0.01.
)))P-0.001.
Table 3
Effect of administration protocol in micronuclei and other nuclear lesions in minnow and mollie
Pho. phoxinus Poe. latipinna
Micronuclei Other abnormalities Micronuclei Other abnormalities
Ž . Ž . Ž . Ž .Group B 0.3 "0.679 4 "3 0 "0 7.75 "5.063
Ž . Ž . Ž . Ž .Group C1 0.5 "0.547 n.s. 4.666 "2.581 n.s. 0.375 "0.517 n.s 3.875 "3.044 n.s.
Ž . Ž . Ž . Ž .Group C2 0.666 "1.032 n.s. 4.166 "2.786 n.s. 0 "0 n.s. 5.125 2.232 n.s.
Ž . Ž . Ž . Ž .Averages and standard deviations "SD in minnows and mollies receiving zero B , one C1 , or two C2 saline solutions. Statistical
Ž . Ž .significance of differences with respect to the control group B . n.s. not significant .
were not found among the experimentalgroups using
Ža single-factor ANOVA F s1.537; F s1.598W L
for minnow and F s1.842; F s1.884 in the caseW L
of mollie, W and L being weight and length, respec-
.tively .
3.1. Induction of nuclear abnormalities by genotoxic
compounds
Mortality associated with the experimental proce-
Ždure was similar for the three compounds around
.25% . The results of the experiment with different
genotoxic compounds are presented in Table 2. There
were evident differences between groups for the
frequencies of both micronuclei and other nuclear
Ž .abnormalities. K–W tests were significant P-0.01
for both micronuclei and other abnormalities. Fish
injected with colchicine showed a significant in-
Ž .crease of micronuclei frequency P-0.05 and also
Ž .of the other nuclear abnormalities P-0.01 . Simi-
lar results were obtained with mitomycin C: the
treated group showed a significant increase for the
frequencies of micronuclei as well as for the fre-
quencies of the rest of alterations with respect to
Ž .control fish P-0.05 in both cases . In contrast, the
group treated with cyclophosphamide did not show
significant differences in micronuclei frequencies
with respect to the control group, although differ-
ences between frequencies of the treated and control
groups for the other nuclear lesions were highly
Ž .significant P-0.001 .
3.2. Effect of administration protocol
In Table 3, we present averages and standard
deviations of micronuclei and other nuclear abnor-
malities obtained in the experiment aimed to investi-
gate the possible induction of micronuclei by the
Ž .administration protocol intraperitoneal injection
used in this work. For the minnow, there were no
Table 4
Frequency of micronuclei and nuclear abnormalities in fish treated with different doses of cadmium
Cadmium dose Pho. phoxinus Poe. latipinna
Micronuclei Other abnormalities Micronuclei Other abnormalities
Ž . Ž . Ž . Ž .Control 0.583 "0.792 4.5 "2.658 0.187 "0.403 4.5 "2.658
Ž . Ž . Ž . Ž .0.17 mgrkg 0.333 "0.516 n.s. 5.142 "4.049 n.s. 0 "0 n.s. 6.833 "4.167 n.s.
Ž . Ž . Ž . Ž .1.7 mgrkg 1.625 "1.060 n.s. 6.875 "3.482 n.s. 0.375 "0.744 n.s. 17.875 "15.724 n.s.
Ž . Ž .3.4 mgrkg 0.875 "0.991 n.s. 6.5 "5.398 n.s. n.a. n.a.
) )Ž . Ž . Ž . Ž .2=1.7 mgrkg 1.333 "0.816 n.s. 12.333 "6.623 0.166 "0.408 n.s. 10.833 "8.207
Ž . Ž .Averages and standard deviations "SD in minnows and mollies injected once and twice 2=1.7 mgrkg with cadmium chloride
solutions. n.a.: not assayed.
Ž .n.s. not significant .
)P-0.05.
Table 5
Frequency of micronuclei and nuclear abnormalities in fish treated with different doses of mercury
Mercury dose Pho. phoxinus Poe. latipinna
Micronuclei Other abnormalities Micronuclei Other abnormalities
Ž . Ž . Ž . Ž .Control 0.583 "0.792 4.5 "2.658 0.187 "0.403 4.5 "2.658
) )Ž . Ž . Ž . Ž .0.17 mgrkg 1 "0.894 n.s. 5.3 "3.444 n.s. 1.333 "1.032 13.5 "10.968
) )Ž . Ž . Ž . Ž .1.7 mgrkg 0.636 "0.809 n.s. 6 "4.560 n.s. 0.666 "0.516 n.s. 9.666 "2.338
Ž . Ž .3.4 mgrkg 1.777 "1.301 n.s. 4.444 " 2.603 n.s. n.a. n.a.
Ž . Ž .2=1.7 mgrkg 1.5 "1.870 n.s. 3.5 "2.167 n.s. n.a. n.a.
Ž . Ž .Averages and standard deviations "SD in minnows and mollies injected once and twice 2=1.7 mgrkg with mercury nitrate solutions.
n.a.: not assayed. n.s.: not significant.
)P-0.05.
))P-0.01.
significant differences between controls and fish re-
ceiving one or two saline injections for micronuclei
Žfrequencies or other nuclear abnormalities K–W
tests: x 2 s1.023 and x 2 s0.257, respectively, both
.n.s. . For the mollie, significant differences were
Ž .detected between groups P-0.05 for micronuclei
frequencies. Nevertheless, pairwise tests a posteriori
have not detected significant differences.
3.3. Induction of micronuclei by heaÕy metals
Mortality was variable depending on treatment
and species, ranging from 5% to 50%. In general
Table 6
Summary of results obtained in both species
Pho. phoxinus Poe. latipinna
Micronuclei Other abnormalities Micronuclei Other abnormalities
Induction of nuclear abnormalities by genotoxic compounds
) ) )Ž .Colchicine 10 mgrkg n.a. n.a.
) )Ž .Mitomycin C 20 mgrkg n.a. n.a.
)) )Ž .Cyclophosphamide 40 mgrkg n.s. n.a. n.a.
Effect of administration protocol
C1 Group n.s. n.s. n.s. n.s.
C2 Group n.s. n.s. n.s. n.s.
Induction of micronuclei by heaÕy metals
Cd 0.17 mgrkg n.s. n.s. n.s. n.s.
Cd 1.7 mgrkg n.s. n.s. n.s. n.s.
Cd 3.4 mgrkg n.s. n.s. n.s. n.s.
) )Cd 2=1.7 mgrkg n.s. n.s.
) )Hg 0.17 mgrkg n.s. n.s.
))Hg 1.7 mgrkg n.s. n.s. n.s.
Hg 3.4 mgrkg n.s. n.s. n.a. n.a.
Hg 2=1.7 mgrkg n.s. n.s. n.a. n.a.
Induction of micronuclei and other nuclear abnormalities in both species. Statistical significance of differences with respect to the control of
each experiment. C1 group: fish receiving one injection of saline solution; C2 group: fish receiving two injections of saline solutions.
Ž . Ž .n.s. not significant , n.a. product not assayed .
)P-0.05.
))P-0.01.
)))P-0.001.
terms, mortality was higher for the mollie. Table 4
presents the frequencies of micronuclei and other
nuclear lesions induced in minnow and mollie by
intraperitoneal injection of cadmium chloride. Signif-
icant differences between controls and fish treated
with a single injection of cadmium chloride at the
doses used in this work were not detected for any
species. Only two injections of cadmium chloride at
a final concentration of 1.7 mg of cadmium per
kilogram of fresh weight induced a significant in-
Ž .crease of the nuclear abnormalities P-0.05 ; but
the apparent increase of micronuclei was not signifi-
cant.
With respect to the mercury, different results were
Ž .obtained for each species Table 5 . In the case of
minnow, no significant induction of micronuclei was
detected at any dose. For mollie, the two doses
assayed induced a significant increase of the fre-
Žquencies of nuclear abnormalities P-0.05 and P
.-0.01 . Nevertheless, only the lowest dose assayed
Ž .0.17 mgrkg induced the formation of micronuclei
Ž .P-0.05 .
The results obtained for all the endpoints are
summarised in Table 6.
4. Discussion
The presence of variations of the nuclear mor-
phology in fish erythrocytes has been previously
w xdescribed in other works 11,25–27 . They have been
interpreted as nuclear lesions analogous to micronu-
clei. Our results show that these abnormalities are
induced by well-known genotoxic compounds
w x24,28 , even if micronuclei are not induced. The
results obtained in this work with colchicine and
mitomycin C show that these two products induce
the formation of micronuclei as well as the other
nuclear abnormalities observed, whereas cyclophos-
phamide induced the formation of other nuclear ab-
normalities but not micronuclei. The genotoxicity of
the products used in this experiment has been previ-
ously demonstrated, therefore they are usually used
as positive controls for genotoxicity experiments. In
other fish species, both mitomycin C and cyclophos-
phamide induce the formation of micronuclei in pe-
w xripheral erythrocytes 5,7,23,26,29 . Both products
have a strong clastogenic action previously described
w x w xin vitro 28,30 and in vivo 31,32 for other species.
Cyclophosphamide can be either metabolised or
w x‘‘auto’’-transformed into its active metabolite 33 .
Previous works have demonstrated that at least 6
exposure days or 3 exposure plus 3 recovery days
were necessary to induce a significant increase of
nuclear abnormalities frequencies with cyclophos-
w xphamide 34,35 . In the case of the minnow a 24 h
Ž .exposure by intraperitoneal injection induced the
formation of nuclear abnormalities, which agrees
w xwith previous investigations 7 . The minnow has
been shown to be a sensitive species to the genotoxic
effects of cyclophosphamide, even more sensitive
w xthan eel 35 .
w xColchicine is an aneuploidy agent 28 due to the
w xinhibition of microtubule assembly 36 . Both clasto-
genic and aneugenic actions induced the formation
of the nuclear lesions of fish erythrocytes described
in this work. Therefore, these lesionsmost probably
are a consequence of a genotoxic event and should
be taken into account when performing the micronu-
cleus test in fish species. However, further investiga-
tions are necessary in order to know the mechanism
of formation of nuclear abnormalities, as well as to
investigate their genotoxic origin.
To our knowledge, the influence of the adminis-
tration protocol in laboratory genotoxicity tests has
not been investigated for fish species. Our results
show that intraperitoneal injection does not affect
significantly neither the frequencies of micronuclei
nor the rest of nuclear abnormalities observed in
minnows and mollies. For other species, available
data indicate that hormonal and metabolic changes
w xare induced under stress conditions 15,16 , and this
might influence the response of treated animals to
genotoxic compounds. But for the two species stud-
ied here, administration protocols based on intraperi-
toneal injections could be considered appropriated
for laboratory genotoxicity tests, as pointed out by
w xAl-Sabti and Metcalfe 14 .
For both species, similar results have been ob-
tained by injection of several cadmium doses. Our
results showed that a repeated intraperitoneal injec-
tion of cadmium induced a significant increase of the
Ž .nuclear abnormalities but not micronuclei in both
minnow and mollie; therefore, cadmium seems to
have a genotoxic effect in these species, supporting
the results obtained for Tilapia, and brown trout
using also intraperitoneal injection of Cd compounds
w xas administration protocol 22,25 . Genotoxic effects
w xof cadmium have also been reported in plants 37
w xand mammals 38,39 , sensitivity to cadmium being
probably common characteristics of all type of or-
ganisms.
In the case of mercury, we obtained different
results for both species. Mercury seems to be geno-
toxic for mollie because it significantly induced mi-
cronuclei and other nuclear abnormalities at low
dose and other nuclear lesions at the highest dose
employed in this work. In contrast, increases of
micronuclei and other nuclear abnormalities were not
detected for minnows treated with different doses of
mercury nitrate. The interspecific difference found in
this work may be attributed to a higher sensitivity of
mollie to mercury. The lack of response of minnow
to the genotoxic effects of mercury nitrate has been
w xrecently reported by Sanchez-Galan et al. 25 . There
are contradictory reports concerning the genotoxic
potential of mercury compounds. For example, sheep
feeding on grounds highly polluted with mercury did
w xnot present any evidence of DNA damage 40 .
However, mercury compounds seem to be genotoxic
w xin vitro for other fish species 41 and also after
w xdirect and indirect exposure in vivo 25,42 for sev-
eral species. For other fish species such as killifish
Fundulus heterociclitus, micronuclei induction was
obtained by exposure to methylmercury derivatives,
more active in short-term than inorganic mercury
w xsalts 21 . In the present case, only the mollie seems
to be sensitive to mercury.
In conclusion, we found an induction of micronu-
clei as well as other nuclear abnormalities, which
were induced even if micronuclei were not; therefore
we suggest to include these anomalies in fish geno-
toxicity analyses based on micronuclei counts. From
our results, intraperitoneal injection can be an ade-
quate routine administration protocol in fish genotox-
icity tests based on micronuclei count. The mollie
Poe. latipinna is presented as a valid target species
to carry out genotoxicity tests of heavy metals. In
contrast, the European minnow seems to be less
sensitive for heavy metals, especially to mercury
salts. However, it was showed to be very sensitive
for detecting the genotoxic actions of several com-
pounds and should be taken into account for future
investigations in fish genotoxicity.
Acknowledgements
The authors are indebted to Dr. Jorge Izquierdo
and Dr. Jose Luis Martinez for helping in field
sampling; to Dr. Miguel A. Comendador for his
helpful comments to our work; to the Consejeria de
Medio Ambiente for allowing the necessary permis-
sions for sampling. This work has been supported by
the NATO Linkage Grant ENVIRLG 974614.
References
w x1 R.N. Hooftman, W.K. de Raat, Induction of nuclear anoma-
Ž .lies micronuclei in the peripheral blood erythrocytes of the
eastern mudminnow Umbra pygmea by ethyl methane-
Ž .sulphonate, Mutat. Res. 104 1982 147–152.
w x2 S.L. De Flora, F. Vigano, A. D’Agostini, M. Camoirano, C.
Bagnasco, F. Bennecelli, F. Melodia, A. Arillo, Multiple
genotoxicity biomarkers in fish exposed in situ to polluted
Ž .river water, Mutat. Res. 319 1993 167–177.
w x3 S. Minissi, E. Ciccotti, M. Rizzoni, Micronucleus test in
Ž .erythrocytes of Barbus plebejus Teleostei pisces from two
natural environments: a bioassay for the in situ detection of
Ž .mutagens in freshwater, Mutat. Res. 367 1996 245–251.
w x4 S. Sanchez-Galan, A.R. Linde, J.I. Izquierdo, E. Garcia-
Vazquez, Micronuclei and fluctuating asymmetry in brown
Ž .trout Salmo trutta : complementary methods to biomonitor
Ž .freshwater ecosystems, Mutat. Res. 412 1998 219–225.
w x5 I.B. Bahari, F.M. Noor, N.M. Daud, Micronucleated eythro-
cytes as an assay to assess actions by physical and chemical
genotoxic agents in Clarias gariepinus, Mutat. Res. 313
Ž .1994 1–5.
w x6 J.C. Nepomuceno, I. Ferrari, M.A. Spano, A.J. Centeno,
Detection of micronuclei in peripheral erythrocytes of Cypri-
nus carpio exposed to metallic mercury, Environ. Mol. Mu-
Ž .tagen. 30 1997 293–297.
w x7 M.A. Campana, A.M. Panzeri, V.J. Moreno, F.N. Dulout,
Genotoxic evaluation of the pyrethroid lambda-cyhalothrin
using the micronucleus test in erythrocytes of the fish
Ž .Cheirodon interruptus interruptus, Mutat. Res. 438 1999
155–161.
w x8 R.N. Hooftman, G.J. Vink, Cytogenetic effects on the eastern
mudminnow Umbra pigmea exposed to ethyl methanesul-
w xfonate, benzo a pyrene and river water, Ecotoxicol. Environ.
Ž .Saf. 5 1981 261–269.
w x9 K.R. Carrasco, K.L Tilbury, M.S. Myers, Assessment of the
piscine micronucleus test as an in situ biological indicator of
chemical contaminant effects, Can. J. Fish. Aquat. Sci. 47
Ž .1990 2123–2136.
w x10 M. Pacheco, M.A. Santos, Induction of EROD and erythro-
cytic nuclear abnormalities by cyclophosphamide and PAHs
Ž .in Anguilla anguilla L., Ecotoxicol. Environ. Saf. 40 1997
71–76.
w x11 C.D. Metcalfe, Induction of micronuclei and nuclear abnor-
Ž .malities in the erythrocytes of mudminnow Umbra limi and
Ž .brown bullheads Ictalurus nebolusus , Bull. Environ. Con-
Ž .tam. Toxicol. 40 1988 489–495.
w x12 J.A. Heddle, M.C. Cimino, M. Hayashi, F. Romagna, M.D.
Shelby, J.D. Tucker, P. Vanprays, J.T. MacGregor, Micronu-
clei as an index of cytogenetic damage: past, present and
Ž .future, Environ. Mol. Mutagen. 18 1991 277–291.
w x13 A.M. Pendas, P. Moran, E. Garcia-Vazquez, Improvements
to Atlantic salmon anterior kidney metaphase yields follow-
Ž .ing phytohemaglutinin injection, J. Fish Biol. 42 1993
801–802.
w x14 K. Al-Sabti, C.D. Metcalfe, Fish micronuclei for assessing
Ž .genotoxicity in water, Mutat. Res. 343 1995 121–135.
w x15 A. Hontella, P. Dumont, D. Duclos, R. Fortin, Endocrine and
metabolic dysfunction in yellow perch, Perca flaÕescens,
exposed to organic contaminants and heavy metals in the St.
Ž .Lawrence River, Environ. Toxicol. Chem. 14 1995 725–
731.
w x16 R.M. Harrell, Stress mitigation by use of salt and anaesthetic
for wild striped bass captured for brood stock, Prog. Fish-Cult.
Ž .54 1992 228–233.
w x17 F.I. Ingel, N.M. Gevorkian, N.A. Iliushina, B.I. Leitina, L.M.
Prikhozhan, E.L. Pereverzeba, I.A. Revazoba, Prolonged
psychoemotional stress as an inducer of mutations in mam-
mals and as a modifier of mutagenesis, Bull. Exp. Biol. Med.
Ž .116 1993 307–309.
w x18 E. Tarifeno-Silva, L. Kawasaki, D.P. Yn, M.S. Gordon, D.J.
Chapman, Aquacultural approaches to recycling dissolved
nutrients in secondarily treated domesticwaste waters: up-
take of dissolved heavy metals by artificial food chains,
Ž .Water Res. 16 1982 59–65.
w x19 T.G. Rossman, Metal mutagenesis, in: R.A. Goyer, G.C.
Ž .Chewrian Eds. , Toxicology of Metals, Springer-Verlag,
New York, 1995, pp. 373–403.
w x20 J.S. Weis, J.S. Weis, Methylmercury teratogenesis in the
Ž .killifish, Fundulus heteroclitus, Teratology 16 1997 317–
326.
w x21 D.M. Perry, J.S. Weis, P. Weis, Cytogenetic effects of
methylmercury in embryos of the killifiss Fundulus hetero-
Ž .clitus, Arch. Environ. Contam. Toxicol. 17 1988 569–574.
w x22 G.K. Manna, A. Sadhukhan, Use of cells of gill and kidney
Ž .of Tilapia fish in micronucleus test MNT , Curr. Sci. 55
Ž .1986 498–501.
w x23 T. Ueda, M. Hayashi, Y. Ohtsuka, T. Nakamura, J.
Kobayashi, T. Sofuni, A preliminary study of the micronu-
cleus test by acridin orange fluorescent staining compared
with chromosomal aberration test using fish erythropoietic
Ž . Ž .and embryonic cells, Water Sci. Tech. 25 11 1992 235–
240.
w x24 J.W. Parton, M.L. Garriott, An evaluation of micronucleus
induction in bone marrow and in hepatocytes isolated from
collagenase perfused liver or from formalin-fixed liver using
4-week-old rats treated with known clastogens, Environ.
Ž .Mol. Mutagen. 29 1997 379–385.
w x25 S. Sanchez-Galan, A.R. Linde, E. Garcia-Vazquez, Brown
trout and European minnow as target species for genotoxicity
tests: differential sensitivity to heavy metals, Ecotoxicol.
Ž .Environ. Saf. 43 1999 301–304.
w x26 R.K. Das, N.K. Nanda, Induction or micronuclei in periph-
eral erythrocyte of fish Heteropneustes fossilis by mitomycin
Ž .C and paper mill effluent, Mutat. Res. 175 1986 67–71.
w x27 J.E. Hose, J.N. Cross, S.G. Smith, D. Diehl, Elevated circu-
lating erythrocyte micronuclei in fishes from contaminated
Ž .sites of Southern California, Mar. Environ. Res. 22 1987
167–176.
w x28 N.L. Rudd, S.E. Williams, M. Evans, U.G.C. Hennig, D.I.
Hoar, Kinetochore analysis of micronuclei allows insights
into the actions of colcemid and mitomycin C, Mutat. Res.
Ž .261 1991 57–68.
w x29 J.C.M. Van der Hoeven, I.M. Bruggeman, G.M. Alink, J.H.
Koeman, The killfish Notobranchus rachowi a new animal
Ž .in genetic toxicology, Mutat. Res. 97 1982 35–42.
w x30 M. Schuler, D.S. Rupa, D.A. Eastmond, A critical evaluation
of centromeric labeling to distinguish micronuclei induced by
chromosomal loss and breakage in vitro, Mutat. Res. 392
Ž .1997 81–95.
w x31 B.M. Miller, I.D. Adler, Application of antikinetochore anti-
Ž .body staining CREST staining to micronuclei in erythro-
Ž .cytes induced in vivo, Mutagenesis 4–5 1990 411–415.
w x32 K.A. Criswell, G. Krishna, D. Zielinski, G.A. Urda, J.C.
Theiss, P. Juneau, M.R. Bleavins, Use of acridine orange in
flow cytometric: assessment of micronuclei induction, Mutat.
Ž .Res. 414 1998 63–75.
w x33 M.A. Santos, M.G. Pacheco, The mutagenicity of cyclophos-
phamide, pulp mill effluents and its sediment on the Eel
Ž .Anguilla anguilla L., Sci. Total Environ. 171 1995 127–130.
w x34 M. Pacheco, M.A. Santos, Induction of micronuclei and
nuclear abnormalities in the erythrocytes of Anguilla an-
guilla L. exposed either to cyclophosphamide or to bleached
Ž .pulp mill effluent, Fresenius Environ. Bull. 5 1996 746–751.
w x35 M. Pacheco, M.A. Santos, Induction of EROD activity and
genotoxic effects by polycyclic aromatic hydrocarbons and
Ž .resin acids on juvenile eel Anguilla anguilla L. , Ecotoxi-
Ž .col. Environ. Saf. 38 1997 252–259.
w x36 M. Wallin, B. Hartley-Asp, Effects of potencial aneuplody
inducing agents on microtubule assembly in vitro, Mutat.
Ž .Res. 287 1993 17–22.
w x37 H. Steinkellner, K. Mun-Sik, C. Helma, S. Ecker, T.-H. Ma,
O. Horak, M. Kundi, S. Knasmuller, Genotoxic effects of¨
heavy metals: comparative investigation with plants bioas-
Ž .says, Environ. Mol. Mutagen. 31 1998 183–191.
w x38 G.C. Jagetia, S.K. Adiga, Cadmium chloride induces dose-
dependent increases in the frequency of micronuclei in mouse
Ž .bone marrow, Mutat. Res. 306 1994 85–90.
w x39 A. Marrazzini, C. Betti, F. Bernacchi, I. Barrai, R. Barale,
Micronucleus test and metaphase analyses in mice exposed
to known and suspected spindle poisons, Mutagenesis 6–9
Ž .1994 505–515.
w x40 T. Gebel, S. Kevekordes, J. Schaefer, H. von Platen, H.
Dunkelberg, Assessment of a possible genotoxic environmen-
tal risk in sheep bred on grounds with strongly elevated
contents of mercury, arsenic and antimony, Mutat. Res. 368
Ž .1996 267–274.
w x41 K. Al-Sabti, Micronuclei induced by selenium, mercury,
methylmercury and their mixtures in binucleated blocked fish
Ž .erythrocyte cells, Mutat. Res. 320 1994 157–163.
w x42 A.P. Krishnaja, M.S. Rege, Induction of chromosomal aber-
rations in fish Boleophtalmus dussumieri after exposure in
vivo to mitomycin C and heavy metals mercury, selenium
Ž .and chromium, Mutat. Res. 102 1982 71–82.