Mycopathologia 158: 233–238, 2004.

233

2004 Kluwer Academic Publishers. Printed in the Netherlands.

Structural and ultrastructural alterations in BALB/c mice: Effects

of Penicillium citrinum metabolites

M.C. Lurá1, M. Fuentes2, M. Cabagna2, A.M. González1, A. Nepote1, M.C. Giugni2,

M. Rico1 & M.G. Latorre1
1
Cátedra de Microbiologı´a General, Facultad de Bioquı´mica y Ciencias Biológicas; 2Cátedra de Morfologı´a
Normal, Facultad de Bioquı´mica y Ciencias Biológicas, Universidad Nacional del Litoral, Argentina

Received 2 December 2002; accepted in revised form 12 April 2004

Abstract
The aims of this work were to determine the effect of feeding BALB/c mice a diet containing culture
materials of a citrinin producing strain of Penicillium citrinum (Thom). Changes in hematological
parameters, serum chemistry and histological changes in liver, kidney and heart were determined. After
60 days, control treated (CT) mice appeared normal in all respects, whereas, the mice fed the feeds sup-
plemented with Penicillium (CMT) showed decreased weight gain, lower hematocrits, increased serum
alanine aminotransferase (ALT) and clear signs of renal and hepatotoxicity based on histological changes.
Changes observed in the liver of CMT mice included portal and lobular infiltration of polymorphonuclear
cells, with concomitant hepatocellular necrosis, hepatic steatosis, prominent Kupffer’s cells, hemosiderin
granules in the cytoplasm of periportal hepatocytes and other lipid inclusions in the surrounding mito-
chondria were also observed. Our findings suggest that in vivo, P. citrinum Thom metabolites, which
contain citrinin, could cause illnesses such as toxic hepatitis or intravascular hemolysis.

Key words: Citrinin, Penicillium citrinum metabolites, ultrastructural alterations

Introduction Penicillium and Fusarium species have been fre-

Fungal contamination of food and feed constitute
important problems, not only to human and animal
health but also can have serious economic impacts.
Fungi can cause crop damage and spoilage, with
important losses in agricultural and animal pro-
duction. Fungal contamination can also reduce the
nutritional value of foods and feeds and is respon-
sible for important mycotoxicoses [1]. Mycotoxins
are secondary products of fungal metabolism pro-
duced by a diverse group of fungi. This wide variety
of chemicals is capable of producing a wide range of
undesirable effects, such as damage to the central
nervous system, the liver, the kidney, etc. An overt
fungal infestation is not essential for the production
of a mycotoxicosis [1, 2].
As in other geographical areas of Argentina, in
Santa Fe and its surrounding area, Aspergillus,
quently identified as food contamination agents.
One of the most frequent species is P. citrinum [3].
Citrinin is one of the secondary metabolites of this
fungus [1, 4, 5]. The most important toxicological
effect caused by citrinin is its ability to disrupt renal
function; however, many nonrenal effects have been
detected, as well such as immunosuppression, tera-
togenicity and mutagenicity [1, 2, 6–9]. We previ-
ously demonstrated that this toxin caused hemolysis
of human erythrocytes [10] and that mice fed with
either commercial citrinin or P. citrinum Thom
metabolites showed structural modifications in
their livers after 30 days exposure [11].
The aim of this study was to determine the
effect of feeding BALB/c mice a diet contain-
ing culture materials of a citrinin producing
strain of Penicillium citrinum (Thom). Changes in
hematological parameters, serum chemistry and
234
histological changes in liver, kidney and heart were
determined.
Materials and methods
Fungal isolate and citrinin detection in P. citrinum
Thom cultures
A cheese isolate of P. citrinum Thom, known to be
a citrinin producer (kept at Universidad Nacional
del Litoral, Argentina) [11] was used. The identi-
fication of the strain to the species level was carried
out according to classical taxonomy, based on
their micro and macro- morphological and physi-
ological characteristics in culture [5].
Commercial feed for mice (Alimento Balance-
ado Cooperación, San Nicolás, Buenos Aires,
Argentina) was used. Its formula was: 23% (min-
imum) crude protein, 6% crude fiber, 10% total
minerals (calcium, phosporous, sodium, iron,
copper, cobalt, selenium, iodine, manganese, zinc,
magnesium, vitamins: A, B1, B2, B6, B12, D3, E, K,
C), nicotinic acid, pantothenic calcium, folic acid
and 12% moisture.
Commercial feed previously kept at approxi-
mately 30% moisture and sterilized, in Erlenmeyer
flasks cotton stoppened, was inoculated with an
aqueous suspension of the fungal conidia (5 ml;
approximately 30 · 108 UFC/ml). After 14 days
incubation at 30 C, the cultures were dried at
50 C until reaching a constant weight and were
stored at )20 C until use [11]. Nitrogen, fat,
moisture, carbohydrate and ash contents of this
diet were determined by AOAC methods [12].
To confirm citrinin and other P. citrinum
metabolites in the culture material, dried culture
material was extracted with chloroform [11] and
analyzed by thin-layer chromatography (TLC) on
Merk Kieselgel 60 by means of Cole and Cox’s [4]
and Hald and Krogh’s method [13], respectively.
Spots were also observed by UV light and those of
interest were sprayed with a commercial boron
triflouride (Voltaix, Inc.) solution (10–15% in
methanol). Blue-green fluorescence under 365 nm
light confirmed the presence of citrinin [13]. Rf
values and semiquantitative analysis of citrinin
were performed in comparison to citrinin results
with an authentic commercial toxin standard
(SIGMA, lot 76H404) (0.4% in acetone:water
solution, by 80:20 vol.) [11].
Two other spots were observed by TLC; UV
spectra of samples were taken with a Metrolab
1700 spectrophotometer in ethanol solution.
Spectra were calibrated with an holmium oxide
standard.
Toxicological assay
Two different groups of BALB/c mice (males and
females), were studied, as described in preliminary
trials [11]. Briefly, twenty four mice (>3 weeks old)
were individually housed, in polypropylene cages
containing bedding made of shredded aspen wood
shavings, in rooms kept under controlled envi-
ronmental conditions (temperature, 23 C; relative
humidity 50% and 12 h light/dark cycle). Mean
weight of different groups was similar (P>0.05).
One of the groups, CMT (Culture material-
treated, n ¼ 12), were given the molded commer-
cial feed mixed with non-contaminated feed
(50:50). The final citrinin dose was approximately
0.065 mg citrinin/100 g food. The remaining
group, CT (control-treated, n ¼ 12), were given
non-contaminated feed.
Feed and water were supplied ad libitum. The
eating average was approximately 6 g/mouse. The
color of urine and feces was monitored daily and
both, weight and hematocrit values were checked
once a week. Mice were bled by periorbital punc-
ture.
At the beginning of the assay and after 60 days,
blood was collected to determine hematological
parameters and biochemical determination of
hepatic functions: aspartate aminotransferase
(AST)1, of alanine aminotransferase (ALT)1 alka-
line phosphatase (ALP)2 and total bilirubin3 and
renal of functions: serum creatinine4 and urea.5 All
determinations were carried out with Merck
Diagnostica reagents (Merck Diagnostica,
Darmstadt, Germany) in a Merck Vitalab Selectra.
Urine was examined for hemosiderin [14].
At the end of the assay (day 60), the animals
were sacrificed by anesthetic overdose, for both
macro and microscopic studies. Livers, kidneys
and hearts were removed and the organ weight/
animal weight rate was calculated. All the organs
were processed as described in Lurá et al. [11]. At
least four sections of each biopsy were used to
evaluate the severity of toxic damage. The severity
of the hepatic lesions was classified according to
the following criteria:
 235

– none: intact liver, with normal cells;

– minimal: necrosis of centrilobular cells;
– moderate: necrotic hepatocytes moving from
portal tract to the centrilobular region;
– severe: complete necrosis of hepatic parenchyma.
For ultrastructural studies, the livers of six
animals in each group, randomly selected, were cut
onto pieces of 1 mm3 each, fixed with glutaralde-
hyde and re-fixed with osmium tetroxide. Once
zones of interest were chosen, they were cut to Figure 1. Comparison of weights (mean value) among treated
(culture material-treated group, CMT) and non treated (con-
80 nm thickness and examined by transmission
trol-treated group, CT) animals during 60 days.
electron microscopy [14].
Ethical rules for the handling of experimental
animals were respected, as indicated in the ÔGuide Neither choluria nor acholia was detected in
for the Care and Use of laboratory animalsÕ either the CT or CMT mice. With the exception of
(NRC, 1996). significantly reduced hematocrits (Figure 2) and
Statistical analysis of the data was carried out significantly elevated serum ALT (Figure 3), in the
using Independent-samples T-test. A P value of CMT mice, all serum biochemical and hemato-
£0.05 was considered significant in all analyses logical parameters measured were normal in both
[15]. SPSS 10.0 for Windows Software (Statistical groups under study (data not shown). Hemosid-
Package Social Science Inc., Chicago, Illinois erin was detected in urine of some CMT mice and
60606) was used. hemorrhagic zones were noted in livers of most of
the CMT mice (n ¼ 10).
Light microscopic examination revealed corti-
Results
cal and medullar congestion of the kidney and
hemosiderin granules were detected in the proxi-
No significant differences among the content of the
mal tubules. No changes were noted in the heart.
main nutrients (nitrogen, fat, moisture, carbohy-
The incidence and relative severity of the patho-
drates and ashes) from both experimental diets
logic changes in the liver is quantified in Figure 4.
were observed.
In the CMT group, all animals exhibited some
UV data of the spot previously identified as
evidence of pathologic changes ranging from
citrinin (TLC assay) were identical to that of a
minimal (n ¼ 1), to moderate (n ¼ 8), to severe
standard (SIGMA, lote 76H404). Data of the
(n ¼ 1). None of the CT animals exhibited any
other spots were different: one of them had Rf 0.80
evidence of liver injury. There was a sharp
(toluene–ethylacetate–90%formic acid (6:3:1;
demarcation of viable and necrotic hepatocytes in
TEF)), and the other Rf 0.69 (TEF). Identity of
the transition from portal tract to the centrilobular
these two spots was not determined.
region. Portal and lobular infiltrate of polymor-
None of the mice in the CT group showed
phonuclear cells, with concomitant hepatocellular
neither macroscopic anomalies nor structural
alterations in any of the microscopically studied
organs (liver, kidney and heart).
Weight gain was significantly reduced in the
CMT group compared to the CT group (mean
weight increase for CMT and CT, respectively,
was 0.84 and 15.5 g; P<0.003) (Figure 1). Simi-
larly, liver weight was significantly reduced in the
CMT group compared to the CT group (mean
weight for CMT and CT, respectively, was 1.77
and 2.30 g; P<0.001). The organ weight/animal Figure 2. Comparison of hematocrits (mean value) among
weight in kidney and heart were not significantly treated (culture material-treated group, CMT) and non treated
different (P>0.05). (control-treated group, CT) animals during 60 days.
236
Figure 3. Comparison of ALT (mean value) among treated
(culture material-treated group, CMT) and non treated (con-
trol-treated group, CT) animals at the beginning of the assay
and after 60 days.
necrosis was observed in CMT group. The com-
bination of hypoperfusion and retrograde con-
gestion acted synergistically to generate
centrilobular hemorrhagic necrosis. We also ob-
served hepatic steatosis, sinusoidal lining cells with
prominent Kupffer’s cells, and iron became evi-
dent as an accumulation of hemosiderin granules
in the cytoplasm of periportal hepatocytes.
Based on transmission electron microscopy,
there was an increased number of swollen and
round mitochondria with denser mitochondrial
matrix in CMT mice compared to CT group. In
the cytoplasm, of the periportal hepatocytes, many
granules and other lipidic inclusions in the sur-
rounding area of the mitochondria were also ob-
served. Control group (CT) showed no
ultrastructural alterations in their livers. Typical
features are illustrated in Figures 5 and 6.
Figure 4. Incidence of the severity of hepatic damage, among
treated (culture material-treated group, CMT) and non treated
(control-treated group, CT) animals during 60 days.
Discussion
Citrinin has been obtained from cultures of several
species of Penicillium. These moulds have been
shown to infest various types of grains, feed and
food, and this toxin has been found in samples of
wheat, rice, barley, rye, oats, etc [2].
Hald [8] and Schlosberg and Bellaiche [16] de-
scribed a delay in the growth and anorexia symp-
toms in animals fed with citrinin. Based on our
experience, body weight loss in mice treated with
the P. citrinum Thom metabolites was similar. Our
earlier studies [17], revealed similar data for
chickens fed with citrinin for 30 days.
The most important toxicological effect caused
by citrinin is its ability to disrupt renal function.
Despite the paucity of information, researchers
agree that the effect of this toxin lies in the prox-
imal tubule and not in the distal tubule or
glomeruli [2, 8]. We noticed vascular changes both
in the glomeruli and renal medulla, although al-
tered parenchyma cells were not observed.
P. citrinum metabolites have not been reported to
produce any effects in the renal medulla.
Some morphologic alterations in the organs
studied of the CMT group were detected. Kidney
and liver congestion might be explained because of
their vulnerability to a wide variety of toxic injury.
An incipient left-sided cardiac failure and a
chronic passive congestion respectively, might
cause slight heart and kidney enlargement [18].
Lurá et al. [11] demonstrated that Mus musculus
fed with commercial citrinin exhibited the same
structural alterations as a different group which
received commercial feed with the addition of the
citrinin-producing mould.
ALTs values in the CMT group were consider-
able higher than the CT group values. This was
likely due to altered hepatocyte permeability. In
cardiac insufficiency, toxic hepatitis and hepatic
steatosis, ALT is increased. In hepatic steatosis
there is a less frequent increase of AST and the other
parameters (bilirubin and ALP) are normal [18].
Our studies detected: small (microvesicular) li-
pid droplets which accumulated in hepatocytes,
subcapsular hydropic degeneration, single or
scattered foci of cells which underwent swelling
(ballooning), a more frequent necrosis in the cen-
trilobular regions of the lobule and prominent
Kupffer’s cells. Alcoholic hepatitis typically leads
to a similar pathological finding [18].
 237

Figure 5. Transmission electron microscopy shows no ultrastructural alterations of the hepatocytes of mice fed with non-contaminated
feed (control-treated group, CT). (a) 7000· (arrows show normal nucleus and two normal mitochondria), (b) 30,000·(arrow shows
normal mitochondria).

Thung and Gerber [19] suggested that a few
vacuoles and small amounts of stainable iron are
common in normal human hepatocytes. Although
hemosiderin is abundant in the cytoplasm of he-
patocytes during the first week of life, there is a
subsequently gradual disappearance and should be
absent. The study found hemosiderin granules in
the livers and proximal tubules of the CMT
animals. These findings could not be explained due
to excessive iron in the diet (hemosiderin was ab-
sent in the CT mice group); hence they are thought
to be caused by an intravascular hemolysis. There
are a few references about hemolysis caused by
citrinin in living animals [20, 21]; however,
hematological changes (decrease in hematocrit
value) confirm that this is to be suspected. On the
other hand, early studies demonstrated that
P. citrinum Thom, a citrinin producer, caused
hemolysis of human erythrocytes [10]. Crawford
[18] described chronic intoxications with hemo-
siderosis without icterus, which could lead to the
absence of choluria.
Iron is usually deposited in periportal hepato-
cytes and being a direct hepatotoxin, inflammation
is typically absent [19].
Chagas et al. [22], studied morphological
alterations in an established cell line of baby
hamster kidney cells. The results showed that the
cells, originally elongated and flattened, became
swollen and round. Electron microscopical exam-
ination demonstrated that citrinin incubated with
those cultured cells, promoted dramatic alterations
of normal mitochondria, leading to mitochondria
swelling and cell death.

Figure 6. Transmission electron microscopy shows an increased number of swollen and round mitochondria of the hepatocytes of the
mice which received commercial feed to which the citrinin-producing mould had been added, mixed with non-contaminated feed, fifty-
fifty (CMT). Note the granules and lipidic inclusions in the surrounding area of the mitochondria, (a) 7000· (arrows show, nucleus,
swollen and round mitochondria and granules), (b) 30,000·(arrows show lipidic inclusions (L), granules (G) and mitochondria (M)
exhibiting abnormal ultrastructural architecture.
238
Similarly, we found an increased number and
alterations of liver mitochondria. This could be
explained because for the cells to keep their nor-
mal function, they must increase the mitochondrial
number to obtain the necessary energy for their
vital processes.
Granules observed showed morphological
similarities to hemosiderin granules detected by
microscopy. Lipid inclusions in the cytoplasm
confirmed hepatic steatosis also detected by
microscopy.
No differences were observed among the main
nutrients. Our findings lead us to speculate that in
vivo, the diet containing P. citrinum Thom
metabolites, which includes citrinin, could cause
illnesses such as a toxic hepatitis or intravascular
hemolysis.
Acknowledgment
This work was partly supported by grants from the
Universidad Nacional del Litoral, Argentina,
CAID’ 2000.
Notes
1
Kinetic and colorimetric determination, based on the refer-
ence method of the IFCC (IFCC: International Federation of
Clinical Chemistry).
2
Optimised kinetic method (DGKC and SSCC) (DGKC:
Deutsche Vereinte Gesellschaft für Klinische Chemie e.V.;
SSCC: Swiss Society of Clinical Chemistry).
3
Colorimetric method.
4
Kinetic and colorimetric Jaffé method.
5
Urease/GLDH (glutamate dehydrogenase) system applied to
a Somogyl deproteinized supernate.
References
1. Council for Agricultural Science and Technology, Myco-
toxins. Economic and Health Risks. Task Force Report,
No 116, Ames IA, Iowa, 1989.
2. Berndt W. Effects of selected fungal toxins on renal
function. In: Hooks JB, Goldstein RS, eds. Toxicology of
the Kidney. New York: Raven Press, 1993: 459–475.
3. Bressler G, Brizzio SB, Vaamonde G. Mycotoxin-produc-
ing potential of fungi isolated from amaranth seeds in
Argentina. Int J Food Microbiol 1995; 25: 101–108.
4. Cole R, Cox R. Handbook of Toxic Fungal Metabolites.
New York: Academic Press, 1981.
5. Pitt Jl, Hodking AD. Fungi and Food Spoilage. Maryland:
Aspen Publishers, 1999.
6. Leifer Z, Kada T, Mandel M, Zeiger E, Stattford R,
Rosenkranz ES. An evaluation of test using DNA repair-
deficient bacteria for predicting genotoxicity and carcino-
genicity. Mut Res/DNA Repair 1981; 87: 211–297.
7. Arai M, Hibino T. Tumorgenicity of citrinin in male F344
rats. Cancer Lett 1983; 51: 281–287.
8. Hald B. Porcine Nephropathy in Europe.: IARC Scientific
Publications 1991; 115: 49–56.
9. Pestka JJ, Bondy GS. Alteration of inmune function
following dietary mycotoxin exposure. Can J Physiol
Pharmacol 1990; 68: 1009–1016.
10. González AM, Lurá MC, Latorre MG, Rico M, Carrera
E, Lound F. Estudio de la actividad hemolitica in vitro de
extractos de hongos productores de citrinina. Rev Fac
Bioquim Cs Biológ (Univ Nac Litoral, Argentina) 1998; 2:
127–130.
11. Lurá MC, Fuentes MB, Cabagna M, González AM,
Nepote AF, Giugni MC, Rico M, Latorre MG. Actividad
de metabolitos de Penicillium citrinum sobre ratones Mus
musculus. Rev Iberoamer Micol 2001; 18: 177–180.
12. Cunniff P. Official Methods for Analysis of Association of
Official Agricultural Chemists. Maryland, USA: AOAC
International, 1999.
13. Hald B, Krogh P. Analysis and chemical confirmation of
citrinin in barley. J AOAC 1973; 56: 1440–1443.
14. Garcia del Moral R. Laboratorio de anatomı́a patológica.
Madrid: Editorial Interameramericana Mc Graw Hill,
1993.
15. Walpole R, Myers R. Probabilidad y Estadı́stica. México:
Editorial Mc Graw-Hill, 1994.
16. Schlosberg A, Bellaiche M. Lead toxicosis in a captive
bottlenose dolphin consequent ingestion of air gun pellets.
J Wildlife Dis 1997; 33: 135–139.
17. Lurá MC, Bası́lico JC. Capacidad toxicogénica de hongos
contaminantes de alimentos para aves. La Industria
Cárnica Latinoamer 1984; 53: 28–32.
18. Crawford JM. The liver and the biliary tract. In: Robbins
SL, Cotran RS, eds. Pathologic Basis of Disease. México:
Nueva Editorial Interamericana, 1994: 831–895.
19. Thung SN, Gerber MA. Liver. In: Sternberg SS, ed.
Histology for Pathologists. Philadelphia: Lippincott-Ra-
ven Publishers, 1997: 575–589.
20. Gupta M, Sasmal D, Bandyopadiyay S, Bagchi G,
Chatterjee T, Dey S. Hematological changes produced in
mice by Ochratoxin A and Citrinin. Toxicology 1983; 26:
55–62.
21. Griffiths IB, Done SH. Citrinin as a possible cause of the
pruritis, pyrexia, haemorragic syndrome in cattle. Vet Res
1991; 129: 113–117.
22. Chagas GM, Kluppel ML, Campello AP, Buchi DF,
Oliveira MB. Alterations induced by citrinin in cultured
kidney cells. Cell Structure and Funct 1994; 19: 103–
108.
Address for correspondence: Marı́a Cristina Lurá, Domingo
Silva 1980, (3000) Santa Fe, Argentina
Phone: +54-342-453-7227
E-mail: eocalafell@ciudad.com.ar; mclura@fbcb.unl.edu.ar