Dietary polyunsatured fatty acids (PUFA) differentially modulate melamine-induced preneoplastic urothelial proliferation and apoptosis in mice

Dietary polyunsatured fatty acids (PUFA) differentially modulate melamine-induced preneoplastic urothelial proliferation and apoptosis in mice

Prostaglandins, Leukotrienes and Essential FattyAcids (2001) 64(3),151^159 & 2001 Harcourt Publishers Ltd doi:10.1054/plef.2001.0255, available online...

362KB Sizes 0 Downloads 57 Views

Prostaglandins, Leukotrienes and Essential FattyAcids (2001) 64(3),151^159 & 2001 Harcourt Publishers Ltd doi:10.1054/plef.2001.0255, available online at http://www.idealibrary.com on

Dietary polyunsatured fatty acids (PUFA) differentially modulate melamine-induced preneoplastic urothelial proliferation and apoptosis in mice D. C. Cremonezzi,1,2 R. A. Silva,1 M. del Pilar D|¤ az,3 M. A. Valentich,1 A. R. Eynard1 1

Instituto de Biolog|¤ a Celular, Facultad de Ciencias Me¤dicas, Universidad Nacional de Co¤rdoba, Co¤rdoba, Argentina Research Fellow of Secretar|¤ a de Ciencia y Tecnolog|¤ a, Universidad Nacional de Co¤rdoba 3 Catedra de Estad|¤ stica y Biometr|¤ a de la Facultad de Ciencias Agropecuarias, Universidad Nacional de Co¤rdoba 2

Summary A numberof experimental and epidemiological studiesindicate that dietary polyunsaturated fattyacids (PUFA) playa modulatory role in the development of several cancers. However, literature on the importance of dietary PUFA in urinary-tract tumourigenesis is scarce, and even contradictory.Therefore, our purpose was to evaluate comparatively, several urothelial cellular parameters linked to neoplasia when180 BALB/c mice were initiated with the tumourigenic agent melamine and fed with two amounts of different PUFA.In experiment1, mice were fed with 6% of fish oil (enriched in n-3 PUFA, FO), corn oil (enriched in n6, CO) and olein (enriched in n-9, an EFA deficiency inducer) formulae plus two chow-fed control lots with (CM) and without (C) melamine treatment.In experiment 2, each of the three varieties of PUFA were offered at10%.Following18^22 weeks of melamine treatment, animals were autopsied. The liver fattyacid profile showed a close correlation with the dietary sources, exhibiting in the O group macroscopic and biochemical EFA-deficient (EFAD) characteristics. The frequency of simple urothelial hyperplasias (H) and dysplasia/carcinoma in situ (D/CIS) was significantly lower in the FO group, whereas both types of lesions increased in the CO and O groups, compared to the C and CM mice. Increased proliferation and abnormal luminal localized mitosis were more frequently recorded in EFAD mice, whereas abnormal apoptotic/mitosis ratio increased in both olein- and corn-oil-fed animals. This study shows that dietary PUFA modulate differentially normal and pre-neoplastic proliferation when induced by the tumorigenic agent melamine. Fish oil, rich in n-3 fatty acids, exhibits a clear antipromoting activity, whereas the role of n-6 and n-9 PUFA derivatives needs further research. & 2001Harcourt Publishers Ltd

INTRODUCTION Epidemiological and experimental results show that high dietary fat intake, mainly of the saturated type, leads to an increased risk of developing breast, colon and prostate cancer. Nowadays, around 35% of all human cancer deaths have been estimated to be linked to diet.1,2

Received 8 August 2000 Accepted 27 October 20000 Correspondence to: A.R. Eynard, Instituto de Biolog|¤ a Celular, Facultad de Ciencias Me¤dicas, UNC, Casilla de Correo 220, 5000 Co¤rdoba, Repu¤blica Argentina. E-mail: [email protected]

& 2001Harcourt Publishers Ltd

The role played by dietary fats in urinary-tract tumourigenesis is unclear. Epidemiological evidence showed that subjects having a high intake of fats had an increased risk of bladder tumours.3 However, a recent study using purified fatty acids as a dietary source of lipids showed no tumour-promoting activities due to saturated (palmitic) and essential fatty acids (EFA) (such as linoleic or linolenic acids) on the urinary tract of N-butyl-N-(4-hydroxybutyl) nitrosamine (BBN)-treated rats.4 On the other hand, our earlier studies showed that rats with chronic EFA deficiency (EFAD) suffered a significantly higher incidence of urothelial tumours, dysplasia and carcinoma in situ (D/CIS) than the cornoil-fed controls.5,6 Several studies support the hypothesis that EFA are beneficial as preventive and therapeutic

Prostaglandins, Leukotrienes and Essential FattyAcids (2001) 64(3), 151^159

152

Cremonezzi et al.

nutrients in the avoidance of cancer.7–13 Thus, it is reasonable to assume that the deprivation of EFA, by inducing an EFAD state, may act as a pro-tumourigenic situation.14–16 Lesions usually associated with the early stages of tumourigenesis, such as enhanced proliferative activity, coupled with the abnormal differentiation of certain cell populations, mainly epidermoid or urinary epithelia, are consistently observed in EFAD animals and humans. These findings indicate that PUFA may have a role in the regulation of urothelial cell proliferation and differentiation.6 Data from both epidemiologic and rodent models show that chronic urinary tract inflamation induced by lithiasis is a significant risk factor in the development of transitional cell tumours.17,18 Melamine (2,4,6-triamino-s-triazine) is a resin widely used in thermosetting domestic plastics. Twenty-four hours after its ingestion, 90% of the administered dose is excreted in urine.19 A high percentage of rodents, chronically treated with melamine, developed uroliths and transitional cell carcinomas.20 The purpose of this study was to test whether normal and neoplastic proliferation and apoptosis of urothelial mucosa, chronically treated with melamine, is modulated by two different amounts of dietary n-3 and n-6 EFA as well as to investigate the role played by an EFAD condition induced in mice by a non-EFA lipidic source belonging to the non-EFA n-9 family.

MATERIALS AND METHODS

Animals and diets A total of 180 homozygous weanling BALB/c mice of both sexes were randomly distributed into eight different dietary groups in two sets of experiments, for 18–22 weeks, as shown in the experimental design depicted in Table 1. The composition of the basic diet21 was, (%): casein (16), sucrose (34), corn starch (39), fibre (2), salt mixture (3.5) and vitamins mixture (0.5). Table 2 shows the fatty acid composition of all the fats used.

Table 2

Table 1 Experimental design Experimental number

Experimental group (n )

Lipids (percentage)

1

CO FO O CM C

(23) (18) (18) (27) (21)

Corn oil, 6 Fish oil, 6 Olein, 6 ^ ^

2

CO FO O

(19) (20) (13)

Corn oil,10 Fish oil,10 Olein,10

CO: corn oil; FO: fish oil; O: olein, CM: commercial plus melamine; C: commercial. (n)= number of animals. All groups, except C, were treated with1.2% melamine. Experiment1 (6% fats /22 weeks) and experiment 2 (10% fats/18 weeks).

The n-3 and n-6 EFA sufficient (EFAS) diets were prepared by adding cod liver oil (Parafarm, Buenos Aires, Argentina) 6% in experiment 1, and 10% in experiment 2, and corn oil, respectively. The EFAD formula was achieved adding 6% or 10% olein (O) refined twice (Retienne, Cordoba, Argentina). Control mice were fed on a commercial diet (NUTRIC-ADIABIC, Co´rdoba), which supplied 6% of fat (only experiment 1). These formulae yielded 3.86 cal/g in experiment 1 and 4.06 cal/g in experiment 2 respectively. Melamine (Sigma Chemical Co.), in powder form, was mixed 10 days after weaning into the diet at a concentration of 1.2%.20,22 Animals from two experimental groups were inspected daily, and food consumption and weight were recorded. Food and water were provided ad libitum.

Autopsy protocol At the end of the experiments the mice were killed under ether anaesthesia. The proximal urethra was dissected and clamped, and the urinary bladder and ureters were gently inflated with 10% buffered formalin, in situ fixed for 60 min and then the block of urinary tract was removed, cleaned and further fixed. Urinary bladders

Fatty acid composition of dietary fats

Diet

Saturated

Mono-UFAs

PUFAs

*

16:0

18:0

20:0

16:1 n-7

18:1 n-9

18:2 n-6

18:3 n-6

18:3 n-3

20:5 n-3

22:5 n-3

22:6 n-3

C CO FO O

25.4 13.5 12.7 3.6

8.1 4.6 2.5 0.3

1.7 1.1 3.4 ^

1.5 0.6 9.4 7.6

41.2 33.1 35.0 79.5

10.1 47.1 2.6 2.9

^ ^ ^ 0.6

3.3 tr 1.5 1.2

^ ^ 5.3 ^

^ ^ 4.8 ^

^ ^ 10.4 ^

Mono-UFAs monosaturated fatty acids; PUFAs polyunsaturated fatty acids; tr traces. *Control (C); corn oil (CO); fish oil (FO); olein (O). Values represent mean of percentage of each fatty acid of no less than three samples run in duplicate. Prostaglandins, Leukotrienes and Essential FattyAcids (2001) 64(3), 151^159

& 2001Harcourt Publishers Ltd

Dietary PUFA modulate proliferation/apoptosis in mice urothelium 153

were divided into two halfs. Samples of pelves, ureters and bladders were routinely processed for histologic examinations, and representative sections were stained with hematoxylin and eosin and periodic acid-Schiff reaction (PAS).

Gas-liquid chromatography (GLC) analysis Considering the unsuitable small amount of material yielded by mouse urothelial mucosa, livers were chosen, since hepatic fatty acids mirror accurately the metabolic changes induced by dietary manipulations.23

Mitotic index and apoptotic counting At the end of the experiments five mice of each dietary group from experiment 1 were injected intraperitoneally at 9 a.m. with colchicine, 0.3 mg/100 g b.w. (Sigma Chemical Co.) in 0.9% sodium chloride.24 Animals were sacrificed 5 h later and bladders were excised and processed for histologic examination. On at least four non-serial sections stained with HE, the number of metaphases and apoptotic figures per 2000 nuclei of cells in interphase of urothelium were counted. As the presence of mitosis at luminal layers is considered a sign of abnormal differentiation and proliferation, a distinction was made between basal and luminal epithelial layers. Apoptotic figures were identified and recorded according to the criteria described by Cummings.25

among dietary treatments through the variance-covariance matrix estimate. Finally, the usual diagnostic techniques were used to check the proposed models.29 RESULTS Diets and water consumption were similar in all groups. As expected, mice fed on the olein diet developed macroscopic signs of EFAD in varying degrees.21 These animals were smaller but weight differences were not significant at the end of experiment. Increased ‘spontaneous’ death (five out of 23 in experiment 1 and seven out of 20 in experiment 2) was observed in this group during the first 4 weeks of experimentation, perhaps linked to the EFAD condition.23 Mortality was rare in the other groups (1/24 CO, 1/19 FO and 2/22 C in experiment 1, and 3/22 CO and 2/22 FO in experiment 2).

Urinary-tract gross lesions Bladder lithiasis ranged from 60 to 85% in all groups treated with melamine, and no significant differences among them were recorded. Thickened whitish areas of the mucosas were observed on bladders, whereas on ureters and pelvices, several degrees of uretero-hydronephrosis were observed in all groups. The size of single or multiple stones ranged from 1 to 5 mm. In the C mice group, without melamine, no lithiasis was observed.

Liver fatty acids Pathological examination The benign and malignant proliferative lesions observed in the urinary epithelia (pelves, ureters, bladder) were evaluated according to Friht et al.26 and Boorman et al.27 Urothelial lesions were classified as: simple hyperplasia without atypia (H), dysplasia/carcinoma in situ (D/CIS) and invasive carcinoma, as illustrated in Figure 1.

As shown in Table 3, the fatty acids of total liver lipids reflected fairly well the fatty acid composition of each formula. The 0.63 triene/tetraene ratio in the O group was consistent with an EFAD condition.30 On the contrary, values were below the 0.2 limit in the other mice: 0.03 in CO, 0.11 in FO, 0.04 in CM and 0.05 in C lots.

Experiment 1 (6% fats) Statistical evaluation

Renal pelves

As data were discrete random variables, dietary effects were analyzed by fitting Generalized Linear Models. These define three components: an error one, a linear predictor, and the link function between them. For the data matrix we selected the Binomial, Extra-Binomial and Poisson error components. An analysis of variance model with the dietary effects parameters as a lineal predictor, and logarithm or logistic link depending on the response variables (number of lesions or percentages, respectively) was used. All these models were constructed using GLIM Release 4. 09 (Generalized Linear Iterative Models-NAG Algorithm).28 After the models were adjusted, some specific contrasts were made to evaluate differences & 2001Harcourt Publishers Ltd

A greater number of transitional H was found in CO animals compared to the others (P 5 0.05). However, there was a significant amount of animals showing D/CIS in the O group (P 5 0.05), whereas the incidence of this lesion in the CO and FO groups was similar (Table 4).

Ureter H lesions were recorded in CO, FO and O fed mice more frequently than in CM and C control groups (P 5 0.05). The D/CIS lesions were more frequently seen in the CO and O lots than in the FO and control groups (P 5 0.05), while no significant differences between the latter experimental groups were seen (Table 4).

Prostaglandins, Leukotrienes and Essential FattyAcids (2001) 64(3), 151^159

154

Cremonezzi et al.

Fig. 1 Stained with H&E.Experiment1. (a) Control diet fed mouse, normal bladder.Epithelium consists of three layers of transitional cells. 6300. (b) Simple hyperplasia (H) seen in FO-fed mouse.There is an increase in the number of layers of transitional cells. 6300. Inset: Cytoplasms become dark and some nuclei are hyperchromatic (arrows). 6500. (c) Example of D/CIS seen in O-fed mouse. Note disorganization within hyperplastic layers of bladder urothelium. Cells and nuclei are not uniform in shape and size. Mitotic figures are frequent (arrows). 6300. Inset: prominent nucleoli and apoptotic figures (arrows). 6500. (d) Transitionalinvasive carcinoma of bladder with areas of squamous metaplasia seen in CO-fed mouse. 6300. Inset: invasion of the lamina propria. 6500.

Bladder Most CM fed animals showed preservation of the normal appearance of the epithelia (Fig. 1a) in comparison to the other dietary groups. Simple H lesions (Fig. 1b) were observed in the urothelium from the CO, FO, O and CM groups. Although not significantly different among semishyntetic fed mice, simple hyperplasia were more frequently seen in these groups than in C dietary lot (P50.05). In contrast, most of the D/CIS (Fig. 1c) lesions were observed in bladders of the CO and O groups (Table 4). Incidentally, one invading transitional carcino-

ma (Fig. 1d) was observed in the CO group. Taken as a whole, there were increased D/CIS lesions from the renal pelvis to the bladder, except in the C group.

Experiment 2 (10% fats) Renal pelvis More than 50% of FO-fed mice showed a preserved urothelial morphology (P 5 0.05). Coincidently, in the CO group a significant frequency of D/CIS was observed (P 5 0.05) (Table 5).

Prostaglandins, Leukotrienes and Essential FattyAcids (2001) 64(3), 151^159

& 2001Harcourt Publishers Ltd

Dietary PUFA modulate proliferation/apoptosis in mice urothelium 155

Triene (n-9)/ tetraene (n-6) ratio

0.06 0.04 0.03 0.63 a 0.11

n-3 22:6

0.22 0.28 2.61 1.36 0.71

Ureter The integrity of the ureteral urothelium was maintained in a significant number of FO-fed mice. H lesions were more frequently observed in the O group than in the FO or CO supplemented diet (P 5 0.05). The CO group exhibited a significant amount of D/CIS lesions in contrast to the O or FO dietary groups (P 5 0.05) (Table 5).

As depicted in Table 5, the urothelial mucosa of FO-fed mice mainly showed either an absence of lesions or just the development of simple H, whereas D/CIS developed more frequently in CO and O animals compared to FO ones (P 5 0.05). Taken as a whole, however, increased percentages of FO doeas not provided better protection.

1.22 1.06 1.54 0.14 4.85 C CM CO O FO

Mitotic index

CO: corn oil; FO: fish oil; O: olein; CM: commercial plus melamine; C: commercial. a Values above 0.2 indicate essential fatty acid deficiency (EFAD).23,29 Values for liver from experiment two were similar.

0.21 0.45 0.9 0.12 0.48 0.15 0.33 0.27 0.11 1 10.5 13.2 13.4 4.8 2.2 3.84 2.57 2.21 6.22 9.7

& 2001Harcourt Publishers Ltd

0 0.20 2.81 0.22 0.45

29.3 33.4 28.0 24.4 28.2

11.2 18.0 12.3 10.2 8.1

37.6 20.9 20.2 44.2 39.3

0.5 0.39 0.42 0.99 0.78

0.5 0.4 0.36 0.45 0.76

0.24 0.19 0.41 1.89 0.32

4.21 4.5 12.1 3.04 2.9

n-6 22:4 n-3 20:5 n-6 20:4 n-9 20:3 n-6 18:2 n-7 16:1 14:0

16:0

18:0

n-9 18:1

n-3 18:3

n-3 18:4

Fatty acids Experiment group

Table 3

Experiment1. Percentage composition of whole liver fatty acids for mice fed diets with 6% of different fats

n-3 22:5

Bladder

Basal mitosis recorded in the bladder epithelium showed that, in the CM group, values were lower than in the CO and FO groups (P 5 0.05) and especially in comparison to the O group (P 5 0.01) (Table 6). Mitotic index in the O group was significantly higher than in the FO group (P 5 0.001). When localization of mitoses, as a measure of cellular atypia, was analyzed, the data from the FO and CM groups did not significantly differ. In the O-fed mice, the number of mitosis located at the luminal layers was significantly higher than in the remnant groups (P 5 0.05). The control group plus melamine (CM) showed values significantly lower than the CO and O groups (P 5 0.05). Taken as a whole, the figures for total mitosis (basal and luminal) indicate that chow-fed animals exhibited values significantly lower than those of the CO, FO and O lots (P 5 0.05) while increased cellular proliferation was recorded in the urothelium of the O-fed, EFAD mice (Table 6).

Apoptosis Apoptotic figures and the apoptosis/mitosis ratio are presented in Table 6. Values for the CO mice were significantly lower than those of O mice (P 5 0.27). On the contrary, chow-fed mice exhibited apoptotic values significantly lower than the other semisynthetic diet-fed animals (P 5 0.01).

Apoptotic/mitotic ratio (A/M ratio) Assuming that both mitotic and apoptotic figures are extremely low in normal urothelium, the A/M ratio recorded for the commercial (C) non-treated mice group may be considered as ‘normal’ (Table 6). CM fed mice showed a ratio greater than 1, whereas values recorded for CO and O, and to a lesser degree FO, were below 1. Prostaglandins, Leukotrienes and Essential FattyAcids (2001) 64(3), 151^159

156

Cremonezzi et al.

Table 4

Experiment1. Proliferative lesions of urothelial tract in mice fed with 6% of different fats

Diet (n)

Renal pelvis NL

CO (23) FO (18) O (18) CM (27) C (21)

H

5 12a 6 21a 19a

15 2 2 2 1

Ureter D/CIS

b

NL

3 4 10c 4 1

2 2 1 17a 21a

H

Bladder D/CIS

b

11 12b 8b 3 0

10 4 9c 7 0

c

NL 1 5 0 11a 21a

H

D/CIS

b

13c 5 11c 9 0

9 8b 7b 7 0

CO: corn oil; FO: fish oil; O: olein; CM: commercial plus melamine; C: commercial. All values express number of mice with proliferative lesions. NL: no lesions. aSignificantly different to the CO, FO and O dietary groups at P50.05 (in renal pelvis minus FO group). H: hyperplasia. bSignificantly different to the CM and C dietary groups (in renal pelvis plus FO and O lot) at P50.05. D/CIS: dysplasia/in situ carcinoma. cSignificantly different to the CM, C and FO dietary groups (in renal pelvis plus CO lot) at P50.05.

Table 5

Experiment 2. Proliferative lesions of urothelial tract in mice fed with10% different fats

Diet (n)

Renal pelvis NL

CO (19) FO (20) O (13)

3 11a 6

Ureter

H

D/CIS

3 2 2

a

NL

13 7 5

0 5a 1

Bladder

H

D/CIS

NL

H

D/CIS

5 5 8a

a

0 4a 1

3 8 5

15a 8 7

14 9 4

CO: corn oil; FO: fish oil; O: olein, NL: no lesions; H: hyperplasia; D/CIS: dysplasia/in situ carcinoma Values represent number of mice with recorded lesions. a Significantly different from the other dietary groups at P50.05. Table 6

Experiment1. Mitotic index and apoptotic figures in bladder urothelium obtained from feeding mice with 6% dietary fats

Diets

Mitotic index Epithelial layer Basal mitosis

CO FO O CM C

a

1.14 1.08a,c 2.2b,c 0.18a,b 0.01

Apoptosis

Apoptosis/total mitosis

0.13h 0.27 0.33h 0.40 0.01i

0.10 0.22 0.12 1.6 1.0

Total mitosis

Luminal mitosis 0.16d,e 0.10d 0.51d,e 0.07d,e 0.0d

1.3f,g 1.18f,g 2.71g 0.25g 0.01f,g

CO: corn oil; FO: fish oil; O: olein; CM: commercial plus melamine, C: commercial. Apoptotic values were average values from five observations. All values with the same superscript letter are significantly different: basal mitosis 0.18, (a) P50.05 and (b) P50.01; 2.2 , (c) P50.001; luminal mitosis 0.51, (d) P50.05; 0.07 (e) P50.05;Total mitosis 0.01(f) P50.05 and 2.71 (g) P50.05. (h) P=0.27 (median). (i) P50.01for group C with respect to the others.

Miscellaneous lesions There were no significant differences among groups, with around 15–20% of animals showing severe necrotizing papillitis. A variable degree of damage at the tip of the papillae, probably mediated by secondary obstruction of the blood flow, along with acute inflammation and haemorrhage, was seen. Scattered inflammatory areas with oedema, vascular congestion and leucocitary infiltration were observed in the bladder wall. Atypical urothelial lesions usually coexisted with ulcers and desmoplastic changes. DISCUSSION The high percentage of both typical and atypical proliferative urothelial lesions induced by 1.2% melamine

corroborates that this is a suitable model for our purpose, in spite of the fact that only one case of invasive transitional carcinoma developed. In the present study we have clearly demonstrated that dietary PUFA in mice differentially modulate urothelial proliferation and apoptosis induced by melamine. The effect of 6% or 10% n-3 and n-6 dietary fats slightly differed in the incidence of D/CIS lesions, and in the O group, no remarkable differences were observed. Therefore, despite the increased frequency of D/CIS when fats were offered at 10% to CO and FO groups, the beneficial effect of fish oil continues. However, a higher dose of FO was not more effective. Under physiological conditions, cells of the mammalian urothelium display a low turnover. Although urothelium normally has a very low rate of proliferation, no previous

Prostaglandins, Leukotrienes and Essential FattyAcids (2001) 64(3), 151^159

& 2001Harcourt Publishers Ltd

Dietary PUFA modulate proliferation/apoptosis in mice urothelium 157

data about normal values of apoptosis for urothelium in mice were found in the literature.31,32 We observed that the apoptotis/mitotic index ratio in normal, non-treated mice urothelium, was 1.0. This is not surprising, insofar as the normal appearance of epithelial tissues is the result of a homeostatic balance among proliferation, differentiation and apoptosis.33 The urothelium of fish-oil supplemented animals treated with melamine exhibited lower values of proliferation. With respect to the CO and O groups and their mitotic indices, the values obtained were closer to those recorded in chow-fed mice. Coincidently, when proliferative lesions developed in FO-fed mice, they were mostly minor and well differentiated in comparison to the other groups fed on a semisynthetic diet. With regard to the apoptosis/mitosis ratio, fish-oil fed animals behaved similarly to melamine-treated chow-fed mice, showing increased apoptosis, perhaps as a reactive mechanism to discard heavily melamine-damaged cells. Most of the FO animals showed mild proliferative lesions with an evident ‘shift to the left’ (Tables 4 and 5). The antipromoting effect of fish oil is in agreement with previous dietary experimental studies, in which an evident reduction in the development of several tumours in other rodent organs was consistently found.34–37 As far as we know, this is the first report which has noted an antipromoting effect exerted by n-3 fatty acids on the urothelial mucosa exposed to tumourigenic chemicals. The reasons for this beneficial activity of FO are unclear. Interestingly, macrophages of mice fed with FO have a diminished protein kinase C activity, this protein being the target for tumoural promoters like phorbol-esters.38 Recently, Kitano et al. reported the lack of promoting effects of 18:2n-6, 18:3n-9 and 16:0 on the urinary bladder mucosa of rats treated with BBN. These discrepancies with our results may be explained by the fact that in their studies, those authors used pure fatty acids, instead of fats or oils, which is not a physiological way of lipid intake per se. Besides, a strong carcinogen (BBN) was employed, yielding 15–30% of invasive carcinoma, whereas in our experiments a mild urothelial irritant was chosen. Moreover, their rats were fed from 35 days old, when fats depots are already formed hence minimizing the effects of dietary manipulation. Although not mentioned in their work, the n-3 and n-6 EFA dietary sources induced the lowest BrdU labelling indices in bladder urothelium.4 Urothelial proliferative lesions of melamine-treated mice fed on corn or olein-supplemented formula behave similarly, both being significantly more severe than those observed in the FO group. On the other hand the O mice, which are EFA deficient, exhibited significantly higher values of mitotic indices as compared to the CO and to the remmant groups. Increased proliferation in the EFAD mice was also significantly higher in the luminal layers. & 2001Harcourt Publishers Ltd

The A/M ratio was heavily disturbed in CO and O, being more severe in the EFAD mice. Differential effects of dietary PUFA on the regulation of normal or neoplastic cellular proliferation are not yet clear. Horrobin et al. and Das et al. proposed that low values of lipoperoxides obtained from oxidation of double chained fatty acids arrest proliferation;9,10,39 i.e. the higher the degree of unsaturation the greater the production of lipoperoxides, explaining the beneficial effects of the long chained, highly unsaturated fatty acids, belonging to n-3 PUFA of fish oil. Alternatively, one of the key mechanisms regulating apoptosis involves the functional permeability and integrity of cellular junctions, mainly nexus or tight junctions.40 In this connection, it has been recently demonstrated that g-linolenic acid (GLA), an EFA derivative from n-6 linoleic acid, modulates the expression of many proteins linked to cell–cell adhesion of several human tumour cell lines.41 In contrast, it has been postulated that the EFA-deficient condition may be deleterious for the functioning of cell– cell and cell-adhesions.42 Also, it has been reported, that in human skin cancer cells, there is a decreased expression of the antimetastic e-cadherin when incubated with 20:3n-9, the typical PUFA which appears in the EFAD condition.29,43 The corn oil (n6) and olein (EFAD) diet had a significant promoting effect on the urothelial lesions induced by melamine, with a greater number of mice developing D/CIS lesions, in comparison with fish oil and control dietary groups. Nonetheless, the role played by linoleic acid on tumourigenic events remains controversial. Indeed, in numerous tumoural breast models, linoleic acid and/or corn oil had a promoting effect,44 while in other breast and squamous cell tumour models these fats exhibited an anti-promoting behaviour.35,45 Human studies have failed to find any evidence of increased cancer risk with linoleic acid. In agreement with Lockniskar et al.,45 there is the possibility that in the urothelium, unusual 18:2n-6 derivatives, belonging to n-3, n-6 and n-9 families, exhibit a kind of ‘specific-organ’ effect, as showed for the role played by these fatty acids on anisotropic properties in rat urothelium.46,47 In conclusion, this research indicates that dietary handling of PUFA-containing fats differentially modulates atypical, preneoplastic urothelial lesions induced by a lithogenic irritant such as melamine. Lesions were more frequent in the bladder, perhaps due to a longer time period of urine contact with the bladder mucosa. Furthermore, the protumorigenic activity of melamine, widely used in domestic life, is clearly attenuated by dietary fish oil and exhibits more aggressive properties when it is offered along with corn oil and olein. Undoubtedly, extended periods of investigation are

Prostaglandins, Leukotrienes and Essential FattyAcids (2001) 64(3), 151^159

158

Cremonezzi et al.

necessary to induce full blown, invasive carcinomas as well as to further study the role of different dietary PUFA on the underlying mechanisms.

ACKNOWLEDGEMENTS Dr Martha Gonzalez Cremer and Paul David Hobson (PhD) are thanked for critical reading of the manuscript, Ne´lida Ramonda, Gustavo Morales and Lucı´a Artino for their skillful technical assistance and Ricardo Matos for animal care. This work was supported by SECyTUNC, CONICET and CONICOR.

REFERENCES 1. La Vecchia C. Cancers associated with high fat diets. J Natl Cancer Inst Monogr 1992; 12: 79–85. 2. Navarro A., Osella A. R., Mun˜oz S. E., Lantieri M. J., Fabro E. A., Eynard A. R. Fatty acids, fibers and colorectal cancer risk in Co´rdoba, Argentina. J Epidemiol Biostat 1998; 3: 415–422. 3. Hebert J. R., Miller D. K. A cross-national investigation of diet and bladder cancer. Eur J Cancer 1994; 3A: 778. 4. Kitano M. K., Mori S., Chen T., Murai T., Fukushima S. Lack of promoting effects of a-linolenic, linoleic or palmitic acid on urinary bladder carcinogenesis in rats. Jpn J Cancer Res 1995; 86: 530–534. 5. Monis B., Eynard A. R. Incidence of urothelial tumors in rats deficient in essential fatty acids. J Natl Cancer Inst 1980; 64: 73–79. 6. Monis B., Eynard A. R. Abnormal cell proliferation and differentiation and urothelial tumorigenesis in essential fatty acid deficiency (EFAD) rats. Prog Lip Res 1981; 20: 691–703. 7. Horrobin D. F., Begin M. E. Essential fatty acids: metabolism, interaction with free radicals and clinical use. In: Miguel J., Quintanilha A. T., Weber H., eds. Handbook of Free Radicals and Antioxidandt in Biomedicine, Vol. II. Boca Raton: CRC Press 1989; 17–26. 8. Horrobin D. F. Gamma linolenic acid. An intermediate in essential fatty acid metabolism with potential as an ethical pharmaceutical and as a food. Rev Contemp Pharmacother 1990; 1: 1–45. 9. Horrobin D. F. Essential fatty acid, lipid peroxidation and cancer. In: Horrobin D. F., ed. Omega 6 Essential Fatty Acid: Pathophysiology and Roles in Clinical Medicine. New York: Alan Liss 1990; 351–377. 10. Horrobin D. F. Is the main problem in free radical damage caused by radiation, oxygen and other toxins the loss of membrane essential fatty acids rather than the accumulation of toxic materials? Med Hypotheses 1991; 35: 23–26. 11. Ip C., Scimeca J. A., Thompson H. J. Conjugated linoleic acid. A powerful anticarcinogen fron animal fat sources. Cancer 1994; 74: 1050–1054. 12. Van Aswegen C. H., Du Plessis D. J. Can linoleic acid and gammalinolenic acid be important in cancer treatment? Med Hypotheses 1994; 43: 415–417. 13. Das U. N., Prasad V. V. S. K., Reddy D. R. Local application of gamma-linolenic acid in the treatment human gliomas. Cancer Lett 1995; 94: 147–153. 14. Eynard A. R. Role of dietary polyunsaturated fatty acids (PUFA) on tumorigenesis. Cancer J 1996; 9: 142–144.

15. Eynard A. R. Does chronic essential fatty acids deficiency (EFAD) constitute a protumorigenic condition? Med Hypotheses 1997; 48: 55–62. 16. Eynard A. R. Is the risk of urinary tract tumorigenesis enhanced by a marginal chronic essential fatty acid deficiency (EFAD)? Nutrition 1998; 14: 1–6. 17. La Vecchia C., Negri E., D’Avanzo B., Savoldelli R., Franceschi S. Genital and urinary tract diseases and bladder cancer. Cancer Res 1991; 51: 629–631. 18. Higgy N. A., Verma A. K., Erturk E., Bryan G. T. Augmentation of N-butyl-N-(4-hydroxybutyl)nitrosamine (BBN) bladder carcinogenesis in Fischer 344 female rats by urinary tract infection. Proc Amer Assoc Cancer Res 1985; 26: 118–124. 19. Mast R. W., Jeffcoat A. R., Sadler B. M., Kraska R. C., Friedman M. A. Metabolism, disposition and excretion of [14C] melamine in male Fischer 344 rats. Food Chem Toxicol 1983; 21: 807–810. 20. Ogasawara H., Imaida K., Ishiwata H., Toyoda K., Kawanishi T., Uneyama Ch., Hayashi S., Takajashi M., Hayashi Y. Urinary bladder carcinogenesis induced by melamine in F344 male rats: Correlation between carcinogenicity and urolith formation. Carcinogenesis 1995; 16: 2773–2777. 21. Eynard A. R., Cejas V., Silva R., Quiroga P., Mun˜oz S. Histopathology of essential fatty acid-deficient mice. Nutrition 1992; 8: 37–40. 22. Heck H. d’A., Tyl R. W. The induction of bladder stones by terephthalic acid, dimethyl terephthalate, and melamine (2,4,6Triamino-s-triazine) and its relevance to risk assessment. Reg Toxicol Pharmacol 1985; 5: 294–313. 23. Aaes-Jorgensen E. Essential fatty acids. Physiol Rev 1961; 41: 1–51. 24. Eynard A. R., Monis B., Kalinec F., Leguizamo´n R. O. Increased proliferation of the epithelium of the proximal alimentary tract of EFA-deficient rat: a light and electron microscopy study. Exp Mol Pathol 1982; 36: 135–143. 25. Cummings M. C., Winterford C. M., Walker N. I. Apoptosis. Am J Surg Pathol 1997; 21: 88–101. 26. Friht C. H., Terrachini B., Turusov V. S. Tumours of the kidney, renal pelvis and ureter. In: Turusov V. S., Mohr U., Ed. Pathologhy of Tumors in Laboratory Animals, Vol II. Tumors of the Mouse. Lyon: IARC Scientific Publications N.111. 1994; 357–406. 27. Boorman G. A, Wood M., Fukushima S. Tumours of the urinary bladder In: Turusov V. S. Mohr U., ed. Pathologhy of Tumors in Laboratory Animals, Vol II. Tumors of the Mouse. Lyon: IARC Scientific Publications N.111. 1994; 357–406. 28. Francis B., Green M., Payne C. The GLIM System. Release 4 Manual. Clarendom Press. Oxford. 1994; p1021. 29. Collett D. Modelling Binary Data Analysis. Editors: Chapman and Hall. 1994. 30. Holman R. T. The ratio of trienoic: tetraenoic acid in tissue lipids as measure of essential fatty acid requirement. J Nutr 1960; 70: 405–410. 31. Hainau B., Dombernowsky P. Histology and cell proliferation in human bladders. Cancer 1974; 33: 126–155. 32. Farsuna T. Cell kinetics of mouse urinary bladder epithelium. I. Circadian and age variations in cell proliferation and nuclear DNA content. Virchows Archiv (Cell Pathol) 1975; 18: 35–49. 33. Roberts R. A., Nebert D. W., Hickman J. A., Richburg J. H., Goldsworthy T. L. Perturbation of the mitosis/apoptosis balance: a fundamental mechanism in toxicology. Fundam Appl Toxicol 1997; 38: 107–115. 34. Carrol K. K., Braden L. M. Different effects of dietary polyunsaturated vegetable and fish oils on mammary tumorigenesis in rats. Prog Lipid Res 1986; 25: 583–585.

Prostaglandins, Leukotrienes and Essential FattyAcids (2001) 64(3), 151^159

& 2001Harcourt Publishers Ltd

Dietary PUFA modulate proliferation/apoptosis in mice urothelium 159

35. Silva R. A., Mun˜oz S. E., Guzma´n C. A., Eynard A. R. Effects of dietary n-3, n-6 and n-9 PUFA’s on benzo-a-pyrene induced forestomach tumorigenesis in C57 BL6J mice. Prostagl Leuk Essent Fatty Acid 1995; 53: 273–277. 36. Narisawa T., Fukawa Y., Yazawa K., Ishikawa C., Isoda Y., Nishizawa Y. Colon cancer prevention with a small amount of dietary perilla oil high in alpha-linolenic acid in an animal model. Cancer 1994; 73: 2069–2075. 37. Pell J. D., Brown J. C., Jhonson I. T. Polyunsaturated fatty acid of the n-3 series influence intestinal crypt cell production in rats. Carcinogenesis 1994; 15: 1115–1119. 38. Erickson K. L., Hubbard E. N., Chakrabarti R. Modulation of signal transduction in macrophagues by dietary fatty acids. J Nutr 1995; 125: 1683–1686. 39. Das U. N., Begin M. E., Ells G. Polyunsaturated fatty acid augment free radical generation in tumor cells in vitro. Biochem Biophys Res Commun 1987; 145: 15. 40. Tsujii M., Dubois R. N. Alterations in cellular adhesion and apoptosis in epithelial-cells overexpressing prostaglandinendoperoxide-synthase-2. Cell 1995; 83: 493–501. 41. Jiang W. G., Bryce R. P., Horrobin D. F. Essential fatty acids: molecular and cellular basis of their anti-cancer action and clinical implications. Oncol Hematol 1998; 27: 179–209.

& 2001Harcourt Publishers Ltd

42. Jiang W. G., Eynard A. R., Mansel R. E. The pathology of essential fatty acid (EFA) deficiency; is it cell-adhesion mediated? Med Hypotheses 2000; 55: 257–262. 43. Eynard A. R., Jiang W. G., Mansel R. E. Eicosatrienoic acid (20 : 3n9) inhibits the expression of E-cadherin and desmoglein in human squamous cell carcinoma in vitro. Prostagl Leuk Essent Fatty Acids 1998; 59: 371–377. 44. Welsch C. W. Enhancement of mammary tumorigenesis by dietary fats: review of potential mechanisms. Amer J Clin Nutr 1987; 45: 192–202. 45. Locniskar M., Belury M. A., Cumberland A., Patrick K. E., Fischer S. M. The effect of dietary lipid on skin tumor promotion by benzoyl peroxide: comparison of fish, coconut and corn oil. Carcinogenesis 1991; 12: 1023–1028. 46. Caldero´n R. O., Glocker M., Eynard A. R. Lipid and fatty acid composition of different fractions from rat urinary transitional epithelium. Lipids 1998; 33: 1017–1022. 47. Caldero´n R. O., Eynard A. R. Fatty acids specifically related to anisotropic properties of plasma membrane from rat urothelium. Biochim Biophys Acta 2000; 1483: 174–184.

Prostaglandins, Leukotrienes and Essential FattyAcids (2001) 64(3), 151^159