Early exposure to restraint stress enhances chemical carcinogenesis in rat liver

Cancer Letters 161 (2000) 215±220

www.elsevier.com/locate/canlet

Early exposure to restraint stress enhances chemical carcinogenesis in rat liver Ezio Laconi a,b,*, Cristina Tomasi a, Francesca Curreli a, Stefania Diana a, Sergio Laconi a, Gino Serra c, Maria Collu d, Paolo Pani a a

Dipartimento di Scienze Biomediche e Biotecnologie, UniversitaÁ di Cagliari, Via Jenner, 09125 Cagliari, Italy b Ospedale Oncologico `A. Businco', ASL 8, Cagliari, Via Jenner, 09125 Cagliari, Italy c Dipartimento di Scienze del Farmaco, UniversitaÁ di Sassari, Sassari, Italy d Dipartimento di Neuroscienze, UniversitaÁ di Cagliari, Via Jenner, 09125 Cagliari, Italy Received 5 July 2000; received in revised form 19 September 2000; accepted 20 September 2000

Abstract This study examines the effect of a stress-associated condition on chemical hepatocarcinogenesis in the rat. Rats were given diethylnitrosamine (200 mg/kg. b.w., i.p.), followed, 1 week later, by three cycles of immobilization at room temperature. Two weeks after the last cycle they were treated according to the resistant hepatocyte protocol. At 4 weeks after selection, mean size of glutathione-S-transferase 7±7 positive foci/nodules was increased in the immobilized group (0:82 ^ 0:22 vs. 0:25 ^ 0:04 mm 2 in controls). Furthermore, at the end of 1 year 10/13 animals (77%) developed hepatocellular carcinoma in the former group, while only 6/14 (43%) incidence of cancer was found in controls. These results indicate that exposure to restraint stress early during carcinogenesis enhances the development of chemically-induced hepatocellular carcinoma in the rat. q 2000 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Restraint stress; Nictemeral inversion; Liver carcinogenesis

1. Introduction The in¯uence of stress-inducing conditions on cancer development has been the subject of several investigations, both at clinical and experimental level [1±13]. Animal experiments have generally documented a positive association between the two variables [9±13], although some studies have led to opposite conclusions [7,8,12]. For example, foot shock enhanced the growth of transplanted neoplastic cells in mice [9] and a similar pattern of results have been reported when psychosocial stress (isolation) was * Corresponding author. Tel.: 139-070-609-5271; fax: 139-070609-5273. E-mail address: [email protected] (E. Laconi).

used in animals bearing grafted tumours [10,14]. However, various types of stress were found to inhibit the growth of both transplanted and chemicallyinduced mammary tumours in rats [8,12]. In addition, the size of virally-induced sarcomas in mice was reduced when stress was elicited prior to virus inoculation, while an enhancing effect on tumour growth was seen when stress sessions followed exposure to the virus [7]. It has also been considered that the ability of the organism to cope with the stress is critical in determining its overall effect. Consistent with this proposition, escapable foot shock had no effect on the growth of transplanted tumours in mice, while inescapable shock increased tumour size under the same experimental

0304-3835/00/$ - see front matter q 2000 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0304-383 5(00)00621-2

216

E. Laconi et al. / Cancer Letters 161 (2000) 215±220

conditions [9]. Furthermore, the effect of inescapable shock was mitigated if mice received long-term shock treatment, suggesting a type of adaptation to stress [9]. The nature of the relationship between stress-associated conditions and cancer has also been investigated in humans. However, human studies have obviously inherent limitations (e.g. inadequate control groups) and are often dif®cult to interpret [1±4,15]. In addition, much attention has been paid to the possible in¯uence of stress on prognosis in patients who already have cancer [1,15,16], while very limited information is available on any possible direct role of stress in the development of neoplasia, at any step(s) [4]. In the present study we examined the effect of two types of stress-associated treatments, given early after carcinogen exposure, on the evolution of hepatocarcinogenesis in the rat. Liver carcinogenesis was induced according to the resistant hepatocyte (RH) protocol, a well characterized model systems for the analysis of experimental cancer development [17]. Stress-inducing regimens consisted of (1) immobilization at room temperature for 3 h or (2) nictemeral inversion. Immobilization is a classic method which has long been used to study the impact of stress on disease processes in experimental animals [18]. Nictemeral inversion causes a disturbance of normal circadian patterns and is considered as a mild stress-inducing regimen [19]. Moreover, it was shown to enhance the growth of transplanted tumour cells [20]. 2. Materials and methods In the ®rst series of experiments, 2-month old male Fischer 344 rats (Charles River Italia, Milan, Italy) were acclimatized for 1 week to a purina rodent chow diet (no. 5001) with food and water available ad libitum and an alternating light±dark cycle of 12 h. Rats were initiated with a single injection of diethylnitrosamine (DENA, 200 mg/kg. b.w., i.p., Sigma Chemical Co., St. Louis, MO). Starting 1 week later they were divided into 3 groups: group 1 was exposed to immobilization at room temperature for 3 h/day, in 3 consecutive days, every 2 weeks, for a total of three cycles. Immobilization performed by individually wrapping animals in a thin metal mesh from 19:30 to 22:30 h each day. The entire body length of the animal was wrapped, from the head to the base of

the tail, with no movements allowed. Timing was selected based on evidence indicating that the stress response to restrain is maximum when it is performed in late evening hours [18]. Rats in group two underwent nictemeral inversion, i.e. the light±dark cycle was inverted for 3 days a week, every 2 weeks, for three cycles. At the beginning of each shift animals were exposed to 24 h of continuous light, while the opposite (i.e. 24 h of continuous darkness) occurred at the end of the inversion period. Group 3 was kept as control under standard housing conditions. Two weeks after the last cycle of either immobilization or light±dark inversion, all three groups were subjected to a modi®ed version of the RH model [21] to selectively stimulate the growth of foci and nodules. Brie¯y, rats were given three consecutive daily doses of 2-acetylamino¯uorene (20 mg/kg b.w., i.g., Sigma) followed, on the 4th day, by a single administration of CCl4 (0.2 ml/kg b.w., i.g., mixed in corn oil, 1:1 v:v.). Animals from all groups were killed 4 weeks after dosing with CCl4. Macroscopic examination was performed; liver samples were taken and processed for standard histological (H&E) analysis and for immunohistochemical detection of glutathione-S-transferase 7-7 (GST 7-7) [22]. The antibody to GST 7-7 was a polyclonal antibody generated in rabbits following injection of puri®ed enzyme isolated from rat liver nodules (a gift from Dr. DSR Sarma, Toronto, Canada). Statistical analysis of results was performed using either Student's t-test or ANOVA followed by Tukey test, or Fischer's Exact Test. In a second study the long term effects of early exposure to restraint stess were examined. Experimental details were similar to those described above (groups 1 and 3); however, animals were killed 1 year after initiation with DENA or when seriously ill. Liver, lung, kidney, pancreas, spleen and other organs that looked grossly abnormal were excised and samples were taken and processed for histological analysis on H&E stained sections. Statistical analysis of results was performed with Fischer's Exact Test.

3. Results 3.1. Experiment 1 Final body weights and liver weights were similar

E. Laconi et al. / Cancer Letters 161 (2000) 215±220

217

Table 1 Final body weight and liver weight in rats exposed to restraint stress or nictemeral inversion a Treatment b

Final body weight (g)

Liver weight (g)

Liver weight (% b.w)

Restraint Nictemeral inversion Control

308 ^ 6 325 ^ 4 321 ^ 6

10.54 ^ 0.34 10.06 ^ 0.28 9.96 ^ 0.25

3.46 ^ 0.15 c 3.06 ^ 0.05 3.09 ^ 0.04

a

Data are mean ^ SE of ten animals. All groups were given DENA, 200 mg/kg b.w., followed by three cycles of either restraint or nictemeral inversion and selection according to resistant hepatocyte model (see Section 2 for details). c Signi®cantly different from control, P , 0:05, (Student's t-test). b

among different groups, except for a slight increase in relative liver weight in group 1 (Table 1), possibly due in part to the presence of larger nodules (Table 2). Incidence of GST 7±7 positive foci/nodules is reported in Table 2. No signi®cant differences were seen in the number of lesions per cm 4 among different groups. However mean size of foci/nodules was increased in the group exposed to immobilization compared to control. Furthermore, percent area occupied by GST 7±7 positive foci/nodules was higher in both groups 1 and 2 compared to group 3, although such increase was statistically signi®cant only in animals exposed to immobilization (group 1). No appreciable differences were noticed at histological level in liver samples obtained from various groups. Hepatocyte nodules were mostly eosinophylic, with pale nuclei and prominent nucleoli. 3.2. Experiment 2 Two animals from group 1 (immobilization) became sick at 9 and 11 months during the experiment and were killed. Both had histologically diagnosed hepatocellular carcinoma (HCC). Towards the end

of the experiment some animals in group 1 began to lose weight and this was re¯ected in a signi®cant difference in ®nal body weight between the two groups (Table 3). Percent liver weight was high in both groups, due to the presence of large hepatic nodules and/or cancers. Incidence of hepatocyte nodules and HCC is reported in Table 4. All animals in both groups had grossly visible tumours and/or hepatocyte nodules. However, 10/13 animals (77%) developed HCC in group 1, compared to only 6/14 (43%) in the controls. Lung metastases were present in three animals in each group. Other organs appeared histologically normal, except for two rats in each group presenting with leukemia and enlarged spleen, a common ®nding in this strain of animals [23]. Histologically cancers were mostly well differentiated HCC with trabecular pattern. Areas of necrosis and haemorrhage were frequent within cancerous lesions, with no inter-group differences. 4. Discussion Taken together, the results of this study indicate

Table 2 Incidence of GST 7-7 positive foci/nodules in rats exposed to restraint stress or nictemeral inversion a Incidence of GST 7-7 positive foci/nodules Treatment

b

Restraint Nictemeral inversion Control a b c d

No./cm 2

Mean area (mm 2)

Mean % area/total

15 ^ 1 17 ^ 1 11 ^ 2

0.82 ^ 0.22 c 0.27 ^ 0.03 0.25 ^ 0.04

11.1 ^ 2.4 d 4.8 ^ 0.8 2.6 ^ 0.3

Data are mean ^ SE of ten animals. See legend to Table 1 and Section 2 for details. Signi®cantly different from control, P , 0:01; (ANOVA followed by Tukey test). Signi®cantly different from control, P , 0:001, (ANOVA followed by Tukey test).

218

E. Laconi et al. / Cancer Letters 161 (2000) 215±220

Table 3 Final body weight and liver weight after 1 year in rats exposed to restraint stress a Treatment b

Final body weight (g)

Liver weight (g)

Liver weight (% b.w)

Restraint Control

318 ^ 16 c 353 ^ 5

21.85 ^ 2.76 25.08 ^ 3.63

7.00 ^ 0.99 7.27 ^ 1.16

a

Data are mean ^ SE of 13±14 animals. All groups were given DENA, 200 mg/kg b.w., followed by three cycles of restraint and selection according to resistant hepatocyte model (see Section 2 for details). c Signi®cantly different from control, P , 0:05, (Student's t-test). b

that exposure to restraint stress early during carcinogenesis enhances the incidence of hepatocyte nodules in rat liver. The effect appeared to be re¯ected mostly on the size (i.e. growth) of nodules, while their number was not signi®cantly different in restrained and control groups. A trend towards an increase in the relative area occupied by nodules was also seen in animals undergoing nictemeral inversion. However, the effect reached statistical signi®cance only in animals undergoing immobilization. Furthermore, in the latter group the incidence of chemicallyinduced HCC at the end of 1 year was nearly doubled compared to appropriate controls. The number of animals with metastasis was similar in both groups, suggesting that a less aggressive cancer phenotype may be promoted by restraint stress. However, studies involving larger numbers of animals are necessary to address this issue. Most of the available information relating stress and neoplastic disease in experimental systems has been derived from studies on transplantable tumours [9±12,14], i.e. cell population which have already Table 4 The incidence of hepatocellular carcinoma in rats exposed to restraint stress a Treatment b

Rats with HN

Incidence of HCC (%)

Rats with metastases

Restraint Control

100% 100%

10/13 (77) c 6/14 (43)

3 3

a Data are mean ^ SE of 13±14 animals. HN, hepatocyte nodules; HCC, hepatocellular carcinoma. b See legend to Table 3 and in Section 2 for details. c P ˆ 0:079 compared to control (Fischer's exact test).

acquired a full malignant phenotype. On the other hand, human studies have focussed mainly on the effect of psychosocial factors, including stress, on prognosis in patients who already have cancer [1,4,16], again limiting the analysis of the potential impact of these factors to very late stages of disease progression. By contrast, relatively less is known on any possible in¯uence of stress on cancer development, when exposure occurs early during neoplastic process [12]. Inhibition of chemically-induced mammary tumours was reported in Sprague±Dawley rats undergoing various types of stress-related treatments during or soon after exposure to the carcinogen [8]. However, another report described a marginal increase in mammary tumour incidence in Fischer 344 rats given dimethylbenz(a)antracene followed by chronic restraint stress [24]. Similarly, formation of carcinogen-induced aberrant crypt foci in the colon was enhanced in rats exposed to restraint stress throughout the experimental period (5 weeks) [13]. Contrasting effects were also observed on the development of virally-induced sarcomas in mice when exposure to stress occurred either prior to virus inoculation (which resulted in growth inhibition), or following treatment with the virus (which enhanced tumour growth) [7]. In a comprehensive review on this topic, Justice [12] proposed that the effect of stress may vary depending on the etiology of cancer. More speci®cally, it was proposed that stress per se enhances cancer of viral origin (which has a relevant immunologic component) while it inhibits carcinogenesis induced by chemicals, (where the role of the immune system is more limited, if any). These considerations add to the complexity of the relationship between stress-associated reaction and neoplastic process and highlight the need for using well characterized systems for the stepwise analysis of neoplasia if one is to examine the effect of stress on each phase of such long term process. In this study cycles of both restraint and nictemeral inversion were started 1 week after carcinogen treatment, in order to avoid interference with the initiation phase. At this stage only small clusters of initiated cells (#10 cells/cross section) are present [25]. On the other hand, stress sessions were stopped about 2 weeks prior to selection with the RH protocol. Thus, both the stress-associated reaction and the subsequent

E. Laconi et al. / Cancer Letters 161 (2000) 215±220

rebound response exerted their effects at very early stages of carcinogenesis, before grossly visible hepatocyte nodules were formed. Nevertheless, this effect was able to extend until very late in the process, when overt cancer developed. These results clearly indicate that stress can modulate the evolution of the entire carcinogenic sequence even when it is experienced very early during the process. Various mechanisms have been considered as possible mediators of the effect of stress on neoplastic process, including alterations in the immune and/or neuroendocrine systems and in the antioxidant defence status [11,24,26±29]. Irrespective of the mediators involved, it is noteworthy that a few cycles of restraint can exert a sizeable effect on the evolution of a chronic process such as cancer development, at least in this system. In a recent study we reported that three cycles of 3-day fasting followed by re-feeding was able to double the incidence of HCC induced by chemicals in rats [25]. These ®ndings give support to the notion that the long term process leading to the appearance of overt cancer is extremely susceptible to modulation by both endogenous and exogenous factors, many of which are within our control capability. Identi®cation of such factors, which may increase or decrease the risk of disease, represents a realistic strategy for the prevention of cancer in humans. Acknowledgements We thank Marinella Boi for her expert technical support and Tiziana Pusceddu for her excellent secretarial assistance. This work was supported in part by funds from Progetto Finalizzato CNR-ACRO, Italy (P.P) and from Sardinian Regional Government (Progetto 5/A) (P.P). References [1] S. Greer, T. Morris, K.W. Pettingale, Psychological response to breast cancer: effect on outcome, Lancet 2 (1979) 785±787. [2] V. Riley, Psychoneuroendocrine in¯uences on immunocompetence and neoplasia, Science 212 (1981) 1100±1109. [3] W.H. Redd, P.B. Jacobsen, Emotions and cancer. new perspectives on an old question, Cancer 62 (1981) 1871±1879. [4] B.H. Fox, The role of psychological factors in cancer incidence and prognosis, Oncology (Hunt.) 9 (1995) 245±253. [5] R.T. Croyle, Depression as a risk factor for cancer: renewing a

[6] [7] [8]

[9] [10] [11]

[12] [13]

[14] [15] [16] [17] [18] [19] [20]

[21]

[22]

[23]

219

debate on the psychobiology of disease, J. Natl. Cancer Inst. 90 (1998) 1856±1857. R.C. La Barba, Experiential and environmental factors in cancer: a review of research with animals, Psychosom. Med. 32 (1970) 259±276. A. Amkraut, G.F. Solomon, Stress and murine sarcoma virus (Moloney)-induced tumors, Cancer Res. 32 (1972) 1428±1433. S.N. Pradham, R. Prabhati, Effects of stress on growth of transplanted and 7,12-dimethylbenz(a)anthracene-induced tumors and their modi®cation by psychotropic drugs, J. Natl. Cancer Inst. 53 (1974) 1241±1245. L.S. Sklar, H. Anisman, Stress and coping factors in¯uence tumor growth, Science 205 (1979) 513±515. L.S. Sklar, H. Anisman, Social stress in¯uences tumor growth, Psychosom. Med. 42 (1980) 347±365. Z. Steplewski, W.H. Vogel, H. Ehya, C. Poropatich, J. McDonald-Smith, Effects of restraint stress on inoculated tumor growth and immune response in rats, Cancer Res. 45 (1985) 5128±5133. A. Justice, Review of the effects of stress on cancer in laboratory animals: importance of time of stress application and type of tumor, Psychol. Bull. 98 (1985) 108±138. Y. Matsukawa, H. Nishino, Y. Okuyama, T. Matsui, T. Matsumoto, S. Matsumura, Y. Shimizu, Y. Sowa, T. Sakai, Effects of quercetin and/or restraint stress on formation of aberrant crypt foci induced by azoxymethane in rat colons, Oncology 54 (1997) 118±121. R. Dechambre, C. Gosse, Individual versus group caging of mice with grafted tumors, Cancer Res. 33 (1973) 140±144. B.H. Fox, Current theory of psychogenic effects of cancer incidence and prognosis, J. Psychosoc. Oncol. 1 (1983) 17±31. B.R. Cassileth, E.J. Lusk, D.S. Miller, L.L. Brown, C. Miller, Psychosocial correlates of survival in advanced malignant disease, N. Engl. J. Med. 312 (1985) 1551±1555. D.B. Solt, E. Farber, New principle for the analysis of chemical carcinogenesis, Nature 263 (1976) 701±703. G.P. Glavin, W.P. PareÁ, T. Sandbak, H.K. Bakke, K. Murison, Restraint stress in biomedical research: an update, Neurosci. Biobehav. Rev. 18 (1994) 223±249. W.J. Kort, J.M. Weijima, Effect of chronic light-dark shift stress on the immune response of the rat, Physiol. Behav. 29 (1982) 1083±1087. J.C. Li, F. Xu, In¯uence of light-dark shifting on the immune system, tumor growth and lifespan of rats, mice and fruit ¯ies as well as the counteraction of melatonin, Biol. Signals 6 (1997) 77±89. E. Semple-Roberts, A. Hayes, D. Armstrong, R. Becker, W. Racz, E. Farber, Alternative methods of selecting rat hepatocellular nodules resistant to 2-acetylamino¯uorene, Int. J. Cancer 40 (1987) 643±645. E. Laconi, L. Tessitore, G. Milia, A. Yusuf, D.S.R. Sarma, P. Todde, P. Pani, The enhancing effect of fasting/refeeding on the growth of nodules selectable by the resistant hepatocyte model in rat liver, Carcinogenesis 16 (1995) 1865±1869. J.M. Ward, C.W. Reynolds, Large granular lymphocytic leukemia. a heterogeneous lymphocytic leukemia in F344 rats, Am. J. Pathol. 11 (1983) 1±10.

220

E. Laconi et al. / Cancer Letters 161 (2000) 215±220

[24] G.A. Tejwani, K.P. Gudehithlu, S.H. Hanissian, I.E. Gienapp, C.C. Whitacre, W.B. Malarkey, Facilitation of dimethylbenz(a)anthracene-induced rat mammary tumorigenesis by restraint stress: role of b-endorphin, prolactin and naltrexone, Carcinogenesis 12 (1991) 637±641. [25] C. Tomasi, E. Laconi, S. Laconi, M. Greco, D.S.R. Sarma, P. Pani, Effect of fasting/refeeding on the incidence of chemically-induced hepatocellular carcinoma in the rat, Carcinogenesis 20 (1999) 1979±1983. [26] G.J. Rowse, J. Weinberg, J.T. Emerman, Role of natural killer cells in psychosocial stressor-induced changes in mouse mammary tumor growth, Cancer Res. 55 (1995) 617±622.

[27] K.J. Wakikawa, M. Utsuyama, K. Hirokawa, Effect of restraint stress on immune system and experimental B16 melanoma metastasis in aged mice, Mech. Ageing Dev. 93 (1997) 107±117. [28] L.T. Harada, K. Tamada, K. Abe, K. Nomoto, Repeated restraint stress impairs the antitumor T cell response through its suppressive effect on Th1-type CD4 1 T cells, Anticancer Res. 17 (1997) 4259±4268. [29] S.C. Boyd, H.A. Sasame, M.R. Boyd, Effect of cold-restraint stress on rat gastric and hepatic glutathione: a potential determinant of response to chemical carcinogens, Physiol. Behav. 27 (1981) 377±379.