Big Green Business

The growth rate of the MAC16 tumour in cachectic animals was significantly enhanced by the hypolipidemic agent bezafibrate, while the growth rate of a histologically similar tumour, the MAC13, which grows without an effect on host body compartments was unaffected. Growth of the MAC16 in vitro was unaffected by bezafibrate, suggesting that it was an in vivo phenomenon only. The stimulatory effect of bezafibrate correlated with the maximum plasma levels of free fatty acids (FFA) arising from the catabolism of adipose tissue. Accumulation of '4C-lipid from 1-'4C-triolein administered by intragastric intubation was enhanced in heart, gastrocnemius muscle and tumour of bezafibrate treated animals, while the total lipid absorption did not differ from solvent treated controls. The increased lipid accumulation in the heart, but not the tumour correlated with an increased tissue lipoprotein lipase level. The increased tumour level may arise from an increased uptake of FFA arising from a weakening of the bonds between FFA and albumin. These results suggest that growth of certain tumours is dependent on maintaining sufficient lipid levels and that the lipid mobilising effect of the tumour may be necessary to sustain tumour growth.

Although tumour cells retain the metabolic capacity for the synthesis of fatty acids (Medes et al., 1953;Mulligan & Tisdale, 1991) the flux through this pathway may be insufficient to meet the tumour's needs and a substantial amount is obtained preformed from the host (Spector, 1975). The host responds to the tumour by releasing increased amounts of lipid into the circulation and either the circulatory free fatty acids (FFA) (Mermier & Baker, 1974), or to a lesser extent triglycerides contained in plasma lipoproteins, are available for the tumour (Lyon et al., 1982). A number of experimental tumours have a high relative in vivo uptake of low density lipoprotein (LDL) (Lombardi et al., 1989). In some cases this results in frank hyperlipidemia (Mider et al., 1949), but if the lipid requirements of the tumour/host are high, then plasma FFA and triglyceride may be decreased despite an enhanced lipid mobilisation (Briddon et al., 1991).
Increases in the mobilisation of host fat stores either by an acute fast (Sauer & Dauchy, 1987a) or by acute strepozotocin-induced diabetes (Sauer & Dauchy, 1987b) lead to an increased tumour growth as measured by an increased size and by incorporation of (methyl-3H)thymidine into tumour DNA. These results indicate that the rate of tumour growth in vivo is limited by the availability of a substance(s) present in the hyperlipemic blood, most probably the polyunsaturated fatty acids linoleic and arachidonic acid (Sauer & Dauchy, 1988). This suggests that it may be possible to modulate tumour growth in vivo by altering the supply of lipids.
Certain fibric acid derivatives (clofibrate and bezafibrate) are effective in lowering plasma lipid levels, although the exact mechanism of action remains unclear (Fallon et al., 1972). One mechanism suggests an increased rate of metabolism of triglyceride-rich lipoproteins due to an increase in the activity of lipoprotein lipase by an inhibition of adenylate cyclase activity (Greene et al., 1970). The resulting decrease in cyclic AMP would be expected to reduce lipolysis and increase tissue lipoprotein lipase activity. Such an effect may modulate lipid uptake by the tumour as well as attenuating the catabolic activity of tumour lipolytic factors which act by increasing cyclic AMP levels in adipose tissue  This study determines the effect of bezafibrate on the growth and lipid metabolism of the MAC16 tumour, which induces cachexia in recipient animals, as compared with the effect on the MAC13 tumour, which has a comparable histology and growth rate, but which has no effect on host lipid stores in order to investigate the relationship between mobilisation of lipids in cachexia and growth of the tumour.

Material and methods
Pure strain NMRI mice bred in our own colony were fed a rat and mouse breeding diet (Pilsbury, Birmingham, UK) and water ad libitum. Fragments of either the MAC16 or MAC13 tumour were implanted into the flank of male NMRI mice (starting weight 24-26 g) by means of a trocar, as described (Bibby et al., 1987). Animals bearing the MAC16 tumour develop weight loss 10-12 days following tumour transplantation and were used when weight loss averaged 2 g (average tumour weight 200 mg). Animals bearing the MAC 13 tumour were used when the tumour volume was comparable with that in animals bearing the MAC16 tumour. The dimensions of the tumours were measured daily by the use of calipers. The tumour volume was calculated from the formula: length x (width)2 divided by two.
Animals bearing either type of tumour were randomised into groups of five animals which either received bezafibrate (100 mg kg-') administered daily by i.p. injection in arachis oil (100 ,Ll), or arachis oil (100 gl) alone. Body weight, food and water intake, and tumour volume were determined daily for each group.
To determine the direct effect of bezafibrate on tumour growth rate, MAC16 cells were grown in vitro in RPMI1640 tissue culture medium containing 10% foetal calf serum (Gibco Europe, Paisley, Scotland) under an atmosphere of 5% CO2 in air. Bezafibrate dissolved in ethanol was added to cells at an initial density of 4 x I04 ml-', such that the final concentration of ethanol in the culture medium was 1 %. Control samples contained ethanol alone. Cell number was enumerated after 72 h continuous incubation with the drug, using a Coulter electronic particle counter. Experiments were performed in triplicate.
Tissue lipid accumulation in the presence of bezafibrate The accumulation of an oral dose of '4C-lipid was determined using the method of Oller do Nascimento and Williamson (1986). 1-'4C-Triolein (50 pCikg-') (Amersham International, Amersham, Bucks, UK) was administered orally by gastric intubation, without anaesthesia and with minimal stress to male NMRI mice bearing the MAC16 tumour 3 days after daily administration of bezafibrate (200 mkg kg-') or arachis oil controls. After 2 h animals were anaesthetised and blood was collected by cardiac puncture. The complete gastrointestinal tract was removed and homogenised in 5 ml of 3% (w/v) HC104. Lipids were extracted from organs and blood by the method of Stansbie et al. (1976). The extracted fatty acids were dissolved in Optiphase scintillation fluid and the radioactivity determined using a Packard Tri-Carb 2000 CA liquid scintillation analyser. Triolein absorption was calculated by subtracting the total gastrointestinal tract radioactivity from that administered.
Determination of tissue lipoprotein lipase (LPL) activity Animals were killed 3 days after administration of bezafibrate (100 mg kg-') or arachis oil control and the heart and tumour were removed onto ice, homogenised in cold acetone and extracted with acetone and ether before being stored at -200C. LPL activity in the resolubilised powder was determined using 3H-triolein as substrate using fasted rat serum as the source of the activator apoprotein C-Il. The 3H-labelled fatty acids released after a 60 min incubation period were extracted and determined by the method of Nilsson-Ehle and Schotz (1976). Lipolysis was identified as LPL activity because it was inhibited by addition of 2 M NaCl to the assay medium. animals (0.83 ± 0.03 mM in bezafibrate treated animals compared with 0.73 ± 0.08 mM in arachis oil controls).
In order to investigate the mechanism of increased lipid intake in bezafibrate treated animals, the effect on lipoprotein lipase (LPL), the main enzyme involved in removing triacylglycerol from the plasma, was investigated. The results presented in Table III show a significantly elevated level of LPL in the heart but not in the tumour of bezafibrate treated animals. This suggests that bezafibrate stimulates lipid uptake into the heart by increasing tissue LPL levels, and that accumulation of lipid in the tumour must occur by some other mechanism. a _:~31 Statistical analysis All results are expressed as mean ± SEM for at least three separate determinations. Differences were evaluated statistically by Student's t-test.

Results
The effect of daily i.p. administration of bezafibrate (100mg kg-') on the growth rate of the MAC16 tumour in animals with weight loss (2-4 g) is shown in Figure la. Tumour growth rate was significantly (P<0.01) enhanced within 2 days of drug administration and the overall increase in tumour volume during the 4 day period in bezafibrate treated animals was three times that of animals administered solvent alone. There was no difference in the growth rate of the tumour in the presence and absence of the arachis oil. The mean food and water intake of animals administered bezafibrate (3.7 ± 0.7 g and 3.7 ± 0.6 ml mouse'l day-', respectively) was not significantly different from solvent alone (5.1 ± 0.6 g and 4.4 ± 0.4 ml mouse-1 day-', respectively), although animals treated with bezafibrate had an increased rate of weight loss (Figure lb), probably due to the more rapid growth of the tumour. Previous studies (Beck & Tisdale, 1987) have shown a linear correlation between weight loss and tumour burden. In contrast with the effect on the growth of the MAC16 tumour, growth of the MAC13 tumour, which does not induce cachexia, was not affected by bezafibrate using the same dose schedule (Figure lc).
Growth of the MAC16 tumour in vitro was unaffected by corlcentrations of bezafibrate up to 1 mM (Table I). This suggests that the in vivo growth stimulatory effect must arise from an indirect action, possibly through the availability of lipids to the tumour, which are not limiting for growth in vitro In order to investigate the effect of bezafibrate on the ability of the animals to deal with administered lipid 1'4Ctriolein was administered by intragastric intubation 3 days after the initiation of bezafibrate administration, and the absorption and tissue accumulation over a 2h period was compared with solvent treated controls. While the absorption of the '4C-lipid was not significantly different between the two groups (94.8 ± 3.8% for bezafibrate treated animals and 95.4 ± 0.5% for arachis oil controls) the pattern of distribution was different, with a significantly elevated accumulation in heart, gastrocnemius muscle and tumour in bezafibrate animals (Table II). This effect occurred without a significant alteration in plasma FFA levels in bezafibrate treated Time (days) Figure 1 Effect of bezafibrate (100mg kg') (0) or arachis oil (0) administered daily to male NMRI mice by i.p. injection on the tumour volume increase, a, or weight loss, b, in animals bearing the MAC16 tumour or the tumour volume increase of the MAC13 tumour, c. Animals bearing the MAC16 tumour had lost 2 to 4 g in weight before the experiment was initiated. Differences a, P <0.05; b, P <0.01; c, P <0.005; and d, P <0.001 between control and bezafibrate-treated animals were evaluated statistically using Student's t-test.

Discussion
In addition to playing a structural role in membrane architecture, lipids are also important regulatory metabolites in cell function. Thus complexes of essential fatty acids such as linoleic acid with bovine serum albumin have been shown to stimulate the growth of human breast carcinoma cells in culture (Rose & Connolly, 1990) and to significantly enhance the growth of transplantable mammary adenocarcinomas in mice (Abraham & Hillyard, 1983). Lipoxygenase metabolites of linoleic acid may be an important element in the epidermal growth factor (EGF)-regulated cascade of biochemical events leading to fibroblast mitogenesis (Glasgow & Eling, 1990), while phospholipids containing arachidonic and linoleic acids can inhibit a guanosine triphosphatase (GTPase) activating protein (GAP) which stimulates the rate at which the 21 kDa gene product of the Harvey ras proto-oncogene converts bound GTP to GDP, and thus inactivates p21 (Tsai et al., 1989). In addition ras proteins must be isoprenylated at a conserved cysteine residue near the carboxyl terminus in order to bind to the inner surface of the plasma membrane and exert their biological activity (Manne et al., 1990). Thus lipids play an essential role in tumour development and the loss of body lipids accompanying cancer cachexia may be an essential prerequisite for. the growth of some tumours. This suggests that it may be possible to modify tumour growth rate by the regulation of the supply of lipids to the tumour. Inhibition of the process of cachexia, with consequent retention of adipose tissue triglycerides by both a ketogenic diet  and by the polyun-saturated fatty acid eicosapentaenoic acid (EPA)  also results in an inhibition of the growth rate of the MAC16 tumour. In the present study we have shown that the lipid lowering agent bezafibrate, which increases peripheral uptake of lipids is also capable of stimulating the growth of the MAC16 tumour, while having no effect on the growth of the MAC13 tumour, which does not induce cachexia in the host. In vitro studies suggest that the stimulatory effect of bezafibrate on tumour growth is not due to a direct effect of the drug on cell growth and that it only occurs in the in vivo situation.
The pattern of stimulation of tumour growth by bezafibrate closely follows the serum FFA levels in cachectic animals with maximum stimulation occurring when the animals had lost between 8 and 16% of body weight which coincided with the peak level of serum FFA in these animals (Briddon et al., 1991). However, bezafibrate did not significantly change the plasma FFA levels in tumour-bearing animals. Bezafibrate stimulated lipid accumulation into heart and gastrocnemius muscle in addition to tumour after an oral dose of 1-'4Ctriolein. The increased lipid in the heart most likely arose from an increased accumulation of plasma 18 triglycerol due to an increased tissue level of LPL. The level of LPL in tumour was low and was not increased by bezafibrate. Thus the increased tumour level probably arose from an increased uptake of circulatory FFA as previously suggested (Mermier & Baker, 1974), since fibrate drugs have previously been shown to stimulate uptake of FFA into Ehrlich ascites cells in incubation medium containing albumin (Spector & Soboroff, 1971). The increase in uptake appears to be due to a weakening of the bonds between the FFA and albumin.
The results are consistent with the hypothesis that certain tumours in vivo are highly dependent on exogenous lipid for their growth and that mobilisation of host fat stores are an essential factor in this process. If the nature of the lipid product and its role in growth are determined, it may be possible to directly attack the growth of slow-growing solid tumours through this pathway.