Influence of dietary fats on cell populations of line 168 mouse mammary tumors: A morphometric and ultrastructural study

cancer Letters, 35 (1987) 281-294 Elsevier Scientific Publishers Ireland Ltd.

281

INFLUENCE OF DIETARY FATS ON CELL POPULATIONS OF LINE 168 MOUSE MAMMARY TUMORS: A MORPHOMETRIC AND ULTRASTRUCTURAL STUDY

NEIL E. HUBBARD and KENT L. ERICKSON Department of Human Anatomy, CA 95616 (U.S.A.)

University

of California, School

of Medicine,

Davis,

(Received 8 October 1986) (Revised version received 4 November 1986) (Accepted 5 November 1986)

SUMMARY

The effect of dietary fat concentration and saturation on cell composition and structure of line 168 mouse mammary tumors in vivo was studied using morphometry and electron microscopy. Both the concentration and saturation of fat fed to mice had a significant influence on the volume ratio of mast cells infiltrating line 168 tumors. Tumors of mice fed diets containing a high concentration (20%) of either safflower oil (SO) or palm oil (PO) had 2-3 times the volume ratio of mast cells than mice fed diets containing a low concentration (5%) of either fat. There were no significant differences among diets with respect to other inflammatory cell populations. Mice fed either one of the high fat diets had tumor cells with inclusions that ultrastructurally, appeared to consist of lipid. Dietary fat, however, had no observable affect on cell junctions or other morphological characteristics. Greater infiltration of mast cells in tumors of mice fed high fat diets and the eventual formation of new blood capillaries may explain the decreased latency of tumor onset and enhanced growth of tumors in mice fed diets with high concentrations of fat.

INTRODUCTION

Extensive evidence from experiments during the last half-century have shown that the incidence and growth rate of mammary tumors in mice and rats may increase by feeding diets with a high fat concentration. Because of Address

COFFt?spOndf?nCe

to: Neil E. Hubbard.

0304-3835/87/$03.50 0 Elsevier Scientific Publishers Ireland Ltd. Printed and Published in Ireland

282

those observations, investigators attempted to determine which component of fat was responsible for enhanced tumorigenesis [l-3]. Since those initial observations, many investigators have used several mammary tumor models to demonstrate that rats or mice fed diets containing polyunsaturated fats (PUF) developed more tumors that appeared and grew faster than animals fed diets containing saturated fats [ 2-71. Many of the dietary PUF used in those experiments contained high levels of linoleic acid (18 : 2), an essential fatty acid. Thus, several investigators have demonstrated its effect in mammary tumorigenesis. For example, the presence of 18: 2 (number of carbons in fatty acid chain: number of double bonds), rather than the amount of fat in the diet was related to the growth of a transplantable mouse mammary adenocarcinoma [ 51. In vitro, 18 : 2 appears to be a necessary growthpromoting factor for a dimethylbenz[a]anthracene (DMBA)-induced rat mammary tumor line [ 81. The actual mechanism(s) by which dietary fat may be involved in mammary tumorigenesis is not completely understood [9]. One view held by several investigators proposes that immunosupression may be causally related to tumorigenesis and that dietary fat can markedly influence the immune response [ 10,111. Cells that have exhibited cytotoxicity toward tumor cells include macrophages [ 12,131, lymphocytes [ 141, and mast cells [ 15-171. The inflammatory cell (IC) population usually contains those cell types; the percentage, however, may change as the tumor grows. For example, transplantable mouse mammary adenocarcinomas which increased in size due to dietary linoleate became progressively more infiltrated with IC and contained progressively larger areas of necrosis with tumor growth [ 181. In that experiment, changes in histologic pattern correlated with tumor size rather than with the type of fat in the diet. In order to investigate one of the possible mechanisms by which dietary fat may be involved in modulation of mammary tumorigenesis, experiments were designed to determine the effects of fat concentration and saturation on line 168 mouse mammary tumors and associated infiltrating cells. For that, morphometric and electron microscopic methods were used. MATERIALS

Animals

AND METHODS

and diets

Seventy-two five-week-old BALB/cAnN mice were obtained from Charles River Laboratories, Inc. (Kingston, NY), immediately segregated by weight and randomly placed into five different groups in order that the mean weight per group was approximately the same. The groups were fed a stock diet (Purina Mouse Chow, Ralston Purina, St. Louis, MO) for 48 h and then switched to the different semipurified diets (Table 1). All diets contained at least the minimum level of recommended nutrients with a constant amount per kilocalorie of casein, salts, vitamins, and fiber. All groups received 1.4% of the gross energy from corn oil which provided at least the minimum

283 TABLE 1 COMPOSITION OF EXPERIMENTAL Ingredient

EFA

DIETS 5% so

20% so

5% PO

20% PO

24.8 6.4 1.7 5.5 41.6 0.7 19.3 0.0

19.8 5.0 1.3 4.4 64.5 0.6 0.0 4.4

24.8 6.4 1.7 5.5 41.6 0.7 0.0 19.3

g/100 g of diet Caseina Salt mixbJ Vitamin mixc*d Cellulosee Cereiose Corn oilf Safflower oil Palm oil

18.4 4.7 1.2 4.1 71.1 0.5 0.0 0.0

19.8 5.0 1.3 4.4 64.5 0.6 4.4 0.0

a AU diets provided 28% of energy from protein. b Provided at 1.4 g/Kcal of gross energy. c See Thomas and Erickson [ 191 for composition. d Provided at 0.3 g/Kcal of gross energy. e Provided at 1.1 g/Kcal of gross energy. f Provided at 1.4% of the gross energy as an essential fatty acid source.

requirement of essential fatty acids [20]. One group received the minimum of essential fatty acids as the only fat source (EFA). Two other groups received additional PUF, safflower oil, for a total fat content of 5% (5% SO) or 20% (20% SO), and the other two groups were fed diets supplemented with saturated and monounsaturated fat, palm oil, for a total fat content of 5% (5% PO) or 20% (20% PO). Safflower and palm oil were selected because they are comparable in carbon chain length but vary in the amount of 18: 2. Analysis of the oils used to make the diets is summarized in Table 2. Corn and safflower oils contained 61.6% and 77.6% 18: 2, respectively, while palm oil had less than 10%. Palm oil consisted mainly of saturated fatty acids (44.4%) and the monounsaturated fatty acid, oleic acid (18 : 1). From the fatty acid analysis of the oils, 18: 2 content in each of the diets was calculated. Levels were 0.3% (EFA), 0.8% (5% PO), 2.3% (20% PO), 3.8% (5% SO), and 15.4% (20% SO). Mice were fed the experimental diets ad libitum throughout the experiment and were weighed weekly to monitor growth. There were no significant differences (P > 0.05) in the growth of mice on the different experimental diets. Fatty acid analysis

A modification of the method previously described [21] was used to determine fatty acid composition of oils used for the diets and of tumor tissue. To analyze the oils, 500-c(g aliquots were suspended in 1 ml of petroleum ether, and 100 gg of margaric acid (17 : 0) was added as internal

284 TABLE 2 FATTY

ACID COMPOSITION OF DIETARY

OILS

Percent (by area) of total fatty acids as determined by gas-liquid chromatography Fatty acid

Corn oil”

Safflower oilb

Palm oilc

12:od 14:o 16:0 18:0 18:l 18:2 Otherse

0.2 9.6 2.3 25.8 61.6 0.5

0.3 6.0 3.1 12.9 77.6 0.2

0.5 0.6 38.2 5.0 45.2 9.7 0.8

%ovided by Best Foods, Inc. (Union, NJ). b Provided by California Fats and Oils (Richmond, CA). c Provided by Proctor and Gamble (Cincinnati, OH). -d Carbon chain length : number of double bonds. eIncludes16:1,18:3,and20:4.

standard. To analyze tumors, tissues were homogenized and the total lipid extracted using methanol/chloroform (2 : 1, v/v) for 90 min then methanol/ chloroform/water (2 : 1: 0.8, v/v). The tumor extract was diluted with equal volumes of chloroform and water, the lower chloroform phase withdrawn after centrifugation then diluted with benzene and brought to dryness under Nz. The residue was suspended in petroleum ether and the internal standard was added. Methyl esters of the dietary oils and tumor extract were formed by first evaporating the solvent under Nz, then adding 1.0 ml of 0.5 M NaOH in methanol by heating to 100°C in a boiling water bath for 5 min. After cooling, 1 ml of methanolic boron trifluoride was added to the mixture, heated again to 100°C for 1 min, then cooled. After the solvent was evaporated, the residues were partitioned using petroleum ether, and the organic phase removed. The ether was then evaporated and the resultant methyl esters were resuspended in 1.0 ml of 2,2,4trimethylpentane (iso-octane). Samples were analyzed in a Hewlett-Packard 5880A gas-liquid chromatograph with a 20-m dimethyl silicone capillary column. The initial temperature was 100°C and a program rate of 25.O”C/min at 200°C and 4.O”C/min to 250°C was used. Tumor

cells

Mouse cell-line 168 was derived from the first in vitro passage of a mammary tumor spontaneously arising in a BALB/cfC3H female [ 221. Latency of this cell line may be influenced by dietary fat concentration and saturation [23]. Tumor cells were maintained in culture (5% COZ, 37°C) with Eagle’s minimum essential medium with Earle’s salts supplemented with 5%

285

fetal bovine serum, 2% essential vitamin mixture, 1% L-glutamine, 1% nonessential amino acids, 1% sodium pyruvate and 50 pg/ml gentamicin. Tumor cells in their exponential growth phase were harvested with 0.05% trypsin and 0.025% EDTA. After mice were fed the experimental diets for 4 weeks, all were injected S.C.with lo5 line 168 cells. Palpable tumors began to appear 7-14 days postinjection with some not appearing for 3-4 weeks. When the tumors reached a minimum size that facilitated removal (50-100 mm3), one mouse per diet was randomly selected for tumor excision. This was done twice weekly for 5 weeks. Light and electron

microscopy

Tumors were removed from mice under light anesthesia, and the volumes of each calculated using the immersion method. Excised tumors were placed in a basket and completely immersed in a beaker filled with 0.9% saline solution. The weight, in grams, was equal to the volume of the tumor in cubic centimeters [ 241. Tumors were then immediately placed in a fixative of 2% glutaraldehyde in 0.2 M phosphate buffer at 24°C for 5 min. The tumors were then cut into slabs 1 mm thick and placed in fresh, 4°C fixative for 90 min. Tissues were rinsed in the phosphate buffer for 30 min, postfixed in 1% osmium tetroxide for 20 min, rinsed again, and dehydrated in a graded series of acetones. After infiltration and embedding in Bojax’s eponaraldite, the tissue blocks were sectioned for light and electron microscopy. Thick sections (1 pm) were stained with toluidine blue and thin sections were stained first with aqueous uranyl acetate for 8-10 min then concentrated Reynold’s lead citrate for 45 s. Thin sections were observed using a Philips-410 electron microscope at 80 kV. Fifty-four tumors were prepared for light and electron microscopy. For each tumor, approximately 25 1 mm X 2 mm pieces were processed up to the infiltration stage and 10 were randomly chosen to be embedded in plastic. Of the 10 blocks, 5-8 were randomly selected for thick sectioning. Light-microscopic observations were based on the examination of the toluidine-blue-stained thick sections of each of the 54 tumors. Of the 5-8 blocks sectioned for light microscopy, 2-3 were selected for thin-sectioning and examination with the electron microscope. Stereology

and statistical analysis

The volume ratios (v/v) of selected components of the tumor tissue were calculated using the point-count method of planimetry [24]. Thick sections of 5 blocks from each tumor were observed using a light microscope equipped with an intraocular grid. Twenty randomly-selected fields from each slide were used to estimate the relative volumes of (a) macrophages, (b) mast cells, (c) white blood cells (WBC), (d) necrosis, and (e) vasculature. The stereological data and the body weight data were analyzed using oneway analysis of variance (Microstat, Ecosoft, Inc., Indianapolis, IN).

286 RESULTS

Light microscopic observations General characteristics of smaller tumors (50-130 mm3) were the absence of necrotic tissue, sparse blood vasculature, and a cellular appearance. Most of the cells appeared to be fibroblast-like, characteristic of line 168 tumor cells [ 221. Organized cell groups were divided by blood vasculature and surrounding connective tissue. Blood vessels of small tumors appeared to be smaller in diameter compared to vessels of the larger tumors. Very little necrotic tissue could be found in these small tumors except for occasional small foci of dead cells or isolated necrotic cells. Large tumors (400-750 mm”) had abundant necrosis but the tissue appeared also to be quite cellular. Cells appeared oriented as in the smaller tumors. Necrotic tissue appeared to be in those areas furthest from any blood supply. There were no differences in tumor tissue of mice fed different diets observable at the light microscope level.

Fig. 1. Line 168 (x 16,400).

tumor

cells with lipid inclusions

(L),

from a mouse

fed a 20% SO diet

Ultrastructural observations Several morphologic features were characteristic for the tumors of all dietary groups. For example, cells possessed nuclei with a thin rim of heterochromatin (Fig. 1) and prominent nucleoli. Mitochondria, Golgi stacks, and cisternae of rough endoplasmic reticulum were also observed. Aside from these organelles, the cytoplasm of the fibroblast-like tumor cells normally possessed a scattering of free ribosomes with some aggregated as polyribosomes. There were no junctional complexes regularly found between tumor cells; however, an occasional gap junction could be observed (Fig. 2). The 168 tumor cells appeared to be connected mostly by transient cell-to-cell contacts. There were no observable differences in junctions among the dietary groups. There were no measurable differences in the ultrastructural morphologies of tumor cells from mice fed different diets except those fed a high level of fat. Tumor cells of mice fed the 20% SO and PO diets had inclusions that

Fig. 2. Two 168 tumor cells joined by a gap junction. Though this junction appears in a tumor from a mouse fed a high concentration of fat (20% PO), there were no quantitative difference among diets (~85,000).

288

Fig. 3. Line 168 tumor cells from a mouse fed a 5% SO diet with no appkent in the cytoplasm (X 15,000).

lipid inclusions

appeared to consist of lipid (Fig. 1). No lipid inclusions were observed in any tumors from mice fed the EFA, 5% SO (Fig. 3), or 5% PO diets. Lipid inclusions, however, were apparent in macrophages of all dietary groups. Inflammatory cells did not appear different qualitatively among the dietary groups. Stereological

analysis

Table 3 lists the v/v of various tissue components of small (86 f 8 mm3) and large (564 f 39 mm3) tumors from mice fed the different diets. Comparing the mean values for tissue components of small and large tumors, the v/v of macrophages and mast cells were significantly greater in small tumors and the v/v of WBC and necrosis were significantly greater in the larger tumors. When small tumors of mice fed the different diets were compared, the v/v of mast cells varied with changes in the concentration and saturation of dietary fat. Tumors of mice fed either of the high fat diets had a significant (P < 0.05) increase in the v/v of mast cells compared to the tumors of mice fed either of the two low fat diets. Also, small tumors of mice fed either of the PO diets had a significantly (P < 0.05) greater v/v of mast cells than mice fed the SO diets. In the group of large tumors, dietary fat significantly influenced the v/v of mast cells and necrosis.

EFA

OF DIETARY

2.6 + 0.97 0.29 ? 0.19 0.18 ? 0.01 0 3.1 ? 2.0

2.9 0.08 0.50 2.3 4.7

f 2 + + +

0.21 0.04 0.17 0.39 2.1

2.0 0.14 0.63 4.1 4.5

? + + + k

0.04 0.04 0.21 0.35 0.95

4.4 + 0.36’ 0.09 + 0.01 0.27 + 0.06’ 0 3.1 2 1.2

5% so

FAT ON VOLUME

2.3 0.12 0.64 1.9 4.4

20.18 2 0.01 + 0.10 to.13 + 1.8

TISSUE

2.3 0.13 0.74 2.3 4.2 + + k + +

0.22 0.01 0.04 0.27 0.20

3.3 + o.2ac 0.30 2 0.01’: 0.33 t O.llC 0 3.7 + 0.16

20% so

OF SELECTED

2.8 ? 1.3 0.21 + O.OIC 0.15 t o.07c 0 2.0 +1.2

5% PO

RATIOS

* Volume ratios (+ S.E.M.) determined by the point-count method of planimetry. b P-values for each tissue component were derived from analysis of variance. c Significantly (P < 0.05) different from same component of larger tumors.

Macrophages Mast cells WBC Necrosis Vasculature

Large tumors (564 ? 39 mm’)

Macrophages Mast cells WBC Necrosis Vasculature

Small tumors (86 + 8 mm”)’

INFLUENCE

TABLE 3

3.7 0.02 0.43 2.1 2.3

f: + + + +

0.92 0.01 0.01 0.74 0.42

4.1 + 0.52 0.49 t 0.16’ 0.21 i: 0.07’ 0 4.3 f. 0.72’

20% PO

COMPONENTS

NS <0.05 NS <0.05 NS

NS <0.05 NS NS NS

P-valueb

OF TUMORS*

2.6 0.10 0.59 2.5 4.0

f 0.15 2 0.03 ?: 0.05 ? 0.38 ? 0.35

3.4 2 0.24c 0.28 2 0.04c 0.23 k 0.03c OC 3.2 + 0.40

Mean

1.0 25.5 2.2 31.0 29.4 1.8

14 :oc 16:0 16:l 18:0 18:l 18:2 20:o 20: 3n6e 20 : 3n9r 20:4 22:o 22~6 24:0

-

? 0.8 ? 2.0 ? 0.0 + 0.9 + 1.9 2 0.3 1.8 1.1 16.0 0.8 4.0 0.8

0.4 19.3 1.9 27.4 20.6 9.5 ? ? ? ? + +

+ k + + + +

5% so

0.5 0.7 5.1 0.1 0.3d 0.1

0.1 2.6 0.5 1.3 9.6 O.gd -

? 0.2 fr 0.8 k 0.0 + 3.7 f 2.0 f 0.4

4.8 ? 1.0 6.9 + 3.0 -

1.3 28.0 0.8 29.0 25.7 3.5

5% PO 0.6 28.4 1.9 22.9 14.4 9.7 1.9 1.2 0.9 14.6 1.9 1.6 -

k + + + + k * + 2 + ? +

20% so

a Percent (by area) of total fatty acids as determined by gas-liquid chromatography. b P-values for fatty acid composition were calculated using one-way analysis of variance. c Carbon chain length : number of double bonds. d Significantly different when compared with EFA diet. e The first double bond for this fatty acid is located 6 carbons from the methyl end (n6). f First double bond for this fatty acid is located 9 carbons from the methyl end (n9).

3.7 f 1.5 5.0 + 0.8 1.0 ?: 0.2 -

EFA

Fatty acid 0.0 9.4 0.3 0.7d 1.8d O.ld 0.3 0.1 O.Od 4.4 0.0 0.0

< 0.05 -

NS NS NS < 0.05 < 0.05 < 0.05 -

2.2 t 0.3 -

0.5 0.2 O.ld 0.9 0.6 0.1 0.1

< 0.01 < 0.05 -

k + ? ? ? * 2

2 0.1

P valueb

1.6 f 0.0 16.0 2 2.6d 2.1 + 0.0

0.5 21.4 1.7 22.5 24.0 5.8 1.3 0.8

20% PO

TOTAL FATTY ACID COMPOSITION OF MOUSE MAMMARY TUMOR TISSUE FROM MICE FED DIFFERENT EXPERIMENTAL DIETSa

TABLE 4

291

Fatty acid composition

of tumors

The fatty acid composition of tumor tissue from mice fed the different diets is summarized in Table 4. Ratios of 14: 0,16 : 0, and 16 : 1 were not significantly (P > 0.05) different when tumor tissue from mice fed the four diets supplemented with SO or PO were compared with tumors of mice fed EFA only. However, there were significant (P < 0.05) differences in the fatty acid composition of tumors from mice fed the other diets (5% SO, 20% SO, and 20% PO) compared to EFA-fed mice. Mice fed 5% SO had tumors with significantly greater level of 18 : 2 and 22 : 6. Levels of 18 : 0, 18: 1, and 20: 3n9 were significantly less and 18: 2 significantly greater in tumors of 20% SO-fed animals compared with EFA-fed animals. Likewise, levels of 18: 0 were less and 20: 4 were greater in tumors from mice fed the high PO diet than animals fed EFA only. DISCUSSION

In these experiments, we have demonstrated that the concentration and saturation of dietary fat have a significant influence on the volume ratio of mast cells infiltrating line 168 tumors in BALB/c mice. Relatively small (86 f 8 mm3) tumors had greater v/v of mast cells than larger (564 rt 39 mm3) tumors and among the small tumors, mice fed 20% dietary fat had tumors with 2-3 times the infiltration of mast cells than did mice fed 5% dietary fat. Mast cells can be involved in the release of substances that aid in neovascularization. For example, investigators showed that mast cells release heparin which stimulated the migration of bovine capillary endothelial cells in vitro [ 251. Endothelial cell migration is important in angiogenesis and is essential for continued growth of solid tumors [ 251. As a working model then, mast cells, which may be chemically attracted to tumor nodules [ 261, may secrete heparin as they move toward the tumor. The heparin released could serve to stimulate and direct the migration of capillary sprouts toward or within the tumor. The greater infiltration of mast cells in smaller tumors of this study may be related to the establishment of vascularization. Larger tumors, which have established their blood supply, may not be as dependent on mast cells for angiogenesis. In this study, greater infiltration of mast cells into tumors of mice fed diets with a high concentration of fat may explain why so many investigators have reported shorter latency and enhanced growth of mammary tumors in animals fed high fat diets. Mast cells may also be involved in tumor-cell cytotoxicity [ 15,161. One observable difference in the tumor cells of mice fed high levels of fat was the presence of inclusions that appeared to consist of lipid. Although 168 tumor cells have a fibroblast-like morphology, they are not true fibroblasts. Normal fibroblasts, however, did accumulate triglycerides and store them in cytoplasmic droplets when they were exposed to an excessive amount of fatty acids, in vitro, relative to their immediate needs [27]. This

could also be the case in vivo where tissues exposed to high levels of fatty acids such as 20% dietary fat, may accumulate triglycerides and form lipid inclusions. One way in which tumor cell growth control processes may be altered involves modulation of intercellular communication [ 28,291. Gap junctions are thought to mediate cell-to-cell communication which plays an important role in the control of differentiation and growth of normal and neoplastic tissue [ 9,301. A breakdown in this normal cellular communication process may be effective in promoting tumor growth and formation. Unsaturated fatty acids appear to have an effect on cell-to-cell interactions. In this study, only a few cells with gap junctions were apparent in the tumor tissue and there were no observable differences among the diets. Other investigators showed in vitro that unsaturated fatty acids, especially 18: 1 and 18: 2, blocked metabolic cooperation between V79 Chinese hamster cells, whereas the saturated fatty acid stearic acid (18 : 0) had no effect [ 91. The content of 18: 2 in dietary fat has been a focus because the growth of tumor cells appears to be directly responsive to 18 : 2 concentration in vitro [8]. More importantly, mammary tumorigenesis has been shown to increase proportionately to the increase of dietary 18: 2 in the range of 0.5-4.4% [31]. Thus, the level of 18: 2 required to elicit a maximal tumorigenic response has been estimated to be approximately 4%. The proportion of 18: 2 in the tumor tissue appears to reflect that of the diet. In this study, we have found differences in mast cell infiltration in 168 tumors. With respect to smaller tumors, the observed increase in infiltration of mast cells into tumors of mice fed diets high in fat, could explain why rodents fed diets high in fat had shorter latency for mammary tumor appearance and enhanced growth once the tumor was established. In addition, we have found no structural changes in the tumor cells which were associated with dietary fat. Therefore, changes in mammary tumorigenesis associated with high levels of dietary fat and particularly linoleic acid are probably due other mechanisms. ACKNOWLEDGMENTS

This study was supported by grants from the National Live Stock and Meat Board and the Gustavus and Louise Pfeiffer Research Foundation. We thank Mrs. Carla McNeil1 for technical assistance, Dr. Mark Bieber of Best Foods, Mr. David Hoffsten of California Fats & Oils, and Dr. Ed Hunter of Proctor & Gamble Co. for providing the corn, safflower, and palm oils. REFERENCES 1 Gammal, E.B., Carroll, K.K., and Plunkett, E.R. (1967) Effects of dietary fat on mammary carcinogenesis by 7,12-dimethylbenz(a)anthracene in rats. Cancer Res., 27, 1737-1742.

293

9

10

11

12

13

14

15 16

17 18 19 20

Carroll, K.K., and Khor, H.T. (1971) Effects of level and type of dietary fat on incidence of mammary tumors induced in female Sprague-Dawley rats by 7,12-dimethylbenz(a)anthracene. Lipids, 6,415-420. Carroll, K.K., and Hopkins, G.J. (1979) Dietary polyunsaturated fat versus saturated fat in relation to mammary carcinogenesis. Lipids, 14, 155-158. Hopkins, G.J., West, C-E., and Hard, G.C. (1976) Effect of dietary fats on the incidence of 7,l%dimethylbenz(or)anthracene-induced tumors in rats. Lipids, 11, 328333. Rao, G.A., and Abraham, S. (1976) Enhanced growth rate of transplanted mammary adenocarcinoma induced in C3H mice by dietary linoleate. J. Natl. Cancer Inst., 56, 431-432. Hillyard, L.A., and Abraham, S. (1979) Effect of dietary polyunsaturated fatty acids on growth of mammary adenocarcinomas in mice and rata. Cancer Res., 39,44304437. Abraham, S., Faulkin, L.J., Hillyard, L.A., and Mitchell, D.J. (1984) Effect of dietary fat on tumorigenesis in the mouse mammary gland. J. Natl. Cancer Inst., 72, 14211429. Kidwell, W.R., Monaco, M.E., Wicha, M.S., and Smith, G.S. (1978) Unsaturated fatty acid requirements for growth and survival of a rat mammary tumor cell line. Cancer Res., 38,4091-4100. Welsch, C.W., and Aylsworth, C.F. (1983) Enhancement of murine mammary tumorigenesis by feeding high levels of dietary fat: a hormonal mechanism? J. Natl. Cancer Inst., 70, 215-221. Mertin, J., Stackpoole, A., and Shumway, S. (1985) Nutrition and Immunity: the immunoregulatory effect of n-6 essential fatty acids is mediated through prostaglandin E. Int. Arch. Allergy Appl. Immunol., 77, 390-395. Thomas, I.K., and Erickson, K.L. (1985) Lipid modulation of mammary tumor cell cytolysis: direct influence of dietary fats on the effector component of cell-mediated cytoxicity. J. Natl. Cancer Inst., 74, 675-680. Loveless, S.E., and Heppner, G.H. (1983) Tumor-associated macrophages of mouse mammary tumors. I. differential cytotoxicity of macrophages from metastatic and nonmetastatic tumors. J. Immunol., 131, 2074-2078. Mahoney, K.H., Fulton, A.M., and Heppner, G.H. (1983) Tumor-associated macrophages from metastatic and nonmetastatic tumors. II. Differential distribution of macrophages from metastatic and nonmetastatic tumors. J. Immunol., 131, 20792085. Rios, A.M., Miller, F.R., and Heppner, G.H. (1983) Characterization of tumorassociated lymphocytes in a series of mouse mammary tumor lines with differing biological properties. Cancer Immunol. Immunother., 15,87-91. Far-ram, E., and Nelson, D.S. (1980) Mouse mast cells as anti-tumor effector cells. Cell. Immunol., 55, 294-301. Henderson, W.R., Chi, E.Y., Jong, E.C., and Klebanoff, S.J. (1981) Mast cellmediated tumor-cell cytotoxicity: role of the peroxidase system. J. Exp. Med., 153, 520-533. Roche, W.R. (1985) Mast cells and tumors: the specific enhancement of tumor proliferation in vitro. Am. J. Pathol., 119, 57-64. Gabor, H., Hillyard, L.A., and Abraham, S. (1985) Effect of dietary fat on growth kinetics of transplantable mammary adenocarcinoma in BALB/c mice. J. Natl. Cancer Inst., 74, 1299-1305. Thomas, I.K., and Erickson, K.L. (1985) Dietary fatty acid modulation of murine T-cell responses in uiuo. J. Nutr., 115, 1528-1534. National Research Council (1978) Nutrient requirements of the mouse. In: Nutrient Requirements of Laboratory Animals, pp. 38-53. Editors: National Research Council. National Academy of Sciences, Washington, DC.

294

21 Erickson, K.L., Adams, D.L., and McNeil& C.A. (1983) Dietary lipid modulation of immune responsiveness. Lipids, 18, 468-474. 22 Dexter, D.L., Kowalski, H.M., Blazar, B.A., Fligiel, Z., Vogel, R., and Heppner, G.H. (1978) Heterogeneity of tumor cells from a single mouse mammary tumor. Cancer Res., 38, 3174-3181. 23 Erickson, K.L., and Thomas, I.K. (1985) The role of dietary fat in mammary tumorigenesis. Food Technol., 39, 69-73. 24 Elias, H., and Hyde, D.M. (1981) An elementary introduction to stereology (quantitative microscopy). Am. J. Anat., 159, 412-446. 25 Azizkhan, R.G., Azizkhan, J.C., Zetter, B.R., and Folkman, J. (1980) Mast cell heparin stimulates migration of capillary endothelial cells in vitro. J. Exp. Med., 152, 931-944. 26 Poole, T.J., and Zetter, B.R. (1983) Stimulation of rat peritoneal mast cell migration by tumor-derived peptides. Cancer Res., 43, 5857-5861. 27 King, M.E., and Spector, A:A. (1981) Lipid metabolism in cultured cells. In: Growth Requirements of Vertebrate Cells in Vitro, pp. 293-312. Editors: C. Waymouth, R.G. Ham, and P.J. Chapple. Cambridge University Press, New York. 28 Trosko, J.E., Yotti, L.P., Warren, S.T., Tsushimoto, G., and Chang, C.C. (1982) Inhibition of cell-to-cell communication by tumor promoters. In: Carcinogenesis-a Comprehensive Survey, pp. 565-585. Editors: E. Hecker, N.E. Fusenig, W. Kunz, F. Marks, and H.W. Thielmann. Raven Press, New York. 29 Aylsworth, C.F., Jones, C., Trosko, J.E., Meites, J., and Welsch, C.W. (1984) Promotion of 7,12-dimethylbenz(cu)anthracene-induced mammary tumorigenesis by high dietary fat in the rat: possible role of intercellular communication. J. Natl. Cancer Inst., 72, 637-645. 30 Freshney, R.I. (1985) Induction of differentiation in neoplastic cells. Anticancer Res., 5, 111-130. 31 Ip, C., Carter, CA., and Ip, M.M. (1985) Requirement of essential fatty acid for mammary tumorigenesis in the rat. Cancer Res., 45, 1997-2001.