The parity-related protection against breast cancer is compromised by cigarette smoke during rat pregnancy: observations on tumorigenesis and immunological defenses of the neonate.

The parity-related protection against breast cancer is compromised

by cigarette smoke during rat pregnancy: observations on tumorigenesis and

immunological defenses of the neonate

Bernard G.Steinetz



_{, Terry Gordon, Salamia Lasano,}

Lori Horton, Sheung Pui Ng, Judith T.Zelikoff,

Arthur Nadas and Maarten C.Bosland

Department of Environmental Medicine, Department of Statistics and Department of Urology, New York University School of Medicine, Tuxedo, NY and New York, NY

_{To whom correspondence should be addressed. Tel:+1 845 731 3517;}

Fax:+1 845 351 4510;

Email: Steinetz@env.med.nyu.edu

Early pregnancy is a powerful negative risk factor for

breast cancer (BCa) in women. Pregnancy also protects

rats against induction of BCa by carcinogens such as

N-methyl-N-nitrosourea (MNU), making the parous rat a

useful model for studying this phenomenon. Smoking

during early pregnancy may lead to an increased risk of

BCa in later life, possibly attributable to carcinogens in

cigarette smoke (CS), or to reversal of the parity-related

protection against BCa. To investigate these possibilities,

50-day-old timed first-pregnancy rats were exposed to

standardized mainstream CS (particle concentration

¼

50 mg/m

3

) or to filtered air (FA) 4 h/day, Day 2–20 of

gestation. Age-matched virgin rats were similarly exposed

to CS or FA. At age 100 days, the CS or FA-exposed, parous

and virgin rats were injected s.c. with MNU (50 mg/kg

body wt), or with MNU vehicle. Mammary tumors

(MTs) first appeared in virgin rats 9 weeks post-MNU

injection. While no MTs were detected in FA-exposed

par-ous rats until 18 weeks post-MNU, MTs appeared in the

CS-exposed parous rats as early as 10 wks (P

< 0.02). As no

MTs developed in CS-exposed rats not injected with MNU,

CS did not act as a direct mammary carcinogen. Serum

prolactin concentration on Day 19 of pregnancy in

CS-exposed dams was reduced by 50% compared with

FA-exposed dams (P

< 0.005). CS exposure during a

preg-nancy may thus ‘deprotect’ rats, enhancing their

vulner-ability to MNU-induced BCa. Prenatal CS exposure had no

detectable effect on the immune responses of the pups

examined at 3, 8 or 19 weeks of age. However, prolactin

concentration in stomach contents (milk) of 3-day-old pups

suckled by CS-exposed dams was decreased when

com-pared with that of FA-exposed dams (P

< 0.032). As

milk-borne prolactin modulates development of the central

nervous and immune systems of neonatal rats, CS exposure

of the dams could adversely affect later maturation of these

systems by reducing milk prolactin.

Introduction

Breast cancer (BCa) is a leading cause of cancer deaths in

women (1). Risks include geographic area of residence, diet

and lifestyle, genetic influences, age at menarche and

meno-pause, and both endogenous and environmental estrogens (c.f.

ref. 1 for review). However, in most cases there are no obvious

causes of BCa (1), supporting the view that a variety of

envir-onmental carcinogens play a major role in initiation of the

disease (2). There is one clearly negative risk factor for BCa:

the incidence of breast tumors in women who carry a

preg-nancy to term prior to age 20 is only about one-third to one-half

that of nulliparous women or women who have children in

their later years (1,3,4). Thus, early pregnancy and childbirth

may confer a lifelong refractoriness to BCa, but the protective

action of a successful pregnancy is lost after the early

repro-ductive years. The mechanism(s) responsible for these

phe-nomena remain a mystery, although they have been widely

studied in both women and laboratory animals (1,3,4).

The powerful protective effects of early pregnancy may be

compromised by environmental factors as suggested by the

observation in a case–control study that smoking during a

first pregnancy was associated with a 3–4-fold increase in

BCa risk (5). Reports of effects of cigarette smoking on

mam-mary cancer incidence in women are inconsistent. Numerous

studies have failed to find a strong association between

smok-ing and BCa in women with no familial history of the disease.

Although cigarette smoking may increase BCa incidence in

women with high familial risk (6), smoking may actually be a

strong negative risk factor in BRCA1 and BRCA2 carriers (7).

These findings present a challenge to researchers. Thus,

cigar-ette smoking could either act directly to initiate tumorigenesis

in the young, undifferentiated mammary gland, or perhaps act

indirectly to prevent or reverse the factors involved in the

protective action of parity against BCa. Elucidation of the

mechanisms whereby cigarette smoking ‘deprotects’

primipar-ous women could provide new and important clues to the

understanding of the nature of the protective action of parity

against BCa and might identify a previously unrecognized

property of tobacco smoke.

To our knowledge, the effect of cigarette smoking during a

first pregnancy on parity-induced mammary cancer resistance

has not been studied previously in animal models.

N-methyl-N-nitrosourea (MNU) is a DNA-methylating agent that induces

estrogen-dependent mammary carcinomas in a high proportion

of Sprague–Dawley rats in 2–6 months (4,8–10). A prior

preg-nancy or even a pregpreg-nancy occurring soon after MNU injection

protects Sprague–Dawley rats against mammary tumor

development (8–10). The latter observation suggests that

pre-neoplastic cells present prior to pregnancy are either

des-troyed or altered during pregnancy by hormone-induced

dif-ferentiation of the glands (11). Thus, pregnancy may somehow

prevent not only initiation of BCa but also tumor progression.

Because the Sprague–Dawley rat is known to be sensitive to

Abbreviations:BCa, breast cancer; CS, cigarette smoke; FA, filtered air; MNU, N-methyl-N-nitrosourea; MTs, mammary tumors; NK, natural killer; PCBS, phosphate-citrate-buffered saline; PGE2, prostaglandin E2; RIA, radioimmunoassays.

many of the carcinogens (e.g., benzo[a]pyrene, nitromethane

and benzene) present in tobacco smoke (12), it offers a suitable

animal model for examining the effects of cigarette smoking

on the protective action of parity against MNU-induced

mammary cancer.

It has been well documented in women and in rodent models

that cigarette smoking during pregnancy can exert harmful

effects on the unborn (c.f. ref. 13 for review). Some of

these untoward effects relate to suppression of the immune

system, which in turn can lead to an increased sensitivity to

toxic,

immunomodulatory

and

carcinogenic

substances

(14–16). Accordingly, in the present study, several parameters

of the immune systems of the pups were measured at various

time intervals after birth following in utero exposure to

cigar-ette smoke (CS) or filtered air (FA). It was of further interest to

investigate the possible effects of CS exposure on the

trans-mission of milk-borne hormonal growth factors from dam to

pup during lactation, as some of these are known to modulate

development of the immune system in rats (17,18).

The present investigation was undertaken specifically to test

the hypotheses that cigarette smoking negates the protective

action of parity against mammary cancer and may also have

deleterious effects on the immune systems of the neonates.

Materials and methods

These experiments were approved by the NYU School of Medicine Institu-tional Animal Care and Use Committee (IACUC).

Experimental protocol and animals

Seventy timed pregnant (mated when 50 ± 2 days old) and 70 age-matched virgin female Sprague–Dawley rats were purchased from Charles River Breed-ing Laboratories (WilmBreed-ington, MA) and shipped to the NYU Institute of Environmental Medicine on the second (±1) day of gestation for the mated rats. The rats were individually housed on corncob bedding in solid bottom plastic cages and fed Purina Rat Chow and water ad libitum. The light–dark schedule was as follows: lights on, 6 a.m.–8 p.m.; lights out, 8 p.m.–6 a.m. The rats were then divided into the experimental groups indicated in Table I: Virgin rats (Groups 1 and 2 and Control groups C1 and C2)

Groups 1 (25 rats) and C1 (10 rats) were placed in a whole-body exposure chamber and exposed only to FA for 4 h/day for 19 days (age 52–70 days). Groups 2 (25 rats) and C2 (10 rats) were placed in a similar chamber and exposed to mainstream CS for 4 h/day for 19 days (ages 52–70 days). Groups 1 and 2 were then injected i.v. with 50 mg MNU in 6.67 ml phosphate-citrate-buffered saline (PCBS)/kg body wt on Day 100 (±2 days), whereas Control Groups C1 and C2 received only the PCBS vehicle at age 100 ± 2 days. Pregnant rats (Groups 3 and 4 and Control groups C3 and C4)

Groups 3 (25 rats) and C3 (10 rats) were exposed only to FA for 4 h/day for 19 days (Day 2–20 of pregnancy). Groups 4 (25 rats) and C4 (10 rats) were exposed to CS for 4 h/day for 19 days (Day 2–20 of pregnancy). Groups 3

and 4 were then injected i.v. with 50 mg MNU in 6.67 ml PCBS/kg body wt on Day 100 ( ± 2 days) as described above, whereas Control Groups C3 and C4 received only the PCBS vehicle on the same schedule.

In each of Groups 3, 4, C3 and C4, 1.5 ml blood samples were drawn from the tail vein under restraint on Day 8, 12 and 19 of gestation using an i.v. cannula, and the serum was harvested and frozen for hormone assays.

At parturition, the size of each litter and birth weight of each pup was recorded. Three pups from each litter were euthanized (by CO2inhalation) on

Day 2 or 3 of lactation and the stomach contents (milk) were removed and frozen for hormone assays. The remaining pups were left with the dams for the remainder of the lactation period and then utilized as described in Table II. Smoke generation and exposure of rats

Mainstream CS was generated with a cigarette-smoking machine (CH Technologies) at the NYU inhalation Core facility. The CS machine is housed in a 2 m3generation box operated under slightly negative pressure (0.1 inches of water). Tobacco and Health Research Institute’s 1R1 non-filtered cigarettes were used to generate the exposure atmosphere. The reference cigarettes were stored as suggested at 4
_{C and acclimated to room temperature and60%}

room humidity prior to burning. The number of cigarettes smoked at a time and the chamber flow rates were adjusted to achieve a target concentration of 50 (±10) mg/m3. Particle concentration was monitored continuously with a real-time aerosol monitor (RAM-1, MIE) and by taking gravimetric filter samples on an hourly basis. The smoke exposure of the rats (50 mg particulate matter/m3/4 h/day/200 g rat) was approximately equivalent to an individual smoking 2.7 packs of cigarettes per day.

Rat Groups 2, 4, C2 and C4 were exposed to smoke for 4 h/day for Days 2–20 of pregnancy [either during pregnancy (Groups 4 and C4) or correspond-ing virgin ages 52–70 days (Groups 2 and C2)]. Groups 1, 3, C1 and C3 were exposed only to FA as a control for effects of the daily confinement. Injections of the carcinogen, MNU or saline vehicle

When the rats (virgin and parous) were 100 ± 2 days old and in proestrus/estrus (the time of maximum sensitivity of breast cells to MNU), each was sedated with a s.c. injection of Ketamine (Aveco, Fort Dodge, IO) and Xylazine

Table I. Treatment groups Group No. of rates Reproductive state Treatmenta during pregnancy state or Days 52–72 Treatmentb

1 25 Virgin Expose to FA MNU 2 25 Virgin Expose to CS MNU 3 25 Pregnant Expose to FA MNU 4 25 Pregnant Expose to CS MNU

C1 10 Virgin Expose to FA PCBS (MNU vehicle) C2 10 Virgin Expose to CS PCBS (MNU vehicle) C3 10 Pregnant Expose to FA PCBS (MNU vehicle) C4 10 Pregnant Expose to CS PCBS (MNU vehicle)

MNU¼ 50 mg/6.67 ml PCBS/kg body wt i.v.; PCBS ¼ 6.67 ml/kg body wt i.v. Please see text for details of PCBS buffer and MNU preparation.

a

Please see text for details of treatment schedule.

b_{Treatments at 100 days of age.}

Table II. Experiments on immune system of pups

Age (weeks) Event/experiment Controls (FA-exposed) CS-exposed Total Males Females Males Females

0 Born

Euthanize one pup per litter on Day 2–3 of lactation Remove stomach contents (milk) for relaxin and

prolactin RIAs

9 9 9 9 36

3 NK cell assay 4 4 4 4

Lymphocyte proliferation and cytokine assay 4 4 4 4 32

8 NK cell assay 4 4 4 4

Lymphocyte proliferation and cytokine assay 4 4 4 4 32

19 NK cell assay 4 4 4 4

(Mobay Shawnee, KS) in a mixture of 10 mg Ketamine/3 mg Xylazine/kg body wt. The sedated rats then received a single injection of either MNU (NCI Chemical Carcinogen Reference Standard Repository, Midwest Research Insti-tute, Kansas City, MO) at a dose of 50 mg/6.67 ml/kg body wt in 0.1 M phosphate-citrate buffer (pH 4.8), diluted 1 : 14 with physiological saline (PCBS) (Groups 1–4) or 6.67 ml/kg body wt PCBS vehicle alone (Groups C1–C4) in the tail vein.

Mammary biopsies

Using Ketamine–Xylazine anesthesia (described above) a biopsy of a cervical mammary gland was taken from half of the rats in each group on Day 140 (40 ± 2 days after MNU injection) and fixed in 10% neutral buffered formalin (NBF). These samples were to be utilized for possible detection of CS and/or MNU-induced changes in estrogen and prostaglandin E2 (PGE2) receptors and p53 expression.

Histology and immunohistochemistry

At necropsy, tissues were fixed in 10% NBF embedded in paraffin, sectioned at 4mm, and stained with H&E unless otherwise specified. Tumors were classified according to criteria established by the 1987 Hanover, Germany Consensus Conference on rat and human breast neoplasms as described by Russo et al. (4) Visualization of estrogen receptors, PGE2 receptors and p53 expression in the mammary biopsies was carried out by immunohistochemistry.

Hormone assays

Serum samples collected during exposure of pregnant rats to CS or FA were assayed for estradiol-17b, progesterone, insulin, prolactin, relaxin and corticos-terone to determine whether CS exposure had affected the concentrations of hormones known to play a role in mammary gland development. Because of the limited amount of serum available from each pregnant rat, the serum samples were diluted 1 : 5 with assay buffer to permit duplicate determinations. Specific radioimmunoassays (RIAs) were used to determine each of the following hormones:

(i) Serum estradiol-17b: specific RIA using Kit no. 07-138015 supplied by ICN, Costa Mesa, CA.

(ii) Serum progesterone: specific RIA using Kit no. 07-17015 supplied by ICN.

(iii) Serum corticosterone: specific RIA using Kit no. 07-12013 supplied by ICN.

(iv) Serum insulin concentrations were determined using a specific rat insulin RIA kit provided by LINCO, St Charles, MO.

(v) Serum prolactin was assayed using a specific rat prolactin RIA kit pro-vided by Albert F. Parlow, Scientific Director of the NIDDK National Hormone and Peptide Program.

(vi) Serum relaxin was measured by a specific rat relaxin RIA (19), the reagents for which were provided by the innovator, Dr O.D. Sherwood.

Experimental protocol for neonatal rat pups

Following parturition, the pups were utilized according to the experimental design outlined in Table II:

RIA of hormones in milk obtained from the stomachs of suckling pups The stomach contents of each pup were diluted 1 : 1 with HOH and centrifuged to sediment the solids. Rat prolactin and rat relaxin in the aqueous phase of the pup stomach milk extracts were assayed using the specific RIAs described for serum hormones under the section Hormone assays. Dilution curves of the immunoactive relaxin and prolactin in milk were parallel to the rat standards in the respective RIAs.

Natural killer (NK) cell activity

NK cell activity was determined using a modification of the protocol initially described by Djeu (20). Briefly, recovered splenocytes were resuspended in RPMI 1640 media (40 ml) and incubated for 1 h at 37
_{C. Non-attached}

cells were recovered and resuspended in RPMI to a final concentration of 2· 107

viable cells/ml. At the same time, 5· 106

cultured YAC1 mouse lymphoma cells (ATCC, Manassas, VA) were incubated (at 37
_{C for 1 h) with}

200mCi 51

Cr and then diluted (in RPMI) to a final concentration of 1· 105_{viable cells/ml. Splenocytes and} 51_{Cr-labeled YAC1 cells (each at 0.1}

ml) were then incubated (at 37
C in 5% CO2) for 4 h in individual wells

of a round-bottom microtiter plate and cytotoxicity calculated as the % cyto-toxicity¼ [(ER-SR)/(TR-SR)] · 100, where ER, SR and TR represent experi-mental release, spontaneous release and total releasable counts, respectively. Lymphocyte proliferation assay

Single cell suspensions of splenic lymphocytes were resuspended to 2· 106 _{cells/ml in RPMI 1640 (containing 5% FBS, 1% non-essential amino}

acids, 1% sodium pyruvate solution, 0.2% gentamycin, 1% penicillin-streptomycin and 0.1%L-glutamine). Aliquots (100ml) of cells were added

in triplicate to each well of a 96-welled microtiter plate along with an additional 100ml of medium, Con A (concanavalin A, 20 mg/ml) or LPS (lipopolysaccharide, 20 mg/ml) to permit assessment of spontaneous and mitogen-induced proliferative activity. The plate was then incubated for 48 h (at 37
C with 5% CO2) before 20ml of MTT

(3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide) was added to each well. Following another 4 h incubation, 50ml of SDS (sodium dodecyl sulfate) was added to each well. The absorbance in all wells was determined at 600 nm on a microtiter plate-reader (Bio-Tek Instruments, Winooski, VT) following over-night incubation (at 37
_{C with 5% CO}

2). The stimulation ratio was calculated

on the basis of the absorbance measured in response to a mitogenic stimulus relative to that in wells receiving medium alone.

Cytokine assays

Concentrations of interleukin-1b (IL-1b) and tumor necrosis factor-a (TNF-a) in pup serum samples were determined by enzyme-linked immunosorbent assays (ELISAs) using commercially available kits purchased from R&D Systems, Minneapolis, MN.

Statistical analyses

The data are presented in tables or graphs as mean values ± standard errors or ± standard deviations as specified in the legends. Differences in continuous variables were statistically evaluated after transformation of the data to their natural logarithms (ln) using the two-tailed Student’s t-Test, or when standard deviations were statistically different, using the alternate Welch t-Test or Mann–Whitney Test. Differences in mammary tumor incidence in response to MNU between CS- and FA-treatment groups were statistically evaluated by modeling the time-to-first-tumor distribution in CS-exposed animals as a neg-ative exponential distribution. It was then shown that according to this fitted distribution, the FA-exposed animals time-to-first-tumor is improbably (P< 0.02) large. A ‘P-value’ of<0.05 was considered statistically significant.

Results

No significant differences were observed in body weight gain

of dams exposed to CS as compared with FA dams during

pregnancy (Table III). Litter size and neonatal pup weights

of CS- and FA-exposed dams did not differ significantly

(Table III).

There were no significant differences between groups in the

concentrations of estradiol-17b, progesterone, corticosterone,

growth hormone, relaxin or insulin measured in the serum

samples obtained during pregnancy (Table IV). However,

there was a highly significant (P

< 0.005) decrease in serum

immunoactive prolactin concentration on Day 19 of pregnancy

in the rats exposed to CS as compared with FA controls,

whereas on Day 8 serum prolactin levels were slightly higher

in FA controls than in CS-exposed dams (Figure 1).

Following injection of MNU at 100 days of age, the rats

were observed five times per week for the appearance of

mam-mary (and other) tumors, and the pattern that emerged is shown

in Figure 2. At 22 weeks post-MNU, 77% of the CS-exposed

AMV rats exhibited mammary tumors (MTs), whereas the

incidence for the FA-exposed controls was only 62%

(Figure 2 and Table V).

Tumors were first detected 10 weeks after MNU injection

in the CS-exposed animals, but not until 18 weeks in the

FA-exposed dams (Figure 2). This difference in time to first

tumor appearance between FA- and CS-exposed parous rats

was significant (P

¼ 0.02). At 22 weeks after MNU injection,

23% of the CS-exposed parous rats had palpable MTs,

com-pared with 8% of the FA-exposed controls. There were no

significant differences between CS- and FA-exposed virgin

rats regarding total numbers of tumors per group or per rat

at 22 weeks post-MNU (Table V).

Histochemical studies of the mammary biopsies taken

40 days after injection of MNU did not reveal any differences

between CS- and FA-exposed parous or virgin rats for

within-tissue distribution and number of positively staining cells, or

for cellular staining intensity of estrogen receptor-alpha, PGE2

receptor and p53 expression in epithelial cells of ducts and

alveoli of mammary tissue (data not shown). The biopsies were

not adequate to determine histological difference in branching

patterns between the four groups of animals.

The MTs observed at necropsy at age 240 days were

pre-dominantly invasive compact tubular carcinomas (non-invasive

compact tubular carcinomas, cribriform-comedocarcinomas

and tubulopapillary carcinomas were also observed, but were

less frequent). The spectrum of specific tumor types and the

histological appearance of the various tumor types were similar

in the mammary glands of virgins and in dams exposed to CS

or FA, and there were no differences in cervical, thoracic,

abdominal and inguinal location of these tumors (data not

shown).

The experiments performed on the pups are summarized in

Table VI and Figure 3. Body weight and relative weights of the

thymus and spleen in pups whose mothers had been exposed to

CS or FA during pregnancy were comparable at all age groups

studied (3, 8 and 19 weeks of age; Figure 3).

Although age- and sex-related variations in immune

function parameters were observed in the prenatally exposed

offspring, no differences were detected that could be ascribed

to CS exposure. The parameters examined included NK

cytotoxicity, lymphocyte proliferation and serum

concentra-tions of IL-1b and TNF-a in samples obtained from pups born

to CS- or FA-exposed dams at 3, 8 or 19 weeks post-partum

(Table VI). However, a significant decrease (P

¼ 0.032) was

observed in the concentration of immunoactive prolactin in the

milk of CS- versus FA-exposed dams (Figure 4) that was

recovered from the stomachs of pups euthanized on Day

2–3 of lactation. No differences in relaxin immunoactivity

were detected in the same samples (data not shown).

Discussion

Exposure of female rats to CS (50 mg PM/m

3

for 4 h/day for

20 days) during a first pregnancy partially reversed the

pro-tective action of parity against MNU-induced BCa by

short-ening tumor latency. Exposure to CS also tended to increase

the tumor response in virgin rats, which were considerably

more sensitive to MNU than parous rats. These results

corrob-orate the observation in the human case–control study by Innes

Table III. Body weights and litter size and weight in rats exposed to CS or FA during gestation

Group Treatment (4 h/day) No. of rats Body wt (g ± SD) Parturition No. of pups Pup body wt (g) Day 12 Day 19

50-day-old pregnant rats FA 35 247 ± 18 308 ± 26 266 ± 22 11.9 ± 1.6 6.9 ± 0.4 50-day-old pregnant rats CS 33 239 ± 19 299 ± 21 258 ± 16 12.1 ± 1.8 7.0 ± 0.4 Age-matched virgin rats FA 35 213 ± 11 232 ± 12 NA NA NA Age-matched virgin rats CS 33 211 ± 15 230 ± 17 NA NA NA

Table IV. Serum hormones in rats exposed to CS or FA from Days 2–20 of pregnancy

Serum hormones (±SEM) FA-exposed (4 h/day) CS-exposed (4 h/day) Day of pregnancy Day of pregnancy

8 12 19 8 12 19 Estradiol-17b (pg/ml) 212 ± 26 162 ± 11 576 ± 128 313 ± 99 196 ± 28 774 ± 196 N 17 17 17 12 16 16 Progesterone (ng/ml) 35 ± 2.7 47 ± 3.3 64 ± 3.0 38 ± 3.6 53 ± 4.0 62 ± 4.6 N 19 19 19 19 17 17 Corticosterone (ng/ml) 938 ± 125 1093 ± 274 871 ± 97 1199 ± 115 999 ± 105 760 ± 103 N 19 19 19 19 17 17 Growth hormone (ng/ml) 26 ± 5.4 21 ± 3.4 69 ± 5.6 15 ± 1.5 31 ± 4.7 81 ± 8.2 N 19 19 19 19 17 17 Relaxin (ng/ml) ND 340 ± 34 768 ± 47 ND 335 ± 10 741 ± 24 N 19 19 16 17 Insulin (ng/ml) 0.26 ± 0.044 0.42 ± 0.096 0.52 ± 0.096 0.27 ± 0.083 0.23 ± 0.048 0.33 ± 0.084 N 22 19 18 21 13 15

8

12

19

0

200

400

600

800

FA-exposed (N = 19)

CS-exposed (N = 17)

Day of Pregnancy

Serum Prolactin ng/ml

*p = 0.005

Fig. 1. Immunoactive prolactin concentration in serum of CS- or FA-exposed rats on Days 8, 12 and 19 of pregnancy.

and Byers (5), supporting the notion that mainstream CS

negatively impacts the protection of parity against BCa.

Pre-sent findings also suggest that CS may enhance—possibly

acting as a weak tumor promoter—the BCa-inducing action

of MNU in the virgin rats, although the difference in tumor

incidence or latency failed to reach statistical significance. In

the absence of MNU, CS-exposed rats did not develop

MTs, indicating that CS, at the concentration used for these

studies, was not by itself carcinogenic.

In this study, there were no significant effects of smoke

exposure on serum concentrations of immunoactive

estra-diol-17b, progesterone, corticosterone, growth hormone,

insulin or relaxin in the pregnant rats (compared with the

FA-exposed controls). However, there was a highly significant

(P

< 0.005) decrease in serum immunoactive prolactin on

Day 19 of pregnancy in the CS-exposed animals. As prolactin

is a likely factor in BCa susceptibility (21), this finding

requires further study. Whether this decrease represents an

actual change in secretion, or is due to proteolytic cleavage

of the prolactin molecule (22) or to increases in receptor

binding requires clarification. It is generally assumed that

increases in prolactin secretion are associated with increased

susceptibility to BCa in both female rodents and women (21).

Thus, the decrease in serum prolactin appears contrary to the

observed increase in mammary tumor response, suggesting

that factors other than hormones are part of the mechanism

that underlies the increased mammary gland sensitivity to

later injection of MNU. However, other factors may modulate

the actions of prolactin on the mammary glands. Evidence is

accumulating that supports the view that local production

(autocrine) of prolactin by mammary tissue may be more

important than its hypophyseal secretion (endocrine) in breast

tumor growth (21,23,24). There is a family of isoforms of

the prolactin receptor that mediate the effects of prolactin in

human (24) and rat (25) mammary tissue. Importantly, these

prolactin isoforms are located in the stroma as well as the

30

20

10

0

0

20

40

60

80

100

Weeks after MNU Injection

Percentage of Rats with Mammary Tumors

Grp 4 (22) Primiparous-CS Grp 1 (24) Virgin-FA Grp 2 (22) Virgin-CS Grp 3 (25) Primiparous-FA

Fig. 2.Mammary tumor progression following MNU injection in primiparous and virgin rats previously exposed to CS or FA.

Sprague–Dawley rats were bred at 50 days of age and exposed to either CS or FA in a chamber 4 h/day Days 2–20 of pregnancy. Age-matched virgin rats were exposed to CS or FA according to the same schedule. At Day 100 the rats were injected with MNU or vehicle. The graph illustrates the appearance of MTs in the MNU-injected groups (the vehicle-injected controls did not develop tumors).

Table VI. Lack of effects of prenatal CS exposure on selected immune parameters of rat pupsa

End-pointb Age (week) Male Female

FA-exposed CS-exposed FA-exposed CS-exposed T-lymphocyte proliferationc _{3 2.77 (0.61) 2.82 (1.08) 5.45 (0.76) 3.42 (1.62)} 8 4.25 (1.24) 3.96 (0.10) 3.69 (0.46) 4.14 (0.73) 19 3.53 (0.99) 2.66 (1.39) 4.39 (1.95) 4.48 (1.33) B-lymphocyte proliferationc _{3 3.17 (0.90) 2.52 (0.10) 4.34 (0.61) 3.59 (0.53)} 8 1.86 (0.27) 2.19 (0.77) 2.71 (0.48) 3.20 (0.89) 19 1.57 (0.30) 1.81 (0.64) 2.18 (0.31) 2.45 (0.50) NK cell activity (% tumor cell lysis) 3 37.53 (9.65) 42.19 (1.72) 36.23 (5.64) 39.70 (0.39) 8 33.40 (7.90) 33.15 (6.05) 35.73 (9.33) 37.39 (3.21) 19 6.77 (3.73) 9.82 (5.40) 11.98 (1.82) 10.76 (3.96)

a

Although there were changes in immune responses related to age and sex, no significant effects could be attributed to CS exposure.

b_{Values represent the mean ± SD (n¼ 3–4 rats/sex/exposure regime/age group).} c

Stimulation index¼ mitogen-stimulated proliferation/unstimulated proliferation.

Table V. Mammary gland tumorigenesis in virgin and primiparous rats injected with MNU Group No. of

rats

Parity and treatment (Days 2–20 of gestation)

MNU No. of rats with tumor count of Total with tumors (%) No. of tumors per group Mean number of tumors per rat ± SD 1 2 3 4 5 6–10

1 24 Virgin (FA for 4 h/day) yes 7 4 2 0 0 2 15 (62) 38 1.58 ± 0.48 2 22 Virgin (CS for 4 h/day) yes 7 3 3 3 1 0 17 (77) 39 1.82 ± 1.5 3 25 Primiparous (FA for 4 h/day) yes 2 0 0 0 0 0 2 (8) 2 0.08 ± 0.28 4 22 Primiparous (CS for 4 h/day) yes 4 1 0 0 0 0 5 (23) 6 0.27 ± 0.55

Only one tumor appeared in the control groups C1–C4 not injected with MNU (Rat C1–2 in PM2. Data not shown)

epithelium of the glands (26); their distribution and occupancy

in this rat model requires determination. Such studies may

provide an explanation for the apparently conflicting effects

of CS on serum prolactin and mammary cancer induction by

MNU. It is also possible that differences in serum prolactin at

other time points during pregnancy than were measured in this

study are more important than the difference at Day 19 that was

observed. This possibility is supported by the observation

of non-significantly higher serum prolactin levels on Day 8

of pregnancy in CS-exposed dams compared with air controls,

which is in line with the hypothesis that prolactin confers

increased BCa risk in rats.

The histochemical studies evaluating p53 expression, as

well as estrogen and PGE2 receptors in the mammary biopsies

obtained from the CS-exposed and control rats 40 days

post-MNU injection (prior to visible tumor appearance) failed to

reveal any notable differences between the groups. Whether

selection of different time periods for biopsy would have

detected CS- or MNU-related effects on these parameters is

unknown. The possible induction of p53 mutations by MNU

also warrants investigation (27,28).

It has been well documented in women and in rodent models

that cigarette smoking during pregnancy can exert harmful

effects on the unborn (13). Some of these untoward effects

relate to suppression of the immune system, which in turn can

lead to an increased sensitivity to toxic, immunomodulatory

and carcinogenic substances (14–16). Thus, effects of CS

exposure during pregnancy on the transmission of

milk-borne hormonal growth factors from dam to pup during

lactation and on some indicators of immune function in the

prenatally exposed pups were investigated. Hormonal analysis

revealed a significant decrease (P

< 0.032) in immunoactive

prolactin concentration in the stomach contents of pups

suckled by CS-exposed dams as compared with that of those

exposed only to FA. As milk-borne prolactin plays a role in

the normal development of the immune system (17,18) and

also the CNS (29,30) in neonatal rats, the results suggest that

0

50

100

150

200

250

300

Spleen Weight mg/100g Body Weight

Spleen Weights

0

20

40

60

80

100

Thymus Weight mg/100g Body Weight

Thymus Weights

3

8

19

0

100

200

300

400

500

600

A

B

C

FA Males CS Males FA Females CS Females FA Males CS Males FA Females CS Females

Pup Age in Weeks

Body Weight, grams

Body Weights

FA Males CS Males FA Females CS Females

Fig. 3. Effects of prenatal exposure to CS on (A) body, (B) spleen and (C) thymus weights of the offspring at 3, 8 and 19 weeks of age. There are four pups per group.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

0

10

20

30

40

50

_{FA (FA; N = 18)}

CS (CS; N = 17)

Animal Number

Milk Prolactin Concentration ng/ml

Mean ± SD

FA = 19.5±1.7 ng/ml

CS = 14.4±2.9 ng/ml

P = 0.032

Fig. 4. Immunoactive prolactin concentration in milk recovered from stomachs of suckling pups on Day 3 of lactation. The stomach contents were extracted as outlined in the text.

CS exposure of the dams could indirectly impair maturation of

these functions in the young by further reducing prolactin

secretion into the milk. However, at the concentration of

smoke employed, no significant changes were detected in

pup thymus or spleen weights or in the particular immune

functions selected for examination in this study. The lack of

CS-induced effects on the pup’s immune response is in contrast

to those immune alterations observed in mice exposed

pren-atally to an even lower dose of CS (31). Other more sensitive

immune end-points such as antibody-forming cell numbers

and/or immune cell-surface markers should therefore be

evaluated in immunological studies utilizing rats.

Finally, it should be mentioned that the concentration of

CS particulates used in the present studies (50 mg PM/m

3

/

200 g rat/4 h/day) was calculated to be roughly equivalent

to an individual smoking 2.7 packs of cigarettes per day.

This is a relevant exposure scenario for humans classified

as ‘heavy smokers’. However, the CS particulate dose used

is below that shown in other toxicological studies to reduce

maternal and neonatal body weight in rodents (see refs 32–35).

In conclusion, the authors suggest that one or more

com-ponents of CS may specifically negate or reverse the action

of the unknown gestational factor that protects against BCa.

Elucidation of the mechanisms involved in these phenomena

will be important in the eventual identification of the

gesta-tional BCa protective factor itself.

Acknowledgements

The reagents for the rat relaxin radioimmunoassay were generously provided by Dr O.D. Sherwood of the University of Illinois. The rat prolactin RIA kit was kindly provided by Dr Albert F. Parlow, Scientific Director of the NIDDK National Hormone and Peptide Program. Research described in this article was supported by the External Research Program of Philip Morris USA and Philip Morris International, and in part by NIEHS Center Grant P30 ES000260 and NCI Center Grant P30 CA016087.

Conflict of Interest Statement: None declared.

References

1. Lippman,M.E. (1998) Endocrine responsive cancer. In: Wilson,J.D., Foster,D.W., Kronenberg,H.M. and Larson,P.R., (eds) Williams Textbook of Endocrinology, 9th edn, WB Saunders, Philadelphia and London, pp. 1675–1692.

2. Coyle,Y.M. (2004) The effect of environment on breast cancer risk. Breast Cancer Res. Treat., 84, 273–288.

3. Russo,J., Calaf,G., Sohi,N., Tahin,Q., Zhang,P.L., Alvarado,M.E., Estrada,S. and Russo,I.H. (1993) Critical steps in breast carcinogenesis. Ann. NY Acad. Sci., 698, 1–20.

4. Russo,J., Gusterson,B.A., Rogers,A.E., Russo,I.H., Wellings,S.R. and van Zwieten,M.J. (1990) Biology of disease. Comparative study of human and rat mammary tumorigenesis. Lab. Invest., 62, 1–32. 5. Innes,K.E. and Byers,T.E. (2001) Smoking during pregnancy and breast

cancer risk in very young women (United States). Cancer Causes Control, 12, 179–185.

6. Couch,F.J., Cerhan,J.R., Vierkant,R.A., Grabrick,D.M., Therneau,T.M., Pankratz,V.S., Hartmann,L.C., Olson,J.E., Vachon,C.M. and Sellers,T.A. (2001) Cigarette smoking increases risk for breast cancer in high risk breast cancer families. Cancer Epidemiol Biomarkers Prev., 10, 327–332. 7. Brunet,J.S., Ghadirian,P., Rebbeck,T.R. et al. (1998) The effect of smoking

on breast cancer incidence in BRCA1 and BRCA2 carriers. J. Natl Cancer Inst., 90, 761–765.

8. Sukumar,S., McKenzie,K. and Chen,Y. (1995) Animal models for breast cancer. Mutat. Res., 333, 37–44.

9. Moon,R.C. (1969) Relationship between previous reproductive history and chemically induced mammary cancer in rats. Int. J. Cancer, 4, 312–317. 10. Thordarson,G., Jin,E., Guzman,R.C., Swanson,S.M., Nandi,S. and Talamantes,F. (1995) Refractoriness to mammary tumorigenesis in parous rats: is it caused by persistent changes in the hormonal

environment or permanent biochemical alterations in the mammary epithelia? Carcinogenesis, 16, 2847–2853.

11. Grubbs,C.J., Hill,D.L., McDonough,K.C. and Peckham,J.C. (1983) N-nitroso-N-methylurea-induced mammary carcinogenesis: effects of pregnancy on preneoplastic cells. J. Natl Cancer Inst., 71, 625–628. 12. Hecht,S.S. (2002) Tobacco smoke carcinogens and breast cancer. Environ.

Mol. Mutagen., 39, 119–126.

13. Haustein,K.O. (1999) Cigarette smoking, nicotine and pregnancy. Int. J. Clin. Pharmacol. Ther., 37, 417–427.

14. Braun,K.M., Cornish,T., Valm,A., Cundiff,J., Pauly,J.L. and Fan,S. (1998) Immunotoxicology of cigarette smoke condensates: suppression of macrophage responsiveness to interferon gamma. Toxicol. Appl. Pharmacol., 149, 136–143.

15. Byrd,R.S. and Howard,C.R. (1995) Children’s passive and prenatal exposure to cigarette smoke. Pediatr. Ann., 24, 640–645.

16. Cederqvist,L.L., Eddey,G., Abdel-Latif,N. and Litwin,S.D. (1984) The effect of smoking during pregnancy on cord blood and maternal serum immunoglobulin levels. Am. J. Obstet. Gynecol., 148, 1123–1126. 17. Grove,D.S., Bour,B., Kacsoh,B. and Mastro,A.M. (1991) Effect of neonatal

milk-prolactin deprivation on the ontogeny of the immune system of the rat. Endocr. Regul., 25, 111–119.

18. Kacsoh,B., Toth,B.E., Avery,L.M., Yamamuro,Y. and Grosvenor,C. (1991) Molecular heterogeneity of prolactin in lactating rats and their pups: biological and immunological activities in pituitary gland, serum and milk. Endocr. Regul., 25, 98–110.

19. Sherwood,O.D. and Crnekovic,V.E. (1979) Development of a homologous radioimmunoassay for rat relaxin. Endocrinology, 104, 893–897. 20. Djeu,J.Y. (1995) Natural killer activity. In: Burlesson,G.R., Dean,J.H. and

Munson,A.E., (eds) Methods in Immunotoxicology. Vol 1, Wiley-Liss, New York, pp. 437–450.

21. Harvey,P.W. (2005) Human prevalence of rodent prolactin-induced non-genotoxic mammary carcinogenesis: prolactin involvement in human breast cancer and significance for toxicological risk assessments. J. Appl. Toxicol., 25, 179–183.

22. Vick,R.S., Wong,V.L. and Witorsch,R.J. (1987) Biological, immunological and biochemical characterization of cleaved prolactin generated by lactating mammary gland. Biochim. Biophys. Acta., 931, 196–204. 23. Goffin,V., Bernichtein,S., Touraine,P. and Kelly,P.A. (2005) Development

and potential uses of human prolactin receptor antagonists. Endocr. Rev., 26, 400–422.

24. Clevenger,C.V. (2003) Role of prolactin/prolactin receptor signaling in human breast cancer. Breast Dis., 18, 75–86.

25. Guillaumot,P. and Cohen,H. (1994) Heterogeneity of the prolactin receptor in the rat mammary gland and liver during various physiological states. J. Endocrinol., 141, 271–278.

26. Van Horn,K.M., Binart,N., Kelly,P.A. and Talamantes,F. (2001) Prolactin receptor expression in the epithelia and stroma of the rat mammary gland. J. Endocrinol., 171, 85–95.

27. Hussain,S.P., Harris,C.C., Hainaut,P. and Vahakangas,K. (1999) p53 mutation spectrum and load: the generation of hypotheses linking the exposure of endogenous or exogenous carcinogens to human cancer. Mutat. Res., 16, 23–32.

28. Hainaut,P. and Vahakangas,K. (1997) p53 as a sensor of carcinogenic exposures: mechanisms of p53 protein induction and lessons from p53 gene mutations. Pathol. Biol., 45, 833–844.

29. Grosvenor,C.E., Shyr,S.W. and Crowley,W.R. (1986) Effect of neonatal prolactin deficiency on prepubertal tuberoinfundibular and tuberohypop-hyseal dopaminergic neuronal activity. Endocrinol. Exp., 20, 223–228. 30. Shyr,S.W., Crowley,W.R. and Grosvenor,C.E. (1986) Effect of neonatal

prolactin deficiency on prepubertal tuberoinfundibular and tuberohypo-physeal dopaminergic neuronal activity. Endocrinology, 119, 1217–1221. 31. Ng,S.P., Silvrstone,A.E., Lai,Z.-W. and Zelikoff,J.T. (2006) Effects of prenatal exposure to cigarette smoke on offspring tumor susceptibility and associated immune mechanisms. Toxicol. Sci., 89, 135–144. 32. Hecht,S.S. (2005) Carcinogenicity studies of inhaled cigarette smoke in

laboratory animals: old and new. Carcinogenesis, 26, 1488–1492. 33. Leichter,J. (1995) Decreased birth weight and attainment of postnatal

catch-up growth in offspring of rats exposed to cigarette smoke during gestation. Growth Dev. Aging, 59, 63–66.

34. Witschi,H., Lundgaard,S.M., Rajini,P., Hendrickx,A.G. and Last,J.A. (1994) Effects of exposure to nicotine and to sidestream smoke on pregnancy outcome in rats. Toxicol. Lett., 71, 279–286.

35. Leichter,J. (1989) Growth of fetuses of rats exposed to ethanol and cigarette smoke during gestation. Growth Dev. Aging, 53, 129–134.

Received October 8, 2005; revised January 7, 2006; accepted January 23, 2006