Suppression of humoral immune responses by dialkylnitrosamines: Structure-activity relationships

FUNDAMENTAL

AND

APPLIED

Suppression

NORBERT

TOXICOLOGY

12,32 l-332 (1989)

of Humoral Immune Responses by Dialkylnitrosamines: Structure-Activity Relationships’

E. KAMINSRI,*

STEPHEN D. JORDAN,* DENNIS PAGE,? BYUNG MICHAEL P. HOLSAPPLE*

*Department of Pharmacology and Toxicology and fDepartment of Oral Pathology, Virginia Commonwealth University, Richmond, Virginia 23298, and *Department Engineering, Korea Advanced Institute of Science and Technology,

Received

February

17, 1988; accepted

SAM KIM,*

AND

Medical College of Virginia/ of Biological Sciences and Seoul, Korea

July 18, I988

Suppression of Humoral Immune Responses by Dialkylnitrosamines: Structure-Activity Relationships. KAMINSIU, N. E., JORDAN, S. D., PAGE, D., KIM, B. S., AND HOLSAPPLE, M. P. ( 1989). Fundam. Appl. Toxicol. 12,32 l-332. Comparisons between chemical structure of N, Ndialkylnitrosamine congeners and their ability to alter the Day 4 IgM antibody response to sRBC, body weights, and organ weights of female B6C3Fl mice were investigated. Short-chain nitrosamine congeners were selected for these studies on the basis of two criteria: (1) congeners with symmetrical aliphatic chain length [N-nitrosodimethylamine (DMN), N-nitrosodiethylamine (DEN), N-nitrosodipropylamine (DPN), N-nitrosodibutylamine (DBN)] and (2) congeners possessing an N-methyl group [N-nitrosomethylethylamine (MEN), N-nitrosomethylpropylamine (MPN), and N-nitrosomethylbutylamine (MBN)]. The immunotoxicity of each congener was evaluated based on the compound’s ability to suppress the in vivo sRE%C antibody response following 7 consecutive days of treatment. An ED50 dose was calculated, using a linear regression analysis, for each congener and represents the millimoles of congener per kilogram body weight required to cause a 50% suppression of the sRBC response. These studies demonstrated two general trends: (1) those dialkylnitrosamine congeners that possessed an N-methyl group were most immunotoxic and exhibited comparable ED50 concentrations (42-183 rmol/kg); and (2) dialkylnitrosamine congeners possessing symmetrical aliphatic chains demonstrated an inverse relationship between aliphatic chain length and immunotoxic potency-DMN (62 pmol/kg) > DEN (276 rmol/kg) > DPN (467 rmol/kg) > DBN (1557 pmol/kg). Comparisons were also made between the immunotoxic potency of various nitrosamine congeners in the whole animal and their potency in an in vitro hepatocyte-spleen cell coculture system. o 1989 Society of Toxicology.

The N,N-dialkylnitrosoamines constitute a well-known group of potent carcinogens that target the liver as one of the primary sites of action. The structurally simplest congener belonging to this chemical group, N-nitrosodimethylamine (DMN), was first reported to be severely hepatotoxic by Barnes and Magee ’ Presented at the 1987 Annual Meeting of the Society of Toxicology, Washington, DC. This work was sup ported by NIH Grant ES-03564, NRSA Grant ES-054 15, and ACS Grant IN- 105L.

(1954) when two laboratory technicians developed cirrhosis of the liver following exposure in the laboratory. Studies by Barnes and Magee (1954) demonstrated that direct exposure to DMN resulted in severe liver damage in rats, mice, rabbits, guinea pigs, and dogs. Subsequent studies by Barnes and Magee (1967) demonstrated that, although animals could survive an acute dose of DMN, a high proportion of those exposed animals would develop neoplastic lesions most characteristitally associated with the liver. Heath (1962)

321

0272-0590/89 $3.00 Copyright 0 1989 by the society of Toxicology. All rights of reproduction in any form reserved.

322

KAMINSKl

ET AL.

reported that DMN undergoes metabolic de- 1977; Lijinsky and Taylor, 1978). Several generation in the liver, which is now known generalizations can be made from the findto result in enzymatic activation of DMN to ings reported in these studies: ( 1) greater Nits active form, a potent alkylator of cellular alkyl chain length results in decreased dialmacromolecules. The metabolic activation of kylnitrosamine toxicity; (2) branched N-alkyl DMN most probably occurs by oxidative de- chains render nitrosamine congeners less methylation. Studies in our laboratory have toxic then if the aliphatic chains are not focused primarily on DMN-induced immubranched; and (3) it appears that the mechanotoxicity based on the premise that alkylanism of carcinogenesis by nitrosamines intors are potent immunosuppressants and that volves initial enzymatic oxidation at an CXDMN-induced carcinogenesis may be at least carbon. Based on these previously reported in part a result of the host’s compromised imrelationships between nitrosamine structure mune system which is unable to fulfill its and toxicity it was our objective to evaluate function in immune surveillance. These stud- whether similar trends could be applied in ies have demonstrated that (1) DMN is a po- predicting the relative immunotoxic potency of various short-chain N,N-dialkylnitrosamtent immunotoxin capable of suppressing both humoral (Holsapple et al., 1984) and ine congeners. Since enzymatic oxidation of cell-mediated immune responses (Holsapple the a-carbon has been reported as being imet al., 1985); (2) the cell target for DMN-meportant in the activation of nitrosamines to a diated immunosuppression is primarily the B carcinogenic form it was hypothesized that cell, however, T cells are also targeted by this would also represent the most immunoDMN but to a lesser extent (Johnson et al., toxic form for nitrosamine congeners. DMN, the simplest dialkylnitrosamine congener, 1987a); (3) DMN-induced immunosuppreswas selected as the prototype to which all sion requires metabolic activation of DMN and is mediated by reactive intermediates of other nitrosamine congeners would be comDMN metabolism (Johnson et al., 1987a); pared. It was our intention to make compariand (4) suppression of in vitro humoral imsons between DMN and structurally similar mune responses by acetoxy-DMN, which N,N-dialkylnitrosamine congeners with symspontaneously breaks down to the active in- metrically greater aliphatic chain length or termediate of DMN without requiring me- congeners with one increasing aliphatic chain tabolism, is reversed by direct addition of ex- and the other aliphatic chain remaining in ogenous DNA to culture but not by addition the N-methyl position. The congeners seof nonspecific proteins such as BSA, suggest- lected for these studies were DMN, N-nitroing that the reactive intermediate of DMN re- sodiethylamine (DEN), N-nitrosodiprosponsible for immunosuppression has a high pylamine (DPN), N-nitrosodibutylamine affinity for DNA (Johnson et al., 1987~). (DBN), N-nitrosomethylethylamine (MEN), Numerous studies have demonstrated that N-nitrosomethylpropylamine (MPN), and NDMN and related N,N-dialkylnitrosamine nitrosomethylbutylamine (MBN). congeners are extremely mutagenic as well as carcinogenic; however, very little is known MATERIALS AND METHODS about the molecular basis for this phenomenon. To gain a better understanding of nitrosamine toxicity a number of structure-activMice. Virus-free (Sendai and hepatitis) female B6C3Fl mice 5-6 weeks of age were purchased from the ity relationship studies have been performed Frederick Cancer Research Center. On arrival, mice were using various endpoints including DNA randomized, transferred to plastic cages containing a damage (Kohda et al., 1982) and relative car- sawdust bedding (four mice per cage), and quarantined cinogenicity (Lee and Guttenplan, 1981; for 1 week. Mice were given food (Purina Certified LaboWishnok and Archer, 1976; Wishnok et al., ratory Chow) and water ad libitum and were not used for

IMMUNOTOXICITY

OF DIALKYLNITROSAMINES

experimentation until their body weights were 17-20 g. Animal holding rooms were kept at 21-24’C and 4060% relative humidity with a 12-hr light/dark cycle. Chemicals. The following nitrosamine congenem were purchased from Sigma (St. Louis, MO.): N-nitrosodimethylamine, N-nitrosodiethylamine, N-nitrosodipropylamine, N-nitrosodibutylamine, N-nitrosomethylethylamine, N-nitrosomethylpropylamine, and N-nitrosomethylbutylamine. Dosing regimen. In all cases,animals were injected (ip) daily for 7 consecutive days with vehicle or the appropriate concentration of nitrosamine congener. Saline was utilized as the vehicle for DMN, DEN, and MEN and corn oil was used as the vehicle for MPN, MBN, DPN, and DBN due to the low aqueous solubility of the latter congeners. In those instances where corn oil was used as a vehicle rather than saline, a naive group was included in the study to assessany potential effects attributable to corn oil. The following concentrations were utilized for the respective congenem and were determined in preliminary range-finding studies: DMN- 1,3,6 mg/kg ( 13,38, 76 pmol/kg); DEN-16, 24, 48 mg/kg (156, 235, 470 rmol/kg); MEN-lo, 15,20 mg/kg(ll3, 170,227 pmol/ kg); DPN-50, 70, 90 m&kg (384, 538, 692 pmol/kg); MPN-5, 10, 15 mg/kg (49,98, 147 rmol/kg); DBN200, 350, 500 mg/kg (1260, 2210, 3160 rmol/kg); and MBN- 1.5,4.7,9.4 mg/kg ( 13,40,8 1 lmol/kg). In vivo antibody responses. For sensitization, mice were injected iv into the tail vein with 1 X 10’ sRBC/ mouse (Colorado Serum Co., Denver) in 0.2 ml of Earle’s balanced salt solution (EBSS; GIBCO, Grand Island, NY) on Day 8, one day after the last treatment with the specific test chemical. The sensitization interval was 4 days, at which time the animals were sacrificed and spleens were removed. Day 4 has consistently represented the optimum day of response for the in vivo IgM antibody response to sRBC. Single-spleen-cell suspensions were prepared from each spleen in 3 ml EBSS, washed and resuspended in 3 ml EBSS, and counted on a Coulter counter. Antibody-forming cells (AFC) were enumerated using a modified version of the Jeme plaque assay. Briefly, a 0.5% melted agar (Difco, Detroit, MI) solution in EBSS was prepared containing 0.05% DEAEdextran (Pharmacia, Piscataway, NJ) and was maintained at 47’C. The melted agar solution was dispensed in 350-~1 aliquots into 12 X 75-mm heated culture tubes. Each spleen cell suspension was diluted 30-fold, by resuspending a lOO-~1 aliquot of each cell suspension in 2.9 ml of EBSS, and held on ice. Each agar tube received 100 ~1 of the diluted cell suspension, 25 ~1 of guinea pig complement (GIBCO), and 25 ~1 of indicator sRBC. The tube was immediately vortex mixed and poured into a 100 X 15-mm Petri dish, and the agar solution was covered with a 45 X 50-mm microscope coverslip. Once the agar solidified, the Petri dishes were incubated at 37°C for 3 hr. Following the incubation, the AFC were enumerated at 6.5X magnification using a Bellco plaque viewer.

323

Histopathology. Livers were isolated from nitrosamine-treated mice and placed immediately into phosphate-buffered formalin. Tissues were subsequently sectioned, stained with hematoxylin and eosin, and examined microscopically for pathological abnormalities. Coculture of primary mouse hepatocytes and mouse splenocytes. Mixed cultures of hepatocytes and spleen cells from B6C3Fl mice were established as previously described (Johnson and et aZ., 1987a). Hepatocytes were isolated by collagenase perfusion and a monolayer was established during a 20-hr incubation. In these experiments hepatocytes and spleen cells were cultured with various nitrosamine congeners for 4 hr at 37°C and 5% COZ. Splenocytes were then gently removed, washed, and cultured in the presence of sRBC (described below) for determinations of the in vitro antibody responses. In vitro antibody assays. Spleens cells were isolated following 4 hr of culture with hepatocytes and the respective nitrosamine congeners. The splenocyte suspensions were isolated from each hepatocyte well, washed with RPMI, and adjusted to 1.O X 10’ cells/ml in RPM1 1640 supplemented with 10% FCS, 5 X lo-’ M 2-mercaptoethanol (2-ME), 2mM L-glutamine, antibiotic-antimycotic (100 units penicillin, 100 pg streptomycin, and 0.25 pg fungizone/ml) (GIBCO). A 500-111 ahquot of each adjusted spleen cell suspension was transferred to a well of a 48well Costar culture plate (Cambridge, MA). Each well was subsequently sensitized with 7.5 X lo6 sRBC and cultured for 5 days in a Bellco stainless-steel tissue culture chamber. In our laboratory, the peak day of response following in vitro sensitization with sheep erythrocytes has consistently been 5 days. AFC were enumerated as previously described, except that a 50-~1 aliquot of cell suspension was taken from each well of the culture plate and added to the agar solution. During the 3-hr incubation period the number of spleen cells per well and viability as previously described (Holsapple et al., 1984) were determined using a Coulter counter. Results from quadruplicate cultures were expressed as the mean AFC/ lo6 recovered splenocytes t- SE. Statistics. The mean f SE was determined for each treatment group of a given experiment. The homogeneity of the results was determined using Bartlett’s test for homogeneity (Bartlett, 1937). Homogeneous data were evaluated by a pan metric analysis of variance. When significant differences occurred, treatment groups were compared to the vehicle controls using Dunnett’s t test (Dunnett, 1955). Nonhomogeneous data were evaluated for significance using Wilcoxon’s rank test (Gehan-Wilcoxon Test, 1975).

RESULTS The immunotoxicity of each nitrosamine congener was evaluated based on the com-

324

KAMINSKl

FIG. 1. In vivo T-dependent antibody response following exposure to IV-nitrosodimethylamine (DMN). Female B6C3Fl mice were treated for 7 consecutive days ip with vehicle (saline) or DMN (1, 3, or 6 mg/kg). On Day 8 mice were sensitized with sRBC by iv injection. On Day 12 the antibody-forming cells (AFC) were enumerated. The bars represent the means + SE derived from five animals per group. *p c: 0.05 as determined by Dunnett’s t test as compared to the vehicle control group.

pound’s ability to suppress the T-dependent antibody response to sRBC following 7 consecutive days of treatment. An ED50 dose was calculated, using a linear regression analysis, for each nitrosamine congener and represents the millimoles of congener per kilogram body weight required to cause 50% suppression of the T-dependent antibody response as compared to the vehicle controls. Measurements of the T-dependent antibody response, following exposure to the nitrosamine congeners, were determined on Day 4, the peak day of response. Previous studies in our laboratory with DMN, the prototype nitrosamine congener, have not shown any evidence of a shift in the peak day of response. Since the molecular weights of the test congeners were different, standardization of the data was necessary for direct comparisons of immunotoxic potency. In addition to antibody responses, measurements of liver, spleen, thymus, and body weights were determined and the histopathology of prepared liver sections was examined. The dosing concentrations for the respective congeners were

ET AL.

determined by first performing range-finding studies using a broad range of concentrations for each nitrosamine compound. From these studies, dose levels were chosen that were anticipated to include concentrations of the respective congeners that would suppress T-dependent antibody responses by approximately 50%. Treatment with DMN resulted in a steep dose-related suppression of the T-dependent antibody response with magnitudes of 20,60, and 87% at 1, 3, and 6 mg/kg, respectively, as compared to the vehicle controls (Fig. I). Based on these results, an ED50 dose of 62 pmol/kg was calculated (Fig. 2). Spleen and liver weights were unaffected; however, a significant decrease in thymus weights was observed at the 6 mg/kg dose level, which represented an 80% reduction as compared to the vehicle controls (Table 1). At 3 mg/kg, livers demonstrated mild chronic hepatitis, and at 6 mg/kg, liver sections showed centrilobular necrosis (Table 2). An increase in body weight was observed at the 6 mg/kg dose level

MEN 2 5 ii 2 5

DEN MPN DPN MBN DBN

0

500

1000 ED50

1500

2000

(pmollkg)

FIG. 2!. The relative nitrosamine concentration required to suppress the in vivo IgM antibody response to sRBC by 50%. Data from dose-response determinations on the Day 4 IgM antibody response to sRBC following the 7-day exposure regimen for each dialkylnitrosamine congener were standardized on a gmol/kg basis. An ED50 dose (rmol of congener/kg body weight required to cause a 50% suppression of the T-dependent antibody response) was calculated, using a linear regression analysis, for each nitrosamine congener which represents the relative immunotoxic potency of each respective congener.

IMMUNOTOXICITY

OF DIALKYLNITROSAMINES

325

TABLE 1 B~DYAND ORGANWEIGHTS’ FOLLOWINGEXPOSURETODIALKYLN~TROSAMINECONGENERS Treatment @&kg) DMN VH 1 3 6 MEN VH 10 15 20 DEN VH 16 24 48 MPN VH 5 10 15 DPN VH 50 70 90 MBN VH 1.5 4.7 9.4 DBN VH 200 300 500

Body weight 6%)

Spleen weight bg)

21.1 f 1.2” 21.1 -co.9 19.8 -e 0.3 28.0 f 1.3**

90r 8 lOok 11 103+ 5 88k 9

Thymus weight (mg)

Liver we&t (mg)

55* 64+ 56+ 9k

2 8 4 6**

1057* 1056+ 881 f 1097 +

59 86 36 54

19.9 + 0.5 19.9-1-0.5 18.1 +0.6* 16.0 -c 0.4**

90r 98+ 96+ 58k

5 4 6 8*

45* 48+ 51Ik 31+

2 3 2 4

1117k 1084+ 989+ 775 f

46 28 52 52**

22.9 + 1.0 21.4kO.4 16.1 + 0.3** 13.6+0.8**

139+ 137* 76 k 33 iz

12 3 5** 12**

61t 40-+ 26 iz 24k

5 5* 2* II**

1089+ 887 + 68Ok 426 k

52 47* 50* 32**

19.9 + 0.6 19.0 f 0.5 17.2 +- 0.5** 17.2 + 0.7**

102k 93+ 56 i61 k

7 6 6** 6**

60+ 71+ 14k 9+

5 8 4** 2**

1081 + 1022f 796+ 746 k

95 63 13* 30**

19.6 + 0.5 18.6 -c 0.5 17.9 + 0.5 16.0 f 0.6**

99+ 68~ 61 -t 53k

6 6** 6’* 4**

63+ 20+ 12* lo+

5 6** 2** 2**

801 f 842 + 706+ 638 f

32 102 51 20**

21.2+0.5 20.3 k 0.4 18.9 k 0.5* 15.6 f 0.6**

1172 105-f98+ 63 f

4 7 6* 4**

50* 50* 47* 26+

2 2 5 8

783-t 689k 652+_ 573 f

15 34 12* 32**

21.2 -r-o.5 18.2 + 0.6 18.3 + 1.0 16.2 + 0.7**

1352 6 86 AZ 6* 66 + 10** 28k l**

49* 42k 46k 6-c

3 2 3 l**

9llk 852k 859+ 812k

33 71 40 7

’ Female B6C3Fl mice were treated for 7 consecutive days ip with vehicle or the indicated nitrosamine congener. On Day 12, mice were sacrificed and organ and body weights were determined. * Means ? SE derived from five animals per group. *p < 0.05 as determined by Dunnett’s I test as compared to the vehicle group. ** p < 0.0 1 as determined by Dunnett’s t test as compared to the vehicle group.

(Table 1) and was attributable to the accumulation of ascites in the peritoneum. Ascites was observed only in animals exposed to DMN and not with any other nitrosamine congener tested in these studies. MEN treatment resulted in a 20, 57, and 92% dose-dependent suppression of the T-de-

pendent antibody response at 10, 15, and 20 mg/kg, respectively (Fig. 3). The calculated ED50 dose for MEN was 183 pmol/kg (Fig. 2). Spleen and thymus weights were affected only at the 20 mg/kg dose level, resulting in a 36% decrease (Table 1). Liver weights were reduced by 23 and 49% at the 15 and 20 mg/

326

KAMINSKI

ET AL.

however, resulted in minimal damage to hepatic tissue as evaluated histopathologically and was characterized by mild hepatitis (Table 2). MPN demonstrated a very steep dose response with respect to the T-dependent antibody response. At the 10 and 15 mg/kg doses, antibody responses were suppressed by 71 and 99%, respectively, as compared to the vehicle controls, yet 5 mg/kg had no apparent effect (Fig. 5). Effects on organ and body weights were also only observed at the 10 and 15 mg/kg dose levels (Table 1). Spleen FIG. 3. In vivo T-dependent antibody response following exposure to IV-nitrosomethylethylamine (MEN). Fe- weights were reduced by 45 and 40%, thymus male B6C3Fl mice were treated for 7 consecutive days weights by 77 and 85%, liver weights by 26 ip with vehicle (saline) or MEN (10, 15, or 20 mg/kg). and 30%, and body weights by 14%, at both On Day 8 mice were sensitized with sRBC by iv injecthe 10 and 15 mg/kg dose levels. Exposure to tion. On Day 12 the antibody-forming cells (AFC) were MPN at all test dose levels resulted in minienumerated. The bars represent the means f SE derived from five animals per group. *p < 0.05 as determined by mal hepatic damage and was characterized Dunnett’s t test as compared to the vehicle control group. as general mild hydropic degeneration (Table 2). Mice treated with DPN demonstrated 42, kg dose levels, respectively. There was no effect on thymus weights at any of the test 54, and 78% suppression of the T-dependent doses. Histopathology of liver sections re- antibody response at 50, 70, and 90 mg/kg, vealed hepatic damage at all dose levels with respectively, as compared to the vehicle controls (Fig. 6). This represents an ED50 dose generalized hydropic degeneration and mild hepatitis occurring at 10 mg/kg and more se- of 467 /*mol/kg (Fig. 2). Spleen and thymus vere damage occurring at the highest dose, re- weights were significantly decreased at all sulting in focal cellular necrosis (Table 2). DPN dose levels and corresponded to 3 1,38, Mitotic figures were also observed at the in- and 46% reductions in spleen weights and 68, termediate and high doses. Body weights 8 1, and 84% decreases in thymus weights, rewere decreased by 10 and 25% at the 1.5 and spectively, at 50, 70 and 90 mg/kg (Table 1). Liver and body weights were also reduced as 20 mg/kg dose levels. Exposure to DEN resulted in 61 and 82% a result of DPN treatment but only at the 90 examinasuppression of the T-dependent antibody re- mg/kg dose level. Histopathologic tion of hepatic tissue showed generalized sponse at 24 and 48 mg/kg, respectively (Fig. 4) which corresponded to an ED50 dose of mild hydropic degeneration at all test doses 276 pmol/kg (Fig. 2). A dose-dependent de- and hepatitis at the 90 mg/kg dose level (Tacrease in spleen, liver, and body weights was ble 2). Treatment with MBN resulted in 3 1, 43, also observed at 24 and 48 mg/kg (Table 1). Spleen weights were decreased by 45 and and 76% suppression of the T-dependent an76%, liver weights were decreased by 38 and tibody response at 1.5, 4.7, and 9.4 mg/kg, 6 1%, and body weights were decreased by 29 respectively (Fig. 7). This nitrosamine congeand 40%, respectively, at 24 and 48 mg/kg ner exhibited the greatest immunosuppresDEN. Thymus weights were decreased dose sive potency on a molar basis of all the test dependently at all doses of DEN and resulted congeners in this study and resulted in an in 34, 56, and 61% decreases at the 16, 24, ED50 dose of 47 pmol/kg (Fig. 2). Effects on and 48 mg/kg dose levels. Exposure to DEN, spleen, liver, and body weights were approxi-

IMMUNOTOXICITY

321

OF DIALKYLNITROSAMINES TABLE 2

HISTOPATHOLOGIC EVALUATION OF LIVER TISSUES FOLLOWING EXPOSURE TO DIALKYLNITROSAMINE CONGENERS Treatment (mg/kg) DMNb I 3 6 MEN 10 15 20 DEN 16 24 48 MPN 5 10 15 DPN 50 IO 90 MBN 1.5 4.7 9.4 DBN 200 350 500

Generalized mild hydropic degeneration

Mild chronic Moderate chronic Focal cellular necrosis hepatitis hepatitis

Centrilobular necrosis

Occasional mitotic figures

+

-

-

+

-

+ + -

+ +

+

+ +

-

+ +

-

+ +

-

-

-

-

+ +

-

-

-

-

-

+ + +

+

-

-

-

-

+

+

-

-

-

-

+ + -

-

-

+ +

-

+ -

-c -

’ Livers were isolated from treated mice, placed in phosphate-buffered formalin, stained with hematoxylin and eosin, and sectioned for mounting on microscope slides. b Female B6C3Fl mice were treated ip for 7 consecutive days with one of the following N,N-dialkylnitrosamine congeners: N-nitrosodimethylamine (DMN), N-nitrosodiethylamine (DEN), N-nitrosodipropylamine (DPN), N-nitrosodibutylamine (DBN), N-nitrosomethylethylamine (MEN), N-nitrosomethylpropylamine (MPN), and N-nitrosomethylbutylamine (MBN). ’ (+) Presence and (-) absence of the designated pathological lesions.

mately the same at the 4.7 mg/kg dose, resulting in approximately a 16% reduction in all three parameters (Table 1). The most marked affects on organ weights were observed on spleen and thymus at the 9.4 mg/kg dose level, resulting in 47 and 48% decreases. Also affected at this dose level were liver weights which were reduced by 27% and body weights which were reduced by 26%. Surprisingly, MBN exposure had little effect on hepatic tis-

sue when assessed by histopathology and could be characterized as generalized mild hydropic degeneration and mild hepatitis (Table 2). Treatment with DBN at 200,350, and 500 mg/kg resulted in 55, 69, and 86% suppression of the T-dependent antibody response as compared to the vehicle controls (Fig. 5), yielding an ED50 dose of 1557 pmol/kg (Fig. 8). Spleen weights were markedly decreased

328

KAMINSKI CH CH

3

-4+No

CH3CH2’

ET AL.

DEN, approximately 10 mM for DPN, and greater than 20 mM for MEN. DISCUSSION

FIG. 4. In vivo T-dependent antibody response following exposure to N-nitrosodiethylamine (DEN). Female B6C3Fl mice were treated for 7 consecutive days ip with vehicle (saline) or DEN (16,24, or 48 mg/kg). On Day 8 mice were sensitized with sRBC by iv injection. On Day 12 the antibody-forming cells (AFC) were enumerated. The bars represent the means + SE derived from five animals per group. **p < 0.01 as determined by Dunnett’s c test as compared to the vehicle control group.

by DBN treatment, 36, 5 1, and 79% at 200, 350, and 500 mg/kg, respectively (Table I). Thymus and body weights were affected only at the highest dose, resulting in 86 and 23% decreases, respectively. There was no significant effect on liver weight attributable to DBN exposure; however, histopathological examination of hepatic tissues revealed generalized mild hydropic degeneration at 200 and 350 mg/kg and focal cellular necrosis at 350 and 500 mg/kg (Table 2). Five nitrosamine congeners-DMN, DEN, DPN, MEN, and MBN-were selected for further in vitro studies. Isolated splenocytes and hepatocytes were cocultured in the presence of various concentrations of nitrosamines and then assayed for their Day 5 IgM antibody response to sRBC. These studies indicated that DMN and MBN were the most potent (i.e., immunosuppressive) ofthe nitrosamines tested utilizing this assay system. The ED50 concentrations in culture for DMN and MBN were similar, 0.7 and 1 mM, respectively (Table 3). The ED50 values for the remaining congeners were 20 mM for

The primary objective of these studies was to determine the relationship between the chemical structure of certain short-chain dialkylnitrosamines and their relative immunotoxic potency. The T-dependent antibody response to sRBC was selected for assessing the relative immunotoxic potency of the respective congeners. This immunological parameter was selected primarily because we have found this to be very sensitive in detecting chemically induced alterations of immune function, particularly by DMN, the prototype of the nitrosamine congeners (Holsapple et al., 1984; Johnson et al., 1987b). We selected two aspects of dialkylnitrosamine chemical structure that have been previously reported as contributing to their toxic potency; (1) aliphatic chain length (Kohda et al.,

Ck$CH CH 2 2-N-No CH3’

FIG. 5. In vivo T-dependent antibody response following exposure to Iv-nitrosomethylpropylamine (MPN). Female B6C3Fl mice were treated for 7 consecutive days ip with vehicle (corn oil) or MPN (5, 10, or I5 mg/kg). On Day 8 mice were sensitized with sRBC by iv injection. On Day 12 the antibody-forming cells (AFC) were enumerated. The bars represent the means f SE derived from five animals per group. **p < 0.01 as determined by Dunnett’s t test as compared to the vehicle control group.

IMMUNOTOXICITY

OF DIALKYLNITROSAMINES

FIG. 6. In viva T-dependent antibody response following exposure to N-nitrosodipropylamine (DPN). Female B6C3Fl mice were treated for 7 consecutive days ip with vehicle (corn oil) or DPN (50,70, or 90 mg/kg). On Day 8 mice were sensitized with sRBC by iv injection. On Day 12 the antibody-forming cells (AFC) were enumerated. The bars represent the means + SE derived from five animals per group. *p < 0.05 and **p < 0.0 1 as determined by Dunnett’s t test as compared to the vehicle control group.

1982; Lee and Guttenplan, 198 1) and (2) the presence of an N-methyl group in the a-carbon position (Wishnok et al., 1978), to determine whether these structural properties increased immunotoxicological potency. Studies with dialkylnitrosamine congeners possessing symmetrical aliphatic chains demonstrated that increased aliphatic chain length resulted in a decrease in immunosuppressive potency, on a ED50 basis. A ranking of immunotoxic potency, following in vivo exposure, for dialkyl congeners with aliphatic chain symmetry is DMN (62 pmol/kg) > DEN (276 pmol/kg) > DPN (467 pmol/kg) > DBN ( 15 57 pmol/kg). These results clearly support previous findings that an inverse relationship exists between aliphatic chain length for N-dialkylnitrosamines and nitrosamine toxicity (Kohda et al., 1982; Lee and Guttenplan, 198 1). Furthermore, it appears that this observation can be extended to immunotoxic potency as well. When N-methyl dialkylnitrosamine congeners were compared, the length of the second aliphatic

329

chains had little effect on the relative immunotoxic potency of these respective congeners. Those congeners that possessed at least one N-methyl group consistently demonstrated the greatest ability to suppress humoral responses, on a micromole-per-kilogram basis. Furthermore, with the exception of DMN, which possesses two N-methyl groups, increased chain length of N-methyl dialkylnitrosamines actually resulted in increased immunotoxicity as can be seen with the following congeners: MBN (47 pmol/kg) > MPN (86 pmol/kg) > MEN (183 pmol/kg). These results strongly suggest that the cu-carbon on dialkylnitrosamine congeners plays a significant role in the generation of the active immunosuppressive metabolite(s). Results from previous metabolic studies with DMN suggested that the very reactive methyldiazonium ion is generated during DMN metabolism which, in turn, alkylates DNA and results in a variety of pathological lesions including hepatotoxicity and altered immune responses. Based on the premise that an N-

FIG. 7. In vivo T-dependent antibody response following exposure to N-nitrosomethylbutylamine (MBN). Female B6C3Fl mice were treated for 7 consecutive days ip with vehicle (corn oil) or MBN (1.5,4.7, or 9.4 mgl kg). On Day 8 mice were sensitized with sRBC by iv injection. On Day 12 the antibody-forming cells (AFC) were enumerated. The bars represent the means f SE derived from five animals per group. *p < 0.05 and **p < 0.0 1 as determined by Dunnett’s t test as compared to the vehicle control group.

330

KAMINSKI

FIG. 8. In viva T-dependent antibody response following exposure to N-nitrosodibutylamine (DBN). Female B6C3Fl mice were treated for 7 consecutive days ip with vehicle (corn oil) or DBN (200, 350, or 500 mg/kg). On Day 8 mice were sensitized with sRBC by iv injection. On Day 12 the antibody-forming cells (AFC) were enumerated. The bars represent the means + SE derived from five animals per group. **p -Z 0.01 as determined by Dunnett’s t test as compared to the vehicle control group.

methyl group is involved in the generation of the reactive intermediate one would expect those congeners (i.e., DMN, MEN, MPN, and MBN) to have the greatest immunotoxic potency. It is, however, unclear why ascites was observed only when mice were treated with DMN and not with the other nitrosamine congeners utilized in these studies. The ascites was attributed to the centrilobular necrosis which was also only produced in mice exposed to DMN. This pathological difference would suggest that a different metabolic profile may in fact exist for DMN than for other N-methyl dialkylnitrosamine congeners. It is interesting to speculate that the liver damage by methylation of DMN is different from that produced by alkylation with longer aliphatic chains. In vitro determinations of nitrosamine potency using a hepatocyte-spleen cell coculture system demonstrated, with the exception of DPN, that immunotoxic potency of the nitrosamine congeners was significantly increased when cultured in the presence of hepatocytes as compared to cultures treated

ET AL.

with nitrosamine congeners in the absence of hepatocytes (data not shown). These findings further suggest that short-chain dialkylnitrosamines undergo enzymatic bioactivation in the liver, resulting in the generation of reactive intermediates and increased immunotoxicity. DMN and MBN, the most potent immunosuppressive nitrosamine congeners tested in vivo, also demonstrated the greatest potency in vitro. The results obtained with MEN, DEN, and DPN, however, were not as conclusive. Based on our findings in vivo, we predicted that MEN would have an ED50 concentration similar to that of MBN and DMN and that DEN would be considerably more immunosuppressive than DPN. In actuality MEN, in our in vitro system, was the least toxic congener of the five tested and DEN was in fact less immunotoxic than DPN. Although the expected immunotoxic potencies of DEN and DPN were reversed, these two congeners still demonstrated lower toxicity than MBN and DMN, as was predicted. It is unclear, specifically, why the reTABLE 3 ED50 CONCENTRATION OF N,N-DIALKYLNITRO SAMINECONGENERSFOLLOWINGA~-~~SPLEENCELLHEPATOCYTECOCULTURE Nitrosamine congener

ED50 b-M

DMN MBN DPN DEN MEN

OS-l.0 1.0 10 20 220

a The ED50 concentration is the concentration of the respective N,N-dialkylnitrosamine congener required to suppress the in vitro IgM antibody response to sRBC by 50%, as compared to the vehicle control, following a 4hr incubation with isolated hepatocytes and spleen cells from female B6C3Fl mice. ’ The following congeners were assayed for their ability to suppress the in vitro IgM antibody response to sRBC: N-nitrosodimethylamine (DMN), N-nitrosodiethylamine (DEN), N-nitrosodipropylamine (DPN), N-nitrosomethylethylamine (MEN), and N-nitrosomethylbutylamine (MBN).

IMMUNOTOXICITY

OF

sults pertaining to the relative potency of several of the congeners did not correlate between studies done in vivo and those done in vitro except to suggest that the in vitro system cannot completely substitute for what is mechanistically occurring in the whole animal. Weights of the spleen and thymus were also determined following nitrosamine treatment and clearly demonstrate that these two lymphoid organs are particularly sensitive to the low-molecular-weight dialkylnitrosamines. Despite the fact that all nitrosamine congeners significantly decreased the size and weight of these two lymphoid organs, with the exception of MEN which had no effect on thymus weight and DMN which had no effect on spleen weight, there was no apparent trends found between nitrosamine chemical structure and lymphoid organ weight. Although decreases in specific lymphoid organs are not necessarily indicative of a specific immunological target, in light of the marked suppression of humoral responses by the various nitrosamine congeners, these findings further suggest that the immune system is significantly compromised following exposure to short-chain dialkylnitrosamines. These differences in pathological lesions observed following treatment with the various congeners further suggest the role of either different metabolic intermediates or differential amounts of specific toxic intermediates being generated as a result of nitrosamine metabolism. Since many of the dialkylnitrosamines have been found to be hepatotoxic, histopathologic evaluation of liver tissues was performed as a secondary indicator of toxicity. These evaluations demonstrated a broad and diverse range of hepatic lesions. The most severe hepatic damage was observed with DMN at 6 mg/kg, MEN at 15 and 20 mg/kg, and DBN at 350 and 500 mg/kg. Exposure to all three congeners, at the doses indicated, resulted in focal cellular necrosis or centrilobular necrosis. The pathological lesions observed with the remaining congeners ranged

331

DIALKYLNITROSAMINES

from generalized mild hydropic degeneration to varying degrees of hepatitis. All congeners that were tested produced some degree of hepatic damage. However, the results with MBN suggest that immunosuppression and hepatotoxicity can be separated, because MBN was the most immunosuppressive and produced one of the mildest forms of liver damage. We have recently reported a similar relationship in a study of interaction between DMN and aminoacetonitrile (AAN), an inhibitor of DMN demethylase (Haggerty et al., 1988). Pretreatment with AAN blocked the severe liver damage but had no effect on suppression of the antibody response. In summary, two generalizations can be made about the relative immunotoxicity of short-chain dialkylnitrosamines: (1) those dialkylnitrosamine congeners that possess an N-methyl group are most immunotoxic; and (2) there exists an inverse relationship between aliphatic chain length and immunotoxic potency for dialkylnitrosamine congeners that possess symmetrical aliphatic chains. REFERENCES BARNES, J. M., AND MAGEE, P. N. (1954). Some toxic properties of dimethylnitrosamine. &it. J. Znd. Med. 11,167-174.

BARNES, J. M., AND MAGEE, P. N. ( 1967). Carcinogenic nitroso compounds. Adv. Cancer Res. 10,163-246. BARTLETT, M. S. (1937). Sub-sampling for attributes. J. R. Stat. Sot. Suppl. 4, 13 1. DUNNE’I-~, C. W. (1955). A multiple comparison procedure for comparing several treatments with a control. J. Amer. Stat. Assoc. 50,1096-l 121. Gehan-Wilcoxon Test. (1975). In Survival Distributions: Reliability Applications in the Biomedical Sciences (A. J. Gross and V. A. Clark, Eds.), pp. 102-123. Wiley, New York. HAGCERTY, H. G., BOISE, L. H., JORDAN, S. D., AND HOLSAPPLE, M. P. (1988). Differential effects of coadministration ofaminoacetonitrile on immunosuppression and hepatotoxicity by dimethylnitrosamine. J. Pharmacol. Exp. Ther., in press. HEATH, D. F. (1962). The decomposition and toxicity of dialkylnitrosamines. Biochem. J. 85,72-9 1. HOLSAPPLE, M. P., BICK, P. H., AND DUKE, S. S. ( 1985). Effects of N-nitrosodiethylamine on cell-mediated immunity. J. Leuko. Biol. 37,367-38 1.

332

KAMINSKI

HOLSAPPLE, M. P., MCNERNEY, P. J., TUCKER, A. N., AND WHITE, K. L., JR. (1984). Effects of N-nitrosodimethylamine on humoral immunity. J. Pharmacol. Exp. Ther. 229,493-500. JOHNSON, K. W., KIM, D. H., MUNSON, A. E., AND HOLSAPPLE, M. P. (1987a). Dependence on intact cells for in vitro activation of dimethylnitrosamine to an immunosuppressive form. Mutat. Res. 182, 211-221. JOHNSON, K. W., MUNSON, A. E., AND HOLSAPPLE, M. P. (1987b). Primary cellular target responsible for dimethylnitrosamine-induced immunosuppression in the mouse. Zmmunopharmacology 13,47-60. JOHNSON, K. W., MUNSON, A. E., KIM, D. H., AND HOLSAPPLE, M. P. (1987~). Role of reactive metabohtes in suppression of humoral immunity by N-nitrosodimethylamine. J. Pharmacol. Exp. Ther. 240,847855.

ET AL. KOHDA, K. H., MOCHIZUIU, M. ANJO, T., AND OKADA, M. ( 1982). Cytotoxicity and DNA damage and repair in the mouse L cells treated with a series of N-alkyl-N(acetoxymethyl)nitrosamines. Gann 73,565-573. LEE, S. Y ., AND GUTTENPLAN, J. B. (1981). A correlation between mutagenic and carcinogenic potencies in a diverse group of N-nitrosamines: Determination of mutagenic activities of weakly mutagenic N-nitrosamines. Carcinogenesis 2, 1339-l 344. LIJINSKY, W., AND TAYLOR, H. W. (1978). Relative carcinogenic effectivenessof derivatives of nitrosodiethylamine in rats. Cancer Res. 38,239 l-2394. WISHNOK, J. S., AND ARCHER, M. C. (1976). Structureactivity relationships in nitrosamine carcinogenicity. Brit. J. Cancer 33,307-3 11. WISHNOK, J. S., ARCHER, M. C., EDELMAN, A. S., AND RAND, W. M. (1978). Nitrosamine carcinogenicity: A quantitative Hansch-Taft structure-activity relationship. Chem.-Biol. Interact. 20,43-54.