- Email: [email protected]
Cytotoxic and morphological effects of phenylpropanolamine, caffeine, nicotine, and some of their metabolites studied In vitro
Toxic. in Vitro Vol. 6, No. 6, pp. 493-502, 1992 Printed in Great Britain
088%2333/92 $5.00+ 0.00 PergamonPress Ltd
CYTOTOXIC AND MORPHOLOGICAL EFFECTS OF PHENYLPROPANOLAMINE, CAFFEINE, NICOTINE, AND SOME OF THEIR METABOLITES STUDIED IN VITRO H. BABICH* and E. BORENFREUND~" *Stern College, Yeshiva University, Department of Biological Sciences, 245 Lexington Avenue, New York, NY 10016 and tThe Rockefeller University, Laboratory Animal Research Center, 1230 York Avenue, New York, NY 10021, USA (Received 4 September 1991; revisions received 14 April 1992)
Abstract--The neutral red cytotoxicity assay was used in vitro to evaluate the potencies of phenylpropanolamine (PPA), nicotine, caffeine, some of their metabolites, and related chemicals.The human cell types used as targets included fibroblast (HFF), melanoma (SK-Mel/27), and hepatoma (HepG2) cell lines and early passage endothelial (ENDO) cells and keratinocytes (NHEK). For all of these cells, nicotine was more cytotoxic than cotinine, its major metabolite; in turn, cotinine was more cytotoxic than chemically related compounds such as nicotinic acid and nicotinamide. Nicotine, but neither cotinine, nicotinic acid, nor nicotinamide, induced cytoplasmic vacuolization in all the cell types tested. Except for the ENDO cells, caffeine and its metabolite, theophylline, showed approximately equivalent cytotoxic potencies. However, for the ENDO cells, caffeinewas more cytotoxic than theophylline. Furthermore, the ENDO cells were 2-3 times more sensitive to caffeine and theophylline than were the other cell types. The HFF, SK-Mel/27, and HepG2 cells were more sensitive than the ENDO and NHEK cells to PPA. Phenylpropanolamine induced cytoplasmic vacuolization only in the ENDO cells. Combinations of caffeine + PPA interacted synergisticallyin their cytotoxicity towards the HepG2 ceils;a similar synergistic interaction was not noted with the ENDO cells.
INTRODUCTION
The past decade has seen the development of a plethora of cytotoxicity assays in vitro using mammalian cells as the bioindicators. Such bioassays have focused primarily on evaluating the potential toxicities of those industrial and environmental chemicals to which humans may be exposed. Less attention has been directed, however, to applying these cytotoxicity assays in vitro to evaluate the safety to human health of those chemicals frequently encountered in everyday activities. The intent of this research was to apply the neutral red cytotoxicity assay in vitro to assess the potencies of nicotine, caffeine and phenylpropanolamine (PPA) to a diverse spectrum of human cell types, both normal and malignant. Owing to the widespread daily exposure to and/or consumption of these chemicals, studies on their toxicological relevance to different human cell types is of importance. Phenylpropanolamine is commonly used in cough medicines, nasal decongestants and over-the-counter diet aids (Johnson, 1991). Caffeine intake comes mainly from coffee, tea, cola drinks, chocolate, and prescription and non-prescription drugs (Stavric, 1988a). Combinations of PPA and caffeine have been used in over-the-counter diet aids (Johnson, 1991). Abbreviations: BPE=bovine pituitary extract; EGF=
epidermal growth factor; FBS = foetal bovine serum; NR = neutral red; NR50= midpoint cytotoxicity value; PPA = phenylpropanolamine.
Also evaluated in this study was theophylline, which is both a metabolite of caffeine and itself a pharmaceutical used as a bronchodilator (Stavric, 1988b). Nicotine is a component of the tobacco leaf. Humans are exposed to nicotine both through the smoking of tobacco products and through smokeless tobacco (chewing tobacco and snuff) (USDHHS, 1986). Also evaluated were cotinine, the major metabolite of nicotine (Kyerematen and Vesell, 1991), and the chemically related agents, nicotinic acid (niacin) and nicotinamide (niacinamide). MATERIALS AND METHODS
Cell cultures. A human foreskin fibroblast cell line, designated HFF, the human hepatoma cell line HepG2, and the human melanoma cell line SKMel/27, were maintained in Dulbecco's minimum essential medium supplemented with 10% Serum Plus (Hazelton Research Products, Lenexa, KS, USA), 2% foetal bovine serum (FBS), 100 units penicillin G/mi, 100 #g streptomycin/ml and 2.5/~g/Fungizone (amphotericin sodium desoxycholate complex)/ ml. Cultures were passaged by dissociation with 0.05% trypsin :0.02% ethylenediaminetetraaceticacid (EDTA). Secondary cultures of human epidermal keratinocytes, designated NHEK, isolated from breast tissue, and obtained from Clonetics Corporation (San Diego, CA, USA), were maintained in serum-free, biochemically defined Keratinocyte Growth Medium. This medium is a modified MCBD
493
494
H. BABICHand E. BORENFREUNO
113 formulation, which was supplemented by the supplier (Clonetics) with 10ng epidermal growth factor (EGF)/ml, 5/~g insulin/ml, 0.5 #g hydrocortisone/ml, 0.15 mM-Ca 2+, 0.4% (v/v) bovine pituitary extract (BPE) and antibiotics (gentamicin and Fungizone). Secondary cultures of human endothelial cells, designated ENDO, isolated from the umbilical vein and obtained from Clonetics, were maintained in Endothelial Growth Medium. This medium is also a modified MCDB 113 formulation which was supplemented by the supplier (Clonetics) with 2% FBS, 10 ng EGF/ml, 1 ng hydrocortisone/ml, 0.4% BPE, and antibiotics. The N H E K and ENDO cells were passaged with 0.025% trypsin:0.01% EDTA. All cell cultures were maintained at 37°C in a humidified atmosphere with 5.5% CO2. Test agents. (_+)-Phenylpropanolamine hydrochloride, caffeine, theophylline, nicotinamide and nicotinic acid were solubilized in water, ( - ) - c o t i n i n e in 2.5% ethanol, and ( + ) nicotine in 50% methanol. Concentrations of ethanol (less than 1%) and methanol (less than 3%) in the exposure medium were at non-toxic levels. All chemicals were obtained from Sigma Chemical Co. (St Louis, MO, USA). Cell culture assay. The cytotoxicities of the test agents to the human cell cultures were determined with the neutral red (NR) assay. This assay, which quantifies the number of viable, uninjured cells after their exposure to test agents, is based on the uptake and subsequent lysosomal accumulation of the supravital dye NR. Quantitation of the dye extracted from the cells has been shown to be linear with cell numbers, both by direct cell counts and by protein determinations of cell populations. The principles of this assay have been extensively described elsewhere (Babich and Borenfreund, 1987; Borenfreund and Puerner, 1985). Individual wells of a 96-well tissue culture microtitre plate were inoculated with 0.2 ml of the appropriate medium containing a number of cells to provide approx. 70% confluence after 24-72 hr of incubation. The medium was then replaced with fresh medium, unamended (control) and amended with varied concentrations of the test agents. Six to eight wells were used per concentration of test agent. After 24hr of exposure, the medium was replaced with 0.2 ml of medium containing 4 0 # g NR/ml. This NR-containing medium had been preincubated overnight at 37°C and then centrifuged for 10 min at 1500 g to remove fine precipitates of dye crystals. The plate was returned to the incubator for another 3 hr to allow for the uptake of the vital dye into the
lysosomes of viable, uninjured cells. Thereafter, the medium was removed and the cells were washed quickly with a fixative (1% CAC12:0.5% formaldehyde), and then 0.2 ml of a solution of 1% acetic acid:50% ethanol was added to each well to extract the dye. After an additional 10 min at room temperature and rapid agitation on a microtitre plate shaker, the plate was transferred to a microtitre plate reader to measure the absorbance of the extracted dye at 540 nm. All experiments were performed at least five times and cytotoxicity curves were constructed with the individual data points presented as the arithmetic mean percentage of the control; values for the standard error of the mean were typically less than 5%. The relative sensitivities of the different cell types to the various agents were compared by computing the concentration of test agent needed to reduce the absorbance of N R by 50% (NRs0) as compared with controls. RESULTS
Table 1 lists the midpoint cytotoxicity (NRs0) values of nicotine and related chemicals for the five cell types tested. Based on NRs0 values, nicotine was about twice as cytotoxic as its metabolite, cotinine, to the NHEK, H F F , ENDO and SK-Mel/27 cells and three times as cytotoxic to the HepG2 cells. Niacinamide (nicotinamide) and nicotinic acid (niacin) were less potent than cotinine or nicotine. The ENDO cells were more sensitive to niacinamide and to nicotinic acid than the other cell types. A different aspect of the cellular response of the cells to these chemicals can be noted from the concentration-response cytotoxicity curves. For example, based on the concentration of test agent causing initial toxicity (i.e. approximately an NRg0), the N H E K cells showed initial sensitivity to nicotine at only 0.25 mM, whereas with the other cell types, initial cytotoxicity occurred at 5-mM-nicotine or greater. More striking was the response of the ENDO and H F F cells to subtoxic concentrations of nicotine. Absorbance measurements of NR above those for controls in the cytotoxicity assay were noted for the ENDO cells exposed to 4-10mM-nicotine, with a peak at 6 mM-nicotine (Fig. i), and for the H F F cells exposed to 5-15 mM-nicotine, with a peak at 10 taranicotine. Such elevated uptake and retention of N R were not noted for the HepG2, SK-Mel/27, and N H E K cells at these subtoxic levels of nicotine (Fig. 1). Similarly, increased NR uptake readings
Table 1. Response of various human cell types to nicotine and related chemicals in the neutral red (NR) cytotoxicity assay Midpoint cytotoxicity (NRs0) values (mM) Test agent Nicotine Cotinine Nicotinic acid Niacinamide
ENDO
NHEK
HFF
SK-Mel/27
HepG2
23 42 52 50
22 48 70 80
28 62 80 88
20 44 90 70
14 42 82 65
Neutral red assay: nicotine and pharmaceuticals
495
150 140 / 0 Iz I.z 0 (_) Lt. 0 F-
z w
130 120 110 100 90
0
n-. i,t (3-
80
t~
70
w
60
0 Z < m nO
t,q m ,(
50 40 30 20
10 0
i
i
i
4O
20 CONCENTRATION IN m M
Fig. 1. Concentration-response cytotoxicity curves obtained for a 24-hr exposure to nicotine as determined with the neutral red assay using cultures of various human cell types. Each point represents the arithmetic mean percentage of the control; values for the standard errors of the mean were less than + 5 % . x ( E N D O ) = e n d o t h e l i a l cells; IIl(HFF)---fibroblasts; O(NHEK)=keratinocytes; + ( S K - M e l / 2 7 ) = melanoma cells; A ( H e p G 2 ) = hepatoma cells.
were not noted for exposures of all cell types to cotinine, nicotinic acid and niacinamide. Marked cytoplasmic vacuolization was observed in all cell types exposed to nicotine, but not to cotinine, nicotinic acid and niacinamide. However, the pattern of nicotine-induced cytoplasmic vacuolization varied among the different cell types. Vacuolization was most extensive in the ENDO and HFF cells exposed to nicotine, at 10-15 mM. In these cells, the enlarged vacuoles filled the entire cytoplasm, pushing the nucleus to the periphery of the cell (Plate la,b). Vacuolization was reversed on exposure of the cells to fresh medium without nicotine. For example, the extensive vacuolization noted in the ENDO cells exposed to 10-15 mM-nicotine was gradually lessened and then reversed after 48-72 hr in medium free of nicotine. The pattern of nicotine-induced cytoplasmic vacuolization was different with the SK-Mel/27 and HepG2 cells: in these cells the vacuoles occupied less of the cytoplasm and were not as large, as noted in the E N D O and HFF cells (Plate lc,d). A third
pattern of nicotine-induced vacuolization was noted with the N H E K cells: the nicotine-treated N H E K cells were generally enlarged and contained swollen vacuoles surrounding a centrally located nucleus. A distinct membrane encircled the entire cluster of vacuoles (Plate le,f). Based on NRs0 values, caffeine and theophylline exhibited similar toxicities to the N H E K keratinocytes, the HFF fibroblasts, the SK-Mel/27 melanoma cells, and the HepG2 hepatoma cells (Table 2). The ENDO cells, however, were more sensitive to both test agents, with caffeine approximately twice as toxic as theophylline (Figs 2 and 3). With regard to PPA, the HFF, SK-Mel/27 and HepG2 cells were about twice as sensitive as the ENDO and N H E K cells (Table 2). PPA, at subtoxic concentrations from 1 to 8 mM, yielded a consistently higher uptake and retention of N R in the ENDO cells, although such increases were not concentration dependent (Fig. 4). Light microscopy of these cells confirmed that this
Table 2. Response of various human cell types to caffeine, theophylline, and phenylpropanolamine in the neutral red (NR) cytotoxicity assay Test agent Caffeine Theophylline Phenylpropanolamine
ENDO 5 9 11
Midpoint cytotoxicity (NRso) value (mM) NHEK HFF SK-M¢I/27 HepG2 18 19 12 15 18 12
20 6
18 6
19 5
496
H. BABICHand E. BORENFREUND
I 100
90 1 8O
70
60
5O
40
-
50
-
20
-
10
-
"-O
\
0
i
I
J
1-
40
2O
0
CONCENTRATION IN m M
Fig. 2. Concentration-response cytotoxicity curves obtained for a 24-hr exposure to caffeineas determined with the neutral red assay using cultures of various human cell types. Each point represents the arithmetic mean percentage of the control; values for the standard errors of the mean were less than + 5%. Symbols as in Fig. 1. increased uptake was related to enhanced vacuolization, as was evident from the presence of enlarged vacuoles evenly distributed within the cytoplasm (not shown). Neither enhanced uptake of NR nor vacuolization was noted in the other four cell types upon their exposures to subtoxic levels of PPA (Fig. 4). A non-toxic concentration of PPA potentiated the cytotoxicity to HepG2 cells of concentrations of caffeine. Thus, the cytotoxicity of 15 and 20 mMcaffeine (Fig. 5), but not of 5 or 10 mM-caffeine, was enhanced on simultaneous exposure of the HepG2 cells to 2.5 mM-PPA. However, the effect of cytotoxic levels of caffeine (0.5-5 mM) to ENDO cells (Fig. 2) was not potentiated by simultaneous exposure to 8 mM-PPA (not shown). DISCUSSION
In humans, nicotine is rapidly and extensively metabolized in the liver and, to a lesser extent, in the lung and kidney. Cotinine, a major metabolite of nicotine, is pharmacologically inactive (USDHHS,
1986). The greater acute cytotoxicity of nicotine than its metabolite, cotinine, as noted with the NR assay in vitro was therefore consistent with studies of acute toxicity in vivo. For example, the LDs0 values for nicotine (Sofia and Knobloch, 1974) and cotinine (Borzelleca et al., 1962) administered to mice by the ip route were 5.9 and 930 mg/kg, respectively. Studies in vitro with the human promyelocytic leukaemia cell line, HL-60, have attributed the cytotoxic effects of nicotine to blockage of the G1 phase of the cell cycle, with a concomitant inhibition of DNA synthesis during the S phase (Konno et al., 1986). Several studies in vitro have focused on the induction of cytoplasmic vacuoles by nicotine. The formation of such swollen vacuoles was described in fibroblasts (Chamson et al., 1980; Raulin et al., 1988), peritoneal macrophages (Ohkuma and Poole, 1981; Thyberg and Nilsson, 1982; Thyberg et al., 1983), and a neuroendocrine cell line derived from a lung carcinoma (Schuller and Hegedus, 1989). The enlarged vacuoles are thought to be swollen lysosomes (Thyberg and Nilsson, 1982) and their induction is
Plate 1. (a) Control, human endothelial cells (ENDO). (b) Endothelial cells incubated for 24 hr with 10 mM-nicotine.Note the laterally displaced nuclei and the large vacuoles evenly distributed throughout the cytoplasm. (c) Control, human melanoma cells (SK-Mel/27). (d) Melanoma ceils incubated for 24 hr with 10 raM-nicotine. Note the very small vacuoles in the cytoplasm. (e) Control, human keratinocytes (NHEK). (f) Keratinocytes incubated for 24hr with 10mM-nicotine. Note the enlarged cells with cytoplasmic vacuoles surrounding the nuclei; vacuoles are encircled by a membrane-like structure. Giemsa stain, x210.
Neutral red assay: nicotine and pharmaceuticals
499
100
901
o or i.z O
80-
o h O
70
I-
60
z h.I O or W 0. ~q
50
40 W
L) Z m or O I/1 m
30 20
10
A 0
I
I
I
I
4O
20
0
CONCENTRATION IN mM
Fig. 3. Concentration-response cytotoxicity curves obtained for a 24-hr exposure to theophylline as determined with the neutral red assay using cultures o f various h u m a n cell types. Each point represents the arithmetic m e a n percentage o f the control; values for the standard errors o f the mean were less than + 5 % . Symbols as in Fig. 1.
120 _J
0 n,,
110
I-
100
o
90
z 0
b.
O I-
z W o or bJ ft. 6'1 < hJ
o z< m n, O 6'1 03
80 70 6O 50 40 3O 20 10 0
c
l
2
,
,
4
i
,
6
,
,
8
i
I
10
,
I
12
,
i
14
,
l
16
CONCENTRATION IN mM
Fig. 4. Concentration-response cytotoxicity curves obtained for a 24-hr exposure to phenylpropanolamine as determined with the neutral red assay using cultures of various h u m a n cell types. Each point represents the arithmetic m e a n percentage of the control; values for the standard errors o f the mean were less than + 5 % . Symbols as in Fig. 1.
500
H. BABICHand E. BORENFREUND
100
\ \ \ x
\ \ x \
03 (/1
90
13
80
,<
I..d rY
x x x \
._1
.< n,"
70
I--
x x x \ x x x x x x x x
D
I..d
Z
60
x x x x x x x x x x x \ x x x x x x x x
..J
0 rY
x x x \
50
I-
Z 0 0
40
b_
0 I--
30
Z
b,.I
tO IZ
20
I.d
n
10
\ x ~ x x x ~ x \ \ x x \ x x \ \ x x \ x x x \ x x x x x x x \ \ x x \ x x x x XNNN \ \ \ x x x \ \ \ N \ N x x x x \ \ x N XNNN \ x x x \ x x \ \ \ X \ \ \ N N \NNN
x x x x x x x x
x x x \ x x x x
xxxx x x x x \ N N \ %NX\ x x \ x NNN\ NNNN NNNN N~N\ XXX\
*\N
o
N N
x x x \ % x x \
0
,x,\\"~
15
N 20
0
15
20
CAFFEINE ( r n M )
Fig. 5. Cytotoxicity of caffeine (15 and 20mM) in the absence ([]) and presence ( I ) of phenylpropanolamine (2.5 mM) to HepG2 hepatoma ceils. The data are expressed as the arithmetic mean percentage of the control (medium containing neither caffeine nor phenylpropanolamine); values for the standard errors of the mean were less than + 5%. not unique to nicotine, but is common to weakly basic amines (Ohkuma and Poole, 1981; Valtolina and Foster, 1991). Presumably, the plasma and lysosomal membranes are highly permeable to the neutral form of weak bases, but are only slightly permeable or impermeable to the protonated forms of weak bases. The pH inside the lysosomes is considerably lower than that of the cytoplasm, so that weakly basic amines--such as nicotine--are trapped by protonation inside the lysosomes. Such protonated bases accumulate and thereby increase the osmotic pressure inside the lysosomes, causing water to enter osmotically and to swell the lysosomes (Ohkuma and Poole, 1981). Ohkuma and Poole (1981), in their study of chemical-induced vacuolization in mouse peritoneal cells, observed that although many amines caused vacuole formation, exceptions were noted. Amines with pK, below neutrality and those that are relatively hydrophilic, even in their neutral forms, failed to induce vacuolization. The pK values of nicotine are 3.1 and 8.0, the latter value accounting for its protonation and trapping inside the lysosomes. Cotinine, although also an amine, failed to induce vacuolization in any of the five human cell types. Cotinine has a pK, of 4.5 (Beckett et al., 1972) and is more polar and less lipophilic than nicotine (Benowitz et al., 1983), factors accounting for its inability to induce vacuolization.
Nicotine-induced vacuolization was neither a permanent cellular alteration nor an indicator of cell death. Thus, endothelial cells exposed to 10-15 mM-nicotine developed swollen vacuoles, which gradually disappeared after 48-72hr incubation in fresh, nicotine-free medium. Hanes et al. (1990) showed that nicotine binds non-specifically to human fibroblasts, is rapidly taken up, accumulating intracellularly, but is then slowly released. Apparently, in the studies reported herein with the endothelial cells, the nicotine was released slowly from the lysosomes over a period of 2-3 days, thereby reducing the osmotic pressure within the lysosomes, resulting in the return of the lysosomes to their normal size. Nicotine-induced vacuolization exhibited different patterns among the various human cell types. With the malignant cell types (i.e. SK-Mel/27 melanoma and HepG2 hepatoma cells) the swollen vacuoles were fewer and smaller than in the nicotine-treated normal cells. With the N H E K keratinocytes, the swollen vacuoles encircled a centrally located nucleus and a distinct membranous structure surrounded the entire group of lysosomes. This discrete membranous structure encircling the swollen vacuoles was similar to a structure, identified as an autophagic vacuole, seen in nicotine-treated macrophages (Thyberg et aL, 1983). In those studies such autophagic vacuoles consisted of groups of lysosomes enclosed by a large
Neutral red assay: nicotine and pharmaceuticals lysosome and were associated with nicotine inhibition of endocytosis. The third pattern of vacuolization occurred in the nicotine-treated fibroblasts and endothelial cells. A 24-hr exposure to 10-15 mM-nicotine induced highly swollen vacuoles that filled the entire cytoplasm, pushing the nucleus to the periphery of the cell. This pattern of nicotine-induced reversible vacuolization apparently accounted for the increased uptake and retention of NR when these cells were exposed to neutral red (150 #M) for 3 hr during the cytotoxicity assay. Phenylpropanolamine, at non-toxic levels, induced enlarged vacuoles in the endothelial cells, but not in the other cell types tested. Such vacuolization may have accounted for the increased uptake and retention of N R noted in endothelial cells exposed to subtoxic levels of PPA. The reason for the failure of the induction of vacuoles in the other cell types exposed to non-toxic levels of PPA is not known. Caffeine and its metabolite, theophylline, exhibited approximately equivalent cytotoxic potencies to the NHEK, HFF, SK-Mel/27 and HepG2 cells. However, the endothelial cells were 2-3 times more sensitive to caffeine and about twice as sensitive to theophylline as the other cell types. The cytotoxic effects of caffeine and theophylline to mammalian cells in vitro has been attributed to adverse effects on DNA synthesis (Lehmann and Kirk-Bell, 1974; Weinstein et al., 1975). Metabolism of caffeine in vivo produces a number of intermediates, in addition to theophylline, which may exert different toxicities (Stavric, 1988a). Borenfreund and Puerner (1987), using BALB/c 3T3 mouse fibroblasts as the indicator in the NR assay, showed that the cytotoxicity of 0.5 mM-caffeine was reduced by co-incubation with an hepatic S-9 microsomal fraction from Aroclorinduced rats. The synergistic cytotoxic interaction between caffeine and PPA towards the HepG2 hepatoma cells is interesting, as acute toxicity studies with rats have shown that caffeine potentiated the lethality of PPA (Johnson, 1991). The studies presented herein demonstrate the utility of cytotoxicity assays in vitro, such as the NR assay, to evaluate the toxic potencies of test agents, including those chemicals to which large segments of the population may be exposed. Such assays in vitro are most suitable for the initial screening (tier I testing) of chemicals, as they provide preliminary data on the relative toxicities of test agents, are cost-effective, and reduce the need for extensive animal experimentation in the early stages of toxicity exploration. Furthermore, these studies in vitro can serve as the basis for more detailed analyses, using biotic test systems of increasing complexity. The findings in vitro of the induction of cytoplasmic vacuolization in nicotine-exposed cells and of the hypersensitivity of endothelial cells to caffeine exemplify such studies. TIV 6/6~B
501
Acknowledgements--ApprecAation is expressed of the assistance provided by Irving Bruck, to Clonetics Corporation for supplying the keratinocytes, the endothelial cells, and the KGM and EGM media, and to Schering-Plough (Bloomfield, NJ, USA) for financial support. REFERENCES
Babich H. and Borenfreund E. (1987) Structure-activity relationship (SAR) models established in vitro with the neutral red cytotoxicity assay. Toxicology in Vitro 1, 3-9. Beckett A. H., Gorrod J. W. and Jenner P. (1972) A possible relation between pKa, lipid solubility, and the amounts excreted in urine of some tobacco alkaloids given to man. Journal of Pharmacy and Pharmacology 24, 115-120. Benowitz N. L., Kuyt F., Jacob P., Jones R. T. and Osman A.-L. (1983) Cotinine disposition and effects. Clinical Pharmacology and Therapeutics 34, 604-611. Borenfreund E. and Puerner J. A. (1985) Toxicity determined in vitro by morphological alteration and neutral red absorption. Toxicology Letters 24, 119-124. Borenfreund E. and Puerner J. A. (1987) Short-term quantitative in vitro cytotoxicity assay involving an S-9 activating system. Cancer Letters 34, 243-248. Borzelleca J. F., Bowman E. R. and McKennis H. (1962) Studies on the respiratory and cardiovascular effects of (-)-cotinine. Journal of Pharmacology and Experimental Therapeutics 137, 313-318. Chamson A., Garrone R., Auger C. and Frey J. (1980) Effects of tobacco smoke extracts on the ultrastructure of fibroblasts in culture. Journal of Submicroscopic Cytology 12, 401--406. Hanes P. J., Schuster G. S. and Lubas S. (1990) Binding, uptake, and release of nicotine by human gingival fibroblasts. Journal of Periodontology 62, 147-152. Johnson D. A. (1991) Pharmacology and safety of phenylpropanolamine. Drug Development Research 22, 197-207. Konno S., Chiao J. W. and Wu J. M. (1986) Effects of nicotine on cellular proliferation, cell cycle phase distribution and macromolecular synthesis in human promyelocytic HL-60 leukemia cells. Cancer Letters 33, 91-97. Kyerematen G. A. and Vesell E. S. (1991) Metabolism of nicotine. Drug Metabolism Reviews 23, 3-41. Lehmann A. R. and Kirk-Bell S. (1974) Effects of caffeine and theophylline on DNA synthesis in unirradiated and UV-irradiated mammalian cells. Mutation Research 26, 73-82. Ohkuma S. and Poole B. (I 981) Cytoplasmic vacuolation of mouse peritoneal macrophages and the uptake into lysosomes of weakly basic substances. Journal of Cell Biology 90, 656--664. Raulin L. A., McPherson J. C., McQuade M. J. and Hanson B. S. (1988) The effect of nicotine on the attachment of human fibroblasts to glass and human root surfaces in vitro. Journal of Periodontology 59, 308-315. Schuller H. M. and Hegedus T. J. (1989) Effects of endogenous and tobacco-related amines and nitrosamines on cell growth and morphology of a cell line derived from a human neuroendocrine lung cancer. Toxicology in Vitro 3, 37-43. Sofia R. D. and Knobloch L. C. (1974) Influence of acute pretreatment with delta9-tetrahydrocannabinol on the LD50 of various substances that alter neurohumoral transmission. Toxicology and Applied Pharmacology 28, 227-234. Stavric B. (1988a) Methylxanthines: toxicity to humans. 2. Caffeine. Food and Chemical Toxicology 26, 645-662. Stavric B. (1988b) Methylxanthines: toxicity to humans. 1. Theophylline. Foodand Chemical Toxicology 26, 541-565. Thyberg J., Hedin U., Stenseth K. and Nilsson J. (1983) Effects of nicotine on the fine structure of cultivated mouse peritoneal macrophages. Acta Pathologica,
502
H. BABICH and E. BORENFREUND
Microbiologica et lmmunologica Scandinavica Section A, 91, 23-30. Thyberg J. and Nilsson J. (1982) Effects of nicotine on endocytosis and intracellular degradation of horseradish peroxidase in cultivated mouse peritoneal macrophages. Acta Pathologica, Microbiologica et Scandinavica 90, 305-310. United States Department of Health and Human Services (USDHHS) (1986) The Health Consequences of Using
Smokeless Tobacco. DHHS Publ. No. (NIH) 86-2874. Public Health Service, Bethesda, MD. Valtolina M. and Foster R. (1991) Induction of cytoplasmic vacuolation by two alkyl-diamine compounds. A TLA 19, 30-38. Weinstein D., Mauer I., Katz M. L. and Kazmer S. (1975) The effect of methylxanthines on chromosomes of human lymphocytes in culture. Mutation Research 31, 57~I.
088%2333/92 $5.00+ 0.00 PergamonPress Ltd
CYTOTOXIC AND MORPHOLOGICAL EFFECTS OF PHENYLPROPANOLAMINE, CAFFEINE, NICOTINE, AND SOME OF THEIR METABOLITES STUDIED IN VITRO H. BABICH* and E. BORENFREUND~" *Stern College, Yeshiva University, Department of Biological Sciences, 245 Lexington Avenue, New York, NY 10016 and tThe Rockefeller University, Laboratory Animal Research Center, 1230 York Avenue, New York, NY 10021, USA (Received 4 September 1991; revisions received 14 April 1992)
Abstract--The neutral red cytotoxicity assay was used in vitro to evaluate the potencies of phenylpropanolamine (PPA), nicotine, caffeine, some of their metabolites, and related chemicals.The human cell types used as targets included fibroblast (HFF), melanoma (SK-Mel/27), and hepatoma (HepG2) cell lines and early passage endothelial (ENDO) cells and keratinocytes (NHEK). For all of these cells, nicotine was more cytotoxic than cotinine, its major metabolite; in turn, cotinine was more cytotoxic than chemically related compounds such as nicotinic acid and nicotinamide. Nicotine, but neither cotinine, nicotinic acid, nor nicotinamide, induced cytoplasmic vacuolization in all the cell types tested. Except for the ENDO cells, caffeine and its metabolite, theophylline, showed approximately equivalent cytotoxic potencies. However, for the ENDO cells, caffeinewas more cytotoxic than theophylline. Furthermore, the ENDO cells were 2-3 times more sensitive to caffeine and theophylline than were the other cell types. The HFF, SK-Mel/27, and HepG2 cells were more sensitive than the ENDO and NHEK cells to PPA. Phenylpropanolamine induced cytoplasmic vacuolization only in the ENDO cells. Combinations of caffeine + PPA interacted synergisticallyin their cytotoxicity towards the HepG2 ceils;a similar synergistic interaction was not noted with the ENDO cells.
INTRODUCTION
The past decade has seen the development of a plethora of cytotoxicity assays in vitro using mammalian cells as the bioindicators. Such bioassays have focused primarily on evaluating the potential toxicities of those industrial and environmental chemicals to which humans may be exposed. Less attention has been directed, however, to applying these cytotoxicity assays in vitro to evaluate the safety to human health of those chemicals frequently encountered in everyday activities. The intent of this research was to apply the neutral red cytotoxicity assay in vitro to assess the potencies of nicotine, caffeine and phenylpropanolamine (PPA) to a diverse spectrum of human cell types, both normal and malignant. Owing to the widespread daily exposure to and/or consumption of these chemicals, studies on their toxicological relevance to different human cell types is of importance. Phenylpropanolamine is commonly used in cough medicines, nasal decongestants and over-the-counter diet aids (Johnson, 1991). Caffeine intake comes mainly from coffee, tea, cola drinks, chocolate, and prescription and non-prescription drugs (Stavric, 1988a). Combinations of PPA and caffeine have been used in over-the-counter diet aids (Johnson, 1991). Abbreviations: BPE=bovine pituitary extract; EGF=
epidermal growth factor; FBS = foetal bovine serum; NR = neutral red; NR50= midpoint cytotoxicity value; PPA = phenylpropanolamine.
Also evaluated in this study was theophylline, which is both a metabolite of caffeine and itself a pharmaceutical used as a bronchodilator (Stavric, 1988b). Nicotine is a component of the tobacco leaf. Humans are exposed to nicotine both through the smoking of tobacco products and through smokeless tobacco (chewing tobacco and snuff) (USDHHS, 1986). Also evaluated were cotinine, the major metabolite of nicotine (Kyerematen and Vesell, 1991), and the chemically related agents, nicotinic acid (niacin) and nicotinamide (niacinamide). MATERIALS AND METHODS
Cell cultures. A human foreskin fibroblast cell line, designated HFF, the human hepatoma cell line HepG2, and the human melanoma cell line SKMel/27, were maintained in Dulbecco's minimum essential medium supplemented with 10% Serum Plus (Hazelton Research Products, Lenexa, KS, USA), 2% foetal bovine serum (FBS), 100 units penicillin G/mi, 100 #g streptomycin/ml and 2.5/~g/Fungizone (amphotericin sodium desoxycholate complex)/ ml. Cultures were passaged by dissociation with 0.05% trypsin :0.02% ethylenediaminetetraaceticacid (EDTA). Secondary cultures of human epidermal keratinocytes, designated NHEK, isolated from breast tissue, and obtained from Clonetics Corporation (San Diego, CA, USA), were maintained in serum-free, biochemically defined Keratinocyte Growth Medium. This medium is a modified MCBD
493
494
H. BABICHand E. BORENFREUNO
113 formulation, which was supplemented by the supplier (Clonetics) with 10ng epidermal growth factor (EGF)/ml, 5/~g insulin/ml, 0.5 #g hydrocortisone/ml, 0.15 mM-Ca 2+, 0.4% (v/v) bovine pituitary extract (BPE) and antibiotics (gentamicin and Fungizone). Secondary cultures of human endothelial cells, designated ENDO, isolated from the umbilical vein and obtained from Clonetics, were maintained in Endothelial Growth Medium. This medium is also a modified MCDB 113 formulation which was supplemented by the supplier (Clonetics) with 2% FBS, 10 ng EGF/ml, 1 ng hydrocortisone/ml, 0.4% BPE, and antibiotics. The N H E K and ENDO cells were passaged with 0.025% trypsin:0.01% EDTA. All cell cultures were maintained at 37°C in a humidified atmosphere with 5.5% CO2. Test agents. (_+)-Phenylpropanolamine hydrochloride, caffeine, theophylline, nicotinamide and nicotinic acid were solubilized in water, ( - ) - c o t i n i n e in 2.5% ethanol, and ( + ) nicotine in 50% methanol. Concentrations of ethanol (less than 1%) and methanol (less than 3%) in the exposure medium were at non-toxic levels. All chemicals were obtained from Sigma Chemical Co. (St Louis, MO, USA). Cell culture assay. The cytotoxicities of the test agents to the human cell cultures were determined with the neutral red (NR) assay. This assay, which quantifies the number of viable, uninjured cells after their exposure to test agents, is based on the uptake and subsequent lysosomal accumulation of the supravital dye NR. Quantitation of the dye extracted from the cells has been shown to be linear with cell numbers, both by direct cell counts and by protein determinations of cell populations. The principles of this assay have been extensively described elsewhere (Babich and Borenfreund, 1987; Borenfreund and Puerner, 1985). Individual wells of a 96-well tissue culture microtitre plate were inoculated with 0.2 ml of the appropriate medium containing a number of cells to provide approx. 70% confluence after 24-72 hr of incubation. The medium was then replaced with fresh medium, unamended (control) and amended with varied concentrations of the test agents. Six to eight wells were used per concentration of test agent. After 24hr of exposure, the medium was replaced with 0.2 ml of medium containing 4 0 # g NR/ml. This NR-containing medium had been preincubated overnight at 37°C and then centrifuged for 10 min at 1500 g to remove fine precipitates of dye crystals. The plate was returned to the incubator for another 3 hr to allow for the uptake of the vital dye into the
lysosomes of viable, uninjured cells. Thereafter, the medium was removed and the cells were washed quickly with a fixative (1% CAC12:0.5% formaldehyde), and then 0.2 ml of a solution of 1% acetic acid:50% ethanol was added to each well to extract the dye. After an additional 10 min at room temperature and rapid agitation on a microtitre plate shaker, the plate was transferred to a microtitre plate reader to measure the absorbance of the extracted dye at 540 nm. All experiments were performed at least five times and cytotoxicity curves were constructed with the individual data points presented as the arithmetic mean percentage of the control; values for the standard error of the mean were typically less than 5%. The relative sensitivities of the different cell types to the various agents were compared by computing the concentration of test agent needed to reduce the absorbance of N R by 50% (NRs0) as compared with controls. RESULTS
Table 1 lists the midpoint cytotoxicity (NRs0) values of nicotine and related chemicals for the five cell types tested. Based on NRs0 values, nicotine was about twice as cytotoxic as its metabolite, cotinine, to the NHEK, H F F , ENDO and SK-Mel/27 cells and three times as cytotoxic to the HepG2 cells. Niacinamide (nicotinamide) and nicotinic acid (niacin) were less potent than cotinine or nicotine. The ENDO cells were more sensitive to niacinamide and to nicotinic acid than the other cell types. A different aspect of the cellular response of the cells to these chemicals can be noted from the concentration-response cytotoxicity curves. For example, based on the concentration of test agent causing initial toxicity (i.e. approximately an NRg0), the N H E K cells showed initial sensitivity to nicotine at only 0.25 mM, whereas with the other cell types, initial cytotoxicity occurred at 5-mM-nicotine or greater. More striking was the response of the ENDO and H F F cells to subtoxic concentrations of nicotine. Absorbance measurements of NR above those for controls in the cytotoxicity assay were noted for the ENDO cells exposed to 4-10mM-nicotine, with a peak at 6 mM-nicotine (Fig. i), and for the H F F cells exposed to 5-15 mM-nicotine, with a peak at 10 taranicotine. Such elevated uptake and retention of N R were not noted for the HepG2, SK-Mel/27, and N H E K cells at these subtoxic levels of nicotine (Fig. 1). Similarly, increased NR uptake readings
Table 1. Response of various human cell types to nicotine and related chemicals in the neutral red (NR) cytotoxicity assay Midpoint cytotoxicity (NRs0) values (mM) Test agent Nicotine Cotinine Nicotinic acid Niacinamide
ENDO
NHEK
HFF
SK-Mel/27
HepG2
23 42 52 50
22 48 70 80
28 62 80 88
20 44 90 70
14 42 82 65
Neutral red assay: nicotine and pharmaceuticals
495
150 140 / 0 Iz I.z 0 (_) Lt. 0 F-
z w
130 120 110 100 90
0
n-. i,t (3-
80
t~
70
w
60
0 Z < m nO
t,q m ,(
50 40 30 20
10 0
i
i
i
4O
20 CONCENTRATION IN m M
Fig. 1. Concentration-response cytotoxicity curves obtained for a 24-hr exposure to nicotine as determined with the neutral red assay using cultures of various human cell types. Each point represents the arithmetic mean percentage of the control; values for the standard errors of the mean were less than + 5 % . x ( E N D O ) = e n d o t h e l i a l cells; IIl(HFF)---fibroblasts; O(NHEK)=keratinocytes; + ( S K - M e l / 2 7 ) = melanoma cells; A ( H e p G 2 ) = hepatoma cells.
were not noted for exposures of all cell types to cotinine, nicotinic acid and niacinamide. Marked cytoplasmic vacuolization was observed in all cell types exposed to nicotine, but not to cotinine, nicotinic acid and niacinamide. However, the pattern of nicotine-induced cytoplasmic vacuolization varied among the different cell types. Vacuolization was most extensive in the ENDO and HFF cells exposed to nicotine, at 10-15 mM. In these cells, the enlarged vacuoles filled the entire cytoplasm, pushing the nucleus to the periphery of the cell (Plate la,b). Vacuolization was reversed on exposure of the cells to fresh medium without nicotine. For example, the extensive vacuolization noted in the ENDO cells exposed to 10-15 mM-nicotine was gradually lessened and then reversed after 48-72 hr in medium free of nicotine. The pattern of nicotine-induced cytoplasmic vacuolization was different with the SK-Mel/27 and HepG2 cells: in these cells the vacuoles occupied less of the cytoplasm and were not as large, as noted in the E N D O and HFF cells (Plate lc,d). A third
pattern of nicotine-induced vacuolization was noted with the N H E K cells: the nicotine-treated N H E K cells were generally enlarged and contained swollen vacuoles surrounding a centrally located nucleus. A distinct membrane encircled the entire cluster of vacuoles (Plate le,f). Based on NRs0 values, caffeine and theophylline exhibited similar toxicities to the N H E K keratinocytes, the HFF fibroblasts, the SK-Mel/27 melanoma cells, and the HepG2 hepatoma cells (Table 2). The ENDO cells, however, were more sensitive to both test agents, with caffeine approximately twice as toxic as theophylline (Figs 2 and 3). With regard to PPA, the HFF, SK-Mel/27 and HepG2 cells were about twice as sensitive as the ENDO and N H E K cells (Table 2). PPA, at subtoxic concentrations from 1 to 8 mM, yielded a consistently higher uptake and retention of N R in the ENDO cells, although such increases were not concentration dependent (Fig. 4). Light microscopy of these cells confirmed that this
Table 2. Response of various human cell types to caffeine, theophylline, and phenylpropanolamine in the neutral red (NR) cytotoxicity assay Test agent Caffeine Theophylline Phenylpropanolamine
ENDO 5 9 11
Midpoint cytotoxicity (NRso) value (mM) NHEK HFF SK-M¢I/27 HepG2 18 19 12 15 18 12
20 6
18 6
19 5
496
H. BABICHand E. BORENFREUND
I 100
90 1 8O
70
60
5O
40
-
50
-
20
-
10
-
"-O
\
0
i
I
J
1-
40
2O
0
CONCENTRATION IN m M
Fig. 2. Concentration-response cytotoxicity curves obtained for a 24-hr exposure to caffeineas determined with the neutral red assay using cultures of various human cell types. Each point represents the arithmetic mean percentage of the control; values for the standard errors of the mean were less than + 5%. Symbols as in Fig. 1. increased uptake was related to enhanced vacuolization, as was evident from the presence of enlarged vacuoles evenly distributed within the cytoplasm (not shown). Neither enhanced uptake of NR nor vacuolization was noted in the other four cell types upon their exposures to subtoxic levels of PPA (Fig. 4). A non-toxic concentration of PPA potentiated the cytotoxicity to HepG2 cells of concentrations of caffeine. Thus, the cytotoxicity of 15 and 20 mMcaffeine (Fig. 5), but not of 5 or 10 mM-caffeine, was enhanced on simultaneous exposure of the HepG2 cells to 2.5 mM-PPA. However, the effect of cytotoxic levels of caffeine (0.5-5 mM) to ENDO cells (Fig. 2) was not potentiated by simultaneous exposure to 8 mM-PPA (not shown). DISCUSSION
In humans, nicotine is rapidly and extensively metabolized in the liver and, to a lesser extent, in the lung and kidney. Cotinine, a major metabolite of nicotine, is pharmacologically inactive (USDHHS,
1986). The greater acute cytotoxicity of nicotine than its metabolite, cotinine, as noted with the NR assay in vitro was therefore consistent with studies of acute toxicity in vivo. For example, the LDs0 values for nicotine (Sofia and Knobloch, 1974) and cotinine (Borzelleca et al., 1962) administered to mice by the ip route were 5.9 and 930 mg/kg, respectively. Studies in vitro with the human promyelocytic leukaemia cell line, HL-60, have attributed the cytotoxic effects of nicotine to blockage of the G1 phase of the cell cycle, with a concomitant inhibition of DNA synthesis during the S phase (Konno et al., 1986). Several studies in vitro have focused on the induction of cytoplasmic vacuoles by nicotine. The formation of such swollen vacuoles was described in fibroblasts (Chamson et al., 1980; Raulin et al., 1988), peritoneal macrophages (Ohkuma and Poole, 1981; Thyberg and Nilsson, 1982; Thyberg et al., 1983), and a neuroendocrine cell line derived from a lung carcinoma (Schuller and Hegedus, 1989). The enlarged vacuoles are thought to be swollen lysosomes (Thyberg and Nilsson, 1982) and their induction is
Plate 1. (a) Control, human endothelial cells (ENDO). (b) Endothelial cells incubated for 24 hr with 10 mM-nicotine.Note the laterally displaced nuclei and the large vacuoles evenly distributed throughout the cytoplasm. (c) Control, human melanoma cells (SK-Mel/27). (d) Melanoma ceils incubated for 24 hr with 10 raM-nicotine. Note the very small vacuoles in the cytoplasm. (e) Control, human keratinocytes (NHEK). (f) Keratinocytes incubated for 24hr with 10mM-nicotine. Note the enlarged cells with cytoplasmic vacuoles surrounding the nuclei; vacuoles are encircled by a membrane-like structure. Giemsa stain, x210.
Neutral red assay: nicotine and pharmaceuticals
499
100
901
o or i.z O
80-
o h O
70
I-
60
z h.I O or W 0. ~q
50
40 W
L) Z m or O I/1 m
30 20
10
A 0
I
I
I
I
4O
20
0
CONCENTRATION IN mM
Fig. 3. Concentration-response cytotoxicity curves obtained for a 24-hr exposure to theophylline as determined with the neutral red assay using cultures o f various h u m a n cell types. Each point represents the arithmetic m e a n percentage o f the control; values for the standard errors o f the mean were less than + 5 % . Symbols as in Fig. 1.
120 _J
0 n,,
110
I-
100
o
90
z 0
b.
O I-
z W o or bJ ft. 6'1 < hJ
o z< m n, O 6'1 03
80 70 6O 50 40 3O 20 10 0
c
l
2
,
,
4
i
,
6
,
,
8
i
I
10
,
I
12
,
i
14
,
l
16
CONCENTRATION IN mM
Fig. 4. Concentration-response cytotoxicity curves obtained for a 24-hr exposure to phenylpropanolamine as determined with the neutral red assay using cultures of various h u m a n cell types. Each point represents the arithmetic m e a n percentage of the control; values for the standard errors o f the mean were less than + 5 % . Symbols as in Fig. 1.
500
H. BABICHand E. BORENFREUND
100
\ \ \ x
\ \ x \
03 (/1
90
13
80
,<
I..d rY
x x x \
._1
.< n,"
70
I--
x x x \ x x x x x x x x
D
I..d
Z
60
x x x x x x x x x x x \ x x x x x x x x
..J
0 rY
x x x \
50
I-
Z 0 0
40
b_
0 I--
30
Z
b,.I
tO IZ
20
I.d
n
10
\ x ~ x x x ~ x \ \ x x \ x x \ \ x x \ x x x \ x x x x x x x \ \ x x \ x x x x XNNN \ \ \ x x x \ \ \ N \ N x x x x \ \ x N XNNN \ x x x \ x x \ \ \ X \ \ \ N N \NNN
x x x x x x x x
x x x \ x x x x
xxxx x x x x \ N N \ %NX\ x x \ x NNN\ NNNN NNNN N~N\ XXX\
*\N
o
N N
x x x \ % x x \
0
,x,\\"~
15
N 20
0
15
20
CAFFEINE ( r n M )
Fig. 5. Cytotoxicity of caffeine (15 and 20mM) in the absence ([]) and presence ( I ) of phenylpropanolamine (2.5 mM) to HepG2 hepatoma ceils. The data are expressed as the arithmetic mean percentage of the control (medium containing neither caffeine nor phenylpropanolamine); values for the standard errors of the mean were less than + 5%. not unique to nicotine, but is common to weakly basic amines (Ohkuma and Poole, 1981; Valtolina and Foster, 1991). Presumably, the plasma and lysosomal membranes are highly permeable to the neutral form of weak bases, but are only slightly permeable or impermeable to the protonated forms of weak bases. The pH inside the lysosomes is considerably lower than that of the cytoplasm, so that weakly basic amines--such as nicotine--are trapped by protonation inside the lysosomes. Such protonated bases accumulate and thereby increase the osmotic pressure inside the lysosomes, causing water to enter osmotically and to swell the lysosomes (Ohkuma and Poole, 1981). Ohkuma and Poole (1981), in their study of chemical-induced vacuolization in mouse peritoneal cells, observed that although many amines caused vacuole formation, exceptions were noted. Amines with pK, below neutrality and those that are relatively hydrophilic, even in their neutral forms, failed to induce vacuolization. The pK values of nicotine are 3.1 and 8.0, the latter value accounting for its protonation and trapping inside the lysosomes. Cotinine, although also an amine, failed to induce vacuolization in any of the five human cell types. Cotinine has a pK, of 4.5 (Beckett et al., 1972) and is more polar and less lipophilic than nicotine (Benowitz et al., 1983), factors accounting for its inability to induce vacuolization.
Nicotine-induced vacuolization was neither a permanent cellular alteration nor an indicator of cell death. Thus, endothelial cells exposed to 10-15 mM-nicotine developed swollen vacuoles, which gradually disappeared after 48-72hr incubation in fresh, nicotine-free medium. Hanes et al. (1990) showed that nicotine binds non-specifically to human fibroblasts, is rapidly taken up, accumulating intracellularly, but is then slowly released. Apparently, in the studies reported herein with the endothelial cells, the nicotine was released slowly from the lysosomes over a period of 2-3 days, thereby reducing the osmotic pressure within the lysosomes, resulting in the return of the lysosomes to their normal size. Nicotine-induced vacuolization exhibited different patterns among the various human cell types. With the malignant cell types (i.e. SK-Mel/27 melanoma and HepG2 hepatoma cells) the swollen vacuoles were fewer and smaller than in the nicotine-treated normal cells. With the N H E K keratinocytes, the swollen vacuoles encircled a centrally located nucleus and a distinct membranous structure surrounded the entire group of lysosomes. This discrete membranous structure encircling the swollen vacuoles was similar to a structure, identified as an autophagic vacuole, seen in nicotine-treated macrophages (Thyberg et aL, 1983). In those studies such autophagic vacuoles consisted of groups of lysosomes enclosed by a large
Neutral red assay: nicotine and pharmaceuticals lysosome and were associated with nicotine inhibition of endocytosis. The third pattern of vacuolization occurred in the nicotine-treated fibroblasts and endothelial cells. A 24-hr exposure to 10-15 mM-nicotine induced highly swollen vacuoles that filled the entire cytoplasm, pushing the nucleus to the periphery of the cell. This pattern of nicotine-induced reversible vacuolization apparently accounted for the increased uptake and retention of NR when these cells were exposed to neutral red (150 #M) for 3 hr during the cytotoxicity assay. Phenylpropanolamine, at non-toxic levels, induced enlarged vacuoles in the endothelial cells, but not in the other cell types tested. Such vacuolization may have accounted for the increased uptake and retention of N R noted in endothelial cells exposed to subtoxic levels of PPA. The reason for the failure of the induction of vacuoles in the other cell types exposed to non-toxic levels of PPA is not known. Caffeine and its metabolite, theophylline, exhibited approximately equivalent cytotoxic potencies to the NHEK, HFF, SK-Mel/27 and HepG2 cells. However, the endothelial cells were 2-3 times more sensitive to caffeine and about twice as sensitive to theophylline as the other cell types. The cytotoxic effects of caffeine and theophylline to mammalian cells in vitro has been attributed to adverse effects on DNA synthesis (Lehmann and Kirk-Bell, 1974; Weinstein et al., 1975). Metabolism of caffeine in vivo produces a number of intermediates, in addition to theophylline, which may exert different toxicities (Stavric, 1988a). Borenfreund and Puerner (1987), using BALB/c 3T3 mouse fibroblasts as the indicator in the NR assay, showed that the cytotoxicity of 0.5 mM-caffeine was reduced by co-incubation with an hepatic S-9 microsomal fraction from Aroclorinduced rats. The synergistic cytotoxic interaction between caffeine and PPA towards the HepG2 hepatoma cells is interesting, as acute toxicity studies with rats have shown that caffeine potentiated the lethality of PPA (Johnson, 1991). The studies presented herein demonstrate the utility of cytotoxicity assays in vitro, such as the NR assay, to evaluate the toxic potencies of test agents, including those chemicals to which large segments of the population may be exposed. Such assays in vitro are most suitable for the initial screening (tier I testing) of chemicals, as they provide preliminary data on the relative toxicities of test agents, are cost-effective, and reduce the need for extensive animal experimentation in the early stages of toxicity exploration. Furthermore, these studies in vitro can serve as the basis for more detailed analyses, using biotic test systems of increasing complexity. The findings in vitro of the induction of cytoplasmic vacuolization in nicotine-exposed cells and of the hypersensitivity of endothelial cells to caffeine exemplify such studies. TIV 6/6~B
501
Acknowledgements--ApprecAation is expressed of the assistance provided by Irving Bruck, to Clonetics Corporation for supplying the keratinocytes, the endothelial cells, and the KGM and EGM media, and to Schering-Plough (Bloomfield, NJ, USA) for financial support. REFERENCES
Babich H. and Borenfreund E. (1987) Structure-activity relationship (SAR) models established in vitro with the neutral red cytotoxicity assay. Toxicology in Vitro 1, 3-9. Beckett A. H., Gorrod J. W. and Jenner P. (1972) A possible relation between pKa, lipid solubility, and the amounts excreted in urine of some tobacco alkaloids given to man. Journal of Pharmacy and Pharmacology 24, 115-120. Benowitz N. L., Kuyt F., Jacob P., Jones R. T. and Osman A.-L. (1983) Cotinine disposition and effects. Clinical Pharmacology and Therapeutics 34, 604-611. Borenfreund E. and Puerner J. A. (1985) Toxicity determined in vitro by morphological alteration and neutral red absorption. Toxicology Letters 24, 119-124. Borenfreund E. and Puerner J. A. (1987) Short-term quantitative in vitro cytotoxicity assay involving an S-9 activating system. Cancer Letters 34, 243-248. Borzelleca J. F., Bowman E. R. and McKennis H. (1962) Studies on the respiratory and cardiovascular effects of (-)-cotinine. Journal of Pharmacology and Experimental Therapeutics 137, 313-318. Chamson A., Garrone R., Auger C. and Frey J. (1980) Effects of tobacco smoke extracts on the ultrastructure of fibroblasts in culture. Journal of Submicroscopic Cytology 12, 401--406. Hanes P. J., Schuster G. S. and Lubas S. (1990) Binding, uptake, and release of nicotine by human gingival fibroblasts. Journal of Periodontology 62, 147-152. Johnson D. A. (1991) Pharmacology and safety of phenylpropanolamine. Drug Development Research 22, 197-207. Konno S., Chiao J. W. and Wu J. M. (1986) Effects of nicotine on cellular proliferation, cell cycle phase distribution and macromolecular synthesis in human promyelocytic HL-60 leukemia cells. Cancer Letters 33, 91-97. Kyerematen G. A. and Vesell E. S. (1991) Metabolism of nicotine. Drug Metabolism Reviews 23, 3-41. Lehmann A. R. and Kirk-Bell S. (1974) Effects of caffeine and theophylline on DNA synthesis in unirradiated and UV-irradiated mammalian cells. Mutation Research 26, 73-82. Ohkuma S. and Poole B. (I 981) Cytoplasmic vacuolation of mouse peritoneal macrophages and the uptake into lysosomes of weakly basic substances. Journal of Cell Biology 90, 656--664. Raulin L. A., McPherson J. C., McQuade M. J. and Hanson B. S. (1988) The effect of nicotine on the attachment of human fibroblasts to glass and human root surfaces in vitro. Journal of Periodontology 59, 308-315. Schuller H. M. and Hegedus T. J. (1989) Effects of endogenous and tobacco-related amines and nitrosamines on cell growth and morphology of a cell line derived from a human neuroendocrine lung cancer. Toxicology in Vitro 3, 37-43. Sofia R. D. and Knobloch L. C. (1974) Influence of acute pretreatment with delta9-tetrahydrocannabinol on the LD50 of various substances that alter neurohumoral transmission. Toxicology and Applied Pharmacology 28, 227-234. Stavric B. (1988a) Methylxanthines: toxicity to humans. 2. Caffeine. Food and Chemical Toxicology 26, 645-662. Stavric B. (1988b) Methylxanthines: toxicity to humans. 1. Theophylline. Foodand Chemical Toxicology 26, 541-565. Thyberg J., Hedin U., Stenseth K. and Nilsson J. (1983) Effects of nicotine on the fine structure of cultivated mouse peritoneal macrophages. Acta Pathologica,
502
H. BABICH and E. BORENFREUND
Microbiologica et lmmunologica Scandinavica Section A, 91, 23-30. Thyberg J. and Nilsson J. (1982) Effects of nicotine on endocytosis and intracellular degradation of horseradish peroxidase in cultivated mouse peritoneal macrophages. Acta Pathologica, Microbiologica et Scandinavica 90, 305-310. United States Department of Health and Human Services (USDHHS) (1986) The Health Consequences of Using
Smokeless Tobacco. DHHS Publ. No. (NIH) 86-2874. Public Health Service, Bethesda, MD. Valtolina M. and Foster R. (1991) Induction of cytoplasmic vacuolation by two alkyl-diamine compounds. A TLA 19, 30-38. Weinstein D., Mauer I., Katz M. L. and Kazmer S. (1975) The effect of methylxanthines on chromosomes of human lymphocytes in culture. Mutation Research 31, 57~I.