Analysis of the Combined Osteolathyritic Effects of ß-Aminopropionitrile and Diethyldithiocarbamate on Xenopus Development

Melissa A. Cotter*, Deidre L. Policz*, Gerald Pöch{dagger} and Douglas A. Dawson*,1

* Department of Biology/Toxicology, 401 College Avenue, Ashland University, Ashland, Ohio 44805; and {dagger} Institute of Pharmacology and Toxicology, University of Graz, Graz, Austria

Received May 30, 2000; accepted July 31, 2000


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
In order to examine the mechanistic basis between combined effects and mechanisms of action, two osteolathyrogens, ß-aminopropionitrile (ßAPN) and diethyldithiocarbamate (DTC), were tested together on Xenopus embryos. In a separate test, DTC was also tested with copper sulfate to determine the importance of copper in DTC-induced osteolathyrism. Frog embryos (Xenopus laevis) were exposed for 96 h, with daily solution removal and replacement. Preserved tadpoles were evaluated for osteolathyritic lesions. For the ßAPN:DTC test, a 1.2-factor matrix design was used, producing two single chemical and seven mixture-response curves. The {chi}2 goodness-of-fit test was used to compare the experimental mixture-response curves with theoretical effects for two combined effects models, dose-addition and independence. All seven mixture curves were consistent with expected results for dose-addition, but the correlations were generally not high. For the DTC:copper test, the three mixture-response curves generated showed that added copper increased the DTC-alone EC50, but there was no corresponding right shift at the top of the response curves, as observed previously with ßAPN and copper. In the ßAPN:DTC and DTC:copper tests, DTC alone showed a biphasic concentration-osteolathyrism curve, and the slope of the response curve for DTC alone in each test was statistically different than the slope for the ßAPN alone response curve. Taken together, the results suggest the potential for a second osteolathyritic effect of DTC that affected the combined toxicity enough to produce a dose-addition correlation without the chemicals necessarily having the same mechanism.

Key Words: chemical mixture toxicity; dose-addition; independence; copper sulfate; FETAX.


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Osteolathyrism is the failure of developing connective tissue fibers to cross-link properly (Selye, 1957Go). A copper-dependent enzyme, lysyl oxidase (LO) (Pinnell and Martin, 1968Go), and a carbonyl cofactor, lysine tyrosylquinone (Wang et al., 1996Go), control normal cross-linking reactions (Smith-Mungo and Kagan, 1998Go). The enzyme oxidizes the amino side chains of peptidyl lysine in both collagen and elastin, producing an aldehyde intermediate that then undergoes nonenzymatic cross-linking (Kagan, 1986Go). Inhibition of LO activity adversely affects proper connective tissue cross-linking.

The complexity of the cross-linking process leads to several possible mechanisms for induction of osteolathyrism. First is inhibition of LO by direct, irreversible binding of the agent to the enzyme, leading to a decrease in enzyme activity (Tang et al., 1983Go). Second, binding of the chemical to aldehyde precursors of developing fibers could inhibit further cross-linking reactions (Deshmukh and Nimni, 1969Go). A third possible mechanism is that the osteolathyrogen may interfere with cofactor generation and/or binding (Bird and Levene, 1983Go; Kagan and Trackman, 1991Go). Another potential mechanism involves the chelation of copper (Harris et al., 1974Go; Osterberg, 1980Go). It has also been proposed that osteolathyrism could be due to oxidative stress (Ghate, 1985Go).

When exposed to sufficient concentrations of osteolathyrogens, frog embryos show abnormal notochord development. At the lowest effective concentrations, osteolathyritic lesions are small dorsoventral bands and/or an outpocketing of the notochord at its ventral margin. At higher concentrations, there is greater disorganization within the notochord, producing a "wavy-tail" appearance (Riggin and Schultz, 1986Go; Schultz et al., 1985Go, 1988Go). Light and electron microscopy have further characterized the effects as disorganization of collagen fibers and, for some chemicals, absence of the elastic externa (Riggin and Schultz, 1986Go; Schultz et al., 1985Go).

ß-aminopropionitrile (ßAPN) is an osteolathyrogen (Levene, 1961Go) that reportedly induced cross-linking defects by binding to the active site of LO, as the amount of ßAPN bound was proportional to the amount of enzyme inactivation (Tang et al., 1983Go). Frog embryos exposed to this nitrile in several labs (Bantle et al., 1996Go; Dawson and Wilke, 1991Go) showed abnormal notochord morphology similar to that observed for ureides, acid and benzoic acid hydrazides, and alkyl carbazates (Dawson et al., 1990Go; 1991Go; Schultz and Ranney, 1988Go; Schultz et al., 1985Go, 1988Go).

In other studies involving frog embryos, certain dithiocarbamates were found to be potent developmental toxicants (Bancroft and Prahlad, 1973Go). At lower concentrations, they disrupted normal notochord development. At higher concentrations, the notochord was severely malformed and other abnormalities could be induced (Ghate and Mulherkar, 1980Go). Diethyldithiocarbamate (DTC) is a copper chelator (Harris et al., 1974Go; Osterberg, 1980Go). Because of its strong affinity for copper, DTC may alter LO activity, thereby adversely affecting connective tissue cross-linking. However, when the interaction between DTC and cysteine was analyzed, it was concluded that the developmental effects caused by DTC were due to oxidative stress (Ghate, 1985Go).

The frog embryo mixture toxicity assay (Dawson and Pöch, 1997Go) determines the toxicity of binary chemical mixtures through statistical comparison of the mixture-response curves with two mechanistic combined effects models. As a result, the approach may be useful in relating the combined toxicity to common or different mechanisms of chemical action. To initially evaluate the mixture toxicity testing strategy and data analysis techniques developed, and to determine the value of the approach for distinguishing between chemicals that work by the same or by different mechanisms, a series of osteolathyrogen combinations is being tested. Osteolathyrism was selected as the toxic effect for evaluation because of the possibility that osteolathyrism results through more than one mechanism (i.e., enzyme inhibition, copper chelation, cofactor disruption, etc.); because frog embryos exposed to osteolathyrogens exhibit characteristic lesions in the notochord that are easily detected in the transparent tadpoles using a dissecting microscope; and because exposed frog embryos exhibit osteolathyritic lesions at chemical concentrations that do not typically induce other malformations or death, thereby reducing the chance that confounding toxicity factors affect data interpretation. In this study, the osteolathyrogens ßAPN and DTC were tested to determine their combined osteolathyritic effects. In addition, DTC was tested with copper sulfate to determine the importance of copper in DTC-induced osteolathyrism. Results of similar tests of ßAPN with copper have been reported (Dawson et al., 2000Go).


    MATERIALS AND METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Animals and animal care.
Adult Xenopus laevis frogs were purchased for breeding (Xenopus I, Dexter, MI). Upon arrival in the lab, each adult was treated for nematode parasites with ivermectin following established procedures (Dawson et al., 1992Go). Adults were housed in 10-gal glass aquaria with 3 gal filtered tap water. Frogs were fed a mixture of raw ground beef liver and adult frog brittle (Nasco, Ft. Atkinson, WI) three to five times per week. Cleaning of aquaria immediately followed feeding. All other care and breeding procedures were as described previously (Dawson and Bantle, 1987Go; Dawson et al., 1992Go). Research was conducted in compliance with the Animal Welfare Act and other federal statutes and regulations relating to animals and experiments involving animals, and all procedures adhered to principles stated in the Guide for Care and Use of Laboratory Animals, NIH Publication 86-23 (1985 ed.)

Assay procedures.
FETAX solution was used for breeding and as the diluent for all other embryo exposures. Following breeding, embryos were collected and the jelly coat was removed by gentle agitation in a 2% (w/v) cysteine solution (pH 8.1) for 1–2 min. After rinsing in FETAX solution (Dawson and Bantle, 1987Go), dead and abnormal embryos were discarded, with remaining embryos microscopically examined for proper development in blastula and early gastrula stages. Those judged to be developing properly were retained for testing.

ß-aminopropionitrile (ßAPN, CAS # 2079-89-2, a monofumarate salt) and diethyldithiocarbamate (DTC, CAS # 20624-25-3, sodium salt trihydrate) were purchased from Sigma Chemical Co. (St. Louis, MO). All reported concentrations of ßAPN and DTC were corrected for fumarate or for sodium and water content, respectively.

Individual stock solutions for ßAPN and DTC were prepared, and the pH was adjusted to 7.3 as needed. Thirty-six treatment solutions at 25 ml were prepared based on a 1.2-factor matrix design (Dawson and Pöch, 1997Go), resulting in 1 control, 12 single-chemical (6 per chemical), and 23 mixture treatments (Fig. 1Go). This design produces 2 single-chemical and 7 mixture-response curves: 2 with the ßAPN concentration fixed and the DTC concentrations increasing, 2 vice versa, and 3 fixed-ratio curves (1:1, 1:3, 3:1) in which the concentrations of both chemicals increase.



View larger version (80K):
[in this window]
[in a new window]
 
FIG. 1. Mixture toxicity testing design with 36 treatments. The design produces two single-chemical and seven mixture-response curves: two with the concentration of chemical A held at a fixed value with increasing chemical B concentrations (shaded columns), two vice versa (shaded rows), and three fixed-ratio curves in which the concentration of both chemicals increase (diagonal lines). The -fold EC50 values are based on a 1.2 factor, with 1.0 representing the EC50 for osteolathyrism induced by each chemical alone.

 
Exposures were conducted for 96 h at 23.5 ± 1°C in covered glass Petri dishes containing 10 ml of the appropriate solutions. These solutions were renewed daily with freshly prepared stock solutions. For the ßAPN:DTC combination, there were 50 embryos per treatment divided equally between two dishes. Preserved tadpoles (1% formalin, pH 7.0) were microscopically examined and the number of tadpoles with osteolathyritic lesions and/or other malformations were recorded.

For the DTC:copper test, a 500 ml stock solution of cupric sulfate-pentahydrade (CAS # 7758-99-8, Fisher Scientific) was prepared in FETAX solution. There were 32 treatments consisting of 1 control, 7 DTC-only, 3 copper sulfate-only (0.5, 1.0, and 1.5 mg/l), and each of those 3 copper sulfate concentrations with each of the 7 DTC concentrations (i.e., 21 mixture treatments). For this combination there were 60 embryos per treatment divided equally between three dishes; with each dish containing 8 ml of solution. The copper sulfate solution was used for treatment preparation throughout the test. Other than the changes noted here, procedures were as described for the ßAPN:DTC test.

Spectrophotometry.
Because DTC is a metal chelator and there was color formation when DTC and copper sulfate were combined, spectrophotometry was employed to examine aspects of the color formation. Spectrophotometric readings of DTC, ßAPN, and copper sulfate solutions, alone and in combination, were conducted using a Milton Roy Spectronic 1201 UV-visible scanning spectrophotometer.

Data analysis.
In order to determine if non-osteolathyritic malformations were related to chemical exposure, two multiple comparison tests were conducted. Dunnett's test (Dunnett, 1955) and the Student-Newman-Keuls test (Steel and Torrie, 1980Go) were used to analyze for differences (p <= 0.05) in rates of non-osteolathyritic malformations between each treatment and the control and across all treatments, respectively.

Experimental data points for osteolathyrism induced by chemical A alone and by chemical B alone were fit to sigmoidal curves using a four-parameter logistic function and SigmaPlot (Jandel Scientific, San Rafael, CA). Curve fit parameters were: a, minimum effect, b, slope, c, EC50 osteolathyrism, and d, maximum effect. Then, for each experimental response curve, theoretical curves for the number of osteolathyritic embryos were calculated for the dose-addition and independence models in the following manners.

Theoretical dose-addition curve calculation for comparison with fixed-dose responses (e.g., Fig. 2Go) were based on the principle described by Pöch et al. (1990). This procedure was conducted using SigmaPlot (Jandel Scientific, San Rafael, CA) as detailed by Holzmann et al. (1999). In essence the theoretical curve is obtained by calculating dose x of chemical A with which the concurrently applied fixed-dose of chemical B was equieffective, using the equation x = {([(a - d)/(a2 - d)] - 1)*(cb)}1/b where a, b, c, and d are the parameters (defined above) for the curve of chemical A alone and a2 is the minimum effect for the curve of chemical A with the concurrently applied fixed-dose of chemical B. From the definition of dose-addition, the effects of chemical A alone are to be expected at the concentrations of A-x. Therefore, the response values for the theoretical dose-addition curves of chemical A correspond to simulated concentrations of A-x.



View larger version (22K):
[in this window]
[in a new window]
 
FIG. 2. Concentration-osteolathyrism curves of ßAPN alone (open circles) and in the presence of a fixed concentration (0.868 mg/l) of DTC (closed circles) versus theoretical response curves for dose-addition (dashed line) and independence (dotted line).

 
Theoretical dose-addition curve calculation for comparison with fixed-ratio responses (e.g., Fig. 3Go) use the left shift from the concentration-osteolathyrism curve of chemical B, calculated as left shift = 1 + (EC50 B/ EC50 A) with a common slope of A and B calculated as (B slope + A slope)/2. The EC50 of the dose-addition curve = 1/left shift*EC50 of the more effective agent.



View larger version (27K):
[in this window]
[in a new window]
 
FIG. 3. Concentration-osteolathyrism curve of ßAPN alone (open circles), DTC alone (open squares), and their 1:1 mixture (closed circles) versus theoretical response curves for dose-addition (dashed line) and independence (dotted line).

 
Theoretical independence curves were calculated based on the effects E of A and B alone: EA+B = EA + EB – (EA x EB), where E is expressed as the fraction of the maximum response and was corrected for osteolathyrism in controls (x) using the equation E = (E – x)/(1 – x).

Fixed-ratio response calculations required that chemical B concentrations be converted to equivalent concentrations of chemical A for each fixed ratio (i.e., 1:3, 1:1, 3:1) using the equation B*factor A/B (Pöch et al., 1997Go).

The {chi}2 goodness-of-fit test was used to compare the experimental responses with the theoretical responses (Pöch, 1993Go), following performance of an F-test to compare slope values generated for each chemical tested alone. The goodness-of-fit analyses were used to determine whether the combined effects differed significantly (p < 0.05) from the theoretical effects for dose-addition and for independence.


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
ßAPN:DTC Test
In the ßAPN:DTC test, no deaths and no osteolathyritic lesions were present in controls. Of the fifty controls, one tadpole had a mouth abnormality (2.0%); otherwise, development was judged as normal.

For all treatments, the frequency of death was 1.6%, whereas the frequency of non-osteolathyritic malformations in surviving tadpoles was 2.4%. Non-osteolathyritic defects were of the eye (1.8%), gut (1.6%), or edema (1.3%). Neither the numbers of dead tadpoles nor tadpoles with non-osteolathyritic malformations were significantly different between each treatment and the control or across all treatments.

The concentration-response data for ßAPN alone and for DTC alone (Table 1Go) were used in calculating expected combined effects for the seven ßAPN:DTC mixture curves. An F-test indicated the slopes for ßAPN alone and DTC alone were significantly different (F = 41.75, p = 0, d.f. = 11). For ßAPN alone, the response curve slope was 5.27. The slope for the DTC-alone response curve was 2.08, but the curve appeared to be biphasic. An analysis of this curve showed the lower phase having a slope of 2.62 [with a minimum effect (Emin) of 0 and a maximum effect (Emax) of 71] and the upper phase a slope of 3.71 (with Emin = 37 and Emax = 98).


View this table:
[in this window]
[in a new window]
 
TABLE 1 Concentration-Osteolathyrism Data for Frog Embryos Exposed to ßAPN Alone and to DTC Alone
 
All seven curves were at least weakly correlated with theoretical effects calculated for the dose-addition model, whereas none of the curves were consistent with independence. The goodness-of-fit data for ßAPN:DTC showed both high (p > 0.90) correlation to dose-addition for the two mixture curves in which the DTC concentrations were in the upper phase of the DTC-alone response curve [i.e., the higher fixed-DTC concentration curve (Table 2Go) and the 1:3 fixed-ratio curve], and low (0.05 < p < 0.40) or moderate (0.50 < p < 0.90) dose-addition correlations for the five curves in which the DTC concentrations either spanned both upper and lower phases [i.e., both fixed-ßAPN concentration curves (Table 3Go) and the 1:1 fixed-ratio curve (Table 4Go, Fig. 3Go)] or were in the lower phase [i.e., the lower fixed-DTC concentration curve (Table 2Go, Fig. 2Go) and the 3:1 fixed-ratio curve (see Table 4Go for all fixed-ratio curves)].


View this table:
[in this window]
[in a new window]
 
TABLE 2 Concentration-Osteolathyrism Data for Frog Embryos Exposed to the Fixed DTC Concentrations with Varying ßAPN Concentrations, Expected Numbers for Dose-Addition and Independence Models of Combined Effects, and Data Fit to the Models
 

View this table:
[in this window]
[in a new window]
 
TABLE 3 Concentration-Osteolathyrism Data for Embryos Exposed to the Fixed ßAPN Concentrations with Varying DTC Concentrations, Expected Numbers for Dose-Addition and Independence Models of Combined Effects, and Data Fit to the Models
 

View this table:
[in this window]
[in a new window]
 
TABLE 4 Concentration-Osteolathyrism Data for Frog Embryos Exposed to Varying Concentrations of ßAPN and DTC at 1:1, 1:3, 3:1 Effect Ratios and Expected Numbers for the Dose-Addition and Independence Models of Combined Effects
 
DTC:Copper Test
In testing DTC with copper, the control treatment had no deaths, and none of the tadpoles had either osteolathyritic lesions or non-osteolathyritic malformations. Across all treatments, the frequency of death was < 0.1%, whereas non-osteolathyritic malformations occurred in 4.0% of the tadpoles. The most common non-osteolathyritic malformations were of the gut (3.2%), eye (2.7%), or edema (1.7%).

As with the ßAPN:DTC test, the DTC-alone response curve in the copper test (slope = 2.01) appeared to be biphasic, with the lower phase having a slope of 2.88 (Emin = 1, Emax = 70) and the upper phase a slope of 3.43 (Emin = 41, Emax = 99).

The EC50 for osteolathyrism induced by DTC alone was 0.80 mg/l, whereas the EC50 values for DTC with 0.5, 1.0, and 1.5 mg/l copper sulfate increased to 1.23, 1.73, and 2.07 mg/l, respectively. Although toxicity was reduced by coadministration of copper at the lower DTC concentrations, this reduction in toxicity was less prominent at the higher DTC concentrations, as evidenced by the lack of a right shift at the top of the DTC-copper response curves (Table 5Go, Fig. 4Go). Previous tests (Dawson, et al., 2000Go) showed an inhibition of ßAPN-induced osteolathyrism upon copper coadministration. As with DTC, the ßAPN EC50 was increased, but in this case there was a right shift of the entire response curve as opposed to that observed with DTC and copper herein.


View this table:
[in this window]
[in a new window]
 
TABLE 5 Concentration-Osteolathyrism Data for Frog Embryos Exposed to DTC Alone and with Three CuSO4 Concentrations
 


View larger version (20K):
[in this window]
[in a new window]
 
FIG. 4. Concentration-osteolathyrism curve of DTC alone (circles) and in the presence of 1 mg/l CuSO4 (squares).

 
Spectrophotometry
Results of UV-visible spectrophotometric analyses (data not shown) of DTC, ßAPN, and copper sulfate solutions indicated that DTC and copper sulfate together produced a peak at about 450 nm. This peak was consistent with color formation when DTC and copper sulfate were combined. The DTC stock solution was clear; the copper sulfate stock solution was pale blue at the concentrations prepared for testing. Although the copper sulfate solution alone had slight absorbance at 450 nm, that absorbance was markedly enhanced (slightly more than one order of magnitude) by addition of DTC. No such coloration or enhancement of absorbance was detected when ßAPN was added to copper sulfate. At the appropriate concentrations, addition of ßAPN to a solution containing DTC and copper sulfate did not alter peak absorbance at 450 nm. Addition of DTC to a solution of ßAPN and copper sulfate produced a peak absorbance at 450 nm that was identical to that observed for DTC and copper sulfate without ßAPN.


    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
In chemical-mixture toxicity testing, it is frequently instructive to compare experimental toxicity with theoretical effects obtained from combined effects models. By using such mechanistic combined effects models in conjunction with a properly designed and conducted toxicity test, it should be possible to distinguish between chemicals working by the same mechanism and those working by different mechanisms (Pöch et al., 1996Go). Two such models, dose-addition and independence, are useful because they relate, theoretically, to events at the molecular sites of action (Pöch et al., 1996Go). In dose-addition, the agents are assumed to share a single molecular site of action, with one chemical acting as a dilution of the other. In cases where dose-addition is observed and the chemicals have a common dose-response curve slope, it is likely that the chemicals have a common mechanism of toxicity. In independence (Bliss, 1939Go), the chemicals act at different sites, such that toxicity of one chemical does not affect the toxicity of the other. Independently acting chemicals, therefore, induce toxicity by different mechanisms.

The frog embryo mixture toxicity assay, a modification of FETAX (ASTM, 1991Go; Dumont et al., 1983Go), was developed to relate the combined effects of chemicals to common or different mechanisms of toxicity (Dawson and Pöch, 1997Go; Mentzer et al., 1999Go). The assay incorporates the fixed-dose (Pöch, 1993Go) and fixed-ratio (Dawson, 1994Go) mixture toxicity testing strategies into a 1.2-factor treatment design. In this design, each chemical is tested alone at six concentrations, treatments 2–7 for chemical A (in this case ßAPN) and 8–13 for chemical B (i.e., DTC). These treatments were selected by assigning the average EC50 for osteolathyrism induced by each chemical alone (determined from previous testing) the value 1.0 (i.e., treatments 5 and 11 for chemicals A and B, respectively). The other treatments for each chemical alone were then determined by multiplying or dividing the EC50 by 1.2x. The 1.2-factor values for generating the concentrations (as given in Fig. 1Go) are rounded (e.g., treatment #2 is 1/1.25 = 0.40). For a chemical with a slope between 2 and 8 (Dawson and Pöch, 1997Go), as is obtained for osteolathyrogens tested on Xenopus, such concentrations cover, at minimum, the linear portion of the dose-response curve. The single-chemical dose-response curves provide the data needed to calculate theoretical mixture-response curves for both the dose-addition and independence models. The 23 mixture treatments, also generated using the 1.2-factor matrix, give seven experimental mixture-response curves (Fig. 1Go), which are compared to the theoretical curves using the {chi}2 goodness-of-fit test. The results of the comparisons, then, indicate whether the chemicals are likely to induce toxicity by the same or by different mechanisms.

In this study, all seven mixture-response curves were consistent statistically with theoretical effects calculated using the dose-addition model. This is the expected result for two chemicals that induce effects by the same mechanism. However, a more thorough analysis of the response curves and osteolathyrogen:copper data suggests a more complex combined effect:

To evaluate the apparent biphasic nature of the DTC-alone response curves, preserved tadpoles were reevaluated by lesion location along the length of the notochord. For ßAPN and DTC, initial lesions appear in different locations, with the ßAPN lesions being in the tail posterior to the cloacal outlet (i.e., approximately the posterior two-thirds of the notochord), and the DTC lesions being at or anterior to the cloacal outlet (i.e., roughly the anterior third of the notochord). However, with increasing ßAPN concentration, anterior lesions become apparent, whereas with increasing DTC concentration, posterior lesions become apparent. Addition of a second lesion type, for either chemical, has the effect of altering the slope of the response curve, as established by the initial lesions, thereby giving the curve a biphasic appearance. The important question resulting from this observation is whether location of the lesions represents a difference in the way (i.e., mechanism) notochordal connective tissue cross-linking is disrupted.

There is support in the recent literature for the idea that a single osteolathyrogen could disrupt connective tissue cross-linking in more than one way (Dawson et al., 2000Go). For example, copper is known to be needed for proper LO activity in connective tissue fiber cross-linking. Recent work indicates that it is likely that copper is also needed in the regeneration of the carbonyl cofactor associated with LO (Dooley, 1999Go; Matsuzaki et al., 1994Go; Smith-Mungo and Kagan, 1998Go; Wang et al., 1996Go), setting up the possibility that copper-sensitive osteolathyrogens could adversely affect connective tissue fiber cross-linking at two different points in the cross-linking process. The exact mechanism by which LO brings about cross-linking of developing collagen and elastin fibers has yet to be discerned (Smith-Mungo and Kagan, 1998Go), so proposed non–copper-dependent osteolathyrogenic mechanisms are also possible.

It should be feasible to assess the hypothesis that differences in location of osteolathyritic lesions represent different mechanisms by which the chemicals act, simply by analyzing the mixture data broken out by lesion location to develop goodness-of-fit comparisons with the dose-addition and independence models of combined effect. For example, for two chemicals A and B, one could examine the goodness-of-fit of the A:B mixture-response curves using posterior lesions only, anterior lesions only, and posterior lesions for A, with anterior lesions for B (and vice versa), to assess similarity or differences in mechanisms. This was not attempted for the ßAPN:DTC data presented herein, because complete concentration-response curves for both lesion locations were not obtained for each chemical alone, and because the n of 50 embryos per treatment was judged to be insufficient for providing a rigorous assessment of the hypothesis. The idea will be evaluated in future tests of osteolathyrogen combinations using n of 80–100 embryos per treatment to provide the desired statistical power. In addition, since several osteolathyrogens show notochord lesions in both locations, it will be necessary to examine this phenomenon across multiple combinations of osteolathyrogens, in osteolathyrogen:copper combinations to determine if one lesion location is inhibited more frequently than another when copper is present, and if coadministration of copper is shown to preferentially inhibit a specific lesion location, in tests in which binary osteolathyrogen combinations are examined in the presence and absence of copper.

A clear mechanistic interpretation of the combined osteolathyritic effects of the ßAPN:DTC data presented here is problematic. Although the mixture-response curves were consistent with dose-addition, strong correlation (p >= 0.90) with dose-addition was obtained for only two of the seven mixture-response curves. These results, coupled with significantly different slopes for the dose-response curves of ßAPN alone and DTC alone and the putative biphasic osteolathyritic response elicited by DTC alone, are considered important indications of a different dose-effect relationship due to differences in the action of the chemicals under study. Such results might be explained by actions of the chemicals that are different with respect to their molecular sites or modes of action but may share a common pathway; in essence similar to the "two-receptor one-transducer" model (Leff, 1987Go; Scaramellini et al., 1997Go). Therefore, the hypothesis presented here to explain the results is that the apparent biphasic nature of the DTC-alone response curve actually represents a second osteolathyritic effect of DTC (and that differences in lesion location should be systematically evaluated as the most likely source of the biphasic response). In such a multiple-mechanism scenario, the additional effect of a second mechanism might enhance the experimental mixture responses enough over the theoretical response calculated for independence to produce a combined effect equivalent to dose-addition.

Whether this hypothesis is subsequently confirmed or not, the results hold some important implications for evaluating chemical-mixture toxicity using this experimental design. The theory for a dose-additive combined effect assumes only one molecular site of action for a chemical. If a chemical such as DTC actually induces osteolathyrism in two ways within the same series of concentrations, the current theoretical models of chemical mixture toxicity would likely be inadequate, necessitating further model development. Even if this is so, data resulting from the current experimental design is amenable to other statistical approaches, such as response surface analysis, and, therefore, when evaluated over numerous data sets, has the potential to advance understanding of chemical mixture toxicity.

Although additional work needs to be done to determine the exact nature of ßAPN- and DTC-induced osteolathyrism and how it relates to the combined effect obtained herein, the results of this study are important for several reasons. First, the osteolathyritic effects of both ßAPN and DTC on Xenopus development are copper sensitive (although perhaps not in the same way), implicating copper as important in the osteolathyrogenic mechanism of these chemicals and providing direction for further mechanistic work at the biochemical level. Second, a combined effect statistically consistent with dose-addition was obtained despite ßAPN and DTC having statistically different slopes for their individual concentration-response curves, pointing out the potential importance of considering other factors such as slope and shape of the response curve in assessing the combined effect obtained using this approach. Third, the experimental design produces multiple mixture-response curves, providing a substantial amount of information to use in assessing chemical mixture toxicity.


    ACKNOWLEDGMENTS
 
The first two authors contributed equally to this research.


    NOTES
 
This publication was made possible by grant 1 R15 ES08019-01 from the National Institute of Environmental Health Sciences (NIEHS), National Institutes of Health (NIH). Its contents are solely the responsibility of the authors and do not represent the official views of the NIEHS, NIH.

1 To whom correspondence should be addressed. Fax: (419) 289-5333. E-mail: ddawson2{at}ashland.edu. Back


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
ASTM (1991). Standard guide for conducting the Frog Embryo Teratogenesis Assay: Xenopus (FETAX). Biological Effects and Fate, E47–01Aquatic Toxicology, American Society for Testing and Materials, Philadelphia.

Bancroft, R., and Prahlad, K. V. (1973). Effect of ethylenebis[dithiocarbamic acid] disodium salt (nabam) and ethylenebis[dithiocarbamato]manganese (maneb) on Xenopus laevis development. Teratology 7, 143–150.[ISI][Medline]

Bantle, J. A., Finch, R. A., Burton, D. T., Fort, D. J., Dawson, D. A., Linder, G., Rayburn, J. R., Hull, M., Kumsher-King, M., Gaudet-Hull, A. M., and Turley, S. D. (1996). FETAX interlaboratory validation study: Phase III-Part 1 testing. J. Appl. Toxicol. 16, 517–528.[ISI][Medline]

Bird, T. A., and Levene, C. I. (1983). The effects of a vitamin B6 antagonist, 4-deoxypyridoxine, on the cross-linking of collagen in the developing chick embryo. Biochem. J. 210, 633–638.[ISI][Medline]

Bliss, C. I. (1939). The toxicity of poisons applied jointly. Ann. Appl. Biol. 26, 585–615.

Dawson, D. A., and Bantle, J. A. (1987). Development of a reconstituted water medium and preliminary validation of the frog embryo teratogenesis assay Xenopus (FETAX). J. Appl. Toxicol. 7, 237–244.[ISI][Medline]

Dawson, D. A., Schultz, T. W., and Schroeder, E. C. (1992). Laboratory care and breeding of the African clawed frog. Lab Animal. 21, 31–36.

Dawson, D. A. (1994). Chemical mixture toxicity assessment using an alternative-species model: Applications, opportunities, and perspectives. In Toxicology of Chemical Mixtures: Case Studies, Mechanisms, and Novel Approaches (R. S. H. Yang, Ed.). pp. 539–563. Academic Press, San Diego.

Dawson, D. A., and Pöch, G. (1997). A method for evaluating mechanistic aspects of chemical mixture toxicity. Toxicol. Methods 7, 267–278.[ISI]

Dawson, D. A., Schultz, T. W., Baker, L. L., and Mannar, A. (1990). Structure-activity relationships for osteolathyrism: III. Substituted thiosemicarbazides. J. Appl. Toxicol. 10, 59–64.[ISI][Medline]

Dawson, D. A., Schultz, T. W., and Baker, L. L. (1991). Structure-activity relationships for osteolathyrism: IV. p-substituted benzoic acid hydrazides and alkyl carbazates. Environ. Toxicol. Chem. 10, 455–461.[ISI]

Dawson, D. A., and Wilke, T. W. (1991). Initial evaluation of developmental malformation as an end point in mixture toxicity hazard assessment for aquatic vertebrates. Ecotoxicol. Environ. Saf. 21, 215–226.[ISI][Medline]

Dawson, D. A., Cotter, M. A., Policz, D. L., Stoffer, D. A., Nichols, J., and Pöch, G. (2000). Comparative evaluation of the combined osteolathyritic effects of two nitrile combinations on Xenopus embryos. Toxicology 147, 193–207.[ISI][Medline]

Deshmukh, K., and Nimni, M. E. (1969). A defect in the intramolecular and intermolecular cross-linking of collagen caused by penicillamine. II. Functional groups involved in the interaction process. J. Biol. Chem. 244, 1787–1795.[Abstract/Free Full Text]

Dooley, D. M. (1999). Structure and biogenesis of topaquinone and related cofactors. J. Biol. Inorg. Chem. 4, 1–11.[ISI][Medline]

Dumont, J. N., Schultz, T. W., Buchanan, M., and Kao, G. (1983). Frog embryo teratogenesis assay: Xenopus (FETAX) – A short-term assay applicable to complex environmental mixtures. In Short-Term Bioassays in the Analysis of Complex Environmental Mixtures III, (M. D. Waters, S. S. Sandhu, J. E. Lewtas, L. Claxton, N. Chernoff, S. Nesnow, Eds.). pp. 393–405. Plenum Press, New York.

Dunnet, C. W. (1955). A multiple comparisons procedure for comparing several treatments with a control. J. Am. Stat. Assoc. 50, 1096–1121.[ISI]

Ghate, H. V., and Mulherkar, L. (1980). Effect of sodium diethyldithiocarbamate on developing embryos of the frog Microhyla ornata. Indian J. Exp. Biol. 18, 1040–1042.[ISI][Medline]

Ghate, H. V. (1985). Dithiocarbamate induced teratogenesis in frog embryo. Riv. Biol. 78, 288–291.[ISI][Medline]

Harris, E. D., Gonnerman, W. A., Savage, J. E., and O'Dell, B. L. (1974). Connective tissue amine oxidase II. Purification and partial characterization of lysyl oxidase from chick aorta. Biochim. Biophys. Acta 341, 332–344.[ISI][Medline]

Holzmann, S., Dittrich, P., and Pöch, G. (1999). Dose-response curves of vasorelaxants: computerized calculation of combination effects using the example of nicorandil. J. Clin. Basic Cardiol. 2, 96–98.

Kagan, H. M. (1986). Characterization and regulation of LO. In Biology of the Extracellular Matrix, Vol 1: Regulation of Matrix Accumulation, (R. P. Mecham, Ed.). pp. 321–398. Academic Press, Orlando, FL.

Kagan, H. M., and Trackman, P. C. (1991). Properties and function of lysyl oxidase. Am. J. Respir. Cell Mol. Biol. 5, 206–210.[ISI][Medline]

Leff, P. (1987). An analysis of amplifying and potentiating interactions between agonists. J. Pharmacol. Exp. Ther. 243, 1035–1042.[Abstract]

Levene, C. I. (1961). Structural requirements for lathyrogenic agents. J. Exp. Med. 114, 295–310.

Matsuzaki, R., Fukui, T., Sato, H., Ozaki, Y. and Tanizawa, K. (1994). Generation of the topa quinone cofactor in bacterial monoamine oxidase by cupric ion-dependent autooxidation of a specific tyrosyl residue. FEBS Lett. 351, 360–364.[ISI][Medline]

Mentzer, R. K., Smith, N. D., Pöch, G. and Dawson, D. A. (1999). Combined osteolathyric effects of ß-aminopropionitrile and penicillamine on Xenopus embryos: Statistical comparison with dose-addition and independence. Drug Chem. Toxicol. 22, 359–374.[ISI][Medline]

Osterberg, R. (1980). Therapeutic uses of copper-chelating agents. Ciba Found. Symp. 79, 283–299.[Medline]

Pinnell, S. R., and Martin, G. R. (1968). The cross-linking of collagen and elastin: enzymatic conversion of lysine in peptide linkage to {alpha}-aminoadipic-{delta}-semialdehyde (allysine) by an extract from bone. Proc. Natl. Acad. Sci. U.S.A. 61, 708–714.[ISI][Medline]

Pöch, G. (1993). Combined Effects of Drugs and Toxic Agents: Modern Evaluation in Theory and Practice. Springer-Verlag, Wein.

Pöch, G., Dittrich, P., and Holzmann, S. (1990). Evaluation of combination effects in dose-response studies by statistical comparison with additive and independent interactions. J. Pharmacol. Methods 24, 311–325.[ISI][Medline]

Pöch, G., Dawson, D. A., and Reiffenstein, R. J. (1996). Model usage in evaluation of combined effects of toxicants. Toxicol. Ecotoxicol. News 3, 51–59.

Pöch, G., Dawson, D. A., and Dittrich, P. (1997). Teratogenic mixtures: Analysis of experimental dose-response curves and statistical comparison with theoretical effects. Arch. Complex Environ. Studies 9, 23–33.

Riggin, G. W., and Schultz, T. W. (1986). Teratogenic effects of benzoyl hydrazine on frog embryos. Trans. Am. Microsc. Soc. 105, 197–210.[ISI]

Scaramellini, C., Bennett, G. and Leff, P. (1997). Analysis of agonist-agonist interactions: the crucial influence of curve shape. J. Pharmacol. Toxicol. Methods 37, 167–178.[ISI][Medline]

Selye, H. (1957). Lathyrism. Rev. Can. Biol. 16, 1–82.[Medline]

Schultz, T. W., and Ranney, T. S. (1988). Structure-activity relationships for osteolathyrism. II. Effects of alkyl-substituted acid hydrazides. Toxicology 53, 147–159.[ISI][Medline]

Schultz, T. W., Dumont, J. N., and Epler, R. G. (1985). The embryotoxic and teratogenic effects of semicarbazide. Toxicology 36, 183–198.[ISI][Medline]

Schultz, T. W., Ranney, T. S., Riggin, G. W., and Cajina-Quezada, M. (1988). Structure- activity relationships for osteolathyrism. I. Effects of altering the semicarbazide structure. Trans. Am. Microsc. Soc. 107, 113–126.[ISI]

Smith-Mungo, L. I., and Kagan, H. M. (1998). Lysyl oxidase: properties, regulation and multiple functions in biology. Matrix Biol. 16, 387–398.[ISI][Medline]

Steel, R. G. D., and Torrie, J. H. (1980). Principles and Procedures of Statistics: A Biometrical Approach, 2nd ed. McGraw-Hill, New York.

Tang, S. S., Trackman, P. C., and Kagan, H. M. (1983). Reaction of aortic lysyl oxidase with ß-aminopropionitrile. J.Biol. Chem. 258, 4331–4338.[Abstract/Free Full Text]

Wang, S. X., Mure, M., Medzihradszky, K. F., Burlingame, A. L., Brown, D. E., Dooley, D. M., Smith, A. J., Kagan, H. M., and Klinman, J. P. (1996). A crosslinked cofactor in LO: redox function for amino side chains. Science 273, 1078–1084.[Abstract]





This Article
Abstract
FREE Full Text (PDF)
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (3)
Disclaimer
Request Permissions
Google Scholar
Articles by Cotter, M. A.
Articles by Dawson, D. A.
PubMed
PubMed Citation
Articles by Cotter, M. A.
Articles by Dawson, D. A.