Toxicity

The state of knowledge regarding the effects of arsenic on microorganisms was summarized very well in a study by Mandel et al. ⁵¹⁷ on the action of arsenic on Bacillus cereus. Trivalent sodium arsenite was found to inhibit growth at a lower concentration (0.4 mM) than pentavalent sodium arsenate (10 mM). The toxicity of the arsenate could be increased by lowering the phosphate concentration of the growth medium, whereas the inhibitory effect of arsenite was independent of phosphate concentration. This inverse relationship between the toxicity of arsenate and the concentration of phosphate might be related to the fact that arsenate can compete with phosphate for transport. ⁶⁸¹ However, Da Costa ¹⁹³ found that phosphate could suppress the inhibitory effects of arsenite, as well as arsenate, on the growth of fungi. Mandel et al. ⁵¹⁷ showed that neither arsenate nor arsenite produced any specific effects in B. cereus on the incorporation of radioactive precursors into ribonucleic and deoxyribonucleic acids, proteins, or cell wall. Radioarsenite was bound by the microorganism much more strongly than radioarsenate, in agreement with the hypothesis that the toxicity of a particular arsenical was related to the binding of it to the tissues. ³⁵⁸ Because no instance of interconversion between As(V) and As(III) could be established, it was concluded that the two compounds inhibited the growth of B. cereus by separate mechanisms. The mechanisms of toxicity of arsenicals for this organism seem to be similar to those proposed for mammalian systems, which are discussed later.

Adaptation

The adaptation of microorganisms to arsenic compounds was of great practical interest during the earlier part of this century, because organic arsenicals were used extensively as trypanocides at that time. The resistance to organic arsenicals was found to depend on the nature of the chemical substituents on the phenyl ring: ⁵ water-attracting groups (such as −OH and −NH₂), less hydrophilic groups (such as −CH₃ and −NO₂), and groups highly ionized at a pH of 7 (such as −COOH). The state of oxidation of the arsenic is of little consequence in inducing resistance to any of these compounds. The mechanism of drug resistance in trypanosomes is usually a decreased permeability to the drug, and the nonarsenical portion of the molecule largely determines the uptake of the drug by the parasite.

A decreased permeability to arsenic appears to be a rather widespread adaptational mechanism, in that a decreased arsenic uptake was observed in Escherichia coli mutants that were resistant to arsenate ⁷⁰ and in Pseudomonas pseudomallei that had adapted to arsenite. ²⁵ In the latter case, no increase in the content of α-ketoglutarate dehydrogenase, total sulfhydryl compounds, or lipoic acid was observed in the resistant bacteria. ⁷² That the total quantity of free thiol groups may be important in some cases of arsenic tolerance, however, was suggested by the work of Harington, ³²⁷ who found that resistant strains of the blue tick contained more total sulfhydryl than sensitive strains.

Novick and Roth ⁵⁹¹ showed that the penicillinase plasmids, a series of extrachromosomal resistance factors in Staphylococcus aureus, carry determinants of resistance to several inorganic ions, as well as resistance to penicillin. Among the inorganic ions were arsenite, arsenate, lead, cadmium, and mercury. Resistance to arsenate was found to be induced in cultures of plasmid-positive strains by prior growth with an uninhibitory concentration of the anion. Dyke et al. ²²⁵ observed that strains resistant to arsenate, mercury, and cadmium were nearly always resistant to multiple antibiotics and produced large amounts of penicillinase. Although the general genetic and physiologic properties of these ion-resistance markers have been studied, hardly any work has been done on the biochemical mechanisms of this sensitivity and resistance.

Biochemical Response to Arsenic Compounds

When arsenic in solution penetrates the cuticle and enters the apoplast system (the nonliving cell-wall phase), it bathes the external surface of the plasmolemma of the symplast. This is the location of at least some of the enzymes of the living plant. One of the first symptoms of injury by sodium arsenite is wilting, caused by loss of turgor, and this immediately suggests an alteration in membrane integrity. Reaction of trivalent arsenic with sulfhydryl enzymes could well explain the effects of membrane degradation—injury and eventually death.

In general, arsenates are less toxic than arsenites. The arsenate symptoms involve chlorosis, but not rapid loss of turgor (at least in the early expression of toxicity), and the contact action of the arsenates is more subtle. Arsenate is known to uncouple phosphorylation. Thus, the coupled phosphorylation of adenosine diphosphate (ADP) is abolished, the energy of adenosine triphosphate (ATP) is not available, and the plant must slowly succumb. ²⁰⁷

Arsenate has other profound effects on plant systems. For example, Figure 5-1 shows the relative effects of arsenite and arsenate on the activation of the enzyme fumarase. Fumaric acid is a constituent of all plants and is involved in the citric acid cycle. Fumarase carries out the conversion of fumarate to L-malate.

FIGURE 5-1

Velocity of activation of fumarase by arsenate and arsenite ions as a function of pH. Adapted from Dixon and Webb. ^{(p. 473)}

The above examples typify the role played by the organic arsenical herbicides in plant metabolism. When one considers the number of reactions in plants that involve sulfhydryl groups and phosphorus, it is easy to appreciate the ways in which arsenites and arsenates may upset plant metabolism and interfere with normal growth. The ability of arsenate to enter into reactions in place of phosphate is probably the most important way in which arsenic acts as a toxicant. Not only does it substitute for phosphate in a number of ways, but work with labeled arsenates and arsonates has indicated that these compounds are absorbed and translocated much as phosphates are. It is difficult to visualize a more effective way in which an herbicide might kill a plant.

Phytotoxicity of Organic Arsenicals

Injury symptoms on crop plants resulting from toxic quantities of arsenic in soils were noted in the 1930's, when it was found that young trees planted in old orchard soils grew slowly and were stunted. ⁷⁴⁸ Young apple trees, in addition to being stunted, had leaf symptoms that indicated water-deficiency stress, which implied injury to the roots; pears showed similar symptoms. ⁷⁸⁵

Peach trees planted on these old orchard soils that have accumulated lead arsenate exhibit by midsummer a red or brown discoloration along the leaf margins and then throughout the leaves. The discolored tissues die and drop out, giving the leaves a shot-hole appearance. Defoliation also occurs and may be complete by late summer. The injury appears first in the older leaves; young leaves on shoot tips may remain normal. Yields of fruit may be reduced, and the trees are usually stunted. Thompson and Batjer ⁷⁸⁵ performed experiments aimed at correcting arsenic injury to peach trees. They found correlations between shot-holing and defoliation and between leaf arsenic content and defoliation; arsenic content varied between 1 and 4 ppm. They found that zinc sulfate applied at 10 lb (4.5 kg) per tree reduced defoliation; nitrogen application at 1.4 lb (0.6 kg) per tree also reduced defoliation; a combination of these two treatments reduced defoliation that had been as high as 81% to around 2–3%, and even eliminated it in some orchards. Repeat treatments with zinc in later years had little or no effect, but maintaining a high nitrogen content reduced defoliation at later years.

Lindner and Reeves ⁴⁷³ explained that arsenic injury was confused with western X-disease, which is caused by a virus. They described the symptoms of the viral disease and arsenic toxicity; both cause shot-holing and defoliation, but arsenic-affected trees have greener leaves, and X-disease causes some chlorosis. Leaves that showed arsenic injury symptoms contained arsenic at 2.1–8.2 ppm; normal leaves, 0.9–1.7 ppm. Viral-diseased trees may produce deformed fruits, which drop prematurely; fruits of arsenic-affected trees are of normal form and remain on the tree for the normal period. Arsenic analyses of leaves provide the most accurate diagnosis.

Woolson ⁸⁷³ studied the uptake and phytotoxicity of arsenic in six vegetable crops grown in a greenhouse. Sodium arsenate was mixed into three soil types; after moistening, the cultures were left for a month for the arsenic to come to equilibrium. Sensitivity to the arsenic decreased in the following order: green beans, lima beans, spinach, cabbages, tomatoes, and radishes. All crops were fertilized at rates indicated by standardized soil tests and crop needs. The yields indicated that arsenic was generally most phytotoxic in the Lakeland soil; no plants grew when the arsenic concentration was 500 ppm. At 10, 50, and 100 ppm, crops survived; growth was proportional to arsenic concentration. The amount of available arsenic in some treatments continued to change throughout the 19-month experimental period; some of the change may have been the result of the addition of phosphate fertilizers, particularly where available arsenic reached a minimum and then increased. Plant growth at any particular degree of available arsenic in the soil may be affected by the amount of available phosphorus in the soil solution. ⁸⁷⁸

The organoarsenical herbicides are not growth-regulators in the way that plant hormones are; they apparently act through or on enzyme systems to inhibit growth. They kill relatively slowly; the first symptoms are usually chlorosis, cessation of growth, and gradual browning, and dehydration and death follow. Rhizomes and tubers may show browning of the storage tissues; buds fail to sprout, and the whole structure eventually decomposes. Cotton plants treated by directed spraying with MSMA show retarded growth from which they may recover. ⁴⁴⁷

When resprouting of tubers or rhizomes does occur, treatment should be repeated when some of the leaves have reached full size; treatment before then will not result in translocation, because movement of the assimilate stream into the underground organs is necessary to carry the toxicant to the proper sites of action.

Rumberg et al. ⁶⁹² reported DSMA-induced loss of chlorophyll in crabgrass within 2–3 days of treatment at 75 or 85 F (24 or 29 C), but it was hardly noticeable after 5 days at 60 F (16 C); the results are summarized in Table 5-1.

TABLE 5-1

Effect of Temperature after Treatment on the Degree of Chlorosis Induced by DSMA in Crabgrass.

Others have observed symptoms on various annual plants, but few detailed descriptions are noted in the literature. Many annual weeds are ultimately desiccated and become necrotic after treatment. Regrowth from axillary buds is usually chlorotic. Cotton and other plants become deeply pigmented (red) in the stem and petiole and even the leaves, if the arsenicals are applied in sublethal dosages.

On perennials, such as Johnsongrass, chlorosis does not develop before necrosis on sprayed foliage, but regrowth usually is chlorotic for some period. ⁴¹⁹ On purple nutsedge (Cyperus rotundus), Holt et al. ³⁶⁶ described symptoms as “a chlorotic appearance which starts at the leaf base and progresses toward the leaf tip until the entire leaf is chlorotic” and as “first visible four days after the initial applications of AMA.”

On hardstem bulrush (Scirpus acutus), an aquatic species, stems became chlorotic; necrosis proceeded from the tip concurrently with development of a tan discoloration over the length of the stem; after a month, the entire stem became brown and collapsed. ⁴¹⁸ Stems regrowing from the bulrush rhizome were chlorotic with necrotic tips and usually died.

Lange ⁴⁵⁰ studied MSMA toxicity symptoms on stone fruits by spraying MSMA at 4 and 16 lb/acre (4.5 and 17.9 kg/ha) on the bottom one-third of the foliage. He observed a spotty chlorosis of the leaf, followed by necrosis of all or part of the leaf and often defoliation. Untreated upper leaves and new growth showed symptoms of injury that indicated movement of the toxic material to the untreated area.

Many factors could affect response; they include herbicide application rate and formulation, surfactant, timing, volume of carrier, quality of diluent, pH, timing of evaluation, ecotypes, senescence, stage of growth, dormant-season disturbance of root systems before treatment, fertility, moisture availability and continuity, plant competition, temperature, light intensity, and insect and mechanical wounding of foliage before treatment. Any of these can have a dominant effect on the response.

Several researchers who have studied methanearsonates have reported temperature-influenced results. Sckerl et al., ⁷¹⁹ Kempen et al., ⁴¹⁹ Laurin and Dever, ⁴⁵⁴ Riepma, ⁶⁶⁹ and Bounds ⁹⁰ all mentioned that toxicity was greater at higher temperatures. Toxicity was less on Johnsongrass in regions of California influenced by cool marine air, and higher rates of application were required than in hotter regions. One controlled-environment study of Rumberg et al. ⁶⁹³ indicated that chlorosis occurred considerably earlier in crabgrass at higher temperature (24 or 29 C, versus 16 C), and injury, as measured by dry weight after 10 days, was greater at higher temperature. With sodium arsenite and cacodylic acid, temperature had no effect.

Kempen ⁴¹⁷ found that relatively high temperature (35 C) and light [2,800 ft-c (30,140 1x)] increased the necrosis of Johnsongrass foliage and the kill of the rhizomes, compared with low temperature (15 C) and light [320 ft-c (3,445 1x)]. He found that only 1 day was required for 50% necrosis of the foliage at the higher temperature and light, but 12 days at the lower temperature and light. McWhorter ⁵³⁵ and Kempen et al. ⁴¹⁹ suggested that droughtiness after application increased weed control.

In studies on the effects of AMA on food reserves in purple nutsedge, Duble and Holt ²¹⁵ found, by tracer experiments and chemical analysis, that starch disappeared and arsenic increased in tubers of plants that were given repeated applications of the herbicide. In general, AMA-treated plants had a higher rate of utilization of the products of photosynthesis than untreated plants. Apparently, carbohydrates were utilized in preference to fats and proteins.

From the results of Woolson et al., it is evident that total arsenic in a soil is not correlated with phytotoxicity; correlation between plant growth and available arsenic is better. ⁸⁷⁶ Some soils can remove arsenic from the soil solution more rapidly and more completely than other soils by fixation on soil colloids. In these experiments. Hagerstown soil removed 2–5 times more arsenic than did the other two soils. Woolson et al. concluded that the large amount of available arsenic in the Christiana soil may have resulted from the high phosphorus content, which prevented formation of insoluble iron arsenates through competition for reaction sites on the surface of soil particles. Also involved is the larger amount of available aluminum in the Hagerstown soil. Woolson, Axley, and Kearney ⁸⁷⁷ have shown that, at high arsenic concentration in some soils, aluminum is more important than iron in removing arsenic from the soil solution.

Arle and Hamilton ²⁶ found that topical applications of MSMA affected growth of cotton more than applications of DSMA. There were usually no deleterious effects of single treatments with DSMA; single applications of MSMA later and at higher rates reduced yields. Repeated treatments with MSMA reduced yields more than DSMA treatments.

Keeley and Thullen ⁴¹⁴ studied the responses of cotton plants to topical applications of MSMA, DSMA, and MAA at 13, 20, and 31 C. DSMA proved to be less injurious to young cotton plants than MSMA. Injury by MSMA was severe at 13 C, intermediate at 20 C, and low at 31 C, and inclusion of 0.4% surfactant with the MSMA increased the injury. Injury by DSMA was intermediate at 13 C, low at 20 C, and lacking at 31 C, and inclusion of surfactant in the spray solution increased injury only slightly.

Cotyledons of cotton seedlings absorbed [¹⁴C]MAA and [¹⁴C]DSMA. The MSMA solution in these experiments was adjusted to a pH of 6.4, and the DSMA to a pH of 10.4. Autoradiographs of cotton plants treated with ¹⁴C-labeled MSMA, MAA, and DSMA showed evidence of greatest absorption and translocation of MSMA at 13 C, slight translocation at 20 C, and no translocation at 29 C. Very little [¹⁴C]MAA or [¹⁴C]DSMA was translocated. These results are contrary to the generalization for translocation of assimilates and other tracers, which normally penetrate and move more readily at high temperatures. (Rumberg et al. ⁶⁹³ found chlorosis from DSMA treatment and the translocation of DSMA to increase with temperature.)

Interactions between Arsenicals and Nutrients

Several studies have been conducted on interactions between phosphorus and arsenic in soils and nutrient solutions. Because these two elements have somewhat similar chemical characteristics, substitution of arsenic for phosphorus might occur in plant metabolic products. Rate trials in soil and nutrient solutions, however, have yielded conflicting results, partially because available phosphorus and arsenic concentrations have not generally been determined.

Schweizer ⁷¹⁵ showed that high phosphorus content increased the toxicity of DSMA to cotton, but there was considerable variation between the two soil types tested.

Little is known of interactions between arsenic and phosphorus in plants. Everett ²³⁹ indicated that phosphorus increased the arsenic content of bluegrass in a turf treated with tricalcium arsenate. However, he found that phosphorus reduced absorption of tricalcium arsenate (measured as arsenic) from nutrient solutions from 246 to 29 ppm. He stated that phosphorus at 100 ppm reduced the soluble arsenic in the nutrient solution from 15.8 ppm to 2.6 ppm. This might account for the lack of increase in arsenic uptake with high phosphorus in nutrient solutions. Everett also indicated that crabgrass absorbed twice as much tricalcium arsenate as did bluegrass; this suggests a species difference. Sckerl, ⁷¹⁶ in his review of the literature, indicated that phosphorus reduced arsenic toxicity.

Webb ⁸⁴³ suggested that arsenates inhibit various phosphatase enzymes about as potently as phosphate and probably combine with the enzymes in a similar manner. Sckerl ⁷¹⁶ related that arsenate competes with phosphorus for uptake and transport in the cell.

That arsenicals might interact with zinc was indicated by work of Batjer and Benson ⁵⁵ and Martin (personal communication, 1968). Batjer and Benson showed that toxicity in peaches (but not apples) grown in arsenic-contaminated soils could be reduced by foliar applications of zinc or iron chelates or soil applications of zinc or iron sulfates. ⁵⁵ Zinc chelates worked best and reduced the symptoms and the arsenic content of peach leaves. Martin related that orchardists in the northwestern United States use zinc sulfate at 5 lb (2.3 kg) per tree plus generous amounts of ammonium sulfate when starting peach trees in high-arsenic soils.

Burleson and Page ¹¹⁹ did root studies with flax that indicated that, with absorption of more than optimal phosphorus, phosphorus and zinc reacted together in a manner that reduced either their mobility or their solubility. Sharma et al. ⁷²² showed that translocation of zinc to shoots was inhibited by high soil phosphorus.

Krantz and Brown ⁴³⁸ published a list of zinc- and iron-sensitive plants; there was no obvious correlation between symptoms of deficiency and susceptibility to methanearsonate sprays.

Mode of Action

Considering the overall action of arsenites as herbicides, it seems important that they are able to penetrate the cuticle and enter into the apoplast phase of the plant system. Here, they may move with transpiration water and bathe the cells of the foliar organs to which they have been applied. At low concentration, it seems possible that arsenites are absorbed into the symplast and then translocated for at least short distances. Under most conditions in which these compounds have been used in the field, their concentrations have been such that rapid contact injury has precluded extensive translocation. This is related at least partly to their rapid effect in membrane degradation.

The arsonates, in contrast, have much lower contact toxicity; they are absorbed and translocated, at least in species that have succumbed to treatment, such as Johnsongrass and nutsedge. In susceptible perennial weeds, the great virtue of MSMA and DSMA has been their ability to penetrate into and destroy underground tubers and rhizomes. Thus, with a few repeated applications, these arsenicals have controlled two of the most serious perennial weed species—species that have resisted control by any other means.

As topical sprays, these compounds are inactivated almost instantaneously on contact with the soil and therefore may be used with impunity in many row crops; cotton is one of the more important of these. Although arsenicals in ordinary herbicidal dosages are rapidly rendered unavailable to plants in the soil, and although most soils have a very great capacity to inactivate and hold arsenic, arsenic residues in soils may eventually become troublesome. For this reason, in any weed control activity involving arsenical herbicides, integrated programs of herbicide rotation should be used. If such programs are used, occasional application of the organic arsenicals in the particular roles in which they are highly effective may not result in soil residues of any significance.

As for the chemical mechanisms by which the organic arsonates kill plants, their relatively slow action involving translocation and producing chlorosis as a primary symptom seems to implicate disturbance of phosphorus metabolism. Not only are they absorbed and translocated in plants much as are phosphates, but also they affect many organelles in the cells, including the chloroplasts, in all of which phosphorus plays important roles. ³⁵

This interpretation is further strengthened by evidence of Schweizer ⁷¹⁵ that the addition of phosphorus to two silt-loam soils increased the toxicity of DSMA to cotton, possibly by saturating sites in these soils on which both arsenate and phosphate are fixed. As early as 1934, Albert ⁶ reported that residues of calcium arsenate became more toxic to several crops where phosphate fertilizer was applied heavily. There is substantial evidence that phosphates and arsonates tend to replace each other chemically, but that arsenic cannot serve the many essential roles of phosphorus in plants.

The uncoupling of oxidative phosphorylation and the formation of complexes with sulfhydryl-containing enzymes may also enter the picture of arsenic phytotoxicity. However, trivalent arsenic is the form commonly associated with these effects, and this would implicate arsenites, rather than arsenates or arsonates.

Factors That Can Influence the Toxicity of Arsenic Compounds

Table 5-2 summarizes information on the toxic and no-effect doses of several arsenic compounds. The variety of systems described in the table suggests some of the factors that can influence the toxicity of arsenic. For example, the studies of Harrison et al. ³²⁹ illustrated effects that the chemical and physical properties of arsenic trioxide can have on its acute toxicity. The toxicity of arsenic trioxide depended on the purity of the substance used; the crude commercial material was less toxic than the purified material. Although the impure arsenic trioxide was less toxic, it caused much more gastric and intestinal hemorrhage than the purified arsenic trioxide. This study and many of those discussed below were carried out in rats, which have a highly peculiar metabolism of arsenic (see Chapter 4). The reader should be aware of this problem and understand that the rat is an animal to be avoided in arsenic research.

TABLE 5-2

Toxic and No-Effect Doses of Some Arsenic Compounds.

In addition to testing the oral toxicity of aqueous solutions of the “crude” and “pure” arsenic trioxide, Harrison et al. ³²⁹ investigated the acute toxic effects of both preparations when given in the dry state mixed in feed. When arsenic trioxide was administered in this manner, the toxic dosage was almost 10 times as high as when it was given in aqueous solution, regardless of whether the crude or the pure material was tested. These results are in accord with those of Schwartze, ⁷¹³ who found that a solution of arsenic trioxide was more toxic than the undissolved compound and that the toxicity of different preparations of solid arsenic trioxide administered orally varied markedly, depending on their coarseness or fineness. The dependence of the toxicity of arsenic trioxide on the physical form in which it is given is probably a result of its rather poor solubility; e.g., Harrison et al. ³²⁹ commented that heating was required to solubilize arsenic trioxide. The practical consequences of the great variation in toxicity of arsenicals due to their different solubilities were recently emphasized by Done and Peart, ²¹⁰ who criticized government regulations that equated the poison hazard of the highly soluble sodium arsenite with that of the less soluble (and thereby less toxic) arsenic trioxide (although the latter does have a toxic potential).

Harrison et al. also demonstrated a species difference in the resistance to acute poisoning with arsenic trioxide: mice were less affected by the arsenic compound than were rats. Similar species differences were shown by Kerr et al., ⁴²² who noted that turkeys and dogs were more susceptible to the toxic effects of the organic arsenical 3-nitro-4-hydroxyphenylarsonic acid than were chickens and rats. McChesney et al. ⁵²⁸ reported that sodium p-N-glycoloylarsanilate was about 20 times as toxic to cats as to mice. The work of Harrison et al. ³²⁹ revealed that even different strains of mice had very different abilities to tolerate arsenic trioxide. These species and strain differences in the toxicity of arsenic could have important implications regarding the use of laboratory animals as predictive models for human response. The average estimated fatal dose of arsenic trioxide for humans is 125 mg. ⁸¹⁵ For a 70-kg man, this is equivalent to about 1.4 mg of arsenic per kilogram of body weight. Thus, a human is much more sensitive to the toxic effects of arsenic on a weight basis than a rat, and it is obviously dangerous to extrapolate results from rodents to humans.

Another factor that can influence the toxicity of arsenic is the valence of the element. Direct comparison of the intraperitoneal LD ₇₅ of sodium arsenite and sodium arsenate in the rat shows that the trivalent form of arsenic is about 4 times as toxic as the pentavalent form. ²⁶³ This difference due to valence is also seen in tissue-culture studies in which many confounding metabolic effects are avoided. ⁷⁰⁴ Differences related to valence apply to the organic arsenicals, as well as to the inorganic. In fact, the greater toxicity of the trivalent form versus the pentavalent is such a good generalization that a microbiologic assay for distinguishing As(III) and As(V) on the basis of their different toxicities has been suggested. ⁴⁸⁹

The toxicity of a number of synthetic aromatic organic arsenicals has been the subject of several investigations, because of the value of these compounds for improving weight gain and feed efficiency in swine and poultry. Generally speaking, the organic forms of arsenic are considered less hazardous than the inorganic forms, and this is shown by the tissue-culture work of Savchuck et al. ⁷⁰⁴ Arsenic concentrations of 84 and 65 ppm in the form of 3-nitro-4-hydroxyphenylarsonic acid and sodium arsanilate, respectively, were needed to cause a 50% growth inhibition of HeLa cells, whereas a concentration of only 2 ppm in the form of sodium arsenate was required to inhibit growth by the same amount. However, feeding arsenic at 114 ppm in the diet of rats as 3-nitro-4-hydroxyphenylarsonic acid caused an 83% mortality in 4 days, ⁴²² whereas arsenic at 125 ppm in the diet as sodium arsenate caused only a mild growth depression after 12 weeks. Furthermore, the intraperitoneal LD ₅₀ of arsenic as 3-nitro-4-hydroxyphenylarsonic acid is 18.8 mg/kg in rats, ⁴²² whereas the LD ₇₅ of arsenic as sodium arsenate is given as 14–18 mg/kg. ²⁶³ To put the above studies in the proper perspective, it should be pointed out that the recommended concentration of 3-nitro-4-hydroxyphenylarsonic acid for feed-additive use is only 25–50 ppm in the diet, or arsenic at 7–14 ppm.

Arsanilic acid is another aromatic organic arsenical that is widely used as a growth-promoter, and Frost et al. ²⁷³ reported that it had no harmful effects on rats over several generations when fed at 500 ppm in the diet. However, Notzold et al. ⁵⁹⁰ noted an incoordinated gait in swine, and Al-Timimi and Sullivan ¹² saw a growth inhibition in turkeys when arsanilic acid was fed at an arsenic concentration of 400 ppm. Again, to put these studies in perspective, it should be emphasized that the recommended concentration of arsanilic acid for feed-additive use is only 50–100 ppm.

Perhaps of more direct concern to consumers is the toxicity of arsenic that occurs naturally in seafood, such as shrimp (such arsenic compounds are commonly referred to as “shrimp arsenic”). Public awareness of this potential hazard was heightened by an article in Consumer Reports. ²⁷⁵ However, many of the allegations in the article were disputed in a press report. ⁵⁴⁸ The work of Coulson et al. ¹⁷¹ showed that rats (and humans) do indeed absorb shrimp arsenic from the gastrointestinal tract readily, but this form of arsenic is excreted rapidly in the urine. These authors also found no evidence of toxic effects of feeding “shrimp arsenic” at 17.7 ppm in the diet of rats for 52 weeks. Criteria of toxicity included growth, physical appearance, activity, and histologic appearance of the liver, spleen, and kidneys.

No toxic symptoms were reported by Morgareidge, ⁵⁶⁰ who supplemented rat diets at 16 ppm with protein-bound arsenic derived from the livers of turkeys whose diets had contained 0.56% p-ureido-benzenearsonic acid (carbarsone). Welch and Landau ⁸⁴⁸ observed no toxic reactions in rats fed a diet containing 1% arsenocholine (the arsenic analogue of choline) for a week. The apparent lack of toxicity of this arsenic compound may be of considerable interest, in light of the work of Lunde, ⁴⁹³ who discovered in fish oils two arsenolipids with chemical properties that resemble those of phospholipids.

Still another class of organic arsenical compounds is the aliphatic arsenicals, such as cacodylic acid and the sodium salts of methanearsonic acid, which are discussed later. Although these compounds are used widely as herbicides, their toxicity is less than that of inorganic arsenical herbicides.

The Problem of Toxic versus “No-Effect” Dosages

In addition to the complexities just discussed, there is a factor that confounds the interpretation of toxicologic data—namely, the criterion used to judge whether a given dosage is toxic. For example, arsenic at 5 ppm in the drinking water as sodium arsenite from weaning until death is not toxic to rats, with respect to growth or life span, ⁷¹¹ and is only slightly toxic to mice. ⁷¹⁰ An identical experiment, however, carried out through three generations of mice revealed that the ratio of males to females born increased in mice exposed to arsenic, compared with controls. ⁷¹² It was concluded that exposure to some trace elements in dosages that do not interfere with growth or survival may affect reproduction. Thus, a more sensitive indicator of toxicity showed the detrimental effects of a dosage that had previously been regarded as “safe.” With sophisticated assessment techniques, such as biochemical and enzyme measurements, even more subtle effects of poisons can be detected. For example, Bencko and Simane ⁶⁶ found that the respiration rate of liver homogenates prepared from mice that had received arsenic at 5 ppm as arsenic trioxide in their drinking water was only 61% of that of the normal controls. However, the dosage of arsenic used by Bencko and Simane was 100 times higher than the currently recommended maximal concentration (0.05 ppm) of arsenic in drinking water. ⁵⁷⁸

An even more sensitive method for determining the toxic effect of arsenic compounds was used by Weir and Hine: ⁸⁴⁶ a conditioned-avoidance technique to assess the deleterious effects of various ions in the aquatic environment of fish. In this study, arsenic (as arsenate) was found to impair conditioned-avoidance behavior of trained goldfish after 48 h of exposure to a concentration of only 0.1 ppm, which is only 1/320 of the lethal concentration for 50% and only 1/15 of the lethal concentration for 1% of the fish. The relevance of this work to mammalian systems is far from clear, but the data suggest at least that similar behavioral experiments should be carried out with animals exposed to various substances suspected of being environmental hazards.

Mechanisms of Toxicity of Arsenic Compounds

The difference in toxicity between trivalent and pentavalent arsenic compounds can best be understood by considering the biochemical mechanisms of action of these two distinct families of compounds. This aspect of arsenic toxicology has been the subject of numerous reviews. ^{l7 , 397 , 634 , 762 , 815} The early work of Ehrlich, Voegtlin, and others suggested that organic arsenicals exert their toxic effects in vivo by first being metabolized to the trivalent arsenoxide form and then reacting with sulfhydryl groups of tissue proteins and enzymes to form an arylbis(organylthio)arsine:

Later work showed that several enzyme systems containing thiol groups could be poisoned in this way and that in most cases the activity of the enzyme could be restored by adding an excess of monothiol.

An important exception to this generalization, however, proved to be the pyruvate oxidase system, which could not be protected against trivalent arsenicals by even a 200% excess of monothiol. Such an apparent anomaly was clarified when it was shown that, under some circumstances, arsenicals can complex with two sulfhydryl groups in the same protein molecule, thereby forming a stable ring structure that is not easily ruptured by monothiols. This finding stimulated the testing of various dithiol compounds for their ability to block the action of arsenicals on pyruvate oxidase and led to the discovery of British antilewisite (BAL), which eventually became a widely used antidote for arsenic poisoning. The simultaneous interaction of arsenic with two thiol groups led Peters and associates ⁵⁹⁴ to postulate the existence of a dithiol-containing cofactor in the pyruvate oxidase system. This idea was later verified experimentally when lipoic acid was identified as a component of pyruvate oxidase. The reaction of an arsenoso compound with lipoic acid to yield a ring structure that can be cleaved by BAL is illustrated in Figure 5-2. This reaction summarizes what is currently felt to be the mode of action of trivalent monosubstituted arsenicals in exerting their toxic effects in biologic systems and illustrates the biochemical rationale for the use of BAL to counteract arsenic poisoning.

FIGURE 5-2

Reaction of lipoic acid with a trivalent monosubstituted arsenical, and regeneration of lipoic acid by addition of BAL.

It should be pointed out, however, that the toxic effects of inorganic trivalent arsenic (arsenite) can often be potentiated by BAL in vitro. Fluharty and Sanadi, ²⁵⁷ for example, showed that an equimolar mixture of arsenite and BAL uncouples oxidative phosphorylation in rat liver mitochondria and drew the conclusion that arsenite is the true active inhibitory species and that the BAL served only as a vehicle for transporting arsenite to a dithiol enzyme site. Siegal and Albers ⁷²⁹ found that addition of equimolar BAL decreased the arsenite concentration necessary to produce 50% inhibition of Electrophorus (electric eel) microsomal (Na⁺/K⁺)-ATPase from 6 mM to 0.1 mM. The authors suggested that the BAL–arsenite complex reacted directly with the enzyme. Wu ⁸⁸² performed a careful kinetic analysis of the dithiol-dependent inhibition of rat liver glutamine synthetase by arsenite and proposed the scheme shown in Figure 5-3. This scheme was thought to account for several facts regarding the dithiol-dependent inhibition by arsenite, including the ready dissociation of the enzyme–arsenite complex and the reversal of the inhibition by cysteine. McDonough ⁵²⁹ suggested that the BAL-arsenite complex can act as an inhibitor of germination, inasmuch as lettuce seeds soaked in mixed solutions of sodium arsenite and BAL yielded lower germination ratios than did seeds soaked in either compound alone.

FIGURE 5-3

Reaction of glutamine synthetase with BAL–arsenite complex.

The mechanism of action of the toxic effects of inorganic pentavalent arsenicals is less clearly understood than that of the trivalent arsenic compounds. It is possible that pentavalent arsenic is reduced to trivalent arsenic before exerting its toxic effects, ^{87 , 292} but whether that happens in vivo is controversial. ^{709 , 868} Unlike the trivalent arsenicals, the pentavalent forms do not appear to react directly with the active sites of enzymes. ³⁹⁷ Rather, arsenate can compete with inorganic phosphate in phosphorylation reactions to form unstable arsenyl esters, which then decompose spontaneously. ²¹¹ Arsenate has also been shown to uncouple oxidative phosphorylation, ¹⁸² presumably by competing with inorganic phosphate at one of the energy-conserving steps. Chan et al. ¹⁴² have isolated an arsenylated component of rat liver mitochondria that they feel may represent the arsenic analogue of a low-molecular-weight phosphorylated mitochondrial constituent that plays a role in oxidative phosphorylation. A nonhydrolytic mode of action of arsenate in inhibiting mitochondrial energy-linked functions has recently been proposed. ⁵⁴⁹

Adaptation to Toxicity of Arsenic Compounds

Most early investigators reported that animals were unable to adapt to the toxic effects of inorganic arsenic compounds, ⁷¹³ although adaptation to some organic arsenicals was readily achieved. ⁴⁴¹ In spite of the early failures to demonstrate adaptation to inorganic arsenic, Bencko and Symon ⁶⁸ have recently shown that the LD ₅₀ for arsenic as arsenic trioxide administered subcutaneously could be increased from 10.96 to 13.98 mg/kg in hairless mice as a result of giving arsenic at 50 ppm as arsenic trioxide in the drinking water for 3 months. Additional evidence that suggested an adaptive response to arsenic was the finding that the decreased metabolic oxygen consumption observed with the liver homogenates from mice given arsenic at 50 ppm in the water for 32 days returned to normal after 64 days in the experiment. ⁶³ However, no such adaptation was seen in mice given arsenic at either 5 or 250 ppm in the water. Moreover, this experiment seems somewhat inconsistent with an earlier report from the same laboratory, which showed that liver homogenates from mice given arsenic at 50 ppm in the water for 256 days exhibited a decreased consumption of oxygen. ⁶⁶ Apparently, the range of experimental conditions under which adaptation to arsenic can be obtained is quite limited.

The mechanisms of these adaptive responses to arsenic are not known, but Bencko and Symon ⁶⁷ found that mice given arsenic at 50–250 ppm in the water accumulated arsenic in the liver and skin until the sixteenth day of the experiment, after which retention decreased. The authors suggested that either decreased absorption or increased excretion of arsenic could account for their results. Studies with [⁷⁴As]arsenate revealed that mice previously exposed to arsenic at 50 ppm as arsenite in the water for 64 days displayed a significant decrease in the retention of a later dose of radioarsenate administered parenterally. ⁶⁵ Although the authors interpreted their results as evidence of an increase in capacity of the excretory mechanism for arsenic due to arsenic exposure, this experiment could also perhaps be explained by a saturation of the tissue binding sites for arsenic by previous arsenic intake, which could then cause an “apparent” increase in the excretion of the element. Clearly, more research is needed to determine whether animals are able to adapt to the toxic effects of inorganic arsenicals.

Experimental Inhalation Toxicity

Air pollution due to arsenic is a particular problem in some parts of Czechoslovakia, because of the high arsenic content of some coal burned in power plants there. Consequently, Bencko and associates of the Institute of Hygiene in Prague have carried out a number of studies concerned with the experimental inhalation toxicity of arsenic. ^{62 , 69} These workers pioneered the use of the hairless mouse for such investigations, because other animal models had several disadvantages for research in arsenic toxicology. ⁶⁸ The rat, whose arsenic metabolism is peculiar (arsenic tends to accumulate in the blood), was ruled out as a test animal. The rabbit and guinea pig were also considered unsuitable for these studies, because in some of the work arsenic was to be administered via the drinking water, and the variable consumption of fresh vegetables by the animals would contribute to an irregular water intake. However, the hairless mouse had a number of experimental advantages for inhalation toxicity determination. First, it could not put its nose into its fur to “filter” the air being breathed. Moreover, there was no hair to trap the arsenic-containing dust; such dust could otherwise be ingested later as a result of cleaning or grooming. Finally, the lack of hair on the mouse enabled the investigators more readily to determine any dermatologic changes caused by the arsenic.

Hairless mice were exposed to fly ash whose particle size was less than 10 µm and that contained 1% arsenic in the form of arsenic trioxide. The exposures were carried out on 5 days/week for 6 h/day in dust chambers specially designed for the application of solid aerosols. The mean arsenic concentration in the dust chamber was 179.4 ± 35.6 µg/m³ of air, which was about 3 times higher than the maximal concentration of arsenic found in the vicinity of the offending power plants. During the first 2 weeks of the experiment, there was a considerable increase in the concentration of arsenic in the livers, kidneys, or skin of the exposed mice. However, there was a significant decrease in the arsenic content of the liver and kidney samples during the fourth week of exposure, although no such decrease was observed in the skin. This decline in the arsenic content of tissues was similar to that seen in animals given arsenic orally and suggested an adaptation to arsenic. Unfortunately, these workers did not carry out physiologic measurements to evaluate effects of the arsenic exposure on the metabolism of the experimental animals.

Zharkova ⁸⁹⁰ studied the effect of continuous 24-h exposures to arsenic trioxide at 25–37 µg/m³ of air on various physiologic characteristics in rats. He found that such treatment resulted in a lag in weight gain, disordered chronaxy ratios of antagonist muscles, suppression of cholinesterase activity, a reduction in concentration of sulfhydryl groups in blood proteins, an increase in the number of reticulocytes, a decrease in blood hemoglobin, porphyrinuria, a reduction in ascorbic acid in all organs and tissues, and accumulation of arsenic in the organs and tissues. Few experimental details were presented in the translated paper, so it is difficult to assess the biologic significance of these results. Although the physiologic importance of disordered chronaxy ratios, suppressed enzyme activities, and reduction in sulfhydryl groups in proteins might be questioned, weight lag, anemia, and porphyrinuria are more difficult to ignore. Moreover, the concentrations of arsenic used in the investigation were very low, although they still exceeded the occupational exposure standard for inorganic arsenic recently recommended by the National Institute for Occupational Safety and Health (NIOSH). ⁸⁰⁹

Rozenshtein ⁶⁸⁸ investigated the effect of continuous exposure to arsenic trioxide aerosols on albino rats. He found that round-the-clock exposure to arsenic trioxide at 60.7 µg/m³ produced inhibition in the central nervous system, reduced the content of sulfhydryl groups, inhibited cholinesterase activity, and raised the concentration of pyruvate in the blood. Similar continuous exposure to a concentration of 4.9 µg/m³ caused disturbances of conditioned reflexes and of the chronaxy ratio of antagonistic muscles and a reduction in the content of sulfhydryl groups in the blood. Exposure to both concentrations resulted in a marked accumulation of arsenic in the body and morphologic alterations in the organs and tissues. Inasmuch as no functional, biochemical, or morphologic alterations were observed when the animals were exposed to arsenic trioxide at 1.3 µg/m³, Rozenshtein recommended 1 µg/m³ as the maximal mean diurnal permissible concentration of this compound in the atmosphere. ⁶⁸⁸ Again, the physiologic implications of these results are not clear.

The main drawback in the research of both Zharkova and Rozenshtein was that the test animal used was the rat, which has a peculiar arsenic metabolism. Also, the rats had considerable body hair, which could trap the arsenic aerosol. The arsenic trapped in this way could then be ingested by the animal as a result of cleaning and grooming. Finally, the rats used in both studies were exposed to arsenic trioxide on a continuous 24-h/day basis, whereas the NIOSH standard was meant to apply only in situations of intermittent exposure. Nonetheless, further research should be carried out to evaluate both experiments. ⁸⁰⁶

Arsenicals and Resistance to Infection

Gainer and Pry ²⁷⁸ showed that virus-infected mice treated with large doses of arsenicals had higher mortality rates than untreated controls. Viral diseases so affected by arsenic included pseudorabies, encephalomyocarditis, and St. Louis encephalitis. Although several experimental protocols were used, arsenical treatment generally consisted of injecting subacute doses of sodium arsenite at the time of inoculation with virus or administering sodium arsenite or 3-nitro-4-hydroxyphenylarsonic acid at rather high concentrations of arsenic (75–150 ppm) in the drinking water for various periods before or after inoculation. In one case (western encephalitis virus), mortality was significantly reduced if the mice were given sodium arsenite at the time of inoculation with virus, but mice treated with 3-nitro-4-hydroxyphenylarsonic acid in the drinking water after inoculation had higher mortality than did controls. British antilewisite did not inhibit, but appeared to stimulate, the mortality-increasing activity of sodium arsenite in pseudorabies infection. This observation is consistent with other reports that under some circumstances BAL may potentiate the toxicity of arsenicals.

The protective effect of a synthetic double-stranded homopolynucleotide complex of polyinosinic acid and polycytidylic acid (poly I/poly C) against mortality in mice due to western encephalitis virus was inhibited by sodium arsenite treatment. Because the protective role of poly I/poly C against viral disease has been associated with the action of interferon, Gainer and Pry ²⁷⁸ hypothesized that the arsenical stimulation of mortality after inoculation with viruses was at least partially explainable by interferon dysfunction. Indeed, a later paper by Gainer ²⁷⁷ showed that the induction of interferon by poly I/poly C in rabbit kidney cell cultures could be inhibited by sodium arsenite. It was found somewhat unexpectedly, however, that, although high concentrations of arsenite inhibited the action of exogenous mouse interferon added to cultures of mouse embryo cells, low concentrations of arsenite increased the antiviral activity of low concentrations of interferon.

The research of Gainer and Pry ²⁷⁸ seems to have two major ecologic consequences. First, exposure to large doses of arsenic clearly impairs a mouse's ability to resist viral disease. However, the doses of arsenic used in these studies were such that any relevance of these data to human pollution problems would have to be limited to outbreaks of massive arsenical toxicosis, such as the Morinaga dry milk incident ⁵⁷⁵ and the Ube soy sauce episode. ⁷¹¹ In this regard, the follow-up study revealed that children poisoned in the Morinaga incident, among other problems, also had a decreased resistance to infection. The mechanism of action of arsenic in decreasing resistance to infection is not known with precision, but the results of Gainer indicate that decreased interferon production or action may be involved. A nonspecific effect of heavy-metal poisoning cannot be ruled out, inasmuch as Hemphill et al. ³⁴³ have shown that mice treated with lead had a greater susceptibility to challenge with Salmonella typhimurium than controls that received no lead. A decrease in resistance to infection might very well be expected in any group of animals additionally stressed by exposure to a metabolic poison. Although Koller and Kovacic ⁴³³ recently found that exposure to lead decreased antibody formation in mice, Gainer and Pry were able to rule out alterations in antibody formation or action as the primary factors accounting for the stimulating effects of arsenicals on the mortality of their virus-infected mice.

A second ecologic consequence of Gainer's experiments is related to the remarkable observation that low concentrations of arsenic appeared to increase the ability of exogenously added mouse interferon to block the infection of cultured mouse embryo cells with vesicular stomatitis virus. ²⁷⁷ Although the author himself had some reservations concerning the proper interpretation of his data, the stimulation of interferon action by arsenic seemed to be real. Gainer suggested that an increased antiviral activity of interferon could provide a rationale for the beneficial “growth-promoting” effects of arsenical feed additives in livestock through reduction in disease incidence or severity. ²⁷⁷ Any conclusions regarding the possible effects of low concentrations of arsenicals in the environment on the ability of humans to resist disease must await further research.

Arsenic as Antagonist to Selenium Poisoning

Moxon ⁵⁶⁵ first demonstrated the protective effect of arsenic against selenium poisoning when he found that arsenic at 5 ppm as sodium arsenite in the drinking water largely prevented liver damage in rats whose diet contained selenium at 15 ppm as seleniferous wheat. Moxon and DuBois ⁵⁶⁶ then showed that arsenic was unique in its ability to prevent selenium toxicity; all other elements tested were unable to protect against all manifestations of chronic selenosis. Sodium arsenite and sodium arsenate were equally effective against seleniferous grain, but the arsenic sulfides were ineffective. ²¹⁸ Arsanilic acid and 3-nitro4-hydroxyphenylarsonic acid, two organic arsenicals used as “growth-promoters” for livestock, also exhibited a beneficial action against selenium poisoning in rats when given in the drinking water. ³⁴⁴ There is evidence that it would be practical to use these two agents to protect swine and poultry in high-selenium regions. ^{130 , 834} Amor and Pringle ¹³ even suggested the use of an arsenic-containing tonic as a prophylactic agent against selenium poisoning in exposed industrial workers.

The metabolic basis for the beneficial effect of arsenic in selenium poisoning remained confused for some time, because arsenic was known to block the biosynthesis of dimethylselenide, a detoxification product in animals that received subacute doses of selenium by injection. ⁵⁹⁹ Moreover, the protective effect of arsenic against dietary selenium was not seen if the arsenic was given in the diet, instead of the drinking water, ²⁸⁰ and Frost ²⁷¹ has shown that the toxicities of arsenic and selenium are additive if both elements are given in the drinking water. These results agree with those of Obermeyer et al., ⁵⁹³ who recently observed an additive toxicity between arsenite and trimethylselenonium chloride or dimethylselenide.

Ganther and Baumann ²⁷⁹ studied the influence of arsenic on the metabolism of selenium when both elements are injected in subacute doses and found that the excretion of selenium into the gastrointestinal tract was markedly stimulated by arsenic. Levander and Baumann ⁴⁶³ observed an inverse relationship in arsenic-treated rats between the amount of selenium retained in the liver and the amount excreted into the gut; and they concluded that the bile might be the route by which selenium was appearing in the gastrointestinal tract. This hypothesis proved correct when it was discovered that in 3 h over 40% of the selenium injected could be recovered in the bile of rats that also received arsenic, whereas only 4% of the selenium was excreted into the bile of rats not given arsenic. ⁴⁶⁴ This effect of arsenic on the biliary excretion of selenium was not confined to subacute-toxicity experiments: A response of selenium to arsenic was seen at dosages approaching a rat's daily intake of selenium when fed some crude commercial diets. Sodium arsenite was the most effective form of arsenic in enhancing the biliary excretion of selenium, but arsenate and 3-nitro-4-hydroxyphenylarsonate were also active to some extent. In experiments with radioactive arsenic, it was found that selenium stimulated the biliary excretion of arsenic, just as arsenic stimulated the excretion of selenium. Initial attempts to characterize the forms of selenium in rat bile suggested that the element is probably present in several forms, including some macromolecularly bound selenium.

Although these studies provide an understanding on a physiologic basis of how arsenic counteracts selenium toxicity, the chemical mechanism of the process is still far from clear. The most logical hypothesis to account for the arsenic–selenium antagonism from the molecular point of view assumes that arsenic combines with selenium—perhaps, in analogy with sulfur chemistry, by reacting with selenol (–SeH) groups—to form a detoxification conjugate that passes readily into the bile.

Nutritional Essentiality of Arsenic

A number of older reports suggested that arsenic in small amounts may play a useful metabolic role in tissues, rather than being merely an accidental contaminant. Underwood ⁸⁰¹ has cited several of these early references that claimed beneficial effects of arsenic, including stimulation of growth in tissue cultures and enhancement of growth and metamorphosis of tadpoles. More recently, Askerov et al. ³⁶ have found that spraying leaves with a 0.002% arsenic solution leads to a 10% increase in viability of silkworm caterpillars and a 29% increase in cocoon yield.

Despite several subjective reports on the effects of arsenic on the appearance of the haircoat or in the prevention or cure of anemia, numerous attempts to induce an experimental dietary deficiency have failed, probably because of the ubiquity of the element. Hove et al. ³⁷¹ fed rats a milk diet fortified with iron, copper, and manganese and found that arsenic caused a slight initial delay in the decrease of hemoglobin when the minerals were withdrawn from the milk. However, these authors concluded that, if arsenic is essential for rats, the requirement must be somewhere below the 2 µg daily that was provided by the milk alone. Schroeder and Balassa ⁷⁰⁹ reported that rats and mice grew well and survived normally when they received only 0.26 µg of arsenic per 100 g of body weight per day in food. Skinner and McHargue ⁷³⁶ found that rats responded to arsenic supplements with increased hemoglobin when fed a ration composed mainly of skim milk powder and sucrose and adequately supplemented with iron and copper.

Sharpless and Metzger ⁷²³ have presented some evidence that arsenate can act as a mild goitrogen when fed in the diet at 5 ppm. To get a significant goitrogenic effect, however, such high concentrations of arsenic were needed that these workers concluded that there was only a remote possibility that arsenic could act as a positive goitrogenic agent in man.

A report by Muth et al. ⁵⁷⁰ suggested that arsenic may have some activity in preventing selenium-deficiency diseases, inasmuch as addition of arsenic at 1 ppm as sodium arsenate to selenium-deficient ration significantly reduced the incidence of myopathy in lambs. This observation has not been confirmed, however (Westwig and Whanger, unpublished data). Attempts to demonstrate a beneficial effect of arsenic in other selenium deficiency diseases, such as liver necrosis in rats ⁷¹⁴ and exudative diathesis in chickens, ⁶²¹ have been unsuccessful.

Using purely theoretical arguments based on the tissue distribution of various trace elements, Liebscher and Smith ⁴⁶⁹ decided that arsenic behaves more like an environmental contaminant than a nutritionally essential mineral. However, a recent preliminary communication has presented evidence of a requirement for arsenic by the rat. ⁵⁸⁶ To demonstrate an arsenic deficiency, the experimental animals had to be housed in plastic cages placed in laminar flow racks. The rats were fed a specially formulated purified diet that contained arsenic at only 30 ppb. The deficiency signs were most striking in male rats and included rough haircoat, low growth rate, splenomegaly, decreased hematocrit, and increased osmotic fragility of red cells. These preliminary results appear to have been verified by recent work describing an arsenic deficiency in goats and minipigs fed semisynthetic rations containing arsenic at less than 50 ppb. ¹⁹ Deficiency signs included impaired reproductive performance, decreased birth weights, increased neonatal mortality, and lower weight gains in second-generation animals. None of these deficiency signs were observed in control animals fed the semisynthetic diet supplemented with arsenic at 350 ppb.

Inorganic and Aliphatic Organic Arsenicals

Arsenic appears to be second only to lead in importance as a toxicant in farm and household animals. ^{115 , 335} Toxicoses caused by inorganic and aliphatic organic arsenicals are generally manifested by a syndrome entirely different from that caused by the phenylarsonic feed additives and therapeutic agents; therefore, the phenylarsonic compounds will be discussed separately.

Some of the more common sources of arsenic poisoning include grass clippings from lawns that have been treated with arsenical crabgrass-control preparations; grass, weeds, shrubbery, and other foliage that have been sprayed with arsenical herbicides; ^{113 , 114} dipping of animals in vats that years before had been charged with arsenic trioxide; and soils heavily contaminated with arsenic, either through the burning of arsenic formulations in rubbish piles or through the application of arsenical pesticides to orchards and truck gardens. ^{153 , 656} The more common sources of arsenic for small animals, especially cats, include ant and snail baits, which usually contain 1–2% arsenic. ¹¹⁵

Man and all lower animals are susceptible to inorganic arsenic poisoning, but poisoning is most often encountered in the bovine and feline species and results from the contamination of their food supply. The incidence of arsenic poisoning in these two species is closely followed in other forage-eating animals, such as sheep and horses. ⁵⁶⁴ Poisoning by inorganic arsenicals occurs only occasionally in dogs and rarely in swine and poultry.

The toxicity of inorganic arsenicals varies with the species of animal exposed, the formulation (e.g., trivalent arsenicals are more toxic than pentavalent), the solubility, the route of exposure, the rate of absorption from the gastrointestinal tract, and the rates of metabolism and excretion. ^{153 , 656} In practice, the most dangerous arsenic preparations are dips, herbicides, and defoliants in which the arsenic is in a highly soluble form. Unfortunately, animals often seek out and eat such materials as insulation, rodent baits, and dirt and foliage that have been contaminated with an inorganic arsenical.

Because so many factors influence the toxicity of arsenic, there is little point in attempting to state its toxicity in terms of milligrams per kilogram of body weight. The lethal oral dose for most species, however, appears to be 1–25 mg/kg of body weight as sodium arsenite, and 3–10 times that range as arsenic trioxide.

That the toxicity of an arsenical is greatly influenced by its solubility and particle size and thus by the extent of its absorption from the intestinal tract or skin is illustrated by an experiment conducted with swine. ¹¹⁵ Sodium arsenite was given in the feed at up to 500 ppm continuously for 2 weeks. The pigs readily ate the contaminated feed, but manifested no signs of acute arsenic poisoning. When the concentration was increased to 1,000 ppm, the pigs refused to eat the feed. When sodium arsenite was added to their drinking water at 500 ppm, severe poisoning and death occurred within a few hours. It was concluded that the lethal dose of sodium arsenite via drinking water was 100–200 mg/kg of body weight.

Experience with field cases of arsenic poisoning has indicated that animals that are weak, debilitated, and dehydrated are much more susceptible to arsenic poisoning than normal animals, probably because renal excretion is reduced.

Arsenic poisoning in most animals is usually manifested by an acute or subacute syndrome. Chronic poisoning, although it has been reported, is seldom seen and has not been clearly documented.

Arsenic affects tissues that are rich in oxidative systems, primarily the alimentary tract, kidneys, liver, lungs, and epidermis. It is a potent capillary poison; although all capillary beds may be involved, the splanchnic area is the most commonly affected. Capillary damage and dilatation result in transudation of plasma into the intestinal tract and sharply reduced blood volume. Blood pressure usually falls to the point of shock, and cardiac muscle becomes weakened; this contributes to circulatory failure. The capillary transudation of plasma results in the formation of vesicles and edema of the gastrointestinal mucosa, which eventually lead to epithelial sloughing and discharge of the plasma into the gastrointestinal tract. ⁶⁵⁶

Toxic arsenic nephrosis is more commonly seen in small animals and man than in farm animals. ^{115 , 148} Glomerular capillaries dilate, allowing the escape of plasma; this results in swelling and tubular degeneration. The anhydremia that results from the loss of fluid through other capillary beds and the low blood pressure contribute to the oliguria that is characteristic of arsenic poisoning. The urine usually contains protein, red blood cells, and casts. ¹¹⁵

After percutaneous exposure, capillary dilatation and degeneration may result in blistering and edema, after which the skin may become dry and papery. The skin may then crack and bleed, providing a choice site for secondary invaders. ⁶⁵⁶

Most textbooks report that arsenic is accumulated in the tissues and slowly excreted, but this appears to be true only in rats. Most species of livestock and pet animals apparently excrete arsenic rapidly. This phenomenon is very important when one considers arsenic content of tissues as a means of confirming suspected poisoning. Experience with field cases in the Veterinary Diagnostic Laboratory, Iowa State University, has indicated that, if an animal lives several days after consuming a “toxic” amount of arsenic, the liver and kidney tissues may contain less arsenic than is ordinarily considered diagnostic of arsenic poisoning. ^{110 , 112} Other authors have reported similar findings. ⁵⁶⁴

Phenylarsonic Feed Additives

Organic arsenical formulations have been used as feed additives for disease control and improvement of weight gain in swine and poultry since the mid-1940's. These compounds are phenylarsonic acids—arsanilic acid, 3-nitro-4-hydroxyphenylarsonic acid, 4-nitrophenylarsonic acid, and 4-ureidophenylarsonic acid—and their salts. The most widely used compounds are arsanilic acid; its sodium salt, sodium arsanilate; and 3-nitro-4-hydroxyphenylarsonic acid. ^{51 , 587} The additives are considered to improve weight gain and feed efficiency and to aid in the prevention and control of some enteric diseases of swine and poultry. ^{76 , 114 , 268 , 558 , 559}

There is still considerable discussion regarding the mode of action of the organic arsenicals. However, it seems certain that the phenylarsonic compounds have an action different from that of inorganic and aliphatic organic arsenicals. The arsenic incorporated in the additives is in the pentavalent form, and it is likely that they have their primary action as pentavalent arsenicals, which may account for their characteristic rapid renal excretion.

There have been several theories as to the possible therapeutic and nutritional effects of phenylarsonic feed additives. First, it is known that some organisms cause a thickening of the intestinal wall; thus, the additives may inhibit these organisms by interfering with their enzyme systems, which would result in a thinner intestinal wall and better nutrient absorption. A second theory is that the additives, by interfering with the development of the bacterial cell wall or by inhibiting normal cellular production of proteins and nucleic acids, lower the harmful bacterial population. A third possibility is that these compounds have a sparing action on one or more of the nutrients required by growing animals. ⁵¹

Some workers have suggested that both the toxicity and the efficacy of these compounds are due to their degradation and reduction to inorganic trivalent forms. ^{227 , 333 , 334 , 825} Eagle and Doak ²²⁷ reported that arsenoso compounds have direct activity, whereas arsonic acid compounds become active when they are converted to arsenoso compounds. Other research, however, clearly established that arsanilic acid and acetylarsanilic acid (4-acetylaminophenylarsonic acid) were excreted unchanged by chickens and that there is no evidence that these compounds are changed to any others or converted to inorganic arsenic. ^{183 , 555 , 613 , 614 , 616} Similar results were obtained in studies with 3-nitro-4-hydroxyphenylarsonic acid and 4-nitrophenylarsonic acid in chickens. Similar experiments by other workers with rats, rabbits, and swine indicated that the phenylarsonic acids for the most part are excreted unchanged by the kidneys, although some apparently undergo a limited amount of biotransformation. ^{556 , 557}

Because pentavalent arsenic compounds do not react readily with sulfhydryl groups and the phenylarsonic acids are apparently excreted unchanged, one must conclude that the mechanism of their action is something other than interaction with sulfhydryl-containing enzymes and proteins. The predominant lesions produced by these compounds in swine and poultry are peripheral nerve demyelination and gliosis, and it has been postulated that the phenylarsonic acids act to produce a vitamin B-complex deficiency, such as a deficiency of vitamin B₆ or B₁. ¹¹⁵ This postulation has not been studied experimentally.

When the phenylarsonic compounds are injected parenterally, they are mostly excreted in the urine within 24–48 h. When they are given orally, a considerable percentage is excreted in the feces. This indicates that they are poorly absorbed by the intestinal tract. The proportion that is absorbed, however, apparently is excreted rapidly by the kidneys. ^{51 , 555 , 556 and 557 , 614}

Recommended Uses and Factors Affecting Toxicity

The registered uses of the phenylarsonic acid feed additives are listed in Table 5-3. Arsanilic acid and sodium arsanilate are recommended at 50–100 ppm (0.005–0.01%) in swine and poultry feeds for improving weight gain and feed efficiency and for other uses. They are recommended at 250–400 ppm (0.025–0.04%) in swine feed for 5–6 days for the control of dysentery. The margin of safety for arsanilic acid and its salt is wide in normal animals. ²⁶⁹ However, the effective concentration and the chronic-toxicity concentration may impinge on one another under some conditions. The health of the exposed animals and management practices, especially those involving availability of water, are important contributing factors for adverse reactions to organic arsenical feed additives. Animals with diarrhea are usually dehydrated and thus are excreting very little urine. Because these arsenicals are excreted via the kidneys, their toxicity is greatly increased when they are given to animals with diarrhea. The morbidity is usually high, and the mortality very low. Experimentally, clinical signs appear after 3–10 days of exposure to high concentrations in the feed (e.g., 1,000 ppm) and within 3–6 weeks at lower concentrations (e.g., 250 ppm). ¹¹⁵

TABLE 5-3

Arsenic Compounds Used as Feed Additives.

The maximal safe dietary concentration of arsanilic acid for young turkeys (up to 28 days old) was reported to be between 300 and 400 ppm (0.03 and 0.04%). ¹²

Roxarsone, 3-nitro-4-hydroxyphenylarsonic acid, is recommended at 25–50 ppm (0.0025–0.005%) for chickens and turkeys and at 25–75 ppm (0.0025–0.0075%) for swine for improving weight gain and feed efficiency. It is also recommended at 200 ppm (0.02%) for 5–6 days for the control of dysentery. ⁵⁸⁷ Swine may exhibit clinical signs after consuming 250 ppm in the feed for 3–10 days and have been chronically poisoned by a concentration of 100 ppm for 2 months. ^{112 , 114}

4-Nitrophenylarsonic acid has been recommended for chickens and turkeys at 188 ppm for prevention of blackhead. It has not been recommended for ducks or geese and has only limited use as a feed additive.

Carbarsone p-ureidophenylarsonic acid) is recommended for the prevention of blackhead in turkeys and increased growth rate at 375 ppm in the feed.

Arsanilic acid and sodium arsanilate are most commonly used as swine feed additives. Toxicoses may occur, however, with any of the phenylarsonics in any of the species. The circumstances usually associated with toxicoses related to the organic arsenicals used as feed additives include:

• Purposeful incorporation of excessive amounts in feed or water. ⁴⁵⁸
• Mistaken feed formulation resulting in excessive amounts in feed.
• Prolonged and excessive administration in combination with other drugs.
• Treatment of animals with severe diarrhea and debilitation, which have increased susceptibility because of reduced renal excretion of the arsenical.
• Limiting of the water available to animals being exposed to therapeutic concentrations of organic arsenicals. ^{115 , 830}

Poisoning by organic arsenicals in swine is not uncommon and probably is second in frequency only to water-deprivation–sodium-iontoxicosis syndrome. ⁴⁵⁸

Signs and Lesions of Toxicosis

Acute clinical signs may appear after 3–5 days of exposure to high concentrations of phenylarsonic compounds in the feed. Signs include incoordination, inability to control body and limb movements, and ataxia. After a few days, swine and poultry may become paralyzed, but will continue to eat and drink (Figure 5-4). Arsanilic acid and sodium arsanilate may produce blindness, but this is rarely seen with 3-nitro-4-hydroxyphenylarsonic acid toxicosis. Erythema of the skin, especially in white animals, and sensitivity to sunlight may also be observed. The clinical signs are reversible up to a point. Removing the excess arsenical will result in recovery within a few days, unless the clinical signs have progressed to partial or complete paralysis resulting from irreversible peripheral nerve degeneration. ^{115 , 598}

FIGURE 5-4

Pig with quadriplegia after 18 days of feeding arsanilic acid at 900 g/ton (992 g/tonne). Reprinted by courtesy of Marcel Dekker, Inc., from Ledet et al. ^{(p. 443)}

Chronic poisoning occurs in swine and poultry when excessive but lower concentrations of phenylarsonic compounds are given in the feed or water for more than a few weeks. Animals will continue to eat and drink and remain alert while progressively developing blindness and partial paralysis of the extremities. The onset of signs is usually insidious and therefore not alarming to the herdsman. Goose-stepping, knuckling of the hock joints, and other manifestations of abnormal locomotion occur. Such animals usually have poor weight gain and feed efficiency.

Poultry usually become incoordinated and ataxic after consuming excessive concentrations of 3-nitro-4-hydroxyphenylarsonic acid, but more commonly exhibit ruffled feathers, anorexia, depression, coma, and death when exposed to excessive concentrations of arsanilic acid or sodium arsanilate. ^{112 , 114 , 539}

Postmortem findings in swine and poultry affected by organic arsenicals include no gross changes, except skin erythema in white pigs and muscle atrophy in chronic cases. ^{112 , 114 , 326} Harding et al. ³²⁶ reported abnormal distention of the urinary bladders in pigs poisoned by arsanilic acid.

Detectable histopathologic changes in swine are confined to the optic tracts, optic nerves, and peripheral nerves. Major lesions noted are necrosis of myelin-supporting cells, degeneration of myelin sheaths and axons, and gliosis of affected tracts (Figure 5-5, Figure 5-6 through Figure 5-7). Damage is first seen after about 6–10 days of feeding on excessive arsenical and is characterized by fragmentation of the myelin into granules and globules and, several days later, by breaking up of the axons. There is an obvious increase in the severity of the lesions with the progression of the toxic syndrome. No microscopic changes are seen in the brain, cord, kidneys, liver, or other organ systems. ^{326 , 458 , 459}

FIGURE 5-5

Sciatic nerve from control pig. An axon appears as a faint gray line between arrows 1. Note darkly stained neurokeratin network, arrow 2. Harris hematoxylin and eosin Y stain. Reprinted by courtesy of Marcel Dekker, Inc., from Ledet et al. ^{(p. 447)}

FIGURE 5-6

Sciatic nerve from pig fed arsanilic acid at 900 g/ton (992 g/tonne) for 16 days. Note contraction of myelin around intact axon, arrow 1; myelin fragment, arrow 2; and myelin ovoid, arrow 3. Harris hematoxylin and eosin Y stain. Reprinted by courtesy (more...)

FIGURE 5-7

Sciatic nerve from pig fed arsanilic acid at 900 g/ton (992 g/tonne) for 27 days. Animal had developed quadriplegia. An estimated 60% of the nerve fibers were damaged. Note axon with myelin contracting around it, arrow 1; fragment of myelin, arrow 2; (more...)

Excretion and Recommended Withdrawal Time

In general, phenylarsonic compounds are rapidly excreted by the urinary system in domestic animals and poultry. Once they are absorbed from the gastrointestinal tract, 50–75% of the material is excreted within 24 h. Excretion of the remaining 25% is much slower and may take 8–10 days. ^{125 , 458 , 555 , 556 and 557} Although nervous tissue tends to accumulate relatively small amounts of the phenylarsonic compounds, their excretion rate from this tissue appears to be relatively low, less than 50% excretion 11 days after withdrawal. ⁴⁵⁸

Ledet et al. ⁴⁵⁹ measured arsenic contents of various organs from swine after their consumption of arsanilic acid at 1,000 ppm in the diet (10 times the recommended concentration for continuous feeding for improving weight gain and feed efficiency) for 19 days. The results are presented in Table 5-4.

TABLE 5-4

Arsenic in Swine Organs after Consumption of Arsanilic Acid.

Evans and Bandemer ²³⁷ measured the arsenic content of eggs from hens fed diets containing arsanilic acid at 100 and 200 ppm for 10 weeks and found concentrations below the tolerance, established by the FDA, of 0.5 ppm.

Baron ⁵¹ reported on the accumulation and depletion of arsenic in tissues of chickens fed a ration containing 3-nitro-4-hydroxyphenylarsonic acid at 50 ppm (0.005%). Medication was started when the chickens were 4 weeks old, and the birds were killed at 1, 2, 3, 4, 5, 7, 9, 11, 14, 28, 56, and 70 days of medication and on day 1-14 after withdrawal of medication. Five birds of each sex from both the medicated and nonmedicated groups were killed on the days indicated. The arsenic concentrations found in kidneys, liver, muscle, and skin are presented in Table 5-5.

TABLE 5-5

Arsenic in Chickens Fed Ration Containing 3-Nitro-4-Hydroxyphenylarsonic Acid at 50 ppm.

FDA regulations require that all labels of feeds containing any of the phenylarsonic compounds include a warning that such feed must be withdrawn from swine and poultry 5 days before slaughter. Elemental arsenic tolerances of 2.0 ppm for uncooked swine liver and kidney tissues and 0.5 ppm for uncooked pork muscle and edible chicken and turkey tissue and eggs have been set.