than arsenic acid; for if the latter were readily reduced to arsenious acid, it would be equally toxic." Again, Meyer and Gottlieb state: "Weder die Arsenige noch die Arsensäure gehen ohne weiters mit Bestandteilen des Protoplasmas erkennbare oder sonstwie nachweisbare Verbindungen ein: ihre Lösungen sind daher zunächst ganz ohne sichtliche morphologische oder funktionelle Wirkung sowohl auf nervöse als auch auf andere organische Gebilde. Nach einiger Zeit aber erlischt das Leben der stark vergifteten Zelle und sie verfällt der postmortalen Zersetzung. Ob diese Wirkung auf katalytischer Hemmung lebenswichtiger Prozesse beruht oder auf chemischer Bindung, irgend eines für das Zelleben notwendigen Minimalstoffes des Protoplasmas durch das Arsen, ist nicht bekannt. Fermente werden durch Arsenic nicht merklich beeinflusst, was nicht gerade auf eine 'katalytische' Wirtkung spricht. Für die Möglichkeit einer specifischen chemischen Bindung von As spricht dagegen die Angabe von Bertrand (1903), dass As sich als integrierender Bestandteil in allen lebenden Zellen findet." These quotations from the latest editions of standard books on pharmacology represent the essential information with regard to the protoplasmic action of arsenic, and a thorough review of the original literature scarcely adds anything of moment. Ehrlich 1909), has used his side chain theory to explain the chemotherapeutic action of arsenic. He assumes that trivalent arsenic is firmly fixed to the cell by a definite chemical group (side chain) of protoplasm, this chemical union causing the death of the cell. The only proof offered for this assumption is his observation on arsenic resistant trypanosome strains. Infected mice were repeatedly treated with subqurative doses of atoxyl, until maximum tolerated. doses of this drug were: no longer capable of clearing the blood of parasites; when it was found that arsacetin was still effective in sterilizing the animals; but if treatment was continued with subcurative doses of arsacetin, the strain gradually became resistant also to this drug; wet arsenophenylglycine was still. capable of curing animals injected with the strain resistant, to arsacetin:. These experiments were interpreted as proof of the theory that the arsenic receptor of the trypanosome had gradually lost, some of its power of reacting with arsenic. In spite of the highly interesting nature of these observations, it is well to realize that they do not disclose either the chemical nature of the hypothetic arseno-ceptor, nor do they furnish any valid proof for the formation of a chemical combination of arsenic with a protoplasmic: constituent.. Plan of Present Investigation. We intend to report in this paper some observations which throw considerable light on the more intimate mechanism of the action of arsenic upon protoplasm and which are not without importance with reference to their bearing upon the more fundamental problem of biological oxidations and reductions. The reasoning which underlies our plan of procedure is briefly as follows: Numerous observations on the relation between chemical constitution and physiological action of a great variety of arsenicals have lead Voegtlin and Smith (1920, 1921) to conclude that arsenic can exert a direct toxic effect only if present in the form R· As—0, where R represents an aliphatic or aromatic radical. The pentavalent arsenicals and the arsenobenzene derivatives must be converted to this form by means of reduction or partial oxidation, respectively, as expressed by the following formula: The conversion of the compounds of Groups I and III into compounds of Group II is accomplished by the tissues of the higher animals but not to any appreciable extent by some of the lower forms of life, such as trypanosomes and Treponema pallidum. These organisms, however, are highly susceptible to the toxic action of R · As—O, a fact which can readily be demonstrated by exposing trypanosomes to high dilutions of these arsenicals in the test tube, the organisms being killed within a few minutes, whereas controls exposed to atoxyl or arsphenamine survive exposure to much higher concentrations of these drugs. The same holds true for the parasiticidal effect produced in the infected animal. Minimum effective doses of Groups I and III compounds show a long latent period before destruction of the organisms in the circulating blood begins, whereas injection of R As O compounds is followed immediately by a rapid disappearance of the parasites, so that within about 20 minutes the blood is cleared. . The fact that R As O compounds are directly toxic explains the great constancy of the results obtained with these compounds in test tube experiments and in the treatment of infected animals, in contrast to the variations observed in work dealing with. the pentavalent arsenicals and the arsenobenzene derivatives, which require for the production of the active R As O modifications, the intermediate action of the tissues involving a number of variable factors. For these reasons it was thought best to select for our experiments a representative of the RAS compounds, namely, the partial oxidation product of arsphenamine, briefly called "arsenoxide," possessing the constitution The questions which have to be answered: are:. How does "arsenoxide produce its toxic effect? How does it kill the parasites and what is the mechanism whereby. it produces the toxic symptoms in and death of the higher animals? If, according to Ehrlich, the toxic action is primarily due to a chemical reaction between "arsenoxide" and a well-defined chemical group of protoplasm, then it should be possible to overcome the toxic effect of arsenic, at least temporarily,. by supplying the parasitic cell or the tissues of animals with an extra amount of the particular protoplasmic constituent containing this. reactive group (side chain in Ehrlich's terminology). Because of the incomplete knowledge regarding the chemical composition of protoplasm generally, it might seem a hopeless task. to attempt an attack of the problem in this manner. The situation, however, was made considerably less complex by a proper appreciation of some well-known chemical properties of arsenic. In qualitative and quantitative analyses, use is made of the great reactivity. of arsenite with HS, as expressed by the equation— 2 As(OH), +3 H2S = 6 H2O+AS2S3 Aromatic arsenious oxides (R AsO), including arsenoxide, also react with the greatest ease with HS according to the equation— A similar reaction takes place when a sulphydryl compound R-SH) is used instead of H2S: = S-CH, COOH As OH + HS-CH2 COOH As S.CH COOH HS.CH COOH S.CH COOH + 3H,0 It was therefore quite logical to consider the possibility that arsenic might react with a protoplasmic constituent containingsulphur, especially as Heffter, several years ago, had demonstrated the fact that tissues contain a substance which, from its behavior against sodium nitroprusside and sulphur, may be considered as a "proteinlike" substance containing a sulphydryl (SH) group. This compound was held responsible for the reducing properties of fresh tissues. It remained for Hopkins (1921) to isolate from yeast and mammalian tissues, glutathione, a dipeptide, containing glutaminic acid and cystein. This remarkable substance is autoxidizable, its sulphydryl group undergoing the following reversible change: Hopkins, and Hopkins and Dixon, by means of experiments on surviving frog's and mammalian muscle, demonstrated that glutathione was concerned in biological oxidations and reductions. According to Hopkins, fresh ox muscle and brewer's yeast contain about 100 to 150 milligrams of glutathione per kilo, and liver is considered to be somewhat richer in this substance. The fact that the absolute mass of glutathione in tissues, and the amount of arsenoxide necessary to produce a toxic effect are both small, a priori speaks in favor of our assumption of a chemical interaction between the two substances. It was therefore decided to subject to experimental proof the hypothesis that the toxic action of arsenic is due to an effect upon glutathione or some closely related sulphydryl compounds. Experimental. For the sake of clearness, the investigation can be divided into two parts: (A) Experiments which deal with the demonstration of the antagonistic action of arsenoxide and certain sulphydryl compounds on trypanosomes, and (B) observations on the antagonism of arsenoxide and sulphydryl compounds in a representative of the higher animals, namely, the albino rat. The arsenoxide (3-amino-4-hydroxy-1-phenylarsenious oxide) was prepared as the hydrochloride by Dr. J. M. Johnson, of the Hygenic Laboratory. It represents a white, amorphous powder, easily soluble in cold water, and stable if kept in an amber-colored glass bottle in a vacuum desiccator. Analysis of this lot of arsenoxide showed that it was unusually pure. Glutathione was prepared from yeast or ox liver, following the directions described by Hopkins (1921). The final product was either the oxidized form or the reduced dipeptide. The latter was obtained in the form of a gum by decomposing the copper hydroxide or final mercury precipitate with HS and allowing the filtrate from Cu,S or Hg,S, respectively, to evaporate in a vacuum desiccator over phosphorus pentoxide. The reduced form, in every instance, gave an intense nitroprusside test, whereas the oxidized form of the substance (R-SS-R) did not yield this test. In order to confirm the results obtained with glutathione prepared in this laboratory, Prof. Gowland Hopkins very kindly supplied us with a small amount of the pure oxidized form of glutathione. A part of this material was converted into the reduced dipeptide by dissolving it in water and adding the mercuric sulphate reagent. The insoluble mercury derivative thus formed was filtered, washed thoroughly with water, suspended in a small amount of water, and treated with HS. The filtrate from the mercurous sulphide was evaporated in a vacuum desiccator over P2O, leaving a colorless, gummy mass. The product gave a strong nitroprusside test, indicating the presence of the SH group. Besides testing the influence of glutathione on the arsenoxide action, it seemed desirable to study also the effect of other sulphydryl compounds and their corresponding disulphide (R.SS.R) modifications. The following substances were prepared: Cystein hydrochloride and cystine, thioglycollic acid, and dithiodiglycollic acid, a-thiolactic acid, glycyl-cystein, thiosalicylic acid, and dithiodisalicylic acid. The latter two compounds were prepared by Dr. M. X. Sullivan, of the division of chemistry, Hygienic Labor atory. Glycyl-cystein was prepared by the method of Emil Fischer and C. Suzuki (1904) for diglyeyleystine, except that in the last stage the material was precipitated with mercuric sulphate reagent and the mercury compound decomposed with H,S; the filtrate of Hg2S, on evaporation in the desiccator, yielded the reduced dipeptide as a thick, yellowish syrup. The microkjeldahl gave 4.495 per cent N: theory for glycyl-cystein, H2SO, 2H,O=4.487 per cent N. Cystein hydrochloride was prepared from pure cystine by tin reduction, removal of excess metal with HS, and evaporation in vacuum. In control experiments the following pure amino-acids were used: d-alanine, 1-aspartic acid, d-glutaminic acid, 1-histidine, 1-leucine, 1-tryptophane, 1-tyrosine, and d-valine. We are indebted to Doctor Jones, of the Bureau of Chemistry, Department of Agriculture, for the samples of valine and aspartic acid. A. TRYPANOSOME EXPERIMENTS. As in previous work, we selected for this investigation our standard strain of Trypanosoma equiperdum, which is propagated in albino rats. In order to obtain simple conditions, the antagonistic action of the sulphydryl compounds was first tested out in vitro. For this purpose, a rat showing about 100,000 to 150,000 trypanosomes per c. mm. was bled into 10 c. c. 2 per cent sodium citrate. A series |