DIATOMIC IODINE. reproduced except very qualitatively. The. reasons for this behavior apparently stem from

Transcription

1 THE RELATIVE GERMICIDAL ACTIVITY OF TRIIODIDE AND DIATOMIC IODINE BENJAMIN CARROLL Newark Colleges of Rutgers University, Newark, New Jersey A freshly prepared aqueous solution of iodine at a ph less than eight will contain mainly two forms of free iodine, namely diatomic iodine, I2, and triiodide, Ii,. The ratio of these two forms is determined by the concentration of iodide which is usually added to bactericidal prepaations for the purpose of increasing the solubility of the iodine. The solubility of diatomic iodine at 25 C is per cent. In the presence of an iodide salt like potasium iodide the solubility of free iodine may be increased several hundredfold, the increase being in the form of triiodide according to the equation: (1) Ta+ 1 - Ir. A quantitative investigation of the relative sporicidal activity of diatomic iodine and triiodide carried out by Wyss and Strandskov (1945) showed that the triiodide ion had negligible activity. Their test organism was Bacillus meties spores. Subsequent publications by Knox et al. (1948) and Marks and Strandskov (1950) have emphasized that the mode of bactericidal action on spores may be quite different from that on vegetative cells and that generalization of the relative sporicidal activity of the two forms of iodine to include other microorganisms is not warranted. Salle (1954) states that both forms of iodine seem to be equally effective. It is the purpose of this paper to present experimental data on the relative lethal effects of I2 and I- toward the vegetative cells of Micrococcus pyogenes var. aureus and Escherichia coli, and to consider the possible relative effects of these forms of iodine on infectious agents in general. A quantitative measure of the activity of I2 and I,- toward vegetative cells was sought by determining the time required to achieve a given level of mortality for various iodine solutions. Our experimental results were frequently erratic. Even the relationship between the time rate of kill and the concentration of diatomic iodine where the concentration of the latter was varied by the simple addition of water could not be Received for publication September 30, reproduced except very qualitatively. The reasons for this behavior apparently stem from the very low concentration of iodine required to kill vegetative cells. In this connection our experience was similar to that adequately described in the recent work of Chang and Morris (1953). They stated that "attempts to evaluate the fundamental bactericidal behavior of iodine in pure water were not successful. A high degree of variability was encountered, probably because the concentrations of iodine required to give measurable rates of kill were so small that they could be completely neutralized by mere traces of reactive impurities. In some tests, concentrations of iodine as low as 0.02 ppm exhibited rapid bactericidal action while in other tests at the same concentration there was no measurable killing". In the present instance, matters were complicated further by the fact that several tenths of a per cent of potassium iodide are required to convert most of the diatomic iodine to triiodide. At these concentrations a slight bactericidal activity was observed for the iodide salt alone. Whether this is due to the intrinsic effect of the iodide ion or possibly the minute quantities of iodine produced by the action of dissolved oxygen in water is not clear. Experimental conditions were sought so as to render the effect of potasium iodide negligible. In order to minimize the side effects due to trace quantities of impurities an attempt was made to increase greatly the dosage of iodine, yet maintaining an experimentally measurable rate of mortality of bacteria. Two possible courses were open. One was to work at low temperatures; the other was to carry out the bactericidal measurements in the presence of high concentration of bacteria or in the presence of organic matter, e.g., human blood serum. The latter course was chosen because of the correlation of such procedure to the therapeutic use of iodine.

2 414 BENJAMIN CARROLL [VOL. 69 MATERIALS AND METHODS All chemicals were of cp grade. Diatomic iodine solutions were formed by shing resublimed iodine crystals in water and decanting the clear solution. The concentration of free iodine was determined by thiosulfate titrations. A Beckman model DU spectrophotometer with temperature controls was used frequently to check the diatomic iodine and triiodide content of iodine solutions. The characteristic peaks at 350 m,u and 460 mju for the triiodide ion and diatomic iodine, respectively, and their extinction coefficients are such as to permit an accurate and rapid determination of the Ii /I2 ratio (Katzin, 1953). All iodine solutions were buffered at a ph somewhat below 6 in order to minimize hydrolysis during the course of a bactericidal run which might have occurred in the presence of blood serum. The test organism, M. pyogews var. aureus strain 209 and E. coli strain 8239, were harvested by centrifuging the saline washings of a 24 hour broth culture. The bacteria were subsequently resuspended in sterile saline solution by vigorous shaking to aasure breaking up of the clumps that might have been formed by the centrifugation. The bacterial suspensions were used immediately in the experiments with iodine. One experimental procedure consisted of adding iodine to the solution of blood serum followed immediately by the addition of 1 ml of bacterial suspension. The latter step was taken as zero time. At specified time intervals 1 ml quantities of the bacteria-iodine mixture were blown into an appropriate solution of sodium TABLE 1 Effect of concentration of diatomic iodine on the nunber of viable bacteria in presence of 6 per cent human blood serum Test organism: Micrococcus pyogenes var. aureus. Initial concentration of viable bacteria = 1.4 X 106 per ml. Time Number of Viable Bacteria per ml X 10- I-0.0 ppm It-- ppm I,-4 ppm I-8 ppm K.I-0.12% KI-0.12%o KI-0.0% KI-0.0% min thiosulfate to reduce the residual iodine. The reduced solution was then counted for viable bacteria by the usual method of serial dilution and plating. Heart infusion (Difco) blood agar was used to pour the plates; incubation was carried out for 48 hours at 37 C. The simpler but somewhat less precise Food and Drug Administration (F.D.A.) type test was used as a check on the relative activity of diatomic iodine and triiodide. With the modified F.D.A. test, 5 per cent human blood serum was used in the case of micrococci, but no organic matter was used for the run using E. coli. The order of mixing was buffer, serum, iodine followed immediately by bacterial Suspension. At specified time intervals a 4 mm loopful of mixture was transferred to tubes containing 10 ml of broth. Two per cent sodium thiosulfate for reducing the residual iodine had been dissolved previously in the broth. The tubes were incubated at 37 C for 48 hours. Some experiments using the modified F.D.A. procedure were duplibated by plating on agar instead of inoculating broth. Results were about the same. All bactericidal activities were determined at C. RESULTS Results on the rate of mortality of M. pyogenes var. aureus are given in table 1. Although the experimental error for the various counts of viable bacteria was about 415 per cent, the data reveal a definite impairment of the antibacterial activity of the diatomic iodine solution upon the addition of potassium iodide. It will be seen from this table that the latter salt alone has a negligible effect. The equilibrium concentration of I2 in the 8 ppm iodine containing 0.12 per cent KI (see table 1) is 1.3 ppm. This calculation is based on the equilibrium constant (Latimer, 1952) for equation (1), which is (2) K x 102. As the test in table 1 was prolonged beyond 10 minutes, the activity of the I, - KI mixture approached the value for I2 system at 4 ppm, an effect that was not entirely unexpected due to the rapid consumption of the diatomic iodine by the organic matter. The data for the same organism using a F.D.A. type procedure are given in table 2. If the end point is taken as that concentration of bactericide that produces a total kill in 10 minutes but

3 19551 GERMICIDAL ACTIVITY OF IODINE 415 TABLE 2 Effect of addition of 0.2 per cent KI on the bactericidal activity of diatomic iodine as indicated by modified F.D.A. procedure Test organism: Micrococcus pyogenes var. aureus 100 Free Iodine Time (minutes) Concentration (ppm) j 10.0 I2 alone 1 sin I2 in presence of 0.2 per cent KI _ not in 5 as is done in the F.D.A. test, then table 2 indicates that 35 ppm of I2 are equivalent to 196 ppm of I, in the presence of 0.2 per cent potassium iodide. Using the equilibrium constant as given in equation (2), I2- KI system yields a calculated value of 41 ppm of free iodine in the I2 state and 175 ppm of free iodine in the Ii state. Since bactericidal activity of this system was found to be the same as 35 ppm of I2, the activity of the I8 ion may be considered as 35-21/175 or 0.08 that of the IL molecule. In the case of the E. coli in the presence of organic matter, graphical representation of the mortality rate was feasible. Rsults are summarized im figure 1. The concentration of diatomic iodine in this figure is altered by dilution and by the addition of potassium iodide. The addition of 0.16 per cent potassium iodide to 10 ppm of diatomic iodine converts 8.7 ppm of the latter to I3, leaving an equilibrium concentration of 1.3 ppm of diatomic iodine as indicated in the figure. A smooth curve has been drawn through all points for a given time interval. The result in the figure indicate a small or negligible activity for the triiodide activity in the case of E. coli. The modified F.D.A. procedure using the same test organism in the absence of organic matter yielded the results given in table 3. The data afford two comparative end points, the one a 10 ppm, DIATOMIC IODM CONCUTRATION Figure 1. Effect of concentration of diatomic iodine on the rate of mortality of Escherichia coli in the presence of 5 per cent human blood serum. Outline hexagon - Solutions of diatomic iodine in the absence of potassium iodide. Solid hexagon = Solution containing initially 10 ppm of diatomic iodine to which 0.16 per cent potassium iodide had been added. Equilibrium concentration is 1.3 ppm of diatomic iodine. TABLE 3 Effect of addition of 0.9 per cent KI on the bactericidal activity of diatomic iodine in the absence of organic matter as indicated by modified F.D.A. procedure Test organism: Escherichia coli Free Iodine Time (minutes) Concentration (ppm) j 10.0 I2 alone I2 in presence of 0.2 per cent KI _ giving a total kill in three minutes but not in one, the other giving a total kill to ten but not in five minutes. The former end point yields an estimated relative activity of Ir/12 of The latter end point yields an estimated ratio of

4 416 BENJAMIN CARROLL [VOL , the average of the two end points being These calculations are based on the fact that 9 per cent of the free iodine in the KI system is in the diatomic state. DISCUSSION Although the complete destruction of bacteria is usually considered an undesirable end point for quantitative work, it was found that the F.D.A. type test yielded results in fair agreement with those obtained from the experiments on counting of viable bacteria. The close numerical agreement of the activity ratio of Ij /I2 for both microorganisms as indicated by the F.D.A. type test may be considered fortuitous. However, all measurements show clearly that the triiodide ion has a small or negligible antibacterial activity compared to diatomic iodine. This result appears to be independent of the level of mortality and of the presence of organic matter. These findings raise two questions. One is, why has the serious impairment of the bactericidal activity of iodine solution by iodide salts gone rather unnoticed in spite of the widespread use of these salts in iodine preparations? The other question concerns the mode of action of iodine. In regard to the first question, it can be seen that in the frequently used procedure in carrying out bactericidal studies on vegetative cells, very high dilutions of iodine are required. At these concentrations according to equation (2), the triiodide ion will dissociate into diatomic iodine and iodide ion unless the relative concentration of the iodide ion is kept very high; even then large experimental inaccuracies may be encountered for reasons discussed earlier. The bactericidal activity of most pharmaceutical iodine preparations may then be expected to depend primarily upon the free iodine content unless appreciable organic matter is used in the bacterial test. Differences should be expected as soon as tests are designed to enable the use of several ppm or more of iodine. It is apparently for this reason that the effect of potassium iodide additions to iodine was observed first in the case of resistant microorganisms like spores. In this connection it is interesting to consider the work of Anson and Stanley (1941). They reported that in the presence of one molar potassium iodide, iodine failed to oxidize even the sulfhydryl groups in the tobacco mosaic virus. However, in low concentrations of potassium iodide the tyrosine as well as the sulfhydryl groups was modified, thus inactivating the virus. Their work may be viewed as the one of the few examples where the presence of potassium iodide completely prevented iodine from inactivating or reacting with an infectious agent. As for the mode of action it is generally thought that the ionic form of a germicide is less active than its neutral counterpart. Examples in line with this view are HOCI versus OC1- (Marks, Wyss, and Strandskov, 1945) and benzoic acid versus benzoate ion (Goshorn et al., 1938). This behavior is ascribed to the greater capacity for the undissociated molecule to penetrate bacteria. However, the permeability theory should not rule out other posible reasons in the case of iodine. Since the inactivity of the triiodide ion is displayed toward such diverse agents as the tobacco mosaic virus, Bacilus metiens spores, M. pyogene8 var. aurew and E. coli, consideration of the oxidation potentials of the I, and Ijmolecules and the mechanism of oxidation and iodination is warranted. The standard oxidation potentials (Latimer, 1952) for the I2-I- and I, -I- couples at 25 C are given as follows: (3) 3I- - Ii- + 2e; E' = volt; (4) 2I- = I2 + 2e; E = volt. Cursory examination of equations (3) and (4) would seem to imply that I, and I, have about the same oxidation potentials. This may be in part responsible for the frequent lack of differentiation between the two forms of iodine in the literature. It should be noted, however, that the molar standard state is inferred for equation (3), whereas a saturated solution of I2 which is 1.3 X 10-' molar is the standard state in equation (4). Thus the oxidation potentials of Ir and I2 are the same provided that there is roughly 103 times as much free iodine in the form of Ia as there is I2. As a matter of fact a one molar solution of Ii in contact with a molar solution of iodide, neglecting activity coefficients, will be in equilibrium with 1.4 x 10-3 M I2 according to equation (2). The close proximity between this numerical value and the molar solubility of diatomic iodine is responsible for the practically identical standard electrode potentials in (3) and (4). Obviously, addition of an iodide salt to a solution of diatomic iodine will lower its oxidation potential. Thermodynamic considerations alone are not sufficient for explaining the efficiency of an oxidant as a germicidal agent. The mechanism

5 ]GERMICIDAL ACTIVITY OF IODINE and kinetics of the action of a halogen on the essential enzymes or proteins of the infectious agent should also be considered. In reacting with iodine, proteins may be modified by oxidation or substitution. One is struck iminediately by the extreme paucity of mechanisms involving the triiodide ion for organic reactions in general and for these two types of reactions in particular (Ingold, 1953). It is known for example that iodination of the phenolic group in the amino acid, tyrosine, will invariably cause a loss of biological activity for the protein (Olcott and Fraenkel-Conrat, 1947). It is a simple matter to demonstrate the complete and instantaneous inhibition of the substitution reaction of iodine with a para substituted phenol, e.g., tyrosine, by the addition of an iodide salt to the reaction mixture (Li, 1942). Oxidation reactions seem to be affected similarly. Thus a glucose solution will be oxidized at an appreciable rate at ph 6 by diatomic iodine. The toxic effect of the iodine solutions toward animal tissue is common experience. Iodine solutions for therapeutic applications, however, are usually prepared with iodide salt for the purpose of increasing the solubility of the iodine. The maxmum concentration of the diatomic form of iodine is fixed nevertheless by its solubility regardless of the quantity of iodide salt that may be present. Secondary salt effects are small in altering this solubility (Lewis and Randall, 1923). The literature appears to give no clue as to the relative toxicity of the various inorganic forms of iodine. When such evidence is obtained, a re-evaluation of the potentialities of iodine as a chemotherapeutic agent will be possible. ACKNOWLEDGMENTS The author wishes to express his appreciation to Dr. Abraham Berliner whose clinical findings suggested this problem, to Dr. Irving Chapman of the Bird S. Coler Memorial Hospital, and to Dr. Henry C. Marks of the Wallace and Tiernan Co. for discussions on some aspects of this work. The technical assistance of Miss Ruth E. Egan is gratefully acknowledged. This work was supported by grants from the Rutgers Research Council and Heliogen Products, Inc. SUMMARY Using both the rate of mortality and Food and Drug Administration (F.D.A.) type tests, it has been shown that the bactericidal activity of the triiodide ion toward Micrococcus pyogenes var. aureus and Escherichia coli is small or negligible compared to diatomic iodine. Oxidation potentials and kinetic factors have been considered to explain the relative inactivity of the triiodide ion toward infectious agents in general. REFERENCES ANSON, M. L., AND STANLEY, W. M Some effects of iodine and other reagents of the structure and activity of tobacco mosaic virus. J. Gen. Physiol., 24, CHANG, S. L., AND MORMUS, J. C Elemental iodine as a disinfectant for drinking water. Ind. Eng. Chem., 45, GOSHORN, R. H., DEGERaNG, E. F., AND TE- TAULT, P. A Antiseptic and bactericidal action of benzoic acid and inorganic salts. Ind. Eng. Chem., 30, INGOLD, C. K Structure and mechanism in organic chemistry. Cornell University Press, Ithaca, N. Y. KATZIN, L. I Note on the absorption spectrum of iodine in oxygenated solvents and the dissociation of iodine water. J. Chem. Phys., 21, KNOX, W. E., STUMPF, P. K., GREEN, D. E., AND AUERBACH, V. H The inhibition of sulfhydryl enzymes as the basis of the bactericidal action of chlorine. J. Bacteriol., 55, LATIMER, W. M Oxidation potentials. 2nd edition. Prentice Hall, Inc., New York. LEWIs, G. N., AND RANDALL, M Thermodynamics. McGraw-Hill Book Co., Inc., New York. Li, C. H Kinetics and mechanism of 2,6-di-iodotyrosine formation.. J. Am. Chem. Soc., 64, MARKS, H. C., AND STRANDSKOV, F. B Halogens and their mode of action. Ann. N. Y. Acad. Sci., 53, MARKS, H. C., WYSS, O., AND STRANDSKOV, F. B Studies on the mode of action of compounds containing available chlorine. J. Bacteriol., 49, OLcorr, H. S., AND FRAENKEL-CONRAT, H Specific group reagents for proteins. Chem. Revs., 41, SALLE, A. J Fundamental principles of bacteriology. Fourth edition. McGraw-Hill Book Co., Inc., New York. WYss, O., AND S1ANDSKOV, F. B The germicidal action of iodine. 6, Arch. Biochem.,