- Email: [email protected]
R-FACTOR-MEDIATED RESISTANCE TO NEW AMINOGLYCOSIDE ANTIBIOTICS
252 be ruled out by the fact that cells of each species not treated with tannic acid are not agglutinated by the same to
TABLE II-FRACTIONATION OF
IgA
BY AGAR-GEL ELECTROPHORESIS
IgA preparations. SOME PROPERTIES OF H.T.-AGGLUTINATION
H.T.-agglutination is a sensitive reaction: a concentration of IgA of 12ng. per ml. causes definite agglutination of sheep red blood-cells, and half this amount consistently gives agglutination of human red blood-cells. In fact, the sensitivity of the test is even higher than this because H.T.-agglutination is produced by only a small fraction of the IgA protein. Similar titres of H.T.-agglutination have been obtained with IgA isolated from either normal or clinically hypersensitive people. The ability to cause H.T.-agglutination is a property shared by nearly all the IgA preparations tested so far (over 60), provided the protein is handled gently during isolation from serum. Several other proteins from human serum, including albumin, haptoglobin, and IgG, are inactive, even at a concentration of 16 mg. per ml.-i.e., more than a thousand times as high as the concentration of IgA sufficient to give H.T.-agglutination. The most interesting point about H.T.-agglutination is the close resemblance between some properties of the factor causing it and the properties of reaginic antibodies. Effects of Heat and 2-mercaptoethanol The H.T.-agglutination factor is inactivated at 560C in a time which depends on the concentration of IgA; a solution containing 200 tLg. IgA per ml. is completely inactivated in 30 minutes. The same solution would be completely inactivated in 18-24 hours at 37"C in 2-3 days at room temperature, and in about 10 days at 4°C. The activity lasts for years when the preparations are kept at -20°C or when they are freeze-dried. The reducing agent, 2-mercaptoethanol, at a 0-1M concentration, has no appreciable effect on the activity of IgA at 4OC; it has a very slow effect at room temperature, more than 19 hours having to elapse before there is a partial but significant TABLE I-FRACTIONATION OF
IgA
BY
PRECIPITATION WITH
1-18 M
SODIUM SULPHATE
reduction in activity;
the other hand, the solution 200 ml. is containing g. IgA per completely inactivated by 2-mercaptoethanol in 60 minutes and almost completely in 30 minutes. Thus, H.T.-agglutinating activity, like the skin-sensitising activity of reagins, is heat-labile at 56C and shows a susceptibility to reduction which can be considered as intermediate between that of IgM and IgG. IgA already fixed to tanned cells withstands heating at 56°C for periods at least eight times as long or reduction with 2-mercaptoethanol at 37’C for periods at least four times as long as those required for the inactivation of IgA in solution. In other words, the agglutination of tanned red blood-cells brought about by IgA is not destroyed by heating or reduction for such prolonged periods. This is also similar to what happens to reagins, which remain active for weeks when fixed to the skin (at the site of a passive-transfer test) while they cannot withstand body temperature for so long a time when they are in a test tube. at
37 °C,
on
Effect of pH IgA preparations retain their H.T.-agglutinating ability unchanged at pH 6-4—86, but are completely inactive at pH 4-2. (The pH
range of
activity of reagins is
between 6 and
11.)
Fractionation When whole sera or IgA preparations endowed with reaginic activity are fractionated by precipitation with 1- 18 A1 sodium sulphate all the skin-sensitising activity is recovered in the soluble fraction, which contains only a small amount of the
IgA originally present: whereas the precipitate, which contains IgA, lacks any reaginic activity.22 When IgA preparations, either from normal or hypersensitive sera, are similarly fractionated, H.T.-agglutinating activity is only found in the fraction soluble in 1-18 M sodium sulphate, whereas the precipitate is inactive. If IgA comes from a clinically hypersensitive patient, the recoveries of skin-sensitising and H.T.agglutinating activities in the soluble fraction closely parallel each other (table I). Further, when IgA preparations are fractionated by agar-gel electrophoresis both activities are found in the same fractions (table II).
the bulk of
CONCLUSIONS
H.T.-agglutination and skin-sensitisation are related in some way. They may be properties of the same substance, but if so I cannot at present identify that substance. All the active preparations contained IgA, although none of them were in an absolutely pure state. Moreover, IgA is present in even greater amounts in fractions devoid of any activity. If H.T.-agglutinating activity and reaginic activity are properties of IgA, it is clear that they are not properties of IgA as a whole, but rather of a subfraction which is different in some way from the bulk of the protein. We do not know enough about IgA to say whether this subfraction should be given a different name. The factor responsible for H.T.-agglutination, although different from reagin, may play some role in the process of tissue sensitisation by reaginic antibodies; and even if it does not the reaction would seem to be an easy-toperform index of the activity of the fraction being handled. Instituto de Investigaciones Médicas, Fundación Jiménez Diaz, Madrid 3
FERNANDO ORTIZ M.D.
Madrid
R-FACTOR-MEDIATED RESISTANCE TO NEW
AMINOGLYCOSIDE ANTIBIOTICS To determine whether the existence and
Summary incidence of R-factor genes
are
always
related to the prevalence of antibiotics in the environment, recent clinical isolates of Escherichia coli and salmonella tested for resistance to 5 relatively new aminoglycoside antibiotics. 19 strains were resistant to bluensomycin, and 11 to spectinomycin, gentamycin, and viomycin: in each case, the resistance genes were associated with R factors. Bluensomycin resistance was mediated by the streptomycin R-determinant; but the resistances to spectinomycin, gentamycin, and viomycin were mediated These by previously unidentified R-determinants. results ate discussed in relation to the origin of R-factor were
genes. INTRODUCTION
R FACTORS are episomes of enteric bacteria that mediate resistance to several commonly employed antibacterial drugs: those defined originally in Japan mediated resist22. Ortiz, F.
Nature, Lond. (in the press).
253 ance to
sulphonamides, chloramphenicol, streptomycin,
and tetracycline 1; others described by workers in Europe also mediate resistance to neomycin-kanamycin,2ampicillin-cephalothin,3 and furazolidine.4 These newer R factors were initially isolated only after the commercial introduction of the drugs; their subsequent prevalence has been attributed to the increasing use of the drug in medicine and agriculture. Since the type of drug resistance engendered may throw light on the origin of R factor genes and may have practical importance as well, cultures of Escherichia coli and salmonella isolated from patients were examined for resistance to five relatively new aminoglycoside antibiotics. A large proportion of strains carrying previously identified R factors were found to be resistant to four of the drugs; in each case, the resistance genes were associated with the R factor.
each test drug was transferred by all potential donors; with the exception of 3 strains that each had two R factors, recipient strains acquired the entire pattern of resistance of the donor strains. Recipient strains transferred resistance to the test drug as part of the R factor in subsequent matings with other enteric bacteria. Moreover, bacteria that lost R factors lost resistance to the test drugs. Finally, resistance to the drugs was cotransduced by phage-PIwith other R-factor genes but was not cotransduced with chromosomal genes. These results clearly indicate that previously defined R factors contain genes (R-determinants) that mediate resistance to bluensomycin, gentamycin, spectinomycin, and viomycin. R-determinants for the test drugs were associated with fi- and fi+ R factors and various other R-determinants TABLE I-GENOTYPES OF R FACTORS CONTAINING R-DETERMINANTS FOR
METHODS
THE TEST DRUGS
The cultures studied were unselected strains that had remained viable since earlier investigations’; the E. coli had been isolated from the urine of children with genitourinarytract infection, the salmonella from stools of children with gastrointestinal disease. The pattern of resistance to each of eight commonly used antibacterial drugs had been determined by disc techniques. Methods of culture, conjugation, and stock bacteria used as conjugation recipients have been described previously.56 The cultures were taken to be resistant to the drugs if they formed colonies on eosin/methylene-blue/ lactose agar containing three or ten times the drug concentration that prevented colony formation of stock bacteria. Minimal inhibitory concentrations of drug were determined by twofold dilutions in tryptic digest broth, using the concentrations of inoculum indicated. Transduction with phage-Pl and penicillin selection of drug-sensitive segregants of resistant bacteria were carried out according to the methods described by Watanabe.’7 RESULTS
tested for resistance
bluensomycin, gentamycin, hygromycin, spectinomycin, and viomycin. Each drug was specifically selected for the study. The mechanism of action of each is under investigation in this department. Hygromycin has been added to agricultural feeds in the United States but has not been used clinically. Bluensomycin has been used only in laboratory investigations, and mutants of E. coli selected for single-step highlevel resistance show cross-resistance with streptomycin. Gentamycin and spectinomycin have undergone restricted clinical trials, but had not yet been officially introduced when these strains were isolated; neither drug had been used in this area. Viomycin was officially introduced for clinical use in 1953 but is not used as extensively as other aminoglycoside drugs. Of 20 cultures carrying R factors, 19 were resistant to bluensomycin, 11 to spectinomycin, and 11 to viomycin and gentamycin; both E. coli and salmonella were resistant to the drugs. 19 of these 20 strains were resistant to streptomycin, but none was resistant to neomycin and kanamycin. The cultures that were sensitive to other drugs were also sensitive to each of the five test drugs. Strains resistant to the test drugs were mated with sensitive E. coli, and recipients were selected for resistance to tetracycline, ampicillin, or a test drug. Resistance to 37 cultures
1. 2. 3. 4. 5. 6. 7.
were
to
Watanabe, T. Bact. Rev. 1963, 27, 87. Lebek, G. Z. Hyg. InfektKrankh. 1963, 149, 255. Anderson, E. S., Datta, N. Lancet, 1965, i. 407. Smith, H. W., Halls, S. Br. med. J. 1966, i, 266. Smith, D. H., Armour, S. E. Lancet, 1966, ii, 15. Smith, D. H. New Engl. J. Med. 1966, 275, 626. Watanabe, T. Meth. med. Res. 1964, 10, 202.
The R-determinants are listed and spaced for does not reflect genetic map positions.
convenience;
the order
presented
(table I). Resistance to streptomycin and bluensomycin was always associated and was not separated by transduction or the selection of clones sensitive to one of the drugs, suggesting that one gene mediates resistance to both drugs. Spectinomycin resistance was mediated by a separate R-determinant, however, since some strains resistant to streptomycin and bluensomycin were sensitive to spectinomycin; segregant clones, resistant only to spectinomycin, were selected from strains originally resistant to all three drugs. Viomycin and gentamycin resistance occurred independently of that to other aminoglycoside drugs and was probably mediated by a single R-determinant: resistance to both drugs occurred in the same strains and was not separated by transduction or segregation. These results were not expected, since the drugs differ TABLE
11—FACTORS
INFLUENCING
THE
LEVEL
OF
RESISTANCE MEDIATED BY R FACTORS
SPECTINOMYCIN
254
substantially in chemical composition and in their in-vitro effects on protein synthesis by extracts of E. coli; although the chemical structure of the drugs is not known, these results strongly suggest that both share a critical structural component. The levels of resistance mediated by the R-determinants of the test drug depended on the episome, species of the host bacterium, inoculum concentration, and test medium (table II). The mechanism by which these R-determinants mediate resistance to the drugs has not yet been completely defined, but it did not depend on a change in the ribosomes-the site of action of these drugs. The test drugs inhibited in-vitro protein synthesis by purified ribosomes of resistant and sensitive E. coli to the same extent. The direct relation between minimal inhibitory concentrations of drug and inoculum concentrations of bacteria suggests that these R-determinants may mediate the synthesis of enzymes capable of inactivating the test drugs; intracellular inactivation of the drugs has not yet been excluded, but extracellular inactivation was not observed. DISCUSSION
The results indicate that the origin and incidence of R-determinants may not always be related to the prevalence of the drug in the environment. And, although the R-determinants for spectinomycin and gentamycinviomycin may have been selected during their limited clinical use, it seems more probable that they have been selected either by natural exposure of enteric bacteria to the streptomyces producing the antibiotics or by selective pressures bearing only a fortuitous relation to the antibiotics. A fortuitous relation is supported by the observation that some R factors alter the susceptibility of bacteria to N-methyl-N’-nitro-N-nitrosoguanidine, an organic chemical used only in laboratory investigations for its strong
lack xanthine oxidase. The enzyme first appears in rat liver some six days after birth. As a result ferritin-bound iron is deposited in ever-increasing quantities in the fcetal liver in the absence of the normal iron-releasing mechanism.’’ The appearance of hepatic xanthine oxidase in the healthy animal soon after birth releases liver-stored iron needed by the animal at a time when it is deprived of its maternal source of iron; when it is subsisting on a diet of mother’s milk that is deficient in iron, and when its needs for iron for haemoglobin biosynthesis are great. Powell and Emerson5 noted a relation between hepatic xanthine-oxidase activity and the extent of iron storage in the rat liver: rats fed the xanthine-oxidase inhibitor allopurinol (4-hydroxypyraxolopyrimidine) with the standard laboratory meal retained a significant amount of hepatic iron. Addition of iron salts to the diet resulted in a hepatic hsmosiderosis that was histologically more pronounced, at all stages, in animals receiving allopurinol. An elevated serum-iron in one of five patients with gout who were receiving allopurinol was also observed.5 Rundles et al.,6 on the other hand, found no serumiron changes in similar patients receiving this drug; but the absence of hypersiderxmia does not exclude the possibility of excessive hepatic iron, since 20% of the patients with haemochromatosis reported by Dreyfus and Schapirahad normal serum-iron values. METHOD AND RESULTS
We have considered the possibility that a failure of xanthine oxidase to make its normal appearance in the liver might, over a period of years, lead to excessive deposits of iron such as is observed in the disease haemochromatosis. We have analysed 18 human liver samples for enzyme xanthine oxidase by a microfluorometric method.The results expressed per unit weight of total sample protein, are listed in the table for three groups of patients :2 control liver samples XANTHINE-OXIDASE-ACTIVITY IN HUMAN LIVER
mutagenic properties. Further studies on the origin of R-factor genes should have importance for both medicine and biology. This work was supported in part by a U.S. Public Health Services grant no. G.M.-14119-01. Requests for reprints should be addressed to D. H. S., Department of Bacteriology and Immunology, Harvard Medical School, 25, Shattuck Street, Boston, Massachusetts 02115, U.S.A.
Departments of Pediatrics and Bacteriology and Immunology, Harvard Medical School, Boston, Massachusetts
DAVID H. SMITH M.D.
Rochester
HÆMOCHROMATOSIS AND HEPATIC XANTHINE OXIDASE IN the normal mechanism for release of hepatic ferritin iron in the rat ferritin-bound ferric iron is reduced to the ferrous state by the reduced form of the enzyme xanthine oxidase during its oxidation of hypoxanthine and xanthine to uric acid.l The iron-ferritin bond is made labile and the iron is transferred to the plasma where it is bound to transferrin. In the healthy adult rat, liver xanthine oxidase is present in excess; the factor limiting the rate of release of ferritin iron is the cellular concentration of hypoxanthine and xanthine. Foetal and newborn
rat
liver2 and human newborn liver3
Green, S., Mazur, A. J. Biol. Chem. 1957, 227, 653; Mazur, A., Green, S., Saha, A., Carleton A. J. clin. Invest. 1958, 37, 1809. 2. Westerifield, A. W., Richert, D. A. J. biol. Chem. 1950, 184, 163. 3. Wells, H. G., Corper, H. J. ibid. 1909, 6, 469. 1.
obtained during surgery or necropsy from patients without liver disease, biopsy samples from patients with cirrhosis, and a third group from patients with idiopathic hamiochromatosis. The range of values in the control samples, from 1-81 to 549 includes one (1-81) from a necropsy specimen. The xanthineoxidase activities in liver samples from patients with cirrhosis or haemochromatosis are all below those in the control group, and in a number of instances there was no evidence for any xanthine-oxidase activity. DISCUSSION
Although there was no overlap of xanthine-oxidaseactivity values between the controls and those with liver disease, interpretation of the data must be tempered by the relatively small number of samples in each group and the fact that expression of enzyme activity per unit weight of protein does not take into account the presence of Mazur, A., Carleton, A. Blood, 1965, 26, 317. Powell, L. W., Emerson, B. T. Lancet, 1966, i, p. 239. Rundles, R. W., Metz, E. N., Silberman, H. R. Archs intern. Med 1966, 64, 229. 7. Dreyfus, J.-C., Schapira, G. in Iron Metabolism (edited by F. Gross). p. 296. Berlin, 1964. 8. Burch, H. B., Lowry, O. H., Padilla, A. M., Combs, A. M. J. brit Chem. 1956, 223, 29. 4. 5. 6.
TABLE II-FRACTIONATION OF
IgA
BY AGAR-GEL ELECTROPHORESIS
IgA preparations. SOME PROPERTIES OF H.T.-AGGLUTINATION
H.T.-agglutination is a sensitive reaction: a concentration of IgA of 12ng. per ml. causes definite agglutination of sheep red blood-cells, and half this amount consistently gives agglutination of human red blood-cells. In fact, the sensitivity of the test is even higher than this because H.T.-agglutination is produced by only a small fraction of the IgA protein. Similar titres of H.T.-agglutination have been obtained with IgA isolated from either normal or clinically hypersensitive people. The ability to cause H.T.-agglutination is a property shared by nearly all the IgA preparations tested so far (over 60), provided the protein is handled gently during isolation from serum. Several other proteins from human serum, including albumin, haptoglobin, and IgG, are inactive, even at a concentration of 16 mg. per ml.-i.e., more than a thousand times as high as the concentration of IgA sufficient to give H.T.-agglutination. The most interesting point about H.T.-agglutination is the close resemblance between some properties of the factor causing it and the properties of reaginic antibodies. Effects of Heat and 2-mercaptoethanol The H.T.-agglutination factor is inactivated at 560C in a time which depends on the concentration of IgA; a solution containing 200 tLg. IgA per ml. is completely inactivated in 30 minutes. The same solution would be completely inactivated in 18-24 hours at 37"C in 2-3 days at room temperature, and in about 10 days at 4°C. The activity lasts for years when the preparations are kept at -20°C or when they are freeze-dried. The reducing agent, 2-mercaptoethanol, at a 0-1M concentration, has no appreciable effect on the activity of IgA at 4OC; it has a very slow effect at room temperature, more than 19 hours having to elapse before there is a partial but significant TABLE I-FRACTIONATION OF
IgA
BY
PRECIPITATION WITH
1-18 M
SODIUM SULPHATE
reduction in activity;
the other hand, the solution 200 ml. is containing g. IgA per completely inactivated by 2-mercaptoethanol in 60 minutes and almost completely in 30 minutes. Thus, H.T.-agglutinating activity, like the skin-sensitising activity of reagins, is heat-labile at 56C and shows a susceptibility to reduction which can be considered as intermediate between that of IgM and IgG. IgA already fixed to tanned cells withstands heating at 56°C for periods at least eight times as long or reduction with 2-mercaptoethanol at 37’C for periods at least four times as long as those required for the inactivation of IgA in solution. In other words, the agglutination of tanned red blood-cells brought about by IgA is not destroyed by heating or reduction for such prolonged periods. This is also similar to what happens to reagins, which remain active for weeks when fixed to the skin (at the site of a passive-transfer test) while they cannot withstand body temperature for so long a time when they are in a test tube. at
37 °C,
on
Effect of pH IgA preparations retain their H.T.-agglutinating ability unchanged at pH 6-4—86, but are completely inactive at pH 4-2. (The pH
range of
activity of reagins is
between 6 and
11.)
Fractionation When whole sera or IgA preparations endowed with reaginic activity are fractionated by precipitation with 1- 18 A1 sodium sulphate all the skin-sensitising activity is recovered in the soluble fraction, which contains only a small amount of the
IgA originally present: whereas the precipitate, which contains IgA, lacks any reaginic activity.22 When IgA preparations, either from normal or hypersensitive sera, are similarly fractionated, H.T.-agglutinating activity is only found in the fraction soluble in 1-18 M sodium sulphate, whereas the precipitate is inactive. If IgA comes from a clinically hypersensitive patient, the recoveries of skin-sensitising and H.T.agglutinating activities in the soluble fraction closely parallel each other (table I). Further, when IgA preparations are fractionated by agar-gel electrophoresis both activities are found in the same fractions (table II).
the bulk of
CONCLUSIONS
H.T.-agglutination and skin-sensitisation are related in some way. They may be properties of the same substance, but if so I cannot at present identify that substance. All the active preparations contained IgA, although none of them were in an absolutely pure state. Moreover, IgA is present in even greater amounts in fractions devoid of any activity. If H.T.-agglutinating activity and reaginic activity are properties of IgA, it is clear that they are not properties of IgA as a whole, but rather of a subfraction which is different in some way from the bulk of the protein. We do not know enough about IgA to say whether this subfraction should be given a different name. The factor responsible for H.T.-agglutination, although different from reagin, may play some role in the process of tissue sensitisation by reaginic antibodies; and even if it does not the reaction would seem to be an easy-toperform index of the activity of the fraction being handled. Instituto de Investigaciones Médicas, Fundación Jiménez Diaz, Madrid 3
FERNANDO ORTIZ M.D.
Madrid
R-FACTOR-MEDIATED RESISTANCE TO NEW
AMINOGLYCOSIDE ANTIBIOTICS To determine whether the existence and
Summary incidence of R-factor genes
are
always
related to the prevalence of antibiotics in the environment, recent clinical isolates of Escherichia coli and salmonella tested for resistance to 5 relatively new aminoglycoside antibiotics. 19 strains were resistant to bluensomycin, and 11 to spectinomycin, gentamycin, and viomycin: in each case, the resistance genes were associated with R factors. Bluensomycin resistance was mediated by the streptomycin R-determinant; but the resistances to spectinomycin, gentamycin, and viomycin were mediated These by previously unidentified R-determinants. results ate discussed in relation to the origin of R-factor were
genes. INTRODUCTION
R FACTORS are episomes of enteric bacteria that mediate resistance to several commonly employed antibacterial drugs: those defined originally in Japan mediated resist22. Ortiz, F.
Nature, Lond. (in the press).
253 ance to
sulphonamides, chloramphenicol, streptomycin,
and tetracycline 1; others described by workers in Europe also mediate resistance to neomycin-kanamycin,2ampicillin-cephalothin,3 and furazolidine.4 These newer R factors were initially isolated only after the commercial introduction of the drugs; their subsequent prevalence has been attributed to the increasing use of the drug in medicine and agriculture. Since the type of drug resistance engendered may throw light on the origin of R factor genes and may have practical importance as well, cultures of Escherichia coli and salmonella isolated from patients were examined for resistance to five relatively new aminoglycoside antibiotics. A large proportion of strains carrying previously identified R factors were found to be resistant to four of the drugs; in each case, the resistance genes were associated with the R factor.
each test drug was transferred by all potential donors; with the exception of 3 strains that each had two R factors, recipient strains acquired the entire pattern of resistance of the donor strains. Recipient strains transferred resistance to the test drug as part of the R factor in subsequent matings with other enteric bacteria. Moreover, bacteria that lost R factors lost resistance to the test drugs. Finally, resistance to the drugs was cotransduced by phage-PIwith other R-factor genes but was not cotransduced with chromosomal genes. These results clearly indicate that previously defined R factors contain genes (R-determinants) that mediate resistance to bluensomycin, gentamycin, spectinomycin, and viomycin. R-determinants for the test drugs were associated with fi- and fi+ R factors and various other R-determinants TABLE I-GENOTYPES OF R FACTORS CONTAINING R-DETERMINANTS FOR
METHODS
THE TEST DRUGS
The cultures studied were unselected strains that had remained viable since earlier investigations’; the E. coli had been isolated from the urine of children with genitourinarytract infection, the salmonella from stools of children with gastrointestinal disease. The pattern of resistance to each of eight commonly used antibacterial drugs had been determined by disc techniques. Methods of culture, conjugation, and stock bacteria used as conjugation recipients have been described previously.56 The cultures were taken to be resistant to the drugs if they formed colonies on eosin/methylene-blue/ lactose agar containing three or ten times the drug concentration that prevented colony formation of stock bacteria. Minimal inhibitory concentrations of drug were determined by twofold dilutions in tryptic digest broth, using the concentrations of inoculum indicated. Transduction with phage-Pl and penicillin selection of drug-sensitive segregants of resistant bacteria were carried out according to the methods described by Watanabe.’7 RESULTS
tested for resistance
bluensomycin, gentamycin, hygromycin, spectinomycin, and viomycin. Each drug was specifically selected for the study. The mechanism of action of each is under investigation in this department. Hygromycin has been added to agricultural feeds in the United States but has not been used clinically. Bluensomycin has been used only in laboratory investigations, and mutants of E. coli selected for single-step highlevel resistance show cross-resistance with streptomycin. Gentamycin and spectinomycin have undergone restricted clinical trials, but had not yet been officially introduced when these strains were isolated; neither drug had been used in this area. Viomycin was officially introduced for clinical use in 1953 but is not used as extensively as other aminoglycoside drugs. Of 20 cultures carrying R factors, 19 were resistant to bluensomycin, 11 to spectinomycin, and 11 to viomycin and gentamycin; both E. coli and salmonella were resistant to the drugs. 19 of these 20 strains were resistant to streptomycin, but none was resistant to neomycin and kanamycin. The cultures that were sensitive to other drugs were also sensitive to each of the five test drugs. Strains resistant to the test drugs were mated with sensitive E. coli, and recipients were selected for resistance to tetracycline, ampicillin, or a test drug. Resistance to 37 cultures
1. 2. 3. 4. 5. 6. 7.
were
to
Watanabe, T. Bact. Rev. 1963, 27, 87. Lebek, G. Z. Hyg. InfektKrankh. 1963, 149, 255. Anderson, E. S., Datta, N. Lancet, 1965, i. 407. Smith, H. W., Halls, S. Br. med. J. 1966, i, 266. Smith, D. H., Armour, S. E. Lancet, 1966, ii, 15. Smith, D. H. New Engl. J. Med. 1966, 275, 626. Watanabe, T. Meth. med. Res. 1964, 10, 202.
The R-determinants are listed and spaced for does not reflect genetic map positions.
convenience;
the order
presented
(table I). Resistance to streptomycin and bluensomycin was always associated and was not separated by transduction or the selection of clones sensitive to one of the drugs, suggesting that one gene mediates resistance to both drugs. Spectinomycin resistance was mediated by a separate R-determinant, however, since some strains resistant to streptomycin and bluensomycin were sensitive to spectinomycin; segregant clones, resistant only to spectinomycin, were selected from strains originally resistant to all three drugs. Viomycin and gentamycin resistance occurred independently of that to other aminoglycoside drugs and was probably mediated by a single R-determinant: resistance to both drugs occurred in the same strains and was not separated by transduction or segregation. These results were not expected, since the drugs differ TABLE
11—FACTORS
INFLUENCING
THE
LEVEL
OF
RESISTANCE MEDIATED BY R FACTORS
SPECTINOMYCIN
254
substantially in chemical composition and in their in-vitro effects on protein synthesis by extracts of E. coli; although the chemical structure of the drugs is not known, these results strongly suggest that both share a critical structural component. The levels of resistance mediated by the R-determinants of the test drug depended on the episome, species of the host bacterium, inoculum concentration, and test medium (table II). The mechanism by which these R-determinants mediate resistance to the drugs has not yet been completely defined, but it did not depend on a change in the ribosomes-the site of action of these drugs. The test drugs inhibited in-vitro protein synthesis by purified ribosomes of resistant and sensitive E. coli to the same extent. The direct relation between minimal inhibitory concentrations of drug and inoculum concentrations of bacteria suggests that these R-determinants may mediate the synthesis of enzymes capable of inactivating the test drugs; intracellular inactivation of the drugs has not yet been excluded, but extracellular inactivation was not observed. DISCUSSION
The results indicate that the origin and incidence of R-determinants may not always be related to the prevalence of the drug in the environment. And, although the R-determinants for spectinomycin and gentamycinviomycin may have been selected during their limited clinical use, it seems more probable that they have been selected either by natural exposure of enteric bacteria to the streptomyces producing the antibiotics or by selective pressures bearing only a fortuitous relation to the antibiotics. A fortuitous relation is supported by the observation that some R factors alter the susceptibility of bacteria to N-methyl-N’-nitro-N-nitrosoguanidine, an organic chemical used only in laboratory investigations for its strong
lack xanthine oxidase. The enzyme first appears in rat liver some six days after birth. As a result ferritin-bound iron is deposited in ever-increasing quantities in the fcetal liver in the absence of the normal iron-releasing mechanism.’’ The appearance of hepatic xanthine oxidase in the healthy animal soon after birth releases liver-stored iron needed by the animal at a time when it is deprived of its maternal source of iron; when it is subsisting on a diet of mother’s milk that is deficient in iron, and when its needs for iron for haemoglobin biosynthesis are great. Powell and Emerson5 noted a relation between hepatic xanthine-oxidase activity and the extent of iron storage in the rat liver: rats fed the xanthine-oxidase inhibitor allopurinol (4-hydroxypyraxolopyrimidine) with the standard laboratory meal retained a significant amount of hepatic iron. Addition of iron salts to the diet resulted in a hepatic hsmosiderosis that was histologically more pronounced, at all stages, in animals receiving allopurinol. An elevated serum-iron in one of five patients with gout who were receiving allopurinol was also observed.5 Rundles et al.,6 on the other hand, found no serumiron changes in similar patients receiving this drug; but the absence of hypersiderxmia does not exclude the possibility of excessive hepatic iron, since 20% of the patients with haemochromatosis reported by Dreyfus and Schapirahad normal serum-iron values. METHOD AND RESULTS
We have considered the possibility that a failure of xanthine oxidase to make its normal appearance in the liver might, over a period of years, lead to excessive deposits of iron such as is observed in the disease haemochromatosis. We have analysed 18 human liver samples for enzyme xanthine oxidase by a microfluorometric method.The results expressed per unit weight of total sample protein, are listed in the table for three groups of patients :2 control liver samples XANTHINE-OXIDASE-ACTIVITY IN HUMAN LIVER
mutagenic properties. Further studies on the origin of R-factor genes should have importance for both medicine and biology. This work was supported in part by a U.S. Public Health Services grant no. G.M.-14119-01. Requests for reprints should be addressed to D. H. S., Department of Bacteriology and Immunology, Harvard Medical School, 25, Shattuck Street, Boston, Massachusetts 02115, U.S.A.
Departments of Pediatrics and Bacteriology and Immunology, Harvard Medical School, Boston, Massachusetts
DAVID H. SMITH M.D.
Rochester
HÆMOCHROMATOSIS AND HEPATIC XANTHINE OXIDASE IN the normal mechanism for release of hepatic ferritin iron in the rat ferritin-bound ferric iron is reduced to the ferrous state by the reduced form of the enzyme xanthine oxidase during its oxidation of hypoxanthine and xanthine to uric acid.l The iron-ferritin bond is made labile and the iron is transferred to the plasma where it is bound to transferrin. In the healthy adult rat, liver xanthine oxidase is present in excess; the factor limiting the rate of release of ferritin iron is the cellular concentration of hypoxanthine and xanthine. Foetal and newborn
rat
liver2 and human newborn liver3
Green, S., Mazur, A. J. Biol. Chem. 1957, 227, 653; Mazur, A., Green, S., Saha, A., Carleton A. J. clin. Invest. 1958, 37, 1809. 2. Westerifield, A. W., Richert, D. A. J. biol. Chem. 1950, 184, 163. 3. Wells, H. G., Corper, H. J. ibid. 1909, 6, 469. 1.
obtained during surgery or necropsy from patients without liver disease, biopsy samples from patients with cirrhosis, and a third group from patients with idiopathic hamiochromatosis. The range of values in the control samples, from 1-81 to 549 includes one (1-81) from a necropsy specimen. The xanthineoxidase activities in liver samples from patients with cirrhosis or haemochromatosis are all below those in the control group, and in a number of instances there was no evidence for any xanthine-oxidase activity. DISCUSSION
Although there was no overlap of xanthine-oxidaseactivity values between the controls and those with liver disease, interpretation of the data must be tempered by the relatively small number of samples in each group and the fact that expression of enzyme activity per unit weight of protein does not take into account the presence of Mazur, A., Carleton, A. Blood, 1965, 26, 317. Powell, L. W., Emerson, B. T. Lancet, 1966, i, p. 239. Rundles, R. W., Metz, E. N., Silberman, H. R. Archs intern. Med 1966, 64, 229. 7. Dreyfus, J.-C., Schapira, G. in Iron Metabolism (edited by F. Gross). p. 296. Berlin, 1964. 8. Burch, H. B., Lowry, O. H., Padilla, A. M., Combs, A. M. J. brit Chem. 1956, 223, 29. 4. 5. 6.