THE DEVELOPMENT OF EXPERiMENTAL MEDEA FOR ISOLATEON AND CULTE‘V‘ATION GF MYCOBACTEREUM TUBERCULOSIS: Thesis for fhc Degree cf Ph. D. MICHIGAN STATE COLLEGE Joseph Wiison Whalan 1955 THE515 This is to certify that the thesis entitled The Development of Experimental Media for Isolation and Cultivation of Mycobacterium Tuberculosis. presented by Mr. Joseph wilson Whalen has been accepted towards fulfillment of the requirements for PH.D. degree in Bacteriology Majoa professor Date September 30, 1954 0-169 THE DEVELOPNENT OF EXPERIMENTAL MEDIA FOR ISOLATION AND CULTIVATION OF MYCOBACTERIUM TUBERCULOSIS By Joseph Wilson Whalen A THESIS Submitted to the School of Graduate Studies of Michigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY twpartment of Bacteriology and Public Health 1955 rHEL‘L‘b‘ m t ng$ ’ w t ACKNOWLEDGMENT The author wishes to express his appreciation to Dr. W. L. Mallmann whose guidance, suggestions, and criticisms contributed extensively toward organization of the experimental studies; to Dr. A. S. Kimball, Jr., Dr. S. A. Yannitelli, and Mr. D. A. Grover who provided laboratory facilities and materials; and to the National Tuberculosis Association for Medical Research Fellowship grants supporting the investigation. THE DEVELOPMENT OF EXPERIMENTAL MEDIA FOR ISOLATION AND CULTIVATION OF MYCOBACTERIUM TUBERCULOSIS By Joseph Wilson Whalen AN ABSTRACT Submitted to the School of Graduate Studies of Michigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Bacteriology and Public Health 1955 Appr0V0d_‘MW .81 Joseph Wilson Whaler The studies described were undertaken to develop simply prepared culture media, the use of which would facilitate recognition and removal of colonies of Excobacterium.tuberculosis, as well as supporting rapid and profuse growth of this organism. Original formulations of experimental media were compared with the Dubos-Davis and Petragnani media. Modifications were effected through addition or removal of l ingredient, and were compared with the above media, as well as others conventionally used for isolation and cultivation of tubercle bacilli. Evaluation was through comparison of both the time required for detection of visible growth and the ultimate extent of growth from inocula consisting of either pathologi- cal specimens or suspensions of E. tuberculosis. Generation time determinations, as well as turbido- metric and micro-Kjeldahl growth rate measurements, a were also used. Among the many types of media deve10ped and studied, several were considered satisfactory. Earlier appearance of growth and a greater number of positive cultures were obtained on an experimental egg medium than on the Petragnani medium. Two agar media supported faster Joseph Wilson Whale: 2 growth of tubercle bacilli than the Dubos-Middlebrook cleats-albumin agar. The highest number of positive cultures and the most rapid growth of E, tuberculosis from.sputum.specimens were obtained with ethyl violet- sodium azide liquid media when they were compared with the Dubos-Davis medium and the Loewenstein-Jensen, . Patrick, Petragnani, and Peizer egg media. A much larger number of positive cultures was secured from sputa of patients receiving streptomycin-~para-amino salicylic acid therapy with ethyl violet-azide than with the other media. Contamination of these experi- mental media was less than with the Dubos-Davis medium, but greater than with the egg media. In the presence of ethyl violet, aggregates and colonies of tubercle bacilli are blue-violet. This typical color aided detection and differentiation of growth, but contami- nating streptococci, forming purple colonies, sometimes masked it. The use of sodium azide in experimental media was subsequently discontinued because of its toxicity for M. tuberculosis. Contamination of ethyl violet media was effectively suppressed by using penicillin and acti-dione as additional selective agents. Tubercle bacilli grow in characteristic blue-violet Joseph Wilson Whalen 5 colonies on the surface of glass wool in ethyl violet culture media. Colonies are light pink in neutral red media and a more brilliant pink in Sudan IV oleate media. The formation of Sudan oleate-albumin complexes made it possible to use fat soluble dyes in media. Ethyl violet, neutral red, ethyl violet-~neutral red, and Sudan IV oleate liquid experimental media in glass wool flasks, and the Petragnani medium were employed to isolate'M. tuberculosis. Colonies of tubercle bacilli were detected on the Sudan IV oleate medium.after an average of 8 days incubation. Earliest detection was on this medium because of the distinctively vivid colony color and more rapid multiplication of the organism. Colonies of saprOphytic mycobacteria can be differentiated from those of M. tuberculosis on neutral red and Sudan IV glass wool media. One grade of activated charcoal, when incorporated in an agar medium, stimulated growth of tubercle bacilli as much as did human serum and oleic albumin. Addition- al generation time studies and routine cultural use established that a 1 to h00,000 concentration of ethyl violet is the optimum selective concentration for this medium. In culturing pathological material, inclusion of 100 units of penicillin and 100 micrograms of Joseph Wifison Whalen actidione per ml further reduced contamination. A greater and more rapid recovery of M. tuberculosis was achieved with the charcoal medium. It is easily prepared, since sterilization is by autoclaving. Addition of filter-sterilized enrichments is unnecessary. VITA Joseph Wilson Whalen candidate for the degree of Doctor of PhilosOphy Oral Examination: September 50, l95h, 2:00 p.m. Room 525, Giltner Hall Dissertation: The DeveloPment of Experimental Media 'for Isolation and Cultivation of Mycobacterium.Tuberculosig Outline of Studies: Major Subject: Bacteriology and Public Health Minor Subjects: Biochemistry, Physical Chemistry Biographical Information: Born: May 27, 1923, Battle Creek, Michigan Undergraduate Studies: Michigan State College, l9h6- l9h9, B. s., Bacteriology and Public Health, l9h9 Graduate Studies: Michigan State College, l9h9-l951: M.S., Bacteriology and Public Health, 1951, sponsor Dr. W. L. Mallmann. Thesis Title: A Comparative Study of Various Digestion Agents Used in the Isolation of Mycobacterium. Tuberculosis From.Sputa Michigan State College 1951-1955; Ph.D. 1955, sponsor Dr. W. L. Mallmann Experience: Assistant Bacteriolo ist, Battle Creek Health Dept. l9hl-l9 3; Medical Labora- tory Technician, Army Air Force 19h}- l9h6; Bacteriologist and Serologist, Leila Y. Post Montgomery Hospital l9h7: Bacteriologist, Arthur S. Kimball Sana- torium.and Calhoun County Health Dept. l9h8-l952, Medical Research Fellow of the National Tuberculosis Association l952-195h; Laboratory Director, Arthur S. Kimball Sanatorium.l95h Affiliations: The Society of American Bacteriologists The Society of the Sigma Xi The American Public Health Association llll I'll ‘ II! ’ TABLE OF CONTENTS Page Introduction 1 Nutritional Requirements of the Tubercle Bacillus 2 Carbon Nutrition Nitrogen Nutrition Inorganic Nutrition 15 Media for Primary Isolation and Study of g. tuberculosis 16 Growth of M. tuberculosis Beneath the Surface of LiquidIMedia 20 Problems in the Rapid Isolation of'M. tuberculosis 50 I Exploratory Studies on Media for Tubercle Bacilli 3h Experimental 3h II Studies of the Toxicity of Ethyl violet and Malachite Green to‘M. tuberculosis 56 Materials and Methods 57 Results 60 Discussion 65 III Primary Isolation of M. tuberculosis in Ethyl Violet Liquid Media 69 Materials and Methods 70 VI VII VIII Results Discussion The Influence of Therapy on Primary Isolation of Tubercle Bacilli Study of Selective Agents in Primary Isolation Materials and Methods Discussion of Results Effects of Selective Agents on the Growth Rate of M. tuberculosis in Liquid Media Primary Isolation of M. tuberculosis on Glass Wool Suspended In Liquid Media Materials and Methods Results Discussion Cultivation and Isolation of E. tuberculo- glg.on Charcoal Agar , Materials and Methods Results Discussion General Discussion Summary Tables Literature Cited 75 78 85 88 88 90 92 95 98 99 102 107 108 111 116 119 151 Introduction The-growth of Mygpbacterium.tuberculosis is relatively slow even when it is cultivated on special bacteriological media. It grows poorly or not at all on many media used for culturing other bacterial species. Multiplication on primary isolation is slower than that of cultural strains of tubercle bacilli. In some instances tubercle bacilli in pathologic inocula do not initiate growth on the accepted isolation media. In diagnosis, isolation of these bacilli sup- plies final proof of tuberculous infection. Rapid recovery of.M. tuberculosis from.sputum.or other discharges of the patient facilitates diagnosis and serves as a means of evaluating his response to therapy. Thus cultural demonstration of the organism.is of prime importance in detection, treatment, and subsequent observation of the tu- berculous infection. The studies described in the experimental section were undertaken in an effort to develOp media permitting more rapid isolation and greater recovery of M, tuberculosis than can be achieved (1) using existing cultural media. Nutritional Requirements of the Tubercle Bacillus Since Koch (55) isolated M, tuberculosis and proved that it is the etiological agent of tuberc- ulosis, many investigations of the growth require- ments and intermediary metabolism.of this organism have been conducted. Many of the results of these studies are in conflict on major or minor aspects. Some of these disagreements are undoubtedly due to strain variances, methods of study, contamination of materials, and unwarranted assumptions. Others may be due to the relative difficulty in the culti- vation of the tubercle bacilli in the undissociated state and the extreme difficulty in studying the enzymatic activity transpiring beneath the unusual slime layer. Removal of this layer without protein denaturation and partial inactivation of enzymes has probably never been achieved. The hydr0phobic character and low permeability of the slime layer raise further barriers to full comprehension of the metabolism of tubercle bacilli. Difficulty in dis- persal of these organisms has probably invalidated many studies from the standpoints of control of (2) inocula and measurement of extent of multiplication. However, the greatest obstacle present in such studies is the near impossibility of evalu- ating the importance of a single constituent in a medium. It has been shown by many workers that the intensity of the action of a favoring substance upon the abundance of a culture, is dependent not only upon its concentration, but also on the gener- al composition of the medium (71). The interaction between constituents is responsible for findings indicating that a substance is stimulatory to growth of‘M. tuberculosis in one medium. but has no effect or is actually inhibitory in another. Therefore, the effect of a chemical added to a new mmdium.can- not be ascertained except by trial in that medium, whose composition may bring about an entirely different response from.that expected. Carbon Nutrition During his original studies of M. tuberculosis, Koch was successful in cultivating this organism on slants of heat-coagulated bovine serum.(55). As an inoculum, he used mascerated tissue contain- ing large numbers of tubercle bacilli. Even with (3) this massive seeding, 15 days incubation was re- quired before colonies were detectable macroscopi- cally. Smaller inocula either grew much more slowly, or failed entirely to produce colonies. Since he had previously failed to obtain multi- plication of this organism on simpler media, Koch realized that its growth requirements were more exacting than those of many other bacteria. The prolonged incubation necessary brought about marked dehydration of the medium. causing it to crack and shrink. Nocard (75) noted that tubercle bacilli grew slowly and sparsely after this dehydration had occurred. Therefore, he added glycerol to the medium.in a concentration of h per cent, only to find that it also served as an excellent carbon.source for these organ- isms 0 Subsequent studies by Nocard and Roux (76) revealed that a 5 per cent concentration of glycer- ol was Optimum, as larger amounts reduced the total yield of tubercle bacilli. Hammerschlag (uh) later reported that the concentration may be lowered to 1.5 per cent without reducing the amount of the harvested culture. However, Frouin (h) and Guillaumie (58) found that the total weight of tubercle bacilli grown in a synthetic medium increased in preportion to the increase in the glycerol content of the medium up to 5 per cent. The weights of bacilli harvested from.media con- taining 6 to 10 per cent were progressively di- minished. Studies based on weights of the final yields of tubercle bacilli when used to evaluate stimm~ latory effects of glycerol are not as valid as are other means such as bacterial nitrogen de- terminations, since tubercle bacilli grown in a medium.with 12.5 per cent glycerol contain approximately 2.5 times as much lipid as those grown in the presence of 0.5 per cent (62). Nevertheless, it is evident that a glycerol con- centration of 10 per cent will reduce the number of tubercle bacilli which can be produced in a specific medium. Kendall, Day, and Walker (50) found that bovine tubercle bacilli are inhibited by glycer- ol concentrations greater than 0.70 per cent. Apparently most strains are unable to utilize (5) glycerol. However a few strains grow well in the presence of this substance, and others which were originally adversely affected, have been adapted so that they will utilize glycerol as their only source of carbon. Media for the isolation of'g. tuberculosis var.‘bgzgg from.pathological materi- al rarely contain glycerol. war has been substi- tuted in a modified Petragnani medium.prepared for this purpose, since the growth of bovine ba- cilli is stimulated by this substance (22). Avian tubercle bacilli utilize glycerol in concentrations roughly equivalent to those stimu- latory to the human variety, but they are able to use many other sources of carbon as effectively (90). Frouin and Guillaumie (57) reported that glucose in equal concentrations produced a greater yield in weight of tubercle bacilli than glycerol. This effect is produced only in certain media, as other workers have reported that glucose is not as good a carbon source (67, 8A). However, in recent growth rate studies Youmans and'Youmans (109) observed that glycerol is a better source of carbon than glucose in similar concentrations. (6) If the concentrations of glucose and glycerol are not larger than 1 and 5 per cent, respectively, a synergistic growth stimulation is obtained.(37). Amounts of glucose greater than 1 per cent reduce the yield of bacilli. Thus the relative preportions of these sources, as well as the nature of the other constituents of the medium, play an important role in Optimal carbon nutrition of tubercle bacilli. Another carbon compound generally used in the culture of these organisms is citric acid and its salts. Bance (5) found that only a relatively small amount is used by tubercle bacilli during growth, but that growth was more luxuriant and rapid when citric acid was present in the Sauton medium. The original purpose of this compound in the medium was to prevent the precipitation of magnesium.phosphate, formed by the reaction be- tween monOpotassium.phosphate and magnesium sulfate. Oxidation of citrate by human tubercle bacilli has never been demonstrated. Edson (50) has expressed the belief that citrate may provide.a means of transporting magnesium and iron to the interior of the bacterial cell. The studies of Frouin and Guillaumie (38) indicate that the citrates, like (7) glucose, stimulate growth when present in some media, while they do not manifest this effect in others. Bretey and Andrejew (12) reported that 0.05 per cent tween 80 in a Dubosrtype medium slightly increased the respiration of‘M.-tuberculosis. At a concentration of 5 per cent it proved toxic. Other esterified fatty acids, in weak concentrations, as well as the sodium.oleate portion of Dubos' oleic albumin complex will also serve as poor carbon sources. Schaefer, Marshak and Burkhart (92) demonstrated that bovine serum.albumin will furnish a small amount of carbon for utilization by tubercle bacilli. Dubos (2h) confirmed this, and showed that with purified bovine albumin (fraction V), which is free of lipids, a source of carbon is not available. The work of Youmans and Youmans (10h) indicates that certain fatty acids, through a limited range of low concentrations, will serve as carbon sources for virulent tubercle bacilli although a protective substance is not present in the medium. The low molecular weight acids were more effective. Caprylic and capric acids stimulated the rate of growth nearly as mmeh as did 2.0 per cent glycerol. (8) Shaefer (91) observed that in the presence of serum.a1bumin, which reduces or nullifies the toxicity of many substances, human tubercle ba- cilli use acetate, lactate, pyryvate, and succinate to a slight extent. However, as carbon sources, they are too poor to be utilized in media for isolation or cultivation of these organisms. However, Youmans and Youmans (105) found that the growth of the H57Rv strain of M, 22333- culosis was supported by pyruvic, acetic, lactic, oxalacetic, oxalosuccinic, and i-ketoglutaric acids, and that the growth rates obtained were similar to the rate observed in the presence of glycerol. Nitrogenous Nutrition Following the work of Bouveault (11), who established that M. tuberculosis will assimilate creatinine and creatine, but not peptone, Kuhne (56) demonstrated that leucine, taurine, and asparagine were also assimilated. Studies by Proskauer and Beck (8h) established that aspara- gine serves as an excellent source of nitrogen in the cultivation of tubercle bacilli. Calmette, (9) Massol and Breton (15) also found it to be much superior to peptone in this respect. Sperry and Rettger (91+) demonstrated that the nitrogen of peptones not containing free amdno acids is not directly available to tubercle bacilli, and Masucci (66) showed that larger yields of tu- berculin were obtained in peptone broth when amino acids were added. The amino acids of the aliphatic series exert a more favorable effect on bacillary growth than do the aromatics, tyrosine and phenyl- alanine (h). Crimm.and Martos (17) reported that amigen, an enzymatic hydrolysate of casein, was an excellent source of amino acids for the cultivation of tubercle bacilli. Acid hydrolysates are less satisfactory. Dubos (28) has found commercial casein hydrolysates to be important ingredients in the nutrition of these organisms. Numerous investigators have found that these preparations will serve as the sole source of nitrogen in various media. These appear to provide necessary amino acids, as well as an unknown factor or factors which are stimulatory to growing tubercle bacilli. (10) While these organisms will utilize inorganic sources of nitrogen, asparagine is much superior to any of the ammonium salts (61). Ammonium succinate, tartrate, and citrate are also less favorable to growth. Gerundo (LO) postulated that asparagine is metabolized by tubercle bacilli to succinic acid, which is then utilized for growth. However, Youmans and Youmans (105) have demonstrated that succinic acid will not serve as the sole source of carbon, whereas it has been found that asparagine will supply although poorly, carbon as well as nitrogen. Recently Youmans and Youmans (105) have found that L-asparagine, of 25 amino acids studied, produced the most rapid and heaviest growth of the virulent strain H57Rv in a modified Proskauer and Beck medium in which it was the sole source of nitrogen. The presence of L-glutamic or DL- aspartic acid effected a slightly slower rate of multiplication. Use of the individual amino acids in conjunction with asparagine in the basal medium.produced no growth stimulatory effect over that of asparagine alone. These investigators (ll) observed that the H57Rv strain of M. tuberculosis grew in the basal synthetic medium without a source of nitrogen. Inocula as small as 10"6 mg initiated growth and the growth rate was nearly as rapid as when asparagine was present. The strain was main- tained on the nitrogen free medium for over three years. In the absence of glycerol no growth occurr- ed from inocula as large as 10-1 mg. This phenomenon casts doubts on the reliability of all studies of nitrogenous nutrition of )3. 1:3- berculosis since it suggests that this organism. may be able to fix atmospheric nitrogen or to utilize the ammonia of the air. If this is true, actual nitrogenous starvation was probably not achieved in the culture media. Youmans and Youmans (105) were able to at least partially circumvent this objection through removal of magnesium citrate from the medium. Without this ingredient minute inocula of tubercle bacilli did not grow in the absence of a nitrogen source. Apparently the bacilli are less able to metabolize extraneous nitrogen from.the unknown source in the absence of this salt. (12) Inorganic Nutrition The inorganic requirements oflg. tuberculosis are not particularly different from.those of other bacteria. Sauton (89), Frouin and Guillaumie (58), and Dubos (28) have established that the indispensa- ble mineral elements are phosphorus, potassium or rubidium. sulfur, and magnesium. In various complex media, calcium. sodium. chlorine, iron, silicon, manganese, copper, and zinc have been used. These elements are not necessary to support growth from minimal inocula. Most of these will, however, enhance and speed growth of tubercle bacilli when they are added to a medium.containing only the basic essentials. Roche (86) reported that sodium. silicate and silicic acid exerted no growth stimu- latory effect. Sodium. potassium.and phosphorus have been amply furnished by the buffer systems used in most media for tubercle bacilli. Too great a potassium. concentration will decrease the growth rate and volume (58). The mechanism.of this partial inhi- bition is unknbwn. Boissevain (9) demonstrated that rubidium (15) will partially substitute for potassium in the nutrition of tubercle bacilli. No reduction in virulence was found after 10 successive transplants in his medium containing rubidium chloride. Only a small amount of magnesium is required for luxuriant growth of‘M. tuberculosis. Tiffeneau and Marie (100) found that 0.6 gm.magnesium.sulfate per liter was sufficient to supply both the sulfur and magnesium requirements of the bacilli. When large yields of tubercle bacilli are desired, how- ever, the latter element soon becomes the limiting growth factor (28). Henley (h?) obtained an increase in cultural yield of Al per cent over that of his controls, when 10 to 20 mg of iron sulfate were added to 100 ml of broth containing 7 per cent glycerol and l per cent peptone. Long and Seibert (65) found that inclusion of ferric ammonium citrate in Long's synthetic medium increased the total weight or the harvested bacilli 7.5 per cent. The work of Frouin and Guillaumie (38) indicates that the favorable influence.of iron is dependent lipon the concentration in the medium of essential :minerals, and the carbon.nutrition of the bacilli. (1h) The stimulatory effect was greater in neutral or acid media with a high glycerol content. Iron also was favorable to growth in the presence of glucose, but only in an acid medium. The fat content of the tubercle bacilli was lower when iron was present if the medium was deficient in mineral salts . More recently Dubos (28) has demonstrated that a very low concentration of ferric ammonium. citrate (0.005 gm per liter) adequately fulfills the requirements of the tubercle bacilli growing in liquid media. In agar media, he recommended that the concentration be increased to 0.05 gm per liter, as he found that a portion of the iron is bound by agar in a form unavailable to the organisms. The inorganic elements calcium. sodium. chlorine, manganese, zinc, and copper are required for growth enhancement only in traces, and may be supplied by tap water and/or by their presence in peptones, amino acid mixtures, and even in chemical reagents used in.media preparation. (15) Media for Primary Isolation and Study of M. tuberculosis A larger number of culture media has been develOped for the primary isolation of M. tubercu- lpglg_from.pathological material since Kech first cultivated this organism _l_._I_1_ EELS. For the most part, solid media have been favored over liquid for this purpose, since it is more difficult to detect growth of tubercle bacilli in the presence of contaminants in broth. Also, until fairly recently, no liquid medium existed in which M. tuberculosis would grow well beneath the surface, so that liquid media in tubes did not offer as great a surface area to support multiplication as did the solid slants. A heat coagulated protein was used as the base of the solid media, as agar was recognized to be toxic to tubercle bacilli (16). Proteins most frequently used were blood serum.(76) and egg albumins (115). Attempts were made to use glycerolated potato as an iso- lation medium. It was found however that it supported sub-cultures of the organism much more readily. The complex protein media contained glycerol (16) glucose, milk, meat or peptones, potato or potato starch, as well as the egg or serum base. Thus they contained unknown and unrecognized inorganic and organic substances which might affect growth of M. tuberculosis in a favorable or unfavorable fashion. It is now recognized that many of the constituents of these media are unnecessary for primary isolation and cultivation of the tubercle bacillus. Although sputa and other pathological material likely to contain saprOphytic or commensal bacteria were treated with acid or basic reagents or with hypochlorite preparations, these bacteria frequently survived this digestion to contaminate or overgrow the media. When the specimens were treated with sufficient concentrations of digestants to completely destroy the undesirable bacteria, the tubercle ba- cilli often failed to multiply unless they were present in large numbers in the inoculum.(h2). Therefore, selective dyes were added to the egg media. Since gramppositive bacteria were among the most frequent contaminants, crystal violet, which effectively inhibits them when in proper concentration, was used (80). Among the media containing this dye (17) were: Petroff's,Corper's, and modifications of Loewenstein's and Dorset's. Subsequently it was found that malachite green was less toxic to M. tuberculosis (79) and this dye was incorporated in a number of new as well as modified versions of some of the older media. Media of Petragnani, Petrik, Loewenstein- Jensen, and the National Tuberculosis Association, contain this selective dye. With these media, a greater number of positive cultures was obtained than with the older media containing crystal violet, while the amount of contamination was approximately the same for both groups (6h). Recently penicillin has been incorporated in both the classical egg media and the newer liquid and solid preparations of Dubos and his co-workers, as a selective agent (1, hi, 116). However, this antibiotic is toxic to tubercle bacilli if the concentrations used are too great (35, 51, 93). It is difficult to state the criti- cal limit above which this toxicity is markedly manifest, since the type of medium and its compo- sition greatly influence it. In the presence of agents which disperse tubercle bacilli, the effect (18) of any toxic agents in the medium is greatly enhanced (11h). Undenatured serum.albumin will protect, to some extent, the bacilli from.those unfavorable substances, including penicillin. Another protective factor is produced by the organism itself. ‘M. tuberculosis elaborates a penicillinase which will inactivate extremely low concentrations of the antibiotic. In using antibiotics as selective agents in media for tubercle bacilli, the same problem is presented as with antibiotic sensitivity determinations on strains of these bacilli: the lack of stability of the antibiotic at 37° C for several days or weeks. Thus no residual selectivity is present as incubation progresses, so inhibited but viable saprophytes may still multiply and contaminate cultures. The sputa of patients receiving streptomycin therapy frequently contain yeast-like fungi and sometimes filamentous forms as well (112). Peni- cillin has a stimulatory effect on these organ- isms, particularly the former. Therefore, this agent sometimes will increase, rather than reduce, the degree of contamination of cultural media. (19) The dyes, crystal violet and malachite green, also do not inhibit these fungi which are also quite resistant to sodium.hydroxide and other chemical digestants commonly employed in the preparation of inocula. Thus these organisms are usually the great- est source of contamination in the isolation of M. tuberculosis from.sputa. Growth of M. tuberculosis Beneath the Surface of LW Virulent tubercle bacilli usually form a pellicle at the surface of liquid media as their multiplication progresses. The tightly knit nature of this pellicle prevents sufficient nutrients and oxygen reaching the majority of the bacilli, and causes them.to be engulfed in the metabolic products of their fellows. Thus, growth of these virulent bacilli is relatively slow, just as it is on the surface of a solid medium. Koch (5h), by first thoroughly breaking up clumps of the inoculum. succeeded in cultivating virulent strains of‘M. tuberculosis beneath the surface of meat infusion broth. The organisms, formed a fine granular deposit at the bottmm of (20) the flask after h to 5 weeks incubation. No turbidity of the medium was observed, indicating that the bacilli were still closely packed to- gether and not dispersed, so that they might better adsorb nutrients. Another reason for the slow growth was the lower oxygen tension in the depths of the medium. Nocard and Roux (76) found 5 per cent glycerol in veal-peptone broth permitted depth growth of tubercle bacilli. The flocculent deposit of organ- isms was readily dispersed by agitation. It was noted that multiplication was more rapid when this was done. Nevertheless, h weeks incubation was required to produce perceptible growth of bacilli from tuberculous material. Besredka and Jupille (6) secured depth cultures of human and bovine strains of tubercle bacilli in a sodium hydroxide clarified liquid egg-broth medium. without glycerol. It is possible that the bacilli bound the egg albumin so that their surfaces were rendered less hydrophobic. Thus the broth wet the coated bacilli, causing them.to sink and multiply in a dispersed state. When the cultures were per- :mitted to remain quiet, the organisms formed large (21) masses which eventually grew into a membrane at the bottom of the flasks. Depth cultures were obtained by Boecker (7) in two types of liquid media, one containing glycerol and rabbit liver and the other milk. The liver medium. if filtered to remove coagula, did not permit sub-surface growth. It may be concluded, therefore, that protein or protein-like constituents were bound to the surface of the bacilli, rendering them.less hydrOphobic. It has been found that p- lactoglobulin of hulk will adsorb fatty acids. It is possible that this substance, and perhaps other milk components, envelOped the tubercle bacilli when they were cultivated in the milk medium. Larson, Cantwell and Hartzell (57) lowered the surface tension of a broth medium.from approxi- mately 59 dynes to 52 dynes and found that depth growth of tubercle bacilli occurred more readily as the surface tension decreased. Larson and Montank (58) confirmed this and postulated that there is less difficulty in wetting the surfaces of the bacilli under these conditions. These workers used castor oil soap which is quite toxic to‘M. tuberculosis. Thus their medium.would not be (22) suitable for primary isolation, or for the growth of minimal inocula. Through inoculation of potato completely immersed in liquid media, Duo (15) observed sub- surface growth of virulent tubercle bacilli. Pleural fluid and spinal fluid were used as inocula. In this instance, the potato surface served not only as a nutrient, but also as a means of holding the multiplying bacilli beneath the surface. Sarnowiec (87) used unheated blood serum.in a small amount of broth and secured depth cultures of graded inocula of tubercle bacilli. The high concentration of serum albumin was probably re- sponsible. Wright (115) and Boissevain (8) added blood plasma to liquid and solid media. They obtained growth beneath the surface of the liquid media and below clots of the plasma. Kirchner (52) also used blood plasma as well as serum.as enrichments for his synthetic medium. He was able to obtain depth growth of tubercle bacilli from.sputa, tuberculous pus, urine, and (25) pleural exudates. Kirchner believed that the enriched medium compared favorably with solid egg media. Proca (85) enriched Sauton's medium with guinea pig or rabbit blood. Tubercle bacilli grew beneath the surface from inocula of tubercu- lous sputa or blood. He stated that with the majority of his positive findings, he observed growth after the hth or 5th day. A study of the action of blood and blood fractions upon acid-fast bacilli (including human tubercle bacilli) was performed by Banks and Evans (AS), who concluded that whole blood and blood albumin enrichments of Long's medium permitted greater recovery of these organisms than a solid egg-glycerol.medium. Boissevain (10) found that unheated blood serum.atimulated and enhanced depth growth of tubercle bacilli in synthetic media from small inocula, 10"6 mg of bacilli producing visible growth. He also reported that horse serum al- bumin stimulated growth strongly, while the globulin fraction inhibited or exerted no effect (21;) on growth. Using 0.1 per cent of the former, he found that growth occurred readily from.inocula as small as 10"7 mg tubercle bacilli. However, when the albumin was autoclaved for 10 minutes at 115° 0, its stimulatory effect was lost. Dubos (25) added a polyoxyalkylene derivative of sorbitan monostearate (tween 60) to Long's medium.which was enriched with 0.1 per cent puri- fied bovine albumin. In this medium he obtained relatively rapid and submerged growth of virulent tubercle bacilli, as well as other mycobacteria. He found that presence of the filter sterilized albumin in media permitted more rapid multipli- cation of small inocula, as well as enhancing growth. If the submerged cultures were transferred to Long's medium. they grew in the characteristic surface pellicle. Subsequently Dubos and Davis (27) described a liquid medium.including a polyoxyethylene ester of sorbitan monooleate (tween 80) and 0.2 per cent bovine serum albumin, electrOphoretic fraction V. Heavy growth occurred in this medium from 8 inocula as small as 10' ml of a suspension of M, tuberculosis, strain H57Rv. Sub-surface, (25) dispersed cultures of tubercle bacilli were ob- tained from sodium.hydroxide treated sputa and from pathological material of animals experimental- ly infected with avian strains. In the medium. without albumin, avian bacilli grew from inocula of 10.1 to 10-7 ml, while the human strain deve10p- 1 ed from inocula of 10" to 10"5 ml. Dubos and Davis (27) found also that the albumin fraction V did not stimulate growth to the same extent as did impure preparations of the protein (2h). However with the virulent strain it permitted initiation of growth from smaller inocula. Shortly thereafter Davis and Dubos (19) showed that fraction V albumin functioned, for the most part, as a protective rather than a nutritive growth factor. It was found to neutral- ize the bacteriostatic effect of several substances, inclucing fatty acids, by adsorbing them. It was established that tween 80 contains as much as 0.6 per cent of unesterified oleic acid, a substance which will inhibit multiplication of small inocula of tubercle bacilli. The purified albumin was found to bind at least 2 per cent of its weight (26) of oleic acid. Thus, in the presence of this protein, growth was initiated from 10.8 mg of tubercle bacilli. These investigators also recommended heating of bovine serum. albumin, and fraction V albumin at 56° C for one-half hour to inactivate lipase present which otherwise hydrolyzes enough tween to liberate oleic acid in excess of the binding capacity of the albumins. Dubos and Middlebrook (28) believed that bovine serum. as well as fraction V albumin, contains a heat stable substance which enhances the growth of tubercle bacilli. Youmans and Youmans (107) found that bovine serum.atimu- lated the rate of growth more than serum.albumin. Subsequently (102) they studied the effect on the growth rate of tubercle bacilli of bovine serum and 12 bovine plasma fractions. Only proanti- thrombin was as stimulatory as the whole serum. Contrary to the findings of Davis and Dubos (21), who found more abundant growth with fraction v albumin, the generation time of the bacilli in the presence of this fraction and with crystalline plasma albumin was the same. The globulin fractions stimulated growth to approximately the same extent (27) as the albumin fraction. Sattler and Youmans (88) found that the growth of the H57Rv strain was slower in a tween- albumin medium than in a medium.anriched with albumin only. YOumans and Youmans (108) reported later that the growth rates of 15 strains of recently isolated tubercle bacilli were also slower in the former medium than in the basal Proskauer and Beck broth alone, or the basal broth enriched with bovine serum or with crystal- line bovine albumin. The strains grsw most rapidly and profusely in the beef serum enriched medium. The mechanism.through which tubercle bacilli are dispersed by tween 80 was explained by Dubos (26). He stated that the long aliphatic chain of this molecule was adsorbed by the lipid portion of the bacillus. The ethylene oxide chain of the ester is exposed to the liquid of the medium.and ionizes. Thus a hydrophilic layer is formed around each bacillus. The organisms then multiply beneath the surface of the medium.and, since they are dis- persed, more readily contact the dissolved nutrients. Since pellicles and granular aggregates are not (28) formed, bacilli are less likely to be engulfed in the metabolic products of their fellows. After further trials using the tween-albumin liquid medium.for primary isolation, Dubos and Middlebrook (28) advised that it not be used for this purpose since it was not sufficiently se- lective and growth of tubercle bacilli was diffi- cult to detect in the presence of contamination. Therefore they used a modification containing less asparagine and no glycerol or glucose, reduced amounts of casein hydrolysate and ferric ammonium. citrate, and lowered pH in order to prevent multi- plication of undesirable bacteria. In their hands this medium.was more selective. Tubercle bacilli grew rapidly from inocula of Sputa and other patho- logical material. The organisms were submerged and dispersed. Mollov, Hill and Oshinsky (72) also used this medium.in routine isolation of'M. tuberculosis. They added 10 units of penicillin per ml to inhibit contaminants. The medium compared favorably in recovery of tubercle bacilli with Petrik solid egg medium. However, it was necessary to streak Petrik's medium.with cultures evidencing growth (29) in order to establish the presence of tubercle bacilli and to demonstrate characteristic colonies of virulent strains. Many other workers also used the liquid medium for primary isolation with varying degrees of success (3h, 35, kl). Dubos and Middlebrook (28) also cultured sputa on an agar medium.af similar composition, excepting that a sodium oleate-bovine plasma fraction V albumin complex was substituted for tween 80 and the albumin fraction. This complex neutralized the toxic effect of fatty acids present in agar. Tubercle bacilli grew from.the pathological inocula nearly as rapidly as they did in the liquid medium, and characteristic colonial morphology indicative of the degree of virulence was observed. Later, Middlebrook recommended incorporation of penicillin in the medium to further reduce contami- nation (68). Roberts, Wallace, and Erlich (85). and Byham (1h), as well as other workers, reported excellent results in isolation of tubercle bacilli on this medium. Problems in the Rapid Isolation of g . tuberculos is Thus the principles of depth cultivation (30) were applied to reduce the time required to recover E. tuberculosis from pathological material. Although the growth time was reduced over that on the classical solid egg media, several problems remain unsolved: l. The most important of these, for which a solution must be found before the Dubos type liquid and solid media can supersede media such as Petragnani's and Jensen's modification of Loewenstein's, is that of contamination of cultures by saprOphytic and come mensal organisms. As discussed previously, penicillin and malachite green, as well as chemical treatment of sputa, do not effectively inhibit or destroy many fungi which will overgrow the medium. 2. A more selective medium would make it unnecessary to employ the present harsh chemical digestants which have been found to be markedly toxic to tubercle bacilli (ha). 3. If a completely selective medium is not develOped, a means of indicating growth of‘g.‘§g' berculosis in the presence of contaminants is needed. A reliable indicator of colonies of tubercle bacilli 'would also save considerable time otherwise spent in preparing and.examining smears from.colonies. (51) h. The Dubos-type liquid and solid media are time-consuming to prepare, since the albumin enrichments must be filter sterilized and added aseptically following autoclaving of the base. The enrichments can be a source of contamination of the entire batch of mediums 5. Tubercle bacilli in depth cultures are growing under conditions of semi-starvation of oxygen. Broth tubes especially do not contain sufficient oxygen dissolved in the medium. In such instances this deficiency becomes the limit- ing factor of the growth rate (hfi, 110). 6. The bacilli are much more susceptible to toxic substances in the liquid media, since they are dispersed. The adsorbancy of albumin or serum is not always sufficient to overcome these detrimental effects. 7. The solid egg media are less satisfacto- ry because their composition is poorly defined and varies from batch to batch. It has been found that the age of eggs used in preparation of these media greatly affects their nutrient qualities for the tubercle bacillus(h8). Some (52) lots of egg media contain lysozyme, which is strongly toxic to these organisms (7h). Egg albumin, like plasma albumin, when denatured by heat sterilization, does not support growth of‘y. tuberculosis as well as when it is unde- natured. Both proteins may actually be toxic to this organism (10). 8. The findings of investigators have been contradictory in regard to which protein enrich- ment is most favorable to rapid growth of tubercle bacilli. In formulating a culture medium.for primary isolation it must be established which of these is most beneficial. 9. Many workers have found it difficult to cultivate E. tuberculosis from sputa of patients receiving streptomycin--para-aminosalicylic acid therapy (59). In some instances, cultures have been negative when the microscOpic examination of concentrate smears revealed typical acid-fast bacilli e 10. In liquid media, colonial morphology cannot be studied, thus presumptive means of identification of E, tuberculosis and ascertain- (55) ing its degree of virulence are unavailable. I Exploratory Studies on Media for Tubercle Bacilli Attempting to solve some of these problems, a study was made of media for the rapid cultivation of tubercle bacilli in order to develop a more satisfactory preparation for primary isolation. In the early stages of the investigation the liquid Dubos-Davis medium.was studied and compared with modified media as well as accepted media. Small inocula of laboratory stock strains of human.tubercle bacilli were cultured before using these media in the cultivation of g. tuberculosis from pathological material. Experimental Changes in the composition of the experimental media were made through addition or removal of a single ingredient. Each resulting modification was compared with the Dubos-Davis medium.(D-D) as to the time required for detection of visible growth and the ultimate extent of growth. The liquid medium, E-lZ, containing constituents of D-D, as well as those of previous modifications, was the (Bit) result of these early studies. 3-12 Qm/L Tmt08°(D1fCO) 00000000000000... 10. Disodium phosphate (N322HP§H°123 O) 2. MonOpotassium.phosphate ( ZPQHI. l Sodium.chloride (NaCl) ........... 1 03891111110 fields (DifCO) 0000000000. 200 V0891 oeeesconce-00000000000000... 1.5 o e momo GlycerOI oeoeeeeeeoeeeoeeeeeeeeeeeeeoeeo 20 Tween 80 (10% SOl'n) 0.0000000000000000. 50 Distilled Water eeeeeeeeoeeeoeeeeeooe-oo 892 Autoclave at 121° C for 15 minutes. Add, aseptically 10 ml of 5% sol'n of bovine serum albumin fraction V, previously sterilized by ultrafiltration. E-12 was found to be an excellent medium.for the prepagation of cultural strains of g. tubercu- losis. The virulence of these strains was not determined. Their growth was rapid and luxuriant; macroscOpically visible submerged aggregates of bacilli were obtained, in some instances, 2h.hours following inoculation with dilute suspensions of tubercle bacilli. The precise weight of the in- oculum.was not determined. However, when strains newly isolated from sputa (suspensions made frmm the Petragnani slant on which the specimen was cultured) were used as (55) WOW E-E-E- inocula, the time required for appearance of visible growth was much greater than had been the case with the cultural strains. Eight to 1h days incubation was required and subsequent incur bation did not produce as generous final yield. Small inocula of the recently isolated tuber- cle bacilli grew slightly faster in D-D. The ultimate growth in this medium.was more profuse than in E-lZ. Larger inocula (approximately 10-1 mg) multiplied at nearly the same rate in both media. MicroscOpic examination of Ziehl-Neelsen stained slide films prepared form shaken cultures, revealed that the recently isolated strains tended to cling together and manifested serpentine patterns of growth, while the more rapidly growing bacilli were oriented in.more random.fashion. The medimm in which the latter strains grew became very turbid, while the newly isolated strains tended to form tightly-knit granules at the bottom of the tube. Upon transferring all strains back to Petragnani's, it was found that the cultural strains produced colo- nies which were smoother and softer than these pro- (56) duced by the newly isolated bacilli. The latter were rough, wrinkled, dry hard and flaky. This colonial morphology indicates that the former were without, or very low in virulence, while the others retained this prOperty (25, 69). The E-lZ medium.was used, in comparison with the Petragnani egg medium.and D-D, to culture chemically treated, concentrated sputa and other pathological material. E-12 cultures of sputa and digested urines were almost universally con- taminated with fungi and undesirable bacteria. These organisms multiplied so much more rapidly than the tubercle bacillus that their presence made macroscOpic detection of the latter impossi- ble. The growth of these contaminants produced metabolic products in the medium which probably hindered the multiplication of‘g. tuberculqgig. Nevertheless, it was possible to detect acid- fast bacilli which had multiplied by preparation of slide films from the culture. Addition of potassium penicillin G, after sterilization of the basal medium, in a final concentration of 100 units per ml to E-12 and to D-D reduced contamination sufficiently to permit (37) a comparative study with the Petragnani medium in the culture of sputa. Isolation of tubercle bacilli from.those specimens was achieved most rapidly in the Dubos medium. The average time required for multi- plication of tubercle bacilli to the extent that they could be detected microscopically was, for each mediums D-D, 12 days; E-12, 18 days; and Petragnani, 26 days. However, more positive cultures with less contamination were obtained on Petragnani than with the other media. The 2 liquid media were frequently nearly or completely overgrown with both bacteria and fungi. The resultant turbidity made it difficult to determine if tubercle bacilli had multiplied. While cultivation of recently isolated strains of M. tuberculosis and modification of media con- tinued, toxicity studies of crystal violet, brilliant green, basic fuchsin, and malachite green were conducted. Recently isolated strains of tubercle bacilli were grown in the presence of varying concentrations of these dyes in liquid media with and without tween 80, and on egg slants containing comparable concentrations. Malachite green was (58) found to be most satisfactory, since it inhi- bited multiplication of the bacilli least in concentrations which effectively inhibited many of the undesirable bacteria present in digested sputa. The most troublesome contaminants while using malachite green in media for the isolation of tubercle bacilli were gram-positive'proteo- lytic bacilli and a variety of fungi. An easily prepared solid experimental medium. which afforded rapid and profuse growth of tubercle bacilli was used tocnlture 153 sputa previously digested with 5 per cent sodium hydroxide. This medium, E-22, contained the following ingredients: E-22 Gm/L Tryptose phosphate broth ......... 2 MonOpotassium.phos hate (KH Poh) . Magnesium sulfate (MgSOhOIHSO) ... Ferric ammonium citrate .......... SEVita eeeeoeeeoeeeeeeeeeeoeeecoco Agar ......OOOOOOOOOOOOOOOO0...... 16.0 GlyceI'Ol ......OOOOOOOOOOOOOCO0.000000... 1000M Malachite green (1% sol'n) .............. 1.25 ml Distilled water eeeeeeeeoeeeeoeeeoeeeeeoe 88800 m1 NOOI—‘VD OHI—‘OU‘I Following sterilization by autoclaving at 121° c, 100 ml of oleic acid-albumin complex (28) was added aseptically. The final concentration of malachite green in this medium is l-h0,000. It was possible to detect the growth of (39) E. tuberculosis on the surface of E-22, using high Gaffky concentrates .of sputa, after 2 days incubation. With low Gaffky concentrates, an average of 5 days incubation was required. Cultures inoculated with negative concentrates showed detectable growth after 12 days. It was possible to detect acid-fast bacilli in shorter periods of time by rubbing the surface of the medium.with a sterile 100p and preparing slide films for microscOpic examination. The colonies of.y. tuberculosis were simi- lar to those described by Dubos (25) as charac- teristic of virulent strains on oleate-albumin agar. As on the Petragnani medium, character- istic pigmentation was present. The extent of contamination was approxi- mately equivalent to that encountered with the Petragnani medium. Most contaminants were readi- ly distinguished from.E, tuberculosis by their colonial morphology. ‘Yeast-like fungi were less prevalent than they were in media containing penicillin. Some of the colonies they produced resembled those of tubercle bacilli. (ho) At this time, a series of modified egg media was also prepared and studied. The results secured cultivating recently isolated strains of g, tuberculosis and culturing pathological materi- al, indicated that EE-18 was the best of these, and was also superior to the Petragnani medium in both number of positive cultures and speed. of multiplication observed. EE-lB Gm/L Potato starCh ......C...‘......... 18.0 A8parag1ne eeoeoeoeeeeoeoeeeeeeeoo 2.0 TI’yptOSQ coco-00000.00.oceeeoeeeeee 5.0 Ferric ammonim Citrate 0000000000 001 Magnesium 3u1fate (MgSOh.7H20) .00 005 Casamino 8.01.6.8 (DifCO) 00.00.0000. 1200 Disodium.phosphate (NagHPO olZH 0) 2.5 MonOpotassium.phosphate ( aPOhs . 1.0 Agar coocoococo-000.000.000.000... 1 .0 Glycerol ooeoeooeeeeeoeeeeeoeeeeeeoeoeeee 10.0 ml Distilled water .....‘..-..-..‘...‘........... 90000 In]. The above portion is sterilized at 121° C for 15 minutes. The following enrichment is then added aseptically: 7 egg yolks BOVine Serum 000000000000oeeoccceeeooooeo (4.0 Sodium 0168.136 coco-e-oeoeeeeeeeoeeoooooee 20 maIQChite green (2% SOl'n) 0000000000.... 5 61110030 eoceeeooeeoeeoeeeoeeeeoeoo 7.0 The ultimate concentration of malachite green is l-10,000. The egg-agar medium EE-lB, was deve10ped (hl) BEE- prior to anouncement of the medium of Peizer and Schecter (78), to which it bears some simu- larity. EE-l8 was not compared with the latter medium, but it was found to support more rapid growth than the Petragnani medium.in primary isolation of M. tuberculosis from.sputa. A higher number of positive cultures was obtained with EE-18. The number of contaminated cultures of each medium.was nearly the same. However, EE-18 is as complex and, conse- quently, as difficult to prepare as Petragnani and the majority of other egg media. Therefore, since one of the primary objectives was the formulation of a good culture medium.which could be easily prepared, the development and study of other new preparations was continued. Another agar medium which compared very favorably with Dubos oleate-albumin agar was A-l-S. In isolation studies of these 2, either penicillin (50 units/ml of medium), or malachite green (1-70,000) was added as selective agent. As in the Dubos-type media, unheated bovine serum.albumin was used as the protective ingredient (he) rather than egg albumin. However, the A-l-S enrichment was more complex than had previously been used in media for tubercle bacilli. A-l-S Gm/L Proteose peptone, No. 3 (Difco) .. . Disodium phosphate (NagHP%~lZH o MonOpotassium.phosphate ( 2P0h§ . Copper sulfate (CuSOh.5H20) oeooeo Ferric ammonium citrate .......... Magnesium.sulfate (MgSOu-7HZD) ... SQVita oeoooeeoeoeeeeeeeeeoeeecoo. o CGSitone (DifOO) cease-0000000.... 00 Agar eoeeeoceeoeeeeeoeeocoeeeeeeee 16.0 G1y09r01 one...coeoeeeeeeeeeeoeeeoeeeeoeo 1000 ml Distilled water oeeeeeoeoeeeeeoeeoeeeeeso 887.0 ml 001 KNUCDCDCH¥FDR) CKflFJCHDUVO This portion is sterilized at 121° C for 15 minutes. Then the following enrichment is added aseptically: 0191C 801d ooeeeceeeeoeeoeeeeeeeoeeoeooeo 0012 N/ZO SOdium.hYdrOXid9 0.0000000000000000. 10.0 5.0 ml of this solution is added to: Bovine serum albumin (fraction V) 2.5 Glucos. coeoeeeeeeeeeeeeeeeeoeeeee 705 B0v1ne serum ............................ Saline (0.85%) once-000.000.000.000...coo uufi CH: 0. CH: 5 HE In A-l-S, proteose peptone No. 5 (Difco) was substituted for tryptose, which was present in all experimental media previously discussed. Work with other liquid and solid media had in- dicated that slightly better growth was obtained (L6) when this substitution was made. Multiplication of minute inocula of tubercle bacilli was more rapid than in the other experimental media, D-D, Petragnani and the Dubos-Middlebrook liquid and solid media. Glucose was used as an additional carbon source, since it was found that final growth was more profuse when it was furnished along with glycerol. If this sugar was autoclaved as part of the basal medium, slower multipli- cation of tubercle bacilli in the medium was observed. Dubos and Davis (27) found that glucose sterilized by heat in the presence of phosphates formed substances toxic to E. tuberculosis. These products were probably responsible for the effect noted in experimental media. The synergistic effect of glycerol and glucose in the prOper proportions, as observed by Frouin and Guillaumie (57), was discussed in the section on carbon nutrition of the tubercle bacillus. Casitone (Difco) was found more favorable than the other Difco casein products, casamino acids and isoelectric casein. This confirms the report of Dubos and Middlebrook (28) that (hit) casitone and N-Z-amine-B were the best commercial products for use in media for tubercle bacilli as amino acid sources. Proteose peptone No. 3 supplies calcium, manganese, and chlorine in sufficient quantities to meet the requirements of the bacilli (2, 28). The addition of trace amounts of zinc ion to media containing this peptone was found unneces- sary. More luxuriant growth of §.-tuberculosis was secured when 0.1 mg of cOpper sulfate was added to agar media. Its effect was not as readi- ly discernible when present in liquid media, but it appeared to produce slight stimulation. Pratt (82) has reported finding trace amounts of capper to be toxic to E. 21111;. Calcium, manganese, and citrate, as well as d-amdno acids neutralized the toxicity. The detoxified copper was not required for rapid growth of the organism. Whether capper is toxic to M. tuberculosis in the absence of these substances is not known, but in their presence, as in A-l-S, it has a favorable effect’on growth. The enrichment added to A-l-S provides the (15) protective preperties of fraction V albumin as well as the greater growth stimulation of bovine serum. As the former is utilized in the fomm of an oleate-albumin complex, the nutritional qualities of detoxified oleic acid are also secured. In a study of enrichments, by means of turbido- metric and micro-Kjeldahl growth measurements, it was established that this protein mixture provided maximum stimulation of the multiplication of the tubercle bacillus. The other enrichments used in the study, in descending order of the growth stimulation observed, were: bovine serum in 5 and 10 per cent concentrations with 7.5 per cent glucose (identical results), Dubos' oleate-albumin complex with 7.5 per cent glucose, 0.5 per cent fraction V albumdn.with 7.5 per cent glucose and Dubos' oleate-albumin complex (approximately identi- cal results), 0.5 per cent fraction V albumin, sodium.oleate-human serum.with 7.5 per cent glucose, 10 per cent human serum, 5 per cent human serwm, sodium oleate-skim.milk with 7.5 per cent glucose, and sodium.oleate-isoelectric casein with 7.5 per cent glucose. (hé) These results confirm the findings of Youmans and Youmans (102) that bovine serum favors growth of E. tuberculosis more than bovine serum fraction V albumin. The Dubos oleate-albumin complex with glucose was nearly as stimulatory as the bovine sera concentrations containing this sugar. All enrichments were filter sterilized excepting the skim.milk and casein preparations. These were autoclaved and the glucose added aseptically after they had cooled. The basal medium to which the enrichments were added was the autoclaved portion of A-l-S without agar. Bordet-Gengou blood agar has been used by several workers for isolation of tubercle bacilli from.sputa and other inocula (70). Most of these investigators found it to be superior to the classical egg media. Tarshis and his associates (96, 97, 98) modified the medium extensively, adding human bank blood, and penicillin, and reported more rapid growth with a higher number of positive cultures when his modification was compared with the Loewenstein-Jensen medium.(99). Along with the experimental egg media, the (1m Bordet-Gengou medium and subsequent modifications were used in the cultivation of recently isolated cultural strains as well as in primary isolation of tubercle bacilli from chemically treated concen- trates of sputa. The results of these studies indicated that the blood media were not as favor- able for primary isolation as the Petragnani medium, EE-18, A-l-S, E-22, and the Dubos-Middlebrook agar medium. The Tarshis modification was not studied. The best blood medium, B-12 was more complex than those developed by other workers, but was easily and inexpensively prepared. B-12 Gm/loo ml Casamino acids (Difco) ............ 0.2 Vegex eeeeeeeeeeeeeeeeeeeeeeeeeeeee 0035 Magnesium sulfate (MgSOhy7HgO) .... 0.2 Agar eeeeeeeeeeeeeeeeeeeeeeeeeeeeee 1. GlycerOI eeoeeeeeooeeeeeeeeeeeeeeeeeeeeeee 0.5 m1 Distilled 'ater coco.oeoeeeeeeeooeeeeoeooe 76.0 ml This basal portion is sterilized at 121° C. Then 20 ml human blood collected in h ml of 5.8% potassium.oitrate is added aseptically. Peni- cillin is added in a final concentration of 50 units /m1 . Growth of E. tuberculosis on 3-12 was slight- ly slower than that on Petragnani's. The number of positive cultures was approximately the same (h8) for each medium. Colonies of tubercle bacilli on B-12 were similar in morphology and pigmen- tation to those deve10ping on the egg media. Evidence regarding the influence of whole blood and hemoglobin on‘y. tuberculosis is contro- versial. 'Yoshida (101) hydrolyzed human blood with 50 per cent hydrochloric acid for 2h hours and obtained a product which stimulated the de- velopment of both E. tuberculosis and M. 251;; in extremely small concentrations. 0n the other hand, Davies (18) reported that both blood and hemoglobin inhibited the growth of tubercle ba- cilli at the surface of liquid media. Depth cultures were enhanced by the presence of either of these substances. Pagel (77) found that the majority of blood samples from individuals with tuberculosis adversely affected growth of tubercle bacilli in Kirchner's liquid medium while most samples from.normal persons aided depth growth in this medium. Hanks and Evans (16) and Boissevain (10) found blood from.various animals stimulatory to sub-surface growth of tubercle bacilli in liquid.media. Frisch (56) observed that nearly all human blood fractions which he separated (1+9) were stimulatory to multiplication of the H57Rv strain. Whole blood probably possesses at least as much lipid binding power as equal amounts of plasma and serum.and in addition may supply additional nutrients and minerals. However, the presence of lipase (21), non-demonstrated anti- bodies, and the ill-defined ”bactericidal power of blood" may all detrimentally affect growth of . tubercle bacilli. The work of Emmart and Seibert (52) indicates that tuberculous sera and gamma globulin from sera sensitized with tuberculin are bacteriostatic to tubercle bacilli. Dubos and Davis (20) reported bovine gamma globulin was toxic to the bacilli, but'Ycumans and Youmans (102) found it to be growth stimulatory. Further study to clarify the effects of blood and its fractions will aid formulation of uniform.media using them.as enrichments. A series of experimental liquid egg yolk media was also developed and studied. The best of these, EY-S, was found to support growth of small inocula of tubercle bacilli. Multipli- cation was nearly as rapid as in A-l-L (A-l-S (50) without agar) and in the Dubos-Middlebrook liquid medium (28). Either penicillin or malachite green was used as selective agent. The Opaque quality of EY-5, however, made it difficult to distinguish and differentiate growth of E. tuberculosis from that of undesirable organisms. The composition of EY-5 was identical to that of A-l-L excepting that 10 egg yolks were added to 900 ml of basal medium.instead of the oleate-albumin serum enrich- ment. Glucose was present in EY-S. Eggerth (51) secured similar results with various concentrations of egg yolk added to the Dubos-Middlebrook liquid m0 dium. Wasiliew and Johnson (111) secured more rapid growth of E. tuberculosis from concentrates of sputa cultured on the Petragnani medium.with 2 ml of "tween-albumin” (concentration not speci- fied) added to the slanted surface of the medium. than they did on the standard Petragnani medium. Prior to publication of their study, during the development of experimental egg medium EE-l8, chemically treated concentrates were cultured ' on Petragnani's with 1 m1 of 1 of 5 sterile mixtures added to the surfaces of different lets (51) of cultures. These mixtures were: (1) a 5 per cent solution of fraction V bovine albumin con- taining 0.05 per cent tween 80, (2) Dubos-Middle- brook oleate-albumin complex, (5) a 5 per cent solution of fraction V bovine albumin contain- ing 0.05 per cent triton A-20. When small in- ocula of recently isolated strains of tubercle bacilli were cultured on the moistened Petragnani's macroscOpically visible growth was secured 2 days earlier than on standard Petragnani slants. No difference in effect on growth was noted between the 5 mixtures. However, when the mixtures were used with Petragnani's for primary isolation of g. tuberculosis fromusodium hydroxide treated concentrates of sputa, so much overgrowth of the slant's surface by contaminants was experienced that the technique was discarded. Another solid medium which supported rapid and luxuriant growth of tubercle bacilli from sputa was A-22-S, the first bovine albumin en- riched agar medium in the experimental series 'not containing a commercial peptone. Savita was the augmented form of yeast extract substi- tuted for vegex. It had been found that this (52) substitution in a medium resulted in slightly shorter generation times of tubercle bacilli. It was necessary to add calcium and zinc ions in trace amounts when the peptone was omitted. A-22-S i Asparagine eeeeeeeeoeeeeeeeeeeeeee 3-0 Disodium phosphate (NagHPO ~12H 0) 2.5 MonOpotassium.phosphate ( 2P0h3 . 1.0 Ferric ammonium citrate .......... 0.1 saV1ta eeoeeeeeeeeeeeeeeeeeeeeeeee 2.0 Magnesium.sulfate (MgSOh-7H20) ... 0.7 Calcium Chloride eeeeeeeeeeeeeeeee 0.0005 Zinc sulfate (ZnSO 07H 0) ........ 0.0001 Copper sulfate (Cu 0h} H20) ...... 0.0001 CEBitODO (DifOO) eeeeeeeeeeeeeeeee 200 Agar ococoeee‘eeeeeeeeeeeeeeeeeeeee 1600 GlyceI'OJ. eeeeeeeeeeeeeeeeoeooeeeeeeeoeeee 2.0 m1 Distilled water eeeeeeeeeeeeeeeeeeeeeeeeo 8911.0 131 After sterilization of the above at 121° 0, the following enrichment was added aseptically: 01910 801d eeeeeeeeeeeeeeoeeeeeeeeeeeoeee 0012 m1 N/20 SOdilm hydrazide eeeeeeeeeeeeeeeeeee 1000 ml 5.0 m1 of this solution is added to: Bovine serum albumin (fraction V) 2.5 Glucose eeoeeeeeeeeeeeeeeeeeeeeeoe 705 BOVine 56mm oeeeeeeeeeeeeeeeeeooeeeeeeee 20.01111 Saline (0.85%) coeeeeeeeeeeeeeeeeoeeecoee 5.0m]. Bacto-penase (DifCO) oeeeeeeeeeeeeeeeeeee 100 m This enrichment is essentially the same as that used with medium.A-l-S, excepting that Bacto-penase, a commercial preparation of a (55) penicillinase 1 ml of which will inactivate up to 1000 units of streptomycin (5) was included. When this preparation was used in culture media, it was possible to isolate tubercle bacilli from.sputa of patients receiving streptomycin-- P.A.S. therapy. Parallel cultures on media not containing Bacto-penase, including Petragnani's, were negative in many instances. It is believed that streptomycin, in low concentrations, is present in sputa of patients who are being treat- ed with this antibiotic. It has been shown that para-amino salicylic acid is present in the sputa of individuals receiving this drug (59). A-22-S was used in the routine cultivation of E. tuberculosis from pathological body fluids, comparing it with Petragnani's, A-l-S, and the Dubos-Middlebrook solid medium. Growth of this organism was most profuse on A-22-S, but a great- er number of positive cultures was obtained on A-l-S. Fewer positive cultures were secured with Petragnani's than with the Dubos-Middlebrook agar. Most rapid growth was also obtained on A-l-S, with A-22-S next in this respect, followed by the Dubos-Middlebrook medium. (5h) As in earlier studies, 50 units of penicillin per ml of A-l-S were used to reduce contamination. Malachite green was not employed as selective agent as it was previously in A-l-S. However, due to the presence of Bacto-penase, penicillin could not be used in A-22-S. Therefore, malachite green, in a final concentration of 1 to h03000 was the selective agent employed. The apparently greater toxicity of the dye to E. tuberculosis might have influenced the results of the studies. A-20-L, a liquid preparation whose compo-. sition was the same as A-22-S excepting for the omission of agar and inclusion of 5 m1 of a 10 per cent solution of triton A-20, was also used in primary isolation studies. Malachite green in 1 to 70,000 concentration was the selective agent. Tubercle bacilli grew more rapidly from inocula of positive and negative concentrate sputa in A-20-L than on A-22-S, but the problem of fungus and bacterial contamination reduced its effectiveness in diagnostic work. It was difficult to identify growth of tubercle bacilli in the presence of these contaminants. (55) Two sets of several different liquid media were studied, 1 set containing tween 80 and the other triton A-20. In each instance, more rapid and profuse growth of tubercle bacilli from.small inocula was obtained in the medium.containing the latter. In the concentration used (0.05%), triton A-20 is not toxic to p. tuberculosis (19). In water solution it will disperse clumps of A tubercle bacilli, but does not serve as a nutri- ent as does tween 80, which liberates small quanti- ties of non-toxic lipids (26). II Studies of the Toxicity of Ethyl Violet and Malachite Green to E. tuberculosgjg Litsky, Mallmann, and Fifield (60) found that ethyl violet 6B, in prOper concentrations, will inhibit gramrpositive bacteria, but exhi- bits less toxicity for gramrnegative bacteria than crystal violet and brilliant green in compa- rable concentrations. It was previously mention- ed that crystal violet, to which ethyl violet bears a close structural resemblance, was used as.a selective agent in early egg media for the isolation of‘y.'tuberculosis, and that, in recent years it has been largely superseded by malachite (56) green which is less toxic for tubercle bacilli in concentrations which inhibit many bacteria present in sputa. Malachite green, in routine culture of sputa, had failed to inhibit multiplication of staphylo- cocci, large bacilli, and fungi. Ethyl violet was therefore used to determine if it would mani- fest less toxicity for‘M. tuberculosis than mala- green and to determine its relative inhibitory effect for undesirable bacteria by using it as a selective agent in media for the primary iso- lation of this organism.from.sputa. Materials and Methods The relative toxicity of malachite green and ethyl violet was compared in liquid, agar base, and egg media. 6 2,.10'8, 10‘ , and Graded inocula of 10- 10"8 mg (wet weight) of‘M. tuberculosis var. hominis, prepared according to the method of Ybumans and Karlson (106), were used in all studies. The homogeneous suspension of tubercle bacilli from which these concentrations were prepared was an approximately equivalent mixture (57) of 5 strains of recently isolated, virulent strains. All were recovered from.sputa on Petragnani medium, and had been transferred into the basic broth enriched with albumin in order to secure actively growing cells in cultures 12 to in days old. The basal medium had the following composition: Gm/L Proteose peptone No. 5 (Difco) .... Disodium.phosphate (NagHPO .12H 0) Monopotassium.phosphate ( 2P0h§ .. Ferric ammonium.citrate ........... . Capper Enlfate (CuSOu'SHZO) eeeeeee 00001 Cobaltous sulfate (CoSOh-7H20) .... 0.0001 88.71153 eeeeeeeeeeeeeeeeeeeeeeeeeeee .0 Triton A-ZO (10% SOl'n) eeeeeeeeeeeeeeeee 5.0m GlycerO]. eeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee 205 ml Distilled water ......C.......O.......... 890.5 m1 OOHNN HOU‘IO ...: Following sterilization at 121° 0 for 15 minutes, 100 m1 of a sterile 5 per cent solution of bovine fraction V albumin was added asepti- cally. The medium was then dispensed in sterile aluminumecapped tubes in 5 m1 quantities. Dye concentrations were achieved through addition of specific amounts of sterile solutions to individual tubes of medium before the heat sterilization. The influence of ethyl violet and malachite green in this medium on the growth of M. tubercu- losis was evaluated through determination of the (58) Optical density of the broth medium. The follow- ing densities were read: (1) That of the unin- oculated medium.with an uninoculated tube of the same dye concentration set at 0. (2) That of the inoculated tube after 16 days incubation, with the same blank set at 0. The difference between densities (2) and (1) was that recorded for the specific dye concentration. Readings were make using a Cenco "Photelometer", Model h1017A. Each tube was shaken vigorously to effect complete suspension of the bacteria before it was inserted in the photometer. In studying the effect of the dyes on tuber- cle bacilli growing on agar and egg solid media, the method of Youmans and Ybumans (107) was used to determine the generation times of the organ- isms. The time at which growth first became macroscOpically visible was recorded in the course of daily examination of the surface of the slanted media. Duplicate generation time determinations were made for each dye concen- tration in each medium. The agar medium.used was identical to the broth, with the exception (59) of addition of 15.6 gm of agar per liter of medium. and omission of triton A-20. The media containing certain dye concen- trations, were also used to culture tubercle bacilli from.sputum.concentrates. These were treated with 5 per cent sodium hydroxide for one-half hour, then neutralized to phenol red with 5 per cent hydrochloric acid. Results The results of a series of studies of growth of tubercle bacilli in ethyl violet and malachite green liquid media are listed in table 1. After preliminary studies were made over a wide range of concentrations of each dye, a narrower range of dilutions within the region of greatest increase of density was inoculated. The concentrations at which a plateau of densities began, were taken to be the strongest dilutions having the minimum toxicity for'M. tuberculosis. For ethyl violet this concentration was 1 to 800,000 and for mala- chite green, 1 to l,h00,000. In table 2 is shown the effect of malachite (60) green and ethyl violet on the rate of growth of M. tuberculosis when the organisms were grown on bovine fraction V albumin enriched agar, and the solid egg medium, EE-18. The table gives the generation time, in hours, of the tubercle bacilli on each medium. The dye concentrations used were those which previous observation had indicated were within a range permitting growth of tubercle bacilli. The dyes were less inhi- bitory in the egg medium, so higher concentrations were used in the study. Generation times for the media without dye were determined as controls. In any study of selective bacterial inhi- bitors, the objective is to find the Optimum selective concentration of each dye. This may be defined as that concentration, in the presence Of which, growth of the desired organism.is near- est that secured in the absence of the selective agent, while the degree of inhibition Of undesira- ble organisms is sufficient to make use of the agent worthwhile. Its determination thus requires incorporation of certain dye concentrations in media for isolation Of the desired bacterial species and utilization Of these media for that (61) purpose. Generation times on the egg and agar media in the dye concentrations studied did not reach the growth rates Of the controls. If lower concen- trations had been used it is possible that these rates would have been attained. However, use Of these media in primary isolation of‘M. tubercu- Ilggig indicated that the lower dye concentrations are ineffective in inhibiting bacterial contami- nants. For malachite green the Optimum selective concentration for the egg medium.was l to 70,000, and l to 600,000 for the fraction V albumin en- riched agar. The primary isolation studies indicated the following Optimum selective concentrations Of ethyl violet: fraction V albumin.enriched agar, l to 600,000, and solid egg, 1 to 80,000. Use Of the liquid media for isolation of M. tuberculosis from chemically treated concen- trates of sputa revealed the Optimum selective concentration for ethyl violet to be 1 to 800, 000, while with malachite green it was 1 to 1,100,000. (62) Discussion The technique of measuring the Optical density of the medium in which tubercle bacilli had been growing for an arbitrary length of time was develOped as a means of most accurately measuring dye effects in liquid media. As far as is known, it has not been described previous- ly. Monod (75), after studying Optical densities of suspensions of microorganisms during growth, concluded that the measurements correspond to the weight of the total bacteria rather than to their number. Thus growth curves based on 'Optical density alone are not as accurate as plate counts. However, with recently isolated virulent tubercle bacilli, plate counts are not feasible due to the tendency of the bacilli to form.aggregates. Concentrations of tween 80 and triton A-20 which prevent formation of, or disperse small clumps, are toxic to M. tuberous lggig. Therefore, growth curves could not be constructed. The size of the inoculum is very important in working with tubercle bacilli, whether the nutrient qualities of a medium or the toxicity (65) of a substance is determined. Larger inocula, because of the protective qualities of the greater amount of tuberculo-lipid, will initiate growth under conditions much more adverse than will smaller inocula. Growth curves secured as the result of multiplication of 10-2 mg Of bacilli are quite different from.those obtained from. 6 inocula of 10‘ or 10-8. For valid represen- tation of dye effect on varying numbers Of tuber- cle bacilli, as is encountered in culturing patho- logical material, one must measure the degree of inhibition of a range of inocula. The Optical densities Of the media provided a very sensitive (means Of securing an expression Of the weight of growth after 16 days from graded inocula in tubes of medium containing varying dye concen- trations. This Optical density technique may also be used to make determinations of growth rates in liquid media by the method of Youmans and YOumans (107) more sensitive. A density to be attained by each inoculum.is selected and the number Of days required is recorded. This number is substituted for the time in days required for (6h) growth to become macroscOpically visible, as specified in the method. The sixteen day interval of incubation in the present study was selected from a knowledge of the approximate density obtained in the medium without dye during this time. If a longer period had been used, the results would have been less precise, since Optical densities become less valid as greater numbers of bacilli accumulate, due to inability to suspend them all by agi- tation. In using the method, one must select a filter intercepting the prOper wave lengths Of light, in order to minimize light interception by the dye concentration in the medium. High concentrations cannot be studied, since trans- missiOn of the uninoculated medium.cannot be set at 100 per cent (0 optical density). Ethyl violet was found to be less toxic tO‘E. tuberculosis than malachite green in both liquid and solid media. Both dyes were much more toxic in liquid media than in solid, since contact Of the individual organisms with them (65) occurred more readily. In the liquid medium the Optimum selective concentration of ethyl violet was 1 to 800,000 while with malachite green it was 1 to 1,100,000. However, the latter dye at this concentration was less ef- fective in inhibiting contaminants than ethyl violet at 1 to 800,000. Thus the superiority of ethyl violet in liquid is greater than these values indicate. Youmans and Youmans have used their method for determination of growth rates to study the effect of both nutrients and toxic agents on the tubercle bacillus (102, 105, 107). It proba- bly has not been used previously to ascertain the effect of dyes on the growth of this organ- ism. It lends itself well to the problem, how- ever, since the growth times of inocula of vary- ing size are required in computing generation times. When the growth rate is determined in the presence of toxic agents, plotting of the growth times is necessary to ascertain the validi- ty of the method. In the present study it was found that the smaller inocula develOped at nearly the same rate as the larger, thus indi- (66) cating that generation times computed from.data secured were valid. The dyes manifested marked differences in effect at the same concentrations in the solid media. The differences are probably the result of binding of the dyes in inactive forms. Although growth was more rapid on the al- bumin agar, the egg medium.aupported growth in the presence Of much higher concentrations of the dyes. This does not indicate that egg al- bumdn exerts a greater protective effect for tubercle bacilli, however, since the egg medium contained a greater proportion of albumin than the bovine albumin agar. Therefore, a greater amount of dye was bound in the inactive state. On the fraction V albumin agar, the short- est generation time, 21.7 hours, was attained with the l to 600,000 concentration of ethyl violet, and remained constant for higher di- lutions. For malachite green in this medium the shortest generation time was also 21.7 hours, but the concentration of the dye permitting this rate of growth was 1 to 900,000. The optimum (67) selective concentrations of ethyl violet and malachite green still further indicated the superiority of the former as a selective agent, as they both were 1 to 600,000. Since the short- est generation time of tubercle bacilli was secured on the albumin medium at this concentration of ethyl violet, little effect is exerted on growth of this organism. However, at a malachite green concentration of l to 600,000, the gener- ation time Of the bacilli was'28.8 hours, thus indicating that at the Optimum selective concen- tration growth is considerably slowed. Ethyl violet was also more efficient in the egg medium, since the shortest generation time, 56.2 hours was attained at the l to 90, 000 concentration and the Optimum selective a concentration was found to be 1 to 80,000, while with malachite green these concentrations were 1 to 90,000 and 1 to 70,000 respectively. With 1 to 90,000 malachite green the generation time was hl.0 hours and with the 1 to 70,000 concen- tration it rose to h6.8 hours. On the other hand, at the Optimum selective concentration of ethyl violet, the generation time was 57.2 hours. (68) Thus, in both the liquid and solid media, studied by means Of Optical densities resulting from growth of tubercle bacilli, through gener- ation times Of the bacilli, and through use in primary isolation of media containing dye concen- trations permitting most rapid growth, ethyl violet was found superior to malachite green as a selective agent in isolation of M. tubercu- ‘lggig. In the concentrations present in media used to culture sputum concentrates, neither dye appeared to reduce fungus contamination appreci- ably. III Primary Isolation of M. tuberculosis in Ethyl Violet Liquid Media Several experimental media, containing ethyl violet as the selective agent, were used to culti- vate recently isolated strains Of tubercle bacilli and in routine diagnostic culture Of sputa. It was found that less contamination occurred when sodium azide in a final concentration of 0.05 per cent was included in the medium, while the rate Of growth was not appreciably affected in non-heated albumin enriched media. In non-enriched media without ethyl violet, this concentration was (69) prohibitively toxic to'M. tuberculosis while in non-enriched ethyl violet media moderate toxicity was manifested. A comparative study was undertaken to come pare ethyl violet-sodium azide experimental media with some classical egg media in the isolation Of‘M. tuberculosis from pathological material. Materials and Methods The following experimental media were used: A-EVA-l Gm/L Proteose peptone NO. 5 (Difco) .... Asparagine ......eoeeeeeeeeeeeeeeee Disodium phosphate (NagHPO '12H 0) MonOpotassium.phos hate ( ZPOh .. Magnesium.sulfate (MgSOhO7H20) .... Ferric ammonium.citrate ......eeeee Copper sulfate (CuSO ~5H20) ....... Cobaltous sulfate (C SQh-7H20) .... SQVita ......eee......eeeeeeeeeeeee SOdium.321dO ......e..............o Casitone (DifOO) ......e........... 2.0 Triton A-ZO (10% 301'n) ......eeeeeeeeee. E. 1 5 . FKfiCMJCDCD .... CDCHD C) C) F: CHUCDCDCHDFJh3FHV . \N Glycerol .....e.e.......eeeeeeeeeeeeeeeee Ethyl ViOlet (0.1% Balln) ......eeeeeeeee Distilled 'ater ......eeeeee......eeeeeee 88 . After sterilization Of the above at 121° C, the following enrichment was added aseptically: 01910 801d ......e....................... 0.12 ml N/2O sodium.hydroxide ......o............ 10.0 ml 5.0 ml of this sol'n is added to: (70) Gm/L Bovine serum albumin (fraction V) . 2.5 Glucose ........................... 7.5 BOVinO serum............................. 0.0 m1 Saline (0.85%) ......e................... E This enrichment is sterilized by filtration. It is almost identical to that used in A-l-S. A-EVA-5 Gm/L Proteose peptone No. 5 (Difco) .... Disodium.phosphate (NaZHPO .123 0) MonOpotassium.phosphate ( 2P0h3 .. Magnesium.sulfate (MgSOhf7H20) .... . Ferric ammonium.01trate ......e.... . SEVitfi ......e..................... 2.0 SOdium.321de .......e.............. 0.3 Triton A‘ZO ...........e................. 5.0 ml Ethyl ViOlet (0.1% BOl'n) ............... 1.25 ml DiStilled water .....................e... 891.0 m1 CDCHJRHU EJNJONflCJ After sterilization at 121° 0, 100 m1 Of the A-EVA-l enrichment was added aseptically. 2 T Gm/L Monopotassium.phosphate (KHZPOh) . 1.0 Tryptose (DifOO) ......e.......... 20.0 Disodium.phosphate (Na HPO -12H20) Magnesium sulfate (Mgsghl7820) ... SQVita ........................... CEBitOne (DifOO) ......o.......... Ferric ammonium citrate .......... . SOdium.az1de ........e............ 0.3 Ethyl violet (0.1% sol'n) ............... 1.25 ml Distillad tater ......................... 897.0 m1 CHURJCHU EJCDCNflkfl After sterilization at 1210 C, the A-EVA-l enrichment was added aseptically. (71) .— 2 A Medium.2 A was identical in compositiOn to 2 T, excepting for the addition of h.ml Of glycerol. A-71 MonOpotassium.phosphate (KHgPO ) . 1 0 Disodium.phosphate (NaaHPOhol 20) 6 3 Asparaglne ....................... 1.0 Tryptose (DifOO) ......e.......... 3 0 Proteose peptone NO. 5 (Difco) ... 1 0 Heat in 100 m1 of distilled water to dissolve. Add the following dissolved in 850 ml dis- tilled water: 7 Gm Casitone (Difco) ................. 0.5 Magnesium.sulfate (MgSOh-7H20) ... 0.05 SOdium 121d° ......e.............. 0.5 Ferric ammonium.citrate .......... 0.05 COpper sulfate (CuSO ~5H20) ...... 0.0001 Cobaltous sulfate (0 so .7H20) ... 0.0001 Zinc sulfate (anOh.7H ) ........ 0.0001 Triton A'ZO (10% 801'n§ .................. 10. Ethyl ViOth (0.1% Bal'n) ......e......... 1 The pH should be adjusted to 6.8. Sterilize at 121° 0. Add 100 ml of the A-EVA-l enrichment aseptically. * The egg media of Loewenstein-Jensen, Petrik, Petragnani, and Peizer were also used, as well as the D~D medium with 50 units Of penicillin per ml. (72) A total of h08 sputa was used in the study. These specimens were classified into 2 groups: (1) those in which it was believed tubercle bacilli were present (82 specimens) and (2) 5 day pool sputa of a miscellaneous character. Approximately h.ml was separated from each specimen and 1 ml of 10 per cent triton A-20 was added. The portion was shaken well, then incubated at room.temperature for 2h hours with occasional shaking. It was concentrated by cen- trifugation at 2,000 rpm for 20 minutes (calcu- lated force times gravity of centrifuge: 59l.h). Approximately equal quantities of the sediment were used to inoculate 1 tube of each of the following media: 2T, 2A, A-EVA-l, A-EVA-5, and A-7l. The remainder of the specimen was processed as follows: An equal volume Of 5 per cent sodium hydroxide was added, and it was incubated at 57° C for 50 minutes with frequent shaking. The alkalinity was neutralized with 1 N hydrochloric acid to phenol red. The specimen was then cen- trifuged at 2,000 rpm.for 20 minutes. Approxi- mately equal quantities of the sediment were (73) . /_. used to inoculate 1 tube of each of the egg media and the D-D medium. A portion of the remaining sediment was used to prepare 2 slide films for microscOpic examination. The number of acid-fast bacilli detected in the stained smears was evaluated in accordance with the Gaffky scale, as modified by L. Brown (55). At least 20 minutes were spent exami- ning the 2 slides. All cultures were incubated at 57°--57.5° C. All experimental media were examined macros sOOpically at h8 hour intervals until a 50 day incubation period had elapsed. If visible growth was detected, a smear was prepared from.the tube for microscOpic examination. All egg media were examined macroscopically at 5 day intervals until they had been incubated for 60 days. Growth was again identified microscOpically. All cultures were discarded after the speci- fied incubation period. The majority of the posi- tive cultures was confirmed as strains of M, £2: berculosis by guinea pig inoculation. It was necessary to transfer positive liquid media cultures (7h) to an experimental agar medium before this could be done. Results Eighty-five per cent of the first group of specimens was positive microscOpically, and 25.8 per cent was positive in the second group. At the conclusion Of the 50 day incubation period the following percentages of the media inoculated with the first group were positive: A-EVA-l, 100; A-EVA-5, 100; A~7l 100; ~D-D, 80.5; Petragnani, 75.6; Peizer, h5; Petrik, 553 Loewenstein- Jensen, 50.5; 2A, 95; and 2T, 80.5. At the end of the 60 day period, the following percentages of the egg media cultures were positive: Petrag- nani, 89; Peizer, 58.5: Petrik, h5.l; and Loewenstein-Jensen, 57.8. After 50 days the following percentages of the media inoculated with the second group were positive: A-EVA-l, 1.8.2; A-EVA-5, Mia; A-7l, h6.0; D-D, 29.6; 2A, h6.h; 2T, hh.l; Petragnani, 5.8; Peizer, h.6; Petrik, 5.5; and Loewenstein-Jensen, h.0. At the end Of 60 days the percentages of positives were: Petragnani, 2h.8; Peizer, 17.h; Petrik, (75) l7.h; and Loewenstein-Jensen, 15.7. It may be noted from.table 5 that most rapid growth Of M. tuberculosis as well as the highest number of positive cultures (259) was secured with A-EVA-l. Growth was slower in A-EVA-5, A-71, and 2A, but the number of positive cultures was nearly as high. A smaller number Of positive cultures was Obtained on the egg media, and multi- plication of tubercle bacilli was slower than that observed in the liquid media. 0f the egg media, the Petragnani medium supported most rapid growth, as well as the highest number of isolations Of the bacilli. Sixty-seven of the patients from.whom the specimens comprising the first group had been obtained were receiving streptomycin-~para-amino salicylic acid therapy. Egg media cultures of only h low Gaffky specimens from.patients re- ceiving such treatment were positive. The other positives were from.eultures of high Gaffky sedi- ments. All A-EVA-l, A-EVA-5, and A-7l cultures of this group were positive In the second group, 218 of the patients (76) a /—‘ were being treated with streptomycin--para-amino salicylic acid. Only 55 of the specimens from patients receiving this therapy were positive on any of the classical egg media. All of these had been inoculated from sediments having a Gaffky greater than III. In A-EVA-l, 157 Of the specimens from.those patients were positive. D—D was the only non-experimental medium in which the growth rate of M. tuberculosis approached that observed in the experimental media. However, frequent contamination of D-D made it difficult to detect and remove samples of early multiplication of this organism. In culturing both groups of specimens, A-7l was contaminated most frequently, while the Petragnani medium had the least contamination. Undesirable organisms multiplied in or on all of the media. The liquid experimental, as well as D-D, were especially heavily contaminated. However, when tubercle bacilli grow in the pres- ence of ethyl violet, the aggregates of bacilli are colored a deep blue-violet color. Formation of these colored clumps aided in detection and removal of samples of growth. (77) The most frequent contaminants of the ethyl violet media were streptococci and fungi. Tubercle bacilli continued to multiply in the presence of these organisms, however. Other bacterial growth appeared to step after 2 to 5 days, but there was no indication of inhi- bition of streptococci and fungi. Discussion Many laboratories make a practice of incu- bating all egg media cultures for periods rang- ing from 6 to 12 weeks. The results of this study indicate that incubation for longer than 6 weeks is necessary to obtain maximum growth on all excepting the Petragnani medium, A larger number of positives as well as more rapid growth, was secured on the latter medium, as compared with the other egg media. The Peizer medium was next in efficiency in these respects. ~ Loewenstein-Jensen's which has been widely recommended recently, was found to be lowest in efficiency. The other non-experimental medium, D—D, yielded more rapid growth and a higher number (78) of positive cultures than the egg media. Since DuD is liquid, containing a dispersing and a protective agent, and also supplies well charac- terized essential ingredients for growth of the tubercle bacilli, it is not surprising that it should be effective in primary isolation. How- ever, more D-D cultures were contamdnated than was the case with any other medium, excepting A-7l. The selective action of 50 units of penie cillin per ml was insufficient to destroy or inhibit undesirable organisms. Greater amounts of penicillin, however, have been found toxic to g. tuberculosis (51). Since ethyl violet was not present in the medium to color growth of this organism, it was difficult to detect it in the face of the extensive contamination. Media A-EVA-l and A-EVA-5 were found superi- or to the others used in the study from.the standpoint of rapid cultivation of E. tubercu- losis from sputa of patients receiving strepto- mycin therapy, as well as those who were not. The higher number of positive cultures secured with A-EVA-l as compared with A-EVA-3, is so slight as to be statistically insignificant. (79) m— The more rapid growth observed in A-EVA-l, however, indicates that this medium.is su- perior, thus further confirming previous find- ings for liquid and solid media with the same ingredients. Although a great deal of contamination occurred in the ethyl violet--sodium azide media, it has been shown that these ingredients, in the concentrations utilized, are effective selective agents since they were presented here with an especially formidable task, in cultur- ing essentially undigested sputa. The triton A-ZO treatment used has little toxic effect on most bacteria; its purpose was to promote lysis of mucin and mucoid portions. This method of preparation of inocula was used in an attempt to achieve a means whereby tubercle bacilli would not be exposed to the relatively harsh effects of sodium.hydrbxide or other harsh alkalis before they were placed in an environ- ment suitable for growth. The results indi- cated that a higher number of positive cultures in a shorter incubation period can be recovered under these conditions. The lesser toxicity (80) may also account for the much higher number of positive cultures on the expefiimental media from sputa of patients receiving streptomycin, as compared with the number of positive egg media cultures. Although tubercle bacilli bind ethyl violet in a color distinctively different from that produced by streptococci and other bacteria when their growth adsorbs the dye, sufficient coloration of such colonies will mask the dis- tinctive color. It was found also that the rela- tively large inocula used subsequently caused coloration of its albuminous portions by ethyl violet, thus occluding the typical color of‘E. tuberculosis colonies, when the latter was pro- duced. These factors reduced the efficiency in detecting multiplication of tubercle bacilli and probably were responsible for a few falsely nega- tive results. In some instances, multiplication of fungi was so extensive that the bacilli had little Opportunity to initiate growth. Fungus contami- nation, therefore, was the most serious problem in the experimental media. (81) The other great difficulty was due to the reduced surface tension of the ethyl violet-- sodium.azide and D-D media. It was nearly imp possible to remove selected growth aggregates from these media through the use of convention- al tools. Cotton swabs were found to be best to pick up the desired solid material, although other material was also transferred to the slide in the process. The technique found most satis- factory for detecting multiplication of tubercle bacilli could not be utilized, since the cultures, after microscOpic examination,_had to be reincu- bated. This technique consists of centrifugation of the entire culture after it has been trans- ferred to a 15 ml centrifuge tube, decanting of the supernatant, and subsequent preparation of a slide film from the sediment. The only way this method can be employed in a study such as was conducted is through use of several cultures of the same medium, with the sacrifice of a culture at a selected interval. This could not be done in the present study since the specimens were barely large enough for inoculation of the number of tubes of different media used. The necessity of dividing the specimen prior (82) to processing permits some chance of error in results, as tubercle bacilli are never homo- geneously distributed in sputa and the mucoid nature of these specimens makes it impossible to evenly suspend bacteria. It is extremely difficult to effect separation of the specimens into 2 equal portions, as they tend to remain one semi-liquid lump unless chemically treated. Isolation in the experimental media of_a much greater number of tubercle bacilli from sputa of patients receiving streptomycin-~para- amino salicylic acid therapy may be explained, at least partially, by the technique of process- ing the inocula. If these bacilli were placed in a resting stage by prior contact within the in- fective focus, the exposure to sodium.hydroxide digestion and then to the malachite green in the medium, might be sufficient to inhibit small numbers so that they would never initiate growth. Some members of larger inocula, by virtue of the protective effect of large numbers of mycobac- teria, are little affected and will thus multi- ply. The superior nutrient qualities of the (83) experimental media, as well as the fact that they are liquid rather than solid, undoubtedly assist affected bacilli to return to the loga- rithmic growth stage and so further increase the favorable environmental conditions presented to the tubercle bacilli. The tremendous amount of time required to microscOpically examine each culture in which evidence of growth was detected makes use of A-EVA-l without chemical digestion difficult, if not impossible, for the routine culture of sputa and other pathological material. If it were not for the marked contamination of this, as well as the other ethyl violet--azide media, multiplication of tubercle bacilli would have been much more readily detected. The basic problem, therefore, is to control the growth of the undesirable organisms either by other inhibitory agents or by better means of chemi- cal digestion of the specimen prior to inocula- tion, and to be sure that the effect of these agents on the tubercle bacillus is no more ad- verse than those to which it was subjected in this study. (814) IV The Influence of Therapy on Primary Isolation of Tubercle Bacilli An attempt was made to determine the imp portance of the effect of streptomycin in sputa on cultivation ofIM. tuberculosis. StreptOmycin added in final concentrations greater than 0.2 units per ml to sputum of positive patients who had not previously received this antibiotic prevented isolation of tubercle bacilli on the Petragnani medium. The 5 per cent sodium hydrox- ide digestion technique previously described was utilized. Concentrations permitting colony production on this medium were: 0.1 unit per ml and 0.05 unit per ml. With A-EVA-l, without sodium hydroxide digestion, tubercle bacilli multiplied from inocula containing 5.5 units of streptomycin per ml of the original unconcen- trated specimen, while cultures of a specimen with a concentration of 5.75 units showed no growth of these organisms. All intact sputum specimens were incubated at room temperature for an hours after the specific streptomycin concentration had been added, before they were concentrated by centrifugation and subsequently processed in accordance with the medium which (85) was to be inoculated. 6 Negative sputum specimens seeded with 10' mg of tubercle bacilli yielded growth in the presence of 0.5 units of streptomycin, but not 0.75 units per ml of the specimen when cultured on the Petragnani medium.after sodium hydroxide digestion. Using A-EVA-l and the triton A—20 treatment previously described,'M. tuberculosis was isolated from specimens containing 5.0 units of streptomycin per ml, but not from those with 5.25 units per ml. These seeded sputa were also incubated at room temperature for 2h hours before they were processed further. At the time therapy with isoniazid was first initiated on an experimental basis it was noted, by qualitative chemical test, that this drug is present, in very low concentrations, in sputa of patients taking it. Cultural results at that time indicated that these low levels in the sputum.did not influence primary iso- lations of.M. tuberculosis. Subsequent control- led studies conducted along with these strepto- mycin investigations, indicated that tubercle bacilli would grow on the Petragnani medium (86) from sputum inocula containing up through 0.2 mg of isoniazid per ml of the original specimen. In A—EVA-l, these bacilli grew from the inocula originally containing up through 0.55 mg per ml. The effect of isoniazid was found to be more dependent upon the character of the sputum speci- men than was that of streptomycin. Mucoid speci- mens appeared to exert a greater protective effect than did more watery sputa. Para-amino salicylic acid was not used in the study since it had previously been shown to be present in sputa and, under such con- ditions, exert a toxic effect on M, tuberculosis (59). The small inocula, which were well distri- buted in the specimens, were employed in order to secure conditions similar to those encountered in primary isolation. The strain of tubercle bacilli utilized was sensitive to 0.1 unit of streptomycin per ml, as determined in a liquid medium without a dispersing agent. Since the character of negative sputa, is not the same as that of positive, the results varied for the positive sputa and the seeded negative specimens. (87) The fact that the size of the inoculum was controlled in the latter case also would lead one to expect slightly different results. V Study of Selective Agents in Primary Isolation Materials and Methods In order to evaluate the advantages of selective ingredients in the A-EV-l medium.as well as to compare it with other media, special- ly prepared sputum specimens were cultured. Twenty 5 day pooled specimens were obtained. Four ml of 10 per cent triton A-ZO was added to each, and they were incubated at room.temper- ature for 2h hours. Then they were concentrated by centrifugation at 2,000 rpm for 20 minutes and slides prepared which were examined micro- scOpically. The positive sediments were come bined with the negative so that subsequent slides prepared from the combined concentrate contained 5 to 10 tubercle bacilli per 10 microsCOpic fields. The concentrate was resuspended in 0.5 M phosphate buffer containing 0.2 per cent triton A-ZO. The suspension was then diluted 1 to 10, 1 to 20, 1 to ho, 1 to 80, and 1 to 160. (88) The media employed were the following: A-EV-l, A-EV-l with 100 units per ml of peni- cillin, A-EV-l with 100 micrograms per ml of acti-dione, A-EV-l with 100 units of penicillin and 100 micrograms of acti-dione per ml, the D-D medium.with 100 units of penicillin per ml, the Dubos-Middlebrook agar with 100 units of penicillin per ml, and the Petragnani medium, The diluted sediments, treated only with the triton A-20--buffer mixture, were cultured, using 2 tubes of each of these media. Twenty additional 5 day pooled sputum.speci- mens were obtained. These were treated for one- half hour with 5 per cent sodium.hydroxide. They were neutralized with 5 per cent hydrochloric acid to phenol red. Concentration was effected by centrifugation at 2,000 rpm for 20 minutes. Slides were prepared from.the sediments. The sediments were again combined, and it was found that subsequent slides revealed 5 to 50 tubercle bacilli per 10 fields examined microscopically. The concentrate was suspended in the same mix- ture, the same dilutions prepared, and the same media inoculated. (89) Several sputa negative by concentrate micro- scOpic examination were pooled, then divided into 8 approximately equal portions. Individual por- tions were seeded with the following amounts of 2 3 h, 10‘7, tubercle bacilli: 10’1, lo' , 10- , lo' and 10"8 mg (wet weight). Efforts were made to fully disperse the bacilli and distribute them throughout the portion. These portions were then digested with sodium.hydroxide and cultured, using 2 tubes of each of the media described. Discussion of Results With all 5 groups of sputa, A-EVel contain- ing penicillin and acti-dione was found to be contaminated less frequently. More positive cultures were secured with this medium.than with any of the others, since multiplication of other organisms was reduced permitting easier detection of growth of tubercle bacilli. Fungus contamination, even in cultures of the first group of specimens, was almost non-existent in the media with acti-dione. On the other hand, in media which did not contain this ingredient, fungus contamination was frequent and sometimes prevented detection of tubercle bacilli. The (90) use of penicillin in conjunction with ethyl violet largely inhibited streptococci, as well as occasional other undesirable bacteria which grew in the media not containing this antibiotic. Nevertheless nearly as many positive cul- tures were secured from the sodium hydroxide digested inocula as from.those which had not been so treated. The further reduction in contamination effected by this digestion facili- tated recognition of the characteristic color of multiplying tubercle bacilli. Therefore, sodium.hydroxide digestion, penicillin, and acti-dione served to enhance the ease in de- tecting positive cultures. Penicillin, in the concentration used, did not eliminate or satisfactorily reduce contami- nation of the D-D and the Dubos-Middlebrook media. Many cultures were completely overgrown. It was not possible to determine its effect on the tubercle bacillus during primary isolation in these media because of the presence of other organisms. More fungus contamination occurred in penicillin containing media without acti- dione than in media without both substances, (91) thus further demonstrating the growth promoting effect of penicillin on many fungi. VI Effects of Selective Agents on the Growth Rate of E, tuberculosis in Liquid Media Turbidometric growth studies were made on the following experimental media: A-EV-l, A-EV-l with 100 units of penicillin, A-EV-l with 100 micrograms of acti-dione, A-EV-l with both peni- cillin and acti-dione in the given concentrations, A-EV-l without ethyl violet (A-l) but with 100 units of penicillin, A-l with penicillin and acti-dione, and A-1 with varying concentrations of acti-dione. The inocula used were small amounts of tubercle bacilli, ranging in weight from 10.1 to 10..6 mg. The results indicated that growth of g, tuberculosis was more rapid when sodium azide was not present in the medium, confirming primary isolation study data. Peni- cillin and acti-dione were found to have a slight inhibitory effect on this organism, in that the lag phase was slightly lengthened when they were present individually or in combination. In A-l with varying concentrations of acti- dione, it was found that growth of g. tuberculosis (92) was unaffected by 50 micrograms per ml. In subse- quent non-turbidometric studies culturing sputa, it was found that tubercle bacilli would grow well in the presence of 100 micrograms acti-dione per ml. However, it was found in liquid media, in the presence of penicillin and ethyl violet, that 50 micrograms per ml were sufficient to eliminate much of the fungus contamination. With A-EV-l containing all of these agents, turbidOe metric studies indicated that acti-dione concen: trations greater than 50 micrograms per ml slowed the rate of growth of tubercle bacilli. Micro- scOpic examination revealed that the bacilli had assumed short, coccoid morphology. These effects probably resulted from the absence of the pro- tection provided by an inoculum of pathological material It was observed while culturing E, tubercu- ‘lggig in alkaline media containing indicator dyes that the hydrogen ion concentration greatly influ- ences the activity of acti-dione. Best cultural results were obtained when the pH was slightly below neutrality. Some deterioration of both penicillin and (95) acti-dione was indicated by the more rapid rate of increase in Optical density, after the fourth day of incubation, of the bacilli growing in A-EV-l containing these substances, as compared with the densities of those in A-EV-l. The in- crease cannot be entirely attributed to complete deterioration, however, as adsorption phenomena of larger masses of bacilli may play a role. Untreated sediments of a series of sputum and gastric specimens were cultured in 125 ml Erlenmeyer flasks containing 50 ml of ethyl violet experimental media. Comparison of re- sults secured with these media and with Petrag- nani's medium indicated that contamination of the former media was so extensive that it was nearly impossible to detect and remove multi- plying tubercle bacilli. The only means of satisfactorily determining the bacterial flora was by removal and centrifugation of a portion of the well mixed medium. The odor of metabolic products of the contaminants was objectionable. Addition of penicillin and acti-dione in suffi- cient strength to retard these other organisms inhibited growth of tubercle bacilli from all (9h) but the high Gaffky inocula. Subsequent studies with untreated specimens cultured in several types of experimental media also led to the conclusion that sputum and gastric sediments must be digested chemically before inocu- lation, as the selective agents employed were not effective against the massive numbers of viable contaminants introduced in untreated sediments. Use of flasks rather than tubes, was con- sidered desirable, however, since the greater volume of medium permitted use of larger inocu- la, thus increasing the likelihood of introducing sufficient tubercle bacilli to initiate rapid growth. Use of flasks also provides greater aeration of the medium than is possible in culture tubes. VII Primary Isolation of'y. tuberculosis on Glass Wool Suspended in Liquid Media~ Sufficient glass wool (fiberglass) was placed in 125 ml Erlenmeyer flasks that, when 50 ml of A-EV—l (final volume after enrichment) was added, it formed a mat at the surface of the medium. Tubercle bacilli grew in characteristic (9S) blue-violet colonies on the surface of this mat when recently isolated cultures and later, chemi- cally digested sputa were used as inocula. Growth was rapid and luxuriant, and was easily detected in the presence of colonies of other organisms. It had been observed previously that colonies of E. tuberculosis were colored pink by neutral red when this dye was present in liquid culture media. When the glass wool technique was employed to culture recently isolated tubercle bacilli in these media, pink colonies develOped on the surface of the glass fibers. After approximately 50 days incubation, the colony color gradually faded until only a faint trace of color remained. In earlier studies, fat soluble dyes were dissolved in oleic acid and various oleate-a1- bumin complexes were prepared using the dyed fatty acid. In liquid media containing these enrichments, the complex was bound to the surface of pathogenic and saprophytic mycobacteria, re- ’sulting in distinctively colored colonies. The fat soluble dyes used were: methyl orange, nigrosin, Sudan I, Sudan III, Sudan IV, and Sudan black B. The Sudan IV--oleate complex was found to impart (96) the most brilliant color to colonies. Those colored by the Sudan black B complex were also distinctive. Colonies of tubercle bacilli grown in methyl orange and nigrosin media were color- less. Although toxicity studies were not made, the Sudan IV--oleate complex appeared to be the least inhibitory to multiplication of tubercle bacilli since growth was most abundant in media containing it. A literature review failed to reveal previ- ous use of fat soluble dyes in media for culti- vation of tubercle bacilli. The insolubility of these dyes in water probably prevented earlier use as indicators of growth of mycobacteria. Previous attempts to color colonies of'fi. tuberculosis with Sudan III dissolved in tween 80 were unsuccessful. Sudan III stained sodium oleate added to a liquid medium.before autoclav- ing did not color colonies grown in the medium. In pure culture studies it was established that pink colonies were formed on the glass wool mat when the liquid medium.used in the flasks was enriched with the Sudan IV--oleate complex. (97) In order to determine its value, the glass wool culture technique was used with ethyl violet, neutral red, ethyl violet and neutral red, and Sudan IV--oleate experimental media to culture (E, tuberculosis from sputum specimens. The Petragnani medium.was also used for comparative purposes. Materials and Methods The sputum specimens, 5 day pools of miscel- laneous character from.tuberculous patients, were divided into 2 equal portions. One of these was digested with 5 per cent sodium hydroxide while the other portion was treated with a triton A-ZO digestion mixture. One glass wool flask was in- oculated with each portion. Two tubes of the Petragnani medium were seeded with the sodium.hy- droxide digested sediment. The cultures were ex- amined macroscOpically at least every 2 days. Growth was identified by microscOpic examination. The triton A-2O digestion mixture was comp posed of the following: Trisodium phosphate ............. 10.0% Triton A-ZO 0.0000000000000000... 008% Penicillin ..........g........... 100 units/ml (added at time of use) Distilled water (98) The mixture was added in equal volume to the sputum specimen. Di cation was carried out at room temperature for to 2A hours. The media employed were the following: A-EV-l, A-EV-l with 1 to 500,000 neutral red added (A-EVNR-l), A-l with 1 to 500,000 neutral red (A-NR-l), and A-1 enriched with Sudan IV-- oleate complex (A-SIV-l) rather than the mixture described for this medium, For the most part, only 1 of these media was compared with Petragnani's in culturing a series of sputa. The Sudan IV-- oleate complex consists of the following: 1 mg of Sudan IV is dissolved in 10 ml of oleic acid. 0.6 ml of this sol'n is added to 50 m1 of N/20 sodium hydroxide in a 50 ml Erlen- meyer flask and the resulting flocculation is dispersed by vigorous rotation of the flask. 5 ml of this sol'n is added to: Bovine serum.albumin (fraction V) ... 2.5 gm Glucosa 00.000...000..0.000.000...... 705 Dissolved in Saline (0.85%) 00......00o000..0.0.00 b.500m1 BOVine Serum 00.0.00000.00.0.000000.0 5000 ml All experimental media contained 100 units per ml of penicillin and 60 micrograms per ml acti- dione. Results A summary of the results secured is presented (99) in table A. Growth of tubercle bacilli was detected in the shortest time in the A-SIV-l flasks, since their colonies were colored pink. It was found that traces of the color could be detected readily on the surface of the glass wool since the medium contained no other dye. Thu. multiplication was perceived earlier. Colonies of E. tuberculosis were identified on an average of 8 days after inoculation with the triton A-20 digested sediments and 12 days after the sodium hydroxide treated sediments were in- oculated. The positive cultures on the Petragnani medium.from this series of specimens were secured after an average of 5h days incubation. Nearly as rapid growth was observed in the A-NR-l flasks. An average or 10 days was re- quired to detect colonies of E. tuberculosis from the triton treated inocula, while 15 days were required in the case of the sodium hydroxide digested sediments. Colonies of this organism were again colored pink, although the color as- sumed through binding of neutral red is not identical to that produced by combination with the Sudan IV--oleate complex. Tiny neutral red- (loo) colored colonies were not as readily detected as were the Sudan dyed colonies, since the medium. is slightly pink in the former instance. Tubercle bacilli multiplied.more slowly in the flasks with media containing ethyl violet, probably due to the additive toxic effect of the 5 selective agents, penicillin, acti-dione, and ethyl violet. With A-EVNR-l, colonies of the bacilli were detected sooner than with A-EV-l, since the ethyl violet and neutral red in the former medium colored them more distinctively than ethyl violet alone. Colonies of tubercle bacilli in A-EVNB-l were purple pink in color and were more readily identified in the presence of ethyl violet dyed albuminous matter from the inoculum and colonies of streptococci and micro- cocci. Colonies of the former are colored by ethyl violet, while the latter were found to be rendered pink by neutral red. The percentage of contaminated cultures was greater with the media without ethyl violet. More contamination was noted in flasks seeded with triton digested sediments. The cultures recorded as contaminated in table h were complete- (101) 1y overgrown with undesirable organisms so that it was impossible to identify colonies of M. £2: berculosis if they were present. Many flasks of the different media showed some contamination in the form of isolated colonies of other bac- teria. These colonies did not impair identifi- cation of tubercle bacilli, however. The number of such partially contaminated cultures was no greater than for the Petragnani medium. The greatest drawback encountered in use of the glass wool technique was the difficulty experienced in removing portions of colonies from the glass fibers for microscOpic exami- nation. They were softer and smoother than those encountered on the classical egg media and tended to cling tenaceously to the fibers. Discussion Positive cultures were detected sooner with the experimental media because of 2 factors: (1) multiplication of tubercle bacilli was more rapid, and (2) colonies were colored distinctive- ly by the dyes present in the media. It is ap- parent that growth was faster than on the Petrag- nani medium, since it was employed to culture (102) each series of sodium hydroxide treated sediments. It is more difficult to directly compare growth rates in the experimental media because with each a different series of specimens was cultured. The series with a larger number of microscopically positive concentrates would likely show a shorter average growth rate. Tubercle bacilli strain differences, as well as physical and chemical characteristics of the specimens, also influence the rate and extent of multiplication. Comparison of the per cent positive cultures and the average number of days required for growth to become visible in experimental media with the results on Petragnani's does furnish a means of evaluation. On this basis, growth was most rapid in A-SIV-l and slowest in A-EV-l. The colors imparted to colonies of‘y..§g- berculosis by Sudan IV--oleate and neutral red are not specific for virulent strains. Strains which failed to produce tuberculous lesions in the guinea pig were indistinguishable from those which proved virulent. However, cultivation of E. smeglatis, I_¢I_. butyricum, fl. hlei, E. lacticola, and E. stercoris in A-EVNR-l, A-NR-l, (103) and A-SIV—l showed that colonies of these acid- fast saprOphytes could be distinguished from M, tuberculosis by their color. It is less vivid than that of colonies of tubercle bacilli. It is believed that the Sudan IV--oleate- albumin complex is bound by the lipid coat of these bacilli in the same fashion as is the lipid portion of the tween 80 molecule and the oleate-albumin complex. Since the oleic acid is colored, the presence of small aggregates of bacilli is revealed by a tiny pink spot or mass. Another point of contact to the tubercle bacillus is afforded by the binding of undenatured bovine albumin to the lipids of this organism. MicroscOpic confirmation of typical colonies of yeast-like fungi which occasionally multiplied despite the presence of acti-dione adso were colored pink. In rare instances, these colonies appeared identical to those of E, tuberculosis. Neutral red has been employed by Dubos and Middlebrook (29) in an i§}zitgg test for viru- lence of tubercle bacilli. They found that, in an alkaline solution of this indicator in which it is yellow, virulent bacilli will bind (10h) it in its red, or acid, form. During develOp- ment of A-NR-l, it was found that tubercle bacilli growing in alkaline neutral red liquid media would produce pink colonies, although the medium was yellow in color. When some of these media were used for primary isolation of the bacilli, however, it was found that they were more readily contaminated by fungi. The acti-dione content of the media had probably deteriorated due to the alkalinity. The phenomenon was demonstrated with recently isolated virulent cultures in such media, however, although growth was less luxuriant, proba- bly due to departure from the optimal pH range for growth of M. tuberculosis. As mentioned previ- ously, colonies of some strains of micrococci are also pink in A-NR-l. The presence of such colonies makes microscOpic confirmation of sus- pected tubercle bacilli imperative. It was possi- ble to macroscOpically differentiate between the acid-fast saprOphytes previously mentioned and colonies of M, tuberculosis by the degree and intensity of color. The glass wool technique of culturing tuber- cle bacilli combines the advantages of liquid and (105) solid media. The growth promoting effect of wetting agents can be employed as in liquid media without total submersion of growth into an environment of oxygen semi-starvation. Softer, and smoother colonies of mycobacteria which grow more rapidly are the result. The growth pro- moting effect of wetting agents is much reduced in agar and other solid media. Colony formation on the surface of the glass wool makes examina- tion and identification much easier than in the case of sub-surface growth in the presence of ‘ contaminants. The colonies are as firmly anchor- ed as on solid media, and are distinctly sepa- rated from.those of other bacterial species, thus facilitating transfer of pure cultures. Because the medium surrounds the inoculum almost as com- pletely as in sub-surface cultures, tubercle bacilli contact nutrients more readily and waste products are diluted and carried away more quick- ly than on solid media. Thus growth is faster and more luxuriant than on the egg and agar media. Some evaporation of the media was noted during incubation. It could have been prevented by placing cellOphane or paper hoods over the (106) necks of the flasks, but this was not done as it might have reduced the oxygen tension within the flasks. Evaporation was not more than 10 ml during a 30 day incubation period. The virulence of many of the strains of tubercle bacilli isolated in the glass wool ex- perimental media was ascertained by guinea pig inoculation. All but 1 proved virulent. The specimens cultured were from patients with diag- nosed tuberculosis and the colonial morphology of the strains when transferred to solid media was identical to that of virulent tubercle bacilli. VIII Cultivation and Isolation of E. tuberculosis on Charcoal Agar The complex enrichment mixture which stimu- lated and enhanced growth of E. tuberculosis to the greatest extent cannot be sterilized by autoclaving without loss of much of its effective- ness. It must be filter sterilized and added aseptically after autoclaving of the basal mediums This affords an Opportunity of accidental contami- nation of the medium as well as making preparation more laborious. (107) Ley and Muller (59) reported that addition of charcoal to agar media eliminates the toxici- ty to gonococci of a fatty acid-like substance contained in agar. Pollack (81) found that charcoal will neutralize the toxic action of fatty acids for Hemophilus pertussis. Stonebrink (95) compared a dextrinized starch--activated charcoal egg medium with that of Loewenstein and found that, with bovine tubercle bacilli, growth was speeded ho per cent on the former medium. In view of these findings, a controlled study of the effects of different grades of activated charcoal on the growth of recently isolated strains of human and avian tubercle bacilli seemed warranted in order to learn if they would produce a stimulatory effect amm- parable to that of undenatured albumin enrich- ments. If such growth stimulation occurred, media containing the charcoals could be compared with other media, including Petragnani's. Materials and Methods Three strains of E, tuberculosis var. hominis, (108) recently isolated, were mixed and graded inocula 1 6 7 of 10' , 10'2, 10'}, 10'”, 10'5, 10' , 10' , and 10-8, mg were prepared from the mixture using the method of Youmans and Karlson (106). Two strains of avian tubercle bacilli were also mdxed and the same series of inocula prepared from this mixture. All strains had been culti- vated on the basal medium enriched with bovine albumin fraction V for 12 to 16 days. The basal medium.amployed was composed of the following ingredients: i Proteose peptone No. 3 (Difco) ... Disodium phosphate (NaZHPO 012B 0) MonOpotassium.phosphate ( 2P0h3 . Copper sulfate (CuSO OSHZO) ...... Cobaltous sulfate (CoSOh°7H20) ... Ferric 8.111111011111131 citrate eeeeeeeeee Glucose ..........O...’........... Glycer01 ......O..............0... sav1ta 0..........CCCCOOCOOCOOOOO. Triton A'ZO eeeeeeeeeeeeeeeeeeeeee Agar eeeeeeeeeeeeeeeeeeeeeeeeeeeee 150 001 001 e e e e OU'IWHOOOU‘O ONN-QOOOI—‘NN Four gm of each of the following grades of nuchar activated charcoal were added to liter portions of this basal medium: C-ll5-N, C-l90-N, WA, and aqua A. The final pH of all the media was 6.8. The media were tubed in 5 ml amounts in screw capped tubes, and sterilized at 121° C for 15 minutes. While the media were cooling, the tubes were slanted. Before utilization of the diluted inocula, (109) 1 tube of each of the media was seeded with -l 10 mg of a recently isolated human strain of M, tuberculosis. Multiplication of this strain was noted within 2 days incubation. This es- tablished that each of the charcoal media would support growth. In generation time studies, 2 tubes of each medium were inoculated with each of the individu- al concentrations of tubercle bacilli. All cultures were examined at daily intervals macroscOpically. Evidence of growth was confirmed by preparation and examination of slide films. Several separate trials were made of each of the media, using the actively growing human and avian tubercle bacilli. As in the dye toxicity studies, gener- ation times of the bacilli were calculated by the method of Youmans and Youmans (107). Ethyl violet was subsequently added to the best charcoal enriched media in concentrations ranging from 1 to 200,000 to l to 1,300,000. Media employed in isolation of E, tuberculosis from sputa also contained 100 units of peni- cillin and 100 micrograms of acti-dione per ml. (110) The sputa were digested and concentrated by means of the sodium.hydroxide technique previ- ously described. Two tubes of each medium were inoculated, h loopfuls of the sediment being used for each tube. During incubation the cultures were examined macroscopically at intervals of 2 days. All colonies were checked microscopically, using the Ziehl-Neelsen stain. After h2 days a final reading was made for all cultures and they were discarded.. The concentrate smears, 2 of which were made for each sediment, were also Ziehl-Neelsen stained and examined for at least 20 miDUtOS 0 Results The generation times in hours of the human and avian tubercle bacilli when growing on the basal medium, the charcoal media, and the basal medium enriched with 0.2 per cent bovine albumin, 10 per cent human serum, and 10 per cent Dubos oleic albumin complex are presented in table 5. The percentage of stimulation of the rate of growth on each over that on the basal medium is also given for-both varieties of tubercle bacilli. (111) Growth was most rapid on the bovine albumin enriched medium. Extreme differences in effect on the growth of tubercle bacilli were noted with the different types of activated carbons; C-l90-N stimulating growth of the human strains 20.8 per cent, as did human serum, while C-115-N accelerated the generation time 16.7 per cent. WA produced only h.l per cent stimulation over the basal medium, and aqua A exhibited no stimu- lation. Although generation times were less regular, the avian tubercle bacilli manifested a similar pattern of response to the h charcoals, C-l90-N again proving the most effective in stimue lating the rate of growth. With both varieties ofIg. tuberculosis, the oleic albumin enrichment was nearly as effective as C-l90-N. In the first 2 trials with human strains, the generation times were identical: 22.8 hours. Twenty-six 5 day pooled sputum.3pecimens were cultured on Petragnani medium, and on C-l90-N and C-llS-N. The charcoal media contained 1 to 1,300,000 ethyl violet, as well as the concen- trations of penicillin and acti-dione previously noted. (112) Table 6 records the number of isolations and average time required for detection of'g. tuberculosis, as well as the contamination en- countered in culturing the 26 specimens, in h of which acid-fast bacilli morphologically identi- cal to tubercle bacilli were found by microscOpic examination. The Petragnani medium.yielded 6 positive cultures as compared with 7 for C-llS- EV. However, the least contamination was experi- enced with the former medium. It was possible to detect colonies of‘g. tuberculgsis on some cultures which were contaminated but not over- grown e The excessive contamination of the experi- mental media, probably due to reduction in active ethyl violet concentration through its adsorption by charcoal, led to a study of generation times of human tubercle bacilli on these 2 media, as well as on different preparations of C-190-EV contain- ing 1 to 200,000, 1 to h00.000. 1 to 600,000, 1 to 800,000, 1 to 1,000,000 and l to 1,200,000 ethyl violet, in order to determine the highest concen- tration which could be used in this medium.without hindering multiplication of tubercle bacilli. As (11;) shown in table 7, the generation times for all concentrations in the C-l90-EV medium were identi- cal until at 1 to u00,000 it rose from 22.8 to 2h.0 hours. In the medium containing 1 to 200,000 ethyl violet the time greatly increased to 31.2 and 32.6 hours. Therefore this latter medium was not included in the study of these media in the culture of 3 day pooled sputa, the results of which are presented in table 8. In culturing h2 specimens, 11 isolations of tubercle bacilli were obtained with C-190-EV containing 1 to h00,000 ethyl violet, as compared with 8 for the Petragnani medium, and the number of contaminated cultures was least of all media. Fewer positive cultures were obtained with the other media, probably due to a greater number of tubes being partially or completely overgrown. As with the Petragnani medium.tubercle bacilli were isolated in some instances on all charcoal media although other organisms had produced colonies. Ethyl violet in a concentration of l to h00,000 was incorporated in the 3 other charcoal media previously studied to determine the effect (111;) of this dye in the presence of C-llS-N, WA, and aqua A on the generation times of a mixture of 2 newly isolated strains of human tubercle bacilli. These media were compared with the basal, and the basal enriched with the substances previously studied. The generation time of the organisms was 27.6 hours on the C-l90-EV medium, as opposed to 2h.0 hours for the human strains previously used in the study of this medium. However, the findings with the other media corresponded to previous results, in that they stimulated growth to approximately the same extent over the gener- ation time on the unenriched medium. Experimental medium C-l90-EV containing 1 to h00,000 ethyl violet was compared in routine culture of 70 sputa, gastric washings, and other pathological material. Seven of these were posi- tive microscOpically. Eighteen isolations of E, tuberculosis were obtained with the charcoal medium, while with Petragnani 12 were positive for tubercle bacilli. Fourteen C-190-EV cultures showed some contamination and 5 were overgrown. With Petragnani medium 26 were partially contami- nated and 6 were overgrown. Thus greater recovery (115) of tubercle bacilli with fewer contaminated cultures was achieved using the charcoal experi- mental medium. Discussion It was anticipated that the avian strains of g. tuberculosis would grow more rapidly on all media than did those of the human variety. However, little difference in the generation time was noted. Perhaps the fact that one of the two strains used had been very recently isolated was responsible for the deviation from behavior previously experienced with such strains. Final growth on all media was more luxuriant with the avian bacilli, however. The nuchar grades C-l90-N and C-llS-N both stimulated growth of tubercle bacilli sufficient- ly to be utilized in.cu1tura1.media as substi- tutes for human and bovine serum.or albumin, C- l90-N proving superior in this respect. Use of this activated carbon simplifies the preparation of cultural media since, unlike the other en- richments studied, it can be added before steri- lization by autoclaving. Thus a saving in time is effected and the danger of accidental contami- (116) nation during preparation is reduced. The peni- cillin--acti-dione inhibitory mixture is not as likely to contain viable bacteria before its addition to the medium. The greatest stimu- latory effect was obtained through addition of bovine albumin to the basal medium. C-l90-N enhanced growth nearly as effectively, however, and was approximately equivalent to the stimu- latory effect of human serum. The activated charcoals used in the media are purified products of cellulose manufacture. Each grade is designed for a particular function. Thus it is not unexpected that differences in effectiveness in promoting growth of E. tubercu- ‘lggig might be found in this study. C-llS-N and C-l90-N will adsorb fat and fatty acids more effectively than will WA and aqua A (M6). WA has a low affinity for dyes which is reflected in the greater effectiveness of more dilute concentrations of ethyl violet in the WA medium in preventing the growth of contaminants. It is believed that the activated charcoals act as protective agents, as do starch and blood and egg constituents, adsorbing toxic substances (117) present in trace amounts in the medium. Whether they serve other purposes in the medium.is not known at present. Oxygen bound to the surface of the charcoal particles may be available for utilization by the highly aerobic tubercle ba- cillus. In culturing pathological material, growth of M. tuberculosis on the C-l90-EV medium was much more abundant at the conclusion of the h2 day incubation period than on the Petragnani medium. The l to h00,000 concentration of ethyl violet used in the final C-190-EV medium is greater than had been found necessary to in- hibit contaminants in agar media enriched with bovine albumin or oleic albumin. In the latter media we have found this amount of dye to be toxic to the tubercle bacillus. In the presence of the charcoals, however, a portion of the added ethyl violet is rendered inactive, proba- bly through adsorption. The concentration of acti-dione used had previously been found, in other agar media, to (118) adversely affect the growth of tubercle bacilli. In C-l90-EV no such toxicity was perceived. However, in the comparative studies this medium was contaminated more frequently than had been the case in the past when penicillin and acti- dione were incorporated in ethyl violet agar media enriched with bovine albumin. Perhaps the activated carbon also reduced the activity of penicillin and acti-dione, as it had been found to do with ethyl violet. IX General Discussion The method of Youmans and Youmans (107) which was utilized to determine growth rates of strains of tubercle bacilli is believed to be more accurate than turbidometric methods or those employing micro-Kjeldahl determinations. For precise turbidometry, complete dis- persal of bacterial aggregates is essential. MicroscOpic examination of portions of triton A-20 media during turbidometric studies of mycobacterial growth revealed the presence of small clumps of bacilli. The concentrations of this wetting agent were the highest that (119) could be employed in culture media without toxicity to tubercle bacilli (h9). These small aggregates were also observed in.media containing tween 80 as dispersive agent, al- though dispersal appeared more complete. How- ever, as was previously mentioned, effective concentrations of tween 80 are toxic to M. tuberculosig, even in the presence of undena- tured albumin (88). Thus, since homogeneous suspensions of growing tubercle bacilli are unobtainable at present, turbidometric methods furnish only approximate growth rates. Sattler and Youmans (98) found that the large inocula of tubercle bacilli required for micro-KJeIdahl determinations grew more slowly than smaller inocula. Youmans and Youmans (107) subsequently observed that conditions in liquid media are more favorable for growth of smaller numbers of tubercle bacilli. Re- duction of available oxygen and the restrictive effect of large clumps of bacilli are probably responsible. The protective effect of large numbers of mycobacteria also will influence their growth rate. (120) The Youmans method is even more influenced by the mass of the individual bacterial cell than are turbidometric determinations, because it is based on detection of earliest macroscOpic growth. Since medium composition strongly influ- ences the size and weight of tubercle bacilli, the media used in generation time studies must be basically similar. Thus growth rates de- termined for bacilli multiplying on egg media with high glycerol concentrations are not di- rectly comparable with those for these organ- isms on oleic albumin agar. In determining growth rates of tubercle bacilli the same concentrations of carbon sources were used in each comparative study and no at- tempt was made to directly compare generation times on entirely different media. Since all influencing factors were constant, use of the Youmans method provided the most reliable growth rate values obtainable. -8 Since the wet weight of 10 mg is believed to contain 2 to 3 tubercle bacilli (19), growth of this inoculum.was not always observed. In some instances this inoculum was very likely (121) sterile, due to clumping of the bacilli in previous dilutions. In other cases the 2 to 3 bacilli may not have been viable. In the dye toxicity studies, however, the occasional failure of the 10.8 mg inoculum.to initiate growth was probably due to the effect of the dye. When this occurred, the 10..6 growth time was used in calculation of the generation time. Use of the method using graded dilutions of heavily positive sputum concentrates as inocula in order to evaluate experimental media would have been desirable. However, it was found that homogeneous suspensions of tubercle bacilli in these sediments could not be ob- tained or prepared. Thus the number of bacilli in each dilution would not be pr0portionally reduced. Contamination of cultural media by sapro- phytic and commensal organisms was reduced through use of acti-dione to inhibit fungi, and ethyl violet and penicillin to prevent growth of undesirable bacteria. Although completely selective media were not develOped, the most troublesome contaminants, gram-positive (122) bacilli and the fungi, were successfully inhi- bited through use of these agents. It was found necessary to employ chemical digestion of sputum.specimens prior to culture. However, presence of the selective agents in the media permitted use of milder digestion mixtures. Using these, tubercle bacilli were isolated from.a greater number of specimens from patients receiving streptomycin--para- amino salicylic acid therapy than was possible using sodium hydroxide digestion and one or more of the classical egg media. The 3 per cent sodium hydroxide digestion technique was used consistently, for compara- tive purposes. Although this technique was found earlier to be less satisfactory and more toxic to tubercle bacilli than various trisodium. phosphate digestion mixtures (112), it is the most widely accepted method. Efforts to simplify the preparation of agar media for the isolation of E. tuberculosis resulted in the discovery that nuchar C-l90-N activated charcoal is nearly as stimulatory to (125) growth of this organism as the Dubos oleate-- albumin complex. Use of the charcoal simpli- fies preparation of these media by making en- richment after autoclaving unnecessary. The possibility of accidental contamination during enrichment is thus eliminated. It is probable that addition of penicillin and acti-dione to the medium after it has been sterilized can be rendered unnecessary through use of these substances in digestion of speci- mens prior to culture. Under such conditions, penicillin and acti-dione would be less toxic to E. tuberculosis because of the protection offered by the albumin content and the mucoid qualities of the specimen. Tubercle bacilli would not then be exposed to penicillin and a surface active agent for a long period of time as they are in media containing this selective agent. The apparent inhibitory effect of auto- claved glucose for E. tuberculosi§_was not observed in the charcoal agar media. Perhaps the degradation products responsible for this effect are adsorbed or partially bound in a (121.1) non-toxic form. Thus it is unnecessary to add filter sterilized glucose following auto- claving of the medium. Recent studies have indicated that acti- vated charcoal will enhance the growth of tuber- cle bacilli in liquid, as well as solid, media. Using the glass wool technique, it is much more difficult to detect small colonies because of the insoluble carbon particles clinging to the glass fibers. In view of the greater suscepti- bility of tubercle bacilli in liquid media to certain toxic substances, it would be worth- while to compare the relative protective ad- sorption of charcoal and bovine albumin in liquids. Since colonies of E. tuberculosis on glass wool mats are exposed to air, the problem.of oxygen deficiency as it is encountered in depth culture is not involved as a growth limiting factor. Because, as was previously pointed out, many of the advantages of liquid medium.culture are provided, multiplication of tubercle bacilli is more rapid. The apparent rapid growth rate indicated by the primary isolation studies (125) should be confirmed by generation time studies. The dyes used in the experimental media provide a presumptive means of detecting colonies of g. tuberculosis. The ethyl violet-~neutral Cred combination is the most reliable indicator of growth of this organism. Ethyl violet and penicillin effectively inhibit bacteria whose colonies also are colored by the combination. The Sudan IV complex is bound by yeast-like fungi and by actinomycetes, as well as by tuber- cle bacilli. While acti-dione will largely elimdnate the fungi, the actinomwcetes often multiply despite the presence of the selective agents. The ingredients of the basal experimental medium are, for the most part, chemically charac- terized and serve a specific function in.sup- porting growth of tubercle bacilli. It was shown experimentally that removal of an ingredi- ent either reduced the amount of growth or slowed the rate of growth. The only ingredients not clearly defined chemically are the yeast extract--ferric citrate preparation, savita, and proteose peptone no. 3. Use of the former (126) increases the citrate and iron content of the medium.by unknown amounts. Augmenting the concen- trations of iron and citrate, however, does not satisfactorily substitute for savita in the medium. Difco yeast extract in the presence or absence of these increases also does not provide the growth enhancing effect of savita. More luxuriant growth of‘y. tuberculosis was obtained in liquid and agar media when asparagine was used in conjunction with proteose peptone no. 3, but growth was as rapid and as many isolations of this organism.were achieved in media without asparagine. As casitone supplies amino acids, including asparagine, for utilization by the tubercle bacillus, addition- al amounts are probably unnecessary unless growth of enormous numbers of bacilli is desired as is the case in the production of tuberculin. Use of proteose peptone no. 5 makes addition of trace amounts of zinc and calcium ions un- necessary. Therefore, the basal medium.is more easily prepared than Dubos-type media. The preparation of the classical egg media is still more difficult and time consuming. The desired (127) goal, simplicity in composition, has been achieved for the basal experimental medium. Care must be exercised in sterilization by filtration of enrichments, such as that used with A-l-S, and in their addition to the medium, in order to prevent contamination. However, the presence of the selective agents previously discussed provides some protection against accidental contamination. Nevertheless, use of activated charcoal in both liquid and solid experimental media would further facilitate their preparation. The A-l-S enrichment which was found to be more stimulatory to growth of tubercle bacilli than other undenatured albumin pre- parations studied provides detoxified oleic acid and glucose as carbon sources, the pro- tective properties of fraction V albumin, as well as the additional protection and growth enhancement of bovine serum. Subsequent modi- fied albumin enrichments (g.g_. the Sudan IV-- oleate complex) also contain all of these growth promoting substances. The use of ethyl violet as a selective (128) agent in media for the isolation of y, tubercu- losis, as was-suggested by Mallmann (65), permits inhibition of many potential contaminants with- out the toxicity for this organism of malachite green. Ethyl violet also serves as an indicator of mycobacterial growth. During studies with this dye, it was found that the petroleum.ether fraction of the lipid layer of tubercle bacilli combines with ethyl violet. It was heped that a relationship between virulence and the extent of ethyl violet combination could be detected, since a portion of this lipid fraction, known as the "cord factor”, has been held responsible for virulence of tubercle bacilli (26). However, saprophytic mycobacteria bound the dye to the same apparent extent as virulent strains of tubercle bacilli. The inhibitory effects of drug residuals in sputum specimens were overcome through neutralization with Bacto-penase. The abolition of chemical treatment of specimens or the use of milder digestants will increase the likeli- hood of isolation of E. tuberculosis from.those specimens, even when Bacto-penase is not present (129) in the medium. Because of the additional time and care required to add Bacto—penase to the sterile medium and since its use prevents inclu- sion of penicillin as selective agent, it is be- lieved that milder digestion is the more feasible procedure. Use of Bacto-penase in culture media might prove valuable when penicillin digestion is used, however, as it would serve to inacti- vate both residual streptomycin and penicillin carried over in the inoculum. The blood and egg experimental media were not developed beyond the status described for each because they were considered complex and relatively difficult to prepare, with some vari- ation in chemical content between lots. From these standpoints, the most promising media for isolation of !, tuberculosis develOped in these studies are the agars, A-l-S and C-l90-EV, and the following liquid preparations used in glass wool flasks: A-EV-l, AéNR-l, A-EVNR-l, and A- SIV-l. The relative worth of each of these media should be evaluated through use of all of them to culture a large series of sputa and other (150) specimens. With the use of more selective, less harsh digestion, each medium.ahould prove efficient in isolation of tubercle bacilli. Further modifications through combination of media ingredients are possible. For example, the use of neutral red or Sudan IV oleate in C-l90-EV should facilitate recognition of colonies of E. tuberculosis. Some of these studies have been undertaken in work now in progress. X Summary In preliminary studies, experimental media were developed and compared with the Dubos-Davis and Petragnani media by cultivating small inocula of laboratory stock strains of‘E. tuberculosis. Penicillin and malachite green were used as se- lective agents in the media. Relatively rapid growth of tubercle bacilli on primary isolation was secured in the agar E-22, and the liquid medium. 3-12. Growth was more rapid and profuse than on the Petragnani medium and nearly as rapid as that observed in the Dubos-Davis medium. Contamination was more extensive with the experimental media. (131) The Dubos-Davis medium.was contaminated more fre- quently than the experimental media. Least con- tamination occurred on Petragnani's. An experimental egg medium, EE-18, was de- ve10ped. Earlier appearance of growth and a greater number of positive cultures was obtained on EE-lB than on the Petragnani medium. An agar medium, A—l-S, supported faster growth of tubercle bacilli than the Dubos-Middlebrook oleate-albumin agar. The enrichment used in A-l-S supplies the protective prOperties of fraction V albumin, as well as the greater growth stimulation of bovine serum.and an oleate-albumin complex. This en- richment was developed as the result of turbido- metric and micro-Kjeldahl growth studies of several enrichments. A whole blood experimental medium, ' B-12, was found to be inferior to EE-18, A-l-S, E-22, the Dubos-Middlebrook and Petragnani media. An experimental liquid egg yolk medium, EY-5, supported submerged growth of small inocula of tubercle bacilli. Multiplication was nearly as rapid as was observed in media containing wetting agents. Because this medium was Opaque it was difficult to distinguish bacterial growth. (132) The addition of 1 ml of wetting agent-albumin solution to Petragnani slants stimulated growth of E. tuberculosis but increased contamination of cultures. Another solid medium, A-ZZ-S, containing asparagine, but no peptone, supported rapid and luxuriant growth of tubercle bacilli. The yeast- vegetable-ferric citrate product, savita, was substituted for the vegex present in earlier media. When Bacto-penase was included in the enrichment, it was possible to isolate tubercle bacilli more frequently from sputa of patients receiving strepto- mycin--para-amino salicylic therapy. Growth of ‘E. tuberculosis was more profuse on A-ZZ-S, but more A-l-S cultures were positive. Penicillin could not be employed as selective agent in A-22-S because of the presence of Bacto-penase. Malachite green was therefore always used. Tubercle bacilli were isolated sooner in A-ZO-L, the liquid version of A-22-S, but excessive contamination reduced the efficiency of this medium. Triton A-20 was found to stimulate the rate and enhance the amount of growth of‘g. tuberculosis (153) in liquid media to a greater degree than tween 80. The relative toxicity of ethyl violet and malachite green for a mixture of 5 recently iso- lated, virulent strains of E, tuberculosis var. hominis was determined by photometric Optical density measurements of a liquid medium.and on agar and egg solid media by use of the method of Youmans and YOumans. In the liquid medium.highest concentrations having minimum toxicities were: ethyl violet l to 800,000, and malachite green 1 to 1,h00,000. With the agar medium.these concen- trations were 1 to 600,000 and l to 900,000, respectively and with the egg medium 1 to 90,000 for both dyes. A new entity, the Optimum selective concentration, was determined for both dyes in each medium.through use of these media for iso- lation of.§. tuberculosis from chemically treated concentrates of sputa. These dye concentrations were those effectively inhibiting most of the undesirable organisms, but permitting nearly Optimum.rates of growth of E, tuberculosis. The The Optimum selective ethyl violet concentrations established were: for the liquid medium, 1 to (15h) 800,000; for the agar, l to 600,000; and for the egg, 1 to 80,000. For malachite green these Optimum selective concentrations were 1 to 1,100,- 000, l to 600,000, and l to 70,000. Generation times of the tubercle bacilli were much longer at the Optimum selective concentrations of mala- chite green. Five ethyl violet-sodium.azide experimental media, the Dubos-Davis medium, and the Loewenstein- Jensen, Petrik, Petragnani, and Peizer egg media were used in the primary isolation of‘g. tubercu- losis from sputum specimens. The portions Of the specimens cultured in ethyl violet-azide media were treated only with 1 ml of 10 per cent triton A-20, while the remainder of each specimen was digested with 3 per cent sodium hydroxide before concentration and inoculation Of the other media. The highest number Of positive cultures as well as the most rapid growth Of g. tuberculosis was ' obtained in ethyl violet-azide medium.A-EVA-l. The Petragnani egg was the best of this type of medium from the standpoints of early growth and number of positive cultures. Other egg media should be incubated for at least 8 weeks before (155) reporting cultural results. A much larger number of positive cultures was secured with ethyl violet- azide media from.sputa of patients receiving streptomycin--para-amino salicylic acid therapy. This may be due to the milder treatment of speci- mens and use in these media Of selective agents less toxic to‘g. tuberculosis. Growth Of this organism in the Dubos-Davis medium.was nearly as rapid as in the ethyl violet-azide media, but much greater contamination of the former medium made detection of typical aggregates difficult. streptococci and fungi multiplied in nearly all experimental media cultures. In the presence of ethyl violet, aggregates and colonies of tubercle bacilli are blue-violet. This typical color aided detection and differentiation Of growth. In some instances, extensive multiplication Of streptococci made it difficult to distinguish this color, be- cause streptococci form purple colonies in ethyl violet media. The inoculated sediment frequently also adsorbed the dye and masked the color Of tuber- cle bacilli aggregates. Therefore microscOpic confirmation of growth at intervals Of 2 days was required. Some difficulty was experienced in removing samples Of growth from the ethyl violet- (156) azide media because of reduced surface‘tension. Because Of the great amount Of time required to examine cultures microscOpically, it was concluded that routine use or ethyl violet-azide media with- out prior chemical digestion of specimens is imp practical. Tubercle bacilli could not be cultured on the Petragnani medium from sodium.hydroxide digested sputa containing 0.2 units of streptomycin per ml. The bacilli were isolated from specimens contain- ing 0.1 and 0.05 units of streptomycin per ml after this chemical treatment. In medium.A-EVA-l, tubercle bacilli grew from.undigested inocula containing 3.5 units of streptomycin per ml of the unconcentrated specimen, but not from.those containing 3.75 units per ml. Seeding Of sputa with known weights Of streptomycin sensitive tuber- cle bacilli confirmed these results, indicating that bacilli exposed to low concentrations Of streptomycin in specimens and subsequently to sodium hydroxide digestion fail to multiply on the Petragnani medium, while they will grow in A-EVA-l frmm non-digested inocula. During further study Of experimental media (157) it was found that 100 units of penicillin and 30 to 50 micrograms Of acti-dione per ml Of ethyl violet experimental media effectively suppressed contamination without detrimental effects on‘g. tuberculosis. The use Of sodium azide in ex- perimental media was discontinued because of its slight toxicity for this organism, Better cultur- al results were secured if the sputum.specimens were treated with a mild digestant before concen- tration. Tubercle bacilli grow in characteristic blue- violet colonies on the surface Of glass wool in ethyl violet culture media. Colonies are light pink in neutral red media and a more brilliant pink in Sudan IV oleate media. The formation of Sudan oleate-albumin complexes made it possible tO use these fat soluble dyes in media. Ethyl violet, neutral red, ethyl violet-- neutral red, Sudan IV oleate experimental media, and the Petragnani medium.were employed to isolate ‘3. tuberculosis from sputum specimens. One-half of each specimen was treated with a trisodium phosphate--triton A-20--penicillin digestion mixture; the remainder was digested with 3 per (138) cent sodium.hydroxide. Each sediment was used to inoculate a container of each medium, Colonies . Of tubercle bacilli were detected on the Sudan IV--Oleate medium seeded with the triton mixture treated sediments after an average Of 8 days incu- bation. Colonies from.sodium.hydroxide digested sediments were macroscOpically visible after an average of 12 days. With the neutral red medium. an average of 10 days was required to detect colo- nies Of g. tuberculosis from the triton treated inocula, while 15 days were required for the sodium. hydroxide digested sediments. Slower growth oc- curred with the 2 ethyl violet media. Visible colonies Of tubercle bacilli were perceived on Petragnani's after an average of 3h days. More contamination occurred in media without ethyl violet. Contamination was also greater in all media seeded with triton digested sediments. However, contamination of the experimental media was no greater than on the Petragnani medium.in- oculated with sodium.hydroxide digested sediments. Colonies Of g. tuberculosis were detected earlier on the Sudan IV oleate medium.because Of their distinctively vivid color and due to more rapid :multiplication of this organism. (139) Colonies of saprophytic mycobacteria can be differentiated from those Of g. tuberculosis On both neutral red and Sudan IV glass wool media. The degree of virulence of the latter organism is not indicated by colony color. The glass wool culture technique combines the advantages for cultivation Of tubercle bacilli of liquid and solid media without the disadvantages Of each. Its chief disadvantage is the difficulty experienced removing portions Of colonies from the glass fibers for microscOpic examination or.sub- culture. The stimulation of growth Of recently isolated strains of M. tuberculosis var. hominis andlg. tuberculosis var. gzigm produced by bovine albumin, oleic albumin, human serum, and h grades of acti- vated charcoal, as compared with growth in the basal agar medium, was demonstrated. Bovine a1- bumin produced the greatest stimulation. The effect of charcoal C-l90-N was equivalent to that Of human serum and oleic albumin. Charcoals C-llfi-N and WA also stimulated growth, but to a lesser extent. Aqua A did not alter the generationtime from.that on the basal medium, Studies of generation (m0) times of tubercle bacilli on agar media containing C-l90-N and varying concentrations of ethyl violet 6B, as well as culturing of 3 day pooled sputa on these media, established that a l to h00,000 concen- tration of this dye does not appreciably affect their growth rate but does aid in inhibiting growth Of undesirable bacteria. The C-l90-N agar, contain- ing 1 to h00,000 ethyl violet, 100 units penicillin per ml, and 100 micrograms of acti-dione per ml, in culturing pathological material, was found to be less frequently contaminated than the Petragnani medium, A greater and more rapid recovery of.g. tuberculosis was obtained with this charcoal medium, The activated charcoal probably adsorbs fatty acids and other substances toxic to tubercle bacilli, as do blood albumins, starch, and egg components, thus enhancing growth of small numbers of these organisms. The medium.is easily prepared, since sterilization is by autoclaving, and only the peni- cillin and acti-diOne need be added subsequently. (lhl) TABLE 1. The effect of dye concentrations on the growth Of tubercle bacilli. Optical densities of medium. after 16 days incubation. Ethyl violet Malachite green Dye Inoculum, Inoouhmm Gone. _1 _ _ _ _ .. 10 2 10 h’ 10 6 10 2 10 h 10 6 -- , , a 1 1-100,000 0.20h 0.1h9 0.069 0.018 0.015 0.007 1-500,000 0.611 0.551 0.u 6 0.0h1 0.0 0 0.052 1-600,000 0.7h5 0.658 0.2 9 0.0u6 0.0 1 0.022 1-700,000 0.7u5 0.210 0. 20 0.051 0.0h1 0.0 1 1-725,000 0.7h5 0. 99 0.620 1-750,000 0.7u5 0.721 0.629 1- 75,000 0.7u5 0.755 0.699 1- 00,000 0.7h5 0.7 5 0.755 0.051 0-0h9 0.0hh 1-825,000 0.7u5 0.7h5 0.755 1-850,000 0.7h5 0.7h5 0.755 . 1-900,000 0.7h5 0.7u5 0.721 0.05h 0.051 0.0u6 1-1,000,000 0-7h5 0.7h5 0.755 0.059 0.056 0.051 1.1,200,000 0.699 0.561 0.hg6 1-1,225,000 0.699 0.6h8 o.h 9 1-1,250,000 0.699 0.688 0.h89 1-1,275,000 0.710 0.6 9 0.552 1-1,500,000 0.7h5 0.710 0.59h 1.1,525,000 0.7h5 0.710 0.59u 1-1,550,000 0.7h5 0.710 0.59h 1-1, 75,000 0.7h5 0.721 ,0.602 1-1, 00,000 0.7h5 0.755 0.620 1-1.h25.000 0.7h5 0.755 0.620 TABLE 2. The effect of dye concentrations on the generation time, in hours, Of recently isolated tubercle bacilli. ‘_. t“ i MEDIUM Dye Malachite green Malachite green Ethyl violet Ethyl concentration Violet agar agar 988 988 1-5,000 1-10 ,000 1-20,000 1" O ’000 1- 0,000 1-50,000 1-60 ,000 1- 0,000 1-90 ,000 1-100,000 1-200,000 1" 00 ,000 1- 00,000 1‘500 ,000 1-600,000 1- 00,000 1- 00,000 1-900,000 1-1,000,000 1-1,100,000 l-l,200,000 Controls 0 -\DOH O e e 0 m0 O\G)GJN NF’F'm wrng-F-rwm F“Ok4 MoMH 0N eeooeeoeeeeeooeeeeeoeoeeeeeeee EN 0>H96WOQ HHI¢>mI< O>H9Hm0m O 0.000000000000000...000000.00. ¢N opfipw on Hld>m04 cpfiflfimom :0“. MH o.cocooooooeooeoooooooooo m|4a9m|¢ opfipdwog QHI¢>M|¢ mpfipfimom NeNm NP eeeeeeoeeeeeeeeeoeoee MH>GQIQOQSQ O>H9dwog QH|<>MI4 m>fipfim0m @000 m oeeoeeeeeoeeeeeeeeeeeeeeee MHHPOM 0>fipwm0n AHI<>MI4 0>fi¥fimom Oewm m H eeeeeeeeeeeeeeeeoeeeeeoeco HONfiom 0>dew0q QHI¢>WI< 0>fiPHmOm N.mw ma .............. nonnohuaaopnaowooq O>fipmwen .H|4>mud O>Hpfinom m. m m OOOOOOOOOOOOOOOOOOOOOO Hgdnmwhpom 0>flpdw0g qflld>m34 Opfipfimom HemN O eeeeeoocoooeeeeeeeoeeeeeeeeeeeeooeeoeeeoe “HUGE HHG OPHHHmom p300 HOW HOggfiz GHUOE PQ®H0%HHU so flpgofiw M0 COflHHdQEOO em mma .......................... meafiuHsO o>apmmoq mo gonads Hmpoe mMN eeeeeeeeeeeeeeeeeoeeeeeeee ”Ohspflzo 0>H9Hmog MO “09555 HNHOB mmm ...................... nonmapaooaoo o>Hpmmed mo gonads Hepoa NmH oeeeeeeeeeoeeeeeeeeeee ”OFGHPQOOCOO 0>Hpfimog MC MODESQ HNPOB mo: eeeeeoeeeeeeeeeoeeeeeeeeeoeeeeeeee MQOEHOCQQ HO HQQES“ H6909 mnOSHOOmn MO mfimhamam Hmaodom .4 .mneafioemm savage Soap manoadoaondp .2 MO noduwHOma hamafipm can :H waves wwo HmOHmanO can mfi>mmamopdm Spas mecca HmpaoEHAomxo MO donfiammSOo .m mqm¢a .xmmv .wm .mu .eaneaaa seam eeeaeee cape» no eaanm ewe mmm mom and ed” Hem emu emm mam Hem oopmnfiampdoo mens» HO nonanz nonumadadpdoo HO acmfinegaoo .n 2. la 9: :3 3A Hm Fe :m :mH mznmm mm med Hm em mm Hem cam emm mmm New mm-mm 4H mma mm Hm Hm mom ewe emu ,mmm 2mm mm-mm. m we on NH m: so we mHH ~:H_ we mmuma m me m .0H NH mm mm a: one we naum o m o o o m 4 :H mm m sue -aaemmmmmmeq -mwmmm aaanea aeaaea «neameanea am am mud>mua H-<>mua as-a. Awmmmv nOHSpHsO o>fipanom HO sonasz npsoam ho open HO aonfismgaoo .o nOdeHdoOnH wan meadow n» B BHHHQOO pow vac H ## 0.:H :H m.:H e m.HH m H.»H NH o.NH NH N.NH m m.e e :.HH m 9m m ma H E m Ha. um an em .mm m 0H NH mH NH mH NH mH o.NN NN m.e: mH o.mN mH m.Nm mN o.~maa>m e.Hm mm m.Hm :N men: Hm a . N.Hm HN e.Hm :N o.O: mN o.mH mH m.mN NH H.HN eH 0.0m HN OOH H: we op R .02 R .02 R .02 R .02 H->Hm-< , Humane Humz>m-a Husmua has nan» connocoam ewe: Ham poz # weapon ma mohfipafio andawmApem uaaQfiHUon condom» omid nopaau npaoaaoon mopnowac medz «noafipano cepdafiadpnoo HO .oz weapon ca nonwpaso anwnmdspom apnofichm condone owud nopanp endeawuon connomfio moez «uoamommd pane“ npwonw when .0: menobd Haddmmnuom no o>HpHnom weaken we .02 o>apamom updeaupom condos» omud confine O>Hpanom manaHUOn copmomfic MOdz haamcaaoomoaOHa o>aufimom poaSHHdo unoafioomm some: nxmmah Hook madam :H eaves HandoEHAomxo : mo nonahsqaou .anHpHodoo cannoawmao oddpaoa .4 mgmqa. .eH >.HN m.ON m.NN N ,NN mJON m,0N mJNN H spate aassn + Heaem . .. m.ON m.NN m m.eN m.mH e.:N .HN N m.NN m.om MJJN ream H dHEfinHw oda>on + Hammm .eH 0.:N m m.mH m.NN m.oN m,NN N H.m o.MN m,ON m.NN H aHsanHe eHeHo + Heaem m.m :. N m.mN N m.m .dJmN m.mm H 4 used + Hdmdm H.: e, N m m.m :.mN H.: m.eN N H.m 0.:N H.: QJNN H as 4 Hanan e.mH 0.:N m e.mH m.NN ~,eH o.nN N H.m 0.:N >.eH o.:N H zumHHuo i Heaem . . m.oN m.NN m w.MH m.NN m.om m.NN N e.mH m.NN m.0N m.NN H auoaH-o + Heeem . m.mN m :.eN m.mN N :.mN meN H Human anllsmaunaw 1ammmnawlli Hanan Hopo mason Hanan Hebo undom noapmazfiapn ma asap noapmdnaapn dd 08H» Heaps .EBHUoE :denoniob L M noapdnonom a i. .EfiH>d .ae>l. anHan .adp i.. mHmOHsannnp .2 mamoafiohonfip am .HHHHOdD cacaonsp ada>e cam amass no nnHenpn cepwaona hapqooon no npsoaw go even onu.do demeanoaano neononm m one nHmOcammo oopd>apom : no poomho 038 .m mnmda m NH m cm econ admawmapcm m mH OH OH OOO.OOm.H-H >m-OmH-O m mH » NH OOO.OOm.H-H >mumHHIO manohdoaonsp doadomam pampnoo QBOpwaopo Uopmnaampnoo .2 H09 pupae flpaoam poaoap . asficos eaHnHeea when dH esHp Hanan 1 omdae>H ’li“ e mhdoam opmapdooaoo HO cOprnflSme OHQOOOOHOHS he popoopoo one; HHHHOmp pnmwnmfiom moans HO : aw .mnoaaoomm Edusam OOHOOQ hep moan» mm wdaafipaso MO muaanom .m mnmdB TABLE 7. The rate of growth of recently isolated strains of‘y. tuberculosis var. hominis on charcoal agar with increasing concentrations Of ethyl violet. W 1} Ethyl violet ‘ Generation time Medium content Trial 7 in hours Cells-EV 1-1,500,000 1 25.h 2 25.h C-l90-EV 1-1,500,000 1 22.8 2 22.8 C-l90-EV 1—1,200,000 1 22.8 2 22.8 C-l90-EV 141,100,000 1 22.8 2 22.8 C-l90-EV 1-800,000 1 22.8 2 22.8 C-l90-EV 1-600,000 1 22.8 2 22.8 C-l90-EV 1-h00,000 1 2u.0 2 2h.0 C-l90-EV 1-200,000 1 51.2 2 52.6 dH Hm ; N anon admnwdapcm MH mm N OOO.OON.H-H >auomHuo :H mm m OOO.OOO.H-H Om-OOH-O m mN a OOO.OOO-H anuomHue m NH m OOO.OO¢-H >m-OOHsO N NH HH . OOO.OO:-H >m-OmH-O .mwnloenons» - naoawaopo oopwauadpnoo .H you unopnoo .asamos L oefipfimom pefloup thpm 1' I I .l‘ till I'lul It‘ll." 1|..- lll'll'ul‘l enfidmam OnGHpQOOQOO HO noapmnfisdxo OHQOOmOHOHa he copoopeo Once HHHHOOQ backspace seen: no m ma .ndoadocmn_fispnmm UOHOOQ men mean» N: wdaanpaso no anSnOm .m Hands (l) (2) (5) (h) (5) (6) (7) (8) (9) (10) Literature Cited Abbott, J. N. 1951 Effective use Of peni- cillin to reduce contamination in sputum. concentrates to be examined for tubercle bacilli. Am. J. Pub. Health, g1, 287-291. Anonymous 1953 Difco Manual, Difco Labo- ratories, Inc. Detroit 1, Mich. 9th Edition. p. 2650 Anonymous 1953 Difco Manual, Difco Labo- ratories, Inc. Detroit 1, Mich. 9th Edition. p. 283. Armand-Delille, P., Mayer, A., Schaeffer, G., and Terroine, E. 1913 Contribution a la biochimie des microorganismes. I Le bacille tuberculeux, culture en milieu chimiquement defini, nutrition azotie. J. Physiol. et Path. G6D., 3.2, 797-811. Bance, J. 19h2 Contribution a l'étude cmmparée du metabolisme des bacilles tuber- culeux, B.C.G., humain, bovin et aviaire cultivés sur milieu de Sauton. Arch. Inst. Pasteur, Tunis, 5;, 71-125 and 222-269. Besredka, A., and Jupilla, F. 1913 Le bouillon 3 l' oeuf. Ann Inst. Pastuer, g7, 1009-1017. ' u Boecker, E. 1923 Uber die submerse Vermehrung der Tuberkelbacillen in flfissigen Nghrbgden. Z. f. Hyg. u. Infektionskr. ‘99, 1.21-1.21}. Boissevain, C. H. 1926 A method for Obtain- ing single colonies Of tubercle bacilli. Am. RGVO T'ubeI'CO, 2.1, 90-910 Boissevain, C. H. 1928 Replacement Of potassium by other elements in culture mediums. J. Infect. Dis., g5, 1911-199. Boissevain, C. H. l9h0 Growth promotion of the tubercle bacillus by serum albumin. Proc. Soc. Expt'l. Biol. 8: Med., Lg, 110-112. (11) (12) (15) (1h) (15) (16) (17) (18) (19) (20) Bouveault, L. 1892 Etudes cliniques sur le bacille de la tuberculose aviaire. These, Faculté Mddecine, Paris. Bretey, J., and Andrejew, A. l9h9 Respi- ration des bacilles tuberculeux dans 1e milieu de Dubos. Tween 80. Compt. rend. soc. biol., gg5, 171-172. Buc, E. l92h Sur la croissance du bacille tuberculeux dans les milieux liquides. Rev. de la Tuberc., 5, 520-529. Byham, B. B. 1950 An evaluation Of certain culture media for tubercle bacilli. Am. J. Cline Pathe, g-Q, 678’6810 Calmette, A., Massol, L., and Breton, M. 1909 Milieux de culture pour le bacille tuberculeux. Compt. rend. soc. biol., él: 580'585 0 Corper, H. J., and Uyei, N. 1927 The cultivation Of tubercle bacilli. An imr proved method for isolation from tubercu- lous material. J. Lab. & Clin. Med., 1 , h69-u80. Grimm, P. D., and Martos, v. P. 19th Effects of amigen and amino acids on the rowth of tubercle bacilli. Am. Rev. Tuberc., is 914310140 Davies, R. 1937 The influence Of haemo- globin and ferrous ammonium.sulfate on the growth Of the tubercle bacillus. J. Path. and Bact-. hi. 773-777- Davis, B. D., and Dubos, R. J. l9h6 Inter- action Of serum albumin, free and esterified oleic acid and lipase in relation to culti- vation Of the tubercle bacillus. Arch. Biochem.,‘ll, 201-203. Dav13, B. DO, and Whoa, R. J. 19,47 The binding of fatty acids by serum albumin, a protective growth factor in bacteriologi- 031 media. Jo Expt'l. Made, §éy 215-2280 (21) (22) (25) (210 (25) (26) (27) (28) (29) (50) (51) (32) Davis, B. D., and Dubos, R. J. 19u8 The inhibitory effect of lipase on bacterial growth in media containing fatty acid esters. J. Bact., 55, 11-23. de Gora, Paul F. 19h2 Bovine tuberculosis. Its incidence in bone, joint and cervical lymph node lesions in Italy. Am. Rev. Tuberc., 42, 576-5850 Dubos, R. J. 19h5 Rapid and submerged growth of mycobacteria in liquid media. Proc. Soc. Expt'l Biol. Med.,.5§, 361-362. Dubos, R. J. 19h? The effect of lipids and serum albumin on bacterial growth. J. Expt']. blade, @5, 9‘22. Dubos, R. J. l9h8 Cellular structures and functions concerned in parasitism. Bact. Rev., 94.2., 173-19he Dubos, R. J. 19h9 Tuberculosis. Scient. £111., 181, 30-14.10 Dubos, R. J., and Davis, B. D. l9h6 Factors affecting the growth Of tubercle bacilli in liquid media. J. Expt'l Med., @5, h09-h25. Dubos, R. J., and Middlebrook, G. 19h7 Media for tubercle bacilli. Am Rev. Tuberc., 16.: 350‘313-5 - Dubos, R. J., and Middlebrook, G. 19h8 Cytochemical reaction Of virulent tubercle bacilli. Am. Rev. Tuberc., 58, 698-699. Edson, N. L. 1951 The intermediary metabo- lism Of the mycobacteria. -Bact. Rev., 15, 1h7-182. Eggerth, A. H. 1950 Growth of Mycobac- terium.tuberculosis in egg yolk mediums. Proc. SOC. Expt'l Biol. Med., 11’ Big-514.50 Emmart, E. W., and Seibert, F. B. l9h5 The effect Of tuberculous and sensitized sera and serum fractions on the develOp-. ment Of tubercles in the chorio-allantoic membrane of the chick. J. Immunol., 59, 1h3-160. (35) (51+) (55) (36) (57) (38) (39) (ho) (El) (142) Fallin, J. R., Patnode, R. A., and Hudgins, A. K. 1952 Toxic effect Of penicillin on the culture of tubercle bacilli. Tech. Bull. Reg. Med. Technol., _2_2_, 1111-1115. Foley, 0. E. 19h6 Submerged growth of tubercle bacilli from patholo ic material in Dubos medium. Proc. Soc. xpt'l Biol. blade, g, 298-5020 Foley, G. E. l9h7 Further Observations on the cultivation of tubercle bacilli from pathologic material in Dubos media. J. Lab. & Clin. Med., 5_2_, Biz-81,6. Frisch, Arthur W. 1952 Multiplication of the H37Rv strain Of M. tuberculosis on blood and its derivatives. Proc. Soc. Expt'l. Biol. Med., 15, 281-285. Frouin, A., and Guillaumie, M. 1923 Influence de la concentration du glucose sur le develOppement et le rendement en pOids du bacille tuberculeux. Compt. rend. soc. biol., 88, 1095-1096. Frouin, A., and Guillaumie, M. 1928 Culture du bacille tuberculeux sur milieux synthetiques. Essai de généralisation de la notion d'équilebre entre les 616menta constitutifs des milieux de culture. Ann. Inst. Pasteur, 42, 667-69h. ' Frublinger, B., and Bela, J. 1953 Para- amino salicylic acid in sputum. Am. Rev. TuberCe’ fl, I’ll—20 Gerundo, M. l9h7 Influence Of sulfasux- idine and succinic acid upon the tubercle bacillus. Am. Rev. Tuberc., 55, 552-555. Goldie, H. l9h7 Use Of Dubos medium for culture Of M. tuberculosis from sputum. Proc. Soc. Expt'l.BiOl. Med., 55, 210-212. Gray, D. F., Clarke, B. L., and Johnstone, W. E. 195M Detection of small numbers of tubercle bacilli in diagnosis. The lethal action of concentrating agents. Am. Rev. Tuberc., £2, 991-1001. (1+3) (Ah) (h5) (h6) (1+7) (h8) (1(9) (50) (51) Halpern, B., and Kirchheimer, W. F. l95h The effect of oxygenation and aeration on the growth pattern of mycobacteria. Bact. Proc 0 , 9’43’95 e Hammerschlag, A. 1891 Bakteriologische- chemische Untersuchungen uber Tuberkel- bacillen. Zentr. f. Klin. Med., gg, 9-16. Hanks, J. H., and Evans, B. 19h0 The action Of serum, cells and blood on acid- fast bacteria in vitro. II Failure to demonstrate bactericidal power in various bloods. The relative advantages Of a solid medium and Of liquid blood in de- monstrating the viability of bacilli in blood. Am. Rev. Tuberc., 5;, 620-636. Hassler, J. w. 19u1 Active Carbon the Modern Purifier.' Industrial Chemical Sales Division, West Virginia Pulp and Paper CO., New York, N. Y. hth edition. 1117-1113 - Henley, R. R. 1925 The influence Of iron on the growth Of the tubercle ba- cillus upon glycerinated beef broth. Am. Rev. Tuberc., 12, 286-259. Hudgins, A. K., and Patnode, R. A. 1952 Effect Of age of eggs on sensitivity of modified Loewenstein medium. Tech Bull. Reg. Med. Technol., 2;, 173-175. Hurivitz, 0., and Miller, J. B. 1950 Effect of Triton A-20 and pH value on the streptomycin sensitivity of a resistant strain of Mycobacterium tuberculosis. Ame RGVO Tuberc., fig) 91'980 Kendall, A. F., Day, A. A., and walker, A. W. 1920 The metabolism of bovine tuber- cle bacilli. Studies in acid-fast bacteria. XII. Jo InfeCto Digep‘gé’ 77‘8h0 Kirby, W. M. M., and Dubos, R. J. l9h7 Effect of penicillin on the tubercle bacillus in vitrO. Proc. Soc. Expt'l. B101. M6d., 66, 120‘1230 (52) (53) (5L1) (55) (56) (57) (58) (59) (60) (61) (62) Kirchner, 0. 1932 Die Leistungsfahigkeit die Tiefenkultur des Tuberkelbazillus bei Verwendung besonders geeingneter flussiger fiflhrdeen. Zentralbl. f. Bakt., ggg, h03- 12. KOCh, Re kulose. 1882 Die Aetiologie der Tuber- Berl klin. Wchnschr., 52, 221-230. ‘KOCh, R e kulose e 188h Die Aetiologie der Tuber- lfiitto a. do K318. Gesundh., _2,’ 1-880 Kolmer, J. A., and Bgerner, F. 19hl Approved Laboratory Technique. D. Appleton-Century CO., New York, N. Y. 5rd edition. u80. Kflhne, W. 1892 Erfahrungen fiber Albumosen und Peptone. z. r. Biol., 2 , l-hO; 508-52h. Larson, W. P., Cantwell, W. F., and Hartzell, T. B. 1919 The influence Of the surface tension Of the culture medium on the growth of bacteria. J. Infect. Dis., 2 , hl-h6. Larson, We PO, and. Montank, To A. 1925 The effect Of wetting on the pathogenicity and viability Of the tubercle bacillus. Proc. Soc. Expt'l Biol. & Med., 20, 229-232. Iey,H. L., and Muller, J. H. 19h6 0n the isolation from agar Of an inhibitor for Neisseria gonorrhoeae. J. Bact., 5g, h53-h60. Litsky, W., Mallmann, W. L., and Fifield, C. W. 1952 Ethyl violet. A selective dye for the isolation of gram-negative bacteria. Stain Techn01., 2 3 229-2320 Long, E. R. bacteria. 1922 The nutrition Of acid-fast Am. Rev. Tuberc., 5, 857-869. Long, E. R., and Finner, L. L. 1927 The relation of glycerol in culture media to the growth and chemical composition Of tUberCIO b801111. Am. 1261’. TuberCe’ ié, 525'529- (65) (6h) (65) (66) (67) (68) (69) (70) (71) (72) Long, E. R., and Seibert, F. B. 1926 The chemical composition of the active principle of tuberculin. I A non-protein mediwm suitable for the production of tuberculin in large quantity. Am. Rev. Tuberc., l3, 395'397- Mac Nabb, A. L. 19h5 The tubercle bacillus. In Diagnostic Procedures and Reagents. 'Ifierican Public Health Association, New York, N. Y. 2nd edition. 506-507. Mallmann, W. L. 19h8 Personal communication. Masucci, P. 1920 A note on the effect of amino acids on growth Of tubercle bacilli. J. Lab. & Clin. Med.‘é, 96-98. Merrill, M. H. 1931 Studies on carbon metabolism of organisms of the genus Mycobacterium. II Utilization of organic compounds in a synthetic medium. J. Bact., .22.: 561‘57LI-0 Middlebrook, G. 19h7 Media for the culti- vation of tubercle bacilli and their clini- cal and experimental applications. 19h? Proc. N. Y. State Assoc. of Public Health Lab., 2 , 28. Middlebrook, G. Dubos, R. J., and Pierce, 0. 19h? Virulence and morphological charac- teristics of mammalian tubercle bacilli. J. Expt'l. Med., 88, 175-18h. Mishulow, L., and Reavin, S. 19h0 Cultur- al diagnosis of tuberculosis using Bordet- Gengou and Loewenstein media. J. Lab. & Cline Med., 2 , 876‘8800 Molliard, M. 1920 Influence d' une dose reduite de potassium sur les caracteres physiologiques des Sterigmatocystis nigra. Compt. rend. acad. sc., $19, 9h94951. Mollov, M., Hill, T., and Oshinsky, A. 1950 The Dubos medium.in routine culti- vation of tubercle bacilli. Am. J. Clin. Path., g9, 1085-1089. (75) (71+) (75) (76) (77) (78) (79) (80) (81) (82) (83) l9h2 La croissance des cultures In Arch. sci. ind. Hermann Monod, J. microbiennes. & 016., Paris. Myrvik, Q. N., Weiser, R. S., and Agar, H. D. 1953 Lethal and cytologic effects Of lysozyme on tubercle bacilli. Am. Rev. Tuberc., 51, 217-231. Nocard, E. 1885 Recherches expgrmentales sur la tuberculose des Oiseaux. Compt. rend. SOC. b101., _2.’ 601-6050 Nocard, E., and Roux, E. 1887 Sur la culture du bacille de la tuberculose. Ann. Inst. Pasteur,‘l, 19-29. Pagel, w. 19h0 The effect or blood serum on the growth Of tubercle bacilli in the depth Of liquid media. J. Path. and Bact., 5Q, 111-119. Peizer, L. R., and Schecter, C. 1950 A new medium for bacteriologic diagnosis Of Egberculosis. Am. J. Clin. Path., g9, 682- 5. Petragnani, G. 1926 Alcune utili modi- ficzioni a1 mio terreno ed alia technica per 1' isolamento in cottura pura dei bacilli di Koch dogli escreati e da altri materiali tubercolari. Boli. d. Inst. sieroterap milanese,'5, 173-182. Petroff, S. A. 1915 Some cultural studies on the tubercle bacillus. Johns HOpkins Hosp. Bull., g8, 276-280. Pollock, M. R. gertussis on media without blood. 0 pt 0 Path., fig, 295-5070 Pratt, Darrell 1953 Growth inhibitory action Of OOpper for Mycobacterium.phlei. J. BaCte, _6ip 157-1590 Proca, G. 1933 Milieux liquides pour 1' isolement du bacille tuberculeux. rend. soc. biol., 11h, 280-282. 19b? The growth Of g. Brit 0 Compt. (8h) (85) (86) (87) (88) (89) (90) (91) (92) (95) Proskauer, B., and Beck, M. l89h Beitrgge zur Ernahrungs-physiologie des Tuberkel- b30111u30 Z. to Hyg., 218-p 128-1520 Roberts, E. G., Wallace, J. L., and Erlich, H. 1950 Methods for isolating tubercle bfiCillie Am. Rev. fiberco’ 91:, 565-5680 Roche, L. 19hh Influence des poussieres minérales siliceuses sur la culture du bacille de la tuberculose. Compt. rend. soc. biol., 158, 555-556. Sarnowiec, W. 1939 Untersuchungen fiber der Einfluss von Kulturfiltraten verschie- dener Bakterien auf das Wachstum von Tuber- kelbazillen in der Tiefenkultur. Zentralbl. r. Bakt., I Abt. 0rig., 1.112: 252-257. Sattler, T. H.“ and Youmans, G. P. l9h8 The effect Of tween 80", bovine albumin, glycerol, and glucose on the growth of M cobacterium tuberculosis var. hominis (1—_H37Rv_7——J—. . Bact., 58, 255-2115. Sauton, B. 1912 Sur la nutrition minérale du bacille tuberculeux. Compt. rend. Acad. Sc 0 , 3.5-5, 860-8610 Saz, A. K. l9h9 The effect of long chain compounds, particularly hydrocarbons, on the metabolism Of tubercle bacilli. Arch. Biochem., 88, 195-203. Schaefer, W. B. 19h8 Recherches sur la croissance du Mycobacterium.tuberculosis en culture homogene. Ann. Inst. Pasteur, 11$: (45848-65 - Schaefer, W. B., Marshak, A., and Burkhart, B. 19h9 The growth of Mycobacterium.tu- berculosis as a function Of its nutrients. 37138013., 18. 5’49‘565° Solotorovsky, M., Bugie, E. J., and Frost, B. M. 19h8 The effect Of penicillin on the growth Of Mycobacterium.tuberculosis in IDu.‘bost medium. To Baote’ SEE, 55i;"5;;0 (9h) (95) (96) (97) (98) (99) (100) (101) (102) Sperry, J. A., and Rettger, L. F. 1915 The behavior of bacteria towards purified animal and vegetable proteins. J. Biol. Chem.) £2: ALB-LL59- Stonebrink, F. 1952 Ontwikkeling van een voedingsbodem voor de primaire cultuur van M cobacterium tuberculosis var. bovis. Tijdschr. Diergeneesk.,'11, 13741EHT- Tarshis, M. S. 1952 Blood media for the cultivation Of Mycobacterium.tuberculosis. IV Effect of adding penicillin. J. Lab. Cline M6d.,'gQ, 628-6550 Tarshis, M. S. 1953 Blood media for the cultivation of Mypobacterium.tuberculosis V. Results with agar-agar basal medium and varying concentrations Of blood, glycerine and penicillin Am. J. Clin. Pathol., 2 , 661-6700 Tarshis, M. 8., and Frisch, A. W. 1951 Blood media for the cultivation of M co- bacterium tuberculosis. Am. J. Clin. Path., 2:, 101-115 . Tarshis, M. S., Kinsella, P. C., and Parker, M. V. 1953 Blood media for the cultivation Of M cobacterium.tuberculosis VII. Comparison of Blood agar-penicillin and Loewenstein-Jensen media under routine fliagnostic conditions. J. Bact.,‘éé, MAB- 52. Tiffeneau, M., and Mario, A. 1912 Sur diverses conditions de culture du bacille tuberculeux. Compt. rend. soc. biol., 12, h8‘500 Yoshida, N. 1938 Experimental studies on the growth of tubercle bacilli. I. On the growth promoting factors for acid-fast bacteria. Kekkaku, _18, 211-25. YOumans, A. 3., and YOumans, G. P. 1950 The effect of bovine plasma fractions on the growth of Mycobacterium.tuberculosis var. hominis. J. Bact., 5Q, 561-568. (105) (1011) (105) (106) (107) (108) (109) (110) Youmans, A. S., and Youmans G. P. 1953 Studies on the metabolism of Mycobacterium guberculosis. II. The effect of compounds related to the Kreb's tricarboxylic acid cycle on the growth of Mycobacterium.tubercu- 10313 var. hominis. Jo Baote’ fig 96'990 Youmans, A. S. and Youmans, G. P. 195M Studies on the metabolism of Mycobacterium tuberculosis. IV. The effect of fatty acids on the growth oflM. tuberculosis var. hominis. J. 83013., _61’ 751-7550 Youmans, A. S., and Youmans, G. P. 1958 Studies on the metabolism of Mycobacterium. tuberculosis. V. The effect of amino acids on the growth Of M. tuberculosis_var. hominis. J. Bact., §1,_73H¥737. Youmans, G. P., and Karlson, A. G. l9h7 Streptomycin sensitivity of tubercle ba- cilli. Studies on recently isolated tuber- cle bacilli and the develOpment Of re- sistance to streptomycin in vivo. Am. Rev. Tuberc., .55, 529-535. Youmans, G. P., and YOumans, A. S. l9h9 A method for the determination of the rate of growth of tubercle bacilli b the use of small inocula. J. Bact., 58, 2 7-255. Youmans, G. P., and Youmans, A. S. 1950 The growth of recently isolated strains Of Mycobacterium tuberculosis var. hominis in liquid media. J. Bact., 89, 5695572. Youmans, G. P., and Youmans, A. S. 1953 Studies on the metabolism of Mycobacterium tuberculosis. I. The effect Of carbo- hydrates and alcohols on the growth Of Mycobacterium tuberculosis var. hominis. J. Bact., Q5, 92-95. Volk, w. A., and Myrvik, Q. N. 1955 An explanation for the arithmetic linear growth Of mycobacteria. J. Bact., 58, 386-588 0 (111) (112) (115) (118) (115) (116) Wasiliew, A. P., and Johnson, L. W. 1951 Tween-albumin-Petragnani medium.for culti- vation of Mycobacterium tuberculosis. Tech. Bull, Reg. Med. Technol., g1, 109-112. Whalen, J. W. 1951 A comparative study Of various digestion agents used in the iso- lation of Mycobacterium tuberculosis from sputa. M.S. Thesis, Mich. State 0011. Willis, H. S. 1928 Laboratory Diagnosis and Experimental Methods in Tuberculosis. Charles C. Thomas, Springfield, Illinois. ist edition. 99-109. Williston, E. H., and Youmans G. P., 19u9 Factors affecting the sensitivity in vitro of tubercle bacilli to streptomycin. Am. Rev. Tuberc., 59, 336-352. Wright, A. E. 192M New methods for the study of the pathology and treatments of tuberculosis disease. Larcet, I, 218-221. Zukerman, A., and Rantz, L. A. 1951 The use Of a Dubos-type solid:nedium.contain- ing penicillin for the primary isolation Of M. figberculosis from human discharges. Am. Re. %b6PC.,__61.£, 518-521. oh I' (:1 1 u‘ \‘l J 143 03143