Upload
vantram
View
226
Download
0
Embed Size (px)
Citation preview
STUDIES ON CHROMOGENIC MYCOBACTERIA
by
Ralph W. Butler, M.Sc., Dip. Bact.
A thesis submltted to the Faculty of Graduate Studies and Research in partial fulfillment of the requirements for the degree of Doctor of Philosophy.
Department of Bacteriology and lmmunology McGill University Montreal August 1965
ACKNOWLEDGEMENTS
The candidate is grateful to Dr. G. G. Kalz who directed this
investigation. Her inspiring counsel and guidance helped during the entire
period of research.
Gratitude is extended to Dr. R. W. Reed, Chairman of the Depart-
ment of Bacteriology and lmmunology for permitting this work to be carried
out in the laboratories of this department.
Dr. J .E. Josephson, Director of Laboratories, Department of Health,
Newfoundland, kindly arranged for special leave from duty. His encourage-
ment and interest is appreciated.
Sorne of the unclassified strains of mycobacterie investigated were
kindly provided by Dr. E.H. Runyon, Veterans Administration Hospital, Salt
LokeCity, Utahi Dr. C.O. Siebenmann, ConnaughtMedical Research Laboro-
tories, University of Torontoi and Dr. E. Mankiewicz, Royal Edward Lourentian
Hosp i ta 1, Montree 1 , Q uebec .
Financiol assistance and leove gronted by the Deportment of Heolth,
Newfoundland, is gratefully ocknowledged. The project was supported in port
by a Federal Heolth Grant and by the Canadien Medical Research Council.
TABLE OF CONTENTS
1. INTRODUCTION AND PURPOSE
Il. HISTORICAL REVIEW
1. SIGNIFICANCE, GEOGRAPHICAL DISTRIBUTION, INCI-
DENCE AND EPIDEMIOLOGY OF CHROMOGENIC
Page
MYCOBACTERIA . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
A. Chromogenic mycobacteria prior to 1950 . . . . . . . . . 6
B. Clinical significance during the post decade........... 6
C. Geograph ical distribution and incidence . . . . . . . . . . . . . . 7
D. Epidemiological aspects . . . . . . . . . . . . . . . . . . . . . . . . . . 9
a) Soil, water, milk and house dust . . . . . . . . . . . . . . . . . 10
b} Animais other thon man . . . . . . . . . . . . . . . . . . . . . . . 11
c) Man . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
E. Origin of chromogenic mycobacteria . . . . . . . . . . . . . . 12
2. CLASSIFICATION OF CHROMOGENIC MYCOBACTERIA 15
3. METHODS PROPOSED FOR THE DIFFERENTIATION OF
CHROMOGENIC MYCOBACTERIA FROM M. TUBERCULO-
SIS, AND FOR THE DETECTION OF SPECIFIC VARIETIES ---,..-
AND SUB-GROUPS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
A. Animal virulence studies . . . . . . . . . . . . . . . . . . . . . 21
B. Antigenic analysis . . . . . . . . . . . . . . . . . . . . . . 22
C. Resistance to drugs and other agents . . . . . . . . . . . . . . . 25
D. Bacteriophage typing . . . . . . . . . . . . . . . . . 26
Page
E. Growth at various temperatures . . . . . . . . . . . . . . . . . . . . . 27
F. 11 Cord11 factor and neutral red test . . . . . . . . . . . . . . . . . 27
G. Lipid content 28
H. Tissue culture 29
1. Morphology and staining . . . . . . . . . . . . . . . . . . . . . . . . . 29
J. Eh potential . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
K. Adansonian classification (Eiectronic computer) . . . . . . . . 31
l. Biochemical tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
M. Miscelloneous tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
N. Characterization of pigments in mycobacterie . . . . . . . . 37
a) M. tuberculosis 37
b) Saprophytes (M.phlei, M.smegmatis, M.lacticola,etc.) 37
c) Unclassified slow growing pathogenic mycobacterie 39
4. FORMATION OF PIGMENT IN MYCOBACTERIA . . . . . . . . 41
A. Introduction . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . 41
B. Oxygen tension . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
C. Temperature of incubation . . . . . . . . . . . . . . . . . . . . . . . . 42
D. pH of medium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
E. Composition of medium including the presence of trace elements and drugs . . . . . . . . . . . . . . . . . . . . . . . 44
F. Exposure to 1 ight . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
G. Exposure to mycobacteriophage . . . . . . . . . . . . . . . . . . . 48
H. Specifie inhibitors of pigment synthesis . . . . . . . . . . . . . . . 49
Ill. MATERIALS AND METHODS
1 . MATERIALS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
A. Strains of mycobacterie . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
B. Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
C. Chemicals and reagents . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
2. METHODS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
A. Pretreatment and culture of clinical specimens . . . . . . . . 54
B. Identification of M. tuberculosis and chromogenic mycobacterie . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
C . Growth of M. kansasi i for pigment stud i es . . . . . . . . . . . 58
D. Pigment extraction and partition . . . . . . . . . . . . . . . . . . . . 59
E. Th in-1 ayer chromatography . . . . . . . . . . . . . . . . . . . . . . . . . 64
F. Spectrophotometric analysis
IV. EXPERIMENTAL RESULTS
PART 1
STUDIES ON CHROMOGENIC MYCOBACTERIA ISOLATED
FROM CLINICAL MATERIAL
71
1. GENERAL ASPECTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
A. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
B. Incidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
C. Age and sex of patients . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
D. Type of specimens yielding chromogenic mycobacterie 75
E. Rate of growth on primary isolation . . . . . . . . . . . . . . . . . . 77
F. Animal virulence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
G. Sensitivity to antituberculous agents . . . . . . . . . . . . . . . . . 78
2. RECOGNITION AND GROUPING OF CHROMOGENIC
MYCOBACTERIA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
A. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
B. Inconstant pigmentation of individual strains . . . . . . . . . . 80
a) lsolates from clinical specimens . . . . . . . . . . . . . . . 80
b) Stock cultures . . . . . . . . . . . . . . . . . . . . . . . . . . 82
C. A special problem with nonphotochromogens . . . . . . . . . 86
D. A simple reliable method for the detection of ali chromogenic mycobacterie . . . . . . . . . . . . . . . 89
3. GROWTH STUD lES .. .. .. . . .. .. .. .. .. .. . .. .. .. . .. .. . 91
A. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
B. Growth in fluid media .. .. .. . .. . .. . . .. .. .. . .. .. .. .. 92
a) Dubos medium 92
b) Sauton medium 97
C. Effect of pH on growth of M. kansasi i . . . . . . . . . . . . . . 99
D. Unidentified growth stimulating factor for M. kansasii 103
4. MORPHO LO GICAL OBSERVATIONS . . . . . . . . . . . . . . . . . 105
A. Pigmented photochromogens as compared with non-pigmented photochromogens . . . . . . . . . . . . . . . . . . . . . 105
B. 11 Aibino11 nonphotochromogenic growth of M. kansasii 107
c. 11 Cord11 formation in chromogen ic mycobacteria . . . . . . . 109
PART Il
PIGMENT STUDIES
1. INVESTIGATIONS ON FORMATION OF PIGMENT
IN CHROMOGENJC MYCOBACTERIA . . . . . . . . . . . . 111
A. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 1
B . Moi sture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1 2
C. Oxygen ... .. .. .. ..... ... .. .. .. .. ..... ..... .. 118
D . Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
E. Age and viabi 1 ity of ce lis .. . . . . . . . . . . . . . . . . . . . . . 128
F. pH . . . . .. . . . . . . . . . . . . . . . . . . .. . . . . . . . . . .. . . . . . . . . . . . 13:>
G . Med i u m . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . lll
2. 11ALBIN0 11 AND 11 SCOTOCHROMOGENIC11 GROWTH
OF M. KANSAS li .. . .. . . .. . .. .. .. .. . . . . .. .. .. .. .. . . 131
3. CHARACTERIZATION OF PIGMENTS . . . . . . . . . . 132
A. Extraction and partition . . . . . . . . . . . . . . . . . . 132
B. Quantitative determination of carotenoids . . . . . . . . 135
C. Thin-layer chromatography and spectrophotometric analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
V. DISCUSSION AND CONCLUSIONS .. .. .. .. .. .. .. .. .. .. 158
VI. SUMMARY .. . ... ... .... .. . ... ... .......... ... .. .... .. .. 181
VIl. CONTRIBUTION TO KNOWLEDGE AND CLAIM TO
ORIGINALITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
VIII. APPENDICES ..............................
A. Media ............................................ .
IX.
B.
c.
Rea gents
Procedures
BIBLIOGRAPHY
v
vi
ix
1. INTRODUCTION AND PURPOSE
INTRODUCTION AND PURPOSE
The term 11 chromogenic mycobacterie" is but one of severa! commonly
used to describe ac id-fast bac ill i which differ from M. tuberculosis, and which
for the past ten years have been isolated from clinical specimens with steadily
increasing frequency. Additional terms which have been used in recent scientific
literature and which are usually, but not necessarily synonymous with "chromogenic
mycobacterie", include: 11atypicaP1 1 11anonymous" ,nparatubercle11 1
11 unclassified 11 1
11 nontuberculous acid-fast ba ci Il i" 1 "unidentified mycobacteriar 1 and "MOTT bac i Il i"
(mycobacteria-other-than tubercle-bacilli). The terms 11yellow bacilli", "orange
bacilliu, and 11 Battey11 refer to specifie varieties within the general group. The
current trend in terminology favars 11 unclassified11 as suggested by the American
Thoracic Society (1961), but the American Review of Respiratory Diseases in recmt
months {1964} has indicated its intention to use only the term "atypical 11 ln this
thesis 11 chromogenic mycobacteria 11 may be considered synonymous with '1unclassified 11 1
11atypi cal" and "anonymousn.
Of primary importance is the fact that certain members of this poorly
defined group can cause disease in man which is clinically, radiologically and patho-
logically indistinguishable from tuberculosis. A positive diagnosis of such infections,
however 1 is most difficult because indistinguishable strains are often isolated from
healthy 1 symptomless persans.
The types of clinical specimens from which chromogenic mycobacterie
have been isolated include: sputum, urine, faeces, exudates from various types of
lesions, skin, tonsils, ear and nasal secretions, mucous membranes, lymph nodes,
- 2-
synovial fluid, bone, spleen, cerebrospinal fluid and blood. Obviously 1 signifi-
cance is attached to isolations from cerebrospinal fluid, blood,or closed lesions
and abscesses; however 1 by far the most commonly encountered isolations are from
pulmonary infections where, in the absence of other recognizable pathogenic organ-
isms, the presence of chromogenic mycobacterie in sputum cannot be relied upon as
positive diagnosis. The same situation exists when members of the unclassified my-
cobacterie are isolated from urine specimens in suspected cases of renal tubercul osis.
Considerable interest has been aroused in these bacterie, especially with
regard to the ir exact relation to true human tubercle bacill i 1 the ir pathogenicity 1
and their sensitivity to antituberculous therapeutic agents.
The candidate flrst became interested in this particular group of myco-
bacterie during 1957 when a chromogenic strain was isolated from a sputum specimen
of a patient with pulmonary disease. On only three previous occasions, ali during
1953, had pigmented mycobacterie been isolated from clinical specimens in the Pro-
vince of Newfoundland. A careful search of the literature in 1957 revealed thot
relatively little was known ab.:>ut slow growin~ chromogenic mycobacterie. Further-
more, no agreement existed among bacteriologists concerning the cl inical significance
and the classification of these organisms. lt was apparent thot most workers ignored
such isolations, especially when made from sputum and urine specimens, on the basis
of their being 11Something other thon M. tuberculosis". ln instances where microscopie
examination of cultural growth revealed the presence of acid-fast bacilli, a few
workers had carried out animal pathogenicity studies. At this level, the conscientious
few were frequently misled, since these mycobacteria are nonpathogenic for the guinea
- 3-
pig. The old concept that the guinea pig parallels man in susceptibility to patho-
genic mycobacterie is still difficult for many to reject despite the fact that chromo-
genie mycobacterie are now known to cause tuberculosis-like disease in man.
ln Newfoundland 1 following the finding of chromogenic mycobacterie
in a clinical specimen, the routine procedure for reading TB cultures was modified to
facilitate a thorough search for these organisms. Almost immediately additional
strains were found. During the years 1957- 1960 unclassified mycobacterie were
isolated from 36 of 35,555 clinical specimens cultured for M. tuberculosis in the
Provincial Public Health Laboratories. The majority of isolations came from patients
suspected of having tuberculosis and from whom repeated attempts to isolate typical
tubercle bacilli had failed.
With increasing experience it became apparent that these organisms could
be easily recognized. The difficulty then arose concerning their classification and
clinical significance. Although some workers suggested that certain specifie varietles
of chromogenic mycobacterie possessed greater clinical significance 1 there were no
reliable methods available to permit strain recognition.
A scheme for the grouping of unclassified mycobacterie was suggested by
Runyon in 1959. lt was based upon the growth rate and the ability (or inability) of
strains to produce pigment under specified conditions. Runyon introduced the terms
photochromogen, scotochromogen 1 nonphotochromogen, and rapid growers for his
Groups 11 11 1 Ill and lV respectively.
At the time this research was started the author had already isolated over
50 strains of chromogenic mycobacterie from patients suspected of having tuberculosis.
-4-
Cultural studies had been undertaken in an effort to group these isola tes but it was
frequently experienced thot repeat 1 or duplicata cultures, under apparently identi-
cal conditions, displayed such marked differences as to affect their grouping within
Runyon 's scheme. ldentical experiences had been reported by others and it was
personally felt that the classification suggested was ineffectual 1 and even potentially
misleading.
ln the absence of specifie biochemical, immunological or pathogenicity
tests to differentiate members of the chromogenic mycobacterie 1 this investigation was
undertaken to determine the cause 1 or causes responsible for the variation so frequent! y
encountered in the pigmentation of these bacterie. To facilitate diagnosis it is
naturally essentiel thot bacteriologists be able to establ ish the identity of strains iso-
lated from clinical specimens and also to be able to determine whether or not repeat
isolations from an individuel patient are the same strain. lt was also initially intended
to extract and characterize the major pigments of these organisms.
Just prior to these studies the Subcommittee on Mycobacterie 1 American
Society for Microbiology {1962), decreed thot photochromogens {Runyon 1s Group 1)
formed a homogenous group regardless of their geographical occurrence. They desig-
nated the species1 nome M. kansas ii for ali photochromogens previously referred to as
Group 1, 11Yellow Bacilli 11 , and M. luciflavum. This species was usually associated
with disease processes and appeared to be a distinct pathogen.
M. kansasii, because of its clinical significance 1 was chosen for most of
the present studies. Fortunately 1 chromogenesis is best studied by employing photo-
chromogens. Pigment formation occurs rapidly 1 following exposure of a mature culture
- 5-
to light. The type species ATCC 12478 (Bostrom) has been used routinely 1 and in
many instances was paralleled by representatives of photochromogens 1 scotochromo-
gens and nonphotochromogens kindly supplied by Dr. Runyon.
Il. HISTORICAL REVIEW
-6-
H ISTORICAL REVIEW
1. SIGNIFICANCE, GEOGRAPHICAL DISTRIBUTION 1 INCIDENCE AND
EPIDEMIOLOGY OF CHROMOGENIC MYCOBACTERIA.
A. Chromogenic mycobacterie prior to 1950
ln 1891, Jess thon ten yeors following the isolation of M. tuberculosts
by Koch 1 Straus and Gamoleyo reported on the chromogenic nature of certain
tubercle bocilli (cited from Xoloborder, 1961). A review of the literoture
indicotes thot prior to 1950 at leost 66 investigotors reported the isolation of one
or more mycobocterio, pothogenic for mon but not for the guinea pig (Xalaborder,
1961).
B. Clinical significonce during the past decade
The year 1952 morked the beginning of a keen ard growing interest in
chromogeni c mycobacterie. ln thot year Tarshis and Frisch (1952 a, b 1 c) reported
cultural, pathological and hypersensitivity studies on twenty-six stroins of chromo-
genie mycobocteria isolated from patients with, or suspected of having tubercu-
losis. The patients from whom these organisms had been isolated were scattered
throughout the United States in Minnesota, Washington, Oregon, Californie and
New York. Also in the same year Prissick and Masson, in Montreal, released a
report on fourteen stroins of pigmented mycobocterio, ten of which hod been ob-
tai ned in pure cul ture from pus aspiroted from suppura ting facial, submoxi llory or
cervical lymph nodes in children (1952i olso 1956 and 1957). ln ali their cases
the cl inical pic ture suggested a tuberculous infection. Because the orgonisms were
in pure culture from closed tuberculous-like lesions they were considered to be the
-7-
actual cause of the lymphadenitis. Since then others have demonstrated unclassi-
fied mycobacterie as the causative agents in cervical adenitis (Weed et al. 1 1956;
Runyon, 1959a; Davis and Comstock, 1961; Marsdenand Hyde, 1962; Hsu, 1962;
Chapmanand Guy,1958, 1959; Noufflardetal., 1961; Wolinsky 1 1963; Gerszten
et al. 1 1964). The role of unclassified mycobacterie in pulmonary tuberculosis-1 ike
disease has 1 ikewise been esta bi ished sin ce the organisms have frequent! y been iso-
lated from resected lung tissue following the Isolation of identical strains over long
periods of time prior to surgery (Buhler and Pollak, 1953; Timpe and Runyon, 1954;
Nassau and Hamilton, 1957; Huppertetal., 1957; Miyamotoetal., 1959; Corpe
and lr.Iing, 1960; Lester etal., 1962;and Lawetal., 1963).
The increased interest in mycobacterie other thon tubercle bacilli has
also revealed that the prevlously recognized and classified M. fortuitum, considered
to be a saprophytic mycobacterium, can also cause infection in man (Wells et al. 1
1955; Gordon and Smith, 1955; Prather, 1960; Corpe et al., 1961; Wa'lle, 1961;
Hartwig et al., 1962; and Dross et al. 1 1964).
Three separate reports support the view thot unclassified mycobacterie may
be etiological agents in cases of sarcoidosis (Chapman, 1961; James, 1961; Mankiewicz
1963). The atypical acid-fast bacilli have also been implicated as the causative agent
in cat scratch fever (Boyd and Craig, 1961; Reikesand Washington, 1962} ..
With present knowledge 1 infection due to unclassified mycobacterie cannot
be differentiated from tuberculosis by any means other thon bacteriological studies.
C. Geographical distribution and incidence
From reports to date 1 a world wide distribution of these organisms is indica-
- 8-
ted. Significant numbers of cases of human disease have been identified in the
United States, in Wales1 the Netherlands, Spain, Remania, Malta, Australie and
Jamaica. lsolated cases of pulmonary disease caused by these organisms have been
reported a Iso from Sweden 1 Finland, France 1 Switzerland, Peru and Canada (Lewis
et al., 1960). Although there may weil be areas of high or law incidence,
sufficient valid information is not available upon which to base an opinion.
Experienced workers, specifically looking for these organisms, have
reported the isolation of atypical mycobacterie from the sputum of approximately 2
per cent of admissions to tuberculosis hospitals in Floride (Lewis et al., 1960).
Approximately 1 percent of 15,180 patients admitted to tuberculosis hospitals in
Georgie during the ten year period 1951 - 1961 had lesions associated with these
organisms (Crow et al., 1961). Chaves in New York City (1960) reported that 8.9
percent of the specimens positive for acid-fast bacilli were atypical, whlle Keltz
et al. in Illinois (1958) reviewed 649 culture-positive consecutive cases and found
164 positive for atypical mycobacterie. This represented an incidence of slightly more
than 25 percent. Dunbar et al., (1963) released figures from the Floride State Board
of Health stating that in one year 28 per cent of the cultures isolated at the Central
Laboratories proved to be mycobacterie ether than M. tuberculosis. Gerszten et al.,
(1963) after a four year study said that the finding of unclassified mycobacterie in
Virginie (5%) was midway between that found in the South (10%) and in the North (2%).
Mercks et al., at the Maya Cl inic (1963) a Iso claimed that 5 per cent of the ir cases
were caused by unclassified mycobacterie. Youmans (1963) in quoting Lester et al.,
(1958) cited an interesting epidemiological study of 49 cases due toM. kansasii.
-9-
These workers found thot the incidence of infection with M. kansasii in the Chicago
suburbs of Cicero, Berwyn, and Oak Park, with a total population of 184,129 was
1 O. 7 per cent 1 whereas the incidence of infection in the remainder of Cook County
with a population of 940,000 was only 1.4 per cent. From Texas, LeMaistre (1963)
reported thot 9 per cent of ali newly discovered mycobacterial disease in Dallas was
due to M. kansasi i .
Foreman (1962) of the Central Tuberculosis Laboratory in Wales felt thot 1
compared with pulmonary tuberculosis due to the tubercle bacillus, disease due to
anonymous mycobacterie was a rority. ln Wales, in the ten years from 1950- 1960
only 59 cases could be traced. This figure is fewer thon 6 cases per year in a popu-
lation of 2 million. Besta {1959) cited the incidence in Jtaly as 0.5 percent.
Apart from the author's work, figures concerning the incidence of un-
classified mycobacterie in Canada ore limited to one report. Mankiewicz {1958)
reported thot over a three year period approximately 4 per cent of the cultures positive
for acid-fast microorganisms were chromogenic. Her study was performed at the Royal
Edward Laurentian Hospital, Montreal. The au thor (Butler and J,:>sephson, 1963} 1
culturing 35,555 specimens during the four year period 1957- 1960, found 2.1 per
cent of the positive cultures examined at the Newfoundland Provincial Public Health
Laboratory to be of the unclassified varieties. Figures for the years 1961 - 1964 ore
included in the body of this thesis.
D. ~idemiological aspects
Epidemiological studies carried out by severa! workers have failed to show
any evidence of contagiousness of infections by the unclassified mycobacterie (Lewis
et al., 1959; Nassau and Hamilton, 1957; Runyon, 1959a; Kubica et al., 1961;
-10-
Crow et al., 1961). A single instance of conjugal disease has been reported
(Beek et al. 1 1963).
What is known concerning the distribution of these organisms is pre-
sented in the fol lowing paragraphs.
a) Soil, water 1 milk and house dust
Kubica et al. 1 {1961) undertook a study of 1200 samples of soil and
water. Preliminary results released on 452 samples in Georgie revealed thot more
thon 45 percent of the samples yielded acid-fast bacilli. They identified photo-
chromogens, scotochromogens, and 11 Battey'' strains, as weil as members of Group IV
rapid growers. Later (1963) they reported thot the marked similarity of the undassi-
fied acid-fast bacilli isolated from soil and animais strengthened the belief thot
these agents of disease in man and lower animais originated in soil. The ubiquity
of the organisms in nature suggests thot they are relatively nonagressive. ln Austral ia
large numbers of atypical acid-fast boeil li were found in swimming-pools, beach sand 1
water tanks, and in mud from dams and creeks (Singer and Rodda 1 1963; Singer, 1964).
Prather et al. 1 (1961) have isolated unclassified mycobacterie from house dust 1 hospi-
tal dust 1 soil 1 vegetables and top water. Others a Iso reported isolations from top
water 1 house dust and soil (Pellman and Runyon, 1964; Jefferies, 1963; Rodda and
Singer 1 1963).
Chapman et al., (1965) in Texas found 261 of 270 samples of raw mi lk to
contain mycobacterie. Representatives of Runyon 1s Groups Il, JI( and IV were identi-
fied among the iso lotes. Biochemi cal, serological, and biological tests suggested thot
a small minority of the organisms shared certain characteristics with M. kansasii but the
- 11 -
findings were inconsistent and inadequate for more precise identification. Also in
Texas, Jones and Jenkins (1965) isolated 101 strains of mycobacteria from 77 of 92
samples of soli. They failed to isolate photochromogens but did find representatives
of Groups Il 1 Ill and IV.
b) Animais other thon man
Anonymous mycobacteria have been isolated from dogs in Japon (Toda et
al. 1 1960) and from swine in California (Froman et al., 1961). The strains isolated
from swine were indistinguishable from the 11Battey" type 1 and it was suggested thot
swine may serve as potential reservoirs for human infection. Scammon et al. 1 {1963)
also isolated Group Ill organisms from swine.
ln a study of the distribution and significance of mycobacteria isolated
from cattle and swine in the United States, 2,244 tissue specimens of cattle and 172
tissue specimens of swine were cultured. Ali were negative for Group l organisms,
but mycobacteria of Groups IJ, Ill and IV 1 as weil as M. avium and M. bovis were
not uncommon isolates (Ellis and Yoder, 1964). Others had earller isolated members
of ali four Runyon Groups from cattle, and had found sorne of the Group Ill organisms
to be more virulent thon those of human origin {Mali man et al. 1 1962). Rodda and
Singer (1963) reported on anonymous mycobacterla isolated from cattle, sheep, pigs,
rats, birds, a toad and a fish.
c} Man
lt is possible thot man himself may act as a symptomless healthy carrier of
these organisms occasionally developing clinical disease when body resistance is low-
ered. ln support of this theory are the reports on the isolation of unclassified myco-
bacteria from healthy individuals (Edwards and Palmer, 1959i Atwell and Pratt, 1960;
- 12 -
Pra th er et a 1 1 1961; Stewart 1 1962).
E. Origin of chromogenic mycobacterie
Several possibilities have been suggested concerning the origin of the
chromogenic mycobacterie. Thus far epidemiological studies coupled with lobera-
tory investigations have not yielded satisfactory answers. One must accept the fact
thot scores of recorded cases of human infection involving mycobacterie thot differed
from Koch 1s bacillus either in cultural morphology or animal virulence, occurred
prior to 1950. lndeed sorne of the early isolates were so weil described thot one
cannet consider them as differing from the chromogens encountered today. We lack,
however 1 a satisfactory explanation for the apparent increase in the incidence of
mycobacterial infections caused by these organisms durlng the post decade.
lt has been suggested thot the chromogenic mycobacterie are variants of
previously existing forms, dissociation occurring either spontaneously 1 or due to the
influence of drug therapy or bacteriophage. As yet there is inconclusive evidence to
faveur either of these as the sole answer 1 however 1 there is sufficient evidence to
suggest each as possible.
Dissociation has been observed in mycobacteria by many bacteriologists.
Petroff reported dissociation of both tubercle ba ci Il i and the BCG strain in 1927 and
1929 (Petroff 1 1927a 1 b; Petroff et al. 1 1929). Orange colonies were isolated from
H37 human tubercle boeil li (Petroff and Steenken, 1930). (See also Miller, 1931;
Pinner, 1935a1 b.) Winn and Petroff {1933) studied dissociated avion bacilli which
had been accidenta li y exposed to increased temperature. Colonies became chromo-
genie and were found to have lost their pathogenicity for chickens.
- 13-
A theory upheld by sorne workers (Burrows and Barclay 1 1959; Tarshis,
1958, 1960, 1962; and Sweany, 1961) is thot antituberculous chemotherapy has
increased the frequency of isolation of chromogenic mycobacterie. Tarshis has
produced chromogenic mycobacteria from H37 Rv organisms by culturing them for
prolonged periods in the presence of either streptomycin or isoniazid. The variants,
on the basis of cultural, pathological, allergenic and antimicrobial susceptibility
studies, appeared genetically related to the parent organisms. Other workers have
reported similar observations {Tirunarayanan et al., 1959). ln a survey, including
tuberculous patients who had received, or were receiving, streptomycin therapy 1
Tarshis isolated chromogenic mycobacterie from 19 per cent of 405 patients compared
to a control series of 203 nontuberculous patients of whom less thon 1 per cent yielded
chromogens. Runyon {1959b) does not accept the drug theory as a source of chromo-
gens; however 1 he points out thot because chromogens are drug reslstant1 they are
favoured by chemotherapy whlch removes other species. This would be a situation
similar to thot of Candida and drug resistant staphylococci which emerge in large
numbers following use of broad spectrum antimicrobial agents.
lt has also been noted thot drug-resistant strains of M. tuberculosis isolated
from patients receiving para-aminosalicylic acid (PAS) and isoniazid (INH) exhibit
either a partial or an absolute growth requirement for oleic ac id. ln this respect 1
Groups 1, Il and Ill of the unclassified mycobacterie resemble drug-resistant strains
of tubercle bacilli more closely thon normal drug-susceptible strains (Hedgecock, 1958;
1962).
Recent studies with mycobacteriophage have proven most interesting, and
- 14 -
speculation concerning the role played by phage is not without support. White
and Knight (1958} reported thot smooth colonies of mycobacterie could be isolated
from previously rough strains after exposure to mycobacteriophages. Mankiewicz
(196la} showed thot a phage immune variant of a strain of chromogenic mycobacterie
differed from the parent strain in rate of growth, colonial morphology and pigmenta-
tion, drug sensitivity, Neutral red reaction, animal virulence and in ability to sensi-
tize guinea pigs to Old Tuberculin. The same yaar (196lb} she was able to demonstrate
the presence of mycobacteriophages in stool specimens of patients with tuberculous
and nontuberculous conditions. Up to then mycobacteriophages active on acid-fast
organisms had been isolated only from ferti 1 ized soil.
Examining 50 stool specimens from tuberculous patients Mankiewicz found
10 to contain mycobacteriophage. Most interesting was the finding thot only 1 myco-
bacteriophage 1 isolated repeatedly from one patient was active on pathogenic myco-
bacterie. Sputum cultures tested in parallel revealed thot the sputum of the patient
excreting mycobacteriophage active against pathogenic mycobacteria contained atypi-
cal chromogenic mycobacteria 1 while the sputum of the 9 patients excreting myco-
bacteriophage active against saprophytic and anonymous mycobacterie contained
M. tuberculosis.
More recent studies reported by Mankiewicz and van Walbeek (1962) re-
vealed further interesting results. Human tubercle bacilli H37 Rv when infected with
the mycobacteriophage found to be active against virulent tubercle bacilli, showed
the emergence of phage-resistant bacteria which differed in colonial morphology
(smooth, rounded colonies) and which reacted to the photochromogenicity test, i.e.
- 15 -
pigmentation increased when1 after 30 minutes exposure to light, incubation
was resumed. Upon repeated exposure to bacteriophage, bacterie from the
smooth colonies underwent further changes involving growth rate, nutritional
requirements, loss of niacin production, and loss of catalase activity. The lyso-
genic bacterie did not elicit tuberculin reaction in guinea pigs. Cytological
changes occurred showing elongated and branching bacillary elements, only the
intracellular granules of which were acid-fast by Ziehi-Neelsen staining.
White et al., (1962) also reported additional results which they felt were
evidence suggesting that lysogeny with certain mycobacteriophages altered colony
morphology.
lt is also an interesting fact that chromogenic mycobacterie are frequently
isolated from patients when they are in a clinically improved phase of their disease.
At times both chromogenic and typical mycobacterie are found together, although
it is more common to isolate chromogenic mycobacterie when M. tuberculosis no
longer can be demonstrated (Huppert et al. 1 1957; Nassau et al. 1 1958; Keltz
et al. 1 1958; Runyon 1 1959b; Tarshis, 1960; Mankiewicz, 196la; Butler and
Josephson 1 1963).
2. CLASSIFICATION OF CHROMOGENIC MYCOBACTERIA
Since the beginning of bacteriology 1 the ability of certain organisms to
produce pigment has been utilized as a criterion in the identification of several bac-
terial species. Although, in most instances, bacterial classification is today based on
combinatlons of morphology, staining reactions, cultural characterlstics, biochemical
tests, animal pathogenicity and antigenic analysis, the ability of microorganisms to
- 16 -
produce pigment still remains important in the identification of certain species.
This is particularly true for this group of slow growing chromogenic mycobacteria
now incriminated as etiological agents of tuberculous-like disease in man. lndeed,
it was through their chromogenicity thot they were first recognized and subsequently
named.
Following the increase in the rate of isolation of chromogenic mycobac-
teria from clinicat specimens during the early fifties, Timpe and Runy'on {1954) pro-
posed a tentative grouping for strains isolated from 120 patients, ali of whom were
thought, or known, to have pulmonary disease. An extended scheme proposed later
by Runyon (1959a) has served as the basis of classification until the present time.
Runyon s groups are described as follows:
GROUP 1, PHOTOCHROMOGENS: ("Yellow Bacilli" of Buhler and Pollak,
1953; M. kansasii 1 Hauduroy 1 Subcommittee
on Mycobacteria, American Society for
Microbiology 1 1962; M. luclflavum of
Middlebrook, 1956). Pigmentation: Little
or none if grown in the dark; bright yellow
ta orange or brick red if grown in continu-
ous light. Young actively growing non-
pigmented colonies will become yellow in
the dark incubator 6 ta 12 hours after expo-
sure for 1 hour 1 45 cm. from a 30-Watt lam p.
Photochromogeniclty is a pronounced and
GROUP Il, SCOTOCHROMOGENS:
- 17-
rapid change. lt is imperative thot tests
be made with actively growing cultures
(5 ta 6 days old).
Growth rate: Almost as for tubercle
bacilli, or slightly more rapid at 37C.
(11 0range Boeil li", of Buhler and Pollak,
1953). Pigmentation: Yellow or orange
from the b.eginning of growth in the dark;
more reddish if grown continuously in light.
Growth rate: About as for tubercle bacilli
or a little more rapid at 37C.
GROUP Ill, NONPHOTOCHROMOGENS: ("Battey" type of Crowetal., 1957).
GROUP IV, RAPID GROWERS:
Pigmentotion: Usuall y weak or none; if
present slowly developing and not as des-
cribed for Groups 1 and Il.
Growth rate: As described for Groups 1
and Il.
Growth rate: Growth within 48 hours at
20- 25C from invisible small inocula.
This Group includes described species of
Mycobacterium, as M. fortuitum, M. phlei,
M. smegmatis or Nocardia species.
- 18 -
The current classification for recognized members of the genus Myco-
bacterium is outlined in Table 1. A description of M. kansasii 1 designated in
1962 by ~he Subcommittee on Mycobacteria, American Society for Microbiology 1
as a type species for Group 1 photochromogens appears in Table Il.
TABLE 1
GENUS MYC08ACTERIUM
(8ergey 1s Manual of Determinative Bacteriology, 7th Edition, 1957)
SAPROPHYTES PARASITES
including potentiaJ parasites; grow on warm-blooded animais
rapldly on most media at 28 C. A. GROWTH ON ORDINARY OR SPECIAL MEDIA
1. M. phJei 7. M. ulcerons
2. M. smegmatis 8. M. tuberculosis
3. M. fortuitum 9. M. bovis '()
4. M, marinum 10. M. microti
5. M. thamnopheos 11. M. avium
6. M. platypoecilus 12. M. paratuberculosis
B. HAVE NOT BEEN GROWN ON NON-LIVING
CULTURE MEDIA
13. M. leprae
14. M. lepraemurium
TABLE Il
MYCOBACTERIUM KANSAS!!
(Subcommittee on Mycobacterie, American Society for Microbiology, 1962)
COMMON NAMES:
GL YCEROL EGG SLANTS:
PATHOGENICITY:
DISTINCTIVE CHARACTER:
TYPE CUL TURE:
PHOTOCHROMOGEN; 11YELLOW BACILLUS11
AFTER TWO WEEKS IN DARK INCUBATOR, RAISED WITH IRREGULAR
SURFACE AND MARGJNS, IVORY, OR OFF-WHITE. IF GROWN IN
LIGHTED INCUBATOR, LEMON-YELLOW BECOMING ORANGE OR
EVEN RED-ORANGE WITH AGE.
IN MAN IT PRODUCES PULMONARY AND EXTRAPULMONARY DISEASE
VERY SIMILAR TO TUBERCULOSIS. VERY SLJGHT OR NO PATHOGEN-
ICITY FOR GUINEA PIGS, RABBITS, AND FOWL.
DEFINITE YELLOW PIGMENTATION WITHIN 24 HOURS AFTER 1 HOUR
EXPOSURE OF AN ACTIVELY GROWING CULTURE TO BRIGHT LIGHT.
AMERICAN TYPE CULTURE COLLECTION STRAIN 12478 (Bostrom).
1
N 0
- 21 -
3. METHODS PROPOSED FOR THE DJFFERENTIATION OF CHROMOGENIC
MYCOBACTERIA FROM M. TUBERCULOSIS, AND FOR THE DETECTION
OF SPECIFIC VARIETIES AND SUB-GROUPS
A. Animal viru1ence studies
Very few of the many strains of acid-fast bacilli which occur in nature
are capable of causing progressive disease in man or animais. Furthermore 1
virulent forms exhibit a well-marked degree of specificity in host range (Dubos,
1948). This characteristic is utilized in the current clssification of human,
bovine and avion strains (Bergey, 1957}.
The most commonly employed animais for virulence testing of myco-
bacterie include guinee pigs 1 rabbits, chickens, hamsters and mice. Untll the
early nineteen fifties much rel iance was placed on the close correlation in
pathogenlcity of mycobacterie for guinee pigs and man. This no longer con be
maintained as many INH-resistant strains of tubercle bactlli, as weil as numerous
unclassified strains of mycobacterie have since been found to possess either very
low or undetectable levels of virulence for the guinee pig.
Animal virulence studies, as related to the unclassified mycobacterie,
have been of little value either in differentlating types within the group, or in
evaluating the pathogenicity of such strains for man. Runyon (1959a), reported
thot unclassified mycobacterie 1 in general 1 foi led to produce progressive disease
in guinee pigs. Photochromogens usually cause disease in mice if 1/100 mg. is
inoculated intravenously, or 3 mg. lntraperitoneally. Scotochromogens are non-
pathogenic for laboratory animais, and Group Ill nonphotochromogens ("Battey")
-22-
are variable in pathogenicity for animais, sorne strains resembling Group 1
by showing limited pathogenicity for mice, and others lacking pathogenicity
for ali laboratory animais.
Hardy et al. 1 (1958) performed virulence studtes on 73 atypical strains
of human origln. Not one produced generalized or progressive disease in gulnea
pigs. They fou nd ali 4 photochromogens tested to be pathogenlc for mi ce 1 as
were 25 of 47 nonphotochromogens tested. None of the 5 scotochromogens
studied infected mi ce. ln contrast 1 others found scotochromogenlc mycobacterie
to be virulent for mice, possessing a special affinity for the kidney and forming
epithelioid tubercles 8 weeks after inoculation (Matsumoto et al. 1 1963). Sorne
workers feel thot differences in results of animal virulence studies for unclassified
mycobacterie are due to too short a period of observation.. Characteristic lesions
have been observed to develop ln guinea pigs and mice thot have been retained
under observation for longer periods of time (Xalabarder, 1961; Kertay et al.,
1962). Chromogenic mycobacterie exhibit limited pathogenicity for the rhesus
monkey (Schmidt 1 1957). 1 t has recent! y been shown thot dogs may be come in-
fected with photochromogenlc mycobacteria (M. kansasit) but the infection is not
serious or progressive (Leon et al., 1964).
B. ~ntigenic analysls
Cross-sensitization experiments indicated an immunological relationship
between hu man tuber cie bac i Il i and the chromogens (Tarsh is and Frisch 1 l952c;
Nassau and Hamilton 1 1957). lt appears thot they not only shore antigenic pro-
parties but thot they also possess specifie components. Using a hemolytic modifi-
- 23-
cation of the Middlebrook-Dubos hemagglutination test it was shown thot sera
from tuberculous patients reacted to antigens prepared from atypical 11yellow11
mycobacterie, and sera from patients with atypical mycobacterie! disease reacted
with human Old Tubercul ln antigens (Nassau et al. 1 1958). Sera from guinea pigs
infected with human strains usually react to Old Tuberculin antigen only 1 whereas
sera of animais infected with strains of the yellow bocillus type cross react with
the Old Tuberculin antigen. Absorption of these sera with heat-killed H37 Rv and
"yellow" bocilli gives a typical absorption pattern. The H37 Rv removes antlbodies
leaving the "yellow titer unchanged, but the 11yellow'1 bocilli remove both human
and 11yellow11 antibodies.
Mankiewicz {1958) using two serological techniques (agar diffusion
precipitation and complement fixation) attempted to classlfy chromogenic acid-fast
bacilli, and found both methods revealed evidence of overlapplng of antlgens be-
tween chromogens, and both typical tubercle boeil li and saprophytic mycobacterla.
This was also the finding of Beek (1959) using culture filtrates of acid-fast bacilli.
He showed cross reaction between the unclassified saprophytes and tubercle baclll i
when performing skln sensitivity tests ln gulnea pigs. There was serologlcal speci-
ficity to sorne degree on a quantitative basis in thot animais infected with unclassi-
fied mycobacterie, or wlth tubercle bacilli, showed strongest reactions wlth homolo-
gous PPD extract (Beek, 1960). The difference ln strength of the reactions wlth
homologous and htHerologous was not marked and consistent enough to distinguish
between various mycobacterie (Beek, 1961i Affrontl 1 1959).
-24-
ln a study uslng mammolian tuberculin (PPD-S) and nonphotochromogen
"Battey'' tuberculin (PPD-B} it was found thot patients infected with tubercle bacilll
had definite sfrong reactions to PPD-S and weaker or no reactions to PPD-B. Patients
infected wlth 11Battey11 organisms had lorger reactions to PPD-B thon to PPD-S. Com-
parative testlng provided a way to separate tubercul in sensitivity produced by the
11Battey11 organism from thot produced by tubercle bac ill i. Using PPD-Y (yellow
bacilli) and PPD-S the sensitivity produced by the two organisms was too similar to
be distingulshed by comparative testing. An interesting epidemiologlcal finding was
made using PPD-S and PPD-B. lt was observed thot two-thirds of the men recruited
for novy services from the states of Georgia and Florida were reactors to PPD antigen
prepared from the 11 Battey11 mycobacterlum while only about 6 per cent were considered
reactors to PPD-S (Edwards and Palmer r 1958}.
The study of Magnusson et al. {1961) suggests antigenic differences among
sfrains isolated in different areas of the world. 11Sensitlns11 prepared from lndian orange-
pigmented stroins were similar to 11sensitins 11 prepared from Danish orange-pigmented
sfrains, but differed from those produced from oronge-pigmented African strains.
lt is proven thot a significant number of non-specifie tuberculin reactions
results from Infections with unclassified mycobacteria (Kendlg, 1962, 1963; Flynn,
1962; Smith and Johnston, 1963).
The amount of protection or immunity conferred upon animais by exposure
to unclassified mycobacterie is under active study. Guinea pigs and mi ce vaccinated
wlth unclassified mycobacteria have been observed to possess sorne degree of immunity
(Wenkle et al. r 1948; Youmans et al. r 1961; Klugh and Pratt r 1962; Larson and
Wicht, 1963). By means of protection tests in mlce, it has been shawn thot immuni-
-25-
zation with atypical mycobacteria representing Runyon 1s Groups 1, Il and Ill gave
partial protection against infection with M. tuberculosis. A photochromogenic
strain equalled or surpassed BCG in protective power. Vaccination of mi ce with
BCG conferred a significant degree of protection against pathogenic photochrome-
genie strains (Siebenmann 1 1964). ln other studies the degree of immunity to
M. tuberculosis H37 Rv induced by a photochromogen in guinee pigs and mlce was
as good as thot produced by BCG (Satake 1 1963).
lmmunological cross reactions between mycobacterial organisms may
eventually be eliminated with purer antigens. Kniker (1961) ln a study of H37 Rv
as weil as representatives of ali groups of unclassified mycobacterie, employed ion-
exchange chromatography and was able to demonstrate thot while sorne antigens were
common to ali or most organisms, others possessed specificlty for lndividual organisms.
A large number of strains have been studied by agglutination and antibody
absorption techniques. This approach may have greater application in epidemiologi-
cal studies thon in cl inical diagnostic laboratorles (Schaefer and Reggiardo, 1963;
Salto et al., 1964). Using agar diffusion methods Chapman {1961) obtained a sero-
logical reaction between sera of patients with sarcoidosis and antigens from photochro-
mogenlc mycobacterie.
C. Resistance to drugs and other agents
Primary drug resistance is a characteristic of the unclassified mycobacterie.
The level of resistance naturally varies with strainsi however, broadly speaking the
photochromogens 1 scotochromogens and nonphotochromogens show parti a 1 resistance
to 1 mcg. per ml. streptomycin 1 10 mcg. per ml. PAS and 1 mcg. per ml. INH
(Runyon, 1959a). Much higher levels of resistance are frequently reported for these
- 26 -
organisms especially when they are isolated from treated cases. Prissick and
Masson (1957) found sorne strains to be inhibited by PAS only in 1000 mcg. con-
centration. Tarshis et al. , (1955) testing a series of chromogens, observed thot
INH failed to inhibit the growth in concentrations up to 100 mcg. per ml. Chro-
mogens have also been found to grow in medium containing 750 mcg. per ml. of
streptomycin (Butler and Josephson, 1963). Results of susceptibility tests performed
on these organisms with the lesser used antituberculous drugs are too scarce and
variable to provide useful information. Resistance to various agents at specifie
concentrations has been suggested for the separation of mycobacteria, but division
on this basis has not been too practical (Gastambide-Odier and S~ith, 1 958; Collins,
1962; Eidus et al., 1959, 1960; Hedgecock and Faucher, 1961; Tsukamura, 1962;
Jonesand Kubica,1963, 1965; Minsley,1964).
D. Bacteriophage typing
The application of bacteriophages for the classification of mycobacteria
was suggested in 1956. Hnatko (1956) phage typed 34 classified a cid-fast micro-
organisms and compared them with 33 unclassified mycobacteria which had been
isolated from tuberculous patients. Seventy~three per cent of the unclassified organ-
isms were lysed by one or more of the 9 phages used.
Mankiewicz (1961b) also studied the phage susceptibil ity of chromogenic
acld-fast bacteria isolated from patients with tuberculosis-1 ike disease and found
thot 22 of 98 were affected by exposure to one or more of the six phages used. Other
workers have since isolated mycobacteriophages from stool specimens of patients with
pulmonary disease (Coter and Redmond, 1963).
- 27-
lt should be mentioned thot non-specifie clearing occurs when heavy
concentrations of phage are employed in the typing of mycobacterie. This find-
ing decreases the value of many results reported from earlier studies in which typing
was performed wh ile uslng heavy suspensions of phage. The routine test dilution
method and lts modifications give promise of more reliable results (Tokunaga et al. 1
1961; T okunaga and Murohash i 1 1961 1 1963; Redmond 1 1963a, b; Red mo nd et a 1. ,
1963). lt would appear 1 however 1 thot ali the technical problems are not yet solved
and more remains to be known before rel labie methods are avallable to glve repro-
ducible results thus permitting the phage typing of mycobacterie on a basis simllar
to thot employed with other groups of b'Jcteria (White et al., 1963; Tokunaga et al.,
1964; Manion and Bradley1 1964; Redmond 1 1964).
E. Growth at various temperatures
Information of 1 imited value can be obtained by preparing subcultures of
mycobacterial isolates and incubating them at 45C. 1 22 - 25C. 1 and 3~C.
Mammalien tubercle bacilli, both human and bovine, grow slowly and only at 37C.
The avion tubercle bacilli grow slowly at ali three temperatures, while the described
species of saprophytic mycobacterie (M .. phlei 1 M. smegmotis) grow rapidly at ali
three temperatures. Most of the unclassified mycobacterie grow slowly at 37C. 1
and at room temperature, but variable results are obtained at 45 C. 1 (American
Thora ci c Society 1 1961).
F. 11 Cord 11 factor and neutra! red test
Both the ability to form ''cords11 in fluid medium and the ability to bind
neutra! red in alkal ine buffer solution have been associated with virulence in tubercle
bac ill i (Middle brook et a 1. 1 1947; Dubos and Middlebrook, 1948). Unclassified
-28-
mycobocterio give variable results in both tests (Huppert et al. 1 1957; Wayne
et al. 1 1 957; Hardy et al., 1958i Torshis, 1960b; Xoloborder 1 1961; Butler and
Josephson, 1963; Korlson et al., 1964).
G. Lipid content
Specifie lipids hove been extrocted from mycobocterio and it is possible
thot this may serve to differentiote species and strains. Each strain possesses severa!
1 ipid fractions 1 and each fraction has on infrared spectrum sufficlently characteristic
ta differentiate lt from other fractions. Using this principle H37 Rv and H37 Ra
strains of M. tuberculosis can be distinguished from eoch other. Bovine and human
stralns have a Iso been spearated. Extending the method to the unclassified myco-
bocterio and using a combination of column chromotography and infrored spectroscopy
it hos been possible to extra ct a specifie glycol ipid from ali 17 photochromogens
exomined. The fraction hos been identified and hos been found to be entirely inde-
pendent of pigment formation and photooctivotion {Randell et al., 1951, 1952; Smith
etal., 1954,1957,1960a,b). ln 1964approximately 150culturesofmycobacterla
were exomined for their lipid content. Basically the results indicated thot among the
20 to 30 substances recognized in the lipids of a given culture, ali but one or at most
two were shared between that culture and most other cultures of mycobacterie regard-
Jess of species. The substance thot is limited in distribution would appear to permit
the recognition of the species or sub-group of mycobacterie to whlch the culture
belongs (Randell and Smith, 1964). lt has a Iso been demonstroted that relative af-
finities of acid-fast bocillt for the fat solvents moy have taxonomie importance. Par-
tition offinity for a given strain of bacillus was found reproducible and characteristic
(Wayne and Jaurez, 1955).
-29-
H Tissue cu 1 ture
ln Hela cells photochromogenic and 11 Battey11 stroins hove been
shown to grow intracellulorly in charocterlstic patterns, the most prominent feature
was a distinct beading which occupied the entire length of the orgonisms {ShepordJ'
1958). Bronching filaments were also observed.
Morigi (1959) studied phagocytosis of atypical mycobacterie belonging
to the chromogenic group. He found thot the chromogens did not show invasive
capacity comparable to thot of virulent tubercle bacflli 1 and their multiplication
in the Hela cells did not produce cytopathological changes. Scotochromogens
were phagocytosed at a faster rote thon photochromogens. From his studies on
scotochromogens and photochromogens it appeared evident thot intracellular multi-
plication of the scotochromogen proceeded more rapidly and thot they showed a
more active invasion. This is contrary to clinical and experimental findings which
indicate thot scotochromogens are genera li y nonpathogenlc. Others (Brosbe et al.,
1962) 1 a Iso using Hela ce lis 1 studied avion and "Battey11 mycobacterie and found
a difference in growth rate, but concluded thot the difference observed did not
permit differentiation of species. Bronching filaments were observed with variable
frequency in ali of the stroins studied, the bronching occurred at right angles and
was seen more often in 11Battey 11 strains. Withfn three to five days, tissue cultures
showed numerous intracellulor organfsms for ali of 9 11 Batteyu and 4 avium stroins
tested.
1. Morphology and stoining
The morphology of chromogenic mycobacterio studied by Tarshis and
Frisch (1952a), voried considerably depending upon the medium used and the age of
-30-
the cu 1 ture.
ln general, most reports agree with Runyon (1959a) in which he
described his Groups as follows:
Photochromogens:
Scotochromogens:
Nonphotochromogens:
Average size larger thon tubercle-
bacill i 1 often qui te long, banded and
beaded; strongl y a cid -fast.
Variable in size 1 strongly acid-fast.
Highly pleomorphic, but often very
short, characteristically containing a
single hyperchromie granule.
Kappler and Janowiec {1963) are of the opinion thot photochromogens
are not strongly acid-fast. This has a Iso been the experience of the author (un-
published data). Others reported thot photochromogens were easily differentiated
from other mycobacteria when stained by the Ziehi-Neelsen method, stating thot
skilled observers would seldom miss them in direct smears or histological sections
(Nassau and Hamilton, 1957).
Periodic acid-Schiff stain has been used to differentiate photochromo-
genic and scotochromogenic mycobacterla from other varieties (Csillag, 1960).
The validity of attempts to dlfferentiate photochromogens on a morpho-
logical basis from other mycobacteria is weakened somewhat by the finding of Gale
(1961) who showed a dramatic difference in the morphological appearance of these
organisms before and after exposure to 1 ight. Prior to exposure to light these or-
- 31 -
ganisms were typical of tubercle boeil li in appearance, but following exposure to
light and subsequent pigment formation the organisms were twice as long, 4- 6
times as wide, and contained more thon three times as many granules. Similar
observations are reported by the author in this thesis.
J. Eh potentiel
Earl y attempts to separate human, bovine and BCG strains by measurlng
their Eh potentiel in buffer solutions showed thot there was no distinct difference
in types (Aksianzew 1 1933; Wilson et al., 1952). Likewise nothing conclusive
was derived from studies involving methylene-blue reduction (Bloch, 1950a 1 bi
Aksianzew 1 1933; Desbordes and Fournier, 1950).
K. Adansonian classification {electronic computer)
Bojal il et al. 1 (1962) using the electronic computer method of Sneath
(1957a 1 b, 1958), studied 229 unclassffied strains of mycobacteria and classified
them into 12 different categories on the basis of physiological properties. Their
study revealed the existence of intermediates between categories (or branches as
they cal led them) which formed a continuous metabol ic spectrum and made difficult
the separation of related categories.
L. Biochemical tests
a) Introduction
For the classification of microorgcmisms, biochemical and metabolic
differences ore naturally preferred to those of cultural morphology and animal
pathogenicity. The slow growth rate of mycobacterie, however, presents technicol
difficulties along these lines.
- 32 -
ln severa! tests thot have been suggested for the detection of bio-
chemical and metabolic differences in members of the mycobacterie, the differ-
ences observed have been quantitative rather th an qualitative in nature. Su ch
tests have limited value in the recognition of specifie types of organisms because
resu 1 ts depend upon the meta bol ic state of the organism und er test, as weil as the
degree of sensitivity and specificity of the test employed. Thus far few useful
quai itative differences for strain recognition have been observed among the myco-
bacterie. Certain tests which distinguish only between rapidly growing saprophytes
and M. tuberculosis are not of much practlcal value.
Biochemical tests which have served to help in dlfferentiating M. tuber-
culosis from chromogenic mycobacterie, and to detect specifie varieties or sub-groups
of the unclassified organisms are listed below under appropriate headings.
b) Niaci n test
One of the more useful tests available for the differentiation of M. tuber-
culosis from other members of the mycobacterie is thot of niacin testing, based on the
principle thot tubercle bacill i con synthesize nicotinic acid from certain ami no ac ids
(e.g. asparagine and glutamic acld) if they are the only source of nitrogen in the
medium. Konno (1953) reported on a marked quantitative difference in the niacin
production of human type tuber cie bacill i and other mycobacterie when grown in a
synthetic culture medium. These workers (Konno, 1956i Konno et al., 1957, 1958a,b)
introduced a test designed to differentiate human from ali other mycobacterie. Bovine
avion, nonpathogenic and unclassified acld-fast badlli were reported to be negative
while the human type gave positive results irrespective of INH susceptibility, cota-
- 33-
Jase content, or pathogenicity for the guinea pig. Rare exceptions have been
reported but it is genera fly accepted that only M. tuberculosis var. hominis pro-
duces enough niacin to give a positive reaction by the Konno test or its modifica-
tions. ft has recently been reported thot a culture of M. fortuitum was isolated
which gave a positive reaction for nlacin (Karl son et al., 1964). As this organism
grows rapidly it is not llkely to be confused with M. tuberculosis. The niacin test
may be considered as reliable as any other test currently available for differentiating
human tubercle bacilli from ali other mycobacteria {Konno and Sbarra, 1959; Konno,
1960).
c) Catalase activity
Unclassified mycobacterla as a group show strong catalase activity
{Runyon, 1959a). ln this respect they differ from strains of INH resistant M. tuber-
culosls which generally show decreased catalase activity (Middlebrook, 1954). lt
is known tht~t catalase of human and bovine stralns, regardless of their virulence,
is inactivated by suspending cultures in a phosphate buffer of pH 7 at 68C for 20
minutes. Unclassified mycobacterie retain their catalase activity under these condi-
tions (Kubica and Pool, 1960). A correlation appears to exist between catalase
activity and animal virulence (Peizer et al. 1 1960). A study of 46 strains of tubercle
bacill i from INH treated patients showed that the organisms could be grouped as
follows:
1. Those of hlgh catalase activity which were virulent for
guinea pigs;
2. Those of moderate catalase activity which were or were not
virulent for guinea pigs;
-34-
3. Those of low catalase activity which were usually non-
virulent for guinea pigs.
Thot catalase activity is in sorne way related to virulence was further supported
by the finding thot the degree of catalase activity in photochromogens paralleled
the clinical significance of the strains in the patients from whom they were isolated
(Wayne 1 1962).
d) Arylsulfatase activity
Early studies on arylsulfatase activity of mycobacterie revealed variable
quantitative differences for most members including unclassified varieties (Whitehead
et al. 1 1953; Engbaek 1 1954; Wayne et al., 1958) 1 but recent reports indicate
thot this test under control led combination of substrate-concentration and bacterial
inoculum may be employed in the differentiation of mammal ian and avion tubercle
bacill i from ali ether mycobacterla. The arylsulfatase test promises to provide a
method of dlstinguishing M. avium from Group Ill nonphotochromogens (nBattey11).
Avion and "Battey'' mycobacteria have been indistinguishable except on the basis
of pathogenicity of the former for the chicken (Kubica and Vesta!, 1961i Kubica
and Bearn, 1961). Arylsulfatase testlng has also been used to separate M. fortuitum
from ether rapidly growing saprophytes (Kubica and Rigdon, 1961; Wayne,1961).
M. fortuitum is the only "rapid grower11 which is capable of causing infection in man.
e) Hydrolysis of Tween uao"
lt was recently noted that certain mycobacteria when grown in media
containing 11Tween 80 11 produced a turbidity out of proportion to the actual number
of viable ce lis present {Wayne, 1962). Sorne organisms could attack "Tween 8011 and
- 35-
form opalescent degradation products. Using this principle as the basis of a test 1
M. tuberculosis and M. bovis con be differentiated from Group 1 photochromogens
(M. kansasii). Photochromogens have the abllity to attack 11 Tween 8011 whereas the
tubercle bacilli and M. bovis as weil as ali Group Ill nonphotochromogens which
produce disease in man, birds or other animais give negative results. Group Il
strains are variable. This principle has been used by Wayne et al., (1964) in a
scheme for the classification and identification of mycobacteria.
f) Growth in the presence of thioglycollate
A useful observation made by Tarshis and Frisch (1952a) (also Tarshis,
1959, 1960b; Potuznik and Matejka, 1962) was thot chromogens would grow slowly
in fluid thioglycollate-containing medium. The saprophytes grew rapidly and luxur-
iantly within 24-48 hours forming thick wrinkled pellicles. M. tuberculosis was
unable to multiply in such a medium. A modified sol id thioglycollate medium con-
taining methylene blue has been used to separate Group 1 and Il organisms from those
of Group Ill, the latter failing to grow on the modified medium. (Smith and Steenken,
1961.)
M. Miscellaneous tests
ln attempts to fi nd rel iable criteria for the recognition and identification
of slow growing mycobacterie a great variety of tests have been proposed. Conven-
tional differentiai ut il ization of carbohydrates 1 al though suggested from ti me to ti me 1
has not proven useful. Recently Sweeny and Jann (1961, 1962) developed a new
technique which may be of possible value. They use a massive inoculum in various
carbohydrates and polyhydric alcohols superimposed on a solid basal medium contain-
- 36 -
ing phenol red indicator. Without multiplication of the organisms specifie metabo-
1 ic activities can be detected. The au thors have used the term "non-growth" test-
ing for this procedure. The organism being tested must first be grown in a rich
medium and harvested in its logarithmic phase of growth to provide cells of high
meta bol ic activity.
Differentiation based on carbon utillzation has been investigated by
sorne workers (Karlson and Ulrich, 1962i Cerbon and Trujillo, 1963) without leading
to practical results. Severa] workers have had partial success in differentiating
groups, species and types of mycobacteria on the basis of amidase activity (Juhlin,
1960i Bonicke, 1960i Halpern and Grossowicz 1 1957i Cerbon and Trujillo, 1963i
Satake, 1963). On the basls of ribonucleic acid-desoxyribonucleic acld ratio,
mycobacteria can be divided lnto two sub-groups (Tsukamura 1 1960).
A test designed to differentiate M. tuberculosis from ali other mycobac-
teria was recently described by Karlson et al., (1964). lt is based upon the appear-
ance of growth within a modified Proskauer and Beek liquid medium. Their test, in
principle 1 is sim il ar to a procedure developed by the candidate in the present research
(Butler and Josephson 1 ] 963).
An indirect approach has been used by McCuiston and Hudglns (1960).
They compared the electrophoretic patterns of sera from patients with sarcoidosis,
tuberculosis and disease due to unclassified mycobacteria. The sera from patients
with sarcoidosis and infection caused by unclassified mycobacteria showed certain
similarities whereas those from patients with tuberculosis gave dissimilar patterns.
The authors studied the sera of seventy-four cases suffering from infections caused by
- 37-
unclassified mycobacterie and in ali instances the changes observed were more
like those in sarcoidosis than tuberculosis. Electrophoresis showed a decrease in
serum albumin and an increase in gamma globulin without an increase in alpha-2
globulin.
N. Characterization of pigments in mycobacteria
a) M. tuberculosis
Anderson and Newman (1933a, b) isolated a yellow crystalline pigment
from saponified acetone-soluble fat of M. tuberculosis var. hominis. They identified
the pigment as a naphthoquinone and named it phthiocol.
Cold acetone extracts of M. tuberculosis var. avium cultivated on
synthetic media containing trace element supplements are often heavily pigmented
(Patterson 1 1960). The pigment involved was isolated and identified as copropor-
phyrin Ill.
M. tuberculosis BCG 1 sensitive to 1 NH 1 produces pigment when exposed
to the drug whereas a resistant strain of the organism &)es not (Youatt 1 1962).
Pterin-1 ike pigments were isolated from tubercle bacill i by Crow and
Walker (1949).
b) Saprophytes- (M. ph1ei 1 M. smeg:natis, M. locticoto, etc.)
Chargaff (1930, 1933) isolated beta and gamma carotenes from M. phlei
by separa ting fractions on columns of al uminum oxide. The same pigments were found
in the ac id-fast organisms 1 Bacillus lombardo pellegrini and Baclllus grassberger,
however the latter also contained lycopene (Chargaff and Lederer, 1935).
From M. phlei lngraham and Steenbock (1935) isolated alpha and beta
carotenes 1 cryptoxanthin, esters of lutein, zeaxanthin and azafrin. They also reported
- 38 -
a fraction resembling the naphthoquinone phthiocol. The ratio of the various
pigments end their absolute amounts varied with age and cultural conditions.
Leprotin 1 a carotenoid hydrocarbon similar to beta carotene was found
by Grundmann and Takeda (1937) in an acid-fast organism isolated from a leprous
lesion. This organism was most likely a saprophyte. Takeda and Ohta {1939 1 1940 1
1944) found leprotin in M. phlei. These workers questioned the flndings of Chargaff 1
1 ngraham and Steenbock and they fel t that what had been reported as be ta carotene
was really leprotin.
M. lacticola (M. smegmatis) when grown on mineral oil media produced
the acidic carotenoid astacin as weil as carotene, but no xanthophylls (Haas and
Bushnell 1 1944). The same organism grown on nutrient agar produced carotenes,
xanthophylls, but not asta ci n.
Turion (1950b) described a yellow pigment 1 acid in character 1 produced
by M. phlei. lt differed slightly from astacin and he proposed the name chrysoflein.
La ter (1951 1 1953) T uri an reported be ta caro te ne 1 xanthophyll 1 leprotene 1 gamma
carotene 1 a 11rhodopin-like11 lycopenoid, an oxycarotenoid, phytofluene and chryso-
flein. The pigments differed quantitatively depending on whether growth occurred
at 30C or 3r>C. Uslng diphenylamine to lnhibit carotenoid biosynthesis in M. phlei 1
Turion and Haxo (1952) isolated two fractions not detectable in uninhibited normal
bacterie. One fraction was a visible yellow and the second was colorless 1 but
possessed a greenish-grey fi uorescence.
Fisher et al. 1 (1955) found leprotene and beta carotene as epiphasic frac-
tions when p:~rtitioning extracted pigments of M. phlei between petroleum ether and
- 39-
90 per cent methanol. A hypophasic red pigment was also observed but not iden-
tified. Goodwin and Jamikorn (1956) investigated three strains of M. phlei and
found thot they ali produced the sorne carotenoids in the sorne relative amounts.
lncluded were beta carotene, zeta caratene, leprotene, lycopene and two uniden-
tified xanthophylls. The xanthophylls were pigments earl ier considered to be cryp-
toxanthin and zeaxanthin. These workers were of the opinion thot Turian's chryro-
flein was a mixture of these two unidentified fractions.
Also working with M. ph lei Schlegel (1959) found phytoene, phyto-
fluene, beta carotene,zeta carotene,gamma carotene, neurosporene, lycopene and
myxoxanthophyll. Quantitatively the oxycarotenoid myxoxanthophyll accounted for
approximately 95 per cent of the total carotenoids.
A group of mycobacterie isolated from the soil and presumed to be sap-
rophytes, was found to forma red product from PAS (Tsukamura, 1961). The colored
matter was not identified. lt could be produced if ce lis of mycobacteria were in-
cubated at 37 C for 12 hours in a phosphate buffer sol ut ion contai ni ng 2 mg. of PAS
per ml.
Rilling (1964) extracted and identified phytoene and phytofluene from
M. lacticola. Other fractions were obtained which were not identified but which
were similar to. zeta carotene, be ta carotene and leprotene. A hypophasi c fraction
was observed but not identified.
c) Unclassified slow growing pathogenic mycobacterio
Pigments of a photochromogenic mycobacterium (Runyon P-8) and a scoto-
chromogenic mycobacterium (Runyon P-6) were investigated by Ebina et al. (1962).
-40-
By chromotography on an alumina column these workers observed four colored
fractions. Near the top of the column was a very narrow brown band. A very
narrow red '::~and was immediately below. These were not identified. A broad
orange band occurred midway down the column, and just below it was a narrow
yellow band. These were identified as beta and alpha carotene respectively.
lt was reported thot exposure of nonpigmented photochromogenic mycobacterie
to light resulted in an increase in the beta carotene content. The appearance
of the chromatogram for P-6 (scotochromogen) was similar to thot for the "light11
culture of P-8. By ultraviolet Illumination five fluorescent bands were observed
but only one was investigated. lt exhiblted the absorption maxima at 348 and
368 mu in petroleum ether and was considered to be phytofluene. Since this
band was broader in the 11dark11 culture of P-8 thon in the "light11 culture, the
workers proposed it as a precursor for beta carotene.
Tsukamura (1962) studied two photochromogenic mycobacterie.
From one strain he isolated beta carotene and lycopene. Only beta carotene
was isolated from the second strain.
Costello et al., (1962) working with representatives of Runyon s
Groups 1, Il and Ill isolated a fraction for wh i ch the absorption spectrum was
a bell shaped curve with a single broad peak at 452 mu in petroleum ether and
462 mu in ethanol. This was similar to Turian s chrysoflein. A second fraction
was considered to be leprotin. Costello (persona) communication, 1963) felt thot
Group 1 and Il organisms contained essentially similar pigments, however they only
studied two strains of each Group. Studies on three Group Ill strains revealed thot
one developed pigments similar to those of Group 1 organisms, but two Group Ill
strains were markedly different.
- 41 -
4. FORMATION OF PIGMENT IN MYCOBACTERIA
A. Introduction
Previous observations reported in the literature concerning pigment
formation in mycobacterie relate to a variety of species and strains. Just how
applicable earlier studies are, in terms of the pathogenic chromogens being
encountered today, is unknown. lt is unfortunate thot practically ali detailed
chromogenicity and pigment studies on mycobacterie, including those involving
pigment extraction and characterization, have been performed on saprophytes.
Pathogens do not lend themselves to the extensive manipulations necessary for
detailed analysis. Furthermore, the association of chromogenic mycobacterie
with human disease has only recently been establ ished.
The factors which have in the post been observed to influence pigment
production in various acid-fast bacilli are reviewed below under appropriate
headings.
B. Oxygen tension
Schwabacher in 1933, when studylng saprophytic chromogenic acid-fast
bacilll concluded thot the most important factor influencing pigment production
appeared to be the oxygen supply. ln tightly sealed tubes little or no pigment
was produced even after incubation for severa! weeks, while in llghtly corked
tubes pigment formation was often weil marked within a few days. Petroff and
Steenken (1935) two years later cultured M. phlei in two series of flasks, one
series plugged with cotton and capped with perforated rubber nipples, and the
second series tightly sealed with melted paraffin. Growth in the sealed tubes
was much smaller and less chromogenic, and they concluded thot chromogenicity
-42-
of M. Ehlei was an unstable characteristic influenced by several factors including
that of available oxygen supply. The following year Reid {1936- 1937) investi-
gated seventy-six strains representing twenty-four genera of chromogenic bacterie.
His collection included M. Ehlei and M. butyricum. No pigment was produced
when bacterie were grown under low oxygen tension. Abundant growth occurred
in freshly inoculated tubes sealed with special tube closures, but lack of oxygen
was reported to have prevented pigmentation. Similarly Baker (1938), using con-
trol led experiments 1 studied pigment formation in a mycobacterium pathogenic for
small tropical killifish. He likewise found thot pigment would not form in the
absence of molecular oxygen. The same has recently been reported for the sapro-
phyte M. lacticola (Rilling, 1962) and it has been postulated within recent months
that a compound capable o inducing a carotenogenic enzyme is formed in the same
organism as a result of exposure to light and oxygen (Rilling 1 1964). At the same
time, and following release of results in this present study (Butler 1 1964) 1 Wayne
and Doubek (1964) reported on the role of air on the photochromogenic behavior
of M. kansasii.
M. tuberculosis BCG has been found to produce pigment when exposed
to isoniazid in the presence of oxygen. This may, however 1 be a straight chemical
reaction since it was not determined whether the pigments were derived from the iso-
niazid, or whether they were a product of disordered cell metobolism. Indirect evi-
dence favoured the view that the pigments were derived from the lls (Youatt,1961).
C. TemEerature of incubation
As in other bacterial genera, the effect of temperature on pigment forma-
tion has been examined in mycobacterie. Petroff and Steenken (1935) reported
-43-
M. phlei deepened in pigment content when left at room temperature. Turion
(1953) has shown quai itative differences in pigment formation by M. ph lei when
grown in glycerol broth at 30C and 3~C. Decrease in pigment production with
increase in temperature was recorded by Reid (1936 - 1937) and a Iso Darzins(l939),
while Baker {1938) found the opposite to be true. He observed greoter chromo-
genicity at 3~C1 a temperature at which the strain studied did not grow. Still
another pecul iar flnding was thot of Winn and Petroff {1933). They accidently
subjected a culture of M. avlum to a high temperature and obtained a dissociated
form having ali the properties of the original, except it became ochre in color and
lost its pathogenlcity. Buhler and Pollak (1953) when first describing their "Yellow
Bacillus", stated thot cultures grown at a temperature of 3~C. could Jose chromo-
genicity and be cream colored. Their "Orange Baclllus11 retained its deep orange
irrespectlve of the temperature of incubation.
D. pH of medium
There is little agreement in the observations recorded concerning the
effect of pH o'1 chromogenesis in mycobacterial species. One study indicated thot
carotenogenesis in a saprophyte, (M. lacticola) was favoured by an alkaline pH
over a very broad maximum, a pH of 8 being considered satisfactory (Rilling, 1962).
ln another study 1 pigment production of M. tubereolosis BCG was diminished at pH 6
compared with thot obtained at pH 7 (Youatt, 1961). Steenken {1935), in contrast,
reported tubercle baci Il i grew with chromogenicity at pH 6 but lacked col or at pH 7 .6.
ln a large survey involving 76 cultures of chromogenic bacteria, includlng representa-
tives of mycobacterie, the reaction of several media was adjusted over a pH range
from 6.0- 8.4. lt was found thot below pH 6.6 or above pH 8.0 pigmentation be-
-44-
came less marked. Within this range differences in reaction had 1 ittle effect upon
the production of pigments although changes in shade or tint were detectable (Reid,
1936- 1937). Bovine tubercle bacilli have been observed to produce definite pig-
ment below pH 8.0, while a more alkaline reaction resulted in growth without pig-
ment (Reed and Rice 1 1929). lngraham and Steenbock (1935), determined the
optimum reaction for pigmentation of M. phlei to be between pH 6.0 and 7.0.
When the pH was above 8.6 pigmentation was very poor in any medium.
E. Composition of medium including the presence of trace elements and drugs
Many sporadic reports have been made on the effect of various media 1
or specifie chemical substances in a particular medium 1 on the pigmentation of myco-
bacterla. lt is of no advantage to cite observations related to cultural systems in
which several variable factors existed. Without controlled pH, temperature, and
oxygen tension, it is obvlous from the foregoing, thot depending upon various com-
binations of variables, pigment may or may not be formed. This was the finding of
Reid (1936- 1937) when 96 media of different composition were investigated to deter-
mine which factors and substances favoured 1 and which inhibited pigment formation.
ln the study of 76 species of chromogenic bacterie 1 results led hlm to conclude thot
pigmentation was governed by a variety of factors, the presence of one or severa! of
which may be adequate to produce sufficient pigment so thot the absence of sorne
other factor or factors would not be noticed. The study indicated 1 however 1 thot
the principal element essential for production of pigment in bacterla was nJtrogen.
Carbohydrates increased pigmentation probably by virtue of stimulating the amount of
growth 1 but ln the absence of nitrogen they were not able to support pigmentation.
-45-
Others studying M. phlei also demonstrated the importance of nitrogen, but further
found that the type of compound which served as a source of nitrogen had little
effect on pigmentation if the hydrogen ion concentration was maintalned around
neutrality {lngraham and Steenbock, 1935). Striking changes were noted, however,
when glycerol was substltuted for glucose. The rate of growth was slower on gly-
cerol than on glucose-containing medium, but cells grown on glycerol became yellow
earl ier in the growth period and pigment during the la ter stage was many times that
obtained on glucose.
Strains of tubercle bacilli have been noted to produce green pigment in
Sauton medium. Kolle {1932) stated that this was first observed by Lange and Pis-
catore at the time of the tragedy in Lubeck caused by contamination of BCG vaccine
with the "KieP' strain of tubercle bacilli. The strain could be readily recognized
by its abundant pigment production after recovery from organs of children who had
died from administration of the contaminated vaccine.
Cultures of M. phlei, M. leprae (?),and M. smegmatis were pigmented
when grown on ordinary media but failed to yield pigments when grown on media con-
taining n-decane, light mineral oil, heavy mineral oil, or paraffin wax. ln contrast
M. lacticola when grown in mineral oil media produced a bright orange pigment
{Hass and Bushnell, 1944).
For certain mycobacteria the enhancement of pigment production ln serum-
containing medium is recognized. The human form of M. tuberculosis generally de-
velops a creamy to yellow or faint red pigment especially on media-containing serum,
while the bovine and murine forms are not pigmented (Bergey, 1957). Griffith {1916) 1
almost 50 years ago, described the effect of bovine serum on pigment formation in
-46-
cultures of tuber cie ba ci Il i.
Reed and Rice {1929) 1 demonstrated the influence of iron on the pig-
mentation of acid-fast bacilli. When grown in the absence of iron, human, bovine,
avion, and saprophytic strains were colorless. When iron-containing medium was
employed, ali strains produced pigment. ln contrast, pigmented non-acid-fast
bacterie tested produced approximately equal amounts of pigment on both iron-free
and iron-containing media. M. tuberculosis var. avium when grown on synthetic
media containing trace-element supplements (cobalt, copper, zinc) formed pigments
which could be extracted from the cells with cold acetone {Patterson, 1960).
Pinner (1935a 1 b) 1 one of