RF-TB-1.23.pdf

Media

extracted text
iviyuvk/uvivi itA

UA. I i v>i

Human Disease
Second edition
John M. Grange MSc, MD
Reader in Clinical Microbiology,
National Heart and Lung Institute,
Imperial College,
London

ARNOLD
A member of the Hodder Headline Group
LONDON • SYDNEY • AUCKLAND
Co-published in the USA by
Oxford University Press, Inc., New York

I

I

I

Introduction

Acknowledgements

1

During my time in the field of ‘mycobacteriology’ I have met so many
delightful, erudite and friendly scientists who have contributed immensely to
this book by imparting their wisdom. I shall mention none by name lest I upset
any whom I might inadvertently omit, but I express my deep gratitude to them
all­
Illustrations have been kindly provided by the Robert Koch Institute (Figs.
1.1 and 1.2), Miss C.A. Dewar (Fig. 2.2), Mr B.W. Allen (Fig. 2.6), Dr J.L.
Stanford (Fig. 3.1), Prof. J. Swanson Beck (Figs. 5.7 and 5.8), Dr P. Ormerod
(Figs. 5.12, 8.3, 8.4, 8.6, 8.10 and 8.18), the Tuberculosis Unit of the World
Health Organization (Figs. 6.2 and 6.3), Dr N.M. Samuel (Fig. 7.3), the late
Dr S.G. Browne (Figs. 7.5,7.6,7.7 and 7.9), The Leprosy Mission (Fig. 7.12),
Dr M.B. Rubens (Fig. 8.11), Dr P.D.O. Davies (Fig. 8.13), Dr K. Schopfer
(Figs. 8.14 and 9.7), Dr J.P. Zellweger (Fig. 8.15), Dr M. Humphries (Figs
8.16 and 8.17), Prof. W.C. Noble (Fig. 9.5) and Mr A.J. Prosser (Fig. 9.6).
Finally I owe a deep gratitude to my wife Helga for her continued support,
encouragement and cordon bleu cooking!

From time immemorial tuberculosis and leprosy have ranked amongst the
most feared and dreaded of the numerous diseases that afflict mankind. The
evangelist John Bunyan dubbed tuberculosis ‘the Captain of all of these men
of Death’, and in India it was known as the King of Diseases. Leprosy may be
termed the Disease of Kings, as Robert the Bruce, King of Scotland, and,
according to legend, the emperor Constantine are numbered amongst its
victims.
The discovery of the causative agent of tuberculosis by Robert Koch in
1882 led to great hopes that this disease would soon be vanquished. In 1908
Leonard Williams wrote: ‘The riddle of the white plague, which had so long
defied solution, had been read at last; the dreary watches of the night were
over; and the dawn, with its promise of victory, peace, and purity, were really
at hand’. Who would, therefore, have predicted that, 111 years after Koch’s
discovery, the World Health Organization, far from celebrating the eventual
conquest of tuberculosis, would take the unprecedented step of declaring it a
global emergency.
Leprosy had become virtually extinct in the industrially-developed nations
by the early twentieth century and, following the introduction of multidrug
therapy, the incidence has dropped throughout the world. Fortunately, many
caring organizations have maintained their interest in this particularly cruel
disease.
Although the classical mycobacterial infections are relatively uncommon
in the industrialized nations, opportunist disease due to mycobacteria that nor­
mally live harmlessly in the environment is becoming an increasingly serious
problem. Once regarded as little more than a curiosity, such disease now fre­
quently complicates the acquired immunodeficiency syndrome (AIDS). The
‘little red rods’ seem determined to continue to inflict misery on the human
race. They are most tenacious and can only be dislodged from individual
patients and from society in general with great difficulty. Regrettably, their
greatest ally is man’s indifference to the sufferings of others.

History
The turning point in the history of tuberculosis occurred at the meeting of the
Berlin Physiological Society on the evening of 24 March 1882, when Robert
Koch (Fig. 1.1) described the isolation of the causative organism of this
disease. This, in fact, was eight years after Armauer Hansen published his
•

•

V▼

•

1

••

1

History

2

i

3

Introduction

Fig. 1.1 Robert Koch (1843-1910): discoverer of the tubercle bacillus in 1882

Fig. 1.2 Jean-Antoine Villemin (1827-92): pioneer of experi­
mental studies on the transmissibility of tuberculosis

not receive the adulation accorded to that of Koch as he was unable to isolate
the organism in pure culture.
The work of Hansen and Koch did not occur in scienti fic isolation. The stage
had been set by the clear establishment of the germ theory of communicable
disease by Louis Pasteur and, in particular, by the experimental demonstration
of the transmissibility of tuberculosis in rabbits by Jean-Antoine Villemin
(Fig. 1.2), a French military surgeon, in 1868. It was therefore considered very
likely that tuberculosis and leprosy were caused by ‘germs’ and many workers
attempted to isolate them. Koch’s critics have remarked that he was only able
to discover the tubercle bacillus because he used methods developed by other
workers, namely Weigert’s stains and Tyndall’s inspissated serum medium. On
the other hand, Koch’s acknowledged industry, patience, tenacity and techni­
cal skill must have contributed overwhelmingly to his success. Indeed, Koch’s
detailed descriptions of his techniques enabled other workers to reproduce his
findings and the few antagonists were rapidly silenced.
Koch’s discovery heralded the era of hope, and serious research soon took

diagnostic purposes, the search for an effective cure, and the development of
a vaccine. Diagnosis required specific stains and methods for the in vzrro cul­
tivation of mycobacteria. Koch stained his preparations with an alcoholic
solution of methylene blue and used vesuvin as a counterstain Very shortly
afterwards, Paul Ehrlich discovered the ‘acid-fastness’ of the tubercle bacillus
and introduced a staining technique which, with minor modifications by Ziehl
and Neelsen whose names the method now bears, is still widely used today.
Originally, tubercle bacilli were grown on heat-coagulated serum, then in
glycerol-beef broth. Egg-based media were introduced by Dorset m 1902 and
were modified by Lowenstein in 1930. Methods for ‘decontaminating chmcal specimens were introduced by Petroff and others around 1915^0 further
significant developments were made for many years so that, in 1954, Dubos
remarked that tuberculosis bacteriology was based on ‘primitive bacterio­
logical techniques worked out decades ago’. Since that time we have wit­
nessed the introduction of rapid radiometric techniques for the detection of
mycobacterial growth and even more rapid techniques for the amplification of

nFf in

main direction^- the iced at inn and culture of the bacillus for

4

History

Introduction

specific DNA in clinical specimens. Nevertheless, the adoption of these new
techniques remains a mere dream for tuberculosis workers in most labora­
tories throughout the developing world.
Koch’s discovery coincided with the birth of the discipline of immunology.
Koch certainly did not regard himself as an immunologist. Indeed, when
Metchnikoff demonstrated the phenomenon of phagocytosis, Koch remarked
‘I am a hygienist and it is of no interest to me where the microbes are,
whether inside or outside the cells’. Nevertheless he attempted to develop an
agglutination test for tuberculosis using the whole bacillus as antigen and
also tried to attenuate a human strain for use as a vaccine for tuberculosis in
cattle. But his main studies centred on the development of a cure for tubercu­
losis, and this led to the extraordinary saga of Old Tuberculin. This, as out­
lined in Chapter 5, followed a meticulous series of experiments on
guinea-pigs that led to the description of the tissue necrotizing reaction sub­
sequently named the Koch Phenomenon. Unfortunately Koch was under con­
siderable pressure from his political overlords to announce a cure for the
disease and his rather hasty use of Old Tuberculin in patients proved disas­
trous and almost ruined his reputation. Nevertheless, this was the first attempt
at ‘immunotherapy’ - a form of treatment that was forced into near oblivion
by the advent of antibacterial chemotherapy but has returned to offer, per­
haps, the only chance of controlling tuberculosis in the face of multidrug
resistance. Koch’s Old Tuberculin, and the reactions elicited by it, may well
have been forgotten if it had not been for Clemens von Pirquet (Fig. 1.3) who
developed the tuberculin test - one of the most widely used yet misunder­
stood of all diagnostic tests.
At the British Congress on Tuberculosis in 1901 Koch made a serious error
that was to have far-reaching consequences; namely his statement that bovine
tuberculosis was of no danger to man. The veterinary surgeons present were
so astounded by this pontification that they persuaded the Minister of
Agriculture to convene a Royal Commission to investigate the issue. In a
period of ten years the scientists employed by the commissioners, notably
Arthur Stanley Griffith and Louis Cobbett, accrued an enormous amount of
information on the epidemiology, bacteriology and pathology of bovine tuber­
culosis and produced irrefutable evidence that humans are susceptible to
tuberculosis of bovine origin. The Commission amply demonstrated the
benefits of state-sponsored medical research and it was the fqrerunner of the
British Medical Research Council. It also laid the foundations for the bovine
tuberculosis eradication programmes which must be hailed as the most effec­
tive control measures ever mounted against a bacterial disease.
The general principles of vaccination were well-established by Pasteur, and
many workers attempted to attenuate the tubercle bacillus for use as a vaccine.
One of the first successful attempts was made by Edward Trudeau who atten­
uated a human strain by repeated passage on coagulated sheep serum for two
years. Although this, the R1 strain, was avirulent in the guinea-pig, no further
development was undertaken. Trudeau established a tuberculosis research
institute, which bears his name, at Saranac Lake, New York State. Years later,
while working at that institute, George Mackaness laid the foundations to the
study of the mechanisms of cell-mediated immunity and the role of the

5

Fig. 1.3 Clemens von Pirquet (1874-1929): the pioneer of the
tuberculin test

A vaccine was eventually produced by Calmette and Guerin after passaging
a bovine tubercle bacillus 230 times on potato slices soaked in bile and glyc­
erol over a period of 13 years. This vaccine, Bacille Calmette-Guerin (BCG),
was first used in 1921 as an oral vaccine for infants. Its early use was delayed
by considerable controversy concerning its safety and by the Lubeck dis­
aster’ in 1930 when many children were accidentally vaccinated with a viru­
lent strain of Mycobacterium tuberculosis and 73 died. After a further delay
caused by the Second World War, a freeze-dried vaccine was introduced and
has been widely used since. Whether it has a future or whether, m the light of
advances in our understanding of the immunology of mycobacterial disease,
we are ready for a totally different approach to vaccination remains to be seen.
The next milestone in the history of mycobacterial disease occurred in the
• ............ :~.u

dicrnvprv nf the first effective drugs. As in the

6 Introduction

case of the discovery of the causative organism, leprosy preceded tuberculosis
though with less fanfare and acclaim. Faget and his colleagues found that
promin was effective against leprosy in 1943 and streptomycin was discov­
ered, after an extensive search, by the Russian bacteriologist Selman
Waksman and his team in 1948. This, and the subsequent discoveries of iso­
niazid and other drugs, at last removed the fear of tuberculosis. The general
public and some physicians were convinced that the disease was conquered
and would soon be extinct. Qtas^including Waksman himselt were not so
optimistic and doubted whether antituberculosis drugs alone would solve the
problem. Sadly they'have been proved correct. Indeed the promises held out
by drugs and the Bacille Calmette-Guerin (BCG) vaccine have probably done
more to eradicate interest in the mycobacterial diseases than to eradicate the
diseases themselves.
The 1950s were a time of great excitement owing not only to the introduc­
tion of effective chemotherapy and the early BCG trials, but also to the seri­
ous interest being taken in disease due to other mycobacterial species. The
first to be described in detail were due to M. ulcerans (Buruli ulcer) and M.
marinum (swimming pool granuloma). In 1954 Ernest Runyon published the
first of his studies on the classification of ‘anonymous’ mycobacteria causing
lung disease in man. This, together with the pioneering investigations of Ruth
Gordon, led to a renewed interest in the taxonomy of the mycobacteria, cul­
minating in the extensive studies undertaken by the International Working
Group on Mycobacterial Taxonomy (IWGMT) established by Dr Larry
Wayne. At present there are 41 ‘approved,’ mycobacterial species (see Chapter
3), and about 20 others. About half the species are known to cause disease in
animals or man.
Before the early 1980s, disease due to environmental mycobacteria was
uncommon and few clinicians would have seen more than one case in a life­
time. As with tuberculosis, the human immunodeficiency virus (HIV) pan­
demic brought about a profound change in interest as these mycobacteria,
notably M. avium (the avian tubercle bacillus) are common causes of oppor­
tunist disease in acquired immune deficiency syndrome (AIDS) patients.
In addition to interest in mycobacteria as pathogens, attention was given to
their place in the environment. It is now clear, contrary to earlier views, that
the genus Mycobacterium is essentially one of environmental saprophytes and
that pathogenicity is not their usual behaviour. Thus, the major pathogens M.
tuberculosis andM. leprae are atypical mycobacteria although, paradoxically,
this term has been applied to the typical saprophytic species! Ecological
studies have proved to be of great relevance to disease as there is little doubt
that immunologically effective contact with environmental mycobacteria has
a profound influence on the way a person subsequently responds to BCG vac­
cination or to infection by a pathogenic species.
The period from 1970 until the late 1980s was one of great fascination,
interest and confidence. Rifampicin was introduced and made it possible at
last to develop short-course curative chemotherapy for both tuberculosis and
leprosy. After a decade or so of extensive clinical trials organized by
Professors Mitchison and Fox of the British Medical Research Council and
their collaborators abroad, 6-month regimens of orally administered drugs

The future and the challenge 7
contrasts sharply with the 3000 doses of drugs over a 2-year period used in the
early days of chemotherapy. During the same period, short multidrug regi­
mens for leprosy were developed and are now in general use with very great
benefit.
,
.
.
Not only was this a period of great advances in chemotherapy, it was also
one in which enormous strides forward were made in mycobacterial ecology,
taxonomy, structural and biochemical studies (especially on the lipid-rich cell
wall) and immunology. This period also saw the beginning of the ‘molecular
era’ of microbiology. It is now possible to clone mycobacterial DNA m alter­
native hosts and to obtain gene products from the new hosts. This enables pure
mycobacterial antigens, even those from the non-cultivable leprosy bacillus,
to be obtained in sufficient quantities for diagnostic tests, immunological
studies and possible incorporation into new vaccines. The polymerase chain
reaction, and related DNA amplification systems and highly specific nucleic
acid probes, are set to revolutionize diagnostic mycobacteriology, provided
that their high cost can be reduced. Recombinant gene technology and mono­
clonal antibodies have also facilitated the ‘fine structure’ analysis of the
immune response in mycobacterial disease by permitting the identification
and isolation of the various cells and mediators involved.
On the less positive side, this was also a period in which it became even
more clear in the collective mind of the medical profession that the days of the
mycobacterial diseases were numbered. Professional societies once dedicated
solely to tuberculosis switched allegiance to other respiratory diseases: even
the International1 Union Against Tuberculosis appended ‘And Lung Disease’
to its title. Worse was to come - in what has been termed the greatest blow
during the 1980s to the fight against tuberculosis, the British Medical
Research Council’s Tuberculosis and Chest Diseases Unit was closed. Interest
likewise waned west of the Atlantic and in 1986 the US Centers for Disease
Control ceased its surveys of drug-resistant tuberculosis. In the research field,
funding and interest declined: young scientists and medical practitioners were
discouraged from entering the discipline of mycobacteriology and only a few
enthusiasts remained. Yet the warning signs were there for those that had eyes
to see, not the least of which was the advent of the HIV pandemic.

The future and the challenge
In the early 1990s there occurred what may, in future times, be depicted as one
of the greatest shifts of awareness and interest in a disease that has ever
occurred in the history of medicine. From being a jaded and exhausted sub­
ject, tuberculosis was suddenly the centre of attention! The trigger events for
this resurgence of interest were the small but definite upsurge of the disease in 1
the USA, after decades of decline, and the occurrence of outbreaks of HIVJ
related multidrug-resistant tuberculosis in New York. Thanks to florid news­
paper articles and television programmes, the general public was left in no
doubt that the ‘white plague’ had returned. The fact that this plague of yester­
year should strike in the centre of a major metropolis in one of the world’s
richest nations struck a chord of fear throughout that society and led to calls

8

Introduction

The medical profession was soon swept up in this wave of renewed interest
and the literature bristled with editorials, of widely varying quality and origi­
nality, on the resurgence of the disease. The World Health Organization’s
Tuberculosis Unit was rapidly expanded and charged with the task of review­
ing the magnitude of the problem and developing a new global control strat­
egy. Funds were made available for laboratory studies and there was a
renaissance in research activity, particularly in molecular microbiology, in the
hope of producing new diagnostic and epidemiological tools, novel vaccines
and new therapeutic approaches.
Although the fruits of this scientific renaissance are very exciting intellec­
tually, they have contributed little so far to the practical problems of the con­
trol of mycobacterial disease. On the contrary, the fact that the World Health
Organization has declared tuberculosis a global emergency eloquently attests
to the inadequacy of the control measures at our disposal or, more likely, our
i abject failure to deploy them responsibly. The high and increasing numbers of
persons dually infected with tubercle bacilli and HIV is, as described in
Chapter 6, a cause for great anxiety and it must be remembered that dual
infection would have been uncommon if tuberculosis had been managed
effectively in the past. Sir John Crofton, a greatly respected pioneer of anti­
tuberculosis chemotherapy, has remarked that:

15f

It is a sad reflection on society’s incompetence that, more than thirty years
after the methods for cure and prevention were evolved, and before the
advent of the HIV epidemic, there were already more patients with active
tuberculosis in the world than there had been in the 1950s.
With the rapidly increasing and devastating effects of the HIV pandemic on
tuberculosis and the emergence of multidrug resistance we need innovative
control measures such as immunotherapy as well as the established ones but,
more than anything else, we need the vision to realize that the global emer­
gency of tuberculosis is growing daily and that to delay the implementation of
adequate control measures is a recipe for disaster. The World Health
Organization has drawn attention to the large discrepancy between the inci­
dence of, and mortality due to, tuberculosis relative to other infectious dis­
eases and the funding made available to combat it. Appeals based on the fact
that tuberculosis is the most cost-effective of all adult diseases to treat and yet
is still the cause of 1 in 4 preventable adult deaths appear to have fallen on
deaf ears. Let us hope that the global community is granted one more chance
to conquer tuberculosis and that this chance will not be lost.
But what of leprosy? Interest in this disease is in danger of being swamped
by that currently given to tuberculosis. Certainly, recent surveys show that,
thanks to the efficacy of multidrug therapy, leprosy is declining in incidence
and there is no evidence that the HIV pandemic will have an adverse effect on
this decline. Nevertheless we must hope that we have learnt from the recent
upsurge in the incidence of tuberculosis that to permit a decrease in interest and
financial investment to minor the decrease in prevalence of a transmissible dis­
ease is most foolhardy. In this context, it is well to remember that, though rarer
than many tropical diseases, leprosy is the cause of physical and mental suffer­
ing well out of proportion to its prevalence. There is no reason why leprosy

Publications of general and historical interest

9

Full scientific, medical and financial cooperation between scientists and
field workers and between the developed and developing nations will be
required for the eventual conquest of mycobacterial disease. There is no
doubt that these are principally afflictions of the socio-economically under- ,
privileged and that the relative freedom of the West from such ills is a direct^
result of its prosperity. Ironically, our failure to eradicate tuberculosis glob­
ally is largely the consequence of our success in the control of the disease in
the developed world and of our parochial indifference to the^ sufferings of
those in distant lands. We are now learning the hard way that ‘none are safe
until all are safe’. It must be evident by now that until the barriers of race,
creed and nationality are broken down, and until mistrust and strife are
replaced by brotherly love, compassion and cooperation, the tyrannical reign
of the King of Diseases and the Disease of Kings will continue unabated.

Publications of general and historical interest
Allen, B.W. and Hinkes, W.F. 1982: Koch’s stain for tubercle bacilli. Bulletin of the
International Union Against Tuberculosis 57, 190-2.
.
,
.
Bishop, P.J. and Neumann, G. 1970: The history of the Ziehl-Neelsen stain. Tubercle
Brothwe^D^and Sandison, A.T. (eds.) 1967: Diseases of antiquity Springfield:
Charles C. Thomas. (Contains chapters on leprosy, tuberculosis and diseases in the
Bible and the Talmud.)
Crawfurd, P. 1911: The King's Evil. Oxford: Clarendon Press
.
Dubos, R. and Dubos, J. 1952: The White Plague: Tuberculosis, Man and Society.
Boston: Little, Brown.
.
Evans CC 1994: Historical background. In Davies, P.D.O. (ed.), ^fnic^L
Tuberculosis. London: Chapman & Hall, 1-17. (See also the foreword by Sir John
Crofton in the same book.)
. .
. .
Francis, J. 1959: The work of the British Royal Commission on Tuberculosis,
1901-1911. Tubercle 40, 124-32.
.
Grange, J.M. 1982: Koch’s tubercle bacillus: a centenary appraisal. Zentralblatt jur
Bakteriologie, Parasitenkunde, Infectioriskrankheiten und Hygiene, Abt. 1. 251,
297—301
Grange, J.M. and Bishop, P. 1982: Uber Tuberkulose - A tribute to Robert Koch’s dis­
covery of the tubercle bacillus, 1882. Tubercle 63, 3-17.
Heifets, L. 1982: Metchnikoff’s recollections of Robert Koch. Tubercle 63, 139-41.
Keers, R.Y. 1978: Pulmonary tuberculosis: a journey down the centuries. London:
Koch R. 1882: Die Aetiologie der Tuberculose. Berliner Klinische Wochenschrift 19,
221-38. Translated by Pinner, B. and Pinner, M. 1932: American Review of
Tuberculosis 25, 285-323.
_ , . .... .
Major, R.H. 1945: Classic descriptions of disease. 3rd edn. Springfield, Illinois.
Charles C. Thomas.
.
Pallamary, P. 1955: Translation of Gerhard Armauer Hansen: Spedalskheden Aarsager
(Causes of Leprosy). International Journal of Leprosy 23, 307-9.
Rosenthal, S.R. 1957: BCG vaccination against tuberculosis. London: Churchill.
(Includes a historical chapter by GuSrin.)
Ryan F 1992: The greatest story never told. Bromsgrove: Swift Publishers.
Stanford, J.L. and Grange, J.M. 1993: New concepts for the control of tuberculosis m

10

Introduction

Villemin, J.A. 1868: Etudes experimentales et cliniqies sur tuberculose. Paris:
Bailli^re et Fils.
Vogelsang, T.M. 1978: Gerhard Henrik Armauer Hansen (1841-1912), the discoverer
of the leprosy bacillus. His life and work. International Journal of Leprosy 46,
257-332.
.
Waksman, S.A. 1964: The conquest of tuberculosis. London: Cambridge University
Press.
Williams, L. 1908: The worship of Moloch. British Journal of Tuberculosis 2, 56-62.
World Health Organization. 1994: TB - A Global Emergency. Geneva: World Health
Organization.

The genus
Mycobacterium

2

The generic name Mycobacterium was introduced by Lehmann
in the first edition of their ‘Atlas of Bacteriology published in 1896. At that
time the genus contained only two species, Mycobacterium tuberculosis and
M leprae. The name Mycobacterium, meaning fungus-bactenum, was
derived from the way that the tubercle bacillus grows as mould-hke pellicles
on the surface of liquid media. The name did not, and should not, imply that
the mycobacteria are related to the fungi. The non-culturable leprosy bacil us
was included in the genus because it shares a staining property with the
tubercle bacillus; namely, resistance to decolounzation by weak mineral acids
after staining with one of the arylmethane dyes. This property, ^d-f^tnes; ,
is the basis of the widely used Ziehl-Neelsen stain, the history of whic i was
reviewed by Bishop and Neumann (1970) and Allen and Hinkes (1982). Agd
fastness, although a useful distinguishing property, is not unique to the
mycobacteria: btterial spores, for example, are often strongly acid-fast and
members of the related genus Nocardia are.weakly acid4ast.
Shortly after the introduction of the generic name acid-fast bacilli were
cultured from birds and cold-blooded animals such as frogs turtles and fish.
Also, at that time, small but constant differences between tubercle bacilli iso­
lated’ from man and cattle were described. Thus, four tubercle bacilli were
recognized; namely, human, bovine, avian and ‘cold blooded .In addition
acid-fast bacilli were isolated from inanimate sources such as hay, compost
and butter. As tuberculosis in man and cattle was such a senous ProDlem,
these other mycobacteria received scant attention, although there were a fe
early reports of their involvement in human disease. Despite the lack of clini­
cal interest, numerous supposedly new species were descnbed and the 1966
edition of Index Bergeyana listed 128 validly published species.
Paradoxically, this plethora of names made identification of individual iso­
lates so difficult that mycobacteria other than M. tuberculosis were often
termed ‘anonymous mycobacteria’.
.
n . .
Interest in the classification of the genus was awakened in the 1950s by the
descriptions of two new mycobacterial diseases of man - swimming pool
granuloma and Buruli ulcer (see Chapter 9) - and by the pioneering studies of

Ruth Gordon and EmestRunyon.
;n
Runyon (1959) drew attention to the role of anonymous mycobacteria in
human lung disease and placed the responsible strains into four groups
according to their speed of growth and pigmentation. These groups are.

I

photochromogens (yellow pigment formed in the light)
/yellow niement formed in the dark)

Antigenic structure
12

13

The genus Mycobacterium

autolysis or mechanical disruption and those that are actively secreted by
III non-chromogens
IV rapid growers
Though now obsolete, this grouping was of great value in that era of taxo­
nomic chaos. Since then, much effort has been devoted to the classification ot
the mycobacteria and, as a result, many species names have been reduced to
synonymity. Indeed, only 16 of the 128 names in the 1966 edition of Index
Bergeyana are now in use (Ratledge and Stanford, 1982). Before 1980, the
correct name for a species was, by international agreement, the first one to be
validly published after 1 May 1753, the publication date ot Linne s Species
Planetarum. It is now only necessary to refer back to the ‘Approved lists ot
bacterial names’ (Skerman et al., 1980) published in the International
Journal of Systematic Bacteriology on 1 January 1980. This list containsjH
species of mycobacteria, but it omits a number of apparently distinct species,
and several others have been described subsequently (see Tables 3.1 and 3.2
in Chapter 3). The introduction of techniques for DNA manipulation has pro­
vided additional powerful tools for the speciation of bacteria and one such
technique known as ‘ribotyping’ appears to be of value in classifying and
identifying mycobacteria, including non-cultivable strains. This technique is
described on page 29.
.
The variation of properties within the genus Mycobacterium is extensive
and is reflected in the range of virulence, habitat, rate of growth, nutntiopal
requirements and antigenicity. There are, in fact, relatively few properties that
are common to all mycobacteria and yet clearly delineate this genus from
related ones. Many of the unique characteristic.s of the mycobacteria are to be
found in their very complg^lipMdjcLceJl. walls.
The mycobacteria appear to have evolved from the group of Gram-positive
aerobic rods which includes the genera Corynebacterium and Nocardia.
Indeed mycobacteria are Gram-positive although they are not easily stainable
by this method. Mycobacteria are afirobic. (although some such as bovine
tubercle bacilli prefer low oxygen tensions), non-sporing and non-motile.
They do not form capsules in the strict sense although some strains are very
smooth, even slimy, owing to a thick coat composed of lipids termed mycosides (see page 21).
Mycobacteria appear to divide by simple binary fissiqn, although some
authors have postulated more complex life cycles, possibly including cell
wall-free, or spheroplast, forms. Although such forms may be produced as
laboratory artefacts, claims that they occur naturally or indeed that they are
the causative agents of certain granulomatous diseases such as sarcoid or
Crohn’s disease require careful substantiation.

llVMycobacterial antigenic determinants (epitopes) are divisible according to
their distribution within the genus. Up to 15 precipitin lines are demonstrable
when ultrasonicates of mycobacteria are tested against homologous antisera
by double diffusion in agar gel. This technique has been extensive!}' st“d1^
for taxonomic purposes by Stanford and his colleagues (see Stanford and
Grange 1974) who described four groups of soluble (diffusible) antigens (Fig.
2 1): those common to all mycobacteria (Group i); those restricted to slow y
crowing species (Group ii); those occurring in rapidly growing species (Group
iii); andgthose unique to each individual species (Group iv). The species delin­
eated by the Group iv antigens correlates very closely with the speciation
obtained by conventional taxonomy and modem molecular techniques.
This antigenic distribution indicates a fundamental difference between the
slowly growing and rapidly growing species, and suggests that these groups
separated early in the evolution of the genus. Furthermore, some of the Group
iii antigens are also found in the genus Nocardta, suggesting a ^Relation­
ship between this genus and the rapidly growing mycobacteria. Many of the
common (Group i) antigens are also found in the noc^rd^e
detectable in related genera such as Corynebacterium and Listeria. This inter
generic sharing of antigens is probably responsible for the notorious lack of

SPe^eCmyorefS^^rossed' immunoelecuophoresis (CIE)
reveals from 60 to 90 antigens in mycobacterial sonicates and some have been
identified as inzymes (Ridell et al., 1987). Data on about 50
antigens, mo-stly from the M. tuberculosis complex and M. leprae, have been

e°Most tf the eariy work on immune reactivity in tuberculosis was based on

the use of crude culture filtrates, notably Koch’s Old Tuberculin.
Antigen groups
iii

iv

Slow
growers

Rapid
growers

Unique
to each
species

M. leprae
M. vaccae

Antigenic structure

I

Mycobacteria, being complex unicellular organisms, contain many antigenic
proteins and sugars. The antigens are conveniently divided intQ_cytpplasmic
(soluble) and cell-wall lipid-bound (insoluble) antigens. Both have proved ol
value for classifying species and typing strains. The soluble antigens are

Related
genera

e:„ o 1 The dictrihution of soluble antigens in the genus Mycobacterium

14

The structure of the mycobacterial cell

The genus Mycobacterium

Subsequently, Seibert (1934) fractionated this filtrate by simple precipitation
with acetone and saturated ammonium sulphate solution, thus producing
Purified Protein Derivative (PPD) which is still the standard reagent for tuber­
culin testing.
Many workers have also used chemical and physical fractionation tech­
niques in attempts io isolate the species-specific (Group iv) antigens for use in
diagnostic tests, but this task has proved very difficult for two reasons. First,
specific epitopes often occur on the same protein molecules as shared epi­
topes. Even purification methods based on binding to highly specific anti­
bodies (affinity chromatography) cannot separate two determinants if they are
on the same molecule. Secondly, a given determinant may be present on a
range of molecules of differing physical and chemical properties. Thus,
preparative techniques based on such differences (gel filtration and ion­
exchange chromatography) have their limitations. Nevertheless, some wellcharacterized antigens have been prepared in this manner. These include the
A60 antigen, the major heat-stable component of PPD, which forms the basis
of a commercially available serodiagnostic test for tuberculosis (Charpin et

a/., 199°).

.

c KTA

In recent years, attention has largely turned to the cloning of DNA coding
for mycobacterial protein antigens in alternative bacterial hosts and thereby
providing a source of such antigens in pure form. This approach has the great
advantage thaF large quantities of proteins coded by DNA from the noncultivable species M. leprae can be produced for research and diagnostic pur­
poses. A limited number of such recombinant antigens are available for
research purposes from the World Health Organization (WHO) and gene
libraries containing DNA for other antigens^have been produced from the
major pathogenic species (see page 28).
Insoluble cell-wall bound antigens are usually demonstrated by direct
agglutination of whole bacilli by appropriate antisera. This technique is
applicable to those species of mycobacteria that form stable, smooth suspen­
sions, and was used extensively by Schaefer and his colleagues (see Wolinsky
and Schaefer, 1973) for typing M. avium, M. intracellulare and M. scrofulaceum.
Serotypes are identifiable in several other species but not, unfortunately, in
M. tuberculosis which is rough and readily auto-agglutinates. The responsible
antigens have been identified as the sugar moieties on a group of pcptidoglycolipids and phenolic glycolipids collectively termed mycosides (see page 21).
Monoclonal antibodies against some mycobacterial antigens have been
produced. In initial studies, many monoclonal antibodies were described but a
workshop organized by WHO showed that the range of epitopes recognized
by them was rather limited. The third WHO workshop was held in 1992 and
the report lists the well-characterized monoclonal antibodies known at that
time (Khanolkar-Young et al., 1992).

Heat-shock proteins
Heat-shock proteins (HSPs) are a class of cytoplasmic proteins that are
II* < L. *

1. .

onrl Or

Jr, oil IJrMnrr nolle

Ttnrlpr nnrmnl

15

conditions, they are mostly involved in the folding and assembly of newly
synthesized, or nascent, proteins and are thus sometimes termed nurse-maid
proteins or chaperonins. Under conditions of stress, such as elevated tempera­
ture or exposure to toxic agents, their synthesis increases considerably and
they are presented on the cell surface where they may elicit immune
responses. Their role in immune phenomena is described in Chapter 5.
Being highly conserved structurally, HSPs of mycobacteria have a close
homology with well-described ones in Escherichia coli (Young and Garbe,
1991). The principal mycobacterial HSPs are listed in Table 2.1. GroEL, the
65 kDa HSP, is the most thoroughly studied of all mycobacterial proteins (see
review by Thole and van der Zee, 1990). Superoxide dismutase (SodA) is
stress-induced in mycobacteria within living host cells. It is also one of the
secreted proteins (see below) and has a function distinct from the others;
namely, to protect the bacteria from toxic reactive oxygen intermediates.
Table 2.1 Examples of mycobacterial heat-shock proteins
Protein

Size
(kDa)

Function

DnaK
GroEL
SodA
GroES

71
65
23-28

Protein folding and translocation
Protein folding and translocation
Superoxide dismutase
Protein folding and translocation

12-14
•I

Actively secreted proteins
These proteins have attracted considerable interest as it it likely that they,
together with cell surface lipoproteins and HSPs, are the first antigens recog­
nized by the immune system after infection. These proteins have been princi­
pally studied in the M. tuberculosis complex and include fibronectin binding
proteins, superoxide dismutase and some proteins of unknown function. For
full details see Abou-Zeid et al. (1988) and Andersen and Brennan (1994).

The structure of the mycobacterial cell
The mycobacterial cell consists of cytoplasm bounded by a plasma membrane
and enclosed by a complex lipid-rich cell wall. The single chromosome is
tightly wrapped into a nuclear body (Fig. 2.2) but is not bounded by a nuclear
membrane Thus, like other bacteria, the mycobacteria are prokaryots (higher
unicellular and multicellular forms of life have nuclear membranes and are
termed eukaryots). In common with many other bacteria, some mycobacteria
contain additional small circles of DNA termed plasmids (see page 30). The
cell membrane consists of a bilayer of polar phospholipids with their
hydrophobic ends facing inwards and their hydrophilic ends facing outwards
"Tl'r»-»orr»Kmnf» ic rlncelv associated with the enzvmes and cofactors involved

16

The cell wall

The genus Mycobacterium

r n

'A

{J

wall shows three distinct layers: inner and outer electron-dense layers sepa­
rated by an electron-transparent layer (Paul and Beveridge, 1992). The exact
structure of the cell wall has not been fully elucidated but the generally
accepted structural model proposed by Minnikin (1982) and elaborated by
McNeil and Brennan (1991) is shown in Fig. 2.3. The inner layer, overlying
the cell membrane, is composed of peptidoglycan (murein). This, as in other
bacteria, consists of long polysaccharide chains cross-linked by short pep­
tide chains, thereby forming a net-like macromolecule that gives the cell
its shape and rigidity. The polysaccharide chains contain N-glycolyl
muramic acid and N-acetyl glucosamine in alternating positions and the
cross-linking peptide chains consist of the four amino acids L-alanine, Disoglutamine, meso-diamino-pimelic acid and D-alanine. An exception is
M. leprae which has glycine instead of L-alanine (Draper, 1976). The
mycobacterial murein is very similar to that in other genera except that it con­
tains N-glycolyl muramic acid instead of the more usual N-acetyl analogue.
Mycobacteria are powerfuLadjuvants: Freund’s complete adjuvant consists
of killed mycobacteria in oil. This activity resides largely in the murein and
also in small water-soluble fragments released from murein by digestion with
lysosome. One such water-soluble adjuvant, N-acetyl muramyl-L-alanyl-Disoglutamine (muramyl dipeptide, MDP) has been synthesized (Lefrancier et
al.. \9TT).
c j ,
External to the murein is a layer of arabinogalactan, a branched polysac­
charide composed of arabinose and galactose (Lederer, 1971). The terminal

Fig. 2.2 Electron micrograph of a thin section of mycobacteria (Bacille Calmette-Gu&rin
(BCG) vaccine) showing nuclear bodies, cell walls, septa and lipid inclusion bodies
(x 42 000)

in energy production. For details of the fine structure of mycobacteria and
their cell membranes see Paul and Beveridge (1992).
The bacterial cells vary in shape from species to species, and even within
an individual strain according to the growth conditions. The cells of the M.
avium complex may be almost coccoid while those of M. xenopi may be fila­
mentous with occasional branching. Cells of M. kansasii and M. marinum are
often elongated and with a distinctive beaded or banded appearance (see page
45). In clinical material, especially comeal scrapings, cells of M. chelonae
may be long, filamentous and weakly acid-fast and are easily mistaken for
Nocardia (Khooshabeh et al., 1994).

The cell wall
The mycobacterial cell wall is the most complex in all of nature and its major
distinguishing characteristic is a very high lipid content. Indeed lipids account
for about 60 per cent of the cell wall weight and they consist of a wide range
of compounds, some being similar to those found in other organisms and

17

Mycosides

Tl 1111 I

Trehalose dimycolates

Lipoarabinomannan

Mycolic acids

LLtLi

Arabinogalactan

Peptidoglycan

Cell membrane

Nutrition and metabolism 25

24 The genus Mycobacterium

Acid-fastness
Acid-fastness is defined as the ability of the bacterial cell to resist decolourization by weak mineral acids after staining with one of the basic arylmethane
dyes. The property is not confined to the mycobacteria: nocardiae, some
corynebacteria and related organisms and bacterial spores are weakly acid­
fast. Nevertheless, the property is widely used for the microscopic detection
of mycobacteria in clinical or environmental specimens.
Despite numerous investigations, the chemical basis of acid-fastness is
poorly understood. Mycolic acids are certainly involved and the degree of
acid-fastness is related to the size of the acids. Thus the corynomycolic acids,
nocardomycolic acids and mycobacterial mycolic acids are progressively
larger (see page 19) and are associated with progressively more intense
staining. Chemical binding of the dye to the mycolic acid occurs but this is not
the whole explanation as disruption of the cell wall by any means reduces its
acid-fastness considerably. It has therefore been postulated that the mycolic
acids are arranged in certain configurations that cause a trapping of the dye.
This view is supported by the finding that acid-fastness is associated with the
mycolic acid that is covalently bound to the layer of arabinogalactan rather
than that lying free within the cell wall (Goren, 1972).

Pathogenicity and virulence
Pathogenicity is the ability of a micro-organism to cause disease. Clearly this
property depends on the susceptibility of the host as well as the aggressiveness
of the invading organism. Some micro-organisms are obligate pathogens, hav­
ing developed a total dependence on a living host for their continued existence.
In the case of the mycobacteria this includes the pathogens in the M. tubercu­
losis complex (M. tuberculosis, M. bovis, M. africanum and M. microti') and,
possibly, M. leprae. Many other mycobacteria are opportunist pathogens, nor­
mally existing as harmless saprophytes but becoming pathogens under certain
permissive conditions. It is possible that M. leprae and other non-cultivable
mycobacteria live in certain inanimate environments and the ability to identify
species-specific mycobacterial nucleic acids after amplification by the poly­
merase chain reaction (see page 29) enables this possibility to be investigated.
Virulence is a quantitative measure of pathogenicity and may vary consid­
erably according to the host species. Thus, although virulence may be assayed
in a standard, preferably inbred, animal, care must be taken when interpreting
such findings in relation to other animals or humans. This is evident within the
M. tuberculosis complex from the following examples: (a) the vole tubercle
bacillus (M. microti') is virulent for voles and some other small animals yet
attenuated in humans; (b) the bovine tubercle bacillus (M. bovis) is much more
virulent than the human type for the rabbit; (c) the South Indian variant of, and
some isoniazid-resistant mutants of, M. tuberculosis are attenuated in the
guinea-pig but are virulent in humans; and (d) M. bovis is virulent for both
cattle and humans yet M. tuberculosis rarely causes progressi ve disease in cattle.
Tt has Inna been recognized that mvcobacteria owe their pathogenicity to

detail in Chapter 5. Although some strains of M. tuberculosis certainly liber­
ate toxic compounds, their virulence is not primarily associated with such
substances. Thus, claims that certain cell wall lipids with toxic properties,
notably cord factor (dimycolyl trehalose) and the sulpholipids are determi­
nants of virulence, have been refuted (Goren et al., 1982). Two other myco­
bacterial lipids, mycolipenic acid and lipoarabinomannan, may be associated
with virulence as described above (see pages 19 and 21).
A further lipid of possible relevance to virulence of M. tuberculosis is the
attenuation indicator lipid, described on page 21. This is characteristic of the
South Indian strains but is not found in isoniazid-resistant classical strains
which, in common with the former are often of diminished virulence for the
guinea-pig, or in laboratory attenuated strains. Thus its association with atten­
uation may also be a fortuitous one.
Mycobacterium ulcerans is the only mycobacterium which owes its
virulence to a cytotoxic substance (see page 103), while the virulence of M.
avium appears to be related to colony morphology, suggesting a protective role
for surface mycosides (Kuze and Uchihira, 1984). Apart from these examples,
the mechanisms of virulence of the opportunist mycobacteria have received
scant attention.
In the past, searches for the determinant(s) of virulence of M. tuberculosis
were based on the comparison of the properties of virulent and attenuated
variants of the same strain, such as strains H37Rv and H37Ra, but such
searches were singularly unproductive. It is now possible to do much more
discriminative' investigations by transferring genes for putative virulence
determinants from one strain to another (Jacobs and Bloom, 1994).
The idea of finding a single determinant of mycobacterial virulence is an
attractive one with great relevance to the possible development of new vac­
cines. On the other hand, the ‘decision’ as to whether infection by a mycobac­
terial pathogen will lead to overt disease may depend more on the nature of
the host’s immune response than on the invasiveness or other properties of the
bacterium. Rook (1991) has suggested that this ‘decision’ is likely to be based
on a complex recognition of the many antigens and adjuvants of the mycobac­
terium rather than on a single factor. This may explain why searches for spe­
cific determinants of virulence have been unsuccessful.

Nutrition and metabolism
Despite the relatively slow growth of mycobacteria and the complexity of
their lipid-rich cell walls, most species have very simple nutritional require­
ments. They do, nevertheless, show an enormous diversity in the substrates
that they are able to use as nitrogen and carbon sources - a diversity exploited
for purposes of identification. Nutritional requirements include oxygen, car­
bon, nitrogen, phosphorus, sodium, potassium, sulphur, iron and magnesium.
A typical simple medium that supplies all these nutrients is Sauton’s medium,
composed of asparagine, glucose, glycerol, Na2HPO4, K2HPO4, MgSO4 and
ferric ammonium citrate. A wide range of nutrients, including lipids and
nucleic acid precursors are available to mycobacteria within cells and host

I

28

Genetics and molecular biology of the mycobacteria

The genus Mycobacterium

antibody. These proteins increase in amount in iron-deprived mycobacteria
and may be used to ascertain the iron status of in vivo grown organisms such
as M. leprae (Sritharam and Ratledge, 1990).

Genetics and molecular biology of the
mycobacteria
The subject of molecular biology of the mycobacteria has become a huge dis­
cipline in recent years. A complete review of the subject is beyond the scope
of this book. For more details see McFadden (1990) and Chapters 12 18 in
Bloom (1994). In this chapter the aspects of molecular biology of actual and
potential clinical significance are outlined.

The mycobacterial genome

In common with all other bacteria, the mycobacteria cqntain a single circular
chromosome (the genome) and some strains also contaiMd^tj^aBmaHcircular unitsof DNA termed plasmids (Crawford et al 1981). The molecular
weights of mycobacterial genomes range from 3x10 to 5.5 x 10 (Baess and
Mansa 1978). For comparison the molecular weight of the genome of
Escherichia coli is 2.5 X109. The ratio of the base pairs adenine (A) and
thymine (T) to guanine (G) and cytosine (C) vane^considerably between
bacterial genera. The G + C content of the mycobacteria is high, from 60 to /1
per cent of the total base content (Baess and Mansa, 1978).
The degree of relatedness of mycobacterial (or any other) genomes may'be
determined by the technique of DNA hybridization. This is based on the abil­
ity of single strands of DNA to associate into double strands provided that the
sequence of the base pairs of the two strands complement each other. The
extent of hybridization between fragments of DNA from two different strains
or species gives an indication of the similarity of the sequence of the bases.
This technique was used as a taxonomic tool to confirm the species bound­
aries determined by other methods (Baess and Bentzon, 1978). The same
technique was used to identify clinical isolates but has been superseded by the
use of specific nucleic acid probes produced by recombinant technology.
These have the advantage that they give a ‘yes or no’ result rather than reveal­
ing various degrees of homology. Nucleic acid probes for the more commonly
encountered mycobacterial species are commercially available.
The development of specific nucleic acid probes was made possible by
advances in molecular biology, enabling fragments of DNA representing the
entire genome of an organism to be replicated in an alternative organism such
as E. coli. These self-perpetuating clones of DNA are termed genome^ibmnes
and have been produced from M. leprae (Clark-Curtiss et J985), M
tuberculosis (Young et al., 1985; Eisenach et al., 1986) and BCG (Thole et al.,
1985). These are not only the source of DNA for use in diagnostic tests and
genetic analysis but the proteins coded by the genes are the source of antigens

29

DNA amplification - the polymerase chain reaction
The polymerase chain reaction (PCR) is a technique for the in vtTr^jnplification of DNA. In the living cell, DNA replication occurs during cell division in
two stages. First, the double helix is split into its two constituent chains by a
gyrase enzyme. Secondly, complementary chains are synthesized by a DNA
polymerase, resulting in two new double helices. This process may be mduced
in vitro by-heating the DNA to separate the double helix and adding DNA
polymerase and the ribonucleotides from which the new chain is synthesized.
This synthesis will only occur in vitro if a short DNA sequence complemen­
tary for part of the single chain is added, thereby forming a short length of
double helix. Synthesis of the complementary chain then commences from
this site. In practice, the requirement foLSU.clL_‘primers’ is of great value as
specific primers can be selected so that DNA amplification only occurs if
DNA containing base sequences complementary to those of the primers is pre­
sent in the specimen.
.
,
... •
The principles and general methods for conducting PCR are described in
detail elsewhere (Brand et al., 1991). In brief, the specimen nbonucleotides
and polymerase are mixed and placed in an automated machine in which the
temperature is raised so that the double helix dissociates and then lowered,
allowing the complementary DNA chain to be synthesized. (The use of a
heat stable polymerase overcomes the problem of heat-inactivation of the
enzyme.) The cycle, which only takes a few minutes is repeated many
times so that- after two hours there may be a milhon-fold replication of the
target DNA. ThT^oduct DNA is then detectable as a band on electro­
phoresis or by hybridization with a chemoluminescent complementary DNA
probe The PCR has been evaluated by many workers for the diagnosis of
tuberculosis and infection by other mycobacteria and a number of different
primers have been developed. Modifications are continually being intro­
duced, such as isothermal methods and amplification techniques for ribo­
somal RNA. The latter has the advantage that each bacterial cell contains
about 2,000 copies of the target. A test for M. tuberculosis based on amplifi­
cation of 16S ribosomal RNA is marketed by GenProbe (Miller et al ,
1994). In view of the rapidity of developments in this field, reference to the
latest current literature is essential. The state-of-the-art in 1994 is clearly
and succinctly reviewed by Shaw (1994). The advantages and problems
encountered in clinical diagnosis are discussed in Chapter 4. In addition o
its use for the diagnosis of tuberculosis in the living, PCR has been used to
detect mycobacterial DNA in ancient skeletons (Spigelman and Lemma,
1993).

Ribotyping
Ribosomes contain molecules of RNA which are genetically highly con­
served One type of ribosomal RNA, 16S rRNA, has minor variations in.its
base sequence which appear to correspond closely with the established
mycobacterial species. Thus, determination of the sequence of the bases in
.i • t»ma rv.n.>
iicaH tn rlaccifv mvcobacteria and to identify clinical

I
30

isolates. This can be achieved rapidly by amplifying the DNA coding for the
16S rRNAby PCR with the appropriate primer and sequencing the amplified
product on an electrophoretic gel (Rogall et al., 1990; Kirschner et al., 1993).
The method can be used to classify and identify mycobacteria that are noncultivable or very slowly growing. The species M. genavense, a culturally
very fastidious pathogen encountered in AIDS patients (see page 56) and
some pet birds (Hoop et al., 1993) was identified by this technique. In addi­
tion, ribotyping confirms the homology of M. tuberculosis, M. bovis and M.
africanum and of M. kansasii and M. gastri determined by immunodiffusion
serology (see page 51). On the other hand, it fails to differentiate between the
quite different species M. marinum and M. ulcerans and it cannot therefore be
taken as a gold standard for speciation.
Very closely related mycobacteria, such as M. avium and M. paratubercu­
losis are more clearly differentiated by cloning and sequencing the DNA
coding for the 23S rRNA and the spacer region between the 16S and 23S
rRNA genes (McFadden et al., 1994).

Genetics of drug resistance
Drug resistance is acquired by mutation affecting cell wall permeability,
enzymes involved in transport or activation of the drug or the susceptibility or
the target molecule. Detailed studies have been conducted on two of the prin­
cipal antituberculosis drugs, isoniazid and rifampicin.
Two genes are involved in susceptibility to isoniazid: katG and inhA. the
former codes for the catalase-peroxidase enzyme and is either deleted or inac­
tivated by mutation in most isoniazid-resistant strains of M. tuberculosis
(Zhang et al., 1992). The inhA gene codes for a target of isoniazid and ethion­
amide and is involved in fatty acid biosynthesis (Banerjee and Dubnau, 1994).
Mutations in this gene have been found in some, but not all, isomazidresistant strains. The discovery of these genes should help to clarify the mode
of action of isoniazid.
Resistance to rifampicin is due to mutations in the rpoB gene which codes
for one of the subunits of the enzyme DNA-dependent RNA polymerase.
Sequencing of PCR-amplified rpoB gene DNA reveals several different muta­
tions and a technique for the rapid detection of such mutations in clinical prac­
tice (PCR-heteroduplex formation assay) has been developed (Williams et
al., 1994).

Mycobacterial plasmids

I

Genetics and molecular biology of the mycobacteria

The genus Mycobacterium

Many environmental mycobacteria contain one or more small extrachromosomal elements of DNA known as plasmids. The evidence for their occur­
rence in M. tuberculosis is less strong: one report indicates that this species
may contain an extrachromosomal element composed of single stranded
DNA (Zainuddin and Dale, 1990). Plasmids are detected by lysing mycobac­
teria in wavs that do not break DNA chains, removing the heavy principal

31

electrophoresis. Plasmids of specific types may also be detected by hybrid­
ization with cloned plasmid-specific DNA probes.
The effect of plasmids on the properties of mycobacteria is poorly under­
stood owing to the great difficulty experienced in ‘curing’ these bacteria of
plasmids. Some supposedly cured strains have been shown, by DNA finger­
printing, to be unrelated to the wild strain, stressing the importance of avoid­
ing mixed cultures. The only conclusive reported direct plasmid-induced
effect is resistance to heavy metal salts in a strain of M. scrofulaceum
(Meissner and Falkinham, 1986). There is indirect evidence that plasmids of
M. avium may facilitate growth at 43°C and in the absence of oleic acid as
only plasmid-containing strains have these properties.
Plasmids have been most thoroughly studied in the M. avium complex
(MAC) since their discovery by Crawford and Bates (1979). Meissner and
Falkinham (1986) found them in 55 per cent of MAC from people, in 75 per
cent of isolates from aerosols but in only 21 per cent of strains from water.
This suggests that plasmids might increase the hydrophobic property of the
cell wall, facilitating entry into aerosols and, possibly, establishment m the
human host. These plasmids weighed between 8.8 and 160 MDa and there
were from one to six in each strain. Crawford and Bates (1986) showed that
all of 26 strains of M. avium from HIV-positive patients contained plasmids.
These strains were either serotype 4 or 8 (see page 47); the former usually
contained two small plasmids and the latter just one. All strains contained a
plasmid that hybridized with one termed pLRT, suggesting that this plasmid
might be a determinant of virulence of M. avium in HIV-positive patients. On
the other hand, plasmids were detected in only 34 of 71 strains from such
patients in London (Hellyer et al., 1991), shedding doubt on a causative role
for plasmids in the pathogenicity of AIDS-related strains of M. wiunr
Plasmids have proved useful as vectors for introducing foreign DNA into
bacterial cells. Such studies on mycobacteria are rather limited as free plas­
mids do not readily enter mycobacterial cells. The entry rate is improved by
incubating the target cells in agents such as cycloserine or glycine that weaken
the cell wall or by a technique called electroporation m which short electrical
pulses of high voltage transiently open pores in the cell envelope through
which DNA may pass (Snapper et al., 1988). A further problem is that, once
within the cell, plasmids do not stably integrate with the host genome
although rare mutants permitting such integration have been isolated. It is
likely that phage/plasmid hybrids (phasmids) will, as described below, prove
to be more useful shuttle vectors for genetic studies on mycobactena.

Mycobacteriophages
Since the first description of a phage lytic for a mycobacterium by Gardner
and Weiser (1947), many have been isolated, mostly from environmental
sources though a few were found in naturally occurring lysogenic strains.
Such strains often differ from their non-lysogenic counterparts: in particular
their colonies on solid media are often very sticky, and not easily dislodged
from the medium, owing to the accumulation of extracellular DNA and are
>•1

rninbnnc nf DNqci* (Grange and Bird. 1978). In

I
32

I
•l

The genus Mycobacterium

contrast to the closely related genus Corynebacterium, there is no evidence
that mycobacterial virulence is associated with lysogeny.
Mycobacteriophages usually have hexagonal or oval heads and long, noncontractile tails (Fig. 2.8(A)), an exception being phage 13 which has a con­
tractile tail (Fig. 2.8(B)). Many of the phages have very wide host ranges that
appear to ignore the usual species boundaries. Thus, for example, some bacte­
riophages isolated and propagated in strains of M. smegmatis may also be
propagated in certain strains of M. tuberculosis.
Some mycobacteriophages are virulent, leading to lysis of the cells that
they infect and some are temperate. In the latter state, the phage DNA inserts
fairly stably into the host genome and persists for many bacterial generations.
Several mycobacteriophages, though not virulent, do not stably integrate into
the host cell genome but exist as extrachromosomal plasmids: a state termed
pseudolysogeny (Baess, 1971).
.
Initially, interest in the mycobacteriophages centred on their use in the
typing of mycobacteria, particularly tubercle bacilli, for epidemiological pur­
poses (reviewed by Grange and Redmond, 1978). In recent years, bacterio­
phage typing has been superseded by the technically easier and much more
discriminative technique of ‘DNA fingerprinting’ (see below). Interest in the
mycobacteriophages has thus moved to the possibility of harnessing their
ability to shuttle genes from one bacterial strain to another. This phenomenon,
termed transduction, was described in M. smegmatis by use of phage 13
(Sundar Raj and Ramakrishnan, 1970). Two other transducing mycobacterio­
phages, LI (Snapper et al., 1988) and TM4 (Jacobs et al., 1987) have been
described.
.
.
.
,
One of these phages, TM4, was used to construct an ingenious hybrid shut­
tle vector (Jacobs et al., 1987). One of the standard techniques used to clone
DNA is to insert the required DNA into a plasmid which is then packaged in a
bacteriophage, such as phage lambda, which infects E. coli. Numerous copies
of the plasmid, and the inserted DNA, are thus produced in cultures of this
host bacterium. The problem was how to get this plasmid to replicate in a
mycobacterium in a way that it could be transferred to other mycobacteria.
This was solved by making a hybrid of the above plasmid with phage TM4.
This phage-plus-plasmid (phasmid) functions as a plasmid in E. cojl ^nd as a
phage in mycobacteria. Thus the plasmid, containing the required foreign
DNA is extracted from E. coli and introduced into M. smegmatis by electro­
poration (see above) where it replicates and forms phages which may be used
to infect various other species, such as M. tuberculosis, thereby introducing
the required foreign DNA.

I

A

•,

Transposons and insertion sequences
Most living cells contain short sequences of DNA that are able to change their
position within the genome. In some cases there is just one such mobile ele­
ment or ‘jumping gene’ in the genome but in others there are many copies at
different sites. There are several different types. The simplest forms only conmmAnciMn fnr

mnvpment and insertion and are termed

B
Pin 9 R MvcohActerionhanfts: A: Mvcobacterium kansasii phaqe (x 120 000): B: Myco-

I
References

34

I

35

The genus Mycobacterium

functional genes, including those determining antibiotic resistance. Some
temperate bacteriophages also function as transposons.
Mobile elements are able to disrupt the function of genes by inserting in
them and disrupting their continuity. Thus they behave as mutagens, a prop­
erty termed transposon mutagenesis and utilized in genetic analysis.
Transposons are used to insert foreign genes into bacterial cells, particularly ,
when many copies of the gene within the cell are required.
Insertion sequences are common in the genus Mycobacterium, borne are
unique to a species, some to a group of strains with related properties within a
species and some are ‘private’ to just one strain. Some examples are listed in
Table 2.2. The number and sites of mobile elements vary greatly within a bac­
terial species, providing a very useful means of subdividing that species lor
epidemiological purposes. The technique used is termed restriction fragment
length polymorphism (RFLP) analysis or, more popularly, DNA fingerprint­
ing. The technique is described below.

IS6110 (IS986)
IS1081
IS1245
IS901
IS900
IS6100

Origin

M. tuberculosis complex
M. tuberculosis complex
M. avium
M. avium, bird pathogens only
M. paratuberculosis
M. fortuitum. single strain

01 A nwchdearer resolution was obtained by performing the digestionand
running the gel and then applying a cte'llolu'’y7^Ce^'t! ?S67l(hs u u-

SSStSSSSSSKfiSiU
References
r Cmilh I Granee J.M. et al. 1988: The secreted antigens of
Mycobacterium tuberculosis and their relationship to those recognized by the avail-

Table 2.2 Examples of mycobacterial insertion sequences (IS)
Sequence

to an electrophoresis gel but this led to a huge number of bands and detection

Number of
copies
0->20
5-6
2-27
2-8
15-20
4

Within the M. tuberculosis complex, the number and position of insertion
sequences is very stable over many cell divisions (Fomukong et al., 19J2,
Cave et al., 1994).
,.
Species-specific insertion sequences can be used as targets for DNA ampli­
fication by PCR for diagnostic purposes. They are particularly useful if there
are multiple copies within the cell as the sensitivity of the test is thereby
greater than when there is only one copy of the specific target sequence.
Some insertion sequences can be transferred experimentally from one
species to another and it is probable that interchange of these genetic elements
occurs between species and even between genera in nature. Thus 1SG1UU
(Table 2.2) found in just one strain of M. fortuitum is identical to an insertion
sequence found in the genus Flavobacterium.

DNA fingerprinting
The DNA fingerprinting technique relies on the availability of enzymes
termed restriction endonucleases that are only able to digest a DNA chain at
points where they recognize particular short sequences of bases, usua ly 4 to
a in nnmhnr ‘such dipp.sfion results in DNA frasments of varying lengths

International Union Against Tuberculosis 57, 190-2

KarXaSss*

s'

lltemmre concerning pseudolysogeny. Acta Pathologica et Microbiologica
B iess^and Bentton4M.W. 1978: Deoxyribonucleic acid hybridization between dif­
ferent species of mycobacteria. Acta Pathologica et Microbiologica Scandinavica
Biesf^and Mansa, B. 1978: Determination of genome size andI base ratio of
^ deoxyribonucleic acid from mycobacteria. Acta Pathologica et Microbiologica
Banerjet A'andfub nau, e'1994: inM a gene encoding, a target for isoniazid and

..
B Ratledge, C„ Stanford, J.L., Grange, J.M. (eds.), The biology of the mycobacteria,
vol 3 I ondon: Academic Press. 37-106.
Barrow PA 1986: Physiological characteristics of the Myco^acterium.^b^'
To°is-M. bovis group of organisms with particular reference to heterogenetty wtthtn
B
GS
D
and c^ohydrates of Mycobactertum
^bereuiosis:
and
control Washington DC, American Society for Microbiology. 285-306.
Bishop, PJ. and Neumann, G. 1970: The history of the Ziehl-Neelsen stain. Tubercle

Bloom,9 KR06’(ed.) 1994: Tuberculosis: pathogenesis, protection, and control.
Bb^cke,"R'f^I^^orto^iden'fificaffion'of mycobacteria by bkKhemical methods.

Bulletin of the International Union Against Tuberculosis 32, 13-6 .

Not viewed