Mycobacteria Protocols

456 Pages·1998·31.537 MB·English

by  Tanya Parish, Neil G. Stoker (auth.), Tanya Parish, Neil G. Stoker (eds.)

1
Mycobacteria
Bugs and Bugbears
Tanya Parish and Neil G. Stoker
1. Introduction
Mycobacteria are gram-posmve, rod-shaped bacteria of the Actmomycete
family, and therefore are most closely related to the nocardia, corynebacteria,
and streptomyces. Their most characteristic feature is their complex cell enve-
lope, containing a high percentage of lipids, which include the large-branched
mycolic acids. This envelope makes the bacteria resistant to breakage and rela-
tively impermeable to antibiotics, and is responstble for the acid-fast staining
property used to identify the organisms. The genomic DNA contains a high
guanme plus cytosine (GC) content, ranging from 58459% (I), which affects
the utility of Escherzchzac oli as a surrogate genetic host.
Mycobacteria can infect most species of animals including rodents, birds,
and fish. However, their importance lies m the fact that they include major
human pathogens. Tuberculosis, caused by Mycobacterwm tuberdons,
remains the most important mfectious cause of mortality in the world, and lep-
rosy, caused by Mycobacterzum leprae, still afflicts large numbers of people.
Other species can be pathogenic, of which perhaps the most concerning is
Mycobacterium unum, which has recently become apparent as a major oppor-
tunist m HIV-infected people in the developed world.
Mycobacteria fall naturally and taxonomically into two main groups: slow-
and fast-growers. The slow-growers include most of the major human and am-
ma1 pathogens, whereas the fast-growers include nonpathogemc species, such
as Mycobactenum smegmatis, which is widely-used as a convenient, if imper-
fect, model organism.
From Methods m Molecular Emlogy, Vol 101 Mycobacterra Protocols
Edrted by T Parish and N G Stoker 0 Humana Press Inc , Totowa, NJ
2 Parish and Stoker
2. Molecular Biology: A Brief History
The use of recombinant-DNA methods to study the molecular biology of the
mycobacterta began m 1985, with the generation of genomlc-DNA libraries
from M tuberculosis and hf. leprue (2-5). These were imtially used to identify
genes encoding antigemc proteins by screenmg with serum or monoclonal
antibodies (MAbs). E. coli IS a gram-negative organism with an average GC
content of 50%, and it was soon realized that the majority of mycobacterial
proteins were not expressed from then own promoters m this cloning host.
Greater successw as therefore obtained with expression libraries (6).
The use of mycobacteria as hosts for recombinant molecules was made pos-
sible by the development of cloning vectors based on mycobacteriophages and
mycobacterial plasmtds. The tdentification of suitable antibiotic-resistance
markers (mmally resistance to kanamycm), and the successo f electroporation
(see Chapter 11) to introduce such vectors into the cell, has led to the wide-
spread use of nonpathogemc mycobacteria as clonmg hosts. The most popular
has been M. smegmatu, because it is a fast-growing, nonpathogenic organism.
Paradoxically, electroporation proved immediately possible with Mycobacte-
rium bows BCG, whereas tt was necessary to isolate an electroporation-compe-
tent mutant of M smegmatis( the most widely used bemg strain mc215 5) (7). Other
mycobactena are now also being developed for use as clonmg hosts such asM yco-
bacterzumv uccae (8,9). Most vectors are still based on pAL5000, a plasmid iso-
lated from Mycobacterium fort&urn (10), although other replicons and integrating
vectors have been used to a lesser extent (see Table 1).
One of the main problems of the vectors for use m mycobacteria is the low-
copy number (approx 2-5 copies per cell). This, coupled with the difficulty of
lysmg mycobacteria, explains why most primary cloning stepsa re carried out in E.
cob using shuttle vectors, whtch carry origins for both E colz and mycobacteria.
A set of vectors are available for mycobacteria that can be transferred
directly from E. colz by conjugation (see Chapter 10). This avoids the use of
electroporation, which always carries the risk of creating aerosols by arcing. In
addmon, the use of phage-based vector systems is gaming popularity because
the efficiency of transformation is much higher than can be achieved for plas-
mid-based vectors (see Chapter 12).
The next obstacle to serious genetic analysis was the production of mutants.
Some work with chemical mutagenesis has been carried out, although made
more difficult by problems of clumping. Transposon mutagenesis has now been
achieved both m M smegmatis and M. bows BCG, as has gene replacement,
although much of the technology and many of the tools are still far from ideal.
These are discussed m Subheading 3.
Table 1
Useful Vectors for Mycobacteria
Vector Type Based on Selection Features Reference
pJRD2 15 Cosmid RSFlOlO KmalSmb Broad host-range plasmtd (36)
Tropist4 Cosmid pAL5000 Km (37)
pus903 Integrating IS900 Km Expression vector (38)
pMV361 Integrating L5 phage Km (39)
pEP3 Plasmid pNG2 HW (40)
pYT937 Plasmid pMSC262 Km Compatible with pAL5000 based vectors (41)
pMB35 1 Plasmid pLR7 Km Rephcon from a Mycobacterzum avzum plasmid (421
pSMT3 Plasmtd pAL5000 Wg Expression vector using hsp60 promoter Chapter 2 1
pAUl51 Plasmid pAL5000 Hm Expression vector; targeted to membrane Chapter 2 1
pDE22 Plasmid pAL5000 Hyg Expression vector to produce secreted protein Chapter 2 1
pJAz6 Plasmid PAL5000 Sm Transfer by coqugation (43)
pJAZl1 Plasnnd pAL5000 Km Temperature-sensitive orzw, transfer by coqugation, carries Tn6 11 (43)
pCG63 Plasmtd pAL5000 Km Temperature-sensitive orzM (44)
pCG79 Plasnud pAL5000 Km&m Temperature-sensitive orzM carrying Tn611 (23)
pMYGFP 1 Plasmtd pAL5000 Km Green-fluorescent protein (GFP) reporter (17)
pMYGFPhsp60 Plasrmd PAL5000 Km hsp60-GFP fusion (17)
pMRRO0 1 Plasmid pAL5000 Hgd Mercury resistance (45)
pENlO Plasmid PAL5000 Hg Expression vector to produce extracellular proteins (46)
pRCX3 Plasmtd pAL5000 Km xyZE reporter gene (15)
pSUM series Plasmid pAL5000 Km Blue/white screening for clonmg (47)
pPE207 Plasmid pAL5000 Ame origin of transfer from RSF 1010 (conJugative) (48)
“Km-kanamycm resistance
*Sm-streptomycin resistance
‘Hyg-hygromycm resistance
dHg-mercury resistance
eAm-apramycm resistance
foorrM-ongm of rephcatlon m mycobactena
Parish and Stoker
3. Molecular Techniques: General Problems
Most of the problems associated with the use of molecular techniques to
study mycobacteria can be related to one of the following problems.
3.1. Slow-Growth Rate
The slow growers can take up to six wk to form colonies on a plate, and even
the fast growers may take up to two wk. This can lead to problems with con-
tammation of cultures, especially with fungi. Cycloheximide can be used to
overcome the problem of fungal contammation somewhat. Most commonly
used media have been developed for the isolation of mycobacteria from clmi-
cal samples; although they contam partially-selective agents, such as malachite
green, many require supplementation with nonautoclavable constituents, com-
pounding the problem of contammation. Other media that do not require supple-
mentation have been successfully used for fast growers (see Table 2). Media
must be selected carefully, depending on the technique. The long incubation
time can also lead to problems with plates drying out, and care must be taken to
ensure that plates are well-wrapped m Parafilm or clmgfilm.
3.2. Clumping
Owing to the nature of the cell wall, mycobacterlal cells tend to stick together
when grown in liquid culture, and form macroscopic clumps even when grown
with shaking. Some species are more prone to this than others. The clumping
leads to problems because many standard techniques require dispersed cul-
tures (e.g., optical-density measurement), or, ideally, single cells (for platmg,
screening for mutants, and infection of tissue culture cells) The problem can
be alleviated by the addition of detergents; Tween-80 is most commonly used.
However, this may be only partially effective, and may not be possible where it
is important not to alter the outermost layer of the cell wall, or where deter-
gent-sensitive procedures such as phage infection are being used. Somcation
can be used to break up clumps of mycobacteria, but for pathogenic species
this must be carried out with a cup-horn sonicator (inside an appropriate safety
cabinet) rather than with a probe, in order to contain the aerosols generated.
Clumps may also be broken up by passing cell suspensions through a 23-gage
needle several times, but this is unacceptable for pathogenic species owing to
the risk of needle stick injuries and infection.
3.3. Resistance to Lysis
Mycobacteria are particularly resistant to normal methods of chemical lysis
owing to the nature of the cell wall. In addition, the wall contains large quanti-
ties of polysaccharides, which can contaminate subsequent preparations of
nucleic acids. This is discussed in more detail in Subheading 4.
Mycobacteria 5
Table 2
Commonly used media for the growth of mycobacteria
Liquid media Appropriate techniques
Glycerol Alanine Salts DNA preparation
Mtddlebrook 7H9 Broth0 General
Dubos Broth” General
Proskauer and Beck Surface-pellicle growth
Sauton’s General defined medium
Lemco Broth General
Tryptrc Soy Brothb Phage mfectlon
M9 Munmal Medium Protein preparation and NTG mutagenesis
Solid media
7Hll agala General
7H 10 agara General
Lemco agar General
Lowenstem Jensen (LJ) slopes Strain maintenance
M9 Mimmal Agar Protein purrficatron/auxotrophy screemng
Tryptrc Soy Agar Phage mfectron
Top agar Phage overlays
BCG top Phage overlays for BCG
BCG agal” Phage mfections for BCG
aRequn-es supplementation
6No Tween.
3.4. Safefy Aspects
One of the most important problems when working with pathogenic myco-
bacteria is the need for containment (see Chapter 2). Any procedure that
mvolves the generation of aerosols IS potentially dangerous and should be mmi-
mized. Thus, the use of nonpathogenic species as model hosts for genetic stud-
ies is of great rmportance and convenience. The lack of established, disabled
host-vector systems means that relatively straightforward recombinant-DNA
experiments using nonpathogenic hosts will have to be carried out m contain-
ment facilities
3.5. Spontaneous Antibiotic Resistance
As mentioned before mycobacterra can be broadly divided mto the slow-
and fast-growing species. Most slow-growers possesso nly one ribosomal RNA
(rDNA) operon; this unusual situatron means that resistance to agents such as
kanamycm can easily be acquired by mutation in the rDNA operon itself (II),
which is not likely to occur where there IS more than one operon. Therefore,
6 Parish and Stoker
care must be taken when working with such anttbtottcs to mclude appropriate
controls when attempting to introduce plasmids, and transformants should
always be checked for the presence of the desired construct. Most of the fast
growers contain two rDNA operons, and therefore have a much lower rate of
spontaneous resistance to kanamycin.
4. Fractionation
As mentioned in Subheading 3.3., mycobacteria are difficult organisms to
lyse, mainly owing to the strength of the cell wall, and methods to isolate sub-
cellular fractions have been developed to cope with this problem. The additton
of glycme to growing cells can be used m order to weaken the cell wall before
attempting lysis. Intact genomic DNA can be isolated relatively easily, and
many methods-chemical, enzymatic and mechanical--exist to achieve this (see
Chapter 3). RNA isolation poses more problems, because mRNA is very un-
stable with an extremely short half-life; therefore, lysis must be rapid m order
to prevent degradation, and methods--generally mechanical-have been de-
veloped for this (see Chapter 6). The isolation of intact mRNA from mycobac-
teria has allowed much more scope with respect to the study of gene regulation,
and analyses of transcrtptton start-sites and transcriptional control of genes are
now possible. Techniques such as differential display, RNA arbitrarily pnmed-
polymerase cham reactton (RAP-PCR) and reverse transcrrption (RT)-PCR are
now all being applied to these organisms (see Chapters 23 and 24). Such studies
have the potential of dissecting out genes that are up-regulated in VIVO,a nd of
examining the effect of phagocytosis on gene expression, both important topics
in virulence.
The preparation of protein fractions from mycobactena IS relatively straight-
forward. Secreted proteins can be recovered simply from the medium supema-
tant and subsequently concentrated. For cell wall-associated or cytoplasmic
protems, the cells are generally broken open using mechanical methods and
then centrtfuged to separatet he sub-cellular fractions (see Chapter 7). Methods for
the isolation and characterization of nonprotein cell wall components such as
mycolic acids and lipoarabinomannan have been well-developed (see Chapter 8).
5. Genetic Tools
5.1. Vectors
There is a general paucity of mycobacterral genetic tools, and the majority
of plasmid vectors for use m mycobacteria have all been developed from one
plasmid, pAL5000, originally isolated from A4 fort&urn (10). This plasmid
has a low copy number m mycobacteria (less than five), so that preparation of
large quantities of plasmid DNA from mycobacteria presents a problem. Rep-
Mycobacferia 7
hcons derived from other natural plasmids have been used to a lesser extent
(see Table 1). An addmonal problem is that these plasmids do not necessarily
function in all mycobacterial species; for example, pAL5000-derived vectors
function m most species including M. smegmatzsb, ut have been unable to trans-
form Mycobactenum intracellular-e (12).
Plasmid and cosmtd isolation has also presented a problem, owing to the
low-copy number and the poor quality of DNA recovered. Therefore, most
plasmid analyses have been conducted after transfer of plasmid DNA isolated
from mycobacterta to E coli for analysis. Plasmrd and cosmid DNA can be
successfully isolated from mycobacteria, but this requires an adaptation to the
standard lysis methods (see Chapter 4).
5.2. Reporter Genes
Several reporter genes have been used successfully in mycobacteria, mclud-
mg those encoding P-galactosidase (13), chloramphenicol acetyl transferase
(14), catechol2,3 dioxygenase (IS), luciferase (16) (see Chapter 19), and green-
fluorescent protein (GFP; 17) ( see Chapter 20). These can be used to assay
promoter activity and provide information about gene regulation and relative
promoter strengths.
5.3. Expression Systems
Mycobacterlal promoters do not function well in E. coli and many do not
possess the standard consensus sequences. The situatron seems much more
similar to Streptomyces, where several different classes of promoters occur
(IS). There IS still a lack of well-characterized promoters for the expressron of
heterologous proteins. The most widely-used promoters for expression m
mycobacteria have been hsp60 and hsp70, which are constitutively expressed
to a high level and can be further induced by heat shock. Targeting signals,
such as the a antrgen-leader peptide or lipoprotein-attachment signal, can be
used to dtrect the protein towards secretion or the cell wall, respectively (see
Chapter 2 1). The use of inducible promoters, such as that of the acetamrdase of
M smegmatis (I4), may be extremely useful in allowing controlled expression
of genes.
5.4. Mutagenesis
Chemical mutagenesis has been used for isolatton of mutants of several
mycobacterral species (18-22) and has the advantage that spectahzed genetic
vectors are not required, and that mutations, with effects ranging from total
gene inactivation to subtle alteratrons of phenotype, may be isolated (see Chap-
ter 13). However, it has several drsadvantages, including the use of dangerous
chemicals, the mabihty to select cells carrying mutattons, the creation of unde-
8 Parish and Stoker
fined mutants, the possibility of multtple mutattons, and difficulty in locating
the defective gene.
Transposon mutagenesrs has been developed for use m M smegmatis, BCG
and M tuberculosis using different insertion sequences and delivery systems
(23,2#). The apphcabrlity of a particular transposon to a mycobacterlal species
depends on an effective delivery system, relatively random msertron, and the
absence of the msertion sequence from the host strain. In M. smegmatis, a
temperature-sensitive delivery system has been used m order to overcome
the low-transposition frequency found with nonreplicating vectors (23) (see
Chapter14); however, this has not been easy to adapt to pathogenic specres
because of the narrow temperature range at which these species are viable. A
BCG transposon-delivery system has been described using a partrally deficient
plasmid orrgm, but the msertron sequence (IS) from whtch the transposon Tn5
was derived is present in M smegmatis, and consequently Tn5 cannot be used
in this species (24). In addition, the transposition frequency wrth thus delivery
system is low. An efficient phage-based delivery system has recently been suc-
cessfully developed for M tuberculosis (25).
5.5. Recombination
The creation of defined-gene knockouts using homologous recombmatron
has not proved to be straightforward m mycobacterra. Although it has been
achieved with relative ease in the fast-grower M. smegmatis (26-28) (see Chap-
ters 15 and 16), nutial attempts m the slow-grower BCG were unsuccessful
(27). This was partly owing to a reported high frequency of rllegitrmate recom-
bination that resulted after transformation wrth a nonreplicating plasmrd. This
situation was unexpected and is more similar to mammalian systems, where
the frequency of illegitimate recombination is much hrgher than that of
homologous recombination. This has frustrated attempts to create gene knock-
outs in species such as BCG and M. tuberculoszs. Recently, homologous
recombmation has been more successful, with both single crossovers and tar-
geted gene replacement being achieved (29-32) (see Chapter 18). It seems that
the combination of more laboratories attempting this procedure and the devel-
opment of more efficient vectors will improve the situation m the future. Other
slow growers such as certain strains of M intracellulare have been much more
amenable to such studies (12) (see Chapter 17).
5.6. Genome Analysis
The study of mycobacterlal genes ~111b enefit greatly from projects to se-
quence the complete genomes of M tuberculosis, M. leprae, and M. avium.
The sequence of M. tuberculosis H37Rv has been completed (49), and 1s avail-
able in sequence databases (accession number AL123456) or from the Sanger
Mycobacteria 9
Centre (http://www.sanger.ac.uk/), who are also completmg the M. leprae ge-
nome sequence. A second stram ofM tuberculosis (CSU93) and M. avzum are
being sequenced by the Institute for Genome Research (http://www.tigr.org/).
The amount of information arising from these projects is immense, and an inte-
grated database--MycDB-is being used to organize the data (see Chapter 9).
MycDB can be accessed on the World Wide Web at http://ww.biochem,kth.se/
MycDB The ready availability of sequence data for all A4. tuberculoszs genes
1sa n exciting prospect and opens up many avenues for future study. The apph-
cation of pulsed-field gel electrophoresis to mycobacteria 1s another welcome
addition to the tools available for analysis of whole genomes (see Chapter 5).
6. Detection and Diagnosis
The slow-growth rate of mycobacterla has meant that traditional tech-
niques for diagnosing infection take a long time. Obviously, culture is a
lengthy process, and subsequent drug susceptibility testing IS even more time
consummg. The advance of molecular tehcniques has allowed the develop-
ment of alternative technologies that have the advantages of sensitivity, specl-
ficlty, and speed. Many PCR tests are being developed and have been
evaluated usmg clinical samples (33); not only 1s PCR rapid, but it can be
designed to identify mycobacterla at the species level (see Chapter 27). Other
DNA-based tests have also been used. Epidemlologlcal studies have been
greatly aided by the development of restriction fragment length polymor-
phism (RFLP) typing (see Chapter 29), most commonly using the insertlon
element ISdllO as a target, although other polymorphic elements have also
been described (31,35). Spohgotyping provides another rapld means of not
only identifying the species, but also typing strains as well (see Chapter 28).
The addition of such techniques as rDNA sequencing adds to the versatility
of speclatlon techniques (see Chapter 26). Biochemical technrques can be
used in conJunctlon with molecular techniques to provide a high degree of
confidence in asslgnmg species (see Chapter 25). Drug-susceptlbllity testing
has also been Improved with the application of such techniques as polymerase
chain reaction-single-strand conformatlon polymorphism (PCR-SSCP) to
detect mutations m the genes coding for drug targets (see Chapter 30)and the
use of the luclferase-phage system to assay for drug resistance phenotypes
(see Chapter 3 1). Many other new and varied techniques are being developed
and evaluated, and it is to be hoped that this will greatly improve both the
detectlon and effective treatment of mycobacterlal diseases.
7. Conclusions
The basic tools required for molecular analysis of mycobacteria are now
available. They are still limited, but at least they provide the necessary founda-
70 Parish and Stoker
tlons for future progress. The imminent completion of several genome
sequences of both M. tuberculoszs and A4 leprae barely 12 yr after the first
gene libraries were constructed is an extraordinary feat, and the challenge over
the next 12 yr will be to improve the tools available m order to make the best
use of the information now in our hands.
References
1 Clark-Curtlss, J. E. (1990) Genome structure of mycobactena, m Molecular Bzol-
ogy of the Mycobacterza (McFadden, J J , ed.), Academic Press Ltd, London,
UK, pp 77-96
2 Clark-Curtlss, J. E., Jacobs, W. R., Docherty, M A , Rltchle, L R , and Cur&s
III, R. (1985) Molecular analysis of DNA and construction of genomlc llbrarles of
Mycobactersum leprae. J. Bacterlol 161, 1093-l 102.
3 Thole, J E R., Dauwerse, H. G., Das, P K., Groothms, D. G , Schouls, L. M., and
van Embden, J. D A (1985) Cloning of Mycobactermm bow BCG DNA and
expresslon of antigens m Escherzchza ~011. Infect Immun 50, 800-806
4 Young, R. A., Mehra, V., Sweetser, D , Buchanan, T., Clark-Curtiss, J , Davis, R
W., and Bloom, B. R (1985) Genes for the maJor protein antigens of the leprosy
parasite Mycobacterwm leprae. Nature 316,450-452
5 Young, R. A, Bloom, B R , Grosskmsky, C M , Ivanyl, J., Thomas, D., and
Davis, R. W (1985) Dissection of Mycobacterzum tuberculoszs antigens using
recombinant DNA Proc Nat1 Acad Scl USA 82,2583-2587.
6 Jacobs, W R , Docherty, M A, Curtlss III, R , and Clark-Curtlss, J. E. (1986)
ExpressIon of Mycobacterium leprae genes from a Streptococcus mutans pro-
moter m Esherlchla co11 K12 Proc Nat1 Acad Scz. USA 83, 1926-1930.
7. Snapper, S B., Melton, R E., Mustafa, S , Kieser, T., and Jacobs, W. R. (1990)
Isolation and characterization of efficient plasmld transformation mutants of
Mycobacterlum smegmatls Mol Mxroblol 4, 19 1 l-1 9 19
8 Garbe, T R , Barathl, J , Barnrm, S., Zhang, Y., Abouzeld, C , Tang, D ,
Mukherjee, R., and Young, D. B. (1994) Transformation ofmycobacterlal spe-
cies using hygromycrn resistance as selectable marker Mzcrobzology 140,
133-138.
9 Houssaml-Iraqm, M., Clavel-Seres, S., Rastogl, N , and David, H L (1992) The
expression of the Mycobacterlum aurum carotenogenesls operon is not repressed
by the repressor of Mycobacterzum vaccae photoinducible carotenogenesls. FEMS
Mw-obzol Lett. 99,233-236
10 Rauzler, J., Momz-Perelra, J., and Glcquel-Sanzey, B. (1988) Complete nucle-
otlde sequence of pAL5000, a plasmld from Mycobacterium fortuztum. Gene 71,
315-321
11 Bottger, E. C (1994) Resistance to drugs targeting protein synthesis m mycobac-
terra Trends Mw-oblol 2,416-421.
12 Marklund, B I , Speert, D P., and Stokes, R W. (1995) Gene replacement through
homologous recombmatlon m Mycobacterzum intracellulare. J. Bacterrol 177,
61Ow105