A Coming of Age Story: Chlamydia in the Post-Genetic Era

Abstract

Chlamydia spp. are ubiquitous, obligate, intracellular Gram-negative bacterial pathogens that undergo a unique biphasic developmental cycle transitioning between the infectious, extracellular elementary body and the replicative, intracellular reticulate body. The primary Chlamydia species associated with human disease are C. trachomatis, which is the leading cause of both reportable bacterial sexually transmitted infections and preventable blindness, and C. pneumoniae, which infects the respiratory tract and is associated with cardiovascular disease. Collectively, these pathogens are a significant source of morbidity and pose a substantial financial burden on the global economy. Past efforts to elucidate virulence mechanisms of these unique and important pathogens were largely hindered by an absence of genetic methods. Watershed studies in 2011 and 2012 demonstrated that forward and reverse genetic approaches were feasible with Chlamydia and that shuttle vectors could be selected and maintained within the bacterium. While these breakthroughs have led to a steady expansion of the chlamydial genetic tool kit, there are still roads left to be traveled. This minireview provides a synopsis of the currently available genetic methods for Chlamydia along with a comparison to the methods used in other obligate intracellular bacteria. Limitations and advantages of these techniques will be discussed with an eye toward the methods still needed, and how the current state of the art for genetics in obligate intracellular bacteria could direct future technological advances for Chlamydia.

INTRODUCTION

Chlamydia spp. are ubiquitous bacterial pathogens that infect a wide variety of animals, including a number of economically important species, as well as humans (1). The Chlamydia spp. of primary concern to human health are C. trachomatis and C. pneumoniae (C. psittaci and C. abortus are serious, but rare, zoonotic pathogens). C. trachomatis serovars D to K and L1 to L3 are the leading cause of reportable, bacterial sexually transmitted infections (STIs) both in the United States and worldwide (2, 3). The number of cases in the United States exceeds one million per year and is likely two- to threefold higher due to asymptomatic infections, particularly in women. Chlamydial STIs in women are highly significant, as infection can lead to pelvic inflammatory disease, life-threatening ectopic pregnancy, or infertility (4). C. trachomatis serovars A to C are responsible for the ocular infection trachoma, which is a leading cause of preventable blindness in underdeveloped regions of the world (5). While C. pneumoniae is primarily a respiratory pathogen that causes pneumonia, infections are associated with an increased risk for the development of cardiovascular disease which remains the leading cause of mortality in the United States (6). Collectively, these pathogens represent a substantial burden on the world economy and human health.

One of the major impediments confronting chlamydial researchers has been a paucity of genetic methods for the creation of mutant strains necessary for the thorough dissection of chlamydial biology in relation to pathogenesis. This limitation has hindered the development of vaccines and therapeutics. The failure to successfully adapt genetic tools widely used in other bacteria for Chlamydia can be largely attributed to its obligate, intracellular lifestyle and unique developmental cycle (7, 8). These bacteria transition between an extracellular, infectious form known as the elementary body (EB) and an intracellular, replicative form termed the reticulate body (RB). Infection is initiated upon binding of the EB to the surface of a cell followed by internalization where the EB resides in a host-derived membrane vesicle termed the inclusion. Within the inclusion, the EB differentiates into the RB, which replicates for approximately 20 h before the population begins an asynchronous transition back into the EB form. After 48 to 72 h, the EBs escape from the cell via cell lysis or a process known as inclusion extrusion (9). During differentiation and growth, gene expression and protein profiles proceed through virus-like early, middle, and late phases of expression. The overall timing of differentiation and cell exit varies between species and is further affected by the cell line used for growth (10–13).

As an obligate, intracellular bacterium, Chlamydia spp. possess reduced genomes ranging from ∼1 to 1.5 Mbp that show little evidence of recent acquisitions of foreign DNA including the lack of restriction modification systems, the presence of few transposons and phages or their remnants, and only a single known horizontally acquired resistance marker (14). The only identified drug resistance marker is tetC, which encodes a tetracycline efflux pump and is found in C. suis (15). Analysis of chlamydial genomes also reveals the presence of Chlamydia-conserved histone-like proteins known to complex with the chromosomal DNA in EBs, an absence of DNA methyltransferases including DNA adenosine methyltransferase (DAM) and DNA cytosine methyltransferase (DCM), and a cryptic plasmid present in most, but not all, species (14, 16, 17). Collectively, the intracellular lifestyle and physiology of Chlamydia present a myriad of road blocks to the development of genetic systems. These road blocks include the following: (i) resolving which bacterial form would be amenable to manipulation, as the EB is relatively metabolically inert and has a highly cross-linked outer membrane presenting a possible barrier to DNA entry, while the RB form is shielded by the host cell, inclusion, and bacterial membranes; (ii) discerning when to apply antibiotic selection in regard to the length of time required for phenotypic conversion after introduction of resistance markers; (iii) determining which antibiotics would provide robust selection, while being cell permeable, safe for the host cell, and not restricted due to clinical usage; (iv) assessing whether Chlamydia would be able to recognize foreign promoters and express foreign genes; and (v) assessing whether plasmid-free species could maintain extrachromosomal DNA. Consequently, the development of genetic methods for Chlamydia has lagged well behind other bacteria though certainly not for a lack of effort. Multiple methods for introduction of chlamydial DNA and cryptic-plasmid-based shuttle vectors were attempted including transfection of infected HeLa cells, conjugal mating with Escherichia coli, calcium phosphate and natural transformation, electroporation, microinjection of RB-stage inclusions, and delivery via a ballistic gun (unpublished).

THE EARLY YEARS

Introduction of foreign DNA into bacteria by horizontal gene transfer occurs via three mechanisms: (i) transduction, in which DNA is delivered by a bacteriophage, (ii) conjugation, in which DNA is directly transferred between bacteria via a conjugal mating bridge, and (iii) transformation, whereby “naked” DNA from the environment is taken up by bacteria. Transformation can occur naturally if the bacteria are naturally competent or artificially via chemical transformation or electroporation. For a review on horizontal gene transfer, readers are referred to Soucy et al. (18).

Genetic work with Chlamydia was initiated by various research groups that characterized the cryptic plasmids found in C. trachomatis and C. psittaci in the mid-1980s (19–22). Variation between the C. trachomatis and C. psittaci cryptic plasmids was identified using hybridization and DNA restriction digestion analysis. These studies also identified the need for an Escherichia coli-Chlamydia shuttle vector, as the cryptic plasmid could not be maintained in E. coli. Full nucleotide sequences of the cryptic plasmids from C. trachomatis serovar E and C. trachomatis serovar L1 determined via Sanger sequencing were reported in 1987 and 1988, respectively (Table 1 summarizes major developments in chlamydial genetics) (21, 22). Such studies typically lay the foundation necessary for the development of shuttle vectors through identification of plasmid replication and maintenance mechanisms. However, due to lack of homology with previously sequenced plasmids, little information on cryptic plasmid replication and maintenance were gained from sequence analysis. More-recent studies, highlighted below, have determined that the cryptic plasmid carries eight open reading frames (ORFs), of which Pgp1 (putative helicase), Pgp2, Pgp6, and Pgp8 (putative integrase/recombinase, only the coding sequence is required) are essential for plasmid maintenance, while Pgp4, which regulates expression of pgp3 and chromosomal genes including glgA, along with Pgp3, Pgp5, and Pgp7 are dispensable (23, 24).

TABLE 1.

Summary of genetic advances for Chlamydia

Genetic advance and species and serovar or strain(s)	Methoda	Comment(s)	Reference
Lateral gene transfer
C. trachomatis L1	Homologous recombination, lateral gene transfer	Lateral gene transfer and homologous recombination between the same strain	37
C. trachomatis, C. muridarum, or C. suis	Homologous recombination, lateral gene transfer	Transfer of the tet(C) allele between species	39
Transformation
C. trachomatis E	Shuttle vector, electroporation, EB	First transformation of Chlamydia; plasmid was unstable	25
C. psittaci 6BC	Homologous recombination, suicide vector, electroporation, EB	Demonstrated feasibility of artificial transformation and homologous recombination	43
C. trachomatis L2 and L2(25667R), a plasmid-free strain	Shuttle vector, chemical transformation, EB	First demonstration of stable plasmid maintenance; use of bla in C. trachomatis serovars D to K is restricted by NIH regulations	26
C. trachomatis L2	Shuttle vectors, chemical transformation, EB	MCSb with and without incDEFG promoter and terminator; fluorescent protein expression	48
	Shuttle vector, chemical transformation, EB	Tet-inducible promoter	53
C. trachomatis L2(25667R), a plasmid-free strain	Shuttle vector, chemical transformation, EB	In vitro deletion analysis of all plasmid ORFs; pgp1, pgp2, pgp6, and pgp8 are essential for maintenance	24
C. trachomatis L2	Group II intron (targeted insertions [TargeTron]), suicide vector, chemical transformation, EB	Development of a targeted gene insertion method for reverse genetic approaches	75
Chemical mutagenesis
C. trachomatis D	Targeting-induced local lesions in genome (TILLING), chemical-based (EMS) mutagenesis, infected cells (RB stage)	Reverse genetic approach; CELI digestion of PCR amplicons used to identify mutant for phenotype analysis	28
C. trachomatis L2	Chemical-based (EMS) mutagenesis, followed by lateral gene transfer and homologous recombination, infected cells (RB stage)	Forward genetic approach; high-content chemical mutagenesis and plaque phenotype screening, followed by genome sequencing and genotype/phenotype linkage via lateral gene transfer and homologous recombination	69
	EMS mutagenesis; lateral gene transfer and homologous recombination; shuttle vector (chemical transformation, EB)	Complementation of tepP mutant using a shuttle vector	56
	Chemical-based (EMS and ENU) mutagenesis approach for forward and reverse genetics	Mutant library construction coupled with full library sequencing; library can be searched for desired gene mutations (reverse genetics) or phenotype screening (forward genetics)	70
Dendrimer-enabled transformation
C. trachomatis K	Gene knockdown via antisense oligonucleotides, dendrimer-enabled transformation, infected cells (RB stage)	Targeted gene knockdown in C. trachomatis using antisense oligomers	79
Other
C. trachomatis L2	Axenic medium	Effort to develop axenic medium; seminal event in Coxiella genetics	84

^aThe method of genetic manipulation and the form of the bacterium (EB or RB) used with the method.

^bMCS, multiple-cloning site.

For Chlamydia, the first publication reporting the creation of a shuttle vector would come in 1994 (25). The chloramphenicol-selectable E. coli-C. trachomatis shuttle vector used by Tam et al. (25) had a C. trachomatis serovar E cryptic plasmid backbone and carried the cat resistance gene. C. trachomatis serovar E transformants created through electroporation of the EB form with the shuttle vector possessed chloramphenicol acetyltransferase activity and were resistant to low levels of chloramphenicol (5 μg/ml). However, the plasmid was transient and was rarely maintained beyond four passages in cell culture. Recent data provide some insight into this failed attempt, as the recombinant plasmid lacked a 500-bp region within pgp1, which we now know to be essential for plasmid maintenance. Had serendipity interceded in choosing a different region for insertion of the E. coli vector into the cryptic plasmid, stable transformation may have been achieved much earlier. It would take another 17 years for a stable shuttle vector system to be developed for Chlamydia (26). In the interim, the small genome size of Chlamydia coupled with rapid developments in sequencing technology would set the stage for the genomics era.

ADOLESCENCE: THE GENOMICS ERA

The genome sequence of a C. trachomatis serovar D strain was published in 1998 and would prove to be the first of many (17). As sequencing technology rapidly improved, a field starved for genetics took advantage of the lone available tool to gain insight into the DNA sequences underpinning chlamydial pathogenesis. Currently, more than 150 genome sequences are available for Chlamydia covering all known species and including numerous clinical C. trachomatis strains. These efforts revealed that while Chlamydia spp. are closely related and exhibit strong genome synteny, key differences were present, and these differences were speculatively linked to tissue and host tropisms displayed both across and within species (14). In particular, researchers sought to explain differences in strain virulence through sequence variations affecting metabolic pathways, putative virulence factors, and overall coding capacity. As a prime example, defects in tryptophan synthesis between C. trachomatis ocular and genital serovars should enable genital serovars to synthesize tryptophan from indole (27). This would allow for rescue of growth in the face of gamma interferon (IFN-γ)-stimulated depletion of tryptophan (if indole is provided) for genital serovars, but not ocular serovars. The necessity of the trpBA genes for survival of genital serovars in the presence of IFN-γ would eventually be demonstrated by Kari et al. using a chemical mutagenesis-based reverse genetic approach (28).

A second major outcome of the genomics era was a more thorough appreciation for the potential virulence gene repertoire of Chlamydia including the type III secretion system (previously reported, but not completely mapped [29]), the presence of a family of polymorphic membrane proteins (Pmps), various inclusion membrane proteins (Incs), proteases, etc. The massive sequencing effort of numerous research groups had presented the field with a slew of potential virulence factors and new hypotheses to test, but no methods to perform targeted mutagenesis experiments that could readily be performed in other pathogens. In comparison, by the late 1990s and early 2000s, gene inactivation by homologous recombination and transposon mutagenesis was being undertaken in other obligate intracellular bacteria (30–34). Table 2 summarizes the genetic tools available for other obligate intracellular bacteria.

TABLE 2.

Genetic tools available for other obligate intracellular bacteria

Organism(s)	Methoda	Reference(s)
Anaplasma phagocytophilum and A. marginale	Transposon mutagenesis (Himar1); electroporation	86–88
Coxiella burnetii	Shuttle vector; homologous recombination (single crossover) and episomal maintenance; Himar1 transposon mutagenesis; electroporation (preaxenic medium)	89, 90
	Axenic medium	83, 85
	Shuttle vector (IncQ broad-host-range plasmid); Himar1 and Tn7 transposon mutagenesis; homologous recombination (sacB counterselection and Cre-lox-mediated genome editing); electroporation	91–96
Ehrlichia chaffeensis	Group II intron (TargeTron) and transposon mutagenesis (Himar1); electroporation	77
Rickettsia prowazekii, R. typhi, R. conorii, R. monacensis, R. rickettsii, R. bellii	Homologous recombination (single and double crossover); EZ::TN and Himar1 transposon mutagenesis; shuttle vectors; group II intron (TargeTron); peptide nucleic acids (antisense; translation inhibition); electroporation	30–34, 78, 97–105

^aMethods have not been validated for all species listed in each category.

Analysis of individual genes, particularly ompA and the pmp family members, had hinted that DNA exchange and recombination occurred within the C. trachomatis serovars, but direct experimental evidence was lacking (35, 36). In 2007, Demars et al. (37) asked a straightforward question—was C. trachomatis capable of lateral gene transfer and homologous recombination? Using marked strains of C. trachomatis L1 carrying spontaneous mutations conferring resistance to ofloxacin, trimethoprim, rifampin, or lincomycin, Demars et al. coinfected cells with strains carrying single markers and then selected for both markers. Dual-resistant strains emerged at rates higher than expected for spontaneous mutants and carried the parental mutations, consistent with recombination. This report was followed shortly by a series of studies using similar coinfection strategies to demonstrate that lateral gene transfer and homologous recombination could occur not only between serovars but also between species (38–40). Of particular note, Suchland et al. found that the tet(C) allele from C. suis could be transferred among C. suis, C. trachomatis, and C. muridarum, indicating that horizontal gene transfer and recombination could lead to the spread of resistance (39). These later studies were particularly convincing, as recombinant strains were mapped via whole-genome sequencing, which has also been used to support recombination between clinical strains in vivo (41).

The recognition that lateral gene transfer and homologous recombination could be performed in cell culture and that recombinants could be selected using resistance markers derived from spontaneous mutations would prove critical for developing chemical mutagenesis-based forward and reverse genetic approaches. It still remains unclear how DNA is transferred between strains, although natural transformation is the best supported mechanism. Chlamydia spp. lack genes for conjugation, and the currently known chlamydiaphages all have genomes of less than 4.8 kbp, making them too small to account for the documented transfer of DNA segments exceeding 100 kbp (38, 42). Of relevance for future genetic methods, the small genome size of the chlamydiaphage also makes them unlikely to be useful for transduction approaches (for a review of chlamydiaphage biology, see reference 42).

THE AGE OF DIRECTED MUTAGENESIS

In 2009, Binet and Maurelli reported the first example of targeted mutagenesis in Chlamydia providing proof of principle that Chlamydia could be artificially transformed with foreign DNA and that integration of DNA into the chromosome could be selected (43). Using a suicide plasmid, a 16S rRNA allele carrying mutations introduced by site-directed mutagenesis to impart resistance to kasugamycin and spectinomycin was introduced into C. psittaci via electroporation, and double-resistant mutant strains were obtained. An important feature of the Binet and Maurelli study (43) was the ability to select for recombinant bacteria immediately after transformation via plaque assay, which permitted method optimization. Variables assessed included DNA methylation status, concentration of DNA, number of bacteria, and DNA form (circular or linear). For example, unmethylated plasmid DNA isolated from an hsdS-, dam-, and dcm-deficient E. coli strain was found to be the most efficient for obtaining recombinants. Most current methods for transformation of shuttle vectors continue to use unmethylated DNA, although the efficiency of methylated versus unmethylated DNA for plasmid maintenance has not been reported. This is partly due to the inability to immediately plaque plasmid transformants using the current selection markers, preventing calculation of transformation efficiency. As the Binet and Maurelli study selected for recombinants and not transformants, transformation frequency could not be directly assessed. Development of selection protocols allowing for first passage mutant isolation such as plaque assay or flow cytometry (43, 44) will facilitate expansion of genetic methods by allowing for optimization approaches as exemplified by Binet and Maurelli (43). For example, when transposon mutagenesis methods are established, saturation of the genome will require either efficient transformation or repeated transformations with poorly efficient methods.

SHUTTLE VECTORS

After the breakthrough report of Binet and Maurelli (43), another 3 years would pass before the landmark study by Wang et al. detailing the development of a stable E. coli-C. trachomatis shuttle vector (26). Two different shuttle vectors were created: pBR325::L2 based on the C. trachomatis L2 cryptic plasmid (45) and pGFP::SW2 based on the cryptic plasmid from a Swedish clinical C. trachomatis E isolate carrying a deletion in CDS1 (pgp7) and a 44-bp duplication in CDS3 (pgp1) (46). These vectors were used to transform cryptic-plasmid-positive and cryptic-plasmid-negative C. trachomatis L2 strains. Significant results from this study include the following. (i) A CaCl₂ chemical transformation method for EBs was described, and this method serves as the basis for the majority of subsequently published transformation protocols. (ii) It was demonstrated that penicillin (via the bla resistance marker) can be used to select for transformants. (iii) Both plasmid-negative and plasmid-positive C. trachomatis isolates can be transformed. (iv) Selection for the shuttle vector will cure the strain of its native cryptic plasmid. (v) Foreign promoters and genes (cat, bla, and gfp) can be expressed in C. trachomatis. It should be noted that while chloramphenicol resistance was demonstrated with the transformants (the pGFP::SW2 vector carries both bla and Φcat-gfp), penicillin was used for initial selection. Xu et al. later demonstrated that chloramphenicol could be used for initial selection using a modified pGFP::SW2 vector lacking the bla gene (47). This finding is relevant, as the use of bla-mediated resistance is not permitted in C. trachomatis serovars D to K in the United States without special approval from the National Institutes of Health.

Following the seminal report from Wang et al. (26), numerous reports of modified shuttle-vector systems began to appear in the literature. These vectors used either the pSW2 or pL2 cryptic plasmid backbone reported by Wang et al. (26) and increased the utility of the shuttle vector through the following: (i) development of an inducible TetR-based promoter system; (ii) introduction of multiple cloning sites with and without promoters to allow for constitutive expression or regulation by native promoters; (iii) expression of a variety of fluorescent proteins including green fluorescent protein (GFP), mCherry, cyan fluorescent protein (CFP), and mKate to facilitate live cell imaging and easier identification of transformants; and (iv) adaptation of systems for measuring protein translocation (23, 48–54). In addition, bsd (conferring resistance to blasticidin) was validated providing a third selection marker (55). These tools have enabled previously impossible research such as complementation of chromosomal mutants and analysis of type III effectors using C. trachomatis rather than surrogate bacterial systems (49, 51, 56). The ability to freely manipulate the cryptic plasmid backbone in shuttle vectors also led to a more thorough understanding of cryptic plasmid genes required for plasmid maintenance as well as virulence (24, 57–59). Vector platforms and transformation approaches are summarized in Fig. 1.

FIG 1.

Summary of transformation methods available for Chlamydia. The vector backbones drawn in light blue represent regions based on E. coli plasmids (pE. coli). The suicide vectors are used for insertional mutagenesis (intron) or allelic exchange (16/23S rRNA vector). Available markers for selection of transformants and markers that provide a trackable phenotype are listed to the left of the cell diagram.

As the shuttle vector repertoire expanded, it became apparent that the pSW2- and pL2-based plasmids were not compatible with all Chlamydia spp. Work by Song et al. demonstrated that C. trachomatis serovar A, C. muridarum, and C. trachomatis serovar L2 could be transformed only with shuttle vectors possessing their own cryptic plasmid as a backbone (60). Subsequently, Wang et al. determined that the plasmid tropism between C. muridarum and C. trachomatis is due, in part, to CDS2 (pgp8) (61). These results indicate that species-specific shuttle vectors may be important for expanding genetic methods into other C. trachomatis serovars and Chlamydia spp. (58). However, some caveats to plasmid tropism exist. pGFP::SW2, a serovar E cryptic-plasmid-based shuttle vector, is stable in C. trachomatis serovars L2, D, and F and a C. trachomatis L2 cryptic-plasmid-based shuttle vector was reported to be stable in C. pneumoniae (26, 50, 52, 55). The latter study used a dendrimer-DNA complex to facilitate selection-free transformation (50). These varied reports make it clear that while our understanding of chlamydial plasmid biology has greatly improved in the age of shuttle vectors, much remains unknown. Consequently, it is crucial that experiments performed with recombinant plasmid strains be matched with strains carrying the empty vector rather than matched to strains lacking a plasmid or carrying the native cryptic plasmid.

While the robust chemical transformation protocol pioneered by Wang et al. (26) and the rapidly expanding shuttle vector repertoire allows the field to push the boundaries of chlamydial genetics, some limitations remain. The identification of multiple resistance markers will be essential for allowing gene complementation along with construction of strains carrying multiple mutations. As these new resistance markers are assessed, it is necessary to be mindful of spontaneous mutation rates that lead to resistance as well as the differential susceptibility of species to antibiotics. These problems are highlighted by failed attempts to utilize the rifampin resistance marker arr-2 for selection in C. trachomatis L2 and the natural resistance of C. muridarum to ofloxacin (62, 63). In addition, the mechanism of drug action may also impact the efficacy of selection. While the bla marker is prohibited for use in certain serovars, it has proven robust and is the marker of choice for most researchers. The robustness of bla-mediated resistance may be due, in part, to the fact that β-lactams appear to be bacteriostatic for Chlamydia inducing a persistent growth state rather than death (64). This may allow for increased time for phenotypic conversion, resulting in an increased rate of transformant recovery. It will also be important to further flesh out how protocols and vectors affect transformation efficiencies, particularly when attempting to maximize transformation efficiencies for random insertional mutagenesis methods such as transposon mutagenesis. The easiest method to measure transformation efficiencies is through plaque assay (65). In this assay, a confluent host cell monolayer is infected with EBs, and then an agarose overlay is added to the infected cells. Upon cell lysis and release of EBs, only cells adjacent to the infected cells can be infected due to the agarose overlay. Repeated rounds of development lead to a localized clearing of the monolayer referred to as a plaque. Plaque assay is also a useful method for obtaining clonal strains, although fluorescence-activated cell sorting methods appear promising for recombinant strains expressing fluorescent proteins (44). Drugs such as blasticidin and zeocin, which are toxic to both host cells and Chlamydia, impede the use of the plaque assay (55). In addition, chloramphenicol at high doses or prolonged treatment periods can also be toxic for the host cell (66).

TARGETING THE CHROMOSOME

Accompanying the description of a shuttle vector for C. trachomatis by Wang et al. (26) were two studies reporting the use of ethyl methanesulfonate (EMS) (an alkylating agent that causes G-to-A transition mutations) for chemical mutagenesis of C. trachomatis. In the first study, Kari et al. used low levels of EMS targeting a mutation rate of one base per genome to generate a library of C. trachomatis mutants (28). The library strains were then pooled and assessed for the desired mutation using PCR, followed by CELI digestion. This method is known as targeting-induced local lesions in genome, or TILLING (Fig. 2A). To detect mutants, PCR amplicons from each pool are heat denatured and then allowed to anneal. If a pool contains wild-type and mutant strains, a heteroduplex will form due to a base mismatch, which allows for cleavage by CELI and detection of the cleaved product by DNA gel electrophoresis. The desired clones can then be obtained from the mutant library by PCR screening of clones obtained through limiting dilution or plaque assays followed by sequencing to map the exact mutation. Mutant phenotypes can then be assessed to determine gene function. Once the EMS mutant library pools are established, one can search for all desired mutants within the pool. Limitations of this method include extensive hands-on work to both create the library and identify mutants along with back-end whole-genome sequencing to confirm the clonal isolate contains a mutation only in the gene of interest. This reverse genetic approach has been used to study the function of trpB and pmpD in C. trachomatis and has been adapted for use with C. muridarum (28, 67, 68).

FIG 2.

Chemical mutagenesis methods used for TILLING (A) or forward genetic approaches (B) are shown. The green and red lines in the wells in the TILLING section represent wild-type and mutant amplicons, respectively (see the text). Both methods require whole-genome sequencing to map mutations and may require lateral gene transfer and homologous recombination between the mutant (yellow) and parental (green) strains to isolate mutant alleles. Chemical mutants (and intron-generated mutants) can be complemented using shuttle vectors (Fig. 1).

EMS has also proven to be a useful tool for forward genetics in C. trachomatis. Using increased amounts of EMS compared to those used by Kari et al. (28) to generate 3 to 20 mutations per genome, Nguyen and Valdivia (69) prepared a library of mutants that were then assessed for growth defects via automated analysis of plaque morphology (Fig. 2B). Strains giving rise to altered plaques were isolated, expanded in cell culture, and genotyped using whole-genome sequencing. As strains contained multiple mutations, lateral gene transfer and homologous recombination as initially described by Demars et al. (37) was then used to link gene lesions with specific phenotypes. This high-content mutagenesis approach has been further expanded by Kokes et al. (70) using either EMS or N-ethyl-N-nitrosourea (ENU) (alkylating agent leading to A-to-T transversion mutation) to mutate C. trachomatis, creating a collection of 934 mutant strains carrying 6 to 25 mutations. Strains were screened via plaque morphology, and whole-genome sequencing was used to map all mutated alleles within the pools of mutant strains. The availability of this sequence data allows mutants of interest to be selected and assessed for their phenotype following segregation of the allele using lateral gene transfer and homologous recombination.

Chemical mutagenesis is clearly a powerful tool for discerning the roles of genes in the biology of Chlamydia. A wide range of mutants have been obtained using these approaches including numerous strains with nonsense mutations that will be particularly useful for studying the impact of gene loss on strain growth in cell culture and animal models. It is important to note that nonsense mutations are less susceptible to polar effects observed with random insertional approaches such as transposon mutagenesis. Chemical mutants can be complemented using the various existing shuttle vectors, and these approaches should be readily adaptable to other Chlamydia spp. The primary weaknesses lie in the cost and effort associated with mutant generation and genotype analysis and the need to perform lateral gene transfer to isolate specific mutant alleles when using large amounts of EMS or ENU. In addition, isolation of chemically generated mutants via plaque assay selects against mutants that have lost the ability to form plaques, and screening for plaque morphology variants may miss strains that are competent for growth in cell culture but would be attenuated in an animal model. The latter limitation would be difficult to overcome with a chemical mutagenesis approach, but it could be dealt with using transposon mutagenesis approaches such as signature-tagged mutagenesis to screen mutant pools for genes required for in vitro growth versus in vivo growth (71, 72).

Currently, the only method available for generating targeted chromosomal mutations in C. trachomatis utilizes a mobile group II intron, Ll.LtrB, originally isolated from Lactococcus lactis and marketed as TargeTron by Sigma-Aldrich and by TargeTronics, LLC (73). Targeting the intron to a gene of interest is performed using PCR to mutate bases within the 5′ region of the intron to allow for hybridization with the target gene. The required mutations are identified using a proprietary algorithm. With the assistance of an intron-encoded protein, LtrA (expressed independently from the intron), the intron is spliced into the target gene creating an insertional gene mutant (74). The intron is introduced into C. trachomatis on a suicide vector, and selection of insertion is performed using ampicillin (the intron carries a bla marker). Johnson and Fisher initially validated TargeTron in C. trachomatis L2 through insertional inactivation of incA (75). Intron insertion is stable in the absence of selection, and at least one insertion site has been identified in each C. trachomatis ORF. Recently, a second selection marker, aadA (spectinomycin resistance), was validated for intron-based mutagenesis in C. trachomatis and has been shown to allow creation of dual insertion mutants (62, 76). As the intron can carry cargo, this approach could also be used for complementation approaches or to insert foreign genes into the chromosome. The major limitation of this approach holds true for all gene insertion methods, namely, the possibility of polar effects. This can make it difficult to distinguish if the resulting phenotype is due to inactivation of a downstream gene (addressable by complementation). Additionally, failure to generate a mutant could result from improper transformation/intron function or polar effects acting on an essential gene, rather than the essentiality of the targeted gene. The Ll.LtrB group II intron has recently been modified to allow for Cre-mediated recombination, which could be adapted for genome editing or marker recycling, and a second intron-based method using an EcI5 intron has been developed that is more efficient than the LtrB intron (not evaluated in Chlamydia) (74). The TargeTron system has also been used in the obligate intracellular pathogens Rickettsia rickettsii and Ehrlichia chaffeensis (77, 78).

FUTURE PROSPECTS

The field of chlamydial genetics has seen major advances over the past 5 years. We can now generate targeted mutants, carry out large-scale chemical mutagenesis for forward and reverse genetic approaches, express foreign and recombinant genes using various shuttle vectors, and we are positioned to begin generating conditional lethal mutants by combining inducible promoter systems with chemical or targeted mutagenesis. The latter approach may prove to be vital for studying many chlamydial genes that may be essential due to the reduced genomes and obligate intracellular lifestyle of these pathogens (17). Alternative approaches available for studying essential genes in Chlamydia include regulated overexpression of wild-type genes or dominant-negative mutant genes along with further development of antisense RNA approaches described by Mishra et al. (79). In addition, validation of fluorescent protein expression in C. trachomatis coupled with detection of fluorescent EBs via flow cytometry (or enumeration in plaque assay) should allow for optimization of transformation procedures as new methods of mutagenesis are explored (44). Along this line, transformation via electroporation as utilized by Binet and Maurelli (and the method of choice for transformation in other obligate, intracellular bacteria) is worth revisiting to assess efficiency versus chemical transformation (43). Collectively, the currently available tools have provided important insights into chlamydial pathogenesis that could not have been ascertained as recently as 4 years ago, such as elucidating the function of the secreted protease, CPAF (chlamydial protease-like activity factor) (80).

However, the chlamydial field still lacks numerous genetic methods available for other bacterial pathogens. Major needs include methods for allelic exchange and identification of broad-host-range plasmids to allow for complementation and/or expression of recombinant genes without loss or alteration of native cryptic plasmids as well as use of a single shuttle plasmid across species/strains. Another major limitation is the absence of transposon mutagenesis systems for both random mutagenesis and targeted delivery of genes to the chromosome as exemplified by the Tn7 platform (81, 82). For many of these approaches, we can look to other obligate intracellular bacteria for inspiration (Table 2). A plethora of tools are available for manipulating Coxiella burnetii, the majority of which were established after the development of axenic growth conditions for the bacterium (83). Axenic growth conditions would likely yield similar fruit for Chlamydia (84). In addition, transposon mutagenesis has been successfully employed in Rickettsia spp., E. chaffeensis, Anaplasma spp., and C. burnetii prior to establishment of axenic growth conditions supporting the feasibility of transposon mutagenesis for Chlamydia (Table 2). While it remains impossible to predict the future, it seems clear that in regard to chlamydial genetics, it will be bright.