Sigma 54-Regulated Transcription Is Associated with Membrane Reorganization and Type III Secretion Effectors during Conversion to Infectious Forms of Chlamydia trachomatis

The factors that control the growth and infectious processes for Chlamydia are still poorly understood. This study used recently developed genetic tools to determine the regulon for one of the key transcription factors encoded by Chlamydia, sigma 54. Surrogate and computational analyses provide additional support for the hypothesis that sigma 54 plays a key role in controlling the expression of many components critical to converting and enabling the infectious capability of Chlamydia. These components include those that remodel the membrane for the extracellular environment and incorporation of an arsenal of type III secretion effectors in preparation for infecting new cells.

There are many aspects of 54 transcriptional initiation that are highly conserved between bacterial phyla and distinct from those of 70 family members (21). A primary difference is that the promoter recognized by 54 is centered at the Ϫ12 and Ϫ24 positions upstream of the transcriptional start site (TSS) rather at than the typical Ϫ10 and Ϫ35 positions (22,23). Unlike 70 -type sigma factors, which are able to spontaneously separate double-stranded DNA and initiate transcription after forming the RNAP holoenzyme (3,24), sigma 54 is incapable of transitioning from the closed complex to the open complex on the DNA without the assistance of an ATPhydrolyzing response regulator (typically referred to as NtrC).
Sigma 54 response regulators are typically composed of three domains: a DNAenhancer binding domain; an effector ATPase domain with a conserved glutamate-242 residue that polarizes the ATP molecule, enabling hydrolysis to occur; and a receiver domain that is phosphorylated by a sensor kinase (NtrB) in response to an environmental cue (4,(25)(26)(27). Phosphorylation of this receiver domain relieves inhibition of the ATPase domain and allows NtrC to hydrolyze ATP, enabling the 54 -RNA holoenzyme to form an open DNA complex and initiate transcription (1,25). Previous studies have shown that in the absence of the receiver domain, the ATPase domain alone can initiate 54 -directed transcription (28)(29)(30). The regulatory cues that trigger the signaling cascade activating 54 differ in various bacterial species, as do the regulons of genes that 54 is responsible for transcribing. Despite the variations in the activating signals and subsets of genes being regulated in different bacteria, this mechanism of regulation provides tight control of the 54 regulon that is expressed only under specific conditions (1).
Chlamydia bacteria are obligate intracellular organisms that code for a 54 homolog, along with two 70 family factors ( 66 and 28 ). 66 and 28 have previously been shown to be important for controlling various stages of temporal gene expression during the developmental cycle of Chlamydia (31,32); however, the role of 54 has yet to be determined. The metabolically active and replicative form of Chlamydia is termed the reticulate body (RB), for which 66 directs transcription of most (ϳ80%) of the encoding genes, and most of the products are associated with metabolism, replication, and maintenance of the intracellular environment (33,34). As the developmental cycle progresses, RBs asynchronously convert into the infectious, spore-like form of Chlamydia, termed the elementary body (EB). During this stage of the developmental cycle, when RB-to-EB conversion is occurring, global gene expression profiling supports the hypothesis that about 20% of the genome is upregulated (33,34). Expression of many of these genes has been shown to be regulated by 66 , with additional factors, such as DNA topology and supercoiling, expected to contribute to this differential expression mechanism (35)(36)(37). 28 has been shown to regulate few (but critical) genes during the RB-to-EB conversion process, including those involved in the formation of the condensed DNA structure that is unique to EBs (38)(39)(40)(41)(42).
Chlamydial EBs are transcriptionally silent due to a highly condensed nucleoid and have limited metabolic activity and an extensive network of disulfide cross-linked outer membrane proteins for osmotic stability. In spite of the minimal metabolic activity, EBs have extensive infectious capabilities that enable host cell adherence, invasion, and the establishment of the intracellular environment (43). EBs have been shown to bind to host cells using multiple membrane proteins (OmcB, OmpA, and Pmps) and subsequently invade and establish an intracellular vacuolar environment (i.e., inclusion) through the use of a type III secretion system and preformed effector proteins, such as TarP, Incs, and others (44,45). Because EBs have minimal metabolic activity, the proteins needed for these critical infection processes need to be prepackaged before the RB-to-EB conversion and exit from an infected cell (46). Therefore, at the late stages of the developmental cycle, transcription of genes for membrane remodeling and structural transition to the EB and the arming of necessities for reinfection must occur. While many aspects of the temporal regulation of the developmental cycle have been determined, there are still gaps in the understanding of important details, including how the chlamydial cells regulate the equipping of the EB with the necessary components for survival in the extracellular environment and the next round of host cell infection.
In this study, the regulon of 54 in the human pathogen Chlamydia trachomatis was determined using the ATPase domain of the NtrC homolog and CtcC response regulator protein. Computational investigation of 54 promoter elements and reporter gene expression analyses supported the accuracy of determination of this regulon. From the proposed regulon, 54 is hypothesized to regulate transcription of genes involved in the RB-to-EB transition, including genes encoding many membrane proteins and type III secretion effector proteins known to be important for early stages of EB infection.

RESULTS
The effector (ATPase) domain of chlamydial 54 activator protein CtcC exhibited ATP hydrolysis in the absence of a regulatory domain. CtcC consists of an N-terminal regulatory domain and a C-terminal effector ATPase domain but lacks a DNA-binding domain (Fig. 1A). Figure S1 in the supplemental material shows an amino acid alignment of the chlamydial CtcC protein with homologous response regulatory proteins from other bacterial species. Without the DNA-binding domain, the chlamydial CtcC protein was found to have the highest sequence identity with AtoC from E. coli at 42.2% (62% sequence similarity) (47). Multiple features are conserved between CtcC and other NtrC homologs, including the E242 residue important for ATP hydrolysis (see Fig. S1 in the supplemental material, red arrow), the predicted 54 interaction site (Fig. S1, red box), and the site of phosphorylation by the sensor kinase protein (Fig. S1, red asterisk) (48).
The response regulator CtcC (CT466) has previously been shown to be phosphorylated in vitro by the sensor kinase CtcB (CT467) (47); however, ATPase activity or the inhibitory effect of the regulatory domain has yet to be investigated. To assess this, in vitro ATPase activity was evaluated for full-length CtcC, CtcC with an ATPase domain (amino acids [aa] 142 to 386), and CtcC with an ATPase domain with an active-site substitution (E242A). The full-length CtcC protein exhibited very low ATPase activity, with less than a 10% depletion of total ATP (Fig. 1). In contrast, the ATPase domain reduced ATP levels by over 40%. When an active-site substitution (E242A) was introduced into the ATPase domain, ATPase activity was severely affected. Overall, these data support the idea that the absence of the regulatory domain enables ATPase activity of the ATP domain of CtcC.
Expression of the ATPase domain revealed the 54 regulon in C. trachomatis. Previous studies expressing only the ATPase effector domain of NtrC homologs have demonstrated constitutive activation of 54 -directed transcription in Salmonella enterica serovar Typhimurium and Sinorhizobium meliloti (28,30). Given the observations revealing that the CtcC ATPase domain exhibited ATP hydrolysis activity, expression of this domain in Chlamydia was hypothesized to activate 54 RNAP and enable the discovery of a cognate regulon. Before evaluating the transcriptional profile, protein induction of full-length CtcC, CtcC with an ATPase domain, or CtcC with an E242A ATPase-inactive variant was evaluated in Chlamydia ( Fig. 2) (49). Limited but equal levels of expression of wild-type CtcC were detected in vector control samples (Fig. 2B). Full-length CtcC was strongly induced; however, CtcC levels were considerably higher than native expression levels prior to induction, indicating leaky expression (Fig. 2B). With the overexpression of the full-length CtcC protein, there appeared to be a faint band of approximately the same size as the ATPase domain band as well, and while it is possible that this might have represented a degradation product, the identity of this band has not been fully investigated. Importantly, induced expression of the ATPase domain and of the E242A variant was evident (Fig. 2B).
To discover a potential 54 regulon in Chlamydia, RNA was isolated from infected samples with vector only or from a construct harboring the ATPase domain after 4 h of induction (16 to 20 h postinfection [hpi]). Stranded RNA sequencing (RNA-seq) analysis was performed on RNA from two biological replicates, with results revealing that transcripts for 64 genes were significantly and reproducibly upregulated in both samples following induced expression of the CtcC ATPase domain (Table 1; see also  Table 2). An initial and striking observation related to this potential regulon was that all of the genes found to be upregulated following ATPase domain induction are normally upregulated during the middle-late stages of the Chlamydia developmental cycle (33,34). These stages coincide with the RB-to-EB conversion, as well as with other late infection and exit events. Twenty-four genes are classified as mid-late-stage genes and are normally upregulated after 16 to18 hpi during the developmental cycle, and 40 genes are late-stage genes, normally upregulated after 24 hpi (33,34). Of the 64 genes, 41 (64%) were found to encode hypothetical proteins with no known or only putative functions, and 23 (36%) were functionally annotated. Nineteen of the genes code for membrane proteins, 6 of which are predicted to be inclusion membrane proteins. Twenty-eight genes encoded type 3 secretion system (T3SS)-associated proteins, including effector proteins such as TarP. To validate this expression profile, droplet digital PCR (ddPCR) was performed on newly isolated RNA for a subset of 19 genes selected from the 54 overactivation RNA-seq data ( Table 1; see also Table 2). All of these genes displayed upregulation following induction of the CtcC ATPase domain (see Table S1 in the supplemental material). Several genes evaluated by ddPCR were found to have expression ratios more than twice as high as those seen with RNA-seq. This difference is possibly due to the additional processing that RNA prepared for RNA-seq must go through, including exonuclease treatment and rRNA depletion steps, although RNA stability and loss have not been evaluated for each step (50). To determine if pleiotropic transcription effects should be considered, the expression ratio was determined for rpoA, which encodes the alpha subunit of RNA polymerase and is transcribed by 66 holoenzyme. Expression levels of rpoA were unchanged following induction of the CtcC ATPase domain. To demonstrate that transcription induction was specific to CtcC ATPase domain activity, ratios were also determined following induction of the E242A ATPase domain. Levels of transcription of the 19 54 regulon genes and rpoA following induction of the ATPasedefective form were unchanged relative to the vector control sample (Table S1).
When the full-length CtcC construct was induced, transcript ratios for most of the 19 genes were found to be considerably lower than those observed with the ATPase domain, supporting the idea that the regulatory domain was still inhibiting ATPase activity in vivo, similarly to the observations of the in vitro ATPase depletion study (Fig. 1). Interestingly, many 54 genes (ct456, ct619, ct620, ct683, ct711, and ct875) were transcribed at ratios twice those observed for the ATPase-defective domain (E242A) whereas rpoA levels were unchanged. This observation suggests that overexpression of the full-length CtcC enabled transcriptional activation for some 54 genes. This limited transcriptional activation might have been contributed by the minimal ATPase activity, as was observed in vitro for the full-length form (Fig. 1B), but amplified by protein overexpression. Alternatively, it might have been due to CtcC autophosphorylation or CtcB phosphorylation of expressed CtcC and subsequent transcriptional activity.
Overall, RNA-seq and ddPCR validation studies support the identification of a putative 54 regulon in C. trachomatis that is associated with mid-late-stage gene expression and RB-to-EB conversion components. To confirm that these components were expressed and present during these stages of the developmental cycle, ddPCR was used to calculate the transcript levels of rpoN ( 54 ), ctcB, and ctcC at 12, 18, 24, and 30 hpi. The transcript levels for 54 were found to be relatively constant, while ctcC and ctcB levels increased throughout the developmental cycle (Fig. S2). Interestingly, the  (Table S2). c Transcript reads from RNA-seq data suggest a different transcriptional start site compared to current C. trachomatis 434/Bu genome (NC010287). Figure S3 shows PCR confirmation of alternative transcriptional start site. d Two standard deviations from average in one of two RNA-seq experiments.
CtcB sensor kinase had very low levels of transcription early in the developmental cycle before they rose above those of 54 after 24 hpi. These results mirror those in prior studies in support of the timing and function of this gene regulatory system (33,34,51). Computational analysis identified 54 promoter sequences upstream of 33 54 regulon genes. One study has investigated the presence of 54 promoters in Chlamydia, with results indicating that nine genes are preceded by a consensus sequence (52). Given that 64 genes appear to be part of the 54 regulon, a broader computational investigation of putative 54 promoters in the C. trachomatis genome was applied. The 54 consensus sequence TGG-N 9 -TGC, allowing one mismatch and a variable length of linker region of between 8 and 10 nucleotides (nt), within 400 bases upstream of a putative 54 gene target was used for the search (23). Of the 64 putative 54 -regulated genes, 28 were found to be preceded by a 54 promoter consensus sequence and are considered to be direct 54 targets (Table 1; see also Fig. S3). Five genes that were included in the 54 regulon were not preceded by 54 promoters but are expected to be contained within an operon that is controlled by 54 promoter (Table 1; operon). These 28 promoters were used to generate a consensus Chlamydia 54 promoter that reflects the diversity of nucleotides at individual sites (WebLogo; Fig. 3A) (53).
Interestingly, four genes (ct105, ct619, ct711, and ct875; all hypothetical proteins) have predicted 54 promoters inside the currently annotated open reading frame; however, based on the RNA-seq transcript reads, the transcriptional start site (TSS) appears to be downstream from the start codon and is in line with the 54 promoter  (Table S2). c Two standard deviations from average in one of the two RNA-seq experiments.
Role and Regulon of Sigma 54 in C. trachomatis  54 or CtcC present (vector control) and with each protein expressed individually in order to assess the system for induction from the E. coli 54 . (C) Additional promoters were tested from the predicted 54 regulon, looking at the difference between the constitutively active CtcC ATPase and the inactivated E242A mutant. The predicted regulatory sigma factor is indicated below the promoter region. Experiments were performed in triplicate. * and **, P values of Ͻ0.05 and Ͻ0.01, respectively, by Student's t test comparing the CtcC ATPase to the E242A ATPase mutant; †, P value of Ͻ0.05 by Student's t test comparing the CtcC E242A ATPase mutant to a promoterless pACYC-lacZ negative control. ns, not significant. (D) Two point mutations were introduced into the ct620 promoter (alignment below graph). The level of LacZ activity was decreased significantly compared to the wild-type promoter but was still increased significantly compared to the inactive conditions. *, P value of Ͻ0.05 by Student's t test. All beta-galactosidase assays were performed in triplicate.
( Fig. S4). In order to investigate if the transcriptional start sites are internal components of the annotated start codon and to validate the 54 promoter for these genes, primers were designed to lie on either side of the alignment of the RNA-seq reads (Table S4). PCR of cDNA confirmed that transcripts occurred only with the internal primers for all four genes, indicating that the transcripts start internally to the annotated start codon for these genes. A new start codon for these genes was selected as the next in-frame ATG, and the location of that start codon is indicated as the distance from the 54 promoter (Table 1). These observations support the idea that 33 genes are direct targets of 54 (Table 1). This also suggests that 31 genes are indirect 54 targets and are regulated by an unknown factor(s) ( Table 2). Reporter gene assays supported 54 -and ATPase-dependent regulation of selected 54 gene promoters. To provide further support for direct targets of 54mediated transcription, ␤-galactosidase assays were performed in E. coli with inducible Chlamydia 54 and CtcC ATPase domain. Induction of Chlamydia 54 and CtcC ATPase proteins resulted in ␤-galactosidase activity when ct084 and ct142 promoter regions preceded lacZ (Fig. 3B). Both of these genes were upregulated with 54 activation and predicted to have 54 promoters (Table 1; see also Fig. S3). 54 or the inactive CtcC ATPase domain, alone or in combination, was unable to initiate transcription (Fig. 3B), supporting the hypothesis of specificity and functionality of this surrogate system. Promoter regions for two genes, ct456 and ct576, did not enable transcription with or without 54 and/or the CtcC ATPase present. These genes were not predicted to have 54 promoters but were upregulated with 54 activation. Interestingly, both ct456 and ct576 promoters result in ␤-galactosidase activity without additional factors, supporting the idea that these promoters are active in E. coli, which corresponds with results of a prior study (37).
Six additional putative 54 promoters were selected for further investigation using this surrogate system and displayed CtcC ATPase-dependent activation (Fig. 3C). The ct456 and ct576 promoters again displayed CtcC ATPase-independent transcriptional activity. Transcription of hctB was upregulated by the CtcC ATPase domain as shown by both RNA-seq and ddPCR (Table 2; see also Table S1). This gene has been shown to be under 28 regulation in late stages of the chlamydial developmental cycle (39)(40)(41)54) and lacks a 54 promoter. The hctB promoter region did not exhibit lacZ activation in the presence or absence of 54 (Fig. 3C); however, when the gene encoding 28 (fliA) was added in conjunction, there was a significant increase in LacZ activity (Fig. S5), showing that the cloned-in promoter region is functional.
The upstream regions for ct394 and ct444 showed a significant increase in ␤-galactosidase activity with the CtcC ATPase E242A mutant compared to the promoterless pACYC-lacZ control (Fig. 3C). While the level of ␤-galactosidase activity seen with CtcC ATPase was significantly greater than that seen with inactive protein, a closer analysis of the upstream regions for ct394 and ct444 revealed that both 54 and 66 promoter sequences were present for these genes (Fig. S6). Interestingly, the 54 promoter for ct394 lies inside the 66 promoter, allowing the possibility that 54 would act as an inhibitor of transcription for this gene when no CtcC activation occurred. For both genes, the presence of these dual promoters could account for the low level of ␤-galactosidase activity seen when no 54 /CtcC ATPase proteins were present.
To further validate the 54 consensus sequence, two conserved bases in the ct620 promoter were mutated and ␤-galactosidase activity evaluated. The central G at position Ϫ24 and the T at position Ϫ13 were both mutated to A. The level of ␤-galactosidase activity promoted by this mutated sequence was significantly lower than that seen with lacZ driven by the wild-type promoter. The ␤-galactosidase levels for the mutated promoter were still higher than those seen with the vector control and with the E242A CtcC ATPase mutant (P value ϭ 0.0339) (Fig. 3D), supporting the idea that the promoter was still functional. These data suggest that the predicted promoter sequence is necessary for 54 -initiated transcription.
Overexpression of CtcC protein variants in Chlamydia resulted in an abnormal phenotype and lower levels of progeny production. Progeny production was eval-uated with Chlamydia-infected L929 cells at 30 hpi either with or without 18 h of protein induction. Addition of anhydrotetracycline (ATc) alone had no effect on progeny formation; nor did the induction of the inactive ATPase domain (Fig. 4A). Fulllength CtcC induction resulted in a 1.4-log-fold decrease in progeny production, while the ATPase domain caused a 1.6-log-fold-lower progeny production level. When no ATc was added to the pL2-ctcC ATPase-only construct chlamydial strain, there was also a significant, 0.5-log-fold decrease in progeny compared to the vector control strain. This decrease was likely due to minimal uninduced production of the CtcC ATPase domainonly protein.
In order to evaluate bacterial cellular morphology and inclusion composition, immunofluorescent microscopy was performed on infected L929 cells with or without 12 h of induction (12 or 24 hpi). Similarly to the progeny production results (Fig. 4A), no observable abnormal bacterial or inclusion morphology was apparent in vector control or ATPase-inactive samples (Fig. 4B). Overexpressing the full-length CtcC protein or ATPase domain resulted in a noticeable decrease in the level of C. trachomatis bacterial cells within the interior of the inclusion, as most organisms were observed almost exclusively around the inclusion membrane (Fig. 4B). Furthermore, the bacterial cells expressing the ATPase domain appeared to be larger, with more diversely sized reticulate bodies. The reduction in the levels of observable organisms in the CtcC full-length and ATPase domain-expressed inclusions (Fig. 4B) matched the reduction of progeny production (Fig. 4A). Importantly, the shared phenotypes observed in CtcC full-length and ATPase domain expression samples are likely a result of pleiotropic effects associated with disrupted signaling and irregular gene expression. Transcriptional analysis of these two samples showed that there was some overlap in gene expression profiles but that the overlap was fairly minimal (Table S1). This could indicate that the induced expression levels of these few shared gene products represent the primary contributors to the disrupted growth phenotypes observed. Finally, the absence of the growth effect with the E242A ATPase-defective mutant supports the idea that ATPase activity and 54 gene expression were the likely causes of these defects, rather than general overexpression of the protein.

DISCUSSION
In this study, the 54 regulon has been determined, and the results support the idea of a role in controlling RB-to-EB conversion processes, especially those involving many components associated with infectious capabilities. As observed with other NtrC homologs, the absence of the CtcC regulatory domain enabled ATPase activity (Fig. 1B). This activity appears to have enabled the induction of 54 transcription in the absence of CtcC phosphorylation by CtcB. This is evidenced by RNA-seq profiling results (Table 1;  see also Table 2), the presence of 54 promoters (see Fig. S3 in the supplemental material), and surrogate reporter gene analyses (Fig. 3). A major observation was that all genes within the resulting 54 regulon were upregulated during the middle and late stages of the developmental cycle, supporting the idea that 54 is critical for controlling gene expression during these developmental phases. Investigation of the products encoded by genes in the 54 regulon further supports the idea of a role in preparing EBs for the next round of infection through remodeling the outer membrane and packaging numerous type III secretion effector proteins. Overall, these observations provide comprehensive support for the previously unknown and critical role of 54 in Chlamydia.
All of the genes making up the chlamydial 54 regulon were normally upregulated during mid-late (16 to 18 hpi) or late (24 hpi) stages of the developmental cycle. Interestingly, omcA and omcB were both found among the genes with 54 promoters and appear to be direct 54 targets. OmcA and OmcB are cysteine-rich proteins that make up a large portion of the highly cross-linked EB cellular envelope whose genes are translated late in the developmental cycle (55). Additionally found to be part of the 54 regulon, ct682 (pdpB), ct455 (murA), and ct841 (ftsH) are all cell wall modification proteins transcribed after 18 hpi (56)(57)(58). This collection of late-stage genes associated with cell membrane composition suggests that the 54 regulon is involved in the RB-to-EB transition. EBs are highly condensed forms of the chlamydial cell with a large degree of disulfide bonds holding the membrane rigid against osmotic pressure outside the host cell, compared to the fragile membrane of RBs (59)(60)(61). Thus, in late stages of the developmental cycle, the transcription of genes involved in this membrane recomposing and restructuring is critical for the transition between chlamydial forms.
Just under half of the 54 regulon contains genes encoding T3SS-associated proteins, including secreted effectors, inclusion membrane proteins, parts of the T3SS apparatus, and chaperone proteins. The T3SS is essential for chlamydial virulence and initially establishing an infection in a host cell (62,63). TARP (for translocated actin recruiting phosphoprotein) associates directly with actin in the host cell at the site of EB invasion and enhances EB uptake (44, 64, 65). CopB and CopD have been shown to make up the translocator of the T3SS apparatus and are thus essential for attachment of the chlamydial cell to the host membrane (37,66). CT229 has been shown to be an inclusion membrane protein that is critical for the subverting of the host cell defenses that would otherwise lead to destruction and therefore is necessary for the establishment of the infection from the initial point of entry (67,68). These are just a few examples of T3SS-associated genes that were found to be regulated by 54 . Because the EBs are less metabolically active than RBs, the many effector proteins needed for establishing the chlamydial inclusion and for commandeering the host cell machinery must be made before the conversion event. With a substantial number of the genes in the 54 regulon coding for T3SS-associated proteins, the data suggest that 54 gene regulation is involved in the preparation of the chlamydial cells for propagation of the infection (46). The 54 regulon is split almost equally between direct and indirect gene targets ( Table 1; see also Table 2). This supports the idea that other regulators participate in controlling the 54 regulon. Currently, only a few transcription factors have been identified for Chlamydia (69,70). More than half (64%) of the genes in the 54 regulon code for hypothetical proteins, several of which have no predicted role or identifiable motifs. Thus, there is a possibility that there is an unknown transcriptional regulator among the genes transcribed by 54 that may contribute to regulating the indirect targets in the regulon. It was interesting to observe that the well-characterized 28 gene target hctB was reproducibly and specifically upregulated upon induction of the intact CtcC ATPase domain ( Table 2; see also Table S1) (41,71). The level of upregulation observed here (ϳ3-fold) is minimal relative to the upregulation observed during temporal expression between 18 and 24 hpi (ϳ60-fold [37]). The absence of a 54 promoter element (Fig. S3) and lack of induction in E. coli ( Fig. 3; see also Fig. S5) support the idea that it is an indirect target of 54 . However, upregulation of this DNA condensation factor could have broader effects on gene regulation due to modified DNA topology and supercoiling, an aspect also well established in Chlamydia and bacterial gene regulation (72,73).
Tandem promoters for both 54 and 66 were found to be present upstream of ct394 (hrcA) and ct444 (omcA) (Fig. S6), potentially allowing for different temporal regulation patterns (38). While hcrA has been shown to have a 66 promoter in vitro and in the E. coli surrogate system (Fig. 3C) (74,75) and a CIRCE operator element, the presence of the 54 promoter directly inside the Ϫ10/Ϫ35 sequence might disrupt the ability of 66 to bind to the DNA and thus additionally inhibit transcription of the downstream gene when 54 -mediated transcription is inactive. These two genes are just a couple of examples of the potential for multiple layers of regulation that could be at play for any number of genes in the chlamydial genome.
The 54 regulon in Chlamydia has previously been investigated only in an in silico promoter mapping study (52). Mathews and Timms predicted only nine genes as being regulated by 54 based on the presence of the canonical 54 consensus sequence. In the current study, a less strict consensus sequence (TGG-N 9 -TGC) was used to look for potential 54 promoters. Two of the nine genes (ct620 and ct683) previously predicted by Mathews and Timms to be regulated by 54 appeared in the RNA-seq analysis as being differentially expressed when 54 was overactivated. The potential for additional regulatory factors affecting the transcription of the 54 -regulated gene may account for the discrepancy between those genes that were computationally predicted to have 54 promoter sequences and those that were found to be experimentally upregulated by 54 overactivation. For instance, CT398 (CdsZ) was previously shown to interact with 54 by Barta et al. (76) and is hypothesized to be a repressor of 54 activity, although its impact on transcription has yet to be evaluated (76).
While this study addressed the general role of 54 in Chlamydia, there are many issues remaining regarding 54 activation. The importance of the timing of 54 activation for the coordinated progression of the developmental cycle is demonstrated by the reduction in infectious progeny produced and the altered morphology of the Chlamydia cells with the overexpression of the active CtcC proteins. However, the importance of the synchronization and interdependence of all three of the chlamydial sigma factors has yet to be fully explored, but they are likely to be crucial for the transitions between RB and EB forms.
The lack of an effect on growth seen with the E242A ATPase-defective mutant supports the idea that ATPase activity and 54 gene expression are the likely causes of defects, rather than the general overexpression of the protein (Fig. 4). The limited change in gene expression seen with the E242A ATPase mutant (see Table S1 in the supplemental material) also supports the idea that this protein variant was not having an effect on 54 activation. The data thus suggest that the native CtcC protein does not interact with the E242A ATPase domain-only protein variant, as opposed to having a dominant-negative effect on the native protein activity. There is a chance that the amino acid substitution causes this version of CtcC to be misfolded although this has not been evaluated. Despite this possibility of misfolding, the data still suggest that overexpression of the protein is not solely the cause of the gene expression and phenotypic changes but rather that the changes are a consequence of the activity of the functional CtcC protein.
The signal that initiates the two-component regulatory cascade starting with the sensor kinase CtcB, stimulating CtcC, and subsequently activating 54 in Chlamydia has yet to be determined. In bacteria such as Salmonella Typhimurium, NtrC becomes phosphorylated by NtrB under nitrogen-limiting conditions (1). The homologous regulatory protein in Rhizobium, DctD, is phosphorylated by DctB in response to the availability of external dicarboxylates (29,30). DctB in Sinorhizobium meliloti and Vibrio cholerae binds to the C 4 -dicarboxylate succinate to signal for the expression of succinate transport proteins (29,30). In E. coli, the sensor kinase AtoS responds to acetoacetate, leading to the transcription of genes encoding enzymes involved in short-chain fatty acid metabolism (77). While homologs of the sensor kinase and response regulators in similar two-component regulatory systems have been identified in various bacteria, the particular environmental signal in many cases has yet to be identified. Supporting the idea of the incorporation of dicarboxylates and metabolic intermediates as a sensor molecule in Chlamydia, recent studies have shown that RsbU binds to tricarboxylic acid (TCA) intermediates to potentially control the main factor ( 66 ) (78). Determining the initial signal sensed by CtcB would provide critical information concerning what Chlamydia specifically responds to in controlling the developmental cycle and a more extensively defined role of 54 . Prior studies have shown that CtcB phosphorylates CtcC in vitro (47); however, direct dependence on phosphorylation for ATPase activity or transcriptional activation has not been assessed. Future studies demonstrating that CtcC phosphorylation enables ATPase activity and transcription are important to experimentally complete the signal Ctc transduction model in Chlamydia. The 54 gene targets and observations reported here provide a sound basis for these future studies.
A 2011 study published by Francke et al. looked for a common theme in 54 regulons across bacterial phyla (2). Their comparative analyses found that genes associated with flagellar and cellular membrane components in response to the external environment were conserved and highly represented in 54 regulons. The chlamydial 54 regulon appears to share the same theme. On the basis of the findings of the current study, the regulon of 54 in Chlamydia includes genes normally upregulated after 16 or 18 hpi during the developmental cycle and genes encoding mainly T3SS-associated and membrane proteins as well as genes which have transcripts enriched in EBs. These data support the idea of a role of 54 in the preparation of the chlamydial cells for RB-to-EB conversion and in the arming for subsequent host cell infection. E242A residue substitution was introduced via site-directed mutagenesis using a Q5 site-directed mutagenesis kit (New England Biolabs, Ipswich, MA). After the sequence was confirmed, the vector was transformed into BL21(DE3) E. coli competent cells, which were then grown at 37°C in Terrific Broth supplemented with 100 g/ml carbenicillin to an optical density at 600 nm (OD 600 ) of approximately 0.8. Protein expression was induced with the addition of IPTG (isopropyl-1-thio-␤-D-galactopyranoside) to reach a final concentration of 0.5 mM for 10 h at 17°C. Following E. coli collection by centrifugation (10,000 ϫ g, 20 min), cells were resuspended in washing buffer (10 mM HEPES [pH 7.2], 5 mM EDTA, 0.1% Triton X-100) with the addition of phenylmethane sulfonyl fluoride to reach a final concentration of 1 mM and 1,000 U of Benzonase endonuclease (EDM Millipore, Burlington, MA). Cells were lysed by sonication and centrifuged for 30 min at 14,000 ϫ g. The supernatant was decanted after this centrifugation, and the insoluble protein pellet was resuspended in wash buffer and centrifuged again. This wash step was repeated one additional time, for a total of three washes and centrifugations. After the final wash, the supernatant was again decanted, and the pellet was resuspended in denaturing buffer (6 M guanidine hydrochloride-phosphate-buffered saline [PBS; pH 8.0]) and rocked overnight at 4°C. The denatured cell mixture was centrifuged at 20,000 ϫ g for 30 min, and then the supernatant was applied to a gravity flow HisPur cobalt resin column (Thermo Fisher, Waltham, MA). After applying the protein to the column, it was washed with three column volumes of the denaturing buffer followed by three volumes of washing buffer (8 M urea, 100 mM NaH 2 PO 4 , 10 mM imidazole at pH 8.0). A total of three column volumes of elution buffer (8 M urea, 100 mM NaH 2 PO 4 , 250 mM imidazole at pH 3.0) was used to elute the protein off the column. Dialysis of the eluted protein into refolding buffer (250 mM NaCl, 50 mM NaH 2 PO 4 , 5 mM dithiothreitol [DTT], 5% glycerol, pH 10.0) was performed with gradually lower concentrations of urea at 6, 3, 2, 1.5, 1, 0.5, 0.25, and 0 M, changing buffers at a minimum of 3 h after each concentration step. Protein concentrations were confirmed using a Bradford assay.

MATERIALS AND METHODS
In vitro ATPase activity assay with purified recombinant CtcC protein variants. Quantification of the ATPase activity was performed using the protocol and reagents provided in an ADP-Glo kinase assay kit (Promega, Madison, WI). All proteins were initially suspended in the same buffer (50 mM NaH 2 PO 4 , 250 mM NaCl, 5 mM DTT, 5% glycerol at pH 10.00) and warmed to 37°C. Each were then added in a 45:45:10 volume ratio to warmed activity buffer ( through L386). The gene was inserted into the pL2-tetO overexpression plasmid using ligaseindependent cloning methods at the AgeI restriction enzyme site and transformed in DH5␣ competent E. coli cells (49). After sequence and protein expression in E. coli cells were confirmed, vector plasmid was isolated using a Qiagen plasmid miniprep kit. The E242A substitution was introduced via site-directed mutagenesis using a Q5 site-directed mutagenesis kit (New England BioLabs, Ipswich, MA).
Purified ctcC expression vector plasmid was transformed into a clonal isolate of L2 C. trachomatis (LGV2 434/Bu; GenBank accession no. CP019386.1 [79]). Briefly, 15 g of vector plasmid was mixed with 100 l 2ϫ CaCl 2 buffer (20 mM Tris [pH 7.5], 100 mM CaCl 2 ) and 25 l of C. trachomatis EBs and diluted with water to reach a final volume of 200 l. This CaCl 2 mixture was incubated at room temperature for 30 min before being diluted into 1ϫ CaCl 2 buffer, added onto a confluent monolayer of L292 cells, and then centrifuged for 30 min at 550 ϫ g and 20°C. Following centrifugation, the CaCl 2 mixture was aspirated off the monolayer and replaced with RPMI media supplemented with 5% tetracycline-free fetal bovine serum (FBS), 10 g/ml gentamicin, and 1 g/ml cycloheximide and incubated at 37°C in 5% CO 2 . After 16 h, 1 g/ml of ampicillin was added to the infected cells. The infection was allowed to incubate for 42 h after the initial infection, at which point the cells were lysed using water lysis and passaged to a new monolayer of cells, with media containing 1 g/ml of ampicillin. After an additional 48 h, the cells were passaged again with 2 g/ml of ampicillin. Two additional passages were performed 48 h after each previous passage, with new media containing 5 g/ml of ampicillin. At that point, the infection was propagated for harvesting. The presence of the vector plasmid was confirmed by PCR, antibiotic selection, and protein expression; in addition, the plasmid was sequenced to ensure the integrity of the construct.
Western blotting of CtcC overexpression in C. trachomatis. L929 host cell monolayers were infected with C. trachomatis transformed with the different tet-inducible ctcC construct variant plasmids and incubated in RPMI media supplemented with 5% tetracycline-free FBS, 10 g/ml gentamicin, and 1 g/ml cycloheximide at 37°C in 5% CO 2 . Infections were induced with 10 ng/ml of anhydrotetracycline (ATc) at 16 h postinfection. At 20 h postinfection, infections were harvested by applying a 1:1 ratio of PBS (Corning, NY) and 2ϫ Laemmli sample buffer with 2-mercaptoethanol (Bio-Rad, Hercules, CA) directly to the cell monolayer and cell scraping was done to collect the contents of the tissue culture plate well. The collected sample was used for Western blot analysis. Rabbit polyclonal primary antibody raised against full-length CtcC (Proteintech, Chicago, IL) and goat anti-major outer membrane protein (anti-MOMP) primary antibody (ViroStat, Westbrook, ME) were used to probe for the protein of interest and a loading control protein, respectively. IRDye donkey anti-rabbit 680LT and donkey anti-goat secondary 800CW antibody (Li-Cor, Lincoln, NE) were used to visualize the Western blot. Western blot analysis was performed upon plasmid transformation into Chlamydia as well as alongside RNA isolation and growth analyses to ensure that the observed phenotypes corresponded to the induced protein expression. Figure 2B shows a representative Western blot of the overexpression of CtcC protein variants.
Progeny assay with induction of CtcC protein variant overexpression. L929 host cell monolayers in 6-well tissue culture plates were infected with C. trachomatis transformed with different tet-inducible ctcC variant plasmids or empty pL2 vector control plasmid using RPMI supplemented with 5% tetracycline-free FBS, 10 g/ml gentamicin, and 1 g/ml cycloheximide. At 12 h postinfection, 10 ng/ml of ATc was added to induce overexpression of CtcC protein variants. At 30 h postinfection, the infected monolayer was lysed using a water lysis technique and passaged onto a new monolayer of host cells in a 96-well plate format to determine infectious titer. At 24 h after the passage, the 96-well plate was fixed and stained using the MicroTrack C. trachomatis culture confirmation test (Syva Co., Palo Alto, CA). Wells were then quantified for the number of inclusions, and the data were compared to the number of inclusion-forming units in the initial infection. This procedure was repeated in triplicate for each condition.
Immunofluorescence microscopy of C. trachomatis-infected cells following expression induction of CtcC variants. L929 cells were grown to ϳ75% confluence in an 8-well ibiTreat -Slide (Ibidi, Martinsried, Germany) and were infected with respective C. trachomatis L2 mutants. At 12 hpi, 10 ng/ml ATc was added to the respective wells, and at 24 hpi, the wells were fixed with methanol for 10 min and then washed with Hanks' balanced salt solution (HBSS). Cells were washed with HBSS and then stained with MicroTrack C. trachomatis culture confirmation test (Syva Co., Palo Alto, CA) and 10 nM 4=,6diamidino-2-phenylindole (DAPI)-PBS. A final overlay of Vectashield antifade mounting medium (Burlingame, CA) was added, and the slides were immediately imaged. Cells were visualized on an Olympus IX81/3I spinning disk confocal inverted microscope at ϫ150 magnification and captured on an Andor Zyla 4.2 scientific complementary metal oxide semiconductor (sCMOS) camera (Belfast, Northern Ireland). Microscope and camera were operated using SlideBook 6 software (Intelligent Imaging Innovations, Denver, USA). Images are representative of at least 5 different fields/inclusions per condition. Exposure time remained consistent for all fields captured, with exposure for DAPI (DNA) at 1 s, FITC (OmpA) for 3 s, and Cy5 (cytoplasm) for 3 s. Three Z-stack images at 0.3 m apart were taken per field imaged. Images were processed in SlideBook 6, and No Neighbors Deconvolution with a subtraction constant of 0.4 was applied to all images. Images represent an average projection over the Z-axis of the three Z-stacks for each field shown.
RNA purification and RNA-seq. L929 cells in a 6-well tissue culture plate with tetracycline-free RPMI media were infected with the pL2-tetO ctcC ATPase-only construct or the pL2-tetO vector control. The infection was induced with a final ATc concentration of 10 ng/ml at 16 hpi. A relatively small window of induction was chosen to reduce the pleiotropic effect of overactivation of 54 . At 20 hpi, the results of the infection was harvested as described in the previous section using TRIzol and subsequent chloroform RNA extraction. Per the procedure for RNA-sequencing analysis of intracellular bacteria described previously by Marsh, Humphrys, and Myers (80), following Turbo DNase treatment (Thermo Fisher, Waltham, MA) and purification performed using an RNeasy kit (Qiagen, Hilden, Germany), both host rRNA and chlamydial rRNA were depleted using Ribo-Zero magnetic core kits (Illumina, San Diego, CA) specific for human/mouse/rat rRNA and Gram-negative bacterial rRNA. A final purification step was performed using the RNeasy minikit (Qiagen, Hilden, Germany). rRNA depletion was assessed by running samples on a 2% agarose gel. RNA quality was assessed using Qubit quantification and TapeStation gel analysis. The RNA library was made using a NEBNext Ultra II directional RNA library prep kit for Illumina (New England BioLabs, Ipswich, MA) and was sequenced with NextSeq 550 high-output single-read 50-bp sequencing (NX-HO-SR50). Geneious Prime (Version 2019.1.1) was used for the data analysis. Reads were aligned to Chlamydia trachomatis 434/Bu genome (NC010287). Transcript levels were normalized to reads per million. Genes with an average differential expression ratio 2 standard deviations away from the mean differential expression ratio for the entire genome were considered to be significantly upregulated and were included as part of the 54 regulon and listed in Table 1 or Table 2, depending on the presence or absence of a predicted 54 promoter region upstream of the open reading frame. The protein roles listed in Table 1 and Table 2 were verified with data from previous studies (37, 39-41, 44, 54-57, 64, 67, 68, 71, 74, 81-104).
Droplet digital PCR quantification of RNA transcripts with CtcC protein variant overexpression. L292 cells in a 6-well tissue culture plate containing tetracycline-free media were infected with the different ctcC construct plasmids or with empty pL2 control vector, and transformed C. trachomatis cells were induced with 10 ng/ml ATc at 16 h postinfection and then harvested with 1 ml per well TRIzol. A 200-l volume of chloroform was added to each 1-ml TRIzol cell lysate, subjected to vortex mixing, and centrifuged for 8 min at 10,000 ϫ g. The aqueous layer of the TRIzol/chloroform mixture was collected in a new tube, and a 1:1 ratio of isopropanol was added before incubation was performed at -20°C for 30 min, followed by incubation at room temperature for 15 min. The precipitated nucleic acids were pelleted by centrifugation for 8 min, the supernatant was decanted, and the pellet was washed with 1 ml 70% ethanol. After another 5-min centrifugation, the pellet was air-dried and resuspended in 80 l of RNase-free molecular-biology-grade water (Corning, NY). DNA depletion was performed by addition of 10 l of Turbo DNase buffer and 10 l of Turbo DNase (Thermo Fisher, Waltham, MA) to each 80 l of sample and incubation in a 37°C water bath for 30 min. Immediately following DNA depletion, the samples were processed through the RNeasy minikit (Qiagen, Hilden, Germany) and eluted into 50 l of RNase-free water.
cDNA was generated using a High-Capacity cDNA reverse transcription kit (Applied Biosystems, Foster City, CA) that included control reaction mixtures excluding reverse transcriptase, which was then used to assess the presence of residual gDNA contamination. Following cDNA generation, droplet digital PCR (ddPCR) was used to determine an absolute quantification for the number of transcripts present for the genes of interest. A QX200 digital droplet PCR system (Bio-Rad, Hercules, CA) with all the suggested Bio-Rad consumables, including QX200 EvaGreen Supermix, was utilized in the setup and reading of the ddPCR run. Primers for specific genes of interest used for ddPCR are listed in Table S4 in the supplemental material. Data were analyzed using QuataSoft Analysis Pro software, version 1.0 (105). The data were normalized to the quantification of secY transcript copies and were converted to log 2 scale for comparisons to the number of transcripts in the empty pL2 vector control induced with ATc (33,34).
Miller assays. ␤-Galactosidase activity was used to assess the ability of 54 to activate transcription of a downstream gene (lacZ) from promoter regions taken from those genes found to be upregulated in the RNA-seq analysis. Promoter regions were amplified using primers listed in Table S3 and cloned into the pAC-lacZ vector at the BamHI restriction site upstream of the lacZ gene (37). The chlamydial genes encoding 54 and the CtcC ATPase domain (residues 142 to 386), with and without the E242A amino acid substitution, were cloned into pRSF-Duet vector at the NcoI and NdeI restriction sites, respectively, putting the transcription of both these genes under the control of IPTG-inducible regulation. All insertions were confirmed by sequencing. E. coli cells were cotransformed with the pAC-lacZ plasmid containing the various promoter regions in combination with the different pRSF-Duet plasmids. Levels of IPTG-induced expression of 54 and the CtcC ATPase domain were assessed via Western blotting.
Overnight cultures were grown in LB broth supplemented with 50 g/ml of chloramphenicol and kanamycin (only chloramphenicol was used when just the pAC-lacZ vector was present). Cultures were then diluted 1:100 in fresh broth and grown at 37°C until the culture reached mid-log phase. Cultures were induced with 1 mM IPTG for 1 h before being centrifuged. Pellets were washed and resuspended in Z-buffer (0.06 M Na 2 HPO 4 , 0.04 M NaH 2 PO 4 , 0.01 M KCl, 0.001 M MgSO 4 , 0.05 M ␤-mercaptoethanol, pH 7.0). Permeabilization was performed with 10% volume chloroform and 5% volume 0.1% sodium dodecyl sulfate for 10 min at room temperature. A 100-l volume of the aqueous layer of the mixture was then aliquoted in triplicate into a microtiter plate, after which 20 l of ortho-nitrophenyl-␤-galactoside (4 mg/ml) was added. A 50-l volume of Na 2 CO 3 was added to each well after 10 min to stop the reactions. OD 420 and OD 550 were then measured for each reaction mixture, and the OD 600 of samples was taken prior to cell permeabilization, in order to calculate Miller units (37,106).

SUPPLEMENTAL MATERIAL
Supplemental material is available online only.   We also appreciate Srishti Baid and Zoe Dimond (University of Kansas) for careful review of and comments on the manuscript.