An early event in the herpes simplex virus type-2 replication cycle is sufficient to induce Chlamydia trachomatis persistence


*E-mail; Tel. (+1) 423 439 6295; Fax (+1) 423 439 8044.


Epidemiological studies have demonstrated that co-infections of herpes simplex virus type 2 (HSV-2) and Chlamydia trachomatis occur in vivo. Data from a tissue culture model of C. trachomatis/HSV-2 co-infection indicate that viral co-infection stimulates the formation of persistent chlamydiae. Transmission electron microscopic (TEM) analyses demonstrated that in both HeLa and HEC-1B cells, co-infection caused developing chlamydiae to exhibit swollen, aberrantly shaped reticulate bodies (RBs), characteristically observed in persistence. Additionally, HSV-2 co-infection suppressed production of infectious chlamydial elementary bodies (EBs) in both host cell types. Co-infection with HSV type 1 (HSV-1) produced similar morphologic alterations and abrogated infectious EB production. These data indicate that virus-induced chlamydial persistence was neither host cell- nor virus strain-specific. Purification of crude HSV-2 stocks demonstrated that viral particles were required for coinfection-induced chlamydial persistence to occur. Finally, co-infection with either UV-inactivated, replication-incompetent virus or replication-competent HSV-2 in the presence of cyclohexamide reduced chlamydial infectivity without altering chlamydial genomic DNA accumulation. These data demonstrate that productive viral replication is not required for the induction of chlamydial persistence and suggest that HSV attachment and entry can provide the necessary stimulus to alter C. trachomatis development.


Treatment of sexually transmitted diseases (STDs) and their sequelae cost the USA billions of dollars each year. Approximately 19 million people in the USA were infected with an STD in 2000. Nearly half of those individuals were between the ages of 15 and 24 years (Weinstock et al., 2004). Given the fact that several STD agents can establish chronic or latent infections, clinical management of STDs remains a challenge of epidemic proportions. Two of the most common STD agents are herpes simplex virus type 2 (HSV-2), causing 200 000–500 000 new infections annually and Chlamydia trachomatis serovars D–K, causing 4 million new infections annually (Butler, 1997; Weinstock et al., 2004).

Herpes simplex virus type 1 (HSV-1) and HSV-2 are enveloped DNA viruses of the viral family, Herpesviridae. Upon infection, herpes simplex viruses bind to the host cell receptor, heparin sulfate. The initial binding event is followed by interaction of viral envelope glycoproteins with one of four known co-receptors that facilitate viral entry (Spear, 2004). Once inside the host cell, viral tegument proteins are released into the host cell and the capsid is transported to the nucleus, where viral genome replication occurs. The viral genome is transcribed and viral proteins are synthesized using host cell machinery. New virions are then assembled and exit the host cell (Roizman and Knipe, 2001).

Herpes simplex virus type-2 is the primary cause of genital herpes infection. HSV-2 infection usually occurs on the mucous membranes and skin surrounding the genitals, causing a characteristic lesion. After primary infection, HSV-2 can establish a latent, lifelong infection in the neurones of the sacral ganglia. On average, latent HSV-2 infections are reactivated five times each year. Upon reactivation, the virus travels through the neurones to the site of initial infection where it can be shed from skin lesions (Corey et al., 1983). Although most genital HSV infections are clinically mild, HSV-2 can also be responsible for serious diseases, such as keratitis and meningitis (Roizman and Knipe, 2001).

Chlamydia trachomatis is a Gram-negative, obligate intracellular bacterium that has a biphasic developmental cycle. During the infection phase of the cycle, chlamydiae exist as small, infectious, metabolically inert, elementary bodies (EBs). The EBs attach and enter a host cell, where they differentiate within an inclusion into the non-infectious, vegetative form of the bacterium, called the reticulate body (RB). Once the RBs have completed several rounds of replication, they condense to form EBs. The EBs are then released and can infect new host cells (Wyrick, 2000). Genital chlamydial infections are often chronic and asymptomatic, leading to severe complications including pelvic inflammatory disease, ectopic pregnancy and infertility (Darville, 2000).

When developing chlamydiae are exposed to certain environmental factors, they deviate from the normal developmental cycle into a state called persistence (Hogan et al., 2004). While in a persistent state, chlamydiae form large, abnormally shaped RBs. Although they remain viable, persistent chlamydiae are no longer infectious. Some of the most well studied inducers of chlamydial persistence include interferon-γ (IFN-γ) exposure, amino acid and iron deficiencies and exposure to penicillin G (Beatty et al., 1994a; Raulston, 1997; Darville, 2000; Gerard et al., 2001).

Several studies have shown that C. trachomatis and HSV-2 co-infections occur in vivo. Both pathogens have been simultaneously isolated from women afflicted with endometritis, salpingitis and cystitis (Paavonen et al., 1985; Tait et al., 1985). Serologic investigations indicate that HSV-2-positive individuals are likely to be C. trachomatis-positive as well (Paroli et al., 1990; Silins et al., 2002). Several studies have documented seropositivity rates of > 50% for both pathogens, implying that some of the individuals had been exposed to both HSV-2 and C. trachomatis (Duncan et al., 1992; Wagner et al., 1994). Although IgG seropositivity does not prove coexisting infection, there is a high probability that some of the individuals were infected simultaneously with both pathogens.

In vitro models of HSV-2/C. trachomatis co-infections have also been established. Transmission electron microscopy (TEM) analyses of Vero cells co-infected with C. trachomatis serovar L2 and HSV-2 revealed swollen inclusions with few RBs or EBs (Pontefract et al., 1989). Chiarini et al. reported that the number of cells positive for chlamydiae by immunofluorescence was reduced when HeLa cells were co-infected with HSV-2/C. trachomatis serovar D (Chiarini et al., 1996). The number of chlamydial inclusions and infectious EBs were also reduced in HSV-2 pre-infected HT1376 human bladder cells (Superti et al., 2001). Although these observations indicate that HSV-2 co-infection alters chlamydial development, these studies did not examine the co-infection process in detail.

Previously, we developed a tissue culture model of C. trachomatis and HSV-2 co-infection. Data from these studies indicate that during C. trachomatis serovar E and HSV-2 co-infection, chlamydiae become persistent. Such data include abnormal ultrastructure and decreased production of infectious EBs. Although less infectious, chlamydiae within HSV-2 co-infected cells remain viable, as demonstrated by continued accumulation of precursor 16S rRNA and chlamydial genomes (Deka et al., 2006). However, the mechanism by which HSV-2 induces chlamydial persistence is currently unknown. As a first step in defining how viral-induced persistence occurs, we have determined whether productive viral replication is necessary for induction of chlamydial persistence. The present study will allow us to begin to unravel the mechanism by which HSV-2 causes C. trachomatis to enter persistence.


HSV-2 induction of chlamydial persistence is not host cell type-specific

Previous studies have shown that host cell interaction with chlamydiae can be a determining factor in the induction of chlamydial persistence. For example, IFN-γ induction of chlamydial persistence has been shown to occur in several host cell types, including HeLa cells (Hogan et al., 2003). However, developing chlamydiae within HEC-1B cells are resistant to IFN-γ because these host cells do not produce indoleamine 2,3-dioxgenase (IDO) in response to IFN-γ (Kane and Byrne, 1998; Wyrick and Knight, 2004). To be certain that induction of chlamydial persistence during HSV-2 co-infection is not a specific property of HeLa cells, HEC-1B cells were also utilized as host cells for co-infection. Duplicate monolayers of HeLa and HEC-1B cells were either mock, singly, or co-infected. Cells were collected 20 h post HSV-2 infection and processed for either TEM or chlamydial EB titration. Electron micrographs demonstrated that chlamydiae within both HeLa and HEC-1B co-infected cells exhibited abnormal RB morphology characteristic of persistence (Fig. 1B and E). EBs were absent from chlamydial inclusions within HSV-2 co-infected HeLa and HEC-1B cells (Fig. 1B and E). In contrast, in both C. trachomatis singly infected cell types, RBs appeared normal and EBs were present (Fig. 1A and D). Furthermore, chlamydial EB titre assays indicated a significant decrease in the production of infectious EBs during HSV-2 co-infection in both host cell types (Fig. 2A and C) when compared with chlamydia singly infected controls. Interestingly, singly infected HEC-1B cells consistently produced higher titres of chlamydiae than did HeLa cells. These data indicate that HSV-2 coinfection-induced persistence is not limited to a specific host cell type.

Figure 1.

HSV co-infection induces persistent chlamydial morphology. Cultures of HeLa (A, B, C and F) or HEC-1B (D and E) cells were C. trachomatis-infected or co-infected with C. trachomatis and HSV-2 strain 333 (A, B, D–F) or HSV-1 strain KOS (C) at 10 moi (A–E) or 1 moi (F). Cells were harvested for TEM analyses. Black arrows on electron micrographs indicate EBs and white arrows indicate RBs.

Figure 2.

HSV co-infection decreases chlamydial infectivity. Cultures of HeLa or HEC-1B cells were mock, C. trachomatis, or HSV singly infected and C. trachomatis/HSV co-infected. Cells were harvested for EB titration assays. A. HeLa cells were co-infected with HSV-2 strain 333 at 10 moi. B. HeLa cells were co-infected with HSV-1 strain KOS at 10 moi. C. HEC-1B cells were co-infected with HSV-2 strain 333 at 10 moi. D. HeLa cells were co-infected with HSV-2 strain 333 at, 0.1, 1, or 10 moi.
EB titres are expressed as ifu ml−1 sample ± SEM; n = 3. Asterisks (*) indicate titres that are significantly different (by T-test) compared with those from C. trachomatis singly infected cells (P < 0.05). The data shown are representative of three independent experiments.

HSV induction of chlamydial persistence is not virus-specific

Although HSV-2 is the primary agent of genital herpes infection, HSV-1 infections also account for many genital infections in vivo (Whitley, 2001). Therefore, the effect of HSV-1 co-infection on C. trachomatis development was examined. HeLa cell monolayers were co-infected with C. trachomatis serovar E and HSV-1 strain KOS as described. C. trachomatis singly infected cells contained abundant EBs and RBs of normal morphology (Fig. 1A). However, in HSV-1 co-infected cells, RBs were enlarged, misshapen and few or no EBs were observed (Fig. 1C). Chlamydial infectivity was also significantly reduced by HSV-1 co-infection compared with that in C. trachomatis singly infected controls (Fig. 2B). These data were indistinguishable from the results observed with HSV-2 in HeLa cells (Figs 1B and 2A), thus indicating that both HSV-2 and HSV-1 co-infection can interfere with normal chlamydial development.

HSV-2 induction of chlamydial persistence is viral dose-dependent

In order to ensure that every chlamydiae-infected cell was also infected with HSV-2, co-infections were originally conducted using a viral multiplicity of infection (moi) of 10 pfu/cell. To examine whether chlamydial persistence could be induced with lower concentrations of virus, co-infection experiments were performed using various moi of HSV-2. Triplicate HeLa cell monolayers were either mock, singly or co-infected with C. trachomatis and HSV-2 at 10, 1, or 0.1 moi. The cells were collected at 20 h post HSV-2 infection and processed for TEM and chlamydial titration. Inclusions in C. trachomatis singly infected cells contained normal developing EBs and RBs similar to those shown in Fig. 1A. In cells co-infected with C. trachomatis and HSV-2 at 1 moi, chlamydiae exhibited morphological characteristics of persistence (Fig. 1F). Subpassage experiments revealed that HSV-2 co-infection at 10, 1 and 0.1 moi significantly suppressed production of infectious EBs compared with that in C. trachomatis singly infected cells (Fig. 2D). Furthermore, the degree to which EB production was reduced correlated directly to increased viral moi.

Purification of HSV-2 does not diminish its capacity to induce chlamydial persistence during co-infection

It is well known that cellular host factors, especially the cytokines IFN-γ, IFN-α, TNF-α, can induce or enhance chlamydial persistence (Hogan et al., 2004). As the viral inocula used in this study were crude stocks, it is possible that these cytokines might be present in the HSV-2 stocks and could be responsible for the observed alterations in chlamydial development during co-infection. To test this possibility, crude lysates from either HSV-2 or mock-infected cells were subjected to centrifugal purification; the resultant purified components were used in co-infection experiments with C. trachomatis. Four separate viral stocks were prepared: HSV-2 crude, mock crude, purified HSV-2 and purified mock. Luminex assays for IFN-γ, IFN-α, TNF-α and IL-6 were performed on both crude and purified stocks to determine: (i) whether these cytokines were present and (ii) whether these cytokines were removed by this purification method. Although IL-6 has not been associated with chlamydial persistence, HSV infection is known to induce this cytokine, thus it serves as a positive control (Kanangat et al., 1996). Of the four cytokines examined, only IL-6 was detected in viral stocks. The HSV-2 crude and mock crude stocks contained IL-6 at 1121 pg ml−1 and 2835 pg ml−1 respectively. IL-6 concentrations were reduced at least 43-fold by centrifugal purification (purified HSV-2, 26 pg ml−1 and purified mock, 38 pg ml−1). These data demonstrate that the purification method utilized efficiently removed cytokines from the stocks and that IFN-γ, IFN-α, TNF-α, cytokines known to alter chlamydial development, were not present in any of the viral inocula used in this study.

Co-infection with crude and purified viral stocks was performed as described. Data from chlamydial titre assays indicated that only HSV-2 crude and purified HSV-2 stocks significantly decreased infectious EB yield during co-infection (Fig. 3). Neither the mock crude nor purified mock stock controls reduced chlamydial infectivity in ‘co-infected’ cells compared with C. trachomatis singly infected cells (Fig. 3). These data indicate that coinfection-induced chlamydial persistence is not mediated by IFN-γ, IFN-α, or TNF-α present in viral stocks and suggest that it is dependent on the presence of viral particles.

Figure 3.

Purified HSV-2 co-infection decreases chlamydial infectivity. Cultures of HeLa cells were infected with C. trachomatis, or co-infected with chlamydia and HSV-2 crude stock, purified stock, mock crude stock, or mock purified stock at 10 moi or the equivalent volume of mock stock. Cells were harvested for EB titration analyses. EB titres are expressed as ifu ml−1 sample ± SEM; n = 3. Asterisks (*) indicate titres that are significantly different (by T-test) compared with those from C. trachomatis singly infected cells (P < 0.05). The data shown are representative of three independent experiments.

Cyclohexamide exposure does not abrogate HSV-induced C. trachomatis persistence

Productive replication of HSV-2 within a host cell is dependent on host cell enzymes and machinery (Roizman and Knipe, 2001). Cyclohexamide is a well known inhibitor of eukaryotic protein synthesis. By preventing host cell protein synthesis, cyclohexamide inhibits synthesis of viral proteins and prevents productive replication of HSV-2 (Swanstrom et al., 1975; Vasquez, 1979). In contrast, cyclohexamide actually enhances C. trachomatis development (Ripa and Mårdh, 1977). Therefore, cyclohexamide exposure was utilized to investigate whether productive viral replication is required to induce chlamydial persistence during co-infection with HSV-2. Monolayers of HeLa cells were either mock or chlamydiae-infected, exposed to cyclohexamide and HSV-2-infected as described in Experimental procedures. Control experiments indicated that productive viral replication and viral protein synthesis were inhibited by > 95% under these exposure conditions, as ascertained by plaque assays and S35-methionine incorporation (data not shown). Co-infected cells exposed to either diluent (ddH2O) or cyclohexamide contained chlamydiae of abnormal morphology (Fig. 4B and E) in electron micrographs. Chlamydial inclusions in singly infected cells contained normal RBs and developing EBs regardless of cyclohexamide exposure (Fig. 4A and D). As previously observed, the quantity of infectious EBs recovered from co-infected cells was significantly lower than that recovered from singly infected cells (Fig. 4C). Cyclohexamide exposure during co-infection did not reverse this result (Fig. 4F).

Figure 4.

Co-infection with HSV-2 in the presence of cyclohexamide induces chlamydial persistence. Cultures of HeLa cells were infected with C. trachomatis. At 23 h post chlamydial infection the cells were refed with medium containing either diluent (ddH2O) (A–C) or 1 μg ml−1 cyclohexamide (D–F) for 1 h. The cells were then infected with HSV-2 inoculum containing cyclohexamide or ddH2O. The cells remained under cyclohexamide exposure until they were harvested for TEM and EB titration analyses. EBs on electron micrographs are indicated by black arrows; RBs by white arrows. EB titres are expressed as ifu ml−1 sample ± SEM; n = 3. Asterisks (*) indicate titres that are significantly different (by T-test) compared with those from C. trachomatis singly infected cells (P < 0.05). The data shown are representative of three independent experiments.

Co-infection with replication-incompetent HSV-2 induces chlamydial persistence

Although cyclohexamide greatly reduces productive viral replication, it does not completely block replication (Swanstrom et al., 1975; Vasquez, 1979; Sanfilippo et al., 2004). Previous studies have demonstrated that UV-inactivation of HSV-2 renders the virus completely replication-incompetent (Moxley et al., 2002). Replication-incompetent HSV-2 (HSV-2UV) was generated using ultraviolet irradiation. An UV dose of 2.5 J cm−2 sufficiently inactivated the virus so that no infectious HSV-2 virions were detectable in plaque assays (data not shown). However, UV irradiation did not affect polymerase chain reaction (PCR) amplification of HSV-2 DNA (data not shown).

Duplicate HeLa cell monolayers were either mock, singly or co-infected with C. trachomatis and HSV-2 or HSV-2UV. Infections with HSV-2UV were performed using a volume of stock equivalent to 10 moi of replication-competent HSV-2. Infected cell monolayers were collected and assayed for chlamydial titre, TEM and PCR analyses at 0 and 20 h post HSV infection. Supernatants were also collected from each sample for plaque assay analyses of viral replication. Infectious viral particles were only recovered from cultures infected with replication-competent HSV-2 (data not shown). Electron micrographs from chlamydia singly infected cells depicted inclusions containing RBs of normal morphology and several developing EBs (Fig. 5A). Alternatively, chlamydiae in HSV-2UV co-infected cells exhibited the same swollen, abnormally shaped RBs that were observed in cells co-infected with replication-competent virus (Fig. 5B and C). Additionally, co-infection with HSV-2UV decreased production of infectious EBs similarly to that observed with replication-competent HSV-2 (Fig. 5D).

Figure 5.

Co-infection with HSV-2UV induces chlamydial persistence. Cultures of HeLa cells were C. trachomatis-infected followed by infection with replication-competent HSV-2 at 10 moi or an equal amount of HSV-2UV. Cells were harvested for TEM (A–C), C. trachomatis titre analyses (D) and total DNA isolation (E) at 0 or 20 h post HSV infection. HSV-2 viral stocks were UV-inactivated by exposure to 2.5 J cm−2 UV irradiation at 4°C; plaque assays demonstrate that HSV-2UV stocks contain no replication-competent virions (data not shown). EBs on electron micrographs are indicated by black arrows; RBs by white arrows. EB titres are expressed as ifu ml−1 sample ± SEM; n = 3. Asterisks (*) indicate titres that are significantly different (by T-test) compared with those from C. trachomatis singly infected cells (P < 0.05). The data shown are representative of three independent experiments.

Total cellular DNA was isolated immediately following (T0) and at 20 h (T20) post HSV-2 infection from singly and co-infected cells from triplicate experiments and the accumulation of human, chlamydial and HSV-2 DNA was determined by PCR. In each experiment, a six log dilution series of known DNA template controls was used to generate amplification standard curves. Experimental samples were only quantified if they fell within the linear range of the PCR. Template-negative samples were also included in every experiment; amplification in template-negative samples was never observed. Accumulation of host cell DNA, as ascertained by human GAPDH amplification, was similar in all samples (data not shown). Chlamydial 16S rRNA was only amplified in those samples infected with C. trachomatis. Likewise, HSV-2 gG DNA was amplified only in virus-infected samples (data not shown). Both chlamydial and HSV-2 genomic DNA were normalized to host cell DNA after quantification. The amount of HSV-2 DNA increased from T0 to T20 in all the samples infected with replication-competent HSV-2 (Fig. 5E). However, the amount of HSV-2 DNA did not increase from T0 to T20 in those samples infected with HSV-2UV, indicating that there was no viral DNA replication during the infection (Fig. 5E). It is important to note that the quantity of cell-associated HSV DNA at T0 is similar in HSV-2- and HSV-2UV -infected cultures (Fig. 5E), demonstrating that UV exposure does not inhibit virus/host cell attachment. Therefore, HSV-2UV viral particles were capable of attaching to the host cells but were unable to productively replicate. These data indicate that chlamydiae become persistent in the presence of replication-incompetent HSV-2.

Co-infection in the presence of cyclohexamide or with HSV-2UV does not alter accumulation of chlamydial DNA

Chlamydiae in a persistent state remain viable and continue to replicate their DNA although they do not divide (Gerard et al., 2001). Therefore, the amount of chlamydial DNA should not differ between persistent and normal chlamydial infections. To ensure that the chlamydiae remained viable and that replication of chlamydial DNA was not halted by co-infection, HeLa cells were co-infected with HSV-2/C. trachomatis and the cells were harvested at both T0 and T20 for PCR analyses. The quantity of chlamydial DNA amplified at T20 is significantly higher than that at T0, demonstrating that continued accumulation of chlamydial DNA occurred during HSV-2 co-infection (Fig. 6A). These data show both that the analysis method is able to detect an increase in chlamydial DNA and that the chlamydiae remain viable.

Figure 6.

Co-infection does not alter accumulation of chlamydial DNA. Total DNA was purified and subjected to PCR analyses using host genome (GAPDH) and chlamydial genome (16S rRNA gene) specific DNA oligonucleotides. Amplimers were electrophoresed and quantified as described. The data in A were quantified after conversion from Macintosh to personal computer format whereas the data in B were quantified using the Macintosh format, thus causing differences in the intensity scales. Chlamydial DNA quantities were normalized to host DNA as ascertained by PCR with human GAPDH-specific primers (data not shown). These data are representative of three independent experiments and are averages of three biological replicates. The average integrated intensity for each amplimer was plotted ± SEM.
A. HeLa cells were co-infected with HSV-2/C. trachomatis. Total cellular DNA was collected at 0 (T0) and 20 h (T20) post HSV-2 infection.
B. HeLa cells were C. trachomatis-infected, or co-infected with HSV-2, HSV-2UV, or HSV-2 in the presence of cyclohexamide. Total cellular DNA was collected at 20 h post HSV infection.
The asterisk (*) indicates that the value obtained at 20 h post HSV co-infection was significantly different (by t-test) from that at 0 h post HSV-2 co-infection (P < 0.05).

Chlamydial DNA accumulation at T20 was also measured in co-infected cultures performed in the presence of cyclohexamide or with HSV-2UV. PCR analyses demonstrated that there were similar amounts of chlamydial DNA in the singly and co-infected cells despite cyclohexamide exposure or HSV-2UV infection (Fig. 6B). HSV-2 genome accumulation was also similar in HSV-2 singly and co-infected cells as previously observed (Deka et al., 2006). As expected, levels of HSV-2 DNA in co-infected cells exposed to cyclohexamide or infected with HSV-2UV were significantly decreased compared with those in diluent-exposed/HSV-2 co-infected cells (data not shown). These results demonstrate that the chlamydiae within co-infected cells continue to accumulate genomic DNA in the presence of cyclohexamide or replication-incompetent virus and, thus, are viable.


Chlamydiae that have deviated from the normal developmental cycle and exist in a viable, yet non-cultivable state are described as being persistent. Studies of chlamydial persistence have revealed several mechanisms by which developing chlamydiae are stimulated to enter this state. The most thoroughly investigated models of persistence include amino acid and iron deprivation, INF-γ exposure and β-lactam antibiotic exposure (Beatty et al., 1994a; Raulston, 1997; Darville, 2000; Gerard et al., 2001). In general, it seems that when chlamydiae are presented with an environmental factor that is unfavourable for their growth, they respond by entering persistence. Previous data from our laboratory indicate that viral co-infection also stimulates the formation of persistent chlamydiae, as evidenced by abnormal RB morphology, and decreased chlamydial titre during co-infection (Deka et al., 2006). In this study, we have further characterized HSV-2 coinfection-induced persistence. In particular, we have demonstrated that this phenomenon is not dependent on a specific host cell type or HSV strain. Additionally, our data indicate that productive viral replication is not required for the induction of chlamydial persistence.

HSV-2/C. trachomatis co-infection has the potential to elicit several unfavourable environmental conditions for chlamydial growth, including amino acid deprivation and induction of IFN-γ. One of the most obvious potential mechanisms for coinfection-induced persistence is that the virus simply exhausts cellular amino acids to synthesize viral proteins, thereby, depriving the chlamydiae of nutrients. UV-inactivation of HSV-2 prevents viral DNA replication and viral protein synthesis (Moxley et al., 2002). Consequently, HSV-2UV would not be expected to deprive chlamydiae of amino acids by using them to manufacture viral proteins. Cyclohexamide also strongly inhibits HSV protein synthesis and enhances C. trachomatis development by inhibiting host cell protein synthesis (Ripa and Mårdh, 1977). However, cyclohexamide exposure during co-infection also did not prevent chlamydial persistence. Therefore, it does not appear that HSV-2 coinfection-induced chlamydial persistence is based upon simple amino acid deprivation.

A second possible mechanism for coinfection-induced persistence is that HSV-2 infection stimulates IFN-γ synthesis/release, which, in turn, upregulates indoleamine 2,3-dioxygenase and depletes cellular tryptophan. Without this source of tryptophan, chlamydiae would be expected to enter a persistent state (Beatty et al., 1994b; Hogan et al., 2004). However, chlamydiae exhibit characteristics of persistence within co-infected HEC-1B cells. Previous studies have demonstrated that chlamydiae do not respond to IFN-γ exposure in HEC-1B cells (Kane and Byrne, 1998; Wyrick and Knight, 2004). Therefore, it is unlikely that IFN-γ-induced IDO activity is responsible for coinfection-induced persistence.

In our studies, the reduction in C. trachomatis infectivity due to HSV co-infection was dose-dependent. Despite the fact that HSV infection at 1 or 0.1 moi does not productively infect all chlamydiae-infected host cells, chlamydial infectivity in these cultures was significantly reduced. These experiments might suggest the involvement of a factor secreted from HSV-infected cells. However, HSV produces 50–200 defective viral particles for every replication-competent virion (Roizman and Knipe, 2001). Given the results obtained with UV-inactivated virions and cyclohexamide-exposed co-infected cells, a more likely explanation is that defective virions in the HSV-2 inocula can alter chlamydial development in a manner similar to replication-competent virions.

Several studies have demonstrated that while UV-irradiation of HSV prevents productive replication, it does not inhibit virion-associated functions such as attachment and entry (Moxley et al., 2002; Sanfilippo and Blaho, 2006). Plaque assays and PCR from co-infection experiments confirm these observations. Interestingly, co-infection with UV-inactivated, replication-incompetent HSV-2 alters chlamydial development similarly to that observed during co-infections with replication-competent virus. Likewise, cyclohexamide exposure decreases HSV-2 productive replication > 95% but does not abrogate HSV-2-induced C. trachomatis persistence. These data indicate that productive HSV-2 replication is not required to stimulate chlamydial persistence. Rather, these results suggest that an early event in the viral replication cycle, most likely host cell attachment and entry, is sufficient to force chlamydiae into a persistent state.

Both cyclohexamide exposure and UV-inactivation strongly inhibit de novo HSV protein synthesis, particularly those of the delayed-early and late kinetic classes. Although a small amount of immediate-early viral protein synthesis occurs in both cases, the magnitude is considerably diminished, particularly in the case of UV-inactivated virus. In contrast, viral attachment and entry are unaffected (Swanstrom et al., 1975; Vasquez, 1979; Moxley et al., 2002; Sanfilippo et al., 2004; Sanfilippo and Blaho, 2006). However, co-infection with HSV-2UV induced chlamydial persistence, suggesting that de novo viral protein synthesis is not required. Data obtained from cyclohexamide exposure experiments also support this interpretation. These experimental results suggest that chlamydial persistence in co-infected cells is induced by: (i) a virus/host cell binding event; (ii) introduction of a virion-associated protein into the host cell or; less likely (iii) expression of a viral immediately early transcript.

During viral attachment and entry, HSV binds to one of several co-receptors using the viral gD envelope glycoprotein (Spear, 2004). These co-receptors include herpes viral entry mediator (HVEM), nectin-1 and nectin-2 and 3-O-sulfated heparan sulfate (Spear, 2004). HVEM is a member of the tumour necrosis factor receptor family (Mauri et al., 1998). When complexed to its natural ligand, LIGHT, HVEM has been shown to stimulate cellular transduction pathways involved in the activation of T-cells (Hsu et al., 1997; Granger and Rickert, 2003). Nectin-1 and nectin-2 are members of the immunoglobulin superfamily and are involved in the formation of cell junctional complexes (Cocchi et al., 1998). When stimulated, nectins interact with cell signalling molecules, Cdc42 and Rac small G proteins, through their cytoplasmic tails to co-ordinate cytoskeletal rearrangements (Nakanishi and Takai, 2004). Given these observations, it is feasible that stimulation of viral co-receptors by attachment of HSV-2 could transmit a cellular signal that has downstream effects on developing chlamydiae.

Herpes simplex virus has also been shown to trigger cellular defence mechanisms by interacting with toll-like receptors (TLR). TLR-2 is host cell surface-exposed and appears to recognize herpes virion envelope glycoproteins. For example, the gB envelope glycoprotein of human cytomegalovirus, a betaherpesvirus, provides sufficient stimulus to activate anti-viral responses through TLR-2 (Compton et al., 2003). Infection with varicella zoster virus, HSV-1 and HSV-2 can all stimulate IL-6 production in culture; this induction is TLR-2-dependent and does not require productive viral replication (Aravalli et al., 2005; Kurt-Jones et al., 2005; Wang et al., 2005). TLR-2 knockout mice challenged with HSV-1 exhibit reduced production of inflammatory cytokines and diminished neuropathogenesis when compared with wild-type control animals, suggesting that HSV-mediated stimulation of the inflammatory response through TLR-2 occurs in vivo as well (Kurt-Jones et al., 2004). Additionally, the intracellular TLR, TLR-9, is stimulated by unmethylated CpG motifs commonly found in HSV genomic DNA (Pyles et al., 2002). Although not definitive, these studies suggest the possibility that HSV-mediated stimulation of TLR-linked anti-viral or anti-chlamydial cascades could affect developing chlamydiae within co-infected cells.

Another possibility is that viral transport within the cell, activity of a virion-associated protein or nuclear entry of viral DNA is responsible for inducing chlamydial persistence. After fusion with the cellular membrane, viral capsids interact with the molecular motor dynein and are transported along microtubules to the nucleus where capsids dock with nuclear pore complexes and translocation of the viral genome occurs (Ojala et al., 2000). Upon viral entry, active viral tegument proteins are released into the cytoplasm. One such protein, VP22, functions as a microtubule-associated protein and induces reorganization of microtubules in the host cell cytoplasm (Elliott and O'Hare, 1998), suggesting the possibility that HSV co-infection might alter or inhibit vesicular transport to the developing inclusion. However, inclusion size analyses revealed that chlamydial inclusions in co-infected cells are actually larger than those in C. trachomatis singly infected cells (Deka et al., 2006). These data suggest that viral-induced changes in host cell trafficking diminish neither vesicular transport to, nor development of, the inclusion. Another virion protein, the HSV-2 virion host shut-off protein (VHS), immediately suppresses host transcription by inducing degradation of host mRNAs (Kwong et al., 1988; Roizman and Knipe, 2001). VP16 is a virion protein that is transported to the nucleus and is involved in activation of both viral and host gene transcription (Roizman and Knipe, 2001). Thus, it is also possible that the activity of one or more virion-associated proteins, such as VHS and VP22, could either activate a cellular anti-chlamydial response or disrupt a cellular pathway that is required for normal chlamydial development, leading to persistence. As stated earlier, another intriguing possibility is that viral nuclear entry or other viral nuclear functions that occur prior to immediate-early gene expression might alter chlamydial development in co-infected cells. Unfortunately, the limited information available regarding host nuclear functions required for normal chlamydial development makes it difficult to speculate upon which virus/host nuclear interactions might be required for this effect.

Previous studies have confirmed the important role that chlamydia/host cell interactions play in the proper development of infectious chlamydial progeny. When these crucial host cell interactions are disrupted, chlamydiae often survive by becoming persistent (Beatty et al., 1994a; Raulston, 1997; Darville, 2000; Gerard et al., 2001). Additionally, it appears that the host has evolved ‘anti-chlamydial’ pathways, such as the IFN-γ/IDO system, which, when stimulated, interfere with chlamydial development. The developing chlamydiae may then respond by becoming persistent (Hogan et al., 2004). This study suggests that attachment and entry of HSV-2 may: (i) disrupt host cellular function, such that chlamydiae cannot complete the developmental cycle and/or (ii) trigger a novel host anti-chlamydial pathway that restricts chlamydial development. Examination of the cellular pathways that are stimulated by HSV-2 attachment and entry during co-infection will further illuminate the mechanisms that C. trachomatis uses to exist within its host cell and the methods by which host cells combat this bacterial invader.

Experimental procedures

Chlamydia, HSV-2 and host cells

A human urogenital isolate of C. trachomatis E/UW-5/CX was originally obtained from S.P. Wang and C.C. Kuo (University of Washington, Seattle, WA). The same standardized inoculum of C. trachomatis serovar E elementary bodies, propagated in McCoy cells, was used for all experiments (Wyrick et al., 1996). HSV-2 strain 333 and HSV-1 strain KOS stocks were obtained from Mary K. Howett and Udayasankar Kumaraguru. Viral stocks were prepared in monolayers of Vero cells (African green monkey kidney cells ATCC No. CCL-81) using standard techniques (Duff and Rapp, 1971).

Co-infection experimental design

HeLa cells, a cervical adenocarcinoma epithelial cell line (ATCC No. CCL2), or HEC-1B cells, an endometrial epithelial cell line (ATCC No. HTB-113) were used for all infection experiments. In each experiment the appropriate host cells were divided into four groups, 1 × 106 cells per 60 mm culture dish, for mock infection, chlamydial infection, HSV infection, and double infection. In some experiments, HEC-1B or HeLa cells were polarized on 3.0 μm, Collagen IV-coated chamber inserts (Biocoat 4544, Becton Dickinson) as previously described (Wyrick et al., 1989). The results obtained from co-infection in polarized host cells were identical to those obtained using non-polarized cells. Host cells were infected with a dilution of crude EB stock (200 μl) calculated to infect > 80% of the cells. For most experiments the cells were infected with HSV-2, HSV-1, or an amount of replication-incompetent HSV-2 (HSV-2uv) equivalent to an moi of 10 pfu/cell. In some experiments, HSV-2 infections were also conducted at 1 or 0.1 moi. Mock infected cells were treated similarly except they were exposed to 200 μl of either 2SPG (0.2 M sucrose, 6 mM NaH2PO4, 15 mM Na2HPO4, 5 mM l-glutamine, pH 7.2; mock C. trachomatis-infected) or growth medium (mock viral infection).

Cyclohexamide exposure during co-infection

In a selected group of co-infections, the infected host cells were exposed to cyclohexamide. Monolayers of HeLa cells were either mock or chlamydia-infected. Twenty-three hours post chlamydial infection, 1 μg ml−1 cyclohexamide or an equivalent volume of ddH2O (diluent exposed controls) was added to the growth medium. Cells were incubated at 37°C for another hour, at which time they were either mock or HSV-2-infected. The viral inoculum also contained 1 μg ml−1 cyclohexamide or diluent (ddH2O). Following HSV-2 infection the cells were refed with medium again containing either 1 μg ml−1 cyclohexamide or ddH2O. The cells remained in the presence of cyclohexamide until they were collected at 20 h post HSV-2 infection.

Generation of HSV-2UV

A UV cross-linker (Spectroline Microprocessor Controlled UV-Crosslinker Spectrolinker XL1500, Spectronics Corporation, New York) was used to generate stocks of UV-inactivated, replication-incompetent HSV-2. Stock HSV-2 was thawed and 200 μl was aliquoted into each well of a 24-well plate. The plates were placed on a 4°C heat sink during UV exposure to prevent heat inactivation of the samples. HSV-2 inactivation was assayed by performing plaque assays with the UV-inactivated virus using the same protocol as detailed for replication-competent HSV-2. A UV dose of 2.5 J cm−2 was found to be sufficient to completely inactivate 109 pfu of HSV-2.

Purification of HSV-2 and mock infected crude stocks

Viral purification was performed as described previously, with the following modifications (Sathananthan et al., 1997). Duplicate flasks of Vero cells were either mock or HSV-2-infected. The cells were monitored and collected when the virally infected cells began to exhibit signs of viral cytopathic effect. Each stock (mock and HSV-2-infected) was split into duplicate aliquots. The stocks were frozen/thawed to lyse the cells and centrifuged for 10 min (1000 g, at 4°C) to remove large cellular debris. The supernatant from one aliquot of the mock and HSV-2 stocks was subaliquoted and stored at −80°C. The supernatant from the second aliquot of each stock was centrifuged again for 1 h at 18 000 r.p.m. (392 000 g) in a Beckman JA-20 rotor. After centrifugation, the supernatants were removed and the pellets were resuspended overnight in 4 ml of growth medium without agitation. The purified mock and HSV-2 stocks were subaliquoted and stored at −80°C. Plaque assays were performed to determine the titre of each stock and to ensure that both the crude mock and purified mock stocks were free of contaminating HSV-2.

Luminex assay

Aliquots of crude, mock crude, pure and mock pure HSV-2 stocks were examined for the cytokines IFN-γ, IFN-α, TNF-α and IL-6 using the BioSource Multiplex Bead Immunoassay (BioSource International, Camarillo, CA) according to the manufacturer's instructions. Samples were UV-irradiated as previously described before the assay was performed such that no infectious virions were present in the samples. A Luminex 100 instrument (Luminex Corporation, Austin, TX) was used to measure the quantity of each cytokine bound to antibody-coupled beads.

Fluorescent and transmission electron microscopy

Fluorescence analyses were performed as described previously (Deka et al., 2006) except FITC-conjugated monoclonal antibodies generated against C. trachomatis MOMP (Pathfinder C. trachomatis monoclonal antibody 30702, Bio-Rad) were used to stain chlamydial inclusions. Duplicate samples of infected HeLa and HEC-1B cells were processed at 20 h post HSV infection for high-contrast TEM as described (Wyrick et al., 1994). Counterstained gold thin sections were examined using a Tecnai 10 (FEI) transmission electron microscope operating at 60–80 kV.

Chlamydial titrations by subpassage

Chlamydial titrations were performed as previously described (Deka et al., 2006) except that Pathfinder anti-chlamydial stain was used to stain chlamydial inclusions formed from subpassaged EBs. Inclusions present on triplicate HeLa cell monolayers were counted and averaged. The number of inclusion-forming units (ifu) in the undiluted inoculum was then calculated and expressed as ifu ml−1.

DNA isolation

Infected HeLa cells were collected at 0 h (T0) and/or 20 h (T20) post HSV infection in 200 μl medium and lysed by freeze/thaw. Total DNA from the resulting cell lysates was isolated using the QIAmp DNA Blood Mini Kit (Qiagen) according to the manufacture's instructions. Total DNA preparations were quantified using optical density (OD) at 260 and 280 nm; all samples had OD260/280 ratios > 1.9.

Polymerase chain reaction

Polymerase chain reaction was performed using purified total cellular DNA as template. Experimental template DNAs were used at dilutions ranging from 1/10 to 1/1000 (in ddH2O) such that each reaction was in the linear amplification range. PCR was performed using identical conditions to those previously described (Deka et al., 2006). Published primer sets included HSV-2 Glycoprotein G (Filen et al., 2004) as well as chlamydial 16S rRNA and human glyceraldehyde-3-phosphate dehydrogenase (Deka et al., 2006). After PCR, all reactions were electrophoresed on 1.5% agarose/TBE gels stained with ethidium bromide. A Bio-Rad Chemi Doc XRS Image Capture System with Quantity One V4.5.0 software (Bio-Rad) was used to visualize and quantify amplimers.

HSV plaque assay

At various times post HSV infection, culture supernatants were centrifuged at 4000 g for 5 min at 4°C to remove cell debris. Plaque assays were carried out on the resulting supernatants as described (Duff and Rapp, 1971). Quadruplicate, infected cultures were incubated for 72 h at 37°C, fixed, stained and counted. Average plaque counts from each set of plates were used to calculate the pfu ml−1 present in the original supernatant.

Statistical analyses

Statistical analyses were performed using Microsoft Excel. Comparison of means was performed by using a two-sample t-test for independent samples. P-values of ≤ 0.05 were considered significant.


The authors would like to thank Dr Priscilla B. Wyrick, Dr Jane Raulston, Dr Sophie Dessus-Babus, Dr John Laffan, Dr Udayasankar Kumaraguru and Dr Michelle Duffourc for helpful discussion of these experiments. Finally, we would like to acknowledge the excellent work of the Electron Microscopy Core Facility, Department of Pathology, James H. Quillen College of Medicine. This work was supported by NIH Grant No. 5R21AI59563 to R.V.S. and ETSU RDC Grant No. 04-024M to R.V.S.