Chlamydia Inhibits Interferon γ–inducible Major Histocompatibility Complex Class II Expression by Degradation of  Upstream Stimulatory Factor 1

The following human cell lines were used: MCF-7 (a mammary epithelium line, provided by Dr. Arnold Greenberg of the Manitoba Cancer Foundation); MRC-5 (fibroblast; American Type Culture Collection [ATCC]); 2C4 (fibroblast; provided by Dr. George Stark of the Cleveland Clinic Foundation); and HeLa (cervical epithelium; ATCC). Cells were infected with Chlamydia trachomatis LGV2 strain at a multiplicity of infection (MOI)¹ of  5 or as indicated and for 24 h or as indicated in individual experiments (24). Cells with or without infection were stimulated with human IFN-γ (PharMingen) at 200 U/ml or as indicated for another 10 h (for reverse transcriptase [RT]-PCR analysis) or 20–24 h (for flow cytometry and Western blot analysis).

Cell samples were stained with mouse anti– HLA-DRα (L243; ATCC), mouse anti–human intercellular adhesion molecule (ICAM)-1 (HA58; PharMingen), or normal mouse IgG (Zymed Labs., Inc.). Primary antibody binding was detected using goat anti–mouse IgG conjugated with FITC (Caltag Labs.) and analyzed with a FACSCalibur™ equipped with CellQuest software (Becton Dickinson). Dead cells were excluded by propidium iodine staining.

Western blot assay was carried out as we previously described (24). Rabbit antibodies were used to detect IFN-γR (SC-700; Santa Cruz Biotechnology), tyrosine-phosphorylated signal transducers and activators of transcription (STAT)1α (9171S; New England Biolabs, Inc.), IFN regulatory factor (IRF)-1 (SC-497), upstream stimulatory factor (USF)-1 (SC-229) and USF-2 (SC-862; all from Santa Cruz Biotechnology). Mouse antibodies were used to detect Janus tyrosine kinase (JAK)-1 (J24320; Transduction Labs.) and STAT1α (SC-464; Santa Cruz Biotechnology), HLA-DRα (DA6.147; provided by Dr. Peter Cresswell, Yale University; reference 27), and a chlamydial major outer membrane protein (MOMP; clone MC22, our unpublished data). Primary antibody binding was detected with horseradish peroxidase–conjugated goat anti–mouse IgG or –rabbit IgG, depending on the source of the primary antibodies, and visualized using an ECL kit (Amersham Corp.).

Cell samples were collected for RNA extraction using the Rneasy Mini Kit from QIAGEN, Inc. 2 μg of total RNA was used for each cDNA synthesis with random primers and the 1^st Strand cDNA synthesis kit from Boehringer Mannheim. Aliquots of the cDNA samples were used as a template for amplifying specific gene fragments by PCR reactions (28, 29). The primers used for amplification of DRα (18), DMα (29), invariant chain p41 (IP41) (29), and IRF-1 (18) were previously described. The other primers used in this study were: for class II transactivator (CIITA) amplification, 5′-GACACGGTGGCGCTGTGGGAGTC-3′ (forward) and 5′-GGCAGCCGTGAACTTGTTGTACTGG-3′ (reverse); for USF-1 amplification, 5′-TGGCACTGGTCAATTCTTTGTG (forward) and 5′-GTTGCTGTCATTCTTGATGAC (reverse); for STAT1 amplification, 5′-TAGAGTTGCTGAATGTCACTG-3′ (forward) and 5′-GGAGTGAAGCTCTTCAGTAAC-3′ (reverse); for indoleamine 2,3-dioxygenase (IDO) gene amplification, 5′-ATGCATCACCATGGCATA-3′ (forward) and 5′-GCTTCCCGCAGGCCAGCATCA-3′ (reverse); and for β-actin amplification, 5′-GTGGGGCGCCCCAGGCACCA-3′ (forward) and 5′-CTCCTTAATGTCACGCACGATTTC-3′ (reverse). β-actin mRNA detection was used as an internal control for the amount of cDNA synthesized. To ensure the specificity of the mRNA detection, all primers were designed to cover at least two exons, and parallel samples without RT were run as negative controls. The amplified DNA products were run on an agarose gel and visualized with ethidium bromide staining.

To investigate whether chlamydia possesses the ability to evade the IFN-γ–induced immune recognition mechanism, we evaluated IFN-γ–inducible MHC class II antigen expression in cells with or without chlamydial infection. IFN-γ significantly upregulated HLA-DR surface expression on uninfected cells, whereas the chlamydia- infected cells displayed a minimal level of  DR, regardless of IFN-γ exposure (Fig. 1 A). However, chlamydial infection did not affect the IFN-γ–induced ICAM-1 surface expression (Fig. 1 A). These observations suggest that chlamydia selectively inhibits IFN-γ–inducible DR expression rather than preventing all IFN-γ dependent signaling or generally suppressing surface protein expression. Furthermore, the total cellular protein level of  IFN-γ–induced HLA-DRα was also significantly diminished in chlamydia-infected cells as compared with uninfected cells (Fig. 1 B), suggesting that the suppression of surface expression of  HLA-DR was not due to an alteration in intracellular trafficking. The chlamydial inhibition of IFN-γ–inducible HLA-DRα was reproduced in many other human cell lines, including HeLa, MRC-5, and 2C4 (Fig. 1 B), demonstrating that the inhibitory effect is not a cell line–specific phenomenon. To determine whether the chlamydial inhibition of HLA-DR expression occurs at the transcription or translation level, MHC class II mRNA levels were evaluated by semiquantitative RT-PCR. IFN-γ dramatically induced the expression of DRα, DMα, and Ip41 mRNA in the uninfected but not the infected cells (Fig. 1 C), suggesting that chlamydial inhibition of MHC class II occurred at the transcription level. Because the genes encoding the MHC class II presentation–related molecules DRα, DMα, and Ip41 share similar promoter structures and CIITA is a master regulator for the expression of these genes (30), we hypothesize that chlamydia may inhibit CIITA function or CIITA gene expression.

Although CIITA is constitutively expressed in professional APCs, such as dendritic cells and B cells, IFN-γ stimulation is often required for the expression of CIITA in nonprofessional APCs, such as epithelial cells (31). CIITA mRNA expression was induced by IFN-γ in uninfected MCF-7 cells (Fig. 2 A). However, CIITA mRNA expression was significantly lower in chlamydia- infected and IFN-γ–treated cells (Fig. 2 A), in accord with chlamydial inhibition of MHC class II gene expression. To investigate how CIITA gene expression was inhibited, we measured mRNA levels for three transcription factors, USF-1, STAT1, and IRF-1, all of which are required for IFN-γ–inducible transcription of the CIITA gene (32). Both USF-1 and STAT1 mRNAs were constitutively expressed, and IRF-1 mRNA was induced by IFN-γ in MCF-7 cells regardless of chlamydial infection (Fig. 2 A), suggesting that chlamydial infection did not affect transcription of the genes for these nuclear factors. Because the three transcription factors are considered to be sufficient and necessary for the IFN-γ induction of CIITA (32), we evaluated the protein levels of these transcription factors as well as upstream molecules in the IFN-γ signaling pathway. We found that IFN-γR, JAK-1, and STAT1 protein levels were not altered by chlamydial infection (Fig. 2 B). Chlamydia did not affect IFN-γ–induced STAT1 tyrosine phosphorylation (Fig. 2 B). Furthermore, IFN-γ induced both IRF-1 and ICAM-1 expression in chlamydia-infected cells (Figs. 1 A and 2 B). As STAT1 is required for the expression of both IRF-1 and ICAM-1 genes (32, 33), we conclude that STAT1 is transcriptionally functional in chlamydia-infected cells. The IFN-γ–induced IRF-1 in chlamydia-infected cells may also be transcriptionally functional, as we found that IFN-γ induced expression of  IDO gene in chlamydia-infected cells (data not shown), and it is known that IRF-1 is required for IFN-γ induction of IDO (34). Therefore, the failure of the IFN-γ–inducible CIITA expression in chlamydia-infected cells is likely due to the deficiency in USF-1. We found that the USF-1 protein was not detectable in chlamydia-infected cells (Fig. 2 B), despite normal USF-1 mRNA expression (Fig. 2 A). We next determined the cause of the USF-1 protein loss.

Because USF-1 mRNA is expressed in chlamydia-infected cells with or without IFN-γ stimulation, the lack of USF-1 protein may be due to either the inhibition of translation of  USF-1 mRNA or the accelerated degradation of USF-1 protein. We tested the protein degradation hypothesis by using protease inhibitors. We found that the proteasome inhibitor lactacystin prevented USF-1 protein degradation in chlamydia-infected cells (Fig. 2 C). Furthermore, the lactacystin treatment also preserved the IFN-γ–inducible HLA-DR expression in chlamydia-infected cells (Fig. 2 C). These observations not only demonstrate that a proteasome-like activity is responsible for the loss of USF-1 protein in chlamydia-infected cells but also suggest that USF-1 degradation may be responsible for the chlamydial suppression of MHC class II expression.

To examine whether the USF-1 protein degradation in chlamydia-infected cells is dependent on chlamydial growth and protein synthesis, we first evaluated the relationship between the chlamydial infection dose and USF-1 degradation. As MOI (ratio of number of organisms versus number of host cells) increased, more chlamydial protein was produced and less USF-1 protein was detected (Fig. 3 A). This effect was selective, as USF-2 was not degraded, regardless of the infection dose (Fig. 3 A). The time course relationship between chlamydial growth and USF-1 degradation was also analyzed (Fig. 3 B). Although the STAT1 protein level was not affected by chlamydial infection at any time points examined, significant USF-1 degradation was detected 17 h after chlamydial infection, when chlamydial protein synthesis approached its maximum (Fig. 3 B). The role of chlamydial and host protein synthesis in USF-1 degradation was examined using antibiotics specifically inhibiting either prokaryotic or eukaryotic protein synthesis. We found that both rifampin (inhibiting prokaryotic transcription) and chloramphenicol (inhibiting prokaryotic translation) blocked chlamydial protein synthesis (Fig. 3 C). More importantly, these antibiotics also prevented USF-1 degradation and preserved HLA-DRα expression in chlamydia-infected cells (Fig. 3 C). However, treatment with penicillin failed to prevent USF-1 degradation and preserve HLA-DRα expression (Fig. 3 C). Penicillin only blocks chlamydial particle assembly without inhibiting chlamydial protein synthesis (35). Penicillin did not alter the constitutively expressed USF-1 protein level and IFN-γ–inducible HLA-DRα expression in uninfected cells (Fig. 3 C). Together, these observations demonstrate that chlamydial protein synthesis, but not particle assembly, is necessary for chlamydia-induced degradation of USF-1 protein and suppression of HLA-DRα expression. Finally, cycloheximide treatment did not affect the chlamydia- induced degradation of  USF-1 (Fig. 3 C). Because cycloheximide did not affect chlamydial protein synthesis but completely inhibited new protein synthesis by the host cell, for example, production of  IFN-γ–induced HLA-DRα (Fig. 3 C), we conclude that newly synthesized host proteins are not required for chlamydia-induced degradation of USF-1.

We have demonstrated that the obligate intracellular bacterial pathogen chlamydia can inhibit IFN-γ–inducible MHC class II expression. This inhibitory effect has also been found with other intracellular pathogens, including leishmania (13), listeria (14), cowdria (15), and cytomegalovirus (16–18). CD4⁺ T cell–mediated immunity plays an important role in host defense against various intracellular infections (36–38). Recognition of the infected cells by CD4⁺ T cells often requires IFN-γ induction of MHC class II expression, because many pathogen-targeted cells, such as epithelial cells, are generally MHC class II–negative. Suppression of IFN-γ–inducible MHC class II expression may represent an efficient immune evasion strategy used by intracellular pathogens to escape host defenses. Thus, chlamydial inhibition of IFN-γ–inducible MHC class II may contribute to the persistent infection caused by chlamydia in humans (22).

It has been demonstrated that cytomegalovirus can prevent IFN-γ–inducible class II expression in infected cells by both IFN-β–mediated inhibition (17, 39) and disruption of IFN-γ intracellular signaling pathways (16, 18). However, mechanisms of IFN-γ–inducible MHC class II inhibition by many other intracellular pathogens are still not clear (13–15). It was recently proposed that chlamydia may suppress IFN-γ–inducible MHC class II expression by stimulating host cells to release IFN-β (40). Here we show that the intracellular bacterial pathogen chlamydia has evolved a more specific mechanism for disrupting IFN-γ signaling pathways and inhibiting MHC class II expression. Chlamydia degrades USF-1, a downstream transcription factor required for IFN-γ–inducible MHC class II but not IRF-1 and ICAM-1 expression (Fig. 4). USF-1 degradation may represent an efficient means of interrupting IFN-γ– inducible MHC class II expression by chlamydia. Although the constitutively and ubiquitously expressed USF-1 is a member of the basic helix-loop-helix family consisting of multiple transcription factors, including Myc and USF-2, only USF-1 is both necessary and sufficient for binding to the E  box within the CIITA promoter IV and cooperating with STAT1 and IRF-1 for promoting transcription of CIITA (32). Therefore, the constitutively and ubiquitously expressed USF-1 may serve as a convenient and efficient target for chlamydia-induced degradation. The correlation between the degradation of USF-1 and the suppression of IFN-γ–inducible MHC class II further confirms that USF-1 plays a critical role in IFN-γ induction of  MHC class II (32). Besides its involvement in MHC class II expression, USF-1 also participates in many other cellular activities, including promoting the transcription of fatty acid synthase in response to insulin regulation (41), interfering with Ras transformation (42), and transactivating the promoter of the p53 tumor suppressor gene (43). Depletion of USF-1 may cause inhibition of host cell lipid biosynthesis and promotion of host cell survival, both of which are likely beneficial to the intracellular chlamydia organisms.

Proteolysis is an important aspect of normal cellular physiology (44–46). Many viruses can take advantage of  host proteolysis for the purposes of evading host defenses (2, 47). For example, human cytomegalovirus infection can induce degradation of JAK-1, a critical upstream kinase required for IFN-γ JAK/STAT signaling pathways, to suppress IFN-γ–inducible MHC class II on the infected cells (18). Furthermore, the cytomegalovirus-induced degradation can be inhibited by the proteasome inhibitor Z-L₃VS (18, 48), suggesting that cytomegalovirus may be able to manipulate host proteasome activity. Because USF-1 degradation by chlamydia is inhibitable by lactacystin and lactacystin is a potent proteasome inhibitor (48, 49), we propose that chlamydia may also produce a factor(s) for manipulating host proteasomes. Efforts to identify the chlamydial factor(s) are underway.

We thank Dr. Ronald N. Germain for helpful discussions and critically reading the manuscript and Drs. Peter Cresswell, George Stark, and Arnold Greenberg for providing useful reagents for this work.

This work was supported by the Medical Research Council (MRC) of Canada. G. Zhong is the recipient of an MRC scholarship.

CIITA

class II transactivator

IDO

indoleamine 2,3-dioxygenase

IRF

interferon regulatory factor

JAK

Janus tyrosine kinase

MOI

multiplicity of infection

MOMP

major outer membrane protein

RT

reverse transcriptase

STAT

signal transducers and activators of transcription

USF

upstream stimulatory factor