Alterations in TLRs as new molecular markers of congenital infections with Human cytomegalovirus?

Authors

  • Wioletta Wujcicka,

    1. Department of Fetal-Maternal Medicine and Gynecology, Polish Mother's Memorial Hospital Research Institute, Lodz, Poland
    Search for more papers by this author
  • Jan Wilczyński,

    1. Department of Fetal-Maternal Medicine and Gynecology, Polish Mother's Memorial Hospital Research Institute, Lodz, Poland
    Search for more papers by this author
  • Dorota Nowakowska

    Corresponding author
    1. Department of Fetal-Maternal Medicine and Gynecology, Polish Mother's Memorial Hospital Research Institute, Lodz, Poland
    • Correspondence

      Dorota Nowakowska, Department of Fetal-Maternal Medicine and Gynecology, Polish Mother's Memorial Hospital Research Institute and Medical University in Lodz, 281/289 Rzgowska Street, Lodz 93-338, Poland.

      Tel.: +48 42 271 10 77

      fax: +48 42 271 10 70

      e-mail: dnowakowska@yahoo.com

    Search for more papers by this author

  • In this article, the existing literature on Human cytomegalovirus is analysed from the viewpoint of the innate immune response, focusing in particular on the role of Toll-like receptors (TLRs) and the role that they might play in transmission and pathogenesis in congenital infection. The relevance of genetic polymorphisms in TLR genes to disease outcome propels the interest for further studies to evaluate their potential as genetic markers to predict disease susceptibility.

Abstract

Toll-like receptors (TLRs) play a crucial role in non-specific immunity against various infections. The most common intrauterine infection, caused by Human cytomegalovirus (HCMV), results in perinatal morbidity and mortality of primary infected fetuses. The induction of immune response by TLRs was observed in HCMV infections in murine models and cell lines cultured in vitro. Studies reported an immunological response in pregnant women with primary HCMV infection and TLR2 activity in collecting of HCMV particles in placental syncytiotrophoblasts (STs) in vivo and cultured ST, and in stimulation of tumor necrosis factor (TNF)-α expression and damage of villous trophoblast. Expression levels of TLRs are associated with cell type, stage of pregnancy and response to microorganisms. We show the effect of HCMV infection on the development of pregnancy as well as the effect of TLR single-nucleotide polymorphisms on the occurrence and course of infectious diseases, immune response and diseases of pregnancy. We report the impact of TLRs on the function of miRNAs and the altered expression levels of these molecules, as observed in HCMV infections. We suggest that the methylation status of TLR gene promoter regions as epigenetic modifications may be significant in the immune response to HCMV infections. We conclude that it is important to study in detail the molecular mechanisms of TLR function in the immune response to HCMV infections in pregnancy.

TLRs contribute to immune response against HCMV

The host organism recognizes the threat of infection mainly through mechanisms of non-specific defense, which detect highly conserved structures of pathogens (Janeway & Medzhitov, 2002; Wilkins & Gale, 2010; Kumar et al., 2011). These structures, called pathogen-associated molecular patterns (PAMPs), play a crucial role in pathogen metabolism and are invariable (Janeway & Medzhitov, 2002; Akira & Hemmi, 2003; Takeda et al., 2003). This means that the identification of pathogens by the non-specific defense mechanisms is also conservative. Host cells identify PAMPs by special receptors known as pattern recognition receptors (PRRs), which are divided into three groups: secreted, surface and intracellular receptors (Medzhitov, 2007; Kumar et al., 2009; Mogensen, 2009; Netea et al., 2011). The group of surface PRRs includes most of the Toll-like receptors (TLRs), transducing signals from PAMPs to the cell interior, activating these cells and representing the first line of host defense against pathogens (Medzhitov & Janeway, 2002; Iwasaki & Medzhitov, 2004; Chaturvedi & Pierce, 2009; Brown et al., 2011). TLRs localize to various cellular compartments such as plasma membrane (TLR1, TLR2, TLR4, TLR5 and TLR6) or endolysosomes (TLR3, TLR7, TLR8 and TLR9; Ahmad-Nejad et al., 2002; Chaturvedi & Pierce, 2009). TLR9, present in fagolysosomes, is activated after fagocytosis of bacteria and release of bacterial ligands such as cytosine-phosphate-guanosine dinucleotides (CpG) for TLR9. TLRs function as monomers and can also heterodimerize, which suggests that more PAMP structures may be recognized by TLR (Chang et al., 2007; Farhat et al., 2008; Govindaraj et al., 2010; Stewart et al., 2010). TLR2 as a monomer recognizes lipoteichoic acid (LTA) in the cell wall of Gram-positive bacteria, whereas zymosan from yeast is recognized by TLR2/TLR6 heterodimer (Ozinsky et al., 2000). In vitro studies with Human cytomegalovirus (HCMV) permissive fibroblasts showed that TLR2 functionally senses HCMV through direct interaction with HCMV envelope glycoproteins (gp), gB and gH (Boehme et al., 2006). TLR2/TLR1 heterodimer was reported to be a functional sensor for HCMV, as anti-gB and anti-gH antibodies blocked the inflammatory cytokine response against HCMV and both glycoproteins coimmunoprecipitated with TLR2 and TLR1 (Boehme et al., 2006).

Recognition of PAMPs by TLRs plays a key role in activation of signaling cascades leading to the synthesis of proinflammatory molecules (Medzhitov & Janeway, 2000; Dalpke & Heeg, 2002; Brown et al., 2011). This interaction induces macrophages and dendritic cells (DCs) to release proinflammatory cytokines and chemokines involved in infection control as components of non-specific immunity (Re & Strominger, 2001, 2004; Schilling et al., 2002; Akira & Hemmi, 2003). At the same time, an increased expression of class I and class II major histocompatibility complex (MHC) and costimulating molecules enables effective induction of specific immunity (Hengel et al., 1995; Hegde et al., 2005; Sinzger et al., 2006; Kessler et al., 2008). TLRs play an important role in immunity through the induction of T regulatory (Treg) cells, which inhibit response, as well as in the induction of T lymphocytes abolishing Treg-cell function (van Maren et al., 2008; Lian et al., 2009; Wang et al., 2010; Oberg et al., 2011). In vivo studies with Mouse cytomegalovirus (MCMV) reported that TLR2, TLR3, TLR7 and TLR9 are involved in the viral sensing of and initiation of innate immunity against MCMV (Bozza et al., 2006; Szomolanyi-Tsuda et al., 2006; Crane et al., 2012; Traub et al., 2012). Similarly, in the case of HCMV infection, a significant role was reported for the same four TLRs (Varani et al., 2007; Iversen et al., 2009; Yew et al., 2010; Yew & Harrison, 2011). Studies showed that TLR alterations at the maternal–fetal interface are involved in innate immunity during gestation, as also observed in the case of HCMV (Chaudhuri et al., 2009) and bacterial infections (Holmlund et al., 2002; Forster-Waldl et al., 2005; Zariffard et al., 2005; Sadeghi et al., 2007; Rose et al., 2011), as well as in the pathogenesis of adverse pregnancy outcomes (Kim et al., 2004; Ilievski et al., 2007; Ilievski & Hirsch, 2010; Sado et al., 2011; Wujcicka et al., 2013).

One of the important causes of perinatal morbidity and mortality is intrauterine infection (Bergstrom, 2003; Romero et al., 2003; Vrachnis et al., 2010, 2012). HCMV infection is the most common intrauterine infection (Cannon, 2009; Syggelou et al., 2010; Cordier et al., 2012; Gaj et al., 2012; Paradowska et al., 2012). The seroprevalence of HCMV infection among pregnant women in Poland was estimated to be c. 76% (Paradowska et al., 2006; Gaj et al., 2012). Current laboratory diagnostics of HCMV infections is based on serological tests to detect specific antibodies of immunoglobulin M (IgM), immunoglobulin G (IgG) and immunoglobulin A (IgA) classes (Gutierrez & Maroto, 1997; Genser et al., 2001; Halwachs-Baumann, 2007; Duan et al., 2012). Primary infections are shown by low IgG avidity and seroconversion (Bodeus et al., 1998; Bodeus & Goubau, 1999; Yinon et al., 2010; Revello et al., 2011). However, diagnosis of these infections by seroconversion determination requires completing screening tests in early pregnancy, which is not practiced routinely. Since the late 1990s, quantitation of HCMV DNA load by real-time PCR assays has become the most widely used diagnostic tool (Sia et al., 2000; Atkinson & Emery, 2011; Mengelle et al., 2011; Khansarinejad et al., 2012). This is a rapid, efficient and sensitive method for determination of HCMV viremia level in newborns and HCMV viral copy number in amniotic fluid and placental cells, as well as urine and saliva samples from pregnant women with active HCMV infections and their newborns within the first week of birth (Paradowska et al., 2006, 2012; Ducroux et al., 2008; Kouri et al., 2010). The assessment of HCMV viremia level in pregnant women and their newborns is significant for early identification of the primary infections in these patients. Currently, the diagnosis of infection does not indicate the proper treatment strategy, as there are still no well investigated and effective medications licensed for application during pregnancy. Hence, it is very important in future to utilize the results of the prenatal tests to predict lesions resulting from cytomegaly in HCMV-infected newborns and children.

Various in vitro studies confirmed the role of TLRs in the non-specific immunity to HCMV infection (Table 1; Tabeta et al., 2004; Harwani et al., 2007; Varani et al., 2007; Iversen et al., 2009; Renneson et al., 2009; Yew et al., 2012). Experiments with human acute monocytic leukemia cell line THP1 and foreskin fibroblast cell lines infected with HCMV showed that within 10 min post exposure the expression of genes encoding TLR2 as well as scavenger receptor A type 1 (SR-A1), tyrosine-protein kinase Lyn and IL-12 p35 subunit was induced; after 1 h, TLR3, TLR9 and TIR domain containing adaptor-inducing interferon-beta (TRIF), interferon regulatory factor 3 (IRF-3), and interferon beta (IFN-β) were induced (Yew et al., 2010). In addition, studies with neutralizing antibodies and morpholino antisense oligonucleotides showed that transcription of tumor necrosis factor alpha (TNF-α) and interleukin 12 (IL-12) observed within 10 min was likely TLR2-dependent, whereas TLR3-mediated IFN-β and TLR9-mediated TNF-α expression appearing within 1 h was dependent on SR-A1 (Yew et al., 2010). Hence, endosomal TLR3 and TLR9 molecules induce transcription of crucial pro-inflammatory cytokines during infection with HCMV, although in the case of these receptors the virus was sensed via SR-A1. Another study showed that SR-A1-dependent induction of TLR3/9 signaling pathway was regulated via Lyn kinase, physically and functionally associated with SR-A1 (Yew & Harrison, 2011). Treatment of THP-1 monocytes with Lyn kinase oligonucleotides decreased expression of TLR9-stimulated TNF-α and strongly increased canonical TLR-3-stimulated IFN-β and non-canonical TLR3-stimulated nuclear factor kappa-B (NF-κB)-dependent IL-12p35 (Yew & Harrison, 2011). In turn, HCMV infection of plasmacytoid DCs (pDCs) resulted in partial maturation of pDCs, increased expression of MHC class II, cluster of differentiation 83 (CD83) and TLR9 molecules, and secretion of cytokines such as interferon-alpha (IFN-α; Varani et al., 2007). However, after inhibition with CpG, the expression of these cytokines was no longer stimulated, suggesting its possible regulation in a TLR7- and/or TLR9-dependent manner (Varani et al., 2007). Studies with monocyte-derived DCs (moDCs) infected with HCMV showed increased expression levels of TLR3 and chemokine ligands 10 (CXCL10), 11 (CXCL11) and 5 (CCL5), and cytokines of TLR3 signaling pathway such as IFN-α and IFN-β (Mezger et al., 2009). The expression of TLR3 was significantly reduced when the receptor was stimulated with TLR3 ligand polyinosinic-polycytidylic acid (poly I:C) after TLR3 silencing with small interfering RNA (siRNA); however, the diminished expression was no longer observed when moDCs were infected with HCMV (Mezger et al., 2009). The influence of specific TLR ligands on the course of HCMV infection was also assessed in foreskin fibroblasts and ectocervical tissue; TLR3 ligand (poly I:C) and TLR4 ligand lipopolysaccharide (LPS) inhibited HCMV and induced secretion of interleukin 8 (IL-8) and IFN-β (Harwani et al., 2007). Ligands for TLR2 (LTA) and TLR9 (CpG) inhibited HCMV infection in ectocervical tissue but not in foreskin fibroblasts (Harwani et al., 2007). In the case of HCMV-permissive human fibroblasts, physical association of HCMV envelope glycoproteins with TLR2 mediated NF-κB activation and inflammatory cytokine response (Boehme et al., 2006). Studies performed with primary fibroblasts infected with HCMV showed a strongly induced TLR9 expression in two of three fibroblast types, suggesting a potential contribution of TLR9 signaling to the development of HCMV infection (Iversen et al., 2009). The elevated level of TLR9 was correlated with a parallel proviral effect of TLR9 ligand CpG-B added to cultures of primary fibroblasts shortly after infection with HCMV (Iversen et al., 2009). The effect of CpG on the outcome of HCMV infection depended on the time of its administration (Iversen et al., 2009). Figure 1a illustrates that cellular receptors including TLRs and other molecules reported in the article play a role in the immune response at a very early stage of HCMV infection, namely, the attachment, entry and early activation phases. The HCMV immune evasive strategies observed after viral entry are illustrated in Fig. 1b. Phosphoprotein (pp) 65 was reported to be a key HCMV protein involved in the impairment of interferon regulatory factor 3 (IRF3) signaling as well as the expression of IRF1 and NF-κB (Abate et al., 2004; Rossini et al., 2012).

Table 1. The role of TLRs in the immune response to HCMV infection revealed from the in vitro studies
TLRCell lines cultured in vitroFunctionReferences
TLR2Syncytiotrophoblast culture derived from primary cytotrophoblastsActivation of the production and secretion of TNF-αChan & Guilbert (2006)
Human monocytoid THP1 cells and foreskin fibroblastsNearly immediate activation of TNF-α and IL-12p35 expressionYew et al. (2010)
Normal human dermal fibroblastsInitiation of inflammatory cytokine responseJuckem et al. (2008)
HCMV permissive human fibroblastsFunctional sensing of HCMV through direct interaction with HCMV envelope glycoproteins (gp), gB and gH. Mediation of NF-κB activation and inflammatory cytokine responsesBoehme et al. (2006)
TLR3Human monocytoid THP1 cells and foreskin fibroblastsIncrease in IFN-β mRNA levels at 1 h after HCMV infectionYew et al. (2010)
Monocyte-derived dendritic cells (moDCs)Lack of significantly altered expression of IFN-α, IFN-β, chemokine ligands 5 (CCL5), 10 (CXCL10), and 11 (CXCL11). No function in the early induction of IFN-β in human moDCsMezger et al. (2009)
Foreskin fibroblasts and ectocervical tissueInhibition of HCMV infection and induction of secretion of IL-8 and IFN-βHarwani et al. (2007)
TLR4Foreskin fibroblasts and ectocervical tissueInhibition of HCMV infection and induction of secretion of IL-8 and IFN-βHarwani et al. (2007)
TLR7 or TLR9Plasmacytoid DCs (pDCs)Induction of IFN-α secretion from pDCsVarani et al. (2007)
TLR9Human monocytoid THP1 cells and foreskin fibroblastsInduction of TNF-α expression at 1 h after HCMV infectionYew et al. (2010)
Neonatal human dermal fibroblastsInvolvement in HCMV infection developmentIversen et al. (2009)
TLRsMyeloid DCsInduction of IFN-β and IFN-λ gene expressionRenneson et al. (2009)
Figure 1.

(a) Cellular receptors and signaling pathways involved in the immune response at the early stage of HCMV infection. Human cytomegalovirus (HCMV) glycoprotein B (gB) as well as epidermal growth factor (EGF) bind specifically to epidermal growth factor receptor (EGFR). gB also binds directly to platelet-growth factor-α receptor (PDGFR-α) and dendritic cell-specific intercellular adhesion molecule-3-grabbing non-integrin (DC-SIGN). Toll-like receptor 2 (TLR2) interacts directly with HCMV glycoprotein gB and glycoprotein H (gH). In ectocervical tissue, the ligand for TLR3 (polyinosinic-polycytidylic acid, poly I:C) inhibits HCMV-induced secretion of IL-8 and IFN-β, and TLR9 ligand cytosine-phosphate-guanosine (CpG) inhibits HCMV infection. The invasion of HCMV into the host cell increases expression of MHC type II and cluster of differentiation 83 (CD83) molecules. In human trophoblast cell line, HCMV unique short region proteins 3 (US3) and 6 (US6) downregulate the surface expression of human leukocyte antigen C (HLA-C) and G (HLA-G) molecules. In experiments with villous explants cultured in vitro, HCMV alters protein expression levels of proto-oncogene c-ErbB-2 (c-erbB-2) and matrix metalloproteinase 2 (MMP-2) and 9 (MMP-9) molecules. In first trimester extravillous cytotrophoblast (EVT) cell line SGHPL4 HCMV infection inhibits secretion of MMP-2 and MMP-9. In pregnant women with primary HCMV infection, the infection induces cluster of differentiation 45RA (CD45RA) re-expression. Syncytiotrophoblast (ST)-like cells utilize cluster of differentiation 14 (CD14) as coreceptor for HCMV to regulate TNF-α expression. Myeloid differentiation primary response protein (MyD88) transmits signals from TLR 1, 2, 4, 5, 7, 8, 9 via activated transcription factors into the nucleus. NF-κB transcription factor stimulates transcription of inflammatory cytokine genes encoding IL-12 p35 subunit (IL-12), TNF-α (TNF-α) as well as IL1-β (IL-1β), IL-6 (IL-6) and IL-8 (IL-8). HCMV-infected fibroblast shows TLR2-dependent NF-κB activation. In human monocytoid THP1 and foreskin fibroblast cell lines, TLR3 transmits a signal via the TIR domain containing adaptor-inducing IFN-β (TRIF) that activates canonical interferon regulatory factor 3 (IRF-3)-dependent IFN-β transcription and non-canonical NF-kB-dependent gene transcription. TLR-3-mediated IFN-β and TLR9-mediated TNF-α expression are scavenger receptor A type 1 (SR-A1)-dependent. Lyn kinase is physically and functionally associated with SR-A1 and regulates TLR3/9 signaling pathways. Molecules considered crucial during HCMV infection are underlined and surrounded by a thick continuous line. (b) HCMV immunoevasion strategies observed after viral entry. HCMV tegument phosphoprotein (pp) 65 inhibits IRF-3 activation by reducing the phosphorylation level of IRF and thereby blocking type I IFN expression. pp65 also impairs the activation of NF-κB by unknown mechanisms.

TLRs activity in immune response to HCMV infections in pregnancy

Molecular mechanisms of immune response to HCMV have been characterized in pregnant women (Landini et al., 1984; Hopkins et al., 1996; Pizzato et al., 2004; Lilleri et al., 2009; Revello & Gerna, 2010). Pregnant women with primary HCMV infection had a cluster of differentiated 45RA (CD45RA) re-expression, cytotoxic T lymphocyte activity and perforin expression as the major components of the adaptive immune response of T lymphocytes that express CD4 (CD4+) and CD8 (CD8+) glycoprotein (Fornara et al., 2011; Fig. 1). The kinetics of V δ 2 γδ T-cells was similar to that of CD8+ T-cells. Additionally, CD45RA re-expression was associated with HCMV transmission to the fetus, suggesting its prognostic significance (Fornara et al., 2011). Another study performed with pregnant women with primary HCMV infection showed that the lymphoproliferative response (LPR) occurred slightly earlier in CD4+ than in CD8+ T-cells (Lilleri et al., 2007). Studies of peripheral blood leukocytes from both pregnant and non-pregnant women with primary HCMV infection showed the occurrence of HCMV-specific CD4+ T cells in both kinds of patients (Revello et al., 2006). However, LPR to HCMV was significantly lowered or delayed in transmitter compared with non-transmitter mothers (Revello et al., 2006). Syncytiotrophoblast (ST)-like cells cultured with HCMV or ultraviolet (UV)-HCMV utilized a cluster of differentiation 14 (CD14) as coreceptor for HCMV to regulate TNF-α expression (Chaudhuri et al., 2009). Both placental ST in vivo and cultured ST expressed CD14 strongly, which functions upstream of TLR2 to gather even transcriptionally inactive HCMV particles. As a result, TLR2 and CD14 stimulated TNF-α expression and villous trophoblast damage. In turn, experiments showed a lack of expression of TLR1 for both cell types, suggesting that this molecule is not involved in HCMV recognition on the ST (Chaudhuri et al., 2009). In a similar culture model (ST derived from primary cytotrophoblasts), ST infected with UV-HCMV elevated the transcription level of TNF-α and IL-8 inflammatory cytokines and doubled the frequency of ST apoptosis (Chan & Guilbert, 2006). The antibody to TLR2 blocked UV-HCMV-stimulated TNF-α transcription and translation and hence inhibited the progression of ST apoptosis (Chan & Guilbert, 2006). In turn, another study showed the role of TLR2 in recognition of HCMV infecting fibroblasts. The interaction between TLR2 and HCMV led to activation of the NF-κB and secretion of inflammatory cytokine (IC; Juckem et al., 2008). Nonetheless, TLR2 receptor did not participate in the IFN response to HCMV (Juckem et al., 2008). Results described in this work showed the role of TLRs in HCMV recognition in murine and human cell lines cultured in vitro with HCMV, although detailed molecular mechanisms of TLR function in innate immunity need to be studied. It seems likely that alterations in the expression of TLR genes occurring during pregnancy in non-immune cells such as trophoblasts, decidual cells and amniotic epithelium might have a significant function in the development of congenital cytomegaly.

TLR expression levels depend on cell type and stage of pregnancy

Many studies show altered TLR genes expression levels that correlate with cell type, stage of pregnancy and response to microorganisms (Ma et al., 2007; Koga & Mor, 2008, 2010; Nitsche et al., 2010; Riley & Nelson, 2010; Rautava et al., 2012; Wujcicka et al., 2013). TLR2, TLR3, TLR4 and TLR9 transcript expression identified in the uterus, cervix and placenta of non-pregnant and across gestation cluster of differentiation 1 (CD1) mice were significantly higher in pregnant uterine and cervical tissues (Gonzalez et al., 2007). The levels of analyzed transcripts differed between tissue types and decreased TLR4 expression was determined in the placenta (Gonzalez et al., 2007). The expression of TLR1-10 genes was observed in term human placental specimens taken from elective caesarean sections (ECS) and normal vaginal deliveries (NVD; Patni et al., 2009). Expression levels of TNF-α induced by TLR4 and TLR7/8 heterodimer agonists were higher in NVD than in ECS specimens (Patni et al., 2009). Differential occurrence of TLR2 and TLR4 transcription was determined between villous cytotrophoblasts, extravillous trophoblasts and syncytiotrophoblasts in the first trimester placenta (Abrahams et al., 2004). This suggested that pathogens enter the placental villous or decidual layers if the TLR-negative outer trophoblast compartment is interrupted (Abrahams et al., 2004, 2006). The expression of TLR2 and TLR6 was typical for trophoblast stem cells, but TLR1 and TLR4 expression was not determined in this cell type (Rose et al., 2011). Additionally, in term trimester placentas, TLR4 expression level was higher than in first trimester specimens (Beijar et al., 2006). Alterations of TLR expression levels observed in trophoblasts suggested that early trophoblasts provide less protection against pathogens compared with term trophoblasts (Rose et al., 2011). In turn, another study reported TLR4 expression and related IL-8 secretion after LPS stimulation as possibly characteristic of leukocytes (Klaffenbach et al., 2011). The analyzed gene expressions were not determined for trophoblast cell fractions purified from leukocytes (Klaffenbach et al., 2011). Paraffin-embedded sections of endometrium and deciduas from first and second trimester elective terminations and third trimester normal deliveries showed changed levels of TLR4 gene expression in various tissues as well (Schatz et al., 2012). Higher TLR4 immunoreactivity in decidual cells than in interstitial trophoblasts indicated the possibility that maternally derived cells are the main protectors against Gram-negative bacteria and factors related to severe inflammation (Schatz et al., 2012). Expression of TLR5, TLR6/2 heterodimer and TLR4 was identified in human amniotic epithelial cells (Gillaux et al., 2011). Active TLR5 and TLR6/2 induced the expression of IL-6 and IL-8 cytokines and NF-κB signaling pathway, whereas TLR4 was engaged in apoptosis of analyzed amniotic cells (Gillaux et al., 2011). TLR ligands were also reported to drive the activation of decidual natural killer (NK) cell and NK cell lines (Negishi et al., 2011).

To summarize, the differential expression of TLRs was shown for the maternal–fetal interface. In addition, the molecular alterations of these genes were reported to be associated with the pathogenesis of adverse pregnancy outcomes (Klaffenbach et al., 2005; Koga & Mor, 2008; Thaxton et al., 2009; Choi et al., 2012; Abrahams et al., 2013). Besides important studies reporting the role of TLRs in the development of HCMV infection, no works have investigated possible alterations in TLR expression levels related to this infection in pregnancy. Therefore, description of TLR gene expression profiles and their genetic determinants in pregnant women with primary HCMV infection might represent a significant new direction in the study of molecular mechanisms driving innate immune response to HCMV infections during gestation.

SNPs in TLR genes are associated with HCMV infections

Single nucleotide polymorphisms (SNPs) are common genetic alterations that impact on the expression levels of genes they are located in, hence their biological role. So far, polymorphisms from TLR genes were broadly investigated regarding their effect on the immune response against various pathogens including hepatitis C virus (HCV), Legionella pneumophila, Plasmodium falciparum, Mycobacterium leprae, Mycobacterium tuberculosis as well as HCMV (Texereau et al., 2005; Wang et al., 2011; Netea et al., 2012). Many studies also showed the involvement of different TLR SNPs in the course of inflammatory diseases and the altered expression of TLR-dependent immune response genes (Lazarus et al., 2003; Ahmad-Nejad et al., 2004; Kormann et al., 2009; Nahum et al., 2012).

In HCMV infections, the role of TLR SNPs was reported for TLR2, TLR3, TLR4 and TLR7 genes (Ducloux et al., 2005; Kijpittayarit et al., 2007; Kruger et al., 2010; Arav-Boger et al., 2012; Table 1). In a cohort of 92 liver transplant recipients, TLR2 2258 G>A SNP was associated with HCMV replication and disease (Kijpittayarit et al., 2007; Table 1). Renal transplant recipients (RTRs) with TLR3 F412L polymorphism (TT/TC allele) had a significantly increased frequency of acute rejection events (ARE) and TLR9 -1237 TT allele was associated with the occurrence of major adverse cardiovascular events (MACE; Kruger et al., 2010). However, in the present study none of the analyzed polymorphisms had a significant impact on the development of HCMV infection, except for a marginal association of TLR9 -1237 SNP with recurrent urinary infection (Kruger et al., 2010). In turn, in a cohort of RTRs, opportunistic infections and HCMV disease were more frequent in subjects with TLR4 1063 A>G and TLR4 1363 C>T SNPs, although these associations were marginal (Ducloux et al., 2005). In a study cohort of healthy women vaccinated with HCMV gB, an association between the occurrence of four SNPs in TLR7 (rs179008, rs179009, rs179013 and rs179018) and the level of antibodies to gB was observed (Arav-Boger et al., 2012). Homozygous carriers of the minor allele at four analyzed TLR7 SNPs had higher levels of vaccination-induced antibodies to gB compared with hetero- or homozygotes for common alleles (Arav-Boger et al., 2012). In a study stimulating patients' peripheral blood mononuclear cells (PBMCs) and fibroblasts with both HCMV and TLR3 ligand (Poly I:C), the cells carrying TLR3 L412F allele showed decreased IFN-γ and TNF-α expression (Nahum et al., 2012). The observed associations confirm results of an earlier study in which TLR3 was reported to be a sensor for MCMV infection (Tabeta et al., 2004). In other study, analyses of 20 SNPs located in TLR2, TLR3, TLR4 and TLR9 genes in a cohort of 336 recipients of hematopoietic cell transplants and their unrelated donors showed an influence of TLR4 haplotype S4 (1063 A/G and 1363 C/T SNPs in TLR4) on increased risk of invasive aspergillosis in donors (Bochud et al., 2008). The seropositivity for HCMV was another risk factor for invasive aspergillosis both in donors and recipients. In that study, TLR4 SNPs were suggested to be associated with other factors for risk of invasive aspergillosis such as acute graft-versus-host disease (GVHD) or HCMV seropositivity. However, donor S4 haplotype, acute GVHD and HCMV seropositivity were three independent risk factors (Bochud et al., 2008). So far, only a few studies have reported on the impact of TLR SNPs on the course of HCMV infection. Hence, we suggest that further studies on the role and genetic alterations of TLRs are necessary as they may represent a significant direction of research to determine mechanisms of cytomegaly pathogenesis.

SNPs in TLR genes associated with pregnancy disorders

In the case of TLR2, TLR4 and TLR9 genes, studies showed correlations of their SNPs with various pregnancy disorders along with preeclampsia (Xie et al., 2010), shorter gestational age (Lorenz et al., 2002; Hartel et al., 2004; Varner & Esplin, 2005; Krediet et al., 2007), premature rupture of membranes (PROM; Lukaszewski et al., 2009; Rey et al., 2008), mother-to-child transmission (MTCT) of Human immunodeficiency virus type 1 (HIV-1; Ricci et al., 2010). In women with preeclampsia the occurrence of TLR2 2258 G>A and TLR4 1063 A>G and TLR4 1363 C>T SNPs was correlated with early-onset disease (Xie et al., 2010). Significantly shorter gestational ages were characteristic for newborns with identified TLR2 -16934 TA/AA and TLR2 2258 GA/AA alleles (Krediet et al., 2007). In the case of TLR2 2258 G allele, the observed frequencies were similar in pregnant women with preterm labor and the control group (Lukaszewski et al., 2009). The analysis of 12 SNPs located in TLR1, TLR 2, TLR 4, TLR 6 and TLR 10 genes performed in cord blood samples from 72 atopic and 128 non-atopic mothers showed correlations of particular polymorphisms with the expression level of Treg marker genes and Th-cell cytokines (Liu et al., 2011). TLR2 rs1898830 GG genotype was associated with a lowered level of Treg marker genes in mothers with atopy and with an elevated level in mothers without atopy (Liu et al., 2011). Studies performed in Finnish mothers and their infants showed the correlation of TLR4 1063 A>G SNP with preterm labor (Lorenz et al., 2002). In a group of preterm infants from Uruguay the occurrence of TLR4 1063 A>G and IL-6, IL-1β and IL-12RB SNPs were not correlated with the disorder described (Pereyra et al., 2012). However, the frequency of analyzed TLR4 allele and observed association was suggested to be population-specific. In other study, TLR4 rs1554973 and rs7856729 SNPs interacted with IL-1R2 rs485127 SNP and were associated with the expression level of cervical cytokine IL-1β (Ryckman et al., 2011). The observed TLR4 and IL1-R2 genotypes were suggested as plausible markers of the levels of cervical cytokines IL-1α or IL-1β associated with the course of pregnancy (Ryckman et al., 2011). The identified correlations were not significant after correction for multiple testing in European and African-American populations (Ryckman et al., 2011). The outcomes are in agreement with other studies that reported various TLR alterations to be population-specific (Barreiro et al., 2009; Ryckman et al., 2009; Velez et al., 2010; Sawian et al., 2013). The occurrence of another TLR4 SNP, rs1554973, with chorionic plate inflammation was observed in mothers and their singleton fetuses (Simhan et al., 2008). In the case of TLR4 1363C>T SNP, the lower frequency of CT genotype and polymorphic T allele, suggested that the analyzed SNP protected against PROM in pregnant women (Lukaszewski et al., 2009). Different TLR9 SNPs were correlated with low birth weight (LBW) of infants, MTCT of HIV-1, higher risk of maternal anemia, the clinical picture of malaria in pregnancy, and cervical cancer (Mockenhaupt et al., 2006; Ricci et al., 2010; Roszak et al., 2012). In P. falciparum-infected Ghanaian pregnant women, the observed TLR9 -1486 T>C SNP was associated with an increased risk of LBW in term infants (Mockenhaupt et al., 2006). Another study showed the involvement of TLR9 c.4-44 G>A and c.1635 A>G SNPs in MTCT of HIV-1 (Ricci et al., 2010). The higher risk of MTCT of HIV-1 was observed with haplotypes G;G: c.4-44 G (rs352139) and c.1635 G (rs352140). The associations remained significant after adjustment for maternal viral load (Ricci et al., 2010). In the Polish population, TLR9 1635 A>G and rs187084 SNPs were correlated with the occurrence of cervical cancer and were suggested as a plausible risk factor for this cancer (Roszak et al., 2012). The associations of particular TLR SNPs with pregnancy disorders as well as with HCMV disease are summarized in Table 2.

Table 2. The effect of single nucleotide polymorphisms (SNPs) located in TLR genes on the occurrence and course of pregnancy disorders and HCMV infections
TLR SNPDiseaseNo. of patientsNo. of controlsP-valueReferences
  1. P-values in italics indicate statistically significant correlations. Although the P-values suggest significance, some studies used multiple comparisons and testing for different TLR SNPs without Bonferroni correction, and therefore their significance remains to be determined.

TLR1 rs4833095Alterations in expression of T regulatory (Treg) marker genes200 < 0.050 Liu et al. (2011)
TLR2 Arg753GlnHCMV disease in liver transplant recipients92 0.003 Kijpittayarit et al. (2007)
Arg753GlnPreeclampsia941760.094Xie et al. (2010)
Arg753Gln (G20877A)Premature rupture of membranes (PROM)3360NSLukaszewski et al. (2009)
Arg753GlnPreterm birth (PTB)3652811.000Hartel et al. (2004)
-16934 TA/AA and 753ArgGln/GlnGlnPTB305 < 0.020 Krediet et al. (2007)
rs4696480Alterations in expression of T helper (Th)-cells cytokine concentrations in maternal atopy200 < 0.050 Liu et al. (2011)
TLR3 L412FReduced IFN-γ and TNF-α expression resulted from stimulation with TLR3 ligand or HCMV14 Nahum et al. (2012)
TLR4 rs10759932Chorionic plate inflammation109 0.005 Simhan et al. (2008)
rs1554973    0.006  
Asp299Gly or/and Thr399IleHCMV disease in renal transplant recipients (RTRs)238 0.020 Ducloux et al. (2005)
Asp299GlyLow birth weight (LBW) of term infants ofwomen with placental malaria pigment1572 0.030 Mockenhaupt et al. (2006)
Asp299GlyMaternal anemia in women with placental malaria pigment24103 0.010  
Asp299GlyPreeclampsia941760.310Xie et al. (2010)
Thr399Ile 941760.140 
Thr399lle (C8993T)PROM3360NSLukaszewski et al. (2009)
896GPTB3652810.180Hartel et al. (2004)
Asp299GlyPTB282351 0.024 Lorenz et al. (2002)
Asp299Gly; genotypes: Asp/Gly or Gly/Gly 282351 0.028 Lorenz et al. (2002)
Asp299GlyPTB2262500.557–1.748Rey et al. (2008)
 Gestational age and PTB108; 118250 < 0.050  
 PROM2262500.890-1.000 
rs4986790PTB53560.229Pereyra et al. (2012)
rs1554973 and IL-1R2 rs485127Alterations in cervical cytokine concentrations < 0.001 Ryckman et al. (2011)
rs7856729 and IL-1R2 rs485127    < 0.001  
TLR7 rs179009gB-specific antibody responses after three doses of HCMV vaccine142 0.001 Arav-Boger et al. (2012)
rs179008    0.019  
rs179018    0.016  
rs179013    0.022  
TLR9 c.4-44 G > A (rs352139)Mother-to-child transmission (MTCT) of HIV-11181820.480Ricci et al. (2010)
c.1635 A > G (rs352140)   0.200 
Haplotype A;A: c.4-44 AA (rs352139) and c.1635 (rs352140)    0.016  
Haplotype G;G: c.4-44 G (rs352139) and c.1635 G (rs352140)    0.004  
T-1486CLBW of term infants of women with placental malaria pigment4147 0.010 Mockenhaupt et al. (2006)
TLR10 rs4129009Alterations in expression of Treg marker genes200 < 0.050 Liu et al. (2011)

Perspectives: towards description of mechanisms driving the involvement of TLRs in the immune response to HCMV infections

Many studies suggested the role of TLR receptors in modulation of the expression of miRNA, the important regulators of gene expression, and also the development, differentiation and apoptosis of immune cells (Baltimore et al., 2008; O'Neill et al., 2011). miRNA molecules influence the development of immune cells and hence indirectly regulate both the innate and the acquired immune response. Innate immunity is regulated, among others, by miR-223, miR-146 and miR-155 (Tili et al., 2007; O'Connell et al., 2008; Perry et al., 2008; Chen et al., 2012). LPS- or poly I:C-activated macrophages showed decreased miR-223 expression (Chen et al., 2012). The downregulation of miR-223 caused the activation of signal transducer and activator of transcription 3 (STAT3), and hence induced expression of IL-6 and IL-1β cytokines, but not TNF-α (Chen et al., 2012). Activation of TLR2, TLR4 and TLR5 by cell wall fragments of bacteria or fungi, or cell exposure to proinflammatory cytokines (IL-1β, TNF-α) leads to increased expression of miR-146a and b. Proteins associated with IL-1 receptors (IRAK1, IRAK2) and TNF (TRAF6) were proposed to be targets of these miRNAs. Thus, increased expression of miR-146 may lead to a decrease in expression of these molecules and, as a result of negative feedback, may inhibit the activation of TLRs. Increased expression of miR-146a was observed only with high concentrations of IL-1 when it was necessary to prevent excessive spread of acute inflammation (Perry et al., 2008). Hence, miR-146a plays a role in the development of tolerance to endotoxin. Exposure to LPS leads to a decrease in the sensitivity of immune cells and reduced secretion of proinflammatory cytokines by subsequent contact with it. Also, stimulation of TLR-2 or TLR-4 with their ligands, lipoprotein and lipopolysaccharide, increases the expression of miR-155 in monocytes and macrophages. The level of miR-155 in these cells increases after recognition of bacterial or viral nucleotides by TLR-3 and TLR-9. In contrast to miR-146a, miR-155 increases the secretion of inflammatory cytokines and susceptibility to septic shock (Tili et al., 2007). This molecule also stimulates proliferation of granulocytes and monocytes. Persistently elevated levels of miR-155 in bone marrow cells after stimulation with LPS can lead to expansion of granulocyte/monocyte cell types and the development of acute myeloid leukemia (O'Connell et al., 2008). Recent studies also showed the involvement of peptidoglycan (PGN)-TLR2 stimulation in miR132/-212-blocked expression of early components of MyD88-dependent pathways (Nahid et al., 2012). So far, studies performed to investigate the influence of HCMV infection on the expression levels of host miRNA molecules confirmed such a relationship with levels of various molecules, including miRNA cluster containing miR-96, miR-182, and miR-183, hsa-miR-92a, miR-17/92 cluster, miR-199a/214 cluster, and miR-132. However, the molecular mechanisms driving the impact of some cellular miRNA expression on the course of HCMV infection remain to be investigated. As both TLR and miRNA molecules are significant regulators of both innate and adaptive immune response as well as playing a role in the progression in HCMV infection, the function of TLRs in congenital infections with HCMV seems plausible. It will be important to assess the role and molecular mechanisms driving TLR function and alterations associated with this infection.

Among important molecular alterations, the most frequent epigenetic alteration that affects gene expression levels, and hence their biological function, is methylation of CpG islands of gene promoter regions. So far, only a few studies showed methylation status of CpG islands located in promoter regions of TLR genes. Significantly elevated levels of methylation of CpG islands in promoter regions of TLR4 and TLR2-1 genes and in the exon of TLR1 gene were observed in chickens sensitive to infections with Salmonella enteritidis compared with chickens that are resistant (De Oliveira et al., 2011; Gou et al., 2012). Although the altered 5′-cytosine methylation status, including hypermethylation of gene promoter regions and hypomethylation of CpG dinucleotides distributed in the genome, are the most common epigenetic changes resulting in the silencing of gene expression, the role of this modification in the determination of TLR expression levels has not been analyzed (Chen et al., 2012). Analyses of the methylation status of TLR genes are undoubtedly needed to describe precisely molecular mechanisms that might have an impact on the role of TLR genes, as well as the immune response to infections with HCMV.

Concluding remarks

The role of TLRs in the innate immunity to HCMV was broadly investigated in murine models and in vitro-cultured cell lines. These receptors are crucial molecules participating in the recognition of pathogen PAMPs such as unmethylated CpG sequences of viral DNA, the specific ligand for TLR9. The altered levels of TLR gene expressions observed in non-immune cells such as trophoblasts, decidual cells and amniotic epithelium dependent on cell type and stage of pregnancy, suggest the likely role of these genes in the immune response against microorganisms during pregnancy. Previous studies did not aim to describe the function of TLRs in non-specific immunity against HCMV in pregnancy or the participation of TLR alterations in HCMV disease. However, the modified TLR gene expression levels and molecular alterations including SNPs from gene sequences and changes in the methylation status of promoter regions seem to play a role in immune response to congenital infections with HCMV. We suggest that investigations including screening of SNPs from some TLR genes, as well as analyses of the methylation status of their promoter regions and TLR expression levels in accordance with primary infection with HCMV during pregnancy, the course of congenital cytomegaly and its clinical picture are fully justified. Recent studies also showed the contribution of other signaling pathways and expression levels of transcription factors and immunomodulatory cytokines in the immune response to HCMV infections, as observed in placental cell lines cultured in vitro and in neonatal DCs. Undoubtedly, further studies are needed to describe the molecular mechanisms and genetic background driving TLR participation in HCMV disease progression.

Acknowledgements

The authors declare that they have no conflict of interest.

Ancillary