OUP user menu

Mycoplasmal infections alter gene expression in cultured human prostatic and cervical epithelial cells

Shimin Zhang, Douglas J. Wear, Shyh-Ching Lo
DOI: http://dx.doi.org/10.1111/j.1574-695X.2000.tb01410.x 43-50 First published online: 1 January 2000

Abstract

To better understand how infections by mycoplasmas affect gene expression in human cells, we quantitatively measured the transcripts of 38 cytokine genes in HPV E6- and E7-immortalized cervical and prostatic epithelial cells before and after infection by four human urogenital mycoplasmas, M. fermentans, M. genitalium, M. hominis and M. penetrans. Using the multi-probe RNase protection assay (RPA), 22 and 23 cytokine gene transcripts were detected in the non-infected control prostatic and cervical epithelial cells, respectively. Although there were no discernible changes in cell morphology and growth kinetics following 72 h of mycoplasmal infection, 55–74% of the cytokine genes expressed in the two human epithelial cell lines were altered. Most changes reflected an increased expression of these cytokine genes, while expression of some cytokine genes significantly decreased. The effects varied with host cell type and species of infecting mycoplasmas. These alterations in gene expression were more profound in the cervical epithelial cells than in the prostatic cells. M. fermentans produced the most significant effects, followed by M. penetrans, M. genitalium and M. hominis. Some alterations in the gene expression were transient, but most persisted over the course of chronic (9 months) mycoplasmal infection. Prolonged gene expression changes induced by chronic mycoplasmal infection may gradually alter important biological properties in the infected mammalian cells and produce a unique form of disease process.

Keywords
  • Gene expression
  • Cytokine
  • Human epithelial cell
  • Mycoplasma

1 Introduction

Interacting with eukaryotic cells, bacteria activate many genes [13] and produce various products [4,5] that facilitate invasion into eukaryotic host cells, protect bacteria from elimination by the host cells, and/or have specific cytotoxic effects [68]. At the same time, the eukaryotic host cells modify their stress response to the bacterial infection and alter expression of a wide variety of genes [9]. The interplay between prokaryotic microbes and eukaryotic host cells is a new focus in recent studies [2,3]. Bacterial infection usually leads to rapid cell death in culture.

Wall-free mycoplasmas are the smallest organisms capable of self-replication [10]. These unusual microbes are among the few prokaryotes that can grow for extended periods of time in close interaction with mammalian cells without producing acute cytotoxic effects. Clinically, many patients are chronically infected or colonized by mycoplasmas in their urogenital or respiratory tracts without apparent illness [11,12]. Compared to bacteria, little is known about the effects of mycoplasmas on gene expression in eukaryotic host cells. Do mycoplasmal infections alter gene expression in mammalian host cells? If so, what are the long-term effects of extended abnormalities in the expression of certain genes on the eukaryotic cells?

We believe that chronic and persistent infection by mycoplasmas with seemingly low virulence may gradually lead to significant changes in many important biological properties of mammalian cells, thereby producing a different form of pathogenesis [13,14]. In a recent study, we showed that chronic infection by Mycoplasma fermentans and M. penetrans induces chromosomal changes and high level expression of H-ras and c-myc proto-oncogenes in C3H mouse embryo cells. Over-expression of both H-ras and c-myc in the C3H cells is associated with malignant cell transformation and tumorigenicity in animals [13,15].

To further study the mycoplasmal effects on gene expression in mammalian cells, we infected human cervical and prostatic epithelial cells in vitro with four different human urogenital mycoplasmas, M. fermentans, M. genitalium, M. hominis and M. penetrans. We then examined the changes in expression for 38 cytokine genes in cells infected by each mycoplasma at different time points. Expression levels of these cytokine genes were studied as representative markers. We were particularly interested in comparing the effects produced by different species of urogenital mycoplasmas as well as the effects found in cells infected by a specific mycoplasma for a short time (12–72 h) versus a prolonged (36 weeks) course.

2 Materials and methods

2.1 Mycoplasmal culture

M. genitalium was provided by Dr. Joseph G. Tully (NIAID, NIH). M. fermentans (PG 18, ATCC 19989) and M. hominis (PG 12, ATCC 23114) were obtained from the American Type Culture Collection (ATCC); M. penetrans (GTU-54) was isolated from a patient with AIDS in this laboratory [16]. The mycoplasmas were cultured in SP4 medium containing 18% fetal bovine serum, 100 U ml−1 of penicillin and 500 U ml−1 of polymyxin [16]. For mycoplasmal stock preparation, mycoplasmas were inoculated in 175-cm2 flasks, each with 100 ml of SP4 medium, and cultured in a 37°C incubator. Mycoplasmas in exponential growth judged by medium color change were aliquotted and stored at −70°C.

Quantification of mycoplasmas was done by measuring the color changing units (CCU) or colony forming units (CFU) for each species. The CCU assay is based on the color change of culture medium due to acid (M. genitalium, M. fermentans and M. penetrans) or alkaline (M. hominis) producing properties. Mycoplasmal cultures were serially diluted 10-fold in SP4 medium on 96-well tissue culture (TC) plates, then incubated at 37°C for 3 weeks for color change. The CFU assay was performed on SP4 agar plates. Fried egg-like colonies were counted 3 weeks after inoculating diluted mycoplasmal culture on the plates. All mycoplasmal cultures were quantified by at least one of the above methods. It is important to note, due to the adhesive properties of mycoplasmas, both assays could only provide an approximate estimation of the mycoplasma titer in the cultures.

2.2 Culture of human epithelial cells

Human prostatic and cervical epithelial cell lines used in this study had been immortalized with E6 and E7 genes of human papillomavirus type 18 (HPV 18). These cell lines were kindly provided by Dr. Michael R. Kuettel, Department of Radiation Medicine, Georgetown University Medical School. The cells were cultured in keratinocyte serum-free medium (SFM) (Gibco BRL) at 37°C in a 10% CO2 incubator. Prostatic cells in passages 35–39 and cervical cells in passages 37–41 were used for this study.

2.3 Infection with mycoplasmas

Infection of the immortalized human prostatic and cervical epithelial cells was done as follows unless specifically indicated in individual experiments. Confluent epithelial cells were split into 175-cm2 TC flasks, each with 107 cells in 30 ml keratinocyte-SFM, and grown overnight. The cells at 70% confluence were re-fed with the same volume of the medium the next morning, followed by the addition of 6 ml mycoplasmal stock. The same volume of SP4 medium was added into control flasks. For the time course experiments, the same amount of cells were placed in each flask at the same time. Mycoplasmas from the same batch of frozen stock were added into different flasks at different time points (0, 24, 48, 58 h). The infected cells were harvested for RNA preparation at the same time post infection (72 h). In this way, possible effects caused by different cell density and different concentration of medium supplements could be minimized. In order to observe effects of long-term mycoplasmal infection on gene expression in human epithelial cells, the epithelial cells were also infected with the four species of mycoplasmas for 36 weeks. The cells were seeded into new flasks once a week at a 1:10 dilution. Prior to harvesting the cells, mycoplasmas in the stock cultures and in conditioned media were quantified by the methods described above. Titers of mycoplasmas in the stocks used in this study were about 108 CCU for M. fermentans, M. hominis and M. penetrans, and about 1010 CCU for M. genitalium.

2.4 Morphological study of infected cells

The human epithelial cells were grown in 96-well TC plates to about 70% confluence and re-fed with 80 µl keratinocyte-SFM 2 h before infection. Twenty microliters of mycoplasma culture with the same titers as described above were added into each well (100 µl final volume). The same volume of SP4 medium was added into each control well. The cells were examined under a phase-contrast microscope before and during infection.

2.5 Study of cell growth kinetics

Cell growth kinetics was determined using the MTT [3-(4,5-dimethylthiazol-2-yl)2,5-diphenyltetrazolium bromide]-based cell proliferation assay kit (Boehringer Mannheim, Indianapolis, IN). Briefly, the human epithelial cells were placed in 96-well TC plates and infected with mycoplasmas in the same way as described in Section 2.4 using the same mycoplasmal stocks. At set time points, a tetrazolium salt MTT solution was added. Lysis buffer was added 4 h later. Absorbance at 540 nm was measured on an ELISA plate reader. In a separate experiment, the mycoplasmas were added into the cell cultures at different times, and the MTT solution was added to all cultures at a set time point. The remaining procedures were the same as described above.

2.6 RNA preparation and quantification

Total RNA was prepared from the cells using TRIzol reagent (Gibco BRL, Gaithersburg, MD) according to the manufacturer's instructions [17]. Briefly, medium was removed completely from the culture flasks, then 6 ml of TRIzol was added into each 175-cm2 flask to lyse the cells. After shaking for a few minutes, 2.5 ml of chloroform was added to denature proteins. The lysate was centrifuged at 12 000×g for 10 min at 4°C. Total RNA was precipitated from the supernatant with isopropanol followed by two ethanol washes. RNAs were dissolved in DEPC (diethyl pyrocarbonate)-treated H2O. The quantity and purity of the RNA was determined by measuring the absorbance at 260 nm and the 260/280 absorbance ratio on a spectrophotometer and by electrophoresis on 1.2% agarose gel in formaldehyde-MOPS (3-[N-morpholino]propanesulfonic acid) buffer. To prepare RNA from mycoplasmas, mycoplasmal culture were centrifuged at 14 000×g for 40 min at 4°C. The mycoplasmal pellet was suspended in a small amount of phosphate buffer saline (PBS), followed by lysis with TRIzol. Later steps are the same as described in the preparation of RNA from eukaryote cells.

2.7 RNase protection assay

Gene transcription (mRNA) levels of cytokines were analyzed by RNase protection assay (RPA) using the RiboQuant multi-probe kits (PharMingen, San Diego, CA) following instructions provided by the manufacturer. In brief, 32P-labeled specific antisense RNA probes were synthesized by in vitro transcription from multiple DNA template sets. An equal amount of total RNA (20 µl per sample) was hybridized overnight to the 32P-labeled RNA probes. The unhybridized free RNAs and RNA probes were digested with RNases A and T1. The protected mRNAs and probes were precipitated with ethanol and resolved on a 6% denaturing polyacrylamide gel [18]. The gel was dried on a gel drier and then exposed to X-ray film as well as a phosphoimaging screen. The expression of mRNA transcript of a specific gene was quantified on a Storm 860 phosphoimaging scanner (Molecular Dynamics, Sunnyvale, CA).

Five multi-probe template sets were used in this study. Each includes eight or nine cytokine genes and two housekeeping genes, L32 and GAPDH (glyceraldehyde 3-phosphate dehydrogenase), for a total of 40 genes. The cytokines included in individual template sets are listed in Table 1.

View this table:
Table 1

Cytokine genes studied

Template setNumber of templatesCytokine genesHousekeeping genes
hCK-111IFN-γ, IL-2, IL-4, IL-5, IL-9, IL-10, IL-13, IL-14, IL-15L32, GAPDH
hCK-210IFN-γ, IL-1α, IL-1β, IL-1Ra, IL-6, IL-10, IL-12p35, IL-12p40L32, GAPDH
hCK-310IFN-β, IFN-γ, LTβ, TNF-α, INF-β, TGFβ1, TGFβ2, TGFβ3L32, GAPDH
hCK-411G-CSF, GM-CSF, IL-3, IL-6, IL-7, LIF, M-CSF, OSM, SCFL32, GAPDH
hCK-510I-309, IL-8, IP-10, Ltn, MCP-1, MIP-1α, MIP-1β, RANTESL32, GAPDH
  • GAPDH: glyceraldehyde 3-phosphate dehydrogenase; G-CSF: granulocyte colony stimulating factor; GM-CSF: granulocyte/macrophage colony stimulating factor; IFN: interferon; IL: interleukin; IP: induced protein; LIF: leukemia inhibitory factor; LT: lymphotoxin; Ltn: lymphotactin; MCP: monocyte chemoattractant protein; M-CSF: macrophage colony stimulating factor; MIP: macrophage inflammatory protein; OSM: oncostatin M; SCF: stem cell factor; TGF: transforming growth factor; TNF: tumor necrosis factor.

3 Results

3.1 Different expression of cytokine genes in human prostatic and cervical epithelial cells induced by mycoplasmas

Cultures of HPV E6- and E7-immortalized human prostatic and cervical epithelial cells were infected with M. fermentans, M. genitalium, M. hominis and M. penetrans for 24 h as described in Section 2. Total RNAs were prepared from mycoplasma-infected cells and non-infected control cells. mRNA transcription of 38 specific cytokine genes was determined by the RNase protection assay (RPA) using five multi-probe sets, hCK1, hCK2, hCK3, hCK4 and hCK5 (Table 1). Although there is no significant difference in growth rate and morphology in this period of mycoplasmal infection (data not shown), the cytokine gene expression profiles of the mycoplasma-infected human epithelial cells and non-infected control cells were different. As an example, different amounts of cytokine transcripts were identified in the human epithelial cells using hCK4 (Fig. 1A) and hCK3 (Fig. 1B) multi-probe sets. Compared to the non-infected prostatic control cells, the M. fermentans-infected prostatic epithelial cells markedly increased expression of M-CSF and GM-CSF by a factor of 6.2 and 3.8, respectively (Fig. 1A). M. penetrans infection also increased M-CSF transcript more than two-fold in the same cell type. In comparison, M. hominis or M. genitalium infection did not significantly change expression of M-CSF and GM-CSF in these prostatic cells. The non-infected cervical control cells produced a large amount of TGFβ1 and TGFβ2 but only small amounts of TGFβ3, TNF-α and LTβ (Fig. 1B). The mycoplasma-infected cervical cells also expressed a large amount of TGFβ1 and TGFβ2. However, infections with all four urogenital mycoplasmas tested up-regulated the expression of LTβ and TNF-α genes by more than 10-fold in the cervical epithelial cells (Fig. 1B).

Figure 1

Effects of mycoplasma infection on expression of cytokine genes in immortalized human prostatic and cervical epithelial cells. Total RNAs from non-infected control cells (C), cells infected by M. fermentans (F), M. genitalium (G), M. hominis (H) and M. penetrans (P) for 24 h were studied for gene expression changes by multi-probe RNase protection assay. A: RNAs were prepared from non-infected and mycoplasma-infected human prostatic cells. Multi-probe set hCK4 (including cytokines G- and GM-CSF, IL-3, -6 and -7, LIF, M-CSF, SCF and OSM) was used in the RPA. B: RNAs were prepared from non-infected and mycoplasma-infected human cervical epithelial cells. The hCK3 set (including IFN-β, LTβ, TGFβ1, -β2 and -β3, TNF-α and -β, and -γ) was used. The undigested probes serving as molecular mass markers are indicated on the left. Protected specific mRNAs are marked on the right. IL-6-like* is the band that hybridized with IL-6 probe, but its molecular mass is larger than the expected IL-6 band. This was confirmed using a single IL-6 probe instead of a multi-probe set.

Table 2 summarizes the changes in expression of the 38 cytokine genes examined by the RPA using all five different multi-probe sets in both of the immortalized human urogenital epithelial cells infected with each of the four mycoplasmas for 24 h. The non-infected control prostatic epithelial cells expressed 22 out of 38 cytokine genes tested (GM-CSF, IFN-γ, IL-1α, -1β and -1Ra, IL-6, -7, -8, -12p35 and -15, IP-10, LIF, LTβ, MCP-1, M-CSF, OSM, RANTES, SCF, TGFβ1, -β2 and -β3, TNF-α). The non-infected control cervical epithelial cells also expressed all the above cytokine genes. In addition, the cervical cells expressed G-CSF gene. Although non-infected prostatic cells did not express G-CSF, infection with any one of the mycoplasma species tested activated the G-CSF gene expression in prostatic epithelial cells. Among the genes expressed, M. fermentans infection changed the expression of 73% of the cytokine genes (16 out of 22; 14 increased and two decreased) in the prostatic cells and increased the expression of 74% of the genes (17 out of 23) in the cervical cells. M. genitalium infection changed the expression of 68% of the cytokine genes (15 out of 22; 12 increased, three decreased) in the prostatic cells and 65% (15 out of 23; 13 increased, two decreased) in the cervical cells. M. hominis infection changed the expression of 55% of the cytokine genes (12 out of 22; 11 increased, one decreased) in the prostatic cells and 65% (15 out of 23; 13 increased, two decreased) in the cervical cells. M. penetrans infection changed the expression of 68% (15 out of 22; 14 increased, one decreased) and 74% of the cytokine genes (17 out of 23; 15 increased, two decreased) in the infected prostatic and cervical epithelial cells respectively (see details in Table 2).

View this table:
Table 2

Change of gene expression in human prostatic and cervical epithelial cells after infection with mycoplasmas

Mycoplasmal infectionGene expressionProstatic epithelial cellsCervical epithelial cells
None (control cells)Genes expressedGM-CSF, IFN-γ, IL-1α, -1β, -1Ra, -6, -7, -8, -12p35 and -15, IP-10, LIF, LTβ, MCP-1, M-CSF, OSM, RANTES, SCF, TGFβ1, β2 and β3, TNF-α22G-CSF, GM-CSF, IFN-γ, IL-1α, -1β, -1Ra, -6, -7, -8, -12p35 and -15, IP-10, LIF, LTβ, MCP-1, M-CSF, OSM, RANTES, SCF, TGFβ1, β2 and β3, TNF-α23
M. fermentansIncreasedG- and GM-CSF, IL-1α, -1β, -1Ra, -8 and -15, IP-10, LTβ, MCP-1, M-CSF, RANTES, TGFβ2, TNF-α14G- and GM-CSF, IL-1α, -1β,-1Ra, -7, -8 and -15, IP-10, LTβ, MCP-1, M-CSF, RANTES, TGFβ1, -β2 and -β3, TNF-α17
DecreasedTGFβ1 and -β320
No changeIFN-γ, IL-6, -7 and -12p35, LIF, OSM, SCF7IFN-γ, IL-6 and -12p35, LIF, OSM, SCF6
M. genitaliumIncreasedG-CSF, IL-1α, -1β, -8 and -15, IP-10, LTβ, RANTES, SCF, TGFβ1 and -β3, TNF-α12GM-CSF, IL-1α, -1β, -8 and -15, IP-10, LTβ, MCP-1, M-CSF, RANTES, SCF, TGF-β1, TNF-α13
DecreasedGM-CSF, IL-1Ra, M-CSF3G-CSF, IL-1Ra2
No changeIFN-γ, IL-6, -7 and -12p35, LIF, MCP-1, OSM, TGFβ28IFN-γ, IL-6, -7 and -12p35, LIF, OSM, TGFβ2 and -β38
M. hominisIncreasedG- and GM-CSF, IL-1α, -1β, -1Ra and -8, IP-10, LTβ, MCP-1, RANTES, TNF-α11G- and GM-CSF, IL-1Ra, and -8, IP-10, LTβ, MCP-1, M-CSF, RANTES, TGFβ1, -β2 and -β3, TNF-α13
DecreasedTGFβ21IL-1α and -1β2
No changeIFN-γ, IL-6, -7, -15 and -12p35, LIF, M-CSF, OSM, SCF, TGFβ1 and -β311IFN-γ, IL-6, -7, -15 and -12p35, LIF, OSM, SCF8
M. penetransIncreasedG- and GM-CSF, IL-1α, -1β, -1Ra, -8 and -15, IP-10, LTβ, MCP-1, M-CSF, RANTES, TGFβ1, TNF-α14G- and GM-CSF, IL-1α, -1β, -1Ra, -7, -8 and -15, IP-10, LTβ, MCP-1, M-CSF, RANTES, TGFβ1, TNF-α15
DecreasedTGFβ21TGFβ2 and -β32
No changeIFN-γ, IL-6, -7 and -12p35, LIF, OSM, SCF, TGFβ38IFN-γ, IL-6 and -12p35, LIF, OSM, SCF6
  • Thirty-eight cytokine genes were studied in non-infected control cells and the mycoplasma-infected human epithelial cells. The housekeeping genes L32 and GAPDH were used as loading controls. Gene expression was compared between non-infected control cells and cells infected with each human urogenital mycoplasma for 24 h by the RPA. The results listed in this table are a summary of 3–5 parallel experiments for each RPA template set, cell line and mycoplasma species. All results were normalized with the housekeeping genes L32 and GAPDH. Increased (or decreased) expression of a cytokine gene is defined as more than 50% increase (or decrease) of mRNA following the mycoplasmal infection. Cytokines in bold represent the differences in the expression between the two human epithelial cells.

In summary, mycoplasmal infections more severely affected cytokine gene expression in cervical epithelial cells than in prostatic epithelial cells in aspects of total number of genes affected and the extent of gene expression changes. M. fermentans and M. penetrans altered the gene expression more profoundly than did the other two mycoplasmas in both infected cell types.

RNAs from M. genitalium, M. fermentans, M. penetrans and M. hominis cultured in SP4 broth and yeast tRNA were hybridized with the multi-probe sets of RPA as controls. No single protected band was detected on the RPA polyacrylamide gels. Therefore, the mycoplasmas did not produce any RNA able to hybridize with the specific probes used in this study.

3.2 Kinetics of changes in the expression of various cytokine genes in human epithelial cells infected by mycoplasmas

Specific cytokine mRNAs from both non-infected cells and cells infected with one of the four species of mycoplasmas for 14, 24, 48 and 72 h were quantified by the RPA. Fig. 2, as an example, shows the expression of G-CSF, GM-CSF, IL-6, IL-7, LIF, M-CSF and SCF genes detected by hCK4 probes in human prostatic and cervical epithelial cells following 14–72 h of infection by M. fermentans. mRNA of M-CSF markedly increased in both prostatic and cervical epithelial cells infected by the mycoplasma. The increase was particularly rapid and dramatic in the cervical cells following the infection. Expression of both GM-CSF and G-CSF also increased significantly in these two epithelial cell lines 14 and 24 h after infection with M. fermentans. However, they returned to the level close to that found in the non-infected control cells 72 h after mycoplasmal infections. Expression of OSM (data not shown) genes showed little change during 72 h of infection by M. fermentans after the data were normalized with both background and housekeeping genes.

Figure 2

Gene expression kinetics of immortalized human prostatic and cervical epithelial cells infected by M. fermentans for 14, 24, 48, 72 h and 36 weeks as described in Section 2. RNAs were prepared from infected and non-infected control cells. The induction of gene expression in the mycoplasma-infected cells was measured by the multi-probe RNase protection assay using the hCK4 probe set.

Four different kinetic patterns were found in the changes of cytokine gene expression in the human epithelial cells within 72 h of infection by the urogenital mycoplasmas. (1) The expression of certain genes continued to increase following mycoplasmal infections, such as M-CSF expression in both prostatic and cervical epithelial cells when infected by M. fermentans (Fig. 2), and in the cervical cells infected by M. hominis (Fig. 3). (2) Gene expression decreased following mycoplasmal infections, such as IL-1Ra expression in M. genitalium-infected prostatic epithelial cells (Fig. 3). (3) Expression initially increased following mycoplasmal infection, but returned to the pre-infection level by 72 h, such as IL-1β expression in M. penetrans-infected prostatic epithelial cells (Fig. 3), or the expression of GM-CSF and G-CSF in M. fermentans-infected prostatic and cervical epithelial cells (Fig. 2). (4) Gene expression remained constant in the 3-day period of mycoplasmal infection, such as the expression of SCF and OSM genes in both prostatic and cervical epithelial cells infected by M. fermentans, and the expression of TGFβ2 in the cervical epithelial cells infected by M. genitalium infection (Fig. 3). In spite of these changes, no significant morphological change was found in either cell line during 72 h of mycoplasmal infection.

Figure 3

Different patterns of alterations in gene expression induced by mycoplasma infections in the human prostatic and cervical epithelial cells. The epithelial cells grown in 175-cm2 flasks had been infected by M. genitalium, M. hominis or M. penetrans, for 14, 24, 48 and 72 h. Cells were harvested at the same time for RNA preparation. The mRNA induction time course for TGF-β2, M-CSF, IL-1Ra and IL-1β genes was studied by the RPA. The graphic curves on the right indicate cytokine gene expression levels normalized to the expressed housekeeping gene L32 in the cells.

3.3 Changes of cytokine gene expressions in cells chronically infected by mycoplasmas

To examine if the changes in expression of cytokine genes persisted in chronically infected epithelial cells, we measured mRNA levels of these genes in the cells chronically infected for 36 weeks by M. fermentans, M. genitalium, M. hominis and M. penetrans. Most of the changes in gene expression found in the epithelial cells infected by the mycoplasmas for 72 h persisted in the chronically infected cells (Fig. 2). However, about 20% of the cytokine genes whose expression increased in the early stage of mycoplasmal infections returned to the original levels of expression found in the non-infected control cells when the cells were chronically infected by the urogenital mycoplasmas. Interestingly, the expression GM-CSF and G-CSF that dramatically increased initially, then rapidly returned to the pre-infection low level after 48–72 h, was highly active in cells chronically infected (36 weeks) by the mycoplasmas (Fig. 2).

4 Discussion

We used the current culture model to study mycoplasmal effects on gene expression in infected mammalian host cells because it more closely resembles the clinical situation. Contrary to studies of cells and mycoplasmas that do not normally interact in clinical conditions, our in vitro model examined the interaction between urogenital epithelial cells and mycoplasmas commonly found in the human urogenital tract. Although the prostatic and cervical epithelial cell lines used in our study have been immortalized with HPV E6 and E7 genes, these cells, differing from other transformed cell lines, behave like normal epithelial cells and depend on epithelial cell growth factor (EGF) for their continued growth. Moreover, HPV infection of the human urogenital tract is a common clinical condition. Studies have long suggested that HPV plays an important, but not sufficient role in causing malignant transformation of cervical epithelial cells [19]. Studying mycoplasmal effects on HPV-immortalized epithelial cells could provide additional insights into the effect of two common infections on human host cells.

In this study, the expression of 38 cytokine genes in human cervical and prostatic epithelial cells was examined as a reporting system. It was somewhat surprising to find that both prostatic and cervical epithelial cells were actively expressing many cytokine genes previously known to be produced mainly in macrophages, hematopoietic or mesenchymal cells [20]. Previous studies have revealed that mycoplasmas and mycoplasmal membrane components can induce expression of several cytokine genes in peripheral blood monocytes, macrophages, amnion cells and in spleen and lungs [2124]. This study demonstrates that urogenital mycoplasmal infections induce significant changes in the expression of many cytokine genes in these cultured human prostatic and cervical epithelial cells. Within 12–24 h, messenger transcripts of certain genes increased more than 10–12-fold. Although these changes could also be due to a selective decrease in messenger degradation, the rapid increase of the messengers was most likely due to a rapid activation of gene transcription.

The same species of mycoplasma could elicit different gene expression responses in different human epithelial cells. M. genitalium increased expression of IL-1α in prostatic cells, but decreased the expression of IL-1α in cervical cells. Depending on the species, mycoplasmal infection could have different or opposite effects on the expression of a particular gene in the same human cells. For example, M. fermentans increased the expression of TGFβ2 in the cervical epithelial cells, whereas M. penetrans decreased TGFβ2 expression in the same cell type. In comparison, there was no significant change in the expression of TGFβ2 in the cervical epithelial cells following infection by M. genitalium. Furthermore, the expression of different genes in the same cells infected by the same mycoplasma could have totally different patterns. In M. fermentans-infected prostatic epithelial cells, the expression level of M-CSF gradually increased with time of infection while the LIF and SCF levels remained relatively unchanged (Fig. 2). When the cervical epithelial cells were infected with M. genitalium, the expression of M-CSF increased 12-fold, whereas the expression of G-CSF decreased 11.6-fold.

Thus, our study concluded: (1) even with no apparent changes in cell growth or cell morphology, mycoplasmal infections rapidly altered expression of many genes in human epithelial cells; (2) different species of mycoplasmas might have completely opposite effects on the expression of a particular gene in a human epithelial cell; (3) the gene expression response is different for each human epithelial cell line after mycoplasmal infection; (4) alterations of gene expression in epithelial cells may be transient, but many changes persisted with chronic mycoplasmal infections.

Clearly, mycoplasmal infections could affect the expression of a large group of different genes in these human epithelial cells in addition to these cytokine genes. Activation or inactivation of these cytokine genes alone can initiate a cascade of reactions in the two epithelial cell types. Many of these cytokine genes' products effectively affect the expression of their downstream target genes [25]. Thus, mycoplasmal infection, of seemingly low virulence, would surely alter many characteristics and biological properties of mammalian host cells. At present, it would be extremely complex to study alteration of expression for thousands of genes, at different time points, during mycoplasmal infections. However, with the rapid development of DNA chip technology, studying the expression of thousands of different genes and their dynamic changes over the course of infection by different mycoplasmas will soon become possible.

It is important to note that live mycoplasmas rapidly decreased in all of the serum-free cell cultures once they were inoculated. The extent of cytokine gene expression alterations in the infected cells did not correlate with the titers of mycoplasmas found in the cultures but did correlate with the species of mycoplasmas infecting the cultures. For example, the titer of M. fermentans was consistently found to be the lowest (101–103 ml−1) among the four species of mycoplasmas tested. However, M. fermentans always had the most significant effect on the alteration of gene expression in both infected prostatic and cervical cell lines.

The true biological significance of the chronic or persistent nature of mycoplasmal infections in human respiratory and urogenital tracts is more complicated than the expression levels of various cytokine genes (or many other genes) in the epithelial cells. Besides epithelial cells, many active players are present, including a heterogeneous spectrum of mesenchymal cells, inflammatory cells and even other microbial flora. These players closely interact and constantly influence one another. We have previously proposed that prolonged alteration of gene expression in the mammalian host cells during chronic and persistent mycoplasmal infections may be associated with various forms of disease processes including malignant cell transformation [13]. Our present study provides additional support for this hypothesis and warrants further investigation.

Acknowledgements

The authors thank Drs. Kaori Iyama, Ivan Ding, and Binxue Zhang for their valuable scientific contributions, Ms. Susan Ditty for her critical review of the manuscript and Mr. Joe Rodriguez for photographic assistance. This study was supported in part by The American Registry of Pathology, Grant 2228.

References

  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
View Abstract