Ureaplasma urealyticum Modulates Endotoxin-Induced Cytokine Release by Human Monocytes Derived from Preterm and Term Newborns and Adults

Abstract

We previously observed that Ureaplasma urealyticum respiratory tract colonization in infants with a birth weight of ≤1,250 g was associated with increases in the tracheal aspirate proinflammatory cytokines tumor necrosis factor alpha (TNF-α) and interleukin-8 (IL-8) relative to the counterregulatory cytokine IL-6 during the first week of life (A. M. Patterson, V. Taciak, J. Lovchik, R. E. Fox, A. B. Campbell, and R. M. Viscardi, Pediatr. Infect. Dis. J. 17:321–328, 1998). We hypothesized that U. urealyticum alters the host immune response in the presence of a coinflammatory stimulus (e.g., bacterial infection or hyperoxia) by shifting the balance of cytokine expression towards the proinflammatory cytokines. To test this hypothesis, we compared the release of TNF-α, IL-8, IL-6, and IL-10 in vitro by unstimulated and U. urealyticum (with or without lipopolysaccharide [LPS])-stimulated human monocytes from adult peripheral blood and from term and preterm cord blood. U. urealyticum alone and in combination with LPS induced concentration- and development-dependent changes in cytokine release. In vitro inoculation with low-inoculum U. urealyticum (10³ color-changing units [CCU]) (i) partially blocked the LPS-stimulated IL-6 release by all cells and reduced LPS-stimulated IL-10 release by preterm cells, (ii) stimulated TNF-α and IL-8 release by preterm cells, and (iii) augmented LPS-stimulated TNF-α release in all cells. In preterm cells, high-inoculum U. urealyticum (10⁶ CCU) (i) stimulated TNF-α and IL-8, but not IL-6 or IL-10, release and (ii) augmented LPS-stimulated TNF-α and IL-8 release. High-inoculum U. urealyticum (i) stimulated release of all four cytokines in term cells and IL-8 release in adult cells and (ii) augmented LPS-induced TNF-α, IL-10, and IL-8 release in term cells but did not significantly affect LPS-induced cytokine release in adult cells. We speculate that U. urealyticum enhances the proinflammatory response to a second infection by blocking expression of counterregulatory cytokines (IL-6 and IL-10), predisposing the preterm infant to prolonged and dysregulated inflammation, lung injury, and impaired clearance of secondary infections.

Bronchopulmonary dysplasia (BPD) is the chronic phase of neonatal lung injury, characterized by delayed alveolarization, chronic inflammation, and fibrosis that occur in susceptible very-low-birth-weight preterm infants (7, 27, 43). Although barotrauma and oxidant injury are important factors in the pathogenesis of BPD (27), there is accumulating epidemiological and biologic evidence that intrauterine and postnatal infections contribute to the initiation and exacerbation of lung inflammation and injury (1, 6, 18, 19, 71). It is now important to determine the roles of specific organisms and the mechanisms of injury involved in the development of BPD.

Ureaplasma urealyticum is the most common organism isolated from infected amniotic fluid and placentas (17, 33, 75) and from the respiratory tracts of preterm infants (1). Respiratory tract colonization of the preterm infant with U. urealyticum has been associated with higher incidences of pneumonia (10, 45), severe respiratory failure and death (44), and BPD (1, 23, 29, 69). Experimental colonization of preterm baboons with U. urealyticum caused an acute bronchiolitis (68). U. urealyticum caused an acute interstitial pneumonia in newborn but not 2-week-old mice, suggesting that there is an age-dependent susceptibility to pulmonary infection with this organism (50). Exposure to 80% oxygen increased the inflammatory response, decreased clearance of U. urealyticum from the lung, and increased mortality in experimental infection in newborn mice (8). Animals with natural or experimentally induced infections with related mycoplasmas are predisposed to secondary bacterial pneumonia with decreased bacterial clearance and higher morbidity and mortality (57). Taken together, these findings suggest that U. urealyticum can cause significant disease in the preterm infant and alter the inflammatory response to secondary injury (hyperoxia or infection). The predisposing host factors that contribute to the age-dependent susceptibility to Ureaplasma-induced lung disease have not been elucidated.

Comparisons of tracheal aspirate cytokine concentrations from infants with self-limited respiratory distress syndrome (RDS) and those who developed BPD suggest that the acute-phase cytokines tumor necrosis factor alpha (TNF-α), interleukin-1β (IL-1β), IL-8, and IL-6 are elevated in infants who develop BPD (2, 22, 35, 36, 42, 48). In contrast, IL-10, a potent anti-inflammatory cytokine, was undetectable in lung lavages of intubated preterm infants with RDS, suggesting that the susceptibility of the immature lung to injury or chronic inflammation is due, in part, to a decreased capacity to down-regulate the inflammatory response (28). The role of cytokines generated in the lung in mediating lung injury is not clearly understood. In particular, IL-6 is generally considered to be a proinflammatory cytokine because it stimulates antigen-dependent immune responses (34) and increases expression of acute-phase proteins (24). However, in the lung, it limits inflammation induced by pneumonia (66, 73) and hyperoxia (70), in part through inhibition of TNF-α and IL-1β expression.

Several lines of evidence now strongly suggest that infection-induced intrauterine-derived cytokines may initiate lung injury in utero. Increased concentrations of IL-6 and IL-1β were detected in tracheal aspirates of preterm infants on the first day of life, associated with prolonged rupture of the membranes (20) and histologic chorioamnionitis (38, 71), respectively. In a related study, infants who were colonized with U. urealyticum or other bacteria had elevated tracheal aspirate IL-1β concentrations on day 1 compared with concentrations in noncolonized infants (21). We previously observed that U. urealyticum respiratory tract colonization in infants with birth weights of ≤1,250 g was associated with increases in the tracheal aspirate proinflammatory cytokines TNF-α, IL-1β, and IL-8 relative to the counterregulatory cytokine IL-6 in the first week of life (46). Taken together, these studies suggest that in some infants, lung inflammation may be initiated prenatally by microbial colonization of the amniotic fluid or fetal membranes and/or airways. Imbalances of proinflammatory and counterregulatory cytokines in the lungs of Ureaplasma-colonized infants may contribute to prolonged pulmonary inflammation and alter the immune response to a secondary inflammatory stimulus.

We hypothesized that U. urealyticum alters the host immune response in the presence of a coinflammatory stimulus by shifting the balance of cytokine expression towards the proinflammatory cytokines, predisposing the preterm infant to dysregulated inflammation and collateral tissue injury. To determine if U. urealyticum modulates cytokine release by monocytes in response to bacterial endotoxin, we compared the release of TNF-α and IL-8 versus IL-6 and IL-10 in vitro by unstimulated and U. urealyticum (with or without lipopolysaccharide [LPS])-stimulated cultured human monocytes from adult peripheral blood and term and preterm cord blood.

MATERIALS AND METHODS

Monocyte isolation.

According to a protocol approved by the Institutional Review Board of the University of Maryland School of Medicine, venous blood was collected in heparinized syringes from umbilical cords of preterm (24 to 32 weeks of gestation; n = 14) and term (n = 16) newborns and from healthy adult volunteers (n = 12). Newborns were excluded if (i) clinical chorioamnionitis was present or (ii) congenital infection was highly suspected or confirmed. Preterm but not term infants were screened for respiratory colonization with U. urealyticum according to the protocol of the neonatal intensive-care unit. All colonized infants were excluded. Mononuclear cells were isolated by Ficoll-Hypaque gradient centrifugation as previously described (4, 14). Mononuclear cells were suspended in CRPMI (RPMI 1640 medium [Gibco BRL, Rockland, N.J.] with 2 mM l-glutamine, 1 mM sodium pyruvate, 10 mM HEPES, 100 U of penicillin per ml, and 100 μg of streptomycin per ml) with 2.5% fetal calf serum (FCS) (HyClone, Logan, Utah) and seeded in a 48-well plate at densities of 1.5 × 10⁶ cells/well for preterm cells and 1 × 10⁶ cells/well for term and adult cells. Cells were incubated for 90 min at 37°C. Nonadherent cells were then removed by washing with phosphate-buffered saline. CRPMI with 10% FCS was added to each well, and the cells were incubated overnight at 37°C in a 5% CO₂–95% air atmosphere.

U. urealyticum.

U. urealyticum serotype 3, the most common clinically isolated serotype (77), was obtained from the American Type Culture Collection (Manassas, Va.), grown to late log phase in 10B medium (56), aliquoted as 10⁷ color-changing units (CCU) (49) per ml, and frozen at −70°C. Serial 10-fold dilutions were made in 10B broth for cell culture inoculation.

Experimental cell culture.

Culture medium was removed and replaced with 0.65 ml of fresh CRPMI–10%FCS without antibiotics. Sterile 10B broth (0.1 ml) was added to control wells. Sample wells were inoculated with 0.1 ml of 10B broth containing U. urealyticum for final concentrations of 10³ to 10⁶ CCU/well. LPS (Escherichia coli 055:B5; Sigma, St. Louis, Mo.) was added to half of the control and sample wells at 100 ng/ml. In preliminary LPS dose-response experiments (10 ng/ml to 1 μg/ml), this dose produced near-maximal cytokine release from newborn and adult cells (data not shown). Triplicate wells for each condition were prepared per experiment. Supernatants were collected at 2, 8, and 24 h after inoculation and stored at −70°C until analysis by enzyme-linked immunosorbent assay. Adherent cells were collected by scraping and counted with a hemacytometer. Cell purity was assessed by alpha naphthyl esterase staining (63), and viability was determined by trypan blue exclusion.

Cytokine measurements.

Human TNF-α, IL-6, IL-10, and IL-8 concentrations were measured in the UMAB Cytokine Core Laboratory using standard two-antibody enzyme-linked immunosorbent assays with commercial antibody pairs and recombinant standards (TNF-α, IL-6, and IL-10 from Endogen, Boston, Mass.; IL-8 from Biosource, Camarillo, Calif.) as previously described (26). A curve was fit to the standards using a computer program (Softpro; Molecular Devices), and cytokine concentrations from each sample were calculated from the standard curve. Samples were analyzed in duplicate. The lower limits of detection were 3.9, 6.25, 1.17, and 3.9 pg/ml for TNF-α, IL-6, IL-10, and IL-8, respectively.

Statistical analysis.

Cytokine data are expressed as the mean ± standard error of the mean (SEM) of picograms per 10⁵ cells or as the percentage of the LPS-positive control value for each experiment to minimize the interexperiment variability. Differences among groups were tested by a Fisher protected least-squares difference test applied to a one-way analysis of variance. A P value of <0.05 was considered significant.

RESULTS

Cell culture characteristics.

The cultured cells from each developmental stage were ≥90% monocytes as determined by alpha naphthyl esterase staining. Plating efficiencies were similar among the groups (19.1% ± 4.3% for preterm, 16.3% ± 3.1% for term, and 15.8% ± 2.9% for adult cells). Plating efficiency and cell viability (≥95%) were unaffected by treatment with LPS or U. urealyticum. After 24 h of incubation, titers of U. urealyticum were reduced, but viable organisms remained.

Basal cytokine secretion.

Isolated monocytes were initially cultured overnight prior to adding U. urealyticum or LPS, to minimize the effects of possible cell activation during monocyte isolation and, for the newborn cells, of intrauterine events. Basal secretions of TNF-α and IL-6 corrected for the number of adherent cells were similar among unstimulated preterm, full-term, and adult monocytes (Table 1). Unstimulated preterm monocytes constitutively released 81 and 94.5% less IL-10 than term and adult monocytes, respectively (P = 0.034). There was a trend towards lower IL-8 release by unstimulated preterm cells than by term or adult cells (P = 0.24).

TABLE 1.

Basal cytokine release and response to LPS by preterm, term, and adult monocytes

Cytokine	Preterm monocytes (n = 14)		Term monocytes (n = 16)		Adult monocytes (n = 12)
	Basal releasea	LPS-stimulated releaseb	Basal release	LPS-stimulated release	Basal release	LPS-stimulated release
TNF-α	26.4 ± 9.2	3.5 (2.7–7.6)	22.6 ± 8.2	8.2 (4.4–28.4)	26.2 ± 11.8	22.7 (13.7–46.6)
IL-6	87.5 ± 26.7	13.7 (2.7–22.3)	96.1 ± 37.7	13.7 (2.75–22.3)	13.2 ± 5.6	192.8 (11.6–518.2)
IL-10	1.26 ± 0.71	16.5 (4.25–196)	6.68 ± 2.81c	29.7 (9.4–36.8)	22.8 ± 11.9c	224 (71.3–339.2)
IL-8	1,413 ± 473	4.97 (2.5–6.5)	3,732 ± 1,060	4.4 (2.05–9.1)	5,075 ± 2,185	6.3 (3.6–15.4)

^aData are expressed as picograms per 10⁵ cells (mean ± SEM). 

^bFold stimulation by LPS (100 ng/ml) compared to medium control, expressed as the median with the 25th to 75th percentiles in parentheses. 

^cP < 0.05 versus preterm cells. 

Response to endotoxin.

LPS alone stimulated marked increases in release of all measured cytokines from neonatal and adult cells, confirming that neonatal cells are capable of responding appropriately to this agonist (Table 1). There was a trend towards less LPS-induced stimulation of TNF-α, IL-6, and IL-10 in preterm cells than in term and adult cells (Table 1).

Dose-response experiments.

To determine whether in vitro inoculation with U. urealyticum has dose-dependent effects on monocyte cytokine release, an initial series of experiments were conducted in which cells were incubated with medium alone or with LPS (100 ng/ml) (positive control) with or without a range of inocula of live U. urealyticum (10³ to 10⁶ CCU/well) for 24 h. This range of inocula is within the range of U. urealyticum cultured from respiratory secretions of preterm infants. The lower numbers of monocytes isolated from preterm cord blood were sufficient only for comparisons of low (10³ CCU) and high (10⁶ CCU) inocula of U. urealyticum with and without LPS. In preterm cells, a high inoculum (10⁶ CCU) of U. urealyticum alone stimulated release of TNF-α and IL-8 comparable to levels induced by LPS but failed to stimulate IL-6 or IL-10 release (Fig. 1). In contrast, the high inoculum alone stimulated all four cytokines in term cells (Fig. 2). In both preterm and term cells, high-inoculum U. urealyticum increased the LPS-induced release of TNF-α (269% ± 64% for preterm and 157% ± 17% for term) and IL-8 (5,004% ± 2,151% for preterm and 140% ± 19% for term) (mean ± SEM, percentage of LPS positive control value; P < 0.05 for each case). In preterm cells, in vitro inoculation with high-inoculum U. urealyticum did not alter the LPS-stimulated release of IL-6 or IL-10. In term cells, high-inoculum U. urealyticum augmented the LPS-stimulated release of IL-10 but not IL-6 (Fig. 2). In contrast with the neonatal cells, in adult cells, high-inoculum U. urealyticum alone stimulated IL-8 release comparable to that with LPS and did not affect LPS-induced cytokine release (Fig. 3). In adult cells, U. urealyticum at 10⁵ CCU, but not 10⁶ CCU, increased LPS-stimulated TNF-α and IL-10 release 1.5- and 1.3-fold, respectively (Fig. 3A and C), suggesting that the optimal dose of U. urealyticum for synergy with LPS varies among the developmental stages.

FIG. 1.

Dose-dependent effects of in vitro infection with U. urealyticum on preterm monocyte cytokine release. Monocytes from preterm cord blood (24 to 32 weeks) were incubated with CRPMI–10% FCS alone, U. urealyticum at 10³ CCU (UU3) (low inoculum) or 10⁶ CCU (UU6) (high inoculum), or LPS (100 ng/ml) with or without U. urealyticum for 24 h. (A) TNF-α; (B) IL-6; (C) IL-10; (D) IL-8. Results are expressed as percentages of the LPS positive control value (mean ± SEM; n = 6). ∗, P < 0.05 versus medium control; †, P < 0.05 versus LPS.

FIG. 2.

Dose-dependent effects of in vitro infection with U. urealyticum on term monocyte cytokine release. Monocytes from term cord blood were incubated with CRPMI–10% FCS alone, U. urealyticum at 10³ CCU (UU3) to 10⁶ CCU (UU6), or LPS (100 ng/ml) with or without U. urealyticum for 24 h. (A) TNF-α; (B) IL-6; (C) IL-10; (D) IL-8. Results are expressed as percentages of the LPS positive control value (mean ± SEM; n = 4). ∗, P < 0.05 versus medium control; †, P < 0.05 versus LPS.

FIG. 3.

Dose-dependent effects of in vitro infection with U. urealyticum on adult monocyte cytokine release. Monocytes from adult blood were incubated with CRPMI–10% FCS alone, U. urealyticum at 10³ CCU (UU3) to 10⁶ CCU (UU6), or LPS (100 ng/ml) with or without U. urealyticum for 24 h. (A) TNF-α; (B) IL-6; (C) IL-10; (D) IL-8. Results are expressed as percentages of the LPS positive control value (mean ± SEM; n = 6). ∗, P < 0.05 versus medium control; †, P < 0.05 versus LPS.

Differential effects of low-inoculum U. urealyticum on monocyte cytokine release.

In initial dose-response experiments, in vitro inoculation with low-inoculum U. urealyticum (10³ CCU) appeared to inhibit the LPS-induced IL-6 release by all cells and IL-10 release by preterm cells. Additional experiments in which monocytes were incubated with low-inoculum U. urealyticum in the presence or absence of LPS for 24 h confirmed these observations (Fig. 4). In preterm cells, U. urealyticum (10³ CCU) reduced both LPS-stimulated IL-6 release and LPS-stimulated IL-10 release (by 52 and 48%, respectively) (Fig. 4B and C). In contrast, in term and adult cells, low-inoculum U. urealyticum reduced only LPS-stimulated IL-6 release (by 56 and 35%, respectively) but did not affect LPS-stimulated IL-10 secretion. Low-inoculum U. urealyticum alone stimulated TNF-α release in preterm monocytes 2.8-fold compared to cells incubated with medium alone and augmented LPS-induced TNF-α release in preterm (262% ± 58%), term (156% ± 24%), and adult (185% ± 35%) cells (percentage of LPS positive control value, mean ± SEM) (Fig. 4A). Low-inoculum U. urealyticum alone stimulated IL-8 release in preterm cells comparable to levels induced by LPS but did not stimulate IL-8 expression in term and adult cells and did not alter LPS-induced IL-8 expression in cells from any developmental stage (Fig. 4D).

FIG. 4.

Comparison of cytokine release by preterm, term, and adult monocytes infected in vitro with low-inoculum U. urealyticum. Monocytes from preterm and term cord blood and adult peripheral blood were incubated with CRPMI–10% FCS medium alone, U. urealyticum (UU) at 10³ CCU (low inoculum), or LPS (100 ng/ml) with or without U. urealyticum for 24 h. (A) TNF-α; (B) IL-6; (C) IL-10; (D) IL-8. Results are expressed as percentages of the LPS positive control value (mean ± SEM; preterm n = 8, term n = 12, and adult n = 6). ∗, P < 0.05 versus LPS.

Time course.

To determine the time required after in vitro infection with U. urealyticum to alter LPS-stimulated cytokine release, preterm monocytes were incubated with LPS with and without U. urealyticum (10³ CCU) for 2 to 24 h. LPS alone stimulated a rapid TNF-α response, with peak release within 2 h (Fig. 5A). U. urealyticum (10³ CCU) alone stimulated TNF-α release similar to levels induced by LPS at 24 h and augmented LPS-induced TNF-α release at 8 and 24 h. LPS-stimulated IL-6 release was detectable at 2 h and continued to increase throughout the 24-h incubation period (Fig. 5B). LPS-stimulated IL-10 release peaked at 8 h (Fig. 5C). U. urealyticum (10³ CCU) reduced the LPS-stimulated IL-6 and IL-10 release at all time points.

FIG. 5.

Time course of effects of LPS and U. urealyticum alone and in combination on preterm (30 weeks) monocyte cytokine release. Preterm monocytes were incubated with medium, U. urealyticum (UU) at 10³ CCU, or LPS with or without U. urealyticum for 2 to 24 h. (A) TNF-α; (B) IL-6; (C) IL-10. Results are from a representative experiment of three experiments.

DISCUSSION

It has been suggested that preterm infants have an increased susceptibility to lung injury and infection in part because their monocytes have a decreased capacity to produce the cytokines required for up-regulation of the inflammatory response (53). Previous studies have reported conflicting findings concerning the responsiveness of LPS-stimulated neonatal monocytes. LPS-stimulated cord blood monocytes from term infants produced amounts of IL-6 similar to (41, 51, 74) or less than (47) those produced by adult monocytes. LPS-stimulated monocytes from preterm cord blood produced significantly less IL-6 (41, 51) than stimulated monocytes from term infants or adults. In contrast, Kavelaars et al. (31) reported similar capacities to produce IL-6 from 26 to 41 weeks gestation. Newborn cells have a reduced capacity to produce other cytokines, including TNF-α (13, 47, 52, 72). Monocytes from term infants stimulated with LPS produce less IL-10 than stimulated adult cells (5, 37). LPS-stimulated preterm monocytes produce less IL-10 than stimulated term or adult cells (39). Although in the present study, there was a trend towards less TNF-α, IL-6, and IL-10 release in response to LPS by preterm cells than by term or adult cells, all cells responded robustly to LPS alone. Individual variability, effects of exposure to intrauterine events (51), and differences in culture conditions and LPS dose and incubation time might contribute to differences in study results. We attempted to minimize the possible effects of in utero cell activation by incubating freshly isolated monocytes in medium alone overnight prior to stimulation with LPS or U. urealyticum.

In vitro inoculation with U. urealyticum elicited dose- and development-dependent effects on cytokine release by cultured monocytes. In the present study, in vitro inoculation with relatively high doses of U. urealyticum was required to stimulate TNF-α and IL-8 in preterm monocytes, all measured cytokines in term cells, and IL-8 in adult cells. It would be desirable to complement these monocyte studies with similar studies of alveolar macrophages, the major source of cytokines in the lung (32). However, limited numbers of alveolar macrophages can be recovered from preterm lung lavage on day 1 of life, and obtaining alveolar macrophages from term infants would be limited by the infrequency of intubation of healthy term infants. Moreover, since most newly recruited alveolar macrophages are blood monocyte derived (25, 58, 76), cultured monocytes are a reasonable alternative for study. The present findings are consistent with results of previous studies using other monocyte/macrophage cell models. U. urealyticum at ≥10⁵ CCU stimulated TNF-α release by RAW 264.7, a murine macrophage cell line (9, 61). Heat-killed U. urealyticum at 10⁸ CCU stimulated TNF-α and IL-6 release by THP-1 (a human monocytic cell line), Nr 8383 (a rat alveolar macrophage cell line), and human lung macrophages isolated from tracheal aspirates from preterm infants (40). In cultured human neonatal fibroblasts, U. urealyticum at 10⁴ CFU stimulated IL-6 and IL-8 release (59), suggesting that effects of U. urealyticum may be cell type specific. These previous studies did not evaluate the response to low-inoculum U. urealyticum or the interaction with LPS.

Mycoplasma species successfully colonize the urogenital and respiratory tract mucosas by evading host defenses. Known mechanisms include inhibition of ciliary motility and neutrophil and macrophage phagocytosis and antigen size variation (57, 77). IL-6 is important in up-regulating antigen-dependent defenses, including synthesis of immunoglobulin A (IgA) by mucosal B cells and synthesis of IgG by differentiated B cells (34). In the present study, low-inoculum U. urealyticum alone stimulated release of the proinflammatory cytokines TNF-α and IL-8 in preterm monocytes but did not affect IL-6 release by preterm cells or any cytokine release by term or adult cells. Failure to stimulate IL-6 might impair generation of Ureaplasma-specific lymphocyte responses and allow persistent U. urealyticum colonization of the respiratory tracts of preterm infants and continued expression of the TNF-α- and IL-8-induced inflammatory cascade.

Previous studies have observed that natural and experimentally acquired infection with other mycoplasma species in animals increases the susceptibility to secondary bacterial pathogens (57), but the mechanisms for the interaction are poorly understood. We used a model of in vitro U. urealyticum inoculation and Escherichia coli LPS to study how U. urealyticum infection might modify the cytokine response to coincident stimulation with a second bacterial pathogen. In the present study, low-inoculum U. urealyticum partially inhibited LPS-stimulated IL-10 release by preterm cells but not that by term or adult cells. This development-dependent regulation of IL-10 by U. urealyticum might contribute to the gestational age-dependent susceptibility of the respiratory tract to Ureaplasma colonization or infection and prolonged inflammation in affected infants. IL-10, a potent and broadly acting anti-inflammatory cytokine, suppresses the innate immune response in part by down-regulating TNF-α and IL-1 (15, 16, 67) and IL-8 (12) expression. It has been shown to reduce the inflammatory response and improve survival in models of endotoxemia (11, 60), bacterial peritonitis (30), and immune complex-induced lung injury (55). IL-10 was undetectable in tracheal aspirates of preterm infants with RDS, but the relationship with Ureaplasma or other bacterial respiratory tract colonization was not examined (28). A reduced capacity to produce IL-10 in the injured preterm lung combined with further suppression of IL-10 by U. urealyticum might contribute to the prolonged pulmonary inflammation that leads to BPD.

In vitro inoculation of monocytes with low-dose U. urealyticum augmented LPS-stimulated TNF-α release but partially blocked LPS-stimulated IL-6 release in all cells. Although IL-6 is an early activator of the local antigen-specific immune response, several lines of evidence suggest that IL-6 also exerts important anti-inflammatory effects in the lung, principally through down-regulating proinflammatory cytokines. Direct intratracheal instillation of IL-6 blocks subsequent lung injury induced by LPS (65) or IgG immune complexes (54) by inhibiting TNF synthesis and TNF-induced cytotoxicity (3, 54, 64). IL-6 induces IL-1 receptor antagonist and soluble TNF receptor 55, which are natural antagonists for IL-1β and TNF, respectively (62). Studies of pneumonia- and LPS-induced lung injury in IL-6-deficient mice confirmed that IL-6 is required for limiting the inflammatory response, bacterial clearance, and survival (66, 73). Transgenic mice overexpressing IL-6 had greater survival than control mice during exposure to hyperoxia, confirming the importance of IL-6 in hyperoxic acute lung injury (70). By partially blocking the IL-6 response to LPS, U. urealyticum might neutralize an important counterregulatory mechanism, leading to an augmented and persistent inflammatory response to a second infection or injury.