Interleukin-6 (IL-6) is a proinflammatory cytokine produced by several tissues as a reaction to stimulation by a number of factors, including bacterial membrane lipopolysaccharide, viruses, and even other cytokines.1 In early gestation, the IL-6 released by trophoblastic cells in response to growth factors induces those same cells to express integrins, thus promoting adequate trophoblast growth.2 Furthermore, IL-6 dose-dependently stimulates human chorionic gonadotropin (hCG) secretion,3 while a decrease of the cytokine level in the maternal plasma is correlated with a higher incidence of first-trimester abortion.4
However, as the pregnancy progresses, IL-6 appears to play an opposite role. Its amniotic fluid concentration is increased in cases of spontaneous abortion secondary to second-trimester amniocentesis.5 In addition, cytokine plays an important role in the pathogenesis of preterm delivery by stimulating prostaglandin (PG) release.
The importance of a well balanced regulation of prostaglandin E2 (PGE2) and IL-6 levels for a successful pregnancy raises the question of whether it is possible to pharmacologically control their release. In this regard, it is worth noting that IL-6 is able to increase PG output from amnion cells6 and that PGs, in turn, can affect IL-6 release.7–9 Moreover, it has been reported that IL-10 is able to decrease both PGE2 10 and IL-611 release from fetal membranes.
Based on the close connection between the cytokine and the prostanoid, one can hypothesize that factors able to decrease PGE2 output could also impair IL-6. Because it has been demonstrated that some classes of antibiotics, of which ampicillin is the most effective, significantly inhibit PGE2 release from amnion tissue,12,13 the decision was made to test the effect of ampicillin on amniotic IL-6 release both in vitro and in vivo. The in vitro experiments were performed on amnion-like Wistar Institute Susan Hayflick cells, a suitable model for the study of amniotic prostanoid metabolism. In these experiments, along with IL-6, PGE2 release upon ampicillin treatment was also tested. The in vivo study consisted of assaying IL-6 amniotic levels before and after ampicillin administration at the 17th week of gestation.
MATERIALS AND METHODS
[5,6(n)-3H]PGE2 (181 Ci/mmol) was purchased from Amersham (Milan, Italy). Interleukin-6, PGE2, anti-PGE-bovine serum albumin serum, and ampicillin were from the Sigma Chemical Co (St Louis, MO). All tissue culture media and sera were purchased from Invitrogen (San Diego, CA). All other chemicals were the highest reagent grades commercially available.
Amnion-like Wistar Institute Susan Hayflick cells were obtained from the American Type Culture Collection (ATCC CCL-25; Manassas, VA) and maintained in the laboratory. Cells were grown at 37°C in an atmosphere of 5% CO2/95% air, in a mixture of Ham’s F12 and Dulbecco’s modified Eagle medium (1:1 vol/vol) supplemented with 10% fetal bovine serum (10%), 30 μg/mL gentamicin, and 0.25 μg/mL amphotericin B.
The cells were seeded in 24-well plates at 2 × 105 cells/well in Ham’s F12 and Dulbecco’s modified Eagle medium + 10% fetal bovine serum and grown to confluence (2–3 days).
For PGE2 and IL-6 determination, the medium was removed and replaced with fresh serum-free Ham’s F12 and Dulbecco’s modified Eagle medium containing ampicillin. Each point of both time-courses and dose-response relationships was performed in duplicate and obtained from 3 or 4 different cell culture plates.
After incubation of samples for the indicated time, the media were collected and stored at −80°C until PGE2 radioimmunoassay was performed.
Cell culture and PGE2 assay were performed at the Laboratory of Cellular Physiology of the Department of Biology of Ferrara University.
The study population consisted of a total of 212 consecutive white patients submitted to genetic amniocentesis during the 17th week of their singleton physiological pregnancy, as revealed by uncomplicated, either spontaneous or elective, cesarean delivery of an healthy fetus at term. Indication for cesarean birth was represented in all the cases by cesarean delivery in the previous pregnancy (12%). The sample was collected between December 2000 and December 2001 at the Section of Obstetrics and Gynaecology of Ferrara University. At this clinic, as part of a protocol aimed at decreasing the incidence of abortion secondary to the procedure, all patients to undergo amniotic fluid sampling are treated with an oral dose of 1 g ampicillin twice daily for 4 days. All the patients gave informed, verbal consent to the trial’s protocol. Approval for this study was obtained from the University of Ferrara institutional review board. The 212 cases were subdivided into the following 3 groups: 1) 92 amniotic fluids sampled before ampicillin administration (untreated group); 2) 70 amniotic fluids sampled 4 hours after administration of an oral dose of 1 g ampicillin (A group); and 3) 50 amniotic fluids sampled 12 hours after administration of 1 g oral ampicillin (B group).
Average maternal age was 36.2 years. The main indication for the diagnostic procedure was represented by advanced maternal age (66.5%); other indications were a positive triple test (21.4%), the presence of echographic markers of increased risk of fetal aneuploidy (10.3%), maternal anxiety (1%), and a family history of Down syndrome or other genetic disorders (0.8%).
Within 6 hours of withdrawal, all the samples were centrifuged at 800g for 10 minutes at 4°C to remove particulate materials and then stored at −80°C.
The assay procedure was performed at the Laboratory of Nuclear Medicine of Ferrara University.
Interleukin-6 was determined in duplicate on each collected medium by commercial enzyme-linked immunosorbent assay (ELISA; Bender Medsystem Diagnostic GmbH, Vienna, Austria) according to the manufacturer’s instructions.
The specific monoclonal antibody in this kit was able to detect IL-6 in cell culture supernatants, human serum, plasma, and body fluids in general. In addition, to ensure correct quantitative determination of the cytokine, all the samples were diluted 1:6 in assay buffer. A reference curve was obtained by the standard IL-6 concentrate (range 1.6–100 pg/mL) (1 ng/mL) adjusted to the International Reference Standard (NBSB 88/514).
Optical density (OD) values were obtained with an automated microplate reader (Model 550; Bio-Rad, Richmond, CA) at 450 nm (620 nm as optional reference wavelength). A linear regression was established between optical density (Y axis) and the corresponding standard concentration (X axis); IL-6 sample concentrations were calculated by interpolating from the reference curve and multiplying by the dilution factor. The results were expressed as picograms per milliliter for the amniotic samples and as picograms per 106 cells for the Wistar Institute Susan Hayflick cells experiments. Assay sensitivity was 1.4 pg/mL.
The amount of PGE2 was assayed in duplicate on each collected medium by a radioimmunoassay procedure, as previously reported.14 A specific antiserum for PGE1 and PGE2 (cross reactions 100% and 165%, respectively) was used. Label [3H] PGE2 (3 nCi) was added to each tube. The level of PGE2 in each serially diluted sample was determined by comparison with a standard displacement curve (15–250 pg/0.1 mL). Incubation was performed for 90 minutes at 4°C. Bound and free radioligands were separated by dextran-coated charcoal, followed by centrifugation of tubes for 10 minutes at 2,000g. Assay sensitivity was 15 pg/tube, and the intra-assay or interassay coefficients of variation were less than 10%. Data were expressed as nanograms of PGE2 produced per 106 cells.
Cell culture and PGE2 assay were performed at the Laboratory of Cellular Physiology of the Department of Biology of Ferrara University.
IL-6 levels measured in amniotic fluid were normalized by using logarithmic transformation.
Kolmogorov-Smirnov goodness of fit test was computed on logarithmic transformation to assess whether the transformed variable was normally distributed (Kolmogorov-Smirnov d = 0.0665; P < .20).
Thus, to assess what effect both antibiotic administration and the amount of time elapsed between antibiotic treatment and amniocentesis have on IL-6 levels in amniotic fluid, one-way ANOVA was applied, followed by the Tukey honestly significant difference post hoc comparison test for unequal sample size. Descriptive statistics in Table 1 refer to the variables before logarithmic transformation. The significance levels for the rejection of the null hypothesis were set at P < .05. These statistical analyses were carried out by using Statistica Release 5.5 software (StatSoft Inc, Tulsa, OK).
As for IL-6 and PGE2 levels measured in Wistar Institute Susan Hayflick cells, statistical analyses were performed with GraphPad Prism software (San Diego, CA). One-way repeated measures analysis of variance (ANOVA), with pair-wise multiple comparisons with Bonferroni correction as post hoc tests, was used for time-course experiments. One-way ANOVA followed by the Dunnet multiple comparison test was used for dose-response relationships of ampicillin.
The time course of IL-6 production by Wistar Institute Susan Hayflick cells (Figure 1) showed a basal cytokine level that increased in time. The addition of 10−6 M ampicillin induced a significant and maximal inhibition after 4 hours (−50%; P < .05), although for longer incubation times this inhibitory effect was no longer observed. For this reason, subsequent dose-response relationships were tested at 4 hours (Figure 2). Concentrations of ampicillin, ranging from 10−7 to 10−4 M, decreased IL-6 levels; the drug effect was already statistically significant (−30%; P < .05) at the lowest concentration tested (10−7 M) and reached the maximum (−50%) at 10−6 M. The higher doses did not further inhibit IL-6 output.
The time course of PGE2 production by Wistar Institute Susan Hayflick cells, incubated in the absence or in the presence of 10−6 M ampicillin, is shown in Figure 3. The basal level of PGE2 increased with time. The addition of 10−6 M ampicillin induced a statistically significant inhibition after 30 minutes (−24%; P < .05) and maximal inhibition (−36%; P < .05) after 4 hours; with longer incubation times, the inhibitory effect disappeared. Therefore, we tested dose-response effects (Figure 4) at 30 minutes (A) and at 4 hours (B). At 30 minutes of incubation, concentrations of ampicillin, ranging from 10−7 to 10−4 M, decreased PGE2 release with significant effect (−26%; P < .05) only at 10−6 M. At 4 hours, the inhibition was significant and reached the maximum effect at 10−6 M (−40%; P < .05).
Table 1 shows the IL-6 levels measured in amniotic fluids along with the descriptive statistics. One-way ANOVA showed highly significant differences in IL-6 levels between the 3 groups (F = 18.04; P < .001). Post hoc comparisons revealed that IL-6 levels in amniotic fluids obtained from women treated with ampicillin 4 hours before amniocentesis were strongly and significantly reduced compared with the levels measured in women who were not treated or were treated 12 hours before the procedure (P < .001 for both comparisons).
No significant difference was observed between IL-6 amniotic levels obtained from women not treated with ampicillin and those obtained from women treated 12 hours before amniocentesis (P = .866).
IL-6 appears to be provided with a dual effect on pregnancy. Indeed, its level was reported to be decreased in the blood of women with spontaneous abortion,4 thus suggesting a protective role in early pregnancy. However, starting from the advanced second trimester, the cytokine appears to work against the physiological progression of pregnancy. Indeed, a large body of evidence indicates that an increase in the amniotic level of IL-6, and of other markers, is the expression of an inflammatory condition that leads to abortion5 or preterm delivery.15 A further aspect of great clinical value is the recent demonstration by Romero et al16 and Gomez et al17 that a fetal IL-6 plasma level greater than 11 pg/mL is the expression of a systemic fetal inflammatory condition that occurs before the clinical onset of premature parturition. In addition, an increased level of the cytokine constitutes an independent risk factor for severe neonatal morbidity. For instance, IL-6, together with other cytokines, is a marker of inflammatory conditions leading to cerebral palsy.18 As for the pathogenic mechanism of Romero’s “Fetal Inflammatory Syndrome,” most cases are related to some clinical or pathological evidence of infection, whereas about 13% of them appear to be the consequence of a different unknown process. At least in some of the noninfected cases, a genetic basis can be hypothesized because it has been demonstrated that, in the presence of fetal chromosomal abnormalities, there is a cytokine imbalance characterized by a decrease in amniotic IL-8 and an increase in IL-6 compared with euploidy.19
On the basis of the above considerations, it is logical to think that, from the second trimester onward, a decrease in the amniotic level of IL-6 could play a protective role in preventing the cytokine-evoked prostaglandin release that leads to abortion, premature delivery, and perinatal morbidity. Several drugs are able to reduce IL-6 output,20–22 but a clinical study on their efficacy in the prevention of adverse events during pregnancy has never been performed.
Our data show that ampicillin has a clear inhibitory effect on IL-6 amniotic release, both in vitro and in vivo. As for the in vitro action, we found that, in amnion-like Wistar Institute Susan Hayflick cells, the maximal effect of the drug is reached at 10−6 M, 4 hours after treatment. Furthermore, we found that, in the same cells, ampicillin significantly inhibits PGE2 release, reaching the maximal effect at the same optimal dose and incubation time as for cytokine inhibition. This suggests a possible correlation between the PGE2 and IL-6 biosynthetic pathways. This correlation is further supported by the report that indomethacin is able to decrease the blood level of IL-6 by inhibiting PGE2 synthesis.20 On the contrary, along with IL-1β and TNFα, the anti-inflammatory drug ibuprofen increases IL-6 levels in the mononuclear cells of preterm newborns.23 Such a dual effect of PGE2 on IL-6 release can be explained if one hypothesizes specific functions for each of the 4 prostanoid receptors. Indeed, it has been reported that, in mast cells, an increase in IL-6 is mediated by the interaction of PGE2 with its EP3 receptor. On the contrary, lack of the EP4 receptor in knockout mice lowers the level of circulating IL-6, as well as its macrophages and liver cell release.8,9
The present study also shows, for the first time, that ampicillin is able to decrease in vivo IL-6 amniotic fluid levels by a mechanism independent from its antibacterial property. Indeed, we found that an oral dose of 1 g of the drug is able to significantly reduce the amniotic cytokine level 4 hours after administration, a response fitting with the time course revealed by the in vitro study. The clinical implication of our finding deals with the management of IL-6–related inflammatory complications during pregnancy. Indeed, ampicillin, which is widely used in the treatment of bacterial-mediated complications, also appears to be indicated in nonbacterial inflammatory conditions. Based on our experience, we believe that its use is the main tool in our preventive arsenal for reducing spontaneous abortion secondary to second-trimester amniocentesis. In fact, in our 3,500 amniocenteses, all performed by the same operator, the total pregnancy loss was 0.3% (unpublished data), a percentage lower than that in the general population between the 17th and the 28th week of gestation. In our opinion, the only explanation for such a finding is that inflammatory conditions marked by increased IL-6 and PGE2 activity are effectively controlled by ampicillin through a mechanism independent of its antibacterial properties. Among other antibiotic candidates that lower the cytokine level, penicillin has been reported to inhibit IL-6 release from decidual cells,22 and erythromycin is known to reduce the Streptococcus pneumoniae–induced cytokine production in human whole blood.21 However, it is logical to believe that the efficacy of the latter is consequent to its antibacterial property, because it is unable to directly inhibit amniotic PGE2 release.13
In conclusion, given ampicillin’s ability to decrease amniotic IL-6 and PGE2 release, its use should be considered in the management of inflammatory complications, such as chorioamnionitis, premature labor, and premature rupture of the membranes mediated by the cytokine and prostanoid interaction.
1. Akira S, Taga T, Kishimoto T. Interleukin-6 in biology and medicine. Adv Immunol 1993;54:1–78.
2. Das C, Kumar VS, Gupta S, Kumar S. Network of cytokines, integrins and hormones in human trophoblast cells. J Reprod Immunol 2002;53:257–68.
3. Carp H, Torchinsky A, Fein A, Toder V. Hormones, cytokines and fetal anomalies in habitual abortion. Gynecol Endocrinol 2001;15:472–83.
4. Koumantaki Y, Matalliotakis I, Sifakis S, Kyriakou D, Neonaki M, Goymenou A, et al. Detection of interleukin-6, interleukin-8, and interleukin-11 in plasma from women with spontaneous abortion. Eur J Obstet Gynecol Reprod Biol 2001;98:66–71.
5. Wenstrom KD, Andrews WW, Tamura T, DuBard MB, Johnston KE, Hemstreet GP. Elevated amniotic fluid interleukin-6 levels at genetic amniocentesis predict subsequent pregnancy loss. Am J Obstet Gynecol 1996;175:830–3.
6. Mitchell MD, Dudley DJ, Edwin SS, Schiller SL. Interleukin-6 stimulates prostaglandin production by human amnion and decidual cells. Eur J Pharmacol 1991;192:189–91.
7. Hatakeyama D, Kozawa O, Otsuka T, Shibata T, Uematsu T. Zinc suppresses IL-6 synthesis by prostaglandin F2α
in osteoblasts: inhibition of phospholipase C and phospholipase D. J Cell Biochem 2002;85:621–8.
8. Nguyen M, Solle M, Audoly LP, Tilley SL, Stock JL, McNeish JD, et al. Receptors and signaling mechanisms required for prostaglandin E2
-mediated regulation of mast cell degranulation and IL-6 production. J Immunol 2002;169:4586–93.
9. McCoy JM, Wicks JR, Audoly LP. The role of prostaglandin E2
receptors in the pathogenesis of rheumatoid arthritis. J Clin Invest 2002;110:651–8.
10. Brown NL, Alvi SA, Elder MG, Bennet PR, Sullivan HF. The regulation of prostaglandin output from term intact fetal membranes by anti-inflammatory cytokines. Immunology 2000;99:124–33.
11. Fortunato SJ, Menon RP, Swan KF, Menon R. Inflammatory cytokine (interleukins 1, 6 and 8 and tumor necrosis factor-alpha) release from cultured human fetal membranes in response to endotoxic lipopolysaccharide mirrors amniotic fluid concentrations. Am J Obstet Gynecol 1996;174:1855–62.
12. Vesce F, Buzzi M, Ferretti ME, Pavan B, Bianciotto A, Jorizzo G, Biondi C. Inhibition of amniotic prostaglandin E release by ampicillin. Am J Obstet Gynecol 1998;178:759–64.
13. Vesce F, Pavan B, Buzzi M, Pareschi MC, Bianciotto A, Iorizzo G, et al. Effect of different classes of antibiotics on amniotic prostaglandin E release. Prostaglandins Other Lipid Mediat 1999;57:207–18.
14. Pavan B, Buzzi M, Ginanni Corradini F, Ferretti ME, Vesce F, Biondi C. Influence of oxytocin on prostaglandin E2
, intracellular calcium, and cyclic adenosine monophosphate in human amnion-derived (WISH) cells. Am J Obstet Gynecol 2000;183:76–82.
15. Harbon S, Tanfin Z, Khae LD, Gourcan O, Leiber D. Receptors and signal transduction in the myometrium. In: Chwalsz K, Garfield RE, editors. Basic mechanisms controlling term and preterm birth. Berlin: Springer Verlag; 1994. p. 29–54.
16. Romero R, Gomez R, Grezzi F, Yoon BH, Mazor M, Edwin SS, et al. A fetal systemic inflammatory response is followed by the spontaneous onset of preterm parturition. Am J Obstet Gynecol 1998;179:186–93.
17. Gomez R, Romero R, Grezzi F, Yoon BH, Mazor M, Berry SM. The fetal inflammatory response syndrome. Am J Obstet Gynecol 1998;179:194–202.
18. Nelson KB, Willoughby RE. Infection, inflammation and the risk of cerebral palsy. Curr Opin Neurol 2000;13:133–9.
19. Vesce F, Scapoli C, Giovannini G, Tralli L, Gotti G, Valerio A, et al. Cytokine imbalance in pregnancies with fetal chromosomal abnormalities. Hum Reprod 2002;17:803–8.
20. Bour AM, Westendorp RG, Laterveer JC, Bollen EL, Remarque EJ. Interaction of indomethacin with cytokine production in whole blood. Potential mechanism for brain-protective effect. Exp Gerontol 2000;35:1017–24.
21. Guchelaar HJ, Schultz MJ, Van der Poll T, Koopmans RP. Pharmacokinetic-pharmacodynamic modelling of the inhibiory effect of erytromycin on tumor necrosis factor-α and interleukin-6 production. Fundam Clin Pharmacol 2001;15:419–24.
22. Dudley DJ, Trautman MS, Araneo BA, Edwin SS, Mitchell MD. Decidual cell biosynthesis of interleukin-6: regulation by inflammatory cytokines. J Clin Endocrinol Metab 1992;74:884–9.
© 2004 The American College of Obstetricians and Gynecologists
23. Sirota L, Shacham D, Punsky I, Bessler H. Ibuprofen affects pro- and anti-inflammatory cytokine production by mononuclear cells of preterm newborns. Biol Neonate 2001;79:103–8.