Cytokines are nonimmunoglobin polypeptides secreted by monocytes and lymphocytes that modulate the magnitude of inflammatory and immune responses. Recent studies have shown that increased cytokine production can lead to adverse pregnancy outcome.1–5 Elevated fetal levels of interleukin (IL)-6, for example, have been associated with fetal injury in a condition known as the fetal inflammatory response syndrome. Several studies have associated the fetal inflammatory response syndrome with adverse neurologic outcomes, including periventricular leukomalacia, intraventricular hemorrhage, and cerebral palsy. These studies have shown strong correlation with both fetal and maternal IL-6 levels and the above adverse neonatal outcomes.
Although an association between elevated IL-6 levels in the fetus and adverse sequelae has been established, the mechanism by which IL-6 becomes elevated in the fetus is not. Depending on the stimulus, cytokine production can be independently stimulated in the mother, fetus, and/or placenta.6 For example, Pierce et al7,8 demonstrated placental production of IL-6 and tumor necrosis factor-alpha (TNF-α) using the ex vivo placenta perfusion model. In 2 separate studies, they showed hypoxia as well as hypoperfusion can lead to placental cytokine production. Stallmach et al9 reported elevated TNF-α and IL-6 in the amniotic fluid of women with chorioamnionitis. In addition, IL-6, but not TNF-α was elevated in the maternal blood, providing indirect evidence for IL-6 transfer across the placenta. In the study now reported, we sought to determine whether the inflammatory cytokines, IL-6, TNF-α, and IL-1α, cross the placenta using the healthy-term ex vivo human placental perfusion model. Because TNF-α and IL-1α can induce synthesis of IL-6,10,11 we also determined whether exposure of the placenta to IL-6, TNF-α, and IL-1α results in additional cytokine production.
MATERIALS AND METHODS
From March 2002 to July 2002, we perfused 10 placentas from women who had uncomplicated term pregnancies undergoing scheduled repeat cesarean deliveries. The placentas were collected at the time of delivery by using an atraumatic technique. The placentas used in this study were collected in accordance with the guidelines set by The University of Texas Southwestern Medical Center Institutional Review Board for Human Studies. The placentas were transferred to the laboratory in isotonic sodium chloride solution within 10 minutes of delivery. On arrival at the laboratory, a fetal artery and vein of a single cotyledon were cannulated with 3F and 5F catheters, respectively. The fetal circulation of the cannulated cotyledon was then gently perfused with Eagle minimal essential medium with Earle's salts and was examined for evidence of loss of membrane integrity. Placental cotyledons that were noted to have fetal–maternal leaks were discarded. If there were no detectable leaks, the perfused cotyledon and the surrounding tissues were transferred to a temperature-controlled chamber and perfused with Eagle minimal essential medium for 20 minutes to remove residual blood and to stabilize the perfusion pressure in the fetal circulation. At the end of the 20-minute stabilization period, the perfusion pressure typically had reached a stable baseline of 35 to 50 mm Hg.12 If a placenta failed to reach a stable baseline pressure, it was discarded. After an additional 30 minutes of stable pressure in the fetal circulation, the placental membrane integrity, transport fraction, and clearance index of the cytokines were determined by using carbon 14-labeled antipyrine and the method of Challier.13 A transport fraction of more than 40% for antipyrine was deemed representative of a maternal–fetal circulatory match.
The cotyledon was perfused initially in an open–open (nonrecirculating) system. The maternal and fetal compartments consisted of 150 mL of Eagle minimal essential medium that was aerated with a gaseous mixture of 95% air and 5% carbon dioxide. Both maternal and fetal compartments were continuously mixed. The fetal and maternal flow rates were constant at 4 to 5 mL/min and 17 mL/min, respectively. The clearance index of the cytokines studied was calculated by comparing the transport fraction of each agent to that of the reference compound antipyrine.
The cytokines to be perfused (Human Recombinant; Sigma, St. Louis, MO) were diluted with sterile water solution to known concentrations and added to the maternal circulation. At 10-minute intervals for 1 hour, samples were collected from the fetal side and cytokine levels were determined. To determine fetal accumulation, the system was converted to a closed–closed circulation (recirculating) as previously described12 60 minutes after the initial cytokines were added. Maternal and fetal levels were determined at both 90 and 120 minutes.
Similarly, with different placentas, the study cytokines were added to the fetal side in an open–open circulation technique for the first hour, and accumulation was measured in a closed–closed circulation during the second hour of perfusion. Samples were collected from the maternal side at 10-minute intervals during the first hour and again at 90 and 120 minutes.
In the 4 placentas collected to serve as controls for endogenous production of cytokines, the placentas were perfused with the exact same methodology described above, and cytokines were collected to determine whether the experiment itself resulted in cytokine production.
Measurement of TNF- α was performed using a solid phase chemiluminescent enzyme-linked immunosorbent assay (ELISA) from R & D Systems (Minneapolis, MN). Intra-assay precision for serum ranged from 1.8% to 6.0% coefficient of variation (CV). The reportable range of TNF-α was from a minimum value of 0.28 to a maximum of 1.7 pg/mL. No significant cross-reactivity was observed. Measurement of IL-6 was also measured by using a chemiluminescent ELISA assay (R & D Systems). The interassay precision for this assay ranged from 7.7% to 1.0% CV, and the intra-assay precision ranged from 2.4% to 3.4% CV. The minimum detectable dose of IL-6 was less than 0.2 pg/mL. IL-1α was measured using a solid-phase ELISA from R & D Systems. The intra-assay variation for this assay ranged from 1.4% to 2.2% CV, whereas the interassay variation ranged from 4.3% to 8.3% CV. The lowest detectable dose of IL-1α was less than 1.0 pg/mL.
The clearance index was used to establish whether the cytokines transfer from the maternal–fetal circulations and from the maternal–fetal circulations. Mean and standard errors are reported for cytokine concentrations. Study and control cytokine concentrations were compared by using the Wilcoxon rank sum test with a P value at < .05 deemed statistically significant.
A total of 18 placentas were collected for this study in a nonrandom fashion from term, Hispanic women during elective cesarean delivery. Of these, 8 were not used secondary to leaks, loss of membrane integrity, or their inability to reach or maintain a stable baseline perfusion pressure. Of the 10 placentas studied, 4 were used as controls to determine endogenous cytokine production. Of the remaining 6 placentas, 4 were used to test maternal–fetal transfer, and 2 were used to test fetal– maternal transfer.
Known amounts of the study cytokines were added to the maternal circulation of the model in concentrations varying from low to high. For TNF-α, this ranged from 165 pg/mL to 337 pg/mL for a mean concentration of 253 ± 64 pg/mL. The clearance index was 0.001, suggesting minimal transfer of TNF-α to the fetal side. For IL-1α, the amount added to the maternal circulation ranged from 47 pg/mL to 388 pg/mL, for a mean concentration of 199.8 ± 153 pg/mL. The clearance index was 0.001, suggesting minimal transfer of IL-1α to the fetal side. As shown in Table 1, IL-6 was added to the maternal circulation in concentrations ranging from 22.4 pg/mL to 283 pg/mL, for a mean value of 139.8 ± 113 pg/mL. The clearance index at 60 minutes was 0.30, suggesting transfer of IL-6 to the fetal circulation.
For the second experiment, known amounts of the study cytokines were added to the fetal circulation of the model. Unfortunately, results from TNF-α transfer are not available because of technical difficulties with the ELISA kit. For IL-1α, the amount added to the fetal circulation ranged from 240 pg/mL to 246 pg/mL, for a mean value of 243 ± 3 pg/mL. The clearance index was 0.001, suggesting minimal transfer of IL-1α to the maternal side. As shown in Table 2, IL-6 was added to the fetal circulation in concentrations of 126 pg/mL and 186 pg/mL, for a mean value of 156 ± 30 pg/mL. The clearance index at 60 minutes was similar to that seen with maternal to fetal transfer of IL-6 at a value of .23, suggesting transfer of IL-6 from the fetal to maternal circulation.
At 60 minutes, a closed–closed circulation was established in each placenta to determine whether accumulation of cytokines occurred on either the maternal or fetal side. TNF-α showed accumulation on the maternal side at 120 minutes (433 pg/mL at 60 minutes, more than 7,000 pg/mL or too high to detect at 120 minutes). As explained previously, no information was available on the fetal side for TNF-α. IL-1α showed no accumulation on the maternal side with 131 ± 65 pg/mL at 90 minutes versus 132 ± 37 pg/mL at 120 minutes. Likewise, there was no accumulation on the fetal side with 230.5 ± .5 pg/mL at 60 minutes versus 90.5 ± 58 pg/mL at 120 minutes. IL-6 showed accumulation on the maternal side. At 60 minutes the level was above the sensitivity of the kit (more than 300 pg/mL), and the other values ranged from 120 pg/mL to 212 pg/mL. At 120 minutes all values were more than 300. Likewise, accumulation of IL-6 was seen on the fetal side with 156 pg/mL ± 35 pg/ml at 60 minutes versus 241 ± 42 pg/mL at 120 minutes.
To determine whether the increases in cytokine levels seen in the model were influenced by the stress of the model itself, IL-6 values were measured in a group of 4 placentas exposed to the same experimental conditions. IL-6 was chosen for this part of the experiment because it was the only compound measured that actually crossed the placenta. As shown in Table 3, on the maternal side, IL-6 levels showed a level of 4.79 ± 4.3 pg/mL at 60 minutes and a level of more than 40 pg/mL at 3 hours. IL-6 showed an increase from 0.7 ± 1.1 pg/mL at 60 minutes to 8.5 ± 3.63 pg/mL at 3 hours on the fetal side. The concentrations in the study placenta were higher in all cases but, secondary to the small sample size, they did not reach statistical significance. The concentrations of IL-6 on the fetal side when cytokines were added to the maternal circulation were not statistically different from the controls (14.2 ± 5.9 pg/mL versus 0.7 ± 1.1 pg/mL, P = .06) at 60 minutes. Also, the maternal concentrations of IL-6 were not statistically different from controls when cytokines were added to the fetal circulation (19.4 pg/mL ± 4.3 pg/mL versus 4.79 pg/mL ± 4.3 pg/mL versus P = .06) at 60 minutes. Clearly, there was a biological difference between study and control placentas.
In this study of placental transfer of inflammatory cytokines, we found both maternal–fetal and fetal–maternal transfer of IL-6 across the placenta. We were unable to demonstrate, however, the transfer of TNF-α or IL-1α across the placenta, and this was true even at concentrations significantly higher than would be expected to occur in vivo. IL-6 is a large molecule, weighing approximately 23 to 30 kd, and is unlikely to passively diffuse across the placenta.14 Although the placenta has the ability to produce IL-6, our controls did not show this to occur under our experimental conditions. This suggests that our findings of bidirectional transfer of IL-6 was most likely the result of active transport and not placental stimulation as a result of our experimental model. The possibility of an adenosine triphosphate–mediated transporter on both the maternal and fetal sides could explain transfer of this large polypeptide. In addition, some placental production was stimulated by exposure of the placenta to the cytokines studied, as evidenced by IL-6 and TNF-α accumulation at 90 or 120 minutes in our closed-circulation experiment.
Recent investigations have provided indirect evidence for placental transfer of IL-6 consistent with our results. Goetzl et al15 reported an association with increased IL-6 levels in maternal, as well as fetal serum. This study was not able to determine whether the origin of the elevated fetal IL-6 levels was from transplacental passage or from a secondary fetal inflammatory response. They recognized the importance of this finding, however, as a potential risk for fetal brain injury. Steinborn et al16 exposed cultured placental tissue to infection and measured cytokine production. In this study, elevated cytokine production was from fetal placental cells, not maternal decidual cells, suggesting that elevated maternal IL-6 levels seen in chorioamnionitis might be caused by IL-6 transfer across the placenta and not maternal production.
The bidirectional transfer of IL-6 across the placenta could help explain maternal response to intrauterine or fetal conditions and vice versa. For example, Romero et al5 showed an association between fetal IL-6 levels and spontaneous preterm labor. They hypothesized a complex set of events initiated by elevated fetal IL-6 levels that lead to myometrial activation, prostaglandin production, and preterm labor. Furthermore, they proposed that the fetus in response to an intrauterine infection releases inflammatory cytokines, such as IL-6, to initiate preterm labor and exit a hostile intrauterine environment. The transport of IL-6 across the placenta may play an important role in maternal decidual cell activation and prostaglandin production consistent with this theory. One limitation of our study is that preterm placentas were not used. Smaller placentas are problematic because of the difficulty in cannulating the placental vessels, fitting the placenta properly in its chamber, and obtaining a working model without leaks. Therefore, we cannot comment on cytokine transfer earlier in gestation.
Maternal infections, such as appendicitis and pyelonephritis, have been associated with fetal injury. Mays et al17 recently described 3 cases in which significant neurologic injury occurred in pregnancies complicated by appendicitis. Jacobson et al18 studied risk factors for cerebral palsy in preterm infants in a population-based case-control study. Both pyelonephritis and clinical chorioamnionitis significantly raised the risk of cerebral palsy. Of the 148 cases of cerebral palsy studied, 18 (12%) were associated with clinical chorioamnionitis or pyelonephritis with an odds ratio 2.02 (95% confidence interval 1.02, 3.99) versus controls. The investigators highlighted the importance of studying inflammatory mediators in the membranes, amniotic fluid, and blood of the newborn in determining the cause of cerebral palsy. Although both studies show a significant association between maternal infection and adverse neonatal outcome, neither study was designed to illustrate the pathway by which this occurs. Although the mechanisms that lead to fetal injury in these conditions are likely complex, our study results suggest that cytokines released by maternal infection, especially IL-6, could lead to ill effect on the fetus through placental transport. Furthermore, exposure of the placenta to inflammatory cytokines, such as TNF-α and IL-1, may further increase IL-6 levels, resulting in even greater transfer to the fetus. Clinicians may have grounds for concern when there is a significant maternal infection or inflammatory response syndrome because it may lead to fetal injury as a result of transfer of IL-6 across the placenta. Further studies are needed to confirm these findings.
1.Gomez R, Romero R, Ghezzi F, Yoon BH, Mazor M, Berry SM. The fetal inflammatory response syndrome. Am J Obstet Gynecol 1998;179:194–202.
2.Adinolfi M. Infectious diseases in pregnancy, cytokines, and neurological impairment: an hypothesis. Dev Med Child Neurol 1993;35:549–53.
3.Yoon BH, Romero R, Yang SH, Jun JK, Kim IO, Choi JH, et al. Interleukin-6 concentrations in umbilical cord plasma are elevated in neonates with white matter lesions associated with periventricular leukomalacia. Am J Obstet Gynecol 1996;174:1433–40.
4.Yoon BH, Romero R, Kim CJ, Koo JN, Choe G, Syn HC, et al. High expression of tumor necrosis factor-α and interleukin-6 in periventricular leukomalacia. Am J Obstet Gynecol 1997;177:406–11.
5.Romero R, Gomez R, Ghezzi 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.
6.Dammann O, Leviton A. Maternal intrauterine infection, cytokines, and brain damage in the preterm newborn [review]. Pediatr Res 1997;42:1–8.
7.Pierce BT, Pierce LM, Wagner RK, Wagner RK, Apodaca CC, Hume RFJr, et al. Hypoperfusion causes increased production of interleukin 6 and tumor necrosis factor α in the isolated, dually perfused placental cotyledon. Am J Obstet Gynecol 2000;183:863–7.
8.Pierce BT, Napolitano PG, Pierce LM, Apodaca CC, Hume RF Jr, Calhoun BC. The effects of hypoxia and hyperoxia on fetal-placental vascular tone and inflammatory cytokine production. Am J Obstet Gynecol 2001;185:1068–72.
9.Stallmach T, Hebisch G, Joller-Jemelka HI, Orban P, Scwaller J, Engelmann M. Cytokine production and visualized effects in the feto-maternal unit: quantitative and topographic data on cytokines during intrauterine disease. Lab Invest 1995;73:384–92.
10.Oberholzer A, Oberholzer C, Moldawer LL. Cytokine signaling: regulation of the immune response in normal and critically ill states [review]. Crit Care Med 2000;28(suppl 4):N3–12.
11.Dinarello CA. Biologic basis for interleukin-1 in disease [review]. Blood. 1996;87:2095–147.
12.Schneider H, Huch A. Dual in vitro perfusion of an isolated lobe of human placenta: method and instrumentation. Contrib Gynecol Obstet 1985;13:40–7.
13.Challier JC. Criteria for evaluating perfusion experiments and presentation of results [review]. Contrib Gynecol Obstet 1985;13:32–9.
14.Santhanam U, Ghrayeb J, Sehgal PB, May LT. Post-translational modifications of human interleukin-6. Arch Biochem Biophys 1989;274:161–70.
15.Goetzl L, Evans T, Rivers J, Suresh MS, Lieberman E. Elevated maternal and fetal serum interleukin-6 levels are associated with epidural fever. Am J Obstet Gynecol 2002;187:834–8.
16.Steinborn A, von Gall C, Hildenbrand R, Stutte HJ, Kaufmann M. Identification of placental cytokine-producing cells in term and preterm labor. Obstet Gynecol 1998;91:329–35.
17.Mays J, Verma U, Klein S, Tejani N. Acute appendicitis in pregnancy and the occurrence of major intraventricular hemorrhage and periventricular leukomalacia. Obstet Gynecol 1995;86:650–2.
18.Jacobsson B, Hagberg G, Hagberg B, Ladfors L, Niklasson A, Hagberg H. Cerebral palsy in preterm infants: a population-based case-control study of antenatal and intrapartal risk factors. Acta Paediatr 2002;91:946–51.