Selective modulation of monocyte and neutrophil responses with activated protein C in preterm infants

Abstract Background Inflammation is associated with many disorders of preterm infants including periventricular leukomalacia, chronic lung disease, and necrotizing enterocolitis. Activated protein c (APC) has shown positive immunomodulatory effects. Objectives We aimed to study neutrophil and monocyte function in response to lipopolysaccharide (LPS) and APC stimulation ex vivo in preterm infants <32 weeks gestation over the first week of life compared to neonatal and adult controls. Methods Peripheral blood was taken on day 1, 3, and 7 and stimulated with LPS in the absence or presence of APC. Expression of toll-like receptor 4 (TLR4) and CD11b and reactive oxygen intermediate (ROI) release from neutrophils and monocytes was examined by flow cytometry. Results LPS induced neutrophil ROI in adults and preterm infants and was significantly reduced by APC. Baseline and LPS-induced monocyte ROI production in preterm neonates was increased compared to adult and term controls. Neutrophil TLR4 baseline expression was higher in term controls compared to preterm infants. Conclusion Increased systemic ROI release in preterm infants may mediate tissue damage, ROI was reduced by APC. However, due to the high risk of hemorrhage further examination of APC mutant forms with anti-inflammatory but decreased anticoagulant properties is merited.


Introduction
Neonates are more susceptible to infections due to an altered immune system [1]. Infections among very low birth weight infants are associated with high mortality, poor growth, and adverse neurodevelopmental outcomes [1,2]. Neutrophils and monocytes are important components of the neonatal immune response. Impairment of neutrophil adherence, chemotaxis, and migration in neonates increases their susceptibility to infection in the first month of life [1,2]. The processes of neutrophil adhesion and diapedesis are mediated by CD11b subunit adhesion molecule of macrophage-1 antigen (Mac-1) [3]. Neutrophil CD11b expression increases with gestational age [4] and also in preterm neonates with respiratory distress syndrome and sepsis [4][5][6]. In addition, the neutrophil and monocyte recognition of microbial invasion via endotoxin is vital for host defense. Toll-like receptor 4 (TLR4) plays an important role in detecting microbial infection and triggering antimicrobial host defense responses and endotoxin signaling [7]. Premature infants express very low levels of monocyte TLR4 which may predispose them to bacterial infection during the neonatal period [8].
Once invading bacteria are recognized, neutrophils and monocytes are recruited to the site of infection [9]. The intracellular respiratory burst is an essential mechanism, by which these cells eliminate invading microorganisms utilizing reactive oxygen intermediate (ROI) production [10]. Neutrophil respiratory burst production is increased in preterm and term neonates with high C-reactive protein (CRP) in the first few days of life and increases with gestational age [10]. We have previously described high CRP as a predictor for histological chorioamnionitis and preterm birth [11]. Ballabh et al. reported increased neutrophil respiratory burst in response to E. coli stimulus in preterm neonates with increasing postnatal age in first four weeks of life [12].
Modulating these immune responses may improve preterm neonatal outcomes. Activated protein C (APC) plays an important role in coagulation and the modulation of inflammation [13]. Case studies of preterm neonates demonstrated that APC improves coagulation and outcomes in neonates with sepsis and disseminated intravascular coagulation [14][15][16]. We hypothesized that protein C would affect neutrophil and monocyte activation and function in response to endotoxin. We aimed to investigate neutrophil and monocyte migration, LPS recognition and ROI production in response to LPS and the effect of APC on these responses ex vivo.

Patient population
This study was approved by the Institutional Research Ethics Committee and fully informed written consent was obtained from the parents of all participants during the study period. The following study groups were enrolled: Adult controls: consisted of healthy volunteers. Neonatal term controls: umbilical cord blood of term infants with uncomplicated normal delivery and postnatal course; Preterm infants: less than 32 weeks gestation admitted to the NICU. Infants were excluded if there was a history of maternal substance abuse or in the presence of a major congenital anomaly.
Clinical details including maternal preeclampsia, histological chorioamnionitis, antenatal steroid administration, Apgar scores, respiratory distress syndrome, and ventilation days were recorded. Neonatal outcomes were recorded as follows: RDS, chronic lung disease, necrotizing enterocolitis, late onset sepsis, and intraventricular hemorrhage. Patent ductus arteriosus on echocardiography receiving ibuprofen therapy [17]. Histological and microbiological examinations of placentas of preterm infant and infants at risk of brain injury were reviewed by consultant perinatal pathologists and graded according to Redline [18].
All infants had cranial ultrasounds performed by a consultant radiologist with special interest in neonatal brain injury on day 1, 3, and 7 of life and on the day of discharge. Preterm infants <30 weeks gestation or with abnormalities on cranial ultrasound had brain magnetic resonance imaging (MRI) at term-corrected age. All the scans were scored independently by a pediatric radiologist who was blinded to the clinical and laboratory results.

Sample processing
Serial blood samples were taken from preterm infants on day 1, 3, and 7 of life during routine phlebotomy from either a central line or peripheral sample. Peripheral blood was collected in sodium citrate anticoagulant blood tubes and were kept on ice and processed within 90 min of collection. Peripheral blood was taken from healthy adults and cord blood from full-term neonates as controls. Whole blood was incubated for 1 h at 37 o C with LPS (10 ng/mL) to mimic an inflammatory response in vitro. In addition, APC (200 ng/mL) was added in the presence or absence of LPS following a dose-response study [19].

Quantification of cell surface antigen expression
Fluorochrome-conjugated monoclonal antibodies (mAb) specific for human CD11b (D12: BD Bioscience, Franklin Lakes, NJ) and TLR4 (HTA125: eBioScience, San Diego, CA) were used. An aliquot of 50 lL of whole blood was stained with mAbs for 15 min. Red blood cells were lysed with BD lysis buffer. Cells were acquired on a BD Accuri 6 C6 flow cytometer with a CFlow Plus software (BD Bioscience). Neutrophils and monocytes were delineated by sorting cell depending on the forwards and sideway scatter profiles. A consultant hematologist and histologist examined the sorted cells under the microscope [20].

Intracellular respiratory burst activity
Generation of ROI was evaluated by flow cytometry using the technique of Smith and Wiedemann [21]. An aliquot of 50 mL of whole blood was incubated with DHR 123 (100 lM) at 37 C for 10 min. Cells were stimulated with 16 mM PMA for 20 min at 37 C. The reaction was then halted by placing samples on ice. Neutrophil and monocyte fluorescence intensity was assessed by flow cytometry and expressed as mean channel fluorescence (MFI). DHR 123 has been shown to detect mainly intracellular H 2 O 2 and OH radical production [21].

Evaluation of protein C activity
Protein C level in plasma from patients was measured by incubating the plasma with protein C activator and the further quantification of APC with a synthetic chromogenic substrate as previously described [22]. HemosIL Normal Control and HemosIL Low Abnormal were used as internal quality control at the start of each working day and subsequently every 4 h throughout the day.

Statistical methods
Matched paired t-test was used to compare conditions within each group. For comparing across groups oneway ANOVA with Tukey post HOC comparison method was used. Statistical analysis was carried out using analysis of variance (ANOVA) using PASW statistical package version 18. Significance was assumed for values of p < .05.

Protein C levels
Protein C levels were available in 26 out of 30 preterm infants as the remaining samples were unsuitable for analysis. Protein C levels were 0.1 (0.04-0.16) mg/L in preterm infants significantly lower in comparison to term neonates (0.94-2.29 mg/L) and adults (0.43-83 mg/L; (p < .0001). Both preterm and term neonatal levels were significantly decreased compared to adults (p < .0001). All adult control and neonatal term control protein c levels were within or above the normal range and respectively (Figure 1(A)).

Monocyte and neutrophil CD11b expression in preterm infants
The effect of APC on neutrophil and monocyte migration and toxin detection was measured by CD11b and TLR4 expression by flow cytometry. There was no significant difference in neutrophil CD11b baseline expression between preterm infants and both adult and neonatal controls. There was significant increase in LPS-induced neutrophil CD11b expression in both control groups, adults and neonates, and preterm infants on day 1, 3, and 7 compared to baseline neutrophil CD11b (p < .001). Although LPS-induced neutrophil CD11b expression was significantly increased in all groups the response was highest in adults. APC had no effect on LPS-induced neutrophil CD11b expression in preterm neonates or controls (Figure 1(B)).
Similar to neutrophil CD11b, no difference was observed in monocyte CD11b expression between adult and neonate controls. Monocyte CD11b baseline expression was significantly higher in preterm infants compared to both control groups (p ¼ .001). Monocyte CD11b LPS response was higher in all the groups compared to unstimulated (p > .0001), no significant difference was observed between controls and preterm infants after LPS stimulation. APC had no effect on LPS-induced monocyte CD11b expression in preterm neonates or controls (Figure 1(C)).

Preterm monocyte and neutrophil TLR4 expression
Neutrophil TLR4 baseline expression was significantly higher in term controls than preterm infants and adult controls (p < .001; Figure 2(A)). There was a significant increased LPS-induced neutrophil TLR4 expression in preterm infants only on day 1 and day 3 of life compared to baseline (p ¼ .01; Figure 2(A)). Monocyte TLR4 baseline expression in preterm infants was similar to both control groups (Figure 2(B)). In preterm infants, monocyte TLR4 expression in response to LPS was significantly higher on day 1 and 7 of life (p < .001 and p ¼ .002, respectively; Figure 2(B)). APC had no effect in reducing LPS-induced neutrophil and monocyte TLR4 expression in preterm infants (Figure 2). There were no significant differences in neutrophil and monocyte TLR4 baseline expression or in response to LPS or APC between the preterm infants with abnormal and normal neuroimaging. These results showed that neutrophil and monocyte migration and endotoxin recognition are not altered in preterm neonates after LPS stimulation, and this is not altered with APC.

Preterm neonatal monocyte and neutrophil ROI production
ROI expression was used to study the effect of APC on neutrophil and monocyte function after LPS stimulation. Neutrophil ROI baseline expression in preterm infants was similar in both control groups (Figure 2(C)). There was a significant increase in neutrophil ROI response to LPS in adults (p ¼ .001) and preterm neonates on day 1 (p ¼ .009) and day 3 (p ¼ .01), however, it was not increased on day 7 and neonatal controls (Figure 2(C)). Monocyte ROI was significantly increased in preterm infants compared to adult and neonate controls (p < .0001; Figure 2(D)). Monocyte ROI was significantly increased in response to LPS in adult controls (p ¼ .015) and preterm neonates on day 1 (p ¼ .003) and day 3 (p ¼ .001), but not in neonatal controls and preterm infants on day 7 (Figure 2(D)).
APC significantly reduced LPS-induced neutrophil ROI in adults (p ¼ .014) and preterm neonates on day 1 (p ¼ .019; Figure 2(C)). APC significantly reduced LPS-induced monocyte ROI in adults (p ¼ .041) but did not alter the monocyte response in preterm neonates on day 1 and 3 of life (Figure 2(D)) or term controls. No alteration of monocyte and neutrophil ROI was observed in the presence of APC alone. These results showed an acute increase in intracellular respiratory burst in preterm infants that was decreased with APC.

Discussion
Protein c levels were significantly reduced in preterm infants compared with adult and neonatal controls. Neutrophil ROI production was significantly increased in preterm neonates on day 1 in response to LPS and Figure 1. Protein c levels in adults compared with term neonatal and preterm infants. (A) protein c activity was measured in adult controls (adult con; n ¼ 9), term neonatal controls (neo con; n ¼ 10) and preterm infants (n ¼ 26). ÃÃÃ p < .0001 using oneway ANOVA. Whole blood was incubated with LPS in the absence or presence of APC for 1 h. Neutrophils and monocytes were stained with mAb anti-CD11b and mean channel fluorescence (MFI) was measured by flow cytometry. The neutrophil and monocyte populations were selected based on their scatter profile: forward scatter and sideway scatter. (B and C) graphs show MFI ± SEM of neutrophil (B) and monocyte (C) CD11b in blood from adult and neonatal controls and preterm infants. ÃÃÃ p < .0001 using one-way ANOVA comparing between LPS stimulated and unstimulated. ### p < .0001 using one-way ANOVA to compared NE unstimulated and adult and neonatal controls. this was significantly reduced by APC. There was no significant difference in LPS-induced neutrophil ROI between preterm infant and adult controls.
Bj€ orkqvist et al. showed a similar baseline neutrophil ROI in term and preterm neonates but did not examine the effect of LPS stimulation [23]. Preterm babies born at less than 32 weeks gestation have lower neutrophil ROI generation in response to Formyl-Methionyl-Leucyl-Phenylalanine (fMLP) compared to neonates >32 weeks without sign of infection and higher fMLP-induced neutrophil ROI was associated with elevated CRP [10].
Different studies have shown similar baseline neutrophil CD11b expression in preterm neonates compared to adults [20,24]. We demonstrated significantly increased LPS-induced CD11b expression in preterm neonates, term neonates, and adult controls [20]. Similar to other reports, this response was significantly higher in adult controls compared to preterm neonates and term cord blood [4,20]. This was in accordance with our findings of relative endotoxin tolerance in both term and preterm neonates [25].
Neutrophil TLR4 baseline expression was lower in preterm infants and adult controls compared to term neonatal controls. Monocyte TLR4 baseline expression was similar in all groups. However, LPS-induced neutrophil and monocyte TLR4 were significantly higher in preterm neonates compared to both control groups. Studies have shown conflicting results for neonatal monocyte and neutrophil TLR4 expression. Some studies have reported neonatal TLR4 expression increased after LPS stimulation, although it was lower than adults [26][27][28][29]. However, Yerkovich et al. found higher TLR4 expression after LPS stimulation in 1-yearold babies compared to adults [30]. We have previously shown similar neutrophil and monocyte CD11b and TLR4 in responses to LPS and APC in a cohort of 15 patients in the pediatric intensive care unit admitted with different medical and surgical conditions [22].
Neonatal phagocytic cells have low ROI production in response to LPS making neonates susceptible to sepsis. We have previously suggested the use of vitamin D to improve ROI production in phagocytic cells as a response to LPS in preterm infants. Vitamin D induced ROI production by these cells [31]. On the contrary, activated monocytes and neutrophils and ROI production may induce tissue damage in neonatal encephalopathy. We found that APC modified neutrophil and monocyte LPS responses in neonates with NE ex vivo [32]. Here, we investigated the effect of APC on ROI production in preterm infants. A reduction in LPSinduced neutrophil ROI in preterm infants was observed after treatment with APC, which may have important implications in neonatal inflammatory diseases. Agents that block ROI release may reduce neonatal inflammation and decrease end-organ injury. In addition brain injury in preterm infants has been linked to sustained inflammation beyond the neonatal period that may be amenable to immunomodulation [33].
However, the Cochrane on the use of APC in neonatal sepsis suggested that it should not be used outside clinical trials and the recent PROWESS-SHOCK trial of adult sepsis found APC provided no benefit. The commercial PAC (Xigirs) has been removed from the market by Eli Lilly. However, newer forms of APC with antiinflammatory and reduced anticoagulant effects may be useful in reducing this early inflammatory response but have not been studied in neonates [34]. This pilot study is limited by the number of preterm infants included. Further studies including a larger number off infants would allow stratification by outcome and more detailed understanding of the relevance of the results in clinical practice. In summary, we found that APC reduced neutrophil ROI production in vitro in preterm neonates on day 1 of life but there was no effect on neutrophil/monocyte CD11b and TLR4 expression.