What Is Known
- Necrotizing enterocolitis remains one of the most devastating intestinal diseases of premature neonates. It is characterized by severe inflammatory response, intestinal ischemia, thrombocytopenia, and disseminated intravascular coagulopathy.
- The association between sepsis and thrombosis has been extensively studied in adults showing a tight link between endothelial damage, inflammation, and coagulation.
What Is New
- This is the first report showing significant alterations in the gene expression of proteins related to the coagulation and anticoagulation systems in neonates with necrotizing enterocolitis.
- The altered genes identified are potentially linked to an overall procoagulant status, reduction in fibrinolysis and worse endothelial regeneration.
Necrotizing enterocolitis (NEC) remains one of the most devastating and poorly understood intestinal diseases of premature neonates (1). In advanced NEC, mortality rate is as high as 50%, and it has not changed in the last decades mainly because of the lack of new medical treatments (2). Despite extensive research into its pathogenesis and prevention, we remain unable to predict which infants will progress to advanced disease and what are the biological mechanisms leading to severe thrombocytopenia, disseminated intravascular coagulopathy (DIC), intestinal ischemia, and death.
Neonatal coagulation is not completely characterized in premature infants, and a state of balance exists between procoagulant and anticoagulant circulating proteins (3). It is known that neonates have higher tendency, compared with older children, to develop thrombosis during sepsis, inflammation, hypotension, hypoxia, and with the use of intravascular catheters (4).
Severe inflammation, thrombocytopenia, and DIC are common findings associated with advanced NEC, which is characterized by extensive intestinal necrosis, multiorgan failure, and high mortality (5). The association between severe inflammation and thrombosis has been extensively studied in adult sepsis, and coagulopathy has been one of the main therapeutic targets to reduce mortality (6). At present, there is absence of data on how the coagulation system reacts in advanced NEC and whether a procoagulant status could be a key mechanism contributing to disease progression and diffuse intestinal necrosis. Our hypothesis is that severe inflammation would lead to anomalies in the coagulation system with a possible procoagulant status in infants with confirmed and advanced NEC (Bell stage 2 and 3) (7). This could explain the consumptive coagulopathy, the mesenteric vessels thrombosis, and the patchy intestinal ischemia that are characteristics of advanced disease and associated with worse outcome. To test this hypothesis, we have conducted a pilot study to identify specific coagulation-related gene expression in NEC compared with healthy infants and septic controls of similar gestational age.
A total of 52 neonates were prospectively recruited from December 2013 to June 2015. Modified Bell classification was used to enroll and stratify NEC neonates (7). Inclusion criteria were infants with NEC stage 2 (confirmed) or 3 (advanced) at admission. Nineteen of 52 patients were excluded from the analysis because they were twins (only 1 kept) or the samples were not of acceptable quality (low RNA concentrations and complementary DNA contamination). Two control groups, matched for birth weight (BW) and corrected gestational age, were prospectively selected within the same study period. The first control group was based on the absence of infection, inflammation, and coagulopathy (n = 10, healthy control). The second control group was based on infants with sepsis (n = 12, septic group) as defined by Oeser et al (8). Clinical end points analyzed for comparisons among groups were hemoglobin, platelet, clotting profile, laparotomy, and survival between NEC and control neonates. This study was performed with approval from the St George's University of London Research and Development Office (protocol number 13/SC/0075), and written parental consent was obtained on behalf of the neonates.
A total of 0.5 mL of venous blood was taken from each infant and incubated with RNAlater solution (Life Technologies, Paisley, UK) according to the manufacturer's instructions. Total RNA was isolated using the RiboPure Blood RNA kit (Life Technologies) according to the manufacturer's instructions and an additional DNase digestion step carried out. The RNA concentration was determined spectrophotometrically (ND-1000; Nanodrop Technologies, Wilmington, DE), and samples were stored at −80°C.
Expression Analysis of Coagulation Genes
A Human Blood Coagulation quantitative real time-polymerase chain reaction (qRT-PCR) array (96 StellArray; Bar Harbor BioTechnology, Trenton, ME) was utilized according to the manufacturer's procedures. The PCR arrays include 94 genes related to the coagulation pathways using a 96-well plate. The complete list of genes analyzed with the Human Blood Coagulation qRT-PCR array can be found in the Supplementary Table (http://links.lww.com/MPG/A654). Total RNA (1.0 μg) from patients and control subjects was reverse transcribed using the GoScript Reverse Transcription System (Promega, Madison, WI) according to the manufacturer's protocol. qRT-PCR was performed using GoTaq quantitative polymerase chain reaction (qPCR) Master Mix (primer design) in a final volume of 20 μin a Bio-Rad CFX96 real-time PCR detection system.
Data and Statistical Analysis
The threshold cycle (Ct) values generated from qPCR reactions were analyzed using the Global Pattern Recognition (GPR) software (Bar Harbor BioTechnology), which went through several iterations to compare the change of expression of a gene normalized to every other gene in the set of genes being analyzed, taking advantage of biological replicates to extract statistically significant changes in gene expression (9). The relative messenger RNA levels were normalized based on the geometric mean of all the genes on the array selected as normalizers by the GPR algorithm (CALR, F2R, SELPLG, MTHFR, VKORC1, and ANXA5). A control for genomic DNA contamination and a conventional 2-ΔΔCt analysis using 18S as a normalizer gene was also included in the output. A Ct value >35 was set as the expression cutoff and not included in the analysis. Data were expressed as the mean of log2 of fold change (FC). The FC threshold was set at ±1.5. A Students t test was used to assess the significance of differential changes for each gene. The χ2 test for categorical data and Kruskal–Wallis H test or Mann–Whitney U test for continuous data were used to compare not normally distributed data among the groups. The threshold for significance for the gene expressions was set at P < 0.05. Statistical analyses were performed using Stata 11 (release 11.2, Stata Corp, College Station, TX) and Stats Direct (version 2.7.8; Stats Direct Ltd, Cheshire, UK) statistical software.
A total of 33 neonates were analyzed in this pilot study: 11 NEC, 12 sepsis, and 10 healthy controls. The NEC group consisted of 6 infants with stage 2 and 5 with stage 3 (Bell classification). Demographic characteristics and outcomes are shown in Table 1. The 3 groups were similar in terms of gestational age at birth, age at sampling, and sex. Babies with NEC were significantly more thrombocytopenic and had higher C-reactive protein compared with the sepsis group. The mortality rates in the NEC and sepsis groups were not significantly different. Five of 11 of the NEC group had an emergency operation for intestinal perforation and severe intestinal necrosis.
Gene Expression Profiles
Using a targeted Human Blood Coagulation quantitative PCR array, 64 of 94 genes showed detectable expression levels across the subjects studied (Supplementary Table, http://links.lww.com/MPG/A654). A total of 12 of 64 genes (19%) were differentially expressed among the NEC neonates (n = 11) when compared with the healthy controls (n = 10) (Table 2). In the comparison between NEC and healthy controls, 5 key genes of the coagulation pathways were significantly upregulated, and 7 were downregulated. For several of these genes, the magnitude of change was the greatest among the subgroup with more severe NEC (Bell stage 3-NEC3, n = 5), suggesting that these may be worth following up as potential biomarkers of disease progression (Fig. 1). The physiologic functions of all the significantly altered genes, related to the coagulation and anticoagulation systems, are illustrated in Table 3.
We also investigated whether there were altered gene expression levels in the patients with sepsis compared with the healthy controls. Only 1 gene was significantly downregulated: HNF4A (GPR FC −5.14, P = 0.006). In the patients with sepsis, although not statistically significant, there was a trend in the same direction for many of the genes altered in the patients with NEC (data not shown). We did observe a significant difference, however, in the expression of factor II (thrombin) receptor-like 1 (F2RL1) comparing sepsis with patients with NEC (GPR FC −2.50, P = 0.01) (Fig. 2). In Figure 3, we have summarized the net procoagulant effect induced by the altered gene expressions we have found in neonates with NEC compared with controls.
Our study found significant alterations in the gene expressions of coagulation and anticoagulation proteins in neonates with NEC. In particular, we have identified 12 potential biomarkers of coagulation pathway abnormalities and disease progression in infants with NEC stages 2 and 3. Upregulation and downregulation of these significant altered genes showed an overall procoagulant status, reduction in fibrinolysis, and impaired endothelial regeneration in NEC.
Our hypothesis was based on evidences, present in adult sepsis, about the tight cross-talk between endothelial damage, inflammation, and coagulation (10–12). Activation of the coagulation pathways in sepsis is associated with fibrin deposition, in small to midsize vessels, with the final effect of microvascular thrombosis, hypoperfusion, and ischemia affecting various organs such as lungs, kidneys, liver, and soft tissue (13). Adult studies have shown that the coagulation system is altered in patients with severe inflammation, following 3 main mechanisms: increased intravascular presence of tissue factor (TF), reduced activity of the anticoagulant pathways, and inhibition of the fibrinolysis (14,15). TF plays a central role in the initiation of inflammation-induced coagulation. TF overproduction is because of the loss of blood vessel integrity or its active production by endothelial cells and circulating blood cells as a defense mechanism to fight microbes (16). Our data suggested that in neonates with NEC there may be a significant production of TF because of an overexpression of neutrophil elastase (ELANE), which encodes an essential protein involved in innate immunity, neutrophil function, and activation of the coagulation pathway. ELANE, a serine protease responsible for local activation of the coagulation cascade, was significantly upregulated in NEC compared with normal infants. The upregulation of ELANE is a common event in adult severe sepsis, and its protein product drives the “immunothrombosis” cascade as part of the host defense to prevent microbial dissemination at the level of the entry site (17). Briefly, immunothrombosis is a process carried out by neutrophils that first accumulate and adhere to the damaged vascular endothelium, and then release neutrophil extracellular traps (NETs) to remove pathogens thanks to the extracellular delivery of TF, by direct activation of factor XII (also significantly upregulated in the NEC group in our data), by platelets adhesions and stimulation, with the consequent activation of the extrinsic pathway of the coagulation and clot formation (17–19). ELANE has also been implicated in the control of platelet responses to vascular injury by enhancing matrix thrombogenicity and by the cleavage of von Willbrand factor (20). This scenario fits well with the clinical picture seen in premature infants with NEC. Gut microbial aggressiveness and their capacity to penetrate into the mesenteric vessels enhance a violent and exaggerated mucosal and endothelial inflammation with the activation of ELANE and local immunothrombosis leading to the formation of clots and various degrees of intestinal ischemia. Moreover, factor XII has a key role in the initiation of blood coagulation via the intrinsic pathway of coagulation (via factors VII and XIIa). The upregulation of the factor XII gene in NEC further supports a significant activation of the coagulation, mainly at the level of the damaged endothelium, leading to clot formation and thrombosis.
A second important role of neutrophil-derived granule proteins, especially ELANE, is their participation in thrombus formation by inhibiting tissue factor pathway inhibitor and anticoagulants such as antithrombin and activated protein C (21). This could be another mechanism leading to both local (ie, mesenteric thrombosis, patchy intestinal ischemia, and perforation) and systemic (ie, organ microthrombosis, DIC, and multiorgan failure) coagulopathy (22).
Inhibition of fibrinolysis is the third key mechanism present in severe inflammation to further imbalance the coagulation cascade. Although the fibrinolytic and coagulation systems perform opposite functions, regulation between these 2 pathways is critical to prevent inappropriate clot formation and to enhance thrombus removal. It is also known that inflammation markedly suppresses fibrinolytic activity via endotoxin-induced tumor necrosis factor (TNF) (15,23). In septic shock, there is a substantial downregulation of tissue plasminogen activator (PLAT) as a response to elevated levels of TNF. The suppressive effect of TNF on PLAT together with its ability to induce plasminogen activator inhibitor (PAI-1) is the reason for TNF being a potent antifibrinolytic agent. Briefly, PAI-1 is the principal inhibitor of PLAT and urokinase, which are the activators of plasminogen, a key protein to breakdown the blood clots (fibrinolysis). Our data showed a downregulation of PLAT in advanced NEC, which could impair fibrinolysis causing inadequate fibrin removal, microvascular thrombosis with reduced gut perfusion, and ischemia. Mice with a deficiency of PLAT expression have more extensive fibrin deposition in their organs when challenged with endotoxin, whereas PAI-1 knockout mice, in contrast to wild type controls, have no microvascular thrombosis upon endotoxin administration (24). Finally, changes in the key players in the protein C pathway, such as PROS1 and PROCR, have been associated with increased risk of thrombosis in sepsis (25).
In infants with NEC, in addition to the presented alterations of the coagulation systems, it is important to consider the significant endothelial damage localized at the level of the mesenteric vessels. An overall procoagulant status together with intestinal endothelial damage could be the main reason for disease progression, intestinal ischemia, and DIC (26). In our patients with NEC, we observed an increase in CD63 gene expression, which encodes a protein expressed on the platelet surface and secreted into plasma when platelets become activated. CD63 levels have recently been shown to be elevated in a mouse sepsis-induced DIC model (27). A second possible biomarker related to the progression of the endothelial dysfunction is MFG-E8 (milk fat globule-endothelial growth factor 8). It has been shown that, in septic mice, intestinal MFG-E8 expression is downregulated leading to intestinal injury, impaired enterocyte migration, and deficient mucosa regeneration (28). This gene was found highly downregulated in our NEC infants, and MFG-E8 should be further investigated as a possible future therapeutic approach to reduce tissue injury and to modulate the overactive immune system. A third important regulator of the endothelial barrier function and blood coagulation is F2RL1, also known as protease activated receptor 2 (PAR2) (29). It has also been shown that PAR2 activation plays a protective role, with less mucosal damage and better gastrointestinal motility, in a mice model of gut ischemia/reperfusion (30). Levels of PAR2 were significantly decreased in patients with NEC, compared with both control and sepsis groups.
STRENGTHS AND LIMITATIONS
Our pilot study is original in its hypothesis, with rigorous methods and a prospective recruitment with 2 comparison groups. The main limitation of the study was the small sample size. It was not possible to compare the results from the gene expression with a complete clotting screening because of the large volume of blood required to be taken in small premature infants. As a pilot study, further research is needed to confirm the data in a multicenter and prospective fashion and to strengthen the findings with protein analysis. A large cohort of patients will allow more accurate subgroup analysis to isolate the most significant peripheral biomarkers of coagulopathy present at an early stage of disease and to define proteins associated with disease progression.
The severe inflammation associated with neonatal NEC leads to alterations in the coagulation and anticoagulation pathways with a net procoagulant effect. In particular, key genes involved in the activation of the coagulation cascade (ELANE and F2RL1), endogenous fibrinolysis (PLAT, PROS1, and PROCR), and endothelial dysfunction (CD63 and MFGE8) should be further investigated as important biomarkers of coagulopathy. This novel research field will unveil new molecular mechanisms and possible treatments to cure NEC.
The authors thank all the participating patients and parents, and the nursing staff in the Neonatal Unit at St George's University Hospitals NHS Foundation Trust for facilitating enrolment and blood sample collections.
1. Neu J, Walker WA. Necrotizing enterocolitis
2. Jacob J, Kamitsuka M, Clark RH, et al Etiologies of NICU deaths. Pediatrics
3. Arnold PD. Coagulation
and the surgical neonate. Paediatr Anaesth
4. Revel-Vilk S. The conundrum of neonatal coagulopathy. Hematology Am Soc Hematol Educ Program
5. Song R, Subbarao GC, Maheshwari A. Haematological abnormalities in neonatal necrotizing enterocolitis
. J Matern Fetal Neonatal Med
6. Levi M, Schultz M, van der Poll T. Sepsis and thrombosis. Semin Thromb Hemost
7. Walsh MC, Kliegman RM. Necrotizing enterocolitis
: treatment based on staging criteria. Pediatr Clin North Am
8. Oeser C, Lutsar I, Metsvaht T, et al Clinical trials in neonatal sepsis. J Antimicrob Chemother
9. Akilesh S, Shaffer DJ, Roopenian D. Customized molecular phenotyping by quantitative gene expression
and pattern recognition analysis. Genome Res
10. Hackam DJ, Afrazi A, Good M, et al Innate immune signaling in the pathogenesis of necrotizing enterocolitis
. Clin Dev Immunol
11. Boisramé-Helms J, Kremer H, Schini-Kerth V, et al Endothelial dysfunction in sepsis. Curr Vasc Pharmacol
12. Levi M, van der Poll T. Two-way interactions between inflammation
. Trends Cardiovasc Med
13. Levi M, van der Poll T, Schultz M. Systemic versus localized coagulation
activation contributing to organ failure in critically ill patients. Semin Immunopathol
14. Levi M, van der Poll T. Inflammation
. Crit Care Med
15. Esmon CT. The interactions between inflammation
. Br J Haematol
16. Osterud B, Rao LV, Olsen JO. Induction of tissue factor expression in whole blood-lack of evidence for the presence of tissue factor expression on granulocytes. Thromb Haemost
17. Van der Poll T, Herwald H. The coagulation
system and its function in early immune defense. Thromb Haemost
18. Engelmann B, Massberg S. Thrombosis as an intravascular effector of innate immunity. Nat Rev Immunol
19. Von Bru[Combining Diaeresis]hl ML, Stark K, Steinhart A, et al Monocytes, neutrophils, and platelets cooperate to initiate and propagate venous thrombosis in mice in vivo. J Exp Med
20. Wohner N, Kovács A, Machovich R, et al Modulation of the von Willebrand factor-dependent platelet adhesion through alternative proteolytic pathways. Thromb Res
21. Iba T, Miki T, Hashiguchi N, et al Is the neutrophil a “prima donna” in the procoagulant process during sepsis? Crit Care
22. Gardiner EE, Andrews RK. Neutrophil extracellular traps (NETs) and infection-related vascular dysfunction. Blood Rev
23. Medcalf RL. Fibrinolysis, inflammation
, and regulation of the plasminogen activating system. J Thromb Haemost
24. Yamamoto K, Loskutoff DJ. Fibrin deposition in tissues from endotoxin-treated mice correlates with decreases in the expression of urokinase-type but not tissue-type plasminogen activator. J Clin Invest
25. Hayashi T, Suzuki K. Changes of expression of the protein C pathway components in LPS-induced endotoxemia—implication for sepsis. Cardiovasc Hematol Disord Drug Targets
26. Boegehold MA. Endothelium-dependent control of vascular tone during early postnatal and juvenile growth. Microcirculation
27. Song J, Hu D, He C, et al Novel biomarkers
for early prediction of sepsis-induced disseminated intravascular coagulation
in a mouse cecal ligation and puncture model. J Inflamm (Lond)
28. Bu HF, Zuo XL, Wang X, et al Milk fat globule-EGF factor 8/lactadherin plays a crucial role in maintenance and repair of murine intestinal epithelium. J Clin Invest
29. Kaneider NC, Leger AJ, Agarwal A, et al “Role reversal” for the receptor PAR1 in sepsis-induced vascular damage. Nat Immunol
30. Cattaruzza F, Cenac N, Barocelli E, et al Protective effect of proteinase-activated receptor 2 activation on motility impairment and tissue damage induced by intestinal ischemia/reperfusion in rodents. Am J Pathol
biomarkers; coagulation; gene expression; infants; inflammation; necrotizing enterocolitis
Supplemental Digital Content
© 2016 by European Society for Pediatric Gastroenterology, Hepatology, and Nutrition and North American Society for Pediatric Gastroenterology,