This study was approved by the Ethical Committee at the Institute for the Care of Mother and Child in Podolí, Prague, Czech Republic. All mothers and healthy volunteers signed a written informed consent before sample collection. Cord and peripheral blood were collected at the Institute for the Care of Mother and Child in Prague, the Czech Republic. Umbilical cord blood (20-30 ml) was drawn from the umbilical vein with a sterile needle into a 50 ml flask with heparin (10U/ml). During the standard pre-labour examination, maternal peripheral blood (2-5 ml) was drawn into K₂EDTA or heparin-containing tubes. Blood samples were stored at RT in a dark box until pick-up. Peripheral blood (2-5 ml) of healthy women volunteers was drawn into K₂EDTA or heparin-containing tubes. Details of the studied cohort of mothers and children are listed in Table 1 .

Whole blood was stained with anti-human CD14 (clone 61D3, Invitrogen, USA), CD15 (clone HI98, BioLegend, USA), CD16 (clone CB16, Invitrogen), CD62L (clone DREG-56, BioLegend, USA), and CD64 (clone 10.6, BioLegend, USA). After 15 min incubation in the dark, erythrocytes were lysed with 10x diluted Lysis Buffer (QIAGEN, Netherlands), and samples were washed twice with PBS. Samples were measured on BD FACS CantoTM II Flow Cytometer (BD Biosciences, USA) equipped with the BD FACS Diva Software v6.1.2 and analyzed with FlowJo v10.10.0. (FlowJo LLC, USA) software. Gating strategy is shown in Supplementary Figures 1 , 2 .

For the isolation of the whole neutrophil population, the EasySepTM Direct Human Neutrophil Isolation Kit (STEMCELL Technologies, Canada) was used. Neutrophils were isolated according to the manufacturer’s instructions with slight modifications to increase cell yield and purity. In the first step, 50 μl of 0.5 M EDTA/1 ml blood was added, and the first incubation was extended to 6 min. Isolated cells were washed once with PBS and counted. Cell viability and purity were assessed by Trypan blue and flow cytometry, respectively. The purity of the isolated neutrophil population was higher than >96%. Mononuclear cell fraction was obtained as follows. The whole cord or peripheral blood was diluted 1:1 with PBS and layered carefully over Ficoll-PaqueTM PLUS (density 1.077 g/l) in a conical tube. Density-gradient centrifugation was performed at 500 x g for 30 min. After centrifugation, blood was separated into CBMC or PBMC-rich layer and granulocyte-rich erythrocyte pellet. Cells from both fractions were washed twice with PBS and counted. LDN were isolated from CBMC using EasySepTM Human CD15 Positive Selection Kit (STEMCELL Technologies, USA) according to the manufacturer’s instructions. HDN were isolated using EasySepTM Direct Human Neutrophil Isolation Kit (STEMCELL Technologies, USA) from the erythrocyte fraction as described above.

Total RNA from isolated neutrophils was extracted using the RNeasy® Mini Kit (QIAGEN, Netherlands) according to the manufacturer’s instructions with the addition of the optional RNase-Free DNase I (QIAGEN) incubation step to prevent genomic DNA contamination. The purity and concentration of the RNA were determined using NanoDrop 1000 Spectrophotometer (Thermo Scientific, USA). Total RNA was reverse transcribed to cDNA using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems) according to the manufacturer’s instructions. The relative expression of selected genes was performed using the Luna® Universal qPCR Master Mix (New England BioLabs, UK) and TaqMan® probes (Applied Biosystems, USA) and measured on the LightCycler® 480 II Real-Time PCR System (Roche Molecular Systems, USA). All tested genes and their respective probe assay IDs are listed in Table 2 .

The phagocytic capacity of neutrophils quantified as relative myeloperoxidase (MPO) activity was tested by the FagoFlowExKit (EXBIO, CZ) according to the manufacturer’s instructions with the addition of anti-human anti-CD15 antibody (clone HI98, BD Biosciences) for precise determination of neutrophils. The method utilizes the oxidation of dihydrorhodamine (DHR) 123 to fluorescent rhodamine 123 by activated MPO. The fluorescence was measured using BD FACS CantoTM II Flow Cytometer (BD Biosciences, USA) equipped with the BD FACS Diva Software v6.1.2 and analyzed with FlowJo 7.6.5. (FlowJo LLC, USA) software. The gating strategy and representative histograms are shown in Supplementary Figure 3 . The relative MPO activity was calculated based on mean fluorescent intensity (MFI) as follows:

(MFI_S – MFI of sample stimulated with E. coli, MFI_P – MFI of positive control, MFI_N – MFI of negative control) to account for the background fluorescence.

Flow cytometry results were standardized using Cytometer Setup and Tracking beads (BD Biosciences, USA) and analyzed by FlowJo software using appropriate compensation controls (single staining, compensation beads, isotype control) and fluorescence minus one (FMO) control to set proper gating strategy. Transcript expression of the target gene was quantified relative to endogenous control (housekeeping gene) actin beta (ACTB) mRNA levels using the 2^-ΔΔCt method as described earlier (42). For flow cytometry analysis and qPCR analysis, parametric statistical tests (unpaired Students t-test) were utilized. Acquired data were statistically evaluated and graphically processed using GraphPad Prism 8 (GraphPad Software Inc.) software. Pearson correlation test was employed for the correlation of gene expression with neutrophil subpopulation. Statistically significant differences among the groups were considered when p-values were lower than 0.05.

Based on previous reports indicating the presence of immature neutrophils in CB (40, 41), we performed phenotype analyses of selected neutrophil cell surface markers reflecting neutrophil maturational and activation status. Neutrophils in CB accounted for only about 40% of total cells compared with proportion of neutrophils in HV (68%) and PB (76%) ( Figure 1A ), whereas their frequency was only slightly elevated in the PB of mothers compared with HV ( Figure 1A ). The frequency of neutrophils identified as CD14^-CD15⁺ granulocytes (43, 44) was lower in CB ( Figure 1B ), mainly due to the enrichment in CD14^-CD15^low granulocytes ( Figure 1D ), a granulocyte population not previously described (gating strategy in Supplementary Figures 1 , 2 ). Monocytes, identified as CD14⁺CD15^- cells (45), did not differ between the study groups ( Figure 1C ). Neutrophils were further gated based on the expression of CD16 (FcγRIIIB) and CD64 (FcγRI), markers considered to reflect neutrophil maturity, and CD62L (L-selectin). As previously reported (41), CD16^lowCD64^high neutrophils were increased in CB ( Figure 1E ), with a similar trend observed in the case of CD16^lowCD62L⁺ ( Figure 1F ). Most of the neutrophils in the control group exhibited CD16⁺CD64^low and CD16⁺CD62L⁺ phenotype ( Figures 1E, F ), suggesting a mature non-activated neutrophil population. Neutrophils in PB of mothers displayed increased expression of CD64 ( Figure 1E ), suggesting neutrophil activation during pregnancy. To address whether CD16^low neutrophils have a low-density characteristics, identical staining and gating strategies were applied to isolated CBMC or PBMC. An increase in the frequency of CD16^low neutrophils in CBMC was observed ( Figures 2B, C ), more prominent when cells were gated as CD16^lowCD64^high ( Figure 2B ). The other subpopulations did not differ between the experimental groups ( Figures 2D, E ), confirming the nature of CD16^low neutrophils as LDN. To better identify the particular subpopulation based on the cell surface presence of CD16, alternative gating was performed (see Supplementary Figure 2 ) and three neutrophil subsets were identified according to CD16 (i.e. CD16 negative (CD16^-); CD16 dim (CD16^int) and CD16 bright (CD16⁺)). Proportion of particular neutrophil subsets based on CD16 distribution is presented in Supplementary Figure 4 . This alternative gating provide us with remarkably elevated proportion of CD16^-CD64^high neutrophils within CD14^-CD15⁺ cells in CB compared with HV. CD16^intCD64^high neutrophils within CD14^-CD15⁺ cells were significantly increased in CB compared with HV. CD16⁺CD64^high neutrophils were the most abundant in PB in comparison with HV and CB. Proportion of CD16⁺CD64^high neutrophils was elevated in CB compared with HV. No difference was observed in proportion of CD16^-CD64^low among HV, PB and CB. On contrary, CD16⁺CD64^low were significantly increased in HV compared to both PB and CB. However, CD16⁺CD64^low was significantly elevated in CB compared to PB, Supplementary Figure 4A . No difference in proportion of CD16^-CD62L⁺ and CD16^intCD62L⁺ among HV, PB and CB has been detected. Subset of CD16⁺CD62L⁺ was significantly higher in HV compared to CB. CD16^-CD62L^- were significantly increased in CB compared with HV, Supplementary Figure 4B .

Next, we determined the expression of genes linked to antimicrobial ( Figure 3 ) or suppressive ( Figure 4 ) functions in all three study groups. The association of human defensin alpha 3 in anti-infectious immunity has been highlighted (46). On the other hand, elevated presence of alpha defensin 3 (DEFA3) in vaginal fluid after premature preterm rupture of membranes pointed to neonatal inflammation (47). In our study, significantly increased gene expression of DEFA3 was observed in cord blood neutrophils compared to neutrophils of adults (both HV and PB). Importantly, PB neutrophils showed promoted gene expression of DEFA3 compared to HV suggesting increased readiness to fight against infection, Figure 3A . The marker linked with the function of neutrophils is myeloperoxidase which stabile levels during pregnancy can be considered as a biomarker of non-complicated course of pregnancy (48). Cord blood neutrophils shows superior gene expression of MPO compared to adult neutrophils (both HV and PB). Interestingly, PB neutrophils exerted elevated gene expression of MPO compared to HV confirming higher anti-microbial capacity, Figure 3B . There was no difference in the expression of MMP9 between the cord and maternal peripheral blood, possibly explained by the specific increase in the expression of metalloproteases during pregnancy and labour (49). Another peptide with antimicrobial capacity secreted by neutrophils is defensin alpha 4 (DEFA4) (50). We have observed significantly increased gene expression of DEFA4 in cord blood neutrophils compared to HV and PB. Neutrophil elastase (ELANE) is a marker of NET formation in health and disease (51). Here, we showed significantly elevated gene expression of ELANE in cord blood neutrophils compared to both HV and PB. Lactotransferrin (LTF) is playing a pivotal role in inflammatory homeostasis (52). Therefore, we have evaluated gene expression of LTF in cord blood neutrophils and confirmed superior expression of LTF in cord blood neutrophils compared with HV and PB highlighting prerequisite of cord blood neutrophils exert antimicrobial functions. PB neutrophils poses still elevated gene expression of LTF compared to HV neutrophils. In summary, cord blood neutrophils display highly increased expression of genes coding for proteins playing an important role in antimicrobial response (e.g., myeloperoxidase, defensins, and neutrophil elastase) compared to neutrophils from the mothers or healthy volunteers ( Figures 3A, B, D–F ).

Successful pregnancies were associated with neutrophil subpopulations with prone expression of arginase (ARG1). Interestingly, in the case of genes associated with immune suppressive properties, there was an increase in the expression of ARG1 ( Figure 4A ) in CB neutrophils but limited expression of CD274 encoding PD-L1, that was highly elevated in neutrophils from maternal PB ( Figure 4 ). PD-L1 is a marker identified on neutrophils and its expression is associated with undergoing inflammatory process (53, 54). Therefore, it is not surprising that only marginal PD-L1 expression has been detected in cord blood neutrophils of healthy neonates. Along with the increase in CD274, maternal neutrophils also displayed elevated expression of PDCD1 ( Figure 4C ), a gene translated to the receptor (Programmed cell death protein 1, PD-1) for PD-L1, potentially indicating a negative feedback loop exerted in maternal neutrophils. Only slight increase in the expression of tumor necrosis factor (TNF) in CB neutrophils was observed in comparison to HV neutrophils ( Figure 4D ), indicating elevated antimicrobial activity. We have focused on gene expression of immunoregulatory markers as well. Interleukin 10 (IL10) is well known cytokine with immunoregulatory/immunosuppressive function preventing undesirable immune responses (55). As another marker suggesting potential immunoregulatory capacity of neutrophils was selected indol amine 2,3 dioxygenase (IDO) which expression has been shown limited in more reactive cells (56). Surprisingly, no difference was observed in the expression of these genes of interest ( Figures 4E, F ). Next, the expression of genes of interest was compared in cord blood of children delivered either by vaginal delivery (VD) or caesarean section (CS) ( Supplementary Figure 5 ). Interestingly, children delivered by CS displayed elevated expression of defensins and myeloperoxidase. In contrast, children delivered by VD displayed higher expression of genes coding for IL-10 and PD-L1 proteins, indicating elevated proinflammatory settings in children born by CS.

Next, we set to elucidate whether the higher relative frequency of LDN in CB might be responsible for the elevated expression of genes associated with antimicrobial function. To this aim, we isolated LDN and HDN from CB and performed gene expression analysis of candidate genes of interest. The data did not reach statistical significance ( Figures 5A, B ), probably due to the low number of samples. Only non-significant tendency toward higher expression of genes associated with antimicrobial response was observed in LDN ( Figure 5A ). In order to overcome the difficulties associated with cell isolation, correlation analysis was performed. The frequency of CD16^lowCD64^high neutrophils in CB was correlated with the relative gene expression of examined genes in neutrophils isolated from whole CB ( Figures 5C, D ). Despite quite low value of correlation coefficient, the correlation reaches significance in three of the candidate genes (DEFA3, DEFA4 and MPO, Figure 5C ) indicating that the presence of CD16^low LDN is responsible for the observed increased expression of genes associated with antimicrobial response.

To elucidate whether the presence of CD16^low neutrophils affects the functional properties of neutrophils in CB, we studied their activation after stimulation with inflammatory stimulus Escherichia coli. To this end, we utilised FagoFlowEx Kit (EXBIO) employing dihydrorhodamine 123 oxidation to fluorescent rhodamine 123 by activated MPO (representative histograms shown at Figure 6A ). Staining for CD15 employed for higher stringency of gating of neutrophil population did not exert any effect on their activation status ( Figure 6B ). As depicted in Figure 6C , the activation of neutrophils by inflammatory stimulus did not differ among the study groups.