Study design

The first study was an in vivo investigation of UCB administration at a time of heightened systemic inflammation induced by brief high VT injurious ventilation in an established fetal ovine model to assess the effects on white matter inflammation and injury. Animal experimental procedures were approved by the Monash Medical Centre Animal Ethics Committee A, Monash University (MMCA/2018/05). All experiments were conducted in accordance with guidelines established by the National Health and Medical Research Council of Australia. All experiments were performed in accordance with the ARRIVE guidelines 2.0 (Supplementary Data File 1) [23].

We then conducted an in vitro study to investigate the characteristic alterations to cells isolated from human-derived UCB following exposure to an inflammatory environment induced by cytokine-spiked media. All human-derived samples were collected with human ethics approval from Monash Health Human Ethics Committee (12387B). Written informed consent was obtained from all participants prior to collection of UCB.

In vivo sheep preparation

The use of these animals and relevant methodology have previously been published [24, 25] and experimental protocols unique to this study are reported in detail below. Prior to surgery, pregnant ewes at 125 ± 1 day gestation (term ~ 148 days) were randomly assigned to a SHAM (n = 5), VENT (n = 7) or VENT + UCB (n = 7) group. Male to female ratio for fetuses within each group was SHAM, 3:2; VENT, 5:2; and VENT + UCB, 4:3. As previously described [24, 25], sterile surgery was performed with a midline laparotomy on pregnant ewes to expose the fetus head and neck for instrumentation of polyvinyl catheters in the left carotid artery (for arterial blood sampling) and jugular vein (for UCB administration), and an ultrasonic flow probe (3PS; Transonic Systems, NY, USA) around the right carotid artery to measure carotid blood flow (CBF). To allow ventilation, the fetus’ chest was exteriorised and the fetus was intubated with a cuffed endotracheal tube (ID 4.0-4.5 mm; Smiths Medical, UK) and lung liquid passively drained.

Ventilation for the VENT and VENT + UCB groups was conducted under sterile conditions using a neonatal positive pressure ventilator (Babylog 8000+, Dräger, Lübeck, Germany) as described previously [24, 25]. Briefly, the fetus was ventilated for 15 min with a high tidal volume (VT) strategy targeting 12–15 mL/kg (normal VT: 5–7 mL/kg). SHAM animals remained intubated and exteriorised for 15 min without mechanical ventilation. CBF and VT were continuously measured for the 15 min and digitised using a PowerLab A-D converter and stored on disk using LabChart7 software (AD Instruments, Bella Vista, NSW, Australia). CBF could not be recorded from one lamb from the VENT group due to technical error. Arterial blood gas was measured at 0 (pre-ventilation), 5, 10 and 15 min of ventilation (ABL80 FLEX, Radiometer Medical ApS, Denmark).

After 15 min of ventilation, the carotid flow probe was removed, and fetuses were extubated and returned to the uterus. The fetal jugular vein and carotid artery catheters were externalised through the ewe’s right flank. All incision sites were sutured closed and the ewe and fetus allowed to recover. Analgesia (buprenorphine; 0.3 mg i.m.; Temgesic; Reckitt Benckiser, UK and transdermal fentanyl patch; 75 µg/h; Janssen-Cilag, NSW, Australia) was administered for post-surgery pain relief. Arterial blood gas measurements and arterial samples were collected in heparinised tubes regularly throughout the 24 h study period.

UCB cells treatment

UCB cells were obtained from healthy term lambs (141 days gestation of term 148 days) undergoing procedures in a separate study, using UCB collection, storage and preparation protocols as previously described [9, 26]. On the day of cell administration, UCB cell samples were rapidly thawed, washed and resuspended in media (Gibco, USA), and cell number and viability measured (minimum 75% viability accepted). At 1 h post-ventilation, ~ 80 million allogeneic UCB cells suspended in 3 mL of phosphate-buffered saline (PBS; Gibco, MA, USA) were administered to VENT + UCB fetuses via the jugular vein catheter, which was then flushed with sterile heparinised saline to ensure all cells were administered [25]. This dosage and administration regime was based on previous studies reporting reduced cerebral inflammation, white matter injury, oxidative stress and cell death after hypoxia-ischemia in fetal sheep [9].

Magnetic Resonance Imaging (MRI) and post-mortem brain collection

Twenty-four hours post-in utero ventilation, ewes were anaesthetised and the fetus exteriorised via the same incision site. All lambs were dried, intubated, and stabilised with a standard protective ventilation strategy (Babylog 8000+, Dräger) using volume guarantee set at 7 mL/kg and PEEP of 5 cmH2O. Once the lambs were stabilized, the umbilical cord was clamped and the lambs delivered. Surfactant was administered (100 mg/kg, CurosurfR, Chiesi Pharma, Italy) and mechanical ventilation was provided targeting 5 mL/kg. Lambs were then transferred to the 3T MRI system (Siemens Skyra, Erlangen, Germany) at Monash Biomedical Imaging (Clayton, Australia), with a 15-channel radio frequency transmitter/receiver knee coil. Lambs were placed in supine position and ventilation maintained using a BabyPAC portable and MR-compatible ventilator (Pneupac, Smiths Medical, UK). Due to technical issues, MRI results from one lamb from the VENT group and two lambs from the VENT + UCB group were not suitable for analysis; thus the revised group numbers for this analysis were VENT, n = 6; VENT + UCB, n = 5. MRI acquisition protocols for T2-weighted image and Diffusion Weighted Imaging (DWI) were conducted as previously published [24, 27, 28]. Total acquisition time was around 40 min. Lambs were then euthanised (sodium pentobarbitone > 100 mg/kg i.v.; Valabarb Euthanasia Solution; Jurox, NSW, Australia) immediately after scanning was completed. Ewes were similarly euthanized immediately after delivery of the lambs.

At post-mortem, the lamb’s brain was removed and the cerebrum hemisected along the medial longitudinal fissure. The periventricular and subcortical white matter (PVWM and SCWM respectively) of the left cerebral hemisphere were dissected and snap-frozen in liquid nitrogen. The right cerebral hemisphere was cut coronally into 5 mm blocks and immersion fixed in 10% neutral buffered formalin (Amber Scientific, WA, Australia) for paraffin-embedding.

MRI data analysis

All data were analysed using the FMRIB Software Library (FSL, FMRIB, Oxford, UK [29]). Pre-processing steps started with extracting brains and creating brain mask using Brain Extraction Tool. All masks were manually brushed and cleaned thoroughly to remove skull and areas other than brain. All DWI data were then corrected for eddy-current distortion using eddy, and then DTI parametric maps of fractional anisotropy (FA), axial diffusivity (AD), radial diffusivity (RD), and mean diffusivity (MD) created with the Diffusion Toolbox of FMRIB. RD is the mean of the second and third eigenvalues (L2 and L3). Region of interests (ROI) were defined on high resolution T2 images to maximise accuracy. Selected ROI: the periventricular white matter (PVWM), the frontal white matter (FWM) was identified using The Sheep Brain Atlas from Michigan State University (Brain Biodiversity Bank, National Science Foundation). High-resolution T2 images of the brains along with the ROIs were co-registered to DWI images using the Linear Image Registration Tool (FLIRT [29]) as previously published [24, 27, 28]. The mean value from 10 adjacent slices within the given ROI volume was calculated; within-animal ROI variability was < 16.8%. MRI examination, image processing and analysis of the structural MR images were performed by persons blinded to group allocations.

Immunohistochemistry and quantitative analysis

Coronal sections (8 μm) at the level of the frontal, parietal, temporal, and occipital lobes (4 slides/animal/antibody) were stained with rabbit anti-ionised calcium binding adapter molecule-1 (Iba-1; 1:1500, Wako Pure Chemical Industries, Osaka, Japan) to identify microglia; rabbit anti-sheep serum (1:1000, Sigma-Aldrich, USA) to identify vascular extravasation of protein; mouse anti-oligodendrocyte transcription factor-2 (Olig2; 1:1000, Merck, Darmstadt, Germany) for oligodendrocytes; rat anti-myelin basic protein (MBP; 1:200; Merck, Darmstadt, Germany) for mature myelin and oligodendrocytes; and rabbit anti-glial fibrillary acidic protein (GFAP; 1:400, Sigma-Aldrich, USA) to identify astrocytes as described previously [30]. Briefly, sections were pre-treated with citrate buffer (pH 6.0) in a microwave oven. Sections were incubated with secondary biotinylated IgG antibody raised against the corresponding animal secondary (1:200; Vector Laboratories, CA, USA) and reacted using the Vectastain Elite ABC Kit (Vector Laboratories, CA, USA). A ‘Terminal deoxynucleotidyl transferase dUTP nick end labelling’ (TUNEL) assay (ApopTag® Peroxidase In Situ Apoptosis Detection Kit, Millipore, USA) was also conducted according to manufacturer’s instructions. Negative controls (primary antibody omitted) demonstrated no positive staining.

Analyses were conducted at equivalent sites within the PVWM and SCWM of sections from the frontal, parietal, temporal, and occipital lobes of each lamb. Slides were coded and the assessors blinded to the groups. Analysis was conducted using either ImageScope (Aperio Technologies; Leica Biosystems, Germany) or ImageJ software (National Institutes of Health, USA) as described previously [30]. Cell density (cells/mm2) of Iba-1+ microglia (resting, ameboid and total), GFAP+ astrocytes, TUNEL+ cells, Olig2+ oligodendrocytes, MBP+ cells and number of blood vessel profiles with protein extravasation were manually quantified in 3 non-overlapping fields in the PVWM and 6 non-overlapping fields in the SCWM from 2 separate gyri in each section (area = 0.14 mm2). The fractional area coverage (%) of Iba-1, GFAP and MBP immunoreactivity was assessed within the same fields. For Iba-1 immunostaining, the total number of microglial aggregations and the fractional area coverage of these aggregations within the PVWM and SCWM were also assessed. Data from all fields were averaged for each brain region then averaged across subjects in each experimental group.

Magpix cytokine analysis

Arterial plasma was collected following ventilation (0 [pre-ventilation], 15 min [end of ventilation] and 1, 3, 6, 12 and 24 h [post-ventilation recovery]) and used to quantify levels of IFNγ, IL-1β, IL-10, TNFα, IL-8 and IL-6 using a Milliplex MAP bovine cytokine magnetic bead panel assay kits according to manufacturer’s instructions (cat#: CYT1-91 K; MerckMillipore, USA). In brief, 96-well plates were coated with samples, assay buffer, serum matrix and antibody-immobilised beads. Plates were incubated overnight at 4 °C, washed and incubated with detection antibodies for 1 h. Streptavidin-phycoerythrin was added to the plates for 30 min. Sheath fluid was added to the plates and cytokine concentrations were quantified using a Bio-Plex MAGPIX® Multiplex reader with xPOTENT® software (Bio-Rad, CA, USA).

In vitro UCB mononuclear cell isolation

For in vitro studies, human term UCB samples were obtained from women with uncomplicated pregnancies undergoing elective caesarean section at term (> 37 weeks gestation). Women gave written, informed consent for the collection of their UCB. After clamping of the cord and delivery of the placenta, UCB was collected from the umbilical vein using blood collection bags containing anticoagulant (Macopharma, Tourcoing, France). On average, ~ 100 mL of UCB was collected. UCB was stored at room temperature for up to 48 h until processing and cell isolation. To obtain the mononuclear layer of cells, UCB was transferred to 50 mL falcon tubes and diluted 1:1 with PBS. This was centrifuged at 3100 rpm for 12 min, with no brake. The mononuclear cells were separated and washed in 20 mL PBS and centrifuged at 800 g for 5 min to isolate a cell pellet. Red blood cell lysis buffer (ammonium chloride, potassium bicarbonate and EDTA dissolved in double distilled water; Sigma-Aldrich, USA) was added for 3–5 min to lyse excess red blood cells. The reaction was stopped with excess media (16.5% fetal bovine serum and DMEM/F12; Gibco, USA), followed by centrifugation at 400 g for 5 min and the supernatant was then aspirated. Cell viability was determined using trypan blue exclusion dye (Gibco), and cells were counted with a hemocytometer. The mononuclear cells were then either used for magnetic bead separation of individual cell types or cryopreserved for later use. For cryopreservation, UCB mononuclear cells were frozen at a density of 20 × 106 cells/mL; in 40% complete media (DMEM/F12, 16.5% FBS, 1% antibiotics), 50% fetal bovine serum (FBS; Gibco) and 10% dimethyl sulfoxide (DMSO; Sigma Aldrich). Cells were then transferred to freezer tubes and left in a freezing container (MrFrosty, Thermo Fisher Scientific) overnight at -80°C, following which they were transferred to liquid nitrogen. To thaw, sample tubes were quickly removed from liquid nitrogen and placed directly into a 37°C water-bath until thawed. Samples were washed to remove DMSO, and cell counts and viability were determined.

In vitro stimulation of cells

For stimulation of cells, three separate donors were used for each cell type and each stimulation experiment was performed in triplicate. Briefly, cells were rapidly thawed and washed with media (DMEM/F12, 10% FBS and 1% Antibiotic-Antimycotic, Gibco), then centrifuged at 480 g for 10 min, the supernatant was aspirated and cells were resuspended in media. Cells were counted using a hemocytometer and viability assessed using trypan blue exclusion dye.

There were four conditions: media alone, media with TNFα, media with IFNγ, and media with TNFα & IFNγ combined. 48-well plates were set up, the four conditions were plated in triplicate for supernatant collection. For RNA, the four conditions were plated in duplicate. After the cells were initially seeded, the plates were incubated at 37°C with 21% O2 and 5% CO2 overnight to allow the cells to recover from thawing after cryopreservation. After 24 h of recovery, supernatant was removed and the cells were exposed to media only, 10 ng/mL of TNFα, 10 ng/mL of IFNγ or both 10 ng/mL of TNFα and 10 ng/mL of IFNγ. Media was added to increase the final volume in the wells to 500 uL. The plates were incubated for 16 h at 37°C with 21% O2 and 5% CO2. Supernatant was collected for protein analysis and cells were harvested for either flow cytometry or RNA isolation.

Luminex protein analysis

Neat cell supernatant was placed into 96-well plates and analyzed with a Magnetic Luminex protein array system (Human Premixed Multi-Analyte Kit; RD Systems, Minneapolis, MN, USA) for the presence of Angiopoietin-1 (Angpt1), SDF-1, IFN-γ, IL-6, VEGFA, BDNF, GDNF, IL-10 and TNFα. This assay is designed for the simultaneous detection of multiple human biomarkers in cell culture supernatants and the assay was performed strictly according to the manufacturer’s instruction (Catalogue Number LXSAHM).

mRNA expression

mRNA expression of genes of interest from frozen tissue are outlined in Table S1. RNA extraction, cDNA preparation and analysis were conducted as described previously [31]. Briefly, mRNA was extracted from the PVWM and SCWM (RNeasy Midi RNA Extraction kit; Qiagen, VIC, Australia) and reverse-transcribed into cDNA according to the manufacturer’s instructions (SuperScript® III First-Strand Synthesis System for RT-qPCR kit; Invitrogen). Genes of interest were measured by quantitative PCR using the Fluidigm Biomark HD system (Fluidigm Corporation, CA, USA). Samples were run in triplicates, averaged and normalised to endogenous housekeeping gene ribosomal protein S18 (RPS18) expression then expressed relative to SHAM group using the 2–∆∆Ct method.

mRNA expression within cultured cells of genes of interest is outlined in Table S2. Cells were harvested for RNA isolation after the 16 h of stimulation. In a modification of the protocol included in the Bioline RNA kit, the cells were first centrifuged after collection, the supernatant was aspirated off and the cells were suspended in 2 µL TCEP and 100 µL RNA Lysis Buffer RL. This solution was then frozen at -80°C until ready for use. After thawing, RNA was isolated according to manufacturer’s instructions (Bioline, Meridian Bioscience, USA). The concentration of the isolated RNA was measured on a Nanophotometer (Implen). RNA was reverse-transcribed into cDNA (SuperScript III reverse transcriptase, Invitrogen; Life Technologies, USA) according to manufacturer’s instructions. Relative mRNA expression was measured by quantitative real-time PCR using Applied Biosystems 7900HT Fast Real-Time PCR system. The expression of all genes was normalized to β-actin for each sample using the 2–∆∆Ct method.

Statistical analysis

Statistical analyses were conducted using GraphPad Prism (version 10.1.1; GraphPad Software, CA, USA). Data were tested for normality by Shapiro-Wilk test and statistical analysis conducted accordingly. Data violating the assumption of sphericity were corrected with the Greenhouse-Geisser method. An alpha of P < 0.05 was adopted to establish statistical significance. All data (Flow cytometry, DTI parameters, proteins, RT-qPCR and immunohistochemistry analyses) were compared using a one-way ANOVA, and significance followed by post-hoc testing using Tukey’s multiple comparisons. For non-parametric data, Kruskal-Wallis test was performed and a Dunn’s post-hoc analysis used. Blood gas, CBF and ELISA data were compared using two-way repeated measures ANOVA. The independent variables assessed were group (PGROUP) and time of measurement (PTIME). Where there was a significant interaction between independent variables (PGROUP X TIME), a post-hoc analysis with Tukey’s multiple comparisons test was undertaken. DTI data are presented as box plots with 5–95% confidence intervals of median, with maximum–minimum error bars. All other data are presented as mean ± SD.