CDH fetuses were induced by nitrofen administration to pregnant rats. Pregnant Sprague-Dawley rats were gavage fed 100 mg of the herbicide nitrofen (Tokyo Chemical Industry Co., LTD., Tokyo, Japan) dissolved in 1 mL of corn oil at embryonic day (E) 9.5, as previously described (Thebaud et al., 1999). Control animals did not receive any manipulation. Pregnant Sprague-Dawley rats were randomized to 2 groups: IA treatment at E13.5 (early treatment; ET) and IA treatment at E15.5 (late treatment; LT).

E13.5 or E15.5 timed-pregnant rats were anesthetized using inhaled isoflurane. Rats were restrained on a heated surgical platform, and a midline laparotomy was performed. Uterine horns were identified and externalized through the laparotomy incision to expose uterine saccules (fetuses). A 31G needle with a syringe was used to penetrate each uterine saccule. We injected either a 0.8 mg/mL sildenafil solution (Revatio; Pfizer, New York, NY, United States) or the same volume of phosphate-buffered saline (PBS) into the amniotic cavity. In order to determine the therapeutic range of IA sildenafil, we delivered three different doses of sildenafil to nitrofen-treated (CDH) fetuses. For the efficacy study, both CDH fetuses and control fetuses were used. The fetuses located in one uterine horn were treated with sildenafil and the fetuses located in the opposite uterine horn were treated with PBS.

The details of the dosing are shown in Table 1. Following injections, the laparotomy incision was closed, and the rats were allowed to recover.

Close to term (E20.5), fetuses were harvested by cesarean section and assessed for various parameters, outlined below, according to our experimental design. A total of 416 fetuses underwent injection and 260 of them survived to harvest. To assess the pulmonary effect of intra-amniotic sildenafil treatment in CDH, fetuses without CDH were excluded from nitrofen-treated fetuses. Overall, a total of 133 fetuses with CDH and 58 control fetuses were analyzed (Supplementary Table S1).

We used immersion fixation for CDH lungs to reduce the artifactual effects of the injection method of lung fixation. Fetal lungs were fixed in 4% paraformaldehyde (PFA), cryoprotected in 30% sucrose overnight, embedded in Tissue-Tek O.C.T., and frozen in liquid nitrogen. Sections (5 um) were cut by cryostat and mounted. Serial sections were taken throughout both lungs and stained with hematoxylin and eosin (H&E). The stained slides were imaged using Nikon Eclipse E800 and the NIS-Elements D microscope imaging software (Nikon, Tokyo, Japan). Images were captured using ×10 and ×20 objectives. The alveolar structure was analyzed via the radial alveolar count (RAC) (Cooney and Thurlbeck, 1982) which was calculated using 3 separate locations of each lung. These 6 counts from both lungs were averaged to calculate the final RAC, and five to eight animals were used per group.

Histologic sections were analyzed via light microscopy. Preacinar resistance arterioles between 30 and 60 µm were included. We measured the external diameter (ED), internal diameter, and proportional medial wall thickness (MWT) using the formula: MWT (ED—internal diameter)/ED (Schmidt et al., 2012). Vascular measurements were obtained from both lungs equally using 5 to 9 different animals per group. A total of 275 vessels were analyzed.

Snap-frozen lungs were homogenized on ice in RIPA buffer containing protease/phosphatase inhibitor cocktail (Cell Signaling Technology, Danvers, MA, United States of America). Samples were sonicated and centrifuged at 11000 rpm for 20 min at 4°C. Protein content in the supernatant was quantified using a BCA protein assay kit (Thermo Fisher Scientific, Waltham, MA, United States of America). Twenty micrograms of protein sample per lane were subjected to SDS-PAGE, and proteins from the gel were transferred to PVDF membranes by electroblotting. Immunodetection was performed with a rabbit anti-protein kinase G (PKG) −1 monoclonal antibody (#3248, Cell Signaling Technology) diluted 1:1000 and a rabbit monoclonal to vascular endothelial growth factor A (VEGFA) antibody (ab214424, abcam, Cambridge, UK) diluted 1:1000 overnight at 4°C. After the blots were washed to remove unbound antibody, secondary antibody, Goat anti-rabbit HRP conjugate (1:2000, Bio-Rad, Hercules, CA, United States of America) detection was applied for 1 h at room temperature. After being washed, bands were visualized using an enhanced chemiluminescence kit (SuperSignal West Femto Maximum Sensitivity Substrate, Thermo Fisher Scientific). Additionally, each gel was stripped and reprobed with actin or tubulin as a housekeeping protein to normalize for protein loading.

The main pulmonary artery of E20.5 rat fetuses was canulated and injected by way of the right ventricle with a mixture of barium sulfate and gelatin at a temperature of 60°C. The lungs were then removed from the body cavity and submerged in 4% PFA. Fixed lungs were imaged using a computed tomography imaging system (Siemens, Munich, Germany). For quantitative assessment, the diameters of the right pulmonary artery (RPA) and left pulmonary artery (LPA) at the bifurcation, as well as the diameter of the aorta were measured using Radiant DICOM Viewer (Medixant, Poznan, Poland) to determine the Modified McGoon Index (MMI): (LPA + RPA)/descending aorta, a prognostic indicator for newborns with CDH (Suda et al., 2000). For the cardiac measurements, the lumenal volumes of the right ventricle (RV) and the left ventricle (LV) and the wall volumes of RV and LV plus septum (S) were measured using Inveon Research Workplace (Siemens Medical Solutions United States). 3D volumetric measurements were performed following segmentation in the three-view axial, sagittal, and coronal planes. The right ventricular wall volume to the left ventricular and septal wall volume ratio (RV/LV + S) and the right ventricular lumen to the left ventricular lumen ratio (LV/RV) were determined.

Prior to fetal harvest, we performed an in utero intracardiac injection of fluorescein-conjugated tomato-lectin (TL) for CDH fetuses in utero as previously described (Shah et al., 2011). An ultrasound microscope probe was used to guide the injection apparatus. Sub-microliter injection system (Nanofil-100, World Precision Instruments, Sarasota, FL, United States of America) was used to inject TL into the heart. Each fetal heart was injected with 40 μL of TL. The injected fluorescent-conjugated lectin was allowed to circulate for 10 min to ensure adequate binding to the vascular endothelial wall of the vasculature, after which the fetus was harvested and fixed in 4% PFA for immunofluorescent analyses.

We used the PACT protocol proposed by Yang et al. (Yang et al., 2014) with minor modifications. Following harvest, the lungs were fixed in 4% PFA at 4°C overnight. Fixed lungs were incubated at 4°C overnight in a hydrogel monomer solution A4P1 (4% acrylamide and 1% paraformaldehyde in PBS) supplemented with 0.25% photoinitiator 2,20-Azobis[2-(2-imidazolin-2-yl) propane] dihydrochloride (VA-044, Wako Chemicals United States of America, Richmond, VA, United States). A4P0-infused samples were degassed and then incubated for 2–3 h at 37°C to initiate tissue-hydrogel hybridization. After removing excess hydrogel, tissue-hydrogel matrices were embedded in a cassette and submerged in 8% sodium dodecyl sulfate (SDS) in 0.1MPBS (pH 7.5), and depending on tissue size, were incubated for 2–3 days at 37°C. For immunostaining, PACT-processed samples were washed in PBS followed by protein blocking with 10% normal donkey serum for 4 h at room temperature. Then, the samples were incubated with anti-PECAM (platelet and endothelial cell adhesion molecule) −1 antibody (Santa Cruz Biotechnology, Dallas, TX, United States) diluted 1:50 in PBST containing 10% normal donkey serum for 2 days at 4°C. Unbound antibody was removed via PBS washes, with 5 buffer changes over the course of a day, and then samples were incubated with secondary antibody (Fabfragment secondary antibody, 1:200) for 2 days at 4°C. Samples were then washed in PBS prior to incubation in imaging media (refractive index matching solution, RIMS), the details of which were described by (Yang et al., 2014). Samples were incubated in RIMS for 1–3 days at 4°C until transparent.

Cleared tissue samples were mounted in RIMS with Nunc^® Lab-Tek^® Chamber Slide System (Millipore Sigma, Burlington, MA, United States). The samples were imaged by STELLARIS 5 (Leica microsystems, Buffalo Grove, IL, United States of America) under a ×10 objective. four to six images were obtained from the peripheral region from both lungs equally for each fetus using 5 different animals per group. Image reconstructions were performed using Imaris imaging software (Bitplane, Belfast, UK) with an image size of 901 μm × 901 μm x 200 μm, and blood perfusion was determined using the following formula: overlapped volume of TL to PECAM-1/total PECAM-1 volume per field (%). A total of 107 images were analyzed.

Statistical analysis was performed using GraphPad Prism software (Prism 9, GraphPad Software, San Diego, California, United States). Values were expressed as mean ± SD. Comparisons of parameters between 2 groups were made by unpaired 2-tailed t-test. Comparisons between multiple groups were made using 1-way ANOVA (with repeated measures where appropriate) followed by the Holm-Šidák test for multiple comparisons. p < 0.05 was considered statistically significant.

The authors had full access to and take full responsibility for the integrity of the data.

Three different doses were used to examine the dose-response effect for both ET and LT. There was no significant dose-response relationship for lung weight-to-body weight ratio (Lw/Bw) which is an indicator of pulmonary hypoplasia, lung weight, and body weight (Figures 1A–C). However, in the ET group, CDH fetuses given the mid-range dose (therapeutic dose; TD) of IA sildenafil had a significantly higher RAC, used as a tool for estimation of lung maturation, than the low dose (LD) or the high dose (HD) of IA sildenafil (Figure 1D; Figure 4A; Supplementary Figure S1A). MWT, used as an indicator of pulmonary hypertension, did not show a significant difference between the three different doses for both ET and LT. (Figure 1E; Figure 5A; Supplementary Figure S1B). These results suggest that earlier intervention with TD is most effective in improving lung development, which is perhaps due to a minor toxicity of the HD and ineffectiveness of the LD.

The mid-range dose (TD) was applied for comparison between IA sildenafil administration and IA PBS administration in order to determine the efficacy of this treatment. The survival rates of fetuses near term (E20.5) were compared in CDH fetuses treated with sildenafil or PBS (placebo) for both ET and LT. In the ET group, IA sildenafil administration led to significantly higher survival in CDH fetuses than IA PBS administration, but there was no difference between the CDH fetuses treated with sildenafil and the CDH fetuses treated with PBS in the LT group (Figure 2A). There was no significant difference in the proportion of CDH (Figure 2B) per cohort. The size of the CDH defect was also determined (Supplementary Table S2). The defect was found on the left side in 96.7% of the fetuses, two fetuses showed bilateral CDH, and only one fetus showed right CDH alone. In all groups, large “Defect C”: <50% hemi-diaphragm present, and “Defect D”: complete or near complete absence of the diaphragm, were detected in more than 70% of the fetuses. These results indicate that the improvement in fetal survival did not correlate with the defect size (Lally et al., 2013). On the other hand, IA sildenafil administration did not affect fetal survival in control fetuses. The survival after IA administration was 95% or higher in both sildenafil-treated and PBS-treated fetuses in both ET and LT in instances when the treated horn had fewer than 8 fetuses.

CDH fetuses had significantly reduced Lw/Bw and lung weights in both ET and LT groups (Figures 3A, C). CDH fetuses in the LT group had significantly reduced body weights, however, CDH fetuses in the ET group did not have reduced body weights (Figure 3B). Early IA sildenafil treatment increased Lw/Bw in CDH fetuses, while late IA sildenafil treatment did not significantly increase Lw/Bw in CDH fetuses. There were no differences in body weight and lung weight between sildenafil-treated and PBS-treated CDH fetuses in both treatment groups. IA sildenafil treatment had no effect on Lw/Bw, body weight, and lung weight in control fetuses. These results suggest that early IA sildenafil treatment may improve lung hypoplasia and fetal survival in CDH fetuses without negative effects in control fetuses.

CDH fetuses treated with PBS showed less complexity ofairway branching and significantly lower RAC than control fetuses in both ET and LT groups (Figures 4A, B). IA sildenafil treatment significantly increased RAC in CDH fetuses in the ET group to levels similar to those of controls. In the LT groups, RAC trended higher in IA sildenafil-treated CDH fetuses compared to IA PBS-treated CDH fetuses but did not show a statistical difference. IA sildenafil treatment led to a 2.5-fold increase in VEGFA levels in CDH fetuses after ET, a crucial factor in the development of the alveolar capillary bed (Thebaud et al., 2005), and a 1.7-fold increase in CDH fetuses after LT (Figures 4C, D). IA sildenafil treatment had no effect in control fetuses on either RAC or VEGFA expression in both ET and LT groups.

The MWT of IA PBS-treated CDH fetuses was significantly higher than control fetuses in both ET and LT groups (Figures 5A, B). IA sildenafil treatment markedly reduced MWT of CDH fetal lungs that were similar to those of control fetuses in both ET and LT groups. There was no difference between IA PBS-treated control fetuses and IA sildenafil treated control fetuses.

The biological activity of sildenafil in fetal lungs was confirmed by measuring PKG -1 (Figure 5C) in fetal rat lungs. IA sildenafil treatment significantly increased PKG-1 expression in CDH fetuses 2.6-fold in the ET group and 2.5-fold in the LT group (Figure 5D). IA sildenafil had no effect on PKG-1 expression in control fetuses in both ET and LT groups. These results indicate that a single injection of IA sildenafil maintained increased PKG-1 expression in the CDH lung for a week and ameliorated the vascular hypermuscularization associated with CDH without negatively affecting control fetuses.

MMI is clinically used to evaluate the risk for pulmonary hypertension in CDH fetuses, with an index of less than 1.3 indicating high risk (Suda et al., 2000). We measured the left and right pulmonary artery diameters and descending aorta diameter using a barium-gelatin angiogram and calculated an MMI in CDH rat fetuses and control fetuses (Figures 6A, B, Supplementary Video S1). IA sildenafil treatment significantly increased the MMI in both ET and LT groups (Figure 6C). IA sildenafil treatment had no significant effect in MMI on control fetuses.

We also determined the RV/LV + S as an indicator of right ventricular hypertrophy (Luong et al., 2011) and the RV/LV as an indicator of pulmonary arterial hypertension (Bax et al., 2020) (Figures 6D, E). In both ET and LT groups, there were no significant differences in RV/LV + S between controls and CDH fetuses. RV/LV trended higher in IA PBS-treated CDH fetuses compared to IA sildenafil-treated CDH fetuses and controls but did not show a statistically significant difference.

For further investigation of IA sildenafil’s effect on CDH fetuses, we visualized pulmonary blood flow in CDH fetuses. We used an in utero intracardiac fetal injection of a fluorescein-conjugated TL to label endothelial cells that are being perfused. The fetal heart pumps this tracer throughout the fetus to delineate which vessels are being perfused with blood, as opposed to existing blood vessels that are not receiving blood flow. By comparing TL-marked vessels (i.e., vessels receiving embryonic blood flow) with immunostaining for a general marker of PECAM-1, we identified numerous pulmonary vessels that were not receiving blood flow (non-perfused) as previously described (Shah et al., 2011). There was a clear distinction in perfused blood vessel distribution (TL+/PECAM-1+ double positive) between IA sildenafil-treated CDH fetuses and IA PBS-treated CDH fetuses (Figure 7A, Supplementary Video S2). IA sildenafil treatment significantly improved the peripheral blood perfusion of CDH fetuses in both ET and LT groups, and the improvement was more strikingly observed in the ET group (Figure 7B).