The RNA-seq datasets generated and analyzed during the current study are available in the GEO repository under the accession number GSE291221. All other relevant data supporting the findings are available from the corresponding author upon reasonable request. All animal experimental procedures were reviewed and approved by the Shanghai Children’s Medical Center Institutional Animal Care and Use Committee (IACUC Approval No. SCMC-LAWEC-2024-1127). All methods were performed in accordance with the relevant guidelines and regulations (e.g., ARRIVE guidelines).

This study was designed to investigate the impact of neonatal IPF on pulmonary small vessel remodeling and its underlying immune-related mechanisms (Figures 1A, B). Neonatal C57BL/6J mice were randomly assigned to either the IPF model group (subjected to aortocaval fistula surgery, ACF) or the sham-operated control group. The primary endpoint was the assessment of pulmonary vascular remodeling via histological and molecular analyses at postnatal day 30 (P30). Secondary endpoints included hemodynamic validation by ultrasound, transcriptomic profiling, and immunophenotyping. An additional interventional arm involved treating a subset of IPF mice with the immunosuppressant Cyclosporin A (CsA, 20 mg/kg/3 day, starting at P14 and continued until P30) or type I IFN receptor blocker MAR1-5A3 (GTX14637, Beyotime, Shanghai, 12.5mg/kg/2 day) to evaluate the functional role of the immune response.

The primary discovery cohort consisted of sham-operated and ACF (IPF) mice used for hemodynamic, histological, and transcriptomic analyses. Interventional studies (CsA and type I IFN receptor blockade) were performed in independent experimental cohorts separate from the primary discovery cohort. In each interventional cohort, IPF mice receiving drug treatment and corresponding IPF control mice receiving vehicle were generated within the same surgical and experimental batch to minimize batch effects. Sham-operated animals from the same time period were used as reference controls when appropriate.

Due to the limited tissue volume of neonatal mouse lungs, cohort usage and tissue allocation were predefined before experiments. Separate animal cohorts were used for major assay categories to ensure adequate sample quality and avoid cross-assay interference. Specifically, whole lung tissue from designated animals was used for RNA sequencing, perfused and fixed lungs were used for histological and immunofluorescence analyses, and freshly dissociated lungs were used for flow cytometry.

qRT-PCR validation was performed using RNA derived either from the RNA-seq cohort or from an independent validation cohort processed in parallel. Each animal was assigned to a primary assay category at the time of sacrifice and was not reused across incompatible downstream assays.

Timed-pregnant C57BL/6J mice were raised in our laboratory. Pups of both sexes were used on postnatal day 7 (P7) for surgery. Animals were randomly assigned to sham or ACF groups at the time of surgery, with sex-balanced allocation performed when feasible within each litter. No animals were excluded based on sex. Perioperative survival was recorded prospectively. Deaths occurring within 24 h after surgery were defined as surgery-related mortality. Early (24–72 h) and late (>72 h to endpoint) mortality were also recorded. Survival and baseline characteristics, including sex distribution and body weight at surgery, are summarized in Supplementary Table 1.

Mice were anesthetized by isoflurane inhalation (2-3% for induction, 1-2% for maintenance) and placed on a heating pad. Under a stereomicroscope, a midline abdominal incision was made. The abdominal aorta and inferior vena cava (IVC) were carefully isolated. An 9–0 needle (0.07 mm diameter) was used to create a puncture between the adjacent vessels, followed by a 11–0 nylon suture (Ethicon) to establish a permanent fistula. Patency was confirmed by visual observation of pulsatile, turbulent flow. For sham operations, vessels were isolated but not punctured. The abdominal wall and skin were closed in layers. Pups were warmed until recovery and returned to the dam. Analgesia (buprenorphine, 0.05 mg/kg) was administered postoperatively.

At P30, mice were anesthetized with 1.5% isoflurane. Abdominal ultrasound was performed using a Vevo 3100 high-resolution imaging system (Fujifilm VisualSonics) with an MX550D transducer to visualize the ACF fistula and measure shunt flow velocity. Transthoracic echocardiography was then conducted. Pulmonary artery outflow was assessed in a parasternal short-axis view to measure the pulmonary artery velocity time integral (PA-VTI) and diameter. Right ventricular (RV) dimensions and function were analyzed from an apical four-chamber view. Right ventricular stroke volume (RVSV) was calculated as: RVSV [mL] =1/4 × πD²× PA-VTI. All measurements were averaged over three consecutive cardiac cycles by an investigator blinded to group allocation.

Hemodynamic measurements were obtained by cardiac catheterization using an open-chest approach under general anesthesia. After tracheal intubation and mechanical ventilation, a left thoracotomy was performed to expose the heart. Millar catheter transducer (ADInstruments, SPR-671NR) was directly inserted into the right ventricle under visual guidance and further advanced into the pulmonary artery for continuous pressure recordings. Right ventricular systolic pressure (RVSP) and pulmonary arterial pressure were recorded and analyzed using PowerLab system (ADInstruments). Hemodynamic parameters were derived from stable waveform segments.

After hemodynamic assessment, lungs were perfused with PBS via the right ventricle, followed by inflation and fixation with 4% paraformaldehyde (PFA) at a constant pressure of 25 cm H₂O. Tissues were paraffin-embedded and sectioned at 5 μm thickness. Sections were deparaffinized, rehydrated, and stained with H&E using standard protocols.

Paraffin sections underwent antigen retrieval in citrate buffer. After blocking with 5% normal donkey serum, sections were incubated overnight at 4 °C with primary antibodies: rabbit anti-α-Smooth Muscle Actin (α-SMA, 1:400, Abcam ab5694). After washing, appropriate fluorescently-labeled secondary antibodies (Alexa Fluor 594, 1:500, Invitrogen) were applied. Nuclei were counterstained with DAPI. Fluorescence images were acquired using a confocal microscope (Zeiss LSM 880). The integrated fluorescence density of α-SMA+ vessels in consistent fields of view was quantified using ImageJ.

Total RNA was extracted from snap-frozen whole lung tissue (n=4–5 per group) using TRIzol reagent. RNA integrity was verified (RIN > 8.0) by Bioanalyzer. Libraries were prepared with the KAPA mRNA HyperPrep Kit and sequenced on an Illumina NovaSeq 6000 platform (150 bp paired-end). Raw reads were quality-trimmed using Fastp and aligned to the mouse reference genome (GRCm39) using HISAT2. Gene expression quantification was performed with StringTie. Differential expression analysis (IPF vs. Sham) was conducted using DESeq2 with thresholds of |log2FoldChange| > 1 and adjusted p-value (q-value) < 0.05. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were performed using the clusterProfiler R package.

qRT-PCR was performed to validate key differentially expressed genes identified by RNA-seq. Total RNA (1 µg) was reverse transcribed into cDNA using the PrimeScript RT reagent Kit (Takara). Quantitative PCR was carried out on a LightCycler 96 System (Roche) using SYBR Green Premix Pro Taq HS (Accurate Biology). Gene expression was normalized to the housekeeping gene Gapdh and calculated using the 2^(-ΔΔCt) method. Primer sequences for target genes are listed in Supplementary Table 2.

Single-cell suspensions were prepared from perfused lung tissue by mechanical dissociation and enzymatic digestion (Collagenase IV/DNase I). Red blood cells were lysed. Cells were blocked with anti-CD16/32 antibody and stained with fluorochrome-conjugated antibodies for surface markers: BV605 anti-CD45 (30-F11), PE-Cy7 anti-CD4 (GK1.5), and APC anti-CD8a (53-6.7). Corresponding isotype controls were used. Cells were analyzed on a Beckman CYTOFLEX flow cytometer. Data were processed using FlowJo software (v10.8.1), gating on live, singlet, CD45+ leukocytes.

All data are presented as mean ± standard error of the mean (SEM). Statistical analyses were performed using GraphPad Prism (v9.0). Normality was assessed using the Shapiro-Wilk test. For comparisons between two groups (Sham vs. IPF), an unpaired, two-tailed Student’s t-test (parametric) or Mann-Whitney U test (non-parametric) was used. For comparisons involving multiple groups (e.g., Sham, IPF, IPF+CsA), one-way ANOVA with Tukey’s post-hoc test or Kruskal-Wallis with Dunn’s test was applied accordingly. A p-value < 0.05 was considered statistically significant.

As depicted in Figures 1A, the neonatal mouse lung undergoes rapid postnatal development, progressing through the alveolar stage (ALV stage, postnatal days 0-18) followed by the lung maturation stage (M stage, days 21-56) (7–10). To create a left-to-right shunt, we performed an aortocaval fistula (ACF) surgery on postnatal day 7 (ALV2 stage). Subsequent analyses, including echocardiography, RNA-seq, and flow cytometry, were conducted on postnatal day 30 (M stage) (Figure 1B). Following ACF surgery (Figure 1C), pulsatile blood flow was clearly visible at the fistula site with a velocity of 550 mm/s (Figures 1D, E). Echocardiographic assessment revealed a significant increase in the velocity time integral of the pulmonary artery (Figures 1F–H). While the diameter of the pulmonary artery showed no significant change (Figure 1I), the right ventricular stroke volume (RVSV) was markedly increased (Figure 1J). These results indicate the successful induction of a left-to-right shunt leading to increased pulmonary blood flow during the neonatal period, consistent with clinical investigation (5, 6). Serial abdominal ultrasound confirmed shunt persistence from P7 to P30, demonstrating the model’s hemodynamic stability (Supplementary Figures 1A, B). Baseline characteristics, including litter size, sex distribution, and perioperative survival, showed no significant intergroup differences at surgery (Supplementary Table 1), supporting balanced allocation and minimizing selection bias.

We next performed RNA-seq analysis on the model. IPF induction resulted in 304 differentially expressed genes (DEGs) (q-value < 0.05 && |log2FC| > 1), among which 112 were upregulated and 192 were downregulated (Figure 2A). Cluster analysis of these DEGs demonstrated clear separation between groups and high similarity within groups (Figure 2B), suggesting a highly consistent pulmonary remodeling phenotype induced by IPF. This finding was further corroborated by PCA analysis, which showed distinct separation between the IPF and sham groups along principal components (Figure 2C).

Although clinical observation identifies IPF as a prerequisite for pediatric pulmonary hypertension (5, 6), the lack of a neonatal IPF animal model has left the direct causal relationship unconfirmed. As shown in Figures 3A, B, immunofluorescence staining revealed that neonatal IPF induced diffuse pulmonary small vessel remodeling, evidenced by a significant increase in the fluorescence density of pulmonary vessels (α-SMA-positive) (Figure 3B). This observation was further confirmed by H&E staining (Figure 3C). Transcriptomic analysis (Figure 4A) showed a significant upregulation of Spp1 (a marker gene for synthetic-phenotype SMCs) and Mmp9 (a key metalloproteinase mediating extracellular matrix remodeling in pulmonary hypertension), alongside a significant downregulation of Myh11 (a marker gene for contractile-phenotype SMCs). Furthermore, the expression of key signaling molecules mediating vascular remodeling, such as those in the TGF-β1–3 and SMAD3/YAP1 pathways, was significantly elevated. qRT-PCR validation confirmed these RNA-seq findings with high consistency (Figure 4B). Collectively, these results indicate that neonatal IPF contributes to the development of pulmonary small vessel remodeling.

To directly address whether pulmonary vascular remodeling represents an early and progressive consequence of neonatal IPF rather than a late adaptive phenomenon, we further performed additional time-course and hemodynamic analyses. The results showed that from postnatal day 14 (P14) to P30, both α-SMA and HE staining revealed a progressive thickening of the medial layer of pulmonary small vessels. In parallel, right ventricular systolic pressure (RVSP) measured by cardiac catheterization also showed a gradual increase (Supplementary Figures 2A–D).

To investigate the mechanisms underlying neonatal IPF-induced remodeling, we performed enrichment analysis on the DEGs. GO enrichment analysis (Figures 5A, B; Supporting Data 1) revealed that among the top ten enriched Biological Process (BP) terms, five were related to immune response: “response to virus,” “defense response to virus,” “response to interferon (IFN)-beta,” “defense response to other organism,” and “cellular response to IFN-beta.” Within the top thirty enriched terms, eight were associated with vascular remodeling, including “regulation of systemic arterial blood pressure mediated by a chemical signal,” “collagen metabolic process,” and terms related to extracellular matrix and collagen structure. Pathway enrichment analysis (Figures 5C, D) indicated that 10 out of the top twenty enriched pathways were immune-related, such as the “TNF signaling pathway,” “IL-17 signaling pathway,” “Cytokine-cytokine receptor interaction,” and “Natural killer cell mediated cytotoxicity.” Enrichment analysis focusing solely on upregulated DEGs showed an even more pronounced association with immune responses (Supplementary Figures 3A–D and Supporting Data 2). These findings suggest that immune activation may represent a key mechanism in IPF-induced pulmonary vascular remodeling.

To experimentally validate the transcriptomic signature indicative of an antiviral-like response, we first examined the expression of key differentially expressed genes (DEGs) within the top-ranked GO term “response to virus.” qRT-PCR analysis confirmed the significant upregulation of all nine tested DEGs in IPF lungs compared to controls (Figures 6A, B). This molecular signature was further corroborated at the cellular level by flow cytometry (Figures 6C, D). While the absolute increase in T cell frequency was modest (~2–3 percentage points), the change was statistically significant and occurred in the context of a globally low inflammatory milieu, suggesting a specific immune activation in response to increased flow.

We next investigated whether broad immunosuppression could mitigate the resultant pathology. Treatment of IPF model mice with the immunosuppressant CsA significantly attenuated IPF-induced remodeling of the pulmonary small vasculature (Figures 7A–D). To specifically probe the contribution of the antiviral-like response, we administered the type I interferon receptor blocker MAR1-5A3. This targeted intervention similarly led to a significant reduction in pulmonary vascular remodeling (Supplementary Figures 4A–C).

Collectively, these data demonstrate that the activation of an antiviral-like immune response is associated with pulmonary small vessel remodeling in neonatal IPF and is functionally involved in the remodeling phenotype, as suggested by attenuation with immunomodulatory interventions.