Influenza A virus (IAV) infection during pregnancy is associated with stillbirth and preterm birth. We hypothesized that maternal IAV infection disrupts placental and fetal immune networks in ways that correlate with disease severity. Pregnant pigtail macaques were inoculated with IAV [A/California/07/2009 (H1N1)] and underwent delivery and fetal necropsy 5 days post-infection (N=11), with comparison to uninfected controls (N=16). Stillbirth occurred in 18% of infected pregnancies and in no controls. Although vertical transmission was not observed, low levels of viral RNA were detected in two placentas. Maternal infection was associated with induction of a type I interferon response in the placenta and altered innate and adaptive immune cell populations in the fetus. Maternal disease metrics (e.g., lung viral load, IFN-a) were rarely associated with placental immune perturbations but frequently linked to fetal immune cell populations (CD4+ and CD8+ T cells, monocyte subsets). Collectively, these data indicate that maternal IAV disease severity is associated with dysregulated fetal immunity.

Research Animals and Ethics Statement 

All animal experiments used in this study were approved by the University of Washington Institutional Animal Care and Use Committee (protocol# 4165-03, last approved 07/22/2025), and in compliance with the U.S. Department of Health and Human Services Guide for the Care and Use of Laboratory Animals and Animal Welfare Act. All animal experiments were performed at the Washington National Primate Research Center (WaNPRC).  

Study Design 

Two groups of female pigtail macaques (Macaca nemestrina) were inoculated using a combination of intranasal, intratracheal, intraocular, and intraoral routes with 107.4 plaque forming units (PFU) of IAV [A/California/07/2009 (H1N1)], reflecting a dose previously used in a study of cynomolgus macaques (Macaca fascicularis). Experimental groups consisted of IAV H1N1 infected-pregnant (N=11) and uninfected pregnant controls (N=18; Table S1). IAV inoculation occurred in the early third trimester to model the time in pregnancy when IAV H1N1 infections are thought to produce the greatest maternal lung disease and pregnancy complications. 

Macaques inoculated with IAV were observed daily for clinical symptoms of influenza disease, which were scored on a 3-point scale to reflect responsiveness of the animal, nasal or ocular discharge, cough, respiratory effort, food consumption, and fecal consistency; higher scores reflected disease with a 0 in each category indicating a normal assessment. We chose a pre-defined study endpoint of 5 days post-inoculation (dpi) to capture immune responses near peak disease. A Cesarean section was performed at the endpoint, and placentas were extracted to sample chorionic villous tissue. The dam was euthanized using a barbiturate overdose while under anesthesia. The fetus was euthanized using a barbiturate overdose, followed by exsanguination. These methods are approved by the 2007 American Veterinary Medical Association Guidelines on Euthanasia. Necropsies were performed on both the dam and the fetus to obtain samples for transcriptomics, immune assays, and histopathology. Random sections of the placental membrane, chorionic disc, and umbilical cord were evaluated for pathology. 

Controls in this study mainly consisted of pregnant adult pigtail macaques used in prior published studies, with the addition of several newer unpublished experimental controls; these control animals received either a subcutaneous inoculation of media or a choriodecidual saline inoculation in the late second or early third trimester (Table S1). Reusing uninfected NHP control pregnancies across studies is vital for ethical reasons and conservation, particularly when immune response profiles indicate that the control cohort had a normal pregnancy and low levels of inflammatory cytokines.  

H1N1 Virus Cultivation and Quantification 

We used an IAV strain [A/California/07/2009 H1N1 influenza virus (CA09)] that was obtained from the World Reference Center for Emerging Viruses and Arboviruses at the University of Texas Medical Branch at Galveston. Before culturing the virus, Madin-Darby canine kidney (MDCK) cells (ATCC, Manassas, Virginia, US) were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 25 mM HEPES, 100 U/mL penicillin, and 100 μg/mL streptomycin. Cells were incubated at 37°C and 5% CO2 and confirmed to be free of mycoplasma with the MycoAlert™ mycoplasma detection kit (Lonza, LT07-318) before use for virus cultivation. Sub-confluent monolayers of MDCK cells in T150 flasks were washed with PBS and infected at a multiplicity of infection (MOI) of 0.5 (5 mL serum-free MEM) at 37°C and 5% CO2. After 1 hour, the media was removed, cells were washed with 1X PBS, and the media was replaced. The supernatant was harvested after 36 hours and centrifuged at room temperature to remove cellular debris (1800 rpm for 5 minutes, followed by 3000 rpm for 10 minutes). Virus stock was frozen at −80°C. The viral titer of each stock was quantified using a TCID50 assay with hemagglutination readout and converted to PFU/mL.64 To transform TCID50/mL into PFU/ml, we multiplied TCID50/mL by 0.7. Reagents were obtained from Thermo Fisher Scientific unless otherwise stated (Waltham, MA).  

RNA Isolation and Reverse Transcription Quantitative Polymerase Chain Reaction (RT-qPCR) 

Tissue samples were kept in RNAlater™ Stabilization Solution (#15596018, Thermo Fisher Scientific) prior to homogenization in Precellys homogenizing tubes (#1011152, Cayman Chemical, Ann Arbor, MI, USA) with the Precellys Evolution machine (Bertin Technologies, Montigny-le-Bretonneux, France). RNA was extracted from homogenates using the RNeasy Mini Kit (Qiagen, #74106). RNA was quantified using the NanoDrop™ 2000 spectrophotometer (Thermo Scientific, #ND-2000) and the Qubit 4 fluorometer (Invitrogen, #Q33238). RNA quality was evaluated by measuring the RNA integrity number (RIN) with the 4200 Tapestation System (Agilent, #G2991BA). 

To quantify IAV vRNA levels from different sample origins, we used reverse transcription quantitative polymerase chain reaction (RT-qPCR) and a WHO-approved primer/probe set.65 cDNA was generated from RNA isolates using an iScript™ Select cDNA synthesis kit (Bio-Rad, #1708897) using H1N1 HA-specific forward primer following manufacturer’s instructions. qPCR was done with H1N1 HA-specific primers and probes (Forward: GGGTAGCCCCATTGCAT; Reverse: AGAGTGATTCACACTCTGGATTTC; Probe: 5‐[FAM] TGGGTAAATGTAACATTGCTGGCTGG [TAMRA]‐3) using the Taqman Fast Advanced Master Mix (#444964, Thermo Fisher Scientific) and a QuantStudio 3 machine (#A28132, Thermo Fisher Scientific) following manufacturer instructions. Samples were run in triplicate. Positive results were defined as any sample with an average Ct value less than 38 cycles. 

Cytokine Quantification 

ProcartaPlex™ NHP multiplex kits were used to quantify cytokines from lysates or fluids (Thermo Fisher Scientific). Tissue lysis and homogenization were performed using an Invitrogen™ ProcartaPlex™ Cell Lysis Buffer (#EPX-99999-000, Thermo Fisher Scientific) with the addition of a protease inhibitor cocktail (Fisher Scientific, #PI78430) using the Precellys Evolution system, as previously described. Type I Interferons (IFN-alpha, IFN-beta) were quantified using the Cynomolgus IFN-beta and Cynomolgus/Rhesus IFN-alpha ELISA kits (#46415-1 and #46100-1, respectively; PBL Biosciences, Menlo Park, CA, USA). All protocols followed the manufacturer’s guidelines.  

Immunophenotyping (IMP) 

Immunophenotyping by flow cytometry was performed to determine the frequencies of immune cells in the placenta and fetal organs. Biopsies from the placental disc were dissociated into single cells and assessed for viability using Acridine Orange/Propidium Iodide (Nexcelom Bioscience LLC). Then, samples were stained with a panel of antibodies in Brilliant Stain Buffer (BD Biosciences) for 20 minutes to identify immune cells as previously described. Antibodies used included BD Biosciences: CD3 (SP34-2)-BV650, CD8 (RPA-T8)-BUV395, CD11b (ICRF44)-AP-Cy7, CD11c (SHCL-3)-PE, CD45 (D058-1283)-PE-CF594, CD86 (IT2.2)-PE-Cy5, TCRγδ (B1)-FITC; BioLegend: CD4 (OKT4)-BV605, CD14 (M5E2)-BV785, CD16 (3G8)-Alexa Flour 700, CD20 (2H7)- BV570, HLA-DR (L243)-BV71; Invitrogen: CD123 (6H6)-eFlour 450, LIVE/DEAD Fixable Aqua for 405nm excitation; Beckman Coulter: CD159a (Z199)-PC7; and Miltenyi: CD66 (TET2)-APC. After the assessment of viability, cells were resuspended in 2% FBS in PBS. All samples were acquired on a Symphony A3 (BD Biosciences) using FACS Diva 8 (BD Biosciences). Samples were analyzed using FlowJo 10.9.0 (FlowJo, LLC). All events were initially gated on Forward Scatter (FSC) singlets, CD45+ leukocytes, live cells, and then on FSC-A and Side Scatter (SSC)-A profiles. Populations were defined using canonical markers as described in Table S4. Populations were excluded from analysis if they did not meet the minimum threshold of >100 cells in their parent gate. 

Total RNA-Seq Library Preparation and Analysis 

Total RNA sequencing was performed on chorionic villous tissue from the placenta. We used 100ng of RNA with RIN>3 as input for RNA library prep using the KAPA RNA HyperPrep kit with RiboErase HMR (#08098131702, Roche Diagnostics Corporation, Indianapolis, IN, USA). The kit was designed for human, mouse, and rat samples, but also effectively reduced ribosomal RNA in our NHP RNA isolates. Libraries were prepared following the manufacturer's standard protocol. Library quantity and quality were evaluated using Qubit Fluorometer (Thermo Fisher Scientific) and TapeStation (Agilent Technologies, Santa Clara, CA, USA). Constructed libraries were sequenced on a NovaSeq 6000 Sequencing System (#20012850, Illumina, San Diego, CA, USA), which produced 2x100 nucleotide stranded paired end reads. 

Raw sequences were aligned to the Macaca nemestrina reference genome from Ensembl (Mnem_1.0, INSDC Assembly GCA_000956065.1) utilizing STAR in Partek Flow, a bioinformatics software suite (Partek, Chesterfield, MO, USA). Ensembl gene symbol conversions were employed for consistency across all gene IDs. As the quality score was greater than 35 across all samples, no bases were trimmed, and 100% of sequencing bases overlapped with read lengths of the reference genome. The raw count matrix was downloaded from Partek. Subsequent statistical processing and analysis of count data were performed within the R statistical computing environment (R version 4.4.0). Gene counts were filtered by a row mean of 10 or greater, then normalized using edgeR to implement TMM normalization. Counts were transformed into log2 counts using voom, a function in the limma package that modifies RNA-Seq data to create a normalized count matrix. Differential Expression (DE) analysis compared placental samples from IAV-infected animals against controls from uninfected animals. Comparisons were based on a linear model fit for each gene using limma. The criteria for all DE analyses included an absolute fold change of 2 and a p-value of less than 0.05. Gene Set Enrichment Analysis was conducted across the comparison group using fgsea with the Broad Institute Molecular Signatures database.  

Hierarchical clustering analysis, an unsupervised machine learning approach, was performed to group total RNA-Seq profiles based on their similarity in Euclidean distance using the hclust function in R. Ward's linkage was used to minimize variance within each cluster. The optimal number of clusters was determined using the elbow method. The resulting hierarchy was visualized using a dendrogram. Finally, an overrepresentation analysis (ORA) was performed to determine if predefined sets of genes in Gene Ontology (GO) pathways were enriched in the cluster of differentially expressed genes (DEGs) obtained from the hierarchical clustering analysis using the clusterProfiler package in R. 

Nanostring nCounter® 

To validate our results from bulk RNA sequencing, we chose a quantitative multiplex method of direct digital counting of gene transcripts using the Nanostring nCounter® Analysis Systems (Nanostring Technologies Inc., Seattle, WA, USA). RNA extracted from the chorionic villous tissues of the placenta was analyzed using the Nanostring nCounter® NHP Immunology V2 gene expression codeset (#11500276, Nanostring Technologies, Inc.). RNA hybridization reactions were performed according to the manufacturer’s protocol. Probe-target RNA complexes from all reactions were processed and immobilized on nCounter® cartridges using the nCounter® Prep Station. Gene transcripts were quantified with the nCounter® Digital Analyzer (#MAN-C0035-07, Nanostring Technologies, Inc.). 

Statistical Analyses 

All statistical analyses were performed in RStudio (using R version 4.4.0) or nSolver (nCounter data only), both of which use the R programming language. A p-value less than 0.05 was considered significant, and 0.5-0.1 was reported considering the modest sample size. Groups were compared using non-parametric tests. Spearman’s rank correlation coefficient was used to calculate correlations between two variables. Data visualizations were done using ggplot2, ggpmisc, and ComplexHeatmaps packages and their derivatives.