Fig 2: The loss of RTF2 increases the IAV infection rate in interferon-treated cells. (A) IAV infection rates based on HA surface protein levels in RTF2-KO cells and RTF2-rescued cells. RTF2-KO cells were transduced with either an empty vector or an sgRNA-resistant RTF2 cDNA to restore RTF2 expression. ****, P ≤ 0.0001, by one-way ANOVA and Tukey’s multiple-comparison test. (B) (Top) IAV infection rates based on HA surface protein levels in A549 cells transduced with nontargeting sgRNA, IFITM3 sgRNA, or RTF2 sgRNA. The three rightmost columns denote three separate RTF2-KO cell clones isolated and cultured from a pool of RTF2 sgRNA 7 polyclonal cells. P values were determined by one-way ANOVA and Dunnett’s multiple-comparison test against the nontargeting sgRNA sample. ****, P ≤ 0.0001; ***, P ≤ 0.001. (Bottom) The Western blot (immunoblot [IB]) image shows RTF2 protein expression in the isolated RTF2-KO cell clones. (C) Quantification of IAV progeny virions produced by infected WT A549, RTF2-KO, and RTF2-rescued cells. Cells were first pretreated with 200-U/ml IFN-β for 24 h, before IAV infection at an MOI of 5 for 16 h. The cell culture supernatant was collected and used to perform plaque assays to quantify the amount of infectious particles released. ****, P ≤ 0.0001, by one-way ANOVA and Tukey’s multiple-comparison test. (D) IAV transcription and replication for WT, RTF2-KO, and RTF2-rescued cells. Cells were pretreated with 200-U/ml IFN-β for 24 h, before IAV infection at an MOI of 5 for 9 h. Total RNA was used for strand-specific RT-PCR and qPCR to quantify the levels of viral NP mRNA, cRNA, and vRNA. ****, P ≤ 0.0001, by one-way ANOVA and Tukey’s multiple-comparison test. (E) Time course of IAV NP mRNA in WT, RTF2-KO, and RTF2-rescued cells. P values were determined by one-way ANOVA and a multiple-comparison test. ***, P ≤ 0.001; **, P ≤ 0.01. vec, vector; sg2, sgRNA 2; sg7, sgRNA 7.

Fig 3: Placental expression of antiviral IFITM proteins. (A) Placental expression of IFITM1–3 mRNAs. Analysis, stratified by disease severity (asymptomatic/mild vs severe) and negative controls, was performed using RT-qPCR as detailed in “Methods”. The red dot denotes SARS-CoV-2-positive placenta. The data were analyzed using Kruskal–Wallis nonparametric test, with post hoc Tukey test for all pairwise comparisons. *Denotes p < 0.01. (B) Placental sections from participants with asymptomatic/mild vs severe COVID disease, stained for either IFITIM1 or IFITM3 immunofluorescence as described in “Methods”, and nuclei visualized with DAPI. Boxed areas are enlarged below. White arrowheads mark positive fetal stem villous vessels. White arrows mark rare less frequent trophoblast staining in terminal villi. Scale Bars: 50 µm. Negative-control immunofluorescence stains are shown in Supplementary Fig. 1. (C) Placental sections from negative-control term placentas, stained for IFITM3 immunofluorescence (green) and co-stained for either EEA1, a marker of early endosome, or LAMP1, a marker of late endosome/lysosome, as described in “Methods”. Nuclei were visualized with DAPI. Boxed areas are enlarged to the right side of each image. White arrowheads denote colocalization of IFITM3 and endosomal markers. Scale Bar: 10 µm.

Fig 4: RTF2 may affect the upregulation of antiviral ISGs in response to IFN exposure. (A) Overexpressing RTF2 does not confer additional protection against IAV infection in WT A549 cells. WT A549 cells were transduced with either the empty vector or RTF2. After 8 days of blasticidin selection posttransduction, cells were treated with 200-U/ml human IFN-β for 24 h before IAV infection at an MOI of 5 for 16 h and sorted by FACS based on HA. ****, P ≤ 0.0001, by two-way ANOVA and Dunnett’s multiple-comparison test. (B to D) IFN pretreatment enhances the differences in infection rates, as measured by HA staining, between RTF2-KO cells and WT A549/RTF2-rescued cells. (B) Cells pretreated with IFN-β or mock treated. (C) Cells pretreated with IFN-β with or without 0.5-μg/ml B18R, an IFNAR decoy protein. (D) Cells pretreated with IFN-β with or without 5 μM ruxolitinib, a JAK inhibitor. Shown are combined data from two independent experiments with three technical replicates each. P values were determined by one-way ANOVA followed by an unpaired t test with Bonferroni’s correction between RTF2-KO and RTF2-rescued cells. ****, P ≤ 0.0001; **, P ≤ 0.01. (E) The RTF2 mRNA level (normalized to the RPS11 mRNA level) appears to be unchanged after 24 h of 200-U/ml IFN-β exposure. (F) Cells were treated with IFN-β (at 200 U/ml for 24 h), infected with IAV (at an MOI of 5 for 16 h), or pretreated with IFN-β for 24 h before IAV infection, before RNA was extracted for qPCR to quantify RTF2 mRNA and RPS11 mRNA levels. P values were determined by one-way ANOVA and Bonferroni’s multiple-comparison test. **, P ≤ 0.01; *, P ≤ 0.05. (G) Western blot showing a reduction of the RTF2 protein level in IAV-infected cells but no such decrease in cells exposed to just IFN-β alone. WCL, whole-cell lysate. (H) Principal-component analysis of gene expression profiles of WT A549, RTF2-KO, and RTF2-rescued cells under different conditions. The top positively weighted genes in PC1 are those for IFITM1, OAS2, and Mx1, and the top positively weighted genes in PC2 are those for NGFR, FOS, and TNFRSF10D. (I) ISG score based on RNA sequencing of WT A549, RTF2-KO, and RTF2-rescued cells with or without IFN (200-U/ml IFN-β for 24 h). ****, P ≤ 0.0001; **, P ≤ 0.01; *, P ≤ 0.05. (J) Comparison of ISGs that are differentially expressed in IFN-β-pretreated RTF2-KO and RTF2-rescued cells. Shown are three technical replicates with the row-normalized values for each gene. (K) Western blot showing phosphorylated STAT1 (pSTAT1; pY701), IFITM1, IFITM3, and STING levels in IFN-β-treated WT A549, RTF2-KO, and RTF2-rescued cells. Cells were mock treated or exposed to 200 U/ml IFN-β for 24 h. (L) Quantification of the pSTAT1 band in panel K.

Fig 5: Early and late dedifferentiation biomarkers predict human chondrocyte plasticity. a Confirmation of the immunostained scRNA-seq-defined markers RBP4, SOD3, IFITM3, and F-actin in human P0-8 chondrocytes, as a human chondrocyte dedifferentiation model; scale bars: 50 μm. b Quantitative data of immunostained scRNA-seq-defined markers, cell/nucleus size and cell relative viability in human P0-8 chondrocytes. c Immunostained RBP4, SOD3, IFITM3, and F-actin in human articular chondrocytes from 2 other donors (Donor 1 and 2) for phenotype evaluation; scale bars: 50 μm. d Heatmap of captured parameters in the human chondrocyte dedifferentiation model and articular chondrocyte samples from 2 donors, including immunostained RBP4, SOD3, IFITM3 and F-actin intensity, as well as cell size and nucleus size. Each cell represents the detection of one biomarker in one cell. Each row represents a round of detection. Twenty cells were detected for one biomarker in each batch. e, f Sample-to-sample distance heatmap and principal component analysis (PCA) plot showing the similarity of Donor-1 and Donor-2 chondrocytes with the human chondrocyte dedifferentiation model (P0-8). g Alcian blue–stained micromasses formed by the human chondrocyte dedifferentiation model (P0-8) and Donor-1 and 2 chondrocytes; scale bars: 3 mm. All data are the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001.