Several signaling pathways associated with cell-intrinsic defense are activated in the intestinal epithelium against Cryptosporidium following its infection, including the TLR/NF-кB signaling and IFN signaling (17, 18). Whereas signaling of all IFN types are activated in the intestine during in vivo Cryptosporidium infection, only IFN-I and IFN-III are activated in vitro infection models using intestinal epithelial cells because these cells in culture themselves cannot produce IFN-γ (10, 17, 18). In our previous studies, we demonstrated that knockdown of the lncRNA Nostrill using siRNA leads to a higher parasite burden of Cryptosporidium in intestinal epithelial cells (24). This suggested that Nostrill might play a role in the cell-intrinsic anti-parasitic defense. Expanding on this, we carried out comprehensive study by examining the transcriptome of cultured mouse intestinal epithelial IEC4.1 cells cells following C. parvum infection at a genome-wide level. These cells were subjected to treatment with a pool of siRNAs to knockdown Nostrill (referred as siR_Nostrill), followed by exposure to C. parvum infection for 24h. Consistent with our previous observations (25), siRNA knockdown of Nostrill in IEC4.1 cells increased the parasite infection burden ( Supplementary Figure 1 ). The infection also triggered substantial changes in the gene expression patterns in IEC4.1 cells. Specifically, while comparing the gene expression between cells treated with siR_Ctrl (scramble siRNA control) with or without C. parvum infection, we identified a significant alteration in the expression levels of 11,761 genes. This shift encompassed both downregulation (5,590 genes) and upregulation (6,171 genes), demonstrating statistical significance at p<0.05 and a fold change >1 ( Figure 1A ). Intriguingly, the extent of gene alterations escalated further in infected cells subjected to siR_Nostrill treatment, revealing a higher count of changes. This encompassed both downregulated (5,791 genes) and upregulated (7,174 genes) entities, all adhering to the same rigorous statistical criteria ( Figure 1B ).

Next, through Gene Set Enrichment Analysis (GSEA), we investigated the gene expression profiles in IEC4.1 cells treated with the siR_Ctrl upon infection. In consistent with results from previous studies (17, 18), our analysis revealed activation of several innate defense pathways at the cellular level, such as IFN-mediated signaling and other cytokine-mediated signaling, in the infected cells ( Figure 1C ). Nonetheless, it was observed that pathways related to cellular defense appeared to be suppressed in cells treated with siR_Nostrill following infection, compared with infected cells treated with siR_Ctrl ( Figure 1D ). Genes whose expression levels were significantly altered and differentially expressed between siR_Ctrl-treated IEC4.1 and siR_Nostrill-treated IEC4.1 cells following infection were listed in Supplementary Table 1 . It’s noteworthy that many of the well-known ISGs exhibited significant changes unique to each sample type. To visualize these changes, we employed unsupervised hierarchical clustering analysis, which generated a heat map illustrating the expression patterns of these differentially expressed immune defense genes, including many ISGs ( Figure 1E ). Intriguingly, genes such as Ifi204, Irgm1/2, Nos2, Gbp3/7, Usp18, Irf7, Stat1, and Isg15, which were upregulated in response to infection in cells treated with siR_Ctrl, showed decreased upregulation in cells treated with siR_Nostrill following infection. These findings highlight distinct gene expression profiles between infected siR_Ctrl-treated and siR_Nostrill-treated IEC4.1 cells, indicating a compromised cell-intrinsic defense response in cells treated with siR_Nostrill.

We then asked whether Nostrill induction acts as a host defense response to Cryptosporidium infection and can modulate IFN-mediated cell-intrinsic anti-parasitic defense in intestinal epithelial cells. Whereas signaling of all IFN types are activated in the intestine during in vivo Cryptosporidium infection, only IFN-I and IFN-III are activated in vitro infection models using intestinal epithelial cells because these cells in culture themselves cannot produce IFN-γ (10). Taken advantage on the absence of IFN-γ in the in vitro infection models, we investigated the impact of Nostrill induction on cell-intrinsic anti-parasitic defense in response to exogenous IFN-γ using our in vitro IEC4.1 cell infection model. We first performed a genome-wide transcriptome analysis of cultured IEC4.1 cells treated with the siR_Ctrl or siR_Nostrill in response to IFN-γ. Upon comparing the gene expression patterns between the siR_Ctrl-treated cells with and without IFN-γ stimulation, we have identified many genes that displayed significant differential expression (p<0.05, fold change > 0.5) ( Figure 2A ). Most of these genes are upregulated, confirming the impact of IFN-γ treatment on the cells in consistent with previous studies (26–28). When comparing the basal gene expression levels between the siR_Ctrl- and siR_Nostrill-treated cells, without IFN-γ treatment, we observed 186 genes that exhibited differential expression ( Figure 2B ). Nevertheless, when comparing expression levels of genes in the siR_Ctrl + IFN-γ-treated cells versus siR_Nostrill + IFN-γ-treated cells, a total of 1,559 genes showed significant differential expression, indicating a notable impact of Nostrill on the gene expression induced by IFN-γ treatment ( Figure 2C ). Out of these genes, 883 were genes whose expression levels were significantly lower in the siR_Nostrill + IFN-γ-treated cells than that in the siR_Ctrl + IFN-γ-treated cells ( Figure 2C ). The normalized transcript expression levels of selected genes out of these 883 genes were shown in Figure 2D , including Igtp, iNos, and Gadd45g. The complete dataset, containing a comprehensive list of genes showing differential expression between these groups, is available in the Supplementary Table 2 .

We then performed gene ontology (GO) analysis to understand the functional enrichment and biological pathways associated with these differentially expressed genes associated with Nostrill in IFN-γ-treated cells. The PANTHER GO-slim tool revealed that signaling pathways associated with cellular process, biological regulation, and immune system process were generally suppressed in siR_Nostrill-treated IEC4.1 cells upon IFN-γ stimulation, compared with that in siR_Ctrl- and IFN-γ-treated cells ( Figure 2E ). To get a deeper understanding of these immune system processes, we integrated GO and Kyoto Encyclopedia of Genes and Genomes (KEGG) databases to perform functional analysis. The GO KEGG analysis revealed various immune system processes that were downregulated in siR_Nostrill-treated cells in response to IFN-γ, including IFN-γ-mediated cellular responses, regulation of MAPK cascade and autophagy mediators ( Figure 2F ). Expression levels of selected genes associated with cell-intrinsic defense were further validated at the RNA level in cells treated with siR_Nostrill or cells overexpressing Nostrill, using quantitative real-time reverse transcription PCR (qRT-PCR) ( Supplementary Figure 1 ).

The canonical IFN-γ signaling utilizes the JAK/STAT pathway to activate Stat1, resulting in the formation of active Stat1 homodimers (5). This complex then binds to gamma-activated sites in the promoters/enhancers of ISGs (5). To determine whether Nostrill and the Stat1 transcription factor have a direct physical relationship with one another, we performed formaldehyde crosslinking RNA immunoprecipitation (RIP) analysis. We observed that the anti-Stat1 immunoprecipitates from IFN-γ-stimulated cells contained a significant amount of Nostrill. In IFN-γ-stimulated cells, immunoprecipitation of Stat1 indicated a 1.8-fold increase in Nostrill enrichment as compared to the untreated control cells ( Figure 3A ).

To determine if Nostrill binds to specific genomic sites in the promoters of ISGs, potentially facilitating the binding of Stat1, we employed the Chromatin Isolation by RNA Purification (ChIRP) technique. We used biotinylated probes specific to Nostrill to isolate chromatin fragments and then measured the enrichment of Nostrill to the promoter regions of selected ISGs, such as, Igtp, iNos, and Gadd45g. A pool of eight biotinylated tiling oligonucleotide probes unique to Nostrill were utilized to isolate the associated chromatin fragments and a pool of six LacZ probes were used as control. Following the Nostrill pull-down, we conducted qRT-PCR assays employing primers designed to target the Stat1 binding sites within the respective gene promoters. The findings revealed enhanced associations of Nostrill with the promoter regions of these genes in response to IFN-γ stimulation ( Figures 3B–D ). Notably, there was an amplified interaction between Nostrill and the Igtp promoter regions 5 and 6 following IFN-γ stimulation ( Figure 3B ). In the case of the Gadd45g promoter site 5, a noticeable increase in Nostrill abundance was observed ( Figure 3C ). Likewise, the iNos promoter region 1 exhibited an increased enrichment of Nostrill upon IFN-γ stimulation ( Figure 3D ).

To further investigate the potential involvement of Nostrill in docking Stat1 at the promoter regions of three genes: Igtp, iNos, and Gadd45g, we employed chromatin immunoprecipitation (ChIP) to assess Stat1 recruitment in IEC4.1 cells treated with siR_Nostrill ( Figures 3E, F ). The results showed that in siR_Ctrl-transfected cells stimulated with IFN-γ, there was a significant increase of Stat1 enrichment at the site 5 and site 6 regions of the Igtp promoter compared to the non-IFN-γ-treated control (p = 0.05) ( Figure 3E ). This suggests that Stat1 plays a role in the transcriptional regulation of Igtp in response to IFN-γ. Furthermore, when Nostrill expression was silenced in IEC4.1 cells and then exposed to IFN-γ stimulation, the enrichment of Stat1 at the site 6 region of the Igtp promoter was significantly reduced (p = 0.02). Of note, Nostrill was also recruited to the same site 6 region in the IFN-γ-treated IEC4.1 cells ( Figure 3B ). This indicates that Nostrill is involved in the docking or recruitment of Stat1 to the Igtp transcriptional region, potentially acting as a co-regulator. Similar observations were made for the Gadd45g and iNos promoter regions. Following IFN-γ stimulation, enrichment of Stat1 at the site 5 region of the Gadd45g promoter showed a significant increase ( Figure 3F ), while at the site 1 region of the iNos promoter, it increased significantly in the siR_Ctrl-transfected IEC4.1 cells ( Figure 3G ). However, silencing of Nostrill in cells significantly reduced the enrichment of Stat1 at these corresponding regions of the Gadd45g and iNos promoters ( Figures 3F, G ). Recruitment of Nostrill was observed to the same regions of the Gadd45g and iNos promoters in IFN-γ-treated IEC4.1 cells ( Figures 3C, D ). These findings suggest that Nostrill interacts with Stat1 and may function as a co-regulator for the chromatin enrichment of Stat1 to promote IFN-γ-stimulated gene transcription in intestinal epithelial cells.

Our lab has previously demonstrated the induction of Nostrill in intestinal epithelial cells in response to C. parvum infection (24). To this end, IEC4.1 cells were infected with C. parvum for up to 24h. At 24h following infection, the Nostrill’s induction was apparent, demonstrating a 72% increase in its expression ( Figure 4A ). To investigate the expression profile of Nostrill in a relevant in vivo model and gain insights into its potential role in the context of intestinal epithelium, neonatal mice (5 days old) were infected with C. parvum via oral gavage. Infection of the small intestinal epithelium by C. parvum was evident by immunostaining ( Figure 4B ). Consistent with previous reports (24), the expression of Nostrill was significantly increased at 24h after the infection in the intestinal epithelium ( Figure 4C ).

Given the significant impact of Nostrill on IFN-γ-mediated gene transcription in intestinal epithelial cells, we sought to ask whether Nostrill induction may influence IFN-γ-mediated intestinal epithelial cell-intrinsic defense against C. parvum infection. To investigate this, we initially primed siR_Ctrl- and siR_Nostrill-treated IEC4.1 cells with IFN-γ and subsequently infected them with C. parvum, followed by an assessment of the resulting infection burden. Consistent with previous reports (25, 29), we detected a decreased infection burden in siR_Ctrl-treated IEC4.1 cells primed with IFN-γ ( Figure 4D ). Intriguingly, knockdown of Nostrill with the siRNA in IEC4.1 cells resulted in a significant suppression of IFN-γ-mediated epithelial defense, reflected by a higher infection burden in siR_Nostrill-treated and IFN-γ-primed cells compared with that in IFN-γ-primed cells treated with the siR_Ctrl ( Figure 4D ). Complementarily, overexpression of Nostrill resulted in enhanced IFN-γ-mediated inhibition of infection burden in IEC4.1 cells ( Figure 4E ). These data suggest that Nostrill promotes IFN-γ-mediated epithelial cell-intrinsic anti-Cryptosporidium defense.

Autophagy has been recognized as one of the immune responses elicited by IFN-γ against intracellular pathogens (30). Several genes, whose expression in response to IFN-γ in intestinal epithelial cells was associated with Nostrill, including Igtp, iNos, and Gadd45g, have been implicated in the regulation of cellular autophagy (30, 31). To examine the role of Nostrill in IFN-γ-induced autophagy, we first assessed the autophagy levels in IEC4.1 cells following exposure to IFN-γ at a dose of 10 ng/mL for 24h. We quantified autophagosome formation using LC3-II release as the marker as previously reported (32, 33). Our results clearly demonstrated the induction of autophagy in IEC4.1 cells following IFN-γ stimulation, as corroborated by both Western blot and qRT-PCR analyses of LC3-II release ( Figures 5A, B , Supplementary Figure 2 ). Subsequently, we sought to determine the influence of Nostrill on this process. To do so, we utilized siR_Nostrill to knockdown its expression in the cells before exposure to IFN-γ. Our findings revealed a significant inhibition of IFN-γ-induced autophagy in cells treated with siR_Nostrill ( Figures 5A, B ).

To further investigate the contribution of Nostrill-mediated genes, including Igtp, iNos, and Gadd45g, to IFN-γ-induced autophagy, we employed a combination of siRNAs (siR_ISGs) to silence their expression in IEC4.1 cells ( Supplementary Figure 3 ). We then examined the impact of siR_ISG knockdown on IFN-γ-induced autophagy. Silencing Igtp, iNos, and Gadd45g with the siR_ISGs led to a significant inhibition of IFN-γ-induced autophagy ( Figure 5C ).

The subsequent question we aimed to address was whether these Nostrill-associated ISGs within the IFN-γ pathway play a role in combating Cryptosporidium infection through autophagy. To investigate this, we first assessed the expression of LC3II in IEC4.1 cells treated with IFN-γ and subsequently infected with C. parvum. As shown in Figure 5D , IFN-γ treatment led to a reduction in C. parvum burden in the cells. Meanwhile, there was a significant increase in the LC3II levels in the IFN-γ-treated cells following C. parvum infection ( Figure 5D ). This suggests that IFN-γ may reduce C. parvum burden in the cells through induction of autophagy. To further substantiate the significance of IFN-γ-induced autophagy in the defense against Cryptosporidium, we assessed both the Cryptosporidium infection burden and the LC3II levels in cells with or without treatment with the siR_ISGs. IEC4.1 cells were first treated with siR_Ctrl or siR_ISGs for 24h, pre-treated with IFN-γ for additional 24h, and subsequently infected with Cryptosporidium for 24h. Similar to our observations in Nostrill knockdown cells, the knockdown of selected ISGs in IEC4.1 cells resulted in a significant suppression of IFN-γ-mediated epithelial defense, as evidenced by a higher infection burden in the siR_ISGs-treated and IFN-γ primed cells compared to those treated with the siR_Ctrl ( Figure 5E ). Furthermore, we observed a reduction in LC3II levels in cells transfected with siR_ISGs, specifically when treated with IFN-γ and subsequently infected with C. parvum ( Figure 5F ). Taken together, the above data suggest that Nostrill facilitates the transcription of Igtp, Gadd45g, and iNos induced by IFN-γ, which in turn positively regulates autophagy, ultimately contributing to the epithelial cell-intrinsic defense against Cryptosporidium infection.

Dr. Pingchang Yang of McMaster University in Hamilton, Canada, generously provided the newborn mouse intestinal epithelial cell line (IEC4.1) (52). Cultures of IEC4.1 cells were maintained in DMEM media with 10% fetal bovine serum (FBS; Gibco; Carlsbad, CA, USA). Streptomycin at 100 μg/ml and penicillin at 100 U/ml (both from Gibco) were added to the growth media. Tissue culture flasks were used to cultivate the cells in a 37°C, 5% CO₂ incubator.

Cryptosporidium parvum oocysts, sourced from the Iowa strain provided by Harley Moon at the National Animal Disease Center (Ames, IA), were purchased commercially from the Bunch Grass Farms in Deary, ID. To purify the oocysts, a modified ether extraction technique was employed. Subsequently, the oocysts were suspended in phosphate-buffered saline (PBS) and stored at a temperature of 4°C. For both in vitro and in vivo infection experiments, the oocysts were subjected to a treatment with 1% sodium hypochlorite while being preserved on ice for 20 minutes. After this treatment, the oocysts were washed three times using Dulbecco’s modified Eagle medium (DMEM) culture media.

For in vitro infection studies using IEC4.1 cells, viable C. parvum oocysts were utilized. These oocysts were treated with 1% sodium hypochlorite and were added to a culture medium consisting of DMEM-F-12 supplemented with 100 U/ml penicillin and 100 μg/ml streptomycin. The infection was established by combining oocysts and host cells in a 1:1 ratio. The cell cultures were then incubated at 37°C for 4h to facilitate parasite attachment and invasion. After this, the cells were thoroughly washed with DMEM-F-12 medium three times to remove any free parasites. The cells were further cultured for varying time periods based on the experimental requirements. For the in vivo experiments, a neonatal murine model of intestinal cryptosporidiosis was employed (25). Neonates, aged 5 days after birth, were orally administered C. parvum oocysts (10^5 oocysts per mouse) through gavage to induce intestinal cryptosporidiosis. Control mice were given PBS via the same method. Following 24, 48, and 72h from the administration of C. parvum oocysts or PBS, the animals were sacrificed, and samples of ileum intestine tissues were collected. Epithelial tissues from the ileum were extracted from at least five animals within each experimental group. The assessment of C. parvum was done using qRT-PCR both in vitro and in vivo context.

To silence the Nostrill gene in murine cells, a pool of three small interfering RNA (siRNA) duplexes (siR_Nostrill) was synthesized using the WI siRNA Selection Program at the Massachusetts Institute of Technology. The specific siRNA sequences used were as follows: 25b_1: sense 5’ CCUGGGAAGAAGCAUUAAU 3’, 25b_2: sense 5’ CCACCAGGAACACACAAAU 3’, and 25b_3: sense 5’ GCCUAUAAGUCGUUCUAAU 3’. The IEC4.1 cells were cultured in 24-well plates at a density of 1.5 x 10^5 cells per milliliter. The next day, the cells were transfected with 120 nM of the siRNA pool (40 nM of each of the siRNA) using Lipofectamine RNAiMax from Life Technologies, Inc., following the manufacturer’s instructions. As a negative control, non-specific scrambled siRNA at a concentration of 120 nM was used as the control siRNA (siR_Ctrl). The transfection process was allowed to continue for 24h, after which the cells were treated with IFN-γ. RNA was then isolated from the cells using a previously described method.

Igtp, iNos, and Gadd45g were targeted for knockdown of ISGs in mouse IEC4.1 cells. The siRNA-mediated knockdown approach was employed to silence these genes. The siRNA to iNos was synthesized using the WI siRNA Selection Program at the Massachusetts Institute of Technology (MIT) and the sequence was sense 5’ GCCGCUCUAAUACUUAGCU 3’. For silencing Igtp and Gadd45g, siRNA duplexes were purchased from Santacruz Biotechnology with the catalogue numbers sc-41792 and sc-37419, respectively. The IEC4.1 cells were cultured in 24-well plates at a density of 1.5 x 10^5 cells/ml. After overnight incubation, the cells were transfected with a pool of the three sets of siRNAs to achieve knockdown of Igtp, Gadd45g, and iNos (as siR_ISGs). The concentrations of gene-specific siRNAs used for transfection were as follows: siR_Igtp at 80 nM, siR_Gadd45g at 120 nM, and siR_Nos at 60 nM. siR_Ctrl at a concentration of 260 nM was used for control. Lipofectamine RNAiMax from Life Technologies, Inc. was used as the transfection reagent, and the transfection process was carried out following the manufacturer’s instructions.

Total RNA from the cells was extracted using the TRIzol reagent from Invitrogen as per the manufacturer’s instructions. The concentration and quality of the RNA were assessed using a spectrophotometer from Beckman in Brea, CA. For subsequent steps, 500 ng of the RNA sample was utilized as a template for cDNA synthesis, which was performed using the M-MLV Reverse Transcriptase kit from Invitrogen, located in Carlsbad, CA. The reaction was carried out in a total volume of 20 µl.

The qRT-PCR was conducted using the Invitrogen™ SYBR GreenER™ qPCR SuperMix Universal from Thermo Fisher Scientific in Waltham, MA. The qRT-PCR was performed on the Bio-Rad CFX96 Touch™ Real-Time PCR Detection System. The expression levels of the RNA samples were determined using the threshold cycle (ΔΔCT) method and normalized to the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (Gapdh). The primer sequences are displayed in Supplementary Table 3 .

RNA-Seq was performed in biological triplicates in IEC4.1 cells treated with siR_Ctrl or siR_Nostrill. Total RNA molecules were isolated using Trizol reagent. The BGI Americas Corporation in Cambridge, MA, carried out transcriptome sequencing and data processing using DNBSEQ™ sequencing technology platforms. To assess RNA quality, a 2100 Bioanalyzer from Agilent Technologies was employed. RNA samples meeting the criteria of consistent visual RNA patterns, an RNA Integrity Number (RIN) greater than 7, and a concentration exceeding 20 ng/μl were categorized as having good quality. The complete RNA was broken down into smaller fragments, and specific mRNA molecules were concentrated using magnetic beads with oligo (dT) sequences. Subsequently, complementary DNA (cDNA) was synthesized from the enriched mRNA. This double-stranded cDNA underwent purification and was amplified using PCR. The sequenced library products were processed on the DNBSEQ-500 platform with paired-end reads of 100 base pairs, generating an average of approximately 4.94 Gb per sample. To ensure data quality, subpar reads were excluded. This included reads with base qualities lower than 10 for over 20% of their bases, reads with adaptors, and those containing more than 5% bases marked as unknown (N bases). This filtration was achieved using internal software called SOAPnuke (version: v1.5.2) on the DNBSEQ-500 platform, with specific parameters (-I 15-q0.5-n0.1). The resultant high-quality reads were stored in the FASTQ format. Reads were aligned to the Mus_musculus genome (GCF_000001635.26_GRCm38.p6) from the NCBI database using two different software tools: HISAT2 (version: v2.0.4) and Bowtie2 (version: v2.2.5). The alignment was guided by parameters tailored to the specifics of the process (-phred64-sensitive-no-discordant-no-mixed-I 1-X 1000). Raw RNA-seq data were further refined by filtering out unreliable sites, accomplished through the application of the GATK program. This step targeted the removal of low-quality reads from each sample. Finally, the gene read counts were normalized to RPKM, which represents the expression level of genes in terms of reads per kilobase of transcript per Million mapped reads. This normalization allowed for meaningful comparison of relative gene expression levels across samples.

Differential gene-expression analyses were conducted using DESeq2. Differences in gene expression, whether increased or decreased, were quantified using logarithmically transformed fold change values. These values were represented as either log2FC (log base 2) or log10FC (log base 10), which were computed as the logarithmic difference between gene expression values under distinct treatment conditions (log2FC = log2[B] - log2[A] or log10FC = log10[B] - log10[A], where A and B denote the gene expression values). GO enrichment analysis was performed using PANTHER. To further analyze the data, Gene Set Enrichment Analysis (GSEA) was conducted using the “clusterProfiler” R package. This analysis was carried out in a pre-ranked manner, where all genes were ranked based on their log2FC values derived from the differential expression analysis. The weighted Kolmogorov-Smirnov test was applied, and p-values were adjusted using the Benjamini-Hochberg method. The findings from GSEA were depicted visually using the “ggplot2” R package, offering insights into patterns of gene set enrichment. Additionally, a volcano plot was generated with the “ggplot2” package, utilizing log2FC and adjusted p-values from the differential expression analysis to visually convey the significance and extent of gene expression alterations. For a comprehensive overview of gene expression trends, heatmaps were generated using the “ComplexHeatmap” R package. These heatmaps visually summarized gene expression patterns across different conditions. Furthermore, the expression z-score was computed based on relative gene expression levels. This score aids in standardizing and comparing gene expression values across diverse samples or conditions, facilitating more meaningful interpretation of expression patterns.

The formaldehyde crosslinking RNA immunoprecipitation (RIP) assay was conducted as described previously (22, 53). To summarize, the process involved the initial cleaning of lysates using 20 μl of Magna ChIP Protein A + G Magnetic Beads (Millipore, Massachusetts) that had been pre-washed with PBS. Subsequently, the pre-cleaned lysate (250 μl) was mixed with an equivalent volume of the whole cell extract buffer (250 μl). This mixture was then combined with beads coated with an antibody to Stat1 (#9172S, Cell Signaling, Danvers, MA). The entire assembly was then rotated at 4°C overnight. Following this, the mixture underwent four rounds of washing using the whole cell extract buffer supplemented with protease and RNase inhibitors. The immunoprecipitated ribonucleoprotein (RNP) complexes and the input material were subjected to digestion in RNA PK Buffer pH 7.0 (comprising 100 mM NaCl, 10 mM TrisCl pH 7.0, 1 mM EDTA, and 0.5% SDS) in the presence of 10 μg of Proteinase K. This mixture was then incubated at 50°C for 45 minutes with continuous shaking at 400 rpm. The formaldehyde cross-links were subsequently reversed through incubation at 65°C with rotation for 4 hours. RNA was isolated from the samples using Trizol, following the protocol provided by Invitrogen. The presence of RNA was evaluated using qRT-PCR, conducted with the CFX96 Touch™ Real-Time PCR Detection System from Bio-Rad. The specific primer pairs for PCR can be found in Supplementary Table 3 .

The detailed protocol is described elsewhere (22, 54). Briefly, cells were treated with 1% formaldehyde for 10 minutes to cross-link the proteins to the DNA, preserving the protein-DNA interactions at the time of fixation. The fixed cells were collected in 1x ice-cold PBS (Phosphate-Buffered Saline) and resuspended in an SDS (Sodium Dodecyl Sulfate) lysis buffer. The genomic DNA was then fragmented into smaller pieces through sonication resulting in fragments ranging from 200 to 500 base pairs in length. One percent of the cell extract was kept aside as input. The cell extracts containing the sheared DNA are incubated overnight at 4°C with anti-Stat1 and added protein G-agarose beads to precipitate the antibody-protein-DNA complexes. The immunoprecipitates (antibody-protein-DNA complexes bound to the beads) were washed sequentially to remove any non-specific interactions. The washing steps involve low-salt buffer, high-salt buffer, LiCl buffer, and Tris EDTA buffer. The DNA-protein complexes were eluted from the beads afterwards, separating the DNA from the proteins. The proteins in the eluted solution were then digested at 45°C for 1hr using proteinase K. To detect and quantify the DNA, qRT-PCR analysis was employed, which allowed for the measurement of the DNA content, enabling the quantification of protein binding levels at particular regions of the genome.

Chromatin isolation by RNA purification (ChIRP) analysis was conducted as described in a previous study (22, 55). In summary, chromatin was isolated after cross-linking the cells with glutaraldehyde. A set of tiling oligonucleotide probes were designed to specifically bind to the Nostrill sequence. The sequences for these probes are available in Supplementary Table 1 . Verification of the DNA sequences in the chromatin immunoprecipitates was performed using qRT-PCR. The same primers that cover the gene promoter regions of interest, similar to the ChIP analysis, were utlized for the ChIRP analyses. As a control, a collection of scrambled oligonucleotide probes targeting LacZ (listed in Supplementary Table 3 ) was used.

Cell Culture and Lysate Preparation: Cells were plated in six-well plates with 2 ml of medium per well to achieve 80%–90% confluency within 48h. After 24h, the culture medium was replaced with treatment (such as, IFN-γ at a dose of 10 ng/ml) or control medium and incubated for an additional 24h. Two hours prior to collection, 1:100 dilutions of the stock solutions of leupeptin (10 mM) and NH4Cl (2 M) were added to the medium to yield final concentrations of 100 μM and 20 mM, respectively. The cell pellet was collected using EDTA/PBS, washed with ice-cold PBS, and lysed with pre-cooled lysis buffer. Lysates were sonicated for 10–15 sec or treated with syringe to complete cell lysis and shear DNA to reduce sample viscosity. The lysate was centrifuged at 12,000 rpm for 10 minutes at 4°C, and the supernatant was transferred to a new tube. Protein concentration was determined, samples were adjusted for equal concentrations, mixed with loading buffer, and boiled before storing at -80°C.

For SDS-PAGE and Western blotting, equal protein amounts (10–20 μg) were loaded onto the SDS-PAGE gel with a molecular weight marker. The gel was run at 95 V for 20 minutes and then at 150 V for approximately 1–2 hours until the dye reached 1/4 inch from the bottom of the gel. LC3-II migrated at ~14 kD. The separated proteins were transferred onto a PVDF membrane, chosen for its enhanced LC3-II binding. Following transfer, the membrane was blocked with 5% BSA blocking buffer for one hour at room temperature and probed with primary anti-LC3II antibody (1:1000, cat#2775S from Cell Signaling) overnight at 4°C. The membrane was washed three times for 15 minutes each with TBST at room temperature, probed with secondary HRP-conjugated anti-rabbit IgG antibody (1:5000) for one hour at room temperature, and washed again. An ECL solution mixture was applied to the membrane, and the emitted light was captured. The levels of LC3-II in each sample were normalized to β-actin (1:1000, cat#A1978 from Cell Signaling) for analysis.

Data are given as mean ± SEM. from at least three independent experiments or biological replicates. The two-tailed unpaired Student’s t test was used for compassion between two groups and one-way ANOVA were used for compassion among three or more groups. p values < 0.05 were considered statistically significant.