SARS-CoV-2 Nsp15 antagonizes the cGAS-STING-mediated antiviral innate immune responses.

Coronavirus (CoV) Nsp15 is a viral endoribonuclease (EndoU) with a preference for uridine residues. CoV Nsp15 is an innate immune antagonist which prevents dsRNA sensor recognition and stress granule formation by targeting viral and host RNAs. SARS-CoV-2 restricts and delays the host antiviral innate immune responses through multiple viral proteins, but the role of SARS-CoV-2 Nsp15 in innate immune evasion is not completely understood. Here, we generate an EndoU activity knockout rSARS-CoV-2Nsp15-H234A to elucidate the biological functions of Nsp15. Relative to wild-type rSARS-CoV-2, replication of rSARS-CoV-2Nsp15-H234A was significantly decreased in IFN-responsive A549-ACE2 cells but not in its STAT1 knockout counterpart. Transcriptomic analysis revealed upregulation of innate immune response genes in cells infected with rSARS-CoV-2Nsp15-H234A relative to wild-type virus, including cGAS-STING, cytosolic DNA sensors activated by both DNA and RNA viruses. Treatment with STING inhibitors H-151 and SN-011 rescued the attenuated phenotype of rSARS-CoV-2Nsp15-H234A. SARS-CoV-2 Nsp15 inhibited cGAS-STING-mediated IFN-ß promoter and NF-κB reporter activity, as well as facilitated the replication of EV-D68 and NDV by diminishing cGAS and STING expression and downstream innate immune responses. Notably, the decline in cGAS and STING was also apparent during SARS-CoV-2 infection. The EndoU activity was essential for SARS-CoV-2 Nsp15-mediated cGAS and STING downregulation, but not all HCoV Nsp15 share the consistent substrate selectivity. In the hamster model, rSARS-CoV-2Nsp15-H234A replicated to lower titers in the nasal turbinates and lungs and induced higher innate immune responses. Collectively, our findings exhibit that SARS-CoV-2 Nsp15 serves as a host innate immune antagonist by targeting host cGAS and STING.

Graph Fourier transform for spatial omics representation and analyses of complex organs.

Spatial omics technologies decipher functional components of complex organs at cellular and subcellular resolutions. We introduce Spatial Graph Fourier Transform (SpaGFT) and apply graph signal processing to a wide range of spatial omics profiling platforms to generate their interpretable representations. This representation supports spatially variable gene identification and improves gene expression imputation, outperforming existing tools in analyzing human and mouse spatial transcriptomics data. SpaGFT can identify immunological regions for B cell maturation in human lymph nodes Visium data and characterize variations in secondary follicles using in-house human tonsil CODEX data. Furthermore, it can be integrated seamlessly into other machine learning frameworks, enhancing accuracy in spatial domain identification, cell type annotation, and subcellular feature inference by up to 40%. Notably, SpaGFT detects rare subcellular organelles, such as Cajal bodies and Set1/COMPASS complexes, in high-resolution spatial proteomics data. This approach provides an explainable graph representation method for exploring tissue biology and function.

A Hitchhiker's guide to high-dimensional tissue imaging with multiplexed ion beam imaging.

Advancements in multiplexed tissue imaging technologies are vital in shaping our understanding of tissue microenvironmental influences in disease contexts. These technologies now allow us to relate the phenotype of individual cells to their higher-order roles in tissue organization and function. Multiplexed Ion Beam Imaging (MIBI) is one of such technologies, which uses metal isotope-labeled antibodies and secondary ion mass spectrometry (SIMS) to image more than 40 protein markers simultaneously within a single tissue section. Here, we describe an optimized MIBI workflow for high-plex analysis of Formalin-Fixed Paraffin-Embedded (FFPE) tissues following antigen retrieval, metal isotope-conjugated antibody staining, imaging using the MIBI instrument, and subsequent data processing and analysis. While this workflow is focused on imaging human FFPE samples using the MIBI, this workflow can be easily extended to model systems, biological questions, and multiplexed imaging modalities.

Cross-domain information fusion for enhanced cell population delineation in single-cell spatial-omics data.

Cell population delineation and identification is an essential step in single-cell and spatial-omics studies. Spatial-omics technologies can simultaneously measure information from three complementary domains related to this task: expression levels of a panel of molecular biomarkers at single-cell resolution, relative positions of cells, and images of tissue sections, but existing computational methods for performing this task on single-cell spatial-omics datasets often relinquish information from one or more domains. The additional reliance on the availability of "atlas" training or reference datasets limits cell type discovery to well-defined but limited cell population labels, thus posing major challenges for using these methods in practice. Successful integration of all three domains presents an opportunity for uncovering cell populations that are functionally stratified by their spatial contexts at cellular and tissue levels: the key motivation for employing spatial-omics technologies in the first place. In this work, we introduce Cell Spatio- and Neighborhood-informed Annotation and Patterning (CellSNAP), a self-supervised computational method that learns a representation vector for each cell in tissue samples measured by spatial-omics technologies at the single-cell or finer resolution. The learned representation vector fuses information about the corresponding cell across all three aforementioned domains. By applying CellSNAP to datasets spanning both spatial proteomic and spatial transcriptomic modalities, and across different tissue types and disease settings, we show that CellSNAP markedly enhances de novo discovery of biologically relevant cell populations at fine granularity, beyond current approaches, by fully integrating cells' molecular profiles with cellular neighborhood and tissue image information.

A Spatial Multi-Modal Dissection of Host-Microbiome Interactions within the Colitis Tissue Microenvironment.

The intricate and dynamic interactions between the host immune system and its microbiome constituents undergo dynamic shifts in response to perturbations to the intestinal tissue environment. Our ability to study these events on the systems level is significantly limited by in situ approaches capable of generating simultaneous insights from both host and microbial communities. Here, we introduce Microbiome Cartography (MicroCart), a framework for simultaneous in situ probing of host features and its microbiome across multiple spatial modalities. We demonstrate MicroCart by comprehensively investigating the alterations in both gut host and microbiome components in a murine model of colitis by coupling MicroCart with spatial proteomics, transcriptomics, and glycomics platforms. Our findings reveal a global but systematic transformation in tissue immune responses, encompassing tissue-level remodeling in response to host immune and epithelial cell state perturbations, and bacterial population shifts, localized inflammatory responses, and metabolic process alterations during colitis. MicroCart enables a deep investigation of the intricate interplay between the host tissue and its microbiome with spatial multiomics.

Epstein-Barr Virus Orchestrates Spatial Reorganization and Immunomodulation within the Classic Hodgkin Lymphoma Tumor Microenvironment.

Classic Hodgkin Lymphoma (cHL) is a tumor composed of rare malignant Hodgkin and Reed-Sternberg (HRS) cells nested within a T-cell rich inflammatory immune infiltrate. cHL is associated with Epstein-Barr Virus (EBV) in 25% of cases. The specific contributions of EBV to the pathogenesis of cHL remain largely unknown, in part due to technical barriers in dissecting the tumor microenvironment (TME) in high detail. Herein, we applied multiplexed ion beam imaging (MIBI) spatial pro-teomics on 6 EBV-positive and 14 EBV-negative cHL samples. We identify key TME features that distinguish between EBV-positive and EBV-negative cHL, including the relative predominance of memory CD8 T cells and increased T-cell dysfunction as a function of spatial proximity to HRS cells. Building upon a larger multi-institutional cohort of 22 EBV-positive and 24 EBV-negative cHL samples, we orthogonally validated our findings through a spatial multi-omics approach, coupling whole transcriptome capture with antibody-defined cell types for tu-mor and T-cell populations within the cHL TME. We delineate contrasting transcriptomic immunological signatures between EBV-positive and EBV-negative cases that differently impact HRS cell proliferation, tumor-immune interactions, and mecha-nisms of T-cell dysregulation and dysfunction. Our multi-modal framework enabled a comprehensive dissection of EBV-linked reorganization and immune evasion within the cHL TME, and highlighted the need to elucidate the cellular and molecular fac-tors of virus-associated tumors, with potential for targeted therapeutic strategies.

Graph Fourier transform for spatial omics representation and analyses of complex organs.

Spatial omics technologies are capable of deciphering detailed components of complex organs or tissue in cellular and subcellular resolution. A robust, interpretable, and unbiased representation method for spatial omics is necessary to illuminate novel investigations into biological functions, whereas a mathematical theory deficiency still exists. We present SpaGFT (Spatial Graph Fourier Transform), which provides a unique analytical feature representation of spatial omics data and elucidates molecular signatures linked to critical biological processes within tissues and cells. It outperformed existing tools in spatially variable gene prediction and gene expression imputation across human/mouse Visium data. Integrating SpaGFT representation into existing machine learning frameworks can enhance up to 40% accuracy of spatial domain identification, cell type annotation, cell-to-spot alignment, and subcellular hallmark inference. SpaGFT identified immunological regions for B cell maturation in human lymph node Visium data, characterized secondary follicle variations from in-house human tonsil CODEX data, and detected extremely rare subcellular organelles such as Cajal body and Set1/COMPASS. This new method lays the groundwork for a new theoretical model in explainable AI, advancing our understanding of tissue organization and function.

MAPS: pathologist-level cell type annotation from tissue images through machine learning.

Highly multiplexed protein imaging is emerging as a potent technique for analyzing protein distribution within cells and tissues in their native context. However, existing cell annotation methods utilizing high-plex spatial proteomics data are resource intensive and necessitate iterative expert input, thereby constraining their scalability and practicality for extensive datasets. We introduce MAPS (Machine learning for Analysis of Proteomics in Spatial biology), a machine learning approach facilitating rapid and precise cell type identification with human-level accuracy from spatial proteomics data. Validated on multiple in-house and publicly available MIBI and CODEX datasets, MAPS outperforms current annotation techniques in terms of speed and accuracy, achieving pathologist-level precision even for typically challenging cell types, including tumor cells of immune origin. By democratizing rapidly deployable and scalable machine learning annotation, MAPS holds significant potential to expedite advances in tissue biology and disease comprehension.

A spatial cell atlas of neuroblastoma reveals developmental, epigenetic and spatial axis of tumor heterogeneity.

Neuroblastoma is a pediatric cancer arising from the developing sympathoadrenal lineage with complex inter- and intra-tumoral heterogeneity. To chart this complexity, we generated a comprehensive cell atlas of 55 neuroblastoma patient tumors, collected from two pediatric cancer institutions, spanning a range of clinical, genetic, and histologic features. Our atlas combines single-cell/nucleus RNA-seq (sc/scRNA-seq), bulk RNA-seq, whole exome sequencing, DNA methylation profiling, spatial transcriptomics, and two spatial proteomic methods. Sc/snRNA-seq revealed three malignant cell states with features of sympathoadrenal lineage development. All of the neuroblastomas had malignant cells that resembled sympathoblasts and the more differentiated adrenergic cells. A subset of tumors had malignant cells in a mesenchymal cell state with molecular features of Schwann cell precursors. DNA methylation profiles defined four groupings of patients, which differ in the degree of malignant cell heterogeneity and clinical outcomes. Using spatial proteomics, we found that neuroblastomas are spatially compartmentalized, with malignant tumor cells sequestered away from immune cells. Finally, we identify spatially restricted signaling patterns in immune cells from spatial transcriptomics. To facilitate the visualization and analysis of our atlas as a resource for further research in neuroblastoma, single cell, and spatial-omics, all data are shared through the Human Tumor Atlas Network Data Commons at www.humantumoratlas.org.

mRNA vaccines against SARS-CoV-2 induce divergent antigen-specific T-cell responses in patients with lung cancer.

BACKGROUND: The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) Omicron variant is highly transmissible and evades pre-established immunity. Messenger RNA (mRNA) vaccination against ancestral strain spike protein can induce intact T-cell immunity against the Omicron variant, but efficacy of booster vaccination in patients with late-stage lung cancer on immune-modulating agents including anti-programmed cell death protein 1(PD-1)/programmed death-ligand 1 (PD-L1) has not yet been elucidated. METHODS: We assessed T-cell responses using a modified activation-induced marker assay, coupled with high-dimension flow cytometry analyses. Peripheral blood mononuclear cells (PBMCs) were stimulated with various viral peptides and antigen-specific T-cell responses were evaluated using flow cytometry. RESULTS: Booster vaccines induced CD8+ T-cell response against the ancestral SARS-CoV-2 strain and Omicron variant in both non-cancer subjects and patients with lung cancer, but only a marginal induction was detected for CD4+ T cells. Importantly, antigen-specific T cells from patients with lung cancer showed distinct subpopulation dynamics with varying degrees of differentiation compared with non-cancer subjects, with evidence of dysfunction. Notably, female-biased T-cell responses were observed. CONCLUSION: We conclude that patients with lung cancer on immunotherapy show a substantial qualitative deviation from non-cancer subjects in their T-cell response to mRNA vaccines, highlighting the need for heightened protective measures for patients with cancer to minimize the risk of breakthrough infection with the Omicron and other future variants.

Silencing of aryl hydrocarbon receptor repressor restrains Th17 cell immunity in autoimmune hepatitis.

Th17-cells play a key role in the pathogenesis of autoimmune hepatitis (AIH). Dysregulation of Th17-cells in AIH is linked to defective response to aryl-hydrocarbon-receptor (AhR) activation. AhR modulates adaptive immunity and is regulated by aryl-hydrocarbon-receptor-repressor (AHRR), which inhibits AhR transcriptional activity. In this study, we investigated whether defective Th17-cell response to AhR derives from aberrant AHRR regulation in AIH. Th17-cells, obtained from the peripheral blood of AIH patients (n = 30) and healthy controls (n = 30) were exposed to AhR endogenous ligands, and their response assessed in the absence or presence of AHRR silencing. Therapeutic effects of AHRR blockade were tested in a model of Concanavalin-A (Con-A)-induced liver injury in humanized mice. AHRR was markedly upregulated in AIH Th17-cells, following exposure to l-kynurenine, an AhR endogenous ligand. In patients, silencing of AHRR boosted Th17-cell response to l-kynurenine, as reflected by increased levels of CYP1A1, the main gene controlled by AhR; and decreased IL17A expression. Blockade of AHRR limited the differentiation of naïve CD4-cells into Th17 lymphocytes; and modulated Th17-cell metabolic profile by increasing the levels of uridine via ATP depletion or pyrimidine salvage. Treatment with 2'-deoxy-2'-fluoro-d-arabinonucleic acid (FANA) oligonucleotides to silence human AHRR in vivo, reduced ALT levels, attenuated lymphocyte infiltration on histology, and heightened frequencies of regulatory immune subsets in NOD/scid/gamma mice, reconstituted with human CD4 cells, and exposed to Con-A. In conclusion, blockade of AHRR in AIH restores Th17-cell response to AHR, and limits Th17-cell differentiation through generation of uridine. In vivo, silencing of AHRR attenuates liver damage in NOD/scid/gamma mice. Blockade of AHRR might therefore represent a novel therapeutic strategy to modulate effector Th17-cell immunity and restore homeostasis in AIH.

MAPS: Pathologist-level cell type annotation from tissue images through machine learning.

Highly multiplexed protein imaging is emerging as a potent technique for analyzing protein distribution within cells and tissues in their native context. However, existing cell annotation methods utilizing high-plex spatial proteomics data are resource intensive and necessitate iterative expert input, thereby constraining their scalability and practicality for extensive datasets. We introduce MAPS (Machine learning for Analysis of Proteomics in Spatial biology), a machine learning approach facilitating rapid and precise cell type identification with human-level accuracy from spatial proteomics data. Validated on multiple in-house and publicly available MIBI and CODEX datasets, MAPS outperforms current annotation techniques in terms of speed and accuracy, achieving pathologist-level precision even for challenging cell types, including tumor cells of immune origin. By democratizing rapidly deployable and scalable machine learning annotation, MAPS holds significant potential to expedite advances in tissue biology and disease comprehension.

Expanded vacuum-stable gels for multiplexed high-resolution spatial histopathology.

Cellular organization and functions encompass multiple scales in vivo. Emerging high-plex imaging technologies are limited in resolving subcellular biomolecular features. Expansion Microscopy (ExM) and related techniques physically expand samples for enhanced spatial resolution, but are challenging to be combined with high-plex imaging technologies to enable integrative multiscaled tissue biology insights. Here, we introduce Expand and comPRESS hydrOgels (ExPRESSO), an ExM framework that allows high-plex protein staining, physical expansion, and removal of water, while retaining the lateral tissue expansion. We demonstrate ExPRESSO imaging of archival clinical tissue samples on Multiplexed Ion Beam Imaging and Imaging Mass Cytometry platforms, with detection capabilities of > 40 markers. Application of ExPRESSO on archival human lymphoid and brain tissues resolved tissue architecture at the subcellular level, particularly that of the blood-brain barrier. ExPRESSO hence provides a platform for extending the analysis compatibility of hydrogel-expanded biospecimens to mass spectrometry, with minimal modifications to protocols and instrumentation.

Morphologic and molecular analysis of liver injury after SARS-CoV-2 vaccination reveals distinct characteristics.

BACKGROUND & AIMS: Liver injury after COVID-19 vaccination is very rare and shows clinical and histomorphological similarities with autoimmune hepatitis (AIH). Little is known about the pathophysiology of COVID-19 vaccine-induced liver injury (VILI) and its relationship to AIH. Therefore, we compared VILI with AIH. METHODS: Formalin-fixed and paraffin-embedded liver biopsy samples from patients with VILI (n = 6) and from patients with an initial diagnosis of AIH (n = 9) were included. Both cohorts were compared by histomorphological evaluation, whole-transcriptome and spatial transcriptome sequencing, multiplex immunofluorescence, and immune repertoire sequencing. RESULTS: Histomorphology was similar in both cohorts but showed more pronounced centrilobular necrosis in VILI. Gene expression profiling showed that mitochondrial metabolism and oxidative stress-related pathways were more and interferon response pathways were less enriched in VILI. Multiplex analysis revealed that inflammation in VILI was dominated by CD8+ effector T cells, similar to drug-induced autoimmune-like hepatitis. In contrast, AIH showed a dominance of CD4+ effector T cells and CD79a+ B and plasma cells. T-cell receptor (TCR) and B-cell receptor sequencing showed that T and B cell clones were more dominant in VILI than in AIH. In addition, many T cell clones detected in the liver were also found in the blood. Interestingly, analysis of TCR beta chain and Ig heavy chain variable-joining gene usage further showed that TRBV6-1, TRBV5-1, TRBV7-6, and IgHV1-24 genes are used differently in VILI than in AIH. CONCLUSIONS: Our analyses support that SARS-CoV-2 VILI is related to AIH but also shows distinct differences from AIH in histomorphology, pathway activation, cellular immune infiltrates, and TCR usage. Therefore, VILI may be a separate entity, which is distinct from AIH and more closely related to drug-induced autoimmune-like hepatitis. IMPACT AND IMPLICATIONS: Little is known about the pathophysiology of COVID-19 vaccine-induced liver injury (VILI). Our analysis shows that COVID-19 VILI shares some similarities with autoimmune hepatitis, but also has distinct differences such as increased activation of metabolic pathways, a more prominent CD8+ T cell infiltrate, and an oligoclonal T and B cell response. Our findings suggest that VILI is a distinct disease entity. Therefore, there is a good chance that many patients with COVID-19 VILI will recover completely and will not develop long-term autoimmune hepatitis.

Diffuse large B-cell lymphomas have spatially defined, tumor immune microenvironments revealed by high-parameter imaging.

Diffuse large B-cell lymphoma (DLBCL) not otherwise specified is the most common aggressive non-Hodgkin lymphoma and a biologically heterogeneous disease. Despite the development of effective immunotherapies, the organization of the DLBCL tumor-immune microenvironment (TIME) remains poorly understood.We interrogated the intact TIME of 51 de novo DLBCLs with triplicate sampling to characterize 337 995 tumor and immune cells using a 27-plex antibody panel that captured cell lineage, architectural, and functional markers. We spatially assigned individual cells, identified local cell neighborhoods, and established their topographical organization in situ. We found that the organization of local tumor and immune cells can be modeled by 6 composite cell neighborhood types (CNTs). Differential CNT representation divided cases into 3 aggregate TIME categories: immune-deficient, dendritic cell-enriched (DC-enriched), and macrophage-enriched (Mac-enriched). Cases with immune-deficient TIMEs have tumor cell-rich CNTs, in which the few infiltrating immune cells are enriched near CD31+ vessels, in keeping with limited immune activity. Cases with DC-enriched TIMEs selectively include tumor cell-poor/immune cell-rich CNTs with high numbers of CD11c+ DCs and antigen-experienced T cells also enriched near CD31+ vessels, in keeping with increased immune activity. Cases with Mac-enriched TIMEs selectively include tumor cell-poor/immune cell-rich CNTs with high numbers of CD163+ macrophages and CD8 T cells throughout the microenvironment, accompanied by increased IDO-1 and LAG-3 and decreased HLA-DR expression and genetic signatures in keeping with immune evasion. Our findings reveal that the heterogenous cellular components of DLBCL are not randomly distributed but organized into CNTs that define aggregate TIMEs with distinct cellular, spatial, and functional features.

A DNA tumor virus globally reprograms host 3D genome architecture to achieve immortal growth.

Epstein-Barr virus (EBV) immortalization of resting B lymphocytes (RBLs) to lymphoblastoid cell lines (LCLs) models human DNA tumor virus oncogenesis. RBL and LCL chromatin interaction maps are compared to identify the spatial and temporal genome architectural changes during EBV B cell transformation. EBV induces global genome reorganization where contact domains frequently merge or subdivide during transformation. Repressed B compartments in RBLs frequently switch to active A compartments in LCLs. LCLs gain 40% new contact domain boundaries. Newly gained LCL boundaries have strong CTCF binding at their borders while in RBLs, the same sites have much less CTCF binding. Some LCL CTCF sites also have EBV nuclear antigen (EBNA) leader protein EBNALP binding. LCLs have more local interactions than RBLs at LCL dependency factors and super-enhancer targets. RNA Pol II HiChIP and FISH of RBL and LCL further validate the Hi-C results. EBNA3A inactivation globally alters LCL genome interactions. EBNA3A inactivation reduces CTCF and RAD21 DNA binding. EBNA3C inactivation rewires the looping at the CDKN2A/B and AICDA loci. Disruption of a CTCF site at AICDA locus increases AICDA expression. These data suggest that EBV controls lymphocyte growth by globally reorganizing host genome architecture to facilitate the expression of key oncogenes.

Robust single-cell matching and multimodal analysis using shared and distinct features.

The ability to align individual cellular information from multiple experimental sources is fundamental for a systems-level understanding of biological processes. However, currently available tools are mainly designed for single-cell transcriptomics matching and integration, and generally rely on a large number of shared features across datasets for cell matching. This approach underperforms when applied to single-cell proteomic datasets due to the limited number of parameters simultaneously accessed and lack of shared markers across these experiments. Here, we introduce a cell-matching algorithm, matching with partial overlap (MARIO) that accounts for both shared and distinct features, while consisting of vital filtering steps to avoid suboptimal matching. MARIO accurately matches and integrates data from different single-cell proteomic and multimodal methods, including spatial techniques and has cross-species capabilities. MARIO robustly matched tissue macrophages identified from COVID-19 lung autopsies via codetection by indexing imaging to macrophages recovered from COVID-19 bronchoalveolar lavage fluid by cellular indexing of transcriptomes and epitopes by sequencing, revealing unique immune responses within the lung microenvironment of patients with COVID.

Glucose dissociates DDX21 dimers to regulate mRNA splicing and tissue differentiation.

Glucose is a universal bioenergy source; however, its role in controlling protein interactions is unappreciated, as are its actions during differentiation-associated intracellular glucose elevation. Azido-glucose click chemistry identified glucose binding to a variety of RNA binding proteins (RBPs), including the DDX21 RNA helicase, which was found to be essential for epidermal differentiation. Glucose bound the ATP-binding domain of DDX21, altering protein conformation, inhibiting helicase activity, and dissociating DDX21 dimers. Glucose elevation during differentiation was associated with DDX21 re-localization from the nucleolus to the nucleoplasm where DDX21 assembled into larger protein complexes containing RNA splicing factors. DDX21 localized to specific SCUGSDGC motif in mRNA introns in a glucose-dependent manner and promoted the splicing of key pro-differentiation genes, including GRHL3, KLF4, OVOL1, and RBPJ. These findings uncover a biochemical mechanism of action for glucose in modulating the dimerization and function of an RNA helicase essential for tissue differentiation.

SARS-CoV-2 replication in airway epithelia requires motile cilia and microvillar reprogramming.

How SARS-CoV-2 penetrates the airway barrier of mucus and periciliary mucins to infect nasal epithelium remains unclear. Using primary nasal epithelial organoid cultures, we found that the virus attaches to motile cilia via the ACE2 receptor. SARS-CoV-2 traverses the mucus layer, using motile cilia as tracks to access the cell body. Depleting cilia blocks infection for SARS-CoV-2 and other respiratory viruses. SARS-CoV-2 progeny attach to airway microvilli 24 h post-infection and trigger formation of apically extended and highly branched microvilli that organize viral egress from the microvilli back into the mucus layer, supporting a model of virus dispersion throughout airway tissue via mucociliary transport. Phosphoproteomics and kinase inhibition reveal that microvillar remodeling is regulated by p21-activated kinases (PAK). Importantly, Omicron variants bind with higher affinity to motile cilia and show accelerated viral entry. Our work suggests that motile cilia, microvilli, and mucociliary-dependent mucus flow are critical for efficient virus replication in nasal epithelia.