Organoid modeling of lung-resident immune responses to SARS-CoV-2 infection

Tissue-resident immunity underlies essential host defenses against pathogens, but analysis in humans has lacked in vitro model systems where epithelial infection and accompanying resident immune cell responses can be observed en bloc. Indeed, human primary epithelial organoid cultures typically omit immune cells, and human tissue resident-memory lymphocytes are conventionally assayed without an epithelial infection component, for instance from peripheral blood, or after extraction from organs. Further, the study of resident immunity in animals can be complicated by interchange between tissue and peripheral immune compartments. To study human tissue-resident infectious immune responses in isolation from secondary lymphoid organs, we generated adult human lung three-dimensional air-liquid interface (ALI) lung organoids from intact tissue fragments that co-preserve epithelial and stromal architecture alongside endogenous lung-resident immune subsets. These included T, B, NK and myeloid cells, with CD69+CD103+ tissue-resident and CCR7− and/or CD45RA− TRM and conservation of T cell receptor repertoires, all corresponding to matched fresh tissue. SARS-CoV-2 vigorously infected organoid lung epithelium, alongside secondary induction of innate cytokine production that was inhibited by antiviral agents. Notably, SARS-CoV-2-infected organoids manifested adaptive virus-specific T cell activation that was specific for seropositive and/or previously infected donor individuals. This holistic non-reconstitutive organoid system demonstrates the sufficiency of lung to autonomously mount adaptive T cell memory responses without a peripheral lymphoid component, and represents an enabling method for the study of human tissue-resident immunity.

organoid system demonstrates the sufficiency of lung to autonomously mount adaptive T cell memory 48 responses without a peripheral lymphoid component, and represents an enabling method for the study of 49 human tissue-resident immunity. 50 INTRODUCTION 52 53 Tissue-resident immunity represents an essential line of defense against infection. In the lung, innate and 54 adaptive immunity coordinately respond to pathogens, with resident memory lymphocytes mediating rapid 55 recall responses 1,2 . However, the study of human lung-resident memory responses has been complicated 56 by the relative inaccessibility of pulmonary tissue for experimental purposes, and lack of in vitro epithelial-57 immune experimental systems, while mouse systems lack human context and possess obligate interchange 58 between tissue and peripheral lymphoid compartments 1-3 . It is further unclear if organs such as lung are 59 themselves sufficient to implement memory immune responses in the absence of interaction with secondary 60 lymphoid organs 4 . In the current COVID-19 pandemic, SARS-CoV-2 prompts a coordinated innate and 61 adaptive immune response involving both peripheral and tissue compartments, characterized by production 62 of cytokines and antibodies, as well as T cell activation 5-7 . Continued pandemic waves have highlighted 63 both the persistence and deficiencies of SARS-CoV-2 adaptive immune responses, typically measured in 64 peripheral blood [8][9][10] . Yet, the ability of lung-resident lymphocytes to mediate mucosal immunity against 65 SARS-CoV-2, to embody crucial memory defenses against repeated infection, and to elicit pathologic 66 inflammation as in acute respiratory distress syndrome and cytokine storm 11 , all emphasize the need for 67 relevant experimental systems. 68 Organoids have recently emerged as ex vivo models for infectious disease research 12 . We and 69 others previously modeled SARS-CoV-2 infection in lung organoids comprised exclusively of epithelium, 70 including alveolar type 2 (AT2), ciliated and club cells 13-19 , but such systems typically lack tissue-resident 71 immune populations, hampering investigations of inflammation during pathogenesis.
Conversely, 72 mononuclear fractions from peripheral blood or lung parenchyma allow study of SARS-CoV-2 memory 73 responses, but omit viral infection of epithelium 3,20-23 , while short-lived lung explants exhibit limited 74 viability 24 . Alternatively, rodent COVID-19 models are limited by interspecies host-pathogen 75 incompatibility 25,26 and immortalized cell lines do not accurately recapitulate the cellular diversity and 76 between fresh tissue and organoid matched known sequences recognizing viral pathogens, including SARS-168 CoV-2, influenza, CMV and EBV (Extended Data Fig. 4b, c). 169 Within the intact lung organ microenvironment, MHC class II is broadly expressed in pulmonary 170 epithelial and immune cell types. Additionally, MHC-II-expressing SPC + AT2 cells can internalize and 171 present peptide and full-length protein antigens, and mouse lung epithelial deletion of H2Ab1 compromises 172 lung-resident TRM function and pathogen response 39,40 . Upon FACS and scRNA-seq analysis, MHC-II was 173 expressed by diverse ALI lung organoid populations, including epithelium (AT2 and club), endothelium, 174 and B lymphocytes, mirroring MHC-II expression in cognate fresh tissue except for additional MHC-II 175 expression in macrophages (Extended Data Fig. 5a, b). Further, lung ALI organoid EPCAM + epithelial 176 cells avidly internalized and proteolytically degraded DQ-Ovalbumin (DQ-OVA) 48 to release the quenched 177 BODIPY fluorophore dye intracellularly, formally demonstrating antigen processing function intrinsic to 178 ALI lung organoids (Extended Data Fig. 5c). TAP1, involved in MHC-I antigen presentation, was also 179 broadly expressed in organoid epithelial and immune compartments (Extended Data Fig. 5b). 180 181 Suspension culture allows epithelial/stromal reorientation and apical ACE2 access 182 The SARS-CoV-2 entry receptor ACE2 is expressed on the apical surface of lung epithelium 41 , which is 183 typically oriented towards the central lumen in epithelial-only organoids and inconveniently precludes 184 access by virus added to the culture medium 13 . In simple epithelium-only lung organoids, removal of the 185 embedding extracellular matrix followed by suspension culture elicits rapid morphologic conversion from 186 an apical-in to an apical-out configuration. The resultant relocation of ACE2 to the organoid exterior then 187 allows facile infection by SARS-CoV-2 13 or entry by bacteria requiring apical access 42 . 188 To better simulate the endogenous host-viral interface in which the pathogens such as SARS-CoV-189 2 interface directly with the alveolar epithelium, we performed organoid eversion of 3D ALI lung organoids 190 by ECM removal and suspension culture (Fig. 3a). In collagen, lung ALI organoids contained epithelial 191 cells in central primary structures surrounded by ramifying mesenchyme (Fig. 1c-d and 3b), accompanied 192 by baseline interior expression of ACE2 (Fig. 3c). Upon collagen removal from 7-day established lung ALI organoids and subsequent suspension culture, the epithelium relocated to the surface within 48 hours 194 (Extended Data Fig. 6a-d) while preserving mesenchymal and immune cells (Fig. 3d). Notably, 195 suspension culture exposed the ACE2-expressing apical surface of epithelial cells on the external surface 196 of the organoids without obstruction by layers of stroma and collagen matrix (Fig. 3e) while retaining major 197 epithelial cell types such as AT1, AT2, and basal cells in distinct domains (Fig. 3f-g). Single cell  sequencing on CD45and CD45 + populations from suspension organoids confirmed preservation of all 199 major epithelial, stromal and immune cell types (Fig. 3h, Extended Data Fig. 6e, f). was readily detectable in ECAD + epithelium (Fig. 4a, b). SARS-CoV-2 induces membrane fusion of 205 infected cells and multinucleate syncytia in lung parenchymal cells, visible in autopsy samples from patients 206 who have succumbed to severe COVID-19 43,44 . Accordingly, syncytia formed within infected organoids at 207 later time points (Fig. 4c). Similar to the suspected tropism of SARS-CoV-2 45 , infected cells mostly 208 consisted of AT2 cells (Fig. 4d, e) and, to a lesser extent, club cells (Fig. 4f, g, Supplementary Table  209 2) 13,18,34 . Rare ciliated cell infection was observed in older organoids (>28 days culture) (Extended Data 210 Fig. 6h) when these cells were present in relative abundance, in agreement with previous studies 33, 34 . 211 SARS-CoV-2-infected NP + -cells progressively expressed the apoptotic marker cleaved caspase-3, 212 consistent with end-stage cytotoxicity 33,46 , beginning at days 10 and 14 post-infection (Fig. 4h-j). The 213 proportion of infected cells within a given organoid increased throughout the course of the incubation 214 period. At later time points, over 80% of cells expressing SARS-CoV-2 NP also expressed cleaved caspase-215 3, indicating frequent progression to cell death (Fig. 4k-l). 216 217

Adaptive immune responses to SARS-CoV-2 in organoids 255
The adaptive memory of lung-resident lymphocytes both defends against repeated infection and can 256 exacerbate pathologic inflammation as in ARDS and cytokine storm 11 . In marked contrast to epithelial-257 only conventional organoids 13,17,18 , the lung ALI organoid co-preservation of diverse epithelial lineages and 258 resident immune subsets afforded a unique opportunity to study tissue-resident SARS-CoV-2 adaptive 259 immunity, including TRM cells. Further, the organ-autonomous nature of this system enabled study of lung-260 intrinsic memory responses in isolation from secondary lymphoid organs. 261 We thus measured SARS-CoV-2-specific adaptive immune responses in T cells from virus-versus 262 mock-infected suspension organoids from seropositive individuals (Supplementary Table 1 as AIM markers in the culture system (Extended Data Fig. 9a-b). Accordingly, SARS-CoV-2 infection 274 increased the prevalence of 4-1BB + CD25 + , OX40 + CD40L + and OX40 + CD25 + AIM marker double-positive 275 CD8 + T cells in lung ALI organoids from seropositive and/or positive infection history donors, but not from 276 the seronegative donor ( Fig. 6a-

503
Human distal lung culture 504 Fresh human tissues and corresponding blood were obtained as deidentified samples from surgical discards 505 from Stanford Health Care (Stanford, CA). All experiments utilizing human tissue were approved by the 506 Stanford University Institutional Review Board. Standard informed consent for research was obtained in 507 writing prior to tissue procurement and all studies followed relevant guidelines and regulations. Human 508 distal lung was defined as peripheral lung tissue within 1 cm of the visceral pleura. For patients with 509 suspected lung cancer, cases with clinical T4 (American Joint Cancer Committee 6th edition) disease (e.g. 510 features such as bronchial invasion or parenchymal satellite nodule/ metastases) were deferred. Normal 511 tissue was harvested from the lung margin most anatomically distal to palpably well-defined lesions, or 512 from uninvolved lobes in the case of pneumonectomies. Samples with tumors containing ill-defined 513 margins were deferred. Tissue was either processed fresh or placed at 4°C overnight and processed the 514 following morning. The distal lung tissue was washed with PBS, minced finely on ice, and resuspended in 515 Cultrex Rat Collagen I (R&D, 3443-100-01). Next, 1 ml of tissue-collagen suspension was layered on top 516 of pre-solidified 1 ml collagen gel within a 30 mm inner transwell, 0.4 µm pore size (Sigma, PICM03050). 517 After fully solidifying, the collagen transwells were placed in a standard 6-well tissue culture plate 518 (Corning, 353046). 1 ml of lung ALI culture media (see below) was added into the tissue culture plate, 519 below the bottom surface of the collagen-containing transwell. Media was changed twice a week. 520 521 Lung ALI culture media 522 Advanced supplemented with 10% fetal bovine serum (R&D Systems, S11550), 10 µM Y-27632 (Peprotech,  529 1293823), and 10 µM CHIR 99021 (R&D Systems, 4423). 530 531 Eversion and suspension culture of human lung ALI organoids 532 Lung ALI organoids were grown as previously described in collagen for 5-10 days. To evert, collagen was 533 removed using collagenase type IV (Worthington, LS004210) for 30 min with shaking at 37°C. Collagenase 534 was washed and quenched with FBS containing media for 3 x 10 min at RT. Organoids were collected by 535 centrifuging at 100 x g for 3 min at RT and resuspended in lung ALI media (above) and plated in 1.5-2mL 536 each in a low-attachment 6-well plate (Corning, 3471). 537 538

SARS-CoV-2 infection of suspension lung ALI organoids 539
All SARS-CoV-2 work was performed in a class II biosafety cabinet under BSL3 conditions at Stanford 540 University. VeroE6 cells were obtained from ATCC and maintained in supplemented DMEM with 10% 541 FBS. SARS-CoV-2 (USA-WA1/2020) was passaged in VeroE6 cells in DMEM with 2% FBS. Titers were 542 determined by plaque assay on VeroE6 cells using Avicel (Sigma, 11365-1KG) and crystal violet (Fisher, 543 C581-25), viral genome sequence was verified, and all infections were performed with passage 3 virus. 544 Human lung ALI organoids were grown in collagen for 5-10 days, placed into suspension for 2-5 days, 545 counted, then infected with SARS-CoV-2 prior to day 14 in culture. Organoids were resuspended in virus 546 media or an equal volume of mock media, at a MOI of 1 relative to total organoid cells in the sample, and 547 then incubated at 37°C under 5% CO2 for 2 h. Organoids were then plated in suspension in lung ALI 548 organoid media. At the indicated timepoints, organoids were washed with PBS for downstream analysis. 549 For remdesivir experiments (RDV), organoids were infected with SARS-CoV-2 as described above and 550 spiked with 10 µM RDV at infection unless otherwise noted. 551 552 Serial brightfield imaging 553 Tissue culture plates containing the ALI organoids in transwells were imaged serially with a Keyence BZ-554 X710 microscope. Images were stitched using BZ-Wide viewer software. 555 556 Histologic analysis of ALI organoids 557 Collagen from transwell containing ALI organoids were fixed in 10% formalin for 1hr at RT, cut into thin 558 slices, and placed into a histology cassette with 70% ethanol. The collagen was then paraffin embedded and 559 sectioned (4-5 mm). Sections were deparaffinized and stained with H&E for histological analysis. 560 561 Whole-mount staining and confocal microscopy of ALI organoids 562 Collagen-containing ALI organoids were cut away from the transwell and fixed in 4% PFA for 1 hour at 563 RT. PFA was neutralized with 1X PBS-glycine (130 mM NaCl, 13.2 mM Na2HPO4, 3.5 mM NaH2PO4, 564 100 mM Glycine, in PBS at pH 7.4) for 30 min at RT, then blocked and permeabilized with 10% donkey 565 serum (Jackson Immunoresearch, 017-000-121) in a permeabilizing solution (130 mM NaCl, 13.2 mM 566 Na2HPO4, 3.5 mM NaH2PO4, 7.7 mM NaN3, 15 μM BSA, 2% Triton X-100, 0.5% TWEEN-20, in PBS 567 at pH 7.4) for 2 hours at RT. Organoids were then stained with primary antibodies at RT for 3 days 568 overnight, followed by 3 X 30 min washes with the permeabilizing solution. Secondary staining used 569 fluorescent donkey secondary antibodies (1:1000, Jackson Immunoresearch) and DAPI for 4 hours at RT, 570 then washed 3 X 30 min with the permeabilizing solution. Organoids were then mounted on slides with 571 mounting buffer (Prolong Gold Antifade mounting media, ThermoFisher Scientific, P36934). Images were 572 acquired using a Zeiss LSM900 confocal microscope and viewed in 3-D using Imaris software. All 573 antibodies, including secondaries, used for immunofluorescence are listed in Supplementary Table 3.

575
Immunofluorescence analysis of infected organoids 576 Infected organoids and corresponding mock-infected organoids were centrifuged and washed with PBS. 577 Organoids were then suspended in 4% PFA for fixation and inactivation of virus for 1 hr at RT. PFA was 578 then removed, and organoids were washed with PBS and removed from BSL3 conditions and subjected to 579 whole-mount immunofluorescence staining as described above. All SARS-CoV-2 work was performed in 580 a class II biosafety cabinet under BSL3 conditions at Stanford University. Human lung ALI organoids 581 infected with SARS-CoV-2 were fixed and stained as described above, and images were acquired on a Zeiss 582 LSM900 confocal microscope. The number of infected cells were quantified using the 'analyze particles' 583 tool of FIJI (Fiji is just ImageJ) software on MacOS. Briefly, for each sample a five-slice image stack was 584 acquired via the ZEN (blue edition) Microscope software and processed with the Z project tool on FIJI. 585 Image channels were separated and converted to grayscale. Threshold and exposure levels were then set 586 based on images of the mock condition and held constant across all images. The analyze particles tool was 587 utilized to count the number of signals present in each channel, with a size restriction set from 5 (particle 588 units) to infinity, and all other parameters set to default. 589    Treg Granulocyte P t 1 o r g P t 2 o r g P t 2 ti s s u e P t 1 o r g P t 2 o r g P t 2 ti s s u e P t 1 o r g P t 2 o r g P t 2 ti s s u e P t 1 o r g P t 2 o r g P t 2 ti s s u e P t 1 o r g P t 2 o r g P t 2 ti s s u e P t 1 o r g P t 2 o r g P t 2 ti s s u e P t 1 o r g P t 2 o r g P t 2 ti s s u e P t 1 o r g P t 2 o r g P t 2 ti s s u e % of total organoid CD45 + cells       IL1A  IL1B  IL1RA  IL2  IL3  IL4  IL6  IL7  IL10  IL12P40  IL12P70  IL13  IL17A  IL17E  IL17F IL22 IL27