An Advanced Human Intestinal Coculture Model Reveals Compartmentalized Host and Pathogen Strategies during Infection.

An Advanced Human Intestinal Coculture Model Reveals Compartmentalized Host and Pathogen Strategies during Salmonella Infection Leon N. Schulte, a,b Matthias Schweinlin, c Alexander J. Westermann, a,d Harshavardhan Janga, b Sara C. Santos, a Silke Appenzeller, e Heike Walles, f,g Jörg Vogel, a,d Marco Metzger c,g a Institute of Molecular Infection Biology (IMIB), University of Würzburg, Würzburg, Germany b Institute for Lung Research, Philipps University, Marburg, Germany c Department of Tissue Engineering and Regenerative Medicine, University Hospital Würzburg, Würzburg, Germany d Helmholtz Institute for RNA-Based Infection Research (HIRI), Helmholtz Centre for Infection Research (HZI), Würzburg, Germany e Comprehensive Cancer Center Mainfranken, University of Würzburg, Würzburg, Germany f Core Facility Tissue Engineering, University of Magdeburg, Magdeburg, Germany g Fraunhofer Institute for Silicate Research ISC, Translational Centre for Regenerative Therapies TLC-RT, Würzburg, Germany Leon N. Schulte and Matthias Schweinlin contributed equally to this article. Author order was determined both alphabetically and in order of increasing seniority. ABSTRACT A major obstacle in infection biology is the limited ability to recapitulate human disease trajectories in traditional cell culture and animal models, which impedes the translation of basic research into clinics. Here, we introduce a threedimensional (3D) intestinal tissue model to study human enteric infections at a level of detail that is not achieved by conventional two-dimensional monocultures. Our model comprises epithelial and endothelial layers, a primary intestinal collagen scaffold, and immune cells. Upon Salmonella infection, the model mimics human gastroenteritis, in that it restricts the pathogen to the epithelial compartment, an advantage over existing mouse models. Application of dual transcriptome sequencing to the Salmonella-infected model revealed the communication of epithelial, endothelial, monocytic, and natural killer cells among each other and with the pathogen. Our results suggest that Salmonella uses its type III secretion systems to manipulate STAT3dependent inflammatory responses locally in the epithelium without accompanying alterations in the endothelial compartment. Our approach promises to reveal further human-specific infection strategies employed by Salmonella and other pathogens. IMPORTANCE Infection research routinely employs in vitro cell cultures or in vivo mouse models as surrogates of human hosts. Differences between murine and human immunity and the low level of complexity of traditional cell cultures, however, highlight the demand for alternative models that combine the in vivo-like properties of the human system with straightforward experimental perturbation. Here, we introduce a 3D tissue model comprising multiple cell types of the human intestinal barrier, a primary site of pathogen attack. During infection with the foodborne pathogen Salmonella enterica serovar Typhimurium, our model recapitulates human disease aspects, including pathogen restriction to the epithelial compartment, thereby deviating from the systemic infection in mice. Combination of our model with state-of-the-art genetics revealed Salmonella-mediated local manipulations of human immune responses, likely contributing to the establishment of the pathogen’s infection niche. We propose the adoption of similar 3D tissue models to infection biology, to advance our understanding of molecular infection strategies employed by bacterial pathogens in their human host. KEYWORDS Salmonella, gene expression, infectious disease Citation Schulte LN, Schweinlin M, Westermann AJ, Janga H, Santos SC, Appenzeller S, Walles H, Vogel J, Metzger M. 2020. An advanced human intestinal coculture model reveals compartmentalized host and pathogen strategies during Salmonella infection. mBio 11:e03348-19. https://doi.org/ 10.1128/mBio.03348-19. Editor Julian Parkhill, Department of Veterinary Medicine Copyright © 2020 Schulte et al. This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International license. Address correspondence to Leon N. Schulte, [email protected], Jörg Vogel, [email protected], or Marco Metzger, [email protected]. This article is a direct contribution from Jörg Vogel, a Fellow of the American Academy of Microbiology, who arranged for and secured reviews by Andreas Bäumler, University of California, Davis, and Petra Dersch, University of Münster. Received 23 December 2019 Accepted 10 January 2020 Published RESEARCH ARTICLE Host-Microbe Biology crossm January/February 2020 Volume 11 Issue 1 e03348-19 ®mbio.asm.org 1 18 February 2020 on March 11, 2020 at GESELLSCHAFT FUR BIOTECHNO-http://mbio.asm.org/Downloaded from Enterobacteriaceae are major commensals of the human gut microflora, but certain members of this bacterial family, particularly Escherichia,Salmonella,Shigella, and Yersinia species, cause a range of different infections that sum up to millions of cases annually (1). Of these latter pathogens, Salmonella enterica serovar Typhimurium (henceforth S. Typhimurium) is a major research model for bacterial virulence strategies and host defense mechanisms during enteric infections. Host cell infection by S. Typhimurium depends on the concerted activity of effector proteins encoded on dedicated genomic virulence loci, referred to as Salmonella pathogenicity islands (SPIs) (2, 3). The two major SPIs (SPI1 and SPI2) additionally encode structural components of type III secretion systems (T3SSs) that deliver the virulence effector cocktail into the host cytosol. While the SPI1 T3SS and associated effectors mediate epithelial cell invasion (4), intracellular survival is promoted by virulence genes associated with the SPI2 cluster (5). Host cell manipulations mediated by SPI1 and SPI2 effectors include rearrangements of the actin cytoskeleton, manipulation of phagosomal maturation, and subversion of host immunity pathways (3). The immune response to S. Typhimurium has been investigated extensively. The innate immune system relies on a variety of pattern recognition receptors (PRRs), which sense conserved pathogen-associated molecular patterns (PAMPs), such as bacterial cell wall components or flagellin, to elicit proinflammatory transcriptional responses. In the intestine, Toll-like receptor 5 (TLR5) recognizes Salmonella flagellin, which activates cytokine and chemokine production for the recruitment and activation of professional immune cells, such as NK cells, T cells, and monocytes (6). These cells, in turn, respond by producing, e.g., gamma interferon (IFN- ␥ ) and interleukin-6 (IL-6), which activate the Janus kinase/signal transducer and activator of transcription (JAK/STAT) pathway on a variety of target cells to promote antimicrobial defense and changes to the cellular survival and metabolic programs (7, 8). Other major cytokines produced by the activated epithelium and professional immune cells include IL-1, which promotes NF- ␬ Bdependent immune gene expression (9), and IL-8, which functions as a major chemoattractant for bacterium-engulfing neutrophils (10). Recently, long noncoding RNAs (lncRNAs) were also implicated in the host response to Salmonella infection (11, 12). NeST lncRNA, for instance, protects from Salmonella-induced lethality in mice by promoting IFN- ␥ expression (13). S. Typhimurium, in turn, may partially evade this host defense through its facultative intracellular lifestyle and adaptation to—and even exploitation of—the inflammatory milieu (14, 15). Responses to Salmonella infections of the human gut necessarily require a finetuned interplay between the gut mucosa, the vascular endothelium, and the cells of the gut-associated immune system (16). These complex interactions have remained difficult to mimic in a human cell culture setting. The current understanding of host subversion by S. Typhimurium and the countermeasures taken by the immune system was largely deduced from studies with immortalized cell lines or mouse models, i.e., infection models with inherent strengths and weaknesses. Cell line monocultures have proven to be invaluable tools to reveal discrete molecular and cellular mechanisms, but they inevitably neglect the complex division of labor and three-dimensional (3D) fine structure within the inflamed tissue. In addition, critical cell components, such as peripheral immune cells, which affect the infection process, are often missing. Likewise, mice have been an important model to study Salmonella infections on a whole-system level, but there are profound differences in antimicrobial immunity and Salmonella pathogenesis between rodents and humans (17–19). For example, while S. Typhimurium induces self-limiting gastroenteritis in immunocompetent humans, it causes systemic infections and even sepsis in mice (6). Therefore, to better understand the molecular mechanisms underlying intestinal Salmonella infection in a human setting, more tailored models are needed. Sophisticated human three-dimensional (3D) in vitro models have recently garnered much attention of scientists in academia, product developers in industry, regulatory authorities, and society in general (20, 21). Examples are primary organoid cultures, rotating-wall vessel approaches, and Transwell-like coculture settings (21). Such models Schulte et al. ® January/February 2020 Volume 11 Issue 1 e03348-19 mbio.asm.org 2 on March 11, 2020 at GESELLSCHAFT FUR BIOTECHNO-http://mbio.asm.org/Downloaded from can deliver important information about drug toxicity and the mode of action, pathophysiology, normal biological tissue function, or immune responses (22, 23), prior to in vivo and clinical extrapolation. Often, however, the available models omit important cellular components, use artificial cellular growth matrices, or are difficult to standardize. To address the particular problem of artificial matrices to support cell growth, tissue models based on recellularized collagen scaffolds, comprising multiple cocultured cell types to more accurately mimic the epithelial barriers, were developed (20). So far, however, such models have neglected the vascular immune cell component and have not been widely adopted in infection research. Here, we present an advanced human intestinal barrier model based upon a recellularized porcine collagen scaffold that was infected with the bacterial model pathogen S. Typhimurium. This coculture model, encompassing both the endothelial and the epithelial intestinal barriers as well as a natural collagen matrix and professional immune cells, allowed us to study reciprocal host and pathogen cell adaptations during acute S. Typhimurium infection. We show that in this model system, in contrast to small-animal models, S. Typhimurium infection is restricted to the epithelial layer and does not spread into the vascular compartment, thereby mimicking human disease. Dual transcriptome sequencing (dual RNA-seq), which comprehensively profiles host and pathogen gene expression during bacterial infections, has been successfully applied to infected cell line-based, two-dimensional (2D) monocultures (reviewed in reference 24) and mouse models of infection (25–27). For the first time, we here applied dual RNA-seq to a 3D tissue model to chart mRNA and noncoding RNA expression changes in the communicating, purified host cell types (intestinal epithelial cells [IECs], endothelial cells, monocytes, NK cells) and in Salmonella. Our data sets determined STAT3 signaling to be a central host pathway targeted by the pathogen. Using CRISPR/Cas9-edited IECs and Salmonella virulence mutants, we show that the T3SSdependent manipulation of STAT3 locally changes the inflammatory milieu to the benefit of the pathogen but leaves the basolateral milieu unaltered. Thus, 3D infection models may reveal compartmentalized pathogen strategies not visible in conventional human cell cultures. Our dual RNA-seq data may serve the community as an important resource for prioritizing Salmonella virulence factors for further investigation and for defining cell type-specific expression signatures of pathogenic attack at the intestinal barrier. RESULTS An engineered human intestinal tissue model to study Salmonella infection. Due to the existing limitations in the currently available 3D in vitro culture models, infection studies with human-pathogenic bacteria typically neglect the tissue microstructure at the primary site of infection. With regard to the intestinal barrier, its major constituents are the epithelial lining and the underlying collagen scaffold of the lamina propria, harboring blood vessels for nutrient exchange and immune cell recruitment (28). To model the intestinal barrier in a commonly used Transwell-like setting, we fixed adecellularized porcine small intestinal submucosa (SIS) collagen scaffold into a cell crown to obtain two separated compartments (Fig. 1A). The apical compartment was populated with a human intestinal epithelial cell (IEC) line (Caco-2) and matured into a tight epithelial lining. The basolateral surface of the matrix was populated with primary human microvascular endothelial cells, and the underlying separated culture compartment was supplemented with peripheral blood leukocytes as a proxy for the vascular immune system (Fig. 1A). To confirm the suitability of our model for infection studies, we conducted a pilot experiment with a S. Typhimurium strain constitutively expressing the green fluorescent protein (GFP) (29) and tracked the bacteria within the tissue construct. Fluorescence microscopy analysis of cross sections visualized epithelial and endothelial cell monolayers, separated by the collagen scaffold, as well as Salmonella-infected cells within the epithelium, but not the endothelium (Fig. 1B to D). Flow cytometry of infected models identified a Salmonella-positive subpopulation of epithelial but not A Coculture Model Mimicking Human Intestinal Infection ® January/February 2020 Volume 11 Issue 1 e03348-19 mbio.asm.org 3 on March 11, 2020 at GESELLSCHAFT FUR BIOTECHNO-http://mbio.asm.org/Downloaded from endothelial cells (Fig. 2A; see also Fig. S1A in the supplemental material), and assays counting the numbers of colony forming units (CFUs) revealed the sterility of the basolateral culture medium (Fig. S1B). In line with these results, no indication for infection of basolateral leukocytes was obtained (Fig. S1C). These results confirm Salmonella to be unable to cross the epithelium. The fluorescence signal intensity emitted by invaded epithelial cells increased over time, indicative of Salmonella intracellular replication (Fig. 2B) at a rate comparable to previous findings from a 2D Caco-2 infection model (30). Furthermore, an increase in the percentage of invaded cells over time indicated spreading of the infection within the epithelium (Fig. 2C). Despite the absence of bacterial transmission across the epithelial barrier, the endothelial cell compartment responded to the infection by release of the major phagocyte attractant IL-8 (Fig. 2D). Thus, our tissue model successfully recapitulates an epithelially retained Salmonella infection and immune signaling across the intestinal barrier and thereby resembles human disease, which usually involves gastroenteritis, but no systemic infection, as is observed in mice. Processing of Salmonella-infected intestinal tissue models for transcriptomics. We sought to utilize our new model to gain an improved understanding of the reprogramming of host immunity by S. Typhimurium during infection. To this end, infections were carried out for 24 h with GFP-positive Salmonella applied to the apical compartment followed by fluorescence-activated cell sorting (FACS)-based separation BC D BC D A B C D FIG 1 Construction of the intestinal tissue model and experimental layout. (A) (Left) Illustration of a cross-section through a cell crown device within a culture dish, with the collagen membrane being fixed between outer and inner metal rings to create the apical and basolateral compartments. (Right) Schematic representation of the engineered intestinal barrier and experimental setup. The epithelium and endothelium are separated by a collagen layer (SIS, small intestinal submucosa), and leukocytes are supplied into the basolateral compartment. Infection is triggered by addition of GFP-positive Salmonella into the apical compartment. (B) Fluorescence microscopy analysis of a cross section through a Salmonella-infected intestinal barrier model (24 h postinfection; MOI, 10). The epithelium and endothelium are visualized by pCK and CD31 staining (green), respectively. Nuclei are stained with DAPI (blue). Salmonella are stained with an anti-LPS antibody and are shown in red. (C) Magnification showing the epithelial layer with Salmonella-infected cells (arrows). (D) Magnification of the endothelial cell layer. Bars, 100 ␮ m (B) or 20 ␮ m (C, D). Schulte et al. ® January/February 2020 Volume 11 Issue 1 e03348-19 mbio.asm.org 4 on March 11, 2020 at GESELLSCHAFT FUR BIOTECHNO-http://mbio.asm.org/Downloaded from of Salmonella-invaded epithelial cells (GFP positive) and noninvaded bystander epithelial cells (GFP negative), RNA extraction, rRNA depletion, and dual RNA-seq (Fig. 3A). To follow the propagation of the immune response across the intestinal barrier, cells of the endothelial lining (CD31 ⫹ ), monocytes (CD14 ⫹ ), and NK cells (CD56 ⫹ ) were FACS purified from the same models and their transcriptomes were sequenced. The corresponding cell types from uninfected models served as host controls and the bacterial inoculum served as the Salmonella preinfection reference. Principal-component analysis (PCA) of host cell RNA-seq data (row Z-scores) (Fig. 3B) revealed that samples primarily clustered according to cell type rather than treatment (infected versus noninfected). In line with this observation, specific expression signatures were revealed for IECs, endothelial cells, monocytes, and NK cells (Fig. 3C). Inspection of the detected host transcript classes (Fig. 3D) proved the intended depletion of rRNAs across all cell types. The mRNA fraction occupied ⬃88% of all mapped reads, followed by small nucleolar RNAs (snoRNAs; 3.9%), small nuclear RNAs (snRNAs; 3.3%), and lncRNAs (2.3%). In the following, we focus on regulated mRNAs and lncRNAs on the host side. Generally, the host response to Salmonella infection (both the number of regulated transcripts and their median fold change in expression) was higher in cells of the basolateral compartment (endothelial cells, monocytes, NK cells) than in cells of the infected epithelium (Fig. 3E and F). This confirms the sensing of the apically retained infection by the vascular components of the tissue construct, as seen in Fig. 2D. Interestingly, the overlap among the infectionregulated host genes between the different cell types was small (Fig. 3F), probably reflecting the nonredundant functions of IECs, endothelial cells, monocytes, and NK cells during bacterial infection. Cell type-specific signatures of the vascular immune response. To characterize the nature of the respective responses by the four interacting host cell types, we closely inspected mRNA and lncRNA expression changes after infection. First, we sought to characterize the propagation of the response of our tissue model to infection across the A Epithelium GFP control 0 h 1 h 4 h 8 h 24 h PE mui lehtodnE 0,0 0,5 1,0 1,5 2,0 2,5 3,0 0 4 8 12162024 fold change hours p.i. intracellular replicaon 0,00 5,00 10,00 15,00 0 5 10 15 124 % infected cells hours p.i. infecvity BCBasolateral IL-8 D 0 500 1000 1500 2000 2500 pg IL8 / ml - + + + + + - S. Tm hours p.i.hours p.i. 3 2.5 2 1.5 1 0.5 0 2.7 ± 0.5 5.7 ± 2.1 11.9 ±11.2 10.2 ±1.5 10.6 ±1.2 FIG 2 Temporal analysis of S. Typhimurium infection of the tissue model. (A) Representative FACS scatterplots showing mockand S. Typhimurium-infected (MOI, 10) epithelial (top row) and endothelial (bottom row) cells in the red (cellular autofluorescence) and green (GFP-expressing Salmonella) channels over the course of 24 h. (B) Quantification of the fold increase in the geometric mean green fluorescence intensity comparing the S. Typhimurium-infected samples postinfection (p.i.) to the control at 0 h. (C) Quantification of the percentage of infected (GFP-positive) cells at 1 and 24 h postinfection. (D) Quantification of IL-8 cytokine levels in the basolateral compartment upon mock control or S. Typhimurium (S. Tm) treatment at the indicated time points via ELISA. A Coculture Model Mimicking Human Intestinal Infection ® January/February 2020 Volume 11 Issue 1 e03348-19 mbio.asm.org 5 on March 11, 2020 at GESELLSCHAFT FUR BIOTECHNO-http://mbio.asm.org/Downloaded from intestinal barrier. Despite the absence of Salmonella transmission into the vascular compartment, endothelial cells upregulated (fold change [FC] ⱖ2; false discovery rate [FDR] ⬍0.05) 344 mRNAs and downregulated (FC ⱕ0.5; FDR ⬍0.05) 392 mRNAs upon apical infection (Fig. 4A). In monocytes, which, in conjunction with lymphocytes, such as NK or T cells, function to orchestrate the peripheral inflammatory response, 427 mRNAs were upregulated and 448 mRNAs were downregulated compared to their regulation in the mock-infected controls (Fig. 4B). In NK cells, which, besides their cytotoxic properties, provide antimicrobial cytokine signals to trigger antibacterial responses and antigen presentation, 372 mRNAs were upregulated and 37 were downregulated upon infection (Fig. 4C). During an acute response to infection, the endothelium functions to transmit the local immune activation signals into the bloodstream to trigger a systemic response (31). In line with this, among the 10 most highly induced mRNAs were those encoding proinflammatory cytokines and immune cell-recruiting chemokines, such as IL-6, CXCL6, and CXCL3L1 (Fig. 4A). Similarly, monocytes upregulated mRNAs encoding major proinflammatory chemokines and systemically acting cytokines, such as CXCL5, CXCL3, IL-1 ␣ , and IL-1 ␤ (Fig. 4B). Among the top induced mRNAs in NK cells were those encoding neutrophil attractant IL-8 (CXCL8), endothelial cell attachment protein TNFSF4, or the IL-1 ␣ /IL-1 ␤ decoy protein IL-1 receptor 2 (IL-1R2) (Fig. 4C), suggesting a vital involvement of NK cells in tuning the vascular innate immune response. The ACIEC Endo Mo NK mock Inf+ Infmock Inf. mock Inf. mock Inf. B 2D or 3D PCA Endo Mono NK IEC PC2 -3 -2 -1 0 1 2 3 PCA PC1 DEndo Mo IEC NK EF 0% 20% 40% 60% 80% 100% mRNA snRNA lncRNA other Read distribuon snoRNA IEC Endo Mo NK Inf+ Inf1000 100 10 1 0.1 0.01 fold-change FACS Dual RNA-Seq: •IECs (GFP+) •IECs (GFP-) •Endothelium •Monocytes •NK cells FIG 3 Overview of RNA-seq of the samples and comparison of the global host responses. (A) Experimental scheme. The indicated cell types were purified from the tissue model by FACS and separately analyzed by RNA-seq. IECs were separated into GFP-positive (GFP ⫹ ;Salmonella-infected) and GFPnegative (GFP ⫺ ; noninfected) populations and subjected to dual RNA-seq analysis. (B) PCA of RNA-seq libraries, based on row Z-scores. (C) Heat map representation of hierarchical clustering result using RNA-seq data tables from all cell types (row Z-scores, color coded according to the key provided at the bottom). Inf, data obtained from Salmonella-infected models; Inf ⫹ and Inf ⫺ , GFP-positive and -negative epithelial cells, respectively, from Salmonella-challenged models. (D) Averaged distribution of RNA-seq reads from all libraries over the main RNA classes. (E) Dot-plot representation of gene expression changes (ⱖ2-fold up or down compared to the level of expression by the mock-treated controls) averaged across both replicates and for the indicated conditions. (F) Venn diagram depicting the overlap of regulated genes for which the results are shown in panel E between the different cell types. IEC, intestinal epithelial cells; Endo, microvascular endothelial cells; Mo or Mono, monocytes; NK, natural killer cells. Schulte et al. ® January/February 2020 Volume 11 Issue 1 e03348-19 mbio.asm.org 6 on March 11, 2020 at GESELLSCHAFT FUR BIOTECHNO-http://mbio.asm.org/Downloaded from 0 10 20 30 40 50 mock inf. fold-change 0 5 10 15 mock inf. fold-change 0 10 20 30 mock inf. fold-change 0 1 2 3 4 5 mock inf. fold-change 0 10 20 30 40 50 mock inf. fold-change 0 2 4 6 8 mock inf. fold-change 0 2 4 6 8 mock inf. fold-change AEndothelium mRNAs BMonocyte mRNAs CNK cell mRNAs Endothelium lncRNAs Monocyte lncRNAs NK cell lncRNAs Validaon Validaon 344 392 LIF DRAXIN CX3CL1 PLA1A CSF2 IL6 CXCL6 NOD2 CCL20 SAA1 HIST1H2BI RHAG SCPEP1 PDK4 CRISP2 EFEMP1 RAMP3 CD34 AGT GIMAP7 31 25 AL031316.1 MSC-AS1 AC025580.2 AC023157.3 LUCAT1 AL132780.1 LINC01137 AC020916.1 AL021937.4 AL137145.1 LINC01876 LINC01116 AC024896.1 LINC02035 AL357507.1 AC019163.1 LINC01094 AC125807.2 AC018616.1 LINC01235 427 448 TCN2 MPEG1 CD36 RNASE1 FUCA1 F13A1 CLEC10A SIGLEC1 COLEC12 FCN1 29 47 MIR3945HG AL031316.1 MSC-AS1 AC083837.2 AC087645.2 AC007384.1 AC025580.2 AC003101.1 MIAT ADORA2A-AS1 AC100830.1 LINC01504 TSC22D1-AS1 AC093227.1 LINC00847 LINC01715 ZBTB11-AS1 AL022311.1 AC007563.2 AL021937.4 CXCL5 CA12 CXCL1 MT1G IL1B IL1R2 IL1A MT1H MT2A CXCL3 372 37 APOBEC3C P4HA1 JAML KIAA1191 PASK WDR73 ALAD PFKFB4 HK2 SLC47A1 30 9 MIAT AC007952.4 LINC00996 LINC02390 AP002433.1 AL157935.1 AZIN1-AS1 AL137145.1 H1FX-AS1 AC020916.1 FP236383.1 AC078777.1 AL035071.1 LINC02453 AL035701.1 CR381653.1 DARS-AS1 AC078883.1 XIST AL021918.3 CXCL8 PID1 ZFY TEX14 TNFSF4 IL1R2 ILDR2 F11R ZNRF1 SYPL1 R1 R2 R1 R2 R1 R2 R1 R2 R1 R2 R1 R2 R1 R2 R1 R2 R1 R2 R1 R2 R1 R2 R1 R2 0.001 0.01 0.1 0.2 0.5 1 2 5 10 100 1000 0.001 0.01 0.1 0.2 0.5 1 2 5 10 100 1000 0 2 4 6 mock inf. fold-change 0 1 2 3 4 5 mock inf. fold-change 0 1 2 3 4 5 mock inf. fold-change CSF2 MSC-AS1 LUCAT1 fold-change Validaon fold-change IL6 CA12 MSC-AS1 MiR3945HG fold-change fold-change CXCL5 0 2 4 6 8 mock inf. fold-change 0 2 4 6 mock inf. fold-change AC007952.4 MIAT fold-change fold-change CXCL8 PID1 mock inf. mock inf. mock inf. mock inf. mock inf. mock inf. FIG 4 Host gene expression changes in the vascular compartment. (A) Heat map (two experimental replicates, R1 and R2) showing changes in mRNA (left) and lncRNA (middle) expression in endothelial cells upon apical Salmonella infection (24 h; MOI, 10) from that obtained by mock treatment of the intestinal tissue construct. The results for the top 10 upand downregulated genes are shown in magnification to the right of each heat map, and the genes are labeled by name. Fold changes in expression are color coded according to the key provided below panel C. (Right) Validation of the induced expression changes of selected mRNAs and lncRNAs upon infection by qRT-PCR. Combined results (mean ⫾SD) from three independent experiments are shown. (B) Same as panel A but for monocyte expression data. (C) Same as panel A but for NK-cell data. A Coculture Model Mimicking Human Intestinal Infection ® January/February 2020 Volume 11 Issue 1 e03348-19 mbio.asm.org 7 on March 11, 2020 at GESELLSCHAFT FUR BIOTECHNO-http://mbio.asm.org/Downloaded from induction of IL-8 was identified to be the common denominator of endothelial, monocytic, and NK-cell responses (Fig. S2). Overall, only a few RNAs were induced in more than one cell type (Fig. S2), illustrating the extensive division of labor during innate immune responses to Salmonella. Noteworthy was the finding that the response to infection by all three basolateral cell types included the differential expression of dozens of lncRNAs which make up a class of transcripts with emerging functions in vertebrate immunity (32, 33). Regulation of selected lncRNAs in all cell types could be confirmed by quantitative real-time PCR (qRT-PCR) analysis (Fig. 4A to C). These measurements also confirmed MSC-AS1 to be ashared lncRNA marker of immune activation in endothelial cells and monocytes. Together, these results demonstrate extensive rewiring of the coding and noncoding transcriptomes of key human cell types involved in vascular immune activation during intestinal Salmonella infection. Host-pathogen transcriptomics of the infected epithelium. Dual RNA-seq simultaneously records the gene expression of a bacterium and its mammalian host, which allowed us to study reciprocal host-pathogen adaptations during the epithelially retained S. Typhimurium infection within FACS-separated IECs. Mapping of RNA-seq reads from the bacterial input and epithelial mock-infected control libraries confirmed almost exclusive alignment to the bacterial or human reference genome, respectively (Fig. 5A). With regard to the infected samples, in the invaded (GFP-positive) but not in the bystander (GFP-negative) epithelial cells, ⬃1% of the total reads mapped to the Salmonella genome (Fig. 5A), verifying successful separation of infected from noninfected host cells at the cell sorting step. Comparison of these intracellular Salmonella transcriptomes to previously recorded expression data for intracellular Salmonella within human 2D monocultures (30) by PCA revealed a segregation according to monocytic/macrophage and epithelial cell lineages (Fig. 5B). Salmonella genes preferentially expressed during epithelial cell (but not monocyte) infections were enriched for Gene Ontology (GO) terms relating to nitrogen compound metabolism (Fig. 5C). Thus, our intestinal human tissue infection model recapitulates an epithelial cell-adapted Salmonella gene expression program, arguing that the intraepithelial environment drives Salmonella gene expression largely independently of the presence or absence of additional host cell types in the culture. Comparison of reads from intraepithelial Salmonella to those from the bacterial input sample revealed the upregulation of 527 Salmonella mRNAs and the downregulation of 145 Salmonella mRNAs (Fig. 5D). Host cell invasion by Salmonella requires the activation of genes encoded by the SPI1 locus (4), whereas intracellular survival depends on the expression of genes encoded by SPI2 (5). The switch from SPI1 to SPI2 gene expression involves the PhoP/Q two-component system (34), activation of which is therefore necessary for intracellular survival (35). Accordingly, intracellular Salmonella upregulated the expression of genes belonging to the PhoP regulon and SPI2-encoded genes and downregulated the expression of SPI1 genes compared to the gene expression of the bacterial input (Fig. 5E). Host cell manipulation by Salmonella occurs through effector proteins secreted through the T3SS encoded on SPI1 and SPI2. In line with the activation of SPI2, expression of SPI2 T3SS-associated effectors, whether encoded on SPI2 itself or within the core genome (except for SseG), was upregulated in Salmonella inside flow-sorted IECs (Fig. 5F). On the other hand, SPI1-associated effectors were largely downregulated (Fig. 5F). Dual RNA-seq also captures the expression of bacterial noncoding transcripts, particularly the class of small noncoding RNAs (sRNAs). Previously, we have uncovered sRNA expression patterns during the intracellular phase of the Salmonella infection cycle (30). Confirming that the bacterial expression patterns detected in FACS-enriched GFP-positive IECs indeed reflect intracellular Salmonella transcriptome signatures, two PhoP-activated sRNAs, PinT and AmgR (30, 36), were highly induced compared to their expression in the bacterial inoculum (Fig. 5D). Additionally, our data reveal the regulation of dozens of further sRNAs in intracellular Salmonella (Fig. 5D; Table S1). For Schulte et al. ® January/February 2020 Volume 11 Issue 1 e03348-19 mbio.asm.org 8 on March 11, 2020 at GESELLSCHAFT FUR BIOTECHNO-http://mbio.asm.org/Downloaded from 0,1 1 10 100 fold-change SPI1 effectors SPI2 effectors 100 10 1 0.1 sseJ sifB pipB spvC sseF sopD2 sifA spvB sseI sspH2 sseG sopA gtgE sopE2 avrA sopD sipC sipB sipA sopE SPI1 SPI1 PhoP PhoP SPI2 SPI2 0.01 0.1 1 10 100 1000 Regulons F fold-change 1000 100 10 1 0.1 0.01 PhoP SPI2SPI1 R1 R2 R1 R2 R1 R2 Salmonella regulons E PhoQ PhoP SPI2SPI1 (PinT) PinT IsrE 5S (bacterial) U6 (human) DapZ InvR 80 98 88 78 in vivoinduced in vivorepressed [nt] G ARead distribuon 0% 20% P_1 P_2 k_1 k_2 s_1 s_2 s_1 s_2 0 % 100 % 1.3 % 1.1 % Salmonella Human Input Mock GFP + GFP - PC1 PC2 PC3 B CD Salmonella mRNAs Salmonella sRNAs 527 145 pagC pspA asr sseB_1 uxuA pspC mgtB SL1344_1190 SL1344_1345 sodCI prgH flgK invF cheW invA invE flaG invG invC orgAa R1 R2 R1 R2 23 19 AmgR PinT STnc510 StyR-3 RyhB RUF_341c.10 IsrE RybA IsrC STnc800 zipA-leader StyR-29 STnc3680 StyR-55 STnc1460 tpk3 DapZ SLnc0014 GcvB InvR R1 R2 R1 R2 0.001 0.01 0.1 0.2 0.5 1 2 5 10 100 1000 I II THP-1_R1 THP-1_R2 dTHP-1 HEK Caco-2_R1 Caco-2_R2 HeLa_R2 HeLa_R1 HeLa_R3 THP-1_R1 THP-1_R2 dTHP-1 HEK Caco-2_R1 Caco-2_R2 HeLa_R2 HeLa_R1 HeLa_R3 FIG 5 Gene expression of intraepithelial Salmonella. (A) Proportion of bacterial reads in the bacterial input sample, mock-treated intestinal epithelial cells (IECs), bystander IECs (GFP negative), and invaded IECs (GFP positive). (B) (Left) PCA of Salmonella intracellular transcriptomes from infected THP1 monocytes and macrophages (dTHP1) or the indicated epithelial cell types. Caco-2 cell replicates R1 and R2 are from this study; the rest are from a previous study (30). (Right) Z-score heat map showing Salmonella gene expression clusters in the PCA data (for the indicated cell types). Gene clusters (I and II) discriminating epithelial from monocyte/macrophage intracellular Salmonella transcriptomes are highlighted. (C) Magnification of cluster I from panel B (right) with enriched GO terms. (D) Heat map (two experimental replicates, R1 and R2) showing changes in mRNA expression (left) and sRNA expression (right) in Salmonella inside infected epithelial cells (24 h) from the expression for the input bacterial control. The results for the top 10 upor downregulated genes are shown in magnification to the right of each heat map, and the genes are labeled by name. Fold changes in expression are color coded according to the key provided on the left. (E) (Top) Schematic illustration of the switch from SPI1 (invasion) to SPI2 (intracellular survival) gene expression upon intracellular activation of the PhoP/Q two-component system. (Bottom) Box plot depicting the regulation of genes belonging to the SPI1, PhoP, and SPI2 regulons in intracellular Salmonella compared to the Salmonella inoculum. (F) Fold changes in the levels of mRNAs encoding Salmonella effectors secreted through the SPI2 or SPI1 T3SS (the mean ⫾SD was calculated from both RNA-seq replicates). (G) Northern blot validation of the differential expression of Salmonella sRNAs prior to infection (inoculum) or 24 h after infection (MOI, 10 or 100). Radioactive signals were absent from mock-infected control samples (mock), supporting the specificity of the selected DNA probes for the intended Salmonella transcripts and excluding cross-reactivity with human RNAs. nt, number of nucleotides. A Coculture Model Mimicking Human Intestinal Infection ® January/February 2020 Volume 11 Issue 1 e03348-19 mbio.asm.org 9 on March 11, 2020 at GESELLSCHAFT FUR BIOTECHNO-http://mbio.asm.org/Downloaded from concentration, 20 ␮ g/ml) into the apical compartment at 1 h postinfection. At the time points after infection indicated above (see text and figures in Results), epithelial and endothelial cells were collected directly from the scaffold by trypsin-Accutase cell detachment solution treatment upon 2 washes with PBS. Leukocytes were collected directly from the basolateral compartment and further purified by cell sorting (see below). For CFU assays, Caco-2 cells were lysed with PBS containing 0.01% Triton X-100. The lysates were serially diluted in PBS and plated onto LB agar, followed by overnight incubation at 37°C. Control samples were mock treated (mock treatment was same treatment used for the infected samples but with the addition of sterile medium instead of the bacterial suspension). CRISPR/Cas9-based genome editing. A synthetic DNA segment (Metabion; see Table S2 in the supplemental material) was cloned into the BbsI site of the pX458 CRISPR vector (from the F. Zhang lab [74] through Addgene) for expression of a guide RNA targeting the STAT3 coding sequence. Caco-2 cells were transfected with 1 ␮ g of plasmid DNA using the Lipofectamine 2000 reagent (Thermo Fisher) according to the manufacturer’s instructions. At 24 h after transfection, single transfected (GFP-positive) cells were spotted into 96-well plates that had been prefilled with complete medium containing 100 ␮ g/ml of the Normocin antibiotic mixture (Invivogen), using a FACSAria III cell sorter (BD) with a 100- ␮ m nozzle size. During clonal expansion in the wells of the 96-well plate, fresh medium was added every 5 days. Knockout success was evaluated by PCR amplification of the STAT3 coding sequence from genomic DNA and confirmed by Sanger sequencing (Seqlab GmbH, Göttingen, Germany). Quantitative real-time PCR. qRT-PCR analyses were carried out using a Power SYBR green RNA-to-Ct 1-step kit (Thermo Fisher) according to the manufacturer’s instructions and a QuantStudio3 real-time PCR machine (Applied Biosystems). RNA was extracted using the TRIzol reagent (Thermo Fisher) method. To remove genomic DNA, the extracted nucleic acids were incubated with DNase I (Thermo Fisher) and an RNase inhibitor (Promega) for 30 min at 37°C and subsequently extracted with phenol-chloroformisoamyl alcohol (Sigma-Aldrich), followed by precipitation with 30:1 ethanol–5 M sodium acetate. The qRT-PCR primers are listed in Table S2. Fold changes based on threshold cycle (C T ) values were calculated using the 2 ⫺ΔΔCT method (75), and human U6 snRNA was used as an internal reference. Immunostaining and flow cytometry. Cells collected from the barrier model were analyzed using a FACSCalibur or a FACSAria III device (BD). To purify monocytes and NK cells from the collected leukocytes, the cells were stained with anti-CD14-FITC (catalog number 11-0149-42; Thermo Fisher) and anti-CD56-allophycocyanin (catalog number 17-0567-41; Thermo Fisher) antibodies in PBS, 0.1% FCS and sorted using a FACSAria III device (100- ␮ m nozzle, single-cell purity setting). FCS3.0 files were analyzed using Flowing software (http://flowingsoftware.btk.fi/). Dual RNA-seq and computational analyses. For RNA-seq analysis, cellular RNA was extracted using amirVana RNA isolation kit (Thermo Fisher) according to the total RNA isolation protocol supplied with the kit. rRNA was depleted using a Ribo-Zero gold (epidemiology) kit (Illumina). Libraries were generated and sequenced on a NextSeq 500 platform at Vertis Biotech (Freising, Germany) as previously described (30). Demultiplexed reads were mapped to the GRCh38 human reference annotation using the CLC Genomics Workbench (Qiagen) with standard settings (mismatch cost ⫽2, insertion cost ⫽3, deletion cost ⫽3, length fraction ⫽0.8, similarity fraction ⫽0.8). The data tables obtained were filtered for genes with a number of reads per kilobase per million (RPKM) value of ⱖ0.5 in both sequenced replicates under at least one experimental condition. Genes exhibiting fold changes in expression of ⱖ2orⱕ0.5 (calculated based on RPKMs) in both replicates were considered differentially expressed. Hierarchical clustering was performed using the Cluster program (Michael Eisen lab) with the correlation (uncentered) similarity metric and the centroid linkage clustering method. Heat maps were generated using the Java TreeView program (76). PCA analysis was done in R software using the script prcomp (stats) and the rgl package. Network plots were generated with Cytoscape software (version 3.7.1). KEGG pathway analysis and induced network analysis were performed using the ConsensusPathDB molecular functional interaction database (77). Bacterial bioinformatics analyses (Fig. 5B and C) were performed as follows. Samples including genes with an RPKM of ⬎1 in at least one sample and a coefficient of variation of ⬎0.5 were subjected to a 3-dimensional principal-component analysis (with the R software scatterplot3d package, version 0.3-41), using log 2 (RPKM) values as the input. Unsupervised complete linkage clustering (with the R software heatmap.2 function from the gplots package, version 3.0.1.1) was performed on rows and columns using the Euclidian distance as a similarity metric and log 2 (RPKM) values as the input. Salmonella GO term enrichment analysis (Fig. 5C) was performed using the ShinyGO tool (version 0.60; http://bioinformatics .sdstate.edu/go/) for the GO term biological process with an FDR cutoff of 0.05. Western blot analysis. For Western blot analysis, samples were collected in radioimmunoprecipitation assay buffer supplemented with Laemmli buffer and boiled for 5 min. Proteins were separated on 10% polyacrylamide-SDS gels and transferred onto nitrocellulose membranes (catalog number 10600015; Amersham) by semidry blotting. Proteins were detected using anti-STAT3 (catalog number 9139; Cell Signaling), anti-phospho-STAT3 (catalog number 9134; Cell Signaling), and anti-actin (catalog number sc-1616; Santa Cruz) primary antibodies, horseradish peroxidase-linked secondary antibodies, and an enhanced chemiluminescence (ECL) reagent (catalog number RPN2232; Amersham). Images were obtained using an Intas Advanced ECL imager system. Histology. Tissue samples were fixed with 4% paraformaldehyde for1hat4°C. Samples were embedded in paraffin and sectioned to a thickness of 5 ␮ m with a microtome (model SM2010 R; Leica). Tissue slices were first deparaffinized using the Roticlear clearing agent (Carl Roth) and rehydrated in a graded series of ethanol according to standard protocols. Characterization of the tissue samples was done by immunofluorescence staining. For antigen retrieval, tissue slices were heat pretreated at 100°C for 20 min in pH 6 citrate buffer (Carl Roth). After blocking unspecific binding by PBS with 0.3% Triton Schulte et al. ® January/February 2020 Volume 11 Issue 1 e03348-19 mbio.asm.org 16 on March 11, 2020 at GESELLSCHAFT FUR BIOTECHNO-http://mbio.asm.org/Downloaded from X-100 (Sigma-Aldrich), 5% bovine serum albumin (BSA; PanReac AppliChem), and 5% donkey serum (Biozol) for 30 min, the slices were incubated with primary antibodies at 4°C overnight. The following primary antibodies were used at a 1:100 dilution: pan-cytokeratin (pCK; specific for epithelial cells; Dako), CD31 (endothelial cell specific; Abcam), and LPS (for the detection of Salmonella; Abcam). After washing, anti-mouse/anti-rabbit immunoglobulin-Alexa Fluor 555 and -Alexa Fluor 647 secondary antibodies were added at a dilution of 1:400 in antibody dilution solution, and the mixture was incubated for1hatroom temperature. Samples were mounted using Mowiol mounting medium with DAPI (4=,6-diamidino-2phenylindole; Sigma-Aldrich) for nuclear staining. Imaging was achieved using an inverted fluorescence microscope (Keyence BZ-9000). ELISA. Enzyme-linked immunosorbent assays (ELISAs) were performed using human IL-6 (catalog number 88-7066-86) and IL-8 (catalog number 88-8086-86) Ready-Set-Go ELISA kits (Thermo Fisher) according to the manufacturer’s instructions. The cell culture supernatants were centrifuged for 1 min at maximal speed to pellet the cell debris, and samples were used at a 1:10 (IL-6) or 1:150 (IL-8) dilution. The samples were analyzed using a Tecan Sunrise plate reader, and absolute quantification was achieved using the cytokine standards supplied with the ELISA kit. Data availability. RNA-seq data have been uploaded to the NCBI GEO repository (GEO accession number GSE136717). SUPPLEMENTAL MATERIAL Supplemental material is available online only. FIG S1, PDF file, 0.1 MB. FIG S2, PDF file, 0.2 MB. FIG S3, PDF file, 0.2 MB. FIG S4, PDF file, 0.2 MB. TABLE S1, PDF file, 0.1 MB. TABLE S2, PDF file, 0.1 MB. ACKNOWLEDGMENTS We thank Konrad Förstner for assistance with the bioinformatics analyses. This work was supported by a grant from the Deutsche Forschungsgemeinschaft (grant DFG GRK 2157, 3D Tissue Models for Studying Microbial Infections by Human Pathogens, to J.V., M.S., and S.C.S.) and by the von Behring-Röntgen-Stiftung (vBR project 63-0036, to L.N.S.). We declare no competing interests. L.N.S. performed experiments and data analysis and participated in manuscript writing, funding acquisition, study design, and supervision. M.S. performed experiments and participated in study design and manuscript writing. A.J.W. performed experiments and data analysis and participated in manuscript writing and supervision. H.J. and S.C.S. performed experiments. S.A. performed bioinformatics analysis. H.W. participated in study design, funding acquisition, and supervision. J.V. and M.M. participated in manuscript writing, funding acquisition, study design, and supervision. REFERENCES 1. Petri WA, Miller M, Binder HJ, Levine MM, Dillingham R, Guerrant RL, Jr. 2008. Enteric infections, diarrhea, and their impact on function and development. J Clin Invest 118:1277–1290. https://doi.org/10.1172/JCI34005. 2. Fabrega A, Vila J. 2013. Salmonella enterica serovar Typhimurium skills to succeed in the host: virulence and regulation. Clin Microbiol Rev 26:308–341. https://doi.org/10.1128/CMR.00066-12. 3. LaRock DL, Chaudhary A, Miller SI. 2015. Salmonellae interactions with host processes. Nat Rev Microbiol 13:191–205. https://doi.org/10.1038/ nrmicro3420. 4. Patel JC, Galan JE. 2005. Manipulation of the host actin cytoskeleton by Salmonella—all in the name of entry. Curr Opin Microbiol 8:10–15. https://doi.org/10.1016/j.mib.2004.09.001. 5. Jennings E, Thurston TLM, Holden DW. 2017. Salmonella SPI-2 type III secretion system effectors: molecular mechanisms and physiological consequences. Cell Host Microbe 22:217–231. https://doi.org/10.1016/j .chom.2017.07.009. 6. Keestra-Gounder AM, Tsolis RM, Bäumler AJ. 2015. Now you see me, now you don’t: the interaction of Salmonella with innate immune receptors. Nat Rev Microbiol 13:206–216. https://doi.org/10.1038/nrmicro3428. 7. Dodington DW, Desai HR, Woo M. 2018. JAK/STAT—emerging players in metabolism. Trends Endocrinol Metab 29:55–65. https://doi.org/10 .1016/j.tem.2017.11.001. 8. Qi Y-F, Huang Y-X, Wang H-Y, Zhang Y, Bao Y-L, Sun L-G, Wu Y, Yu C-L, Song Z-B, Zheng L-H, Sun Y, Wang G-N, Li Y-X. 2013. Elucidating the crosstalk mechanism between IFN-gamma and IL-6 via mathematical modelling. BMC Bioinformatics 14:41. https://doi.org/10.1186/1471-2105 -14-41. 9. Cohen P. 2014. The TLR and IL-1 signalling network at a glance. J Cell Sci 127:2383–2390. https://doi.org/10.1242/jcs.149831. 10. Russo RC, Garcia CC, Teixeira MM, Amaral FA. 2014. The CXCL8/IL-8 chemokine family and its receptors in inflammatory diseases. Expert Rev Clin Immunol 10:593–619. https://doi.org/10.1586/1744666X.2014.894886. 11. Munschauer M, Vogel J. 2018. Nuclear lncRNA stabilization in the host response to bacterial infection. EMBO J 37:e99875. https://doi.org/10 .15252/embj.201899875. 12. Imamura K, Takaya A, Ishida Y᎑I, Fukuoka Y, Taya T, Nakaki R, Kakeda M, Imamachi N, Sato A, Yamada T, Onoguchi᎑Mizutani R, Akizuki G, Tanu T, Tao K, Miyao S, Suzuki Y, Nagahama M, Yamamoto T, Jensen TH, Akimitsu N. 2018. Diminished nuclear RNA decay upon Salmonella infection upregulates antibacterial noncoding RNAs. EMBO J 37:e97723. https://doi.org/10.15252/embj.201797723. A Coculture Model Mimicking Human Intestinal Infection ® January/February 2020 Volume 11 Issue 1 e03348-19 mbio.asm.org 17 on March 11, 2020 at GESELLSCHAFT FUR BIOTECHNO-http://mbio.asm.org/Downloaded from 13. Gomez JA, Wapinski OL, Yang YW, Bureau J-F, Gopinath S, Monack DM, Chang HY, Brahic M, Kirkegaard K. 2013. The NeST long ncRNA controls microbial susceptibility and epigenetic activation of the interferongamma locus. Cell 152:743–754. https://doi.org/10.1016/j.cell.2013.01 .015. 14. Haraga A, Ohlson MB, Miller SI. 2008. Salmonellae interplay with host cells. Nat Rev Microbiol 6:53–66. https://doi.org/10.1038/nrmicro1788. 15. Rivera-Chavez F, Baumler AJ. 2015. The pyromaniac inside you: Salmonella metabolism in the host gut. Annu Rev Microbiol 69:31–48. https:// doi.org/10.1146/annurev-micro-091014-104108. 16. Patel S, McCormick BA. 2014. Mucosal inflammatory response to Salmonella typhimurium infection. Front Immunol 5:311. https://doi.org/10 .3389/fimmu.2014.00311. 17. Zschaler J, Schlorke D, Arnhold J. 2014. Differences in innate immune response between man and mouse. Crit Rev Immunol 34:433–454. 18. Mestas J, Hughes CCW. 2004. Of mice and not men: differences between mouse and human immunology. J Immunol 172:2731–2738. https://doi .org/10.4049/jimmunol.172.5.2731. 19. Coburn B, Grassl GA, Finlay BB. 2007. Salmonella, the host and disease: a brief review. Immunol Cell Biol 85:112–118. https://doi.org/10.1038/sj .icb.7100007. 20. Gordon S, Daneshian M, Bouwstra J, Caloni F, Constant S, Davies DE, Dandekar G, Guzman CA, Fabian E, Haltner E, Hartung T, Hasiwa N, Hayden P, Kandarova H, Khare S, Krug HF, Kneuer C, Leist M, Lian G, Marx U, Metzger M, Ott K, Prieto P, Roberts MS, Roggen EL, Tralau T, van den Braak C, Walles H, Lehr C-M. 2015. Non-animal models of epithelial barriers (skin, intestine and lung) in research, industrial applications and regulatory toxicology. Altex 32:327–378. https://doi.org/10.14573/altex .1510051. 21. Barrila J, Crabbé A, Yang J, Franco K, Nydam SD, Forsyth RJ, Davis RR, Gangaraju S, Ott CM, Coyne CB, Bissell MJ, Nickerson CA. 2018. Modeling host-pathogen interactions in the context of the microenvironment: three-dimensional cell culture comes of age. Infect Immun 86:e00282 -18. https://doi.org/10.1128/IAI.00282-18. 22. Basketter DA, Clewell H, Kimber I, Rossi A, Blaauboer B, Burrier R, Daneshian M, Eskes C, Goldberg A, Hasiwa N, Hoffmann S, Jaworska J, Knudsen TB, Landsiedel R, Leist M, Locke P, Maxwell G, McKim J, McVey EA, Ouédraogo G, Patlewicz G, Pelkonen O, Roggen E, Rovida C, Ruhdel I, Schwarz M, Schepky A, Schoeters G, Skinner N, Trentz K, Turner M, Vanparys P, Yager J, Zurlo J, Hartung T. 2012. A roadmap for the development of alternative (non-animal) methods for systemic toxicity testing. Altex 29:3–91. https://doi.org/10.14573/altex.2012.1.003. 23. Leist M, Hasiwa N, Rovida C, Daneshian M, Basketter D, Kimber I, Clewell H, Gocht T, Goldberg A, Busquet F, Rossi A-M, Schwarz M, Stephens M, Taalman R, Knudsen TB, McKim J, Harris G, Pamies D, Hartung T. 2014. Consensus report on the future of animal-free systemic toxicity testing. Altex 31:341–356. https://doi.org/10.14573/altex.1406091. 24. Westermann AJ, Barquist L, Vogel J. 2017. Resolving host-pathogen interactions by dual RNA-seq. PLoS Pathog 13:e1006033. https://doi.org/ 10.1371/journal.ppat.1006033. 25. Thanert R, Goldmann O, Beineke A, Medina E. 2017. Host-inherent variability influences the transcriptional response of Staphylococcus aureus during in vivo infection. Nat Commun 8:14268. https://doi.org/ 10.1038/ncomms14268. 26. Kusmierek M, Heroven AK, Beckstette M, Nuss AM, Dersch P. 2019. Discovering Yersinia-host interactions by tissue dual RNA-Seq. Methods Mol Biol 2010:99–116. https://doi.org/10.1007/978-1-4939-9541-7_8. 27. Nuss AM, Beckstette M, Pimenova M, Schmühl C, Opitz W, Pisano F, Heroven AK, Dersch P. 2017. Tissue dual RNA-seq allows fast discovery of infection-specific functions and riboregulators shaping host-pathogen transcriptomes. Proc Natl Acad SciUSA114:E791–E800. https://doi.org/ 10.1073/pnas.1613405114. 28. Luissint AC, Parkos CA, Nusrat A. 2016. Inflammation and the intestinal barrier: leukocyte-epithelial cell interactions, cell junction remodeling, and mucosal repair. Gastroenterology 151:616–632. https://doi.org/10 .1053/j.gastro.2016.07.008. 29. Papenfort K, Said N, Welsink T, Lucchini S, Hinton JCD, Vogel J. 2009. Specific and pleiotropic patterns of mRNA regulation by ArcZ, a conserved, Hfq-dependent small RNA. Mol Microbiol 74:139–158. https:// doi.org/10.1111/j.1365-2958.2009.06857.x. 30. Westermann AJ, Förstner KU, Amman F, Barquist L, Chao Y, Schulte LN, Müller L, Reinhardt R, Stadler PF, Vogel J. 2016. Dual RNA-seq unveils noncoding RNA functions in host-pathogen interactions. Nature 529: 496–501. https://doi.org/10.1038/nature16547. 31. Mai J, Virtue A, Shen J, Wang H, Yang XF. 2013. An evolving new paradigm: endothelial cells—conditional innate immune cells. J Hematol Oncol 6:61. https://doi.org/10.1186/1756-8722-6-61. 32. Atianand MK, Caffrey DR, Fitzgerald KA. 2017. Immunobiology of long noncoding RNAs. Annu Rev Immunol 35:177–198. https://doi.org/10 .1146/annurev-immunol-041015-055459. 33. Schulte LN, Bertrams W, Stielow C, Schmeck B. 2019. ncRNAs in inflammatory and infectious diseases. Methods Mol Biol 1912:3–32. https://doi .org/10.1007/978-1-4939-8982-9_1. 34. Palmer AD, Kim K, Slauch JM. 2019. PhoP-mediated repression of the SPI1 T3SS in Salmonella enterica serovar Typhimurium. J Bacteriol 201: e00264-19. https://doi.org/10.1128/JB.00264-19. 35. Groisman EA. 2001. The pleiotropic two-component regulatory system PhoP-PhoQ. J Bacteriol 183:1835–1842. https://doi.org/10.1128/JB.183.6 .1835-1842.2001. 36. Lee EJ, Groisman EA. 2010. An antisense RNA that governs the expression kinetics of a multifunctional virulence gene. Mol Microbiol 76: 1020–1033. https://doi.org/10.1111/j.1365-2958.2010.07161.x. 37. Masse E, Vanderpool CK, Gottesman S. 2005. Effect of RyhB small RNA on global iron use in Escherichia coli. J Bacteriol 187:6962–6971. https:// doi.org/10.1128/JB.187.20.6962-6971.2005. 38. Padalon-Brauch G, Hershberg R, Elgrably-Weiss M, Baruch K, Rosenshine I, Margalit H, Altuvia S. 2008. Small RNAs encoded within genetic islands of Salmonella typhimurium show host-induced expression and role in virulence. Nucleic Acids Res 36:1913–1927. https://doi.org/10.1093/nar/ gkn050. 39. Pfeiffer V, Sittka A, Tomer R, Tedin K, Brinkmann V, Vogel J. 2007. A small non-coding RNA of the invasion gene island (SPI-1) represses outer membrane protein synthesis from the Salmonella core genome. Mol Microbiol 66:1174–1191. https://doi.org/10.1111/j.1365-2958.2007.05991.x. 40. Chao Y, Papenfort K, Reinhardt R, Sharma CM, Vogel J. 2012. An atlas of Hfq-bound transcripts reveals 3=UTRs as a genomic reservoir of regulatory small RNAs. EMBO J 31:4005–4019. https://doi.org/10.1038/emboj .2012.229. 41. Ma H, Han P, Ye W, Chen H, Zheng X, Cheng L, Zhang L, Yu L, Wu X, Xu Z, Lei Y, Zhang F. 2017. The long noncoding RNA NEAT1 exerts antihantaviral effects by acting as positive feedback for RIG-I signaling. J Virol 91:e2250-16. https://doi.org/10.1128/JVI.02250-16. 42. Ko YP, Flick MJ. 2016. Fibrinogen is at the interface of host defense and pathogen virulence in Staphylococcus aureus infection. Semin Thromb Hemost 42:408–421. https://doi.org/10.1055/s-0036-1579635. 43. Hannemann S, Gao B, Galan JE. 2013. Salmonella modulation of host cell gene expression promotes its intracellular growth. PLoS Pathog 9:e1003668. https://doi.org/10.1371/journal.ppat.1003668. 44. Hannemann S, Galan JE. 2017. Salmonella enterica serovar-specific transcriptional reprogramming of infected cells. PLoS Pathog 13:e1006532. https://doi.org/10.1371/journal.ppat.1006532. 45. Grivennikov S, Karin E, Terzic J, Mucida D, Yu G-Y, Vallabhapurapu S, Scheller J, Rose-John S, Cheroutre H, Eckmann L, Karin M. 2009. IL-6 and Stat3 are required for survival of intestinal epithelial cells and development of colitis-associated cancer. Cancer Cell 15:103–113. https://doi .org/10.1016/j.ccr.2009.01.001. 46. Hodge DR, Hurt EM, Farrar WL. 2005. The role of IL-6 and STAT3 in inflammation and cancer. Eur J Cancer 41:2502–2512. https://doi.org/10 .1016/j.ejca.2005.08.016. 47. Leonard F, Collnot EM, Lehr CM. 2010. A three-dimensional coculture of enterocytes, monocytes and dendritic cells to model inflamed intestinal mucosa in vitro. Mol Pharm 7:2103–2119. https://doi.org/10.1021/ mp1000795. 48. Higbee RG, Byers AM, Dhir V, Drake D, Fahlenkamp HG, Gangur J, Kachurin A, Kachurina O, Leistritz D, Ma Y, Mehta R, Mishkin E, Moser J, Mosquera L, Nguyen M, Parkhill R, Pawar S, Poisson L, Sanchez-Schmitz G, Schanen B, Singh I, Song H, Tapia T, Warren W, Wittman V. 2009. An immunologic model for rapid vaccine assessment—a clinical trial in a test tube. Altern Lab Anim 37(Suppl 1):19–27. https://doi.org/10.1177/ 026119290903701S05. 49. Forbester JL, Goulding D, Vallier L, Hannan N, Hale C, Pickard D, Mukhopadhyay S, Dougan G. 2015. Interaction of Salmonella enterica serovar Typhimurium with intestinal organoids derived from human induced pluripotent stem cells. Infect Immun 83:2926–2934. https://doi.org/10 .1128/IAI.00161-15. 50. Co JY, Margalef-Català M, Li X, Mah AT, Kuo CJ, Monack DM, Amieva MR. 2019. Controlling epithelial polarity: a human enteroid model for hostSchulte et al. ® January/February 2020 Volume 11 Issue 1 e03348-19 mbio.asm.org 18 on March 11, 2020 at GESELLSCHAFT FUR BIOTECHNO-http://mbio.asm.org/Downloaded from pathogen interactions. Cell Rep 26:2509–2520.e4. https://doi.org/10 .1016/j.celrep.2019.01.108. 51. Yin YB, Zhou DG. 2018. Organoid and enteroid modeling of Salmonella infection. Front Cell Infect Microbiol 8:257. https://doi.org/10.3389/fcimb .2018.00257. 52. Bein A, Shin W, Jalili-Firoozinezhad S, Park MH, Sontheimer-Phelps A, Tovaglieri A, Chalkiadaki A, Kim HJ, Ingber DE. 2018. Microfluidic organon-a-chip models of human intestine. Cell Mol Gastroenterol Hepatol 5:659–668. https://doi.org/10.1016/j.jcmgh.2017.12.010. 53. Salerno-Goncalves R, Fasano A, Sztein MB. 2011. Engineering of a multicellular organotypic model of the human intestinal mucosa. Gastroenterology 141:e18–e20. https://doi.org/10.1053/j.gastro.2011.04.062. 54. Kurtz JR, Goggins JA, McLachlan JB. 2017. Salmonella infection: interplay between the bacteria and host immune system. Immunol Lett 190: 42–50. https://doi.org/10.1016/j.imlet.2017.07.006. 55. Kerneis S, Bogdanova A, Kraehenbuhl JP, Pringault E. 1997. Conversion by Peyer’s patch lymphocytes of human enterocytes into M cells that transport bacteria. Science 277:949–952. https://doi.org/10.1126/science .277.5328.949. 56. Blanco LP, DiRita VJ. 2006. Bacterial-associated cholera toxin and GM1 binding are required for transcytosis of classical biotype Vibrio cholerae through an in vitro M cell model system. Cell Microbiol 8:982–998. https://doi.org/10.1111/j.1462-5822.2005.00681.x. 57. Jaslow SL, Gibbs KD, Fricke WF, Wang L, Pittman KJ, Mammel MK, Thaden JT, Fowler VG, Hammer GE, Elfenbein JR, Ko DC. 2018. Salmonella activation of STAT3 signaling by SarA effector promotes intracellular replication and production of IL-10. Cell Rep 23:3525–3536. https:// doi.org/10.1016/j.celrep.2018.05.072. 58. Lu R, Wu S, Zhang Y-G, Xia Y, Zhou Z, Kato I, Dong H, Bissonnette M, Sun J. 2016. Salmonella protein AvrA activates the STAT3 signaling pathway in colon cancer. Neoplasia 18:307–316. https://doi.org/10.1016/j.neo .2016.04.001. 59. Kuhn KA, Manieri NA, Liu TC, Stappenbeck TS. 2014. IL-6 stimulates intestinal epithelial proliferation and repair after injury. PLoS One 9:e114195. https://doi.org/10.1371/journal.pone.0114195. 60. Westermann AJ, Venturini E, Sellin ME, Förstner KU, Hardt W-D, Vogel J. 2019. The major RNA-binding protein ProQ impacts virulence gene expression in Salmonella enterica serovar Typhimurium. mBio 10:e02504 -18. 61. Radtke AL, Wilson JW, Sarker S, Nickerson CA. 2010. Analysis of interactions of Salmonella type three secretion mutants with 3-D intestinal epithelial cells. PLoS One 5:e15750. https://doi.org/10.1371/journal.pone .0015750. 62. Schweinlin M, Wilhelm S, Schwedhelm I, Hansmann J, Rietscher R, Jurowich C, Walles H, Metzger M. 2016. Development of an advanced primary human in vitro model of the small intestine. Tissue Eng Part C Methods 22:873–883. https://doi.org/10.1089/ten.TEC.2016.0101. 63. Nickerson CA, Goodwin TJ, Terlonge J, Ott CM, Buchanan KL, Uicker WC, Emami K, LeBlanc CL, Ramamurthy R, Clarke MS, Vanderburg CR, Hammond T, Pierson DL. 2001. Three-dimensional tissue assemblies: novel models for the study of Salmonella enterica serovar Typhimurium pathogenesis. Infect Immun 69:7106–7120. https://doi.org/10.1128/IAI.69.11 .7106-7120.2001. 64. Song J, Willinger T, Rongvaux A, Eynon EE, Stevens S, Manz MG, Flavell RA, Galán JE. 2010. A mouse model for the human pathogen Salmonella typhi. Cell Host Microbe 8:369–376. https://doi.org/10.1016/j.chom.2010 .09.003. 65. Mathur R, Oh H, Zhang D, Park S-G, Seo J, Koblansky A, Hayden MS, Ghosh S. 2012. A mouse model of Salmonella typhi infection. Cell 151:590–602. https://doi.org/10.1016/j.cell.2012.08.042. 66. Manz MG. 2007. Human-hemato-lymphoid-system mice: opportunities and challenges. Immunity 26:537–541. https://doi.org/10.1016/j.immuni .2007.05.001. 67. Hansmann J, Groeber F, Kahlig A, Kleinhans C, Walles H. 2013. Bioreactors in tissue engineering—principles, applications and commercial constraints. Biotechnol J 8:298–307. https://doi.org/10.1002/biot.201200162. 68. Bricks T, Paullier P, Legendre A, Fleury M-J, Zeller P, Merlier F, Anton PM, Leclerc E. 2014. Development of a new microfluidic platform integrating co-cultures of intestinal and liver cell lines. Toxicol In Vitro 28:885–895. https://doi.org/10.1016/j.tiv.2014.02.005. 69. Maschmeyer I, Hasenberg T, Jaenicke A, Lindner M, Lorenz AK, Zech J, Garbe L-A, Sonntag F, Hayden P, Ayehunie S, Lauster R, Marx U, Materne E-M. 2015. Chip-based human liver-intestine and liver-skin cocultures—a first step toward systemic repeated dose substance testing in vitro. Eur J Pharm Biopharm 95:77–87. https://doi.org/10.1016/j.ejpb .2015.03.002. 70. Kim HJ, Ingber DE. 2013. Gut-on-a-chip microenvironment induces human intestinal cells to undergo villus differentiation. Integr Biol (Camb) 5:1130–1140. https://doi.org/10.1039/c3ib40126j. 71. Saliba AE, Westermann AJ, Gorski SA, Vogel J. 2014. Single-cell RNA-seq: advances and future challenges. Nucleic Acids Res 42:8845–8860. https://doi.org/10.1093/nar/gku555. 72. Groeber F, Kahlig A, Loff S, Walles H, Hansmann J. 2013. A bioreactor system for interfacial culture and physiological perfusion of vascularized tissue equivalents. Biotechnol J 8:308–316. https://doi.org/10.1002/biot .201200160. 73. Schulte LN, Eulalio A, Mollenkopf HJ, Reinhardt R, Vogel J. 2011. Analysis of the host microRNA response to Salmonella uncovers the control of major cytokines by the let-7 family. EMBO J 30:1977–1989. https://doi .org/10.1038/emboj.2011.94. 74. Ran FA, Hsu PD, Wright J, Agarwala V, Scott DA, Zhang F. 2013. Genome engineering using the CRISPR-Cas9 system. Nat Protoc 8:2281–2308. https://doi.org/10.1038/nprot.2013.143. 75. Livak KJ, Schmittgen TD. 2001. Analysis of relative gene expression data using real-time quantitative PCR and the 2(⫺delta delta C(T)) method. Methods 25:402–408. https://doi.org/10.1006/meth.2001.1262. 76. Saldanha AJ. 2004. Java Treeview—extensible visualization of microarray data. Bioinformatics 20:3246–3248. https://doi.org/10.1093/bioinformatics/ bth349. 77. Kamburov A, Wierling C, Lehrach H, Herwig R. 2009. ConsensusPathDB—a database for integrating human functional interaction networks. Nucleic Acids Res 37:D623–D628. https://doi.org/10.1093/nar/ gkn698. A Coculture Model Mimicking Human Intestinal Infection ® January/February 2020 Volume 11 Issue 1 e03348-19 mbio.asm.org 19 on March 11, 2020 at GESELLSCHAFT FUR BIOTECHNO-http://mbio.asm.org/Downloaded from