key: cord-0296658-rca49jzr
authors: Mora, Darcy S.O.; Cox, Madeline; Magunda, Forgivemore; Williams, Ashley B.; Linke, Lyndsey
title: An optimized live bacterial delivery platform for the production and delivery of therapeutic nucleic acids and proteins
date: 2021-10-17
journal: bioRxiv
DOI: 10.1101/2021.10.17.464697
sha: 37c4059c3084e095454dfb61fb2efbf40fd2185e
doc_id: 296658
cord_uid: rca49jzr

There is an unmet need for delivery platforms that realize the full potential of next-generation therapeutic and vaccine technologies, especially those that require intracellular delivery of nucleic acids. The in vivo usefulness of the current state-of-the-art delivery systems is limited by numerous intrinsic weaknesses, including lack of targeting specificity, inefficient entry and endosomal escape into target cells, undesirable immune activation, off-target effects, a small therapeutic window, limited genetic encoding and cargo capacity, and manufacturing challenges. Here we present our characterization of a delivery platform based on the use of engineered live, tissue-targeting, non-pathogenic bacteria (Escherichia coli strain SVC1) for intracellular cargo delivery. The SVC1 bacteria are engineered to specifically bind to epithelial cells via a surface-expressed targeting ligand, to escape the endosome upon intracellularization, and to have minimal immunogenicity. Here we report findings on key features of this system. First, we demonstrated that bacterial delivery of a short hairpin RNA (shRNA) can target and silence a gene in an in vitro mammalian respiratory cell model. Next, we used an in vivo mouse model to demonstrate that SVC1 bacteria are invasive to epithelial cells of various tissues and organs (eye, nose, mouth, stomach, vagina, skeletal muscle, and lungs) via local administration. We also showed that repeat dosing of SVC1 bacteria to the lungs is minimally immunogenic and that it does not have adverse effects on tissue homeostasis. Finally, to validate the potential of SVC1 bacteria in therapeutic applications, we demonstrated that bacterial delivery of influenza-targeting shRNAs to the respiratory tissues can mitigate viral replication in a mouse model of influenza infection. Our ongoing work is focused on further refining this platform for efficient delivery of nucleic acids, gene editing machinery, and therapeutic proteins, and we expect that this platform technology will enable a wide range of advanced therapeutic approaches.

of E. coli-mediated nucleic acid delivery for development in other clinical applications 10,11,49 .

Linke and colleagues developed an shRNA-based approach for the prophylaxis and treatment of avian influenza virus 9 . In this system, two E. coli strains were engineered to independently silence the expression of two essential influenza viral proteins: an RNA polymerase subunit (polymerase acidic protein; PA) and a capsid protein (nucleoprotein; NP). When administered together to the respiratory tract, these bacterial strains can target most known influenza A strains while minimizing the development of therapeutic resistance, especially that resulting from genetic drift [50] [51] [52] [53] . The safety and efficacy of this approach were robustly validated in a chicken model of avian influenza 9 , demonstrating that E. coli-mediated nucleic acid delivery can indeed be used to mitigate the replication and shedding of a clinically important virus.

In this study, we have established our engineered strain, E. coli SVC1, as a viable system for in vivo applications, particularly via validation in a murine model. We first characterized the duration of the gene silencing it can mediate in vitro. We then examined the feasibility of directly administering SVC1 to specific tissues and organs and assessed its biodistribution in clinically relevant sites in vivo. Next, we investigated the safety profile of SVC1 upon repeated dosing to the lungs via histological analysis and monitoring of expression level changes in innate and adaptive immunity-related genes. Finally, we validated the ability of SVC1 to deliver therapeutic shRNAs, functioning as an antiviral, using an in vivo influenza virus infection model. Taken together, the results presented here support the potential of SVC1 to serve as a powerful delivery platform for therapeutic nucleic acids and support translational research to drive its future clinical development.

To assess the RNAi activity of shRNAs expressed from the SiVEC plasmid (pSiVEC) and delivered to eukaryotic cells by invasive, non-pathogenic E. coli cells, we measured green fluorescent protein (GFP) depletion in human alveolar basal epithelial cells (A549 cells) stably expressing GFP. The host E. coli strain (SVC1) was previously engineered to be invasive to eukaryotic epithelial cells ( Figure 1A) . Via an interaction between Yersinia pseudotuberculosis invasin on the bacterial surface and b1 integrin on the surface of the target cells, SVC1 bacteria invade eukaryotic cells via receptor-mediated endocytosis. Upon entry, the bacteria lyse in the endosome to release their cargo, including the shRNA and LLO, the product of Listeria monocytogenes hlyA, which perforates the endosome allowing the shRNA to enter the cytoplasm of the invaded cells. We treated A549 cells with SVC1 carrying pSiVEC encoding a GFP shRNA (GFP-shRNA) or pSiVEC encoding a non-targeting small RNA (scramble) for two hours. Subsequently, we removed the bacteria, and the cells were further incubated for an additional 96 hours. We measured GFP expression 6, 24, 48, 72, and 96 hours after bacterial removal ( Figure 1B) . As shown in Figure 1C , GFP expression was robustly reduced in A549 cells treated with GFP-shRNA compared with A549 cells treated with scramble. GFP depletion persisted over 96 hours at both a low dose (Figure 1D , left) and at a high dose of bacteria ( Figure   1D , right). As would be expected, the higher dose of invasive bacteria resulted in more robust GFP depletion (i.e., a dose-dependent response), suggesting that the level of depletion achieved via shRNA delivery can be controlled by varying the number of invasive bacteria. The length of the cell cycle of A549 cells under the growth conditions used here is approximately 20 hours (Supplemental Figure 1) ; thus, some cell division likely occurred over the course of this experiment. The observation that GFP expression did not increase during at least the first 72 hours of the time course suggests that the abundance of delivered shRNA is sufficient to be inherited by the daughter cells of the originally invaded cells, which has been previously reported for siRNA 54 . Taken together, the data presented in Figure 1 confirm that SVC1 can invade eukaryotic epithelial cells and deliver a cargo of shRNA that is then processed via the RNAi pathway to robustly and persistently silence the expression of a target gene.

Invasive SVC1 E. coli can be administered to various epithelial tissue types SVC1 binds to target cells via an interaction between its surface-expressed invasin ( Figure 1A) and b1 integrin on the surface of the target eukaryotic cells. b1 integrin is expressed by epithelial cells in multiple tissue types, including those in the cornea, respiratory tract, reproductive tract, digestive tract, and skeletal muscle. This ubiquity suggests that SVC1 can be used as a flexible approach for delivering therapeutic moieties to various tissues. To examine the versatility of SVC1 as a delivery system, we used various methods (Table 1) to apply invasive, fluorescently labelled SVC1 bacteria to the eye, upper (nasal cavity) and lower (lungs) respiratory tract, vagina, digestive tract, and skeletal muscle. As shown in Figure 2A , SVC1 bacteria remained localized following administration, suggesting that they can indeed be used for organ-and tissue-specific delivery of therapeutic cargo. These results validate the versatility of SVC1 for use as a delivery vehicle in various tissue types. We aim to administer the fewest SVC1 bacterial cells required to achieve the desired effect. While our ongoing work suggests that target cell invasion is efficient, our current routes of in vivo administration likely introduce a surplus of bacteria. In light of this consideration, we were interested in the fate of any excess bacterial cells. The clearance of the system is especially important with regard to trafficking of the SVC1 delivery vehicle to the liver, which is often undesirable in drug delivery applications due to associated toxicity [55] [56] [57] To explore this feature of SVC1, we administered SVC1 intramuscularly in the hind limb and then collected a section of muscle tissue that received the localized injection, liver, and proximal draining lymph node after 20 and 72 hours. We then tested for the presence of the delivery vehicle in the injected muscle as well as the liver and lymph nodes using PCR with primers that detect the SVC1-borne plasmid (pSiVEC) in total DNA isolated from each tissue sample as shown in the table in Figure 2B , the PCR results suggest that the excess bacteria are cleared via the lymphatic system rather than being trafficked to the liver.

To explore the potential of SVC1 as a tissue-targeting nucleic acid delivery platform, we focused on the respiratory tract. By virtue of their ability to silence expression of target proteins, therapeutic shRNAs have broad usefulness in the treatment of a variety of infectious diseases of the lungs, including respiratory viruses (e.g., influenza and SARS viruses); however, because of the intrinsic instability of shRNAs in the cytoplasm and the transience of the SVC1 bacteria at the delivery site (see above), a robust therapeutic effect would likely require repeat administration. Various issues with repeat dosing (e.g., undesirable immune responses and acquired resistance) limit the usefulness of some current delivery systems, especially adenoassociated virus (AAV)-based systems 33-37 ; therefore, we were interested in whether SVC1 could overcome such limitations, especially detrimental immunogenicity. To this end, we examined the effects of repeat dosing of SVC1 to the respiratory tract. We administered six doses of 50 µL containing 1x10 9 SVC1 bacteria suspended in sterile phosphate-buffered saline (PBS)

intranasally to mice over the course of 60 hours ( Figure 3A) . As a control, we also treated mice with PBS alone (sham). Twenty-four hours after the sixth dose, we euthanized the mice and collected tissue samples for analysis. We monitored posture, grooming, interest in food, and behavior, and all were normal prior to euthanasia, and all of the SVC1-treated mice appeared healthy. Upon dissection, we found no gross abnormalities or differences in the lungs, liver, kidneys, heart, abdominal cavity, nasal cavity, spleen, or other internal anatomy, and we did not observe any gross lesions (data not shown). Nasal sections from bacteria-and sham-treated mice had normal morphology with no inflammation, no bacteria, no increase in mucus secretion, and no alteration in the mucociliary apparatus (representative nasal sections are shown in Figure   3 , B, C). We also examined the lungs for three pathologies: inflammation, immune cell infiltration, and injury (Figure 3 , D-F). The lung tissue samples all had minimal to mild increase in alveolar and intracapillary macrophages affecting predominantly one or two lobes that did not vary significantly among the treated and control, suggesting a baseline responsive mild lung infiltrate that was not distinctly due to the treatment nor made worse by the treatment. In rare cases, the blood vessels were minimally reactive and surrounded by edema and rare fibrin, supporting the possibility of a hematogenous antigen source unrelated to either treatment (SVC1 or sham). A single SVC1-treated mouse had severe lymphoplasmacytic and histiocytic pneumonia that varied significantly from the other mice. There was subjectively a moderate lymphoid aggregate (BALT) hyperplasia, supporting a mild baseline immune response; however, this pathology was uniformly present and could not be ascribed to SVC1 bacterial treatment. The immunogenicity of SVC1 was then examined objectively as described in detail below. We did not observe SVC1 bacteria on hematoxylin and eosin-stained sections in any tissue sample, suggesting that the bacteria were rapidly cleared and/or that invasion was rapid and robust. Taken together, these results establish SVC1 as a safe bacterial delivery vehicle for repeated intranasal delivery to the respiratory tissues.

To evaluate the systemic immunogenicity of SVC1 in the respiratory tract, we collected the spleens from the mice in the repeat dosing experiment, purified total RNA, and examined gene expression differences between sham and SVC1-treated mice using a Qiagen RT 2 Profiler PCR Array for mouse adaptive and innate immune responses ( Figure 4A ). This array allows simultaneous monitoring of 84 immune-related genes, which taken together are representative of key innate and adaptive immune responses. As SVC1 was developed from a commensal, highly attenuated E. coli strain, we expected that it would be minimally immunogenic in the respiratory tract, even after repeated dosing. Among the 84 immune-related genes analyzed, only 7 had statistically significant (α=0.05) expression level changes in the SVC1 bacteria-treated mice compared with the sham-treated mice: Ccr4, Ccr5, IL1a (Interleukin 1a), H2-Q10, Il5

(interleukin 5), Nlrp3, and Rorc ( Figure 4B ). The upregulation of these genes likely reflects the presence of bacteria, as each has been linked to responses to bacterial lipopolysaccharide (LPS) [58] [59] [60] [61] [62] [63] [64] [65] . The upregulation of LPS-related genes is not unexpected because while the LPS of SVC1 is truncated (lacking the O-antigen), it remains recognizable, though only mildly immunogenic, in mammals 66-68 . In addition, the gene encoding myeloperoxidase (MPO) was upregulated over 2fold. While this change was not statistically significant, it could also reflect an effect of bacterial presence as MPO has been linked to a response to pro-inflammatory agents in the lung epithelium 69 . Repeated treatment did not result in significant non-specific or specific changes in the expression levels of pattern recognition receptors (PRRs), cytokines/chemokines, innate/adaptive markers, inflammatory or bacterial defense markers. Taken together, these data indicate that repeat dosing with SVC1 to the respiratory tract in mice does not induce a robust immune response compared to PBS (sham) dosed mice, suggesting that SVC1 is minimally immunogenic and safe for repeat dosing in a mammal.

Finally, we wanted to demonstrate the in vivo therapeutic potential of SVC1 to deliver anti-viral shRNAs. To this end, we designed two therapeutic strains: an SVC1 derivative expressing an shRNA against the influenza A virus (IAV) PA protein (RNA polymerase complex subunit) (SVC1-PA) and an SVC1 derivative expressing an shRNA against the influenza NP protein (nucleocapsid) (SVC1-NP). These strains are mixed prior to administration to produce the SiVEC-IAV cocktail.

Upon simultaneous delivery of the shRNAs to the cytoplasm of a respiratory epithelial cell (the site of IAV replication) via administration of SiVEC-IAV, they are processed via the RNAi pathway into siRNAs that silence PA and NP expression, thereby inhibiting IAV replication and reducing viral shedding. To test the efficacy of these shRNAs delivered via SVC1, we dosed mice with a cocktail of SVC1-PA and SVC1-NP at three doses (low, medium, high) twice prior to viral challenge and then four times after the mice were exposed to H1N1 IAV (PR8 strain) as described in Figure 5A . On days 3, 5, 7, and 9 post-challenge, we collected the nasal turbinates, purified total RNA, and determined the viral titers (as EID50 equivalent/mL) via reverse transcription quantitative PCR (RT-qPCR). As shown in Figure 5B , reductions in viral titer were observed at the low, medium, and high bacterial doses, with a clear dose-response trend. As expected, the highest dose was the most effective at reducing viral titer in the nasal turbinates. These results demonstrate that SVC1 can be used as an effective vehicle for the delivery of therapeutic shRNAs to the lungs in a respiratory disease model.

A significant factor limiting the translation of therapeutic nucleic acids, proteins, and gene editing technologies from bench to bedside is the absence of safe and robust vehicles for targeted delivery to affected cells and tissues. In the case of nucleic acids, this limitation is imposed by the negative charge, instability, immunogenicity, and in some cases the large size of nucleic acids, particularly relative to typical small-molecule drugs. The SVC1-based platform described here offers an elegant solution to the targeted delivery conundrum and holds promise for utility in ameliorating a range of diseases and disorders. SVC1 was engineered 1) to constitutively express nucleic acids, 2) to target clinically relevant cell types (i.e., mucosal epithelial cells), and

3) to escape the endosome allowing the release of the nucleic acids into the target cell cytoplasm. SVC1 can be tailored to produce different types of therapeutically relevant nucleic acids. Furthermore, based on the targeted administration results shown in Figure 2 , SVC1 might be useful for treating a plethora of oral, respiratory, gastrointestinal, ocular, vaginal, and rectal diseases. While not investigated here, SVC1 can be used for delivery of proteins, eukaryotetranslatable mRNA, and gene-editing systems (e.g., CRISPR/Cas) to targeted mucosal tissues, providing cellular uptake and, where appropriate, nuclear translocation of the gene-editing nuclease system, without the need for host genome integration.

A key feature of any drug delivery system is the amount of therapeutic moiety that can be The intracellular delivery modalities currently in use in the clinic, e.g., lipid nanoparticles and viral vectors, suffer from undesirable immune effects that can limit their utility, particularly for repeated dosing. The preliminary analysis of the immunogenicity of SVC1 delivered to the lungs (after 6 repeated doses over three days) presented here suggests that the bacteria are minimally immunogenic, as no significant immune cell infiltration ( Figure 3F ) nor statistically significant changes in the expression level of any screened immune-related gene (2-fold or greater) ( Figure   4 ) were observed. Interestingly, of the systemically upregulated genes (n=7) detected after repeated respiratory administration, genes related to cellular responses to LPS were overrepresented (7 out of 7) [58] [59] [60] [61] [62] [63] [64] [65] . This observation suggests that the array analysis used here is sufficiently sensitive to detect subtle gene expression changes in response to the presence of the bacterial LPS but that the amount of LPS delivered via this administration scheme was sufficiently low to not induce a robust systemic immune response. Finally, tissue damage was not observed even in tissue directly exposed to the bacterial vehicle (Figure 3B-E) . We are currently further modifying SVC1 to express a less-immunogenic LPS to further mitigate immunogenicity concerns. Work is also underway to explore whether the SVC1 system is affected by acquired immunity. However, an acquired immune response to the SVC1 bacteria is not anticipated based on the gene expression analysis described here as well as the additional genetic modifications we are making to further attenuate the LPS of SVC1.

To demonstrate the potential of SVC1 as an in vivo therapeutic delivery vehicle, we demonstrated that simultaneous delivery of shRNAs designed to silence two essential influenza genes (the SiVEC-IAV cocktail) could ameliorate viral replication in a mouse model of influenza infection ( Figure 5) . To our knowledge, this work represents the first demonstration of the use of a bacteria-based delivery system in a mammalian antiviral application. The robust reductions in viral replication shown in Figure 5B confirm that SVC1 can indeed be used to deliver therapeutic RNA molecules. Importantly, our data also revealed a dose-dependent reduction in viral replication when different numbers of SVC1 cells were intranasally administered. This dose responsiveness demonstrates that the number of bacteria delivered can be modulated to achieve different therapeutic outcomes, which might be advantageous in some applications of such a platform (e.g., delivery of gene editing components).

Finally, a distinguishing advantage of the SVC1 bacterial delivery platform (in comparison to other available delivery platforms) is that the bacteria themselves can produce the therapeutic moieties that they deliver, as demonstrated here by the bacterial transcription of shRNAs that feed into the host cellular RNAi pathway. This feature of the system eliminates RNA manufacturing steps and production costs 73-76 . As it is simple and fast to generate large quantities of bacteria using widely available manufacturing approaches, SVC1-based therapeutic products could be readily generated in massive quantities from a small stock, and if properly stored, could have a long shelf life. With manufacturing in mind, we are currently working on characterization of potency, including developing methods to quantitate the number of shRNA molecules generated per SVC1 bacterial cell.

Our ongoing research and development efforts are focused on optimization of SVC1 as a platform for the production (via bacterial transcription) and delivery of both linear and circular eukaryote-translatable mRNAs and for the production and delivery of gene editing proteins and RNAs (i.e., CRISPR/Cas machinery). Due to the vast genetic coding capacity and transcriptional flexibility of E. coli, SVC1 can express and deliver high molecular weight RNA molecules.

Furthermore, our vast knowledge of E. coli molecular genetics enables further applicationspecific optimization (e.g., additional modulation of RNase activities) to improve its performance as a highly versatile delivery platform. We expect that the advantages offered by live bacteria, and SVC1 in particular, will lead to future studies that further enable and validate the usefulness of bacteria as a powerful multi-application delivery platform.

The invasive E. coli strain SVC1 is a K-12 derivative [F -endA1 hsdR17 (rK -mK + ) glnV44 thi-1 relA1 rfbD1 spoT1 Δrnc ΔdapA]. The cells are auxotrophic for diaminopimelic acid (DAP) due to a deletion of dapA. E. coli cells were cultured in brain-heart infusion (BHI) medium supplemented with DAP (100 µg/mL) and appropriate antibiotics at the following concentrations: kanamycin, The statistical significance of the differences between the SVC1-scramble control and SVC1 -GFP experimental groups was assessed using two-way ANOVA (p < 0.005).

In vivo biodistribution assays. To characterize the biodistribution of the SiVEC vehicle following localized administration to mouse mucosal epithelial and skeletal muscle tissues (Figure 2A ) were anesthetized with inhaled isoflurane in an anesthesia chamber and then transferred to the Perkin Elmer IVIS Spectrum system for in vivo imaging. Mice were treated with SVC1-scramble (invasive) and an untreated control mouse was included for calibration of background signal due to autofluorescence. The route of administration and dose per tissue type is shown in Table 1 .

To demonstrate localized delivery to mucosal epithelia (Figure 2A) , mice were imaged at 0, 2, 4, 6, and 18 hours post-administration to the eyes, lungs, nose, vagina, oral cavity, and stomach as described in Table 1 In vivo IAV challenge assays. To demonstrate the therapeutic potential of the SiVEC delivery vehicle, SVC1-PA and SVC1-NP were constructed to express shRNAs targeting the influenza A viral PA and NP mRNAs, respectively, for delivery to the respiratory tissues in an established murine influenza disease model [78] [79] [80] . These strains were generated as previously described 9 and were mixed 1:2 (SVC1-PA:SVC1-NP) to create an antiviral cocktail referred to as SiVEC-IAV.

Eight-week-old female BALB/c mice (n=200) were anesthetized with inhaled isoflurane and dosed with SiVEC-IAV or PBS-sham by intranasal instillation. Mice were treated with 50 µL of PBS or SiVEC-IAV in high (1x10 8 CFU/mL), medium (1x10 7 CFU/mL), or low (1x10 6 CFU/mL) doses twice prior to and four times after infection with 1x10 6 EID50 per 50 µL dose of influenza A virus, A/Puerto Rico/89VMC3/1934 (H1N1) (BEI Resources, NIAID, NIH, NR-29028) (see Figure   5A ). Three, 5, 7, and 9 days post infection (DPI), mice (n=6 per treatment/dose and time point)

were euthanized, and nasal turbinates were collected and placed in RNALater.

H1N1 virus titers in the nasal turbinates of mice treated with the high, medium, and low SiVEC-IAV doses were measured via RT-qPCR and fold-reductions in viral titer were calculated relative to the PBS-sham control group. Briefly, total RNA was extracted from approximately 5 mg of nasal turbinate tissue using the Omega Mag-Bind® Total RNA 96 RNA extraction kit with a KingFisher Flex purification system, and RNA concentration and purity was determined using a Nanodrop system. RNA was diluted to 3 ng/µL and reverse transcribed and amplified using the seconds and 95 °C for 5 seconds to visualize the melting curve for each RT-qPCR assay. The standard curve for virus quantification was generated in triplicate using a series of 10-fold dilutions from 1x10 1 to 1x10 10 of the H1N1 stock virus from which the EID50 equivalent per mL (EID50 eq/mL) of each sample was calculated. The limit of detection was determined to be 10 1 EID50/ml (1 log10 EID50/ml) per reaction. Statistical significance in fold reduction in viral titer between treated and PBS sham mice was calculated using the two-sample Wilcoxon rank-sum (Mann-Whitney) test (p < 0.05). Isberg RR, Barnes P. Subversion of integrins by enteropathogenic Yersinia. J Cell Sci 

First-in-human Phase I study of bacterial RNA interference therapeutic CEQ508 in patients with familial adenomatous polyposis (FAP)

Bacteria as gene delivery vectors for mammalian cells

Functional gene transfer from intracellular bacteria to mammalian cells

Phagocytosis: Our Current Understanding of a Universal Biological Process

ß1 integrins as therapeutic targets to disrupt hallmarks of cancer

Integrin Signaling in Cancer: Mechanotransduction, Stemness, Epithelial Plasticity, and Therapeutic Resistance

Biological containment and cloning vector transmissibility

Listeriolysin O: From bazooka to Swiss army knife

Adeno-associated viral vectormediated immune responses: Understanding barriers to gene delivery

Innate immune responses to AAV vectors

Humoral immunity to AAV vectors in gene therapy: Challenges and potential solutions

Toxicological considerations when creating nanoparticle-based drugs and drug delivery systems

Effective tools for RNA-derived therapeutics: siRNA interference or miRNA mimicry

Molecular mechanisms of RNA interference

RNAi Mechanisms and Strategies to Reduce Off-Target Effects

Preclinical and Clinical Studies on Safety of CEQ508 Bacteria Engineered to Deliver Short-Hairpin RNA to

Mediate RNA Interference Against β-catenin in the GI Tract of Patients With Familial Adenomatous Polyposis

Towards a universal influenza vaccine: different approaches for one goal

The Doctrine of Original Antigenic Sin: Separating Good from Evil

Memory T cells established by seasonal human influenza A infection cross-react with avian influenza A (H5N1) in healthy individuals

Making Better Influenza Virus Vaccines?

Insights into the kinetics of siRNA-mediated gene silencing from live-cell and live-animal bioluminescent imaging

Toxic potential of materials at the nanolevel. Science (80-)

Assessing nanoparticle toxicity

A key role for CC chemokine receptor 4 in lipopolysaccharide-induced endotoxic shock

CCR4 Is a Key Modulator of Innate Immune Responses

CCR5 signaling promotes lipopolysaccharide-induced macrophage recruitment and alveolar developmental arrest

IL-1 Transcriptional Responses to Lipopolysaccharides Are Regulated by a Complex of RNA Binding Proteins

SILAC-MS based characterization of LPS and resveratrol induced changes in adipocyte proteomics -Resveratrol as ameliorating factor on LPS induced changes

Regulation of allergic airway inflammation through Toll-like receptor 4-mediated modification of mast cell function

NLRP3 inflammasome activation involved in LPS and coal tar pitch extract-induced malignant transformation of human bronchial epithelial cells

Bronchial epithelial cells of young and old mice directly regulate the differentiation of Th2 and Th17

Characterization of lipopolysaccharides from Escherichia coli K 12 mutants

Modulating the innate immune response by combinatorial engineering of endotoxin

The structural basis of lipopolysaccharide recognition by the TLR4-MD-2 complex

Myeloperoxidase modulates lung epithelial responses to pro-inflammatory agents

Transgenerational Inheritance: Perpetuating RNAi

Molecular mechanisms of ZD1839-induced G1-cell cycle arrest and apoptosis in human lung adenocarcinoma A549 cells

In vitro cultivation of human tumors: Establishment of cell lines derived from a series of solid tumors

Matrix metalloproteinase 2-sensitive multifunctional polymeric micelles for tumor-specific co-delivery of siRNA and hydrophobic drugs

Animal models for influenza virus pathogenesis and transmission

Modeling human influenza infection in the laboratory

Development of a murine nose-only inhalation model of influenza: comparison of disease caused by instilled and inhaled A/PR/8/34. Front Cell Infect Microbiol

We thank Alan Schenkel for helpful discussions of the immune response data and Stephanie Morphet-Tepp for technical assistance. The research reported in this study was supported by the National Center for Allergy and Infectious Disease of the National Institutes of Health under award number 1R43AI140243-01A1.