Parrett, Brian J.

Background/Objectives: Antigen presenting cells (APCs) and immune cells have unique properties to drive or suppress immune responses. They are therefore key targets for the expression of vaccine antigens or transgene proteins. To better determine the utility of different molecular therapies to modify these cells, mRNA and DNA-based molecular therapy vectors were compared for their ability to genetically modify immune cells after intradermal injections in mice. DNA-based vectors included naked plasmid DNA, plasmid packaged in lipid nanoparticles (LNPs), and replication-defective adenovirus (Ad) vectors. mRNA delivery was mediated by packaging into LNPs like those used in COVID-19 vaccines. Methods: Each vector was used to deliver Cre recombinase into Cre reporter mice whose cells were activated to express green fluorescent protein (GFP) and firefly luciferase after Cre recombination. The mice were injected intradermally (ID) near the base of their tail at a site that drains into the inguinal lymph node. Luciferase activity was imaged in the living mice 1 or 4 days after vector injection. The animals were then euthanized, and luciferase activity was imaged in the draining inguinal lymph node. Cells were prepared from the intradermal injection site and from the draining lymph node to determine which immune cells were genetically modified by phenotyping CD45, CD3, and CD11b GFP-positive cells by flow cytometry. Given that the skin uniquely contains Langerhans dendritic cells, these CD207+ cells were also phenotyped in skin samples and in the draining lymph node. Results: In both the skin and in the draining lymph node, the rank order of luciferase and GFP activation by the vectors were: (1) Ad; (2) mRNA-LNP; (3) DNA-LNP; and (4) naked DNA. Only mRNA-LNP and Ad vectors mediated obvious luciferase activity in the living animals and in the draining lymph nodes by imaging. Notably, both vectors appeared to leak from the ID injection site and not only modify the draining lymph node but also strongly modify the livers of the mice. Naked DNA and DNA-LNP mediated detectable GFP activation in the skin and draining lymph node in some mice, but this activity was low and did not reach statistical significance when compared to PBS-treated animals. mRNA-LNPs and Ad both mediated significant Cre delivery in CD45+, CD3+, CD11b+, and CD207+ immune cells in the skin and in the lymph node, with adenovirus mediating consistently higher levels of expression in all of the tested cells. Conclusions: These data indicate that mRNA-LNP and Ad vectors mediate stronger modification of skin and lymph node immune cells after intradermal injections. Naked DNA and DNA-LNPs were markedly less potent at this activity than the other vectors. These data are consistent with the higher vaccine potency of mRNA-LNP and Ad vectors and suggest that approaches that increase targeting of immune cell subsets may have utility to increase efficacy while also reducing off-target modification of tissues like the liver.

Abstract Alport syndrome is caused by mutations in the type IV collagen genes COL4A3, COL4A4, and COL4A5 which are expressed in podocytes in the glomerulus of the kidney. Mutation of these genes disrupts blood filtration by the kidney leading to proteinuria and kidney failure. There is no cure for Alport Syndrome. Current therapies do not treat the genetic cause of the disease, but instead aim to delay the disease by reducing blood pressure in the kidney. Here we tested the ability of mRNA therapy to treat the disease. Mice with X-linked Alport syndrome (XLAS) due to mutation of Col4A5 were treated intravenously with lipid nanoparticles (LNPs) carrying three mRNAs encoding COL4A3, COL4A4, and COL4A5 to produce trimeric collagen IV repair proteins. This intravenous mRNA therapy significantly reduced proteinuria and blood urea nitrogen. Protection against the syndrome was maintained provided that LNP-mRNA injections were continued. However, efficacy was lost when therapy was terminated. These data provide proof of principle to apply a genetic mRNA therapy to halt progression of Alport syndrome in a mouse model of XLAS. These data also demonstrate that damage to the kidney filtration barrier can be leveraged to deliver large molecular therapies to podocytes and other cells within the kidney to treat Alport syndrome and other kidney genetic diseases.

Abstract Pulmonary fibrosis is increasingly understood to involve dysfunction within and across multiple cellular compartments, with recent attention highlighting the involvement of pulmonary vascular dysfunction in failed repair and progression of fibrosis. Formulation and delivery of lung-targeting lipid nanoparticles (LNPs) may provide a means to selectively target the lung but not systemic vasculature. However, the feasibility and efficacy of such approaches in the fibrotic lung are unknown. We sought to test whether intravenously administered lung-targeting LNPs can safely deliver mRNA to the healthy and fibrotic lung vasculature in young and aged mice and whether delivery of mRNA encoding a matricellular protein could promote fibrosis resolution. We used a selective organ targeting LNP formulation and characterized cell-specificity of delivery after bleomycin-induced lung fibrosis. We then delivered Ccn3 mRNA (encoding cellular communication network factor 3) to aged mice in the setting of established lung fibrosis and evaluated fibrotic regression and vascular repair. The matricellular protein encoded by Ccn3 was previously identified by our group as an important regulator of lung endothelial function. We found that LNP delivery was lung specific and predominantly endothelial targeting in the setting of lung fibrosis. Delivery of Ccn3 mRNA to aged mice via LNPs modestly reduced fibrosis and improved microvascular density in the lungs. Our results support the concept that cell-specific and repair-promoting cargos delivered via lung-targeting LNPs may have utility for treatment of established fibrosis.

Abstract Most adenovirus vaccines are replication-defective RD-Ads. In contrast, replication-competent adenovirus (RC-Ad) replicate antigen genes 10,000-fold in each cell to amplify immune responses. While RC-Ads vaccines are more potent than RD-Ad, they risk causing adenovirus infections in humans. We engineered “single-cycle” Ad (SC-Ad) vectors that still amplify transgenes, but that do not produce infectious progeny. RD-, SC-, and RC-Ads were tested by single immunization against influenza, Ebola virus, HIV-1, Zika virus, HCV, CMV, Clostridium difficile, and cancers in mice, hamsters, and macaques and antibody and T cell responses were measured. SC-Ad expressing GFP, influenza HA, or HIV gp140 generated antigens with 33 to 100-times less virus than RD-Ad. SC-Ad generated vaginal antibodies that persisted beyond 6 months after single intranasal immunization. Intramuscular injection of 1010 vp of SC-Ad6 expressing HIV-1 gp160 generated env antibodies in sera within 4 weeks of a single immunization in rhesus macaques. SC-Ad expressing a C. difficile toxin A and B immunogen generated toxin-binding antibody reciprocal titers above 105 that persisted in excess of a half of a year after a single immunization. Single immunization with SC-Ad protected vaccinated animals against toxin and against challenge with C. difficile spores. When SC-Ads were used to amplify the expression of immune stimulatory proteins in the context of cancer, single intratumoral injection of SC-Ad expressing 4-1BBL or GMCSF mediated significant reductions in tumor growth and extension of survival without the need for exogenous cancer antigen expression. These data suggest that SC-Ads may have utility against infectious diseases and cancer.

Most gene-based severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) vaccines are nonreplicating vectors. They deliver the gene or messenger RNA to the cell to express the spike protein but do not replicate to amplify antigen production. This study tested the utility of replication in a vaccine by comparing replication-defective adenovirus (RD-Ad) and replicating single-cycle adenovirus (SC-Ad) vaccines that express the SARS-CoV-2 spike protein. SC-Ad produced 100 times more spike protein than RD-Ad and generated significantly higher antibodies against the spike protein than RD-Ad after single immunization of Ad-permissive hamsters. SC-Ad–generated antibodies climbed over 14 weeks after single immunization and persisted for more than 10 months. When the hamsters were challenged 10.5 months after single immunization, a single intranasal or intramuscular immunization with SC-Ad-Spike reduced SARS-CoV-2 viral loads and damage in the lungs and preserved body weight better than vaccination with RD-Ad-Spike. This demonstrates the utility of harnessing replication in vaccines to amplify protection against infectious diseases.

The human species D Ad26 is being investigated as a low-seroprevalence vector for oncolytic virotherapy and gene-based vaccination against HIV-1 and SARS-CoV-2. However, there is debate in the literature about its tropism and receptor utilization, which directly influence its efficiency for certain applications. This work was aimed at determining which receptor(s) this virus uses for infection and its role in virus biology, vaccine efficacy, and, importantly, vaccine safety.

Oncolytic viruses (OVs) can have utility for direct killing of cancer cells, but may also serve to activate the immune system against cancer cells. While viruses alone can serve as immune stimulators, there is great interest in arming OVs with genes encoding immune stimulatory proteins to amplify their effects. In this work, we have tested the efficacy of conditionally-replicating adenoviruses (CRAds) with and without selected immunostimulatory payloads, murine CD40L (mCD40L) or 4-1BBL (m4-1BBL), in an immune competent mouse model of melanoma. When CRAd657-m4-1BBL and CRAd657-mCD40L were injected into B16-hCAR murine melanoma tumors, both single vectors delayed tumor growth and prolong survival compared to empty CRAd657. However, combined injection of both CRAd-m4-1BBL and CRAd-mCD40L mediated significantly better control of tumor growth. All of the payloads increased immune cell infiltration into tumors and notably reduced expression of PD-1 exhaustion marker on T cells. Tumor volumes were negatively associated with total infiltrating T cell population. We found that the payloads increased immune cell infiltration into tumors with some specificities: recruitment of CD8 + T cells was higher with m4-1BBL expression, while mCD40L expression induced more CD4 + T cell infiltration. Importantly, the combination of CRAd657-m4-1BBL and CRAd657-mCD40L induced the highest immune cells/T cell infiltration and the highest anti-TRP-2 tumor-associated antigen T cell responses than empty or single gene vector. This combination also caused depigmentation in areas adjacent to the tumor sites in more animals. These data indicate that driving two axes of the immune system with combined immune stimulatory payloads can lead to improved anticancer immune responses and better tumor control in an immune competent model of cancer.