Periodic Reporting for period 1 - MISTRO (Identification of the cellular sites of cytomegalovirus latency that contribute to the induction of inflationary CD8 T cell memory against the virus.)
Reporting period: 2018-04-01 to 2020-03-31
Cytomegalovirus (CMV), a β-herpes virus, is an opportunistic pathogen that chronically infects the majority of human population. The virus poses a significant health risk for neonates and immune-compromised individuals. Similar to other herpesviruses, CMV persists in host's cells in a state of latency, from which it reactivates in association with immune suppression. Molecular pathways regulating the state of latency remain ill defined, at least partially due to incomplete characterization of the non-hematopoietic cell types latently infected by CMV in vivo. A further rationale for better understanding of the non-hematopoietic cellular targets of CMV latency is the development of improved vaccination strategies. Accordingly, CMV and CMV-based vaccine vectors elicit robust and durable populations of inflationary memory CD8+ T cells, which confer strong protective immunity, but whose translational application suffers from the lack of insight about the identity of latently infected non-hematopoietic types that promote their generation. This project aimed to: (Work Package 1, WP1) identify and characterize the non-hematopoietic cell types that are latently infected with CMV in vivo; (Work Package 2, WP2) assess the role of such latently infected subsets in driving CMV-specific inflationary memory CD8+ T-cells. During my fellowship, I completed all main project tasks and reached milestones ascribed to each WP. In the frame of WP1, I identified PDGFR+ mesenchymal stromal cells as a novel site of CMV latency in lymphoid tissues. I further characterized the distribution of latent CMV genomes in this cell compartment with unprecedented depth, which revealed that Tcf21+ stromal cells of the splenic red pulp are the most abundant source of latent CMV in lymphoid tissues. In the frame of WP2, I developed a novel genetic tool to ablate viral antigen processing in latently infected cells to study their role in memory inflation. Using this model, I demonstrated that (1) memory inflation of CMV-specific CD8+ T cells is dependent on viral antigen processing by latently infected stromal cells; (2) Tcf21+ stromal cells of the splenic red pulp, despite showing the highest latent virus load, are dispensable for the generation of memory inflation.
During this fellowship, I characterized the sites of CMV latency in lymphoid tissues and assessed the role of identified latently infected subsets in driving the generation of CMV-specific inflationary memory CD8+ T-cells. Specifically, by combining qPCR-based viral genomic load quantification in sorted cell populations and virus reactivation assays, I showed that PDGFR+ mesenchymal stromal cells are a novel site of CMV latency in lymphoid tissues. In-depth phenotypic mapping of stromal cells in lymphoid organs revealed that latent CMV genomes are highly abundant in Tcf21+ stromal cells of the splenic red pulp, while being significantly less frequent in PDGFR+ mesenchymal stromal cells of the lymph nodes. Using recombinant virus-based and single-cell transcriptomic analyses, I further generated insights about latency in mesenchymal stromal cells at molecular level, which indicated that entry into latency is not associated with significant upregulation of virus immediate early genes, but rather follows a program of minimal transcriptional activation of the CMV genome. To study the role of newly identified latently infected stromal cell subsets in driving memory inflation, I devised a novel genetic tool allowing me to ablate viral antigen processing in specific cell types. Using this model, I demonstrated that memory inflation of CMV-specific CD8+ T cells relies on viral antigen processing by latently infected stromal cells. Finally, I showed that Tcf21+ stromal cells of the splenic red pulp, which show the highest latent virus load, are dispensable for the generation of CMV-specific inflationary CD8+ T cells, arguing that memory inflation depends on viral antigens provided by redundant stromal sources. Dissemination of the results occurred through (1) participation in national and international scientific conferences; (2) outreach activities (Falling Walls Lab Marie Sklodowska-Curie Actions Contest) as well as through publications, such as co-authorship on two peer-reviewed articles and first authorship on a manuscript reporting the main findings of this action. Regarding the latter, submission of this work is planned for autumn 2020, awaiting finalization of complementing experiments, which address the overall importance of PDGFR+ mesenchymal stromal cells in driving memory inflation. In summary, the MSCA fellowship allowed me to develop a unique portfolio of scientific and technological merits, which I regard crucial in my ability to initiate research that is distinctive and independent from my former supervisors.
The results of this action show progress beyond state-of-the-art in several areas:
(1) The data provide the first formal demonstration that cytomegaloviruses establish latency in PDGFR+ mesenchymal stromal cells, a cell type previously known to be permissive to lytic, but not latent CMV infection, thus arguing against the field’s paradigmatic view that CMV induces lytic and latent infections in distinct cell types or differentiation states.
(2) The data characterize the distribution of latent CMV genomes in stromal cell subsets of lymphoid tissues with unprecedented depth and describe Tcf21+ splenic red pulp stromal cells as the primary site of CMV latency in lymphoid organs.
The findings (1-2) extend our understanding of the mechanisms of viral latency and are anticipated to prompt in-depth mechanistic studies focusing on identification of candidate molecular pathways harnessing the ability of CMV to reactivate in immune suppressed patients.
(3) The data establish that viral antigens that sustain memory inflation must be processed by latently infected stromal cells, whereas antigen processing by professional antigen presenting cells of hematopoietic origin cells is not sufficient.
(4) The data rule out that memory inflation is driven by the stromal cell population with the highest viral genomic load in lymphoid tissues, the splenic red pulp cells, but rather is sustained through redundant contribution of various latently infected cell subsets.
The findings (3-4) may lay basis for future development of improved vaccination vectors that directly target stromal cells.
(1) The data provide the first formal demonstration that cytomegaloviruses establish latency in PDGFR+ mesenchymal stromal cells, a cell type previously known to be permissive to lytic, but not latent CMV infection, thus arguing against the field’s paradigmatic view that CMV induces lytic and latent infections in distinct cell types or differentiation states.
(2) The data characterize the distribution of latent CMV genomes in stromal cell subsets of lymphoid tissues with unprecedented depth and describe Tcf21+ splenic red pulp stromal cells as the primary site of CMV latency in lymphoid organs.
The findings (1-2) extend our understanding of the mechanisms of viral latency and are anticipated to prompt in-depth mechanistic studies focusing on identification of candidate molecular pathways harnessing the ability of CMV to reactivate in immune suppressed patients.
(3) The data establish that viral antigens that sustain memory inflation must be processed by latently infected stromal cells, whereas antigen processing by professional antigen presenting cells of hematopoietic origin cells is not sufficient.
(4) The data rule out that memory inflation is driven by the stromal cell population with the highest viral genomic load in lymphoid tissues, the splenic red pulp cells, but rather is sustained through redundant contribution of various latently infected cell subsets.
The findings (3-4) may lay basis for future development of improved vaccination vectors that directly target stromal cells.