Expanded Lymphoid Organ Chip model captures mRNA vaccination from injection to protective antibody production with unprecedented complexity, with unique potential for mRNA vaccine research
By Benjamin Boettner
(BOSTON) — The protective abilities of the Moderna and BioNTech mRNA vaccines developed against COVID-19 saved millions of lives during the pandemic and put this platform technology in the spotlight of vaccine research. By now, mRNA vaccines for other infectious diseases, such as influenza, RSV, and HIV are being tested in clinical trials and highly anticipated.
The breakneck speed with which the COVID-19 mRNA vaccines were developed during the pandemic required decades of prior research: effective antigens needed to be predicted, antigen-encoding mRNA molecules had to be stabilized and stripped of unwanted inflammatory potential, and they had to be made safely and effectively deliverable via specifically composed lipid nanoparticles (LNPs). But mRNA vaccine developers still face a number of challenges.
It is well-known that immunization with mRNA vaccines starts at the muscle injection site and that it requires antigen-uptake and presentation by local antigen-presenting cells (APCs) and the secretion of pro-inflammatory cytokine molecules by APCs as well as muscle cells. “Despite the powerful immune responses induced by COVID-19 mRNA vaccines, including the production of protective antibodies, there are critical gaps in our knowledge of how mRNA-LNP vaccines are sensed by cells at the muscle injection site, as well as the factors and pathways driving their immunogenicity and other physiological reactions in the body,” said Girija Goyal, Ph.D., the Director in Bioinspired Therapeutics and co-senior author of a new study with significant potential to impact mRNA vaccine research.
Currently approved vaccines induce transient limited antigen expression only as long as the injected mRNA remains. In contrast, self-amplifying mRNA vaccines (SAMs) allow their mRNA to be amplified inside host cells to further boost antigen production in the body. This could lead to more durable vaccine responses at significantly lower doses, but also to more inflammation, which could pose risks for recipients. How exactly the immune requirements and responses of non-amplifying mRNA vaccines (NAMs) and SAMs differ between each other is not well understood. Importantly, since animal models can’t accurately recapitulate human immune responses and current human in vitro systems fall short of mimicking the complexities of vaccination, more powerful approaches are necessary to answer these questions and accelerate the development of live-saving mRNA vaccines.
Now, a team of researchers at the Wyss Institute at Harvard University, GlaxoSmithKline (GSK)’s Rockville Center for Vaccine Research, and Duke University School of Medicine has advanced a human in vitro model that recapitulates both, the initial immunological responses at the injection site in human muscle tissue and the complex adaptive immune responses created in nearby lymph nodes, enabling more comprehensive pre-clinical human mRNA vaccine testing. Led by Goyal and Wyss Founding Director and Bioinspired Therapeutics Lead, Donald Ingber, M.D., Ph.D., the team emulated the muscle injection site by combining muscle cells with antigen-presenting immune cells (APCs) in a co-culture system to allow the initial vaccine uptake and antigen processing as well as proinflammatory responses that are required for subsequent immune responses. After vaccinating this “intra-muscular injection site module” (injection module) with differently composed mRNA vaccines, they re-isolated the vaccine-primed APCs and mixed them with B and T cells derived from the same donors – B cells produce protective antibodies in immunized individuals and are supported in this function by T cells. This mixture was then transferred into a microfluidic Lymphoid Organ Chip along with cell culture media from the injection module that contained immune-stimulating, pro-inflammatory cytokines produced by the muscle cells and APCs, which was flowed through the device. This enabled the formation of mRNA vaccine-primed lymph node-like tissue, so-called lymphoid follicles in which B cells actively produce protective antibodies.
Using this integrated system, the researchers recapitulated the complex process initiated by mRNA vaccines injected into muscle tissue, culminating in somatic hypermutation within the B cells and resulting production of protective antibodies against pathogens like SARS-CoV-2 and rabies virus. The study opens the door to investigating the efficacies of alternative LNP compositions and mRNA designs in inducing initial “prime” and memory “booster” immune responses, as well as inflammatory consequences.
This all-human alternative for evaluating vaccine-induced immunity has vast potential for guiding mRNA vaccine developments for multiple infectious diseases at the preclinical stage, and for limiting the need for animals such as non-human primates, which has been both, a bottleneck and ethical concern.
“This all-human alternative for evaluating vaccine-induced immunity has vast potential for guiding mRNA vaccine developments for multiple infectious diseases at the preclinical stage, and for limiting the need for animals such as non-human primates, which has been both, a bottleneck and ethical concern,” said Goyal.
The findings are published in iScience.
Driving advances with industry and foundation support
In 2022, with support from the DARPA and the Bill and Melinda Gates Foundation, Ingber and Goyal had published a human Lymphoid Follicle Chip (LF Chip) and successfully used it to recapitulate immune responses to vaccines against seasonal influenza that are observed in people, including the production of pro-inflammatory cytokine molecules and influenza antigen-specific protective antibodies. In this study, the Wyss group continued to collaborate with the Gates Foundation and worked with GSK’s Rockville Center for Vaccine Research to expand the pre-clinical utility of the LF Chip as a vaccine testing platform with a first focus on a clinical COVID-19 mRNA vaccine and mRNA vaccines against Rabies virus provided by GSK.
Upon their injection into muscle tissue, the primed APCs, as well as APC and muscle cell-produced cytokines then travel to nearby “draining lymph nodes” where immunity is built up. Thus far, this complex process had not been modeled in any human in vitro system. As an important advance, Goyal’s team engineered a muscle injection module and combined it with their LF Chip technology. To form vaccine-primed lymphoid follicular structures, they mixed APCs from the injection module with B and T cells that they obtained from blood samples from the same donors, transferred them to one of two parallel running channels of a small microfluidic chip that are separated by a porous membrane. Through the other channel they flowed cell culture medium from the injection module that contained cytokines produced by the co-cultured APCs and muscle cells which then could traverse the membrane and stimulate cells in the LF structures. “The LF Chip can be maintained for at least 28 days, which allowed us to investigate longer-term responses as compared to traditional in vitro models,” said first-author Yunhao Zhai, Ph.D., who worked as a Wyss Technology Development Fellow with Goyal and Ingber. “Our engineered platform comprehensively mimics all steps from mRNA vaccine uptake to the formation of protective immune responses much closer than other human in vitro systems.”
From design to validation and new insights
Using their system, the team compared the immune responses induced by NAM and SAM mRNA vaccines encoding a rabies antigen in their system and found an interesting difference: while for the NAM vaccine being taken up directly by APCs in the muscle injection module was sufficient, the SAM vaccine needed direct cell-cell contact between APCs and muscle cells to trigger maximum cytokine production and antigen-priming of the APCs. They found that the SAM vaccine potently induced the formation of LFs in the chips and resulted in more robust pro-inflammatory cytokine responses and antigen-specific antibody responses than the NAM vaccine. “This was interesting to see with our platform since, first of all, the rabies antigen was a “naïve” antigen, meaning that the cell donor had never been confronted by it and the system had to establish an immune response from scratch,” said Goyal.
Diving deeper into these differences, the researchers performed the first proteomic analysis by comparing SAM and NAM-induced protein expression in LFs. Surprisingly, they showed only a small overlap, but perhaps less surprisingly, SAMs had a significantly greater capacity to induce a protein expression pattern reflecting a strong viral defense and inflammation. Currently, multiple pharmaceutical and biotech companies are advancing SAM mRNA vaccine candidates for seasonal influenza, as well as highly lethal and emerging zoonotic threats like, for example, the Nipah virus, and the team thinks that such developments could benefit from their newly developed platform and analytical capabilities.
As opposed to the tested rabies antigen, SARS-CoV-2 antigens such as the one encoded by the Moderna bivalent COVID-19 mRNA vaccine tested in this study are not naïve anymore because large parts of the population have been vaccinated with it. However, they offered the team the opportunity to investigate memory responses, booster effects, and strain specificities in their system. Indeed, the Wyss team’s collaborators at the Center for Human Systems Immunology at Duke University School of Medicine confirmed that the Moderna vaccine induced three to ten times more antibodies against the original Wuhan strain compared to two Omicron variants of SARS-CoV-2. The Wyss team confirmed the neutralizing potential of the antibodies produced by the vaccinated chips using SARS-CoV-2 pseudoviruses. In addition, the Wyss team was able to detect somatic hypermutation (SHM) in immunity-developing B cells in the LF Chips in response to mRNA vaccines. SHM signals the chromosomal changes that enable antibody diversity and, eventually, the selection of B cells producing protective antibodies, and could not be observed before in blood-derived B cells that have been exposed to mRNA vaccines in vitro.
Our expansion and validation of this human LF Chip-based platform demonstrates its enormous potential for future mRNA vaccine developments, and how the Wyss advances solutions in collaboration with partners from industry and foundations. We are in a much better position now to dive deeper into the immune responses to mRNA vaccines against a variety of infectious diseases.
“Our expansion and validation of this human LF Chip-based platform demonstrates its enormous potential for future mRNA vaccine developments, and how the Wyss advances solutions in collaboration with partners from industry and foundations. We are in a much better position now to dive deeper into the immune responses to mRNA vaccines against a variety of infectious diseases and because these chips include cells from humans, they also can be used to examine mRNA effects on individuals with different age, sex, or ethnicity,” said Ingber, who is also the Judah Folkman Professor of Vascular Biology at Harvard Medical School and Boston Children’s Hospital, and the Hansjörg Wyss Professor of Biologically Inspired Engineering at Harvard John A. Paulson School of Engineering and Applied Sciences.
Other authors on the study are Josie McAuliffe, who led the GSK group, Steven Gygi and Ka Yang, who performed the proteomics analysis, Min Wen Ku, Aditya Patil, Pranav Prabhala, Liqun He, Yuncheng Man, Mizanur Rahman, Kavya Menon, Sven B. Spörri, Supriya Gharpure, Abdul Rahman Isaacs, Shay Ferdosi, Andrzej Pitek, Giulietta Maruggi, Sylvie Bertholet, Jessica Firestone, Kambiz Mousavi, Eric Miller, Kate Luisi, Caleb Hellman, LaTonya Williams, Georgia Tomaras, and Peng Yin. The study was funded by GlaxoSmithKline Biologicals SA, the Bill and Melinda Gates Foundation (awards #INV-002274 and INV-002164), an in-kind Women in Science Award from Adaptive Biotechnologies to Goyal, the Biomedical Advanced Research and Development Authority (contract #75A50121C00075), and the Biomedical Advanced Research and Development Authority-U.S. Food and Drug Administration (contract #75A50123D00004).