THE GORDON WILSON LECTURE: RAPID COVID-19 VACCINE DEVELOPMENT AND THE FUTURE OF VACCINOLOGY

INTRODUCTION

We face an ongoing threat from emerging viral diseases and have learned over the last three years that traditional public health measures are effective but limited in scope. We have also witnessed and relearned the value of vaccines for reducing disease severity caused by viral infections. Vaccine development is needed for emerging viral diseases, unmet needs, and improvement of existing vaccines, not to mention malaria, tuberculosis, and a variety of other untreatable or virulent bacterial, fungal, and parasitic diseases.

NEW TECHNOLOGIES SPARK A NEW ERA OF VACCINOLOGY

Fortunately, we now have a large and growing number of new technologies that provide vaccine solutions to safely establish immunity prior to exposure to microbial pathogens. Many of these new tools have only been widely available for about 15 years, and their accessibility and sophistication are increasing exponentially. Solving atomic-level structures of proteins that are targets for protective immune responses can now be done in days rather than months or years because of advances in crystallography and electron microscopy. Understanding the precise antigenic structure and interface of antigen-antibody complexes has turned some vaccine development projects into an engineering exercise rather than the historical iterative, empirical process associated with most prior vaccines. The availability of protein structures has fostered a field of computational design and protein engineering that can preserve optimal antigenic sites on candidate vaccines whether displayed on self-assembling nanoparticles or anchored in lipid membranes. The large accumulation of structural data has made possible the use of artificial intelligence for predicting structures from amino acid sequences and designing proteins de novo that can interact with a particular surface.

Single-cell analysis using multi-parameter flow cytometry or other sorting techniques combined with sequencing of transcriptomes can provide exquisite details about the phenotypes of B or T lymphocytes responding to an antigenic stimulus and is the basis for rapid identification of human monoclonal antibodies and the analysis of immune repertoires. These types of data for difficult pathogens with high antigenic variability can identify rare cross-reactive antibodies and provide molecular targets for vaccine-induced B cell responses.

Rapid gene sequencing is fundamental to single-cell analysis and is also the basis for new surveillance approaches to identify pathogens in wastewater or in other environmental niches that might warn of a looming outbreak. Rapid pathogen identification is key to triggering the development of medical countermeasures. Once pathogen sequences are known, gene synthesis is readily available to create the plasmids needed for protein production to solve structures, develop immune assays, design probes for isolating B cells and monoclonal antibodies, and devise diagnostics. Importantly, chemical gene synthesis is also the basis upon which mRNA, recombinant vectors, and cell lines to produce recombinant proteins can be initiated for candidate vaccine development.

The discovery tools described above and advances in gene-editing have made it possible to rapidly develop animal models that are susceptible to the emerging pathogen, and other reagents needed to evaluate candidate interventions. The technologies can generally be organized under two major categories. Some improve precision of antigen design, antigen display, and immune analysis, and others improve the speed of development. Precision and speed are the properties needed for pandemic preparedness (1) and response (2).

THE NIAID VACCINE RESEARCH CENTER AT NIH

The Vaccine Research Center (VRC) in the National Institute of Allergy and Infectious Diseases (NIAID) on the campus of the National Institutes of Health (NIH) was established to develop an HIV vaccine. Announced on May 18, 1997, by President Clinton at the Morgan State University commencement, Building 40 on the NIH campus was completed in 2000 and the initial cohort of investigators was assembled. The VRC includes not only basic research, but also process development, pilot plant manufacturing, a clinical trials unit, and a Good Laboratory Practice (GLP) laboratory to analyze clinical samples. It is a multidisciplinary group with access to all the regulatory and technical support needed to develop and evaluate vaccines. Although an HIV vaccine is not yet available, the efforts to understand HIV pathogenesis and immunity, and determination to develop preventive countermeasures, have driven the evolution of technologies mentioned above that have been used to address unmet public health needs including the response to emerging pathogens.

Since its inception, VRC investigators have created candidate vaccines in response to outbreaks of SARS-CoV-1, H5N1 avian influenza, West Nile virus, H1N1 2009 pandemic influenza, Chikungunya, MERS-CoV, Ebola, Zika, and SARS-CoV-2. In addition to HIV, VRC has also advanced vaccine programs for malaria, tuberculosis, respiratory syncytial virus, Marburg, seasonal and pandemic influenza, other alphaviruses, paramyxoviruses, and nonenveloped viruses. Discovery and evaluation of human monoclonal antibodies for prevention or treatment of infectious diseases has been achieved for HIV, Ebola, SARS-CoV-2, malaria, and other pathogens. Nucleic acid vaccines have been central to many of the programs and beginning with SARS-CoV-1 were used for outbreak response. In 2003, it took 20 months from sequence selection to initiation of Phase I, 11 months in 2006, 4 months in 2009, and 100 days in 2016 when responding to Zika. In 2014, a chimpanzee-derived adenovirus vector (ChAd3) expressing the Ebola Zaire surface glycoprotein (GP) was already vialed and approved for Phase I testing when the West African Ebola outbreak occurred (3). Despite rapidly evaluating this candidate vaccine in Phase I and Phase II trials and initiating a Phase III study in Liberia within one year (4), the outbreak receded before an efficacy result was obtained. In Guinea, where the outbreak lingered, a VSV-Ebola Zaire vaccine candidate was evaluated by another group of investigators and achieved an efficacy result which led to the vaccine ERVEBO^® (5). In 2016, in response to the Zika outbreak in South and Central America and the Caribbean, a candidate DNA vaccine was rapidly deployed (6) and advanced to Phase IIb efficacy evaluation within 16 months, but again the outbreak waned before an efficacy result could be obtained (7). These experiences motivated a new way of thinking about future pandemics that would require more proactive advanced preparation and greater speed of response.

A PROTOTYPE PATHOGEN APPROACH TO PANDEMIC PREPAREDNESS

Observational studies cataloguing human virus infections in the twentieth century showed that while there was a linear increase in the number of new viruses, the number of virus families associated with human infection began to plateau (8). This suggests that a pandemic threat from new viral diseases is a finite and tractable problem particularly if generalizable solutions can be found for vaccines, antivirals, and diagnostics that are applicable within or across virus families. Prompted by repeated pandemic threats during the twenty-first century, we proposed a prototype pathogen approach to pandemic preparedness (1,9). There are viruses from 26 taxonomic families known to infect humans. There are licensed vaccines for some of these families, but in 2016 there were no licensed vaccines manufactured by a platform technology that could be rapid enough to effectively respond to a pandemic. Even during the 2009 influenza pandemic for a disease that has been vaccine preventable for 80 years, the new vaccines matched to the H1N1 strain became available after the peak of the outbreak had passed. Platform manufacturing for vaccines refers to a process that reliably yields products using the same upstream and downstream methods regardless of the antigen being expressed and that shortens development timelines. To act rapidly when threatened by emerging viral diseases, we need to accumulate the detailed knowledge of potential pathogens from each viral family using the technologies referenced above to understand protein structures, entry and replication mechanisms, tropism, and other aspects of pathogenesis.

PROOF-OF-CONCEPT FOR STRUCTURE-BASED VACCINE DESIGN

The feasibility of the prototype pathogen approach was informed in part by our work on respiratory syncytial virus (RSV), an orthopneumovirus in the family Pneumoviridae, and extended by studies on coronaviruses and paramyxoviruses. These enveloped viruses all share in common Class I fusion proteins on their surface that mediate membrane fusion and are required for virus entry. The RSV F (fusion) protein had been used in five previous Phase III efficacy trials, always with the same result, boosting neutralizing activity 2- to 4-fold and not meeting primary objectives for protecting from symptomatic RSV infection. Between 2008 and 2012, working in collaboration with Jason McLellan and Peter Kwong at the VRC, we were able to solve the atomic-level structure of RSV F in its prefusion conformation, revealing new sites of vulnerability to neutralizing antibodies that are not present on the rearranged postfusion form of F (10). Stabilizing F in its prefusion conformation using a C-terminal trimerization domain, foldon, and internal disulfide and cavity-filling mutations resulted in a protein that was much more immunogenic than prior RSV vaccines (11) inducing a 10- to 20-fold rise in serum neutralizing activity (12). These findings suggested that stabilizing Class I fusion proteins from other enveloped viruses in their prefusion conformation might also result in better vaccine antigens.

While solving the prefusion RSV F structure, MERS-CoV emerged in the Middle East. Since there were no available structures for a coronavirus spike and since Jason McLellan was beginning a new faculty position, we decided to work together and extend the RSV findings to coronaviruses, another family of enveloped viruses with Class I fusion proteins. The MERS-CoV and SARS-CoV-1 spike proteins are more than twice the size of RSV F, and the structures proved difficult to solve by X-ray crystallography. We therefore engaged Andrew Ward at the Scripps Institute to attempt cryo-electron microscopy (EM) which also proved difficult because the MERS-CoV and SARS-CoV-1 spike proteins were inherently unstable. After three years through a series of serendipitous events, we began evaluating endemic coronaviruses and shortly after sending the HKU1 coronavirus spike protein to the Ward lab, a cryo-EM structure was obtained (13). Next, a series of spike ectodomain constructs were produced to define stabilizing mutations. Ultimately, a two-proline substitution at the top of the central helix was found to stabilize spike proteins across multiple coronaviruses including MERS-CoV, SARS-CoV-1, veterinary coronaviruses, and α-coronaviruses. Prolines were predicted to break alpha-helical domains 60 years ago (14) and have been previously used as -stabilizing mutations for Class I fusion proteins from RSV (15) and HIV-1 (16). By stabilizing a spike in the prefusion conformation, the -2-proline substitution preserved neutralization-sensitive apical epitopes in S1. Importantly, it also significantly increased protein expression from transduced cells (17), which provided a potential dose-sparing advantage for gene-based vaccine delivery or an increase in protein production from cells in vitro.

ZIKA OUTBREAK AND COLLABORATION WITH MODERNA

During the 2016 Zika outbreak, while testing the DNA vaccine expressing a non-infectious subviral particle (6), the VRC evaluated other candidate Zika vaccines. One of those was an mRNA vaccine from Moderna expressing the same prME construct. It was more immunogenic at a much lower dose than the VRC DNA (18). Because of our interest in platform manufacturing technologies that could be applied to pandemic preparedness (2,19), we initiated a formal collaboration with Moderna in which the VRC would design antigens and perform the evaluation of candidate products and Moderna would provide the mRNA. The plan was to do a demonstration project focused on two large families of enveloped viruses that require Class I fusion proteins for entry. Nipah virus was the prototype for the Paramyxoviridae, and MERS-CoV-2 was the prototype for the Coronaviridae. The VRC designed and did the preclinical testing of candidate vaccines for both prototypes. The chimeric Nipah antigen used a stabilized prefusion F with three monomers of the attachment protein G connected through their N-termini to the C-termini of the foldon trimerization domain (20). The spike protein was used as the antigen for MERS-CoV and was stabilized with the 2-proline substitution at the top of the central helix of S2. By 2019, we knew that both the protein and mRNA versions of the candidate vaccines for each prototype virus protected against lethal challenge in appropriate animal models. NIAID and Moderna had reciprocal site visits in 2019 and agreed to initiate a clinical trial of the Nipah mRNA vaccine candidate in the first quarter of 2020 as a demonstration project.

COVID-19 OUTBREAK AND RESPONSE

On December 31, 2019, there were reports from Wuhan China about an outbreak of respiratory virus disease. On January 6, 2020, we received information that it was likely to be a beta-coronavirus. Because we already had agreements in place, but had not started manufacturing the Nipah mRNA, VRC and Moderna agreed to switch the demonstration project to the new coronavirus and delay the Nipah project. Because the 2-proline substitution had been used successfully to stabilize several beta-coronavirus spike proteins, we agreed to initiate manufacturing at Moderna without any additional experimentation using the antigen designed by the VRC as soon as the viral sequences were released from China. This occurred on the evening of January 10, and the VRC, in conjunction with the McLellan lab, designed and ordered the plasmids on January 11 that would be needed to construct pseudoviruses for neutralization assays and make protein for solving structures, develop antibody binding assays, and make probes for isolating B cells. The VRC also shared the proposed amino acid sequence for mRNA expression with Moderna to initiate manufacturing. The McLellan lab had a cryo-EM structure of the SARS-CoV-2 spike protein by the end of January (21). Moderna sent the initial research grade mRNA to VRC investigators in 24 days, and Dr. Kizzmekia Corbett, a senior research fellow in the Viral Pathogenesis Laboratory at the VRC, performed the preclinical animal studies and developed serological assays to show the product was immunogenic by Day 42. She worked with others to prepare the preclinical data for the investigational new drug application (IND) to the FDA needed to initiate the Phase I clinical trial on Day 65. The IND was held by the NIAID Division of Microbiology and Infectious Diseases (DMID), and the clinical trial was led by a site in Seattle that was part of a NIAID-sponsored network (22). While the Phase I clinical trial was progressing, VRC investigators continued preclinical evaluation of mRNA-1273 in mice (23) and nonhuman primates (24) that was needed to support the IND for advanced clinical evaluation of the candidate vaccine product. These studies included collaborators from the University of North Carolina (Ralph Baric's lab) and Vanderbilt University School of Medicine (Mark Denison's lab) who had been part of the academic consortium led by the VRC Viral Pathogenesis Laboratory to study the immunogenicity of coronavirus spike antigens over the preceding six years.

A new Coronavirus Prevention Network (CoVPN) was formed by combining the NIAID DMID Infectious Diseases Clinical Research Consortium (IDCRC) with the HIV Vaccine Trials Network (HVTN) sponsored by the NIAID Division of AIDS (DAIDS) and other clinical trials sites from the Department of Defense (DoD). This group in collaboration with contract research organizations (CROs) working with Moderna initiated Phase III testing of the Moderna mRNA-1273 on July 27, 2020, 199 days after sequences became available. More than 30,000 subjects were enrolled in the Moderna mRNA-1273 Phase III clinical trial, and the CoVPN conducted additional large Phase III trials for other candidate COVID-19 vaccines in 2020 and 2021. The initial report from the independent Data and Safety Monitoring Committee was released on November 15, 2020, and revealed that there was 94% efficacy against symptomatic COVID-19 leading to Emergency Use Authorization of the Moderna mRNA-1273 on December 18, 2020, 343 days from the time of sequence availability.

The SARS-CoV-2 spike protein with proline substitutions at positions 986 and 987 was eventually used in authorized COVID-19 vaccines delivered as mRNA (Moderna, Pfizer-BioNTech), viral vectors (Johnson & Johnson), and subunit proteins (Novavax, Medigen) and used in at least 150 countries. The Commonwealth Fund estimated that more than 3 million lives and $1 trillion were saved in the United States by vaccines during the first two years of availability (25). Globally, it was estimated based on excess mortality that more than 19 million deaths were averted in 2021 by COVID-19 vaccines, and there was a large disparity in lives saves based on country income (26). While access to life-saving vaccines was a problem in low- and middle-income countries, there was also a problem with vaccine hesitancy among people with access to the vaccine. Some explained the reason for their reluctance was that they thought a one-year timeline was too fast for vaccine development, and historically vaccine development timelines had been measured in decades. However, rapid COVID-19 vaccine development was really a consequence of successive advances in multiple scientific disciplines and technologies over 40 years driven largely by ongoing efforts to develop an AIDS vaccine (Figure 1).

Fig. 1.

Vaccine-related technology development. Rapid COVID-19 vaccine development was a consequence of many years of cumulative basic research advances driven largely by efforts to make an AIDS vaccine over the last 40 years. Abbreviations: RSV, respiratory syncytial virus; VAERD, Vaccine-associated enhanced respiratory disease.

THE FUTURE OF VACCINOLOGY AND PANDEMIC PREPAREDNESS

Existing technologies can provide the basic knowledge and reagents needed to develop candidate vaccines for viral threats from most viral families using a combination of gene-based, protein, and whole virus modalities. Structure-based antigen design, new advances in artificial intelligence to inform protein engineering, single-cell analysis for human monoclonal antibody discovery, B cell repertoire analysis, and mRNA delivery technology have made research and development of biologics more of an engineering challenge than the historical iterative empirical process. Products can be conceived, produced, evaluated, and manufactured with much more precision and speed than was possible even 10 years ago. The technological advances also make it more feasible to accomplish vaccine research and development in in low- and middle-income settings. Chemical synthesis of mRNA is ideal for a rapid design cycle process to identify biologics for medical countermeasures. It is also a small footprint, small batch approach that may eventually be cost-effective for solving small market, regional problems.

While many things are possible with current knowledge, some cross-cutting issues in immunology will require breakthroughs to achieve optimal vaccine approaches for disease prevention. These include avoiding immunodominant responses, meaning that it is sometimes subdominant epitopes that need to be targeted to achieve broad immunity against pathogens with high antigenic diversity and variation like HIV or hepatitis C. Avoiding immunodominant responses is also an issue in the presence of preexisting immunity for pathogens like influenza where repeated immunization with current vaccines fails to amplify the subdominant cross-protective responses and also fails to elicit sufficient responses to new antigenic targets. Understanding this well enough to design better vaccine solutions will require more comprehensive B cell repertoire analysis on a large scale combined with immunizations using antigens designed with atomic-level precision. Ultimately, we need to be able to identify and target specific B cell lineages that make antibodies recognizing defined epitopes at an optimal angle and having the right chemical and Fc properties. In addition, those B cells need to have the right memory phenotype to sustain immunity and ideally be located at the site of potential exposure to the pathogen of interest. Parallel work is needed to induce T cells with properties that achieve rapid clearance of pathogens while minimizing immunopathology.

Preparing for all future viral threats is a large, but finite, task if approached by finding generalizable solutions for vaccines, diagnostics, and antivirals for all viral families and genera known to infect humans and that have shared properties. Prioritization of tasks within functional groups is reasonable, but there are sufficient global resources so that with coordination and cooperation work on all relevant families should proceed in parallel. There are plausible scenarios for each viral family in which they could pose a threat to global public health. Based on the WHO landscape analysis of COVID-19 vaccines, there were more than 350 vaccine development programs globally and more than 150 of those reached clinical evaluation. This not only created a precedent and safety database for several new vaccine modalities but also demonstrated a global capacity for vaccine development that is sufficient to encompass the 26 viral families known to infect humans (Figure 2).

Fig. 2.

Prototype pathogen approach for pandemic preparedness. There are currently 26 virus families known to be associated with human infection. Here they are organized by entry mechanisms and class of fusion protein used by enveloped viruses. Arteriviridae are added here in parentheses because they are not known to infect humans, but they are in the Nidovirales order (like coronaviruses) and can cause a hemorrhagic fever syndrome in nonhuman primates. They have distinct characteristics compared to the other virus families listed and not much is known about basic aspects of virology and immunity. It is recommended that prototypic viruses selected from viral families and functional groups are studied in depth to prepare candidate surveillance tools, vaccines, monoclonal antibodies, antivirals, diagnostics, and reagents prior to any future threats. The basic research findings would inform approaches to other family members and guide generalizable solutions that may be applicable across virus families. Centers with the capabilities listed in the middle should be established to support process development, pilot scale manufacturing, and preclinical and clinical evaluation of candidate countermeasures identified by pathogen-directed research.

LESSONS LEARNED AND OTHER CONSIDERATIONS

As the world emerges from three years of pandemic disruption, it is a good time to take account of what has been learned and how it may influence policy and practice going forward. A key lesson is the value of investment in basic research. Several lines of scientific investigation converged to make rapid COVID-19 vaccine development feasible. The pandemic revealed deficiencies in the public health infrastructure and coordination needed for an efficient emergency response. There is a need to develop a better system and strategy for communicating health messages to the public when the world is not in crisis mode. If there was a better understanding of the principles of respiratory virus transmission and immunity by the general public, messaging during the pandemic and the mitigation approaches recommended would have been more readily adopted. Some of this could be accomplished by emphasizing biology education in primary and secondary schools. There is a unique opportunity to engage children in learning about biology because of the way they have been impacted by the pandemic. Better biology understanding in the general population may also help counter the spread of vaccine misinformation that became so prevalent during the pandemic. There is an urgent need to build faith in biomedical science and restore trust and respect for health care providers and public health officials. Many lives were lost because of the failure to take advantage of the immunity afforded by vaccination. Deficiencies in global health equity became more apparent during the pandemic, and the consequences of inequity negatively impacted both access to and uptake of vaccines. Improving research and manufacturing capacity in low- and middle-income countries and diversification of the scientific and public health workforce should be top priorities going forward.

DISCUSSION

Zeidel, Boston: Wonderful talk and such a great example of how the NIH with scale and ingenuity sets things up for a lot of other people. At Beth Israel, Dan Barouch and his group were involved with the adenoviral vaccines. What is so striking is the impact of the AIDS pandemic that led to many collaborations. There were some sharp elbows at times in the past, but with the COVID pandemic, I really saw everyone collaborating more closely. Do you think that we should continue to develop adenoviral vaccine vectors as well as mRNA vectors, or do you think that we should be focused more exclusively on the mRNA ones going forward because the mRNA vaccines seem to have worked quite well? Although I will say the adenoviral vaccines seem to have done pretty well, too.

Graham, Atlanta: Right. What I’d like to say is that vaccines need to be fit for purpose and we can't just rely on one modality. We're still planning to deploy a chimpanzee adenoviral vector for Ebola Sudan in central or western Uganda where there's an outbreak of Ebola right now. Adenoviral vectors are very potent and useful, but there may be reluctance to use them in general population programs. People are going to be reluctant to immunize all babies or all adults with an adenovirus vector because of rare thrombotic side effects that were noted during the COVID-19 pandemic. There is somewhat of a movement for generalizable vaccines that require a gene-based delivery to use mRNA because of the side effect profile. Such a simple vaccine, and the components, are gone within a few hours, and it's probably the easiest vaccine to make. So, I think for some purposes adenoviral vectors are still going to be useful, but maybe not for general population widespread use unless the molecular mechanisms underlying the rare episodes of thrombotic disease can be defined. I also meant to say that Dr. Corbett is now in Boston; she's an assistant professor at the Harvard T.H. Chan School of Public Health.

Simari, Kansas City: With the proline substitutions causing stabilization of the prefusion structure, what caused the increase in protein expression from those substitutions?

Graham, Atlanta: I don't think I can answer that question. If we understood that, we might be able to design better mRNA vaccines to increase protein expression. I think there's a lot of things we don't understand about both codon modifications and mRNA secondary structure and protein folding and how things can get out of the endoplasmic reticulum more efficiently. There's still a lot to learn about very basic aspects of cell biology that could improve what we're doing dramatically. I wish I had a better answer.

Hsu, San Francisco: Thank you for that great talk. How do you propose we deal with misinformation to enhance trust in vaccines?

Graham, Atlanta: How do we solve misinformation? Well, misinformation has been going on since the beginning of time. It's just more effective now because of social media and all the tools we've developed to spread it around. After print media was started several hundred years ago and it led to misinformation problems, some agreements about the ethics of publishing were established. Journalism, at least up until recently, had a strict code of ethics. Misinformation wasn't purposely leaked out and facts were really checked. We clearly need a different approach on how to manage information on social media. I believe that a lot of controversies and conspiracy theories and other things are coming from bad state actors (for example, Russia, North Korea, or Iran), and people don't even realize how they're being led by the nose into wrong ideas. I think the solution is up to the young people, people who are under 35 years of age, who really understand social media, who understand how to communicate in that way, like Dr. Corbett who's very effective on social media as opposed to me. This is the generation that needs to come up with some better solutions to manage information accuracy on social media.

Sharma, San Antonio: Thank you for that amazing and inspiring presentation. This is a great example of visionary science and anticipating problems. Could you briefly comment on use of technology for cancer vaccines?

Graham, Atlanta: BioNTech and Moderna were both founded based largely on making cancer vaccines. Because RNA is fast, personalized cancer vaccines became feasible. Moderna and BioNTech learned how to make small batches of GMP-quality RNA, and they would obtain patient tumor tissue, do whole genome sequencing, look for tumor-associated polymorphisms, create constructs expressing a string of tumor-exclusive epitopes from those changed residues in the cancer, encode them in mRNA, and then give it back to the oncology investigators within 40 days to immunize subjects against their malignancies. I think there's reason to hope that if we seriously tackle these 26 virus families known to infect humans and figure out how to make vaccines for these different virus families there will be findings applicable to cancer biology and a better understanding of how to wake up relevant T-cells. Hopefully, as we invest in more basic research, more cell biology, and more understanding of induction and regulation of immune responses, it will be applicable to cancer as well as infectious diseases.

Humphrey, New York: I want to also thank you for that very elegant and inspiring talk. I especially would like to underline the way in which you concluded the talk. Through your presentation of the incredible advances in science, you acknowledged that it takes a whole team to do the kind of work that you illustrated, and you emphasized the most junior members of your team. It causes me to think back to all the amazing scientific discoveries over time, such as Banting and Best who discovered insulin and won the Nobel Prize. That, of course, is a subject for another talk on another day. Because of the way in which you intentionally ended your talk, I wish 20,000 medical students were in the room to have heard that. You responded to an earlier question about the important role of social media. Yesterday, we had a really exciting talk about the pipeline for our physician-scientist training programs, and I’m wondering if you have any other specific steps and ideas for how we nurture and build that pipeline so that we will have the scientific workforce for the future to do the kind of work that you so elegantly shared with us this morning.

Graham, Atlanta: I went to Morehouse School of Medicine to figure out how to work on global health equity. I’m an immunologist and a virologist, and I haven't figured it out yet. I’m still trying to formulate ideas on how to achieve genuine equity, how to create trust, and how to nurture young people who are inspired to be involved in medicine, science, and public health. I think these are the big questions. It became so painful to me during the pandemic that there was help available in the form of COVID-19 immunity that people chose not to accept. The first steps toward solving this problem are better educating our children in biology and finding ways to regain the trust of patients through community engagement and better, consistent communication when not in crisis mode. We also have to find ways to share intellectual property and “know-how” around the world because it's in everybody's best interest—how to do that and how to achieve equity is still a work in progress. The Morehouse School of Medicine's mission statement includes the word “equity.” Its whole purpose for being a medical school is to create and promote more equity in all aspects of health care. A primary reason I went to Morehouse School of Medicine was to learn about how it is doing this and hopefully I can contribute.

Abbas, San Francisco: That was a real tour de force! I have an immunological question. What strategies are being used to enhance the memory response especially to mRNA vaccines?

Graham, Atlanta: I’m glad you asked that. You should probably be the one to answer it, but I think mRNA immunity or durability of immunity has been mischaracterized in the public and press because mRNA vaccines actually induce a very durable response if you measure memory B-cells instead of serum antibody. mRNA vaccination established a very long sustained B-cell memory response that can rapidly make antibodies during anamnestic responses to infection or another vaccination within just a few days. Since COVID-19 was found to be amenable to treatment with therapeutic antibodies, it is more likely that an anamnestic antibody response even when you can't measure antibody in serum ahead of time can be effective. Post-infection therapeutic antibodies have not worked for other viruses like influenza or RSV, suggesting that SARS-CoV-2 may have a relatively slower progression from upper to lower airways infection. This may be part of the reason why people who don't have antibody to the Omicron variant in their serum are still being protected from severe lower airway disease because there's enough antibody with enough cross-reactivity from the immediate memory B cell response to provide some lower airway immunity. I think the durability of immunity from mRNA vaccination has been mischaracterized and it's actually pretty good, especially if you measure the B-cell memory instead of serum antibody.

Arnaout, Boston: Thank you for a very nice presentation and work. It may change some minds if people know that this is 40 years of work rather than what the public thought was only one year. Do you have or do you know of data comparing the durability of the immune response to mRNA versus the protein stabilized protein?

Graham, Atlanta: A protein vaccine will also work. The problem with protein-based vaccines is that it takes around 16 months to make them instead of 40 days to make mRNA vaccines in a rapid response to a pandemic, so mRNA was a better choice for the pandemic. The better choice for boosting going forward may be protein. Once we've been primed with a gene-based approach to create a platform of CD8 T-cell memory and other things, then boosting with protein may be preferred, so I think protein would have worked. It may not have been quite as good because it doesn't induce that primary CD8-T cell response, but as a booster protein may end up being better.

Golden, Baltimore: Thank you again for a very inspirational talk. I love the slide about the 40 years because everything we've heard from the general community was that it happened too fast so actually that's not true. I wish I had your 40 years slide; I think we used 10 years in our community education, but I also wanted to acknowledge that last slide. Dr. Corbett actually did one of our largest community events at Hopkins. I think 1,800 people joined and that really helped to launch our community education initiatives. I wanted to comment about that because the fact that she was a Black basic scientist was very impactful, particularly in our Baltimore community, for helping people to really understand the vaccine biology and why it was important for them to be vaccinated. We need to think about the pathway to medicine and science and getting those who are underrepresented engaged. We actually have to not just tell them about science, but we have to actually invest in the educational system in the country because many of these children are living in cities where they're not even learning basic math and reading. I've been in Baltimore for 28 years and until we actually fix our overall school system, we're not going to be able to make progress. High school is too late. I was very fortunate that I grew up in an area where my parents were able to move to a better school district so I could go to a science and tech high school. Every family doesn't have that option. There's a much bigger issue that we need to think about than fortifying our educational system, particularly for children who are in lower-class communities. We need to expose them to physician-scientists. By the time they get that exposure, it's too late if they haven't had basic reading and math.

Graham, Atlanta: I totally agree. That's why I keep bringing up the seven- and eight-year-old children who lived through this pandemic. If they can just see that it's possible to have a career as a scientist, I think it makes a big difference.

Tomaselli, New York: Thank you for the absolutely amazing talk. I can't think of a better example for promotion of the infrastructure that supports the development of physician-scientists and scientific knowledge in the community. I wish there was some way to broadcast this widely, not just to our community but to the community in general and to our government agencies to make sure that the pipeline remains full. I did have a scientific question though. In our addressing of the COVID-19 pandemic, we were part of a group that was using convalescent plasma. I wanted to get your take on what the role of convalescent plasma might be in the hyper-acute phase of a pandemic like this, whether or not there are better ways that you might utilize plasma, or any downside of its use.

Graham, Atlanta: Yes, convalescent plasma was tested during the pandemic -response. The problem is that the concentration and potency of neutralizing activity was not great enough to have much of an effect. Without the highly selected potent monoclonal antibodies, you just couldn't get the neutralizing activity needed, but using convalescent plasma in the early stages of a pandemic should be safe. It may be very informative, and it does work against other viruses like Argentinian hemorrhagic fever, which is caused by Junin virus. You can use convalescent plasma and have a big impact. Professor Jean-Jacque Muyembe used convalescent plasma successfully in the Kikwit Ebola epidemic back in 1995. One of the survivors of that epidemic provided the B-cell that led to the antibody that now is useful for treating Ebola. Ebanga^® was shown in 2018 to effectively treat Ebola so I think using convalescent plasma is not a bad step to take early in a pandemic like this. But now we can rapidly discover human monoclonal antibodies, so I see its value as largely a test of concept, not necessarily a final solution.