Reverse Vaccinology and Structural Vaccinology

Reverse vaccinology and structural vaccinology are the two computational and biophysical methodologies that replaced Louis Pasteur's century-old empirical paradigm ("isolate, inactivate, inject") with rational, sequence-to-structure vaccine engineering. Rather than cultivating a pathogen and fractionating its most abundant biochemical components, Reverse Vaccinology 1.0 (pioneered by Rino Rappuoli and Craig Venter in 2000) starts in silico from the pathogen's sequenced genome, using bioinformatics algorithms to predict the complete surface-exposed proteome and identify conserved protective antigens that are scarce or unexpressed in in vitro culture.

Building upon genomic discovery, Reverse Vaccinology 2.0 mines the memory B-cell repertoire of convalescent human donors via single-cell sequencing to isolate potent neutralizing monoclonal antibodies and map the quaternary epitopes they recognize. Finally, Structural Vaccinology (pioneered by Barney Graham, Jason McLellan, and Peter Kwong) uses X-ray crystallography and cryo-electron microscopy (cryo-EM) to engineer disulfide bonds, cavity-filling hydrophobic residues, and proline substitutions (such as DS-Cav1 in RSV and the 2P/6P mutations in SARS-CoV-2 Spike) that lock metastable viral fusion proteins in their high-energy prefusion conformation, boosting neutralizing antibody titers by orders of magnitude.

The Neisseria meningitidis Serogroup B (MenB) Impasse and 4CMenB (Bexsero)

For decades, developing a vaccine against Neisseria meningitidis Serogroup B (MenB)—the leading cause of invasive meningococcal meningitis and fulminant meningococcemia in Europe and the Americas—defied conventional vaccinology for two biological reasons:

1. Autoimmune Molecular Mimicry of the MenB Capsule: Whereas meningococcal serogroups A, C, W, and Y were conquered by polysaccharide–protein conjugate vaccines, the MenB capsular polysaccharide is a homopolymer of α(2→8)-linked N-acetylneuraminic acid (polysialic acid). This carbohydrate structure is chemically identical to the polysialic acid post-translational modification expressed on human fetal and adult Neural Cell Adhesion Molecule (NCAM). Consequently, the human immune system is immunologically tolerant to the MenB capsule, and forcing an anti-capsular response risks autoimmune neuropathology.
2. Hypervariability of Major Outer-Membrane Proteins: Outer-membrane vesicle (OMV) vaccines relies on the immunodominant porin PorA, which is hypervariable across circulating MenB strains and confers only strain-specific protection.

In 1998–2000, Rino Rappuoli (at Chiron/Novartis) partnered with Craig Venter (at The Institute for Genomic Research) to pioneer Reverse Vaccinology:

1. Whole-Genome Sequencing and In Silico Filtering: Starting from the 2,272,351-base-pair genome of virulent MenB strain MC58 (containing 2,158 predicted open reading frames), bioinformaticians ran secretory signal peptide (SignalP), lipoprotein lipobox, transmembrane α-helix, and outer-membrane β-barrel algorithms (PSORT) to identify 570 putative surface-exposed or secreted proteins.
2. High-Throughput Cloning and Serum Bactericidal Screening: Of the 570 candidates, 350 ORFs were successfully cloned and expressed in Escherichia coli as His-tagged or GST-fusion proteins, purified, and used to immunize mice. Immune sera were screened via flow cytometry for binding to the surface of intact encapsulated MenB bacteria and tested in the Serum Bactericidal Antibody (SBA) assay (measuring complement-mediated bacterial lysis, the gold-standard correlate of protection).
3. Formulation of 4CMenB (Bexsero): From 28 bactericidal antigens, Rappuoli's team selected three conserved, surface-exposed virulence factors previously unknown to microbiologists:

• Factor H Binding Protein (fHbp): A surface lipoprotein that binds human complement regulator Factor H to evade alternative complement lysis (also the sole antigen in Pfizer's bivalent Trumenba vaccine).
• Neisserial Adhesin A (NadA): A trimeric autotransporter adhesin that mediates epithelial invasion.
• Neisserial Heparin-Binding Antigen (NHBA): A lipoprotein that binds heparan sulfate proteoglycans and increases serum resistance.

Combining recombinant fHbp, NadA, and NHBA fusion proteins with New Zealand strain outer-membrane vesicles (providing PorA P1.4) yielded 4CMenB (Bexsero), licensed in 2013–2015 and achieving >75%–85% effectiveness against invasive MenB disease in infants.

Reverse Vaccinology 2.0: Human B-Cell Repertoire Mining

Whereas Reverse Vaccinology 1.0 screened linear bacterial ORFs in mice, Reverse Vaccinology 2.0 (advanced by Antonio Lanzavecchia, James Crowe, and Dennis Burton) uses the human immune system itself to identify complex quaternary viral epitopes formed only when multiple glycoprotein subunits assemble on an intact virion. Peripheral blood memory B cells or plasmablasts from convalescent or elite-neutralizer human donors are isolated by fluorescence-activated cell sorting (FACS) or high-throughput B-cell microculture; paired heavy-chain (V_H) and light-chain (V_L) immunoglobulin genes are sequenced and expressed as recombinant human monoclonal antibodies. For example, screening human memory B cells against human cytomegalovirus (HCMV) revealed that the most potent neutralizing antibodies (1,000-fold more potent than anti-gB antibodies) did not bind monomer gB, but targeted quaternary epitopes on the endothelial-entry gH/gL/UL128/UL130/UL131A pentameric complex; similarly, cloning the human monoclonal antibody D25 from a convalescent donor unlocked the prefusion structure of Respiratory Syncytial Virus (RSV) F glycoprotein.

Metastable Class I Viral Fusion Proteins and RSV DS-Cav1 Engineering

Enveloped viruses that enter cells via Class I viral fusion glycoproteins—including Respiratory Syncytial Virus (RSV F), paramyxoviruses, human metapneumovirus, influenza (HA), HIV-1 (Env gp120/gp41), and coronaviruses (Spike S)—face a biophysical constraint. On the surface of an infectious virion, the trimeric fusion protein exists in a compact, spring-loaded, high-energy metastable prefusion conformation. Upon receptor binding or endosomal acidification, the central α-helical coiled-coil undergoes an irreversible >100-Å jackknife refolding transition, thrusting the hydrophobic fusion peptide into the host cell membrane and collapsing into a thermostable, rod-like postfusion six-helix bundle (6HB).

For over 50 years, RSV subunit vaccine trials failed because purified RSV F spontaneously triggered in solution into the low-energy postfusion F state (which displays antigenic sites I, II, and IV, but loses the apex surfaces of the infectious virion). In 2013, Jason McLellan, Barney Graham, and Peter Kwong at the NIH Vaccine Research Center (VRC) used the human neutralizing antibody D25 to trap RSV F in its prefusion state and solve its atomic crystal structure. They discovered antigenic site Ø (zero) at the membrane-distal apex of the prefusion F trimer: site Ø is the target of >60% to 80% of the neutralizing activity in convalescent human sera, and it is completely obliterated when the protein refolds into postfusion F.

To lock RSV F permanently in the prefusion conformation without requiring bound D25 antibody, McLellan and Graham engineered the DS-Cav1 immunogen via two structure-guided modifications:

1. Disulfide Bond ("DS", S155C–S290C): Covalently links the F₁ fusion peptide region to the central stalk, preventing the refolding of the α2–α3–β3–β4 structural elements into a single continuous postfusion α-helix.
2. Hydrophobic Cavity-Filling Mutations ("Cav1", S190F and V207L): Replaces small serine and valine side chains inside a hollow internal cavity of the prefusion head with bulky phenylalanine and leucine aromatic/aliphatic side chains, increasing van der Waals packing interactions and raising the activation energy barrier against triggering.

In human clinical trials, a single dose of prefusion-stabilized RSV F boosted serum neutralizing titers 10- to 40-fold compared to postfusion vaccines that had failed in Phase III. This breakthrough directly enabled the 2023–2024 FDA approvals of the first RSV vaccines in history: Arexvy (GSK, RSVPreF3 + AS01_E, 82.6% efficacy against RSV lower respiratory tract disease in older adults), Abrysvo (Pfizer, bivalent RSV-A/B preF for older adults and maternal immunization at 32–36 weeks gestation to protect newborns), and mRESVIA (Moderna, mRNA-1345 encoding membrane-anchored prefusion F).

Coronavirus Spike "2P" (K986P/V987P) and "HexaPro (6P)" Stabilization

Following the 2012 MERS-CoV outbreak, Jesper Pallesen, Nianshuang Wang, Jason McLellan, and Barney Graham investigated how to stabilize the massive trimeric Coronavirus Spike (S) glycoprotein (~1,273 amino acids per monomer). In 2017, after solving the cryo-EM structure of the human betacoronavirus HKU1 and MERS-CoV Spike trimers, they identified the conformational hinge: during fusion triggering, the loop connecting the first heptad repeat (HR1) and the central helix (CH) straightens into a single continuous 100-Å central coiled-coil helix.

Because the amino acid proline contains a rigid pyrrolidine ring covalently bonded to its backbone amide nitrogen, it lacks a backbone amide proton for α-helical hydrogen bonding and introduces a steric kink that energetically forbids α-helix extension. Substituting two consecutive proline residues ("2P") at the apex of the turn between HR1 and the central helix (V1060P and L1061P in MERS-CoV; V986P and L987P in SARS-CoV-1) locked the Spike trimer in its prefusion conformation and increased recombinant protein expression yields >50-fold.

Because this HR1–CH hinge sequence is evolutionarily conserved across all betacoronaviruses, when Chinese researchers posted the genome sequence of SARS-CoV-2 (Wuhan-Hu-1) on January 10, 2020, Graham and McLellan immediately aligned the S2 subunit and substituted the homologous residues K986P and V987P (Spike-2P) within hours. This 2017 structural vaccinology blueprint formed the exact antigenic payload of BNT162b2 (Pfizer/BioNTech), mRNA-1273 (Moderna), Ad26.COV2.S (Janssen), and NVX-CoV2373 (Novavax). Subsequently, in mid-2020, McLellan's laboratory engineered HexaPro (6P) by adding four additional proline substitutions (F817P, A892P, A899P, and A942P) in flexible loops of the S2 subunit, increasing Spike expression yield another 10-fold and raising thermal denaturation stability by +5 °C.

Why did conventional conjugate vaccinology fail against Neisseria meningitidis Serogroup B?

Whereas the capsular polysaccharides of meningococcal serogroups A, C, W, and Y are foreign bacterial sugars that elicit strong bactericidal antibodies when conjugated to carrier proteins, the Serogroup B (MenB) capsule is composed of α(2→8)-linked polysialic acid. This exact carbohydrate polymer coats human fetal and adult Neural Cell Adhesion Molecule (NCAM) in the central nervous system. Because B and T cells reactive to self-polysialic acid are deleted or silenced by central and peripheral self-tolerance, MenB polysaccharide conjugates are non-immunogenic and carry a theoretical risk of inducing autoantibodies against human neural tissue.

Why does locking RSV F in its prefusion conformation (DS-Cav1) increase neutralizing titers up to 40-fold?

On the surface of an infectious RSV virion, the F glycoprotein exists in a metastable prefusion state displaying antigenic site Ø (zero) at its apex, which accounts for over 60%–80% of the neutralizing potency of human antibodies. When unstabilized F protein is purified or expressed without stabilizing mutations, it spontaneously refolds into the rod-like postfusion conformation, destroying site Ø and exposing non-neutralizing or weakly neutralizing epitopes (sites I, II, and IV). Introducing the S155C–S290C disulfide bond and S190F/V207L cavity-filling mutations locks F in the prefusion state so that B cells are primed against site Ø.

How do the K986P and V987P (2P) mutations stabilize the SARS-CoV-2 Spike glycoprotein?

During coronavirus membrane fusion, the S1 receptor-binding subunit sheds and the S2 subunit refolds by straightening the hinge loop between the first heptad repeat (HR1) and the central helix (CH) into one long, continuous α-helical coiled-coil. Substituting lysine-986 and valine-987 with two consecutive proline residues (K986P/V987P) at the exact apex of this hinge loop introduces steric rigidity from proline's cyclic pyrrolidine ring and eliminates backbone α-helical hydrogen-bond donors, creating a thermodynamic barrier that prevents the HR1–CH spring from extending into the postfusion state.

1. Pizza, M., Scarlato, V., Masignani, V., Giuliani, M. M., Aricò, B., Comanducci, M., Jennings, G. T., Baldi, L., Bartolini, E., Capecchi, B., Galeotti, C. L., Luzzi, E., Manetti, R., Marchetti, E., Mora, M., Nuti, S., Ratti, G., Santini, L., Savino, S., ... Rappuoli, R. (2000). Identification of vaccine candidates against serogroup B meningococcus by whole-genome sequencing. Science, 287(5459), 1816–1820. https://doi.org/10.1126/science.287.5459.1816
2. McLellan, J. S., Chen, M., Joyce, M. G., Sastry, M., Stewart-Jones, G. B., Yang, Y., Zhang, B., Chen, L., Srivatsan, S., Zheng, A., Zhou, T., Graepel, K. W., Kumar, A., Moin, S., Boyington, J. C., Chuang, G. Y., Soto, C., Baxa, U., Bakker, A. Q., ... Kwong, P. D. (2013). Structure-based design of a fusion glycoprotein vaccine for respiratory syncytial virus. Science, 342(6158), 592–598. https://doi.org/10.1126/science.1243283
3. Pallesen, J., Wang, N., Corbett, K. S., Wrapp, D., Kirchdoerfer, R. N., Turner, H. L., Cottrell, C. A., Becker, M. M., Wang, L., Shi, W., Kong, W. P., Andres, E. L., Kettenbach, A. N., Denison, M. R., Chappell, J. D., Graham, B. S., Ward, A. B., & McLellan, J. S. (2017). Immunogenicity and structures of a rationally designed prefusion MERS-CoV spike antigen. Proceedings of the National Academy of Sciences, 114(35), E7348–E7357. https://doi.org/10.1073/pnas.1707304114
4. Rappuoli, R., Bottomley, M. J., D'Oro, U., Finco, O., & De Gregorio, E. (2016). Reverse vaccinology 2.0: Human immunology instructs vaccine antigen design. Journal of Experimental Medicine, 213(4), 469–481. https://doi.org/10.1084/jem.20151960