Combinatorial and computational design of bnAb mRNA vaccines for HIV

Many HIV vaccine candidates have failed clinical trials, as they were unable to elicit a potent and durable response to
HIV viral challenge. Broadly neutralizing antibodies (bnAbs) have been identified in a number of HIV+ individuals with
well-controlled viral levels, and these bnAbs target epitopes that contain residues that are relatively conserved across viral
strains. It is thought bnAbs may have efficacy against various strains of HIV pathogen. It is therefore widely believed that
systems which induce a potent immune response that includes the generation of broadly neutralising antibodies (bnAbs) in
humans could be effective HIV vaccines, and help to mitigate the wide genetic diversity in envelope proteins and
relatively high mutation rate of HIV.
However, developing a vaccine which can elicit the production of these bnAbs in vivo has proven to be extremely
challenging. This is likely due to the complex affinity maturation process that is required to produce bnAbs. Immunization
protocols typically administer a single dose of antigen (prime dose), which is sometimes followed by a “boost” dose
delivered several weeks later. In a traditional bolus immunization, the half-life of the antigen present in lymph nodes is
generally shorter than the time scale over which germinal centres start producing higher affinity IgG antibodies relative to
the initial IgM response (~18 hrs). In contrast, natural infections expose the immune system to escalating antigen and
inflammation over days to weeks, resulting in the formation of a germinal centre with dynamic antigen presentation. This
germinal centre niche also supports activation of antigen presenting cells, T follicular helper cells, and appropriate
cytokine signalling to generate bnAbs. It is likely that to develop effective bnAbs, sophisticated vaccination techniques
which can more closely mimic natural infections and natural bnAb formation may be required.
We believe that to develop a successful HIV vaccine, researchers must aim to engineer more sophisticated and biomimetic
vaccines. Bioengineered vaccines should therefore consider three key parameters in parallel; 1) delivery of an
appropriately selected antigen, with 2) favourable kinetics of antigen expression, and 3) control of the immune response in
the germinal centre. We believe lymph node targeted delivery of computationally designed mRNA antigens inside
immunostimulatory lipid nanoparticles (mRNA LNPs) administered with computationally optimized immunization
protocols will address these three aspects in a unique way. Additionally,Translate Bio will provide expertise in
manufacturing considerations for mRNA therapeutics. As modifications to mRNA structure may impact the mRNA
antigen translation, stability, and immunogenicity, the input of our translational partner (Translate Bio) will allow us to
develop vaccines with a potential avenue for commercial development. This R61/R33 proposal combines our expertise in
computational antigen design, HIV immunology, combinatorial chemistry, and the commercialisation of mRNA
therapeutics to develop a new class of HIV mRNA vaccine candidates.

Grant Number: 5R33AI161805-05
NIH Institute/Center: NIH
Principal Investigator: DANIEL ANDERSON