In recent years, we have focused on using a PIV3-based vector to express RSV antigen, providing a bivalent vaccine against the two most important pediatric viral respiratory pathogens. The vector is one that we previously developed, called rB/HPIV3, which consists of bovine PIV3 in which the F and HN genes have been replaced by those of HPIV3. This results in a chimeric virus that is attenuated in non-human primates and humans due to the BPIV3 backbone, and which bears the neutralization and major protective F and HN antigens of HPIV3. Both the empty B/HPIV3 vector and B/HPIV3 expressing the unmodified RSV F protein were previously shown to be well-tolerated in infants and young children. Therefore, this vector appears to be in the desired range of attenuation. In work continuing from previous years, we continued to focus on expressing the RSV fusion F glycoprotein because it generally is considered to be the most important RSV neutralization and protective antigen. RSV F also is much more highly conserved among RSV strains than the attachment G protein, which is the other neutralization antigen and the second most important protective antigen. We continued to evaluate a number of strategies to optimize the immunogenicity of rB/HPIV3 expressing RSV F protein. This involved increasing both the quantity and the quality of the expressed RSV F antigen. Evaluation of several different positions for the RSV F gene in the vector genome identified the second gene position as generally being optimal. Evaluation of several versions of codon-optimization of the RSV F ORF identified the most efficient one, provided by GenScript (GS). The RSV F protein was modified with two missense mutations (called HEK) to be identical to an early-passage isolate of this RSV strain, which reduced fusion and stabilized the trimer. Two additional modifications in particular substantially increased the immunogenicity of vector-expressed F protein: (i) one modification was to increase the stability of the pre-fusion conformation of the F protein - the conformation that is the most effective in inducing RSV-neutralizing antibodies - by introducing mutations that have been reported by colleagues in the NIH Vaccine Research Center and elsewhere. The most successful mutations involved addition of a disulfide bond (called the DS mutation) in combination with two cavity-filling missense mutations (called Cav1). (ii) The other modification was to engineer RSV F to be efficiently packaged in the B/HPIV3 vector particle. This was done by replacing the transmembrane and cytoplasmic tail (TMCT) domains of RSV F with those of BPIV3 F. Each of these two modifications, DS-Cav1 and TMCT, resulted in a substantial increase in the induction of serum RSV-neutralizing antibodies, and in particular antibodies that neutralized RSV efficiently in vitro without added complement and thus are highly effective in neutralization. In work continuing from previous years, we constructed and evaluated more than 35 versions of rB/HPIV3-RSV-F in pre-clinical studies. These constructs included a variety of missense mutations identified by workers in the field as well as with deletion of the F cleavage site, resulting in a single chain protein. These studies resulted in the identification of two lead versions. One is called rB/HPIV3-F2/HEK/GS-opt/DS-Cav1 and has the following characteristics: insertion of RSV F at the second gene position (F2), an early-passage amino acid sequence (HEK), GenScript optimization (GS-opt), and the DS-Cav1 pre-F stabilization. The second lead version, called rB/HPIV3-F2/HEK/GS-opt/DS-Cav1/B3TMCT, is identical except that it also contains the TMCT modification. These candidates presently are being manufactured into clinical trial maternal for pediatric clinical evaluation. We also used the B/HPIV3 vector to express the RSV attachment glycoprotein G, which is the second RSV neutralization antigen and a major protective antigen. G contains a conserved fractalkine-like CX3C motif and is expressed as membrane-anchored (mG) and secreted (sG) forms. The CX3C motif and sG are widely thought to interfere with host immune responses, and it has been suggested that elimination of these features would improve the immunogenicity of an RSV vaccine. We used the rB/HPIV3 vector to express wt RSV G and nine modified forms, including sG, mG, mutants with ablated CX3C motif, and mutants bearing the TMCT of BPIV3 HN to achieve enhanced packaging into vector virions. Using a hamster model, we evaluated the effects of these individual factors on immunogenicity and protective efficacy induced against both the vector and the RSV insert. Furthermore, since PIV3 and RSV have similar tropisms and patterns of replication in vivo, evaluation of the replication of the B/HPIV3 vector provided a means to detect possible changes in the pulmonary immune milieu (e.g., due to sG or the CX3C motif) that might affect RSV/PIV3 replication. In hamsters, all of the vector constructs replicated to similar titers in the upper and lower respiratory tracts, allowing direct comparison of immune responses. Ablation of sG did not affect the RSV- or PIV3-neutralizing antibody (NAb) responses. Increased packaging did not affect the immunogenicity of RSV G, in contrast to previous findings with RSV F. Mutation of the CX3C motif drastically reduced the G-specific serum NAb response and protection against RSV challenge, indicating the importance of the integrity of the CX3C motif for the immunogenicity of RSV G. In human airway epithelium (HAE) cultures, post-immunization sera specific to wt RSV G, but not sera specific to wt RSV F, completely blocked RSV infection in the absence of added complement. The induction of antibodies with this blocking activity was reduced if the CX3C motif in the G immunogen was ablated. This suggests that NAbs induced by wt RSV G conferred more complete protection of the epithelium than NAbs induced by wt RSV F. In addition, vector expressing wt G was more protective in hamsters than that expressing wt F against RSV challenge. Increased expression of wt RSV G by codon-optimization increased immunogenicity and protective efficacy. This study showed that ablation of the CX3C motif in an RSV vaccine, as has been suggested by some, would be ill-advised, and ablation of sG would have no benefit. This study also suggests that RSV G is a more effective protective antigen than previously appreciated and could be an important component of an RSV vaccine. Previous studies comparing G and F employed purified proteins or vaccinia virus recombinants administered parenterally, and the comparison in the present study may be more relevant because it involved intranasal infection with a vector with a tropism and biology similar to those of RSV.
Showing the most recent 10 out of 17 publications
