Biomedical Engineering Reference
In-Depth Information
challenge infection in a nonhuman primate model [17, 18]. Although the specific
function of AMA1 is unknown and still under investigation, it is believed to have an
erythrocyte binding function based on different assays [14, 19, 20]. At this time all three
proteins (EBA-175 RII, AMA1, and Pfs25) have been cGMP manufactured, at pilot
scale, for human clinical trials (AMA-1 [21, 22]; Pfs25 [23]; EBA-175 RII-NG, Hall
et al., unpublished). The QbD approach for the molecular development and successful
expression in P. pastoris of each of these three recombinant proteins will be discussed.
3.2 THE MALARIA GENOME AND PROTEOME
The proteome of P. falciparum is estimated to encode more than 5000 proteins [24], of
which approximately 5% or 250 proteins may be of interest for vaccine development. An
analysis of the literature would suggest that approximately 100 malarial proteins have
been recombinantly expressed in some form and analyzed to a limited extent. The
number of malarial proteins under investigation for P. falciparum vaccine development
using the various vaccine delivery systems described above is 47 candidates, including
31 in preclinical development and 16 in clinical development [3]. Of the 16 in clinical
development, the number of recombinant-protein-based vaccines is about 10 due in part
to the difficulty in expressing malarial proteins with relevant structure. Part of the
difficulty in expressing malarial proteins in recombinant expression systems is due to the
unique content of the genome of P. falciparum. Analysis of the P. falciparum genome
shows that the A
T content is 76.2%, compared to 57.3% in P. pastoris and to 47.7% in
humans (Homo sapiens) (http://www.kausa.or.jp/codon/). The A
þ
T genomic bias of
P. falciparum negatively impacted expression of EBA-175 RII in P. pastoris using the
native gene. Efforts to produce EBA-175 RII in P. pastoris using the native A
þ
T-rich
gene failed apparently due to premature termination of mRNA transcripts as observed
by Northern blot [25]. Similar observations were reported for AMA1 FVO [5]
and the human immunodeficiency virus type 1 envelope glycoprotein, gp120 [26],
when expressed in P. pastoris using native genes. The native gp120 gene sequence was
identified to contain a Saccharomyces cerevisiae polyadenylation consensus sequence
that acted as a premature polyadenylation site, resulting in the production of a truncated
mRNA transcript [26]. In the case of EBA-175 RII, stretches of polyA's may be observed
within the native sequence such that these may act as premature polyadenylation sites
(Fig. 3.2).
þ
3.3 EXPRESSION OF TWO MALARIA ANTIGENS IN P. pastoris
In two independent preclinical development efforts, EBA-175 RII and AMA1 were
shown to be malaria vaccine candidates. Analyses of baculovirus-expressed proteins
demonstrated for each that antigen-specific antibodies inhibited parasite development
in vitro [16, 27], in vivo, or both [17]. Despite an interest in pursuing human clinical
studies, progression was hindered due to a bottleneck in identifying a manufacturing
process for the production of these two proteins in a scalable manner with suitable
