Var gene diversity and their serological recognition by naturally exposed individuals

Plasmodium falciparum causes the worst form of human malaria and leads to 1-2
million deaths annually, most of them children below the age of 5 living in subsaharan
Africa. Morbidity varies from asymptomatic infections with no symptoms to
severe malaria accompanied by organ failure, severe anemia and coma. Most of
these clinical presentations are associated with sequestration of infected red blood
cells (iRBC) on host endothelium. By attaching the parasitized erythrocyte to host
receptors such as CD36, ICAM or CSA the parasite prevents the cell from being
cleared by the spleen and therefore prolongs its own survival.
A key protein involved in this process is the variant surface antigen Plasmodium
falciparum erythrocyte membrane protein 1 (PfEMP1) which is a parasite derived
protein transported to the RBC surface to mediate cytoadherence. With this process
exposes the parasite itself to the host immune system leading to the production of
specific antibodies. In order to evade this host immune response the parasite
undergoes antigenic variation by switching to another member of the same protein
family. PfEMP1 is encoded by approximately 60 var genes per haploid genome and
is expressed at the surface in a mutually exclusive manner, i.e. only 1 of the 60
proteins is expressed and exposed at any one time whilst the others remain silenced.
Protection against severe malaria is thought to be mediated to a large degree by the
piecemeal acquisition of anti-PfEMP1 antibodies during early childhood, since adults
still get infected but rarely develop severe malaria symptoms.
Recent observations suggest that not all PfEMP1 proteins expressed by a parasite
are equally virulent, but only a subset of distinct var genes might render a parasite
more pathogenic than parasites expressing different var gene variants. To generate
potential anti-severe disease interventions members of this particular subset need to
be identified. To date, only 6 studies have been published investigating the repertoire
of expressed var genes in vivo. We have further used samples collected in Papua
New Guinea from a case control study and analyzed var transcripts by RT-PCR
followed by cloning and sequencing. We determined the 3 main groups of 5’UTR and
analysed the data with respect to the clinical presentation of the children they were
collected from.
The detected number of different var group B and C transcipts was not significantly
different between asymptomatic, mild or severe malaria cases, whereas an increase of group A var genes was observed in symptomatic cases when compared to
children without any malaria symptoms. We identified an amino acid substitution
mainly occurring in asymptomatic children with high parasitemia that might influence
the binding affinity of parasites expressing these variants. However, using
phylogenetic analyses we were not able to identify other distinct var genes or subsets
associated with severe malaria. Blasting DBL1α domains against the 3D7 genome to
obtain information on the upstream region was found to be suitable for group A var
genes only, whereas 28% of group B and 62% of group C sequences were assigned
to the wrong subgroup using this method. Even though we observed a 7% sequence
overlap, bioinformatic analyses estimated the var gene repertoire in this region of
PNG to be unlimited.
It has previously been shown, that isolates causing severe disease are recognized
more frequently than those causing mild malaria. In the second part of this thesis, we
wanted to obtain information on the importance of distinct PfEMP1 domains in the
recognition by the host immune system. For that purpose, fragments of 2
representative var genes shown to be associated with severe malaria were
recombinantly expressed in E.coli and analyzed for their recognition by naturally
exposed sera of different origin. Analysis of synthetic peptides using the same sera
served to complement the results of ELISAs using recombinant proteins if expression
of distinct domains was not possible. ELISA and Western blot analysis determined
that 3 recombinant fragments and 2 synthetic peptides harbor epitopes that might
play a role in the generation of protective antibodies. However, since sample size
was small further investigations are required to confirm these findings.
In the third part of this thesis, we tested the usefulness of the GeneMapper® analysis
software to genotype var genes. It has been successfully established for genotyping
the polymorphic marker gene msp2 and since var genes also show some length
polymorphism it was investigated whether this technique could replace tedious
cloning and sequencing approaches, used so far to dissect var gene diversity.
Therefore, purified PCR products of UTR-DBL domains generated during the
sequence analysis were reamplified with fluorescently labeled DBL-specific primers
and analyzed by GeneMapper®. The results were then compared to the sequencing
data. GeneMapper® sizing was highly accurate with a mean deviation of 1bp and
showed a high consistency with sequencing data. Furthermore, GeneMapper®
detected 141 sequences which were not identified with the sequencing approach, whereas vice verca, this was only the case for 16 sequences. However, a significant
proportion of var genes could not be distinguished because the analyzed DBL
domains were identical in size. Despite this shortcoming, we belive that
GeneMapper® would greatly facilitate the analysis of expressed var genes and their
dynamics.