Single chain antibodies against the 37 kDa/67 kDa laminin receptor as tools for prion diseases therapy

Prions are unconventional pathogens that cause transmissible
spongiform encephalopathies (TSEs). According to the "protein only"
hypothesis, prions consist of an infectious protein that is capable
of converting a normal host protein termed PrPc into a protease
resistant form termed PrPSc. PrPSc is poorly degraded by the host
and accumulates in the CNS. Normal biological functions of PrPc and
mechanisms involved in neurodegeneration remain obscure. During the
past two decades, considerable efforts have been made to elucidate
prion diseases and in particular to identify PrP interactors for a
better understanding in prion biology. A major break-through was
the identification of the 37 kDa laminin receptor (LRP), which
represents the precursor of the human 67 kDa high-affinity laminin
receptor (LR), as the cell surface receptor for the cellular prion
protein. We investigated the role of LRP/LR in the propagation of
PrPSc in chronically infected cells by different approaches. Three
strategies resulted in downregulation or blocking of LRP and
prevented PrPSc accumulation in different scrapie infected neuronal
cell lines (i) transfection with an antisense LRP RNA expression
plasmid (ii) transfection with small interfering (siRNAs) specific
for the LRP mRNA and (iii) incubation with the polyclonal anti-LRP
antibody, W3. We observed that the treatment with W3 abolished
PrPSc deposition and reduced PrPc levels after one week of
incubation. PrPSc did not reappear in cells being cultured for 14
additional days without therapeutic antibody treatment. Taken
together, these results indicate that LRP is not only required for
PrPc metabolism under non-pathological conditions but also has a
pivotal role in prion propagation in a cell culture model. LRP/LR
appears then to be a promising potential target for the development
of therapeutics for the treatment of prion disease. Due to these
encouraging cell culture data, we decided to select single chain
antibodies (scFv) encompassing a suitable format for therapy. ScFvs
are composed of variable parts of heavy and light chains of an
immunoglobulin that are connected by a peptidic linker. The
antibodies were screened on recombinant GST::LRP employing a phage
display strategy. Two scFvs termed N3 and S18 were screened and
selected by ELISA. Both antibodies were further characterized by
western blotting and FACS analysis: both N3 and S18 specifically
recognized mouseLRP and humanLRP overexpressed in mammalian cells
under denaturating conditions (western blot) and under native
conditions at the cell surface (FACS). Epitope mapping revealed
that as expected both scFvs are directed against the extracellular
part of LRP: S18 and N3 recognized amino acid residues 225-233 and
273-278, respectively. The ability of N3 and S18 to interfere with
LRP/PrP interaction was tested by pull-down assays. In contrast to
the control scFv C9 directed against the pre-S1 coat-protein of
hepatitis B virus, both anti-LRP scFvs were able to block the
specific LRP/PrP binding. In order to investigate a potential
curing effect of scFv S18 in vivo, this scFv was tested in a
scrapie mouse model by passive immunization. The application of S18
by intra-peritoneal injection was able to reduce PrPSc deposition
in the spleen in comparison to mice injected with PBS or C9.
However the survival times of S18 treated animals was not
increased. Anti-LRP scFv S18 seems to contribute to block prion
propagation in the periphery but it is likely that this effect was
not enough strong to have an impact on the CNS invasion. Thus, we
hypothesized that a strategy targeting directly the brain should be
more effective. In this context, an approach based on the
expression of single chain antibodies as secretory molecules in the
brain via an adeno-associated virus (AAV) vector was initiated. To
assure secretion of the scFv expressed in mammalian cells, a signal
sequence was fused to the scFvs. Tranfection experiments
demonstrated that neuronal cells were able to express and secrete
high quantities of both scFvs. Furthermore, the generated scFvs
were still functional as shown by western blotting. To find the
appropriate AAV serotype for scFv expression, neuronal cells were
transduced with varying serotypes carrying a GFP. AAV serotype 2
was chosen due to (i) its good transduction performance in two
neuronal cell lines and (ii) the possibility of its purification by
affinity chromatography. The sequences encoding for the scFvs N3,
S18 and C9 have been cloned in an AAV-based vector. The AAV system
was also able to drive high expression of scFvs into the
supernatant by transfection or transduction. rAAV-scFv particles
were produced and purifed for further stereotaxic injections into
mice. Although the investigation of this therapeutic strategy is
still in progress in a murine scrapie model, we already proved that
a single injection of rAAV led to the expression of scFvs into the
brain of mice 30 days post injection. This study represents the
first gene therapeutic approach for the treatment of prion
diseases.