| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
FJ
EXPRESS SUMMARY ARTICLE The Full-length version of this article is also available, published online May 9, 2001 as doi:10.1096/fj.00-0645fje. |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,2,3,24
* Institut für Biotechnologie, Martin-Luther-Universität Halle-Wittenberg, 06120 Halle (Saale), Germany; and
ACGT ProGenomics AG, 06120 Halle (Saale), Germany
4Correspondence: Institut für Biotechnologie, Martin-Luther-Universität Halle-Wittenberg, Kurt-Mothes-Str. 3, 06120 Halle (Saale), Germany. E-mail: constanze.guenther{at}biochemtech.uni-halle.de
SPECIFIC AIMS
A system for the delivery of peptides and proteins into eukaryotic cells based on the directed encapsidation of polyproline-tagged compounds into modified polyomavirus-like particles was developed. The properties of capsomeres fused to a WW domain were investigated with respect to their physicochemical properties, ligand encapsidation, and uptake into mammalian cells.
PRINCIPAL FINDINGS
1. Polyomavirus-like particles can encapsidate protein ligands
The inner surface of capsids was modified in order to achieve a
specific interaction of the capsomeres with the ligands. During in
vitro assembly, this interaction results in encapsidation of ligands
like proteins and peptides into the forming polyomavirus-like
particles. The first WW domain of the mouse Formin binding protein 11
(FBP11) was fused to the NH2 terminus of VP1
(which faces the interior of the capsid) as a module to facilitate this
interaction. WW domains are small protein domains that were named after
two conserved tryptophan residues that are essential for maintenance of
the native fold and for ligand binding. The FBP11 WW domains
specifically bind proline-rich ligands that bear the consensus motif
PPLP.
Protein modeling studies of the VP1-WW fusion protein revealed there is sufficient space inside the capsid to allow proper folding and arrangement of the WW domains. Given that there are 360 ligand binding sites (72 pentameric VP1 molecules per capsid), it was calculated that 360 globular proteins with an average size of up to 17 kDa could theoretically be encapsidated.
2. VP1-WW capsomeres obtain a native fold and bind PPLP sequences
The fusion protein VP1-WW, expressed and purified from recombinant
Escherichia coli in a soluble form, was characterized with
respect to correct protein folding and ligand binding of the WW domain.
To confirm the native fold, far UV circular dichroism spectra of the
VP1-WW fusion protein were recorded and compared to the spectrum of the
wild-type protein. The VP1-WW spectrum had a significantly increased
negative ellipticity difference (
) below 207 nm, indicating an
additional portion of ß-sheet secondary structure as expected from
the addition of the three-stranded antiparallel ß-sheet of the WW
domain. The difference spectrum of VP1-WW minus VP1 represents the
spectrum of the single WW domain. The integrity and functionality of
the fused WW domain were further verified by surface plasmon resonance
using an immobilized PPLP peptide ligand. Determination of the kinetic
parameters yielded an equilibrium dissociation constant
Kd = 40 ± 5 nM. These results are in good
agreement with earlier surface plasmon resonance data using a fusion
protein of glutathione-S-transferase and the FBP11 WW domain. The
observed features (i.e., correct protein fold and specific binding of
PPLP ligands) are important prerequisites for a directed encapsidation
of PPLP-tagged peptides and recombinant proteins into the virus-like
particles.
3. VP1-WW capsomeres encapsidate PPLP-tagged peptides and proteins
during in vitro assembly
Either fluorescence-labeled PPLP peptides or PPLP-tagged green
fluorescent protein (GFP) were used as model ligands for optimization
of their encapsidation. Encapsidation rates were determined by size
exclusion chromatography, which separates virus-like particles from
free capsomeres and free ligands (Fig. 1
). The absorptions of the proteins at 280 nm and of the chromophores at
490 nm and 583 nm, respectively, were recorded simultaneously; after
integration of the respective peak areas, the in vitro assembly
efficiencies and the encapsidation rates were calculated. It was found
that the maximum encapsidation rate was 230 peptide molecules and 260
GFP molecules per virus-like particle.
|
4. Model peptides and proteins are efficiently delivered into NIH
3T3 cells
The ability of polyomavirus-like particles to transfer
encapsidated molecules into eukaryotic cells was examined in tissue
cultures in vitro. Cultures of NIH 3T3 mouse fibroblasts were incubated
with virus-like particles containing either fluorescence-labeled
peptides or PPLP-tagged GFP and subsequently analyzed using confocal
laser scanning microscopy. Significant fluorescence was detected within
all cells even after 30 min incubation time (Fig. 2
). The fluorescent molecules were localized near the cellular membrane
and eventually were distributed more throughout the entire cells with
longer incubation times, indicating that encapsidated ligands were
rapidly and efficiently delivered.
|
To determine the distribution of capsids and encapsidated ligands
individually, the capsid protein VP1 was labeled with a different
fluorescence dye at a specific cysteine residue in addition to the
fluorescent ligands. During uptake experiments into NIH 3T3 cells, the
capsid proteins and ligands could be clearly detected and
distinguished. Within the time frame of the experiments, most of the
ligands were still colocalized with the capsid proteins (Fig. 2)
.
However, there was also a portion of noncolocalized fluorescence
indicating that at least partial release of the ligands from the
capsids occurred within the period of the experiment. Although capsids
and ligands were also found in endocytotic vesicles, simultaneous
staining of lysosomes revealed only weak colocalization with the
peptides. Therefore, lysosomal degradation of the ligands seems
unlikely.
CONCLUSIONS
With an increasing understanding of the pathogenesis of many diseases at a molecular level, it is possible to use recombinant proteins for the functional substitution of defective or missing proteins or for the modulation of metabolic or signal transduction pathways. However, most proteins are unable to enter cells and reach intracellular targets, and are therefore limited to extracellular applications. Recently, fusions of target proteins with viral protein sequences of the HIV TAT and the Herpes simplex virus VP22 were found to mediate transfer of the target protein across the plasma membrane into cells.
In a novel approach, we used polyomavirus-like particles that are able to enter mammalian cells for the delivery of peptides and proteins. The icosahedral viral shell of murine polyomavirus is composed of 72 pentameric subunits of the major capsid protein VP1 and an inner layer of the minor capsid proteins VP2 and VP3. The outer capsid protein VP1 can be expressed and purified from recombinant E. coli and spontaneously forms virus-like particles in vitro in the presence of Ca2+ ions without the need for VP2 or VP3.
The major focus of the present study was the directed encapsidation of
model ligands into the particles. We focused on peptides and
recombinant proteins as prospective therapeutic molecules and
demonstrated that capsomeres fused to a WW domain can be used for an
efficient directed encapsidation of peptides and proteins tagged with a
proline-rich sequence (Fig. 3
). The observed maximum encapsidation rates of fluorescence-labeled PPLP
peptide (230) and GFP (260) were significantly higher than the
statistical inclusion of less than three molecules calculated for the
capsid and ligand concentrations used here, although the theoretical
limit of encapsidation was not reached, probably due to a fast
association/dissociation equilibrium between the proline-rich ligands
and the WW domain.
|
Cell culture experiments demonstrated an efficient uptake of the vector and the delivery of the encapsidated substances. There is approximately an equal distribution of the ligands within the cytoplasm and ligands remaining in endosomes. Since there is not much known about endosomal release of polyomavirus-like particles, these observations cannot be fully explained yet. For the native virus, myristylation of the minor coat protein VP2 may have a crucial function for endosomal release; empty wild-type polyomavirus-like particles consisting only of VP1 are mostly targeted to lysosomes.
The drug delivery system presented here should be applicable in general
for the delivery of a wide range of peptides, small PPLP-tagged
compounds, and proteins. Possible limitations for the encapsidation of
proteins could arise for large proteins, which may decrease the in
vitro assembly of the capsids due to steric hindrance. This effect
could possibly be avoided by reducing the number of binding sites per
particle using mixed capsids (Fig. 3)
.
Compared to fusion proteins with viral sequences, encapsidation into
virus-like particles has the advantage that the ligands are protected
inside a stable protein shell from external proteases. The strategy of
mixed capsid assembly may also be exploited to combine different
functions within the vector (Fig. 3)
. For example, capsomeres that
facilitate encapsidation of the ligands could be combined with
capsomeres modified on the outer surface so as to allow targeting of
specific cell types. It is also possible to fuse other WW domains to
the VP1 capsomeres that recognize different proline-rich sequences.
During in vitro assembly with varying VP1-WW fusion capsomeres,
different ligands could be encapsidated and administered simultaneously
with an exactly defined ratio. For example, emerging resistance of
tumor tissue to a single drug could be circumvented by applying a
mixture of diverse drugs.
In summary, it is envisaged that engineered virus-like particles as a basis of a modular delivery system may greatly enhance the use of therapeutic peptides and recombinant proteins. Such multifunctional drug delivery systems may be useful for treating human diseases. Future studies will focus on the delivery of biologically active compounds and their distribution and effects in tissue and animal models.
FOOTNOTES
1 To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.00-0645fje ; to cite this
article, use FASEB J. May 9, 2001) 10.1096/fj.00-0645fje 
2 These authors contributed equally to this work.
3 Present address: Western Australian Institute for Medical Research (WAIMR), B Block, Hospital Avenue, Nedlands WA 6009, Australia.
This article has been cited by other articles:
|
|
 |
|
  A. Abbing, U. K. Blaschke, S. Grein, M. Kretschmar, C. M. B. Stark, M. J. W. Thies, J. Walter, M. Weigand, D. C. Woith, J. Hess, et al. Efficient Intracellular Delivery of a Protein and a Low Molecular Weight Substance via Recombinant Polyomavirus-like Particles J. Biol. Chem., June 25, 2004; 279(26): 27410 - 27421. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
