| 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 November 14, 2001 as doi:10.1096/fj.01-0537fje. |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Institute for Cancer Research and Treatment, University of Torino Medical School, 10060 Candiolo-Torino, Italy
3Correspondence: IRCC, Str. Prov. 142, 10060 Candiolo-Torino, Italy. E-mail: cboccaccio{at}ircc.unito.it
SPECIFIC AIM
We built a chimeric tyrosine kinase receptor made up of the conditional dimerization domain FKBP and the intracellular domain of the HGF receptor (FKBP-HGFR). We aimed to demonstrate that upon dimerization, FKBP-HGFR can mimic the biochemical and biological activities of the corresponding wild-type receptor in hepatocyte precursors. This is preliminary to application in ex vivo hepatocyte gene therapy, which can be exploited to promote liver regeneration and hepatocyte functions.
PRINCIPAL FINDINGS
1. FKBP-HGFR expression and activation in liver progenitor cells
The human HGF receptor (HGFR) was engineered in order to obtain a fusion protein (FKBP-HGFR), which includes, starting from the NH2 terminus: 1) a membrane-targeting myristoylation signal; 2) three copies of FKBP12 protein; and 3) the entire cytoplasmic domain of HGFR. A control construct was generated in which the HGFR cytoplasmic domain is affected by a single point mutation that causes substitution of a lysine (K1126 in wild-type HGFR), critical for ATP binding, into alanine. The resulting molecule (FKBP-HGFR-K-) is devoid of tyrosine kinase activity. It is known that the FKBP12 moiety binds with high affinity lipophilic molecules such as FK 506. Synthetic homodimers of FK 506 and related molecules (FK1012, AP1510, indicated as ‘dimerizers’) freely diffuse across plasma membranes and couple in trans adjacent FKBP12 domains, resulting in dimerization and signal transduction activation.
We expressed FKBP-HGFR and FKBP-HGFR-K- into MLP29 (mouse liver progenitor 29) cells, which express HGFR and respond to HGF with cell proliferation and differentiation. The latter includes cell spreading and migration (‘scatter’), protection from apoptosis, acquisition of cell polarity, and tubule formation (branching morphogenesis). Administration of dimerizer to intact cells induced tyrosine phosphorylation of FKBP-HGFR (but not of FKBP-HGFR-K-) in a dose-dependent manner, reaching a plateau at a concentration of 1 µM.
2. FKBP-HGFR activation induces cytoskeleton rearrangement, cell scattering, branching morphogenesis, and protection from apoptosis but not cell proliferation
We thus compared the biological outcomes of FKBP-HGFR activation with those of wild-type HGFR in MLP-29. We observed that cells grown as tightly packed colonies on plastic culture dishes are equally stimulated to dissociate and to scatter by the dimerizer (1 µM) or by HGF (100 ng/ml). To evaluate the effect of the dimerizer on long-term differentiation, we grew cells in suspension in a collagen I extracellular matrix, where they form solid spheroids. Long-term culture in the presence of dimerizer (1 µM) as well as of HGF (100 ng/ml) indeed caused formation of ramified tubules that began to sprout a couple of days after stimulation and reached complex development within a week. During morphogenic movements, remodeling of interactions between cells and the extracellular matrix could induce a particular form of apoptosis called ‘anoikis’. However, by immunoenzymatic detection of free nucleosomes, we showed that dimerizer protects cells from apoptosis induced by staurosporine as efficiently as HGF.
When stimulated by dimerizer, we observed that FKBP-HGFR mediates differentiative activities corresponding to those of the wild-type HGFR when stimulated by its ligand. HGFR and FKBP-HGFR behaves differently in mediating cell proliferation, which by definition is an elective activity induced in hepatocytes by HGF. In fact, we observed that the dimerizer was not able to stimulate MLP29, expressing FKBP-HGFR, to incorporate 3H-thymidine and synthesize new DNA.
3. The FKBP-HGFR signal transduction profile is different from that of HGFR
Thus, we investigated the signaling mechanisms that could explain this difference in the biological outcomes of FKBP-HGFR and HGFR. Since FKBP-HGFR contains the entire cytoplasmic domain of HGFR, which as far as we know is solely responsible for recruitment of effector molecules that lead to cellular responses, we concentrated on quantitative aspects of signal transduction. We compared FKBP-HGFR and HGFR tyrosine phosphorylation kinetics in time course experiments. As shown by densitometric analysis of a representative immunoblot (Fig. 1
), HGFR phosphorylation peaks 15 min after stimulation with 100 ng/ml HGF, then progressively decreases toward basal level within 4 h. After decreasing, HGFR phosphorylation has a second, less intense peak 6 h after stimulation (Fig. 1A
). When we analyzed the FKBP-HGFR phosphorylation pattern in response to dimerizer, we found a striking difference. FKBP-HGFR tyrosine phosphorylation also reaches a maximum between 15 and 30 min, showing only a twofold increase (vs. the eightfold of HGFR) with respect to basal levels. Phosphorylation then decreases toward basal levels in 
4 h and peaks again 6 h after stimulation, similar to HGFR (Fig. 1A
). Thus, we conclude that upon dimerizer treatment, the relative increase of FKBP-HGFR phosphorylation is lower than that of HGF-stimulated HGFR. However, both receptors are able to sustain tyrosine phosphorylation over long periods (hours).
|
We thus hypothesized that differences in receptor tyrosine phosphorylation could result in quantitative differences in downstream signal transduction. It is known that among transducers, Ras/MAP Kinase (MAPK) and phosphatidylinositol-3 kinase (PI-3K) play a prominent role, both being required for proliferation on one hand and cell scatter, morphogenesis, and protection from apoptosis on the other. We studied the ability of FKBP-HGFR to activate ERK1 and 2 MAPK, and protein kinase B/AKT (or, more simply, AKT), which in turn is a target and a marker of PI-3K activity. We observed that ERK2 kinase activation by HGF peaks 15 min after stimulation (with a 25-fold increase with respect to basal levels), then decreases to a fivefold increase level within 2 h. This activation is sustained for at least 6 h. In the presence of dimerizer, ERK2 phosphorylation is as prolonged as in the presence of HGF, but the early activation peak shows only a fivefold increase in basal levels. However, at later times (2–6 h), the amount of activated ERK2 detected in the presence of dimerizer is comparable to that found in the presence of HGF (Fig. 1B
). We reported a similar activation pattern for ERK1. On the contrary, the kinetics of AKT activation by HGF or dimerizer are similar (Fig. 1C
). Thus, we conclude that FKBP-HGFR and HGFR differ in their ability to activate MAPK but not AKT. FKBP-HGFR fails to induce the early peak of MAPK activation, likely as a consequence of its relatively weak tyrosine phosphorylation. This offers a mechanistic explanation for the failure of FKBP-HGFR in inducing hepatocyte proliferation.
CONCLUSIONS
A promising strategy for the pharmacological control of signal transduction consists in fusing the signaling domain of membrane receptors such as tyrosine kinases to proteins that mediate conditional dimerization. FKBP dimerization is strictly dependent on the presence of lipophilic homodimeric molecules, namely, FK1012 and its derivatives, known as dimerizers. Dimerization in turn drives activation of domains fused to FKBP and signal transduction. Some growth factor receptors (e.g., PDGF receptor, erythropoietin receptor) have already been engineered by fusion of the intracellular domain with FKBP domains, and display a biochemical and biological behavior largely comparable to that of the corresponding wild-type receptors.
An exciting prospective would be to use these chimeras to engineer cells ex vivo or in vivo for therapeutical purposes. A general advantage is that, in vivo, dimerizers can selectively induce proliferation or other biological activities on engineered cells, circumventing the problem of the broad distribution of natural growth factor receptors. In the case of HGF, the clinical applications in gene and cell therapy are well recognized. HGF is considered the principal stimulator of liver regeneration, which can be therapeutic in a wide spectrum of diseases. Moreover, HGF receptor activation efficiently protects a variety of cell types from apoptosis, in vitro and in vivo, in the presence of toxic or apoptotic stimuli, including chemotherapeutic agents. On the other hand, the pharmacological use of HGF as such raises problems. First, HGF is a complex molecule that is difficult to handle in vitro and in vivo. It is a huge (92 kDa) disulfide-linked heterodimer, which is synthesized and secreted as a single polypeptide that must undergo proteolytic cleavage in order to become biologically active. Moreover, cell surface-associated proteoglycans bind HGF with high avidity, forming a diffuse and high-capacity reservoir that heavily affects HGF biodistribution and modulates its activity. Second, wild-type HGF has potentially adverse effects for its involvement in cell transformation and stimulation of cancer invasion and metastasis.
Here we show that in murine hepatocytes, the construct composed by FKBP and the intracellular domain of HGF receptor in the presence of the dimerizer protects hepatocytes from apoptosis and induces the full phenotype of hepatocyte differentiation. We also found that in contrast to the HGF receptor, FKBP-HGFR activation does not induce hepatocyte proliferation. We explain this functional divergence with the difference in MAP kinase activation. Both HGFR and FKBP-HGFR cause long-term activation of ERK1 and 2 MAP kinases, unlike pure mitogen such as EGF. Indeed, the ability to sustain MAP kinase activation has been correlated with the ability of growth factors to promote cell differentiation, as shown for NGF. However, the HGF receptor also induces an early peak of MAP kinase activity corresponding to the so-called ‘transient’ activation, which has been associated with induction of mitosis. This early peak is not induced by FKBP-HGFR, suggesting a causal correlation with its failure in stimulating cell proliferation (Fig. 2
). The absence of the early peak of MAP kinase activation is likely due to the different kinetic of activation of the FKBP-HGFR tyrosine kinase by dimerizer.
|
Instead of being a limit, the lack of mitogenic activity makes FKBP-HGFR potentially valuable for therapeutic application in conditions requiring induction and maintenance of cell differentiation combined with long-term survival. A good example is treatment of acute hepatic failure by cell transplantation or with extracorporeal devices (artificial liver). In both instances, it is mandatory to sustain specialized hepatocyte functions while avoiding the risk of cell expansion.
FOOTNOTES
1 To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.01-0537fje; to cite this article, use FASEB J. (November 14, 2001) 10.1096/fj.01-0537fje 
2 The first two authors equally contributed to this work.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
