Differential endocytic routing of homo- and hetero-dimeric ErbB tyrosine kinases confers signaling superiority to receptor heterodimers
1998; Springer Nature; Volume: 17; Issue: 12 Linguagem: Inglês
10.1093/emboj/17.12.3385
ISSN1460-2075
Autores Tópico(s)Cell Adhesion Molecules Research
ResumoArticle15 June 1998free access Differential endocytic routing of homo- and hetero-dimeric ErbB tyrosine kinases confers signaling superiority to receptor heterodimers Anne E.G. Lenferink Anne E.G. Lenferink Department of Cell Biology, University of Nijmegen, Toernooiveld 1, 6525 ED Nijmegen, The Netherlands Search for more papers by this author Ronit Pinkas-Kramarski Ronit Pinkas-Kramarski Department of Biological Regulation, The Weizmann Institute of Science, Rehovot, 76100 Israel Search for more papers by this author Monique L.M. van de Poll Monique L.M. van de Poll Department of Cell Biology, University of Nijmegen, Toernooiveld 1, 6525 ED Nijmegen, The Netherlands Search for more papers by this author Marianne J.H. van Vugt Marianne J.H. van Vugt Department of Cell Biology, University of Nijmegen, Toernooiveld 1, 6525 ED Nijmegen, The Netherlands Search for more papers by this author Leah N. Klapper Leah N. Klapper Department of Immunology, The Weizmann Institute of Science, Rehovot, 76100 Israel Search for more papers by this author Eldad Tzahar Eldad Tzahar Department of Biological Regulation, The Weizmann Institute of Science, Rehovot, 76100 Israel Search for more papers by this author Hadassa Waterman Hadassa Waterman Department of Biological Regulation, The Weizmann Institute of Science, Rehovot, 76100 Israel Search for more papers by this author Michael Sela Michael Sela Department of Immunology, The Weizmann Institute of Science, Rehovot, 76100 Israel Search for more papers by this author Everardus J.J. van Zoelen Everardus J.J. van Zoelen Department of Cell Biology, University of Nijmegen, Toernooiveld 1, 6525 ED Nijmegen, The Netherlands Search for more papers by this author Yosef Yarden Corresponding Author Yosef Yarden Department of Biological Regulation, The Weizmann Institute of Science, Rehovot, 76100 Israel Search for more papers by this author Anne E.G. Lenferink Anne E.G. Lenferink Department of Cell Biology, University of Nijmegen, Toernooiveld 1, 6525 ED Nijmegen, The Netherlands Search for more papers by this author Ronit Pinkas-Kramarski Ronit Pinkas-Kramarski Department of Biological Regulation, The Weizmann Institute of Science, Rehovot, 76100 Israel Search for more papers by this author Monique L.M. van de Poll Monique L.M. van de Poll Department of Cell Biology, University of Nijmegen, Toernooiveld 1, 6525 ED Nijmegen, The Netherlands Search for more papers by this author Marianne J.H. van Vugt Marianne J.H. van Vugt Department of Cell Biology, University of Nijmegen, Toernooiveld 1, 6525 ED Nijmegen, The Netherlands Search for more papers by this author Leah N. Klapper Leah N. Klapper Department of Immunology, The Weizmann Institute of Science, Rehovot, 76100 Israel Search for more papers by this author Eldad Tzahar Eldad Tzahar Department of Biological Regulation, The Weizmann Institute of Science, Rehovot, 76100 Israel Search for more papers by this author Hadassa Waterman Hadassa Waterman Department of Biological Regulation, The Weizmann Institute of Science, Rehovot, 76100 Israel Search for more papers by this author Michael Sela Michael Sela Department of Immunology, The Weizmann Institute of Science, Rehovot, 76100 Israel Search for more papers by this author Everardus J.J. van Zoelen Everardus J.J. van Zoelen Department of Cell Biology, University of Nijmegen, Toernooiveld 1, 6525 ED Nijmegen, The Netherlands Search for more papers by this author Yosef Yarden Corresponding Author Yosef Yarden Department of Biological Regulation, The Weizmann Institute of Science, Rehovot, 76100 Israel Search for more papers by this author Author Information Anne E.G. Lenferink1, Ronit Pinkas-Kramarski2, Monique L.M. van de Poll1, Marianne J.H. van Vugt1, Leah N. Klapper3, Eldad Tzahar2, Hadassa Waterman2, Michael Sela3, Everardus J.J. van Zoelen1 and Yosef Yarden 2 1Department of Cell Biology, University of Nijmegen, Toernooiveld 1, 6525 ED Nijmegen, The Netherlands 2Department of Biological Regulation, The Weizmann Institute of Science, Rehovot, 76100 Israel 3Department of Immunology, The Weizmann Institute of Science, Rehovot, 76100 Israel *Corresponding author. E-mail: [email protected] The EMBO Journal (1998)17:3385-3397https://doi.org/10.1093/emboj/17.12.3385 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Both homo- and hetero-dimers of ErbB receptor tyrosine kinases mediate signaling by a large group of epidermal growth factor (EGF)-like ligands. However, some ligands are more potent than others, although they bind to the same direct receptor. In addition, signaling by receptor heterodimers is superior to homodimers. We addressed the mechanism underlying these two features of signal tuning by using three ligands: EGF; transforming growth factor α (TGFα); and their chimera, denoted E4T, which act on cells singly expressing ErbB-1 as a weak, a strong, and a very strong agonist, respectively. Co-expression of ErbB-2, a developmentally important co-receptor whose expression is frequently elevated in human cancers, specifically potentiated EGF signaling to the level achieved by TGFα, an effect that was partially mimicked by ErbB-3. Analysis of the mechanism underlying this trans-potentiation implied that EGF-driven homodimers of ErbB-1 are destined for intracellular degradation, whereas the corresponding heterodimers with ErbB-2 or with ErbB-3, dissociate in the early endosome. As a consequence, in the presence of either co-receptor, ErbB-1 is recycled to the cell surface and its signaling is enhanced. This latter route is followed by TGFα-driven homodimers of ErbB-1, and also by E4T-bound receptors, whose signaling is further enhanced by repeated cycles of binding and dissociation from the receptors. We conclude that alternative endocytic routes of homo- and hetero-dimeric receptor complexes may contribute to tuning and diversification of signal transduction. In addition, the ability of ErbB-2 to shunt ligand-activated receptors to recycling may explain, in part, its oncogenic potential. Introduction A large group of polypeptide growth factors mediates intercellular signaling by binding to, and activation of, transmembrane allosteric kinases with specificity to tyrosine residues (van der Geer et al., 1994). As in other allosteric systems, the monomeric form of the receptor tyrosine kinase (RTK) is inactive, but upon ligand-induced oligomerization (primarily dimerization) it initiates a plethora of intracellular events ranging from stimulation of ion fluxes to cytoskeletal alterations, and culminating in regulation of gene expression. The underlying biochemical mechanism involves autophosphorylation of specific tyrosine residues of the activated receptor. These are turned into docking sites for cytoplasmic signaling proteins containing Src-homology 2 (SH-2) domains (Koch et al., 1991), such as the adapter molecules SHC, Sem-5/Grb-2 and the p85 subunit of phosphatidylinositol 3′ kinase (Eagan and Weinberg, 1993). As a consequence thereof, several linear cascades of protein kinases are triggered, including the mitogen-activated protein kinase (MAPK) pathway (Seger and Krebs, 1995) and the S6-kinase pathway (Ming et al., 1994). In addition to this ‘vertical’ transduction pathway, lateral propagation of growth factor signals is made possible within subgroups of homologous RTKs by means of receptor heterodimerization. The best characterized example of ‘lateral’ signaling is provided by the type I RTKs (also named ErbB or HER family) (Carraway and Cantley, 1994; Alroy and Yarden, 1997). This subfamily comprises four members whose prototype is ErbB-1, a receptor that binds several ligands, including epidermal growth factor (EGF) and transforming growth factor (TGFα). Likewise, ErbB-3 and ErbB-4 bind three groups of alternatively spliced growth factors, collectively called neuregulins (Burden and Yarden, 1997). The fourth member, ErbB-2, binds no known ligand with high affinity. Nevertheless, impairment of ErbB-2 function by gene targeting resulted in a phenotype shared with that of neuregulin- and ErbB-4-deficient embryos (Lee et al., 1995), and a mutant form of this receptor promotes cancer in rodents (Bargmann et al., 1986). Overexpression of the wild-type human protein leads to phenotypic transformation of cultured cells (Di Fiore et al., 1987; Hudziak et al., 1987), and is frequently observed in several types of human carcinomas (Slamon et al., 1987, 1989). Moreover, ErbB-2 overexpression predicts poor prognosis and resistance to certain therapeutic modalities, implying that the orphan receptor contributes to tumor virulence (reviewed in Hynes and Stern, 1994; Stancovski et al., 1994). Despite the absence of a direct ligand, ErbB-2 plays a central role in a network of inter-receptor interactions; although the four ErbBs can form all 10 possible homo- and hetero-dimeric combinations, ErbB-2-containing heterodimers are preferred over other combinations (Tzahar et al., 1996; Graus-Porta et al., 1997). Each dimeric receptor complex has a distinct signaling potency, resulting in diversification and fine-tuning of signaling (Riese et al., 1995; Pinkas-Kramarski et al., 1996a). In general, signaling by homodimeric complexes is relatively weak, whereas heterodimers, and especially those containing ErbB-2, are more potent transmitters of signals. The collaborative action of two different ErbBs is best exemplified by the potent combination of ErbB-2, the ligandless receptor, with ErbB-3, whose kinase function is defective, and is reflected by the synergistic effect on cell transformation of certain co-expressed pairs of ErbBs (Kokai et al., 1989; Alimandi et al., 1995; Wallasch et al., 1995). In addition to the receptor level, combinatorial signaling by the ErbB network is further diversified at two additional levels. First, multiple EGF-like ligands exist and they differentially induce certain receptor combinations (Pinkas-Kramarski et al., 1996b), probably because each ligand carries not only a high affinity site, but also a ‘low affinity/broad specificity’ site that recruits the dimer's partner (Tzahar et al., 1997). Interestingly, some ligands induce more potent signals than others although they bind to the same receptor. For example, on certain cellular systems, such as keratinocytes (Barrandon and Green, 1987) and endothelial cells (Schreiber et al., 1986), TGFα is more potent than EGF, although both ligands bind to ErbB-1 with comparable affinity (Kramer et al., 1994). Another level of signal diversification is comprised of the multiple substrates of RTKs; members of this large group of SH-2 domain-containing proteins are differentially recruited to certain ErbBs. Examples include the phosphatidylinositol 3′-kinase and c-Cbl that preferentially engage with ErbB-3 (Soltoff et al., 1994) and with ErbB-1 (Levkowitz et al., 1996), respectively. Despite differences in second messenger activation, signaling by all ErbBs feeds into the MAPK pathway, raising the question of how signal specificity is maintained intracellularly. One potential answer is provided by results obtained with other growth factors in pheochromocytoma cells, indicating that the kinetics of MAPK activation, and especially its inactivation, may critically determine signal identity (reviewed in Marshall, 1995). Unlike the activation process which has been extensively studied, the inactivation phase of RTK signaling is poorly understood. One obvious candidate is the process that leads to endocytosis, down-regulation and degradation of ligand-activated receptors. Indeed, individual ErbB proteins differ remarkably in their rate of endocytosis and down-regulation (Baulida et al., 1996; Pinkas-Kramarski et al., 1996a). Our present study addressed the hypothesis that the multiple ligands of ErbBs differ in their potencies because they differentially recruit certain heterodimeric receptor combinations (Beerli and Hynes, 1996; Pinkas-Kramarski et al., 1996b; Gulliford et al., 1997). To this end we compared signaling by EGF and TGFα, a pair of ligands that display respectively weak and strong signaling in most tissues, in a well-defined cellular system expressing combinations of ErbB-1 with either ErbB-2 or ErbB-3. In contrast to our working hypothesis, differences in potency were observed even in the absence of either co-receptor, namely ErbB-2 or ErbB-3. However, to our surprise, the co-receptors potentiated the effect of EGF without significantly affecting TGFα signaling. In subsequent experiments we investigated the mechanism of potentiation and found that the co-receptors, by forming heterodimers with ErbB-1, redirected this receptor to an endocytic route that allows receptor recycling and, therefore, enhanced signaling. These results imply that EGF-like ligands whose ErbB specificity is shared are functionally distinct, and suggest that alternative endocytic routing may be critical for controlled inactivation and fine-tuning of signal transduction. Results ErbB-2 and ErbB-3 potentiate EGF mitogenicity but not TGFα signaling To examine possible functional relationships between the multiplicity of EGF-like ligands and the extensive inter-receptor interactions within the ErbB family of receptors we used the two best characterized ligands of the family, namely EGF and TGFα, in combination with a series of cell lines co-expressing ErbB-1 with either ErbB-2 (D12 cells), or with ErbB-3 (D13 cells) (Pinkas-Kramarski et al., 1996a). A third cell line that singly expresses ErbB-1 (D1 cells) was used for comparison of ErbB-1 homodimers with heterodimers of this receptor. In addition, a chimeric EGF/TGFα molecule, designated E4T, comprised of the A and B loops of EGF, and the C loop of TGFα, was used because of its superior mitogenic activity to that of other chimeric molecules and the parental ligands (Lenferink et al., 1997). Due to their dependence on interleukin-3 (IL-3), the cell lines we employed are extremely sensitive to EGF-like ligands when tested in the absence of IL-3. Thus, TGFα exerted mitogenic stimuli that were at least 10-fold more active than EGF-induced signals when tested on D1 cells (Figure 1A). However, E4T was even more potent in inducing cell proliferation. This pattern of relative potency was also reflected in long-term survival experiments in which IL-3 was replaced by the corresponding ErbB-1 ligand and cell survival monitored daily (Figure 1B). Introduction of ErbB-2 into D1 cells elevated the basal proliferation rate of the resulting cell line, D12, in agreement with previous reports (Kokai et al., 1989; Cohen et al., 1996; Tzahar et al., 1996; Zhang et al., 1996). Thus, whereas maximal stimulation of D1 cells by IL-3 was 5.5-fold, only a 2-fold activation was displayed by D12 cells. Interestingly, however, co-expression of ErbB-2 together with ErbB-1 (D12 cells) resulted in remarkable potentiation of the mitogenic action of EGF; whereas half maximal mitogenic effect was induced by 10 ng/ml of this ligand on D1 cells, only 0.7 ng/ml was necessary to stimulate the D12 cells (Figure 1A, compare D1 with D12 panels). In contrast, ErbB-2 co-expression only slightly improved the mitogenic action of TGFα and E4T. In fact, in the presence of ErbB-2, EGF almost approached the high mitogenic activity of TGFα, a phenomenon that was reflected, in part, also in a long-term survival assay (Figure 1B, D12 panel). Interestingly, ErbB-3 only partially potentiated EGF activity in D13 cells (compare the EC50 of EGF on D13 cells, which is 2 ng/ml, with that on D1 cells, which is 10 ng/ml). Once again, co-expression exerted no significant effect on the potency of either TGFα, or E4T (D13 panels in Figure 1). In conclusion, ErbB-2, and to some extent also ErbB-3, specifically enhance the EGF-induced mitogenic action of ErbB-1, probably by forming heterodimeric complexes with this receptor. Figure 1.Ligand-induced proliferation and survival of ErbB-expressing 32D-cells. (A) The following derivatives of 32D cells were examined for cell proliferation by using the MTT assay: D1 cells that singly express ErbB-1, D12 cells expressing a combination of ErbB-1 with ErbB-2, and D13 cells expressing a combination of ErbB-1 with ErbB-3. Cells were washed free of serum factors and IL-3, and seeded at a density of 5×105 cells/ml in RPMI-1640 medium containing serial dilutions of EGF (□), TGFα (▪), or E4T (●). Following 24 h of incubation, the MTT assay was performed as described in Materials and methods. (B) The indicated sublines of 32D-cells were plated as described above in the presence of 100 ng/ml EGF, TGFα or E4T [symbols are as in (A)]. Cell proliferation was measured daily using the MTT assay. As a negative control cells were plated in serum- and IL-3-free medium (○). The data from both experiments are given as the means of three determinations. Bars in (A) represent standard deviations. The experiments were repeated three times. The responses to IL-3 (fold induction) of D1, D12 and D13 were 5.54±0.63, 1.96±0.67 and 3.03±0.81, respectively. Download figure Download PowerPoint Binding parameters may explain superiority of E4T, but not the difference between EGF and TGFα Perhaps the simplest explanation for the observed differences in mitogenic potencies of EGF, TGFα and E4T might be parallel differences in receptor binding affinities. To examine this possibility we labeled the three ligands with 125I and determined their apparent binding affinities to D1, D12 and D13 cells using ligand displacement analysis. The results of this experiment are shown in Figure 2A. Evidently, the apparent affinities of EGF, TGFα and E4T were not remarkably different when tested on D1 cells, in agreement with a similar analysis that was performed with fibroblasts (Lenferink et al., 1997). Co-expression of ErbB-2 (or ErbB-3) only slightly improved the affinity of D12 cells (or D13 cells) to EGF or TGFα (Figure 2A, D12 and D13 panels). Notably, ligand binding assays performed with derivatives of 32D cells usually yield affinities that are consistently lower than those measured with adherent cell types such as fibroblasts or epithelial cells. For example, the Kd values of EGF and TGFα binding to adherent cells are in the range of 0.1–5 nM (Tzahar et al., 1994; Lenferink et al., 1997), whereas D1 cells bind these ligands with apparent Kd values of 30–50 nM. This may be due to the relatively prolonged washing procedure required in the case of the 32D myeloid cells, which results in an overall reduction in assay sensitivity. We used a ligand dissociation assay as an alternative to partly overcome this limitation. Cells were loaded with the various radiolabeled ligands under saturating conditions, then the unbound ligand was removed and the rates of release of radioactivity were monitored. Clearly, the rates of release of E4T from the surfaces of all three cell lines examined were higher than the dissociation rates of EGF and TGFα (Figure 2B). In addition, the co-expressed co-receptors, namely ErbB-2 and ErbB-3, comparably decelerated the rate of dissociation of EGF and TGFα from ErbB-1, in agreement with previous reports (Kokai et al., 1989; Karunagaran et al., 1996; Tzahar et al., 1996). Taken together, rapid dissociation from the cell surface may be involved in the mitogenic superiority of E4T over EGF and TGFα. However, neither the enhancement of EGF signaling by the co-receptor, nor the superiority of TGFα over EGF may be attributed to binding parameters. Figure 2.Ligand displacement and dissociation analyses. (A) Displacement analysis was performed with 1.0×106 cells of the indicated subclones of the 32D cell line. Cells were washed free of IL-3 and serum factors using binding buffer, and subsequently incubated for 2 h at 4°C with [125I]EGF (1 ng/ml) in the presence of serial dilutions of unlabeled EGF (□), TGFα (▪) or E4T (●). Unbound ligand was removed by sedimenting the cells through a cushion of calf serum. The results are presented as the mean ±SD of two determinations. Experiments were repeated three times with similar results. (B) The indicated cell lines were incubated for 2 h at 4°C with [125I]EGF (□), [125I]TGFα (▪), or [125I]E4T (●), each at 60 ng/ml. Then, the unbound ligand was replaced by an excess of the unlabeled growth factor (3 μg/ml), and cell-bound radioactivity was monitored at the indicated time intervals. Results are expressed as the fractional ligand binding (mean ±SD) relative to the amount of ligand that bound at t=0. The experiment was performed in duplicate and repeated twice with similar results. Download figure Download PowerPoint Co-receptors decelerate ligand depletion and internalization, but clearance of the E4T superagonist is defective Because E4T is released from the cell surface at a much faster rate then EGF or TGFα, we expected that these latter ligands would be depleted from the medium at a much faster rate than E4T. This possibility was tested by incubating D1, D12 and D13 cells with serial dilutions of the ligands for 24 h, thereby allowing their depletion from the medium. Then we determined the relative concentration of each ligand in the conditioned medium by employing a bioassay that uses serum-starved HER-14 fibroblasts overexpressing ErbB-1. As predicted, the rate of ligand depletion inversely correlated with mitogenic potency; the weakest and the strongest mitogens of D1 cells, namely EGF and E4T, respectively displayed rapid and slow depletion from the medium (Figure 3A). For example, when D1, D12 and D13 cells were incubated for 24 h with a low concentration of EGF (1 ng/ml) and the resulting conditioned media compared with medium similarly incubated in the absence of cells, we observed a 63, 28 and 47% reduction, respectively, in mitogenic activity. The corresponding numbers for TGFα were 28, 36 and 43%, and for E4T, 14, 16 and 24%. Thus, the presence of ErbB-2 significantly decelerated the rate of EGF depletion, but it less efficiently affected removal of E4T or TGFα from the medium. The relative rates of cell-mediated removal of the three ligands correlated with their mitogenic potency, implying that an endocytic mechanism is responsible for the observed differences in signaling potency. Consistent with this model, co-expression of the less potent co-receptor, ErbB-3, together with ErbB-1 only partly extended the half life of EGF (D13 panel in Figure 3A). Figure 3.Receptor-mediated depletion and uptake of ligands. (A) Increasing concentrations of the following ligands were incubated for 24 h at 37°C with the indicated derivatives of 32D cells (open symbols): EGF (diamonds), TGFα (squares) or E4T (circles). For control, ligands were similarly incubated in the absence of cells (closed symbols). The capacity of the resulting conditioned media to stimulate DNA synthesis in HER-14 fibroblasts was then determined as described in Materials and methods. Results are given as the mean ± SD of three individual experiments carried out in duplicate. (B) For determination of ligand internalization rates, radiolabeled forms of the indicated ligands (each at 1 ng/ml) were incubated for 2 h at 4°C with the following derivatives of 32D cells: D1 (▴), D12 (●) or D13 cells (▪). Following incubation on ice, cells were washed free of unbound ligand and incubated at 37°C for various time intervals with excess of the corresponding unlabeled ligand (at 3 μg/ml). Cellular uptake of radioactivity was monitored by removing surface-bound ligand with an acidic ligand-strip buffer. Data are presented as the mean ± SD of duplicate determinations. Each experiment was repeated at least twice. Download figure Download PowerPoint To test directly a model involving endocytosis, we comparatively analyzed the internalization rates of the various ligands of ErbB-1, and also determined their dependence on the presence of a co-receptor, either ErbB-2 or ErbB-3. It is notable that our previous experiments, which used a standard ligand internalization assay, detected only minor differences between the rates of ligand internalization through homo- and hetero-dimeric receptors (Pinkas-Kramarski et al., 1996a). Therefore, we tested several ligand internalization protocols for their ability to discriminate between the rates of endocytosis of homo- and hetero-dimeric receptors and selected the following assay. Cells were first incubated in the cold with a moderately low concentration of the respective radiolabeled ligand, then the unbound ligand was removed, cells chased at 37°C with a saturating ligand concentration and the ligand distribution between the cell surface and the cytoplasm was determined using an acid wash. This protocol differs from that previously employed (Pinkas-Kramarski et al., 1996a) in two aspects. First, a 10-fold lower ligand concentration was used in order to avoid saturation of the coated pit-mediated internalization pathway (reviewed in Sorkin and Waters, 1993). Secondly, other protocols do not include a step that removes unbound ligand prior to initiation of endocytosis. Therefore, continuous uptake of the radiolabeled ligand may mask differences in endocytosis rates. The results of this experiment presented in Figure 3B confirmed that internalization of E4T is significantly slower than that of EGF or TGFα. More importantly, the rate of EGF uptake was remarkably decelerated by a co-expressed ErbB-2, but less so in the presence of ErbB-3 (EGF panel in Figure 3B). The rate of TGFα internalization was similarly affected by the presence of ErbB-2 or ErbB-3 (hTGFα panel in Figure 3B), implying that receptor heterodimers endocytose more slowly than homodimers, irrespective of ligand identity. Because both homodimers and heterodimers of ErbB-1 apparently exist in D12 and in D13 cells, the net kinetics of heterodimer internalization is expected to be even slower than the rates reflected in Figure 3B. Taken together, the data presented in Figure 3 suggest that signaling superiority of E4T is due to the slow rates of internalization and clearance of this ligand from the medium. Possibly, rapid dissociation of E4T from ErbB-1 (Figure 2) prevents efficient internalization. On the other hand, the relatively weak signaling capacity of EGF through the singly expressed ErbB-1 is attributed by our results to the efficient rate of cellular uptake of this ligand. Moreover, the potentiating effect of ErbB-2 is probably due to its ability to decelerate both the rate of internalization (Figure 3B) and the rate of clearance of EGF from the medium (Figure 3A), in line with the relatively slow down-regulation and endocytosis of ErbB-2 (Sorkin et al., 1993; Baulida et al., 1996). Despite these consistencies, our results cannot provide a satisfactory explanation for the relatively high potency of TGFα; although this ligand is more potent than EGF on D1 cells, and it is almost equipotent to EGF on D12 cells (Figure 1), its rates of internalization (Figure 3B), depletion from the medium (Figure 3A) and dissociation from the cell surface (Figure 2B), are only slightly different than those of EGF, and they apparently cannot account for the EGF-specific 10–15-fold mitogenic enhancement effect of ErbB-2 (Figure 1A). EGF and TGFα are comparably degraded, but E4T degradation is limited According to one possibility, EGF and TGFα are similarly endocytosed, but whereas the former is efficiently degraded in lysosomes, the other escapes intracellular degradation. To test this model we treated cells with each of the radiolabeled ligands under conditions that prevent receptor recycling and retard targeting to the degradative pathway. Upon transfer of chilled cells to 37°C ligand degradation was allowed and monitored using acid precipitation. The results presented in Figure 4 indicate that E4T is degraded at a slower rate than EGF and TGFα, as expected on the basis of its slower rate of uptake (Figure 3B), but intracellular degradation of EGF and TGFα were comparable in kinetics and extent. Remarkably, expression of a co-receptor together with ErbB-1 only slightly affected the rates of ligand degradation. In experiments not shown we confirmed a previous report (Hamel et al., 1997) that degradation of both ligands was significantly inhibited by chloroquine, a drug known to inhibit degradation in both endosomal (prelysosomal) and lysosomal compartments, but leupeptin, a tripeptide whose inhibitory action is specific to lysosomes (Cardelli et al., 1989), did not affect TGFα degradation. Conceivably, EGF is destined for lysosomal degradation after endocytosis (Renfrew and Hubbard, 1991), whereas TGFα is degraded in a non-lysosomal compartment whose identity is only partly characterized (Hamel et al., 1997). Independent of its exact intracellular location, endocytic degradation of EGF and TGFα cannot provide an explanation for the superiority of TGFα and the potentiating effect of ErbB-2. Figure 4.Kinetics of ErbB-mediated ligand degradation. The indicated radiolabeled ligands (each at 1 nM) were incubated for 1 h at 20°C with the following derivatives of 32D cells: D1 (▴), D12 (●) or D13 cells (▪). Thereafter, the cells were spun through a cushion of serum to remove unbound ligand, and then incubated at 37°C for various time intervals. Media were then collected and cells solubilized. The fraction of acid-soluble (degraded) ligand in the medium was determined by counting the acid-soluble radioactivity in the medium and the total cell-associated radioactivity. The results are expressed as the average percentage of acid-soluble radioactivity, relative to the sum of cell-associated and medium-released radioactive counts. Bars represent standard deviations. The experiment was performed in dupli
Referência(s)