| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Kimmel Cancer Center, Thomas Jefferson University, Philadelphia, Pennsylvania 19107
Address all correspondence and requests for reprints to: Renato Baserga, M.D., Kimmel Cancer Center, Thomas Jefferson University, 233 South 10th Street, Philadelphia, Pennsylvania 19107. E-mail: B_lupo{at}mail.jci.tju.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Nuclear translocation of IRS-1 in cells expressing the SV40 T antigen was not totally unexpected, as Zhou-Li et al. had demonstrated that IRS-1 coprecipitates with the SV40 T antigen (and vice versa) both in MEFs (9) and in 32D cells (10). In 32D cells, neither IRS-1 nor the T antigen, singly, can protect cells from apoptosis caused by IL-3 withdrawal. However, a combination of the two results in IL-3 independence (10). Therefore, in these cells, the association between T antigen and IRS-1 has also a functional significance. More difficult to explain is the mechanism(s) that results in nuclear translocation of IRS-1 in cells not expressing the T antigen. We explored this question by using mutants of IRS-1 to determine the IRS-1 domains required for IGF-IR-mediated nuclear translocation. Deletion of the phosphotyrosine binding (PTB) domain abrogates nuclear translocation of IRS-1. Deletion of the Pleckstrin domain, on the contrary, does not alter IGF-I-mediated translocation (5). Interestingly, a plasmid expressing only the pleckstrin homology (PH)/PTB domains of IRS-1 (11) translocated, at least partially, to the nuclei (5).
Since IRS-1 and IRS-3 can translocate to the nuclei, it is reasonable to ask whether the same may be true for IRS-2. One would also like to investigate the signals from the IGF-IR itself that regulate IRS translocation, as they may give us clues on the mechanisms of translocation. In the present experiments, therefore, we have asked two questions: does IRS-2 also translocate to the nuclei? And, if so, which are the domains of the IGF-IR required for nuclear translocation of IRS-1 and IRS-2? Our results indicate that IRS-2 translocates to the nuclei of MEFs stimulated by IGF-I, but not of MEFs devoid of IGF-IR and expressing the SV40 T antigen. As to the domains of the IGF-IR required for translocation, the tyrosine kinase (TK) domain is essential. A triple mutation at Y1131, Y1135, and Y1136 abrogates the ability of the IGF-IR to induce nuclear translocation of IRS-1 and IRS-2. Other mutations, detailed below, result in decreased translocation (but not abrogation) of both proteins. In addition, we find that IRS-2, like IRS-1, seems to translocate to the nucleolus, but to a lesser extent. Nucleolar translocation is decreased also in cells expressing IGF-IR mutants that are mitogenic but nontransforming. IRS-1 and, to a lesser extent, IRS-2 immunoprecipitate the upstream binding factor (UBF) from nuclear lysates. UBF is a major regulator of RNA polymerase I activity and, therefore, of rRNA synthesis (12, 13, 14). In agreement with this finding, the presence of IRS-1 in 32D cells significantly increases the activity of the ribosomal DNA (rDNA) promoter. Conversely, the decreased translocation in cells expressing the mutant IGF-I receptors is accompanied by a decrease in rRNA synthesis. The interaction of nuclear IRS proteins, especially IRS-1, with UBF provides a direct link between nuclear IRS proteins and the transcription machinery of the rDNA promoter. It also provides an explanation for the role of IRS proteins in determining cell and body size (see Discussion).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
We have previously shown by confocal microscopy that IRS-1 is not translocated to the nuclei of R- cells (6). Figure 1A
shows an immunohistochemistry picture of R- cells, stained with an antibody to IRS-2 and counterstained with hematoxylin. IRS-2, like IRS-1, is essentially localized to the cytoplasm of R- cells (see also the subcellular fractionation in Fig. 4). There are some specks of IRS protein over the nuclei, but, in immunohistochemistry, there is always a thin layer of cytoplasm over the nuclei. For a proper evaluation, compare the color of the nuclei in panel A with the color of the nuclei in panel D, where the nuclei are stained by an antibody to IRS-1 (see below).
|
|
(immunohistochemistry in panel B and confocal microscopy in panel C). For confocal microscopy, we present only the merged pictures of cells in serum-free medium, but the results were essentially the same when the cells were stimulated with IGF-I. IRS-2 is essentially cytoplasmic (panel B). When the cells are examined by confocal microscopy, IRS-2 is also detectable only in the cytosol (panel C). The nuclei appear as punched out empty spaces dotted with the green specks of nucleolin-stained nucleoli. In contrast, notice the color of the nuclei (and the nucleoli) in the immunohistochemistry picture of R-/T cells stained with an antibody to IRS-1 (panel D).
These results were confirmed by subcellular fractionation and Western blots (Fig. 2). IRS-1 is present in the nuclear fraction of R-/T cells, but IRS-2 is not. Nuclear extracts were immunoprecipitated with antibodies to either IRS-1 or IRS-2, and the resultant blots were stained with an antibody to T antigen. T antigen was detectable only in the IRS-1 antibody immunoprecipitates (data not shown). The combined results of subcellular fractionation, immunohistochemistry, confocal microscopy, and immunoprecipitation indicate that in R-/T cells, IRS-1, but not IRS-2, translocates to the nuclei.
|
). IRS-1 translocation confirms the report of Tu et al. (6). We used as a control a cell lysate of parental 32D cells, which do not express IRS-1 or IRS-2 (17, 18). This lysate is negative for both IRS-1 and IRS-2 (not shown here but shown in Fig. 4). Grb2 and c-Jun antibodies were used to monitor the purity of the subcellular fractions. The IRS proteins in R+ cells are translocated also in cells kept in serum-free medium. R+ cells, like most R- derived cells, produce a detectable amount of IGF-I (19). When they express exogenous IGF-IRs, R- derived cells grow, albeit slowly, in serum-free medium. Incidentally, the 8-h point was chosen after a time course (not shown, but shown in Ref. 6) indicated that 8 h is the most reliable point.
The nuclear translocation of IRS-2 in R+ cells was confirmed by confocal microscopy (Fig. 3). The cells were stained for IRS-2 (rhodamine) and nucleolin [FITC (fluorescein isothiocyanate)]. Although there is some IRS-2 in the cytoplasm, a substantial fraction of the protein is now detectable in the nuclei of R+ cells, as already shown for IRS-1 (5, 6). The nucleoli, in the merged picture, show mix staining for IRS-2 and nucleolin (greenish-yellow nucleoli), but less so than with IRS-1. The conclusion, though, is that IRS-2 can also localize to the nucleoli of R+ cells, but not as strongly as IRS-1.
|
is a Western blot of lysates from the cytosol and nuclear fractions of R- cells and R- 3Y cells expressing the TK mutant receptors. Neither IRS-1 nor IRS-2 translocate to the nuclei of R- cells or of R- cells expressing the TK mutant of the IGF-IR. This was not surprising because the TK domain mutant is essentially a dead receptor, which fails to phosphorylate IRS-1 (20). The experiment was repeated, and the results were again so clear that this mutant receptor was no longer pursued.
We then focused on three other mutants of the IGF-IR that are functional but not completely so. These mutants are the 
1245 mutant (in which the receptor is truncated at residue 1245), the Y950F mutant (tyrosine 950 mutated to phenylalanine), and the 4-serine (4-ser) mutant (in which the serines at 1280–1283 have been mutated). The R- cells expressing these mutants have been described in previous papers (16, 21, 22). A common characteristic of MEFs expressing these mutant receptors is that they respond to IGF-I with mitogenesis, but they can no longer transform R- cells (colony formation in soft agar), whereas R+ and R-/GR15 cells (both of them expressing wild-type receptors) are transformed (16).
1245.
Figure 4B is a Western blot of cytosolic and nuclear fractions obtained from the R-/1245 cells (receptor truncated at residue 1245). With these cells, both IRS-1 and IRS-2 can be found in the nuclear fractions. Notice that lysates of parental 32D cells are completely negative for both proteins (left lane). Nuclear translocation was confirmed by immunohistochemistry and confocal microscopy (Fig. 5). In R-/1245 cells stimulated for 8 h with IGF-I, IRS-1 definitely translocates to the nuclei of these cells, and also to the nucleoli, as in R+ cells (again compare the color of the nuclei with the color of the nuclei in Fig. 1). There is partial translocation of IRS-2, but less than IRS-1, and the nucleoli in the merged picture conserve their green staining. Repeated experiments, even in different culture conditions, confirmed a reduced translocation of IRS proteins in cells expressing the 1245 receptor and the lack of nucleolar localization of IRS-2.
|
and 7. By subcellular fractionation, in R- cells expressing the Y950F mutant receptor, IRS-2 translocates to the nuclei, but in lesser amounts than IRS-1. This was confirmed by immunohistochemistry and confocal microscopy (Fig. 7). There is some IRS-2 in the nuclei of these cells, but less than IRS-1, and some of the nuclei are negative. The nucleoli in cells stained for IRS-2 are green-colored, as in R-/1245 cells, and contrast with the yellow color of nucleoli in the same cells stained for IRS-1.
|
|
and 8. The results are very much the same as with the previous mutant receptor. Both proteins translocate, IRS-2 less than IRS-1 (Fig. 6B), and the nucleoli do not seem to stain for IRS-2 (Fig. 8). The differences between IRS-1 and IRS-2 are concordant by immunohistochemistry, confocal microscopy, and subcellular fractionation. The differences in nucleolar staining are especially striking.
|
. Whether in R+ cells or in R-/4 ser cells, an antibody to IRS-1 is more efficient than an antibody to IRS-2 in immunoprecipitating UBF from nuclear extracts. The main band is UBF1, which is the active form (14). The lower band in some of the lanes is UBF2, the inactive form of UBF, which is 37 amino acids shorter than UBF1 (14). A small amount of UBF is detectable also in blots from R-/4 ser cells immunoprecipitated with IRS-2. It is difficult to explain this result as a contamination, because the IRS antibodies do not immunoprecipitate UBF in 32D cells (no IRS-1 or -2). The antibody to IRS-2 fails to immunoprecipitate UBF in R-/T cells, where IRS-2 does not translocate to the nucleus.
|
shows a Western blot, where the proteins obtained by immunoprecipitation with the anti-IRS antibodies were run on a 7.5% gel. There are two bands, but the predominant form, both with IRS-1 and IRS-2 antibodies, is UBF1. There is also an increase in UBF amounts in stimulated cells, in agreement with previous reports from the literature (14). We repeated the experiment with another cell line, 32D IGF-IR PHPTB IRS-1, derived from 32D cells. This cell line (5) expresses a mutant IRS-1 that comprises only the first 300 amino acids. This mutant IRS-1 translocates to the nucleus, but less than the wild-type protein (5). Panel D shows that in these cells, an antibody to IRS-1 coprecipitate UBF but to a lesser extent than in cells expressing the wild-type protein. It seems therefore that the amount of UBF1 interacting with IRS-1 is a reflection of the extent of nuclear translocation. The left lane of panel D is the control, from cells (32D IGF-IR) that do not express IRS-1.
rRNA Synthesis
In a previous paper (6), we showed that the presence of IRS-1 in the nuclei of 32D-derived cells markedly increases rRNA synthesis. Because the cells expressing three of the mutant receptors have nuclear translocation, but a diminished nucleolar translocation, we have asked whether these mutations have any effect on rRNA synthesis. The cells indicated in Fig. 9C, stimulated with IGF-I, were labeled with 32P for 4 h, followed by a 2-h incubation in growth medium (no radioactivity). The RNA was extracted, and the gel was autoradiographed (panel C). There is a marked decrease in rRNA labeling in cells expressing the mutant receptors, in comparison to R+ cells, that express the wild-type IGF-IR. The decrease showed by autoradiography was confirmed by counting the radioactivity in the bands (not shown). In Fig. 9E, we show rRNA synthesis in the cell lines used in panel D, which are derived from 32D cells. The cells expressing the wild-type IRS-1 have the highest rate of rRNA synthesis, whereas the PH/PTB mutant, which translocates to the nucleus, but to a lesser extent, synthesizes rRNA at a rate intermediate between the cells expressing the wild-type IRS-1 and the cells not expressing IRS-1.
Activation of the rDNA Promoter
Activation of rRNA synthesis requires an active UBF (23, 24), in agreement with our findings in Fig. 9, where the levels of rRNA synthesis seem to correlate with the amount of UBF immunoprecipitated by an antibody to IRS-1. We asked next whether the activation of the IRS-1 pathway could directly increase the activity of the rDNA promoter. For this purpose, we used the miniribosome gene of Grummt and co-workers (23, 25), in which a specific rDNA sequence is driven by the rDNA promoter. The miniribosome gene was transfected transiently in the appropriate 32D-derived cell lines, and the amount of transcription was determined by Northern blots and counting of the bands (Fig. 10). The activation of the rDNA promoter is highest in 32D IGF-IR IRS-1 cells, clearly above 32D IGF-IR cells, that do not express IRS-1. This is a significant finding because, at this time after shifting from IL-3 to IGF-I, both cell lines are growing exponentially (18, 26). Despite the fact that both cell lines are growing, the presence of IRS-1 markedly increases rDNA promoter activation. The result is confirmed by the fact that mutant IRS-1 proteins, although they can also translocate to the nuclei (see Ref. 5 ; albeit less efficiently) are not as effective as the wild-type protein in activating the rDNA promoter. These results indicate that the activation of the IGF-IR signaling pathway increases transcription from the rDNA promoter.
|
Pleckstrin Domain of IRS-1 shows that the FLAG epitope is detected only in the cytosol fraction of R-/T cells. This was confirmed by confocal microscopy (panel B), where the cells are stained for FLAG and the nuclei are stained with propidium iodide. Again, the FLAG-tagged mutant IRS-1 is limited to the cytoplasm of R-/T cells. We repeated the experiment, and this time we compared R-/T cells to R+ cells, also transfected with the PH mutant of IRS-1. Figure 11C shows the dramatic difference between the two cell lines. Both by immunohistochemistry and by confocal microscopy, the FLAG-tagged PH IRS-1 is in the cytoplasm of R-/T cells (1, 2) and (largely) in the nuclei of R+ cells (3, 4). In this experiment, the nuclei were stained with an antibody to Id1.
|
, right panel, shows that tyrosyine-phosphorylated bands are detectable at the levels of IRS-1 and IRS-2. The blots were stripped and stained with the respective antibodies to the IRS proteins. Figure 12, left panel, shows that the two bands detected in the right panel are indeed IRS-1 and IRS-2. This experiment was repeated with the same results. These data indicate that the IRS proteins in the nuclei of R+ cells can be tyrosyl phosphorylated. They do not state whether all of the nuclear IRS proteins are phosphorylated, nor do they indicate whether phosphorylation is necessary for translocation. However, an IRS-1 with a deletion of the PTB domain does not translocate to nuclei (5).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
1245,Y950, and 4 ser) show nuclear translocation of both IRS proteins, albeit somewhat reduced in comparison to the wild-type receptor. 4) IRS-1 translocates to the nucleoli in all instances, except in cells expressing the 3Y mutant receptor. 5) IRS-2 localization to the nucleoli is modest in R+ cells, and below levels detectable by confocal microscopy in the cell lines expressing the mutant receptors. 6) Both anti-IRS-1 and anti-IRS-2 antibodies coprecipitate UBF1 from nuclear lysates of R+ and R-/4 ser cells, but the IRS-1 antibody is more efficient. In R-/T cells, only the anti-IRS-1 antibody can coprecipitate UBF1. 7) Wild-type IRS-1 increases transcription from the rDNA promoter. Mutant IRS-1 proteins are much less effective. 8) The decreased nuclear and nucleolar translocation in cells expressing the mutant IGF-I receptors is accompanied by a decrease in rRNA synthesis. 9) The Pleckstrin domain of the IRS proteins is determinant for T antigen-dependent translocation, but not for IGF-I-mediated translocation. In comparing the nuclear translocation of IRS-1 and IRS-2, two differences are striking: 1) With the mutant IGF-IR, IRS-2 translocation is more impaired than in the case of IRS-1, especially translocation to the nucleoli. 2) IRS-2 does not translocate to the nuclei of R-/T cells, while IRS-1 does.
The effects of IGF-IR mutations on nuclear translocation of IRS proteins indicate that the TK domain is an absolute requirement. The other mutations tested cause an inhibition, but not an abrogation, of nuclear translocation. The three cell lines expressing these mutant IGF-I receptors all respond with mitogenesis to the addition of IGF-I (16, 22), but they do not form colonies in soft agar. MEFs expressing the wild-type IGF-IR (R+ or R-/GR15 cells) form colonies in soft agar. We do not pretend to place a precise quantitative assessment of nuclear translocation in experiments based on immunohistochemistry and confocal microscopy, and, for this reason, the following statement is limited to the Discussion. When we look at the percentage of cells with nuclear translocated IRS proteins, we can state that 99% of the cells expressing the wild-type IGF-IR (R+ cells) had some IRS-1 or IRS-2 in the nuclei. In cells expressing the mutant receptors (except for the 3 Y receptor), IRS-1 was nuclear in about 90% of the cells examined, and IRS-2 was nuclear in about 40–60% of the cells. The subcellular fractionations are roughly in agreement with this estimate. However, it should be understood that none of the three techniques used to localize the IRS proteins in the nuclei are strictly quantitative. Even with subcellular fractionation, leakage of proteins from the nuclei and the partial breakdown of proteins that inevitably occurs in a lengthy procedure make quantitation elusive.
A very important question raised by our experiments is the nucleolar localization of IRS-2. From a previous paper (6) and from this one also, there is no question that IRS-1 can localize to the nucleoli of R+ or R-/T cells, where it interacts with nucleolin and UBF (Ref. 6 and this paper). The interaction of IRS-2 with the nucleolus is not as convincing. There is some IRS-2 colocalizing with nucleolin in R+ cells, but in cells expressing the mutant receptors, the nucleoli are green, as if there were no or little IRS-2 in the nucleoli. An antibody to IRS-2 does coprecipitate a small amount of UBF even in R-/4 ser cells. UBF is a nucleolar protein (14), and its localization to the nucleoli remains unchanged in a variety of growth conditions (24). A reasonable explanation for the interaction of IRS-2 with UBF is that there is a modest amount of IRS-2 in the nucleolus, below the level detectable by confocal microscopy. In support of this interpretation is the complete absence of UBF from nuclear extracts of R- T cells immunoprecipitated with an antibody to IRS-2. Alternatively, one could attribute the small amount of UBF to the fact that, during nuclear lysis, UBF and IRS-2 coexist in the lysis buffer. This may allow coprecipitation, even though the two proteins may be predominantly localized in different compartments. Both antibodies to IRS-1 and IRS-2 largely coprecipitate UBF1, which is the active isoform, while the lower form UBF2 is inactive (14).
In previous papers, nuclear IRS-1 seemed to be responsible for a marked increase in rRNA synthesis (5, 6). A decrease in nucleolar localization is accompanied by a substantial reduction in rRNA synthesis (present paper). This occurs in cells expressing the mutant IGF-I receptors, which are mitogenic but nontransforming (16). There are two possible explanations for this finding. One possibility is that nucleolar IRS-2 is necessary for the full activation of rRNA synthesis. This is unlikely, because in 32D cells (which do not express IRS-2), ectopic expression and nuclear translocation of IRS-1 are sufficient to increase rRNA synthesis (5, 6) and to increase cell size (26), which depends on increased ribosome biogenesis (27). A second explanation is that the mutations examined, especially the Y950F mutation, affect the binding of the IGF-IR to Shc, and the subsequent activation of the Ras/Raf/ERK pathways (16, 28). This is true also of the 4 ser mutation (28), and of the 1245 mutation. Truncation at residue 1245 eliminates not only the 4 serine group, but also tyrosines 1250/1251, which interact with Shc (18). These mutations cause an attenuation of ERK activation (29, 30). Interestingly, UBF is phosphorylated and activated by ERK in response to growth factors (31). This explanation is in agreement with our data, indicating that attenuation of the ERK pathway in this model reduces rRNA synthesis, even when IRS proteins are present. It suggests that full activation of rRNA synthesis requires both a nuclear IRS-1 and a signal from the activated ERK proteins. This interpretation is supported by the finding that a combination of overexpressed IRS-1 and Ha-ras results in transformation of parental 32D cells (32). Even accepting a contribution from the ERK proteins, it is fair to say that our present findings and those reported by Tu et al. (6) suggest that the nuclear translocated IRS proteins play an important role in rRNA synthesis.
This brings us to the biological significance of our findings. Cell size is known to correlate strictly with protein amounts (33) and the amount of rRNA (34, 35, 36). As mentioned previously (27), genetic studies in yeast have confirmed that cell size largely depends on ribosome biogenesis. In turn, ribosome biogenesis is controlled by the rate of rRNA synthesis (37), which is dependent on the activity of RNA polymerase I (14, 38, 39). Cell size is strongly controlled, in vivo, by the IGF-IR, the IRS proteins, and their downstream effectors. Thus, cell size is regulated in Drosophila by the homologs of either IRS-1 (40) or S6K1 (41) or Akt (42). Particularly instructive is the Drosophila IRS homolog, called chico, which is Drosophila-only IRS protein and therefore equivalent to IRS-1 to -4. Deletion of chico reduces fly weight by 55–65%. The reduction in body and organ size is due both to a reduction in cell number and cell size. The role of IGF-IR and IRS signaling has also been demonstrated in mice with a targeted disruption of the IRS-1 (43) or the S6K1 (44) genes. Similarly, 32D IGF-IR cells expressing IRS-1 are larger than their parental cells (26). Finally, deletion of the IGF-IR genes and their ligand results in embryos that are only 30% in size in respect to wild-type littermates (45). Thus, direct evidence supports a major (albeit not exclusive) role of IRS proteins in determining cell size and, therefore, rRNA and protein production. Since UBF is a key regulator of RNA polymerase I activity (14, 24, 31, 34, 38), the interaction of IRS proteins, but especially of IRS-1, with UBF provides a mechanistic explanation for the effect of IRS-1 and chico on cell and body size. The interaction of IRS-1 (and IRS-2) with UBF1 constitutes the first direct link between nuclear IRS proteins and the transcription machinery of the rDNA promoter, which is indeed activated by wild-type IRS-1 in 32D cells (present paper). Less clear, at the moment, is the biological significance of IRS-1 interaction with nucleolin (6, 46).
The failure of IRS-2 to translocate to the nuclei of R-/T cells was confirmed by all three methods. In addition, an antibody to IRS-2 fails to coprecipitate the T antigen, which is instead precipitated by an antibody to IRS-1 (9, 10). It seems that IRS-2 does not interact with the SV40 T antigen, which explains its failure to translocate in R-/T cells. The PH domain is required for interaction of IRS-1 with JCV T antigen (4), the human counterpart of the SV40 T antigen (present paper). The PH domain, however, is not required for IGF-I-mediated translocation of IRS-1 in R+ cells (5). Our data show that the PH IRS-1 mutant, which translocates to the nuclei of cells with an activated IGF-IR (5), does not translocate to the nuclei in R-/T cells. Thus, IRS-1 with a deletion of the PH domain behaves, in respect to T antigen, like the wild-type IRS-2. It suggests that the PH domain of IRS-1 has a sequence for T antigen binding that is absent in wild-type IRS-2. An in silico analysis shows that the sequences of the PH domains of IRS-1 and IRS-2 have similarities and differences. While the ß-sheets and the 
-helices are very well conserved, the interdomains are quite divergent (47). It is reasonable to hypothesize that the binding to T antigen will depend on these interdomains, a hypothesis that can be tested. This hypothesis could also apply to the difference in nucleolar localization between the two proteins. It may be mentioned that the PH domains of both IRS proteins contain a sequence that is similar to published nucleolar localization signals (48). At any rate, two conclusions are acceptable: IRS-2 seems to require an active IGF-IR for translocation, while IRS-1 can translocate also with the T antigen, in cells without IGF-IR, and is less affected by mutations in the receptor.
The evidence that the antibodies used for IRS-1 in these experiments actually detect IRS-1 and only IRS-1 has been given in previous papers from this laboratory and others (4, 5, 6). The results obtained by subcellular fractionation, confocal microscopy, and immunohistochemistry are all concordant. Furthermore, the nuclear localization of IRS-1 was confirmed using an IRS-1 with a FLAG epitope and competition with the peptide used to produce the antibody. We believe that the anti-IRS-2 antibody is also reliable for the following reasons. By subcellular fractionation and by confocal microscopy, the same antibody recognizes the presence of IRS-2 in the nuclei of R+ cells, and in the cytosol of R-/T cells. No IRS-1 or IRS-2 is detected by these antibodies in the nuclei of R- cells expressing the 3Y mutant of the IGF-IR or in lysates of 32D cells, which do not express these two proteins (17, 18).
Nuclear IRS proteins are tyrosyl phosphorylated. Given the unavoidable semiquantitative aspects of these experiments, we cannot say that all nuclear IRS proteins are phosphorylated, but this possibility is suggested by the failure of an IRS-1 with a deletion of the PTB domain to translocate to the nuclei (5). Translocation seems to be ligand independent. However, this is not quite true. As already mentioned, R- cells are known to secrete detectable amounts of IGF-I. When they are expressing the human IGF-IR (wild type or certain mutants), the cells grow, albeit slowly, in serum-free medium. Addition of IGF-I to the medium does increase growth rates (19).
In conclusion, we have demonstrated that in MEFs, IRS-2 also translocates to the nuclei of cells expressing the wild-type IGF-IR. For translocation of both IRS-1 and IRS-2, the TK domain of the IGF-IR is an absolute requirement. Mutations in the IGF-IR that abrogate the transforming activity, but not the mitogenic activity, of the IGF-IR result in decreased translocation, especially to the nucleoli. The nuclear IRS proteins interact with UBF1, a key regulator of rRNA synthesis, and decreased nucleolar localization correlates with decreased rRNA synthesis. These findings establish a link between nuclear IRS proteins and the regulation of rDNA transcription, with its implications in the regulation of cell growth in vivo and in vitro. It goes without saying that a possible role of IRS proteins in rRNA regulation does not exclude that the IRS proteins, in the cytosol or in the nucleus, may be also responsible for regulating the expression of RNA polymerase II-dependent genes.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
1245 mutant (in which the receptor is truncated at residue 1245), and 4-ser mutant (in which the serine at 1280–1283 has been mutated). These cell lines have been described in detail in previous papers (9, 15, 16, 20, 21, 22). 32D cells, a murine hemopoietic cell line that does not express either IRS-1 or IRS-2 (17), were used as a negative control in some experiments. In some experiments, 32D IGF-IR/IRS-1, 32D IGF-IR, 32D-IGF-IRPH, and 32D IGF-IR/PHPTB cells were used for immunoprecipitation of the UBF, rRNA synthesis, and rDNA promoter activation. These cell lines have been described previously (5, 18, 26).
Immunofluorescence/Confocal Microscopy and Immunohistochemistry
Cells seeded at a density of 1 x 104 cells/ml were attached to glass coverslips at least overnight, and then shifted to serum-free medium for 48 h and stimulated with IGF-I (50 ng/ml) for 8 h. After fixation with 3% paraformaldehyde for 25 min and washing three times with PBS at room temperature, cells on coverslips were permeabilized with 0.2% Triton X-100 in PBS for 5 min, blocked with 10% normal donkey serum (sc-2044, Santa Cruz Biotechnology, Inc., Santa Cruz, CA) for 20 min, and incubated for 1 h with the appropriate primary and secondary antibodies. Confocal analysis was carried out on a MRC-600 Ar/Kr laser scanning confocal microscope (Bio-Rad Laboratories, Inc., Hercules, CA) using a 40x objective (Carl Zeiss, Thornwood, NY). The immunostaining was processed by the use of the Histomouse (AEC) kit (Zymed Laboratories, Inc., South San Francisco, CA) according to the manufacturer’s protocol. The magnifications for the figures presented were either x400 or x1000.
Subcellular Fractionation, Western Blotting, and Immunoprecipitation
Before stimulation with IGF-I, cells were serum starved for 48 h. The quiescent cells were stimulated with IGF-I (50 ng/ml) for 8 h. The cells were rinsed with cold PBS, detached with trypsin, and pelleted at 1500 rpm for 5 min at 4 C. The cells were then lysed in homogenization buffer [10 mM Tris-HCl, pH 7.4; 15 mM NaCl; 60 mM KCl; 1 mM EDTA; 0.1 mM EGTA; 0.5% Nonidet P-40; 5% sucrose; and the protease mixture from Roche (Indianapolis, IN)] for 10 min, and then homogenized by 15 strokes in a tightly fitting Dounce homogenizer and examined under a microscope to ascertain that the majority of nuclei had been freed from the cytoplasm. The homogenate was centrifuged at 6000 rpm for 1 min to sediment the nuclei. The supernatant was then resedimented at 10,000 rpm for 10 min, and the resulting supernatant formed the cytoplasmic fraction. The nuclear pellet was passed through 5 ml sucrose buffer (10 mM Tris-HCl, pH 7.4; 15 mM NaCl; 60 mM KCl; 10% sucrose), washed three times with wash buffer (10 mM Tris-HCl, pH 7.4; 15 mM NaCl; 60 mM KCl), and resuspended in the lysis buffer containing 0.5 M NaCl to extract nuclear protein. The extract was centrifuged at 10,000 rpm for 10 min at 4 C, and the supernatant was termed the "nuclear fraction." Cytoplasmic and nuclear fractions (50 µg) were separated on a 4–15% gradient gel (Bio-Rad Laboratories, Inc.) and transferred to a nitrocellulose membrane. In one experiment, the proteins were separated on a 7.5% gel. For immunoprecipitation, 200 µg of nuclear or cytoplasmic lysate were incubated for 2 h at 4 C with the corresponding antibodies coupled to 20 µl of packed protein G-sepharose beads (Oncogene Science, Inc.). Immunocomplexes were resolved by means of SDS-PAGE and immunoblotted with the indicated antibodies.
Antibodies
The antibodies used in this study included: rabbit polyclonal anti-IRS-1 antibody (Upstate Biotechnology, Inc., Lake Placid, NY), rabbit polyclonal anti-IRS-1 antibody recognizing the N terminus of IRS-1 (Santa Cruz Biotechnology, Inc.), rabbit polyclonal anti-IRS-2 antibody antibody (Upstate Biotechnology, Inc.), goat polyclonal anti-IRS-2 antibody (Santa Cruz Biotechnology, Inc.), mouse monoclonal anti-SV 40 T antigen antibody (Santa Cruz Biotechnology, Inc.), mouse monoclonal antinucleolin antibody (Santa Cruz Biotechnology, Inc.), mouse monoclonal anti-Grb2 antibody (Transduction Laboratories, Inc., Lexington, KY), mouse monoclonal UBF antibody (Santa Cruz Biotechnology, Inc.), anti-C-jun polyclonal antibody (Santa Cruz Biotechnology, Inc.). Monoclonal anti-FLAG-fluorescein isothiocyanate conjugates and anti-FLAG-peroxidase conjugate were from Sigma (St. Louis, MO).
Metabolic Labeling of rRNA
R- derived cells were seeded at a density of 5 x 104 cells/ml in growth medium and eventually transferred to serum-free medium for 24 h. The cells were stimulated with 50 ng/ml of IGF-I (Life Technologies, Inc., Gaithersburg, MD) for the appropriate times. The cells were then labeled for 4 h with [32P]orthophosphate at a final concentration of 250 µCi/ml (ICN Biochemicals, Inc., Cleveland, OH) in phosphate-free medium (Life Technologies, Inc.). After labeling, the cells were washed and incubated in fresh medium for 2 h. Total RNA was isolated using RNeasy MiniKit (QIAGEN, Chatsworth, CA) and separated by electrophoresis on 1% agarose formaldehyde gels. After drying, the 32P-labeled rRNA was visualized by autoradiography. The bands also were counted in a liquid scintillation counter.
Activation of the rDNA Promoter
The activation of the rDNA promoter by IRS-1 was carried out by the transient transfection of the miniribosome gene described by Grummt and co-workers (23, 24, 25). It has almost 2000 residues before the starting site, a brief stretch of transcribed rDNA, and termination sequences. Just before the termination sequence, a unique in-frame viral sequence provides the specific sequence for Northern blots. The cells were transfected for 24 h, and the amount of transcribed miniribosome gene was determined by Northern blots and subsequent counting of the bands.
FLAG Tagging of Mutant IRS-1
The tagging of IRS-1 was described in detail by Prisco et al. (5) as well as the procedure for transfection and staining.
|
FOOTNOTES |
|---|
Abbreviations: IGF-IR, Insulin-like growth factor I receptor; IRS, insulin receptor substrate; MEFs, mouse embryonic fibroblasts; PH, pleckstrin homology; PTB, phosphotyrosine binding; rDNA, ribosomal DNA; 4-ser, 4-serine; SV40, Simian virus 40; TK, tyrosine kinase; UBF, upstream binding factor.
Received for publication August 7, 2002. Accepted for publication November 21, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  I. Villalpando, E. Lira, G. Medina, E. Garcia-Garcia, and O. Echeverria Insulin-Like Growth Factor 1 Is Expressed in Mouse Developing Testis and Regulates Somatic Cell Proliferation Experimental Biology and Medicine, April 1, 2008; 233(4): 419 - 426. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. Dalmizrak, A. Wu, J. Chen, H. Sun, F. E. Utama, D. Zambelli, T. H. Tran, H. Rui, and R. Baserga Insulin Receptor Substrate-1 Regulates the Transformed Phenotype of BT-20 Human Mammary Cancer Cells Cancer Res., March 1, 2007; 67(5): 2124 - 2130. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Yan, C.-T. Yu, M. Ozen, M. Ittmann, S. Y. Tsai, and M.-J. Tsai Steroid Receptor Coactivator-3 and Activator Protein-1 Coordinately Regulate the Transcription of Components of the Insulin-Like Growth Factor/AKT Signaling Pathway. Cancer Res., November 15, 2006; 66(22): 11039 - 11046. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Camarata, B. Bimber, A. Kulisz, T.-L. Chew, J. Yeung, and H.-G. Simon LMP4 regulates Tbx5 protein subcellular localization and activity J. Cell Biol., July 31, 2006; 174(3): 339 - 348. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M L Panno, L Mauro, S Marsico, D Bellizzi, P Rizza, C Morelli, M Salerno, F Giordano, and S Ando' Evidence that the mouse insulin receptor substrate-1 belongs to the gene family on which the promoter is activated by estrogen receptor {alpha} through its interaction with Sp1 J. Mol. Endocrinol., February 1, 2006; 36(1): 91 - 105. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Wilson, M. Hargreaves, and K. F. Howlett Exercise does not alter subcellular localization, but increases phosphorylation of insulin-signaling proteins in human skeletal muscle Am J Physiol Endocrinol Metab, February 1, 2006; 290(2): E341 - E346. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Z. Sheng, Y. Liang, C.-Y. Lin, L. Comai, and W. J. Chirico Direct Regulation of rRNA Transcription by Fibroblast Growth Factor 2 Mol. Cell. Biol., November 1, 2005; 25(21): 9419 - 9426. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Chen, A. Wu, H. Sun, R. Drakas, C. Garofalo, S. Cascio, E. Surmacz, and R. Baserga Functional Significance of Type 1 Insulin-like Growth Factor-mediated Nuclear Translocation of the Insulin Receptor Substrate-1 and {beta}-Catenin J. Biol. Chem., August 19, 2005; 280(33): 29912 - 29920. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Wu, X. Tu, M. Prisco, and R. Baserga Regulation of Upstream Binding Factor 1 Activity by Insulin-like Growth Factor I Receptor Signaling J. Biol. Chem., January 28, 2005; 280(4): 2863 - 2872. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Sun and R. Baserga Deletion of the Pleckstrin and Phosphotyrosine Binding Domains of Insulin Receptor Substrate-2 Does Not Impair Its Ability to Regulate Cell Proliferation in Myeloid Cells Endocrinology, November 1, 2004; 145(11): 5332 - 5343. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Drakas, X. Tu, and R. Baserga Control of cell size through phosphorylation of upstream binding factor 1 by nuclear phosphatidylinositol 3-kinase PNAS, June 22, 2004; 101(25): 9272 - 9276. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Prisco, A. Maiorana, C. Guerzoni, G. Calin, B. Calabretta, R. Voit, I. Grummt, and R. Baserga Role of Pescadillo and Upstream Binding Factor in the Proliferation and Differentiation of Murine Myeloid Cells Mol. Cell. Biol., June 15, 2004; 24(12): 5421 - 5433. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Sciacca, M. Prisco, A. Wu, A. Belfiore, R. Vigneri, and R. Baserga Signaling Differences from the A and B Isoforms of the Insulin Receptor (IR) in 32D Cells in the Presence or Absence of IR Substrate-1 Endocrinology, June 1, 2003; 144(6): 2650 - 2658. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
