Effects of Mutations of a Phosphorylation Site in an Exposed Loop in Acidic Fibroblast Growth Factor*

Acidic fibroblast growth factor (aFGF) contains a phosphorylation site recognized by protein kinase C. A non-mitogenic mutant growth factor is devoid of this phosphorylation site. We have changed amino acids in and close to the phosphorylation site and studied the consequences of this for binding of the growth factor to high affinity receptors as well as to heparin. We have also studied the ability of the mutants to stimulate DNA synthesis and cell proliferation as well as phosphorylation of mitogen-activated protein kinase and the ability of the growth factor mutants to be transported to the nucleus. The results indicate that while the mutations strongly affect the ability of the growth factor to bind to heparin, they do not affect much the binding to the specific FGF receptors, activation of mitogen-activated protein kinase or transport of the growth factor to the nucleus. The mutations affect to various extents the ability of the growth factor to stimulate DNA synthesis and to induce cell multiplication. We find that phosphorylation of aFGF is not required for mitogenic activity. The data suggest that altered interaction of the growth factor with a cellular component different from the receptor, possibly a component in the nucleus, is the reason for the different mitogenicity of the different growth factor mutants.

Acidic fibroblast growth factor (aFGF) contains a phosphorylation site recognized by protein kinase C. A non-mitogenic mutant growth factor is devoid of this phosphorylation site. We have changed amino acids in and close to the phosphorylation site and studied the consequences of this for binding of the growth factor to high affinity receptors as well as to heparin. We have also studied the ability of the mutants to stimulate DNA synthesis and cell proliferation as well as phosphorylation of mitogen-activated protein kinase and the ability of the growth factor mutants to be transported to the nucleus. The results indicate that while the mutations strongly affect the ability of the growth factor to bind to heparin, they do not affect much the binding to the specific FGF receptors, activation of mitogen-activated protein kinase or transport of the growth factor to the nucleus. The mutations affect to various extents the ability of the growth factor to stimulate DNA synthesis and to induce cell multiplication. We find that phosphorylation of aFGF is not required for mitogenic activity. The data suggest that altered interaction of the growth factor with a cellular component different from the receptor, possibly a component in the nucleus, is the reason for the different mitogenicity of the different growth factor mutants.
Acidic fibroblast growth factor (aFGF or FGF-1) 1 belongs to large family of growth factors that bind to four closely related receptors, FGFRs 1-4 (1). The receptors contain, in their cytoplasmic part, a split tyrosine kinase domain that is activated upon binding of FGF to the extracellular receptor domain (2). Signaling through this tyrosine kinase with the activation of downstream signaling molecules is an important part in the signal transduction from FGF (3)(4)(5)(6)(7). However, there is also evidence that some growth factors of this family, in particular aFGF and basic FGF (bFGF), enter the cytosol and the nucleus and act directly on intracellular targets (8 -18).
aFGF contains an exposed loop that is involved in binding to heparin (19). Mutation in this region of lysine 132 to glutamic acid (K132E) reduces heparin affinity and essentially abolishes the ability of the growth factor to stimulate DNA synthesis in cells (15,20). aFGF(K132E) binds to FGFR with similar affinity as wild-type aFGF (15) and activates the tyrosine kinase of the receptor (15,20) and stimulates MAP kinase. 2 Furthermore, the K132E mutant stimulates expression of several immediate-early genes (15,20,21), and it is as potent as the wild-type aFGF in inducing mesoderm formation in Xenopus animal caps (22). The mutant was found to be transported to the nuclear fraction similarly as the wild-type (15). We recently cloned a novel, intracellular protein, FIBP, which binds the wild-type aFGF, but not aFGF(K132E) (23).
The K132E mutation destroys a consensus phosphorylation site for protein kinase C (PKC) in aFGF, Ser 130 -Cys-Lys. aFGF was previously shown to be phosphorylated by PKC on serine in vitro, but not by protein kinase A or casein kinase I and only negligibly by casein kinase II (24). We found that wild-type aFGF was phosphorylated in vitro in a cell lysate while aFGF(K132E) was not. In vitro phosphorylation in cell lysate was augmented in the presence of diacylglycerol and phosphatidylserine and inhibited by staurosporine (15). Furthermore, externally added wild-type aFGF, but not aFGF(K132E), was phosphorylated in intact cells (15). Phosphorylation in vivo depended on translocation of aFGF to the cytosol or nucleus, because phosphorylation was not observed when aFGF was incubated with cells transfected with FGFR4 constructs incapable of mediating translocation to these locations. 3 Taken together, these data indicate that after translocation to the cytosol or nucleus aFGF becomes phosphorylated on serine 130 by PKC.
In attempts to elucidate the role of the phosphorylation site in the exposed loop, we made several mutations in this region and measured the ability of the mutated growth factors to bind to heparin and FGFR and to stimulate MAP kinase, DNA synthesis, and cell proliferation.

EXPERIMENTAL PROCEDURES
Plasmid Construction and Protein Purification-Constructs for in vitro transcription and translation encoding aFGF with different mutations (Fig. 1A) were generated by polymerase chain reaction-directed mutagenesis with the plasmid paFGF(K132E) (15) as template. Polymerase chain reaction products were cut with StuI and EcoRI and ligated into paFGF(K132E) between the same sites. Each construct contained a new restriction site for colony screening, and the sequence of each construct was verified by DNA sequencing. Constructs encoding maltose-binding protein (MBP) fusion proteins were generated by inserting the SphI-EcoRI fragment from each of these plasmids between the same sites in pMal-cN-aFGFCAAX, which had been previously constructed by inserting the NcoI-EcoRI fragment of pHBGF-cax (13) into the polylinker of pMal-cN (25). In some of the constructs a stop-codon had not been generated by polymerase chain reaction and was instead made by cutting the plasmids with EcoRI, filling in with T4 DNA polymerase and religating the plasmids. This procedure introduced an in-frame stop-codon and an AseI restriction site. MBP fusion proteins were affinity-purified on an amylose resin column as described by the manufacturer (New England Biolabs, Beverly, MA).
Cell Culture and Polyacrylamide Gel Electrophoresis in the Presence of Sodium Dodecyl Sulfate (SDS-PAGE)-Cells were propagated in DulbeccoЈs modified essential medium with 7.5% (v/v) fetal calf serum in a 5% CO 2 atmosphere at 37°C. SDS-PAGE was carried out with 7.5 or 10% gels as described by Laemmli (26) and processed as described (8).
In Vitro Transcription and Translation-Plasmids were linearized with EcoRI and transcribed in vitro with T3 RNA polymerase. Ethanolprecipitated transcripts were translated for 1 h 30 min at 30°C in a nuclease-treated rabbit reticulocyte lysate system (Promega, Madison, WI) in the presence of 1 M [ 35 S]methionine or unlabeled methionine. To be able to compare concentrations of translated growth factor in different unlabeled translation mixtures, labeled translations were always run in parallel. After translation the lysates were dialyzed against dialysis buffer (20 mM HEPES, pH 7.0, 140 mM NaCl, 2 mM CaCl 2 ) (27).
Measurement of DNA Synthesis and Cell Proliferation-DNA synthesis was measured by incorporation of [ 3 H]thymidine in serum-starved NIH 3T3 cells stimulated with in vitro translated, unlabeled growth factors for 24 h. Cell proliferation was measured in serum-starved NIH 3T3 cells after stimulation for 48 h with in vitro translated growth factors by counting the cells in a cell counter.
Heparin Affinity-In vitro translated, [ 35 S]methionine-labeled growth factors diluted in PBS:water ϭ 1:1 were incubated under rotation with 100 l of heparin-Sepharose at 4°C for 2 h. The beads were washed, and bound proteins were eluted with stepwise increasing concentrations of NaCl (1 ml of each concentration), and radioactivity in each fraction (1 ml) was measured.
Binding of Wild-type and Mutant aFGF to Cells-Calf pulmonary artery endothelial cells were incubated for 2 h at 4°C in HEPES medium supplemented with 10 units/ml heparin and 5 ng/ml [ 35 S]methionine-labeled wild-type or mutant aFGF in the absence or presence of excess unlabeled wild-type aFGF. The cells were washed three times, lysed, and the postnuclear supernatant was analyzed by SDS-PAGE and fluorography. In other experiments, 125 I-labeled wild-type aFGF (5 ng/ml) was incubated with NIH 3T3 cells for 3 h at 4°C in HEPES medium containing 10 units/ml heparin in the absence or presence of increasing concentrations of MBP fusion proteins with aFGF mutants or wild-type. In some cases, the fusion proteins had been cleaved with factor Xa. The cells were washed three times, lysed in 0.1 M KOH, and the radioactivity was measured.
Fractionation of Cells into Cytosol/Membrane and Nuclear Fraction-After lysis in lysis buffer (0.1 M NaCl, 10 mM Na 2 HPO 4 , 1% Triton X-100, 1 mM EDTA, pH 7.4, 1 mM phenylmethylsulfonyl fluoride, 4 g/ml aprotinin) cells were centrifuged for 15 min at 720 ϫ g. The supernatant was centrifuged again for 5 min at 15,800 ϫ g and designated the cytosol/membrane fraction. The pellet was washed once by resuspension in lysis buffer and once by resuspension in lysis buffer containing 0.3 M sucrose that was layered on lysis buffer containing 0.8 M sucrose and centrifuged at 720 ϫ g for 15 min at 4°C, then sonicated in lysis buffer containing additional 0.5 M NaCl and centrifuged for 5 min at 15,800 ϫ g. The supernatant of this centrifugation was designated the nuclear fraction.
Transport of Externally Added Growth Factor to the Nuclear Fraction-Serum-starved NIH 3T3 cells were incubated in DulbeccoЈs modified essential medium containing 1 mM unlabeled methionine, 10 units/ml heparin, and 2 l/ml [ 35 S]methionine-labeled mutant or wildtype growth factors for different periods of time. In some experiments 50 M LY294002 was also present. Cells were washed and treated with 5 mg/ml Pronase at 37°C for 5 min, transferred to an Eppendorf tube, and 1 mM phenylmethylsulfonyl fluoride and 4 g/ml aprotinin were added. Cells were washed once by resuspension in HEPES medium with the same protease inhibitors, lysed, and fractionated. Trichloroacetic acid precipitable material was analyzed by SDS-PAGE and fluorography.
MAP Kinase Activity-Serum-starved NIH 3T3 cells were incubated at 37°C for different periods of time in HEPES medium with different concentrations of in vitro translated wild-type or mutant aFGF or with a translation mixture without added mRNA in the presence of heparin (5 units/ml). The cells were washed twice with HEPES medium on ice and lysed in P-lysis buffer (10 mM Tris-HCl, pH 7.4, 50 mM NaCl, 5 mM EDTA, 50 mM NaF, 30 mM sodium pyrophosphate, 100 M sodium orthovanadate, 1% Triton X-100) with protease inhibitors. The postnuclear supernatant was subjected to immunoprecipitation with an anti-MAP kinase (ERK1 and ERK2) antibody (Zymed Laboratories Inc.), and the immunoprecipitates were subsequently incubated for 15 min at 37°C in 25 l of kinase buffer (20 mM Tris-HCl, pH 7.4, 10 mM MgCl 2 , 2 mM MnCl 2 , 1 mM dithiothreitol) with 1 Ci of [␥-32 P]ATP and 20 g of myelin basic protein. SDS-PAGE sample buffer was added, and the samples were analyzed by SDS-PAGE and autoradiography (28).

RESULTS
Generation of Mutants in the Exposed Loop of aFGF-Since aFGF(K132E) exhibits strongly reduced ability to stimulate DNA synthesis in cells (15,20,22) and is incapable of being phosphorylated in vitro and in vivo (15), we carried out a number of additional mutations in this region of the molecule (Fig. 1A). Serine 130 in the consensus sequence for phosphorylation by PKC was mutated to glutamic acid (S130E) or to alanine (S130A) to mimic phosphorylated and unphosphorylated growth factor, respectively. Similar mutations have been made previously in other proteins to study the role of phosphorylation (29 -32). To study if lysine 132 as such is required, we mutated it to alanine (K132A). Since this mutation destroys the PKC phosphorylation site, we also made the double mutant S130E/K132A as a "phosphorylated" control. We further made the conservative mutation of lysine 132 into arginine (K132R). Burgess et al. (22) made the double mutant cysteine 131 into serine and lysine 132 into glutamic acid (C131S/K132E) and reported that it was mitogenic to the same extent as the bovine wild-type aFGF. In this double mutant a new, potential PKC phosphorylation site is created (Ser 131 -Glu-Arg), and therefore we also made this double mutant to investigate if it would become phosphorylated.
In Vitro Phosphorylation and Heparin Affinity of the Mutants-Some of the mutants were expressed as fusion proteins with maltose-binding protein in bacteria and affinity-purified. These fusion proteins were then cleaved with the protease, factor Xa, to separate the MBP part from the growth factor part, and analyzed by an in vitro kinase assay with purified PKC. As shown in Fig. 1B, lower panel, lane 4, the wild-type growth factor was strongly phosphorylated. The mutants where serine 130 had been mutated to alanine (lane 1) or glutamic acid (lane 2) were not or negligibly phosphorylated, consistent with the notion that serine 130 is the major phosphorylation site in aFGF. The mutant C131S/K132E, where we had introduced a possible PKC site one amino acid away from the normal site, was also not phosphorylated (lane 3). In lanes 5 and 6 we used growth factor expressed as such in bacteria and purified as described previously (14,15) as controls. The pure, wild-type growth factor was phosphorylated to a similar extent as wild-type growth factor expressed as MBP fusion protein (compare lanes 4 and 6). With pure aFGF(K132E) a weak phosphorylated band was detectable (lane 5) similarly as with the other mutants. Since aFGF(K132E) is not phosphorylated in vitro in a cell lysate or in vivo in intact cells (15), the weak phosphorylation observed in lanes 1-3 and 5 probably represents background phosphorylation in the in vitro assay with pure PKC. In aFGF(K132R) the PKC consensus site is intact, since both lysine and arginine are tolerated at this position, and we found that aFGF(K132R) was indeed substrate for PKC (not demonstrated). aFGF(K132E) exhibits strongly reduced heparin affinity (20) as compared with wild-type aFGF. To measure heparin affinity of the different aFGF mutants, we bound in vitro translated, [ 35 S]methionine-labeled growth factors to heparin-Sepharose columns, washed, and eluted with stepwise increasing concen-trations of NaCl. The results, shown in Fig. 1C, demonstrate that the different mutants varied considerably in their affinity for heparin. The S130E mutant exhibited virtually identical heparin affinity as the wild-type, while the S130A mutant was eluted in a broad peak. This result was reproducible in four different experiments, but the reason for the unusual broadness of the peak is unclear. The mutants where lysine 132 had been changed all exhibited reduced heparin affinity. The most conservative mutation, K132R, reduced the affinity least, while the K132A mutation reduced heparin affinity considerably. We found that aFGF(K132E) eluted at 0.3 M NaCl (not demonstrated). It appears that the heparin affinity of aFGF is very sensitive to mutations of lysine 132 and that more drastic mutations have a more pronounced effect than conservative ones.
Ability of Mutant Growth Factors to Bind to Cells-To test the ability of the different mutants to bind to cells, in vitro translated, [ 35 S]methionine-labeled growth factors were incubated with calf pulmonary artery endothelial cells at 4°C in the presence of heparin and in the absence or presence of excess unlabeled wild-type aFGF. All mutants were bound to a similar extent and the binding could be completely competed out with excess aFGF in all cases ( Fig. 2A).
To measure relative binding affinities to cells of the wild-type and mutant growth factors, we incubated NIH 3T3 cells at 4°C with 125 I-labeled, pure wild-type aFGF in the presence of hep- arin and different concentrations of unlabeled wild-type or mutant aFGF as MBP fusion proteins. After washing, the cellassociated radioactivity was measured. Most of the tested mutants were able to compete with 125 I-labeled wild-type aFGF for binding to cells with a similar potency as the wild-type itself (Fig. 2B). However, it appears that the S130E mutant was somewhat more potent than the wild-type in this respect. Similar data were obtained when the MBP fusion proteins had been cleaved off with factor Xa prior to the experiment (not demonstrated).
Ability of the Different Mutants to Stimulate DNA Synthesis and Cell Proliferation-To study the mitogenic activity of the different mutants, we added increasing amounts of unlabeled, in vitro translated growth factors to serum-starved NIH 3T3 cells and measured the ability of the cells to incorporate [ 3 H]thymidine during 24 h of stimulation (Fig. 3A). The concentration of unlabeled growth factor was for each mutant estimated from parallel translation reactions containing [ 35 S]methionine, which were analyzed by SDS-PAGE and densitometric scanning. The data in Fig. 3A are corrected for differences in translation efficiency. The different mutants varied strongly in their ability to stimulate DNA synthesis. Wildtype aFGF and the S130E and S130A mutants were equally potent in this matter. The K132A and the S130E/K132A mutants were both approximately 10-fold less potent, while the K132R mutant was 3-fold less potent than wild-type to stimulate DNA-synthesis. For comparison, we also tested the K132E mutant, which was at least 100-fold less active than wild-type aFGF. aFGF(K132E) could not elicit a full mitogenic response within the limits of this assay using in vitro translated growth factors. As control, addition of translation mixture to which no mRNA had been added did not stimulate [ 3 H]thymidine incorporation at all (not demonstrated).
To study the ability of the different mutants to support proliferation, we incubated serum-starved NIH 3T3 cells for 48 h with different concentrations of wild-type and mutant growth factors that were made in vitro and then counted the cells in a cell counter (Fig. 3B). In this assay, one l/ml of translation mixture gave optimal stimulation of proliferation for all constructs tested, and increasing the concentration to 3 l/ml did not increase the effect reproducibly. Therefore, only results with the concentration of 1 l/ml is shown in Fig. 3C. Wild-type aFGF and the S130E mutant stimulated proliferation most effectively, while the S130A and the K132R mutants were somewhat less effective. The K132A and the S130E/ K132A mutants gave the weakest stimulation of proliferation. Together, these data indicate that phosphorylation of aFGF is not required for stimulation of DNA synthesis. It appears that lysine 132 is important for the mitogenic effect of aFGF, independent of phosphorylation of serine 130.

Stimulation of MAP Kinase by the Mutant Growth
Factors-Upon binding of aFGF to the FGFR, the tyrosine kinase of the receptor is activated and several signal transduction cascades are initiated (33)(34)(35). Activation of the Ras/Raf/Mek/MAP kinase cascade has been suggested to be important for stimulation of cell proliferation by FGF (34,36,37). Therefore, we measured the ability of the different mutants to stimulate MAP kinase. Serum-starved NIH 3T3 cells were stimulated with different concentrations of in vitro translated wild-type or mutant growth factor for 5, 30, or 60 min, lysed, and MAP kinase (ERK1 and ERK2) was immunoprecipitated. MAP kinase activity in the immunoprecipitates was measured in an in vitro kinase assay with myelin basic protein as substrate.
There was virtually no MAP kinase activity immunoprecipitated from cells treated with a translation mixture to which no mRNA had been added (Fig. 4, lane 1). Ten percent fetal calf serum (lane 2) and the higher concentration used of wild-type aFGF (lane 4) stimulated MAP kinase activity to a similar extent. The aFGF induced MAP kinase activation was completely prevented when the cells were treated with genistein, a tyrosine kinase inhibitor (lanes 15 and 16, upper panel). Importantly, no detectable differences in MAP kinase activity could be measured when the cells were stimulated with the  3-14). The apparent differences after 5-min stimulation in the experiment shown were not reproducible between experiments. Also the kinetics of MAP kinase activation was similar for the different aFGF constructs (compare lanes 3-14 in upper, middle, and lower panels). Together, all the aFGF constructs tested stimulated MAP kinase activation to a similar extent.
Transport of Mutant Growth Factors to the Nuclear Fraction-We and others have previously provided evidence suggesting a role of aFGF in the nucleus for stimulation of DNA synthesis (8, 10, 14 -18, 38). Therefore, we examined the ability of the different aFGF mutants to be transported to the nuclear fraction. [ 35 S]Methionine-labeled, in vitro translated wild-type and mutant growth factors were added to cells together with heparin and incubated at 37°C for different periods of time. The cells were washed, treated with Pronase to digest growth factor at the cell surface or bound to the plastic, then the cells were lysed and fractionated into cytosol/membrane and nuclear fractions as described under "Experimental Procedures." In Fig. 5 only the nuclear fractions are shown. After 9-h incubation of the cells with growth factor, all mutants as well as the wild-type aFGF were found in the nuclear fraction (Fig. 5A). Transport to the nuclear fraction could be completely inhibited by the phosphoinositide 3-kinase inhibitor LY294002 (Fig. 5B), which we have found recently, is able to inhibit transport of externally added aFGF to the cytosol and nucleus of cells. 4 It was reported recently that in retinal cells, phosphorylated aFGF was more stable intracellularly than unphosphorylated aFGF (39). We also tested the different mutants for presence in the nucleus after longer incubation times, but no reproducible differences could be detected. However, in the cytosol/membrane fraction a rather stable degradation product of approximately 14 kDa was reproducibly obtained from cells incubated with wild-type aFGF and with the S130E mutant, but not from cells incubated with the other mutants (not demonstrated). The degradation product was probably enclosed within a vesicular compartment, since it was also obtained in the presence of LY294002. The degradation product was never obtained from the nuclear fraction.
Based on the ability to bind to wild-type aFGF, but not to the K132E mutant, we recently cloned a novel intracellular protein, which we designated FIBP (aFGF Intracellular binding protein) (23). We tested the ability of several of the aFGF mutants as MBP fusion proteins to bind to FIBP in vitro (23). In contrast to the K132E mutant, all of the other mutants that we tested in this way (S130E, K132A, S130E/K132A) bound to FIBP, although somewhat less effectively than did the wildtype (not demonstrated).

DISCUSSION
In this paper we report the generation and characterization of several mutations in an exposed loop of aFGF, known to be of importance for heparin affinity and mitogenic activity of the growth factor. The mutants varied 10-fold in mitogenic activity, but were equally potent in stimulating MAP kinase. A correlation, however, could be observed between mitogenic activity and affinity to heparin.
High affinity receptor binding varied little among the constructs, with the exception of the S130E mutant, which apparently bound with higher affinity. Interestingly, bFGF phosphorylated by protein kinase A on threonine 112 (corresponding to Thr 120 according to the numbering used in Ref. 1) was found to be three to eight times more potent at displacing 125 I-labeled FGF from the high affinity receptors (24). This threonine is located in a part of bFGF reported to be involved in both heparin binding and high affinity receptor binding (40). Possibly, phosphorylation of aFGF may modulate its binding to FGFR.
The Ras/Raf/Mek/MAP kinase cascade is considered to be important for mitogenic activity of several growth factors, FGF included (41). Microinjection of an antibody against FRS2, a tyrosine-phosphorylated protein that links FGFR activation to activation of the Ras/Raf/Mek/MAP kinase cascade, inhibited stimulation of DNA synthesis by aFGF (34). A cell-permeable peptide derived from the Grb2-binding sequence of epidermal  5. Transport of the different mutants to the nuclear fraction. A, serum-starved NIH 3T3 cells were incubated at 37°C for 9 h in the presence of 10 units/ml heparin and 2 l/ml [ 35 S]methionine-labeled growth factors, washed, and treated with 5 mg/ml Pronase at 37°C for 5 min. The cells were washed in the presence of protease inhibitors, lysed, and the nuclear fraction was collected by centrifugation. After sonication, the trichloroacetic acid-precipitable material in the nuclear fraction was analyzed by SDS-PAGE and fluorography. B, serumstarved NIH 3T3 cells were treated with wild-type aFGF as in A for 10 h or 28 h. In lanes 1 and 2, 50 M LY294002 was present. The cells were further processed and analyzed as in A.
growth factor receptor was shown to inhibit MAP kinase activation in cells stimulated with epidermal growth factor or with platelet-derived growth factor, but not in cells stimulated with bFGF. The peptide also inhibited the mitogenic response of the cells to epidermal growth factor and to platelet derived growth factor, but not to bFGF (42). However, even though the different aFGF mutants described in the present paper differed 10-fold in mitogenic activity, they were equally potent in stimulation of MAP kinase. We have also found that the K132E mutant, which exhibits at least a 100-fold lower mitogenic activity than wild-type aFGF, shows a similar potency as the wild-type to induce MAP kinase activity. 2 It therefore appears that activation of the Ras/Raf/Mek/MAP kinase cascade is not sufficient for stimulation of DNA synthesis by aFGF.
We previously provided data suggesting that the strongly reduced mitogenic activity of the K132E mutant is due to a defect in its intracellular action (15). We also found that the K132E mutation rendered the growth factor incapable of being phosphorylated in vivo (15). One of the aims of the present study was to investigate whether there could be a cause-effect relationship between these observations. To use phosphorylated wild-type growth factor in experiments that last for 24 h or more, may not be optimal (24). Therefore, we took the strategy of mutating the serine in the phosphorylation site to the negatively charged glutamic acid or to the uncharged alanine to mimic phosphorylated and unphosphorylated growth factor, respectively. Similar strategies have previously been used successfully with other proteins (29 -32). However, we have no positive control to tell us whether the S130E mutant really mimics phosphorylated aFGF. Anyhow, we can conclude that inability of being phosphorylated can not be the only reason for the mitogenic defect of the K132E mutant. Rather, it appears that lysine 132 as such is important, independent of phosphorylation.
Interaction of aFGF (5) and bFGF (43) with heparin or heparan sulfate proteoglycans has been proposed to play a pivotal role in stimulation of FGFR. In this study, we have found a correlation between mitogenic potency and heparin affinity of the different mutants. The leading hypothesis on how heparin or heparan sulfate proteoglycans exert their effect is that these molecules bind several FGF molecules at the same time as each FGF molecule binds to an FGFR molecule, thereby dimerizing or oligomerizing FGFR molecules (5,44,45). If the reason for the reduced mitogenic activity of the mutants found in this and previous studies (15,20) were reduced heparin affinity, a proportional reduction in activation of FGFR and its downstream effectors should be expected from this hypothesis. This is not the case. One possible explanation could be that interaction with intracellular targets of aFGF also depends on the ability of aFGF to interact with negatively charged (parts of) molecules. In this context, it is interesting that glypican, a heparan sulfate proteoglycan, and biglycan, a chondroitin sulfate proteoglycan, have been found to contain functional nuclear localization sequences and to exist in the nucleus (46).
In conclusion, phosphorylation of aFGF is not required for its mitogenic activity. Lysine 132 of aFGF is important both for mitogenic effect and for heparin affinity. We cannot detect a correlation between mitogenic activity and ability to activate MAP kinase.