Identification of the amino acids 300-600 of IRS-2 as 14-3-3 binding region with the importance of IGF-1/insulin-regulated phosphorylation of Ser-573

Abstract

Phosphorylation of insulin receptor substrate (IRS)-2 on tyrosine residues is a key event in IGF-1/insulin signaling and leads to activation of the PI 3-kinase and the Ras/MAPK pathway. Furthermore, phosphorylated serine/threonine residues on IRS-2 can induce 14-3-3 binding. In this study we searched IRS-2 for novel phosphorylation sites and investigated the interaction between IRS-2 and 14-3-3. Mass spectrometry identified a total of 24 serine/threonine residues on IRS-2 with 12 sites unique for IRS-2 while the other residues are conserved in IRS-1 and IRS-2. IGF-1 stimulation led to increased binding of 14-3-3 to IRS-2 in transfected HEK293 cells and this binding was prevented by inhibition of the PI 3-kinase pathway and an Akt/PKB inhibitor. Insulin-stimulated interaction between endogenous IRS-2 and 14-3-3 was observed in rat hepatoma cells and in mice liver after an acute insulin stimulus and refeeding. Using different IRS-2 fragments enabled localization of the IGF-1-dependent 14-3-3 binding region spanning amino acids 300-600. The 24 identified residues on IRS-2 included several 14-3-3 binding candidates in the region 300-600. Single alanine mutants of these candidates led to the identification of serine 573 as 14-3-3 binding site. A phospho-site specific antibody was generated to further characterize serine 573. IGF-1-dependent phosphorylation of serine 573 was reduced by inhibition of PI 3-kinase and Akt/PKB. A negative role of this phosphorylation site was implicated by the alanine mutant of serine 573 which led to enhanced phosphorylation of Akt/PKB in an IGF-1 time course experiment. To conclude, our data suggest a physiologically relevant role for IGF-1/insulin-dependent 14-3-3 binding to IRS-2 involving serine 573.

Figure 1. Sequence alignments of IRS-1 and -2 species.

Protein sequences from mouse IRS-2 (NP_001074681.1), human IRS-2 (NP_003740.2), mouse IRS-1 (NP_034700.2) and human IRS-1 (NP_005535.1) were aligned. The sequences were searched for homologue residues of mouse IRS-2 that were identified by mass spectrometry and marked in bold red. Shown are only parts of the sequences that contain an identified phosphorylated serine/threonine residue. The term “numbering” indicates the amino acid position of the first amino acid in every line.

Figure 2. Co-immunoprecipitation and overlay assays indicate interaction of 14-3-3 and IRS-2 upon IGF-1/insulin stimulation.

A. HEK293 cells were transiently transfected with GFP or GFP-IRS2 and after serum starvation cells were incubated with 50 ng/ml IGF-1 for 30 min or after preincubation with 100 nM wortmannin for 30 min. 400 µg total protein was used for immunoprecipitation with 14-3-3 antibody (C-17) and samples were separated on 5–15% gradient gel. Upper membrane was incubated with IRS-2 antibody, lower membrane with 14-3-3 antibody (K-19). B. HEK293 cells were transfected with GFP-IRS2 and stimulation was carried out after starvation for serum overnight with 50 ng/ml IGF-1 for 30 min or subsequently after preincubation with 1 µM PI-103 for 30 min. 250 µg of total protein was pulled down using GFP-Trap®. SDS-PAGE followed transfer onto nitrocellulose membranes. Overlay assay followed stripping of the membrane and reprobing with GFP antibody as loading control. C. Extent of interaction was quantified by scanning densitometry of blots and normalization for GFP-IRS2 serum starved condition (mean ± SEM; n = 4; *p<0.05 serum starved vs. IGF-1 or IGF-1 vs. PI-103/IGF-1). D. Male C57Bl/6 mice were fasted overnight and injected intravenously with 2 IU (international units) insulin. After 10 min liver was taken and 500 µg of total protein was immunoprecipitated with IRS-2 antibody. After performing overlay assay membrane was stripped and reprobed with IRS-2 antibody as loading control. Two mice of each group are shown. E. Densitometric analyses of 14-3-3 interaction with IRS-2. Overlay signal was normalized for total IRS-2 protein content (mean ± SEM; n = 4; *p<0.05 fasted vs. insulin). F. Male C57Bl/6 mice were fasted overnight, refed for 4 hours or injected intraperitoneally with insulin for 30 min. Procedure as in G. Four mice of each group are shown. G. 14-3-3 interaction with IRS-2 was quantified by scanning densitometry of immunoblots and normalization for IRS-2 protein (mean ± SEM; n = 4; *p<0.05 fasted vs. refed and insulin stimulation).

Figure 3. The area spanning amino acids 301–600 on IRS-2 is the 14-3-3 binding region.

A. Schematic illustration of truncated IRS-2 constructs to identify the 14-3-3 binding region. B. 20 µg of protein was separated on a 5–15% gradient gel and membrane was probed with GFP antibody to check expression and molecular weight of truncated IRS-2 versions. The arrow indicates a longer exposure time. C. HEK293 cells were transfected with either GFP-IRS2 or truncated versions of IRS-2 (GFP-IRS2-1-300, GFP-IRS2-1-600, GFP-IRS2-301-1321, GFP-IRS2-601-1321, GFP-IRS2-301-600), starved for serum overnight and stimulated for 30 min with 50 ng/ml IGF-1 or subsequently after preincubation with 1 µM PI-103 for 30 min. With GFP-Trap® 250 µg protein was pulled down and samples were subjected to overlay assay. For loading and expression control membranes were stripped and reprobed for GFP.

Figure 4. Ser-573 of IRS-2 is an IGF-1-dependent 14-3-3 binding site.

A. HEK293 cells were transiently transfected with GFP-IRS2, GFP-IRS2-S303A, GFP-IRS2-T401A, GFP-IRS-2-T517A, GFP-IRS2-S556A or GFP-IRS2-S573A and stimulated with 50 ng/ml IGF-1 for 30 min or after preincubation with 1 µM PI-103 30 min prior IGF-1 stimulation. 250 µg of total protein were used for GFP pulldown and overlay assay was performed. Stripping followed reprobing of the membrane with GFP antibody as loading and expression control. B. Transiently with GFP-IRS2 or GFP-IRS2-S573A transfected HEK293 cells were stimulated with 50 ng/ml IGF-1 alone for 30 min or after preincubation with 100 nM wortmannin for 30 min. 100 µg of total protein was used for GFP pulldown, separated on 5–15% SDS gels and overlay assay was performed. Corresponding GFP reblot as expression and loading control is shown C. Flp-In HEK293 cells stably expressing GFP-IRS2 or GFP-IRS2-S573A were stimulated as in A and cell lysis was followed by pull down of 400 µg total protein. Samples were divided into two equal volumes and separated on SDS-PAGE. One membrane was used for overlay assay, the other one was directly incubated with GFP antibody. D. Interaction of 14-3-3 with IRS-2 from B was quantified scanning densitometry of Western blots (mean ± SEM; n = 3; *p<0.05 IRS-2 wild type IGF-1 vs. IRS-2 wild type wortmannin/IGF-1).

Figure 5. Sequence alignments and characterization of a polyclonal antibody raised in sheep against position Ser-573 on IRS-2.

A. Sequence alignment of the amino acids adjacent to the 14-3-3 binding site Ser-573 of IRS-2 in different species. B. Shown is the amino acid sequence surrounding Ser-573 in mouse IRS-2 and the sequence surrounding the homologue position Ser-522 in mouse/rat IRS-1. C. Flp-In HEK293 cells stably expressing GFP-IRS2 or GFP-IRS2-S573A were starved for serum followed by stimulation for 30 min with 50 ng/ml IGF-1 or after preincubation with 100 nM wortmannin for 30 min. 200 µg of lysate was separated on a 7.5% SDS gel, membrane was probed with p-Ser-573 antibody. Expression was checked by stripping the membrane and reprobing with GFP antibody. D. HEK293 cells were co-transfected transiently with mouse IRS-2/IR (insulin receptor), human IRS-2/IR and rat IRS-1/IR and stimulated with 10 nM insulin for 30 min. Membranes were incubated with p-Ser-573 antibody and reprobed with the corresponding protein antibodies. IR expression was also checked.

Figure 6. Inhibition of Akt/PKB reduces IGF-1-induced phosphorylation of serine 573 and 14-3-3 binding.

A. Flp-In HEK293 cells stably expressing GFP-IRS2 were starved for serum and incubated for 30 min with 50 ng/ml IGF-1 or 1 µM Akti-1/2 alone, or Akti-1/2 preincubation followed IGF-1 stimulation. 100 µg of total protein was separated on 7.5% SDS gels and membranes were probed with p-Ser-573 and p-Thr-308 of Akt/PKB. Membranes were stripped and reprobed with respective antibodies for detection of protein levels. B. Effect of Akt/PKB inhibition on serine 573 phosphorylation was assessed by scanning densitometry of blots and normalization for protein (mean ± SEM; n = 3; *p<0.05 IGF-1 vs. Akti/IGF-1). C. GFP pulldown from Flp-In HEK293 cells stably expressing GFP-IRS2, stimulated as in A. 200 µg protein was pulled down and overlay assay followed reprobing with IRS-2 antibody. 100 µg of total protein was separated on 7.5% SDS gel and membrane was checked for p-Thr-308 phosphorylation and reprobed with Akt/PKB protein antibody.

Figure 7. Insulin induces binding of 14-3-3 to endogenous IRS-2 and phosphorylates Ser-573 on IRS-2.

A. Fao cells were starved for serum overnight and incubated for 30 min with either 10 nM insulin or 50 ng/ml IGF-1. 250 µg protein was immunoprecipitated with IRS-2 antibody and separated on a 5–15% gradient gel. Overlay assay followed stripping and reprobing with IRS-2 antibody as loading control. Successful stimulation is shown as phosphorylation of p-Thr-308 and corresponding Akt/PKB reblot. B. Fao cells were starved for serum overnight and stimulated with 10 nM insulin for the indicated time points. 100 µg of protein was separated on a 7.5% gel and membranes were probed with specific antibodies against p-Ser-573 of IRS-2 and p-Thr-308 of Akt/PKB. For loading control membranes were stripped and reprobed with protein antibody.

Figure 8. Ser-573 influences phosphorylation of Akt/PKB.

A. HEK293 cells transiently expressing GFP-IRS2 or GFP-IRS2-S573A were stimulated with 50 ng/ml IGF-1 for the indicated time points. 40 µg of total protein was separated on 7.5% SDS gels and membranes were incubated with GFP antibody to ensure equal expression levels and with p-Thr-308 and p-Ser-473 antibody respectively. Corresponding Akt/PKB reblots are shown. B. Densitometric analyses of Akt/PKB phosphorylation. Black diamonds represent IRS-2 wild type, white squares IRS2-S573A mutant. Phosphorylation was normalized against total protein and IRS-2 wild type stimulated with IGF-1 for 5 min was set as 1 (mean ± SEM; n = 3; *p<0.05 IRS-2 120 min vs. S573A 120 min; IRS-2 240 min vs. S573A 240 min).

Figure 9. GST pulldown of transiently transfected IRS-2 and mutants.

A and B. GFP-IRS2, GFP-IRS2-S573A and GFP-IRS2-S556A/S573A were expressed transiently in HEK293 cells and cells were stimulated with 50 ng/ml IGF-1 for 30 min or after preincubation with 100 nM wortmannin for 30 min. Lysates were incubated with 2 µg GST-14-3-3β or ε for 2 hours. Samples were separated on 7.5% SDS gels and membranes were probed with GFP and GST antibodies.