Novel nuclear localization and potential function of insulin-like growth factor-1 receptor/insulin receptor hybrid in corneal epithelial cells

Abstract

Background: Type I insulin-like growth factor receptor (IGF-1R) and insulin receptor (INSR) are highly homologous molecules, which can heterodimerize to form an IGF-1R/INSR hybrid (Hybrid-R). The presence and biological significance of the Hybrid-R in human corneal epithelium has not yet been established. In addition, while nuclear localization of IGF-1R was recently reported in cancer cells and human corneal epithelial cells, the function and profile of nuclear IGF-1R is unknown. In this study, we characterized the nuclear localization and function of the Hybrid-R and the role of IGF-1/IGF-1R and Hybrid-R signaling in the human corneal epithelium.

Methodology/principle findings: IGF-1-mediated signaling and cell growth were examined in a human telomerized corneal epithelial (hTCEpi) cell line using co-immunoprecipitation, immunoblotting and cell proliferation assays. The presence of Hybrid-R in hTCEpi and primary cultured human corneal epithelial cells was confirmed by immunofluorescence and reciprocal immunoprecipitation of whole cell lysates. We found that IGF-1 stimulated Akt and promoted cell growth through IGF-1R activation, which was independent of the Hybrid-R. The presence of Hybrid-R, but not IGF-1R/IGF-1R, was detected in nuclear extracts. Knockdown of INSR by small interfering RNA resulted in depletion of the INSR/INSR and preferential formation of Hybrid-R. Chromatin-immunoprecipitation sequencing assay with anti-IGF-1R or anti-INSR was subsequently performed to identify potential genomic targets responsible for critical homeostatic regulatory pathways.

Conclusion/significance: In contrast to previous reports on nuclear localized IGF-1R, this is the first report identifying the nuclear localization of Hybrid-R in an epithelial cell line. The identification of a nuclear Hybrid-R and novel genomic targets suggests that IGF-1R traffics to the nucleus as an IGF-1R/INSR heterotetrameric complex to regulate corneal epithelial homeostatic pathways. The development of novel therapeutic strategies designed to target the IGF-1/IGF-1R pathway must take into account the modulatory roles IGF-1R/INSR play in the epithelial cell nucleus.

Figure 1. IGF-1 induces the canonical IGF-1R/Akt signaling pathway.

(A) Akt phosphorylation induced by IGF-1 and insulin. hTCEpi cells were starved for 24 h in KBM-2 culture medium and then stimulated with IGF-1 (100 ng/ml) or insulin (100 ng/ml) for 15 min. Whole cell lysates were immunoblotted (IB) with anti-phospho-IGF-1Rβ, anti-IGF-1Rβ, anti-INSRβ, anti-phospho-Akt, anti-Akt, or anti-β-actin (loading control) as indicated. (B) IGF-1-induced IGF-1R phosphorylation. Protein lysates obtained as above were immunoprecipitated (IP) with anti-IGF-1Rβ (CST#3027) or anti-INSRβ (C-19) and phosphorylation of both receptors was detected on immunoblots with anti-PY20. Blots are representative of at least two repeated experiments.

Figure 2. Presence of IGF-1 receptor and insulin receptor (IGF-1R/INSR) hybrid in corneal epithelium.

(A) Reciprocal immunoprecipitation (IP) of IGF-1 receptor and insulin receptor in the hTCEpi cell line. Whole cell lysates were subjected to IP with antibodies against either the IGF-1 receptor (IGF-1R) β-subunit (CST#3027) or the insulin receptor (INSR) β-subunit (C-19). Immunoprecipitated products were then immunoblotted with the same antibodies. Input represents the whole cell lysates before IP. (B) Co-precipitation of IGF-1R/INSR hybrid in primary corneal epithelial cell cultures. Protein lysates from passage 3 of human corneal epithelial cells were equally immunoprecipitated with anti-IGF-1Rβ (CST#3027) or anti-INSRβ (C-19), and detection of IGF-1R/IGF-1R or IGF-1R/INSR was performed with immunoblots with the same antibody against IGF-1Rβ.

Figure 3. IGF-1 activates the IGF-1R/INSR hybrid.

(A) Existence of IGF-1R and INSR αβ-dimers after reducing immunoprecipitation. hTCEpi lysates were reduced with DTT to break the receptors into αβ-dimers. Reduced lysates were then immunoprecipitated with rabbit polyclonal anti-IGF-1Rβ (CST#3027) or anti-INSRβ (C-19). Top panel: Immunoblot of INSRβ immunoprecipitates from reduced lysates with anti-IGF-1Rβ showed no or faint bands corresponding to the IGF-1Rβ. Bottom panel: Immunoblot of IGF-1Rβ immunoprecipitates from reduced lysates with anti-INSRβ showed no or faint bands corresponding to the INSRβ. (B) Activation of IGF-1R and INSR in hTCEpi cells by IGF-1 stimulation. Protein lysates of hTCEpi cells after stimulation by IGF-1 (100 ng/ml) or insulin (100 ng/ml) for 15 min were collected and reduced by DTT and then subjected to reducing IP. IGF-1, but not insulin, induced phosphorylation of IGF-1 receptor (Top panel) and insulin receptor (Bottom panel) detected on immunoblots with anti-PY20.

Figure 4. IGF-1 promotes hTCEpi cell proliferation.

(A) Neutralization of IGF-1 receptor and inhibition of IGF-1R/PI3K/Akt pathway. hTCEpi cells were starved for 24 h and then cultured with 5 µg/ml of IGF-1R neutralizing antibody (αIR3) or mouse IgG (negative control) for 4 h or with 50 µM of PI3K inhibitor (LY294002, denoted by LY) for 1 h, followed by stimulation with IGF-1 (100 ng/ml) for 15 min. Whole cell lysates were immunoblotted with antibodies as indicated. Antibody against β-actin was used as a loading control. (B) IGF-1 promoted the growth of hTCEpi cells. hTCEpi cells were starved for 24 h and then cultured with 10 ng/ml of IGF-1 or insulin in the absence or presence of either αIR3 (1 µg/ml) or IGFBP-3 (2 µg/ml) for 4 days. Total DNA content was then assayed. Results are the mean fluorescent signal ± s.e. of three independent experiments. *The mean value was significantly different from that obtained in the control group (P<0.05); **the mean fluorescent signal with αIR3 or IGFBP-3 was significantly lower than that with IGF-1 (P<0.05); the mean value with insulin was not significantly higher than that obtained in the control group.

Figure 5. Co-localization of IGF-1R and INSR.

(A) Confocal images showing hTCEpi cells labeled with antibodies against IGF-1Rβ (CST#3027) (panel I, revealed by green), INSRβ (CT-3) (panel II, revealed by red), or both (panel III, merged by panels I and II); nuclei were counterstained with PI (panel IV, revealed by blue) and overlap of the three channels is shown in panel V. Scale: 20.41 µm. (B) Confocal images showing hTCEpi cells labeled with antibodies against IGF-1Rα (24–31) (panel I, revealed by red), INSRβ (C-19) (panel II, revealed by green), or both (panel III, merged by panels I and II); nuclei were counterstained with PI (panel IV, revealed by blue) and overlap of the three channels is shown in panel V. Scale: 15.89 µm. Orange or Yellow color in merged panels III denotes the presence of both IGF-1R and INSR immunoreactivity; white color in merged panels V represents the co-localization of both IGF-1R and INSR in the nucleus. (C) Imaris analysis of co-localization. Confocal images of hTCEpi cells labeled with antibodies as described above were further analyzed by Imaris software. The 3D reconstruction of a two-dimensional Z-stack of images illustrates the subcellular distribution of IGF-1Rβ (green) and INSRβ (red) (Panel I) or IGF-1Rα (red) and INSRβ (green) (Panel IV) in hTCEpi cells. Panel II and V, co-localized voxels of panel I and IV, respectively. Scale: 3 or 5 µm. Panel III or VI shows the representative two-dimensional histogram of both signals for either panel I or IV. The red signal is on the x axis and the green is on the y axis. Automatic selected co-localized areas are shown by a rectangular yellow overlay. Data are representative of three independent experiments with 8 to10 cells imaged per experiment.

Figure 6. Subcellular localization of IGF-1R/IGF-1R and IGF-1R/INSR hybrid.

(A) Presence of IGF-1R/INSR hybrid in the nucleus. Cytosolic (Cyto) and nuclear (Nuclear) fractions of hTCEpi cells were immunoprecipitated with anti-IGF-1Rβ (CST#3027) or anti-INSRβ (C-19) and then immunoblotted with the same antibodies. Bottom left: Immunoblot of IGF-1R immunoprecipitates with anti-INSR indicated the presence of hybrid in the nuclear fraction. (B) Exclusive presence of hybrid in the nucleus. Whole cell lysates (WCL) and nuclear fractions (Nuclear) of hTCEpi cells were subjected to immunoprecipitation with antibodies against IGF-1Rα (αIR3, which reacts with IGF-1R/IGF-1R) or INSRβ (C-19, which reacts with IGF-1R/INSR and INSR/INSR). Immunoprecipitates were then immunoblotted with anti-IGF-1Rβ. Hybrid-R was shown in the WCL and nuclear fractions but IGF-1R/IGF-1R was only detected in the WCL. Controls for cross-contamination between each compartment were confirmed by immunoblotting with antibodies against GAPDH (cytosolic extract) and SP1 (nuclear extract).

Figure 7. Knockdown of IGF-1R/INSR.

Control or INSR siRNA were transfected into hTCEpi cells. Two to three days after transfection, WCL, cytosolic (Cyto) and nuclear fractions (Nuclear) were collected. (A) Immunoblots of WCL with antibodies against INSR (C-19), IGF-1R (CST#3027), or β-actin (loading control). (B) WCL of hTCEpi cells were immunoprecipitated with antibodies αIR3, C-19, or MA-20 (against INSRα and only reacts with INSR/INSR). Immunoprecipitates were then immunoblotted with anti-IGF-1Rβ (CST#3027) or anti-INSRβ (C-19). (C) Cytosolic and nuclear fractions of hTCEpi cells were subjected to immunoprecipitation with antibodies αIR3 or C-19. Immunoprecipitates were then immunoblotted with anti-IGF-1Rβ (CST#3027). GAPDH is the marker for cytosolic extract; and SP1 is the marker for nuclear extract.