Research paperInsulin-like growth factor binding protein-2: NMR analysis and structural characterization of the N-terminal domain
Highlights
► Determined NMR peak assignments for N-domain of human IGFBP-2 (N-BP-2). ► NMR data indicate N-BP-2 structure is similar to N-domain of other IGFBPs. ► N-BP-2 has a Pro/Ala-rich loop that differs from other IGFBPs. ► NMR data show that the loop adopts two distinct conformations. ► The loop is a potential SH3 domain binding site.
Introduction
IGF signalling plays a key role in cellular proliferation, survival, differentiation and senescence, and dysregulation of the signalling pathway has been implicated in the development and progression of many human diseases including cancer, diabetes and atherosclerosis. There is therefore considerable interest in the development of IGF-based therapeutics for these diseases [1], [2], [3], [4]. For example, IGFBP-2 is markedly over-expressed in malignancies such as glioma [5], prostate cancer [6], lung cancer [7], [8], colorectal cancer [9], [10], ovarian cancer [11], adrenocortical tumours [12], breast cancer [13], [14] and leukaemia [15], although it is unclear what role IGFBP-2 plays in the development and progression of human cancers [16]. This is complicated by the fact that IGFBPs act through both IGF-dependent and IGF-independent mechanisms [17], [18]. The pathways that mediate IGF-independent activities have not been fully elucidated but can involve interactions with other proteins [19], [20].
The pro-survival and mitogenic affects of IGFs are modulated by a family of six IGF-binding proteins (IGFBPs), which bind tightly to IGF-I and -II (Kd ∼ 10−9 M), inhibiting their interaction with the IGF type 1 receptor (IGF-IR). The affinity of IGFBPs for IGFs is regulated by several mechanisms, including proteolysis, phosphorylation and binding to the extracellular matrix [1], [21], [22]. IGFBPs consist of conserved N- and C-terminal domains (33–46% identity) connected by a non-conserved linker region (l-domain) [1], [22]. IGFBPs 1–5 have 18 conserved cysteines and a common disulfide bond pattern, while IGFBP-6 lacks two of the first eight N-terminal cysteines conserved in other IGFBPs [23].
Both the N- and C-domains of IGFBPs are involved in high affinity IGF binding. The isolated N- and C-domains of IGFBPs bind IGFs with 3-fold to 1000-fold lower affinity than full-length IGFBPs, and IGF affinity varies among the N- and C-domains of different IGFBPs [24], [25], [26], [27], [28], [29]. The C-domain of IGFBP-2 binds IGFs with higher affinities and lower dissociation rates than the N-domain [24]. In contrast, the C-domains of IGFBP-4 [30], -5 [31] and -6 [26] exhibit lower affinities than their N-domains. Co-incubation of the N- and C-domains of IGFBP-3 with IGF-I significantly enhances IGF-binding affinity to about 5–10-fold lower than that of the intact IGFBP [27]. In contrast, enhanced binding was not observed for the N- and C-domains of IGFBP-2 [24] and 6 [26]. The L-domain is predicted to be unstructured and proteolytic cleavage within this domain reduces binding affinity and releases IGF, which can then bind to the IGF-IR [32].
The three-dimensional (3D) structure of a full-length IGFBP has not yet been determined, although structures are available for individual N- or C-domains of several IGFBPs. The structures of the free and IGF-bound forms of the N-domain of IGFBP-4 [33], [34], mini-BP-5 (residues 40–92 of the N-terminal domain of IGFBP-5) [31] and the unique N-terminal region of IGFBP-6 [35], as well as the C-domains of IGFBP-1 [33], 2 [36], 4 [33], [34] and 6 [37], have been solved. The crystal structure of the ternary complex comprising the N- and C-domains of IGFBP-4 bound to IGF-I has also been reported [33]. The individual N- and C-domains of various IGFBPs display a high degree of sequence and structural homology despite significant differences in their affinity and specificity for IGFs.
To gain a more detailed understanding of the factors underlying these differences, we have undertaken a detailed study of IGFBP-2 and its interactions with IGF-I. We aim to determine the contribution made by each domain (N-, C- and linker domains) to the affinity and specificity of IGFBP-2 for IGF-I. Toward this goal, we previously reported the structure of the C-domain of human IGFBP-2 (C-BP-2) and identified key residues of IGF-I and C-BP-2 that were involved in binding [36], [38]. Herein, we describe detailed studies of the N-domain of IGFBP-2 (N-BP-2).
In order to probe the interaction with IGF-I by NMR, resonance assignments are required for N-BP-2. This proved to be challenging because of the significant variation of peak intensities in the two-dimensional heteronuclear single quantum coherence (HSQC) spectrum, attributed largely to conformational averaging in certain parts of the protein and to non-specific aggregation. Recording spectra at lower N-BP-2 concentrations resulted in more uniform peak intensities in the HSQC spectrum. The acquisition of a complete dataset required for structural characterization of the protein at such low protein concentrations, however, would have required excessive NMR acquisition times. To overcome this we employed non-uniform sampling (NUS) to acquire the NMR spectra; data acquired in this way are not amenable to conventional Fourier transform analysis and were therefore processed using the maximum entropy method (MEM) [39], [40], [41].
Section snippets
Preparation of 15N/13C-labelled N-domain of IGFBP-2 (N-BP-2)
Uniformly 15N- and 13C/15N-labelled N-domain of human IGFBP-2 (N-BP-2, residues 1–98) were prepared as described previously [38]. Briefly, E. coli BL21 (λDE3) cells were transformed with a pET32a(+) vector containing cDNA encoding the N-domain of IGFBP-2 (residues 1–98) and an N-terminal HRV 3C protease cleavage site, and were grown in M9 minimal medium containing ampicillin (50 mg/ml) with 15NH4Cl and U-13C6 d-glucose as the sole nitrogen and carbon sources, respectively. Cells were grown at
Characterization of N-BP-2
The N-domain construct (N-BP-2) described in these studies consisted of residues 1–98 of the mature full-length IGFBP-2 protein (UniProt P18065) and included N-terminal Gly-Pro residues resulting from cleavage with 3C protease. The protein was purified to homogeneity and ran predominately as a single band (>95% purity) on an SDS-PAGE gel under reducing conditions (Supplementary material, Fig. S1A). The experimental mass for 13C/15N-labelled N-BP-2 (10 761.8 Da) determined by matrix-assisted
Discussion
The development of anti-cancer drugs based on modulating the activity of individual IGFBPs depends on gaining a detailed knowledge of their mechanism of action. In this study we have used NMR spectroscopy to gain further insights into the solution properties and structure of N-BP-2. Purified human N-BP-2 bound IGF-I and IGF-II with a lower affinity (417 and 87 nM, respectively) compared to bovine N-BP-2 (36 and 52 nM, respectively) [24]. The reason for this difference was not clear since human
Acknowledgements
We thank Gary Shooter and Zee Upton for valuable discussions and acknowledge support received from the staff and facilities of the Queensland NMR Network (QNN). This work was supported in part by grants from the Australian Research Council (LP0776825) and the National Health and Medical Research Council, Australia (Program grant 461219), as well as by Tissue Therapies Ltd. R.S.N. acknowledges fellowship support from NHMRC. This research was partly undertaken on the SAXS/WAXS beamline at the
References (73)
- et al.
Defining the pathway to insulin-like growth factor system targeting in cancer
Biochem. Pharmacol.
(2010) - et al.
Growth regulation via insulin-like growth factor binding protein-4 and -2 in association with mutant K-ras in lung epithelia
Am. J. Pathol.
(2006) - et al.
Neuroendocrine cancer-specific up-regulating mechanism of insulin-like growth factor binding protein-2 in small cell lung cancer
Am. J. Pathol.
(2009) - et al.
IGF-dependent and IGF-independent actions of IGF-binding protein-1 and -2: implications for metabolic homeostasis
Trends Endocrinol. Metab.
(2009) - et al.
IGF-binding proteins–the pieces are falling into place
Trends Endocrinol. Metab.
(2005) - et al.
BIAcore analysis of bovine insulin-like growth factor (IGF)-binding protein-2 identifies major IGF binding site determinants in both the amino- and carboxyl-terminal domains
J. Biol. Chem.
(2001) - et al.
Isolation of a biologically active fragment from the carboxy terminus of the fetal rat binding protein for insulin-like growth factors
Biochem. Biophys. Res. Commun.
(1988) - et al.
Structural basis for the regulation of insulin-like growth factors by IGF binding proteins
Structure
(2005) - et al.
Structure, dynamics and heparin binding of the C-terminal domain of insulin-like growth factor-binding protein-2 (IGFBP-2)
J. Mol. Biol.
(2006) - et al.
Accelerated acquisition of high resolution triple-resonance spectra using non-uniform sampling and maximum entropy reconstruction
J. Magn. Reson.
(2004)