Article Text


Signalling pathways involved in antiproliferative effects of IGFBP-3: a review
  1. R C Baxter
  1. Kolling Institute of Medical Research, University of Sydney, Royal North Shore Hospital, Sydney, NSW 2065, Australia
  1. Dr Baxter robaxter{at}


Insulin-like growth factor binding protein-3 (IGFBP-3), the major circulating carrier protein for IGFs, is also active in the cellular environment as a potent antiproliferative agent. It appears to function both by cell cycle blockade and the induction of apoptosis. Transfection of p53 negative T47D breast cancer cells to express IGFBP-3 leads to induction of the apoptotic protein bax and an increase in sensitivity to ionising radiation. IGFBP-3 can be transported to the nucleus by an importin β mediated mechanism, where it has been shown to interact with the retinoid X receptor α and possibly other nuclear elements. Expression of oncogenic ras is associated with resistance to exogenous IGFBP-3, the effect being reversible by inhibition of mitogen activated protein (MAP) kinase phosphorylation. IGFBP-3 antiproliferative signalling appears to require an active transforming growth factor β (TGF-β) signalling pathway, and IGFBP-3 stimulates phosphorylation of the TGF-β signalling intermediates Smad2 and Smad3. These recent findings all point to a complex intracellular mode of action of IGFBP-3, which will need to be better understood if anti-cancer treatments are to take advantage of the antiproliferative activity of IGFBP-3.

Statistics from

Insulin-like growth factor binding protein-3 (IGFBP-3) is one of a family of six IGFBPs, multifunctional proteins that transport and stabilise IGFs in the circulation, modulate the effects of IGF on a variety of cellular functions, and may also regulate cells by IGF independent mechanisms. Initially purified from human plasma,1 IGFBP-3, a 40–45 kDa glycoprotein, is by far the most abundant IGFBP in the circulation, where it has a central role in regulating IGF bioavailability to the tissues by forming heterotrimeric complexes with IGF-I or IGF-II and an 85 kDa glycoprotein, the acid labile subunit.2

IGFBP-3 is also active in the cellular environment. A decade ago, work in our laboratory showed that IGFBP-3 production is potently induced by transforming growth factor β (TGF-β), and we proposed a role for IGFBP-3 in mediating TGF-β inhibitory activity.3 This mechanism is known to operate in some breast cancer cells.4 Subsequently, as recently reviewed, a substantial body of literature has similarly implicated the induction of IGFBP-3 in the mode of action of a variety of growth inhibitory proteins and agents, including the tumour suppressor protein p53, retinoic acid, vitamin D, anti-oestrogens, and tumour necrosis factor α.5 In some cases at least, IGFBP-3 is proposed to act by inhibiting the access of IGF to the type I IGF receptor, which mediates most of the actions of IGF.

However, the effects of IGFBP-3 on IGF dependent cellular functions are complex, with both stimulatory and inhibitory actions reported, even within the one cell type.6 In recent years, important studies with cells lacking the type I IGF receptor have revealed that IGFBP-3 can be inhibitory to cell growth even in the absence of IGF signalling,7 raising the possibility of truly IGF independent effects of IGFBP-3. The mechanisms underlying such effects are poorly understood because, to date, no cell surface IGFBP-3 receptor that initiates an intracellular signalling cascade has been identified.

This review will summarise studies, both from our laboratory and from other researchers, that contribute to the current understanding of the mechanisms underlying these growth inhibitory effects. Many of the studies cited refer to investigations in breast cancer cells, although important contributions also come from other areas of cell biology.

Nuclear transport and function of IGFBP-3 and IGFBP-5

The structural resemblance between a highly basic C-terminal domain of IGFBP-3 (residues 215–232) and typical nuclear localisation signals on nuclear targeted proteins led to studies published in 1997 showing that exogenous IGFBP-3 could pass to the nucleus of opossum kidney cells,8 and that endogenous IGFBP-3 could be detected in the nuclei of A549 human lung cancer cells.9 Nuclear IGFBP-3 has now been identified in a variety of cell types. The sequence necessary for nuclear translocation was confirmed as the basic C-terminal domain,10 and recently β importin, a known nuclear transport protein, has been implicated in the transport process.11 The key IGFBP-3 residues necessary for nuclear localisation, 228–232, are identical to those shown previously to be necessary for cell association of IGFBP-3.12 The plasma membrane appears to present a major barrier to nuclear transport of IGFBP-3 from outside the cell because only a small proportion of cells are labelled in intact cells, but virtually all cells can transport IGFBP-3 to the nucleus when the plasma membrane is detergent permeabilised, as long as an ATP generating system is provided.11

When the nuclear envelope is permeabilised, IGFBP-3 is retained within the nucleus, suggesting that it interacts with nuclear components.11 Although the identification of these interacting species remains a key area for future research, the recent observation that IGFBP-3 is a ligand for the nuclear retinoid X receptor α (RXR-α) is a major advance in this area.13 RXR-α was shown to be necessary for IGFBP-3 induced apoptosis, and RXR ligands were additive with IGFBP-3 in inducing apoptosis. These data suggest an essential role for RXR-α in mediating the nuclear effects of IGFBP-3. IGFBP-5 has recently been shown to have an opposite role to IGFBP-3 in the induction of apoptosis, leading to an enhancement of cell survival.14 Because IGFBP-5 appears to use a similar nuclear transport mechanism to IGFBP-3,10 the question of whether IGFBP-5 might oppose IGFBP-3 induced apoptosis by competitively inhibiting IGFBP-3 interaction with the importin nuclear transport mechanism, the nuclear RXR system, or some other cellular components will be important to determine.

The role of MAP kinase in regulating cell sensitivity to IGFBP-3

Recent studies from our laboratory have demonstrated that a ras dependent pathway is involved in regulating cell sensitivity to the inhibitory effect of IGFBP-3 on DNA synthesis. MCF-10 mammary epithelial cells, when transfected with an oncogenic H-ras gene, became refractory to inhibition by IGFBP-3, although the wild-type cells are inhibited by IGFBP-3 concentrations as low as 10 ng/ml.15 Blockade of mitogen activated protein kinase (MAPK)/extracellular signal regulated kinase (ERK) phosphorylation by MAPK kinase (MEK), using the inhibitor PD98059, restored sensitivity to inhibition by IGFBP-3. Hs578T is a tumour derived breast cancer cell line that is relatively refractory to inhibition by IGFBP-3, requiring concentrations of up to 1000 ng/ml to inhibit DNA synthesis. In the presence of PD98059, this cell line, like the ras transfected MCF-10 cells, showed a 10 fold increase in sensitivity to inhibition by IGFBP-3.15

MCF-10 cells expressing oncogenic ras remain sensitive to inhibition by TGF-β, despite their refractoriness towards IGFBP-3. However, in many other cell types, constitutive activation of ras signalling pathways leads to TGF-β resistance.16 A possible mechanism involves the phosphorylation by MAPK/ERK of sites on the TGF-β signalling intermediates, Smad2 and Smad3, that are distinct from the sites phosphorylated by the type I TGF-β receptor. This phosphorylation prevents the nuclear translocation of the Smads, leading to resistance to the action of TGF-β.16 Is it possible that in some cells a similar mechanism might lead to resistance to IGFBP-3?

Involvement of Smads in IGFBP-3 inhibitory signalling

Although early passage T47D cells transfected to express IGFBP-3 are growth inhibited, with blockade in the G1–S transition,17 we found paradoxically that T47D cells that are exposed to exogenous IGFBP-3 are resistant to its growth inhibitory action. In the course of investigating the mechanism accounting for this resistance to exogenous IGFBP-3, we observed that if an active TGF-β signalling pathway could be restored to T47D cells, which normally lack this pathway, the cells also became sensitive to growth inhibition by exogenous IGFBP-3.18 TGF-β signalling is initiated in most cells by the interaction of TGF-β with the type II TGF-β receptor (TGF-βRII) located on the cell surface, and association of this receptor with the cell surface type I receptor (TGF-βRI), which becomes phosphorylated and activated. Active TGF-βRI then serine phosphorylates intracellular receptor associated intermediates, Smad2 and Smad3, which subsequently form complexes with cytosolic Smad4, and heteromeric Smad complexes translocate to the nucleus, where they influence transcriptional events.19 An alternative, Smad independent TGF-β signalling pathway, involving an approximately 400 000 kDa protein named the type V TGF-βR, has also been proposed, and is reported to mediate IGFBP-3 inhibitory signalling in mink lung cells.20 However, its possible role in other cell types, in either TGF-β or IGFBP-3 action, is unknown.

The Smad mediated pathway of TGF-β signalling cannot normally be initiated in T47D breast cancer cells because they lack the TGF-βRII, although they express TGF-βRI and Smad2, Smad3, and Smad4.21 Restoration of TGF-βRII by transfection restores TGF-β signalling. Recently, we have shown that IGFBP-3, in the absence of exogenous TGF-β, leads to Smad2 and Smad3 phosphorylation in TGF-βRII transfected T47D cells and that TGF-β mediated Smad phosphorylation is enhanced by IGFBP-3 (fig 1).18 This is the first demonstration of known growth regulatory intracellular signalling intermediates being phosphorylated in response to IGFBP-3. Therefore, it might be important for our understanding of the mechanisms by which IGFBP-3 leads to the growth inhibition of some cancer cells, and why some cancer cells appear to be resistant to its actions.

Figure 1

In T47D breast cancer cells transfected to express the type II transforming growth factor β (TGF-β) receptor, exogenous insulin-like growth factor binding protein-3 (IGFBP-3) causes the rapid phosphorylation of the TGF-β signalling intermediates, Smad2 and Smad3, and enhances Smad phosphorylation by TGF-β. Cells were treated with 500 ng/ml human IGFBP-3 for the times indicated, which included a final 10 minute (Smad2) or 15 minute (Smad3) co-treatment with 2.5 ng/ml TGF-β1. Reproduced from Fanayan et al,18 with publisher's permission.

Induction of apoptosis by IGFBP-3

The proliferation of cell populations represents the balance between cell division and cell death. IGFBP-3 has the potential to modulate apoptosis in addition to its effects on DNA synthesis. One way in which this might occur is by sequestering IGFs, thus blocking their antiapoptotic activity.22 However, apoptotic effects of IGFBP-3 have been reported in cells lacking the type I IGF receptor,23 and under conditions where IGF-I could not elicit a survival effect,24 pointing to the existence of an IGF independent mode of IGFBP-3 action. The tumour suppressor protein p53 might initiate some IGFBP-3 mediated effects because IGFBP-3 gene expression is activated by p53 in response to DNA damaging stimuli, such as ionising radiation.25 However, IGFBP-3 appears to be capable of inducing apoptosis by a p53 independent pathway because it is active in p53 negative cell lines.23, 26

Apoptosis may be induced by an alteration in the ratio of proapoptotic and antiapoptotic proteins of the bcl-2 family—for example, the bax : bcl-2 ratio. In T47D breast cancer cells, which lack wild-type p53 and are relatively resistant to ionising radiation, transfection with IGFBP-3 cDNA causes an increase in apoptosis as determined by nuclear fragmentation or terminal deoxynucleotidyl transferase mediated dUTP nick end labelling. In clonogenic survival assays, IGFBP-3 transfected cells have a significantly reduced survival rate compared with controls, when exposed to increasing doses of ionising radiation (fig 2), and indices of apoptosis are significantly increased above values seen in control cells.26 Thus, IGFBP-3 can sensitise the radiation resistant cell line T47D to radiation induced apoptosis even in the absence of wild-type p53. The mechanism appears to involve a post-transcriptional increase in bax because concentrations of bax protein but not mRNA are increased by IGFBP-3 and further by ionising radiation. Antiapoptotic Bcl-xL values are correspondingly decreased (T47D cells do not express Bcl-2).26

Figure 2

T47D breast cancer cells transfected to express insulin-like growth factor binding protein-3 (IGFBP-3) show increased sensitivity to ionising radiation. Cell survival was measured 14 days after irradiation. Reproduced from Butt et al,26 with publisher's permission.


This brief review has attempted to provide an overview of currently proposed mechanisms by which IGFBP-3 exerts its antiproliferative effects. The unequivocal evidence for nuclear accumulation of IGFBP-3, and the recent demonstration of its interaction with retinoid and Smad dependent TGF-β signalling pathways, all point to cellular effects of IGFBP-3 that are clearly not dependent on simple blockade of the interaction between IGF-I and its receptor, although these effects may yet prove to be modulated by IGFs. Beyond these demonstrated pathways, there remains the possibility that interactions between IGFBP-3 and other cellular ligands might be important as modulators, or even initiators, of the effects of IGFBP-3.27 The delineation of these signalling pathways, and understanding their inter-relations, will be essential to the development of future anti-cancer treatments that might take advantage of the potent antiproliferative actions of IGFBP-3.


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.