Elsevier

Cellular Signalling

Volume 17, Issue 10, October 2005, Pages 1183-1193
Cellular Signalling

Review
Autocrine, paracrine and juxtacrine signaling by EGFR ligands

https://doi.org/10.1016/j.cellsig.2005.03.026Get rights and content

Abstract

Receptor and cytoplasmic protein tyrosine kinases play prominent roles in the control of a range of cellular processes during embryonic development and in the regulation of many metabolic and physiological processes in a variety of tissues and organs. The epidermal growth factor receptor (EGFR) is a well-known and versatile signal transducer that has been highly conserved during evolution. It functions in a wide range of cellular processes, including cell fate determination, proliferation, cell migration and apoptosis. The number of ligands that can activate the EGF receptor has increased during evolution. These ligands are synthesized as membrane-anchored precursor forms that are later shed by metalloproteinase-dependent cleavage to generate soluble ligands. In certain circumstances the membrane anchored isoforms as well as soluble growth factors may also act as biologically active ligands; therefore depending on the circumstances these ligands may induce juxtacrine, autocrine, paracrine and/or endocrine signaling. In this review, we discuss the different ways that EGFR ligands can activate the receptor and the possible biological implications.

Introduction

In a biological system, cells are continuously exposed to diverse stimuli. Differential gene expression and other cell functions during development are often controlled by signals from one cell to another. Information flow from the extracellular environment into the cell is required for appropriate functioning of the system and is mediated by signaling molecules (ligands), receptors, and machinery for transducing signals through the cell membrane and for communicating with the nucleus and cytoskeleton (Fig.1). It is important that these extracellular signals are correctly interpreted by the cell, especially considering that cells are exposed to stimuli that range from soluble endocrine and paracrine factors to signaling molecules on neighboring cells. Receptor tyrosine kinases (RTKs) are cell surface allosteric enzymes consisting of a single transmembrane domain that separates an intracellular kinase domain from an extracellular ligand-binding domain and are primary mediators of many of these signals [1], [2].

The prototypal example of RTKs is the EGF receptor (EGFR; also referred to as HER1 (human EGF receptor) and c-erbB1) which was the first RTK to be discovered (reviewed in [2], [3], [4]). Excitement about EGFR activation and its ligands was first generated by the initial observations of Stanley Cohen in the 1960s that in vivo EGF administration to newborn mice can induce effects such as precocious eyelid opening and tooth eruption [5].

Activation of RTKs induces kinase activity directed against tyrosine residues located both within the receptor itself (autophosphorylation) and on target downstream molecules [4], [6]. Ligand binding activates the kinase, which, with a few minor exceptions, is required for all cellular responses. The pleiotropic cell responses include cell proliferation, migration, and differentiation, as well as homeostatic functioning [7], [8], [9], [10]. However, understanding the biological functions of EGF receptor is complicated considering that a wide range of biologically active ligands can bind and activate the receptor [11]. Furthermore, while most of EGF family of growth factors are biologically active (capable of activating the receptor) only in their soluble secreted form, some are also active even as transmembrane precursor molecules [11], [12], [13].

Therefore, the question arises: what is the biological significance of the multitude of ligands for a single receptor and what are the differences in their mode of action? In this review, we have concentrated on the various ways a ligand may be presented to the EGF receptor, which includes autocrine, paracrine and juxtacrine modes of signaling. We have further discussed the potential differences in biological outcome due to differential signaling of the EGFR.

  • 1.

    Epidermal Growth Factor Receptor (EGFR)

  • 2.

    EGFR and Signaling

  • 3.

    Structure and processing of EGFR ligands

  • 4.

    Endocrine, autocrine and paracrine signaling through EGFR

  • 5.

    Juxtacrine signaling through EGFR

  • 6.

    Conclusion

The epidermal growth factor receptor is a 170-kDa plasma membrane glycoprotein. It contains an extracellular ligand-binding domain, a single transmembrane region, an intracellular domain and a C-terminal tail with multiple phosphorylation sites [Fig. 2 [14]]. The extracellular domain is composed of four subdomains that contain two cysteine-rich domains and comprise the ligand-binding domain [2], [15]. The other parts of the extracellular domain mediate receptor dimerization and interactions with other membrane proteins [16]. The transmembrane domain primarily serves as a site for feedback attenuation by PKC (protein kinase C) and erk MAP kinases (extracellular signal-regulated kinase, mitogen-activated protein kinase) [14], [17]. However, there is some evidence that a motif within this region may possibly link to the heterotrimeric G proteins [18], [19]. The C-terminal tail contains five autophosphorylation motifs that link to proteins containing SH2 or PTB (phospho-tyrosine binding) domains, three internalization motifs comprised of a tight turn, and sites for transphosphorylation as well as proteolytic activation and degradation [3], [14]. Studies suggest that this tail also functions as an autoinhibitory substrate, as either deletion of the cytoplasmic tail or mutation of the autophosphorylation sites renders the ligand-activated EGFR unable to phosphorylate downstream substrates [20], [21], [22], [23]. In contradistinction to many other receptor tyrosine kinases, EGFR autophosphorylation motifs are structurally similar [2], [14], [24] and functionally redundant. This simple architecture and flexible interchange of redundant motifs supports the designation of EGFR as an archetypal gene.

The EGFR is a member of a family of four structurally similar tyrosine kinase receptors (ErbB family; [25]) that also includes HER2/neu (erbB2; [26]), erbB3 [27], and erbB4 [28], [29]. ErbB receptors are expressed in a variety of tissues of epithelial, mesenchymal and neuronal origin, where they play fundamental roles in development, proliferation and differentiation. Furthermore, dysregulated expression (overexpression, constitutive activation and/or genetic mutations) of ErbB receptors, in particular ErbB1 and ErbB2, has been implicated in the development and malignancy of numerous types of human cancers [2], [30], [31], [32], [33]. Cumulative results from studies using genetic knockout in mice of members of the ErbB family of receptors suggest that these receptors play critical roles during the development of many organs [34], [35], [36], [37], [38], [39]. However, EGFR appears to contribute primarily to epithelial development whereas the other ErbB receptors regulate the formation of other tissues including neural and skeletal muscle tissues [34], [35], [40]. An important role for the EGFR is reflected by its conservation in organisms from C. elegans to mammals. In general, the size of the receptor as well as the extracellular/ transmembrane/cytoplasmic domain structure has been maintained [41], [42].

In normal tissue, the EGF receptor is activated by a variety of receptor-specific ligands [43]. Upon ligand binding, the intrinsic kinase is activated and EGFR tyrosyl phosphorylates itself and thereafter numerous effector molecules [7], [30]. EGFR interacts with most members of the c-ErbB subfamily of RTKs [10]. While no known ligand from the EGF family of growth factors binds to erbB-2, the erbB-3 and erbB-4 serve as heregulin and neuregulin receptors [43], [44]. However, irrespective of their ligand specificity, a major function of these other receptors appears to be as downstream effectors of each other. It adds to their interdependence that binding affinity of the EGF towards EGFR is modulated by co-expression of ErbB2 or ErbB3 in the same cells, even though these latter receptors do not directly bind to EGF. Indeed, Garrett et al. [45] demonstrated that deletion of the dimerization loop reduces ligand binding affinity and ligand-induced tyrosine phosphorylation of the mutated EGFR. The ErbB family of receptors is known to homo and hetero-aggregate, cross-phosphorylate, and modulate signaling from each other in specific pairings. For instance, EGFR (erbB-1) can interact with erbB-2 and erbB-3 but does not usually dimerize with erbB-4, whereas erbB-4 will pair with erbB-2 [6]. The integrated biological responses to EGFR signaling are pleiotropic, including mitogenesis or apoptosis, enhanced cell motility, protein secretion, and differentiation or dedifferentiation.

The activation of EGFR kinase activity triggers numerous downstream signaling pathways similar to other RTK (Fig. 3). These pathways include those that involve PLCγ (phospholipase C-γ) [46], [47] and its downstream calcium- and PKC-mediated cascades, ras activation leading to various MAP kinases [8], other small GTPases such as rho and rac, multiple STAT (signal transducer and activator of transcription) isoforms [48], and heterotrimeric G proteins, as well as other pathways including the phospholipid-directed enzymes PI3 kinase (phosphatidylinositol 3′-OH kinase) and PLD (phospholipase D), and the proto-oncogene cytoplasmic tyrosine kinase src [24]. This profusion of cascades, downstream of the EGFR, has often prevented the logical elucidation of the biochemical links to the appropriate biological responses; however a few principles are becoming clear. The first emerging concept is that a number of signaling pathways can be shown to be required, but may not be sufficient for a particular response. For example, PLCγ-mediated hydrolysis of PIP2 (phosphatidylinositol (4,5) bisphosphate) and mobilization/activation of actin-modifying proteins is required for EGFR-mediated motility, but motility is blocked if MEK (MAP kinase kinase) signaling is abrogated [49], [50], [51]. The second concept is that a multiplicity of the biological responses may be elicited by an individual signaling pathway, as for example, after activation of the erk MAP kinases [24].

The question arises as how the EGFR signaling is regulated and whether presentation of different ligands can induce differential signaling? Considering the pleiotropic effects of EGFR dependent signaling, it would be reasonable to argue that the duration and strength of signals are tightly regulated in the cell by the action of numerous negative regulatory mechanisms. These regulatory mechanisms constitute 1) availability of the ligand to the receptor, 2) negative regulation by phosphatases that may interfere with the amplitude and kinetics of the growth factor receptor signals, thereby modulating the biological responses and lastly 3) terminal signal inactivation by regulating the availability of the receptor itself for activation through receptor internalization and degradation.

In polarized epithelial cells EGFR is largely restricted to the basolateral aspects, allowing for epithelial–stromal communication from fibroblast-derived TGF-α and other matrix-associated EGFR ligands. This asymmetric presentation of EGFR limits autocrine signaling, as many of these epithelial organs, particularly throughout the genitourinary system, secrete copious amounts of EGF into the lumen (apical aspect of the cell) [52], [53]. High concentrations of EGF are found in the urine, and high concentrations of prepro-EGF mRNA have been detected in the kidney, localized to thick ascending limb of Henle (TALH) and distal convoluted tubules, although at the apical surface [54], [55], [56].

Following ligand binding, EGFR receptors are rapidly internalized from the cell surface via several pathways, including clathrin-coated pits, which depend on specific adaptins and sorting nexins complexing with EGFR carboxy-terminal motifs [57], [58]. Internalized receptors are initially delivered to early endosomes, which in turn mature into late endosomes and multivesicular bodies. The Cbl family of ubiquitin ligases plays a pivotal role in these processes [58]. Cbl can bind directly to phosphorylated EGFR receptors via its tyrosine kinase-binding domain, while the RING finger domain of Cbl recruits ubiquitin-conjugating enzymes (E2, Ubc) and mediates the transfer of ubiquitin to the receptor [58], [59]. The fate of the receptor upon internalization depends on continued occupancy in the endosomal compartments, as seen with EGF binding, or ligand dissociation and recycling of the receptor, as is the case with TGF-α binding [14]. Ligand-induced internalization and degradation results in signal attenuation with net removal of both ligand and receptor (in the case of nondissociative ligands like EGF) or ligand alone (for dissociative ligands such as TGF-α) (Fig. 4). Thus, different ligands can dictate the strength and temporal lifespan of EGFR signals, thereby providing a rationale for the existence of multiple genetically distinct ligands.

The number of ligands that can activate the EGFR has increased during evolution. There is a single ligand in C. elegans, four in Drosophila and seven ligands have been identified in mammals [60], [61]. The mammalian ligands that activate EGF receptor include EGF, transforming growth factor-α (TGF-α), heparin-binding EGF-like growth factor (HB-EGF), amphiregulin (AR), betacellulin (BTC), epiregulin (EPR) and epigen (EPI) [11], [62], [63]. Each of the mature peptide growth factors is characterized by a consensus sequence consisting of six spatially conserved cysteine residues (CX7 CX4–5 CX10–13 CXCX8 C) that form three intramolecular disulfide bonds, with the following interactions: C1−C3, C2−C4, C5−C6 [11]. This consensus sequence is known as the EGF motif and is crucial for binding members of the HER receptor tyrosine kinase family. In addition to binding EGFR, HB-EGF, BTC, and EPR are reported to also bind HER4. Mature HB-EGF and AR also contain an amino-terminal heparin-binding domain (HBD). All of the EGFR ligands exist in a proform as type I transmembrane proteins consisting of an EGF motif flanked by an N-terminal extension and a C-terminal membrane-anchoring domain, and soluble ligands are produced through extracellular cleavage of the integral membrane precursor proteins [11]. There is no obvious homology in the predicted cleavage sites among the different EGFR ligands, although many studies have shown that metalloprotease activity is required for their release [64], [65].

TACE (TNF alpha converting enzyme), also known as ADAM-17, has been shown to be responsible for cleaving TGF-α, and TACE-/-cells showed a 95% reduction in TGF-α release [66], [67], [68]. TACE has also been implicated in the cleavage of amphiregulin and HB-EGF [69], [70]. ADAM-10 and ADAM-12 have also been implicated in HB-EGF shedding in different cells [69], [71], [72]. In addition, studies have implicated MMP-2, 3 and -9 in the processing of HB-EGF [73], [74]. The identification of more than one proteolytic enzyme responsible for processing of an EGFR ligand could be simply due to differences between the cell lines used in these studies. Furthermore, substantial evidence indicates that proteolytic release of EGFR ligands is a regulated process, which can be influenced by protein kinase C (PKC) activity, calcium influx, and phosphatase activity [75], [76], [77].

Proteolytic release of EGFR ligands appears to be an important regulatory step in activating the receptor. For example, mice lacking TACE showed a phenotype similar to that of EGFR-/-mice, probably because of defective processing of TGF-α and possibly other EGFR ligands [78], [79]. Similarly, inhibiting release of EGF and/or TGF-α inhibits growth as well as migration in several EGFR dependent cell lines [80]. Furthermore, proteolytic release of Spitz, a Drosophila TGF-α homologue, is a limiting step in the activation of the Drosophila EGFR [81]. However, many EGFR ligand precursors including HB-EGF, TGF-α, AR, and BTC are biologically active even when they are tethered to the plasma membrane, suggesting their capability of functioning as juxtacrine factors [12], [82], [83], [84]. These precursors can induce tyrosine phosphorylation of EGFR expressed on juxtaposed cells without the release of detectable ligand, and studies have suggested that the biological outcome of juxtacrine activation of the EGFR is different than when there is autocrine/paracrine activation of the receptor [85], [86], [87].

In general, cell–cell communication produces spatially non-uniform patterns in the expression of genes that guide the development of tissues and organs. Typically, a ligand can activate receptors on the cell of its origin (autocrine) or can be released into the cell milieu where it interacts with the extracellular matrix and cell surface receptors as it spreads through the tissue (so-called paracrine stimulation; may be activation of same cell type or a different cell type, i.e. mesenchymal derived factors activating epithelial or endothelial cells or vice versa) or can be released systemically and thereby induce endocrine mode of activation, a frequent mechanism in hormone dependent stimulations. All of the above mechanisms described for receptor activation are important in development and various other cellular functions and have a common mode of receptor activation in that the ligand is released from the cell that generates it and acts in the fluid phase (Fig. 5).

Epidermal growth factor (EGF) and TGF-α are the two most extensively studied ligands that specifically bind and activate EGFR. Recently enormous enthusiasm has been given to HB-EGF as an important molecule in EGFR activation since the initial observations of Prenzel et al. (2000) that G-protein coupled receptor-mediated EGFR transactivation requires metalloproteinase dependent cleavage of proHB-EGF, suggesting a role for autocrine/paracrine activation [88]. However, HB-EGF can also activate HER-4 (ErbB-4, EGFR 4) [12], [89]. In vitro studies using exogenous administration of EGFR ligands including EGF, TGF-α, HB-EGF etc. have demonstrated a role for EGFR activation in cell proliferation, migration, apoptosis, differentiation and dedifferentiation [3], [8], [65]. In the developing animal pro-EGF mRNA, immunoreactive EGF, immunoreactive TGF-α and EGF receptors are present in many tissues. EGF also is produced and secreted by the maternal mammary gland, and mammary derived EGF appears to be important in gut development in the neonatal rodent [90], [91]. Exogenous administration of EGF accelerates recovery from nephrotoxic injuries [92], [93]. HB-EGF is present in human amniotic fluid and breast milk [94]. EGF, TGF-α, HB-EGF and neuregulin play important roles in the proper development of nervous system [95]. It is interesting that although EGF has been identified in most body fluids of several mammalian species, neither EGF antibody administration to newborn animals nor passive immunization of pregnant rodents against EGF has caused major deleterious effects (except the delay in epidermal maturation events). Also, to date, no pathological EGF deficiency disorder has been characterized. This may be due to redundancy in the ligands available to activate the EGFR in a defined cell milieu. Available evidence suggests that TGF-α and HB-EGF may subserve the growth factor family roles in fetal development.

Each of the EGF receptor ligands is capable of autocrine/paracrine activation of EGFR receptor, and studies have demonstrated distinct biological activities in response to the receptor activation by different EGFR ligands; however, the mechanism responsible for their diverse actions is unknown. Biochemically, their ability to activate the EGFR appears to be identical, however the expression of the EGFR ligands varies between tissues and also among different cells of the same tissue. There is also redundancy in EGFR ligand expression temporally and spatially considering the basolateral localization of the EGFR. Therefore, the question arises as how these ligands, which interact in identical fashion, can induce distinct biological responses. As mentioned before, EGF receptor is known to form homo- and heterodimers upon ligand binding, which has been postulated to affect the ligand-binding affinity to the EGFR. Therefore, differential dimerization of the receptor/s with each other could be hypothesized to be one possible way to induce ligand specific signaling and hence biological responses; however, there are few studies at present that have compared the details of the differences in receptor dimerization in response to various EGFR specific ligands. Another possible way to discriminate between signaling induced by different EGFR ligands could be dictated by the strength of the ligand–receptor interaction and differential Trafficking of the EGFR (the compartment-restricted signaling).

As mentioned above, following ligand binding, the EGFR is rapidly internalized by a mechanism that requires intrinsic receptor kinase activity and specific motifs in the carboxy terminus domain of the receptor [57], [58]. Following internalization, the EGFR is sorted into early endosomes and late endosomes. This initial sorting step appears to be mediated by di-leucine motifs in the juxtamembrane region of the receptor. In contrast to the requirement of kinase activity for occupancy-induced recruitment into coated pits, regulated endosomal sorting appears to be independent of receptor kinase activity [58], [59]. Possibly, the occupancy-induced conformational changes in the receptor are sufficient for endosomal sorting. Molecules such as c-Cbl can modify the lifetime of activated EGFR-ligand complexes, but how would this affect signaling? Internalized EGFRs have been shown to be enzymatically active, hyperphosphorylated, and associated with Ras-GAP, Shc, Grb2, and mSOS [96]. The tyrosine kinase adaptor protein Shc appears to be strongly associated with active EGFRs at both the cell surface and during endocytic Trafficking [97]. This would suggest similar signaling through internalized or surface-localized EGFRs; however; some studies have also suggested that specific EGFR signaling pathways are triggered within endosomes.

It is known that TGF-α binds poorly to EGFR within endosomes whereas EGF binds well. Recently, a number of studies have compared the differences in the endosomal signaling between TGF-α and EGF by exploiting the differential pH sensitive binding of TGF-α and EGF to the EGFR [98], [99], [100]. The application of a mild acid mileau results in the removal of surface-localized TGF-α and a total loss of TGF-α-activated receptors. However, because EGF binds well to its receptor within endosomes, the removal of surface-localized EGF leaves a significant population of activated, endosome-localized receptors. Such studies have shown that endosomally localized EGFR can activate Ras as efficiently as surface-localized receptors, but that activation of phospholipase C (PLC)-γ is restricted to the cell surface [101], [102]. A similar approach has been used to demonstrate that internalized EGFRs are responsible for enhanced expression of the cell cycle regulatory protein p21:CIP1 [103]. Interestingly, internalized EGFRs were unable to activate the PLC-γ pathway due to a lack of the appropriate lipid substrate in the endosomal compartment. Together, these results suggest that signaling from endosomes is qualitatively different from that generated at the cell surface. Thus, because TGF-α rapidly dissociates within the acidic environment of endosomes, its signaling pattern would be biased toward the cell surface. In contrast, ligands such as EGF that remain receptor-associated within endosomes would display a pattern that was biased toward the endosomes. For example, it is known that PLC-γ activation is necessary for EGFR-stimulated cell migration [51]. The observation that TGF-α has a more pronounced effect on cell migration than EGF may in part be explained by the restriction of PLC-γ signaling to the cell surface. A role for compartment-specific signaling in ligand discrimination becomes more attractive where ligands can function as soluble (autocrine/paracrine) or membrane-anchored growth factors (juxtacrine). As discussed below, membrane-anchored HB-EGF induces different biological responses than soluble HB-EGF.

The term “juxtacrine” was coined by Anklesaria, Massague and colleagues to describe binding of pro-TGF-α, expressed on the plasma membrane of a mouse bone marrow stromal cell line, to EGFR on an adjacent hematopoietic progenitor cell [82]. This molecular interaction led to both adhesion and activation (DNA replication, cell division) of the target progenitor cell. Therefore juxtacrine intercellular signaling is described as “where the molecule that induces the functional changes in the target cell remains associated with the plasma membrane of the signaling cell, rather than acting in the fluid phase”. Intercellular interactions in which one cell sends a signal to another cell, inducing a change in function of the second cell, are common in morphogenesis, development, inflammation, and repair of various organs. Various molecules including cytokines and growth factors have been reported to act in a juxtacrine fashion and include C-Kit and Ephrins and TGF-α, Amphiregulin and HB-EGF among the EGFR ligands, [12], [82], [83], [84], [104], [105].

The modes of actions of membrane-anchored growth factors (juxtacrine factors) could be distinct from those of soluble growth factors in the following respects: (1) the juxtacrine factors are “nondiffusible” i.e. they are tethered to the plasma membrane and do not diffuse in fluids. Therefore, these molecules can transmit signals only to neighboring cells in contrast to a soluble growth factor, which can transduce signal to any surrounding cell that expresses the respective receptor. Needless to say, this feature would provide a great advantage for restricting the signals to only those cells that are in direct contact with the cells bearing the membrane-anchored growth factor, even if surrounding cells also possess their receptors (as illustrated in Fig. 5). (2) Bidirectional signaling between juxtacrine factors and their respective receptors may occur, as is the case with ephrin and the Eph receptor [106], and (3) Membrane-anchored juxtacrine factors can form a complex with other membrane proteins, and their activity can be regulated by associating molecules, as has been shown for pro-HB-EGF-associating CD9 (see below). These unique features of membrane-anchored growth factors imply that they have different biological activity from their soluble counterparts and that the juxtacrine signaling is likely to be a physiologic event that requires tight regulation, and disruption of juxtacrine signaling may lead to pathologic outcomes.

As mentioned before, TGF-α was the first EGFR ligand suggested to activate EGFR in a juxtacrine fashion [82]. Recently TGF-α has also been shown to interact with CD9, a tetraspanin [107]. Interestingly, increased CD9 expression decreased the growth factor-induced release of the TGF-α ectodomain and enhanced the transmembrane TGF-α-induced EGFR stimulation. However, these studies, in general, were performed in systems with endogenous EGFR expression and in the absence of inhibition of endogenous metalloproteinase activity and therefore the question remains unanswered, whether the observed effect is true juxtacrine activation or a tightly regulated autocrine/paracrine activation of the EGFR. This notion further gains strength from the fact that studies over-expressing TGF-α failed to detect any immunoreactive TGF-α in the medium unless receptor binding was blocked by an EGFR-specific monoclonal antibody suggesting a rapid uptake of the secreted ligand [108].

Amphiregulin (AR) has also been identified as an EGFR ligand capable of juxtacrine activation of EGFR [83]. Amphiregulin has been localized in human placenta and decidua throughout gestation, suggesting a role in trophoblast growth that involves localized cell–cell interaction and requires tight regulation of signal transduction [109]. Using an antibody that cross-reacts only with the transmembrane form of (HB-EGF-TM), HB-EGF has also been shown to be expressed in mouse uterine luminal epithelium temporally, just prior to implantation, and is spatially localized to the site of blastocyst apposition [110].

CD9 was co-immunoprecipitated with HB-EGF in human keratinocytes and co-expression of CD9 was shown to induce juxtacrine activity in cell proliferation [83]. As has been seen with HB-EGF, CD9 can be co-immunoprecipitated with AR in human keratinocytes and co-expression of CD9 will also increase juxtacrine activation by AR [111], [112]. Co-expression of CD9 in renal epithelial cells increases proHB-EGF's cytoprotective capacity [113] and administration of a CD9 blocking antibody abrogates the ability of membrane-anchored HB-EGF molecule to inhibit Hepatocyte Growth Factor (HGF/SF) induced cell scattering in MDCK cells [86]. CD9 does not interact with soluble HB-EGF but only associates with membrane-associated proHB-EGF.

Tetraspanins are accessory molecules that stabilize and/or facilitate interactions of other proteins [114]. CD9 is expressed in hematopoietic cells but is also expressed in mesenchymal and epithelial cells. CD9 expression in carcinoma cell lines inhibits motility and metastasis [115], [116], and decreased CD9 expression is correlated with poor prognosis in breast cancer [117]. In addition to CD9, other tetraspanins, including CD63, CD81, and CD82, also associate with HB-EGF, but only CD9 upregulates mitogenic and diphtheria toxin binding activities of HB-EGF [118], [119]. CD9 interacts with selected integrins, especially β1 integrins [115], which are important in cell–cell and cell–ECM (extracellular matrix) adhesive properties. In Vero cells expressing HB-EGF and CD9, HB-EGF-CD9 complexes localize to the cell–cell contact interface in association with α-catenin and vinculin [120], [121], and CD9 interacts specifically with α3β1 integrins at the adherens junctions [120]. In polarized cells, EGFR is also localized to sites of cell–cell contact at the zona adherens in association with E-cadherin and α- and β-catenin [122], [123], [124]. β1 integrins appear to be important for epithelial cell–cell interactions [125], including renal epithelial cells, in which α2β1 and α3β1 integrins have been localized to sites of cell–cell contact [126]. In contrast to TGF-α, the majority of protein delivered to the plasma membrane remains as proHB-EGF.

The question still remains whether the biological responses of the juxtacrine mode of receptor activation are different than the autocrine/paracrine mode of receptor activation. We hypothesize that autocrine–paracrine versus juxtacrine signaling of EGFR signaling has developed as a means to ensure signaling in proper temporal and spatial context. For example, under normal physiological conditions, when the cell layer is intact, the ligands capable of juxtacrine signaling are in close apposition with the receptor and thereby help in maintaining the integrity of the cell layer. Conversely, when the cell layer is damaged, repair requires cell proliferation and migration. Damage to the cell layer induces metalloproteinase activity resulting in cleavage of the ligand, which would then initiate the autocrine/paracrine signaling and induce proliferation and migration and ultimately tissue repair (Fig. 6). Although there are still relatively few studies that differentiate juxtacrine vs. tightly coupled autocrine/paracrine signaling, there is growing evidence to suggest that juxtacrine signaling may indeed induce a differential biological response than the autocrine/paracrine mode of signaling. Studies using a non-cleavable HB-EGF mutant have suggested differences in growth, branching morphogenesis, and trans-epithelial resistance and apoptosis [86], [127]. It remains to be seen whether differences in ligand induced receptor dimerization (homo- and/or heterodimer formation between different receptors of the erbB family) and/or lack of receptor endocytosis due to binding to a membrane-teetered ligand or formation of a protein–protein complex is involved in these differences. Further studies are clearly required to answer these questions.

The spatial pattern formation is an important issue in the developmental of an embryo. Some fine patterning arising in early development has a very small spatial scale and the natural explanation is that they arise by direct cell–cell interactions and thereby signaling in epithelial and mesenchymal context and emphasizes the importance of juxtacrine signaling. Further, a corollary of the idea that juxtacrine signaling provides another dimension of biological regulation and a dysregulation of this interaction may be the way the physiological cell–cell interaction becomes pathological, which may justify the importance of the ErbB receptors during development but also targets them as a mediator of carcinogenesis, when the natural interactions are disrupted. This picture is further complicated by the different cellular responses due to differential Trafficking of activated EGFR, which leads to the question: unless specific mechanisms exist to remove and degrade the ligand, will signaling continue after endocytosis? Although studies in the past decade have made great advances towards understanding the complexities of the EGFR dependent signaling, further studies with carefully designed models with engineered systems are needed to answer the specific questions regarding the differences between different modules of signaling and also the possible differences between the biological potentials of various ligands.

Section snippets

Acknowledgements

This work was supported by National Institutes of Health grant DK51265 (R.C.H.), funds from the Department of Veterans Affairs (R.C.H.) and American Heart Association Southeast Affiliate grant 4043756112 and National Scientist Development Grant 0435471N (A.B.S.).

References (127)

  • G. Carpenter et al.

    J. Biol. Chem.

    (1990)
  • A. Ullrich et al.

    Cell

    (1990)
  • S. Cohen

    Dev. Biol.

    (1965)
  • J. Schlessinger

    Cell

    (2002)
  • I. Alroy et al.

    FEBS Lett.

    (1997)
  • R.C. Harris et al.

    Exp. Cell Res.

    (2003)
  • R. Iwamoto et al.

    Cytokine Growth Factor Rev.

    (2000)
  • A. Wells

    Int. J. Biochem. Cell Biol.

    (1999)
  • A.W. Burgess et al.

    Mol. Cell

    (2003)
  • H. Sun et al.

    J. Biol. Chem.

    (1997)
  • J.R. Glenney et al.

    Cell

    (1988)
  • P.J. Bertics et al.

    J. Biol. Chem.

    (1988)
  • P.O. Hackel et al.

    Curr. Opin. Cell Biol.

    (1999)
  • D.S. Salomon et al.

    Crit. Rev. Oncol./Hematol.

    (1995)
  • C. Arteaga

    Semin. Oncol.

    (2003)
  • R.S. Herbst et al.

    Semin. Oncol.

    (2002)
  • R. Schweitzer et al.

    Trends Genet.

    (1997)
  • T.P. Garrett et al.

    Cell

    (2002)
  • M. David et al.

    J. Biol. Chem.

    (1996)
  • M.S. Murphy

    Nutrition

    (1998)
  • R.C. Harris

    Am. J. Kidney Dis.

    (1991)
  • E.C. Salido et al.

    J. Ultrastruct. Mol. Struct. Res.

    (1986)
  • R.C. Kurten

    Adv. Drug Deliv. Rev.

    (2003)
  • S. Bogdan et al.

    Curr. Biol.

    (2001)
  • L. Strachan et al.

    J. Biol. Chem.

    (2001)
  • S.W. Sunnarborg et al.

    J. Biol. Chem.

    (2002)
  • M. Suzuki et al.

    J. Biol. Chem.

    (1997)
  • S.M. Le Gall et al.

    J. Biol. Chem.

    (2003)
  • W. Shi et al.

    Dev. Biol.

    (2003)
  • J.R. Lee et al.

    Cell

    (2001)
  • R. Iwamoto et al.

    J. Biol. Chem.

    (1999)
  • M.P. Michalsky et al.

    J. Pediatr. Surg.

    (2002)
  • A.R. French et al.

    J. Biol. Chem.

    (1995)
  • K. Maeda et al.

    J. Control. Rel.

    (2002)
  • B.M. Sefton et al.

    Adv. Cycl. Nucleotide Protein Phosphoryl. Res.

    (1984)
  • D.M. Thompson et al.

    Cancer Surv.

    (1985)
  • K.D. Brown

    Eur. J. Gastroenterol. Hepatol.

    (1995)
  • L. Lu et al.

    Exp. Biol. Med.

    (2001)
  • P. Casalini et al.

    J. Cell. Physiol.

    (2004)
  • M.R. Bush et al.

    Biol. Reprod.

    (1998)
  • I. Lax et al.

    EMBO J.

    (1989)
  • X. Li et al.

    EMBO J.

    (1998)
  • H. Daub et al.

    EMBO J.

    (1997)
  • K. Helin et al.

    Oncogene

    (1991)
  • M. Magni et al.

    Biochem. J.

    (1991)
  • A. Ullrich et al.

    Nature

    (1984)
  • C.I. Bargmann et al.

    Nature

    (1986)
  • M.H. Kraus et al.

    Proc. Natl. Acad. Sci. U. S. A

    (1989)
  • G.D. Plowman et al.

    Proc. Natl. Acad. Sci. U. S. A

    (1990)
  • G.D. Plowman et al.

    Proc. Natl. Acad. Sci. U. S. A

    (1993)
  • Cited by (311)

    • Growth factors and their receptors: multitasking functionality in health and disease

      2023, Receptor Tyrosine Kinases in Neurodegenerative and Psychiatric Disorders
    • Regulation of receptor tyrosine kinase gene expression

      2023, Receptor Tyrosine Kinases in Neurodegenerative and Psychiatric Disorders
    • Applications of exosomes in nanomedicine

      2023, Nanomedicine: Technologies and Applications
    View all citing articles on Scopus
    View full text