Serum integrin-linked kinase (sILK) concentration and survival in non-small cell lung cancer: a pilot study.

BACKGROUND: Integrin-linked kinase (ILK) is an intracellular signaling protein critically involved in cellular growth and motility. In non-small cell lung cancer (NSCLC), increased ILK expression has been associated with decreased recurrence-free and overall survival. Recently, ILK has also been detected in the serum of NSCLC patients. OBJECTIVE: To assess the prognostic impact of preoperative serum ILK (sILK) concentration on overall survival in surgically amenable NSCLC. PATIENTS AND METHODS: Preoperative sILK was quantified by ELISA in 50 newly diagnosed NSCLC patients. After surgery, patients were followed-up for a median interval of 2.5 years. RESULTS: Serum ILK concentrations ranged from 0 to 2.44 ng/ml. Mean sILK was around 2.3 times higher in the 16 patients who died as compared to the 34 patients who survived (1.04 vs. 0.45 ng/ml, p = 0.001). In univariate time-to-event analysis, increased sILK was associated with adverse survival [Hazard ratio (HR): 4.03, 95 % CI: 2.00-8.13, p < 0.001]. This association prevailed after multivariable adjustment for several clinical, demographic, and laboratory parameters (HR: 3.85, 95 % CI: 1.53-9.72, p = 0.004). CONCLUSIONS: Serum ILK shows potential as a novel strong and independent prognostic marker for postoperative survival in surgically amenable NSCLC.

Selective regulation of p38ß protein and signaling by integrin-linked kinase mediates bladder cancer cell migration.

Integrin-linked kinase (ILK) and p38(MAPK) are protein kinases that transduce extracellular signals regulating cell migration and actin cytoskeletal organization. ILK-dependent regulation of p38(MAPK) is critical for mammalian kidney development and in smooth muscle cell migration, however, specific p38 isoforms has not been previously examined in ILK-regulated responses. Signaling by ILK and p38(MAPK) is often dysregulated in bladder cancer, and here we report a strong positive correlation between protein levels of ILK and p38ß, which is the predominant isoform found in bladder cancer cells, as well as in patient-matched normal bladder and tumor samples. Knockdown by RNA interference of either p38ß or ILK disrupts serum-induced, Rac1-dependent migration and actin cytoskeletal organization in bladder cancer cells. Surprisingly, ILK knockdown causes the selective reduction in p38ß cellular protein level, without inhibiting p38ß messenger RNA (mRNA) expression. The loss of p38ß protein in ILK-depleted cells is partially rescued by the 26S proteasomal inhibitor MG132. Using co-precipitation and bimolecular fluorescent complementation assays, we find that ILK selectively forms cytoplasmic complexes with p38ß. In situ proximity ligation assays further demonstrate that serum-stimulated assembly of endogenous ILK-p38ß complexes is sensitive to QLT-0267, a small molecule ILK kinase inhibitor. Finally, inhibition of ILK reduces the amplitude and period of serum-induced activation of heat shock protein 27 (Hsp27), a target of p38ß implicated in actin cytoskeletal reorganization. Our work identifies Hsp27 as a novel target of ILK-p38ß signaling complexes, playing a key role in bladder cancer cell migration.

Integrin-linked kinase: not so 'pseudo' after all.

Integrin-linked kinase (ILK) is a highly evolutionarily conserved intracellular protein that was originally identified as an integrin-interacting protein, and extensive genetic and biochemical studies have shown that ILK expression is vital during both embryonic development and tissue homeostasis. At the cellular and tissue levels, ILK regulates signaling pathways for cell adhesion-mediated cell survival (anoikis), apoptosis, proliferation and mitosis, migration, invasion, and vascularization and tumor angiogenesis. ILK also has central roles in cardiac and smooth-muscle contractility, and ILK dysregulation causes cardiomyopathies in humans. ILK protein levels are increased in several human cancers and often the expression level predicts poor patient outcome. Abundant evidence has accumulated suggesting that, of the diverse functions of ILK, some may require kinase activity whereas others depend on protein-protein interactions and are, therefore, independent of kinase activity. However, the past several years have seen an ongoing debate about whether ILK indeed functions as a protein serine/threonine kinase. This debate centers on the atypical protein kinase domain of ILK, which lacks some amino-acid residues thought to be essential for phosphotransferase activity. However, similar deficiencies are present in the catalytic domains of other kinases now known to possess protein kinase activity. Numerous studies have shown that ILK phosphorylates peptide substrates in vitro, corresponding to ILK-mediated phosphorylations in intact cells, and a recent report characterizing in vitro phosphotransferase activity of highly purified, full-length ILK, accompanied by detailed enzyme kinetic analyses, shows that, at least in vitro, ILK is a bona fide protein kinase. However, several genetic studies suggest that, not all biological functions of ILK require kinase activity, and that it can function as an adaptor/scaffold protein. Here, we review evidence for and against ILK being an active kinase, and provide a framework for strategies to further analyze the kinase and adaptor functions of ILK in different cellular contexts.

Integrin-linked kinase (ILK) is highly expressed in first trimester human chorionic villi and regulates migration of a human cytotrophoblast-derived cell line.

The placenta represents a critically important fetal-maternal interaction. Trophoblast migration and invasion into the uterine wall is a precisely controlled process and aberrations in these processes are implicated in diseases such as preeclampsia. Integrin-linked kinase (ILK) is a multifunctional, cytoplasmic, serine/threonine kinase that has been implicated in regulating processes such as cell proliferation, survival, migration, and invasion; yet the temporal and spatial pattern of expression of ILK in human chorionic villi and its role in early human placental development are completely unknown. We hypothesized that ILK would be expressed in trophoblast subtypes of human chorionic villi during early placental development and that it would regulate trophoblast migration. Immunoblot analysis revealed that ILK protein was highly detectable in placental tissue samples throughout gestation. In floating branches of chorionic villi, from 6 to 15 wk of gestation immunofluorescence analysis of ILK expression in placental tissue sections demonstrated that ILK was highly detectable in the cytoplasm and membranes of villous cytotrophoblast cells and in stromal mesenchyme, whereas it was barely detectable in the syncytiotrophoblast layer. In anchoring branches of villi, ILK was highly localized to plasma membranes of extravillous trophoblast cells. Transient expression of dominant negative E359K-ILK in the villous explant-derived trophoblast cell line HTR8-SVneo dramatically reduced migration into wounds compared to cells expressing wild-type ILK or empty vector. Therefore, our work has demonstrated that ILK is highly expressed in trophoblast subtypes of human chorionic villi during the first trimester of pregnancy and is a likely mediator of trophoblast migration during this period of development.

Modulation of integrin signal transduction by ILKAP, a protein phosphatase 2C associating with the integrin-linked kinase, ILK1.

ILKAP, a protein serine/threonine (S/T) phosphatase of the PP2C family, was isolated in a yeast two-hybrid screen baited with integrin-linked kinase, ILK1. Association of ILK1 and ILKAP was independent of the catalytic activity of either partner, as assayed in co-precipitation and two-hybrid experiments. Condi tional expression of ILKAP in HEK 293 cells resulted in selective inhibition of ECM- and growth factor-stimulated ILK1 activity, but did not inhibit Raf-1 kinase activity. A catalytic mutant of ILKAP, H154D, did not inhibit ILK1 kinase activity. Two cellular targets of ILK1, glycogen synthase kinase 3 beta (GSK3beta) and protein kinase B (PKB)/AKT, were differentially affected by ILKAP-mediated inhibition of ILK1. Catalytically active, but not mutant ILKAP, strongly inhibited insulin-like growth factor-1-stimulated GSK3beta phosphorylation on Ser9, but did not affect phosphorylation of PKB on Ser473, suggesting that ILKAP selectively affects ILK-mediated GSK3beta signalling. Consistent with this, active, but not H154D mutant or the related PP2Calpha, selectively inhibited transactivation of a Tcf/Lef reporter gene, TOPFlash, in 293 cells. We propose that ILKAP regulates ILK1 activity, targeting ILK1 signalling of Wnt pathway components via modulation of GSK3beta phosphorylation.

Mapping of the gene encoding the integrin-linked kinase, ILK, to human chromosome 11p15.5-p15.4.

We have recently reported the identification and cloning of the gene encoding p59ILK, a novel protein ser/thr kinase that is found in physiologic complexes with beta integrin subunits. ILK is a potential protoonocogene that appears to function in mediating signal transduction by beta 1 family integrins. Fluorescence in situ hybridization analysis of metaphase and decondensed free chromatin fibers localized ILK to 11p15.5-p15.4. This position was also confirmed by relational mapping using well-characterized translocations with breakpoints in chromosome band 11p15. Our results indicate that ILK maps between HBBC and CALC loci, in the 11p15.5-p15.4 band interval. This location may be important in evaluating the potential role of p59ILK in tumorigenesis since it has been shown that this region is associated with both genomic imprinting and loss of heterozygosity in certain types of tumor.

Rescue of a Drosophila NF1 mutant phenotype by protein kinase A.

The neurofibromatosis type 1 (NF1) tumor suppressor protein is thought to restrict cell proliferation by functioning as a Ras-specific guanosine triphosphatase-activating protein. However, Drosophila homozygous for null mutations of an NF1 homolog showed no obvious signs of perturbed Ras1-mediated signaling. Loss of NF1 resulted in a reduction in size of larvae, pupae, and adults. This size defect was not modified by manipulating Ras1 signaling but was restored by expression of activated adenosine 3', 5'-monophosphate-dependent protein kinase (PKA). Thus, NF1 and PKA appear to interact in a pathway that controls the overall growth of Drosophila.

Protein kinase mediators of integrin signal transduction.

Protein kinases are important mediators of signal transduction initiated by soluble growth factors and cytokines. Cellular interactions with the extracellular matrix are mediated largely by members of the integrin class of cell adhesion molecules, which also subsume signal transduction functions required for cell growth, differentiation, and survival. Here we review the involvement of protein kinases in mediating integrin intracellular signal transduction and the possible role for these molecules in regulating integrin adhesion. Although in most cases mechanistic details are incomplete, the emerging theme of protein kinases mediating cross-talk between growth factor receptor and integrin signalling systems provides a timely backdrop against which to present new developments in this area. The contribution of the actin cytoskeleton to integrin signal transduction is discussed, with respect to the concept of 'solid-state' signalling providing a mechanism for imposing order on the protein-protein interactions which underlie signal discrimination. Moreover, we review evidence that dysregulated integrin signalling contributes to pathological processes including arthritis, thrombasthenia, leucocyte adhesion deficiencies, and tumour angiogenesis and invasion.

Integrin cytoplasmic interactions and bidirectional transmembrane signalling.

Integrins are heterodimeric integral plasma membrane proteins containing extracellular, transmembrane, and cytoplasmic domains. These highly versatile receptors mediate not only cell adhesion and migration, but also the bidirectional transfer of information across the plasma membrane. The cytoplasmic domains of integrins are required for the transduction of this bidirectional information, and have recently been shown to participate in direct interactions with some novel cytoplasmic proteins, such as an ankyrin repeat containing serine/threonine protein kinase (integrin-linked kinase) and beta3 endonexin. New evidence also suggests that, via interactions with focal adhesion kinase, the integrin cytoplasmic domains can coordinate actin cytoskeletal organization and responses to growth factors. The elucidation of the signal transduction pathways activated by integrins is an intense area of investigation that has shown that integrins have some unique properties as signal transducing receptors.

Regulation of cell adhesion and anchorage-dependent growth by a new beta 1-integrin-linked protein kinase.

The interaction of cells with the extracellular matrix regulates cell shape, motility, growth, survival, differentiation and gene expression, through integrin-mediated signal transduction. We used a two-hybrid screen to isolate genes encoding proteins that interact with the beta 1-integrin cytoplasmic domain. The most frequently isolated complementary DNA encoded a new, 59K serine/threonine protein kinase, containing four ankyrin-like repeats. We report here that this integrin-linked kinase (ILK) phosphorylated a beta 1-integrin cytoplasmic domain peptide in vitro and coimmunoprecipitated with beta 1 in lysates of mammalian cells. Endogenous ILK kinase activity was reduced in response to fibronectin. Overexpression of p59ILK disrupted epithelial cell architecture and inhibited adhesion to integrin substrates, while inducing anchorage-independent growth. We propose that ILK is a receptor-proximal protein kinase regulating integrin-mediated signal transduction.

Mouse neurofibromatosis type 1 cDNA sequence reveals high degree of conservation of both coding and non-coding mRNA segments.

To identify evolutionary conserved domains and facilitate the recognition of potentially significant mutations in NF1 patients or tumors, we have determined the complete approximately 12 kb sequence of mouse neurofibromatosis type 1 mRNA. The sequence predicts a 2841 amino acid protein that is more than 98% identical to human neurofibromin. All but 9 of the 45 amino acid differences between mouse and human neurofibromin occur in the N-terminal half of the protein, with 16 changes clustered just upstream of the IRA-related segment. Given the high degree of sequence identity, virtually any sequence alteration in NF1 patients or tumors is potentially significant. We have also found that the 3' untranslated segment of NF1 mRNA is highly conserved, suggesting that this region may also be a target for mutations in NF1 patients.

Neurofibromatosis type 1 gene mutations in neuroblastoma.

The introduction of human chromosome 17 suppresses the tumourigenicity of a neuroblastoma cell line in the absence of any effects on in vitro growth and the neurofibromatosis type 1 (NF1) gene may be responsible. Here we report that 4 out of 10 human neuroblastoma lines express little or no neurofibromin and that two of these lines show evidence of NF1 mutations, providing further proof that NF1 mutations occur in tumours that are not commonly found in NF1 patients. We also show that NF1 deficient neuroblastomas show only moderately elevated ras-GTP levels, in contrast to NF1 tumour cells, indicating that neurofibromin contributes differently to the negative regulation of ras in different cell types.

Complete human NF1 cDNA sequence: two alternatively spliced mRNAs and absence of expression in a neuroblastoma line.

Neurofibromatosis type 1 (NF1) is caused by mutations in a large gene on chromosome 17q11.2. Previously described partial cDNAs for this gene predicted a protein related to yeast IRA1/IRA2 and the mammalian RAS GTPase activator protein GAP. To initiate a detailed study of the role of this gene in NF1, we have characterized a set of overlapping cDNAs that represent its complete coding sequence. Our results show that two differentially expressed human NF1 mRNAs differ by a 63-bp insertion in the GAP-related domain. These mRNAs predict two 2,818- and 2,839-amino acid proteins with calculated molecular masses of approximately 317 and 319 kD. Extensive similarity to IRA proteins is evident in a 1,450-amino-acid central segment, roughly between amino acids 900 and 2,350. However, the remainder of the NF1 protein is not significantly similar to other proteins. Interestingly, the SK-N-SH human neuroblastoma line expresses no detectable NF1 mRNA, indicating that expression of NF1 is not essential for viability of this neural crest-derived tumor cell line.

Interferon-alpha activates binding of nuclear factors to a sequence element in the c-fos proto-oncogene 5'-flanking region.

Interferon-alpha (IFN-alpha) can regulate the expression of the c-fos proto-oncogene in different cell types. Here we show IFN-alpha-activated binding of murine and human fibroblast nuclear factors to a DNA sequence element located in the 5' upstream region (nucleotides -351/-337) of the c-fos gene. This element, like the conserved enhancer element, the IFN-stimulated response element (ISRE), that mediates transcriptional induction of IFN-alpha-inducible genes, also binds factors in response to platelet-derived growth factor (PDGF) and v-sis-conditioned medium (SCM). The IFN-inducible ISRE shares an 8-bp stretch of sequence homology with the IFN-responsive c-fos SCM element, and competes efficiently for binding of factors to the SCM. Protein-DNA cross-linking experiments with the SCM binding site identified an IFN-modulated nuclear protein of approximately 98 kD. This protein does not appear to be involved in transcription activation, since IFN-alpha failed to stimulate c-fos transcription in nuclear run-off assays, or the c-fos promoter in transient transcription assays of 3T3 fibroblasts. Our data nonetheless suggest the c-fos promoter may be an early target for signal transduction triggered by IFN alpha-receptor interaction.

Enhancement by retinoic acid and dibutyryl cyclic adenosine 3':5'-monophosphate of the differentiation and gene expression of human neuroblastoma cells induced by interferon.

Human neuroblastoma cell lines are induced to differentiate and display neuronal phenotypes when treated with interferon (IFN)-alpha 2, retinoic acid (RA), or dibutyryl cyclic AMP (dbcAMP). We investigated the effects of combinations of these agents in induction-differentiation in the neuroblastoma cell line, NUB-6. The inductive effect of IFN-alpha 2 was markedly enhanced when used in combination with RA or dbcAMP. In parallel, RA or dbcAMP also enhanced the level of 2'-5'-oligoadenylate (2-5A) synthetase, and enzyme induced by IFNs and implicated in their biological action. The levels of another IFN-inducible enzyme, p68 kinase, were not enhanced by the combination treatments. The enhancement effects appeared to be exerted largely at the posttranscriptional level as both RA and dbcAMP stabilized IFN-induced 2-5A synthetase mRNA, resulting in increased enzyme activity. Thus, the 2-5A synthetase system is likely involved in mediating the IFN-alpha 2-induced differentiation of neuroblastoma cells and may also mediate the enhancement effects of RA and dbcAMP on IFN activity in these cells. These results also provide a rational basis for establishing a combination therapeutic approach for the treatment of neuroblastoma.

Induced factor binding to the interferon-stimulated response element. Interferon-alpha and platelet-derived growth factor utilize distinct signaling pathways.

Interferon-alpha (IFN alpha) and platelet-derived growth factor (PDGF) each rapidly stimulate binding of nuclear factors from Balb/c 3T3 fibroblasts, to a 29-base pair regulatory sequence derived from the 5' upstream region of the murine 2-5A synthetase gene. This regulatory sequence contains a functional interferon-stimulated response element (ISRE) and also functions as a PDGF-responsive sequence. We show that IFN alpha induces binding of a protein of molecular mass 65 kDa to the ISRE. Constitutively expressed ISRE-binding proteins of 98 and 150 kDa are also demonstrated. Binding of inducible factors to the ISRE increases significantly within 15 min of IFN alpha or PDGF treatment. PDGF-induced binding is not mediated by IFN beta. The protein kinase inhibitors, staurosporine and K252a, block PDGF-induced ISRE binding and 2-5A synthetase gene expression. IFN alpha-induced ISRE binding and gene activation are not blocked by these inhibitors. Treatment of cells with 12-O-tetradecanoyl-13-acetate or dibutyryl cyclic AMP does not activate ISRE binding factors or 2-5A synthetase gene expression. PDGF responsiveness of the ISRE in vivo is also sensitive to staurosporine, indicating that inhibition of a protein kinase activity blocks the PDGF-specific transcriptional signal. Our data indicate the signal transduction pathway for IFN alpha-induced, ISRE-dependent transcription is distinct from the PDGF-induced ISRE response and is likely independent of cyclic AMP-dependent protein kinase and protein kinase C activities.

Signal transduction by interferon-alpha through arachidonic acid metabolism.

Molecular mechanisms that mediate signal transduction by growth inhibitory cytokines are poorly understood. Type I (alpha and beta) interferons (IFNs) are potent growth inhibitory cytokines whose biological activities depend on induced changes in gene expression. IFN-alpha induced the transient activation of phospholipase A2 in 3T3 fibroblasts and rapid hydrolysis of [3H]arachidonic acid (AA) from prelabeled phospholipid pools. The phospholipase inhibitor, bromophenacyl bromide (BPB), specifically blocked IFN-induced binding of nuclear factors to a conserved, IFN-regulated enhancer element, the interferon-stimulated response element (ISRE). BPB also caused a dose-dependent inhibition of IFN-alpha-induced ISRE-dependent transcription in transient transfection assays. Specific inhibition of AA oxygenation by eicosatetraynoic acid prevented IFN-alpha induction of factor binding to the ISRE. Treatment of intact cells with inhibitors of fatty acid cyclooxygenase or lipoxygenase enzymes resulted in amplification of IFN-alpha-induced ISRE binding and gene expression. Thus, IFN-alpha receptor-coupled AA hydrolysis may function in activation of latent transcription factors by IFN-alpha and provides a system for studying the role of AA metabolism in transduction of growth inhibitory signals.

Interferon and growth factor modulation of nuclear factors binding to 5' upstream elements of the 2-5A synthetase gene.

We assayed fragments of the 5' flanking sequence of the human 2-5A synthetase gene for their ability to respond to interferon-alpha (IFN) and platelet-derived growth factor (PDGF). Transient transfection assays identified a 40-base pair fragment, which, regardless of orientation, could confer IFN-inducibility on the thymidine kinase promoter. This same fragment was active in monkey and mouse cells and in the latter was responsive to PDGF. The effect of PDGF could be inhibited by anti-interferon antibodies. Gel retardation assays, using the 40-base pair probe, detected the presence of IFN-modulated DNA-binding factors in nuclear extracts from monkey cells. In mouse cells both IFN and PDGF induced the binding of nuclear factors to a synthetic 2-5A synthetase response sequence. Thus, both IFN and growth factors directly or indirectly modulate the binding of nuclear factors to the same region of the 2-5A synthetase gene.

The interaction of interferon-alpha and -gamma: regulation of (2-5)A synthetase activity.

The interaction of IFN-alpha and IFN-gamma on the induction of antiviral, growth inhibitory, and (2-5)A synthetase activities was investigated in T98G and A549 cells. Synergistic or inhibitory effects which were host-cell dependent were seen with both simultaneous and sequential IFN treatments. Interestingly, IFN-gamma 1 pretreatment affected the levels of IFN-alpha-induced (2-5)A synthetase activity differently in the two cell types. In T98G cells, sequential treatment resulted in an overall decrease in induction of enzyme activity that was not observed in A549 cells. Receptor binding assays and measurements of steady-state mRNA levels indicated that this effect of IFN-gamma pretreatment does not occur at the level of either receptor expression or induction of transcription of the low-molecular-weight form of (2-5)A synthetase.

Interferon-induced binding of nuclear factors to promoter elements of the 2-5A synthetase gene.

Fragments of the 5'-flanking sequence of a human 2-5A synthetase gene were assayed for their ability to respond to interferon-alpha (IFN). Transient transfection assays in monkey cells demonstrated that the 5' boundary of the sequence required for IFN-regulated transcription is, at most, 155 nucleotides upstream from the presumed translational initiation codon. The 3' boundary of this sequence lies within a region of multiple transcription start sites preceded by no obvious TATA box. Binding assays, using a 40-bp probe derived from this IFN-responsive sequence, demonstrated the presence of three IFN-modulated, DNA-factor band shifts using nuclear extracts prepared from human and monkey cells. The induction of these complexes in human cells by IFN occurs with kinetics which closely parallel those previously observed for the transcriptional activation of the 2-5A synthetase gene by IFN. In vivo competition assays showed that the same 40-bp region which bound IFN-modulated factors could decrease the IFN-induced activity of a co-transfected 2-5A synthetase promoter; this fragment, regardless of its orientation, could confer IFN-inducibility on a heterologous promoter.