BACKGROUND

Fibroblast growth factor (FGF)-23 levels have been associated with impaired vasoreactivity, increased arterial stiffness, and cardiovascular morbi-mortality, whereas a protective function of KLOTHO against endothelial dysfunction has been reported. Since expression of the FGF23-KLOTHO system in human vascular tissue remains unproved, we aimed to study the expression of FGF23, FGF receptors (FGFR) and KLOTHO in human aorta. In addition, we analyzed the FGF23-KLOTHO expression in occlusive coronary thrombi.

METHODS

Thoracic aorta specimens from 44 patients underwent elective cardiac surgery, and thrombus material from 2 patients with acute coronary syndrome (ACS), were tested for FGF23-KLOTHO system expression.

RESULTS

Expression of KLOTHO (mean expression level 4.85 ± 5.43, arbitrary units) and two of the three cognate FGFR (FGFR-1 and -3) were detected and confirmed by RT-PCR, sequencing and qRT-PCR. KLOTHO expression was confirmed within occlusive coronary thrombi from patients with ACS. However, expression of FGF23 and FGFR4 was not observed. We also detected the aortic expression of membrane-anchored A Desintegrin and Metalloproteinases (ADAM)-17, the enzyme responsible for the shedding of KLOTHO from the cell surface, and the anti-inflammatory cytokine interleukin (IL)-10. Interestingly, in aortic samples there was a direct association between KLOTHO mRNA levels and those of ADAM-17 and IL-10 (r = 0.54, P<0.001; r = 0.51, P<0.01, respectively).

CONCLUSIONS

Human vascular tissue expresses members of the FGF23-KLOTHO system, indicating that it can be a direct target organ for FGF23. In addition, KLOTHO expression is also detected in occlusive coronary thrombi. These findings suggest a putative role of FGF23-KLOTHO axis in human vascular pathophysiology and cardiovascular disease.

OBJECTIVE

This study objectively characterized the microenvironment of indolent, chronic wounds by developing a method by which minute quantities of cytokines could be extracted from chronic wounds and separately identified.

BACKGROUND

Recombinant DNA technology and the ability to clone compounds such as cytokines allow new management schemes for the treatment of acute and chronic wounds. Before treatment with an exogenous cytokine is started, it would be helpful to know the endogenous level of that cytokine in the wound. Although various methods of extracting cytokines from acute wounds have been reported, no techniques have existed to reliably measure endogenous levels of cytokines in chronic wounds.

METHODS

Porous, inert hydrophilic dextranomer beads were tested for their ability to absorb or adsorb protein and cytokines in vitro with either albumin or albumin laced with various known amounts of cytokines, and then from chronic human pressure ulcers. The Bradford protein assay was used to determine protein levels. Enzyme-linked immunosorbent assay (ELISA) techniques were used to determine levels of platelet-derived growth factor (PDGF)-AB, basic fibroblast growth factor (bFGF), epidermal growth factor (EGF), and transforming growth factor-Beta (TGF-beta) extracted by the beads.

RESULTS

Between 88% and 98% of known amounts of albumin could be recovered. Similarly, more than 90% of the laced cytokines could be recovered. In 20 grade III/IV pressure ulcers, although protein concentrations were remarkably similar, endogenous levels of cytokine growth factors varied tremendously. Platelet-derived growth factor-AB ranged from 49 to 867 pg/mL; bFGF from 47 to 697 pg/mL; and EGF from nondetectable to 247.5 pg/mL. TGF-B was not detected in 17 of the 20 pressure ulcers.

CONCLUSIONS

This new technique appears useful for measuring endogenous levels of cytokines. Levels of cytokines found in these chronic wounds are much lower than those reported from acute wounds. The marked variation found among the 20 wounds may help to explain the differences reported in recent wound healing trials with exogenous cytokines.

OBJECTIVE

Advances in bone tissue engineering with mesenchymal stromal cells (MSC) as an alternative to conventional orthopedic procedures has opened new horizons for the treatment of large bone defects. Bone marrow (BM) and trabecular bone are both sources of MSC. Regarding clinical use, we tested the potency of MSC from different sources.

METHODS

We obtained MSC from <em>17</em> donors (mean age 64.6 years) by extensive washing of trabecular bone from the femoral head and trochanter, as well as BM aspirates of the iliac crest and trochanter. The starting material was evaluated by histologic analysis and assessment of colony-forming unit-<em>fibroblasts</em> (CFU-F). The MSC populations were compared for proliferation and differentiation potential, at RNA and morphologic levels.

RESULTS

MSC proliferation potential and immunophenotype (expression of CD49a, CD73, CD90, CD105, CD146 and Stro-1) were similar whatever the starting material. However, the differentiation potential of MSC obtained by bone washing was impaired compared with aspiration; culture-amplified cells showed few Oil Red O-positive adipocytes and few mineralized areas and formed inconsistent Alcian blue-positive high-density micropellets after growth under adipogenic, osteogenic and chondrogenic conditions, respectively. MSC cultured with 1 ng/mL fibroblast growth factor 2 (FGF-2) showed better differentiation potential.

CONCLUSIONS

Trabecular bone MSC from elderly patients is not good starting material for use in cell therapy for bone repair and regeneration, unless cultured in the presence of FGF-2.

We have shown that basic <em>fibroblast</em> <em>growth</em> <em>factor</em> (FGF) is heterogeneously expressed by human pituitary adenomas and may be implicated as a <em>growth</em> stimulus for these tumors. There are four mammalian FGF receptor (FGFR) genes encoding a complex family of transmembrane tyrosine kinases. The prototypic receptor is composed of three Ig-like extracellular ligand-binding domains, a transmembrane domain, and a cytoplasmic split tyrosine kinase. Multiple forms of cell-bound or secretable isoforms of FGFR-1, -2, and -3 can be generated by cell- and tissue-specific alternative splicing, resulting in tissue-specific FGF function. Shifts in isoform expression accompany tumor progression in some systems. We examined the normal human adenohypophysis and 40 pituitary adenomas to determine the pattern of FGFR expression by reverse transcription-PCR; all tumors were characterized clinically and morphologically. Ribonucleic acid (RNA) was extracted from frozen tumor tissue and primers were used to distinguish messenger RNA of the secretable first Ig-like domain (I) and those of the transmembrane and kinase domains (K) of each FGFR subtype. The normal pituitary-expressed mRNAs for FGFR-1 I and K, FGFR-2 I and K, FGFR-3 I and K, and FGFR-4 I but not FGFR-4 K; this represents the first report of a truncated isoform of FGFR-4, indicating possible alternative polyadenylation sites in this receptor. Only 3 tumors had the same pattern of expression of the 4 FGFRs as the normal gland. Although all tumors expressed FGFR-1 I, 1 tumor did not express FGFR-1 K, suggesting the production of only a secretable form of FGFR-1 by this tumor. Four tumors were negative for FGFR-2 I and K; 6 expressed the secretable form only, and <em>17</em> expressed FGFR-2 K but not I. All tumors expressed FGFR-3 I; 14 had secretable forms only, and no tumors expressed FGFR-3 K alone. As in the normal gland, 13 tumors expressed only the secretable I form of FGFR-4. Unlike the normal pituitary, however, 22 expressed FGFR-4 I and K, indicating a possible tumor-specific transmembrane receptor. Five tumors were negative for FGFR-4 I and K. Expression of FGFR proteins was confirmed by immunohistochemical localization of the C-terminal portion of FGFR-1, -2, -3, and -4; the results correlated with the RNA data in each case. There was no correlation between tumor type, size, or aggressiveness and the expression pattern of FGFRs. Our study suggests that pituitary adenomas have altered FGFR subtype and isoform expression, which may determine their hormonal and proliferative responses to FGFs.

OBJECTIVE

To study the role of some soluble factors in the process of angiogenesis that accompanies multiple myeloma (MM).

METHODS

The concentrations of three well-known angiogenic peptides, vascular endothelial growth factor (VEGF), basic fibroblast growth factor (bFGF), and hepatocyte growth factor (HGF) were evaluated by an ELISA method. All of these factors were measured in the plasma obtained from peripheral blood (PB) and bone marrow (BM) aspirates of 34 patients affected by plasma cell disorders. This series included one patient with a solitary extramedullary plasmacytoma, 17 patients with MM at diagnosis, and 16 with previously treated MM.

RESULTS

In all the patients, the concentration of each angiogenic factor was higher in bone marrow than in peripheral blood. Mean values of the three angiogenic factors in BM or in PB were lower in stage I than stage II-III. One patient with extramedullary solitary myeloma had high levels of VEGF and bFGF but this increase was not found in the other 6 patients with extramedullary disease when compared with patients without extramedullary disease. VEGF and bFGF did not correlate with each other while HGF showed a weak correlation with VEGF and a stronger one with bFGF. Moreover, VEGF correlated with features of disease activity, such as C-reactive protein, and 2-microglobulin, while both bFGF and HGF showed an inverse correlation with albumin level. No correlation was found between VEGF, bFGF and HGF levels and age, M protein level, osteolytic lesions, or percentage of BM plasma cells. Since angiogenic factors may be released by normal cells in response to hypoxia, we also evaluated erythropoietin (EPO) levels (which correlate with the hypoxic stimulus) both in PB and BM plasma of these patients but none of the measured angiogenic factors correlated with EPO levels. Interpretation and Conclusions. Several soluble factors may play a role in the angiogenic activity described in MM but their contribution to the progression of disease may be different. The finding of higher levels of these factors in BM than in PB might indicate that the bone marrow environment is their major source. Concentrations of angiogenic factors parallel the activity of disease and are independent of the hypoxic stimulus.

<em>Growth</em> <em>factors</em> and hormones may play an autocrine/paracrine role in mechanical stress-induced cardiac hypertrophy. Using an in vitro model of mechanical stress, i.e. stretch of cardiomyocytes and cardiac <em>fibroblasts</em>, we tested the involvement of <em>growth</em> <em>factors</em> and hormones in this process. We found that conditioned medium (CM) derived from 4 h cyclicly (1 Hz) stretched cardiomyocytes increased the rate of protein synthesis in static cardiomyocytes by 8 +/- 3%. Moreover, CM derived from 2 h stretched <em>fibroblasts</em> increased the rate of protein synthesis in static <em>fibroblasts</em> as well as in static cardiomyocytes by 8 +/- 2 and 6 +/- 2%, respectively. Analysis of CM using size-exclusion HPLC showed that cardiomyocytes and <em>fibroblasts</em> released at least three <em>factors</em> with MW < or = 10 kD, their quantities being time-dependently increased by stretch. Subsequent analyses using immunoassays revealed that cardiomyocytes released atrial natriuretic peptide (ANP) and transforming <em>growth</em> <em>factor</em>-beta1 (TGFbeta1) being increased by 45 +/- <em>17</em> and 21 +/- 4% upon 4 h of stretch, respectively. <em>Fibroblasts</em> released TGFbeta1 and very low quantity of endothelin-1 (ET-1). The release of TGFbeta1 was significantly increased by 18 +/- 4% after 24 h of stretch in <em>fibroblasts</em>. Both cell types released no detectable amount of angiotensin II (Ang II). In conclusion, upon cyclic stretch cardiomyocytes and <em>fibroblasts</em> secrete <em>growth</em> <em>factors</em> and hormones which induce <em>growth</em> responses in cardiomyocytes and <em>fibroblasts</em> in an autocrine/paracrine way. TGFbeta secreted by cardiomyocytes and <em>fibroblasts</em>, and ANP secreted by cardiomyocytes are likely candidates. We found no evidence for the involvement of Ang II and ET-1 in autocrine/paracrine mechanisms between cardiac cell types.

Tumor necrosis <em>factor</em> alpha (TNF-alpha) plays an important role in the host immune response to infection with the intracellular pathogen Mycobacterium tuberculosis. It is essential for the formation of protective tuberculous granulomas and regulates the expression of other cytokines which contribute to a protective immune response. Interleukin-12 (IL-12) is known to promote a Th1 response, which is essential for antimycobacterial resistance. Recombinant guinea pig TNF-alpha (rgpTNF-alpha) protein (<em>17</em> kDa) was purified, and its bioactivity was confirmed by its cytotoxicity for L929 <em>fibroblasts</em>. High titers of polyclonal anti-gpTNF-alpha antibody were obtained by immunization of rabbits. Resident alveolar and peritoneal macrophages were isolated from guinea pigs and infected with either the H37Ra or H37Rv strain of M. tuberculosis. The mRNA levels for TNF-alpha and IL-12 p40 were measured using real-time PCR. IL-12 p40 mRNA was up-regulated in a dose-dependent manner by rgpTNF-alpha alone. In infected macrophages, a lower dose of rgpTNF-alpha intensified the mRNA levels of TNF-alpha and IL-12 p40. However, higher doses of rgpTNF-alpha suppressed TNF-alpha and IL-12 p40 mRNA. The antimycobacterial activity of macrophages was assessed by metabolic labeling of M. tuberculosis with [3H]uracil. Resident alveolar and peritoneal macrophages treated with anti-gpTNF-alpha antibody to block endogenous TNF-alpha exhibited increased intracellular mycobacterial <em>growth</em>. These data suggest that the dose of TNF-alpha is crucial to the stimulation of optimal expression of protective cytokines and that TNF-alpha contributes to the control of mycobacterial replication to promote host resistance against M. tuberculosis.

Thrombospondin (TSP) is an extracellular matrix glycoprotein whose expression has been associated with a variety of cellular processes including <em>growth</em> and embryogenesis. The recent discovery of the existence of a second mouse TSP gene necessitates careful examination of the discrete biochemical and functional properties associated with each molecule. In this report, the primary structures of human TSP, mouse TSP1 (mTSP1), mouse TSP2 (mTSP2), and chicken TSP are compared; and the expression of mTSP1 and mTSP2 during embryogenesis and <em>growth</em> <em>factor</em>-mediated cell proliferation is examined. The cloning and sequencing of the entire coding regions of mTSP1 and mTSP2 revealed considerable conservation of residues critical for TSP structure and function; these data suggest that TSP2 is capable of trimer formation and many of the same cell-surface and ligand interactions that mediate TSP function. Comparison of the various TSP sequences also allowed the assignment based on sequence homology of previously reported human TSP as TSP1 and chicken TSP as TSP2. mTSP2, like mTSP1, was shown to be a primary response gene when quiescent Swiss 3T3 cells were stimulated with serum, platelet-derived <em>growth</em> <em>factor</em> BB, basic <em>fibroblast</em> <em>growth</em> <em>factor</em>, or interleukin-1 beta. Interestingly, TSP1 and TSP2 exhibited markedly different tissue- and stage-specific patterns of mRNA expression during mouse embryogenesis, implying that the two TSP molecules possess discrete functional properties important for development. Additionally, the TSP genes (Thbs1 and Thbs2) were mapped to single loci on mouse chromosomes 2 and <em>17</em>, respectively.

MCF-7 human breast cancer cells have been studied for hormonal regulation of secretion of an insulin <em>growth</em> <em>factor</em>-I (IGF-I)-related <em>growth</em> <em>factor</em>. <em>17</em> beta-Estradiol, which is required for tumorigenesis of the cell line in the nude mouse and which stimulates proliferation in vitro, was able to significantly induce IGF-I secretion at 10(-13) M, with maximal induction at 10(-11) M. Under optimal conditions IGF-I could be induced 4-fold after 4 days. Demonstration of estrogenic stimulations required removal of phenol red, a weak estrogen, from the cell culture medium. In addition to estrogen, insulin, epidermal <em>growth</em> <em>factor</em>, and transforming <em>growth</em> <em>factor</em> alpha induce both cellular proliferation and IGF-I secretion, while <em>growth</em> inhibitory antiestrogens, transforming <em>growth</em> <em>factor</em> beta, and glucocorticoids have the opposite effect. In each case, modulations in IGF-I secretion preceeded effects on cellular proliferation. IGF-I was not regulated by human GH, basic <em>fibroblast</em> <em>growth</em> <em>factor</em>, platelet-derived <em>growth</em> <em>factor</em>, or PRL, none of which affected proliferation rate. Thus, regulation of IGF-I secretion in human breast cancer is controlled by different hormones from those previously reported in human <em>fibroblasts</em>. Regulation of IGF-I by neither estrogen nor antiestrogen was associated with changes in steady-state mRNA levels; thus regulation may occur at a step beyond mRNA. We conclude that IGF-I production is tightly coupled to <em>growth</em> regulation by estrogens, antiestrogens, and other hormones and may contribute to autocrine and/or paracrine <em>growth</em> regulation by these agents in breast cancer.

Basic <em>fibroblast</em> <em>growth</em> <em>factor</em> (bFGF) is a secreted multifunctional cytokine and a potent stimulator of angiogenesis. We measured bFGF concentrations from serum samples taken from 103 patients with small cell lung cancer at the time of diagnosis. Serum concentration of bFGF (S-bFGF) ranged from undetectable to 54 pg/ml (median, 6 pg/ml). S-bFGF was not associated with age, sex, performance status, or stage. A high pretreatment S-bFGF was associated with poor overall survival. The 1- and 2-year survival rates of the patients within the highest quartile of S-bFGF >>or=<em>17</em> pg/ml) were only 26% and 11%, respectively, in contrast to the 49% and 20% 1- and 2-year survival rates of those patients with S-bFGF < <em>17</em> pg/ml (P = 0.013). The 1- and 2-year survival rates of the patients with extensive-stage disease were 33% and 10%, respectively (P = 0.0091). Interestingly, S-bFGF provided additional prognostic information to the stage because the 1- and 2-year survival rates of patients with extensive-stage disease and a high S-bFGF >>or=<em>17</em> pg/ml) were as low as 16% and 5%, respectively (P = 0.0026). Similarly, in the multivariate model of survival analysis, patients with both extensive-stage disease and a high S-bFGF >>or=<em>17</em> pg/ml) were found to have a particularly poor prognosis (relative risk of death, 2.1; 95% confidence interval, 1.2-3.6; P = 0.0057). We conclude that a high S-bFGF at diagnosis is associated with poor outcome in small cell lung cancer, possibly reflecting active angiogenesis and rapid tumor <em>growth</em>, and may complement prognostic information obtained by staging.

Dermatofibrosarcoma protuberans (DFSP) and its juvenile form, giant-cell fibroblastoma (GCF), are uncommon infiltrative tumors of the dermis, which present unique cytogenetic features, such as the reciprocal translocation t(<em>17</em>;22) or, more commonly, supernumerary ring chromosomes containing sequences from chromosomes <em>17</em> and 22. We have recently shown that these aberrations are cytogenetic manifestations of gene fusions between the platelet-derived <em>growth</em> <em>factor</em> B-chain gene (PDGFB), the cellular equivalent of the v-sis oncogene, and the collagen type 1 alpha 1 gene (COL1A1), the major protein constituent of the extracellular matrix in connective tissue of skin. We now report characterization of COL1A1/PDGFB chimeric genes at the RNA and DNA sequence levels in a series of DFSPs and GCFs. All 16 tumors studied contained the COL1A1/PDGFB gene. The location of breakpoints within COL1A1 varied greatly, but was always limited to the region encoding the alpha-helical domain. The PDGFB segment of the chimeric transcript always starts with exon 2, placing PDGFB under the control of the COL1A1 promoter and removing all known elements negatively controlling PDGFB transcription and translation. Production of these aberrant transcripts in <em>fibroblasts</em>, the suspected cell of origin of DFSP/GCF, likely causes autocrine stimulation and cell proliferation. No specific function has yet been assigned to exon 2 of PDGFB, and this exon does not encode for the mature <em>growth</em> <em>factor</em>. Its retention in all chimeric COL1A1/PDGFB genes suggests that it is important for the normal processing of the PDGFB polypeptide.

The identities of extracellular <em>growth</em> <em>factors</em> that regulate skeletal muscle development in vivo are largely unknown. We asked if FGFs, which act as repressors of myogenesis in culture, play a similar role in vivo by ectopically expressing in the developing limb a truncated FGF receptor 1 (dnFGFR1) that acts as a dominant negative mutant. Hind limbs and the adjacent somites of Hamburger and Hamilton (HH) stage <em>17</em> chickens were infected with a replication-competent RCAS virus encoding dnFGFR1. By ED5, the virus had spread extensively within the limb and the adjacent somites with little rostral or caudal expansion of the infection along the axial midline. Viral infection and mutant receptor expression were coincident as revealed by the distribution of a viral coat protein and an HA epitope tag present on the carboxy terminus of dnFGFR1. Within 48 h following injection of dnFGFR1, we could detect no obvious changes in skeletal muscle precursor cell migration into the hind limb as compared to control limbs infected with an empty RCAN virus. However, by 3 days following infection of RCAS-dnFGFR1 virus, the level of skeletal muscle-specific myosin heavy chain was decreased and the expression pattern altered, suggesting disruption of skeletal muscle development. Two striking muscular phenotypes were observed in dnFGFR1-expressing limbs, including an average loss of 30% in skeletal muscle wet weight and a 50% decrease in myofiber density. At all ages examined the loss of skeletal muscle mass was accompanied by a loss of myoblasts and an unexpected concomitant loss of <em>fibroblasts</em>. Consistent with these observations, explants of infected cells revealed a reduction in the number of myonuclei in myotubes. Although the myofiber density per unit area was decreased over 50% compared to controls there were no detectable effects on myofiber diameter. The loss in myofiber density was, however, accompanied by an increase in the space surrounding individual myofibers and a generalized loss of myofiber integrity. It is noteworthy that long-bone development was unaffected by RCAS-dnFGFR1 infection, suggesting that FGFR2 and FGFR3 signaling was not disrupted. Our data provide conclusive evidence that FGFR1 signaling is necessary to maintain myoblast number and plays a role in myofiber organization.

Fibroblasts that interact with cancer cells are called cancer-associated fibroblasts (CAFs), which promote progression of different tumor types. We investigated the characteristics and functions of CAFs in diffuse-type gastric cancers (DGCs) by analyzing features of their genome and gene expression patterns.

We isolated CAFs and adjacent non-cancer fibroblasts (NFs) from 110 gastric cancer (GC) tissues from patients who underwent gastrectomy in Japan from 2008 through 2016. Cells were identified using specific markers of various cell types by immunoblot and flow cytometry. We selected pairs of CAFs and NFs for whole-exome and RNA sequencing analyses, and compared expression of specific genes using quantitative reverse transcription PCR. Protein levels and phosphorylation were compared by immunoblot and immunofluorescence analyses. Rhomboid 5 homolog 2 (RHBDF2) was overexpressed from a transgene in fibroblasts or knocked down using small interfering RNAs. Motility and invasiveness of isolated fibroblasts and GC cell lines (AGS, KATOIII, MKN45, NUGC3, NUGC4, OCUM-2MD3 and OCUM-12 cell lines) were quantified by real-time imaging analyses. We analyzed 7 independent sets of DNA microarray data from patients with GC and associated expression levels of specific genes with patient survival times. Nude mice were given injections of OCUM-2MD3 in the stomach wall; tumors and metastases were collected and analyzed by immunohistochemistry.

Many of the genes with increased expression in CAFs compared with NFs were associated with transforming growth factor beta 1 (TGFB1) activity. When CAFs were cultured in extracellular matrix, they became more motile than NFs; DGC cells incubated with CAFs were also more motile and invasive in vitro than DGC cells not incubated with CAFs. When injected into nude mice, CAF-incubated DGC cells invaded a greater number of lymphatic vessels than NF-incubated DGC cells. We identified RHBDF2 as a gene overexpressed in CAFs compared with NFs. Knockdown of RHBDF2 in CAFs reduced their elongation and motility in response to TGFB1, whereas overexpression of RHBDF2 in NFs increased their motility in extracellular matrix. RHBDF2 appeared to regulate oncogenic and non-canonical TGFB1 signaling. Knockdown of RHBDF2 in CAFs reduced cleavage of the TGFB receptor 1 (TGFBR1) by ADAM metallopeptidase domain 17 (ADAM17 or TACE) and reduced expression of genes that regulate motility. Incubation of NFs with in interleukin 1 alpha (IL1A), IL1B or tumor necrosis factor, secreted by DGCs, increased fibroblast expression of RHBDF2. Simultaneous high expression of these cytokines in GC samples was associated with shorter survival times of patients.

In CAFs isolated from human DGCs, we observed increased expression of RHBDF2, which regulates TGFB1 signaling. Expression of RHBDF2 in fibroblasts is induced by inflammatory cytokines (such as IL1A, IL1B, and tumor necrosis factor) secreted by DGCs. RHBDF2 promotes cleavage of TGFBR1 by activating TACE and motility of CAFs in response to TGFB1. These highly motile CAFs induce DGCs to invade extracellular matrix and lymphatic vessels in nude mice.

OBJECTIVE

Matrix metalloproteinases (MMPs) are involved in tissue remodelling, which is one of the important aspects of inflammatory disease. To assess the balance between the matrix degradation and production, we analysed the in situ expression of MMP-1, -3, -8 and -9, tissue inhibitor of metalloproteinases (TIMP)-1 and -2, and type I procollagen (PC-I) in inflammatory bowel disease.

RESULTS

Immunohistochemistry using frozen sections was performed in <em>17</em> patients with ulcerative colitis (UC) and 16 with Crohn's disease (CD). In both UC and CD, MMPs and TIMPs were expressed by inflammatory cells as well as by fibroblastic cells most prominently in actively inflamed areas in ulcer bases, but sparsely in intact inflamed mucosa in both UC and CD. In UC, inflamed mucosa with erosions expressed these substances focally. <em>Fibroblasts</em> also expressed PC-I. We identified that vascular smooth muscle cells of venules in ulcer bases expressed MMP-1 and -9, TIMP-1 and PC-I. These venules also expressed E-selectin, a cell adhesion molecule to facilitate the leucocyte extravasation, and vascular endothelial <em>growth</em> <em>factor</em> (VEGF) receptor 2, consistent with their property of newly formed vessels.

CONCLUSIONS

Our results suggest that MMPs are involved in the tissue remodelling, angiogenesis and promotion of leucocyte extravasation in the actively inflamed area in the ulcer base in both UC and CD. MMP-1 expression in the mucosa may be related to the initial step of ulceration in UC. Therapeutic manipulation of extracellular matrix turnover would be an effective therapy to alleviate active inflammation and accelerate ulcer healing.

Attempts to stimulate therapeutic angiogenesis using gene therapy or delivery of recombinant <em>growth</em> <em>factors</em>, such as vascular endothelial <em>growth</em> <em>factor</em> (VEGF), have failed to demonstrate unequivocal efficacy in human trials. Bioactive glass stimulates <em>fibroblasts</em> to secrete significantly increased amounts of angiogenic <em>growth</em> <em>factors</em> and therefore has a number of potential applications in therapeutic angiogenesis. The aim of this study was to assess whether it is possible to encapsulate specific quantities of bioactive glass and <em>fibroblasts</em> into alginate beads, which will secrete <em>growth</em> <em>factors</em> capable of stimulating angiogenesis. Human <em>fibroblasts</em> (CCD-18Co) were encapsulated in alginate beads with specific quantities of 45S5 bioactive glass and incubated in culture medium (0-<em>17</em> days). The conditioned medium was collected and assayed for VEGF or used to assess its ability to stimulate angiogenesis by measuring the proliferation of human dermal microvascular endothelial cells. At <em>17</em> days the beads were lysed and the amount of VEGF retained by the beads measured. <em>Fibroblasts</em> encapsulated in alginate beads containing 0.01% and 0.1% (w/v) 45S5 bioactive glass particles secreted increased quantities of VEGF compared with cells encapsulated with 0% or 1% (w/v) 45S5 bioactive glass particles. Lysed alginate beads containing 0.01% and 0.1% (w/v) 45S5 bioactive glass contained significantly more VEGF (p<0.01) compared with beads containing no glass particles. Endothelial cell proliferation was significantly increased (p<0.01) by conditioned medium collected from alginate beads containing 0.1% (w/v) 45S5 bioactive glass particles. The results of this study demonstrate that bioactive glass and <em>fibroblasts</em> can be successfully incorporated into alginate beads for use in delivering angiogenic <em>growth</em> <em>factors</em>. With further optimization, this technique offers a novel delivery device for stimulating therapeutic angiogenesis.

Numerous studies of the proliferative effects of basic <em>fibroblast</em> <em>growth</em> <em>factor</em> (bFGF) in culture, including neonatal and adult hippocampal precursors, suggest that the <em>factor</em> plays a ubiquitous and life-long role in neurogenesis. In contrast, in vivo, bFGF is devoid of effects on neurons in mature hippocampus, raising the possibility that bFGF exhibits developmental stage-specific activity in the complex animal environment. To define neurogenetic effects in the newborn, a single subcutaneous injection of bFGF (20 ng/gm) was administered to postnatal day 1 (P1) rats, and hippocampal DNA content was quantified: bFGF elicited an increase in total DNA throughout adulthood, by 48% at P4, 25% at P22, and <em>17</em>% at P180, suggesting that bFGF increases hippocampal cell number. To define mechanisms, bromodeoxyuridine (BrdU) was injected at P1 and mitotically labelled cells were assessed at P22: there was a twofold increase in BrdU-positive cells in the dentate granule cell layer (GCL), indicating that bFGF enhanced the generation of neurons, or neuronogenesis, from a cohort of precursors. Moreover, enhanced mitosis and survival led to a 33% increase in absolute GCL neuron number, suggesting that neuron production depends on environmental levels of bFGF. To evaluate this possibility, bFGF-knockout mice were analyzed: hippocampal DNA content was decreased at all ages examined (P3, -42%; P21, -28%; P360, -18%), and total GCL neuron and glial fibrillary acidic protein (GFAP)-positive cell number were decreased by 30%, indicating that bFGF is necessary for normal hippocampal neurogenesis. We conclude that environmental levels of bFGF regulate neonatal hippocampal neurogenesis. As adult hippocampal neuronogenesis was unresponsive to bFGF manipulation in our previous study [Wagner, J.P., Black, I.B. & DiCicco-Bloom, E. (1999) J. Neurosci., 19, 6006], these observations suggest distinct, stage-specific roles of bFGF in the dentate gyrus granule cell lineage.

The effect of basic <em>fibroblast</em> <em>growth</em> <em>factor</em> (bFGF) on the healing of partial anterior cruciate ligament (ACL) lacerations was investigated in <em>17</em> adult mongrel canines. The defect was created in the midsubstance of both ACLs using a biopsy punch. In the bFGF group, a bFGF-impregnated pellet was sutured to the infrapatellar fat pad close to the defect. In the control group, the same pellet without bFGF was used. The healing process was evaluated macroscopically and histologically at 1, 3, 6, and 24 weeks postoperatively. In the bFGF group, a pannus-like tissue which contained abundant blood vessels extended into the defect from the adjacent synovial tissue in the early postoperative phase. Histological examination of the tissue which filled the defect revealed initial remodeling processes with a decreased number of cells and better orientation of the collagen fibers at 6-24 weeks. On the other hand, in the control group, poor healing processes were observed at each examination period. This study demonstrated that the application of a bFGF-impregnated pellet may enhance the healing potential of a partially injured ACL. Enhanced neovascularization and the formation of granulation tissue induced by bFGF early in the healing process accounted for the increased healing response.

Besides vascular endothelial <em>growth</em> <em>factor</em> (VEGF) and basic <em>fibroblast</em> <em>growth</em> <em>factor</em> (bFGF), matrix metalloproteinases (MMPs) play critical roles in angiogenesis, tumor invasion and metastasis. Increased angiogenesis is observed in chronic B lymphocytic leukemia (B-CLL) and published data reported VEGF and bFGF production in this disease. The purpose of this study was to investigate MMP expression in early stage B-CLL. Elevated MMP-9 concentrations were detected by ELISA in the sera of B-CLL patients (median level 250 ng/ml) compared with healthy donors (67 ng/ml) (P < 0.0001), and immunostaining with antibodies against MMP-9 and B cell antigens (CD19, CD23) substantiated the presence of MMP-9 in tumoral B lymphocytes. By using RT-PCR, ELISA and zymography experiments, we confirmed that B-CLL cells expressed and released the pro-form of MMP-9 with Mr 92 kDa (158-1300 pg/ml/10(6) cells/48 h), p-aminophenylmercuric acetate generating a 82 kDa active form. In contrast, the production of MMP-9 by normal counterpart B cells was significantly low (28-169 pg/ml/10(6)cells/48 h). Moreover, B-CLL culture supernatants contained bFGF (median levels <em>17</em> pg/ml/10(6) cells/48 h), VEGF (1.4 pg/ml/10(6) cells/48 h) and TNF-alpha (0.2 pg/ml/10(6) cells/48 h). TNF-alpha and VEGF antibodies blocked MMP-9 at the mRNA and protein levels. Interferons (IFNs) type I or type II repressed MMP-9 gelatinolytic activity in a dose and time dependency, and this was reflected by a parallel inhibition of MMP-9 mRNA and protein. IFNs however did not affect the production of bFGF, VEGF and TNF-alpha. Together, our data show that B-CLL lymphocytes synthesize MMP-9 and emphasize the specific inhibitory actions of IFNs on its expression.

In vertebrates, a number of <em>fibroblast</em> <em>growth</em> <em>factors</em> (FGFs) have been shown to play important roles in developing embryos and adult organisms. However, the molecular relationships of the vertebrate FGFs are not yet completely understood, partly due to the divergence of their amino acid sequences. To solve this problem, we have identified six FGF genes in a basal chordate, the ascidian Ciona intestinalis. A phylogenetic analysis confidently assigned two of them to vertebrate FGF8/<em>17</em>/18 and FGF11/12/13/14, respectively. Based on the presence of the conserved domains within or outside of the FGF domains, we speculate that three of the other genes are orthologous to vertebrate FGF3/7/10/22, FGF4/5/6 and FGF9/16/20, respectively, although we cannot assign the sixth member to any of the vertebrate FGFs. A survey of the raw whole genome shotgun sequences of C. intestinalis demonstrated the presence of no FGF genes other than the six genes in the genome. The identification of these six FGF genes in the basal chordate gave us an insight into the diversification of specific subfamilies of vertebrate FGFs.

OBJECTIVE

Effective systemic therapy options for carcinoid tumors are lacking. We conducted in vitro studies and a phase II clinical trial to explore the activity of imatinib in carcinoid tumors.

METHODS

Cells of the human bronchial carcinoid cell line NCI-H727 and the human pancreatic carcinoid cell line BON-1 were treated with increasing concentrations of imatinib using standard procedures to assess in vitro growth-inhibitory activity. A clinical trial using a two-stage phase II design to assess the response rate and safety profile of imatinib at a dose of 400 mg given twice daily in patients with advanced carcinoid tumors was completed.

RESULTS

In both cell lines, there was a dose- and time-dependent cytotoxic effect. The clinical trial enrolled 27 evaluable patients. Median duration on trial was 16 weeks. One patient had a partial response, 17 had stable disease, and 9 had progressive disease by the Response Evaluation Criteria in Solid Tumors criteria. Median progression-free survival time was 24 weeks. Median overall survival is 36 months. Seven patients who achieved a biochemical response had a superior progression-free survival time compared with patients without biochemical response (115 weeks compared with 24 weeks; P = 0.003). An increase in plasma basic fibroblast growth factor was associated with a shorter progression-free survival duration (P = 0.02).

CONCLUSIONS

Our data suggest that imatinib is active in vitro and has a modest clinical activity in carcinoid patients. Changes in tumor markers may help select patients who are likely to benefit from therapy.

The heat shock protein HSP90 serves as a chaperone for receptor protein kinases, steroid receptors, and other intracellular signaling molecules. Targeting HSP90 with ansamycin antibiotics disrupts the normal processing of clients of the HSP90 complex. The platelet-derived <em>growth</em> <em>factor</em> receptor alpha (PDGFRalpha) is a tyrosine kinase receptor up-regulated and activated in several malignancies. Here we show that the PDGFRalpha forms a complex with HSP90 and the co-chaperone cdc37 in ovarian, glioblastoma, and lung cancer cells. Treatment of cancer cell lines expressing the PDGFRalpha with the HSP90 inhibitor <em>17</em>-allylamino-<em>17</em>-demethoxygeldanamycin (<em>17</em>-AAG) promotes degradation of the receptor. Likewise, phospho-Akt, a downstream target, is degraded after treatment with <em>17</em>-AAG. In contrast, PDGFRalpha expression is not affected by <em>17</em>-AAG in normal human smooth muscle cells or 3T3 <em>fibroblasts</em>. PDGFRalpha degradation by <em>17</em>-AAG is inhibited by the proteasome inhibitor MG132. High molecular weight, ubiquitinated forms of the receptor are detected in cells treated with <em>17</em>-AAG and MG132. Degradation of the receptor is also inhibited by a specific neutralizing antibody to the PDGFRalpha but not by a neutralizing antibody to PDGF or by imatinib mesylate (Gleevec). Ultimately, PDGFRalpha-mediated cell proliferation is inhibited by <em>17</em>-AAG. These results show that <em>17</em>-AAG promotes PDGFRalpha degradation selectively in transformed cells. Thus, not only mutated tyrosine kinases but also overexpressed receptors in cancer cells can be targeted by <em>17</em>-AAG.

Locally produced <em>growth</em> <em>factors</em> may have important modulatory roles in final ovarian follicular <em>growth</em>. The aim of this study was to investigate the possible participation of vascular endothelial <em>growth</em> <em>factor</em> (VEGF) and basic <em>fibroblast</em> <em>growth</em> <em>factor</em> (FGF2) in bovine follicles during final <em>growth</em>. Ovaries were collected from a slaughterhouse within 10-20 min after exsanguination. A classification of follicles into five groups (<0.5; >0.5-5; >5-20; >20-180; >180 ng/ml) was performed according to the follicular fluid (FF) oestradiol-<em>17</em> beta content. For a better characterisation of classes the mRNA expressions of FSH receptor, LH receptor and aromatase cytochrome P450 in theca interna (TI) and granulosa cells (GC) were determined. Analysis of VEGF transcript by RT-PCR showed that GC and theca cells express predominantly the smallest isoforms (VEGF(121) and VEGF(165)). VEGF mRNA expression in both tissues (TI and GC) and VEGF protein concentration in total follicle tissue increased significantly (and correlated) with developmental stages of follicle <em>growth</em>. The expression of mRNA for VEGF receptor (VEGFR)-1 and VEGFR-2 was very weak in GC, without any regulatory change during final follicle <em>growth</em>. In contrast, TI showed strong expression of mRNA for both receptors in all follicle classes examined. VEGF protein concentrations in FF increased significantly and continuously to maximum levels in preovulatory follicles. As shown by immunohistochemistry, VEGF protein was clearly localised in TI and GC of preovulatory follicles. FGF2 and FGF receptor (FGFR) mRNA expression in TI increased significantly during final <em>growth</em> of follicles. In contrast, the FGF2 and FGFR mRNA expression in GC was very weak and without any regulatory change during follicle <em>growth</em>. Histological observation revealed that FGF2 protein was localised in theca tissue (cytoplasm of endothelial cells and pericytes) but not in GC. Our results suggest that VEGF and FGF families are involved in the proliferation of capillaries that accompanies the selection of the preovulatory follicle resulting in an increased supply of nutrients and precursors, and therefore supporting the <em>growth</em> of the dominant follicle.

Perlecan, a heparan sulfate proteoglycan, is widely distributed in developing and adult tissues and plays multiple, important physiological roles. Studies with knockout mouse models indicate that expression of perlecan and heparan sulfate is critical for proper skeletal morphogenesis. Heparan sulfate chains bind and potentiate the activities of various <em>growth</em> <em>factors</em> such as <em>fibroblast</em> <em>growth</em> <em>factor</em> 2 (FGF-2). Previous studies indicate that important biological activities are associated with the heparan sulfate-bearing domain I of perlecan (PlnDI; French et al. J. Bone Miner. Res. <em>17</em> , 48, 2002). In the present study, we have used recombinant, glycosaminoglycan-bearing PlnDI to reconstitute three-dimensional scaffolds of collagen I. Collagen I fibrils bound PlnDI much better than native collagen I monomers or heat-denatured collagen I preparations. Heparitinase digestion demonstrated that recombinant PlnDI was substituted with heparan sulfate and that these heparan sulfate chains were critically important not only for efficient integration of PlnDI into scaffolds, but also for FGF-2 binding and retention. PlnDI-containing collagen I scaffolds to which FGF-2 was bound sustained <em>growth</em> of both MG63, an osteoblastic cell line, and human bone marrow stromal cells (hBMSCs) significantly better than scaffolds lacking either PlnDI or FGF-2. Collectively, these studies demonstrate the utility of PlnDI in creating scaffolds that better mimic natural extracellular matrices and better support key biological activities.

Cutaneous infantile hemangioma progresses through proliferation and involution phases. Since treatment with interferon, a negative regulator of angiogenesis, accelerates the involution phase, we hypothesized that cutaneous infantile hemangioma is associated with an imbalance between endogenous positive and negative regulators of angiogenesis. We examined 30 specimens of cutaneous hemangioma [proliferative phase (n=15), involuting phase (n=8), and involuted phase (n=7)] and control human skin (n=<em>17</em>), fixed in formalin and embedded in paraffin. Routine histology, immunohistochemistry, and an mRNA in situ hybridization technique were used to measure expression of the positive angiogenic molecules basic <em>fibroblast</em> <em>growth</em> <em>factor</em> (bFGF) and vascular endothelial <em>growth</em> <em>factor</em>/vascular permeability <em>factor</em> (VEGF/VPF), and an endogenous inhibitor of angiogenesis, interferon-beta (IFN-beta). Proliferative phase hemangiomas expressed high levels of bFGF and VEGF/VPF but not IFN-beta (mRNA and protein). The epidermis directly overlying proliferating hemangiomas was hyperplastic, contained numerous dividing cells, and expressed bFGF and VEGF/VPF but not IFN-beta. Epidermis from normal individuals and epidermis directly overlying involuted tumors or at sites distant to the proliferating hemangioma was not hyperplastic and expressed normal levels of bFGF, VEGF/VPF, and IFN-beta. These data suggest that the proliferation of cutaneous hemangiomas and adjacent epidermis is associated with an imbalance between positive and negative angiogenic <em>factors</em> expressed by the neoplasm and adjacent normal tissue.