[Experimental study on compound of xenogenic inorganic bone and auto-marrow encapsuled by pedicled muscular flap].

OBJECTIVE: To explore the possibility of prefabricate a vascularized artificial bone-muscular flap, a compound of xenogenic inorganic bone and auto-marrow, with capability of osteogenesis. METHODS: Twelve 6-month New Zealand rabbits were divided into two groups, compound of xenogenic inorganic bone and auto-marrow implanted into the muscle near to radia and ulna of left forefoot as experimental group. Simple xenogenic inorganic bone implanted into the same site of right side as control group. After 2, 8, 12 weeks, x-ray examination, gross evaluation and histological observation were carried out. RESULTS: X-ray film showed that the implant had a cancellus-like density, and no variance with time. Since 2 weeks after implantation, gross observation showed a complete fusion formed between the implant and the host muscular tissue, with obvious blood vessels on the surface of the muscular flap in which the compound was encapsulated. No necrosis was observed in the following 30 minutes after the compound was dissected from its surrounding tissues. The histological investigation showed a contact connection between implant and its surroundings. In the experimental group, no new bone formed but in-growth of blood vessels was observed at the end of the 2nd week, and a little new bone formed along the edge of xenogenic inorganic bone at the end of the 8th week, while at the end of 12th week, more new bone formed in the compound with osteocytes in bone lacuna, plenty blood vessels in bone matrix, and lots of osteoblasts surrounded by un-differentiated mesenchymal cells at the fringe of the new bone. While in the control group, there were only loose connective tissue with blood vessels grew into xenogenic inorganic bone with no new bone formation until 12 weeks. CONCLUSION: The compound of xenogenic inorganic bone and auto-marrow can promote the formation of vascularized myo-bone flap with new bone formation.

Silencing of the VHL tumor-suppressor gene by DNA methylation in renal carcinoma.

Mutational inactivation and allelic loss of the von Hippel-Lindau (VHL) gene appear to be causal events for the majority of spontaneous clear-cell renal carcinomas. We now show that hypermethylation of a normally unmethylated CpG island in the 5' region provides another potentially important mechanism for inactivation of the VHL gene in a significant portion of these cancers. This hypermethylation was found in 5 of 26 (19%) tumors examined. Four of these had lost one copy of VHL while one retained two heavily methylated alleles. Four of the tumors with VHL hypermethylation had no detectable mutations, whereas one had a missense mutation in addition to hypermethylation of the single retained allele. As would be predicted for the consequence of methylation in this 5' CpG island, none of the 5 tumors expressed the VHL gene. In contrast, normal kidney and all tumors examined with inactivating VHL gene mutations but no CpG island methylation had expression. In a renal cell culture line, treatment with 5-aza-2'-deoxycytidine resulted in reexpression of the VHL gene. These findings suggest that aberrant methylation of CpG islands may participate in the tumor-suppressor gene inactivations which initiate or cause progression of common human cancers.

[Effect of lymphokine-activated killer cells of head and neck tumors in vitro and in nude mice].

Human lymphokine-activated killer (LAK) cells were prepared by culturing normal peripheral blood mononuclear cells (PBMC) for 4 days with or without added recombinant interleukin 2 (rIL-2) and assayed for anti-tumor activity against established squamous cell carcinoma lines of the head and neck (SCCHN) in vitro and in nude mice. The results showed that LAK cells had a stronger cytotoxicity than control cells to different SCCHN in vitro. In vivo, LAK cells almost completely inhibited the growth of HEP-2 tumor in nude mice and resulted in an inhibition rate of 81.6% (P < 0.01 with control cells). Our data indicate that LAK cells may be useful in the treatment of patients with head and neck tumors.

A naturally occurring secreted form of fibroblast growth factor (FGF) receptor 1 binds basic FGF in preference over acidic FGF.

Alternatively spliced variants of fibroblast growth factor receptor 1 mRNA are predicted to encode secreted forms and membrane-bound forms of receptors. The predicted amino acid sequences of these receptor variants differ in a portion of the extracellular region. In this study, we characterized the function of one of these splice variants which was predicted by its cDNA to be a secreted FGF receptor. We expressed this secreted form of the human FGFR1 (sFGFR1) in Chinese hamster ovary cells. The sFGFR1 protein oligomerized upon ligand binding. Surprisingly, the sFGFR1 preferentially bound basic FGF over acidic FGF. In cross-linking experiments, the sFGFR1 showed higher binding affinity for basic FGF (Kd approximately 30 nM) than for acidic FGF (Kd greater than 300 nM). These results suggest that this secreted form of FGF receptor has an unusual ligand binding specificity that may be important for its biological role in vivo.

Characterization of three T-lymphoid cell lines with distinct sensitivities to deoxyadenosine plus deoxycoformycin.

Three established human T-cell lines, HPB-MLT, HPB-ALL and PEER, were characterized and tested for their sensitivity to deoxyadenosine (dAdo) plus deoxycoformycin (dCoF). Phenotypic characterizations showed that all three cell lines had receptors for peanut agglutinin (PNA) and soybean agglutinin (SBA) while HPB-MLT and HPB-ALL, but not PEER, expressed the cortical thymocyte-specific marker, CD1. The majority of HPB-MLT cells (88%) expressed only CD4 but not CD8 antigen while most HPB-ALL cells (81%) co-expressed CD8 and CD4 antigens. PEER cells were negative for both CD8 and CD4. These three T cell lines showed differential sensitivity to dAdo plus dCoF in consequent tests. dAdo or dAdo plus dCoF (1 microM) had no effect on the growth, or DNA and RNA synthesis of HPB-MLT cells while the combination of dAdo and dCoF partially inhibited cellular growth and DNA and RNA synthesis of HPB-ALL cells. Further, the growth and DNA and RNA synthesis of PEER cells were strongly inhibited by the combination of dAdo and dCoF. This high sensitivity to dAdo plus dCoF reflected an immature phenotype of PEER cells despite its expression of CD3. Flow cytometric analysis of PEER cells demonstrated disappearance of the G2/M phase cells from the cell cycle after treatment with dAdo plus dCoF. Measurements of adenosine deaminase (ADA) and purine nucleoside phosphorylase (PNP) activities in all three cell lines, however, did not establish correlations between purine metabolizing enzyme activities and the differential sensitivities to dAdo plus dCoF. In sum, we report here three T-cell lines of different phenotypes that displayed significantly different sensitivities to dAdo plus dCoF which may facilitate investigations on the mechanisms of ADA deficiency.

Differential splicing in the extracellular region of fibroblast growth factor receptor 1 generates receptor variants with different ligand-binding specificities.

We have cloned a genomic region of the murine fibroblast growth factor (FGF) receptor 1 (FGFR1) gene that includes three alternative exons for the third immunoglobulinlike domain in the extracellular region of the receptor. The mRNA of one of these splice variants encodes a secreted receptor that lacks transmembrane and cytoplasmic sequences as well as a portion of the third immunoglobulinlike domain. Highest levels of mRNA encoding this variant were found in brain, skeletal muscle, and skin. We expressed this form of FGFR1 in CHO cells and showed that the recombinant secreted protein binds acidic FGF. We also discovered a novel alternative exon in the third immunoglobulinlike domain that encodes part of a transmembrane FGFR1 mRNA. This exon is highly homologous to the corresponding region of the keratinocyte growth factor receptor. Transcripts including this exon were present at highest levels in the skin. We cloned an FGFR1 cDNA which includes this exon and expressed this receptor variant in L6 rat skeletal muscle myoblasts. The new receptor variant had a 50-fold-lower affinity for basic FGF than does the published FGFR1 variant, whereas both forms of receptor bound acidic FGF with high affinity. These results show that the third immunoglobulinlike domain plays an important role in determining the binding specificities for different FGFs. Our data provide the first evidence that differential splicing in the extracellular region of a receptor gene generates receptor variants with different ligand-binding specificities.

A functional soluble extracellular region of the platelet-derived growth factor (PDGF) beta-receptor antagonizes PDGF-stimulated responses.

The platelet-derived growth factor beta-receptor (PDGFr) is a 180-kDa transmembrane glycoprotein which binds BB-PDGF with high affinity. We have expressed the extracellular region of the receptor in Chinese hamster ovary cells using an expression vector that carries a dihydrofolate reductase gene as an amplifiable marker. Upon amplification of the receptor cDNA sequences by methotrexate a 110-kDa soluble form of the receptor extracellular region (XR) was secreted at 12 mg/liter. The soluble XR protein fully retained the high affinity specific binding of the intact PDGFr for BB-PDGF (apparent dissociation constant, 0.4 nM). In the presence of ligand the soluble XR protein formed complexes that migrated on sodium dodecyl sulfate gels at the size expected for dimers of the protein. When added to fibroblast cultures the soluble XR protein blocked the ability of BB-PDGF to stimulate DNA synthesis but did not alter the mitogenic effect of AA-PDGF. The XR fragment also inhibited the binding of BB-PDGF to PDGFr and the activation of PDGFr tyrosine kinase by BB-PDGF. Thus, the soluble extracellular region protein of the PDGFr binds BB-PDGF with high affinity and functions as a specific antagonist of BB-PDGF actions.

Regulation of cyclic AMP by the mu-opioid receptor in human neuroblastoma SH-SY5Y cells.

The human neuroblastoma clonal cell line SH-SY5Y expresses both mu- and delta-opioid receptors (ratio approximately 4.5:1). Differentiation with retinoic acid (RA) was previously shown to enhance the inhibition of adenylyl cyclase (AC) by mu-opioid agonists. We tested here the inhibition of cyclic AMP (cAMP) accumulation by morphine under a variety of conditions: after stimulation with prostaglandin E1 (PGE1), forskolin, and vasoactive intestinal peptide (VIP), both in the presence and in the absence of the phosphodiesterase inhibitor 3-isobutyl-1-methylxanthine (IBMX). Morphine inhibition of the forskolin cAMP response (approximately 65%) was largely unaffected by the presence of IBMX. In contrast, deletion of IBMX enhanced morphine's inhibition of the PGE1 and VIP cAMP response from approximately 50 to approximately 80%. The use of highly mu- and delta-selective agents confirmed previous results that inhibition of cAMP accumulation by opioids is mostly mu, and not delta, receptor mediated in SH-SY5Y cells, regardless of the presence or absence of IBMX. Because of the large morphine inhibition and the high cAMP levels even in the absence of IBMX, PGE1-stimulated, RA-differentiated SH-SY5Y cells were subsequently used to study narcotic analgesic tolerance and dependence in vitro. Upon pretreatment with morphine over greater than or equal to 12 h, a fourfold shift of the PGE1-morphine dose-response curve was observed, whether or not IBMX was added. However, mu-opioid receptor number and affinity to the mu-selective [D-Ala2, N-Me-Phe4, Gly5-ol]enkephalin were largely unaffected, and Na(+)- and guanyl nucleotide-induced shifts of morphine-[3H]naloxone competition curves were unchanged.(ABSTRACT TRUNCATED AT 250 WORDS)

Mechanisms of 2'-deoxyguanosine toxicity in mouse T-lymphoma cells with purine nucleoside phosphorylase deficiency and resistance to inhibition of ribonucleotide reductase by dGTP.

Purine nucleoside phosphorylase (PNP; EC 2.4.2.1) deficiency is thought to cause T-lymphocyte depletion by accumulation of dG and dGTP, resulting in feedback inhibition of ribonucleotide reductase (RR; EC 1.17.4.1) and hence DNA synthesis. To test for additional toxic mechanisms of dG, we selected a double mutant of the mouse T-lymphoma S-49 cell line, dGuo-L, which is deficient in PNP and partially resistant to dGTP feedback inhibition of RR. The effects of dG on dGuo-L cells (concn. causing 50% inhibition, IC50 = 150 microM) were compared with those on the wild-type cells (IC50 = 30 microM) and the NSU-1 mutant with PNP deficiency only (IC50 = 15 microM). Fluorescence flow cytometry showed that equitoxic dG concentrations arrested wild-type and NSU-1 cells at the G1-S interface while allowing continued DNA synthesis in the S-phase, whereas the double mutant dGuo-L cells progressed through the cell cycle normally. dGuo-L cells accumulated high levels of dGTP in G1-phase, but not in S-phase cells, because of the utilization of dGTP for DNA synthesis and limited capacity to synthesize dGTP from dG. These results support the hypothesis that dG/dGTP toxicity occurs in the G1-phase or at the G1-S interface. Failure of dG to arrest the double mutant dGuo-L cells at the G1-S interface allows these cells to escape into S-phase, with an accompanying drop in dGTP levels. Thus the partial resistance of dGuo-L cells to dG toxicity may result from their shorter residence time in G1, allowing them to sustain higher dGTP levels. Hence RR inhibition by dGuo may not be the primary toxic mechanism in S-49 cells; rather, it may serve as an accessory event in dG toxicity by keeping the cells in the sensitive phase of the cell cycle. Among the possible targets of dG toxicity is RNA synthesis, which was inhibited at an early stage in dGuo-L cells.

Cytotoxicity of lymphokine-activated killer cells against human neuroblastoma cells: modulation by neuroblast differentiation.

The cytolytic activity of lymphokine-activated killer (LAK) cells against human neuroblastoma (NB) cells was investigated using the continuous NB cell lines, IMR-32, Kelly, and two subclones of SK-N-SH, SH-SY5Y (neuroblastic phenotype), and SH-EP (non-neuronal phenotype). NB cells were found to be sensitive targets of LAK. Of the SK-N-SH subclones, the neuroblasts, SH-SY5Y, were more susceptible to LAK killing than were the non-neuronal cells, SH-EP. Pretreatment of the targets SH-SY5Y and SH-EP with the differentiating agents, retinoic acid (RA, 10 microM), herbimycin A (236 nM), or nerve growth factor (10 ng/ml), did not substantially alter LAK killing. Furthermore, these differentiating agents did not measurably affect LAK activity during the cytolysis assay or with 1-h preincubation of the LAK effectors. However, co-incubation of the LAK cultures over the 3-day activation period with RA (1 microM) or PGE2 (1 microM) inhibited cytolysis by 80%, suggesting that these agents interfere with an early activation step of LAK. These results support the potential use of LAK treatment for neuroblastoma, in combination with differentiation agents that do not affect neuroblastoma sensitivities toward LAK cells. However, some differentiation agents, (e.g., RA) and endogenous prostaglandins (e.g., PGE2) may interfere with LAK activation.

Retinoic acid enhances VIP receptor expression and responsiveness in human neuroblastoma cell, SH-SY5Y.

Retinoic acid (RA) induces partial differentiation of neuroblastoma (NB) cells in vitro. In the human NB line, SH-SY5Y (a neuroblastic subclone of SK-N-SH), RA was previously shown to enhance the stimulatory (PGE1) and inhibitory (opioid) regulation of adenylyl cyclase. Since these cells are also sensitive to cAMP stimulation by vasoactive intestinal peptide (VIP), we have tested the effects of RA on VIP receptor expression and function. Pretreatment of SH-SY5Y cells with 10 microM RA over 6 days dramatically increased VIP receptor number from approximately 3,000 to approximately 70,000 sites per cell and enhanced threefold the cAMP accumulation after external VIP addition, while VIP immunoreactive content in the cells increased 2-3-fold. In the light of the recently proposed autocrine function of VIP in this cell lineage, the strong enhancement of the VIP system may contribute to the differentiation effects of RA.

Preoperative intermittent manual traction in congenital dislocation of the hip.

To avoid prolonged hospitalization and the complications of skin and skeletal traction, we have devised a method of intermittent preoperative manual traction in the treatment of congenital dislocation of the hip. Of 60 hips that were followed for more than 1 year, 88.3% had Class I or Class II results according to Severin's criteria. No hip in this series was complicated by avascular necrosis of the proximal femoral epiphysis.

Guanine for DNA synthesis. A compulsory route through ribonucleotide reductase.

Two alternative pathways for the synthesis of dGTP and its incorporation into DNA were studied: guanine (Gua)----GMP----GDP----dGDP----dGTP----DNA and dG----dGMP----dGDP----dGTP----DNA. To determine the contribution of each pathway to DNA synthesis independently of each other, [14C]Gua and [3H]dG tracer experiments were performed in a double-mutant S-49 mouse T-lymphoma cell line, dGuo-L, with purine nucleoside phosphorylase (EC 2.4.2.1)-deficiency and dGTP-feedback-resistant ribonucleotide reductase (RR, EC 1.17.4.1). In this cell line, dGTP pools can be selectively elevated by exogenous dG without affect RR and DNA synthesis. Although [3H]dG, but not [14C]Gua (up to 200 microM), readily expanded the cellular dGTP pool in a dose-dependent fashion in asynchronous cells, only a small fraction of the Gua flux into DNA was derived from [3H]dG, with the major fraction coming from [14C]Gua. H.p.l.c. analysis of G1- and partially enriched S-phase cells revealed that [3H]dGTP only accumulates in G1- but not in S-phase cells because of a rapid turnover of the dGTP pool during DNA synthesis. These results fail to provide evidence for cellular dGTP compartmentation and suggest that the pathway dG----dGMP----dGDP----dGTP alone has insufficient capacity to maintain DNA synthesis.

Distinct effects of adenine and guanine starvation on DNA synthesis associated with different pool sizes of nucleotide precursors.

Guanine nucleotide depletion primarily causes a drastic inhibition of DNA synthesis, while adenine nucleotide depletion interferes with other vital functions before inhibiting DNA synthesis (M. B. Cohen and W. Sadée, Cancer Res., 43: 1589-1591, 1983). This study addresses the hypothesis that the presence of a large adenine nucleotide pool with direct access to DNA synthesis prevents immediate cessation of DNA synthesis under conditions of adenine starvation, while the small guanine-DNA precursor pool is readily exhausted under guanine starvation. Adenine, guanine, and deoxyadenosine tracers were incubated with asynchronized or synchronized S-49 cells, and tracer progression into cellular nucleotide pools and nucleic acids was measured. Compartmentation of the dATP pool into a functional DNA precursor pool and a general cellular pool could not be demonstrated with [3H]dAdo tracer experiments with S-phase cells. While guanine tracer was incorporated into DNA without delay (less than 5 min), consistent with a small functional guanine-DNA precursor pool, adenine tracer incorporation into DNA was associated with a substantial delay period (approximately 30 min) indicative of a large functional adenine-DNA precursor pool. These results suggest that the different size of the functional nucleotide precursor pools with rapid access to DNA synthesis accounts for the dramatic difference in the effects of purine antimetabolites that cause either adenine or guanine starvation.