Autocrine and juxtacrine effects of amphiregulin on the proliferative, invasive, and migratory properties of normal and neoplastic human mammary epithelial cells

Amphiregulin (AR) autocrine loops have been associated with several types of cancer. We demonstrate that SUM149 breast cancer cells have a self-sustaining AR autocrine loop. SUM149 cells are epidermal growth factor (EGF)-independent for growth, and they overexpress AR mRNA, AR membrane precursor protein, and secreted AR relative to the EGF-dependent human mammary epithelial cell line MCF10A. MCF10A cells made to overexpress AR (MCF10A AR) are also EGF-independent for growth. Treatment with the pan-ErbB inhibitor CI1033 and the anti-EGF receptor (EGFR) antibody C225 demonstrated that ligand-mediated activation of EGFR is required for SUM149 cell proliferation. AR-neutralizing antibody significantly reduced both SUM149 EGFR activity and cell proliferation, confirming that an AR autocrine loop is required for mitogenesis in SUM149 cells. EGFR tyrosine phosphorylation was dramatically decreased in both SUM149 and MCF10A AR cells after inhibition of AR cleavage with the broad spectrum metalloprotease inhibitor GM6001, indicating that an AR autocrine loop is strictly dependent on AR cleavage in culture. However, a juxtacrine assay where fixed SUM149 cells and MCF10A AR cells were overlaid on top of EGF-deprived MCF10A cells showed that the AR membrane precursor can activate EGFR. SUM149 cells, MCF10A AR cells, and MCF10A cells growing in exogenous AR were all considerably more invasive and motile than MCF10A cells grown in EGF. Moreover, AR up-regulates a number of genes involved in cell motility and invasion in MCF10A cells, suggesting that an AR autocrine loop contributes to the aggressive breast cancer phenotype.     

Differential sensitivity of specific and nonspecific antigen-presentation by B cells to a protein synthesis inhibitor

Specific and nonspecific Ag-presentation by B cells was examined for the sensitivity to the treatment with emetin, an irreversible protein synthesis inhibitor. For this aim, A20-HL B lymphoma cells expressing surface IgM receptors specific for TNP were used as APC. OVA and TNP-OVA were used as nonspecific and specific Ag, respectively. The treatment with emetin greatly impaired the ability of A20-HL cells to present specific Ag, but not nonspecific Ag, to 42-6A cloned T cells specific for OVA. The ability of the emetin-treated A20-HL cells to present nonspecific Ag indicates that the treated cells are able to process nonspecific Ag and to present processed Ag. Ag binding and the internalization by A20-HL cells through surface receptors were not affected by the emetin treatment. A20-HL cells took up specific Ag for stimulation of 42-6A cells in the presence of cycloheximide, a reversible protein synthesis inhibitor. These results suggest that the action of emetin is localized to the intracellular processing of specific Ag, not of nonspecific Ag. Thus, the processing pathway for specific Ag seems to be different from that for nonspecific Ag.     

Isolation and Characterization of Subpopulations of Rat Ascites Hepatoma Cell Line of AH109A with Different Metastatic Potentials

Rat ascites hepatoma cell line of AH109A proved to be divided into two subpopulations with different invasive and metastatic potentials, when cultured in the medium containing allogeneic rat sera. One population adheres to the culture dish, actively extending pseudopodia, and the other remains in a floating state. Utilizing this character, we have separated these two populations. After three successive separation steps, adhesive AH109A cells and floating AH109A cells were obtained. Adhesive AH109A cells proliferated more rapidly and invaded more actively than did floating AH109A cells. Adhesive AH109A cells metastasized mainly to lung, while floating AH109A cells to mesentery, when intravenously injected into tail veins. Histological studies revealed that adhesive AH109A cells showed lymphatic metastases to lung. These results suggest that the two populations separated from parental AH109A cells provide good models for the study of tumor invasion and tissue-specific metastasis and that adhesive AH109A cells can be used for the creation of lymphatic metastasis model of rats.     

Requirements for eIF4A and eIF2 during translation of Sindbis virus subgenomic mRNA in vertebrate and invertebrate host cells

We have examined the requirements for the initiation factors (eIFs) eIF4A and eIF2 to translate Sindbis virus (SV) subgenomic mRNA (sgmRNA) in the natural hosts of SV: vertebrate and arthropod cells. Notably, this viral mRNA does not utilize eIF4A in SV‐infected mammalian cells. However, eIF4A is required to translate this mRNA in transfected cells. Therefore, SV sgmRNA exhibits a dual mechanism for translation with respect to the use of eIF4A. Interestingly, SV genomic mRNA requires eIF4A for translation during the early phase of infection. In sharp contrast to what is observed in mammalian cells, active eIF2 is necessary to translate SV sgmRNA in mosquito cells. However, eIF4A is not necessary for SV sgmRNA translation in this cell line. In the SV sgmRNA coding region, proximal to the initiation codon is a hairpin structure that confers eIF2 independence only in mammalian cells infected by SV. Strikingly, this structure does not provide independence for eIF4A neither in mammalian nor in mosquito cells. These findings provide the first evidence of different eIF requirements for translation of SV sgmRNA in vertebrate and invertebrate cells. These observations can help to understand the interaction of SV with its host cells.     

Induction of Apoptosis and Tyrosine Phosphorylation of Cellular Proteins in T Cells and Non-T Cells by Stimulation with Concanavalin A

A high concentration (30 μg/ml or more) of Con A caused the death of not only thymocytes but also splenic cells of BALB/c mice, whereas a moderate concentration (3 μg/ml) of Con A induced proliferation of these cells. A high concentration of Con A also induced the death of splenic cells of athymic BALB/c-nu/nu mice and the bone marrow cells of BALB/c mice which mainly consist of non-T cells. However, any concentration (1-30 μg/ml) of Con A failed to induce the proliferation of these cells. Specific binding of tetrameric Con A to mannose-containing receptors was required for the induction of cell death. DNA fragmentation was observed by both laser flow cytometry and electrophoresis in Con A-stimulated T cells and non-T cells. This indicated that the mechanism of induction of apoptosis with Con A is not necessarily TCR-dependent. Con A induced tyrosine phosphorylation of a number of proteins in various types of cells. Interestingly, phosphorylation of the 40 kDa protein developed only in the thymocytes and spleen cells that contain T cells, whereas phosphorylation of the 80 and 120 kDa proteins appeared in both T cells and non-T cells. These results suggested that the Con A-induced apoptosis of T cells and non-T cells involves different but possibly mutually related protein tyrosine phosphorylation-linked signals.     

Differential TLR recognition of leptospiral lipid A and lipopolysaccharide in murine and human cells

Leptospira interrogans is a spirochete that is responsible for leptospirosis, a zoonotic disease. This bacterium possesses an unusual LPS that has been shown to use TLR2 instead of TLR4 for signaling in human cells. The structure of its lipid A was recently deciphered. Although its overall hexa-acylated disaccharide backbone is a classical feature of all lipid A forms, the lipid A of L. interrogans is peculiar. In this article, the functional characterization of this lipid A was studied in comparison to whole parental leptospiral LPS in terms of cell activation and use of TLR in murine and human cells. Lipid A from L. interrogans did not coagulate the Limulus hemolymph. Although leptospiral lipid A activated strongly murine RAW cells, it did not activate human monocytic cells. Results obtained from stimulation of peritoneal-elicited macrophages from genetically deficient mice for TLR2 or TLR4 clearly showed that lipid A stimulated the cells through TLR4 recognition, whereas highly purified leptospiral LPS utilized TLR2 as well as TLR4. In vitro experiments with transfected human HEK293 cells confirmed that activation by lipid A occurred only through murine TLR4-MD2 but not through human TLR4-MD2, nor murine or human TLR2. Similar studies with parental leptospiral LPS showed that TLR2/TLR1 were the predominant receptors in human cells, whereas TLR2 but also TLR4 contributed to activation in murine cells. Altogether these results highlight important differences between human and mouse specificity in terms of TLR4-MD2 recognition that may have important consequences for leptospiral LPS sensing and subsequent susceptibility to leptospirosis.     

Natural selection of Varroa jacobsoni explains the different reproductive strategies in colonies of Apis cerana and Apis mellifera

In colonies of European Apis mellifera, Varroa jacobsoni reproduces both in drone and in worker cells. In colonies of its original Asian host, Apis cerana, the mites invade both drone and worker brood cells, but reproduce only in drone cells. Absence of reproduction in worker cells is probably crucial for the tolerance of A. cerana towards V. jacobsoni because it implies that the mite population can only grow during periods in which drones are reared. To test if non-reproduction of V. jacobsoni in worker brood cells of A. cerana is due to a trait of the mites or of the honey-bee species, mites from bees in A. mellifera colonies were artificially introduced into A. cerana worker brood cells and vice versa. Approximately 80% of the mites from A. mellifera colonies reproduced in naturally infested worker cells as well as when introduced into worker cells of A. mellifera and A. cerana. Conversely, only 10% of the mites from A. cerana colonies reproduced, both in naturally infested worker cells of A. cerana and when introduced into worker cells of A. mellifera. Hence, absence of reproduction in worker cells is due to a trait of the mites. Additional experiments showed that A. cerana bees removed 84% of the worker brood that was artificially infested with mites from A. mellifera colonies. Brood removal started 2 days after artificial infestation, which suggests that the bees responded to behaviour of the mites. Since removal behaviour of the bees will have a large impact on fitness of the mites, it probably plays an important role in selection for differential reproductive strategies. Our findings have large implications for selection programmes to breed less-susceptible bee strains. If differences in non-reproduction are mite specific, we should not only look for non-reproduction as such, but for colonies in which non-reproduction in worker cells is selected. Hence, in selection programmes fitness of mites that reproduce in both drone and worker cells should be compared to fitness of mites that reproduce only in drone cells. © Rapid Science Ltd. 1998     

Quantitative Binding of 125I-Concanavalin A to Normal and Transformed Cells

We have measured the quantitative binding of the radioactively labeled agglutinin (125)I-concanavalin A to normal mammalian cells and simian virus 40- and polyoma virus-transformed cells from tissue culture. Parallel measurements of the amount of (125)I-concanavalin A necessary to cause agglutination of the cells in suspension were carried out. The transformed and nontransformed cells used for these experiments show large differences in their ability to be agglutinated by (125)I-concanavalin A. However, these cell lines have the same number of specific binding sites and similar affinities for the agglutinin whether transformed, trypsinized, or nontransformed. We conclude that the differential capacity of concanavalin A to agglutinate transformed cells relative to normal cells does not result from differences in the number of binding sites between the two types of cells.     

Concanavalin a binding and cytodifferentiation in pancreatic acinar carcinoma of rat

The binding of concanavalin A to the plasmalemma of acinar carcinoma cells was characterized by electron microscopy utilizing horseradish peroxidase. Heavy labeling due to specific concanavalin A binding was detected on the plasmalemma of undifferentiated carcinoma cells lacking zymogen maturation, neoplastic cells of intermediate differentiation with only occasional zymogen granules, and highly differentiated acinar carcinoma cells containing numerous cytoplasmic zymogen granules. The plasmalemma of acinar carcinoma cells was also compared to the normal pancreatic acinar cell plasmalemma by measurement of specific ¹²⁵I-labeled concanavalin A binding. Although only about one-third of pancreatic acinar carcinoma cells demonstrate mature zymogen differentiation, the acinar carcinoma had a full complement of normal plasmalemma receptors for ¹²⁵I-labeled concanavalin A. It is concluded that, unlike normal pancreas, the presence of concanavalin A receptors on the plasmalemma of acinar carcinoma cells is not a specific membrane marker for differentiated cells containing zymogen granules.     

Membrane ATPases and acid tolerance of Actinomyces viscosus and Lactobacillus casei

Lactobacillus casei ATCC 4646 and Actinomyces viscosus OMZ105E were found to differ markedly in acid tolerance. For example, pH profiles for glycolysis of intact cells in dense suspensions indicated that glycolysis by L. casei had an optimal pH of about 6.0 and that glucose degradation was reduced by 50% at a pH of 4.2. Comparable values for A. viscosus cells were at pHs of about 7.0 and 5.6. The difference in acid tolerance appeared to depend mainly on membrane physiology, and the addition of 40 microM gramicidin to cell suspensions increased the sensitivity of the glycolytic system by as much as 1.5 pH units for L. casei and up to 0.5 pH unit for A. viscosus. L. casei cells were inherently somewhat more resistant to severe acid damage than were A. viscosus cells, in that Mg release from L. casei cells in medium with a pH of 3.0 occurred only after a lag of some 4 h, compared with rapid release from A. viscosus cells. However, the major differences pertinent to the physiology of the organisms appeared to be related to proton-translocating ATPases. Isolated membranes of L. casei had about 3.29 U of ATPase per mg of protein, compared with only about 0.06 U per mg of protein for those of A. viscosus. Moreover, the ATPase of L. casei had a pH optimum for hydrolytic activity of about 5, compared with an optimal pH of about 7 for that of A. viscosus.(ABSTRACT TRUNCATED AT 250 WORDS)     

Thermotolerance of pulp cells and phagocytosis of apoptotic pulp cells by surviving pulp cells following heat stress

Apoptosis is known to be associated with wound healing and regeneration of dental pulp. We examined the effects of heat stress on clonal dental pulp cell line (RPC-C2A cells) to clarify the pulp wound healing process. RPC-C2A cells were exposed to heat stress at 43 degrees C for 45 min. After several time intervals, the inhibition of cell proliferation and apoptosis induction were analyzed by cell viability assay, DNA gel electrophoresis, nuclear staining, and terminal deoxynucleotidyl transferase mediated labeling assay. RPC-C2A cells showed the thermotolerance following heat stress. We found that apoptosis was induced in some RPC-C2A cells, whereas others remained alive, and observed the engulfment of apoptotic cells by scavenger-like RPC-C2A cells following heat stress. We also analyzed the phagocytotic activity of RPC-C2A cells and found that they had an ability to engulf apoptotic RPC-C2A cells, which was stimulated by heat stress. These results suggest that heat stress induces apoptosis of RPC-C2A cells, which are phagocytosed by the surviving RPC-C2A cells.     

Differential Expression of Syntaxin 1A and 1B by Noradrenergic and Adrenergic Chromaffin Cells

The expression and localization of syntaxin isoforms 1A and 1B in adrenergic and noradrenergic chromaffin cells were examined by both immunoblot analysis and confocal immunofluorescence microscopy. Syntaxin 1A was found in higher levels in noradrenergic cells, whereas syntaxin 1B was similarly expressed in most noradrenergic and adrenergic cells. However, some heterogeneity was observed within each catecholaminergic phenotype. Although the majority of adrenergic cells appeared to express low levels of syntaxin 1A, about 7% was strongly stained for syntaxin 1A. A subpopulation of noradrenergic cells, about 17%, expressed greater levels of syntaxin 1B. Syntaxin 1B labeling showed a punctate appearance in the cytoplasm, whereas syntaxin 1A appeared predominantly localized to the plasma membrane. These data show differences in the exocytotic machinery of the two subtypes of chromaffin cells that may underlie some of the distinct characteristics of adrenaline and noradrenaline secretion.     

A rapid and sensitive flow cytometric method for the detection of multidrug-resistant cells

Multidrug-resistant (MDR) cells are characterized by a defect in drug accumulation caused by activity of an energy-dependent rapid drug efflux pump. The action of this drug pump can be inhibited by specific agents, referred to as membrane transport modulating agents (MTMAs), resulting in a restoration of the intracellular drug accumulation. This paper presents a flow cytometric assay for the detection of MDR cells, which is based on the ability of these cells to respond to MTMAs. Daunorubicin net-uptake kinetics were measured of anthracycline-sensitive (A2780/S) and -resistant (A2780/R) human ovarian carcinoma cells in vitro. A2780/R cells accumulated significantly less (about a factor of 5) daunorubicin as compared to A2780/S cells. Addition of verapamil or cyclosporin A to A2780/R cells at steady-state daunorubicin uptake led to a dose-dependent increase in cellular daunorubicin accumulation. The sensitivity of the assay was determined by testing mixtures of A2780/S and A2780/R cells. Analysis of A2780/S cells contaminated with A2780/R cells showed that as few as 2.5% MDR cells could readily be detected in the mixture. In conclusion, this functional assay enables the detection of MDR cells in a heterogeneous cell suspension and is ideally suited for the study of the occurrence of typical MDR in human cancer.     

The decrease of mitochondrial NADH dehydrogenease and drug induced apoptosis in doxorubicin resistant A431 cells

Doxorubicin (DOX) resistant A10A cells derived from human squamous carcinoma A431 cells were found to exhibit a smaller degree of apoptosis after DOX treatment as compared to their parent cells. Induction of reactive oxygen species (ROS) formation and mitochondrial depolarization by DOX were more pronounced in the parent cells than in the A10A cells. The fact that catalase suppressed the DOX effect on ROS induction, mitochondrial depolarization and apoptosis in both cell lines suggests an involvement of ROS in the DOX-induced apoptosis. To investigate the underlying mechanisms for DOX resistance in A10A cells, RT-PCR based differential display was used. One of the clones, which was down-regulated in the A10A cells, had sequence homology with part of the mitochondrial NADH dehydrogenase III (ND3) gene. NADH dehydrogenase plays an important role in generating ROS during DOX treatment. The results indicate that down-regulation of ND3 may at least in part contribute to the mechanism for A10A cells resistant to DOX-induced apoptosis.     

Effect of T- and B-lymphocyte mitogens on interactions between lymphocytes and macrophages.

The mitogenic response of murine T cells ² to Con A, S-Con A and PHA was found to be macrophage-dependent. Optimal mitogenic responses were obtained when macrophage-depleted T-cell populations were reconstituted with 5% normal peritoneal macro-phages. Studies were carried out to investigate the effect of T- and B-cell mitogens on in vitro physical interactions between murine lymphocytes and macrophages. This was done by determining the number of T- or B cells binding to macrophages in the absence and in the presence of T- and B cell mitogens, and comparing the results of these experiments with the induction of lymphocyte proliferation. Con A increased the binding of T cells to macrophages when used in mitogenic doses (1–5 μg/ml). Dose response experiments showed that the same dose of Con A which produced maximal mitogenic stimulation also induced the greatest number of T cells to bind to macrophages. Nonmitogenic doses of Con A (20–50 μg/ml) did not enhance the binding of T cells, while identical doses of S-Con A both induced T cell mitogenesis and increased the number of T cells bound to macrophages. Similar results were obtained with PHA. None of the B-cell mitogens tested (LPS, EPO 127 and LAgl) increased the binding of either T or B cells to macrophages. PWM, which is mitogenic for both T and B cells, increased the binding of T cells to macrophages, but not that of B cells. In brief, the four T-cell mitogens tested (Con A, S-Con A, PHA, and PWM) induced specific physical interactions between T cells and macrophages, while none of the B-cell mitogens had any effect on the physical interactions between either B or T cells and macrophages when used in mitogenic doses.     

Biochemical and molecular properties of cisplatin-resistant A2780 cells grown in folinic acid

In this study, A2780 human ovarian carcinoma cells were grown in folinic acid in contrast to folic acid, and the molecular and biochemical properties of cisplatin-resistant A2780 cells were analyzed for changes in the dTMP synthase cycle. At concentrations of folinic acid that were optimal for cell growth (10(-8) M), the ED50 for cisplatin was 2.5 and 43 microM in the A2780S and A2780DDP cells, respectively. Resistance to cisplatin was associated with a 2-fold cross-resistance to 5-fluorodeoxyuridine and 5-fluorouracil as well as a 3-fold increase in both dTMP synthase activity and mRNA. The ED50 for methotrexate was similar in both A2780S and A2780DDP cells (1.2 microM). When both the A2780S and A2780DDP cells were grown in folinic acid, there was no significant difference in the level of dihydrofolate reductase activity. This data would suggest that cisplatin resistance is associated with changes in folate metabolism.     

Membrane ATPases and acid tolerance of Actinomyces viscosus and Lactobacillus casei.

Lactobacillus casei ATCC 4646 and Actinomyces viscosus OMZ105E were found to differ markedly in acid tolerance. For example, pH profiles for glycolysis of intact cells in dense suspensions indicated that glycolysis by L. casei had an optimal pH of about 6.0 and that glucose degradation was reduced by 50% at a pH of 4.2. Comparable values for A. viscosus cells were at pHs of about 7.0 and 5.6. The difference in acid tolerance appeared to depend mainly on membrane physiology, and the addition of 40 microM gramicidin to cell suspensions increased the sensitivity of the glycolytic system by as much as 1.5 pH units for L. casei and up to 0.5 pH unit for A. viscosus. L. casei cells were inherently somewhat more resistant to severe acid damage than were A. viscosus cells, in that Mg release from L. casei cells in medium with a pH of 3.0 occurred only after a lag of some 4 h, compared with rapid release from A. viscosus cells. However, the major differences pertinent to the physiology of the organisms appeared to be related to proton-translocating ATPases. Isolated membranes of L. casei had about 3.29 U of ATPase per mg of protein, compared with only about 0.06 U per mg of protein for those of A. viscosus. Moreover, the ATPase of L. casei had a pH optimum for hydrolytic activity of about 5, compared with an optimal pH of about 7 for that of A. viscosus.(ABSTRACT TRUNCATED AT 250 WORDS)     

Agglunation of a transformed mouse cell line and a variant subline with concanavalin A: Effect of temperature and time of lectin incubation

Colchicine treatment enhanced Con A-mediated agglutination of erythrocytes to LM cells (LM is a “spontaneously” transformed mouse line) incubated for brief periods with Con A at 22° C. Longer incubations with Con A at 22° C rendered colchicine treated cells less agglutinable than untreated cells. Even short incubation times with Con A at higher temperature (37° C) rendered colchicine treated LM cells less agglutinable than their untreated counterparts. Below 15° C, colchicine treated cells remained more agglutinable than untreated cells even after long periods of Con A treatment. Cells of a variant clone (Rl) isolated from LM by negative selection with concanavalin A exhibited increased substratum adhesiveness and an absolute serum requirement. LM and variant cells exhibited a differential reponse to colchicine treatment, the variant subline reguiring longer periods of colchicine treatment to elicit changes in morphology and agglutinability.     

Autocrine and paracrine regulation of tissue inhibitor of metalloproteinases, transin, and urokinase gene expression in metastatic and nonmetastatic mammary carcinoma cells.

Acquisition of metastatic competence by tumor cells is frequently accompanied by increased expression of extracellular proteases capable of degrading basement membrane and extracellular matrix. However, very little is known about how the genes encoding these enzymes and their inhibitor proteins are regulated in metastatic versus nonmetastatic cells. In this report, we have compared autocrine and paracrine regulation of tissue inhibitor of metalloproteinases (TIMP), transin, and urokinase plasminogen activator (uPA) genes in genetically related nonmetastatic SP1 and metastatic A3a cell lines. Compared to SP1 cells, metastatic A3a cells showed 15-20-fold higher transin, 3-5-fold less TIMP mRNA, and comparable levels of uPA mRNA. A qualitatively similar shift in expression of these genes was rapidly (i.e., 4-8 h) induced in nonmetastatic SP1 cells following the addition of conditioned medium from A3a cells. The gene-regulating activity present in A3a conditioned medium was heat-labile, suggesting that it was protein in nature. The responsiveness of SP1 cells to the factor(s) secreted by A3a conditioned medium was inhibited by cycloheximide. Basic fibroblast growth factor mimicked the effect of the A3a conditioned medium as an inducer of transin expression in the tumor cells. Although medium conditioned by the tumor cells did not affect uPA expression, addition of epidermal growth factor to the tumor cells transiently induced expression of uPA with a biphasic response that differed in SP1 and A3a cells. Initial induction of uPA at 2-4 h was similar for both cell lines, but after 24 h of exposure to epidermal growth factor, SP1 cells showed a net reduction in uPA, whereas metastatic cells returned to the unstimulated levels.(ABSTRACT TRUNCATED AT 250 WORDS)     

鳞癌细胞增殖能力与岩藻糖基转移酶表达水平及糖蛋白岩藻糖基化修饰有关

岩藻糖基转移酶（fucosyltransferases,FUTs）是一类催化糖蛋白和糖脂发生岩藻糖基化（修饰）酶,主要包括FUT1~FUT9。已有研究证明,很多癌组织中都有不同FUT基因表达升高的现象。本研究证明,表皮鳞癌细胞的增殖能力与几种FUT基因表达水平有关。本文比较研究了人表皮鳞癌A431和SCC12细胞的增殖速度和几种FUT的表达状况,以揭示鳞癌细胞增殖能力与几种FUT基因表达水平的关系。细胞倍增时间结合MTT法揭示,鳞癌A431细胞的倍增时间约为26 h,而鳞癌SCC12细胞的倍增时间约为33 h（P < 0.05）,提示A431细胞增殖速度比SCC12细胞明显加快。与增殖速度一致的是,Western 印迹显示,A431细胞中与DNA合成相关的增殖细胞核抗原（proliferating cell nuclear antigen,PCNA）蛋白表达水平比SCC12细胞高。实时定量PCR（qPCR）检测FUT1-9基因 mRNA转录本,揭示A431细胞中几种FUT基因的mRNA水平均显著高于SCC12细胞。凝集素免疫印迹法和Western 印迹法进一步证明,A431细胞中总蛋白的岩藻糖基化水平比SCC12细胞中的明显升高。敲低FUT4基因表达后,A431细胞中LeY寡糖的表达水平下调,细胞增殖被明显抑制。这些结果证明,较强的表皮鳞癌细胞增殖能力可能与几种FUT基因的高表达,以及糖蛋白的岩藻糖基化（修饰）相关。岩藻糖基转移酶表达水平与临床表皮鳞癌的恶性增生的相关性有待进一步证明。