Glycotargeting: influence of the sugar moiety on both the uptake and the intracellular trafficking of nucleic acid carried by glycosylated polymers

Nucleic acids (plasmids as well as oligonucleotides) used to specifically express or modulate the expression of a gene, must reach the cytosol and/or the nucleus. Several systems have been developed to increase their uptake and their efficiency. Glycosylated polylysines have been shown to specifically help nucleic acids to be taken up in cells expressing a given cell surface membrane lectin. However, it appeared that the efficiency of the imported nucleic acid was not directly related to the extent of the uptake. Indeed, some glycosylated polylysines bearing sugar moities which are poor ligands of the cell surface lectins of a given cell were found to be more efficient than those bearing better sugar ligands. The interpretation of this paradoxal result is discussed with regards to the nature of the compartment allowing the nucleic acid to cross the membrane and to be delivered in the cytosol on the one hand, and to the presence of intracellular lectins on the other hand.

Optimized conditions to couple two water-soluble biomolecules through alkylamine thiolation and thioetherification

A simple method for introducing, in buffered saline, a reactive sulfhydryl group on water-soluble molecules bearing an alkyl-amino group is described. This method is based on the use of two water-soluble reagents: 2-iminothiolane and 6,6'-dithiodinicotinic acid. The first one is open upon reaction with an amino group, and the generated thiol group is immediately protected by action of the second reagent. The optimal conditions were determined by taking into account the stability and the reactivity of both reagents with regards to pH and temperature. This method was validated through two applications, the substitution of bovine serum albumin with a bromoacetyl peptide and the substitution of an amino link at the 5' end of an oligonucleotide by reaction with either a fluorescent tag, iodoacetamidofluorescein, or a bromoacetyl peptide, upon reduction of the protected disulfide bridge with a third water-soluble reagent, namely tris(2-carboxyethyl)phosphine.

Glycofection in the presence of anionic fusogenic peptides: a study of the parameters affecting the peptide-mediated enhancement of the transfection efficiency

Gene delivery mediated by polyplexes such as DNA complexed with polylysine conjugates is limited by the low efficiency of escape of DNA from the endosomes. One of the strategies which favors the transmembrane passage of polyplexes consists of adding anionic amphipathic peptides capable of destabilizing membranes in an acidic medium. Although less efficient than replication-defective adenoviruses, fusogenic peptides increase the expression of the reporter gene by a factor between 100 and 1000 depending on the cell line. However, the activity of a given peptide depends on the composition of the lipid bilayer. We were interested in developing a polyplex (glycoplex) formulation comprising a glycosylated polylysine, a fusogenic peptide and a plasmid which would be useful for efficient transfection (glycofection) of a large panel of cells, even in the presence of serum. We synthesized several peptides and tested their efficiency in combination with different glycoplex formulations. We found that glycofection with a quaternary complex (called one pot formulation) made of lactosylated-polylysine, polylysine, DNA, and the dimeric peptide (E5-WYGG)2-KA was less cell-type dependent than other peptide-based formulations. In addition, its efficiency was not affected by the presence of serum (up to 20%).

Sugar-mediated uptake of glycosylated polylysines and gene transfer into normal and cystic fibrosis airway epithelial cells

We have examined the membrane lectin expressed by immortalized normal and cystic fibrosis (CF) airway epithelial cells, using fluorescein-labeled neoglycoproteins; the uptake of plasmid DNA using fluoresceinylated glycoplexes (plasmid/glycosylated polylysine complexes); and the efficiency of gene transfer when glycosylated polylysines and glycosylated, partially gluconoylated polylysines were used as vectors. The most efficient uptake of neoglycoproteins by normal and CF cells was obtained with mannosylated BSA (bovine serum albumin). Similarly, the most efficient uptake of plasmid DNA was obtained with glycoplexes bearing alpha-D-Man residues. Surprisingly, glycoplexes bearing alpha-D-Man residues were poorly efficient for gene transfer into normal and CF cells. The highest luciferase activity was achieved with lactosylated polylysine- and beta-D-GlcNAc-substituted gluconoylated polylysine as vectors. Gene transfer efficiency obtained with gluconoylated polylysine bearing beta-D-GlcNAc residues was similar to that observed with polyethylenimine (PEI; 25 and 800 kDa) and 10-fold higher than that observed with lipofectin and LipofectAMINE. These results suggest that the transfection efficiency with glycoplexes is not determined only by the specificity of the lectin expressed at the cell surface membrane but also by intracellular trafficking of the glycoplexes, which could be mediated by lectins present inside the cells.

Intracellular localization of oligonucleotides: influence of fixative protocols

In many studies reporting the use of antisense oligonucleotides (ODN), the intracellular localization was investigated by using fluorescent-labeled oligonucleotides (F-ODN). More often, cells were fixed on uptake of F-ODN before microscopic analysis. We report here the influence of various methods of cell fixation on the intracellular localization of ODN. By confocal microscopy, we show that with unfixed cells, endocytosed peptides, oligonucleotides (Mr around 10,000), and endocytosed proteins were mainly localized in vesicular compartments. On mild fixation with paraformaldehyde, an identical intracellular localization was observed repeatedly after fixation, from immediately up to several days. In contrast, with methods based on the use of strong fixatives, such as methanol or acetone, the small molecules diffuse into the cytosol and in the case of oligonucleotides into the nucleus. These results point out the importance of the fixation protocol in the study of intracellular localization of ODN and their derivatives.

A Plasmodium falciparum aminopeptidase gene belonging to the M1 family of zinc-metallopeptidases is expressed in erythrocytic stages

A new single copy gene has been isolated from Plasmodium falciparum, by immunoscreening a genomic DNA expression library. The gene appears devoid of introns, displays the classical A + T richness and codon usage of P. falciparum genes, and is transcribed into a 4 kb mRNA in erythrocytic stages. The deduced amino acid sequence corresponds to a 1056 residue protein (122 kDa) containing the canonical HExxHx18E signature of zinc-metallopeptidase active sites of the M1 family at position 467-490, a downstream conserved tyrosine residue involved in catalysis in position 551, and the GAMEN conserved motif characteristic of aminopeptidases in the M1 family, at position 431-435. The greatest similarities were found with aminopeptidases N of Escherichia coli and Haemophilius influenza (more than 80% identical residues in the canonical signature of the active site) but significant similarities centred on the active site region exist with all other members of the M1 family such as other prokaryotic aminopeptidases, eukaryotic aminopeptidases A and N and leukotriene A4 hydrolases (40-50% identical residues in the canonical signature of the active site). A polyclonal serum raised to a synthetic peptide deduced from the gene labelled schizont proteins of 96 and 68 kDa purified to homogeneity and both displaying aminopeptidase activity, as well as cytoplasmic structures in schizont stages.

Cellular vaccines

This project is devoted to the development of novel cellular vaccines designed to treat cancer patients. These cellular vaccines present and enhance immunogens, which will elicit a potent immune response. The goal is to achieve safe and effective immune reaction against the patient's own tumour. (1) Autologous cellular vaccines are prepared by processing circulating blood mononuclear cells outside of the patient's body (ex vivo) to differentiate them into antigen-presenting cells (APCs). Monocyte-derived APCs (MD-APCs) are then grown in the presence of exogenous target antigens (tumour cell debris, or apoptotic bodies) to become fully mature APCs. (2) Functionality for antigen presentation to T cells of ex vivo MD-APCs is evaluated in vivo. (3) Cellular vaccines are tested in selected rodent animal models. Efficiency and immune response are monitored in pertinent experimental systems for cancer. Pharmacological data are generated for clinical investigation. Tolerance and biologic effects are documented in primates. (4) The first clinical trials on cancer patients are taking place in 1998 on melanoma and prostate cancer to validate the concept. Specialized cell processors with dedicated software and standardized controls are being developed and used for the preparation of cellular vaccines. (5) The evaluation of new non-viral vectors and the validation of new non-viral transfection methods of mononuclear cells with marker genes is in progress and will lead to the ex vivo transfection of genes coding for immunostimulating cytokines or for tumour antigens in MD-APCs. Efficiency will be validated in vitro and in animal models. The ex vivo and animal model studies validate the clinical relevance of this new cellular immunotechnology. Clinical validation of individual autologous cellular vaccines in specific indications for which no treatment is presently available will allow the development of cellular and gene immunotherapy for other types of cancers.

Galectin-3 expression in human atherosclerotic lesions

The expression of galectin-3, a beta-galactoside-binding lectin, was studied in atherosclerotic lesions from specimens obtained from carotid endarterectomies, lower limb amputations, and thoracic aortas from autopsies of young adult trauma victims. Immunohistochemical staining with the monoclonal antibody M3/38 demonstrated the presence of galectin-3 in advanced atherosclerotic lesions from each of 13 cases of carotid endarterectomy and 16 lower limb amputations and in the thoracic aorta of 4 of 20 cases of trauma victim adults. Immunostaining did not detect galectin-3 in umbilical cord and normal thoracic aorta arteries and limb veins. Dual immunostaining with monoclonal antibodies M3/38 for galectin-3 and clone 1A4 for smooth muscle alpha-actin or HAM56 for human macrophage antigen showed that galectin-3 was localized predominantly in foam cells and macrophages and rarely (<5%) in the smooth muscle cells of atherosclerotic lesions. The incidence of galectin-3-positive cells was higher in the carotid artery atherosclerotic lesions, which are richer in foam cells, than in the lower limb atherosclerotic lesions, which are more fibrotic. Reverse transcription polymerase chain reaction showed a significantly higher ratio of galectin-3/beta-actin transcripts in 20 atherosclerotic arteries compared with that of 5 umbilical cord arteries. Western blot analysis confirmed a higher level of galectin-3 in atherosclerotic carotid and lower limb arteries compared with that of umbilical cord arteries. The increased expression of galectin-3 in atherosclerotic lesions suggests the involvement of this multifunctional protein in atherogenesis.

Novel glycosynthons for glycoconjugate preparation: oligosaccharylpyroglutamylanilide derivatives

The reducing sugar of an oligosaccharide reacting with the alpha-amino group of an amino acid is converted to an N-oligosaccharylamino acid which can then be stabilized by N-acylation. Oligosaccharides in solution in N,N-dimethylformamide reacted with alpha-glutamyl-p-nitroanilide at 50 degrees C for a few hours, leading to an N-oligosaccharylglutamyl-p-nitroanilide. Then, the gamma-carboxylic group of the glutamyl moiety, activated by adding (benzotriazol-1-yloxy)tris(dimethylamino)phosphonium hexafluorophosphate (BOP), reacted with the substituted alpha-amino group of the glutamyl residue, leading to an N-oligosaccharylpyroglutamyl-p-nitroanilide within 0.5 h. Such a one-pot two-step reaction was shown to be very efficient in the case of a disaccharide such as lactose, or pentasaccharides such as lacto-N-fucopentaoses, Lewis(a) or Lewis(x). The glycosynthons were characterized by chromatography (HPAEC and HPLC); their molecular mass was determined by electrospray ionization mass spectrometry, and the glycosylamides were shown to have a beta-anomeric configuration on the basis of their proton NMR. The N-oligosaccharylpyroglutamyl-p-nitroanilides are quite stable at room temperature over a large pH range. They are easily converted to N-oligosaccharylpyroglutamyl-p-isothiocyanatoanilides which can be used to prepare glycoconjugates such as cationic glycosylated polylysines suitable for specifically delivering genes or oligonucleotides in a sugar-dependent manner.

Membrane permeabilization and efficient gene transfer by a peptide containing several histidines

We designed a peptide, H5WYG (GLFHAIAHFIHGGWHGLIHGWYG), that permeabilizes cell membrane at a slightly acidic pH but not at neutral pH. Absorbance, fluorescence, and circular dichroism spectra showed that H5WYG undergoes a dramatic conformational change between pH 7.0 and 6.0 that correlates with the protonation of the histidyl residues. Cell permeabilization studies monitored by flow cytometry on living cells showed that H5WYG permeabilizes the cell membrane with a great efficiency at pH 6.4 but was not active at neutral pH; at pH 6.8, the peptide permeabilized 50% of the cells at 20 degrees C within 10 min. H5WYG increased the expression of genes transferred to cells as glycosylated polylysine-DNA complexes, and the transfection efficiency was not impaired in the presence of serum. Therefore, this peptide containing several histidines that become positively charged when the pH decreased to less than 7.0 is a suitable helper for delivering molecules into the cytosol upon either permeabilization of the plasma membrane induced by lowering the extracellular medium to pH 6.4 or permeabilization of the endosomal membrane induced by acidification of endosomes.

Glycotargeting: the preparation of glyco-amino acids and derivatives from unprotected reducing sugars

Lectins are present on the surface of many cells. Many lectins actively recycle from membrane to endosomes and efficiently take up glycoconjugates in a sugar-dependent manner. On this basis, glycoconjugates, specially those obtained by chemical means, are good candidates as carriers of drugs, oligonucleotides or genes. In this paper, we present a panel of methods suitable to transform unprotected reducing oligosaccharides into glycosynthons designed to be easily linked to therapeutic agents. All the glycosynthons presented here are glycosylamines or derivatives, mainly glyco-amino acids or glycopeptides. Glycosylamines are easy to obtain, but they are very labile in slightly acidic or neutral medium; they must be stabilized, by acylation for instance. The coupling efficiency of a reducing sugar with ammonia as well as an alkylamine or an arylamine is higher at high temperature, however, because of the Amadori rearrangement, special conditions have to be selected to prepare the expected glycosylamine derivative with a high yield. Glycosylamines are easily acylated by N-protected amino acids, or by halogeno acids which can then be transformed into amino acids. Alternatively, unprotected reducing oligosaccharides may very efficiently be transformed into N-glycosyl-amino acids and then protected by N-acylation. With a glutamyl derivative having both the alpha-amino and the gamma-carboxylic groups free, the coupling and the acylation, which is intramolecular, are roughly quantitative. N-oligosaccharyl-amino acid derivatives are interesting glycosynthons, because their sugar moiety bears the specificity towards membrane lectins while the amino acid part has the capacity to easily substitute a therapeutic agent.

Purification of two lectins from a nopalin Agrobacterium tumefaciens strain

Lectins were evidenced on the surface of one Agrobacterium tumefaciens wild strain (82,139) by agglutination test and neoglycoprotein labelling. Bacteria were incubated in the presence of various fluorescein-labelled neoglycoproteins and the binding was assessed by a fluorimetric method. Among the fluorescein-labelled neoglycoproteins tested, the one bearing alpha-D-galactosyl residues was the most efficient. The labelling was optimal at pH 5.0 and naught at pH above 7. The binding was specifically inhibited by homologous fluorescein-free neoglycoproteins. A galactoside-specific lectin was purified to homogeneity by affinity chromatography on agarose-A4 substituted with alpha-D-galactopyranosyl residues. Upon polyacrylamide gel electrophoresis, a single band (M(r) 58,000) was detected. This alpha-D-galactoside-specific lectin agglutinated preferentially human B red blood cells at pH 5.0. Another lectin specific for alpha-L-rhamnoside (M(r) 40,000) not retained on the immobilised galactose was purified by affinity chromatography on alpha-L-rhamnosyl substituted agarose-A4. This L-rhamnoside-specific lectin preferentially agglutinated horse erythrocytes. On the basis of their M(r) and on their sugar specificity, these two lectins are novel lectins with regard to the known sugar-binding proteins present in the Rhizobiaceae family: Agrobacterium, Rhizobium or Bradyrhizobium strains.

Modulation of the expression of the rabbit galectin-3 gene by p53 and c-Ha-ras proteins and PMA

Galectin-3 is a galactose-binding lectin that has been found in several mammalian tissues. Galectin-3 gene is expressed in a wide range of normal and tumoral cells. In the case of myeloid cells, its expression correlates with the differentiation of monocytes to macrophages. In the case of cancer cell lines, its expression correlates with tumorigenicity and metastatic potential. The regulation of the expression of this gene is still largely unknown. The rabbit galectin-3 gene has been isolated and characterized. Its structure revealed an organization similar to that of the murine galectin-3 gene. The genomic sequences located upstream from its 5' end, upon insertion upstream from a promoter-free reporter gene, exhibited a strong promoter activity. This activity was upregulated upon treatment of transfected smooth muscle cells with phorbol 12-myristate 13-acetate (PMA) as well as upon transfection with a EJ/ras encoding plasmid. Conversely, it was downmodulated upon transfection with wild-type p53 but not with mutated p53. The regulatory sequences involved in the positive regulation of the gene were located upon serial deletion experiments.

Lack of modulation of the galectin-3 and asialoglycoprotein receptor transcription in hepatocarcinoma of transgenic mice

Galectin-3 and asialoglycoprotein receptor are lectins belonging to the classes of soluble lectins and of membrane C type lectins respectively. Conflicting results have been reported concerning their transcription level in the time course development of tumours. In the present study we investigated the abnormalities and the transcription levels of galectin-3 and asialoglycoprotein receptor genes in liver-targeted SV40 large T transgenic mice related to normal mice. In the strain expressing the highest level of large T, 100% of the male mice reproducibly developed an hepatocarcinoma. We provide evidence that the galectin-3 and asialoglycoprotein receptor genes are stable in such mice. The galectin-3 gene is weakly transcribed and its level is identical and constant in normal and transgenic mice, suggesting a lack of involvement in the development of large T-induced hepatocarcinoma. The asialoglycoprotein receptor gene is actively transcribed and its level remains high all along the development of the tumour; therefore, in such an hepatocarcinoma the asialoglycoprotein receptor could be used to take up drugs, genes or oligonucleotides associated with glycosylated carriers bearing galactose residues in a terminal non-reducing position.

Increased alpha2,6 sialylation of N-glycans in a transgenic mouse model of hepatocellular carcinoma

Liver cancer is one of the most frequent and lethal malignancies worldwide. Early detection is hampered by the absence of reliable markers. Mice transgenic for the SV40 large T antigen under the control of a liver-specific promoter spontaneously develop well-differentiated hepatocellular carcinomas between 8 to 10 weeks of age. They are excellent models to investigate the alterations of protein expression in the early stages of tumor development and to follow these changes during tumor progression. In the present study, we analyzed the glycosylation changes occurring during tumor development in transgenic mice expressing the SV40 T antigen under the control of the antithrombin III promoter. The analysis of serum and liver glycoproteins by an ELISA type assay, using the lectin from Sambucus nigra (SNA) as a probe, revealed the presence of increased levels of Neu5Ac alpha2,6Gal beta1,4GlcNAc on N-glycans in the tumor-bearing transgenic mice as compared to controls. On serum glycoproteins the increase in alpha2,6 sialylation followed tumor progression, reaching up to 10 times control levels. However, significantly higher SNA binding (2-fold) could already be observed on serum glycoproteins from mice exhibiting only microscopically small neoplastic foci. On liver membrane glycoproteins, the increase in alpha2,6 sialylation was less pronounced, reaching two to three times control values in 6-month-old mice. Western blotting of serum and liver proteins with radiolabeled SNA showed that all glycoproteins that bind the lectin in controls exhibit larger amounts of Neu5Ac alpha2,6Gal beta1,4GlcNAc on N-glycans in the tumor-bearing mice. This general increase in alpha2,6 sialylation on all glycoproteins is due to the increased activity of the galactoside:alpha2,6 sialyltransferase (ST6Gal I), which specifically transfers Neu5Ac residues in alpha2,6 linkage to Gal beta1,4GlcNAc units on N-glycans. As for the structures synthesized by the enzyme, the increase of ST6Gal I activity in the serum as well as in liver microsomes of the transgenic mice followed tumor progression. Interestingly, the activity of the galactoside:alpha2,3 sialyltransferase (ST3Gal III), which uses the same acceptor substrate (Gal beta1,4GlcNAc), was unchanged in the earlier stages of tumor development but decreased in the serum and in liver microsomes from later stages. Using a rat ST6Gal I cDNA as a probe, Northern blots of total RNA extracted from the livers of control and transgenic mice revealed an increased (4-fold) expression of the ST6Gal I gene. The single transcripts detected in both normal and cancerous liver showed identical size.

The direct agglutination test: a non-specific test specific for the diagnosis of visceral leishmaniasis?

Serology has an important role to play in the diagnosis of the severe clinical syndrome of visceral leishmaniasis (VL). The direct agglutination test (DAT), a simple agglutination test which requires no laboratory facilities, has become the preferred test, particularly for field studies. The nature of the antigens responsible for the agglutination of leishmanial promastigotes by the serum of VL patients is not known. A series of experiments which provide some clues to the molecular basis for the test and which indicate that there might be more in DAT than meets the eye is reported.

Cytosolic and nuclear delivery of oligonucleotides mediated by an amphiphilic anionic peptide

Antisense oligonucleotides (ODN) were easily introduced into the cytosol of mammalian cells on permeabilization of the plasma membrane by an amphiphilic anionic peptide. The E5CA peptide (GLFEAIAEFIEGGWEGLIEGCA) is an E5 peptide analog derived from the N-terminal segment of the HA2 subunit of influenza virus hemagglutinin. This peptide undergoes a conformational change when the pH shifts from neutral to around 6.0, inducing a transient permeabilization of the plasma membrane. In the presence of the E5CA peptide at pH close to 6.0, fluoresceinylated ODN were rapidly taken up by cells and diffused into the nucleus. The uptake of ODN was dependent on the E5CA peptide concentration and on the duration of the incubation at low pH, as shown by confocal microscopy and flow cytometry analyses. This procedure is suitable for loading adherent cells as well as nonadherent cells with single-stranded or double-stranded ODN. Under optimal conditions, a high percentage of cells were nuclei loaded, and the viability was not affected. This method makes use of a well-defined chemical product without the requirement of any special equipment. It will be useful to study the interactions of single-stranded or double-stranded ODN used as antisense, antigenes, or decoys.

Ultraviolet rays induced expression of lectins on the surface of a squamous carcinoma keratinocyte cell line

Human keratinocytic cells from squamous carcinoma (SCL-1) present, under resting conditions, relatively low amounts of endogenous lectins (sugar-binding proteins). Upon uv irradiation, they express on their cell surface large amounts of endogenous lectin molecules able to bind neoglycoproteins bearing either alpha-L-rhamnosyl or alpha-D-glucosyl residues. A similar binding specificity was found with normal human keratinocytes under the same culture conditions. At sunlike doses, uv.A (365 nm) was more efficient than uv.B (312 nm) in the expression of such receptors on the surface of SCL-1 cells. The increased presentation of lectins by SCL-1 cells was transient and reached a maximum 4 h after irradiation. Such a specific modulation of receptor expression upon uv irradiation might be biologically significant, considering the numerous intercellular recognition phenomena in skin biology. alpha-L-Rhamnose-specific receptor on SCL-1 could not be distinguished from alpha-D-glucose-specific receptor on the basis of neoglycoproteins binding, uptake, and related inhibitions. Lectin expression was mainly detected on the cell surface, and its overexpression due to uv rays required a de novo protein synthesis process.

Intracellular routing and inhibitory activity of oligonucleopeptides containing a KDEL motif

On internalization, oligonucleotides (ODN) remain mostly sequestered in endocytic compartments. To increase their delivery into the cytosol and/or nucleus, which contain their targets, we attempted to guide them into compartments containing the KDEL receptor. Antisense ODN, phosphodiester protected at both ends, that are complementary to the AUG initiation site of gagHIV-1 mRNA (odn) were linked to a peptide ending with the Lys-Asp-Glu-Leu (KDEL) motif in a carboxyl-terminal position (odn-p-KDEL) or with the Lys-Asp-Glu-Ala (odn-p-KDEA) as a control. The effect of odn substitution with a peptide was examined with regard to its accumulation, subcellular location, and activity in HepG2 cells. Although odn-p-KDEL was internalized 4-fold less than the corresponding peptide-free odn, it was 5-fold more efficient in inhibiting gagHIV-1 gene expression in HepG2 cells. The internalization of odn-p-KDEA was as low as that of odn-p-KDEL, but its biological activity was lower, close to that of the peptide-free odn. On endocytosis at 37 degrees, both conjugates as well as the peptide-free odn were found in a neutral environment. However, the substitution of an odn with a KDEL motif altered its intracellular trafficking; most of the odn-p-KDEL was found in the endoplasmic reticulum and in the intermediate compartment as identified by colabeling with either anti-ERGIC-53 or anti-KDEL receptor antibodies. Conversely, odn-p-KDEA and peptide-free odn were localized in vesicular compartments not labeled with these antibodies. In addition, pulse-chase experiments showed that odn-p-KDEL and odn-p-KDEA had a lower efflux than peptide-free odn. Therefore, the large increase in efficiency was due to the KDEL motif.

The reduction of the positive charges of polylysine by partial gluconoylation increases the transfection efficiency of polylysine/DNA complexes

A polylysine partially substituted with polyhydroxyalkanoyl residues and specially with gluconoyl residues was developed in order to increase the transfection efficiency by decreasing the strength of the electrostatic interactions between the DNA and the cationic polymer. Partially gluconoylated polylysine/DNA complexes were more easily dissociated in solution and their transfection efficiency in the presence of chloroquine, evaluated with HepG2 cells, a human hepatocarcinoma line, was higher when 43 +/- 4% of the epsilon-amino groups of polylysine were blocked with gluconoyl residues. Partially gluconoylated polylysine/plasmid complexes were efficient in transfecting different adherent as well as non-adherent cell lines. Partially gluconoylated polylysine formed highly soluble (above 100 micrograms/ml in DNA) complexes with DNA plasmids. In addition, partially gluconoylated polylysine bearing few lactosyl residues increased the transfection efficiency of HepG2 cells which express a galactose-specific membrane lectin.

Selective induction of peripheral and mucosal endothelial cell addressins with peripheral lymph nodes and Peyer's patch cell-conditioned media

Vascular endothelial cell addressins play an important role in lymphocyte homing in secondary lymphoid organs and in chronic inflammatory areas. A SV40 large T antigen-immortalized cell line from peripheral lymph nodes, HECa1O [Bizouarne et al., 1993a], was used to characterize the location of addressins with regard to environmental factors and cytokines. For this purpose, two monoclonal antibodies, MECA 79 and MECA 367, specific for peripheral lymph node vascular addressin and for mucosal addressin (Peyer's patches), respectively, were bound to unstimulated HECa1O cells. Both mucosal and peripheral addressins were detected inside the cells and in cellular extracts of the resting cells. On the cell surface, both addressins could be evidenced on the same cells at a moderate level of expression. They partly mediate the EL4/EL4IL2 lymphoma cells' adhesion to HECa1O cells. Supernatants of cultured peripheral lymph node or Peyers' patch cells induced expression of MECA 79 or MECA 367 antigens, respectively, on the surface of HECa1O cells. Interleukins, IL-7, IL-3, and IL-8, induced the cell-surface appearance of MECA 79 but not of MECA 367 antigen. Therefore, the same cell type synthesizes both antigens, but the expression of these antigens on the cell surface is independently regulated, thus uncovering a characteristic tissue type-specific as well as environment-sensitive properties of microvascular endothelial cells.

Enhanced biological activity of antisense oligonucleotides complexed with glycosylated poly-L-lysine

We sought to exploit glycosylated poly-L-lysine (pLK) to increase the uptake and biological antisense activity of a phosphorothioate oligonucleotide (pt-odn) [pt-odn complementary to the 3' noncoding region of intercellular adhesion molecule-1 (ICAM-1) (odn(ICAM-1))] complementary to the 3'-noncoding region of ICAM-1 in A549 cells. Dose-dependent inhibition of ICAM-1 expression was obtained (IC50 = 500 nM) through treatment of cells with odn(ICAM-1) complexed with pLK carrying fucose residues in the presence of 100 microM chloroquine. Alteration in the charge ratio between fucosylated pLK and pt-odn had a significant effect on the efficacy of inhibition (optimal conditions, charge ratio = 1.1). This effect was also dependent on the number of fucose moieties per pLK. Free pt-odn or pt-odn complexed with nonglycosylated pLK gave no inhibition at concentrations of < or = 2 microM. Two control pt-odn (one was targeted against an unrelated gene not present in these cells, gag(HIV), and the other had a randomized sequence) gave no inhibition of ICAM-1 expression in the presence or absence of pLK carrying fucose residues at concentrations of < or = 2 microM. When complexed with pLK carrying 100 fucose residues, the amount of cell-associated pt-odn was increased by 15-fold compared with the free pt-odn. Nongycosylated pLK also increased the amount of cell-associated pt-odn by >10 fold but did not alter the biological activity. These results demonstrate clearly the potential of glycosylated pLK as a pt-odn transporter.