First safety and response results of a randomized phase III study with liposomal platin in the treatment of advanced squamous cell carcinoma of the head and neck (SCCHN).

BACKGROUND: Cisplatin is one of the most active chemotherapeutic agents used in the treatment of advanced squamous cell carcinoma of the head and neck (SCCHN). However, its clinical efficacy is limited by its renal and hematotoxicity profile. In a randomized, multicenter phase III trial, we replaced conventional cisplatin by a liposomal formulation of cisplatin (lipoplatin) and compared the safety and efficacy profiles of patients in the two treatment arms. PATIENTS AND METHODS: Main inclusion criteria were: histologically confirmed SCCHN, age between 18-75 years with sufficient renal function. Main endpoints for this interims analysis were hemato- and nephrotoxicity. First response data were collected. RESULTS: Forty-six patients were evaluable for outcome and toxicity. Grade III and IV hematotoxicity were more frequent in the cisplatin arm (31.7% vs. 12%), with grade IV leucopenia occurring in 22.2%. However, 16% of the patients in that treatment arm experienced grade III anemia compared to only 9.5% treated with the cisplatin regimen. A total 4% of the patients in the lipoplatin arm developed grade IV neuropathy, whereas in the cisplatin arm, 19% developed grade III neuropathy and none developed grade IV. The renal toxicity profile of both drugs also showed marked differences. In the cisplatin arm, 23.8% of patients suffered grade III toxicity. In contrast, no grade III or IV renal toxicity occurred in patients treated with lipoplatin. The efficacy results showed 38.8% objective partial remission in the cisplatin arm vs. 19% in the lipoplatin arm. However 64% of the patients achieved stable disease while being treated with lipoplatin/5-fluorouracil (5-FU), vs. 50% in the cisplatin/5-FU arm. CONCLUSION: Liposomal cisplatin seems to reduce both the renal and hematological toxicity to a clinically relevant extent as compared to conventional cisplatin. The clinical benefit rate is similar for both regimens.

Pharmacokinetics of liposomal cisplatin (lipoplatin) in combination with 5-FU in patients with advanced head and neck cancer: first results of a phase III study.

BACKGROUND: Lipoplatin, a novel liposomal formulation of cisplatin, is composed of cisplatin and liposomes based on dipalmityl phosphatidyl glycerol (DPPG), soy phosphatidyl choline (SPC-3), cholesterol and methoxypolyethylene glycol-distearoyl phosphatidylethanolamine (mPEG2000-DSPE). Liposomal encapsulation of cisplatin is designed to increase safety and tolerability by decreasing, e.g., nephrotoxicity through decreased exposure of organs to cisplatin, while effectively delivering the drug to the tumor. In an ongoing phase III trial comparing cisplatin to lipoplatin (both in combination with infusional high-dose 5-Fluoruracil) in advanced head and neck cancer (HNC), a sub-study to determine the pharmacokinetic profile of lipoplatin in comparison to conventional cisplatin was undertaken. MATERIALS AND METHODS: In total, twelve patients with advanced HNC received a combination chemotherapy with either lipoplatin/5-FU or cisplatin/5-FU. Plasma samples were analyzed for concentration of total platinum in patients from both arms. RESULTS: All twelve patients from the pharmacokinetic sub-study were male Caucasians at a mean age of 60 years. There was no difference in age or kidney function between the two treatment groups. The total body clearance for cisplatin was 1.25 L/(hxm2) for the liposomal formulation, compared to 0.62 L/(hxm2) for conventional cisplatin. The terminal half life was half as long for lipoplatin (10.98 h) as compared to cisplatin (24.5h). Even though the maximum observed concentration in the plasma (C(max) was greater for lipoplatin than for cisplatin, the area under the concentration time-curve (AUC) was less (6.5 microg/ml vs. 4.07 microg/ml and 66.85 microg/h/ml vs. 130.33 microg/h/ml, respectively). CONCLUSION: The pharmacokinetic profile of lipoplatin (in combination with 5-FU) suggests that the liposomal formulation results in a greater body clearance and shorter half life than conventional cisplatin, which confirms the clinical observation of decreased taxicity, especially renal deterioration.

Chromatin domains and prediction of MAR sequences.

Polynuceosomes are constrained into loops or domains and are insulated from the effects of chromatin structure and torsional strain from flanking domains by the cross-complexation of matrix-attached regions (MARs) and matrix proteins. MARs or SARs have an average size of 500 bp, are spaced about every 30 kb, and are control elements maintaining independent realms of gene activity. A fraction of MARs may cohabit with core origin replication (ORIs) and another fraction might cohabit with transcriptional enhancers. DNA replication, transcription, repair, splicing, and recombination seem to take place on the nuclear matrix. Classical AT-rich MARs have been proposed to anchor the core enhancers and core origins complexed with low abundancy transcription factors to the nuclear matrix via the cooperative binding to MARs of abundant classical matrix proteins (topoisomerase II, histone H1, lamins, SP120, ARBP, SATB1); this creates a unique nuclear microenvironment rich in regulatory proteins able to sustain transcription, replication, repair, and recombination. Theoretical searches and experimental data strongly support a model of activation of MARs and ORIs by transcription factors. A set of 21 characteristics are deduced or proposed for MAR/ORI sequences including their enrichment in inverted repeats, AT tracts, DNA unwinding elements, replication initiator protein sites, homooligonucleotide repeats (i.e., AAA, TTT, CCC), curved DNA, DNase I-hypersensitive sites, nucleosome-free stretches, polypurine stretches, and motifs with a potential for left-handed and triplex structures. We are establishing Banks of ORI and MAR sequences and have undertaken a large project of sequencing a large number of MARs in an effort to determine classes of DNA sequences in these regulatory elements and to understand their role at the origins of replication and transcriptional enhancers.

A compilation and classification of DNA binding sites for protein transcription factors from vertebrates.

The field of protein transcription regulators and their DNA sequence specificity has been the most rapidly expanding in the last few years. The concerted interplay of protein transcription factors on the regulatory regions of eukaryotic genes (promoters, enhancers, origins of replication, silencers, and matrix-attached regions) regulates transcription levels; the differential activity of genes during development and the cell cycle, between cell types, and in response to physiological stimuli results from interdigitation of regulatory circuits controlling transcription initiation, finely tuned by the relative amounts of protein factors synthesized in a cell type, their phosphorylation, isoforms within factor families, the way protein regulators are brought in contact with one another through the patchwork of their cognate sites on the regulatory regions of genes, and by regulation of their nuclear import. The varying affinity of the same factor for its cognate DNA in different promoters can also be modulated by the type of proteins it is brought into contact with, by one or more nucleotide changes in its binding sites among promoters, and by the chromatin structure. The classification of protein transcriptional regulators attempted here according to their DNA binding specificity into those that bind AT-, GC-, GA, TG-rich and mixed motif has one obvious advantage: different protein factors that bind to the same DNA sequence will be found within the same class. In addition, this classification has allowed us to discern a class of transcriptional regulators whose binding site consists of a GA- and a CT-rich moiety; no other two pairs of dinucleotides compose a major class of factor sites.(ABSTRACT TRUNCATED AT 250 WORDS)

Phosphorylation of transcription factors and control of the cell cycle.

Protein phosphorylation has evolved as the most versatile posttranslational modification widely used by cells. Signal transduction pathways mediated by activation of MAP kinases and protein kinase C trigger the exit of cells from the quiscence (Go-->G1 transition). Indeed, binding of growth factors at the cell surface triggers their receptors, usually possessing a tyrosine kinase on the cytoplasmic side, to phosphorylate other molecules passing on the information sequentially to GRB2 protein, to p21ras, to c-Raf-1, to MAP kinase kinase, to MAP kinase, to p90rsk, to transcription factors. Activated PKC, MAP kinase, and pp90src can translocate to the nucleus where they phosphorylate a number of protein transcription regulators in a cell cycle-dependent manner or in response to cell stimulation for exit from quiescence. The cell cycle is mainly regulated by p34cdc2 or otherwise called cdc2 in association with cyclins B at G2/M and by Cdk2 in association with cyclins A, D1, and E at G1/S checkpoints; phosphorylation of histone H1 and lamins by cdc2 triggers chromosome assembly and nuclear envelope breakdown, respectively, as a prelude to mitosis. Cdc2 activities functioning as a G2/M regulator are controlled by its phosphorylation and dephosphorylation at Ser/Thr residues. MAP kinases might be the missing link in the chain connecting the Go to G1 transition with the cell cycle regulation, whereas phosphorylation of replication protein factors, retinoblastoma, and p53 might link the G1 to S transition with the control of DNA synthesis. A number of transcription factors are known to stimulate DNA replication, including p53, c-Myc, AP-1, Oct-1, T-antigen; the DNA binding activities of all these proteins and their interaction with other transcription factors are controlled by phosphorylation. The nuclear import of several proteins including NF kappa B, Dorsal, glucocorticoid receptor, ISGF3, rNFIL-6, T antigen, and the kinases PKC, MAP, and p90rsk, are dependent on their phosphorylation at specific sites. Histone phosphorylation stimulated at discrete stages of the cell cycle or in response to cAMP or other stimuli might induce profound changes in chromatin organization.

Nuclear localization signals (NLS).

Available data strongly suggest that simple karyophilic clusters of arginines and lysines in nucleus-targeted proteins signal the anchoring of these proteins to specialized transporter molecules found on the pore complex or in the cytoplasm. These peptides have been termed nuclear localization signals (NLS). Other nuclear proteins contain "split" or "bipartite" NLS hypothesized to be reconstituted by protein folding or conformational change induced by glucocorticoids. The active NLS needs to be exposed to the protein surface, the cell has invented mechanisms to expose a hidden or cryptic NLS by protein phosphorylation, dephosphorylation, dissociation of an inhibitory subunit that masks the NLS, processing of a larger precursor, and binding of a hormone to regulate the nuclear import of a protein transcription factor at a certain stage of development, or cell cycle. It is proposed (1) that a hexapeptide with four arginines and lysines (and histidines, in some cases) is a good candidate for a "core NLS," (2) that acidic domains on proteins to be imported may participate in anchoring them to the transporter cytoplasmic or pore complex NLS-receptor, and (3) that NLS have both a cytoplasmic and a nuclear function. The interaction between nuclear proteins and transporter proteins in the pore appears to be largely electrostatic and to be disrupted by the binding of mRNA molecules to the same transporter protein, functioning also for the cytoplasmic export of RNA.

Poly (adp-ribosyl)ation of variant histones and possible implications in carcinogenesis.

Poly(ADP-ribose) is a third species of nucleic acids. The rapid synthesis and degradation of poly(ADP-ribose) is a dynamic nuclear process that seems to be involved in carcinogenesis via its implication in repair, replication, gene expression, recombination and chromatin remodeling in development. In the present study, the ADP-ribosylated species of histones synthesized in human erythroid K562 cell nuclei were examined by high-resolution two-dimensional polyacrylamide gel electrophoresis. Modification of core histone variants appears to be highly specific. Variants of histone H3 are modified with one or more than five ADP-ribosyl groups, whereas the concentrations of di-, tri- and tetra(ADP-ribosyl)ated H3 is extremely low. This is in contrast to the presence of a continuous ladder of mono- to poly(ADP-ribosyl)ated species for the variants of H2B, H2A, H1 and H4. The order of poly(ADP-ribosyl)ation of core histones in isolated nuclei is H2B > H2A.X > H3.2,3 approximately H4 > H2A.Z approximately H3. 1. A model is proposed to explain how mono- and oligo(ADP-ribosyl)ated core histones containing 1 - 7 ADP-ribosyl groups might unfold higher-order chromatin structures and open individual nucleosomes into transcriptionally poised structures. The relevance of these studies to molecular carcinogenesis are discussed.

Myb proteins talking to their DNA (review).

DNA sequence-specific proteins called transcription factors found in all multicellular organisms control the expression of genes and are involved in the regulation of the cell cycle. Myb oncoproteins are transcription factors with a distinct DNA binding domain lacking zinc fingers or a basic region. Found originally in the avian myeloblastosis virus (v-Myb), then in the genome of chickens (c-Myb), the DNA binding domain of Myb occurs in regulatory proteins from mammals including humans, plants, flies, Dictyostelium, and yeast. Myb proteins show a preference for the AACGTT, AACnGTT and the [GRAPHICS] motifs with AAC (or its complementary GTT) as the core recognition site. Our searches through a dictionary of DNA recognition sequences of vertebrate transcription factors for Myb sites suggests that a subset of regulatory sequences of defined genes that are targets of the transcription factors XBP, ATF, LRF-1, GR, NF-AT, HNF-1, GATA-1, H1P1, alpha-CP-1, AP-3, RF-X, the nuclear matrix protein SATB1, and the ETS proteins PEA3, E74A, and GABPalpha share the Myb binding site and might thus 'crosstalk' with Myb on selected genomic targets.

Nuclear localization signal peptides for the import of plasmid DNA in gene therapy (review).

Failure of cells to import specific proteins into nuclei can lead to carcinogenesis. Trafficking of nuclear proteins from the site of their synthesis in the cytoplasm to the sites of function in the nucleus through pore complexes is mediated by the presence of nuclear localization signals (NLSs) on proteins to be imported into nuclei. mRNA is exported through the same route as a complex with nuclear proteins. During gene therapy the foreign DNA needs to enter nuclei for its transcription in order to express the therapeutic protein; a pathway is proposed involving the complexation of plasmids and oligonucleotides with nascent nuclear proteins possessing NLSs as a prerequisite for their nuclear import. Covalent linkage of NLS peptides to oligonucleotides and plasmids or formation of complexes of plasmids with proteins possessing multiple NLS peptides is proposed to increase their import rates and the efficiency of gene expression. We suggest that cancer cells import more efficiently foreign DNA into nuclei compared with terminally differentiated cells because of their increased rates of proliferation and protein import and are easier targets for expression of foreign genes.

Cancer gene therapy and immunotherapy (review).

Gene therapy is a newly hatched field of biomedical research aimed at introducing therapeutically important genes into somatic cells of patients for the treatment of human disease. Whereas for inborn errors of metabolism transfer of a single gene can correct the disorder, cancer is a complex disease involving mutations in a number of protooncogenes and tumor suppressor genes as well as an imbalance and disarray in phosphorylation events and regulatory circuits of the cell cycle; transfer of the wild-type p53 or p21 tumor suppressor genes is a successful gene therapy approach leading to apoptotic death of cancer cells or in restrain of their chaotic growth. A different promising approach is transfer of the herpes simplex virus thymidine kinase (HSV-tk) gene (suicide gene) and systemic treatment with the prodrug ganciclovir which is converted by HSV-tk into a toxic drug killing dividing cells. Expression of suicide genes, p53, and other therapeutic genes preferentially in cancer cells can be achieved by regulatory elements from tumor-specific genes such as carcinoembryonic antigen, BRCA1, and PSA.

Gene therapy to human diseases.

The identification of defective genes associated with a number of human disorders (tyrosine hydroxylase for Parkinson's disease, aspartylglucosaminidase in lysosomal storage disease, CFTR in cystic fibrosis, and LDL receptor in familial hypercholesterolemia) has promoted the development of strategies aimed at transferring to the somatic cells of the patient or of animal models vectors carrying the corrected gene. The obstacles to overcome include targeting the specific cell type or organ (liver for Factors VIII and IX in hemophilia), enhancing entry to cells into non-lysosomal compartments, nuclear import, percentage of cells transduced with the therapeutic gene, sustained expression of the transgene in human tissues, and immunogenicity of the transduced cells expressing the recombinant or viral proteins. Improvements in each single of these steps are likely to enhance enormously the potential of gene transfer for the treatment of human diseases. A number of human diseases including HIV infections and hypertension are approached by somatic gene transfer. VEGF regulating vascular permeability, growth of endothelial cells and angiogenesis, and TGF-B implicated in wound healing and in stimulation in synthesis of extracellular matrix, are potential targets for restenosis, atherosclerosis, and cancer.

A unified model explaining the preferential repair of active over inactive genes and of the transcribed over the nontranscribed strand.

DNA lesions are removed more efficiently and at higher rates from active than from inactive genes and from the transcribed compared with the nontranscribed strand. A unifying model is presented explaining the heterogeneity in DNA repair activities through the genome in terms of chromatin structure, nuclear matrix anchorage, transcription factor binding, protein phosphorylation mechanisms, and the linkage of repair mechanisms to other nuclear functions including transcription and replication. Transcription factors, preferentially assembled into complexes on the regulatory regions of active but not of inactive genes are proposed to contribute to repair differences.

The non-uniform repair of active and inactive chromatin domains (review).

Substantial evidence demonstrates that DNA repair processes are not distributed homogeneously throughout the genome but that lesions are removed more efficiently and at higher rates from active than from inactive genes. Transcription-coupled repair (TCR) appears to be a sophisticated subpathway of nucleotide excision repair (NER) that targets the repair machinery to lesions present in the actively transcribed strand; TCR could explain the preferential repair of the transcribed over the nontranscribed strand in active genes and involves ERCC2 and ERCC3 helicases in human cells, integral components of TFIIH, as well as ERCCG. XPC protein is responsible for the repair of inactive regions. In S. cerevisiae RAD16 helicase as well as RAD9 and RAD24 proteins are involved in repairing nontranscribing regions whereas RAD3 and SSL2 helicases, homologs of the human ERCC2 and ERCC3, function for the repair of active regions. Unraveling the mechanisms that govern the heterogeneous repair rates among active and inactive parts of the genome is important in understanding mechanisms of carcinogenesis.

The phosphorylation connection to cancer (review).

Despite our incomplete comprehension of how growth factor-stimulation of cells is linked to the cell cycle and of how the G(1)/S checkpoint is linked to initiation in DNA replication there is an unparalleled wealth of experimental evidence to connect protein phosphorylation to molecular mechanisms of carcinogenesis. Many growth factors, growth factor receptors with tyrosine kinase activity (insulin receptor, EGFR, PDGFR, CSF1R, NGFR, HGFR), nonreceptor serine/threonine or tyrosine kinases (c-Raf-1, cMos, c-Abl, c-Src) and cyclin D1 are encoded by oncogenes mutated or overexpressed in a variety of human tumors; the physiological functions of oncoproteins that are involved in gene expression and replication (c-Jun, T-antigen, c-Myc, c-Myb) as well as p53, RB and CDK4 tumor suppressor proteins and replication factor A are also regulated by phosphorylation, ms genes transduce growth factor receptor signals to protein kinase C (PKC) or to c-Raf-1 triggering two different cascades of protein kinases, the PKC and MAPK signaling pathways both targeting nuclear proteins. Thus cancer can be considered as a disease of the signaling pathways.

How enhancers work - juxtapositioning of DNA control elements by synergistic interaction of Mars - (review-hypothesis).

The sites of attachment of chromatin loops to the nuclear matrix (MARs) seem to harbor transcriptional enhancers, promoters and origins of replication (ORIs). According to the model proposed, the cooperative interactions among classical nuclear matrix proteins which are abundant (topoisomerase II, histone H1, HMG-I(Y), lamins A, B1, SAF-A, ARBP and others) bring together distant AT-rich classical MAR sequences causing looping of DNA. This process juxtaposes enhancers, ORIs, promoters, and other control elements that cohabit with MARs loaded with the less abundant transcription factors (TFs) facilitating productive interactions between enhancers and promoters or enhancers and core ORIs. The implications of the model in the interactions of oncoproteins with regulatory DNA elements and their integration into a chromatin and nuclear matrix environment are discussed.

Chromatin and nuclear matrix in development and in carcinogenesis - a theory.

Sufficient evidence indicates that the nuclear matrix which organizes chromatin into loops or domains harbors origins of replication, homeodomain protein binding sites, and transcriptional enhancer sequences. Thus nuclear matrix might sculpture the crossroads of the differential gene expression and differential activation of origins of replication occurring during development. Nuclear matrix is also the site of transcription, replication, repair and recombination. These properties of the nuclear matrix are explored in the context of their implications in carcinogenesis. Many tumors result from abnormal translocations of DNA segments at other genomic sites via a recombination process involving some unusual DNA sequences at the breakpoints, such as Alu repetitive sequences or Z-DNA. Translocations are predicted to occur more frequently on the nuclear matrix and to involve DNA sequences at the bases of chromatin loops. Other tumors may result from the overexpression of distinct genes whose transcription regulation occurs at their nuclear matrix attachment sites. Tumorigenesis connected with mutations at coding DNA sequences is also related to the nuclear matrix, since nuclear matrix governs repair activities in the nucleus and might also harbor the most vulnerable DNA sites to damage and the most refractory to repair. These novel, and the previously known, features of the nuclear matrix reveal a new approach in understanding carcinogenesis.

A novel class of matrix attached regions (Mars) identified by random cloning and their implications in differentiation and carcinogenesis.

MARs, i.e., the attachment points of chromatin loops to the insoluble nuclear suprastructure called nuclear matrix, define the borders between domains, are the sites of initiation of DNA replication, and display transcriptional enhancer activity. MARs play an important role in differentiation and in carcinogenesis. MAR-protein complexes were proposed to interact with proteins bound to the immediate upstream regulatory region of active genes and bring the 5' end of active genes onto the nuclear matrix where transcription factors and RNA polymerases are located. We have isolated the nuclear matrix from human K562 cells, comprising 1-2% of nuclear DNA, after removal of the bulk of nuclear polynucleosomes with micrococcal nuclease. The MAR DNA was cloned in E. coli. We report here the sequence of a MAR fragment of 480 bp (clone Hum. MAR 19.2a). This sequence contains homopurine stretches of 5-10 nt alternating with homopyrimidine stretches of 4-17 nt, favoring formation of cruciform structures on the DNA. We have identified seven potential cruciform structures, characteristic elements of origins of replication, on the 480 bp 19.2a fragment. In addition this MAR clone contains three ATTA-type motifs representing homeodomain protein binding sites and five TG-rich (or its complementary CA-rich) stretches that we call TG boxes. TG boxes are found at recombination sites. in chromosome telomeres, and in the binding sites of a class of protein factors that regulate transcription and replication. The present study showing a multitude of potential cruciform structures characteristic of origins of replication in the Hum. MAR 19.2a clone gives further support to the idea that a subset of human MARs may function as origins of DNA replication. The relevance of these studies to carcinogenesis and differentiation are discussed.

Histones, protamine, and polylysine but not poly(E:K) enhance transfection efficiency.

Liposomal gene delivery has a great potential for the treatment of cancer and other human diseases. In this work we have investigated the optimal conditions for liposome-mediated transfer of the luciferase gene to human erythroleukemia K562 cells. DDAB:DOPE liposomes were more efficient than lysyl-DOPE:DOPE (1:2) and DDAB:cholesterol for transfection. Total histones from bovine thymus, salmon sperm protamine, and polylysine at an optimal ratio of 0.5 mg protein/mg DNA enhanced up to 7-fold the transfection efficiency of luciferase plasmids; on the contrary, the synthetic polymer poly(E:K), containing glutamic acid and lysine residues in a random order at a ratio 1:4, diminished luciferase expression. Transfection was nearly zero at high histone:DNA ratio by reversal of the charge of the particle from negative to positive leading:to its inability to interact with cationic liposomes. The increase in luciferase gene expression by DNA-binding proteins might arise from an increased transfer across the cell membrane of the liposome-DNA-protein complex but also by an increase in nuclear import of the DNA-protein complex because of the presence of nuclear localization signals on the protein molecule used for DNA condensation.

The 3' untranslated region of the human poly(ADP-ribose) polymerase gene is a nuclear matrix anchoring site.

The nuclear matrix displays the most dramatic changes among all cellular structures during carcinogenesis. Matrix attachment regions (MARs) organize chromatin into domains, harbor origins of replication and display a notable transcriptional enhancer activity. To understand the nature of MARs and their involvement in gene expression, replication, and carcinogenesis, we have cloned the MAR DNA fragments, of a size of 0.1-5.0 kb, isolated from human cells in culture. Over 150 clones have been sequenced. One MAR clone was identified as a stretch of 393 bp from the 3' untranslated region (3' UTR) of the human poly(ADP-ribose) polymerase (PARR) gene (100% homology). The 393 bp MAR fragment contains several repeats of TTGTTTTGT and related sequences (the TG boxes) and motifs with similarity to the binding site of the general yeast transcription factor GFI and to the ARS origins of replication in yeast. In addition, the 3' UTR of the PARP gene harbors MAR-type sequences found in other genes, kinked and curved DNA, two imperfect inverted repeats, and short alternating GA- and CT-rich motifs. The presence of TG-, GA-, and CT-rich motifs and of potential cruciforms is proposed to identify a novel type of MAR sequence. This report suggests that MAR sequences may reside in the 3' untranslated region of other genes and has important implications for a potential role of the nuclear matrix in transcription termination.

Liposome DNA delivery and uptake by cells (review).

Human gene therapy is a rapidly emerging field; therapeutic genes are engineered into adenovirus, retrovirus, or into plasmid-liposome complexes for their delivery into cells in culture or in vivo. Steps to improve for successful liposome-mediated gene delivery to somatic cells include persistence of the plasmid in blood circulation, port of entry and transport across the cell membrane, release from endosomal compartments into the cytoplasm, nuclear import by docking through the pore complexes of the nuclear envelope, expression driven by the appropriate promoter/enhancer control elements, and persistence of the plasmid in the nucleus for long periods. A number of strategies for enhancing the efficiency of uptake by the cells and release from endosomal compartments of liposome-plasmid or liposome-oligonucleotide complexes are reviewed here; emphasis is given to the direction of liposomes to caveolae vesicles.