A comprehensive analysis of human gene expression profiles identifies stromal immunoglobulin κ C as a compatible prognostic marker in human solid tumors.

PURPOSE: Although the central role of the immune system for tumor prognosis is generally accepted, a single robust marker is not yet available. EXPERIMENTAL DESIGN: On the basis of receiver operating characteristic analyses, robust markers were identified from a 60-gene B cell-derived metagene and analyzed in gene expression profiles of 1,810 breast cancer; 1,056 non-small cell lung carcinoma (NSCLC); 513 colorectal; and 426 ovarian cancer patients. Protein and RNA levels were examined in paraffin-embedded tissue of 330 breast cancer patients. The cell types were identified with immunohistochemical costaining and confocal fluorescence microscopy. RESULTS: We identified immunoglobulin κ C (IGKC) which as a single marker is similarly predictive and prognostic as the entire B-cell metagene. IGKC was consistently associated with metastasis-free survival across different molecular subtypes in node-negative breast cancer (n = 965) and predicted response to anthracycline-based neoadjuvant chemotherapy (n = 845; P < 0.001). In addition, IGKC gene expression was prognostic in NSCLC and colorectal cancer. No association was observed in ovarian cancer. IGKC protein expression was significantly associated with survival in paraffin-embedded tissues of 330 breast cancer patients. Tumor-infiltrating plasma cells were identified as the source of IGKC expression. CONCLUSION: Our findings provide IGKC as a novel diagnostic marker for risk stratification in human cancer and support concepts to exploit the humoral immune response for anticancer therapy. It could be validated in several independent cohorts and carried out similarly well in RNA from fresh frozen as well as from paraffin tissue and on protein level by immunostaining.

Human epidermal Langerhans cells express the tight junction protein claudin-1 and are present in human genetic claudin-1 deficiency (NISCH syndrome).

Claudin-1 (CLDN1) is a structural tight junction (TJ) protein and is expressed in differentiating keratinocytes and Langerhans cells in the epidermis. Our objective was to identify immunoreactive CLDN1 in human epidermal Langerhans cells and to examine the pattern of epidermal Langerhans cells in genetic human CLDN1 deficiency [neonatal ichthyosis, sclerosing cholangitis (NISCH) syndrome]. Epidermal cells from healthy human skin labelled with CLDN1-specific antibodies were analysed by confocal laser immunofluorescence microscopy and flow cytometry. Skin biopsy sections of two patients with NISCH syndrome were stained with an antibody to CD1a expressed on epidermal Langerhans cells. Epidermal Langerhans cells and a subpopulation of keratinocytes from healthy skin were positive for CLDN1. The gross number and distribution of epidermal Langerhans cells of two patients with molecularly confirmed NISCH syndrome, however, was not grossly altered. Therefore, CLDN1 is unlikely to play a critical role in migration of Langerhans cells (or their precursors) to the epidermis or their positioning within the epidermis. Our findings do not exclude a role of this TJ molecule once Langerhans cells have left the epidermis for draining lymph nodes.

T cell division and death are segregated by mutation of TCRbeta chain constant domains.

We have studied the role of the T cell receptor (TCR) beta chain transmembrane and cytoplasmic domains (betaTM/Cyto) in T cell signaling. Upon antigen stimulation, T lymphocytes expressing a TCR with mutant and betaTM and Cyto domains accumulate in large numbers and are specifically defective in undergoing activation-induced cell death (AICD). The mutant TCR poorly recruits the protein adaptor Carma-1 and is subsequently impaired in activating NF-kappaB. This signaling defect leads to a reduced expression of Fas ligand (FasL) and to a reduction in AICD. These beta chain domains are involved in discriminating cell division and apoptosis.

Microglia/macrophages migrate through retinal epithelium barrier by a transcellular route in diabetic retinopathy: role of PKCζ in the Goto Kakizaki rat model.

Diabetic retinopathy is associated with ocular inflammation, leading to retinal barrier breakdown, macular edema, and visual cell loss. We investigated the molecular mechanisms involved in microglia/macrophages trafficking in the retina and the role of protein kinase Cζ (PKCζ) in this process. Goto Kakizaki (GK) rats, a model for spontaneous type 2 diabetes were studied until 12 months of hyperglycemia. Up to 5 months, sparse microglia/macrophages were detected in the subretinal space, together with numerous pores in retinal pigment epithelial (RPE) cells, allowing inflammatory cell traffic between the retina and choroid. Intercellular adhesion molecule-1 (ICAM-1), caveolin-1 (CAV-1), and PKCζ were identified at the pore border. At 12 months of hyperglycemia, the significant reduction of pores density in RPE cell layer was associated with microglia/macrophages accumulation in the subretinal space together with vacuolization of RPE cells and disorganization of photoreceptors outer segments. The intraocular injection of a PKCζ inhibitor at 12 months reduced iNOS expression in microglia/macrophages and inhibited their migration through the retina, preventing their subretinal accumulation. We show here that a physiological transcellular pathway takes place through RPE cells and contributes to microglia/macrophages retinal trafficking. Chronic hyperglycemia causes alteration of this pathway and subsequent subretinal accumulation of activated microglia/macrophages.

Characterization of MAP1B heavy chain interaction with actin.

Microtubule-associated protein 1B is an essential protein during brain development and neurite outgrowth and was studied by several assays to further characterize actin as a major interacting partner. Tubulin and actin co-immunoprecipitated with MAP1B at similar ratios throughout development. Their identity was identified by mass spectrometry and was confirmed by Western blots. In contrast to previous reports, the MAP1B-actin interaction was not dependent on the MAP1B phosphorylation state, since actin was precipitated from brain tissue throughout development at similar ratios and equal amounts were precipitated before and after dephosphorylation with alkaline phosphatase. MAP1B heavy chain was able to bind actin directly and therefore the N-terminal part of MAP1B heavy chain must also contain an actin-binding site. The binding force of this interaction was measured by atomic force microscopy and values were in the same range as those of MAP1B binding to tubulin or that measured in MAP1B self-aggregation. Aggregation was confirmed by negative staining and electron microscopy. Experiments including COS-7 cells, PC12 cells, cytochalasin D and immunocytochemistry with subsequent confocal laser microscopy, suggested that MAP1B may bind to actin but has no obvious microfilament stabilizing effect. We conclude, that the MAP1B heavy chain has a microtubule-stabilization effect, and contains an actin-binding site that may play a role in the crosslinking of actin and microtubules, a function that may be important in neurite elongation.

Exovesicles from human activated dendritic cells fuse with resting dendritic cells, allowing them to present alloantigens.

Dendritic cells (DCs) can release microvesicles, but the latter's numbers, size, and fate are unclear. Fluorescently labeled DCs were visualized by laser-scanning microscopy. Using a Surpass algorithm, we were able to identify and quantify per cell several hundred microvesicles released from the surface of stimulated DCs. We show that most of these microvesicles are not of endocytic origin but result from budding of the plasma membrane, hence their name, exovesicle. Using a double vital staining, we show that exovesicles isolated from activated DCs can fuse with the membrane of resting DCs, thereby allowing them to present alloantigens to lymphocytes. We concluded that, within a few hours from their release, exovesicles may amplify local or distant adaptive immunological response.

Roles of transient guidepost neurons in corpus callosum formation and guidance

RESUMENeurones transitoires jouant un rôle de cibles intermédiaires dans le guidage des axones du corps calleuxLe guidage axonal est une étape clé permettant aux neurones d'établir des connexions synaptiques et de s'intégrer dans un réseau neural fonctionnel de manière spécifique. Des cellules-cibles intermédiaires appelées « guidepost » aident les axones à parcourir de longues distances dans le cerveau en leur fournissant des informations directionnelles tout au long de leur trajet. Il a été démontré que des sous-populations de cellules gliales au niveau de la ligne médiane guident les axones du corps calleux (CC) d'un hémisphère vers l'autre. Bien qu'il fût observé que le CC en développement contenait aussi des neurones, leur rôle était resté jusqu'alors inconnu.La publication de nos résultats a montré que pendant le développement embryonnaire, le CC contient des glies mais aussi un nombre considérable de neurones glutamatergiques et GABAergiques, nécessaires à la formation du corps calleux (Niquille et al., PLoS Biology, 2009). Dans ce travail, j'ai utilisé des techniques de morphologie et d'imagerie confocale 3D pour définir le cadre neuro-anatomique de notre modèle. De plus, à l'aide de transplantations sur tranches in vitro, de co-explants, d'expression de siRNA dans des cultures de neurones primaires et d'analyse in vivo sur des souris knock-out, nous avons démontré que les neurones du CC guident les axones callosaux en partie grâce à l'action attractive du facteur de guidage Sema3C sur son récepteur Npn- 1.Récemment, nous avons étudié l'origine, les aspects dynamiques de ces processus, ainsi que les mécanismes moléculaires impliqués dans la mise en place de ce faisceau axonal (Niquille et al., soumis). Tout d'abord, nous avons précisé l'origine et l'identité des neurones guidepost GABAergiques du CC par une étude approfondie de traçage génétique in vivo. J'ai identifié, dans le CC, deux populations distinctes de neurones GABAergiques venant des éminences ganglionnaires médiane (MGE) et caudale (CGE). J'ai ensuite étudié plus en détail les interactions dynamiques entre neurones et axones du corps calleux par microscopie confocale en temps réel. Puis nous avons défini le rôle de chaque sous-population neuronale dans le guidage des axones callosaux et de manière intéressante les neurones GABAergic dérivés de la MGE comme ceux de la CGE se sont révélés avoir une action attractive pour les axones callosaux dans des expériences de transplantation. Enfin, nous avons clarifié la base moléculaire de ces mécanismes de guidage par FACS sorting associé à un large criblage génétique de molécules d'intérêt par une technique très sensible de RT-PCR et ensuite ces résultats ont été validés par hybridation in situ.Nous avons également étudié si les neurones guidepost du CC étaient impliqués dans son agénésie (absence de CC), présente dans nombreux syndromes congénitaux chez 1 humain. Le gène homéotique Aristaless (Arx) contrôle la migration des neurones GABAergiques et sa mutation conduit à de nombreuses pathologies humaines, notamment la lissencéphalie liée à IX avec organes génitaux anormaux (XLAG) et agénésie du CC. Fait intéressant, nous avons constaté qu'ARX est exprimé dans toutes les populations GABAergiques guidepost du CC et que les embryons mutant pour Arx présentent une perte drastique de ces neurones accompagnée de défauts de navigation des axones (Niquille et al., en préparation). En outre, nous avons découvert que les souris déficientes pour le facteur de transcription ciliogenic RFX3 souffrent d'une agénésie du CC associé avec des défauts de mise en place de la ligne médiane et une désorganisation secondaire des neurones glutamatergiques guidepost (Benadiba et al., submitted). Ceci suggère fortement l'implication potentielle des deux types de neurones guidepost dans l'agénésie du CC chez l'humain.Ainsi, mon travail de thèse révèle de nouvelles fonctions pour ces neurones transitoires dans le guidage axonal et apporte de nouvelles perspectives sur les rôles respectifs des cellules neuronales et gliales dans ce processus.ABSTRACTRole of transient guidepost neurons in corpus callosum development and guidanceAxonal guidance is a key step that allows neurons to build specific synaptic connections and to specifically integrate in a functional neural network. Intermediate targets or guidepost cells act as critical elements that help to guide axons through long distance in the brain and provide information all along their travel. Subpopulations of midline glial cells have been shown to guide corpus callosum (CC) axons to the contralateral cerebral hemisphere. While neuronal cells are also present in the developing corpus callosum, their role still remains elusive.Our published results unravelled that, during embryonic development, the CC is populated in addition to astroglia by numerous glutamatergic and GABAergic guidepost neurons that are essential for the correct midline crossing of callosal axons (Niquille et al., PLoS Biology, 2009). In this work, I have combined morphological and 3D confocal imaging techniques to define the neuro- anatomical frame of our system. Moreover, with the use of in vitro transplantations in slices, co- explant experiments, siRNA manipulations on primary neuronal culture and in vivo analysis of knock-out mice we have been able to demonstrate that CC neurons direct callosal axon outgrowth, in part through the attractive action of Sema3C on its Npn-1 receptor.Recently, we have studied the origin, the dynamic aspects of these processes as well as the molecular mechanisms involved in the establishment of this axonal tract (Niquille et al., submitted). First, we have clarified the origin and the identity of the CC GABAergic guidepost neurons using extensive in vivo cell fate-mapping experiments. We identified two distinct GABAergic neuronal subpopulations, originating from the medial (MGE) and caudal (CGE) ganglionic eminences. I then studied in more details the dynamic interactions between CC neurons and callosal axons by confocal time-lapse video microscopy and I have also further characterized the role of each guidepost neuronal subpopulation in callosal guidance. Interestingly, MGE- and CGE-derived GABAergic neurons are both attractive for callosal axons in transplantation experiments. Finally, we have dissected the molecular basis of these guidance mechanisms by using FACS sorting combined with an extensive genetic screen for molecules of interest by a sensitive RT-PCR technique, as well as, in situ hybridization.I have also investigated whether CC guidepost neurons are involved in agenesis of the CC which occurs in numerous human congenital syndromes. Aristaless-related homeobox gene (Arx) regulates GABAergic neuron migration and its mutation leads to numerous human pathologies including X-linked lissencephaly with abnormal genitalia (XLAG) and severe CC agenesis. Interestingly, I found that ARX is expressed in all the guidepost GABAergic neuronal populations of the CC and that Arx-/- embryos exhibit a drastic loss of CC GABAergic interneurons accompanied by callosal axon navigation defects (Niquille et al, in preparation). In addition, we discovered that mice deficient for the ciliogenic transcription factor RFX3 suffer from CC agenesis associated with early midline patterning defects and a secondary disorganisation of guidepost glutamatergic neurons (Benadiba et al., submitted). This strongly points out the potential implication of both types of guidepost neurons in human CC agenesis.Taken together, my thesis work reveals novel functions for transient neurons in axonal guidance and brings new perspectives on the respective roles of neuronal and glial cells in these processes.

Cytosolic sensing of extracellular self-DNA transported into monocytes by the antimicrobial peptide LL37.

The intracellular location of nucleic acid sensors prevents recognition of extracellular self-DNA released by dying cells. However, on forming a complex with the endogenous antimicrobial peptide LL37, extracellular DNA is transported into endosomal compartments of plasmacytoid dendritic cells, leading to activation of Toll-like receptor-9 and induction of type I IFNs. Whether LL37 also transports self-DNA into nonplasmacytoid dendritic cells, leading to type I IFN production via other intracellular DNA receptors is unknown. Here we found that LL37 very efficiently transports self-DNA into monocytes, leading the production of type I IFNs in a Toll-like receptor-independent manner. This type I IFN induction was mediated by double-stranded B form DNA, regardless of its sequence, CpG content, or methylation status, and required signaling through the adaptor protein STING and TBK1 kinase, indicating the involvement of cytosolic DNA sensors. Thus, our study identifies a novel link between the antimicrobial peptides and type I IFN responses involving DNA-dependent activation of cytosolic sensors in monocytes.

Waddlia chondrophila enters and multiplies within human macrophages

Waddlia chondrophila is an obligate intracellular bacterium of the Chlamydiales order. W. chondrophila has been isolated twice from aborted bovine foetuses and a serological study supported the abortigenic role of W. chondrophila in bovine species. Recently, we observed a strong association between the presence of anti-Waddlia antibodies and human miscarriage. To further investigate the pathogenic potential of W. chondrophila in humans, we studied the entry and the multiplication of this Chlamydia-like organism in human macrophages. Confocal and electron microscopy confirmed that W. chondrophila is able to enter human monocyte-derived macrophages. Moreover, W. chondrophila multiplied readily within macrophages. The proportion of infected macrophages increased from 13% at day 0 to 96% at day 4, and the mean number of bacteria per macrophage increased by 3logs in 24h. Intracellular growth of W. chondrophila was associated with a significant cytopathic effect. Thus, W. chondrophila may enter and grow rapidly within human macrophages, inducing lysis of infected cells. Since macrophages are one of the major components of the innate immune response, these findings indirectly suggest the possible human pathogenicity of W. chondrophila.

Remorins form a novel family of coiled coil-forming oligomeric and filamentous proteins associated with apical, vascular and embryonic tissues in plants.

Remorins form a superfamily of plant-specific plasma membrane/lipid-raft-associated proteins of unknown structure and function. Using specific antibodies, we localized tomato remorin 1 to apical tissues, leaf primordia and vascular traces. The deduced remorin protein sequence contains a predicted coiled coil-domain, suggesting its participation in protein-protein interactions. Circular dichroism revealed that recombinant potato remorin contains an alpha-helical region that forms a functional coiled-coil domain. Electron microscopy of purified preparations of four different recombinant remorins, one from potato, two divergent isologs from tomato, and one from Arabidopsis thaliana , demonstrated that the proteins form highly similar filamentous structures. The diameters of the negatively-stained filaments ranged from 4.6-7.4 nm for potato remorin 1, 4.3-6.2 nm for tomato remorin 1, 5.7-7.5 nm for tomato remorin 2, and 5.7-8.0 nm for Arabidopsis Dbp. Highly polymerized remorin 1 was detected in glutaraldehyde-crosslinked tomato plasma membrane preparations and a population of the protein was immunolocalized in tomato root tips to structures associated with discrete regions of the plasma membrane.

Impact of mycelia on the accessibility of fluorene to PAH-degrading bacteria.

Mycelia have been recently shown to actively transport polycyclic aromatic hydrocarbons (PAH) in water-unsaturated soil over the range of centimeters, thereby efficiently mobilizing hydrophobic PAH beyond their purely diffusive transport in air and water. However, the question if mycelia-based PAH transport has an effect on PAH biodegradation was so far unsolved. To address this, we developed a laboratory model microcosm mimicking air-water interfaces in soil. Chemical analyses demonstrated transport of the PAH fluorene (FLU) by the mycelial oomycete Pythium ultimum that was grown along the air-water interfaces. Furthermore, degradation of mycelia-transported FLU by the bacterium Burkholderia sartisoli RP037-mChe was indicated. Since this organism expresses eGFP in response to a FLU flux to the cell, it was also as a bacterial reporter of FLU bioavailability in the vicinity of mycelia. Confocal laser scanning microscopy (CLSM) and image analyses revealed a significant increase of eGFP expression in the presence of P. ultimum compared to controls without mycelia or FLU. Hence, we could show that physically separated FLU becomes bioavailable to bacteria after transport by mycelia. Experiments with silicon coated glass fibers capturing mycelia-transported FLU guided us to propose a three-step mechanism of passive uptake, active transport and diffusion-driven release. These experiments were also used to evaluate the contributions of these individual steps to the overall mycelial FLU transport rate.

Spatio-temporal sequence of cross-regulatory events in root meristem growth.

A central question in developmental biology is how multicellular organisms coordinate cell division and differentiation to determine organ size. In Arabidopsis roots, this balance is controlled by cytokinin-induced expression of SHORT HYPOCOTYL 2 (SHY2) in the so-called transition zone of the meristem, where SHY2 negatively regulates auxin response factors (ARFs) by protein-protein interaction. The resulting down-regulation of PIN-FORMED (PIN) auxin efflux carriers is considered the key event in promoting differentiation of meristematic cells. Here we show that this regulation involves additional, intermediary factors and is spatio-temporally constrained. We found that the described cytokinin-auxin crosstalk antagonizes BREVIS RADIX (BRX) activity in the developing protophloem. BRX is an auxin-responsive target of the prototypical ARF MONOPTEROS (MP), a key promoter of vascular development, and transiently enhances PIN3 expression to promote meristem growth in young roots. At later stages, cytokinin induction of SHY2 in the vascular transition zone restricts BRX expression to down-regulate PIN3 and thus limit meristem growth. Interestingly, proper SHY2 expression requires BRX, which could reflect feedback on the auxin responsiveness of SHY2 because BRX protein can directly interact with MP, likely acting as a cofactor. Thus, cross-regulatory antagonism between BRX and SHY2 could determine ARF activity in the protophloem. Our data suggest a model in which the regulatory interactions favor BRX expression in the early proximal meristem and SHY2 prevails because of supplementary cytokinin induction in the later distal meristem. The complex equilibrium of this regulatory module might represent a universal switch in the transition toward differentiation in various developmental contexts.

VAMP-2 and cellubrevin are expressed in pancreatic beta-cells and are essential for Ca(2+)-but not for GTP gamma S-induced insulin secretion.

VAMP proteins are important components of the machinery controlling docking and/or fusion of secretory vesicles with their target membrane. We investigated the expression of VAMP proteins in pancreatic beta-cells and their implication in the exocytosis of insulin. cDNA cloning revealed that VAMP-2 and cellubrevin, but not VAMP-1, are expressed in rat pancreatic islets and that their sequence is identical to that isolated from rat brain. Pancreatic beta-cells contain secretory granules that store and secrete insulin as well as synaptic-like microvesicles carrying gamma-aminobutyric acid. After subcellular fractionation on continuous sucrose gradients, VAMP-2 and cellubrevin were found to be associated with both types of secretory vesicle. The association of VAMP-2 with insulin-containing granules was confirmed by confocal microscopy of primary cultures of rat pancreatic beta-cells. Pretreatment of streptolysin-O permeabilized insulin-secreting cells with tetanus and botulinum B neurotoxins selectively cleaved VAMP-2 and cellubrevin and abolished Ca(2+)-induced insulin release (IC50 approximately 15 nM). By contrast, the pretreatment with tetanus and botulinum B neurotoxins did not prevent GTP gamma S-stimulated insulin secretion. Taken together, our results show that pancreatic beta-cells express VAMP-2 and cellubrevin and that one or both of these proteins selectively control Ca(2+)-mediated insulin secretion.

VP22 light controlled delivery of oligonucleotides to ocular cells in vitro and in vivo.

PURPOSE: To study VP22 light controlled delivery of antisense oligonucleotide (ODN) to ocular cells in vitro and in vivo. METHODS: The C-terminal half of VP22 was expressed in Escherichia coli, purified and mixed with 20 mer phosphorothioate oligonucleotides (ODNs) to form light sensitive complex particles (vectosomes). Uptake of vectosomes and light induced redistribution of ODNs in human choroid melanoma cells (OCM-1) and in human retinal pigment epithelial cells (ARPE-19) were studied by confocal and electron microscopy. The effect of vectosomes formed with an antisense ODN corresponding to the 3'-untranslated region of the human c-raf kinase gene on the viability and the proliferation of OCM-1 cells was assessed before and after illumination. Cells incubated with vectosomes formed with a mismatched ODN, a free antisense ODN or a free mismatched ODN served as controls. White light transscleral illumination was carried out 24 h after the intravitreal injection of vectosomes in rat eyes. The distribution of fluorescent vectosomes and free fluorescent ODN was evaluated on cryosections by fluorescence microscopy before, and 1 h after illumination. RESULTS: Overnight incubation of human OCM-1 and ARPE-19 cells with vectosomes lead to intracellular internalization of the vectosomes. When not illuminated, internalized vectosomes remained stable within the cell cytoplasm. Disruption of vectosomes and release of the complexed ODN was induced by illumination of the cultures with a cold white light or a laser beam. In vitro, up to 60% inhibition of OCM-1 cell proliferation was observed in illuminated cultures incubated with vectosomes formed with antisense c-raf ODN. No inhibitory effect on the OCM-1 cell proliferation was observed in the absence of illumination or when the cells are incubated with a free antisense c-raf ODN and illuminated. In vivo, 24 h after intravitreal injection, vectosomes were observed within the various retinal layers accumulating in the cytoplasm of RPE cells. Transscleral illumination of the injected eyes with a cold white light induced disruption of the vectosomes and a preferential localization of the "released" ODNs within the cell nuclei of the ganglion cell layer, the inner nuclear layer and the RPE cells. CONCLUSIONS: In vitro, VP22 light controlled delivery of ODNs to ocular cells nuclei was feasible using white light or laser illumination. In vivo, a single intravitreal injection of vectosomes, followed by transscleral illumination allowed for the delivery of free ODNs to retinal and RPE cells.

Cell-permeable peptides induce dose- and length-dependent cytotoxic effects.

We have explored the threshold of tolerance of three unrelated cell types to treatments with potential cytoprotective peptides bound to Tat(48-57) and Antp(43-58) cell-permeable peptide carriers. Both Tat(48-57) and Antp(43-58) are well known for their good efficacy at crossing membranes of different cell types, their overall low toxicity, and their absence of leakage once internalised. Here, we show that concentrations of up to 100 microM of Tat(48-57) were essentially harmless in all cells tested, whereas Antp(43-58) was significantly more toxic. Moreover, all peptides bound to Tat(48-57) and Antp(43-58) triggered significant and length-dependent cytotoxicity when used at concentrations above 10 microM in all but one cell types (208F rat fibroblasts), irrespective of the sequence of the cargo. Absence of cytotoxicity in 208F fibroblasts correlated with poor intracellular peptide uptake, as monitored by confocal laser scanning fluorescence microscopy. Our data further suggest that the onset of cytotoxicity correlates with the activation of two intracellular stress signalling pathways, namely those involving JNK, and to a lesser extent p38 mitogen-activated protein kinases. These responses are of particular concern for cells that are especially sensitive to the activation of stress kinases. Collectively, these results indicate that in order to avoid unwanted and unspecific cytotoxicity, effector molecules bound to Tat(48-57) should be designed with the shortest possible sequence and the highest possible affinity for their binding partners or targets, so that concentrations below 10 microM can be successfully applied to cells without harm. Considering that cytotoxicity associated to Tat(48-57)- and Antp(43-58) bound peptide conjugates was not restricted to a particular type of cells, our data provide a general framework for the design of cell-penetrating peptides that may apply to broader uses of intracellular peptide and drug delivery.