Amyloid precursor protein (APP) belongs to a conserved gene family, also including the amyloid precursor-like proteins, APLP1 and APLP2. The function of these three proteins is not yet fully understood. One of the proposed roles of APP is to promote neurite outgrowth. The aim of this study was to investigate the regulation of the expression levels of APP family members during neurite outgrowth. We observed that retinoic acid (RA)-induced neuronal differentiation of human SH-SY5Y cells resulted in increased expression of APP, APLP1 and APLP2. We also examined the effect of the NFκB, AP-1 and c-Jun N-terminal kinase inhibitor curcumin (diferuloylmethane) on the RA-induced expression levels of these proteins. We found that treatment with curcumin counteracted the RA-induced mRNA expression of all APP family members. In addition, we observed that curcumin treatment resulted in neurite retraction without any effect on cell viability. Surprisingly, curcumin had differential effects on the APLP protein levels in RA-differentiated cells. RA-induced APLP1 protein expression was blocked by curcumin, while the APLP2 protein levels were further increased. APP protein levels were not affected by curcumin treatment. We propose that the sustained levels of APP and the elevated levels of APLP2, in spite of the reduced mRNA expression, are due to altered proteolytic processing of these proteins. Furthermore, our results suggest that APLP1 does not undergo the same type of regulated processing as APP and APLP2.

The amyloid precursor protein (APP) and its proteolytic cleavage products, the amyloid P peptides, have been implicated as a cause of Alzheimer's disease. Peptide nucleic acids (PNA), the DNA mimics, have been shown to block the expression of specific proteins at both transcriptional and translational levels. Generally, the cellular uptake of PNA is low. However, recent studies have indicated that the effect of unmodified antisense PNA uptake is more pronounced in nervous tissue. In this study we have shown that biotinylated PNA directed to the initiator codon region of the APP mRNA (-4 - +11) was taken up into the cytoplasm of primary rat cerebellar granule cells and cortical astrocytes, using fluorescence and confocal microscopy studies. Uptake of PNA was faster in neurons than in astrocytes. Western blotting analysis showed that APP was strongly down-regulated in both neurons and astrocytes. Thus, unmodified PNA can be used for studies on the function of APP in neurons and astrocytes.

Cellular machinery is protected from the surrounding by two-layer lipid membrane that is impermeable for most substances unnecessary for cellular metabolism. Unfortunately, from a cellular point of view, most new generation drugs, designed to act on gene regulation and transcription, are also considered to be unnecessary for metabolism and therefore showing poor, if any, intracellular localization. To overcome this obstacle, several chemical and physical methods have been developed, improving the uptake, but, on the other hand, also showing some unwanted side effects or limitations for in vivo applications. This dictates the continuing need for improved drug delivery and one way seems to be the relatively new class of compounds – cell-penetrating peptides (CPPs). Discovered approximately a decade ago, the content of this class is growing rapidly, containing now more than 100 compounds, which shows the intensity of work in this field. CPPs have already been shown to translocate cellular membranes in an unknown, seemingly receptor-independent and non-endocytotic manner. Moreover, they are able to deliver cargoes exceeding their own size up to 100-fold into a cellular milieu both in vitro and in vivo. The variety of different cargoes includes, but is not limited to: DNA, antisense PNA, oligonucleotides and small proteins. Recent data argues though that endocytosis is involved and contributes in some cases to the main part of the translocation. This review summarizes data on mechanisms of cell-penetrating peptides.

One approach to investigate gene function, by silencing the activity of certain proteins, is the usage of double stranded decoy oligodeoxynucleotides (ds decoy ODNs). Decoy, in this sense, is ds ODNs bearing the consensus binding sequence for a DNA-binding protein. This can be used in clinical settings to attenuate the effect of overexpressed transcription factors in tumor cells. We here choose to target the oncogenic protein Myc. Since oligonucleotides are poorly internalized to cells, a cell-penetrating peptide, TP10, was coupled to the Myc decoy, using two different strategies. Either TP10 was simply mixed with ds decoy ODNs forming complexes through non-covalent electrostatic interactions, or by having a nona-nucleotide overhang in one of the decoy strands, and adding a complementary PNA sequence coupled to an NLS sequence and TP10, which could hybridize to the Myc decoy.

By using these strategies, uptake was significantly enhanced, especially with the co-incubation approach. Interestingly, various endocytosis inhibitors had no effect on the uptake pattern, suggesting that uptake of these complexes is not mediated via endocytosis. Finally, a decreased proliferative capacity was observed when treating the neuroblastoma cell line N2a with TP10–PNA conjugate hybridized to Myc decoy compared to naked Myc decoy and untreated cells. A dose-dependent decrease in proliferation was also observed in MCF-7 cells, when using both strategies. These results suggest an alternative way to efficiently deliver ds ODNs into cells using the cell-penetrating peptide TP10 and prevent tumor growth by targeting the oncogenic protein Myc.

Inflammatory processes, mediated by the pro-inflammatory cytokines interleukin (IL) -1a/b, tumor necrosis factora(TNFa), IL-6 and excitotoxic actions of glutamate, are proposed to be implicated in neurodegenerative diseases, such as Alzheimer's disease.

A neurotoxic fragment ofb-amyloid,bA25-35, was shown to induce a reactive phenotype of primary astrocytes and microglial cells in a time-dependent manner. The morphological changes were found to coincide with a transient expression of IL-1band a sustained increased IL-1aand IL-6 expression as determined by reverse transcription and polymerase chain reaction analysis of RNA extracted from treated and untreated cells.

bA25-35 induced cytokine expression in astroglial cells from IL-1 receptor type I (IL-1RI)-deficient cultures, showed a hypersensitive IL-1aresponse and a decreased IL-6 response. Thus it was concluded that the IL-1RI is necessary for full scale induction of IL-6. Expression of IL-6 and TNFain the absence of a signalling IL-1 receptor may be induced by reactive oxygen species and/or by activation of NF_kB. The IL-1aincrease may reflect a missing transcriptional feed-back regulation involving the IL-1RI.

The role of IL-1 receptor accessory protein (IL-1RAcP) in IL-1binduced signalling was investigated in astroglial cultures from IL-1RAcP deficient mice. IL-1bhad no effect on nuclear NF_kB binding activity suggesting that IL-1RAcP is necessary for NF_kB activation. By use of specific antibodies it was shown that the activated NF_kB complex consisted of p50 and possibly p52 or RelB, NF_kB related proteins, respectively.

The excitatory amino acid glutamate was shown to stimulate respiratory activity in primary astroglial cell cultures. This effect was found to be correlated to the Na⁺-dependent high affinity glutamate uptake and an increased Na⁺/K⁺-ATPase activity since it could be blocked by ouabain and mimiced by gramicidin.

Carbon tetrachloride (CCl₄) and n-hexane, two neurotoxic organic solvents, were shown to interfere with cAMP formation in primary rat astroglial cells. In addition, both solvents decreased basal and glutamate induced increase in respiratory activity an observation which may reflect both a decreased Na⁺/K⁺-ATPase activity and a reduction of high-affinity glutamate uptake.

The use of powdered metallic zinc in acidic solution for the reduction of disulfide bonds in peptides and proteins has been investigated. The method has several advantages over the traditional mereapto based reducing methods currently used; the reducing agent is readily available and inexpensive; reduction can be performed in weakly acidic solutions of water and/or acetonitrile; work up simply consists of a centrifugation step followed by pipeting the supernatant from the metal pellet, thereby greatly diminishing the risk of reoxidation as a more elaborate work up procedure could result in. As no mercapto compounds are added, there is no risk that the reducing agent will interfere in subsequent modification of the thiol functionality. Disulfides in a model peptide are reduced within 5 min in any mixture of water/acetonitrile containing 1% TFA, all disulfides in insulin is reduced within I h in any mixture of water/acetonitrile containing 5% acetic acid. To stress the convenience of the metallic zinc reduction method, the resulting thiol compound was subjected to two commonly employed reactions in peptide chemistry: Cys(Npys) directed disulfide formation (70% yield) and native chemical ligation between the reduced model peptide and Boc-Ala-p-metylthiobenzyl ester (65% yield of the ligation product plus disulfide formation between Cys and p-thiocresol).

The activation of nuclear factor B (NFB) is a key event in immune and inflammatory responses. In this study, a cell-penetrating transport peptide, transportan (TP) or its shorter analogue TP 10, was used to facilitate the cellular uptake of an NFB decoy. Peptide nucleic acid (PNA) hexamer or nonamer was linked to the transport peptide by a disulfide bond. NFB decoy oligonucleotide consisted of a double-stranded consensus sequence corresponding to the B site localized in the IL-6 gene promoter, 5'-GGGACTTTCCC-3', with a single-stranded protruding 3'-terminal sequence complementary to the PNA sequence was hybridized to the transport peptide–PNA construct. The ability of the transport peptide–PNA–NFB decoy complex to block the effect of interleukin (IL)-1-induced NFB activation and IL-6 gene expression was analyzed by electrophoretic mobility shift assay and reverse transcriptase-polymerase chain reaction in rat Rinm5F insulinoma cells. Preincubation with transport peptide–PNA–NFB decoy (1 M, 1 h) blocked IL-1-induced NFB-binding activity and significantly reduced the IL-6 mRNA expression. The same concentration of NFB decoy in the absence of transport peptide–PNA had no effect even after longer incubations. Our results showed that binding of the oligonucleotide NFB decoy to the nonamer PNA sequence resulted in a stable complex that was efficiently translocated across the plasma membrane.

The uptake of different cell-penetrating peptides (CPPs) in two yeast species, Saccharomyces cerevisiae and Candida albicans, was studied using fluorescence HPLC-analyses of cell content. Comparison of the ability of penetratin, pVEC and (KFF)(3)K to traverse the yeast cell envelope shows that the cellular uptake of the peptides varies widely. Moreover, the intracellular degradation of the CPPs studied varies from complete stability to complete degradation. We show that intracellular degradation into membrane impermeable products can significantly contribute to the fluorescence signal. pVEC displayed highest internalizing capacity, and considering its stability in both yeast species, it is an attractive candidate for further studies.

This thesis is devoted to the deeper understanding of the molecular details of the interaction between the neuropeptide galanin and its target protein, the G-protein coupled 7TM galanin receptor. The enzymatic control of galanin degradation and second messenger systems triggered by galanin receptor activation are also studied.

Galanin analogs with substitutions of amino acids by units of L-Ala residues have been developed as selective ligands to galanin receptor subtypes in the rat hypothalamus and in the rat jejunum. It is shown that galanin receptors in the jejunum require also the C-terminal segment KHGLT^25-29for receptor recognition in contrast to receptors in the hypothalamus which only interact with the N-terminal portion (1-13) of galanin.

It is confirmed that the N-terminal portion (1-13) of galanin is essential for receptor recognition in the rat hypothalamus and that the C-terminus of the ligand can be extended with sequences of amino acids or other organic structures to increase the affinity of the N-terminal (1-13) fragment of galanin. It is also concluded that a minimum motif required for galanin receptor recognition should contain the indol structure of a Trp residue and the aromatic side chain of a Tyr residue plus one additional bulky highly hydrophobic structure as well as some positively charged structures.

The galanin binding domains of the receptor, GalR1 have been studied by site directed mutagenesis and a structural model for the interaction of the major pharmacophores of galanin is proposed. The charged N-terminus of galanin is suggested to interact with Phe¹¹⁵ in TM III by cation-p interaction, the Trp² residue of galanin is proposed to interact with a pair of His residues located at the top of TM VI and the Tyr9 of galanin is interacting with another phenylalanine, Phe²⁸² in the EC III of the receptor. One of the two histidine residues (264,267); His²⁶⁷is believed not only to be important for ligand binding but also is stabilising the active conformation of the GalR1 receptor, important for the G-protein coupling of the agonist occupied receptor.

It is shown that galanin binding sites are present at significantly higher numbers in the ventral part of the rat hippocampus compared to the dorsal area. Furthermore it is determined that the efficiency of the inhibitory coupling to basal and forskolin stimulated cAMP production by galanin receptors is 250 fold less potent in the dorsal region of the rat hippocampus. The differences in regulation of adenylate cyclase suggests an unequal receptor subtype distribution in the dorsal and ventral hippocampus with higher levels of G_icoupled GalR1 receptors in the ventral segments of the hippocampal formation.

An internally quenched fluorescent substrate for galanin degrading enzymes has been developed and used for purification of a galanin inactivating 70 kDa membrane associated metallo-endopeptidase from bovine spinal cord. The enzyme that has a inhibitor profile different from NEP-24.11 primarily cleaves galanin between Trp² and Thr³ and recognises both galanin and the galanin mimicking fluorescent artificial substrate with high affinity. This cleavage generates products that are inactive at both GalR1 and GalR2 receptors.

Solid-phase synthesis of small non-polymeric molecules has become increasingly important as a tool in the development of pharmacologically active compounds. Of particular interest are the syntheses of compounds with low molecular weight, low polarity and a reduced flexibility of the functional groups, i.e. physiochemical properties typical for the majority of drugs currently in use.

The first part of this thesis describes the synthesis of polycyclic structures with constrained conformation, by the application of multi-component condensations on solid-phase. The 8-azabicyclo[3.2.1]octan-3-one structure of the tropane alkaloids was synthesized on solid-phase by a modification of the well known Robinson tropinone synthesis. The 3-component reaction was performed with the amino component anchored to a solid support, consisting of polyethyleneglycol-grafted polystyrene. Treatment of the primary amine with 1,3-acetonedicarboxylic acid and excess of succinaldehyde, resulted in high purity of the corresponding tropane derivative. Further derivatization of the resin-bound tropane derivative was performed by reduction of the keto-group and acylation of the hydroxyl-group.

The oxygen-bridged tetrahydropyridones are relatively complex polycyclic structures, which previously have been synthesized in solution by condensation of primary amines, coumarin-3-carboxylic acid and ketones. In the evaluation of a solid-phase approach of this 3-component condensation, several members of this class of compounds were synthesized with the amine anchored to a solid support. The yield and purity of the products were dependent on the ketones used in the reaction, but the expected products were obtained using both acyclic and cyclic ketones and substituted acetophenones.

The second part of this thesis describes the development of a solid-phase polyamine synthesis. The polyamines are a class of compounds with a wide range of pharmacological and physiological effects, which are constituents in many types of venoms of wasps and spiders. Synthesis of polyamines in homogenous solution is accompanied with several problems concerning selectivity, work-up and yield. By employing a solid-phase approach, a rapid and convenient method for synthesis of polyamines is achieved. In the developed protocol, acid labile benzhydryls are used as amino-protecting groups. The polyamine backbone is assembled sequentially by reductive alkylation of the protected secondary amine with Fmoc-amino aldehydes. The protecting groups allow the use of excess of aldehyde, to avoid underivatized amines, without the risking dialkaylation of the amines, which occurs during reductive alkylations of primary amines with unhindered aliphatic aldehydes. The versatility of the method is displayed by a convenient synthesis of four analogues of a wasp toxin, philanthotoxin.

A solid-phase approach for the synthesis of oxygen-bridged tetrahydropyridones has been developed. A diamine is attached to Trt-Cl resin and condensed with different aliphatic or aromatic ketones and coumarin-3-carboxylic acid for 20 h and cleaved with 5% TFA in DCM, resulting in tri or tetracyclic products in moderate to high yield.

During apoptosis (also called programmed cell death), the chromatin condenses and the DNA is cleaved into oligonucleosomal fragments. Caspases are believed to play a major role in nuclear apoptosis. However, the relation between dismantling of nuclear pores, disruption of the nucleocytoplasmic barrier, and nuclear entry of caspases is unclear. We have analyzed nuclear import of the green fluorescent protein fused to a nuclear localization signal (GFP-NLS) in tissue culture cells undergoing apoptosis. Decreased nuclear accumulation of GFP-NLS could be detected at the onset of nuclear apoptosis manifested as dramatic condensation and redistribution of chromatin toward the nuclear periphery. At this step, dismantling of nuclear pores was already evident as indicated by proteolysis of the nuclear pore membrane protein POM121. Thus, disruption of nuclear compartmentalization correlated with early signs of nuclear pore damage. Both these events clearly preceded massive DNA fragmentation, detected by TUNEL assay. Furthermore, we show that in apoptotic cells, POM121 is specifically cleaved at aspartate-531 in its large C-terminal portion by a caspase-3-dependent mechanism. Cleavage of the C-terminal portion of POM121, which is adjoining the nuclear pore complex, is likely to disrupt interactions with other nuclear pore proteins affecting the stability of the pore complex. A temporal correlation of apoptotic events supports a model where caspase-dependent disassembly of nuclear pores and disruption of the nucleocytoplasmic barrier paves the way for nuclear entry of caspases and subsequent activation of CAD-mediated DNA fragmentation.

Cell-penetrating peptides (CPPs) are novel high-capacity delivery vectors for different bioactive cargoes. We have evaluated the CPP transportan 10 (TP10) as a delivery vector in different in vitro plasmid delivery assays. Tested methods include: TP10 crosslinked to a plasmid via a peptide nucleic acid (PNA) oligomer, TP10 conjugation with polyethyleneimine (PEI), and addition of unconjugated TP10 to standard PEI transfection assay. We found that without additional DNA condensing agents, TP10 has poor transfection abilities. However, the presence of TP10 increases the transfection efficiency several folds compared to PEI alone. At as low concentrations as 0.6 nM, TP10–PNA constructs were found to enhance plasmid delivery up to 3.7-fold in Neuro-2a cells. Interestingly, the transfection efficiency was most significant at low PEI concentrations, allowing reduced PEI concentration without loss of gene delivery. No increase in cytotoxicity due to TP10 was observed and the uptake mechanism was determined to be endocytosis, as previously reported for PEI mediated transfection. In conclusion, TP10 can enhance PEI mediated transfection at relatively low concentrations and may help to develop future gene delivery systems with reduced toxicity.

In this work, we have targeted positions 18–38 of the human galanin receptor type 1 (GalR1) mRNA coding sequence with different peptide nucleic acid (PNA) oligomers. This region has previously been shown to be a good antisense region and therefore we aimed to identify the subregions and/or thermodynamic parameters determining the antisense efficacy. Nine different PNA oligomers were conjugated to a cell-penetrating peptide, transportan, to enhance their cellular uptake. Concentration-dependent down-regulation of GalR1 protein expression in human melanoma cell line Bowes was measured by radioligand binding assay. No reduction of GalR1 mRNA level was observed upon PNA treatment, thus, the effect was concluded to be translational arrest. Judging from the EC₅₀ values, antisense PNA oligomers targeting regions 24–38 (EC₅₀ = 70 nM) or 27–38 (EC₅₀ = 80 nM) were the most potent suppressors of protein expression. No parameter predicted by M-fold algorithm was found to correlate with the measured antisense activities. Presence of some subregions was found not to increase antisense efficiency of PNA. Presence of a short unpaired triplet between nucleotides 33 and 35 in the target region was, on the other hand, found to be the most critical for efficient GalR1 down-regulation. Thus, the results are of high impact in designing antisense oligomers. Specific results of this study demonstrate 20-fold more efficient antisense down-regulation of GalR1 as achieved before.

Cellular translocation into a human Bowes melanoma cell line was investigated and compared for penetratin and pIsl, two peptides that correspond to the third helices of the related homeodomains, from the Antennapedia transcription factor of Drosophila and the rat insulin-1 gene enhancer protein, respectively. Both biotinylated peptides internalized into the cells with similar efficacy, yielding an analogous intracellular distribution. When a large cargo protein, 63 kDa avidin, was coupled to either peptide, efficient cellular uptake for both the peptide−protein complexes was observed. The interactions between each peptide and SDS micelles were studied by fluorescence spectroscopy and acrylamide quenching of the intrinsic tryptophan (Trp) fluorescence. Both peptides interacted strongly and almost identically with the membrane mimicking environment. Compared to penetratin, the new transport peptide pIsl has only one Trp residue, which simplifies the interpretation of the fluorescence spectra and in addition has a native Cys residue, which may be used for alternative coupling reactions of cargoes of different character.

Studies on the hemoproteins of the hagfish Myxine glutinosa L. (Cyclostomata) are of great interest with respect to the evolution of vertebrate hemoglobins.

Hagfish hemoglobin is heterogeneous. Three main components, Hb I, Hb II and Hb III, as well as ten minor ones, have been isolated by isoelectric focusing. All main components are monomers, their molecular weights decreasing from Hb I to Hb III. Two of them, Hb I and Hb III, have been crystallized.

The amino acid composition of the main components have been determined. They differ considerably from each other and from the hemoglobins of the lampreys, members of another subclass of Cyclostomata. The N-terminal residue of Hb III is proline, whereas those of Hb I and Hb II are blocked.

The amino acid sequence of Hb III from Myxine has been elucidated. The globin was subjected to tryptic digestion, as well as to cleavage with cyanogen bromide and BNPS-skatol. The sequences of the separated peptides were determined by automated Edman degradation. Globin Hb III consists of 148 amino acid residues and contains no cysteine, in contrast to other vertebrate hemoglobins.

A novel heme linkage has been demonstrated, which has only been found in opossum hemoglobin and globin from the invertebrate Chironomus. The regularly occurring distal histidine E7 is substituted by glutamine, and valine E11 by isoleucine. It has been proposed that Gin E7 is directed towards the outside of the heme pocket and Ile E11 towards the inside, replacing histidine in the distal heme contact.

Globin Hb III has an additional segment of nine amino acid residues at the N-terminus in contrast to mammalian hemoglobins. This N-terminal nona- peptide is also present in globins from other Cyclostomata.

The oxygen binding characteristics of Myxine hemoglobin have been studied. The main components show no cooperative oxygen binding and do not exhibit any Bohr effect, having however, different oxygen affinities. Although all hemoglobins of Cyclostomata are monomers, their oxygen binding characteristics vary. ESR studies of the nitrosyl complex of hemoglobin Hb III have shown that the critical substitutions in the heme linkage at E7 and E11 do not disrupt the proximal histidine-iron bond, even though substitutions of the same type are known to cause destabilization of the R quaternary structure in tetrameric mammalian nitrosylhemoglobins.

The primary structure of Hb III differs by more than 50% from that of hitherto analysed lamprey hemoglobins. More than 400 million years were required for the development of this difference in the course of evolution. Nevertheless, a comparison of the hemoglobin sequences from the two branches of Agnatha, hagfish and lamprey, suggests a monophyletic origin, distinct from that of Gnathostomata. The primary structure of Myxine hemoglobin is compatible with the theory that this hemoglobin is a link between invertebrate and vertebrate globins, and that the divergence of hemoglobins occurred more than 500 million years ago.

Cell-penetrating peptides have the remarkable ability to carry hydrophilic macromolecules over the cellular membrane in an energy and protein independent manner. To synthesise and characterise new variants of cell-penetrating peptides, mainly aimed at finding analogues with minimised side effects and efficient cellular delivery, was one of the two main objectives of this thesis. The other principal objective was to study cell-junction interactions by the protein vascular endothelial cadherin. In addition, a study where the two objectives were combined was aimed at investigating the possibility of utilising cell-penetrating peptides for trans-barrier delivery.

In order to gain more information about the cell-junction interactions of vascular endothelial cadherin, monoclonal antibodies directed to the extracellular domain were tested for their activity. Three of the antibodies were able to increase paracellular permeability, inhibit VE-cadherin reorganisation, and block angiogenesis in vitro. In addition, epitope mapping of the antibodies located their binding sequences to EC1, EC3 and EC4 of VE-cadherin ectodomain. These results support the concept that VE-cadherin protein sequence consists of multiple biologically important domains.

The aim of paper II was to compare transportan, developed by our group, with the first discovered cell-penetrating peptide, penetratin. Novel analogues were synthesised and tested for cellular uptake and also in molecular modelling. The transportan analogues were different chimeric constructs, elaborating with the two different parts of the peptide, while the penetratin analogues were modifications of the original sequence. Furthermore, the translocation ability of fluorescently labelled parent peptides, transportan and penetratin, were quantitatively measured. The results imply that the two peptides and their analogues do not enter the cells by the same mechanism.

Since it was clear that the mastoparan part of transportan was necessary for cellular penetration while the galanin part was more adaptable, several shorter analogues were synthesised in order to define the essential sequence for transportan penetration. In order to reduce the unwanted side effects, the biological effects of transportan and its deletion analogues were studied by GTPase activity measurements. It was concluded that the deletion of six amino acids from the N-terminus did not significantly impair the cell penetration, while truncation of the C-terminus or in the middle of the peptide decreased or even abolished the cellular uptake of transportan.

pVEC is a new type of cell-penetrating peptides derived from the murine sequence of VE-cadherin. In this study we show that pVEC is a non-toxic, efficient cell-penetrating peptide that can be used for cellular delivery of both PNA and large proteins, such as streptavidin. The uptake of pVEC was quantified, by direct fluorescence labelling in both mouse and human endothelial cell lines. Furthermore, no significant effect could be detected on the parent protein VE-cadherins cell-junction clustering.

In paper V it is shown that transportan and the analogue, transportan 10, enter and pass across a human colon cancer Caco-2 epithelial cell layer. However, the peptides decreased the trans-epithelial electric resistance of the barrier model, but not the dextran passage to the same extent. Taken together, these data demonstrate that transportan and transportan 10 are conveyed over the epithelial cell layer, mainly by the transcellular pathway, but at higher peptide concentration, the paracellular pathway may contribute to the passage. To conclude, cell-penetrating peptides may be a new possible strategy for drug delivery over a tight junction barrier, such as the blood-brain barrier. 

The N-terminal (1-28) part of the mouse prion protein (PrP) is a cell penetrating peptide, capable of transporting large hydrophilic cargoes through a cell membrane. Confocal fluorescence microscopy shows that it transports the protein avidin (67 kDa) into several cell lines. The (1-28) peptide has a strong tendency for aggregation and P-structure formation, particularly in interaction with negatively charged phospholipid membranes. The findings have implications for how prion proteins with uncleaved signal peptides in the N-termini may enter into cells, which is important for infection. The secondary structure conversion into beta-structure may be relevant as a seed for the conversion into the scrapie (PrPSc) form of the protein and its arnyloidic transformation.

IL-1 is a proinflammatory cytokine which causes systemic responses such as fever, increased neuroendocrine activity, sympathetic outflow and increased immune responses. Activation of the transcription factor NFkB is one of the molecular mechanisms involved in IL-1 signalling.

We have shown that the IL-1binduced fever response is regulated the hypothalamic - pituitary - adrenal axis (HPA-axis), as either the nonpeptidic CRF receptor antagonist CP154,526, or a subchronic glucocorticoid treatment, respectively, blocked the ability of peripheral IL-1bto induce fever in rats. The subchronic glucocorticoid treatment caused a dampened IL-1aand IL-1bmRNA expression in the hypothalamus whereas IL-6 mRNA was induced by the same glucocorticoid treatment. Both the glucocorticoid altered cytokine mRNA levels and IL-1 induced fever responses, respectively, were reversed to normal upon the removal of exogenous glucocorticoids. Both IL-1 receptor antagonist (IL-1ra) mRNA and protein were detectable in adrenal chromaffin cells. Adrenal IL-1ra mRNA levels were furthermore rapidly induced by a peripheral lipopolysaccharide (LPS) challenge.

Type II IL-1R (IL-1RII) is a negative regulator of IL-1 bioactivities by being a nonsignaling, IL-1 binding protein. Here we present evidence that IL-1RII also can interact with IL-1R accessory protein (IL-1RAcP), thus extending the previously known IL-1 suppressing activity of IL-1RII to encompass the sequestering of IL-1RAcP from the signaling IL-1RI complex. IL-1RAcP was shown to be a necessary component of the murine (m) IL-1RI complex in transducing the IL-1 signal to NFkB activation in murine fibroblast C127 cells and in primary mouse astrocyte cultures. In addition, heterologous murine IL-1RAcP and human IL-1RI complexes formed functional IL-1bsignaling complexes.

Transgenic mice deficient in IL-RAcP did not respond with fever to either a peripheral injection of IL-1aor to IL-1b, thus implicating IL-1RAcP as a necessary component of the IL-1RI complex in vivo.

To examine the role of central IL-1ra in suppressing IL-1signaling, the cDNA encoding the human secretable form of IL-1ra (hsIL-1ra) was expressed under the control of the glia fibrillary acidic protein (GFAP) promoter in transgenic mice. The expression of hsIL-1ra mRNA and protein were exclusively restricted to CNS, as measured by RT-PCR and ELISA, respectively. The GFAP-hsIL-1ra (GILRA) mice did not exhibit fever upon a central injection of IL-1b, while a peripheral LPS injection caused a hypersensitive fever response in these mice. These mice strains provide a model for studies on IL-1R occupancy and systemic responses.

Galanin(2–11) has been introduced as a receptor subtype selective ligand for the GalR2 subtype of the galanin receptors, and has gained use in pharmacological studies of galaninergic signaling in the past two years. By introducing l-Ala substitutions in the galanin(2–11) sequence, we have examined the amino acid residues which are of importance for binding to the GalR2 receptor. Our study shows that Trp², Asn⁵, Gly⁸ and Tyr⁹ are of great importance for high affinity binding. When placed in an α-helical conformation, the side chains of these residues are, with the exception of Tyr⁹, displayed on the same “side” of the peptide. This information is useful in the rational design of non-peptide type GalR2 receptor ligands.

The chimeric peptide M617, galanin(1–13)-Gln¹⁴-bradykinin(2–9)amide, is a novel galanin receptor ligand with increased subtype specificity for GalR1 and agonistic activity in cultured cells as well as in vivo. Displacement studies on cell membranes expressing hGalR1 or hGalR2 show the presence of a high affinity binding site for M617 on GalR1 (K _i=0.23±.12 nM) while lower affinity was seen towards GalR2 (K _i=5.71±1.28 nM) resulting in 25-fold specificity for GalR1. Activation of GalR1 upon stimulation with M617 is further confirmed by internalization of a GalR1-EGFP conjugate. Intracellular signaling studies show the ability of M617 to inhibit forskolin stimulated cAMP formation with 57% and to produce a 5-fold increase in inositol phosphate (IP) accumulation. Agonistic effects on signal transduction are shown on both receptors studied after treatment with M617 in the presence of galanin. In noradrenergic locus coeruleus neurons, M617 induces an outward current even in the presence of TTX plus Ca²⁺, high Mg²⁺, suggesting a postsynaptic effect. Intracerebroventricular (i.c.v.) administration of M617 dose-dependently stimulates food uptake in rats while, in contrast, M35 completely fails to affect the feeding behavior. Spinal cord flexor reflex is facilitated by intrathecal (i.t.) administration of M617 as well as galanin with no significant change upon pre-treatment with M617. M617 dose dependently antagonizes the spinal cord hyperexcitablility induced by C-fiber conditioning stimulus and does neither enhance nor antagonize the effect of galanin. These data demonstrate a novel galanin receptor ligand with subtype specificity for GalR1 and agonistic activity, both in vitro and in vivo.

Certain short peptides, which are able to translocate across cell membranes with a low lytic activity, can be useful as carriers (vectors) for hydrophilic molecules. We have studied three such cell penetrating peptides: pAntp (‘penetratin’), pIsl and transportan. pAntp and pIsl originate from the third helix of homeodomain proteins (Antennapedia and Isl-1, respectively). Transportan is a synthetic chimera (galanin and mastoparan). The peptides in the presence of various phospholipid vesicles (neutral and charged) and SDS micelles have been characterized by spectroscopic methods (fluorescence, EPR and CD). The dynamics of pAntp were monitored using an N-terminal spin label. In aqueous solution, the CD spectra of the three peptides show secondary structures dominated by random coil. With phospholipid vesicles, neutral as well as negatively charged, transportan gives up to 60% α-helix. pAntp and pIsl bind significantly only to negatively charged vesicles with an induction of around 60% β-sheet-like secondary structure. With all three peptides, SDS micelles stabilize a high degree of α-helical structure. We conclude that the exact nature of any secondary structure induced by the membrane model systems is not directly correlated with the common transport property of these translocating peptides.

Interleukin-1 (IL-1) is an endogenous mediator of the inflammatory response. IL-1 executes its actions by binding and subsequent activation of the IL-1 receptor (IL-1R) proteins. Among the many systemic effects of IL-1, the fever response was the first to be investigated.

We characterised, by cDNA sequence analysis, the primary structure of the type II IL-1 receptor (IL-1RII) expressed in the rat insulinomab-cell line Rinm5F. In addition, we found that IL-1R agonists upregulated the mRNA concentrations of IL-1RI and IL-1RII in Rinm5F cells.

By generation of IL-1R subtype selective mutants of human IL-1bwe created pharmacological tools to investigate IL-1R subtype specific signalling in vivo. We showed that the type I IL-1R (IL-1RI) mediates the induction of fever, activation of the hypothalamic-pituitary-adrenal axis and induction of IL-6 expression by IL-1bin the rat.

We presented data on an interaction between the IL-1RII and the IL-1 receptor accessory protein (IL-1RAcP) in the presence of IL-1b, suggesting a novel mechanism of regulating the cellular responsiveness to IL-1bi.e., by a competition between IL-1RII and the signalling IL-1RI for interaction with IL-1RAcP.

We showed that mice deficient in IL-1bresponded with exacerbated fever when challenged with IL-1aorbor lipopolysaccharide (LPS) of E. coli, as compared to wild type mice. We found increased concentrations of IL-6 mRNA in the hypothalamus from untreated IL-1bdeficient mice as compared to wild type mice, suggesting the IL-1bdeficient mice to have compensated for the lack of IL-1bwith an increased expression of the functionally related pro-inflammatory cytokine IL-6.

Using mice deficient in IL-1RAcP we showed that IL-1RAcP is involved in IL-1R signalling, as measured by lack of IL-1binduced NFkB activation and IL-1aorbinduced fever. In the same manner as mice deficient in IL-1b, the IL-1RAcP deficient mice responded with a fever when challenged with LPS suggesting that IL-1 is not essential for fever per se.

Cells are protected from the surrounding environment by plasma membrane which is impenetrable for most hydrophilic molecules. In the last 10 years cell-penetrating peptides (CPPs) have been discovered and developed. CPPs enter mammalian cells and carry cargo molecules over the plasma membrane with a molecular weight several times their own. Known transformation methods for plant cells have relatively low efficiency and require improvement. The possibility to use CPPs as potential delivery vectors for internalisation in plant cells has been studied in the present work. We analyse and compare the uptake of the fluorescein-labeled CPPs, transportan, TP10, penetratin and pVEC in Bowes human melanoma cells and Nicotiana tabacum cultivar (cv.) SR-1 protoplasts (plant cells without cell wall). We study the internalisation efficiency of CPPs with fluorescence microscopy, spectrofluorometry and fluorescence-activated cell sorter (FACS). All methods indicate, for the first time, that these CPPs can internalise into N. tabacum cv. SR-1 protoplasts. Transportan has the highest uptake efficacy among the studied peptides, both in mammalian cells and plant protoplast. The internalisation of CPPs by plant protoplasts may open up a new effective method for transfection in plants.

Cell-penetrating peptides (CPPs) are carriers developed to improve mammalian cell uptake of important research tools such as antisense oligonucleotides and short interfering RNAs. However, the data on CPP uptake into non-mammalian cells are limited. We have studied the uptake and antimicrobial effects of the three representative peptides penetratin (derived from a non-mammalian protein), MAP (artificial peptide) and pVEC (derived from a mammalian protein) using fluorescence HPLC in four common model systems: insect cells (Sfg), gram-positive bacteria (Bacillus megaterium), gram-negative bacteria (Escherichia coli) and yeast (Saccharomyces cerevisiae). We demonstrate that non-mammalian cells internalize CPPs and a comparison of the uptake of the peptides show that the intracellular concentration and degradation of the peptides varies widely among organisms. In addition, these CPPs showed antimicrobial activity.

Proteins with molecular masses ranging from 30 kDa. (green fluorescent protein, GFP) to 150 kDa (monoclonal and polyclonal antibodies) were coupled to the cellular translocating peptide transportan. We studied the ability of the resulting protein-peptide constructs to penetrate into Bowes melanoma, BRL, and COS-7 cells. After 0.5-3 h incubation with recombinant GFP coupled to transportan, most of the GFP fluorescence was found in intracellular membranes of BRL and COS-7 cells, which suggests that transportan could internalize covalently linked proteins of about 30 kDa in a folded state. Transportan could internalize covalently coupled molecules of even larger size; that is, avidin and antibodies, (up to 150 kDa). The covalent bond between the transport peptide and its cargo is not obligatory because streptavidin was translocated into the cells within 15 min as a noncovalent complex with biotinylated transportan. Inside the cells, the delivered streptavidin was first located mainly in close proximity to the plasma membrane and was later distributed to the perinuclear region. Most of the internalized streptavidin was confined to vesicular structures, but a significant fraction of the protein was distributed in the cytoplasm. Our data suggest that transportan can deliver proteins and other hydrophilic macromolecules into intact mammalian cells, and this finding demonstrates good potential as powerful cellular delivery vector for scientific and therapeutic purposes.

The anchoring the first amino acid in Boc chemistry to a 4-(3-hydroxy-4-methylpentyl)phenylacetic acid linker is described and compared to the conventional Pam resin. The peptidyl-4-(4-methyl-3-pentoxy)phenylacetamide linkage is slightly more stable to TFA than the Pam linker but in contrast to the Pam linker stable to cleavage of benzylic protective groups with TFMSA/DMS/TFA mixtures. This allows a mild and convenient two step deprotection procedure using the “low TFMSA-high HF”. In HF this new linker reacts preferentially in an intramolecular reaction forming a tetrahydronaphthalene derivative.

Cyclic peptides with alternating d- and l-amino acid residues containing tert-leucine residues in every second position can form peptide nanotubes only when both enantiomers of the peptide are present in the solution. These results strongly indicate the formation of peptide nanotubes that assemble with one enantiomer in every second position, thereby forming a lamellar structure.

An increasing body of evidence suggests that β-amyloid (Aβ) and activated glial cells play a crucial part in the pathogenesis of Alzheimer's disease (AD). Activated glial cells surrounding the senile plaques, formed by Aβ peptides, have been proposed to promote neurodegeneration by producing putatively toxic factors, including the inflammatory cytokine interleukin-1β (IL-1β). Elevated levels of both IL-1β and activated nuclear factor κB (NF-κB), a key transcription factor regulating a wide variety of inflammatory genes, have been found in the brains of AD patients. In this study, we have investigated the ability of the Aβ(25-35) peptide and IL-1β, either alone or together, in activating NF-κB in glial cells. Mixed primary glial cells from rat were treated with IL-1β and/or Aβ(25-35), and NF-κB binding activity was analyzed by electophoretic mobility shift assay. We observed that the induction of NF-κB binding activity induced by either IL-1β or Aβ(25-35) showed a peak at 30 min, and significantly declined after 2 h. The induced NF-κB activation persisted after 24 h and even seemed to increase in cells treated with Aβ(25-35). The activation of NF-κB by Aβ(25-35) was shown to be dose-dependent. In addition, Aβ(25-35) potentiated the effect of IL-1β in a dose-dependent manner when co-stimulating the cells. The potentiating effect of Aβ(25-35) on IL-1β-induced NF-κB binding activity was observed after 30 min, 2 h and 24 h, and did not significantly differ over time. A possible explanation is that when glial cells are stimulated by inflammatory factors in the presence of Aβ peptides or senile plaques, the NF-κB negative feedback regulation is no longer functional.

Two novel metalloproteases from Arabidopsis thaliana, termed AtPrePI and AtPrePII, were recently identified and shown to degrade targeting peptides in mitochondria and chloroplasts using an ambiguous targeting peptide. AtPrePI and AtPrePII are classified as dually targeted proteins as they are targeted to both mitochondria and chloroplasts. Both proteases harbour an inverted metal binding motif and belong to the pitrilysin subfamily A. Here we have investigated the subsite specificity of AtPrePI and AtPrePII by studying their proteolytic activity against the mitochondrial F₁β pre-sequence, peptides derived from the F₁β pre-sequence as well as non-mitochondrial peptides and proteins. The degradation products were analysed, identified by MALDI-TOF spectrometry and superimposed on the 3D structure of the F₁β pre-sequence. AtPrePI and AtPrePII cleaved peptides that are in the range of 10 to 65 amino acid residues, whereas folded or longer unfolded peptides and small proteins were not degraded. Both proteases showed preference for basic amino acids in the P₁ position and small, uncharged amino acids or serine residues in the P^′₁P′1

position. Interestingly, both AtPrePI and AtPrePII cleaved almost exclusively towards the ends of the α-helical elements of the F₁β pre-sequence. However, AtPrePI showed a preference for the N-terminal amphiphilic α-helix and positively charged amino acid residues and degraded the F₁β pre-sequence into 10–16 amino acid fragments, whereas AtPrePII did not show any positional preference and degraded the F₁β pre-sequence into 10–23 amino acid fragments. In conclusion, despite the high sequence identity between AtPrePI and AtPrePII and similarities in cleavage specificities, cleavage site recognition differs for both proteases and is context and structure dependent.

Fever, a 1-4 °C elevation of the body temperature, is part of the systemic inflammatory response to infection or tissue damage, and it is believed to be a potent host defence. The underlying physiological mechanisms of the febrile response are one of the two main objectives in this thesis, where studies conducted in whole animals (rats and mice), chiefly aimed at clarifying the cytokine cascade in fever. The other principal objective of this thesis was to gain a general understanding on the signalling properties of neurons in the thermoregulatory region of the brain, the preoptic area, through electrophysiological investigations in vitro of ion channels and impulse firing behaviour of medial preoptic neurons of rat. From these studies it was concluded that:

The endogenous pyrogen tumour necrosis factor _ (TNF_) injected intraperitoneally causes fever independent of interleukin-1_ (IL-1_), but dependent on interleukin-6 (IL-6) in the central nervous system. The TNF__induced fever is triggered through release endoperoxides, since pre-administration of the cyclooxygenase inhibitor indomethacin, efficiently blocks the increased body temperature.

An intraperitoneal injection of lipopolysaccharide (LPS) induces fever that is independent on central IL-1 binding to its receptors, since occupancy of central IL-1 receptors of the IL-1 receptor antagonist (IL-1ra), is unable to block the febrile response.

The temperature in the central nervous system and in the peritoneum are similar during the febrile response with respect to onset, temporal characteristics and fever amplitude.

The neurons of the medial preoptic nucleus respond to glutamate application with currents that can be attributed to ion channels of the AMPA-receptor type as well as of the NMDA-receptor type. The functional characteristics of these channels comprise fast activation and desensitization of the AMPA-receptor channel, as well as glycine dependency, Mg²⁺dependent outward rectification and slow desensitization kinetics of the NMDA-receptor channel.

The medial preoptic neurons display two types of Ca²⁺ spikes, that result in two types of firing behaviour. First, low-threshold spikes depend on T-type Ca²⁺ channels and are generated from membrane potentials < -75 mV. They may induce short bursts of fast Na⁺ spikes. Second, high-threshold spikes can be generated from more depolarized levels, and depend on Ca²⁺channels that are mainly of the L, N and P -types. One role of these channels is to sustain long-burst firing.

Medial preoptic neurons spontaneously fire with several types of temporal firing patterns. Apart from burst firing, neurons are silent or discharge regularly as well as irregularly. The type of firing pattern can be manipulated with steady current injection and possibly also by PGE2.

Insults to the brain as well as neurodegenerative diseases are known to elicit inflammatory responses. Inflammation in the brain can on one hand initiate processes that are harmful to the injured tissue and exacerbate the damage, leading to neuronal degeneration and glial activation, and on the other hand activate processes that may be necessary for repair mechanisms and regeneration. Among the mediators of inflammatory response in the brain are the inflammatory cytokines. The most studied are interleukin-1 (IL)-1, IL-6 and tumor necrosis factor-alfa (TNF- ). Although the expression of these cytokines is low under normal conditions in the brain, it can be rapidly induced in response to injury.

This thesis is focused on the role of IL-1 family of proteins, namely the agonists IL-1 and and the endogenous IL-1 receptor antagonist (IL-1ra), and IL-6, in different experimental models of neurodegeneration. In order to study the role of IL-1 family of proteins during inflammation in an excitotoxic model of brain injury, adult rats were injected systemically with kainic acid, a glutamate analogue known to evoke seizures and neuronal cell loss in the rat brain. Using the combined technique of reverse-transcriptase coupled to PCR (RT-PCR) and in situ hybridization histochemistry, an upregulation of microglial mRNA expression of IL-1 and IL-1ra was found in brain areas with neuronal degeneration, such as the hippocampus and amygdala. IL-1ra mRNA was induced at later time point than IL-1 mRNA and was identified as the transcript coding for the secreted isoform of IL-1ra. This suggets that upregulation of these cytokines is a part of an inflammtory response associated with neurodegeneration and that the effect of IL-1 may be regulated by the expression of IL-1ra in this model. In order to study the role played by IL-1 in inflammation associated with traumatic brain injury (TBI), an experimental model was inflicted on transgenic mice. Heterozygous overexpression of the human secreted isoform of IL-1ra in the brain decreased the induction of IL-1 and IL-6 after injury. Using a neurological severity score (NSS), which mainly reflects motor recovery, we found that these animals recovered faster as compared to their non-transgenic littermates.

Furthermore, the proinflammatory cytokine expression was studied by RT-PCR in a mouse model of Alzheimer's disease (AD). The Tg2576 mice strain overexpress -amyloid (A ) precursor protein with the "Swedish" mutation linked to familiar AD and exhibits some of the neuropathology associated with AD, such as the deposition of insoluble extracellular amyloid fibrils (amyloid plaques) in specific brain regions. Analysis of expression of cytokines in the brain of Tg2576 mice revealed an early induction of IL-6 in the hippocampus and cerebral cortex that precedes the formation of amyloid plaques. This finding is interesting since in AD brain IL-6 is detected in microglia in the vicinity of diffuse plaques (non-fibrillar). Thus, the result from this study suggests that increased IL-6 expression may be an early event in AD inflammation.

The main constituent of amyloid plaques in the AD brain is the A peptide. The synthetic peptide A (25-35), a neurotoxic fragment of the full-length A peptide was studied for its ability to activate glial cells in culture and induce cytokine expression, as well as for its influence on G-protein coupled signalling in rat brain tissue. A (25-35) treatment of mixed astroglial cultures resulted in marked induction of IL-6 mRNA as studied by RT-PCR. Together with the results from the Tg2576 mice these results suggest a role for IL-6 in AD pathogenesis.

Alteration in cellular signal transduction has also been reported in AD brain. A (25-35) was shown to stimulate the enzymatic activities of GTPase and adenylate cyclase in membrane preparations from rat hippocampus and cerebral cortex, which are particularly affected regions in AD brain. Using Pertussis toxin treated membranes, the stimulatory effect on GTPase activity was totally abolished, suggesting that Gi/Go type of G-proteins mediated the effect of the A peptide. 

The proinflammatory cytokines interleukin (IL) -1α, IL-1β and IL-6 are key mediators in the host's response to injury and infection. One of the systemic responses elicited by these proinflammatory cytokines, when injected peripherally or centrally, is fever. Thus, these proteins may act as endogenous pyrogens.

We have shown that for a functional IL-1 mediated febrile response in vivo, both IL-1 receptor type I (IL-1RI) and IL-1 receptor accessory protein (IL-1RAcP) are necessary components, as knockout (KO) mice, lacking either IL-1RI or IL-1RAcP proteins, do not mount febrile responses to rrIL-1 (25-50 μg/kg, i.p.) injections, whereas they responded with fever when challenged by LPS (50 μg/kg, i.p.). These results suggest that IL-1 induces fever via signalling through a receptor complex, consisting of both IL-1RI and IL-1RAcP. We have also shown that IL-1β cannot induce NFκB translocation to the nucleus in primary astrocyte cultures derived from IL-1RAcP KO mice, suggesting a signalling role for this protein in mouse astrocytes.

When the regulation of interleukin-1β converting enzyme (ICE, caspase-1) was studied in rat after peripheral LPS treatment (2 mg/kg, i.p.), elevations were seen, on both mRNA levels and enzyme activity in the pituitary, whereas in adrenal glands, an increase of mRNA levels did not coincide with upregulated enzyme activity. However, LPS was further shown to differentially regulate ICE isoforms, some of which are supposed to act as endogenous inhibitors of ICE activity. Thus, ICE activity seems to be tightly controlled on a transcriptional level, where regulation of ICE isoforms play an important role, as well as at the post transcriptional level.

We have also shown that treatment with a neurotoxic fragment of β-amyloid, βA(25-35), induces a reactive phenotype of rat primary astrocyte cultures. The progression of this reactive phenotype was shown to coincide with elevated mRNA levels of IL-1α and IL-6. In addition, βA(25-35) treatment of primary astrocyte cultures derived from IL-1RI KO mice, induced a hypersensitive IL-1α response and a decreased IL-6 response. IL-1a and IL-6 are therefore proposed to be important molecules for development of reactive gliosis, and we also propose that signalling via IL-1RI is necessary for a full-scale induction of IL-6 mRNA.