Refine
Year of publication
- 2008 (85) (remove)
Document Type
- Doctoral Thesis (85) (remove)
Language
- English (85) (remove)
Has Fulltext
- yes (85)
Is part of the Bibliography
- no (85)
Keywords
- ALICE <Teilchendetektor> (2)
- Angiogenese (2)
- CERN (2)
- Dissertation (2)
- LHC (2)
- Quark-Gluon-Plasma (2)
- Schwerionenkollisionen (2)
- Strukturaufklärung (2)
- ABC-Transporter (1)
- AKBA (1)
Institute
- Biochemie und Chemie (24)
- Biowissenschaften (11)
- Physik (11)
- Pharmazie (10)
- Geowissenschaften (7)
- Medizin (4)
- Neuere Philologien (4)
- Frankfurt Institute for Advanced Studies (FIAS) (3)
- Psychologie (3)
- Informatik (2)
The respiratory chain is composed of protein complexes residing in the inner mitochondrial membrane of eukaryotes or in the cytoplasmic membrane of prokaryotes. This cellular energy converter transforms a redox potential stored in low potential substrates into an electrochemical potential across the respective membrane. Typical respiratory chains contain the complexes I, II, III and IV named according to their sequence in the respiratory chain reaction. Electrons of low potential substrates enter at complex I or II and are passed via complex III to complex IV where they are transferred to oxygen. The transport of electrons between the complexes is mediated by small electron shuttles like quinol or cytochrome c. Two different models describe their exchange either by (1) random collision of freely diffusible electron shuttles and membrane protein complexes or (2) arrangement of the complexes in supercomplexes enabling direct channeling of electron shuttles. In the Gram positive bacterium Corynebacterium glutamicum, the complex III to complex IV electron shuttle cytochrome c is not diffusible but a covalently bound part of the diheme cytochrome subunit QcrC of complex III. Therefore, the complexes III and IV have to form a supercomplex for electron transduction. The aim of this thesis was to purify and characterise this obligatory supercomplex III/IV of C. glutamicum. To gain sufficient biomass of C. glutamicum as starting material for purification, a phosphate buffered minimal medium was developed that enabled yield of total 120 g wet cell mass (38 g dry mass) in 12 L (6×2 L) shaking cultures. The determined conversion factor of glucose into biomass was 0.46 g/g indicating an intact respiratory chain. The yield was increased by bioreactor cultivation to ~690 g wet cell mass (~220 g dry mass) in ~10 L culture volume. A previously described homologous expression system was applied that produces the complex IV subunit CtaD with a fused Strep-tag II to facilitate purification. Affinity purifications using the Strep-tag II affinity to Strep-Tactin resin yielded a mixture of complexes and supercomplexes. Two supercomplex III/IV versions named supercomplex A and B and free complex IV were identified in this mixture by size exclusion chromatography, redox difference spectroscopy and two dimensional polyacrylamide gel electrophoresis including blue native polyacrylamide electrophoresis. The here presented downscaled blue native polyacrylamide electrophoresis method with analysis times of ~1 h enabled efficient screening of factors influencing the stability of supercomplex III/IV. The screening resulted that the integrity of supercomplex III/IV is preserved by using neutral detergents at minimal detergent to protein ratios for solubilisation and low detergent concentrations for purification and storage slightly above the required critical micellar concentration. Furthermore, pH <=7.5 is required for stability of supercomplex III/IV. Large biomass yields enabled upscaling of supercomplex III/IV affinity purification. Application of the identified stability conditions resulted in affinity purified samples free of supercomplex B. The major component supercomplex A was efficiently separated from residual free complex IV by preparative size exclusion chromatography. Concentration of purified supercomplex A by ultracentrifugation resulted in integrity of the supercomplex for several days at 4 °C. Purified supercomplex A contains ten different previously described subunits. The heme content of supercomplex A relative to the protein mass is heme A: 6.0 μmol/g, heme B: 6.5 μmol/g, and heme C: 5.8 μmol/g determined by redox difference spectroscopy and biochemical protein quantification. This indicates an equimolar ratio of complex III and complex IV in supercomplex A. Supercomplex A has quinol oxidase activity that is inhibited by stigmatellin or sodium azide. The turnover number of transferred electrons per complex III monomer is 148 s−1 at 25° C. The homogeneity and stability of the prepared supercomplex A enabled the growth of threedimensional crystals of up to 0.1 mm in length. Their composition of supercomplex A was verified by redox difference spectroscopy of intact crystals and blue native polyacrylamide electrophoresis of dissolved crystals. The crystals diffracted X-rays corresponding to a resolution of ~10 Å. Electron microscopy of negative stained samples revealed the uniform shape of purified supercomplex A particles with dimensions of 22 × 9 nm in the view plane. Combined heme quantification, size determination, determined activity, symmetry considerations, and particle shape indicate that supercomplex A has a central dimer of complex III and two monomers of complex IV on opposite sides. This conformation is functionally reasonable because it provides each complex III monomer with one complex IV monomer as electron acceptor. Therefore, the stoichiometry of supercomplex A is most likely III2IV2. The sensitivity of supercomplex A to detergents indicated a role of phospholipids in its stability. Therefore, a method for phospholipid identification and quantification was developed that is suitable for detergent solubilised crude and purified membrane protein samples. The analysis combines separation of phospholipid classes according to their head group by normal phase high performance liquid chromatography with evaporative light scattering detection. Calibration with external standard allows quantification of phospholipid amount in the range of 0.25-12 μg. The method is verified by analysing the phospholipid content of the well characterised complex III of Saccharomyces cerevisiae. The reduction of its phospholipid content during its purification steps is monitored. The complex III sample purified to crystallisation quality contains the phospholipid content that was also observed in previously reported structures determined by X-ray crystallography. Purified stable supercomplex A from C. glutamicum revealed a large content of bound phospholipids. The main differences between intact supercomplex A and a mixture of potentially disintegrated smaller complexes is that intact supercomplex A has a doubled phosphatidic acid content and an increased phosphatidyl glycerol content. The importance of the small anionic phosphatidic acid for mediation of contacts between complexes in a supercomplex is discussed. The total phospholipid content of stable supercomplex A is sufficient for a complete belt surrounding the supercomplex in the membrane plane. This indicates that also all essential internal phospholipid binding positions are occupied and potentially stabilise supercomplex A.
Hypoxic pulmonary vasoconstriction (HPV) redistributes pulmonary blood flow from areas of low oxygen partial pressure to areas of normal or relativity high oxygen availability, thus optimising the matching of perfusion to ventilation and preventing arterial hypoxemia. Generalised alveolar hypoxia results in a sustained increase in pulmonary artery pressure which in turn leads to structural changes in the walls of the pulmonary vasculature (pulmonary vascular remodelling). Recent findings have indicated a role for cytochrome P450 (CYP) epoxygenase-derived epoxyeicosatrienoic acids (EETs) in hypoxia-induced pulmonary vasoconstriction. Given that the intracellular concentration of EETs is determined by the soluble epoxide hydrolase (sEH), which metabolises EETs to their less active dihydroxyeicosatrienoic acids (DHETs), we assessed the influence of the sEH and EETs on pulmonary artery pressure, acute and chronic HPV, and pulmonary vascular remodelling in the mouse lung. In isolated lungs from wild-type mice, acute HPV was significantly increased by sEH inhibition, an effect abolished by pre-treatment with CYP epoxygenase inhibitors and the EET antagonist 14,15-EEZE. The acute hypoxia-induced vasoconstriction and EET production were greater in lungs from sEH-/- mice than from wild-type mice and sEH inhibition had no further effect on HPV in lungs from the former animals, while MSPPOH (CYP epoxygenase inhibitor) and 14,15-EEZE decreased the response. Exogenous application of 11,12-EET increased pulmonary artery pressure in a concentration-dependent manner and enhanced acute HPV in wild-type lungs, while 14,15-EET and 11,12-DHET were without significant effect on pulmonary artery pressure. 5-HT2A receptor antagonism or Rho kinase inhibition shifted the EET concentration-response curve to the right and abrogated the EET- and sEH inhibition-induced potentiation of acute hypoxic vasoconstriction. In lungs from wild-type and sEH-/- mice, hypoxic preconditioning (hypoxic ventilation for 10 minutes) enhanced the 5-HT response. 1-Adamantyl-3-cyclohexylurea (ACU), a sEH inhibitor, further amplified the hypoxia-induced 5-HT-hypersensitivity in wild-type mice. However, after hypoxic preconditioning, the sEH-/- lungs displayed a striking leftward shift in the 5-HT response. 11,12-EET can activate TRPC6 channels in endothelial cells by eliciting its translocation to the plasma membrane, more specifically to membrane domains enriched with the caveolae marker caveolin-1. This effect was also observed in rat pulmonary artery smooth muscle cells overexpressing the channel. Exposure of the latter cells to acute hypoxia also stimulated the intracellular translocation of TRPC6 to caveolae, an effect that was sensitive to the EET antagonist. The EET-induced translocation of TRPC6 channels was prevented by a 5-HT2A receptor antagonist but not by a Rho kinase inhibitor. Moreover, while acute hypoxia and 11,12-EET increased pulmonary pressure in lungs from TRPC6+/- mice, lungs from TRPC6-/- mice did not respond to either stimuli. These results indicate that the sEH and CYP-derived EETs are involved in acute HPV and that EET-induced pulmonary contraction under normoxic and hypoxic conditions involves a TRPC6 channel, a 5-HT2A receptor-dependent pathway and Rho kinase activation. In the second part of the study the role of the sEH in the development of pulmonary hypertension and vascular remodelling induced in mice by exposure to hypoxia (10% O2) for 21 days was analysed. In wild-type mice, chronic hypoxia decreased the pulmonary expression/activity of the sEH, induced right heart hypertrophy and erythropoiesis, and increased the number of partially and fully muscularised pulmonary resistance arteries (by 3-fold). Moreover, in HEK 293 cells, hypoxia (1% O2 up to 24 h) decreased sEH promoter activity by 50%. In isolated lungs, pre-exposure to chronic hypoxia significantly increased baseline perfusion pressures and potentiated the acute HPV. While an sEH inhibitor, ACU, potentiated acute HPV in lungs from mice maintained in normoxic conditions, it had no effect on HPV in lungs from mice exposed to hypoxia. The EET antagonist, 14,15-EEZE, abolished the sEH inhibitor-dependent increase in acute HPV in normoxic lungs and decreased HPV in chronic hypoxic lungs. Hypoxia-induced right heart hypertrophy and erythropoiesis were more pronounced in sEH-/- than in wild-type mice. Under normoxic and hypoxic conditions the muscularisation of resistance pulmonary arteries was greater in lungs from sEH-/- mice than in lungs from wild-type mice. sEH-/- mice also displayed an enhanced acute HPV, compared to that observed in wild-type mice and chronic exposure to hypoxia did not further potentiate acute HPV. However, in the presence of 14,15-EEZE responses returned to levels observed in normoxic lungs from wild-type animals. Furthermore, immunohistochemistry demonstrated an extensive expression of the sEH in the medial wall of pulmonary arteries from human donor lungs. Whereas sEH expression was not detectable in samples from pulmonary hypertension patients, indicating that the sEH is involved in hypoxia-induced pulmonary vascular remodelling and hypoxic pulmonary vasoconstriction. Taken together, the results presented in this thesis indicate that the expression/activity of the sEH is an important determinant of the magnitude of acute and chronic hypoxia-induced pulmonary vasoconstriction and pulmonary vascular remodelling by inactivating vasoconstrictor CYP-derived EETs. As sEH inhibitors are currently being developed for the treatment of human systemic hypertension, it should be noted that these compounds may even promote the development of pulmonary hypertension.
RNA interference (RNAi) is triggered by recognition of double-stranded RNA (dsRNA), and elicits the silencing of gene(s) complementary to the dsRNA sequence. RNAi is thought to have emerged as a way of safeguarding the genome against mobile genetic elements and viral infection, thus maintaining genomic integrity. dsRNA is first processed into small interfering RNAs (siRNA) by the enzyme Dicer. siRNAs are ~21 to 25 -nt long, and contain a signature 5’ phosphate group and a two nucleotide long 3’ overhang (Bernstein et al., 2001). The siRNA is then loaded into the RNA-induced si-lencing complex (RISC), of which Argonaute is the primary catalytic component (Liu et al., 2004). Energetic asymmetry of the siRNA ends allows for its directional loading into RISC (Khvorova et al., 2003; Schwarz et al., 2003). Argonaute cleaves the passen-ger strand of the siRNA, leaving the guide strand of the siRNA bound to RISC (Gregory et al., 2005; Matranga et al., 2005; Rand et al., 2005). This single-stranded guide strand siRNA bound to Argonaute is able to recognize target mRNA in a sequence-specific manner, and cleaves the mRNA. Argonaute 2 in complex with single-stranded siRNA is sufficient for mRNA recognition and cleavage, thus forming a minimal RISC (Rivas et al., 2005). miRNAs, endogenously expressed small RNA genes which typically contain mismatches and non-Watson-Crick base pairing, are processed by this general pathway, although typically modulate gene expression by translational repression as opposed to cleavage of their target mRNA. The number of Argonaute genes is highly variable between species, ranging from one in S. pombe to twenty-seven in C. elegans. Earlier crystal structures of Argonaute apoen-zymes show the architecture of Argonaute to be a multidomain protein composed of N terminal, PAZ, MID, and PIWI domains (Song et al., 2004; Yuan et al., 2005). These multi-domain proteins are present in both prokaryotic and eukaryotic organisms. The role of Argonaute proteins in prokaryotes is still unknown, but based similarity to eu-karyotic Argonautes, they may also be involved in nucleic acid-directed regulatory pathways. These proteins have served as excellent models for learning about the struc-ture and function of this family of proteins. RNAi has found a widespread application for the simple yet effective knockdown of genes of interest. The catalytic cycle of RISC requires the binding of a number of different nucleotide structures to Argonaute, and we expect Argonaute to undergo a number of conforma-tional changes during the cycle of mRNA recognition by RISC (Filipowicz, 2005; Tom-ari and Zamore, 2005). Nevertheless, it remains unclear how the multi-domain ar-rangement of Argonaute recognizes and distinguishes between single-stranded and dou-ble-stranded oligonucleotides, which correspond to the Dicer-processed siRNA product, guide strand siRNA, and the guide strand / mRNA duplex. The Argonaute protein from Aquifex aeolicus was cloned, expressed, crystallized and solved by molecular replacement. Relative to earlier Argonaute structures, a 24° reorientation of the PAZ domain in this structure opens a basic cleft between the N-terminal and PAZ domains, exposing the guide strand binding pocket of PAZ. A 5.5-ns molecular dynamics simulation of Argonaute showed a strong tendency of the PAZ and N-terminal domains to be mobile. Binding of single-stranded DNA to Argonaute was monitored by total internal reflection fluorescence spectroscopy (TIRFS). The experi-ments showed biphasic kinetics indicative of large conformational changes, and re-vealed a hotspot of binding energy corresponding to the first 9 nucleotides, the so-called “seed region” most crucial for sequence-specific target recognition. As RNAi may have evolved as a way of safeguarding the genome viral infection, it is not surprising that viruses have evolved different strategies to suppress the host RNAi response in the form of viral suppressor protein. (Hock and Meister, 2008; Lecellier and Voinnet, 2004; Rashid et al., 2007; Song et al., 2004; Vastenhouw and Plasterk, 2004). These viral suppressors are widespread, having been identified in a number of different viral families. Not surprisingly, they generally share little sequence homology with one another, although they appear to exist as oligomers built upon a ~ 100-200 amino acid protomer. Tomato aspermy virus, a member of the Cucumoviruses, encodes for protein 2B (TAV 2B, 95 a.a., ~11.3 kDa) that acts as an RNAi suppressor. Intriguingly, a similar genomic arrangement is seen in RNAi suppressors in the Nodaviruses, a family of viruses that can infect both plants and animals, such as Flock house virus b2 (FHV b2). The 2B and b2 proteins are both derived from a frameshifted ORF within the RNA polymerase gene (Chao et al., 2005). In spite of this genomic similarity, the 2B and b2 proteins share little sequence identity, and it is not well understood how the Cucumovirus 2B proteins suppress RNAi. To address how TAV 2B suppresses RNAi, the oligonucleotide-binding properties of TAV 2B were studied. TAV 2B shows a preference for double-stranded RNA oligonucleotides corresponding to siRNAs and miRNAs, and also binds to single-stranded RNA oligonucleotides. A stretch of positively charged residues between amino acids 20-30 are critical for RNA binding. Binding to RNA oligomerizes and induces a conformational change in TAV 2B into a primarily helical structure. These studies sug-gest that suppression of RNAi by TAV 2B may occur by targeting different stages of the RNAi pathway. TAV 2B falls under the category of more general RNAi suppres-sors, with potentially multiple targets for suppression.
The µ-opioid receptor is the primary target structure of most opioid analgesics and thus responsible for the predominant part of their wanted and unwanted effects. Carriers of the frequent genetic µ-opioid receptor variant N40D (allelic frequency 8.2 - 17 %), coded by the single nucleotide polymorphism A>G at position 118 of the µ-opioid receptor coding gene OPRM1 (OPRM1 118A>G SNP), suffer from a decreased opioid potency and from a higher need of opioid analgesics to reach adequate analgesia. The aim of the present work was to identify the mechanism by which the OPRM1 118A>G SNP decreases the opioid potency and to quantify its effects on the analgesic potency and therapeutic range of opioid analgesics.
To elucidate the consequences of the OPRM1 118A>G SNP for the effects of opioid analgesics, brain regions of healthy homozygous carriers of the OPRM1 118A>G SNP were identified by means of functional magnetic resonace imaging (fMRI), where the variant alters the response to opioid analgesics after painful stimulation. Afterwards, the µ-opioid receptor function was analyzed on a molecular level in post mortem samples of these brain regions. Finally, the consequences of the OPRM1 118A>G SNP for the analgesic and respiratory depressive effects of opioids were quantified in healthy carriers and non-carriers of OPRM1 118A>G SNP by means of experimental pain- and respiratory depression-models.
To identify pain processing brain regions, where the variant alters the response to opioid analgesics after painful stimulation, we investigated the effects of different alfentanil concentration levels (0, 25, 50 and 75 ng/ml) on pain-related brain activation achieved by short pulses (300 msec) of gaseous CO2 (66% v/v) delivered to the nasal mucosa using a 3.0 T magnetic head scanner in 16 non-carriers and nine homozygous carriers of the µ-opioid receptor gene variant OPRM1 118A>G. In brain regions associated with the processing of the sensory dimension of pain (pain intensity), such as the primary and secondary somatosensory cortices and the posterior insular cortex, the activation decreased linearly in relation to alfentanil concentrations, which was significantly less pronounced in OPRM1 118G carriers. In contrast, in brain regions known to process the affective dimension of pain (emotional dimension), such as the parahippocampal gyrus, amygdala and anterior insula, the pain-related activation disappeared already at the lowest alfentanil dose, without genotype differences.
Subsequently, we investigated the µ-opioid receptor-expression ([3H]-DAMGO saturation experiments, OPRM1 mRNA analysis by means of RT-PCR), the µ-opioid receptor affinity ([3H]-DAMGO saturation and competition experiments) and µ-opioid receptor signaling ([35S]- GTPγS binding experiments) in post mortem samples of the human SII-region, as a cortical projection region coding for pain intensity, and lateral thalamus, as an important region for nociceptive transmission. Samples of 22 non-carriers, 21 heterozygous and three homozygous carriers of OPRM1 118A>G SNP were included into the analysis. The receptor expression and receptor affinity of both brain regions did not differ between non-carriers and carriers of the variant N40D. In non-carriers, the µ-opioid receptors of the SII-region activated the receptor bound G-protein more efficiently than those of the thalamus (factor 1.55-2.27). This regional difference was missing in heterozygous (factor 0.78-1.66) and homozygous (factor 0.66-1.15) carriers of the N40D variant indicating a reduced receptor-G-protein-coupling in the SII-region.
Finally, the consequences of the alteration of µ-opioid receptor function in carriers and noncarriers of the genetic variant was investigated using pain- and respiratory depression-models. Therefore, 10 healthy non-carriers, four heterozygous and six homozygous carriers of the µ- opioid receptor variant N40D received an infusion of four different concentrations of alfentanil (0, 33.33, 66.66 and 100 ng/ml). At each concentration level, analgesia was assessed by means of electrically (5 Hz sinus 0 to 20 mA) and chemically (200 ms gaseous CO2 pulses applied to the nasal mucosa) induced pain, and respiratory depression was quantified by means of hypercapnic challenge according to Read and recording of the breathing frequency. The results showed that depending on the used pain model, both heterozygous and homozygous carriers of the variant N40D needed 2 – 4 times higher alfentanil concentrations to achieve the same analgesia as non-carriers. This increase seems to be at least for homozygous carriers unproblematic, because to reach a comparable respiratory depression as non-carriers, they needed 10-12 times higher alfentanil concentrations.
The results of this work demonstrate that the µ-opioid receptor variant N40D causes a regionally limited reduction of the signal transduction efficiency of µ-opioid receptors in brain regions involved in pain processing. Thus, the painful activation of sensory brain regions coding for pain intensity is not sufficiently suppressed by opioid analgesics in carriers of the variant N40D. Due to the insufficient suppression in hetero- and homozygous carriers of the variant N40D, the concentration of opioids has to be increased by a factor 2 - 4, in order to achieve the same analgesia as in non-carriers. At the same time, the respiratory depressive effects are decreased to a greater extent in homozygous carriers of the N40D variant as they need a 10 - 12 times higher opioid concentration to suffer from the same degree of respiratory depression as non-carriers. Due to the increased therapeutic range of opioid analgesics, an increase of the opioid dose seems to be harmless, at least for homozygous carriers of the N40D variant.
Platelets are anucleate cells that play a major role in hemostasis and thrombosis in the vasculature. During primary hemostasis platelets adhere to sites of vascular damage and the initial platelet coat is reinforced by additional platelets forming a stable aggregate. At the same time platelets secrete their intracellular granules containing substances that further activate platelets in an autocrine and paracrine fashion and affect local coagulation and endothelial smooth muscle cell function. The small guanine nucleotide binding protein Rap1 regulates the activity of the platelet integrin alphaIIbbeta3 and thus platelet aggregation. Rap1 activity is controlled by guanine nucleotide exchange factors and GTPase activating proteins. In platelets, Rap1GAP2 is the only GTPase activating protein of Rap1. In order to identify Rap1GAP2-associated proteins, a genetic two-hybrid screening in yeast was performed and synaptotagmin-like protein 1 (Slp1, also called JFC1) was found as a new putative binding partner of Rap1GAP2. Slp1 is a tandem C2 domain containing protein and is known to bind to Rab27, a small GTPase involved in platelet dense granule secretion. The direct interaction between Rap1GAP2 and Slp1 was confirmed in yeast and in transfected cells. More importantly, Slp1 is expressed in platelets and binding of endogenous Rap1GAP2 and Slp1 was verified in these cells. The Rap1GAP2 and Slp1 interaction sites were mapped by mutational analysis. Rap1GAP2 binds through the -TKXT- motif within its C-terminus to the C2A domain of Slp1. Moreover, the Slp1 binding -TKXT- motif of Rap1GAP2 was confirmed by complementary approaches using short synthetic Rap1GAP2 peptides. The C2A domain of Slp1 is a phospholipid binding domain and thus mediates binding of Slp1 to the plasma membrane. Phospholipid overlay assays revealed that simultaneous binding of Slp1 via its C2A domain to Rap1GAP2 and to phospholipids can occur. In addition, the interaction between Rap1GAP2 and Slp1 is regulated by cAMP-dependent protein kinase (cAK or PKA), and kinase activation in platelets enhanced binding of endogenous Rap1GAP2 to Slp1. In-vitro phosphorylation assays revealed that Slp1 is a substrate of PKA, and serine 111 was identified as phosphorylation site. Since Slp1 is a Rab27 binding protein, a trimeric complex of Slp1, Rab27 and Rap1GAP2 is conceivable. The association of Slp1, Rab27 and Rap1GAP2 was investigated by immunofluorescence and co-immuno-precipitation experiments in both, transfected cells and platelets. By Slp1 affinity chromatography and subsequent mass spectrometric analysis additional Slp1 binding proteins were identified in platelets, and binding of Slp1 to Rab8 was confirmed in pull-down assays. To investigate the functional significance of the interaction between Rap1GAP2 and Slp1, an assay system was established to determine serotonin secretion of streptolysin-O permeabilized platelets. Addition of recombinant Slp1 protein to permeabilized platelets strongly inhibited platelet dense granule secretion, whereas addition of recombinant Rap1GAP2 protein or synthetic Rap1GAP2 peptide enhanced secretion. Deleting the Slp1 binding -TKXT- motif abolished the stimulatory effect of Rap1GAP2 on secretion. Addition of Rap1 to permeabilized platelets had no effect on secretion. These findings indicate that the Rap1GAP2 effect on platelet secretion does not depend on the GTPase activating function of Rap1GAP2, but is rather dependent on the -TKXT- mediated interaction of Rap1GAP2 with Slp1. In addition, in-vitro GAP assays revealed that Slp1 binding to Rap1GAP2 does not affect the Rap1GAP activity of Rap1GAP2, and adhesion assays excluded a role for the Rap1GAP2/Slp1 interaction in cell adhesion. Altogether, the results of the present study demonstrate that besides its function in platelet aggregation by controlling the activity of the small guanine nucleotide binding protein Rap1, Rap1GAP2 is involved in platelet dense granule secretion by the new -TKXT- mediated interaction with the Rab27 and membrane binding protein Slp1. In addition, the interaction between Rap1GAP2 and Slp1 is embedded into an elaborate network of protein-protein interactions in platelets which appear to be regulated by phosphorylation. Future studies will in particular aim to dissect the molecular details of Rap1GAP2 and Slp1 action in platelet secretion and investigate the potential biochemical and pharmacological value of the unique protein binding -TKXT- motif of Rap1GAP2.