Defect in the Protein Trafficking Machinery in Human Diseases

Séminaire FSER organisé par Bruno Goud et David Sabatini du 29 mai au 3 juin 2006

Participants

William E. Balch, Juan Bonifacino, Véronique Chauvet, Esteban Dell’Angelica, Bruno Goud, Daniel Hanau, Martin Lowe, Jean-Paul Luzio, Ralph Nixon, Graça Raposo, Mindong Ren, Enrique Rodriguez-Boulan, David D. Sabatini, Frédéric Saudou, Miguel Seabra, Geneviève de St-Basile, Linton Traub, Peter van der Sluijs, Thomas E. Willnow

Review

Recent advances in the understanding of the molecular mechanisms of intracellular transport highlight that many human pathologies are directly linked to defaults in the transport machinery. In most cases, this leads to the formation of a defective organelle unable to perform its function within the cell. In particular, defaults in the transport machinery have been associated with Hermansky-Pudlak, Chediak-Hisgashi et Griscelli syndromes, that affect the biogenesis of lysosomes or lysosome-related organelles such as melanosomes in melanocytes or cytotoxic granules in cytotoxic T cells. Others pathologies, including Alzheimer and Huntington diseases and the Lowe syndrome also result from alterations in various components of the transport machinery.
This timely colloquium brought together specialists in the field of intracellular transport working on animal models of several pathologies or with cells derived from patients. Below, we summarize the main findings discussed by each participant in sequential order of presentation.

William Balch (Scripps Research Institute, La Jolla, USA)
Bill Balch opened the conference with a provocative presentation on aspects of processes that in the endoplasmic reticulum (ER) promote folding of cargo proteins and contribute to their incorporation into the COPII coded vesicles that mediate transport to the Golgi apparatus. Balch emphasized the tissue-specific role of chaperones and the ER environment in regulating folding and exit from the organelle. In studying the energetics of protein folding and exit he has concluded that the energetic threshold for transport of proteins out of the ER is very permissive and, therefore, proteins do not need to achieve a native state folding stability in order to be exported from the ER. Balch is concentrating on chemical interactions that may compensate for defective folding of mutant proteins, such as the mutant CFTR in cystic fibrosis that must reach the plasma membrane to exert their function.

Linton Traub (University of Pittsburg School of Medicine, Pittsburgh, USA)
The internalization of extracellular ligands by receptors that are incorporated into clathrin coated vesicles requires the participation of adaptor proteins that recognize signals in the cytoplasmic domains of the receptors and also promote assembly of clathrin triskelions into the polymeric cages that promote vesicle budding. For many years the only two heterotetrameric adaptors, AP1 and AP2 were known and only the latter was known to promote the assembly of clathrin coated pits at the plasma membrane. Numerous other adaptors are now known and Linton Traub has carried out extensive functional and structural studies of two of them, the Autosomal Recessive Hypercholesterolemia (ARH) and Disabled-2 (Dab2), both of which recognize the FXNPXY sorting signals found in the cytoplasmic domains of members of the LDL receptor superfamily. These proteins are known as CLASPS, Clathrin Associated Sorting Proteins, and they interact with AP2 to extend the repertoire of the latter’s potential endocytic cargos. CLASPS bind to the b2 appendage platform of AP2 via a specific motif that is also found in other CLASPS, notably the b arrestins that govern the internalization of G protein coupled receptors (GPCRs). Traub has carried out a crystallographic analysis of the complex formed between an ARH peptide and the b2 appendage which has illuminated the molecular mechanism of the ligand-dependent triggering of GPCR uptake in clathrin coated vesicles. Although Dab2 and ARH are functionally redundant in non-polarized cells, individuals with two mutant ARH alleles present with a clinically manifested hypercholesterolemia resembling that in patients with a defective LDL receptor. This phenotype is likely to be accounted for by the low level of expression of Dab2 in hepatocytes.

Enrique Rodriguez-Boulan (Cornell University, New-York, USA)
Enrique Rodriguez-Boulan discussed various aspects of the trafficking mechanisms that operate in polarized cells to sort in the biogenetic and recycling pathways plasma membrane proteins to the distinct surfaces of those cells.
AP1 adaptors have been classically known to mediate the intracellular sorting of the manose-6-phosphate receptor and other proteins into vesicles formed in the trans region of the Golgi apparatus that are ultimately destined to late endosomes.
The discovery of an AP1 adaptor variant that contain an isoform of the m1B subunit that appears to be specific to polarized epithelial cells was followed by the demonstration that polarized cells lacking this AP1 isoform fail to properly localize certain basolateral proteins, such as the transferrin receptor, unless they are transfected to express the missing isoform.
In previous work Rodriquez-Boulan showed that the AP1 isoform functions and effects trancytosis in the endocytic pathways to resort interiorized basolateral plasma membrane proteins that had initially been delivered to the apical surface. He has shown, using RNAi to deplete AP1B from MDCK cells, that the postendocytic pathway operates only in newly polarized MDCK monolayers. Cells within mature monolayers lacking m1B can, nevertheless, accurately deliver newly synthesized tranferrin receptors to the basolateral membrane, presumably through a still unidentified adaptor. They, however, missort the protein after its endocytosis, reflecting the requirement of m1B for postendocytic sorting.

Juan Bonifacino (NIH, Bethesda, USA)
Juan Bonifacino presented studies that enhance our understanding of the mechanisms by which the HIV Nef protein promotes down regulation in the levels of CD4, the HIV co-receptor that binds the infecting virus at the cell surface of lymphocytes. This downregulation prevents superinfection of infected cells, and facilitates release of newly produced virions. The Nef protein is essential for virus pathogenecity, since individuals infected with a virus with Nef mutations show long term non-progression to AIDS and do not manifest the characteristic depletion of CD4+ T cells.
Nef binds to the cytoplasmic domain of CD4, and itself contains a dileucine motif that can be recognized by AP2, supporting a role of Nef as a mediator of endocytosis of CD4, which could lead to its degradation. On the other hand, Nef can bind directly to AP1 and AP3, arguing for a role of Nef in promoting intracellular retention of CD4.
Bonifacino developed a system of cultured Drosophila S2 cells that constitutively express human CD4 and can express Nef from the metalothionine promoter in a copper inducible manner. He tested in RNAi interference experiments the effect of knocking down individually 75 candidate genes that encode proteins involved in protein trafficking on the capacity of Nef to promote CD4 down regulation. He found that the knock down of clathrin and AP2 levels counteracted the Nef down regulating effect, while the knock down of AP1, AP3 and other proteins involved in biosynthetic trafficking had no inhibitory effect. He concluded that the main mechanism by which Nef mediates CD4 down regulation requires internalization of the coreceptor from the plasma membrane.

J. Paul Luzio (University of Cambridge, Cambridge, UK)
Luzio reviewed his studies on traffic events between endosomes and lysosomes. Using live cell imaging techniques to track differentially labeled organelles and in vitro systems that he developed to detect fusion he demonstrated that kiss and run interactions that permit transfer of luminal content, as well as direct fusion events, take place between the organelles, but ruled out vesicular transport between them. Homotypic fusion of late endosomes requires the R-SNARE VAMP8, whereas their heretotypic fusion with lysosomes involves the R-SNARE VAMP7. Once fusion of lysosomes with late endosomes has taken place, the hybrid organelle undergoes a process of condensation of content and removal or recycling of membrane proteins through the budding of vesicles, which leads to the reformation of a lysosome.
VAMP7 plays a critical role in mediating fusion of late endosomes with lysosomes. The results of antibody inhibition experiments suggest the requirement of a trans SNARE complex composed of syntaxin7, Vti1b, syntaxin8, and VAMP7. Luzio considered the question of how VAMP7 is correctly localized to late endosomes. Earlier reports have shown that VAMP7, which is a member of the “longin” subfamily of R-SNAREs, interacts through the longin domain with the delta subunit of the AP3 adaptor and this may explain the recruitment of VAMP7 to late endosomes where it plays its critical role in heterotypic fusion with lysosomes. He also explained that the VAMP7 longin domain interacts with the protein Hrb (HIV Rev binding protein), which contains a motif that is known to bind to the alpha appendage of AP2. In NRK cells Hrb co-localizes with clathrin and AP2.
Luzio also discussed the disease mucolipidosis IV, which results from mutations in the lysosomal membrane protein mucolipin 1 and suggested that this protein may be involved in the reformation of lysosomes after fusion events.
His laboratory has also studied aspects of the sorting of membrane proteins into the luminal vesicles characteristic of multivesicular bodies (MVB), a process that leads to the degradation of interiorized receptors such as EGF. The sorting of proteins into the lumen of MVBs has been studied in yeast where it was shown to involve the participation of four ESCRT protein complexes (Endosomal Sorting Complexes Required for Transport), whose components are encoded by a subclass of VPS genes. These studies with mammalian cells indicate that Spastin, a protein mutated in Hereditary Spastic Paraplegias, interacts with an ESCRT associated protein.

Martin Lowe (University of Manchester, Manchester, UK)
Lowe presented a broad discussion of the structure and function of the OCRL1 protein and its role in membrane traffic. A defect in this protein leads to an X-linked disorder characterized by congenital cataracts, mental retardation and renal Fanconi syndrome. OCRL is a type 2 inositolpolyphosphate-5-phosphatase that hydrolyzes the phosphoinositides PI(4, 5)P2 and PI(3, 4, 5)P3. The protein contains a 5-phosphatase domain separated by a linker region from a C-terminal domain with homology to a Rho GAP. The linker domain contains a peptide motif that binds to the clathrin heavy chain and OCRL1 promotes clathrin assembly in vitro. OCRL1 is localized to the Golgi apparatus, endosomes and plasma membrane ruffles to which it translocates upon activation of RAC. OCRL1 is present in clathrin-coated vesicles that operate between the TGN and endosomes and binds to the alpha subunit of the AP2 adaptor. The presence of OCRL1 in the TGN may ensure that low levels of PIP2 are maintained in that organelle through the generation of P14P, which together with ARF recruits a number of effectors, including the clathrin adaptor AP1.
Lowe discussed the role of OCRL1 in endosome to TGN traffic and showed that overexpression of the protein results in a redistribution of clathrin and the cation independent mannose-6-phospate receptor (CI-M6PR) and TGN46 to enlarged endosomal structures that are defective in retrograde transport to the TGN, as demonstrated employing the B-subunit of the Shiga toxin.
Depletion of OCRL1, also causes an accumulation of CI-M6PR and TGN46 to early endosomes, where the transferrin receptor is also found, as expected from a block in traffic to the TGN.
OCRL1 was found to bind to the GTP form of various Rab proteins, including Rab1, Rab5 and Rab6. It was proposed that OCRL1 is brought to the Golgi by Rab6, which stimulates its phosphatase activity and generates P14P through hydrolysis of PIP2, which together with ARF (and AP1) recruits clathrin. Inpp5b is another 5-phosphatase with 45% homolog to OCRL-1, for which it is likely to substitute in many tissues not affected in Lowe’s syndrome. Indeed, the knockout of either OCRL1 or Inpp5b gives viable mice but the double knockout is lethal. OCRL1 also plays a role in maintaining the actin cytoskeleton and regulating actin dynamics. The Rho GAP domain may serve to target OCRL to the plasma membrane.
The defects in protein reabsorption in the kidney in Lowe syndrome are likely to be due to a diminution in the levels at the apical membrane of proximal tubules of the scavenger receptors, megalin and cubulin, which in the absence of OCRL1 may be unable to reach the cell surface from either the TGN or recycling endosomes.

Veronique Chauvet (Institut Curie, Paris, France)
V. Chauvet, a post-doc in B. Goud’s laboratory, presented results on the interaction between Rab6 and OCRL1, which confirmed those reported by M. Lowe (see above). She also gave an overview of the pathologies associated with the Lowe syndrome and Dent’s disease, considered so far as distinct entities both on clinical and genetic basis. Dent’s disease is a renal disorder associating proteinuria, hypercalciuria and nephrolithiasis, mostly due to mutations in CLCN5 encoding a voltage-gated chloride channel. Recent findings indicate that several cases with the isolated renal phenotype of Dent’s disease are due to OCRL1 mutations.

Miguel Seabra (Imperial College School of Medicine, London, UK)
Seabra discussed the role of several Rab proteins in the biogenesis and transport of melanosomes, the pigment containing granules that are present in epidermal melanocytes and in cells of the retinal pigmentary epithelium (RPE).
The “chocolate” mouse mutant is defective in Rab38 and he studied the role of the Rab in primary cultures of melanocytes in which the different stages of melanosome development can be observed. Rab38 has substantial sequence homology (62%) with Rab32 and the two proteins co-localize in melanocytes with the melanosomal proteins TRP1 and tyrosinase. The two Rabs appear to be functionally equivalent, which may account for the observation that “chocolate” cells maintained in culture eventually recover their pigmentation. Moreover, knockdown of Rab32 in chocolate melanocytes with RNAi to generate a functional null for both Rab proteins blocks exit of the melanosomal proteins TRP1 and tyrosinase from the Golgi apparatus, where they accumulate in the TGN together with TGN38. Under these conditions, stage 2 melanosomes, marked by the protein pmel17 become abundant in the cultured cells.
Seabra is also studying melanogenesis in the RPE, where he found that Rab32 expression could not compensate for the defective Rab38 in “chocolate” cells. This may reflect the limited time window of expression of Rab32 during development.
Seabra did not present his work on Choroideremia, a retinal degenerative disease which he has shown results from a defect in a subunit (REP1) of the geranyl-geranyl transferase that in the eye specifically prenylates Rab27a. Instead, he discussed his studies on the mechanism by which Rab27a, which is a melanosome-binding Rab, functions in the transport of melanosomes from the perinuclear regions of melanocytes to the tips of the dendritic process of the cells, where transfer of the granules to keratinocytes occurs. It is known that a tripartight complex, consisting of Rab27a linked via the protein melanophilin to myosin Va, mediates movement of the granules through the action of the myosin motor on actin filaments. Different forms of Griscelli Syndrome result from mutations in each one of these three proteins. Corresponding mouse mutants have also been identified: the “ashen” mouse for Rab27a, “dilute” for myosin Va, and “leaden” for melanophilin. Seabra explained that in RPE cells the role of melanophilin can be carried out by another Rab27a interacting protein, MyrIP, which also interacts with myosin Va. He regards melanophilin as a modular protein and has been able using a novel melanosome assay with melanophilin null cell lines, to dissected the melanophilin protein into its modules and identified domains that interact specifically with the various partners.

Daniel Hanau (Etablissement Français du Sang-Alsace, Strasbourg, France)
Daniel Hanau is a specialist of Langerhans cells, a distinct population of immature dendritic cells, located in epidermis and stratified mucosal epithelia. He presented a recent study on the biogenesis of Birbeck granules (BGs), unusual rod-shaped ultrastructural organelles, of unknown function, specific to Langerhans cells.
D. Hanau and coll. have previously investigated the intracellular location and fate of Langerin, a type II transmembrane Ca2+-dependent lectin and the major component of BGs. In Langerhans cells, as well as in transfected cells, Langerin continuously recycles between the early endosomal compartments and the plasma membrane, with dynamic retention in the Rab11+ endosomal recycling compartment (ERC). Moreover, BGs, where Langerin accumulates, appear to be subdomains of the ERC. Since Rab11 regulates receptor recycling, D. Hanau and coll. investigated its role in Langerin traffic and distribution, and in the cellular localization of BGs. Expression of the dominant negative Rab11 mutant dramatically affects the internalization of yeast invertase, a Langerin ligand, and led to the formation of a Langerin positive tubular network which lacks BGs. Moreover, at higher levels of expression of the dominant negative Rab11 mutant, Langerin became un-detectable. Remarkably, the complete extinction of Rab11 with small interfering RNA induced the specific lysosomal degradation of Langerin. Despite rescue of Langerin accumulation by lysosomotropic agents, no BGs were found when Rab11 molecules were knock-down. Therefore, the formation of BGs is dependent on the presence of an “active” Rab11 molecule, enlightening a role for Rab proteins in compartment biogenesis.
At the end of his presentation, D. Hanau emphasized that this particular traffic might have consequences on the immunological functions of Langerhans cells. In these cells, CD1a share a common cellular pathway with Langerin and, in particular, accumulates in BGs. CD1a is involved in the presentation of microbial lipidic antigens to T cells. Therefore, BG might represent a specialized membrane domain of the ERC, devoted to the loading of CD1a with antigens internalized and routed to the BGs by Langerin. Altogether, these observations suggest that Rab11 not only participates in the architecture of the “housekeeping cellular compartment” ERC, but also of specialized immunological compartments, BGs.

Esteban Dell’Angelica (University of California, Los Angeles, USA)
Since his work with Bonifacino that first identified the subunits of the AP3 adaptor and demonstrated that mutations in these genes were responsible for different forms of the Hermansky Pudlak Syndrome (HPS), Dell’Angelica has gone on to discover that other forms of HPS are due to mutations in components of multimeric complexes known as BLOCS (for Biogenesis of Lysosome-related Organelle Complexes). Three BLOCS have been defined and BLOC-1 contains eight subunits, several of which correspond to proteins mutated in the mouse hair color mutants “cappuccino”, “muted”, and “sandy”. The “sandy” gene product is known as “dysbindin”, a protein that has been reported to interact with and proposed to exist as a complex with “dystrobrevin”, which itself is deficient in Duchenne and Becker muscular dystrophies.
Dysbindin is encoded by a gene (DTNBP1) whose polymorphisms in non-coding regions appear to affect expression levels of the protein, and in epidemiological studies dysbindin has been associated with susceptibility to schizophrenia. Dell’Angelica showed that, as had been reported, dysbindin can interact with dystrobrevin in vitro, as well as in a yeast two hybrid system, and in transfected cells co-expressing both proteins. On the other hand, in both native muscle and brain tissue, he found that the association of the two proteins was negligible. He concluded that the interactions previously reported result from the artifactual association of coiled coil regions of the proteins, whereas in natural systems dysbindin binds only to the multiple subunits of BLOC1, via the same regions that could artifactually bind it to dystrobrevin. Moreover, he concluded that dysbindin within the BLOC1 complex plays no role in the formation of the so-called Dystrophin-Glycoprotein Complex (DGC), which includes dystrobrevin and dystrophin and is impaired in muscular dystrophies.
It is not quite clear if these observations in any way weaken the validity of the reported association of dystrophin polymorphisms with a predisposition to schizophrenia.
Dell’Angelica mentioned that a new HPS form, HPS7, was recently identified in a Portuguese patient with an in frame deletion in DTNBP1, the gene that encodes dysbindin.

Fredéric Saudou (Institut Curie, Orsay, France)
Huntington’s disease (HD) is a fatal neurodegenerative disorder that affects 1 in 10000 individuals of European origin. It is characterized by involuntary movements, personality changes and dementia. The neuropathology of HD involves neuronal dysfunction and the selective death of striatal neurons in the brain. The mutation that causes disease is an abnormal expansion of a polyglutamine (polyQ) stretch in the N-terminus of the 350 kD protein huntingtin. The mechanisms by which huntingtin induces dysfunction and death of neurons in the brain are not clearly understood. These could involve the gain of a new toxic function as well as the loss of the beneficial activities intrinsic to wild type huntingtin. Indeed, huntingtin possesses anti-apoptotic properties as observed in cell culture and animal models.
F. Saudou and coll. have used 3D fast video microscopy techniques to study the intracellular dynamics in normal and pathological situations. Using this approach they have unravelled a function of huntingtin in the microtubule-based transport of neurotrophic factors such as BDNF. In the pathological situation, huntingtin-stimulated BDNF transport is altered. PolyQ-huntingtin via the huntingtin-associated protein-1, HAP1 disrupts the association of key components of the motor machinery to microtubules. Reduced BDNF transport leads to a decrease in neurotrophic support and to neurotoxicity that are both rescued by wild-type huntingtin. These results demonstrate that the anti-apoptotic properties of huntingtin are linked to the ability of huntingtin to promote transport of BDNF in the brain.
There is so far no treatment for HD. However, one of the most promising candidate drugs for HD is cystamine, a compound described as an in vitro transglutaminase (TGase) inhibitor. F. Saudou presented recent results demonstrating that cystamine positively regulates BDNF release by inhibiting TGase. In addition, cysteamine, the FDA-approved reduced form of cystamine, is neuroprotective in HD mice by enhancing BDNF levels in brain.

Ralph Nixon (NYU School of Medicine, Orangeburg, USA)
Nixon has had a long standing interest in Alzheimer’s Disease (AD) and the cellular events that lead to the development of amyloid plaques, intracellular neurofibrilary tangles, and ultimately to synaptic impairment and neuronal death. His laboratory first observed the presence of enlarged Rab5+ early endosomes in pyramidal neurons of the neocortex in AD patients at very early stages of the disease and he made similar observations in brain samples from Down Syndrome fetuses. He also found increased levels of lysosomal enzymes and in particular cathepsin in those endosomes. The affected neurons also had higher levels of Rab4 and the enlarged endosomes, which contained the amyloid precursor polypeptide (App) also had higher levels of Rabaptin 5 and the early endosome antigen A1 (EEA1). Nixon concluded that both endocytic uptake and recycling are activated in the early stages of pathogenesis of AD and the dysfunction drives the increased production of Aß from internalized App in early endosomes.
The inheritance of the e 4 allele of apolipoprotein E (ApoE4), which is known to promote an early onset of AD, and a more severe amyloid pathology also accentuates the endosomal changes observed at early stages of AD. The endosomal dysfunction was also found to be accompanied by an impairment in the retrograde transport of endosomes containing ligand-bound neurotrophin receptor, which is necessary for neurotrophin signaling at the cell body, and this led to a degeneration of cholinergic neurons which could be reversed by direct delivery of NGF by infusion.
Down Syndrome patients, who have three copies of chromosome 21 and, therefore, and increased doses of the App gene, which resides in that chromosome, are known to develop amyloid plaques and if they live long enough to develop the pathological features and the dimensia of AD. In the mouse chromosome 16 contains many of the genes in human chromosome 21, but the trisomy 16 is lethal in mice. However, a mouse strain with a partial trisomy 16, encompassing ~140 genes, including the App gene, is available and manifests features of Down Syndrome, as well as the enlarged neuronal early endosomes. Nixon and his collaborators have shown that a triplicated App gene is essential for these animals to develop the abnormalities, although overexpression of App alone is not sufficient to produce them. Hence, it seems clear that the triplication of other neighboring genes is also required.
With the mouse model Nixon has shown that the increased doses of App markedly decreases nerve growth factor retrograde transfer and causes degeneration of basal forebrain cholinergic neurons. Indeed, in these animals NGF and App containing early endosomes were also enlarged.
Nixon also discussed the emerging notion that autophagy is a neuroprotective cellular response to damage and that defective autophagy promotes neuronal cell death. Thus, presenilin 1 (PS-1) is required for macroautophagy and PS-1 mutations, which are one of the major causes of early onset AD, impair this phenomenon and cause Aß accumulation.

Thomas Willnow (Max-Delbrueck-Center for Molecular Medicine, Berlin, Germany)
SorLA is a novel neuronal type 1 membrane protein structurally related to intracellular sorting receptors with vacuolar protein 10 protein (Vps10p) domain. The physiological role of SorLA has been unclear so far. However, its ability to shuttle between plasma membrane, endosomes and Golgi compartments suggested a function in neuronal trafficking processes. Because expression of SorLA is absent in the brain of patients with Alzheimer’s disease, T. Willnow and coll. tested involvement of this receptor in the intracellular transport and processing of the amyloid precursor protein (APP), the etiological agent of Alzheimer’s disease. APP follows a complex trafficking pathway that determines processing into the soluble sAPP fragment (non-amyloidogenic pathway) or into the Aß peptide (amyloidogenic pathway). The cellular mechanisms that direct APP into either intracellular processing pathways remain elusive. In a recent study, they demonstrated that SorLA directly interacts with APP in vitro and in cultured cells and that the two proteins colocalize in the endosomal and Golgi compartments. Overexpression of SorLA in cultured neurons results in sequestration of APP in the Golgi and in decreased processing into Aß. The significance of SorLA activity for APP processing in vivo was confirmed in a mouse model genetically deficient for the receptor. Loss of SorLA expression in knockout mice results in increased levels of Aß in the brain similar to the situation seen in patients with Alzheimer’s disease. This suggests that SorLA as a novel sorting receptor for APP that regulates intracellular transport and proteolytic processing of the precursor protein.
T. Willnow proposed a model whereby SorLA protects APP from processing into Aß and thus reduces the burden of amyloidogenic peptide production. Consequently, reduced expression of the receptor in the brain increases Aß production and amyloid plaque formation, and constitutes a major cause of Alzheimer’s disease.

Graca Raposo (Institut Curie, Paris, France)
Graça Raposo focused her research on the organisation of the endocytic pathway in specialized cell types such as immune cells and skin melanocytes. Studies on antigen presenting cells evidenced the exocytic capacity of endocytic multivesicular bodies and the release of their intraluminal vesicles called exosomes. Exosomes may constitute potential vehicles of intercellular communication in health and disease. G. Raposo and coll. have recently shown that transmissible pathogens such as prions make use of the secretory capacities of endosomal organelles and use an exosomal route for their propagation. Cells may thus exploit the nature of endosome-derived exosomes to communicate with each other in normal and pathological situations, providing for a novel route of cell-to-cell communication and therefore of pathogen transmission. These findings open the possibility that methods to interfere with trafficking of such unconventional pathogens could be envisioned from insights on the mechanisms involved in exosome formation, secretion and targeting.
The secretory capacity of endosomal compartments is also exemplified by the existence in specialized cell types of so called lysosome related organelles. To get further insights on the mechanisms underlying the biogenesis of these organelles, G. Raposo and coll. have been investigating in detail the sorting and structural events leading to the generation of melanosomes, the pigment producing organelle of melanocytes. She presented a model highlighting the complexity of the early endocytic system of these specialized cells. Distinct endosomal subdomains are involved in the sorting of melanosomal proteins and in the generation of premelanosomes and mature melanosomes. Several components, including proteins with unknown roles bearing mutations in lysosomal diseases associated with hypopigmentation (Hermansky-Pudlak syndrome) are likely to be involved in the regulation and specialization of the endocytic pathway in melanocytes.

Geneviève de St-Basile (Hôpital Necker, Paris, France)
G. de St-Basile is a specialist of a group of human inherited immune disorders characterized by the onset of an excessive and often fatal response of the immune system, known as hemophagocytic syndrome (HS). HS is the only feature of the various forms of familial hemophagocytic lymphohistiocytosis (FHL), while it is associated with hypopigmentation in Griscelli syndrome (GS) and Chediak-Higashi syndrome (CHS). HS results from a defect in the granule dependent cytotoxic function of lymphocytes. The characterization of the molecular causes leading to these various conditions allowed to identify several effectors critical for the function of the lytic machinery in lymphocytes and for the transport of melanosome in melanocytes. G. de St-Basile et coll. as well as several other groups have shown that myosin Va, Rab27a, and melanophilin form a tripartite protein complex required for the transport of melanosome along microtubules and actin filaments. Defects in each of these proteins define the three genetic forms of Griscelli syndrome (GS1-3, respectively). Rab27a and Munc13-4 are two effectors of the cytotoxic apparatus. The function of Rab27a seems required for the transport and or docking of cytotoxic granules to the plasma membrane and therefore the release of lytic granule contents at the immunological synapse (IS). Defect in Munc13-4 causes FHL by affecting the priming of the cytotoxic granules at the IS, which do not fuse with the plasma membrane and release their contents.
G. de St-Basile presented recent data showing that the cytotoxic function of lymphocytes requires the cooperation of two types of organelle, the lysosomal cytotoxic granule and the endosomal “exocytic vesicle”. hMunc13-4, independently of Rab27a, mediates the assembly of Rab11+ recycling and Rab27+ late endosomal vesicles, constituting a pool of vesicles destined for regulated exocytosis. hMunc13-4 also primes cytotoxic granule fusion, potentially via interaction with active Rab27a. CTL-target cell recognition induced a rapid polarization of both types of organelle which coalesce near the cell-cell contact area.

Peter van der Sluijs (University Medical Center, Utrecht, The Netherlands)
P. van der Sluijs showed that Munc13-4 is a direct partner of Rab27a. The two proteins are highly expressed in CTLs and mast cells where they colocalize on secretory lysosomes. The region comprising the Munc13 Homology Domains is essential for the localization of Munc13-4 to secretory lysosomes and also for the function since the FHL3 mutant Munc13-4delta608-611 fails to localize to lytic granules and does not bind Rab27a due to misfolding. Transfection of Munc13-4 enhanced degranulation, showing that it has a positive regulatory role in secretory lysosome fusion. These results show that the secretion defects seen in GS2 and FHL3 have a common origin and P. van der Sluijs proposed that the Rab27a/Munc13-4 complex is an essential regulator of secretory granule fusion with the plasma membrane in haematopoietic cells.
P. van der Sluijs also presented recent experiments on the dynamics of Rab27a and Munc13-4 using FRET/FLIM. Munc13-4 appears more mobile than Rab27a. The interaction of Rab27a with Munc13-4 leads to a decreased motility of Munc13-4 positive structures. In stimulated CTLs, Munc13-4 dissociates from granules whereas Rab27a remains associated with the membranes of the granules. In resting cells, Rab27a is immobile and behaves like a GTP-hydrolysis deficient mutant.

Mindong Ren (NYU School of Medicine, New-York, USA)
The biogenesis of lysosomal related organelles (LRO) is affected in individuals with the Hermansky-Pudlak Syndrome, a group of recessive disorders characterized by pigmentation and coagulation defects. Ren has chosen the Drosophila eye pigment granules as a model system to study the biogenesis of LRO. Although many genes encode proteins involved in the synthesis of the eye pigments, 11 genes that affect the eye color in the fruitfly, have been recognized to play important roles in the trafficking and sorting processes involved in granule biogenesis and have been designated “granule group” of genes. Four of these genes had been previously identified as encoding the four subunits of the AP3 adaptor complex and three others correspond to the Drosophila orthologues of specific yeast vacuolar protein sorting (Vps) genes. Ren has identified the gene products of the four remaining genes. He initially focused on the “claret” and “lightoid” granule group mutations and was able to identify the “claret” product as a 1961 amino acid protein with domains characteristic of RCC1, a protein that in mammalian cells serves as exchange factor for the small GTPase Ran. Indeed, Ren showed that the “lightoid” gene encodes a Rab protein (RP1) with homology to the mammalian Rabs, Rab 32 and Rab 38. He found that “claret” protein interacts with RP1, as well as with the two RP1 mammalian homologues. However, only Rab 38, which is a melanosomal protein mutated in the “chocolate” mouse pigmentation mutant, can replace RP1 in defective flies and appears to be the true orthologue of RP1.
Using double mutants, Ren showed that “claret” and “lightoid” function in the same LRO biogenetic pathway, which is different from the AP3 dependent pathway. Ren also identified the Drosophila “pink” gene as encoding the fly orthologue of the human HPS5p gene, which is a BLOC-2 component defective in patients with Hermansky-Pudlak Syndrome 5, and identified the Drosophila eye color “purpleoid” mutant as an orthologue of the yeast Vps 34 gene, which encodes a PI3 kinase.
Ren presented a novel Drosophila genetic screen to identify genes that function together with known HPS genes in the generation of LROs. His strategy involves producing phenocopies of the granule group mutants by expressing the corresponding RNAis and using the fly phenocopies to screen for suppressors of the dominant RNAi knockdown in a collection of Drosophila lines that overexpress specific candidate genes. He demonstrated the feasibility of this approach for the Rab-RP1/claret pair, showing that overexpression of RP1 suppresses the phenocopy produced by the RNAi for “claret”, which strengthens the notion that “claret” is indeed an exchange factor for the RP1 Rab protein. Ren has screened 127 candidate genes for suppression of the “lightoid” phenotype and has identified three potential genetic interactors that might serve as downstream effectors of RP1, and he is now investigating the specific roles that these proteins might play in the biogenesis of LROs.