Linking Albinism and Immunity: The Secrets of Secretory Lysosomes

See allHide authors and affiliations

Science  02 Jul 2004:
Vol. 305, Issue 5680, pp. 55-59
DOI: 10.1126/science.1095291


Lysosomes are membrane-bound organelles that are found in all mammalian cells and contain hydrolases and lipases required for protein and membrane degradation. In many cells of the immune system, lysosomes also contain secretory proteins that can be released by regulated exocytosis in response to an external stimulus, providing different cell types with a wide range of effector functions. Melanosomes also use a lysosome-related organelle to secrete melanin for pigmentation. Links between albinism and immunity in patients have uncovered a number of key proteins required for lysosomal secretion and have revealed a versatile secretory mechanism that can be fine-tuned by distinct interactions in different cell types.

The term “lysosome” was coined to convey the idea of a membrane-bound lytic organelle that contains hydrolases active at acid pH within cells. Lysosomes are thought to be the endpoint of the endocytic pathway to which proteins and extracellular particles are delivered for degradation by a number of proteases and lipases (1). However, several studies, such as those in Tetrahymena (2), hinted that lysosomes might also function as secretory organelles. Recent data has revealed that both the degradative and secretory functions of lysosomes can be finely controlled (3), providing important regulatory mechanisms in the immune system and a number of other cell types. It now emerges that the use of lysosomes as secretory organelles may be even more widespread; endocytic compartments have been identified in repairing membranes in fibroblasts (4), facilitating entry of trypanosomes into cells (5), and secreting viruses [including human immunodeficiency virus (HIV)] (6), and potentially in secreting morphogen gradients in Drosophila (7, 8). What enables lysosomes to take on the dual role of secretory granule and degradative organelle? The molecular details of the mechanisms controlling lysosomal secretion are now beginning to emerge.

Secretory lysosomes are modified lysosomes that can undergo regulated secretion in response to external stimuli. Secretory lysosomes are found in many different cell types of the immune system (9) and contain specialized secretory proteins required for the specialized function of that cell type in addition to lysosomal hydrolases required for protein degradation. Cytotoxic T lymphocytes (CTL) and natural killer (NK) cells, for example, secrete the pore-forming protein perforin required to initiate cell death by means of secretory lysosomes, whereas mast cells secrete the inflammatory mediator serotonin from modified lysosomes. Morphological studies reveal the colocalization of lysosomal hydrolases and secretory proteins in many cells of the immune system, whereas biochemical and functional studies point to an important role for pH in controlling the dual functions of these organelles. Lysosomal hydrolases function at acidic pH, whereas secretory proteins, such as perforin, function at neutral pH after secretion.

Steps Leading to Lysosomal Secretion

Lysosomes move back and forth along microtubules by means of kinesin- and dynein-based motors, but they often cluster around the microtubule organizing center (MTOC) (10). Lysosomal secretion generally occurs at the plasma membrane, although the fusion between lysosomes and phagosomes in phagocytic cells can be regarded as an intracellular secretory event. In CTL, lysosomal secretion is triggered by recognition of a target cell by means of the T cell receptor. This triggers formation of the immunological synapse, a distinct topological rearrangement of cell surface proteins formed by a ring of adhesion proteins (leukocyte function associated antigen–1 and talin) surrounding a central domain containing a patch of signaling proteins and a distinct secretory domain in which granule exocytosis takes place (11). Within CTL, the MTOC moves from a perinuclear region to the contact site, repolarizing the microtubule network toward the target cell (Fig. 1). Several other organelles also polarize with the MTOC, including the associated Golgi complex and mitochondria. Secretory lysosomes move in a minus direction along the microtubules toward the MTOC, accumulating at the immunological synapse where fusion occurs within the secretory domain. Released material, including partially degraded membranes from the lysosome, is occasionally seen within the cleft formed between CTL and the target by electron microscopy (EM) (12).

Fig. 1.

Secretory lysosomes in wild-type CTL immunofluorescence and illustrations showing the distribution of secretory lysosomes (cathepsin D, green) on microtubules (red) in human CTL (A). Secretory lysosomes move along microtubules (B) to the MTOC, which migrates to the immunological synapse (C). Each picture is a composite of multiple z-series confocal images. (D) Electron micrograph showing secretory lysosomes (SL), mitochondria (m), and Golgi (G) polarized near to the MTOC (*) at the immunological synapse in wild-type CTL. Arrowheads indicate the area of membrane contact between CTL and the target. Scale bar, 1 μm.

The use of lysosome-related organelles for secretion is not restricted to cells of the immune system. Melanosomes are also lysosomal-related organelles, serving to release melanin to keratinocytes and produce pigmentation (13). One important difference between melanocytes and CTL is that, in melanocytes, the MTOC remains perinuclear and melanosomes move in a plus direction along microtubules to reach the plasma membrane before secretion.

Diseases of Secretory Lysosomes

The finding that melanosomes are lysosome-related organelles involved in the secretion of the melanin required for skin and hair pigmentation has provided some powerful insights into the mechanisms controlling lysosomal secretion. A number of autosomal human genetic diseases give rise to defects in both pigmentation and immune dysfunction, suggesting that melanocytes and immune cells might use a common secretory machinery, shared by cells with secretory lysosomes. Importantly, patients with diseases combining immunodeficiency and albinism, or their mouse models, show no primary neurological defects, suggesting that conventional secretion is intact and that the defective proteins are required specifically for lysosomal secretion (Table 1). These diseases not only support the idea that specialized mechanisms might be required for secretion of lysosome-related organel-les, but they have helped in identifying these proteins by mapping the defective genes.

Table 1.

Diseases of secretory lysosomes. Human diseases and their mouse models giving rise to defects in secretory lysosomes are shown. The defective gene and the cell types affected by loss of this protein are given, as well as the protein complexes involved.

Mouse mutantDefective geneDefective cellsComplex
Chediak-Higashi syndrome
beige LYST Immune cells and melanocytes Unknown
Griscelli syndrome 2
ashen RAB27A Immune cells and melanocytes Rab27a/MyoV/melanophilin complex in melanocytes only
Elejalde syndrome (Griscelli 1)
dilute MYOV Melanocytes and neurons Rab27a/MyoV/Melanophilin
Griscelli syndrome 3
leaden MLPH Melanocytes Rab27a/MyoV/melanophilin
Hermansky-Podlak syndrome
gunmetal RABGGTA Immune cells, melanocytes, and platelets Transient with Rabs
pale ear HPS1 Melanocytes and platelets BLOC-3
light hear HPS4 Melanocytes and platelets BLOC-3
pearl AP3B1(HPS2) Immune cells, melanocytes, and platelets AP-3
mocha AP3D1 Melanocytes and platelets AP-3
cocoa HPS3 Melanocytes and platelets BLOC-2
ruby-eye2 HPS5 Melanocytes and platelets BLOC-2
ruby-eye HPS6 Melanocytes and platelets BLOC-2
sandy DTNBP1(HPS7) Melanocytes and platelets BLOC-1
pallid PLDN Melanocytes and platelets BLOC-1
muted MU Melanocytes and platelets BLOC-1
cappuccino CNO Melanocytes and platelets BLOC-1
buff VPS33A Melanocytes and platelets Unknown

Albinism and Immunodeficiency

The first gene to be mapped came from patients with Chediak-Higashi syndrome (CHS) and its mouse model, the beige mouse (14). Both patients and mice are characterized by a marked hypopigmentation and the presence of enormous lysosomes in all cell types. However, the enlarged lysosomes function well as degradative organelles and the majority of cell types function normally. The phenotype is restricted to immune cells and melanocytes, all of which depend upon secretion of a lysosomal organelle for proper function. The Chediak gene encodes a previously unknown cytosolic protein of 429 amino acids, with little homology to known proteins (15, 16). The enlarged lysosomes seem to be the result of too much membrane fusion or insufficient membrane fission during lysosomal biogenesis, but the CTL defect arises because of a loss of secretion at the immunological synapse (17). These paradoxical results are best explained by the CHS protein, known as Lyst, playing a role in regulating membrane fusion (18).

The second gene to be cloned from an autosomal recessive disease combining immunodeficiency and albinism was RAB27A, from patients with Griscelli syndrome (19) and the ashen mouse (20). Like CHS, both patients and ashen mice exhibit a marked hypopigmentation, but unlike CHS, lysosomes are normal in size. In CTL and melanocytes, Rab27a is required at a late stage of secretion in order to leave the microtubule cytoskeleton and dock at the plasma membrane (Fig. 2) (12, 2123). However, the precise function of Rab27a differs in melanocytes and CTL. Rab27a associates with the melanosomal membrane, and in melanocytes recruits melanophilin, a synaptotagmin-like protein, which in turn interacts with myosin Va, an unconventional myosin motor that moves along the actin cytoskeleton and tethers the melanosome at the plasma membrane ready for secretion (24, 25).

Fig. 2.

CTL from patients with genetic defects. EM and illustrations show secretory lysosomes (SL) in CTL from patients lacking AP3B1, RAB27A, LYST, or MUNC13-4, conjugated with target cells. *, MTOC. Scale bar, 1 μm.

In CTL, Rab27a does not interact with either melanophilin or myosin Va. Myosin Va–deficient patients exhibit severe neurological problems as well as albinism, but no immunodeficiency (26, 27). Patients lacking melanophilin show the characteristic albinism resulting from melanocyte dysfunction, but no immune or neurological defects (28). CTL lacking myosin Va or melanophilin function normally (29). These results demonstrate a central role for Rab27a in secretory lysosome release from both melanocytes and CTL. However, they also show that Rab27a is interacting with different effector proteins in different cell types.

Although CHS and Griscelli syndrome patients show grossly impaired CTL and NK activity, this in itself is not the most serious aspect of the disease. The fatal “acute” phase is characterized by hemophagocytic infiltrates of activated T cells and macrophages in many different organs and tissues. When this occurs in the brain, neurological problems result.

Given the strong correlation between immunodeficiency and albinism, a third group of patients and their mouse models presented themselves as interesting candidates for understanding the mechanisms regulating secretory lysosome release. Hermansky-Pudlak syndrome (HPS) is a disease characterized by partial albinism and bleeding disorders resulting from platelet dysfunction (30). Seven separate genes, termed HPS1 to HPS7, have been identified in humans (31, 32), and 16 separate loci have been identified in mice (33), many of which have not yet been identified in humans. The proteins encoded by several of the HPS genes form complexes, termed biogenesis of lysosomal-related organelle complexes (BLOCs). Although loss in any one of the BLOCs gives rise to defective platelet and melanosome function, it is not clear whether these complexes play a role in lysosomal secretion. Patients lacking components of BLOCs 1 to 3 do not appear to be immunodeficient (34), and CTL derived from mice lacking components of these BLOCs exhibit normal activity (35). These results indicate that although some proteins are involved in secretion from both melanocytes and immune cells, the exact composition of the secretory machinery varies in different cell types.

Two genes that give rise to HPS do appear to control lysosomal secretion in CTL as well as in melanocytes and platelets. The first mutation has so far only been found in gunmetal mice and results in decreased function (but not complete absence) of Rab geranylgeranyl transferase (RGGT), a protein that controls the prenylation, and hence membrane association, of a number of Rab proteins, including Rab27a (36). Secretion from gunmetal-derived CTL is reduced, but not ablated as in CTL lacking Rab27a, suggesting that sufficient Rab27a is prenylated for some secretion to occur (12). However, although secretory lysosomes in Rab27a-deficient CTL cluster tightly at the MTOC, many of the secretory lysosomes from gunmetal CTL fail to polarize and are left around the periphery of the cells (12). This implies that other Rab proteins, such as Rab7, which is involved in microtubule-mediated movement of lysosomes (37), may require RGGT activity and play a role in CTL secretion. The pigmentary and platelet defects of these mice may also involve Rab proteins in addition to Rab27a.

HPS2 patients lack the beta subunit of the lysosomal adaptor protein, AP3 (38). Patients with HPS2 are not only characterized by bleeding disorders and partial albinism, but also show signs of immunodeficiency (38). AP-3 is involved in the lysosomal sorting of proteins from an early endosomal compartment and is expressed in all cell types (39, 40). The selective defects in the immune system and melanocytes suggest that AP-3 plays a critical role in secretory lysosome function in these cell types. Loss of pigmentation may be entirely explained by the essential role that AP-3 plays in sorting tyrosinase (41), a protein required in the synthesis of melanin, to melanosomes. The role of AP-3 in controlling secretory lysosome release in CTL is different. CTL killing is severely impaired because of the loss of secretory lysosome polarization. Although the MTOC, Golgi complex, and mitochondria of AP-3–deficient CTL all polarize toward the immunological synapse, the secretory lysosomes remain in the periphery of the cell, apparently stuck at the ends of microtubules (42) (Fig. 2). Because AP-3 is localized to endosomal structures (40) and not to the secretory lysosomes, AP-3 is most likely crucial for sorting a protein required for lysosomal polarization in CTL. In addition, the secretory lysosomes of AP-3–deficient CTL and platelets (43) are greatly enlarged and more vesiculated. The reasons for this are not fully understood.

Immunodeficiency Without Albinism

A major advance in the understanding of Familial Hemophagocytic Lymphochistiocytosis (FHL) came when 30% of patients were found to lack perforin, the central protein of CTL-mediated killing (44). Although FHL patients are immunodeficient, they do not share the albinism of CHS and Griscelli syndrome patients. The finding that the loss of perforin results in FHL suggests a critical role for perforin in immune homeostasis (45) and explains that CHS and Griscelli syndrome might also result in hemophagocytosis because the defect in each of these diseases prevents the release of perforin-containing granules. It also raised the possibility that other FHL patients might lack proteins required for the delivery of perforin from the secretory lysosomes of CTL.

The picture that has emerged reveals that in addition to central proteins such as Rab27a, the exocytosis of secretory lysosomes involves additional proteins, which are required in CTL but not melanocytes. Munc13-4 protein, a homolog of the Munc13-1 protein that is required for secretion from neurons, is required for secretion from CTL and is defective in a group of patients with FHL (46). In these patients, the secretory lysosomes of CTL are able to dock at the plasma membrane but cannot secrete their contents. The lack of primary neurological disorders in these patients demonstrates that although Munc13-4 is required for CTL secretion, neuronal secretion is unaffected.

Munc proteins are mammalian homologs of the uncoordinated (UNC) proteins, identified initially in mutants of Caenorhabditis elegans (47). Munc proteins regulate exocytosis (48) by controlling the formation of the soluble N-ethylmaleimide–sensitive fusion factor attachment protein receptor (SNARE) complex required for membrane fusion (49). In mammalian neurons, Munc18-1 binds selectively to the closed form of syntaxin 1, a SNARE protein found on the target membrane (t-SNARE), preventing interaction with SNAREs on the vesicle membrane (v-SNAREs) and formation of the complex that mediates membrane fusion. Syntaxin 1 is released from Munc18-1 by interaction with two large proteins localized on the presynaptic membrane: Munc13-1 and RIM (Rab3-interacting protein, effector of Rab3), allowing syntaxin 1 to adopt the open or active conformation and initiate membrane fusion. In this way, Munc13-1 regulates the state of “priming” by binding syntaxin 1 and RIM.

The finding that the loss of Munc13-4, a nonneuronal homolog of Munc13-1, prevents exocytosis of secretory lysosomes suggested that it might play a similar role in priming for secretion in CTL. Munc13-4 is a cytosolic protein which, by immunofluorescence, localizes around the secretory lysosomes in CTL. When Munc13-4 in CTL is lost, vesicles dock at the membrane in the immunological synapse but are not released (46). The picture is very similar to that seen in neuronal cells lacking Munc13-1, in which case the synaptic vesicles are docked at the neuronal synapse but not secreted (50).

Munc13-4 plays a similar role in platelets. Munc13-4 regulates Ca2+-mediated release of alpha granules in platelets by binding directly with the active, guanosine 5′-triphosphate–bound form of Rab27a (51). In permeabilized platelets, Munc13-4 enhances serotonin secretion from platelets. Thus, in platelets (and possibly also CTL) Munc13-4 binds a Rab directly, whereas in neurons Munc13-1 binds Rab3A by means of an interaction with RIM. In platelets, Munc13-4 associates with the plasma membrane, whereas Rab27a associates with the secretory lysosomes, suggesting that Munc13-4 might mark the target site for docking. This contrasts with colocalization of Rab27a and Munc13-4 on the secretory lysosomes of CTL (46).

CTL lacking Rab27a contain secretory lysosomes of normal size and morphology, which appear to polarize toward the MTOC normally (Fig. 2). However, EM reveals that secretory lysosomes remain aligned along the microtubules leading to the MTOC, one behind the other. Secretory lysosomes in Rab27a-deficient CTL are unable to dock at the plasma membrane, and together these observations suggest that Rab27a is required to detach from microtubules before they can dock at the plasma membrane.

Secretory lysosomes in CTL lacking Munc13-4 are able to both polarize to the MTOC and dock at the plasma membrane (Fig. 2). However, whereas docked secretory lysosomes are rarely seen in electron micrographs of wild-type CTL, numerous secretory lysosomes are seen associated with the plasma membrane, supporting a role for Munc13-4 in exocytosis subsequent to docking. Notably, although Munc13-4 and Rab27a are able to interact, the loss of Rab27a disrupts secretion earlier than the loss of Munc13-4. This suggests that Rab27a may interact with other effector proteins in order to detach secretory lysosomes from the microtubules.

The Central Role of Rab27a

One important lesson to emerge from studies on HPS and Griscelli syndrome is that although some key proteins, such as Rab27a, play a role in lysosomal secretion in many cell types, the precise composition of the secretory machinery varies from one cell to another. A broader role for Rab27a is suggested by evidence that Rab27a has been found associated with the chromaffin granules (52) and the secretory granules of insulin-secreting cells (53), as well as the proacrosomal vesicles of spermatids (54). A comprehensive study of Rab27a expression, which used the Rab27a promoter to control expression of a Rab27a–green fluorescent protein transgene, revealed expression of Rab27a in goblet and other intestinal cells, as well as exocrine and endocrine secretory cells in a number of other tissues, including the ovary (55). Patients lacking Rab27a are immunodeficient but lack signs of intestinal or hormonal problems, suggesting that other mechanisms may play a role in secretion from these cell types. Interestingly, the litter number of ashen mice appears to be decreased, suggesting that Rab27a may play a role in exocytic events in the ovary and sperm and function during fertilization (55).

Neuronal-Related Proteins Regulating Lysosomal Secretion

The central role played by Rab27a in lysosomal secretion can be explained by the plethora of effector proteins with which Rab27a can interact. Interactions between Rab27a and several different synaptotagmin-like proteins (Slps) have been demonstrated in vitro (56). Many of these Slps show cell-specific expression, which could allow Rab27a to regulate secretion in a number of cell types. Synaptotagmins form the minimal protein machinery required for calcium-mediated secretion with SNARE complexes (57). The SNARE complex, which drives membrane fusion, has been best characterized for synaptic vesicle exocytosis in nerve terminals with one coil of syntaxin and two coils of synaptosomal-associated protein of 25 kD (SNAP-25), localized on the t-SNAREs, forming a complex with vesicle-associated membrane protein (VAMP), located on the v-SNARE (49). These proteins form a fourhelix bundle coiled-coil core complex that brings together opposite membranes and induces their fusion.

Synaptotagmins have been linked to lysosomal secretion in other cell types. Synaptotagmin VII has been shown to play a crucial role in lysosome-mediated membrane repair in fibroblasts (58), interacting with the ubiquitously expressed SNAP of 23 kD (SNAP-23) and forming a complex including toxin-insensitive (TI)–VAMP/VAMP7, SNAP-23, and syntaxin 4 (59). In contrast, synaptotagmins I and II regulate lysosomal secretion from mast cells (60), and synaptotagmin VIII has been proposed to play a role in acrosomal secretion (61).

Another group of Munc proteins identified as playing a role in neuronal secretion also appear to play a role in secretory lysosome release. Munc18-2 and Munc18-3, two isoforms of the neuronal protein Munc 18-1, are found in nonneuronal cells. Both are expressed in mast cells, and Munc18-2 overexpression inhibits FcϵRI-stimulated exocytosis (62). In resting mast cells, Munc18-2 localizes to the secretory lysosomes, which align along microtubules upon stimulation.

What Makes a Secretory Lysosome?

A comparison of cells that use lysosomal secretion suggests that the core secretory apparatus bears many similarities to neuronal secretion, involving SNARE interactions controlled by Munc and Rab proteins (Fig. 3). The picture which emerges with the identification of the proteins required for secretion in cells with “secretory lysosomes” is that many of these proteins are involved in delivery of these lysosomes or, in the case of Munc13-4, in regulating fusion at the site of secretion. Because the manner of delivery varies between cell types, so do the proteins involved. For example, melanosomes move in a plus direction along microtubules to reach the actin cytoskeleton next to the plasma membrane, whereas in CTL the MTOC polarizes toward the site of secretion and secretory lysosomes move in a minus direction to reach the site of secretion. In contrast to the long-distance microtubule-mediated movement used in CTL and melanocytes, only the membrane-proximal conventional lysosomes are recruited during wound repair in fibroblasts (4), suggesting that enhanced delivery to the site of secretion is a key difference between conventional and secretory lysosomes. Overall, these findings suggest that it is the highly developed delivery and regulatory machinery that allow some cell types to use secretory lysosomes as a major route of regulated secretion.

Fig. 3.

Variations on a theme of lysosomal secretion. Summary of the protein machinery identified in lysosomal secretion in spermatid (acrosomal granule, orange) (54), platelets (7173), fibroblasts (59), CTL (12, 21, 42), melanocytes (19, 20, 2224, 28, 29), and mast cells (74). Secretion of lysosomes (white) is required for secretion in some cell types and can involve movement along microtubules (black) in a plus or minus direction.

It is not clear at what stage during their biogenesis lysosomes become fully capable of secretion. SNARE complex proteins are present on endosomes and lysosomes (63), but not all of these compartments are capable of polarized secretion at the plasma membrane. In CTL, mature lysosomes become polarized during contact with a target cell, but early endosomes do not, suggesting that secretion-competent lysosomes acquire secretory machinery during their biogenesis. Weibel-Palade bodies, the secretory lysosomes of platelets, appear to recruit Rab27a during maturation in a cargo-dependent fashion (64), suggesting that the content of the lysosomes may regulate their secretory potential.

The Versatility of the Secretory Lysosome

The fine details involved in lysosomal secretion in each cell type also play an important role in the proteins required for effective lysosomal secretion in different cells. Mast cells undergo compound secretion, releasing all granules at once, whereas lysosomes from CTL are released in small numbers, or possibly one by one (11). There are yet more mechanisms by which lysosomes can undergo secretion. For example, the lysosomes of dendritic cells can form tubules that extend to the plasma membrane and deliver MHC II to the plasma membrane (6567). Thus, although many cells use lysosomes as secretory organelles, the exact composition of the secretory machinery required varies according to the way in which the lysosomes are delivered to the site of secretion and the way in which secretion occurs. All in all, this provides an enormously versatile secretory organelle. Dendritic cells are able to secrete multivesicular bodies, resulting in the release of small inner vesicles, termed “exosomes” (68). The finding that viruses such as HIV can bud into multivesicular bodies in macrophages (69, 70) and use the endocytic machinery to bud at the cell surface raises the possibility that viruses may also take advantage of secretory lysosomes to spread from cell to cell.

The secretion of endocytic organelles is involved in many diverse roles in many different organisms and pathogens. In Drosophila, endocytic vesicles appear to be used to secrete ligands and transfer them from cell to cell, generating morphogen gradients between cells (7, 8). Complete lipid bilayers, perhaps similar to exosomes, are transferred between cells. This route of secretion from cell to cell also allows for the use of the degradative properties of lysosomes, degrading the signal to create morphogenic gradients (8).


The curious link between albinism and immunity has played a role in identifying many of the proteins involved in lysosomal secretion in melanocytes and different cells of the immune system. The marked defects that result when proteins, such as Rab27a, are lost demonstrate the crucial role of secretory lysosomes in immunity and pigmentation. Understanding the molecular machinery involved in lysosomal secretion also reveals that the basic secretory machinery apparatus is very similar to that observed in neurons, with nonneuronal homologs such as Munc13-4 playing critical roles in some but not all cells that secrete lysosomal organelles. The versatility of the mechanisms leading to lysosomal secretion provides a range of functions from albinism and immunity to morphogenesis and wound repair.

References and Notes

Stay Connected to Science

Navigate This Article