The Hermansky-Pudlak Syndrome 3 (cocoa) protein is a component of the biogenesis of lysosome-related organelles complex-2 (BLOC-2)

3 BLOC-2 is abolished by the 3 amino acid deletion in the Hps6 ru mutant allele, indicating that these three amino acids are important for normal BLOC-2 complex formation.


INTRODUCTION
The cocoa (coa /Hps3), ruby-eye-2 (ru2 /Hps5) and ruby-eye (ru /Hps6) mouse pigment genes encode novel proteins, which regulate the synthesis of lysosome-related organelles including melanosomes and platelet dense granules (1,2). Hps3, Hps5 and Hps6 mutant mice have morphologically abnormal melanosomes and decreased quantities of intragranular components of platelet dense granules (3)(4)(5)(6). Organellar trafficking abnormalities lead in turn to hypopigmentation of both coat and eyes and prolonged bleeding times. All three mutants are appropriate animal models for the inherited human disease Hermansky-Pudlak Syndrome (HPS) (MIM 203300) (7,8), which presents with similar abnormalities of subcellular organelles. Associated clinical symptoms of HPS include loss of visual acuity, prolonged bleeding and lung disease due to abnormalities of melanosomes, platelet dense granules and lysosomes respectively.
Hps3, Hps5 and Hps6 mutant mice are among at least 16 mouse models of HPS (2,5). Human HPS patients with mutations in seven mouse HPS genes by guest on March 24, 2020 http://www.jbc.org/ Downloaded from have been identified (2,7,9). One class of five HPS genes (2) encodes proteins with established functions in vesicle trafficking to lysosome-related organelles in both lower and higher eukaryotes. In contrast, the second class of nine genes (2,10,11), which includes the Hps3, Hps5 and Hps6 genes of this report, are expressed only in higher eukaryotes and encode novel proteins with no recognizable structural motifs and whose functions are unknown. Most recently (9) the sandy (sdy/Hps7/Dtnbp1) gene was identified as encoding dysbindin, a dystrobrevin interacting protein (12).
The Hps5 and Hps6 proteins directly interact in a multiprotein complex termed biogenesis of lysosome organelles complex-2 (BLOC-2) (2). Hps3 mice have coat color (13) similar to that of Hps5 and Hps6 mutants which in turn are mimic mutants regarding coat and eye colors (2,14). These facts suggested that the function(s) of the Hps3, Hps5 and Hps6 genes are related and that they might be residents of a common protein complex. To test this hypothesis and to better understand the novel proteins of the BLOC-2 complex, we tested for epistatic interactions of the Hps3, Hps5 and Hps6 genes in doubly mutant mice and for complex formation by their protein products.

Mice
Mutant mice together with normal C57BL/6J mice were obtained from The Jackson Laboratory (Bar Harbor, ME). Mice were subsequently bred and maintained in the animal facilities of Roswell Park Cancer Institute. Unless indicated otherwise, the particular alleles utilized in these studies are as follows.
The Hps3 coa allele contains a splice site mutation resulting in a frameshift and loss of expression of the Hps3 mRNA (1); the Hps5 ru-2J allele contains a frameshift mutation which causes loss of the C-terminal third of the Hps5 protein (2); the Hps6 ru allele contains a small in-frame deletion which results in loss of 3 amino acids at positions 187-189 (2). The Hps6 ru-6J mutation contains a 5.3 kb IAP element insertion which causes loss of transcript expression (2). The Hps3 coa mutation arose and is maintained on the C57BL/10J background (15). Both the Hps5 ru-2J and Hps6 ru mutations arose on the C3H inbred strain background and were subsequently transferred to and maintained as congenic mutants on the C57BL/6J inbred strain background. All mice utilized in these experiments were 2-5 months old. All procedures (mouse protocol 125M) were reviewed and approved by the Roswell Park Institutional Animal Care and Use Committee and adhered to the principles of the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Construction of double mutant mice
Heterozygous F 1 offspring (Hps3/+ and Hps6/+ or Hps5/+ and Hps6/+) were produced by mating of Hps3/Hps3 and Hps6/Hps6 or Hps5/Hps5 and Hps6/Hps6 mice and were, as expected, of normal black coat and eye color. F 1 offspring were mated to produce an F 2 generation. F 2 mice doubly homozygous (Hps3/Hps3,Hps6/Hps6 or Hps5/Hps5,Hps6/Hps6) for mutant genes were verified by molecular diagnoses of genotype at each gene by PCR amplification and sequencing of normal and genomic tail DNA using appropriate primers (1,2) (sequences available upon request). Since double mutants appeared healthy and had no obvious reductions in breeding efficiency, they were mated among themselves to maintain the double mutant colonies. Since all single and double mutants are on the C57BL/6J strain background (or on the closely related C57BL/10J strain background in the case of the Hps3 mutant), contributions of background genes are essentially identical in all.

Antibodies
The peptide sequence, CNQERRGKPERIHVSSE, located near the amino-terminus of the Hps5 protein (2) was conjugated to carrier protein KLH, and a polyclonal antiserum was prepared in rabbits by Covance, Denver, PA .
To prepare antisera to the Hps6 protein, an expression plasmid pET 15b Rabbits were initially injected with 250 µg of purified His-Ruby protein followed by boosting with 125 µg biweekly before final collection of antisera at 100 days.

Platelet collection and platelet serotonin analyses
Platelets were harvested from the peripheral blood of normal and mutant mice in the presence of sodium citrate (17). Washed platelets were lysed in 1 ml distilled water, counted in a Coulter Z2 particle count and size analyzer and assayed fluorometrically for serotonin (17). confirmed that blots were exposed within a linear range.

Yeast Two-hybrid Analyses
The Matchmaker GAL4 Two-Hybrid System 3 kit (Clontech) for two hybrid analyses was used at low and high stringency as described .

Electron microscopy
Eyes were fixed in glutaraldehyde, postfixed in osmium tetroxide, and embedded in spur resin as described (1) before viewing on a Siemens 101 Electron microscope at an accelerating voltage of 80 kV.

Coimmunoprecipitation
Open reading frames of Hps3, Hps5, Hps6, Hps7, Ap3b1 and pa cDNAs were fused in-frame to pCMV-Tag vectors (Myc and Flag), and the resulting Horseradish peroxidase-linked donkey antibody against rabbit immunoglobulin G (1:5000 dilution; Amersham Life Sciences) was used as a secondary antibody for Flag blots, and horseradish peroxidase-linked bovine antibody against goat immunoglobulin G (1:5000 dilution; Santa Cruz Biotechnology) was used as a secondary antibody for Myc blots. Blots were treated with the enhanced chemiluminescence reagent (ECL +plus; Amersham, Piscataway, NJ) and exposed for 1 min.

Size-exclusion Chromatography and Sedimentation Velocity Analysis
Cytosolic extracts from the liver of normal and mutant mice were prepared , using a Dounce homogenizer, followed by centrifugation at 5000 ✕ g for 5 min and then at 120,000 ✕ g for 90 min, at 4ºC. Size-exclusion chromatography was performed as described (18). Sedimentation velocity analysis was carried out by ultracentrifugation of cytosolic extract (0.2 ml, ~5 mg total protein) loaded on top of a linear 5-20% (w/v) sucrose gradient prepared in Detergent-free Tris Buffer (12 ml), for 13 h at 39,000 rpm on a SW41 rotor (Beckman Coulter).
Fractions were collected from the bottom of the tube. Fractions resulting from both the size-exclusion chromatography and sedimentation velocity experiments were analyzed by immunoblotting using antibodies to Hps5 and Hps6 proteins.
Cytosolic liver extracts from Hps5 ru-2J and Hps6 ru-6J null mutant mice were analyzed in parallel to confirm the identity of the Hps5 and Hps6 protein bands, respectively.

In vivo physiological interactions of Hps3, Hps5 and Hps6 proteins in the production of lysosome-related organelles in doubly mutant mice
To test for interactions between a) the Hps5 and Hps6 gene products and b) the Hps3 and Hps6 gene products at the physiological level, appropriate double mutant mice (i.e. homozygous for mutant genes at two of these HPS loci) were bred, verified (see Methods) and analyzed for abnormalities of melanosomes and platelet dense granules, the lysosome-related organelles most severely affected in Hps5 and Hps6 mutants.
The coat and eye colors of Hps5/Hps5,Hps6/Hps6 double mutants are hypopigmented in comparison to C57BL/6J controls and identical (Fig. 1a) to those of the Hps5/Hps5 and Hps6/Hps6 single mutants (which themselves exhibit [Fig 1a] mimic phenotypes) suggesting a common abnormality in melanosomes, the subcellular organelle which imparts coat and eye coloration.
In both single and double mutants, coat colors are the classical (14) ruby color.
Eye color in all is a light pink in offspring less than 1 week of age (not shown).
This deepens in adults to a dark ruby eye color, distinguishable from the black eyes of normal C57BL/6J mice only when closely observed with intense light.
The detailed mimicry in pigmentation and melanosomal properties of the  Combined, these several mimic effects on melanosomes and platelet dense granules suggest that the Hps3, Hps5 and Hps6 genes regulate the synthesis of lysosome-related organelles by a common mechanism at the physiological level.

Test for destabilization of HPS5 and HPS6 proteins in extracts of other Hps mutants
The above mimic effects of the Hps3, Hps5 and Hps6 genes suggested possible co-residence of their protein products within a common protein complex.
An indication of residence of two proteins within a common protein complex is destabilization of the partner protein within cells derived from mutants lacking one of the proteins, as loss of one member of a protein complex often leads to destabilization of other members of that complex (18,20). Accordingly, polyclonal antibodies to the Hps5 and Hps6 proteins were produced, and levels of these proteins were analyzed by western blotting in tissues of fourteen mouse HPS mutants and alleles to determine if mutations in other HPS genes affected their concentrations (Fig. 3). Consistent with its residence, together with the Hps6 protein, within the BLOC-2 complex (2), the Hps5 protein exhibits destabilization within spleen and lung extracts of the Hps6 ru-6J null mutant (Fig. 3). Significant destabilization of Hps5 protein is also apparent in extracts of the Hps3 null mutant. Its concentration is also, as expected, depressed in extracts of single and double mutants containing the Hps5 ru2-J allele, which have undetectable Hps5 protein levels, an expected result given the frameshift null mutation within this allele (2). Similar results were observed in heart extracts (not shown). No significant destabilization was apparent in any of the remaining 11 HPS mutants or the misty hypopigmentation mutant.
Similarly, in regard to possible residence within a common protein complex, there was a notable loss of Hps6 protein within brain and lung of the Hps3 (coa) and Hps5 mutants (Fig. 3). The latter result is consistent with evidence that both Hps5 and Hps6 proteins reside within the BLOC-2 complex (2). As expected, levels of the Hps6 protein were undetectable (Fig. 3)

Coimmunoprecipitation of the Hps3, Hps5 and Hps6 proteins
To directly test for co-residence of the Hps3, Hps5 and Hps6 proteins within a common complex, epitope-tagged constructs of each gene were expressed within transfected cells, and immunoprecipitates were tested with appropriate antibodies (Fig. 4). Immunoblots of FLAG precipitates analyzed with the Myc antibody demonstrated that the Hps3, Hps5 and Hps6 proteins coprecipitated in all combinations, but did not interact with either the BLOC-1 proteins Hps7 (dysbindin) (9) or the Ap3b1 subunit of the AP-3 adaptor complex.
A second method of detection of interacting proteins, the yeast two-hybrid approach (Fig. 5) revealed the previously reported (2) interaction of the Hps5 and Hps6 proteins. There was, however, no direct interaction of the Hps3 protein with either the Hps5 or Hps6 proteins (Fig. 5), suggesting that additional proteins bridging Hps3 with Hps5 and Hps6 are present in BLOC-2.

Size-exclusion chromatography and Sedimentation Velocity Analyses
The Hps5 and Hps6 proteins were found in common fractions in both size-exclusion chromatography (Fig. 6a) and sedimentation velocity (Fig. 6b) analyses of cytosolic liver extracts of C57BL/6J, consistent with coresidence within BLOC-2. The gel filtration results indicate that BLOC-2 has a Stoke's radius of 98 +/-5 Angstroms, and the sedimentation coefficient, from the sucrose gradient is 8.3 +/-0.5 S. It is an asymmetric complex with a frictional ratio (f/fo) of ~2. The calculated molecular mass of the complex is 350 kDa +/-60 kDa.

An Hps6 mutation (Hps6 ru ) abolishes interaction of the Hps5 and Hps6 proteins
Previous studies (2)  Nevertheless, this small alteration produces the abnormalities of lysosomerelated organelles typical of the HPS phenotype. Therefore, we tested the hypothesis that this 3 amino acid region of the Hps6 protein is functionally important in interacting with the Hps5 protein. Consistent with this possibility, the Hps6 ru allele causes no loss of expression of the Hps6 mRNA (2) and maintains significant expression of the Hps6 protein (Fig. 3). Indeed, yeast two-hybrid analyses (Fig. 5) indicated no interaction of the Hps5 and Hps6 ru proteins.
Interaction between the Hps5 and Hps6 ru proteins was undetectable even after testing for long periods (7 days) of yeast growth. To determine if a particular residue among these three amino acids was critical for Hps5/Hps6 interaction, corresponding constructs in which only one of these three amino acids in the wild type Hps6 protein was substituted with alanine were tested. Each of these Hps6 mutant proteins (Fig. 5) retained full ability to interact with the Hps5 protein.
Together these results suggest that the histidine-cysteine-proline motif indirectly mediates interaction between the Hps5 and Hps6 proteins. Loss of these 3 amino acids likely produces secondary structural alterations within the HPS6 protein to abolish its interaction with the HPS5 protein within BLOC-2.
The findings that the Hps3, Hps5 and Hps6 proteins are members of a common BLOC-2 complex provide potentially useful information for future therapeutic interventions. A therapy stabilizing the BLOC-2 complex or correcting abnormal BLOC-2 function in any one of the HPS-3, HPS-5 or HPS-6 syndromes may likewise be applicable to the remaining two.

Fig. 1. Mimicry of coat and eye hypo-pigmentation in (A) single and double Hps5
and Hps6 mutants and (B) single and double Hps3 and Hps6 mutants.    (38). Spleen extracts were utilized for Hps5 protein expression as it is expressed at very low levels in brain.

B.
Lung extracts (20 µg protein) of Hps3, Hps5 and Hps6 mutants were analyzed with Hps5 (above) and Hps6 (below) antiserum.   Size-exclusion chromatography. Liver cytosol from C57BL/6J mice was fractionated on a Superose 6 column as described (18). Fractions were analyzed for the presence of