View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

High Frequency of Hermansky–Pudlak Syndrome Type 1 (HPS1) Among Japanese Albinism Patients and Functional Analysis of HPS1 Mutant

Shiro Ito,Ã1 Tamio Suzuki,Ã1 Katsuhiko Inagaki,à Noriyuki Suzuki,à Kenji Takamori,w Tomoko Yamada,z, Mitsuru Nakazawa,y Michihiro Hatano,z Hirotsugu Takiwaki,k Yumi Kakuta,Ãà Richard A. Spritz,ww and Yasushi Tomitaà ÃDepartment of Dermatology, Nagoya University Graduate School of Medicine, Nagoya, Japan; wDepartment of Dermatology, Juntendo University Urayasu Hospital, Urayasu, Japan; zDepartment of Dermatology, Jichi Medical School, Minamikawamachi, Japan; yDepartment of Ophthalmology, Hirosaki University School of Medicine, Hirosaki, Japan; zDepartment of Pediatrics, Saiseikai-Kanagawaken Hospital, Yokohama, Japan; k Department of Dermatology, The University of Tokushima Faculty of Medicine, Tokushima, Japan; ÃÃDepartment of Medicine, Tokyo Women’s Medical University, Tokyo, Japan; wwHuman Medical Genetics Program, University of Colorado Health Sciences Center, Denver, Colorado, USA

Hermansky–Pudlak syndrome (HPS) is an autosomal recessive disorder characterized by oculocutaneous albinism (OCA), bleeding tendency, and lysosomal accumulation of ceroid-like material. Seven genetically distinct subtypes of HPS are known in humans; most are rare outside of Puerto Rico. Here, we describe the analysis of the HPS1 in 24 Japanese OCA patients who lacked mutations in the four known to cause OCA (TYR/OCA1, P/OCA2, TYRP1/OCA3, and MATP/OCA4), and the identification of eight different HPS1 mutations in ten of these patients, four of which were novel (W583X, L668P, 532insC, 1691delA). An IVS5 þ 5G ! A splice consensus mutation was particularly frequent, the result of a founder effect for this allele in Japanese patients. Functional analysis by transfection of the L668P variant into Hps1-mutant melan-ep mouse melanocytes showed that this missense sub- stitution is pathologic, resulting in an Hps-1 protein that is unable to assemble into the biogenesis of lysosome- related organelles complex-3. Key words: albinism/biogenesis of lysosome-related organelles complex/founder effect/haplotype analysis/ Hermansky–Pudlak syndrome J Invest Dermatol 125:715 –720, 2005

Hermansky–Pudlak syndrome (HPS; MIM# 203300) is an (Witkop et al, 1990), where essentially all HPS1 patients are autosomal recessive disorder characterized by oculocuta- homozygous for a 16 bp frameshift duplication at codons neous albinism (OCA), bleeding tendency because of a lack 491–496 (Oh et al, 1996), a result of a founder effect and of platelet-dense granules, and lysosomal accumulation of occult inbreeding (Fukai et al, 1995). Among Japanese pa- ceroid-like material (Spritz, 1999; Huizing et al, 2000; Star- tients, four different HPS1 gene mutations (IVS5 þ 5G ! A, cevic et al, 2002). Life-threatening manifestations are fre- 962insG, 441insA, and 332delC) have been reported (Oh quent, and death typically results from restrictive lung et al, 1996, 1998; Horikawa et al, 2000). Altogether, more disease in patients aged 30–50 y, or in some cases from than 18 different pathological mutations of the HPS1 gene granulomatous colitis, although the clinical course can be have been described (Gahl et al, 1998; Oh et al, 1998; Her- quite variable (Gahl et al, 1998). mos et al, 2002; reviewed in Oetting and King, 1999). Seven genetically distinct subtypes of HPS have been The specific function of the HPS1 protein remains un- defined in humans, resulting from mutations in seven dif- known. It is associated with the product of the HPS4 locus ferent genes: HPS1, ADTB3A, HPS3, HPS4, HPS5, HPS6, in a protein complex named biogenesis of lysosome-related and DTNBP1 (Oh et al, 1996; Dell’Angelica et al, 1999; An- organelles complex (BLOC)-3 (Oh et al, 2000; Chiang et al, ikster et al, 2001; Suzuki et al, 2002; Li et al, 2003; Zhang 2003; Nazarian et al, 2003). Absent expression of the HPS4 et al, 2003). HPS type 1 (HPS1) is the most common, ac- protein destabilizes BLOC-3, resulting in loss of HPS1 as counting for perhaps half of all cases (Oh et al, 1998). Al- well (Chiang et al, 2003). though rare in most populations, HPS1 occurs with a We previously showed that, of 93 unrelated Japanese frequency of 1 per 1800 on the island of Puerto Rico patients with OCA, 24 lacked apparent mutations in the four known OCA genes (TYR/OCA1, P/OCA2, TYRP1/OCA3, and MATP/OCA4) (Tomita et al, 2000; Suzuki et al, 2003; Abbreviations: BLOC, biogenesis of lysosome-related organelles Inagaki et al, 2004). Here, we describe analysis of the HPS1 complex; HPS, Hermansky-Pudlak syndrome; HRP, horseradish gene in these 24 patients and the identification of eight dif- peroxidase; OCA, oculocutaneous albinism; SNP, single nucleotide polymorphism; SSCP, single-strand conformation polymorphism ferent HPS1 mutations, four of which were novel (W583X, 1The first two authors contributed equally to this work. L668P, 532insC, 1691delA). An IVS5 þ 5G ! A splice con-

Copyright r 2005 by The Society for Investigative Dermatology, Inc. 715 716 ITO ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Table I. Mutations and clinical phenotype of Japanese patients with HPS1

Mutation Clinical phenotype

Patient Sex Age Maternal Paternal Hair color Iris color Nystagmus Bleeding time 1 F 3 mo IVS5 þ 5G ! A 1691delA Brown Gray þ ND 2 F 10 mo W583X 974insC Light brown Brown þ 20 min 3 F 6 mo 974insC 288del T Blond Gray þ ND

4 M60yIVS5 þ 5G ! A IVS5 þ 5G ! A Blond Blue þ ND

5a M 2 y IVS5 þ 5G ! A IVS5 þ 5G ! A Blond Blue þ 10 min

6M1mo532insC 1323insA Blond Reddish brown þ 13 min 7F1yL668P IVS5 þ 5G ! A Yellow Blue þ 9.5 min 8 M 39 y (IVS5 þ 5G ! A)b (L668P)b Light brown Blue-brown þ ND 9 M 1.5 mo 974insC IVS5 þ 5G ! A Blond Hazel þ 12 min 10 F 10 mo IVS5 þ 5G ! A IVS5 þ 5G ! A Blond Reddish brown 12 min

Novel mutations are in bold. aA sample of patient 5 was used for the analysis for the IVS5 þ 5G ! A mutation causing aberrant pre-mRNA splicing (Suzuki et al, 2004). bIt was not determined which mutations were of maternal or paternal origin in this patient. The normal value for the bleeding time is o6 min. ND, not done; HPS, Hermansky–Pudlak syndrome. sensus mutation was particularly frequent, a result of a nary fibrosis, confirmed by radiography and high-resolution founder effect for this allele in Japanese patients. Functional computerized tomography of the chest, was found only in analysis of the L668P variant showed that this missense patient 4, who was 60 y old and had severe dyspnea. Pa- substitution is pathologic, lacking the ability to assemble tient 8, who was 39 y old, had no symptoms of pulmonary into the BLOC-3. fibrosis or colitis. The other eight patients were not exam- ined for these complications, as they appeared too young to exhibit symptoms. Results Functional analysis of the L668P substitution by Mutations of the HPS1 gene and phenotypes of the complementation of HPS1-null mouse melanocytes The Japanese patients To assess the occurrence of HPS1 in L668P missense substitution involves an amino acid residue the 24 patients who lacked mutations in the TYR (OCA2), P conserved among most known species, including human, (OCA2), TYRP1 (OCA3), or MATP (OCA4) gene among 93 mouse, rat, Xenopus, Fugu, and Ancylostoma, although unrelated Japanese OCA patients, we screened for muta- divergent in Drosophila (Y) and Mosquito (F). Furthermore, tions in the HPS1 gene. As shown in Table I, among ten of the L668 substitution was not detected in any of the 101 these patients we detected a total of eight different patho- logic mutations (W583X, L668P, 288delT, 532insC, 974insC, 1323insA, 1691delA, and IVS5 þ 5G ! A), four of which were novel (W583X, L668P, 532insC, and 1691delA). Anal- Table II. Frequencies of mutations in Japanese patients with ysis of DNA of 101 unrelated normally pigmented Japanese HPS1 and normally pigmented Japanese subjects individuals detected only one of these variants, IVS5 þ No. of alleles identified in 5G ! A, in a single heterozygous individual (Table II). This IVS5 þ 5G ! A mutation, which we previously showed as Japanese Normally pigmented being pathologic with a sample of patient 5, causing aber- patients with Japanese subjects Mutation HPS1 (n 20) (n 202) rant pre-mRNA splicing with skipping of exon 5, and con- ¼ ¼ sequent frameshift in the resulting HPS1 mRNA (Suzuki IVS5 þ 5G ! A 10 (50%) 1 (0.5%) et al, 2004), was observed in seven of the ten patients, three c.974insC 3 (15%) 0 of whom were homozygous for this mutation (Table I). Thus, c.288delT 1 (5%) 0 among these 93 Japanese patients thought to have OCA, mutations in HPS1 occurred in 11% of the patients. c.1323insA 1 (5%) 0 Clinical phenotypes of these ten patients with HPS1 are p.W583X 1 (5%) 0 summarized in Table I. All of the patients showed some p.L668P 2 (10%) 0 pigmentation, which, although variable, was generally more than in patients with tyrosinase-negative OCA (OCA1A). c.532insC 1 (5%) 0 Prolonged bleeding time was confirmed in six of ten pa- c.1691delA 1 (5%) 0 tients. The other four patients all had clinical evidence of a bleeding diathesis, e.g. purpura on lower limbs or frequent n, number of alleles examined; HPS1, Hermansky–Pudlak syndrome type 1. epistaxis, but their bleeding times were not tested. Pulmo- 125 : 4 OCTOBER 2005 HERMANSKY–PUDLAK SYNDROME 1 IN JAPANESE 717

Japanese control individuals tested. Together, these data Chiang et al (2003) showed that overexpression of hu- indicated that residue L668 was functionally important and man HPS4 protein in melan-le melanocytes restores stabil- that L668P was not a common non-pathological poly- ity of endogenous HPS1 protein, most likely by allowing morphism. assembly of intact BLOC-3. We used an analogous ap- To assess the function of L668P-mutant HPS1 protein proach to test whether the L668P mutant HPS1 protein experimentally, we tested the ability of the wild-type versus could stabilize endogenous HPS4 protein in transfected mutant HPS1 polypeptides to constitute the BLOC-3 pro- melan-ep melanocytes. We transfected cells with cDNA tein complex in mutant mouse melanocytes. We previously expression plasmids encoding either wild-type or L668P showed that cultured melanocytes derived from ho- mutant human HPS1 protein, and analyzed HPS1 and HPS4 mozygous Hps1-null (melan-ep) and Hps4-null (melan-le) protein expression by western blot assay. As shown in Fig 2, mutant mice both lacked the HPS1 protein, the apparent overexpression of normal human HPS1 protein in trans- result of protein destabilization by the absence of either one fected melan-ep cells fully restored the stability of endog- of the complex components (Suzuki et al, 2002). To deter- enous HPS4 protein (lane 1). Overexpression of L668P- mine whether lack of the HPS1 protein likewise results in mutant HPS1 protein, however, failed to restore the stability destabilization of the HPS4 protein, we raised antisera to of endogenous HPS4 (lane 2). This result indicated that the mouse HPS4 protein. As shown in Fig 1, western blot L668P-mutant HPS1 was functionally incapable of assem- analysis showed that both melan-ep and melan-le me- bling with HPS4 in the BLOC-3 protein complex. lanocytes lack detectable 80 kDa HPS4 protein, indicat- ing that the HPS4 protein was destabilized in Hps1-mutant Haplotype analysis As discussed above, we found the (melan-ep) melanocytes, just as the HPS1 protein was de- HPS1 IVS5 þ 5G ! A mutation in seven of the ten Japa- stabilized in Hps4-mutant (melan-le) cells (Suzuki et al, nese HPS1 patients, three of whom were homozygous for 2002). These data thus demonstrated that lack of either the this mutation. This allele thus appears to be common HPS1 or HPS4 protein destabilized the other member of the among Japanese HPS1 patients. To determine whether the BLOC-3. Unexpectedly, we also found that the steady-state multiple occurrences of this allele might derive from a com- quantity of HPS4 protein was reproducibly greater in Hps3- mon founder, we carried out haplotype analysis of patients mutant melan-coa melanocytes in wild-type (melan-a) cells and their parents for five single-nucleotide polymorphisms (Fig 1). (SNP) located within the HPS1 gene (Fig 3). As shown in Table III, all alleles carrying the IVS5 þ 5G ! A mutation appear to share the SNP haplotype C–C–G–G–A. In con- trast, among 101 unrelated normally pigmented Japanese controls, the C–C–G–G–A haplotype occurred with a fre- quency of 6.3%. Thus, the relatively high frequency of the HPS1 IVS5 þ 5G ! A mutation in Japanese patients ap- pears to represent a founder effect for the mutation on this island nation.

Mutational analysis of the HPS4 gene We additionally screened for mutations in the HPS4 gene in the group of 14

Figure 2 Expression of exogenous human Hermansky–Pudlak syndrome type 1 (HPS1) and endogenous HPS4 in melan-ep cells. Figure 1 Melan-ep melanocytes were transfected using pIREShyg3-HPS1nor- Presence of Hermansky–Pudlak syndrome (HSP)4 protein in cul- mal (lane 1), pIREShyg3-HPS1mutantL668P (lane 2), and by pIREShyg3 tured mouse melanocyte lines. Cell extracts of wild-type melan-a as a mock (lane 3). Then the cell extracts were immunoblotted using the (lane 1), Hps3-mutant (melan-coa; lane 2), Hps4-mutant (melan-le; lane antibodies. Top: anti-HPS1 antibody; Middle: anti-HPS4 antibody; bot- 3), and Hps1—mutant (melan-ep; lane 4) melanocytes were immuno- tom: anti-b-actin antibody as a control for protein loading. Exogenous blotted using affinity-purified polyclonal antisera to mouse HPS4 pro- normal and L668P-mutant HPS1 protein were abundantly expressed in tein. The mouse HPS4 protein reproducibly migrates at 80 kDa on transfected cells. Endogenous HPS4 protein, however, was restored SDS-PAGE. Anti-b-actin antibody was used as a control for protein only in melan-ep cells transfected by the normal HPS1 expression loading. plasmid. 718 ITO ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Only patient 4 developed pulmonary fibrosis. His dyspnea was severe. No other patients had symptoms or clinical evidence of either pulmonary fibrosis or colitis. Nev- ertheless, it is inappropriate to conclude that there is a low frequency of these complications among Japanese HPS1 patients, as eight of the ten patients studied here were younger than 2 y of age. The age of onset of pulmonary fibrosis in HPS is typically in the third or fourth decade (Davies and Tuddenham, 1976), and the age of onset of colitis in HPS is typically between 12 and 30 y (Schinella et al, 1980). It will therefore be important to follow these Japanese pediatric HPS1 patients carefully over several Figure 3 decades to assess the occurrence of common HPS com- Mutations of Hermansky–Pudlak syndrome type 1 (HPS1) gene. plications. Five single-nucleotide polymorphisms (SNP) used for the haplotype analysis (upper) and eight mutations of HPS1 gene identified in this In Puerto Rico, HPS is relatively frequent, a specific study (lower). Asterisks indicate novel mutations identified in this study. HPS1 gene mutation having been amplified via the founder effect because of occult inbreeding and genetic drift (Fukai et al, 1995). An analogous situation pertains to the OCA patients who remained undiagnosed. But we found no IVS5 þ 5G ! A mutation that we found in HPS patients pathological mutations of the HPS4 gene, indicating that from Japan, where a founder effect appears to account for a HPS4 may be rare among Japanese OCA patients. high frequency of this mutant allele among patients with OCA. We also identified the 974insC mutation in three of ten Japanese patients with HPS1. This mutation has been pre- Discussion viously found in a number of European patients, with mul- tiple independent mutational origins attributed to a This is a molecular analysis that documents a significant frameshift hotspot associated with a polycytosine tract frequency of HPS among non-Puerto Rican patients with (Oh et al, 1998). Our findings extend this mutation to an OCA. Remarkably, as many as ten of 93 Japanese patients Asian population, most likely derived from at least one ad- with OCA were found to have HPS1, indicating that this ditional independent origin at this mutational hotspot. disorder, previously considered rare, is actually one of the The L668P missense substitution described here is only most common types of OCA among Japanese. Non-pig- the second HPS1 missense substitution reported thus far, mentary features of HPS in these patients seem to be rel- and subjected to functional analysis. The HPS1 protein in- atively mild, and thus attracted little clinical attention, teracts with the HPS4 protein in the BLOC-3 protein com- resulting in these patients’ prior classification as having plex, which is required for the biogenesis of various non-syndromic ‘‘OCA’’. The diagnosis of HPS1 is very im- lysosome-related organelles, including melanosomes. We portant, because it implies significant risk of major clinical have shown that L668P-mutant HPS1 protein, whereas sta- complications that do not occur in non-syndromic OCA, ble, lacks the ability to stabilize the HPS4 protein, suggest- including bleeding, pulmonary fibrosis, and colitis. ing that L668P-HPS1 protein cannot assemble into BLOC-3.

Table III. Identified haplotype with the IVS5 þ 5G ! A mutation

Patient Mutation SNP1 SNP2 SNP3 SNP4 SNP5 1 IVS5 þ 5G ! AC C G G A c.1691delA C A C C G 4 (Homozygote) IVS5 þ 5G ! AC C G G A 5 (Homozygote) IVS5 þ 5G ! AC C G G A 7 IVS5 þ 5G ! AC C G G A p.L668P C C G C A 8 IVS5 þ 5G ! A C C G C/G A p.L668P C C G C/G A 9 IVS5 þ 5G ! AC C G G A c.974insC C A C C G 10 (homozygote) IVS5 þ 5G ! AC C G G A Frequency (%) in normal Japanesea C: 91 C: 49 G: 57 G: 13 A: 57

aTwo hundred and two normally pigmented Japanese alleles were analyzed. SNP, single-nucleotide polymorphism. 125 : 4 OCTOBER 2005 HERMANSKY–PUDLAK SYNDROME 1 IN JAPANESE 719

This may implicate L668 in stabilization of or even binding to Japan. HRP-conjugated secondary antibodies were from Da- HPS4. Interestingly, patients with the L668P mutation (pa- koCytomation (Glostrup, Denmark). tients 7 and 8 in Table I) manifested a relatively moderate Cells were harvested by trypsin and centrifugation, washed twice with cold PBS, and were resuspended in CF-PBS lysis buffer HPS1 clinical phenotype that included congenital nystag- (0.25 M sucrose, 25 mM KCl, 2.5mM MgCl2, 0.01% SDS). Cell mus, OCA with some pigmentation, and mild bleeding ten- extracts from melanocytes were separated in 7.5% SuperSep po- dency (bleeding time; 9.5 min), suggesting that this variant lyacrylamide gels (Wako, Osaka, Japan) and transferred to may not be functionally null in vivo. Immobilon-P (Millipore, Billerica, Massachusetts). Membranes Finally, we unexpectedly observed that the amount of were blocked in 5% fat-free milk in PBS plus 0.05% Tween-20 steady-state HPS4 protein is considerably greater in Hps3- for 1 h at room temperature, and were then incubated in blocking null (melan-coa) melanocytes than in wild-type mouse me- solution containing the appropriate primary antibody for 1 h at room temperature. Membranes were then washed extensively, in- lanocytes (see Fig 1). The HPS3 protein is a component of cubated with appropriate HRP-conjugated secondary antibody for the BLOC-2 (Gautam et al, 2004), whereas HPS4, like 1 h at room temperature, and washed again. Signals were detected HPS1, is a component of BLOC-3. These observations using ECL plus western blotting detection reagents (Amersham suggest that there may be some functional relationship be- Biosciences, Piscataway, New Jersey). tween BLOC-2 and BLOC-3, although its nature remains to Haplotype analysis Figure 3 shows the positions of five SNP de- be elucidated. rived from the genomic sequence of human 10 (Gen- bank accession no. NT_030059): SNP1 (IVS1-9C ! T; SNP accession no. rs2296429) is located in intron 1 of HPS1 gene, Materials and Methods SNP2 (nucleotide position 162C ! A; accession no. rs1804689) in Patients Twenty-four unrelated Japanese patients with OCA (six exon 2, SNP3 (IVS14 þ 7G ! C; accession no. rs2296433) in intron females and 18 males) who had no mutations in the TYR, P, TYRP1, 14, SNP4 (1472C ! G; accession no. rs2296434) in exon 15, and and MATP genes were included in this study. Patients’ clinical SNP5 (IVS16-15A ! G; accession no. rs2296435) in intron 16. phenotypes are described in Table I. This study was approved by SNP1 and SNP2 were detected with a primer set for exon 2. SNP3 the ethics committee of the Nagoya University School of Medicine. was detected with a primer set for exon 14. SNP4 was detected Informed consent was obtained from each patient, or in the case of with a primer set for exon 15. SNP5 was detected with a primer set children from the patient’s parents. The study was conducted ac- for exon 17. SSCP analysis and direct sequencing were used for cording to the Declaration of Helsinki Principles. genotyping. A haplotype of the allele with the mutation was con- structed, based either on homozygosity or on the analyses of pa- Mutation screening of the HPS1 gene DNA segments spanning rental DNA. We also analyzed haplotypes of 101 unrelated normally each of the 20 HPS1 genomic exons were amplified using primers pigmented Japanese as controls by the PCR-SSCP/heteroduplex and PCR conditions described by Bailin et al (1997) and screened method and direct sequencing, using the Estimating Haplotype- for mutations by the simultaneous single-strand conformation po- frequencies (EH) software program (Terwlliger and Ott, 1994; Curtis lymorphism (SSCP)/heteroduplex method (Lee et al, 1995). Two and Sham, 1995) to estimate the frequencies of five SNP in the kinds of SSCP gels, with glycerol concentrations of 7% and 10%, controls. were used to elevate the sensitivity of our mutation screening sys- tem. PCR products showing aberrant patterns were reamplified and sequenced directly using the genetic analysis system The authors are grateful to the patients and volunteers for donating CEQ2000XL (Beckman Coulter, Fullerton, California). blood samples. We also thank Drs A. Morita (Nagoya City University, Japan) for introducing a patient to our project; D. Bennett (St George’s DNA constructs Full-length normal human HPS1 cDNA was Hospital Medical School, UK) for kindly providing melan-a, melan-coa, cloned in pCR2.1-TOPO (Invitrogen, Carlsbad, California), and melan-ep, and melan-le; and K. Fukai (Osaka City University, Japan) for subsequently thymine at nucleotide position 2003 was substituted the PCR primer sets for the HPS1 gene. This work was supported by by cytidine to create the L668P substitution by the site-directed grants (numbers 16591095, 16044221, and 16390315) from Ministry of Education, Sports, Culture, Science and Technology of Japan to T. mutagenesis method (Giebel and Spritz, 1990). The cDNA were Suzuki and Y. Tomita and AR39892 from the National Institutes of then sequenced to confirm accuracy, and were inserted into mam- Health to R. Spritz. malian expression plasmid pIREShyg3 (BD Biosciences, San Jose, California), creating pIREShyg3-HPS1normal and pIREShyg3- DOI: 10.1111/j.0022-202X.2005.23884.x HPS1mutantL668P, respectively. Manuscript received March 3, 2005; revised April 26, 2005; accepted Cell cultures and cell transfection Melan-a, melan-coa, melan- for publication June 6, 2005 ep, and melan-le were cultured in RPMI-1640 medium with 10% fetal calf serum, 200 nM 12-O-tetradecanoyl phorbol 13-acetate, Address correspondence to: Tamio Suzuki, MD, PhD, Department of and 200 nM cholera toxin, in 10% CO as described (Suzuki Dermatology, Nagoya University Graduate School of Medicine, 65 2 Tsurumai, Showa-ku, Nagoya, Aichi 466-8550, Japan. Email: tasuzuki et al, 2001). Melan-ep mouse melanocytes were cultured for @med.nagoya-u.ac.jp 24 h and transfected with 1.6 mg pIREShyg3, pIREShyg3-HPS1- normal, or pIREShyg3-HPS1mutantL668P per 4 cm2 flask using Lipofectamine 2000 reagent (Invitrogen). Stable tranformants were selected in culture media containing 500 mg per mL hygromycin B References (Invitrogen). Anikster Y, Huizing M, White J, et al: Mutation of a new gene causes a unique Antibody production and immunoblotting Polyclonal antisera form of Hermansky–Pudlak syndrome in a genetic isolate of central against mouse HPS4 protein were raised by immunizing rabbits Puerto Rico. Nat Genet 28:376–380, 2001 with a synthesizetic peptide (486-DRTGFKPSPSGRH-498) as an- Bailin T, Oh J, Feng GH, Fukai K, Spritz RA: Organization and nucleotide se- quence of the human Hermansky–Pudlak syndrome (HPS) gene. J Invest tigen, and were then affinity purified using a smaller peptide (490- Dermatol 108:923–927, 1997 FKPSPS-495) within the primary antigen peptide by Sigma Gen- Chiang PW, Oiso N, Gautam R, Suzuki T, Swank RT, Spritz RA: The Hermansky– osis Japan (Hokkaido, Japan). Anti-HPS1 monoclonal antibody Pudlak syndrome 1 (HPS1) and HPS4 proteins are components of two (mAb hHPS5) was previously described (Oh et al, 2000). Anti-b- complexes, BLOC-3 and BLOC-4, involved in the biogenesis of lyso- actin monoclonal antibody was purchased from Sigma Genosis some-related organelles. J Biol Chem 278:20332–20337, 2003 720 ITO ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Curtis D, Sham PC: Model-free linkage analysis using likelihoods. Am J Hum Oetting WS, King RA: Molecular basis of albinism: Mutations and polymorphisms Genet 57:703–716, 1995 of pigmentation genes associated with albinism. Hum Mutat 13:99–115, Davies BH, Tuddenham EGD: Familial pulmonary fibrosis associated with oculo- 1999 cutaneous albinism and platelet function defect: A new syndrome. Q J Oh J, Bailin T, Fukai K, et al: Positional cloning of a gene for Hermansky–Pudlak Med 45:219–232, 1976 syndrome, a disorder of cytoplasmic organelles. Nat Genet 14:300–306, Dell’Angelica EC, Shotelersuk V, Aguilar RC, Gahl WA, Bonifacino JS: Altered 1996 trafficking of lysosomal proteins in Hermansky–Pudlak syndrome due to Oh J, Ho L, Ala-Mello S, et al: Mutation analysis of patients with Hermansky– mutations in the beta-3A subunit of the AP-3 adaptor. Mol Cell 3:11–21, Pudlak Syndrome: A frameshift hot spot in the HPS gene and apparent 1999 locus heterogeneity. Am J Hum Genet 62:593–598, 1998 Fukai K, Oh J, Frenk E, Almodovar C, Spritz RA: Linkage disequilibrium mapping Oh J, Liu ZX, Feng GH, Raposo G, Spritz RA: The Hermansky–Pudlak syndrome of a gene for Hermansky–Pudlak syndrome to chromosome 10q23.1– (HPS) protein is part of a high molecular weight complex involved in q23.3. Hum Mol Genet 4:1665–1669, 1995 biogenesis of early melanosomes. Hum Mol Genet 9:375–385, 2000 Gahl WA, Brantly M, Kaiser-Kupfer MI, et al: Genetic defects and clinical char- Schinella RA, Greco MA, Cobert BL, Denmark LW, Cox RP: Hermansky–Pudlak acteristics of patients with a form of oculocutaneous albinism (Herman- syndrome with granulomatous colitis. Ann Intern Med 92:20–23, 1980 sky–Pudlak syndrome). N Engl J Med 338:1258–1264, 1998 Spritz RA: Multi-organellar disorders of pigmentation: Intracellular traffic jams in Gautam R, Chintala S, Li W, et al: The Hermansky–Pudlak syndrome 3 mammals, flies and yeast. Trends Genet 15:337–340, 1999 (cocoa) protein is a component of the biogenesis of lysosome- Starcevic M, Nazarian R, Dell’Angelica EC: The molecular machinery for the bio- related organelles complex-2 (BLOC-2). J Biol Chem 279:12935–12942, genesis of lysosome-related organelles: Lessons from Hermansky– 2004 Pudlak syndrome. Semin Cell Dev Biol 13:271–278, 2002 Giebel LB, Spritz RA: Site-directed mutagenesis using a double-stranded DNA Suzuki T, Ito S, Inagaki K, Suzuki N, Tomita Y, Yoshino M, Hashimoto T: Inves- fragment as a PCR primer. Nucleic Acids Res 18:4947, 1990 tigation on the IVS5 þ 5G ! a splice site mutation of HPS1 gene found in Hermos CR, Huizing M, Kaiser-Kupfer MI, Gahl WA: Hermansky–Pudlak syn- Japanese patients with Hermansky–Pudlak syndrome. J Dermatol Sci drome type 1: Gene organization, novel mutations, and clinical-molecular 36:106–108, 2004 review of non-Puerto Rican cases. Hum Mutat 20:482, 2002 Suzuki T, Li W, Zhang Q, et al: The gene mutated in cocoa mice, carrying a defect Horikawa T, Araki K, Fukai K, Ueda M, Ueda T, Ito S, Ichihashi M: Heterozygous of organelle biogenesis, is a homologue of the human Hermansky–Pudlak HPS1 mutations in a case of Hermansky–Pudlak syndrome with giant syndrome-3 gene. Genomics 78:30–37, 2001 melanosomes. Br J Dermatol 143:635–640, 2000 Suzuki T, Li W, Zhang Q, et al: Hermansky–Pudlak syndrome is caused by mu- Huizing M, Anikster Y, Gahl WA: Hermansky–Pudlak syndrome and related dis- tations in HPS4, the human homolog of the mouse light-ear gene. Nat orders of organelle formation. Traffic 1:823–835, 2000 Genet 30:321–324, 2002 Inagaki K, Suzuki T, Shimizu H, et al: Oculocutaneous albinism type 4 is one of the Suzuki T, Miyamura Y, Matsunaga J, et al: Six novel P gene mutations and most common types of albinism in Japan. Am J Hum Genet 74:466–471, oculocutaneous albinism type 2 frequency in Japanese albino patients. J 2004 Invest Dermatol 120:781–783, 2003 Lee ST, Park SK, Lee KH, Holmes SA, Spritz RA: A non-radioactive method for Terwlliger J, Ott J: Handbook for Human Genetic Linkage. Baltimore: Johns detection of single-strand conformation polymorphisms (SSCPs) and Hopkins University Press, 1994 heteroduplexes. Mol Cell 5:668–672, 1995 Tomita Y, Miyamura Y, Kono M, Nakamura R, Matsunaga J: Molecular bases of Li W, Zhang Q, Oiso N, et al: Hermansky–Pudlak syndrome type 7 (HPS-7) results congenital hypopigmentary disorders in humans and oculocutaneous al- from mutant dysbindin, a member of the biogenesis of lysosome-related binism 1 in Japan. Pigment Cell Res 13 (Suppl. 8):130–134, 2000 organelles complex 1 (BLOC-1). Nat Genet 35:84–89, 2003 Witkop CJ, Babcock MN, Rao GHR, et al: Albinism and Hermansky–Pudlak Nazarian R, Falcon-Perez JM, Dell’Angelica EC: Biogenesis of lysosome-related syndrome in Puerto Rico. Bol Assoc Med P R 82:333–339, 1990 organelles complex 3 (BLOC-3): A complex containing the Hermansky– Zhang Q, Zhao B, Li W, et al: Ru2 and Ru encode mouse orthologs of the genes Pudlak syndrome (HPS) proteins HPS1 and HPS4. Proc Natl Acad Sci mutated in human Hermansky–Pudlak syndrome types 5 and 6. Nat USA 100:8770–8775, 2003 Genet 33:145–154, 2003