MATERIALS AND METHODS

DNA sequencing

DNA sequencing was performed with the fmol cycle sequencing kit (Promega, Madison, WI). Primers were labeled in reactions containing 10 pmol primer, 2.5 U T4 kinase, 0.025 Ci -³²P-dATP, and 1 T4 PNK buffer (70 mM Tris-HCl, pH 7.6, 10 mM MgCl₂, 5 mM dithiothreitol), and incubated for 1 h at 37°C. Cycling conditions were 95°C for 1 min followed by 20 cycles of 95°C for 25 s, 58°C for 25 s, and 72°C for 20 s.

Southern analysis

Probes for Southern analysis were derived from three separate regions of the HPS gene by polymerase chain reaction (PCR) amplification of human genomic DNA or by restriction digestion of the 1.5 kb cDNA clone. Following separtation on 1% SeaKem GTG agarose (FMC BioProducts, Rockland, ME), restriction digested DNA was blotted onto a Zetabind membrane (Cuno, Meriden, CT), fixed by UV cross-linking, and prehybridized for 4 h at 55°C in 5 sodium citrate/chloride buffer (SSC), 10 Denhardt's reagent, 50 mM Tris, 0.1% Na₄P₂O₇, 1% sodium dodecyl sulfate (SDS), and 667 g salmon sperm per ml. Hybridization was carried out in 5 SSC, 1 Denhardt's reagent, 50 mM Tris, 0.1% Na₄P₂O₇, 1% SDS, 15% dextran sulfate, 50% formamide, 100 g salmon sperm DNA per ml, and 25 ng of -³²P-dCTP labeled probe at 42°C for 15–18 h. Post-hybridization washes were 15 min in 2 SSC and 0.2% SDS at 42°C and 30 min at room temperature in 0.2 SSC and 0.2% SDS.

Reverse transcriptase polymerase chain reaction (RT-PCR)

RNA was isolated from lymphoblastoid cells lines, melanoma, or megakaryocytic leukemia cells lines by a guanidinium isothiocyanate procedure (^{Sambrook et al. 1989}). Reverse transcription was carried out with 100 pmol oligo dT₁₅, 200 units SuperScript II reverse transcriptase (Gibco BRL, Gaithersburg, MD), 250 mM Tris-HCl (pH 8.3), 375 mM KCl, 15 mM MgCl₂, 10 mM dithiothreitol, 30 units RNasin ribonuclease inhibitor (Promega, Madison, WI), 125 M of each dNTP, and 1–5 g of total RNA at 42°C for 2 h. Reverse transcription products were then amplified in reactions containing 1.5 mM MgCl₂, 1PCR buffer (50 mM KCl, 10 mM Tris-HCl pH 8.3), 20 pmol of each amplification primer, 200 M of each dNTP, 2.5 U Ampli-Taq (Perkin Elmer, Branchburg, NJ), and 10 l of the reverse transcription reaction. The cycling conditions consisted of 35 cycles of 95°C for 1 min, 55°C for 1 min 10 s, and 72°C for 1 min 30 s. For sequence analysis, PCR products were cloned into pCR 2.1 plasmid using the TA cloning kit (Invitrogen, Carlsbad, CA).

Sequences of primers used for PCR amplifications are as follows:

Primer 1: 5'-GGTGGCAGTGAGGAAGTATC-3'

Primer 2: 5'-GTCCACTGGAGCCTAAGACA-3'

Primer 3: 5'-GACTGTTAAGACCACTCCAC-3'

Primer 4: 5'-CCTTCCTTCAGCTTCTTCTC-3'

Primer 5: 5'-GAGACGGAGACAGACAGCTT-3'

Primer 6: 5'-CAGTGGAGTGGTCTTAACA-3'

Primer 7: 5'-CCGGTCATCATCTCCTCCAT-3'

Northern analysis

Ten micrograms of total RNA from lymphocytes or 1 g of polyadenylated RNA from melanoma or bone marrow cells (Clontech, Palo Alto, CA) was separated on 1% SeaKem GTG Agarose (FMC BioProducts, Rockland, ME) containing 1 2-[N-Morpholino]-ethanesulfonic acid and 5.92% formaldehyde, blotted onto a Zetabind membrane (Cuno, Meriden, CT) in 20 SSC, and fixed to the membrane by UV cross-linking. Blots were prehybridized for 4 h at 55°C in 5 SSC, 10 Denhardt's reagent, 50 mM Tris, 0.1% Na₄P₂O₇, 1% SDS, and 40 g salmon sperm DNA. Hybridization was carried out in 5 SSC, 1 Denhardt's reagent, 50 mM Tris, 0.1% Na₄P₂O₇, 1% SDS, 10% dextran sulfate, 50% formamide, 0.8 g salmon sperm DNA, and 25 ng of -³²P-dCTP labeled probe at 42°C for 16–18 h. Post-hybridization washes were 15 min at room temperature in 2 SSC, 0.2% SDS and 0.2 SSC, 0.2% SDS at 42°C for 30 min.

Allele specific oligo analysis

DNA for allele specific oligo analysis was blotted onto a Zetabind membrane (Cuno, Meriden, CT) using a 96 well Bio-Dot apparatus (Bio-Rad, Hercules, CA) and fixed by UV cross-linking. Oligonucleotide probes specific to normal or variant alleles were end-labeled with -³²P-dATP by polynucleotide kinase (Promega, Madison, WI). Blots were prehybridized for 4 h at 55°C in 5 SSC, 10 Denhardt's reagent, 50 mM Tris, 0.1% Na₄P₂O₇, 1% SDS, and 100 g salmon sperm DNA per ml, and hybridized with 3 M TMAC (Sigma, St. Louis, MO), 0.6% SDS, 1 mM ethylenediamine tetraacetic acid, 10 mM Na₃PO₄, 5 Denhardt's reagent, 1 mg salmon sperm DNA per ml, 26.25 pmol of unlabeled competitive primer, and 2.6 pmol of labeled primer at 55°C for 15–18 h. Post-hybridization washes were 20 min at room temperature in 3 M TMAC, 1 mM ethylenediamine tetraacetic acid, 10 mM Na₃PO₄, and 0.6% SDS followed by a second wash at 52°C for 20 min.

Probes used were as follows:

1259CT alteration:

normal: 5'-AGTGGACCCCCGTGCCTCC-3'

mutant: 5'-AGTGGACCTCCGTGCCTCC-3'

1328CT alteration:

normal: 5'-GCGGCTCCCCATCTTCCCT-3'

mutant: 5'-GCGGCTCCCTATCTTCCCT-3'

RESULTS

Isolation of a 1.5 kb cDNA in the HPS locusAs previously described (^{Gardner et al. 1997}), the Puerto Rican HPS locus was mapped to a 1 cM region between D10S58 and D10S1433, and very close to D10S184 on chromosome 10q2, by recombination analysis and multipoint disequilibrium analysis (^{Terwilliger 1995}), and YAC containing these markers were identified (943f9 and 811d8).¹ A single exon bearing fragment, 7RV-H, isolated from these YAC was used to isolate a human cDNA clone. This cDNA consisted of four EcoRI fragments whose lengths were 1.5, 1.3, 0.9, and 0.344 kb. Sequencing of these fragments revealed that the clone was a chimera consisting of two separate cDNA. The 1.3, 0.9, and 0.344 kb fragments all contained sequences identical to portions of the mRNA for human BiP protein (GenBank accession #X87949), and were not analyzed further. The sequence of the remaining 1.5 kb fragment was unique and contained partial sequence identity to the previously published HPS cDNA sequence (^{Oh et al. 1996}). The complete DNA sequence of this 1.5 kb cDNA (GenBank accession #U96721) and the predicted amino acid sequence are shown in Figure 1. The gene responsible for the mouse pale ear phenotype was isolated by homology to this HPS cDNA (^{Gardner et al. 1997}).

Figure 1.

DNA and predicted protein sequence of the 1.5 kb cDNA (GenBank accession #U96721). Underlined bases were not reported in the 3.6 kb cDNA sequence. Dashed underlined bases indicate the position of a consensus 5' splice site within the exon containing the sequences unique to the 1.5 kb cDNA. The vertical line indicates a splice site. Boxed sequences represent regions removed in alternatively spliced transcripts (490–632, 739–900, and 1003–1101) that were detected by RT-PCR analysis.

Full figure and legend (81K)

The 1.5 kb cDNA has partial identity with the HPS cDNA and a unique 3' end

In a separate study, Oh et al. published the sequence of a 3.6 kb cDNA associated with HPS (^{Oh et al. 1996}). Comparison of the published cDNA sequence with the sequence of the 1.5 kb cDNA suggested that both were from the same gene, but with significant differences. Bases 1–1145 of the 3.6 kb cDNA are identical to bases 17–1173 of the 1.5 kb cDNA, except for additional bases found at positions 74–76 and 119–127 in the 1.5 kb cDNA. Bases 1–16 of the 1.5 kb cDNA are absent from the 3.6 kb cDNA sequence. These additional bases are underlined in Figure 1, and their presence in the gene has been confirmed by sequencing genomic DNA. The most significant difference occurs after base 1173 of the 1.5 kb cDNA (base 1145 of the 3.6 kb cDNA). The two cDNA sequences diverge completely after this point. The 1.5 kb cDNA continues for an additional 319 bases, as shown in Figure 1, whereas the 3.6 kb cDNA continues for 2528 bases after the divergence point (^{Oh et al. 1996}). The 319 bases unique to the 1.5 kb cDNA have no overlap with the 2528 bases unique to the 3.6 kb cDNA, and no homology to other cDNA could be found on a BLAST search. These bases are also excluded from the recently reported exon sequences of the HPS gene (^{Bailin et al. 1997}). Sequence analysis of the region unique to the 1.5 kb cDNA in primate genomic DNA showed 98% homology between the human, gorilla, and chimpanzee sequences, and 96% homology between the human and orangutan sequences.

The 1.5 kb and 3.6 kb cDNA are both derived from the HPS gene

Three probes were used to demonstrate that the two cDNA are derived from the same gene. One probe was derived from the region common to both cDNA, and the other probes were derived from the regions unique to each. For each probe, hybridization to YAC DNA (clone 811d8) and human genomic DNA identified the same specific fragments. Furthermore, PCR amplification of the common region and each unique region verified their presence in the YAC and in a 130 kb BAC clone (145 M16) reported to contain the HPS gene (^{Oh et al. 1996;Bailin et al. 1997}) (data not shown). These results suggested that all three cDNA regions are likely to be part of a single gene.

Amplification of genomic DNA further revealed that the region unique to the 1.5 kb cDNA and a portion of the common region are contained within a single exon. Human genomic DNA was amplified with primers 5 and 6 shown in Figure 2(a). The sequence of the resulting PCR product was identical to the sequence found in the 1.5 kb cDNA Figure 3. Furthermore, the sequences around the junction between the common region and the 1.5 kb cDNA unique region contained a consensus 5' splice site. The sequence of bases 1169–1177 of the 1.5 cDNA is 5'-CAGGTGAGG-3' as shown in Figure 3. This consensus splice site could allow splicing from the common region of the gene to the region unique to the 3.6 kb cDNA. The results of these analyses suggested a model for the genomic structure of the gene and the post-transcriptional processing that would produce the transcripts with alternative 3' ends Figure 2a.

Figure 2.

RT-PCR analysis and model for processing of the HPS gene. (a) RT-PCR analysis of the alternative HPS gene transcripts. Numbered arrows represent locations of primers used in the analysis. Primer sequences are given in the Materials and Methods. (b) Model for processing of the HPS gene. Differently shaded sections represent the three distinct regions of the gene, and the putative protein products from each transcript are depicted at the bottom of the figure. The 1173 bp region is the region common to both cDNA, the 319 bp region is the region unique to the 1.5 kb cDNA, and the 2528 bp region is the region unique to the 3.6 kb cDNA.

Full figure and legend (13K)

Figure 3.

Sequence analysis of the HPS gene. RT-PCR products (a and c) and genomic DNA (b) were sequenced at the junction between the common region and the regions unique to the 1.5 kb cDNA (a and b) or the 3.6 kb cDNA (c). The sequence of the genomic DNA and the 1.5 kb cDNA matches the consensus 5' splice site sequence: AG GT(A/G)AG.

Full figure and legend (49K)

Both HPS transcripts are expressed in cell types related to the disorder

Messenger RNA from melanoma cells and bone marrow was hybridized to probes from the common region or from each of the unique 3' regions. The probe from the unique region of the 1.5 kb cDNA hybridized to a major band in both tissues of 1.5 kb Figure 4a. The probe from the unique region of the 3.6 kb cDNA hybridized to a major band of about 3.0 kb in both tissues Figure 4a, as expected (^{Oh et al. 1996}). The probe from the common region hybridized to both of these transcripts Figure 4a, and the bands identified were of approximately equal intensity as determined by densitometric analysis.

Figure 4.

Northern analysis of alternative HPS transcripts. Lymphocyte total RNA (L), bone marrow mRNA (B), or melanoma mRNA (M) was probed with DNA from either the region unique to the 1.5 kb cDNA (a), the region unique to the 3.6 kb cDNA (b), or the region common to both cDNA (c). Non-specific hybridization to 28S and 18S rRNA bands is indicated in the lymphocyte lane.

Full figure and legend (71K)

The expression of the 1.5 and 3.6 kb cDNA was examined by RT-PCR in lymphoblastoid cell lines of HPS affected and unaffected individuals, normal melanocytes, and megakaryocytic leukemia cells. Total RNA was isolated, reverse transcribed with an oligo-dT primer, and the reverse transcription products were amplified with primers designed to amplify from the common region to each of the two unique 3' regions. The 1.5 kb cDNA was amplified using primers 5 and 2 in Figure 2(a), and the 3.6 kb cDNA was amplified with primers 5 and 4. In all tissue types examined, these reactions produced fragments of the expected size Figure 5 and sequence Figure 3, indicating that both alternative transcripts are produced and polyadenylated in these cell types.

Figure 5.

Results of RT-PCR analysis. Reverse transcribed RNA from normal melanocytes (lanes 1, 3, 5, and 7) or leukemic megakaryocytes (lanes 2, 4, 6, and 8) was amplified with primers shown in Figure 2(a). Lanes 1 and 2, primers 2 and 5 (specific to the 1.5 kb cDNA);lanes 3 and 4, primers 4 and 5 (specific to the 3.6 kb cDNA);lanes 5 and 6, primers 3 and 4 (1.5 cDNA kb unique region to 3.6 cDNA kb unique region);lanes 7 and 8, primers 1 and 2 (3.6 kb cDNA unique region to 1.5 kb cDNA unique region);lane 9 is a negative control.

Full figure and legend (53K)

To determine if a transcript containing both of the unique 3' regions could be detected, RNA from normal melanocyte and leukemic megakaryocyte cells was reverse transcribed and amplified using primers 1 and 2 or 3 and 4 Figure 2a. These reactions did not produce any amplification products Figure 5, implying that transcripts can only contain one 3' region or the other.

Gene alterations are present in both transcripts among individuals with HPS

Oh et al. identified a 16 base duplication mutation of the HPS gene in affected Puerto Rican individuals (^{Oh et al. 1996}). This mutation is located in the unique region of the 3.6 kb cDNA. In addition to this 16 base duplication, we identified two other base changes in affected Puerto Ricans within the 3'-UTR unique to the 1.5 kb transcript: 1259CT and 1328CT. DNA from 81 individuals was screened for the presence of these alterations to determine their correlation with the HPS phenotype (Table 1). The 16 base duplication was analyzed by PCR amplification with the primers reported by^{Oh et al. (1996}), and the alterations in the 1.5 kb cDNA unique region were analyzed by allele specific oligo hybridization. The 16 base duplication and the 1328CT alteration segregated together completely in Puerto Rican individuals with HPS or those who were obligate heterozygotes. Neither of these alterations was found in non-Puerto Rican individuals. The 1259CT alteration was present in DNA from many Puerto Rican and non-Puerto Rican individuals and was found in the homozygous state in five unaffected individuals.

Table 1 - Table 1. Results of HPS gene mutation analysis.

Full table (26K)

DISCUSSION

We report here the identification of a novel 1.5 kb cDNA sequence associated with the HPS locus. The transcript represented by this cDNA shares 1173 bases with a previously reported cDNA and genomic sequence but diverges into a unique 319 base 3' end. The region common to both cDNA and the two unique 3' regions are all found within a 130 kb fragment of genomic DNA, suggesting that they are all part of a single gene, and expression studies have shown that both transcripts are produced in normal melanocytes, melanoma cells, bone marrow cells, leukemic megakaryoctes, and lymphoblastoid cell lines. Previous studies of this HPS associated gene, however, have not revealed the presence of transcripts containing the sequences unique to the 1.5 kb cDNA (^{Oh et al. 1996;Bailin et al. 1997}). Our analysis now shows that complex post-transcriptional processing of this gene may play a role in the production of two alternative transcripts.

Analysis of genomic DNA has suggested a model for the post-transcriptional processing used to produce the two alternative transcripts of this gene Figure 2a. The region unique to the 1.5 kb cDNA and a portion of the common region were found to be parts of a single exon. A consensus 5' splice site was also found to exist in this exon at the partition between the common region and the region unique to the 1.5 kb cDNA. Because the divergence point between the two alternative cDNA lies within this exon, it appears that the transcript for the 1.5 kb cDNA is produced by polyadenylation after this exon and cleavage of the exons unique to the 3.6 kb cDNA. The 3.6 kb cDNA would then be produced by activation of the 5' splice site within this exon to allow splicing to the sequences unique to that product.

This type of RNA processing has been observed in the genes for the immunoglobulin M (IgM) heavy chain (^{Peterson and Perry 1989}) and for the Chediak–Higashi Syndrome (CHS) (^{Barbosa et al. 1997}). The IgM gene produces membrane bound (m) or secreted (s) protein products whose relative levels are regulated during B cell maturation. More relevant to our studies are the recent findings showing that the gene responsible for CHS produces two major transcripts using the same type of processing we have described for the HPS gene (^{Barbosa et al. 1997}). Although disimilar in size, it is intriguing that both of these genes produce alternate transcripts in an identical manner and that HPS and CHS both exhibit oculocutaneous albinism and platelet storage pool deficiencies.

The role played by the two alternative transcripts in the HPS phenotype deserves further study. It is clear that the 3.6 kb cDNA contributes to the phenotype because frame-shift mutations have been reported in the region unique to the 3.6 kb cDNA in patients from Puerto Rico, Japan, Switzerland, and Ireland (^{Oh et al. 1996}). These alterations all result in aberrant translation and premature termination of the protein product. The recently cloned ep mouse mutant also contains a mutation within the region homologous to the sequences unique to the 3.6 kb cDNA (^{Feng et al. 1997;Gardner et al. 1997}).

The role of the 1.5 kb cDNA in HPS is, however, uncertain. We have identified two base changes in the region unique to this transcript among affected Puerto Rican individuals. One of these alterations (1259CT) appears to be a nonpathologic base change. The other alteration (1328CT) segregates completely with the HPS phenotype in Puerto Rican individuals and has not been observed among non-Puerto Ricans. It is not likely to be a pathogenic mutation because it is found within the 3' untranslated region of the 1.5 kb cDNA. Its correlation with the HPS phenotype may simply be due to the fact that the Puerto Ricans we have examined all carry a single founder haplotype. The effects of this alteration have not been directly tested, however, and studies have suggested that the 3'-UTR may play a role in the proper expression of some genes (^{Jackson 1993;Tsukamoto et al. 1996}). Furthermore, the sequences unique to the 1.5 kb cDNA have not yet been analyzed in non-Puerto Rican HPS individuals, so it is unknown whether there are pathologic alterations in this region among other affected individuals. It is also possible that mutations within the common region or within the region unique to the 1.5 kb cDNA lead to an as yet undetermined phenotype. The high degree of homology seen between human, gorilla, chimpanzee, and orangutan for the sequences unique to the 1.5 kb cDNA and for the consensus 5' splice site in the exon containing these sequences, suggests that this region is required for proper function of the gene.

Although a gene associated with HPS has been isolated, it appears to have a complex expression pattern. In addition to producing two transcripts with different 3' ends, there is evidence suggesting alternative splicing within the region common to both transcripts. The biologic significance of these alternative forms remains to be seen, and knowledge gained from the study of this gene will provide further insight into the molecular systems affected by HPS.

Notes

¹ YAC clones were identified by searching the Whitehead Institute's human genome map (http://www-genome.wi.mit.edu/cgi-bin/contig/physÈmap) or by physical mapping data developed by the Genome Therapeutics Corporation (Waltham, MA).

References

1. Bailin, T, Oh, J, Feng, GH, Fukai, K, Spritz, RA Organization and nucleotide sequence of the Hermansky-Pudlak (HPS) gene. J Invest Dermatol, 1997 108, 923–927,  | Article | PubMed | ISI | ChemPort |
2. Barbosa, Mdfs, Barrat, FJ, Tchernev, VT et al. Identification of mutations in two major mRNA isoforms of the Chediak–Higashi syndrome gene in human and mouse. Hum Molec Genet, 1997 6, 1091–1098,
3. Feng, GH, Bailin, T, Oh, J, Spritz, RA Mouse pale ear (ep) is homologous to human Hermansky–Pudlak syndrome and contains a rare "AT-AC" intron. Hum Molec Genet 1997 5, 793–797,
4. Fukai, K, Oh, J, Frenk, E, Almadovar, C, Spritz, RA Linkage Disequilibrium Mapping of the Gene for Hermansky–Pudlak Syndrome to Chromosome 10q23. 1-q23 3 Hum Molec Genet, 1995 4, 1665–1669,
5. Gardner, JM, Wildenberg, SC, Keiper, NM et al. The mouse pale ear (ep) mutation is the homologue of human Hermansky–Pudlak Syndrome (HPS). Proc Natl Acad Sci USA, 1997 94, 9238–9343,  | Article | PubMed | ChemPort |
6. Jackson, RJ Cytoplasmic regulation of mRNA function: the importance of the 3' untranslated region. Cell, 1993 74, 9–14,  | Article | PubMed | ISI | ChemPort |
7. Novak, EK, Hui, SW, Swank, RT Platelet storage pool deficiency in mouse pigment mutations associated with seven distinct genetic loci. Blood, 1984 63, 536–544,  | PubMed | ISI | ChemPort |
8. O'Brien, EP, Novak, EK, Keller, SA, Poirier, C, Guenet, JL, Swank, RT Molecular map of chromosome 19 including three genes affecting bleeding time: ep, ru, and bm. Mamm Genome, 1994 5, 356–360,
9. Oh, J, Bailin, T, Fukai, K et al. Positional cloning of a gene for Hermansky–Pudlak syndrome, a disorder of cytoplasmic organelles. Nat Genet, 1996 14, 300–306,  | Article | PubMed | ISI | ChemPort |
10. Peterson, ML & Perry, RP The regulated production of the mu m and mu s mRNA is dependent on the relative efficiencies of the mu s poly (A) site usage and the c mu 4-to-M1 splice. Molec Cell Bio, 1989 9, 726–738,
11. Sambrook, J, Fritsch, EF, Maniatis, T Extraction of RNA with guanidium thiocyanate followed by centrifugation in cesium chloride solutions. In: Ford N, Nolan C, Ferguson M, Ockler M (eds). Molecular Cloning: A Laboratory Manual, 1989 2nd edn. Cold Spring Harbor Laboratory Press, Cold Spring Harbor. pp. 7, 19–722,
12. Summers, CG, Knobloch, WH, Witkop, CJ, King, RA Hermansky–Pudlak syndrome: ophthalmic findings. Ophthalmol, 1988 95, 545–554,
13. Terwilliger, JD A powerful method for the analysis of linkage disequilibrium between trait and one or more polymorphic marker loci. Am J Hum Genet, 1995 56, 777–787,  | PubMed | ISI | ChemPort |
14. Tsukamoto, H, Boado, RJ, Pardridge, WM Differential expression in glioblastoma multiforme and cerebral hemangioglastoma of cytoplasmic proteins that bind to different domains within the 3 prime-untranslated region of the human glucose transporter 1 (GLUT1) messenger RNA. J Clin Invest, 1996 97, 2823–2832,
15. Wildenberg, SC, Oetting, WS, Almadovar, C, Krumwiede, M, White, JG, King, RA A gene causing Hermansky–Pudlak syndrome in a Puerto Rican population maps to chromosome 10q2. Am J Hum Genet, 1995 57, 755–765,  | PubMed | ISI | ChemPort |

Acknowledgments

We would like to thank Dr. Harry Orr, Dr. Timothy Behrens, and Dr. Lynne Maquat for their helpful discussions and advice, and the HPS Network for their continued support. We also gratefully acknowledge Jen-I Mao from the Genome Therapeutics Corporation for providing contig and physical mapping information. This research was supported by NIH Grants #GM22167, #GM/AR56181, and #CA06927.