SCIENCE@home

Familial Pulmonary Fibrosis and Hermansky-Pudlak Syndrome Rare Missense Mutations In Context.

R. Stearman,A. Cornelius,L. Young,David S Conklin,E. Mickler,X. Lu,Naoko Hara,L. M. Fettig,T. Phang,M. Geraci

Published 2019 in American Journal of Respiratory and Critical Care Medicine

ABSTRACT

Hermansky-Pudlak syndrome (HPS) is an autosomal-recessive disease characterized by a constellation of phenotypes related to the proper functioning of lysosome-related organelles, including dense platelet granules. Currently, 10 HPS genes (HPS1–HPS10) have been identified in humans. Interestingly, HPS-associated pulmonary fibrosis (PF) is associated with complete penetrance in HPS1, HPS2, and HPS4, but lung disease has not been reported in other HPS subtypes (1). An initial study genetically mapped the HPS1 locus in Puerto Rican families and discovered a foundereffect mutation (16 bp duplication) (2); however, other HPS1 variants have been identified worldwide. The prevalence of PF in HPS carriers is an open question, as a detailed study examining this relationship, especially focused on milder phenotypes, has not been completed. We reasoned that HPS1, HPS2, or HPS4 mutations in heterozygote carriers may be important in PF presentation. We focused on HPS1 and HPS4 as target genes because they form the BLOC-3 complex, which is required for proper protein trafficking in cells (3). Defects in protein trafficking can be used as a model for the induction of PF in vulnerable epithelial cells (4). HPS1 and HPS4 genes are z30–35 kb in length. We developed a long-range PCR strategy (6–7 kb products) to generate end-to-end gene sequencing for HPS1 and HPS4 (including z500 bp of the promoter and flanking the 39untranslated region [UTR]), which allowed us to examine other imbedded alterations (splice sites, microRNA sites, enhancer/regulator sites, etc.). Genomic DNA from 87 proband individuals from a collection of patients with familial PF (FPF; obtained from Dr. James E. Loyd, Vanderbilt University) was sequenced. The R package haplo.stats (5) was used to assemble the predicted protein haplotypes for each individual with FPF. We identified five patients with FPF and rare HPS1 or HPS4 missense mutations (three in HPS1 and one in HPS4), and a novel HPS4 mutation containing a frameshift mutation (Table 1). In addition, 10 patients with FPF had the HPS1:G283W mutation (Table 1), which has a 3% haplotype frequency from the 1000 Genomes Project (1KGP) database. This suggests a slight enrichment for HPS1:G238W in our patients with FPF (Fisher’s exact test P value = 0.03). Table 1 shows different functional predictions of the consequence of the missense mutations, identifying HPS1-2 and HPS1-3 as probably damaging (Exome Variant Server, http://evs.gs.washington.edu/evs/). HPS1-1 and the novel HPS4-1 frameshift were not found in the Exome Variant Server database. In addition, the data for HPS1:V4A and HPS1:G283W are shown. Using amino-terminal tagged reference sequence cDNAs (HA3-HPS1 wild-type [HA3-HPS1wt] and MYC3-HPS4wt [6]), we introduced each mutation by site-directed mutagenesis. Cycloheximide treatment of each cell line was completed after it was transfected either singly or cotransfected with its wild-type partner (e.g., HPS1-1/HPS4wt). We used Western blot quantitation to measure tagged protein levels over time, approximating protein stability compared with the wt. When they were singly transfected, we found that the HPS1-1 and HPS1-2 proteins were less stable than HPS1wt, although HPS1-3 appeared to be more stable. HPS4-1fs was less stable than HPS4wt, and the protein stability of HPS4-2 was similar to that of HPS4wt. When the mutant forms were cotransfected with their wt counterpart, a similar pattern emerged (preliminary data collected in several cell lines). We built each patient’s predicted HPS1 and HPS4 protein haplotypes using a subset of SNPs that we annotated as being within the promoter, mRNA (including the 59UTR and 39UTR), and within 50 bp of the exon/intron splice site. In addition, we wanted the average heterozygosity to be between 10% and 30% (from the dbSNP database; https://www.ncbi.nlm.nih.gov/snp/) for maximum utility in building haplotypes. In most cases, haplo.stats built one unique pair of protein haplotypes with high post hoc probability (0.9–1.0). In cases where more than two pairs were predicted, the changes were either in noncoding regions or were silent mutations. Table 2 summarizes the predicted protein haplotype for each missense mutation found in our FPF cohort, along with the most common and/or reference genome sequence protein haplotype. The predicted protein haplotypes for HPS1-1, 1-2, and 1-3 mutations indicate that HPS1-1 and HPS1-3 were not tested in the appropriate protein haplotype (Table 2). Even though HPS1-2 was tested in the correct protein haplotype, that particular patient, FPF47, has the HPS1:G283W mutation in his/her other allele, suggesting a potential partial loss of function from both HPS1 alleles. HPS4 protein haplotypes presented a different problem. The reference genome sequence (tested in a wt cDNA clone) is actually a minor haplotype (13.5%) within the 1KGP, and the HPS4 mutations identified in our FPF cohort are in fact on the major HPS4 protein haplotype (43.7%; Table 2). This demonstrates that the correct protein context for the HPS1 and HPS4 mutations found in our cohort have not been adequately tested in their protein haplotype context. With the increased use of next-generation-sequencing platforms, the identification of rare and novel missense mutations occurs frequently in large disease-specific studies. HPS is a recessive disorder with a high penetrance for PF in the affected HPS subtypes. However, in the context of PF disease, heterozygous carriers, as identified in this study, may confer an added risk in a two-hit hypothesis, akin to BMPR2 (bone morphogenetic protein receptor 2) mutations in pulmonary arterial hypertension. The initial approach is to test these missense mutations by computational methods to suggest the potential degree of maladaptation of a given change (for a review of current methods, see Reference 7). The various approaches used to obtain the data shown in Table 1 tested the mutation either in isolation (how biochemically similar is glycine to tryptophan?) or in its relative evolutionary conservation (local sequence preservation over evolutionary time). None of these methods take into account that a particular missense mutation is in fact part of a specific protein haplotype. Author Contributions: R.S.S., L.R.Y., and M.W.G. conceived the project and designed the experiments. R.S.S., A.R.C., L.R.Y., D.S.C., E.A.M., X.L., N.H., L.M.F., and T.L.P. conducted experiments or helped with analysis. R.S.S., L.R.Y., and M.W.G. wrote the manuscript. All authors read and approved the manuscript.

PUBLICATION RECORD

CITATION MAP

EXTRACTION MAP

CLAIMS

  • No claims are published for this paper.

CONCEPTS

  • No concepts are published for this paper.

REFERENCES

Showing 1-7 of 7 references · Page 1 of 1

CITED BY

Showing 1-5 of 5 citing papers · Page 1 of 1