Background
Hermansky Pudlak syndrome is a rare hereditary multisystem disorder first described in 1959. More than 50% of all worldwide cases are identified in individuals from Puerto Rico where HPS has an estimated frequency of 1:1800 [
1,
2]. Clinically the syndrome is characterized by oculocutaneous albinism, a bleeding diathesis due to platelet storage deficiency [
1,
2], and other manifestations which may include neutropenia, a granulomatous colitis, or pulmonary fibrosis.
Genotypic analysis over the last decade allowed differentiating at least 10 separate forms of HPS, due to mutations in different genes [
3]. All entities have in common defects in intracellular protein trafficking and the biogenesis of lysosome-related organelles like melanosomes or platelet dense granules [
2].
Pulmonary fibrosis has not been described in HPS-3 and HPS-5 through HPS-10, which are all very rare. HPS-1 is the most common subtype and characteristically develops a severe and progressive pulmonary fibrosis in almost all cases. Usually middle-aged adults and rarely late adolescents are affected by fibrosis; however, so far, children are not described [
2‐
4]. HPS-4 has been documented in less than 10 patients, few of which had pulmonary fibrosis [
5,
6].
The HPS-2 subtype is also very rare, with less than 40 cases reported worldwide [
4,
7‐
10]. HPS-2 is caused by mutations in the
AP3B1 gene, inherited in an autosomal recessive way and distinguished from the other forms of HPS by the presence of neutropenia that can lead to severe respiratory infections and that is responsive to granulocyte colony-stimulating factor [
1]. Among the few patients described, development of an interstitial lung disease (ILD) has been mentioned in 30 to 50% [
1,
7]; details on the pulmonary phenotype have been described in four cases [
4]. Potential mechanisms causing pulmonary disease in HPS-2 are poorly understood. It has been suggested that altered
AP3B1 gene product within alveolar epithelial type II cells leads to defective intracellular processing of surfactant proteins B and C (SP-B, SP-C), endoplasmic reticulum-stress, apoptosis, and a fibrotic lung phenotype [
1].
The aim of this study was to describe the pulmonary phenotype of HPS-2 in children and to further investigate the presence and the possible role of cellular stress and apoptosis in patient-derived material.
Methods
Patients, diagnosis, and follow up
Patients were recruited from the chILD-EU register and biobank and the kids’ lung register collecting diffuse parenchymal lung diseases [
11,
12]. Among the children included between 2009 and 2017, seven children were diagnosed with HPS. A 0.4-year old infant with HPS-1 referred for the assessment of potential pulmonary involvement had no pulmonary symptoms and was excluded from this study. All the other cases were HPS-2.
The diagnosis of HPS-2 was based on typical clinical symptoms and proven by genetic analysis (Table
1). Mutational analysis was performed by Sanger sequencing. Routine clinical evaluation in different European centers was performed; data were collected retrospectively and prospectively following the inclusion into the study.
Table 1
Baseline demographics and genetics
Gender | female | female | male | female | female | female | 5: 1, female: male |
Consanguinity by history | yes | no | yes | no | yes, sibling of patient 1 | no; genealogy demonstrated common ancestors, i.e. very distant relation | 3: 3, yes: no |
Allelesa | homozygous | homozygous | homozygous | compound eterozygous | homozygous | homozygous | 5: 1, homozygous: compound heterozygous |
AP3B1 mutation 1 | c.3222-3223delTG (b) | g.151312_159483del8172bp (c) | c.2546 T > G (b) | c.177delA (b) | c.3222-3223delTG (b) | c.2944delC (b) | |
AP3B1 mutation 2 | c.3222-3223delTG (b) | g.151312_159483del8172bp (c) | c.2546 T > G (b) | c.1839_1842delTAGA (b) | c.3222-3223delTG (b) | c.2944delC (b) | |
Previously described; predicted pathogenic effect of mutations | known [ 10]; likely pathogenic variant (frame shift) | known [ 28]; likely pathogenic variant (exon skipping) | unknown; likely pathogenic variant (stop-mutation; early termination in exon 22 instead of 27) | unknown; likely pathogenic variant (frameshift mutations and subsequent early terminations in exons 2 and 17) | known [ 10]; likely pathogenic variant (frame shift) | unknown, likely pathogenic variant (frame shift; early termination in exon 26) | 2 known, 4 previously unknown variants |
Lung function testing was done according to standards set previously in children old enough to perform spirometry [
13]. CT images of the chest were evaluated for the presence of parenchymal abnormalities (like mosaic attenuation, ground glass opacity, consolidation, linear opacity, septal thickening, reticular opacity, nodular opacity, honeycombing, emphysema, cysts, bleb or bulla) and airway abnormalities (tree-in-bud, bronchiectasis, bronchial wall thickening) on a lobar basis, counting lingula as the separate lobe [
14]. Also the presence of pneumothorax, pleural thickening, pleural effusion and enlarged hilar or mediastinal lymph node were evaluated. The image analysis was performed blinded by a pediatric radiologist with expertise in chest imaging.
Bronchoscopy and bronchoalveolar lavage (BAL)
Flexible Bronchoscopy including BAL (mostly of the middle lobe) were performed if clinically indicated using 3 times 1 ml warmed normal saline per kilogram body weight. BAL was examined cytologically and microbiologically.
Lung biopsies and histological investigations
Lung biopsies available were peer-reviewed independently and blinded by a pathologist specialized in pulmonary pathology. Lung tissue of patient 3 was analyzed by Western blotting under reducing and denaturing conditions using sodium dodecyl sulfate–polyacrylamide gel electrophoresis followed by electroblotting and immunostaining for pro-SP-C (Merk Millipore, Darmstadt, Germany), ATF6, β-actin (abcam, Cambridge, UK), and cleaved caspase-3 (Cell Signaling, Gaithersburg, USA). Blotted membranes were developed with the ECL Plus chemiluminescent detection system (Amersham Biosciences, Amersham, UK). Immunohistochemistry was performed on lung tissue fixed in 4% formaldehyde on serial sections with the AP Fast Red kit (Zytochem Systems, Berlin, Germany) after antigen retrieval by microwaving in 10 mM sodium citrate buffer, pH 6.0. Hemalaun was used as counter-stain. Slides from patient 2 were also available for immunostaining for pro-SP-C and cleaved caspase-3, as described above. As controls, lung sections from 3 different organ donor lungs were used.
Ethics, consent and permissions, consent to publish, declarations and statement
Informed consent to report individual patient data was obtained by all patients old enough to consent, and their parents or guardians. The study was approved by the ethics committee of the Ludwig-Maximilian University of Munich (EK 111-13).
All supporting data have been entered in the additional material (Additional file
1).
Discussion
Children suffering from HPS-2 may present with a severe and progressive chronic pulmonary phenotype. Severe lung fibrosis may develop until early adulthood; associated complications including pneumothorax, recurrent infections, and the development of scoliosis were key features identified. Together with few forms of ATP-binding cassette sub-family A member 3 (ABCA3) [
17‐
19], SFTPC [
20,
21], and MARS deficiency [
22], this condition operates under the few clearly and rapidly fibrosing diffuse parenchymal lung diseases in childhood.
Clinically HPS-2 in children is diagnosed by the combination of albinism, bleeding diathesis, and neutropenia. The evolution of symptoms in 4 children started with epistaxis or bleeding diathesis as initial symptoms during the first 2 years of life. However, respiratory symptoms were present in all our children but obviously were rated as too non-specific or developed too insidious to contribute to the diagnosis at age 5 years. At that time, half of our patients already had clubbing, dyspnea, and oxygen demand. Seventeen of the 22 cases under 18 years of age in the largest series on HPS-2 so far had respiratory symptoms, although not further specified [
7]. Tachypnea and wet coughing developed somewhat later and suggest secondary chronic bronchitis or suppurative lung disease. The latter may be due to additional immune deficiency from neutropenia in HPS-2 subjects, which may aggravate early respiratory affections in these children. Three of our 6 patients developed pneumothoraces, adding substantially to pulmonary morbidity. Subpleural lung fibrosis, in particular cysts or honeycombing, predisposes to such a complication, which is very unusual in children at this young age. Among 4 children with HPS-2, Gochuico et al. described one child with 6 recurrent pneumothoraces [
4].
The natural history of HPS-2 lung disease differs from that of HPS-1 pulmonary fibrosis, which usually affects middle-aged adults and not children [
23]. Based on published reports, patients generally first manifest symptoms of HPS pulmonary fibrosis in middle age, however, anecdotal experience includes rare patients with HPS-1 beginning to develop ILD in late adolescence [
24]. Carmona-Rivera described a 16-year-old boy with no pulmonary symptoms in HPS-1 [
25]. Characteristic pulmonary CT findings known in adults with HPS-1 are increased reticular opacities, thickened interlobular septa, and ground-glass infiltrates in addition to fibrotic changes, including traction bronchiectasis, subpleural cysts and honeycombing. These imaging findings evolve over time, starting in early adulthood, as in HPS-1 patients < 20 years usually no CT changes are noted, in those between 20 and 29 years minimal changes were identified, with increasing symptoms in patients 30 years and older [
26,
27]. In HPS-1, high-resolution CT abnormalities inversely correlated with percentage of forced vital capacity and were useful in defining the progression of interstitial disease [
27]. In our patients with HPS-2, patchy ground-glass opacity predominated at diagnosis during early childhood and a reticular pattern evolved rapidly over time. At follow-up most of the children developed the CT findings characteristic for HPS fibrosis in adults. Thus, compared to patients with HPS-1, in HPS-2 not only pulmonary symptoms as described above but also CT abnormalities were detected very early. Of interest and in contrast to our observations, two children with HPS-2 have been described in literature before with bilateral ground-glass opacity, thickening of interlobular septa, and interstitial reticulations (4 and 8 vs. 14 years of age) and 3 more children suffering from HPS-2 had changes in HRCT all with improvement over time [
4,
9].
In all children, the diagnosis of HPS-2 was verified genetically (Table
1). A pair of siblings had the same mutation (patient 1 and 5), however their clinical course was different. This was also due to the fact that the pulmonary phenotype of the second child was much more and earlier in focus after the other died. A frameshift mutation in the C-terminal region of
AP3B1 resulted in loss of the stop codon, prolonging translation into the 3’UTR region. Although an alternative in-frame stop codon is available further downstream, increased distance to the original stop codon may predispose transcripts to nonstop-mediated decay mechanisms [
28]. Alternatively, the translated protein product might be subject to proteolytic breakdown due to misfolding, defective assembly, or intrinsic conformational instability [
29]. Patient 2 displayed a larger genomic deletion which includes parts of introns 14, 15, and exon 15. This specific region has shown to be essential for correct assembly of the AP3-complex [
30]. In all other individuals, single point mutations (patient 3) or frameshift mutations caused by base pair deletions about 10-120 base pairs upstream (patient 4, 6) lead to premature stop codons, translation termination, and possibly activation of the nonsense-mediated decay pathway (Fig.
1).
The pulmonary fibrosis in patients with HPS may be preceded by a macrophage-mediated alveolar inflammation, as BAL fluid contains increased numbers of constitutively activated macrophages [
26]. Only one of the subjects with BAL had increased macrophage counts (Additional file
1: Table S3). All our patients had progressively fibrosing lung disease already during childhood. Our histological data prove that pulmonary fibrosis affects children with HPS-2. This is in contrast to HPS-1 and 4, where the development of pulmonary fibrosis starts in middle-age adults and children are only rarely affected [
24]. Overall rapidly progressive fibrosing lung disease in childhood is extremely rare. The development of fibrosis might be facilitated by the patients’ neutropenia and natural killer - and T-cell dysfunction and the resulting susceptibility to severe recurrent chest infections. Such an observation is consistent with the HPS-2 animal model where environmental lung injury by silica or bleomycin aggravates fibrosis [
26]. Therefore, preventive measures as vaccination and aggressive antibiotic treatment are warranted. Taking our limited observation length into account, the clinical course was not stable in most children. Despite intense symptomatic treatments after diagnosis, we saw deterioration and development of complications in several patients. Four patients remained unchanged, one patient improved, one patient (subject 5), however, died from respiratory insufficiency.
The histology of HPS-2 interstitial lung disease is not widely explored, as the diagnosis may now be done genetically. Patient 2 had a combination of NSIP and UIP-like pattern with dense fibrosis in peribronchiolar and subpleural distribution, as well as a DIP- like areas with intraalveolar aggregates of alveolar macrophages, very similar to other descriptions published [
4]. Lung biopsy of patient 3 was dominated by a patchy dense fibrosis with UIP-like pattern comparable to patient 2. Lung biopsy of patient 6 did not show areas of dense fibrosis but demonstrated a cellular NSIP pattern. Additionally, there was a mild lymphoid hyperplasia with few lymphoid follicles with germinal centers that could probably be interpreted as post-infectious changes.
Apoptosis of type II pneumocytes, in addition to ER stress and defective autophagy, was observed in a HPS-2 patient lung. This finding is in line with the previously reported observations of cellular stress and apoptosis of type II pneumocytes in several interstitial lung diseases. More studies are needed to determine if defective autophagy or ER stress underlie type II pneumocyte apoptosis and are subsequently responsible for fibrotic remodeling in the HPS-2 patient lung.