Main

Gaucher disease, the inherited deficiency of lysosomalβ-glucocerebrosidase (EC 3.2.1.45), is a multisystem disorder resulting from storage of glucosylceramide within cells of the reticuloendothelial system(1, 2). In the past, the only reports of skin manifestations in this disease were limited descriptions of pigmentary alterations in adult type 1 Gaucher patients(3). However, recently neonates have been described with severe type 2 (acute neuronopathic) Gaucher disease who also presented at birth with a collodion baby phenotype(4–7). Moreover, a mouse model of Gaucher disease, homozygous for a null allele created by targeted disruption of the murine glucocerebrosidase gene, also displays ichthyosiform skin abnormalities(8), as well as an increased glucosylceramide to ceramide ratio in skin(9). The skin scaling abnormality, both in Gaucher infants and in the Gaucher mouse model, is of particular interest because ceramides, as major components of the intercellular bilayers in the stratum corneum of normal epidermis, are critical for permeability barrier homeostasis(9–13). Glucocerebrosidase is particularly abundant in mammalian epidermis, with the highest levels found in the stratum corneum(12). Recent studies have demonstrated that severe deficiency of this enzyme results in morphologic and functional abnormalities in the epidermis(9, 10).

The early and presymptomatic discrimination of type 2 (acute neuronopathic) patients from the milder and more treatable, type 1 (nonneuronopathic), and type 3 (chronic neuronopathic) patients remains an ongoing diagnostic challenge for those treating Gaucher disease(1, 2). Tests which would selectively identify type 2 Gaucher patients would greatly facilitate genetic counseling and management decisions for those treating very young Gaucher patients. Ultrastructural examination of skin biopsy specimens has been advocated in the past as a means to screen for lysosomal storage disorders(14, 15). To better characterize the epidermal abnormalities in type 2 Gaucher infants, and to determine whether these changes are unique to type 2 patients and therefore diagnostically useful, we compared: 1) the epidermal ultrastructure in skin biopsy samples from type 2 patients to the other types of Gaucher patients and to normal individuals and 2) alterations in lipid content in the epidermis of Gaucher patients. Our results show that patients with type 2 Gaucher disease have epidermal abnormalities which are virtually identical to those seen in type 2 (null allele) Gaucher mice. These abnormalities may prove to be useful markers for the early identification of type 2 Gaucher disease.

METHODS

Patient case reports. Samples from 10 patients with Gaucher disease were included in these studies. Their case histories are summarized inTable 1. The diagnosis of Gaucher disease was established by enzymatic and/or genotypic analyses.

Table 1 Patient case histories
Full size table

Sources of human skin samples for ultrastructural studies. Skin samples from three type 2 Gaucher patients (cases 1, 2, and 3) and two type 1 Gaucher patients (cases 4 and 5) were obtained at autopsy and preserved in 10% formaldehyde before postfixation and processing for ultrastructural analysis(see below). Punch biopsy samples were obtained from patients 6, 7, 8, 9, and 10 with informed consent, and preserved either in formalin or quick frozen in liquid nitrogen until analysis. In addition, fresh normal skin samples were obtained from the surgical margins of unaffected adult control samples(n = 5). All animal and human studies were approved by an Institutional Review Board.

Sources of samples for ultrastructural studies on Gaucher mice. Fresh full-thickness skin samples were obtained from newborn type 2 (null allele) Gaucher mice(8). These mice have a devastating clinical course with decreased movement, respirations, and feeding resulting in death within 24 h of birth. The affected animals can be identified after birth by their dry, scaly skin and dark coloration. The diagnosis is confirmed by genotypic analysis performed by Southern blotting of tail DNA(8). Skin biopsies from euthanized animals were minced to<0.5 mm3, and processed for electron microscopy, as described below. Samples for light microscopy were fixed in buffered formalin, paraffin-embedded, sectioned (5 μm), and stained with hematoxylin and eosin or periodic acid-Schiff. Tissue sections were examined and photographed with a Leitz Ortholux II microscope (Leica, Inc., Deerfield, IL).

Histology and transmission electron microscopy. Full-thickness skin samples from type 1, 2, and 3 Gaucher patients, human controls, and Gaucher and normal mice were minced to <0.5 mm3, and fixed overnight in modified Karnofsky's solution. Samples were then divided for: 1) routine paraffin embedding followed by hematoxylin and eosin staining and2) postfixation in the dark in both 0.5% ruthenium tetroxide(16) and 2% aqueous osmium tetroxide, each containing 1.5% potassium ferrocyanide(6). After 2 h at room temperature, samples for ultrastructural analysis were dehydrated in increasing grades of ethanol solution and embedded in an Epon-epoxy mixture(13). Thin sections were examined, with or without further contrasting with lead citrate, in a Zeiss 10A electron microscope, operating at 60 kV.

Lipid biochemistry of human and murine epidermis. Skin samples from Gaucher patients were either immediately frozen in liquid N2 and stored at -70°C or fixed with buffered formalin or one-half strength Karnofsky's solution. Epidermal sheets were obtained from newborn type 2 Gaucher mice by removing skin samples from euthanized animals and submerging these samples in calcium and magnesium-free Dulbecco's PBS containing 10 mM EDTA (pH 7.4) for 35-45 min (37°C), followed by gentle scraping with a scalpel blade. Human epidermis was separated from dermis using EDTA in the same manner. Epidermal samples were minced and placed into chloroform:methanol:water (4:2:1.6, vol/vol) overnight at 4°C, to obtain a total lipids extract that was subsequently dried, weighed, and stored in chloroform at -70°C until use(17). Separation of individual sphingolipid species was performed by HPTLC as described previously(11), and quantitation was performed by spectrodensitometry. After HPTLC fractionation, the dried plates were dipped in charring solution (1.5% cupric sulfate in acetic acid:sulfuric acid:orthophosphoric acid-:distilled water, 50:10:10:95, vol/vol), as described previously(10), dried (40°C, 10 min), and then baked at 180°C for 15 min. The HPTLC plates were then scanned with a variable wavelength scanning densitometer (Camag; Muttenz, SWI), and the lipid fractions were quantitated by comparison to known standards run in parallel with the experimental samples. Quantitation of individual lipid species was performed using CATS II software (Camag).

RESULTS

Histology. Light microscopy of skin from the patients with type 1 and 3 Gaucher disease did not reveal any epidermal abnormalities. In contrast, skin samples from type 2 Gaucher patients (cases 1-3) revealed dense hyperkeratosis, epidermal hyperplasia, and inflammation(Fig. 1). Moreover, in some areas the epidermis showed parakeratosis with an absence of the granular layer. The inflammatory infiltrate was localized to the superficial dermis and consisted predominately of lymphocytes and histiocytes in a perivascular distribution. These results show that skin from patients with type 2 Gaucher disease displays histologic changes which are not present in skin from type 1 and 3 patients.

Figure 1
figure 1

Skin histology of type 1 and 2 Gaucher disease.(A) Normal epidermis and stratum corneum in a type 1 Gaucher patient(case 4). (B) Hyperkeratosis and epidermal hyperplasia in a sample from a type 2 Gaucher patient (case 2). In addition, the upper dermis displays a moderate perivascular inflammatory infiltrate. Hematoxylin and eosin stain; original magnification, ×125.

Full size image

Ultrastructure of epidermis in Gaucher patients. The ultrastructure of human epidermal samples from type 1, 2, and 3 Gaucher patients and normal subjects were compared. In the samples from all three type 2 patients, abnormal arrays of loosely packed, lamellar body-derived sheets(Fig. 2) replaced the normal lamellar bilayer unit structures of the stratum corneum extracellular domains. Moreover, the morphologic features in samples from both an ichthyotic neonate (case 1) and the two older, nonichthyotic type 2 infants (cases 2 and 3) were virtually identical. In contrast, the structure of the extracellular lamellae in two type 1 Gaucher patients (cases 4 and 5) and three type 3 patients (cases 8, 9, and 10) appeared identical to those in normal human stratum corneum, revealing a normal lamellar unit bilayer pattern (Fig. 3). Despite the abnormal extracellular lamellar pattern in the type 2 Gaucher patients, the structure and contents of lamellar bodies in these patients was similar to those from both the type 1 and type 3 patients and normal subjects(Fig. 2). In some areas, the stratum corneum cells from the type 2 patients showed retention of nuclei and the presence of intracytoplasmic lipid droplets (Fig. 2C). These studies demonstrate an abnormality in stratum corneum membrane structure in type 2 Gaucher patients that is attributable to a failure in the extracellular processing of secreted lamellar body contents.

Figure 2
figure 2

Ultrastructure of human type 2 Gaucher epidermis.(A and C) All three type 2 Gaucher patients had unprocessed, elongated lamellar body-derived sheets in loosely packed arrays through the mid-to-outer stratum corneum (arrows) (case 3 is shown). Corneocytes also display abnormal cytosolic lipid droplets (L).(B) Secreted lamellar body contents at the stratum granulosum(SG) and stratum corneum (SC) interface appear completely normal (arrow). Ruthenium tetroxide postfixation; bar = 1 μm.

Full size image
Figure 3
figure 3

Ultrastructure of normal human and type 1 Gaucher epidermis. (A) Lamellar body-derived sheets are present in the intercellular spaces at the stratum granulosum/stratum corneum interface in normal human stratum corneum. At the apical margins of the intercellular spaces, transformation of lamellar body-derived sheets (lower arrowhead) into lamellar bilayer structures (upper arrowheads) can be observed. (B and C) Stratum corneum from an individual with type 1 Gaucher disease (case 4) contains lamellar bilayers exhibiting normal basic unit structures. Ruthenium tetroxide postfixation; bars = 0.1 μm.

Full size image

Ultrastructure of stratum corneum in Gaucher mice. We also compared the ultrastructure of the epidermis from type 2 Gaucher patients with that of type 2 (null allele) Gaucher mice, which have a total absence of epidermal glucocerebrosidase activity. As was seen in type 2 Gaucher patients, epidermis from homozygous Gaucher mice revealed immature, partially processed lamellar body-derived sheets at all levels of the stratum corneum interstices(Fig. 4, A and B). Again, as described in epidermis from type 2 Gaucher patients, the stratum granulosum/stratum corneum interface appeared engorged with normal-appearing, newly secreted lamellar body contents, and the structure of lamellar bodies appeared normal. In contrast, heterozygous carrier mice which display partial glucocerebrosidase deficiency showed normal lamellar unit structures (Fig. 4C), comparable to those in the stratum corneum of type 1 and type 3 Gaucher patients (Fig. 3).

Figure 4
figure 4

Ultrastructure of stratum corneum in type 2 Gaucher mice. (A and B) Tissues from the type 2 (null allele) Gaucher mice reveal immature lamellar body-derived sheets (arrows) throughout the stratum corneum interstices, arranged in loose arrays.Asterisks indicate sites of possible phase separation during tissue processing. This appearance is indistinguishable from stratum corneum samples from type 2 Gaucher patients (cf. Fig. 2, A and C). (C) Stratum corneum from carrier mice reveal normal intercellular bilayers, with typical basic bilayer unit structure(arrows). Ruthenium tetroxide postfixation; bar = 0.1 μm.

Full size image

Alterations in epidermal lipid content in type 2 Gaucher patients and Gaucher mice. Inasmuch as deficient epidermal glucocerebrosidase activity results in an altered glucosylceramide:ceramide ratio in the stratum corneum of Gaucher mice(9) and in mice treated with the glucocerebrosidase inhibitor, bromoconduritol B epoxide(10), we investigated whether similar changes in lipid distribution occur in the epidermis of type 2 Gaucher patients. Although ceramides predominate over glucosylceramides in normal human and murine epidermis, HPTLC analysis of epidermal samples from two type 2 patients (cases 1 and 3) demonstrated increased glucosylceramides and decreased ceramides(Fig. 5). The ratio of glucosylceramide to ceramide was 2.7 ± 0.4 and 0.7 ± 0.3 in the type 2 samples versus normal epidermal samples (n = 4), respectively. No abnormalities of glucosylceramide to ceramide ratios were observed in epidermis from either type 1 (case 6 and 7) or type 3 (case 8) Gaucher patients, where the ratios were 1.2 ± 0.2 and 0.5 (a single determination), respectively. Epidermis from the Gaucher mice was also shown to have a reversed glucosylceramide:ceramide ratio, with a nearly 10-fold increase in glucosylceramides compared with epidermis from normal and carrier mice(9). These results demonstrate that severe deficiency of epidermal glucocerebrosidase results in an accumulation of glucosylceramides and an alteration in the glucosylceramide to ceramide ratio in the epidermis of type 2 Gaucher patients, an alteration which is not observed in skin from either type 1 or type 3 Gaucher patients.

Figure 5
figure 5

Epidermal sphingolipid content of type 2 patients. The content of total glycosylceramides was significantly increased (p< 0.01) and that of total ceramides decreased (p < 0.01) in epidermis from a patient with type 2 Gaucher disease (GD) (case 3) when compared with normal human epidermis. Each value represents the mean percent of total extracted lipids ± SEM for triplicate determinations of single patient samples.

Full size image

DISCUSSION

Although more than a century has elapsed since the first description of Gaucher disease in 1882(18), the full spectrum of sequelae from glucocerebrosidase deficiency continues to unfold. Gaucher disease has classically been divided into 3 types; type 1 (nonneuronopathic), type 2 (acute neuronopathic), and type 3 (chronic neuronopathic) based upon the presence and degree of neurologic involvement. Our description of skin ultrastructural abnormalities in type 2 Gaucher patients may potentially serve as a means to discriminate between Gaucher phenotypes. The observation that the type 2 (null allele) Gaucher mouse, which has a selective, profound deficiency of glucocerebrosidase, also displays skin abnormalities comparable to those found in type 2 Gaucher patients confirms that these skin lesions are related directly to the deficiency of this enzyme.

The finding that the type 2 Gaucher patients examined in this study, both with and without clinical evidence of skin disease, had distinctive ultrastructural and biochemical abnormalities of skin, is particularly noteworthy. In case 1, skin pathology was already suspected prenatally and immediately confirmed by biopsy. In contrast, in cases 2 and 3 epidermal abnormalities were identified only retrospectively from examination of autopsy specimen. Congenital ichthyosis has also been observed before neurologic deterioration in two siblings(19) who several months later developed neurologic abnormalities and subsequently were confirmed to have type 2 Gaucher disease.

In this report we demonstrate that the lack of epidermal glucocerebrosidase activity leads to glucosylceramide accumulation with a concomitant decrease in ceramide levels in epidermis from type 2 Gaucher patients, analogous to the similar alterations in the stratum corneum of type 2 (null allele) Gaucher mice(9). Prior studies have linked an increase in glucosylceramide to abnormal stratum corneum membrane structure resulting in compromised permeability barrier function(10, 13). The occurrence of ichthyosis in both type 2 Gaucher patients and in the type 2 Gaucher mice suggests that an alteration of epidermal glucosylceramide to ceramide ratios results not only in abnormal epidermal permeability barrier function, but also in altered cutaneous desquamation.

Although one could speculate that the extracellular lamellar bilayer abnormality may be responsible for the abnormality in desquamation, at least three other mechanisms should be considered. First, an increased glucosylceramide and decreased ceramide content in the stratum corneum could alter membrane cohesive properties leading to increased corneocyte retention. Second, the light microscopic histology of epidermis from both human type 2 Gaucher patients and the type 2 (null allele) Gaucher mice display not only hypercornification but also epidermal hyperplasia, as seen in a variety of other hyperproliferative disorders of cornification(20). Thus, the increased scale seen in certain type 2 Gaucher patients could result primarily from excess keratinocyte production due solely to hyperproliferation. In Gaucher disease, both tissue hypertrophy and hyperproliferation have been observed and attributed either to stimulation of cellular proliferation by accumulated glucosylceramide or to an, as yet, unidentified metabolite(21). A third pathogenic mechanism that could lead to abnormal desquamation is a permeability barrier abnormality. Barrier dysfunction is a prominent feature of both type 2 (null allele) Gaucher mice and topical bromoconduritol B epoxide-treated hairless mice(9, 11). Barrier function is an important regulator of epidermal DNA synthesis, and barrier disruption per se stimulates epidermal DNA synthesis, leading to epidermal hyperplasia(21, 22). Thus, severely depressed or totally absent glucocerebrosidase activity, leading to marked alteration of the glucosylceramide to ceramide ratio, may cause either abnormal intercellular cohesion resulting in clinical ichthyosis and/or induction of an abnormal barrier that leads to hyperproliferation.

The observation that type 2, but not type 1 or 3, Gaucher patients have distinctive epidermal ultrastructural and lipid biochemical abnormalities(Table 2) could allow more rapid and accurate assignment of prognosis for Gaucher patients diagnosed prenatally or in infancy. Neither enzyme activity nor genotypic analysis are adequate to determine whether an infant or fetus is affected with type 2 rather than types 1 or 3 disease(1, 23). Currently, the most useful conclusion from studies of genotype-phenotype correlation is that patients with mutation N370S are unlikely to develop neurologic disease. However, the prediction of prognosis from other genotypes is more problematic(24). The ability to distinguish type 2 from type 1 or 3 Gaucher disease becomes particularly relevant with the availability of successful enzyme therapy(25–27), where early, aggressive therapeutic intervention may be of benefit for type 1 and 3 patients, but is unlikely to alter the devastating neurologic deterioration that occurs in type 2 disease(28). In addition, parents might also be reassured by counseling that could ascertain that an affected infant did not have type 2 disease. The ultrastructural and/or lipid biochemical evaluation of skin should provide an invaluable addition to enzymatic assays and genotyping to identify type 2 Gaucher disease in young patients. Analyses on additional Gaucher patients will help to confirm the general usefulness of this approach.

Table 2 Epidermal ultrastructure and lipid abnormalities
Full size table