Search

We have previously shown that intracellular trafficking and extracellular assembly of tropoelastin into elastic fibers is facilitated by the 67-kD elastin-binding protein identical to an enzymatically inactive, alternatively spliced variant of beta-galactosidase (S-Gal). In the present study, we investigated elastic-fiber assembly in cultures of dermal fibroblasts from patients with either Morquio B disease or GM1-gangliosidosis who bore different mutations of the beta-galactosidase gene. We found that fibroblasts taken from patients with an adult form of GM1-gangliosidosis and from patients with an infantile form, carrying a missense mutations in the beta-galactosidase gene-mutations that caused deficiency in lysosomal beta-galactosidase but not in S-Gal-assembled normal elastic fibers. In contrast, fibroblasts from two cases of infantile GM1-gangliosidosis that bear nonsense mutations of the beta-galactosidase gene, as well as fibroblasts from four patients with Morquio B who had mutations causing deficiency in both forms of beta-galactosidase, did not assemble elastic fibers. We also demonstrated that S-Gal-deficient fibroblasts from patients with either GM1-gangliosidosis or Morquio B can acquire the S-Gal protein, produced by coculturing of Chinese hamster ovary cells permanently transected with S-Gal cDNA, resulting in improved deposition of elastic fibers. The present study provides a novel and natural model validating functional roles of S-Gal in elastogenesis and elucidates an association between impaired elastogenesis and the development of connective-tissue disorders in patients with Morquio B disease and in patients with an infantile form of GM1-gangliosidosis.

We have identified and characterized three missense mutations in a patient with type 1 G(M1) gangliosidosis, namely a substitution of G for A at nucleotide position 1044 (G1044-->A; in exon 10) on one allele, which converts Asp(332) into asparagine, and both a mutation (C492-->A in exon 4, leading to the amino acid change of Arg(148)-->Ser) and a polymorphism (A1644-->G in exon 15, leading to a change of Ser(532)-->Gly) on the other allele. This patient had less than 1% residual beta-galactosidase activity and minimally detectable levels of immunoreactive beta-galactosidase protein in fibroblasts. To account for the above findings, a series of expression and immunolocalization studies were undertaken to assess the impact of each mutation. Transient overexpression in COS-1 cells of cDNAs encoding Asp(332)Asn, Arg(148)Ser and Ser(532)Gly mutant beta-galactosidases produced abundant amounts of precursor beta-galactosidase, with activities of 0, 84 and 81% compared with the cDNA clone for wild-type beta-galactosidase (GP8). Since the level of vector-driven expression is much less in Chinese hamster ovary (CHO) cells than in COS-1 cells, and we knew that exogenous beta-galactosidase undergoes lysosomal processing when expressed in these cells, transient expression studies were performed of Arg(148)Ser and Ser(532)Gly, which yielded active forms of the enzyme. In this case, the Arg(148)Ser and Ser(532)Gly products gave rise to 11% and 86% of the control activity respectively. These results were not unexpected, since the Arg(148)Ser mutation introduced a major conformational change into the protein, and we anticipated that it would be degraded in the endoplasmic reticulum (ER), whereas the polymorphism was expected to produce near-normal activity. To examine the effect of the Asp(332)Asn mutation on the catalytic activity, we isolated CHO clones permanently transfected with the Asp(332)Asn and Asp(332)Glu constructs, purified the enzymes by substrate-analogue-affinity chromatography, and determined their kinetic parameters. The V(max) values of both mutant recombinant enzymes were markedly reduced (less than 0.9% of the control), and the K(m) values were unchanged compared with the corresponding wild-type enzyme isolated at the same time. Both the Arg(148)Ser beta-galactosidase in CHO cells and Asp(332)Asn beta-galactosidases (in COS-1 and CHO cells) produced abundant immunoreaction in the perinuclear area, consistent with localization in the ER. A low amount was detected in lysosomes. Incubation of patient fibroblasts in the presence of leupeptin, which reduces the rate of degradation of lysosomal beta-galactosidase by thiol proteases, had no effect on residual enzyme activity, and immunostaining was again detected largely in the perinuclear area (localized to the ER) with much lower amounts in the lysosomes. In summary, the Arg(148)Ser mutation has no effect on catalytic activity, whereas the Asp(332)Asn mutation seriously reduces catalytic activity, suggesting that Asp(332) might play a role in the active site. Immunofluorescence studies indicate the expressed mutant proteins with Arg(148)Ser and Asp(332)Asn mutations are held up in the ER, where they are probably degraded, resulting in only minimum amounts of the enzyme becoming localized in the lysosomes. These results are completely consistent with findings in the cultured fibroblasts. Our results imply that most of the missense mutations described in G(M1) gangliosidosis to date have little effect on catalytic activity, but do affect protein conformation such that the resulting protein cannot be transported out of the ER and fails to arrive in the lysosome. This accounts for the minimal amounts of enzyme protein and activity seen in most G(M1) gangliosidosis patient fibroblasts.

GM1 gangliosidosis and Morquio B disease are distinct disorders both clinically and biochemically yet they arise from the same beta-galactosidase enzyme deficiency. On the other hand, galactosialidosis and sialidosis share common clinical and biochemical features, yet they arise from two separate enzyme deficiencies, namely, protective protein/cathepsin A and neuraminidase, respectively. However distinct, in practice these disorders overlap both clinically and biochemically so that easy discrimination between them is sometimes difficult. The principle reason for this may be found in the fact that these three enzymes form a unique complex in lysosomes that is required for their stability and posttranslational processing. In this review, I focus mainly on the primary and secondary beta-galactosidase deficiency states and offer some hypotheses to account for differences between GM1 gangliosidosis and Morquio B disease.

Our previous studies showed immunological and functional similarities, as well as partial sequence homology, between the enzymatically inactive alternatively spliced variant of human beta-galactosidase (S-gal) and the 67-kDa elastin/laminin-binding protein (EBP) from sheep. To define the genetic origin of the EBP further, a full-length human S-gal cDNA clone was constructed and subjected to in vitro transcription/translation. The cDNA was also transfected into COS-1 cells and into the EBP-deficient smooth muscle cells (SMC) from sheep ductus arteriosus (DA). In vitro translation yielded an unglycosylated form of the S-gal protein, which immunoreacted with anti-beta-galactosidase antibodies and bound to elastin and laminin affinity columns. S-gal cDNA transfections into COS-1 and DA SMC increased expression of a 67-kDa protein that immunolocalized intracellularly and to the cell surface and, when extracted from the cells, bound to elastin. The S-gal-transfected cells displayed increased adherence to elastin-covered dishes, consistent with the cell surface distribution of the newly produced S-gal-encoded protein. Transfection of DA SMC additionally corrected their impaired elastic fiber assembly. These results conclusively identify the 67-kDa splice variant of beta-galactosidase as EBP.

Lysosomal beta-galactosidase precursor is processed to a mature form and associated with protective protein in lysosomes. In this study we used two cysteine protease proinhibitors, E64-d for cathepsins B, S, H, and L, and CA074Me for cathepsin B. They are converted intracellularly to active forms, E-64c and CA074, respectively. Both active compounds inhibited maturation of the exogenous beta-galactosidase precursor, but E-64c did not inhibit further degradation to an inactive 50-kDa product. We concluded that cathepsin B participated exclusively in maturation of beta-galactosidase, and a non-cysteine protease was involved in further degradation and inactivation of the enzyme molecule.

Human lysosomal beta-galactosidase catalyzes the hydrolysis of beta-galactosides via a double displacement mechanism involving a covalent glycosyl enzyme intermediate. By use of the slow substrate 2,4-dinitrophenyl-2-deoxy-2-fluoro-beta-D-galactopyranoside, a glycosyl enzyme intermediate has been trapped on the enzyme. This has allowed the catalytic nucleophile to be identified as Glu-268 by peptic and tryptic digestion of the inactivated enzyme followed by high performance liquid chromatography-electrospray ionization tandem mass spectrometry of the peptide mixture. This glutamic acid is fully conserved in a sequence-related family of enzymes (Family 35), consistent with its essential role.

Processing of human beta-galactosidase (beta-GAL) was studied in permanently transfected Chinese hamster ovary (CHO) cells and compared with that in normal cells and in cells from subjects with GM1-gangliosidosis, galactosialidosis and I-cell disease. Biosynthesis of beta-GAL in CHO cells results in the synthesis of an 88 kDa glycosylated and phosphorylated monomer precursor which is enzymically active and is secreted into the medium. Post-translational processing begins at the C-terminal end of the protein and gives rise to structurally related 67 and 64 kDa mature forms. These are subsequently degraded to give several inactive products of which a 50 kDa and a 18 kDa species are prominent. In normal fibroblasts only the 84 kDa precursor is readily detected inside cells, while the 88 kDa precursor is the only form secreted from cells in the presence of ammonium chloride. Processing of the precursor in normal fibroblasts results in the appearance of both the 67 and 64 kDa mature forms, which are also degraded to give 50 and 18 kDa products, as in transfected CHO cells. As affected controls, GM1-gangliosidosis cells showed a general loss of all forms of the enzyme, while in I-cell fibroblasts only the 84 kDa precursor and an 18 kDa degradation form were prominent. In galactosialidosis fibroblasts, taken from two different subjects, processing of beta-GAL was characterized by the respective appearance of intermediate 80 and 72 kDa enzymically inactive polypeptides, at levels lower than the normal amounts of the 67 and 64 kDa mature forms and higher than the normal amounts of degradation products, one of which is of 45 kDa and arises by endoproteolytic cleavage of the 80 kDa polypeptide. Incubation for up to 72 h in medium containing leupeptin, a potent inhibitor of thiol-dependent proteases, resulted in a significantly increased level of beta-GAL activity to near normal levels in fibroblasts from one galactosialidosis subject. Concordant with this, the abundance of the 84 kDa precursor was increased and the levels of the 80 kDa, 45 kDa and 18 kDa digestion products were diminished. However, in fibroblasts from the second galactosialidosis subject, the amount of the abnormal 72 kDa polypeptide was not influenced by leupeptin treatment. Leupeptin treatment did not increase enzymic activity levels in normal, GM1-gangliosidosis or I-cell disease fibroblasts, despite the fact that the production of the 50 kDa and 18 kDa degradation products was blocked in the presence of leupeptin. We concluded that in galactosialidosis the leupeptin-inhibitable proteolytic cleavage of a small fragment causes a conformational change of the precursor that precludes its further normal processing and results in its enzymic deficiency. This early abnormal trimming of beta-GAL is ascribable to a deficiency in the functional protective protein, the function of which is absolutely essential to render beta-GAL cryptic from at least two distinct and separate proteolytic attacks that together remove at least 12 kDa from the C-terminal end of the enzyme.

Chinese hamster ovary cell clones permanently transfected with the cDNA for human lysosomal beta-galactosidase secrete the enzyme precursor into the cell medium, from which it is purified to apparent homogeneity in a single step by affinity chromatography. The purified precursor is fully active, displays the same pH optimum and Km values as the mature placental enzyme, and has an intact C-terminus. The intact enzyme when chromatographed on a Sephacryl S-200 molecular-sieve column elutes as a 105,500 Da monomer, whereas on SDS/PAGE gels the polypeptide migrates as an 88 kDa polypeptide. A time course of digestion with glycopeptide-N-glycanase shows the gradual conversion of the precursor from an 88 to a 72 kDa protein, suggesting the presence of five N-linked oligosaccharides in the protein. The precursor is readily taken up in a mannose-6-phosphate-dependent manner into beta-galactosidase-deficient, GM1-gangliosidosis fibroblasts, and the enzyme activity is returned to normal levels. We show that the stereochemical course of enzymic hydrolysis involves the retention of the beta-configuration at the anomeric centre, suggesting a double-displacement mechanism. Furthermore, the enzyme is rapidly and irreversibly inactivated in the presence of the mechanism-based inactivator 2,4-dinitrophenyl-2-deoxy-2-fluoro-beta-D-galactopyranoside, which implicates a covalent intermediate. The enzyme is also inactivated by 1-ethyl-3(3-dimethylamino-propyl)carbodi-imide and by phenylglyoxal, which implicates carboxylate and arginine residues respectively in the active site. We conclude that the beta-galactosidase precursor is functionally identical to the mature lysosomal form of the enzyme and serves as an excellent enzyme source for investigation of structure-function relationships in the protein.

Galactosialidosis is a heterogeneous disorder that is manifested in infantile, late infantile, juvenile/adult, and atypical forms. In every instance the primary defect is in the ability of protective protein to associate with beta-galactosidase and neuraminidase to protect them from intralysosomal proteolysis. The protective protein is in reality a serine protease that displays both cathepsin A and C-terminal deamidase activity. We summarize the major clinical features of each form, and the range of storage products accumulated. The concept of an intralysosomal complex containing beta-galactosidase and neuraminidase in addition to protective protein seems irrefutable but major gaps exist in our understanding of how the complex is formed and in what subcellular organelles, how it is sustained, and the protein domains contributed by the constituent enzymes that play a pivotal role in this process.

We and others have previously shown that a 67-kD cell surface elastin/laminin-binding protein (EBP) is responsible for cell adhesion to elastin and laminin and for mediating the process of elastin fiber assembly, but the nature of this protein was unknown. In this report we provide evidence that a 67-kD catalytically inactive form of beta-galactosidase produced by alternative splicing demonstrates immunological and functional similarity and sequence homology to the 67-kD EBP, suggesting that the two might be the same. Antibody prepared to a synthetic peptide, N-Ac-GSPSAQDEASPL, corresponding to a frame-shift-generated sequence unique to the alternatively spliced form of human beta-galactosidase, also recognized sheep EBP both on Western blotting and in aortic tissue. Furthermore, this synthetic peptide (S-GAL) binds to elastin and laminin, but not to fibronectin, collagen I, or collagen III. Moreover, both tropoelastin and laminin which bind to S-GAL peptide affinity columns can be specifically eluted from them with an excess of free S-GAL peptides. In addition, sequence homology among this splice variant of human beta-galactosidase, sheep EBP, and NH2-terminal sequences of some elastases suggests that these proteins share a common ligand-binding motif that has not been previously recognized.

Lysosomal beta-galactosidase (beta-Gal) occurs either alone in monomeric and dimeric forms, or in a high-M(r) complex with at least two additional proteins. One is neuraminidase and the second is the protective protein, which has also been shown to possess carboxypeptidase activity. beta-Gal activity is deficient in GM1-gangliosidosis as a primary defect, and is secondarily affected in galactosialidosis (GS), where the primary defect is the absence of protective protein activity. Fibroblasts from three patients with GM1-gangliosidosis, type 1, showed markedly reduced amounts of beta-Gal cross-reacting material (CRM), and a fourth appeared to have normal levels. A patient with type 2 GM1-gangliosidosis was also found to be CRM-normal. These findings demonstrate that patients with GM1-gangliosidosis type 1 are heterogeneous with respect to the level of residual beta-Gal protein. Fibroblasts from four patients with GS were strongly CRM-positive with an anti-beta-Gal antibody, as was a sample of brain from one of these patients, suggesting that the loss of beta-Gal activity is linked to a subtler change in the primary structure of the enzyme than has been previously thought. While three GS cell lines displayed reduced carboxypeptidase activity (to 32-42% of the control), one cell line was completely devoid of activity, demonstrating that while carboxypeptidase activity is a property of the protective protein this action is distinct and separate from its protective role. On direct immunoprecipitation with anti-beta-Gal antibody, a portion of the total carboxypeptidase activity co-precipitated with beta-Gal from extracts of normal and GM1-gangliosidosis cells, consistent with the presence of the complex in these cells. However, no carboxypeptidase activity was precipitable with this antibody from GS fibroblasts, suggesting the absence of complex from these cells. To examine this further, the various forms of beta-Gal were resolved by h.p.l.c. molecular-sieve chromatography. Three forms of beta-Gal activity were resolved in normal cells: a complex, a dimer and a monomer. Residual beta-Gal activity of GS cells resolved into two of these forms, the complex and the monomer. In normal and GM1-gangliosidosis cells a portion of the total carboxypeptidase activity co-chromatographed with the complex while the bulk of the activity occurred in a single 36,000-M(r) peak. Only the low-M(r) carboxypeptidase activity was detected in GS cells. This confirms our results on immunoprecipitation indicating that portions of the beta-Gal and the carboxypeptidase activities exist outside the complex in normal, GM1-gangliosidosis and GS cells. In summary, the loss of protective protein function from GS cells results in disproportionate loss of the dimeric and monomeric forms of beta-Gal activity, but does not result in the complete degradation of the protein.

GM1 ganglioside beta-galactosidase (beta-Gal) is deficient in the autosomal recessive disorder GM1 gangliosidosis. A portion of the enzyme occurs in a complex with neuraminidase and an additional glycoprotein, protective protein, but the nature of the interactions conferring the stability of the complex is unknown. Affinity chromatography of beta-Gal on p-aminophenylthiogalactose-Sepharose (PATG-Sepharose) at pH 4.3, the pH optimum of beta-Gal, resulted in a 260-fold enrichment of beta-Gal, but the major protein in the fraction had an M(r) value of 74,000. Affinity chromatography on PATG-Sepharose at pH 5.2 showed substantial enrichment (4000-fold) of beta-Gal, and the mature form of the enzyme (M(r) 64,000) was the major protein in the preparation. Using h.p.l.c. molecular-sieve chromatography, we found that about 15% of the total beta-Gal occurred in a high-M(r) form (greater than 600,000), the presumptive complex, with 85% eluting at M(r) 150,000, suggestive of a dimer. This distribution was independent of both high (60 mg/ml) and low (5 mg/ml) protein concentration and the pH (pH 4.3 or 5.2) of the sample applied to the column. Furthermore, incubation for 90 min at 37 degrees C, conditions which had previously been suggested as optimal for formation of the complex, had no effect on this distribution. Further fractionation by anion-exchange chromatography and a second affinity column step yielded a beta-Gal preparation that contained a single polypeptide chain (M(r) 64,000), was devoid of neuraminidase and protective protein (absent carboxypeptidase activity), and when injected into rabbits gave rise to monospecific rabbit antisera. We conclude that the protein composition of the complex is variable (i.e. it is different when isolated at pH 4.3 and 5.2) and that the amount of beta-Gal tightly associated with the complex constitutes a small fraction of the total beta-Gal activity. The more prevalent form of the enzyme is a beta-Gal homodimer that is stable and devoid of either neuraminidase activity or protective protein.

beta-Galactosidase has been purified from rabbit brain by a procedure which gives substantially greater purification and yield than the previously reported method. The enzyme was solubilized in 4% aqueous butanol and purified by a procedure which included affinity chromatography on columns of concanavalin A - Sepharose (ConA-Sepharose) and aminophenylthiogalactoside Sepharose (APT galactoside-Sepharose). The activity was resolved by DEAE-Sepharose chromatography into two fractions; the first did not bind to the column and was eluted in the unbound fraction, while the second bound to the column and could be eluted with a salt gradient. The unbound and bound fractions were purified 7600-fold and 4900-fold, respectively, by the newly developed procedure. Both gave two closely migrating protein bands on polyacrylamide gel electrophoresis and the major contaminating protein was beta-hexosaminidase in each case. The enzyme in the unbound fraction had an isoelectric point (pI) of 6.7 and an apparent molecular weight by gel filtration of 114 000 +/- 10 000. The enzyme that bound to DEAE-Sepharose at pH 8.0 and comprised about 60% of the total acid beta-galactosidase activity of rabbit brain eluted from Sephacryl S-200 as a single peak with apparent molecular weight 140 000 +/- 10 000. The two fractions displayed similar relative specific activities towards the synthetic substrates and ganglioside GM1 and lactosyl ceramide. Neither enzyme hydrolyzed galactosyl ceramide. We have immunological evidence which suggests that the two fractions of beta-galactosidase are also structurally related.

beta-Galactosidase purified to apparent homogeneity from human placenta occurred in two separable fractions. A low molecular mass form (relative mass (Mr) 170 000) is composed of a single polypeptide chain (Mr 70 000). This was derived from a larger form by molecular sieve chromatography at both low (10 mM) and high (500 mM) NaCl concentration. The larger form of beta-galactosidase also contains small amounts of two polypeptides with apparent Mr values of 23 000 and 35 000 daltons. Both forms of the enzyme hydrolyze synthetic aryl galactosides and natural glycolipid substrates at comparable rates. Antibodies raised in rabbits against the low Mr beta-galactosidase also cross-reacts with the high Mr enzyme. The antibody preparation also cross-reacted with beta-hexosaminidase even though the latter was found at very low levels in the antigen, as judged by lack of detection of representative protein bands on sodium dodecyl sulfate - polyacrylamide gel electrophoresis and enzyme activity measurements. A portion of this cross-reactivity (35%) against beta-hexosaminidase could not be absorbed from the preparation without the simultaneous loss of beta-galactosidase activity, suggesting that the two enzymes show a degree of antigenic identity.

Neuraminidase can be activated by incubation of crude glycoprotein fractions at acidic pH for 90 minutes at physiological temperature. This activation is inhibited by leupeptin. Incubation of the purified neuraminidase-beta-galactosidase-protective protein complex under the same conditions used for crude glycoprotein fractions did not lead to enhanced neuraminidase activity, but incubation in the presence of exogenous Cathepsin C at 4 degrees C resulted in marked enhancement of neuraminidase activity. This activation was again inhibited by leupeptin. Cathepsin D treatment resulted in destruction of neuraminidase under the same conditions and this effect was again inhibited by leupeptin. beta-galactosidase in crude glycoprotein fractions and in the complex was resistant to both Cathepsin C and D, while homogeneous beta-galactosidase was inactivated by these enzymes. We suggest that in vitro activation of neuraminidase may mimic the in vivo intralysosomal conversion of the neuraminidase precursor into the mature form of the enzyme.

Human placental hexosaminidase B and beta-galactosidase are taken up very poorly by human fibroblasts in culture. However, if fibroblasts manifesting genetically determined deficiencies of these lysosomal hydrolases are first treated with concanavalin A, then enzyme uptake is markedly increased. Enzyme activity which becomes associated with concanavalin A-treated fibroblasts maintained at 4 degrees C can be greatly removed by treatment with haptene sugar, while enzyme activity which becomes associated with cells maintained at 37 degrees C is refractory to haptens treatment. These results are interpreted as an initial binding of enzyme to concanvalin A molecules located at the cell surface, followed by an active cellular process leading to internalization of the lectin-enzyme complexes.