Immunoelectron microscopical localization of lysosomal beta-galactosidase and its precursor forms in normal and mutant human fibroblasts

Immunoelectron microscopy was performed to study the biosynthesis of lysosomal beta-galactosidase (beta-gal) in normal and mutant human fibroblasts. Using polyclonal and monoclonal antibodies we show in normal cells precursor forms of beta-gal in the rough endoplasmic reticulum (RER) and in the Golgi apparatus throughout the stack of cisternae. In the lysosomes virtually all beta-gal exists as a high molecular weight multimer of mature enzyme. In the autosomal recessive disease GM1-gangliosidosis caused by a beta-gal deficiency and in galactosialidosis, associated with a combined deficiency of lysosomal neuraminidase and beta-gal, precursor forms of the latter enzyme are found in RER, Golgi and some labeling is present at the cell surface. The lysosomes remain unlabeled, indicative for the absence of enzyme molecules in this organelle. In galactosialidosis fibroblasts also no mature beta-gal is found in the lysosomes but in these cells the presence of the monomeric form can be increased by leupeptin (inhibition of proteolysis) whereas addition of a partly purified 32 kDa "protective protein" results in the restoration of high molecular weight beta-gal multimers in the lysosomes.     

The relation between human lysosomal beta-galactosidase and its protective protein

Cultured skin fibroblasts from patients with the lysosomal storage disease galactosialidosis lack a 54-kDa protein which is a precursor of 32-kDa and 20-kDa proteins, which immunoprecipitate with human anti-beta-galactosidase antiserum. The lack of a 32-kDa "protective protein" results in a combined deficiency of beta-galactosidase and sialidase. The mechanism of protection of lysosomal beta-galactosidase against proteolytic degradation is elucidated by sucrose density gradient centrifugation and immunoprecipitation studies. In normal fibroblasts at the low intralysosomal pH, more than 85% of beta-galactosidase exists as a high molecular weight (600-700 kDa) multimer and about 10% as a monomer of 64-kDa. In mutant cells from galactosialidosis patients, the residual enzyme activity, about 10%, is present as a monomer and no multimer exists. After addition of the 54-kDa precursor form of the protective protein, the density pattern of beta-galactosidase in galactosialidosis cells is normalized. Immunoprecipitation studies after sucrose density gradient centrifugation on homogenate and on purified beta-galactosidase from normal fibroblasts show that the protective protein is associated only with the multimeric form of beta-galactosidase. We propose that intralysosomal protection against proteolysis of beta-galactosidase and sialidase is accomplished by aggregation into a high molecular weight complex consisting of multimeric beta-galactosidase, sialidase, and protective protein. The genetic deficiency of the latter, as in galactosialidosis, results in a rapid degradation of monomeric beta-galactosidase and a loss of sialidase activity.     

Novel mutations (Asn 484 Lys,Thr 500 Ala,Gly 438 Glu) in Morquio B disease

Primary deficiency of beta-galactosidase results in GM1 gangliosidosis and Morquio B disease. Of the more than 40 disease-causing mutations described in the Gal gene to date, about 75% are of the missense type and are scattered along the length of the gene. No single, major common mutation has been associated with GM1 gangliosidosis. However, a Trp 273 Leu mutation has been commonly found in the majority of patients with Morquio B disease defined genotypically to date.We now report three new mutations in three Morquio B patients where the Trp 273 Leu mutation is absent. Two of the mutations, C1502G (Asn 484 Lys) and A1548G (Thr 500 Ala), were found in twins (one male, one female) who display a mild form of Morquio B disease and keratan sulfate in the urine. In their fibroblasts, residual activity was 1.9% and 2.1% of controls. On Western blots, the 84-kDa precursor and the 64-kDa mature protein were barely detectable. The occurrence of a 45-kDa degradation product indicates that the mutated protein reached the lysosome but was abnormally processed. In the third case, we identified only a G1363A (Gly 438 Glu) mutation (a major deletion on the second allele has not been ruled out). This female patient too displays a very mild form of the disease with a residual activity of 5.7% of control values. In fibroblasts from this case, the 84-kDa precursor and the 45-kDa degradation product were present, while the mature 64-kDa form was barely detectable. The occurrence of these three mutations in the same area of the protein may define a domain involved in keratan sulfate degradation.     

Expression of cDNA encoding the human "protective protein" associated with lysosomal beta-galactosidase and neuraminidase: homology to yeast proteases

The "protective protein" is a glycoprotein that associates with lysosomal beta-galactosidase and neuraminidase and is deficient in the autosomal recessive disorder galactosialidosis. We have isolated the cDNA encoding human "protective protein". The clone recognizes a 2 kb mRNA in normal cells that is not evident in fibroblasts of an early infantile galactosialidosis patient. The cDNA directs the synthesis of a 452 amino acid precursor molecule that is processed in vivo to yield mature "protective protein," a heterodimer of 32 kd and 20 kd polypeptides held together by disulfide bridges. This mature form is also biologically functional since it restores beta-galactosidase and neuraminidase activities in galactosialidosis cells. The predicted amino acid sequence of the "protective protein" bears homology to yeast carboxypeptidase Y and the KEX1 gene product. This suggests a protease activity for the "protective protein."     

Mouse "protective protein". cDNA cloning, sequence comparison, and expression

The "protective protein" is the glycoprotein that forms a complex with the lysosomal enzymes beta-galactosidase and neuraminidase. Its deficiency in man leads to the metabolic storage disorder galactosialidosis. The primary structure of human protective protein, deduced from its cloned cDNA, shows homology to yeast serine carboxypeptidases. We have isolated a full-length cDNA encoding murine protective protein. The nucleotide sequences as well as the predicted amino acid sequences are highly conserved between man and mouse. Domains important for the protease function are completely identical in the two proteins. Both human and mouse mature protective proteins covalently bind radiolabeled diisopropyl fluorophosphate. Transient expression of the murine cDNA in COS-1 cells yields a protective protein precursor of 54 kDa, a size characteristic of the glycosylated form. This cDNA-encoded precursor, endocytosed by human galactosialidosis fibroblasts, is processed into a 32- and a 20-kDa heterodimer and corrects beta-galactosidase and neuraminidase activities. A tissue-specific expression of protective protein mRNA is observed when total RNA from different mouse organs is analyzed on Northern blots.     

Galactosialidosis: molecular heterogeneity in biosynthesis and processing of protective protein for β-galactosidase

Summary Biosynthesis and processing of the protective protein for -galactosidase in normal and galactosialidosis fibroblasts were investigated using specific antiserum preparations. A 45-kd precursor was processed to a mature 30-kd protein in normal fibroblasts. The mature protective protein was not detected in any of the twelve galactosialidosis fibroblast strains examined in this study. The precursor was not detected in two cases and in the others was of heterogeneous molecular weight, i.e., normal, abnormally low, or abnormally high. These molecular abnormalities were not correlated with clinical manifestations of the patients.     

Incorporation and degradation of GM1 ganglioside and asialoGm1 ganglioside in cultured fibroblasts from normal individuals and patients with β-galactosidase deficiency

The uptake and degradation of G_M1 ganglioside (G_M1) and asialoG_M1 ganglioside (G_A1) were studied in cultured fibroblasts from normal individuals and patients with β-galactosidase deficiency, using the lipid-loading test. The glycolipids were incorporated from the media into the fibroblasts and the terminal galactose was hydrolyzed in normal cells. The hydrolysis rates of G_A1 were 80–86% of normal on the 3rd day after loading, while G_M1 was hydrolyzed slowly; 35–54% on the 14th day. In infantile G_M1 gangliosidosis and I-cell disease, little G_M1 and G_A1 was hydrolyzed on any day of culture, while fibroblasts from patients with adult G_M1 gangliosidosis, Morquio disease type B and galactosialidosis hydrolyzed the lipids at nearly normal rates. The intracellular accumulation of the glycolipids, on the basis of protein content, was abnormally high in the case of infantile G_M1 gangliosidosis and I-cell disease, but normal in the other disorders examined. These observations indicate that the in situ metabolism of G_M1 and G_A1 is probably normal in fibroblasts from patients with adult G_M1 gangliosidosis, Morquio disease type B and galactosialidosis, although in vitro β-galactosidase activities in these disorders are very low. The results are compatible with findings that G_M1 and G_A1 do not accumulate in the somatic organs of patients with adult G_M1 gangliosidosis and galactosialidosis. In I-cell disease, however, the results of the loading test did not agree with the finding that there is little accumulation of glycolipids in postmortem tissues.     

Characterization and sequence of a Thermomonospora fusca xylanase.

TfxA is a thermostable xylanase produced by the thermophilic soil bacterium Thermomonospora fusca. The enzyme was purified to homogeneity from the culture supernatant of Streptomyces lividans transformed by plasmid pGG92, which carries the gene for TfxA, xynA. The molecular mass of TfxA by sodium dodecyl sulfate-polyacrylamide gel electrophoresis is 32 kDa. TfxA is extremely stable, retaining 96% of its activity after 18 h at 75 degrees C. It has a broad pH optimum around pH 7 and retains 80% of its maximum activity between pH 5 and 9. The native enzyme binds strongly to both cellulose and insoluble xylan even though it has no activity on cellulose. Treatment of TfxA with a T. fusca protease produced a 24-kDa catalytically active fragment that had the same N-terminal sequence as TfxA. The fragment does not bind to cellulose and binds weakly to xylan. The Vmax values for TfxA and the fragment are 600 and 540 mumol/min/mg, respectively, while the Kms are 1.1 and 2.3 mg of xylan per ml, respectively. The DNA sequence of the xynA gene was determined, and it contains an open reading frame that codes for a 42-amino-acid (42-aa) actinomycete signal peptide followed by the 32-kDa mature protein. There is a 21-aa Gly-Pro-rich region that separates the catalytic domain from an 86-aa C-terminal binding domain. The amino acid sequence of the catalytic domain of TfxA has from 40 to 72% identity with the sequence of 12 other xylanases from seven different organisms and belongs to family G.(ABSTRACT TRUNCATED AT 250 WORDS)     

GM1-gangliosidosis. Defective recognition site on beta-galactosidase precursor

Cultured fibroblasts from different variants of GM1-gangliosidosis synthesize normal amounts of 88-kDa beta-galactosidase precursor. Yet the amount of the mature 64-kDa form is reduced to 5-15% of normal values. In this communication it is shown that the mutation in the infantile and adult form of GM1-gangliosidosis interferes with the phosphorylation of precursor beta-galactosidase. As a result the precursor is secreted instead of being compartmentalized into the lysosomes and further processed. The impaired phosphorylation might be due to conformational changes of the precursor molecule.     

β-Galactosidase-deficient human fibroblasts: uptake and processing of the exogenous precursor enzyme expressed by stable transformant COS cells

Summary COS-1 cells were transfected by electroporation with a cDNA for human acid -galactosidase cloned in our laboratory and stable transformants expressing the enzyme activity were selected. The precursor form of the enzyme was secreted in large quantities into the culture medium. The fibroblasts from patients with G_M1-gangliosidosis or Morquio B disease showed a remarkable increase of enzyme activity, up to the normal level, after culture in this medium for 2 days; the amount of uptake was essentially the same as that for the precursor form in human fibroblasts. After endocytosis, the precursor molecules were processed normally to the mature form and remained as stable as those produced by human fibroblasts. On the other hand, cells from galactosialidosis patients did not show any increase of enzyme activity in a similar experiment. It was concluded that the transformants are useful as the source of precursor proteins for the study of intracellular turnover of enzyme molecules in mutant cells.     

Incorporation and degradation of GM1 ganglioside and asialoGM1 ganglioside in cultured fibroblasts from normal individuals and patients with beta-galactosidase deficiency

The uptake and degradation of GM1 ganglioside (GM1) and asialoGM1 ganglioside (GA1) were studied in cultured fibroblasts from normal individuals and patients with beta-galactosidase deficiency, using the lipid-loading test. The glycolipids were incorporated from the media into the fibroblasts and the terminal galactose was hydrolyzed in normal cells. The hydrolysis rates of GA1 were 80-86% of normal on the 3rd day after loading, while GM1 was hydrolyzed slowly; 35-54% on the 14th day. In infantile GM1 gangliosidosis and I-cell disease, little GM1 and GA1 was hydrolyzed on any day of culture, while fibroblasts from patients with adult GM1 gangliosidosis, Morquio disease type B and galactosialidosis hydrolyzed the lipids at nearly normal rates. The intracellular accumulation of the glycolipids, on the basis of protein content, was abnormally high in the case of infantile GM1 gangliosidosis and I-cell disease, but normal in the other disorders examined. These observations indicate that the in situ metabolism of GM1 and GA1 is probably normal in fibroblasts from patients with adult GM1 gangliosidosis, Morquio disease type B and galactosialidosis, although in vitro beta-galactosidase activities in these disorders are very low. The results are compatible with findings that GM1 and GA1 do not accumulate in the somatic organs of patients with adult GM1 gangliosidosis and galactosialidosis. In I-cell disease, however, the results of the loading test did not agree with the finding that there is little accumulation of glycolipids in postmortem tissues.     

C-terminal processing of tyrosinase is responsible for activation of Pholiota microspora proenzyme

Tyrosinase is expressed as a 67-kDa protein in Pholiota microspora (synonym Pholiota nameko), whereas the same enzyme purified from fruiting bodies of P. microspora is a 42-kDa protein that is cleaved with a C-terminal 25-kDa polypeptide from the 67-kDa protein. To confirm the role of C-terminal processing in enzyme activity, we expressed a recombinant 67-kDa tyrosinase in Escherichia coli cells. To obtain a soluble protein, the recombinant tyrosinase was expressed as a thioredoxin fusion protein with an enterokinase-cleavable site. Enterokinase digestion of the fusion protein produced a recombinant 67-kDa tyrosinase that did not have any catalytic activity. However, chymotrypsin digestion of the fusion protein produced a recombinant 44-kDa tyrosinase that was catalytically active and had a 25-kDa cleaved C-terminal. Kinetic parameters of the 44-kDa tyrosinase were similar to those of the 42-kDa tyrosinase purified from the fruiting bodies. These results suggest that tyrosinase is expressed in P. microspora as a latent 67-kDa proenzyme and is converted to the mature active 42-kDa enzyme by proteolytic processing of the C-terminal.     

Identification and properties of the catalytic domain of mammalian DNA polymerase beta

Rat DNA polymerase beta (beta-pol) is a 39-kDa protein organized in two tightly folded domains, 8-kDa N-terminal and 31-kDa C-terminal domains, connected by a short protease-sensitive region. The 8-kDa domain contributes template binding to the intact protein, and we now report that the 31-kDa C-terminal domain contributes catalytic activity. Our results show that this domain as a purified proteolytic fragment conducts DNA synthesis under appropriate conditions but the kcat is lower and primer extension properties are different from those of the intact enzyme. A proteolytic truncation of the 31-kDa catalytic domain fragment, to remove a 60-residue segment from the NH2-terminal end, results in nearly complete loss of activity, suggesting the importance of this segment. Overall, these results indicate that the domains of beta-pol have distinct functional roles, template binding and nucleotidyltransferase, respectively; yet, the intact protein is more active for each function than the isolated individual domain fragment.     

Molecular characterization of a vacuolar processing enzyme related to a putative cysteine proteinase of Schistosoma mansoni.

Proproteins of various vacuolar proteins are post-translationally processed into mature forms by the action of a unique vacuolar processing enzyme. If such a processing enzyme is transported to vacuoles together with proprotein substrates, the enzyme must be a latent form. Immunocytochemical localization of a vacuolar processing enzyme, a 37-kD cysteine proteinase, in the endosperm of maturing castor bean seeds places the enzyme in the vacuolar matrix, where a variety of proproteins is also present. To characterize a molecular structure of vacuolar processing enzyme, we isolated a cDNA for the enzyme. Deduced primary structure of a 55-kD precursor is 33% identical to a putative cysteine proteinase of the human parasite Schistosoma mansoni. The precursor is composed of a signal peptide, a 37-kD active processing enzyme domain, and a propeptide fragment. Although the precursor expressed in Escherichia coli has no vacuolar processing activity, a 36-kD immunopositive protein expressed in E. coli is active. These results suggest that the activation of the vacuolar processing enzyme requires proteolytic cleavage of a 14-kD C-terminal propeptide fragment of the precursor.     

Folding and unfolding of the protoxin from Bacillus thuringiensis: evidence that the toxic moiety is present in an active conformation

The action of trypsin or papain on the 130-kDa crystal protein (protoxin) from Bacillus thuringiensis subsp. kurstaki HD-73 yields a 67-kDa proteinase-resistant toxic fragment (toxin) which is derived from the N-terminal half of the molecule. Sensitivity to proteolysis and fluorescence emission spectroscopy showed that the toxin unfolded to a much greater extent in 6 M guanidinium chloride (GuHCl) than in 8 M urea. Protoxin also unfolded extensively in 6 M GuHCl, whereas in 8 M urea only the C-terminal half of the molecule had unfolded extensively. Both unfolded protoxin and unfolded toxin refolded to their native and biologically active conformations. The biphasic unfolding observed for protoxin suggests that the C-terminal half of the molecule unfolded rapidly, whereas the N-terminal toxic moiety unfolded at a much slower rate, similar to that of the free 67-kDa toxin. A 67-kDa fragment, derived from the N-terminal half of the molecule, could be generated from the protoxin in the presence of either urea or GuHCl by treatment with proteinases. Compared to toxin in denaturants, this fragment was found to be more sensitive to proteolysis. However, on removal of the denaturants the fragment had the same proteinase resistance and cytolytic activity as native toxin. The increased proteinase sensitivity of the fragment generated in the presence of denaturants appears to be due to a perturbation in the conformation of the N-terminal toxic moiety. This perturbation is attributed to the unfolding of the C-terminal region of the protoxin prior to its proteolysis to yield the 67-kDa fragment.(ABSTRACT TRUNCATED AT 250 WORDS)     

Precursor of mitochondrial aspartate aminotransferase synthesized in Escherichia coli is complexed with heat-shock protein DnaK.

On expression of the cDNA encoding the precursor of chicken mitochondrial aspartate aminotransferase (pmAspAT) in Escherichia coli, the bulk of pmAspAT was found to be associated with the 70-kDa heat-shock protein DnaK which is closely related to mitochondrial 70-kDa heat-shock protein (HSP70). Purification protocols for the DnaK/pmAspAT complex and its individual components were elaborated. The complex dissociated on treatment with MgATP or at pH 5.5. Like the mature enzyme, pmAspAT is a dimer (2 x 47 kDa) and exhibits about a third of its enzyme activity. In the DnaK/pmAspAT complex, one DnaK molecule is bound to each subunit of pmAspAT; this tetramer may further aggregate to an octamer. The complex is catalytically almost as active as free pmAspAT. It could be reconstituted from isolated DnaK and pmAspAT. No complex was formed with mAspAT. Apparently, DnaK binds to the solvent-exposed presequence of folded pmAspAT without significantly changing the structure and functional properties of its mature moiety.     

Distinct roles of the N-terminal and C-terminal precursor domains in the biogenesis of the Bordetella pertussis filamentous hemagglutinin.

The 220-kDa Bordetella pertussis filamentous hemagglutinin (FHA) is the major exported protein found in culture supernatants. The structural gene of FHA has a coding potential for a 367-kDa protein, and the mature form constitutes the N-terminal 60% of the 367-kDa precursor. The C-terminal domain of the precursor was found to be important for the high-level secretion of full-length FHA but not of truncated analogs (80 kDa or less). The secretion of full-length and truncated FHA polypeptides requires the presence of the approximately 100-amino-acid N-terminal domain and the outer membrane protein FhaC, homologous to the N-terminal domains of the Serratia marcescens and Proteus mirabilis hemolysins and their accessory proteins, respectively. By analogy to these hemolysins, it is likely that the N-terminal domain of the FHA precursor interacts, directly or indirectly, with the accessory protein during FHA biogenesis. However, immunogenicity and antigenicity studies suggest that the N-terminal domain of FHA is masked by its C-terminal domain and therefore should not be available for its interactions with FhaC. These observations suggest a model in which the C-terminal domain of the FHA precursor may play a role as an intramolecular chaperone to prevent premature folding of the protein. Both heparin binding and hemagglutination are expressed by the N-terminal half of FHA, indicating that this domain contains important functional regions of the molecule.     

Two-site autoinhibition of the ADP-ribosylating mosquitocidal toxin (MTX) from Bacillus sphaericus by its 70-kDa ricin-like binding domain

The mosquitocidal toxin (MTX) from Bacillus sphaericus SSII-1 is an approximately 97-kDa arginine-specific ADP-ribosyltransferase that is activated by proteolytic cleavage, thereby releasing the active 27-kDa enzyme (MTX(30-264)) and a 70-kDa C-terminal fragment (MTX(265-870)). In solution, the cleaved 70-kDa fragment is still a potent inhibitor of the ADP-ribosyltransferase activity of MTX. Here we studied the interaction of the 70-kDa fragment with the enzyme domain of MTX. Several C-terminal deletions of the 70-kDa fragment inhibited the enzymatic activity of MTX(30-264). However, the IC(50) values were about 2 orders of magnitude higher for the deletions than for the 70-kDa fragment. A peptide covering amino acid residues 265-285 of the holotoxin exhibited the same inhibitory potency as the C-terminal deletions of the 70-kDa fragment. MTX(265-285) contains several acidic residues, of which D273 and D275 were found to be essential for the inhibitory effect. Exchange of these residues in the 70-kDa fragment (MTX(265-870)) reduced its inhibitory potency. Kinetic analysis showed that the peptide MTX(265-285) had no effect on the V(max) of MTX(30-264) but increased the K(m) for NAD. By contrast, the 70-kDa fragment deleted of residues Ile265 through Asn285 inhibited the enzyme activity of MTX(30-264) mainly by decreasing the V(max) of the enzyme. A second binding site for interaction of MTX(265-870) with MTX(30-264) was localized to the C-terminus within the region of residues 750-870. The data support a two-site binding model for inhibition of the ADP-ribosyltransferase activity of MTX(30-264) by the 70-kDa fragment MTX(265-870) with an interaction of amino acid residues 265-285 at the active site and an allosteric inhibition by the C-terminal part of the 70-kDa fragment.     

Identification of an autoinhibitory domain in the C-terminal region of the plant plasma membrane H(+)-ATPase.

Proteolytic (trypsin) treatment removes a small terminal segment from the 100-kDa plant plasma membrane H(+)-ATPase. This results in activation of H+ pumping across the plasma membrane, suggesting that an inhibitory domain is located in one of the terminal regions of the enzyme (Palmgren, M.G., Larsson, C., and Sommarin, M. (1990) J. Biol. Chem. 265, 13423-13426). In order to identify the origin of the fragment released by trypsin, polyclonal antibodies were raised against the first 55 amino acids (N-terminal region), the last 99 amino acids (C-terminal region), and a portion of 150 amino acids in the central part of the enzyme as deduced from one of the H(+)-ATPase genes (PMA2) of Arabidopsis thaliana. The native, 100-kDa H(+)-ATPase was recognized by all three antisera in Western blots. By contrast, the approximately 90-kDa polypeptide appearing after trypsin treatment was only recognized by the antisera against the N-terminal and central region, but not by the antiserum against the C-terminal region, suggesting that the inhibitory domain is located in this part of the enzyme. To more closely determine the position of the inhibitory domain, three peptides representing conserved parts of the C-terminal region were synthesized (residues 861-888, 912-943, and 936-949 of the Arabidopsis (PMA2) sequence). Only one of the peptides (residues 861-888) affected H+ pumping by the trypsin-activated (approximately 90-kDa) enzyme. This peptide of 28 amino acids inhibited H+ pumping with an IC50 of about 15 microM, suggesting that the auto-inhibitory domain is located within the corresponding part of the C-terminal region.