Neuraminidase assay in cultured human fibroblasts: in situ versus in vitro procedures

ELSEVIER

Clinica Chimica Acta 251 (1996) 163-171

Neuraminidase assay in cultured human fibroblasts: in situ versus in vitro procedures M. Beck *a, E. Scheuring a, H.-U. Voelter b, J. Brandt a, K. Harzer b aKinderklinik, Universitiit Mainz, Langenbeckstrasse 1, 55101 Mainz, Germany bInstitut J~r Hirnforschung, Universitiit Tiibingen, Schwiirzlocherstrasse 79, 72070 Tiibingen, German), Received 20 June 1995; revised 31 January 1996; accepted 2 February 1996

Abstract Further investigations have been carried out to characterize a published procedure of neuraminidase assay, in which the activity is measured directly on the cell culture layer. The pH optimum was 4.0. A Vma~ value of 130 nmol/mg/h and a K m of 0.3 mmol/l were found. During incubation in the acid buffer, arylsulphatase activity was released into the medium, whereas neuraminidase activity remained attached to the cells. The in situ method allowed an unequivocal diagnosis of primary and secondary neuraminidase deficiencies. Because of its simplicity and reliability, the method appears useful as a routine method in clinical laboratories.

Keywords: Neuraminidase; Sialidosis; Mucolipidosis; l-cell disease; Galactosialidosis; In situ method

1. Introduction Lysosomal neuraminidase (sialidase, EC 3.2.1.18) plays an essential role in the hydrolysis of sialic acid containing oligosaccharides, glycoproteins and glycolipids. Deficiency in neuraminidase is known in two groups of lysosom~Ll storage disorders: sialidosis (isolated neuraminidase defect) and galactosialidosis (combined neuraminidase/fl-galactosidase defect due to a *Corresponding author, Fax: + 49 6131222332. 0098-8981/96/$15.00 ~ 1996 Elsevier Science B.V. All rights reserved SSD1 S0009-8981(96)06304-8

164

M. Beck et al. / Clinica Chimica Acta 251 (1996) 163-171

malfunction of the protective protein). In both groups a broad clinical heterogeneity is observed, the courses varying from severe hydrops fetalis to adult forms where myoclonus epilepsy (together with macular cherryred spot) is a leading symptom. In mucolipidosis II (I-cell disease) and mucolipidosis III, lysosomal neuraminidase deficiency is due to a lack of N-acetylglucosamine- 1-phosphotransferase (E.C.2.7.8.17). In the absence of the transferase activity the common phosphomannosyl recognition marker of acid hydrolases is not generated, and the enzymes are not targeted to the lysosomes [1,2]. Lysosomal neuraminidase is a membrane-bound enzyme that is difficult to characterize and to purify because of its marked lability [3]. It is associated with fl-galactosidase to form a high molecular weight complex that is stabilized by a 32 kDa protective protein. Several methods have been developed to assay neuraminidase activity in human tissue. The enzyme activity can be monitored by measuring sialic acid released from fetal calf serum fetuin [4]. A chromogenic substrate (methoxyphenyl-N-acetyl-neuraminic acid) was synthesized by Tuppy and Palese [5]. Omichi and Ikenaka [6] used high-performance liquid chromatography to separate aglycones from unhydrolyzed substrates. A fluorogenic substrate described by Warner and O'Brien [7] and Potier et al. [8] is usually applied in clinical laboratories. For activity measurement of the labile enzyme some precautions must be regarded. The enzyme is susceptible to freezing or sonication [9]. Furthermore, fl-galactosidase deficiency destabilizes the fl-galactosidase/ neuraminidase/protective protein complex and may lead to a reduction of neuraminidase activity in GMl-gangliosidosis. These difficulties can be overcome by an in situ method that has been described by Paton et al. [10] and was used by Ledvinova et al. [11]. However, since the enzyme properties were not described in these papers, we have analyzed the assay procedure in more detail. The pH optimum and kinetic properties have been compared to the values described in the literature. The suitability of the method for clinical purposes has been demonstrated in cells of patients with different forms of primary and secondary neuraminidase deficiency.

2. Patients and methods 2.1. Case reports

Control strains were derived from skin biopsies from patients (28 children and 12 adults) who suffered from diseases unrelated to lysosomal enzyme deficiency. Fibroblasts of the patients were taken from our own stock. Patient 1 was a child who at birth showed oedema, ascites and

M. Beck et al. I Clinica Chimica Acta 251 (1996) 163-171

165

hepatosplenomegaly: an extensive clinical description was given by Beck et al. [12]. Patient 2 (a 5 year-old boy) presented with dysmorphic signs, organomegaly and severe mental retardation. Patients 3 and 4 were sisters (24 and 25 years old, respectively), in whom severe myocloni, cerebellar ataxia and macular cherry-red spots were the predominant clinical symptoms. Patient 5 (an 18 month-old girl) could be classified with the early infantile form of galactosialidosis. Patient 6 and patient 7 presented with disproportionate dwarfism, coarse facial features and mental retardation. Detailed clinical data were published on both patients [13]. Patient 8 was a 21 year's old, mentally normal woman who had short stature, joint stiffness and corneal clouding. Patients 9, 10 and 11 showed clinical features of infantile GMx-gangliosidosis with neurological, skeletal and cardiac involvement, they died before the age of 2 years. 2.2. Fibroblasts Skin biopsies were obtained and fibroblasts were cultured using standard methods [14]. The cells were maintained in Eagle's minimum essential medium supplemented with L-glutamine, fetal calf serum and antibiotic's. Proteins were determined by the method of Lowry et al. [15] with bovine serum albumin as standard. 2.3. Neuraminidase assay Neuraminidase activity in fibroblast homogenate (in vitro method) was measured with 4-methylumbelliferyl-0~-D-N-acetylneuraminic acid as described by Kleijer et al. [16]. The in situ assay of neuraminidase was performed as described by Paton et al. [10]: the substrate (0.12 mmol/1 4-methylumbelliferyl-~-D-N-acetylneuraminic acid, dissolved in 1.6 ml of 0.05 mol/1 sodium acetate buffer, pH 4.0) was applied to 75 cm / flasks of cultured fibroblasts (0.5-0.8 mg protein) that had been washed twice with iso-osmotic NaC1 and once, for some seconds, with water. After incubation at 37°C for 30 or 60 min, 200 #1 of the reaction mixture was withdrawn. The 4-methylumbelliferone released from the substrate was determined after adjusting the withdrawn samples to pH 10.0 with 500 #1 of 85 mmol/1 glycine-carbonate buffer. For blank assays culture flasks without any cells were used. Homogenates of the cell layer that remained attached to the flask at the end of the assay were used for protein measurement. To assess diagnostic reliability several determinations were performed in separate subcultures from the same biopsy; the reproducibility of values was better than 5%. The Michaelis-Menten constant (Km) was determined by varying the substrate concentration between 0.12 and 0.019 mmol/1 and constructing a Lineweaver-Burk plot.

166

114. Beck et al. I Clinica Chimica Acta 251 (1996) 163-171

To examine whether neuraminidase activity was released into the medium during incubation, cultured fibroblasts were incubated for 20, 40 and 60 min with acetate buffer (pH 4.0) without substrate. Thereafter, 200 #1 of medium was withdrawn and incubated for 60 min with 50/~1 of the substrate solution described above.

2.4. Arylsulphatase A assay Arylsulphatase A activity was assayed in the cells and in the reaction medium with p-nitro-catechol sulphate 1-17]. The incubation mixture contained 10 mmol/1 p-nitro-catechol sulphate, 0.5 mol/1 sodium acetate buffer (pH 5.0), 0.5 mmol/1 NaaP20 7 and 1.7 mol/1 NaC1. Then, 50 #1 of cell homogenate was added to 100 #1 of this mixture. After incubation for 30 min at 37°C the reaction was stopped by adding 300 /tl of 1 mol/1 NaOH. The absorbance was measured against a zero-time control at 515 nm. Determination of arylsulphatase A activity released from the cells was performed by adding 50 #1 of medium instead of cell homogenate to the reaction mixture. 3. Results

During incubation of the fibroblasts with acetate buffer for 60 min, no neuraminidase activity was released into the medium indicating that the enzyme remained bound to the cell or lysosome membrane. However, a time-dependent increase of arylsulphatase A activity in the medium was observed; after the 60 min reaction period about 60-70% of the total arylsulphatase activity was released from the cells. When 0.5-1.0 mg of cell protein was used for the neuraminidase assays, hydrolysis of 4-methylumbelliferyl-~-o-N-acetylneuraminic acid increased proportionally. The time curve was linear for the first 40 min; thereafter a non-linear increase in activity was observed (Fig. 1). As shown in Fig. 2, the pH optimum of the enzyme reaction against 4-methyl-umbelliferyl-eD-N-acetylneuraminic acid was 4.0. Enzyme-substrate saturation kinetics were observed with an apparent g m of 0.3 mmol/1 (Fig. 3). The Vmax value was 130 nmol/mg protein per h. Routine assays were carried out at 0.12 mmol/1 substrate concentration (see Methods, Sections 2.2-2.4) to avoid high blank effects. Normal range was 26-71 nmol/mg protein per h (n = 40). To evaluate the use of the described method for routine diagnoses, fibroblasts of patients with several lysosomal storage disorders were analyzed. Neuraminidase activity was measured both with the in situ procedure and in vitro in cell homogenates. As shown in Table 1, with the in situ assay a distinct diagnosis of primary and secondary neuraminidase defciencies could be made, whereas normal

167

M. Beck et al. I Clinica Chimica Acta 251 (1996) 163-171

30 A

25

.=_ tl)

2 ca. 20 E -~ O E ---.

15 10

~.

5 0 0

I

I

I

I

I

!

10

20

30

40

50

60

Time (rain) Fig. 1. Enzymatic hydrolysis of 4-methylumbelliferyl-~t-D-N-acetylneuraminicacid in cultured fibroblasts as a function of time.

50 t-

t-

40

ca.

30

{:7)

E O

E t,-

20

0

I

I

!

I

I

|

I

I

3,1 3,4 3,7 4,0 4,3 4,6 4,9 5,2 5,5 5,8 pH Fig. 2. Effect of pH on rate of hydrolysis of 4-methylumbelliferyl-ct-D-N-acetylneuraminic acid (range and mean value, n = 3).

M. Beck et al. / Clinica Chimica Acta 251 (1996) 163-171

169

activity was detected in fibroblasts of patients with GMl-gangliosidosis. However, when the in vitro procedure was used, in /~-galactosidase deficient cells a reduced neuraminidase activity was found.

4. Discussion The fluorogenic substrate 4-methylumbelliferyl-e-D-N-acetylneuraminic acid allows rapid and sensitive assay of neuraminidase activity in cultured fibroblas~ts. However, neuraminidase (sialidase) is a very labile membrane-bound enzyme; therefore, certain conventional methods for cell disruption, e.g. freezing, thawing, sonication or Triton X-100 treatment cannot be used before the enzyme assay. Moreover, as neuraminidase forms a complex with /~-galactosidase, a destabilized complex with reduced neuraminidase activity may be present in G M 1-gangliosidosis, also under optimal assay conditions. Hence, it seems convenient to measure neuraminidase activity directly on the cultured cell layer as described by Paton et al. [10]. In the present study, this in situ method was characterized in more detail. Conventional techniques for cell culturing could be used; in confluent cells, substrate hydrolysis increased in a linear manner with reaction time (Fig. 1). In our neuraminidase assay the pH optimum was at pH 4.0, whereas other authors found it in the range of 4.2-4.4, using the: same substrate [7,8]. Frisch and Neufeld [18] have synthesized tritium-labeled sialyllactitol for the measurement of neuraminidase. With this subs~Irate a pH optimum of 4.0 was determined. Den Tandt and Leroy [19] have compared the fluorogenic substrate and sialyllactose, and found the pH optimum at 4.0 for both. In the assay procedure described here the determined kinetic properties of neuraminidase were only partially similar to those reported by other authors. We have found a K m value of 0.3 mmol/1 and a Vmax value of 130 nmol/mg protein per h. In human fibroblasts Warner and O'Brien [7] have measured a K m value of 0.13 mmol/1 and a Vmax of 115 nmol/mg/h. A Km value of 0.13 mmol/1 was found by Potier et al. [8]. The Km value of 0.3 mmol/1 described by Myers et al. [20] was the same as our value. Cell lines of patients with primary and secondary neuraminidase deficiency were tested with the in situ method (Table 1). A clear distinction of normal and deficient values could be made, although one patient with the red-spot-myoclonus syndrome exhibited a rather high residual activity. Furthermore, in the mucolipidosis II (I-cell disease) cases a much lower activity was found than in the patient with mucolipidosis III. This result confirms the assumption of den Tandt and Leroy [19] that both diseases may have different residual enzyme activities. A clear distinction of combinecl Neur-//~-Gal- deficiency and GMl-gangliosidosis (/~-galac-

170

M. Beck et al. I Clinica Chimica Acta 251 (1996) 163-171

tosidase deficiency) could be made by the in situ procedure (Table 1). The in vitro method with cell homogenates in general showed more equivocal results (Table 1). It could be demonstrated by our study that the in situ neuraminidase assay is a simple procedure which is useful for clinical purposes. The difficulties of activity measurement that arise from the lability and membrane-bound character of the enzyme are avoided, since no cell disruption procedures are applied. The assay should be performed in a flask that contains confluent cells with a total protein of about 0.5-0.8 mg. The release of arylsulphatase A into the medium during the reaction period may indicate that cellular (including lysosomal) membranes are increased in their permeability under the assay conditions used in these experiments. From the observation that the neuraminidase activity cannot be recovered from the medium even after incubation with the acid buffer for 60 min, it may be concluded that the cellular membranes remain intact. However, the fluorogenic substrate as well as its hydrolysis products are able to diffuse between the site(s) of active neuraminidase and the reaction medium allowing reliable determination of enzyme activity.

Acknowledgements This study is part of the thesis of E. Scheuring

References [1] Hasilik A, Waheed A, von Figura K. Enzymatic Phosphorylation of lysosomal enzymes in the presence of UDP-N-acetylglucosamine. Absence of the activity in I-cell fibroblasts. Biochem Biophys Res Commun 1981;98:761-767. [2-1 Reitmann ML, Varki A, Kornfeld S. Fibroblasts from patients with I-cell disease and pseudo-hurler polydystrophy are deficient in uriddine 5'-diphosphate-N-acetylglucosamine: glycoprotein N-acetylglucosaminyl-phospotransferase activity. J Clin Invest 1981;67:1574-1579. I-3] Verheijen FW, Palmeri S, Hoogeveen AT, Galjaard H. Human placental neuraminidase. Activation, stabilization and association with fl-galactosidase and its "protective" protein. Eur J Biochem 1985;149:315-321. [4] Thomas GH, Tiller GE, Reynold LW, Miller CS, Bace JW. Increased levels of sialic acid associated with a sialidase deficiency in I-cell (mucolipidosis 1I) fibroblasts. Biochem Biophys Res Commun 1976;71:188-195. [51 Tuppy H, Palese P. A chromogenic substrate for the investigation of neuraminidases. FEBS Lett 1969;3:72-75. [6] Omichi K, Ikenaka T. Rapid and sensitive sialidase assay by high-performance liquid chromatography and its application to detection of sialidase in human urine. J Chromatogr 1982;230:415-419. I-7] Warner TG, O'Brien JS. Synthesis of 2'-(4-Methylumbelliferyl)-~-D-N-acetyl°

M. Beck et al. I Clinica Chimica Acta 251 (1996) 163-171

[8]

[9] [10]

[11]

[12]

[13]

[14]

[15] [16] [17] [18] [19] [20]

171

neuraminic acid and detection of skin fibroblast neuraminidase in normal humans and in sialidosis. Biochemistry 1979;18:2783-2787. Potier M, Mameli L, B61isle M, Dallaire L, Melanqon SB. Fluorometric assay of neuraminidase with a sodium (4-methylumbelliferyl-~-o-N-acetylneuraminate) substrate. Anal Biochem 1979;94:287-296. Thomas GH, Reynolds LW, Miller CS. Characterization of neuraminidase activity of cultured human fibroblasts. Biochim Biophys Acta 1979;568:39-48. Paton BC, Schmid B, Kustermann-Kuhn B, Poulos A, Harzer K. Additional biochemical findings in a patient and fetal sibling with a genetic defect in the sphingolipid activator protein (SAP) precursor, prosaposin. Evidence for a deficiency in SAP-1 and for a normal lysosomal neuraminidase. Biochem J 1992;285:481-488. Ledvinova J, Poupetova H, Elleder M, Tichy J, Penickova V, Harzer K. Sialidosis type 1: first report in the Czech population of two siblings with cherry-red spot myoclonus syndrome but without sialyloligosacchariduria. J Inherit Metab Dis 1994;17:118-119. Beck M, Bender SW, Reiter HL, Otto W, Bassler R, Dancygier H, Gehler J. Neur~tminidase deficiency presenting as non-immune hydrops fetalis. Eur J Pediatr 1984;1.43:135-139. Beck M, Barone R, Hoffmann R, Kratzer W, Rakowsky T, Nigro F, Fiumara A. Inter- and intrafamilial variability in mucolipidosis II (I-cell disease). Clin Genet 1995;47:191-199. Cantz M, Kresse H, Barton RW, Neufeld EF. Corrective factors for inborn errors of mucopolysaccharide metabolism. In: Ginsburg V, ed. Methods in Enzymology. New York/London: Academic Press, 1972;884-897. Lowry OW, Rosenbrough N J, Farr AL, Randall RJ. Protein measurement with the folin !phenol reagent. J Biol Chem 1951;193:265-275. Kleijcr WJ, Hoogeveen A, Verheijen FW et al. Prenatal diagnosis of sialidosis with combined neuraminidase and fl-galactosidase deficiency. Clin Genet 1979;16:60-61. Baum H, Dodgson K S, Spencer B. The assay of arylsulphatase A and B in human urine Clin Chim Acta 1959;4:453-455. Frisch A, Neufeld EF. A rapid and sensitive assay for neuraminidase: application to cultured fibroblasts. Anal Biochem 1979;95:222-227. den Tandt WR, Leroy JG. Deficiency of neuraminidase in the sialidoses and the mucolipidoses. Hum Genet 1980;53:383-388. Myers RW, Lee RT, Lee YC, Thomas GH, Reynolds LW, Uchida Y. The synthesis of 4-methylumbelliferyl ct-ketoside of N-acetylneuraminic acid and its use in a fluorometric assay for neuraminidase. Anal Biochem 1980;101:166-174.