Journal list menu
An investigation of the activity of recombinant rat skeletal muscle cytosolic sialidase
Abstract
Rat cytosolic sialidase is expressed at elevated levels in skeletal muscle and is believed to play a role in the myogenic differentiation of muscle cells. Here, we observed varying levels of enhancement of sialidase activity in the presence a range of divalent cations. In particular, a significant enhancement of activity was observed in the presence of Ca2+. Conversely, inhibition of the sialidase activity was found when the enzyme was incubated in the presence of Cu2+, EDTA, and a range of carbohydrate-based inhibitors. Finally, an investigation of the enzymatic hydrolysis of a synthetic substrate, 4-methylumbelliferyl N-acetyl-α-d-neuraminide, by 1H NMR spectroscopy revealed that the reaction catalysed by rat skeletal muscle cytosolic sialidase proceeds with overall retention of anomeric configuration. This result further supports the notion that all sialidases appear to be retaining enzymes.
1 Introduction
In addition to being implicated in lysosomal catabolism, mammalian sialidases are also thought to be responsible for the modulation of functional biomolecules involved in many biological processes [1]. Miyagi et al. [1-3] have identified four types of mammalian sialidases that differ not only in their subcellular location, but also their catalytic and immunological properties. These sialidases are found within the lysosomes, the lysosomal membrane, the cytosol, and the plasma membrane. The distinct locations and differing catalytic properties of these sialidases suggest that each may perform a unique role. In general, however, little is understood regarding the function of these enzymes in part due to their apparent instability and low activity, which makes their purification and characterisation extremely challenging.
Miyagi et al. [4, 5] have successfully cloned, expressed and purified to homogeneity the cytosolic sialidase from rat skeletal muscle. Moreover, it was shown that the cytosolic sialidase had a higher activity and expression in the skeletal muscle cytosol, compared to other subcellular sialidases in other tissues. The cytosolic sialidase has been shown to have activity towards α2,3-sialylglycoconjugates, including sialyloligosaccharides, sialylglycopeptides and gangliosides [5-7]. In the context of cytosolic resident asparagine-linked glycopeptides it is of value to note that a cytosolic peptide:N-glycanase (PNGase, EC 3.5.1.52) has been identified and characterised [8-11]. This enzyme appears to play a key role in deglycosylation of misfolded glycoproteins that have been exported out of the ER to the cytosol [8-11]. The fact that free N-linked glycans and PNGase-released glycans are resident in the cytosol may also provide a rich source of substrates for the cytosolic sialidase and may suggest that the enzyme plays a role in the turnover of such glycans.
The explicit reasons for the high expression and the physiological function of the cytosolic sialidase in skeletal muscle are still not entirely clear. Miyagi and coworker [12] and more recently Fanzani et al. [13] have suggested that this enzyme may play an important role in myoblast differentiation by desialylation of glycoconjugates involved in the process.
Evidence has emerged to implicate cytosolic sialidase in muscle metabolism [14-23]. These findings suggest that the agrin-signalling pathway can be activated by sialidase or calcium treatment. Of particular interest is the observation that applied N-acetylneuraminic acid not only decreases acetyl choline receptor (AchR) clustering but also diminishes the tyrosine phosphorylation of muscle-specific kinase (MuSK) and the β-subunit of AChR [23]. It therefore appears that N-acetylneuraminic acid may play a regulatory role in the signal transduction events of neuromuscular synapse formation.
Clearly, further characterisation of this enzyme will help to elucidate the biological function and putative role of this sialidase in muscle differentiation. Herein, we report the use of a N-acetylneuraminic acid-based affinity chromatography matrix to generate partially purified cytosolic skeletal muscle sialidase with high sialidase activity. Furthermore, as a result of the apparent poor stability of the enzymatic activity that had previously been a significant problem plaguing studies of this enzyme [1-3], we thought it of value to undertake a stabilisation study of the sialidase activity using additives such as calcium. We have also investigated the susceptibility of the enzyme to inhibition by a number of N-acetylneuraminic acid-based compounds and sialylmimetics (Fig. 1 ). All of the carbohydrates evaluated are either known inhibitors of other sialidases or are potential sialidase inhibitors. Finally, we [24-27] and others [28, 29] have explored the catalytic reaction of sialidases from various sources by 1H NMR spectroscopy. In light of these previous studies, we have taken the opportunity to further characterise the sialidase activity from cytosolic skeletal muscle by using 1H NMR spectroscopy to monitor the hydrolysis of a synthetic sialoside.
2 Materials and methods
2.1 Expression and purification of cytosolic sialidase
The cDNA clone (pGEX-2T-SD) of rat muscle cytosolic sialidase was kindly provided by Dr. Taeko Miyagi (Miyagi Prefectural Cancer Centre, 47-1 Nodayama, Japan) and was expressed as previously described [4]. An affinity matrix derived from the neuraminic acid analogue 2-S-(aminoethyl) 5-acetamido-2-thio-d-glycero-α-d-galacto-2-nonulopyranosidonic acid was prepared using a previously described method [31, 32]. The sialidase purification by affinity chromatography was performed at 4 °C according to published methods [31, 32].
2.2 Sialidase activity assay
Sialidase activity was measured by the modified fluorometric method [33] using the fluorescent substrate 4-methylumbelliferyl N-acetyl-α-d-neuraminide (MUN). The hydrolysis reaction was performed in 20 mM NaOAc buffer (pH 6.1) containing 4 mM CaCl2. One unit of sialidase activity is defined, as the amount of enzyme needed to catalyse the hydrolysis of 1 μmol of MUN/hour under the conditions of the fluorometric assay.
2.3 Sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) and immunoblot analysis
SDS separating gels (12%) were run according to the published method of Laemmli [34] using silver staining for protein visualisation. For immunoblot analysis protein samples fractionated by SDS–PAGE were transferred to a PVDF membrane and were subjected to immunodetection. The primary antibody was polyclonal rabbit anti-rat skeletal muscle sialidase antibody [2] (kindly provided by T. Miyagi). The secondary antibody used was sheep polyclonal anti-rabbit IgG conjugated with horseradish peroxidase (Silenus). Bands were visualised using an enhanced chemiluminescence reagent (Boehringer–Mannheim), according to the manufacturer's instructions.
2.4 Kinetic studies
The influence of various additives such as metal ions, BSA and EDTA on sialidase activity of expressed and partially purified enzyme samples was investigated. Additionally, a number of carbohydrate-based compounds (Fig. 1) synthesised in our laboratory [27, 35-38] were tested as potential inhibitors of the sialidase activity. Thus, the additives were incubated at various concentrations (see Table 1 ) with the sialidase and the activity measured using the modified fluorometric assay [33] described above.
| Additive a | Concentration (mM) | Residual activity (%) |
|---|---|---|
| CaCl2 | 7 | 1018 |
| CuCl2 | 7 | 50 |
| LiCl | 7 | 142 |
| MnCl2 | 7 | 784 |
| MgCl2 | 7 | 382 |
| KCl | 7 | 109 |
| EDTA | 1 | 48 |
| BSA | 0.3% | 158 |
| (1) | 0.2 | 82 |
| (2) | 0.2 | 102 |
| (3) | 0.2 | 103 |
| (4) | 0.2 | 4 |
| (5) | 0.2 | 38 |
| (6) | 0.2 | 98 |
| (7) | 0.2 | 95 |
| (8) | 0.2 | 96 |
- a Compound numbers refer to Fig. 1.
For the inhibition studies of the carbohydrate-based compounds the assay solution contained 0.2 mM MUN, 0.2 mM inhibitor, and 8 × 10−3 units of partially purified enzyme in 20 mM NaOAc buffer (pH 6.1) containing 4 mM CaCl2.
2.5 MUN hydrolysis reaction
1H NMR spectroscopy was used to follow the hydrolysis of MUN catalysed by sialidase. MUN (0.55 mg, 1.18 μmol) was incubated with 24 U of partially purified recombinant rat cytosolic sialidase in 600 μL of NaOAc-d 4/DCl (pH 6.1, uncorrected) containing 4 mM CaCl2 for 102 min. 1H NMR spectra (600 MHz, 298 K) were acquired periodically during this time period to monitor the progress of the hydrolysis reaction. A control spectrum (time = 0) of a sample containing MUN (1.18 μmol) in 600 μL of NaOAc-d 4/DCl (pH 6.1, uncorrected) was also acquired.
3 Results
3.1 Affinity chromatography
Partial purification, as determined by SDS–PAGE, of the recombinant rat cytosolic sialidase was achieved by using 10 mL of the affinity matrix based on the N-acetylneuraminic acid analogue 2-S-(aminoethyl) 5-acetamido-2-thio-d-glycero-α-d-galacto-2-nonulopyranosidonic acid [26, 27]. An immunoblot analysis (Fig. 2 ) of the partially purified recombinant rat skeletal muscle cytosolic sialidase showed only two immunoreactive bands with approximate molecular weights of 43 and 76 kDa in fractions containing sialidase activity. A molecular weight of 43 kDa is consistent with the size of the cytosolic sialidase isolated from rat skeletal muscle by Miyagi et al. [2]. The 76 kDa band has been previously ascribed to a contaminant arising from the original protein used to raise antibodies [2]. As can be seen from the immunoblot analysis (Fig. 2) a sialidase active band at 43 kDa was obtained after a final elution with 0.5 M NaCl (lane 6) (see Fig. 2).
3.2 Kinetic studies of rat cytosolic sialidase
Table 1 summarises the effect of additives on the activity of the partially purified recombinant rat cytosolic sialidase. It is evident from Table 1 that CuCl2 and EDTA markedly reduced the sialidase activity while KCl had little effect on the activity. Increases in sialidase activity were noted on addition of 7 mM CaCl2, MnCl2, MgCl2, LiCl or 0.3% BSA. An explanation for the observed increase in sialidase activity as a result of BSA addition is not clear and could be either non-specific effects of higher carrier protein concentration or some specific BSA effect. Strikingly, addition of CaCl2 lead to over a 1000% increase in sialidase activity for partially purified enzyme and 555% for the crude enzyme extracts. K m values for the binding of Ca2+ were calculated using the direct linear plot method [39] and were found to be 0.2 mM for the enzyme and 1.1 mM for the crude enzyme extracts. The response to calcium addition displayed saturation kinetics.
A naturally occurring known μM inhibitor of a number of sialidases [36] Neu5Ac2en (4, Fig. 1), inhibited rat cytosolic sialidase activity by over 95% at a concentration of 0.2 mM. These results are in good agreement with those of Miyagi et al. [3]. Interestingly, at the same concentration, iso-propyl 2-acetamido-2-deoxy-Δ4-β-d-glucopyranosiduronic acid (5), a Neu5Ac2en mimetic in which the glycerol sidechain of Neu5Ac2en is replaced by an aliphatic ether [37], resulted in ca. 60% inhibition of the sialidase activity. None of the other carbohydrate compounds listed in Table 1 or shown in Fig. 1 produced significant inhibition of the recombinant sialidase activity.
3.3 Rat cytosolic sialidase-catalysed hydrolysis of MUN
The recombinant rat cytosolic skeletal muscle sialidase-catalysed hydrolysis of MUN, an N-acetylneuraminic acid-based synthetic substrate, was monitored using 1H NMR spectroscopy at 25 °C as a time course reaction (Fig. 3 ). To the deuterated NaOAc buffer (pH 6.1, uncorrected) incubation medium containing MUN and 24 U of expressed partially purified rat cytosolic skeletal muscle sialidase was added 4 mM CaCl2. The hydrolysis of MUN catalysed by the sialidase was monitored by acquiring spectra periodically and examining the signals of the H3eq and H3ax protons of the sialic acid portion of MUN that resonate at 2.9 and 1.65 ppm, respectively, in the 1H NMR spectrum (Fig. 3). As the hydrolysis reaction proceeds, the intensities of the resonances at 2.9 and 1.65 ppm decrease with a concomitant increase in the intensities of signals at 2.75 and 1.65 ppm corresponding to the H3eq and H3ax protons of free α-Neu5Ac. The appearance of these resonances, due to α-Neu5Ac as the first product of release in the hydrolysis reaction (T = 2 min), indicate that rat skeletal muscle cytosolic sialidase is a retaining sialidase, i.e., the sialic acid portion of MUN binds in the active site of rat cytosolic sialidase in the α-form and is released after hydrolysis of the aglycon unit as α-Neu5Ac. The appearance of other signals in the spectrum at 2.24 and 1.83 ppm, with the concomitant decrease in the intensities of peaks due to α-Neu5Ac is due to the mutarotation of α-Neu5Ac to the thermodynamically preferred β-Neu5Ac in solution. From the spectra it is possible to determine that the hydrolysis reaction is complete at T = 17 min and that equilibrium between the α- and β-anomers of Neu5Ac has been reached within ca. 102 min.
4 Discussion
The emergence of evidence that suggests the involvement of cytosolic sialidase in myoblast differentiation [12, 13], together with the observed influence of sialidase, sialic acids and calcium, in the tyrosine phosphorylation of muscle-specific kinase (MuSK) and the aggregation of acetylcholine receptors [23] is an exciting development in this field. Clearly, any data that can consolidate, or further develop an understanding of this body of information is highly desirable.
The influence of metal ions on enzyme activity is of great importance and may have implications on the enzyme's function. In the present study, the effect of calcium on recombinant rat skeletal muscle cytosolic sialidase activity was found to be dramatic with a 1000% increase in the observed sialidase activity. These results clearly demonstrate an increased influence of added calcium on the partially purified enzyme compared to the crude extracts. This may suggest that there are other components in the crude extracts such as chelating agents or endogenous inhibitors that either modulate the effect of the added calcium or inhibit the enzyme. Presumably, these components are removed during the purification procedure. It is interesting to note that the further addition of calcium into soluble partially purified enzyme not only restores enzyme activity, it also appears to increase activity. A possible explanation for these observations is that calcium plays a crucial role in either structural maintenance of the enzyme's architecture or in catalysis itself. This is not unlike the important role of calcium in stabilisation of the binding region of Vibrio cholerae sialidase [40].
Our investigation of other divalent metal ions led us to conclude that the sialidase is also stimulated by the presence of divalent magnesium and manganese ions. It is interesting to note that purified rat liver cytosolic sialidase activity is apparently inhibited by the divalent ions of Ca2+, Mn2+ and Mg2+ [5]. Cu2+ was found to have an inhibitory effect (50% inhibition at 7 mM) on the sialidase activity. This is in good agreement with a similar observation by Miyagi et al. [3].
The recombinant sialidase activity was inhibited by Neu5Ac2en 4 (95%), a known sialidase inhibitor, and by the previously reported [37] mimetic of Neu5Ac2en, iso-propyl 2-acetamido-2-deoxy-Δ4-β-d-glucopyranosiduronic acid 5 (62%). These compounds are designed to structurally resemble the putative transition state, a sialosyl cation, of the sialidase-catalysed reaction. We have previously found that 5 inhibits other microbial sialidases such as influenza virus sialidase [37]. It is clear from this data that the cytosolic sialidase can accommodate hydrophobic groups in the C-6 glycerol binding domain. We have observed a similar hydrophobic group accommodation in microbial sialidases such as V. cholerae sialidase [37, 41, 42] and influenza virus sialidase [37]. Modest inhibition (18%) was also observed with the α2,3-linked substrate-like compound Neu5Ac-2-S-(α2,3)-Galβ1Me 1, a thiosialoside metabolically stable to the action of sialidases [27]. Apart from Neu5Ac-2-S-(α2,3)-Galβ1Me 1, all other substrate-like compounds tested that is, 6, 7, and 8, did not inhibit the sialidase. This outcome combined with the fact that 6–8 are α2,6-sialylmimetics further supports the notion that the cytosolic sialidase has an α2,3-linkage preference [5-7].
Investigation of the recombinant sialidase-promoted hydrolysis reaction revealed that it was a retaining sialidase and that hydrolysis of the synthetic substrate proceeds with retention of anomeric configuration. The general findings of this study and others [24-30] suggest that irrespective of the source of the enzyme all sialidases appear to be retaining enzymes.
The present study provides some further useful characterisation of the recombinant enzyme activity.
Acknowledgements
We gratefully acknowledge the financial support of the Australian Research Council and the Monash University Research Fund. MvI acknowledges the award of an Australian Federation Fellowship. We thank Dr. T. Miyagi for providing the rat muscle cytosolic sialidase cDNA clone and anti-rat skeletal muscle sialidase antibody.
