OUP user menu

trans-Sialidase activity for sialic acid incorporation on Corynebacterium diphtheriae

A.L Mattos-Guaraldi , L.C.D Formiga , A.F.B Andrade
DOI: http://dx.doi.org/10.1111/j.1574-6968.1998.tb13269.x 167-172 First published online: 1 November 1998


A rapid and sensitive assay for neuraminidase using peanut lectin hemagglutination was used to study the prevalence of neuraminidase activity among sucrose-fermenting and non-sucrose-fermenting toxigenic Corynebacterium diphtheriae strains. Neuraminidase activity was found in 15 (100%) isolates regardless of biotype, hemagglutinating activity and site of isolation of bacteria. Besides expressing the neuraminidase activity that hydrolyzes sialic acid from glycoconjugates, C. diphtheriae was also capable of transferring sialic acid residues from a sialyl-lactose donor. A single molecule probably expresses both neuraminidase and trans-sialidase activity. The trans-sialidase activity was documented by observations of the interactions of bacterial cells with wheat germ agglutinin and peanut lectins. C. diphtheriae expressed a trans-sialidase activity located on the cell surface that produced asialoglycoconjugates from a sialyl donor substrate and at the same time generated bacterial sialyl derivatives of β-Gal acceptors.

Key words
  • Corynebacterium diphtheriae
  • Diphtheria
  • Neuraminidase
  • trans-Sialidase

1 Introduction

Bacterial neuraminidases are considered to be virulence factors in many pathogenic organisms that colonize mucosal surfaces, such as Corynebacterium diphtheriae[1], Vibrio cholerae, Streptococcus pneumoniae, group B streptococci [2], Bacteroides sp. and Porphyromonas sp. [3]. Many studies have demonstrated that microbial neuraminidases have the capacity to modify the ability of the host to respond to infection (for review see [2]). The finding of neuraminidase activity in crude diphtherial toxin preparations was reported by Blumberg and Warren [4]. The enzyme was characterized by Warren and Spearing [5] and Jamieson [6] suggesting a molecular mass of about 50 000 and 100 000, respectively. The authors found difficulty in separating toxin and enzyme activities. Later, Moriyama and Barksdale [1] compared various C. diphtheriae strains with respect to the properties of toxigenicity and neuraminidase production. The conditions of optimal production of toxin and enzyme and the distribution of enzyme activity between cell and the medium in culture were also investigated. In each of these respects toxigenicity and neuraminidase production behave as independent attributes of the diphtherial cell. The authors suggested that the molecular mass of the enzyme was similar to that proposed for diphtherial toxin, 64 500. The action of C. diphtheriae neuraminidase on the Sendai virus receptor of human erythrocytes was also investigated. The bacterial enzyme did not easily remove Sendai virus receptors, nor did it adsorb to the erythrocyte surface [7].

Pathogenic organisms may also express a trans-sialidase [8, 9]. The trans-sialidase reaction produces asialoglycoconjugates from sialyl donor substrates, and at the same time generates sialyl derivatives of acceptor substrates. In theory, the trans-sialidase could function either on the surface of microorganisms or as a soluble mediator. The trans-sialidase may behave as a microbial adhesive component, on one hand, directly participating in host-parasite interactions, and on the other hand indirectly contributing to the infectivity process by altering physiological parameters of the mammalian host [8]. The acquisition of sialic acid residues by the catalytic action of trans-sialidase may confer infective properties on some parasites [10].

The present study was undertaken to determine the trans-sialidase activity for sialic acid incorporation on C. diphtheriae. A rapid and sensitive peanut lectin (PNA) hemagglutination assay was also used to compare toxigenic C. diphtheriae strains of both sucrose-fermenting and non-sucrose-fermenting biotypes with respect to neuraminidase enzyme production.

2 Materials and methods

2.1 Bacterial strains and culture conditions

Neuraminidase activity of C. diphtheriae was evaluated using nine sucrose-positive and five sucrose-negative toxigenic strains previously isolated in Rio de Janeiro, Brazil and identified according to criteria previously described [11]. Eleven strains were isolated from throat, two from skin lesions and one from blood.

The trans-sialidase assays and the additional studies of the effect of storage temperature and time of reaction upon the neuraminidase activity were carried out with the toxigenic sucrose-negative strain C. diphtheriae var. mitis CDC-E8392 from Centers for Disease Control, Atlanta, GA, USA.

Microorganisms were stored at room temperature in CTA medium and in GC-glycerol at a temperature of −20°C [11]. Bacterial hemagglutinating properties using human erythrocytes (0.5%) were determined by methods previously described [12].

Microorganisms were grown in trypticase soy broth (TSB, Difco Laboratories, Detroit, MI, USA) medium supplemented with 5% fetal calf serum (v/v) at 37°C for 24/48 h. For determination of specific neuraminidase and/or trans-sialidase activity, bacterial cells were collected by centrifugation at 4°C and washed four times in cold 0.01 M Na-phosphate-buffered saline (PBS), pH 7.2. Bacterial cells were suspended in PBS to a density which gave a reading of 0.8 OD in λ 570 nm, and held on ice until assayed. Supernatant fluid was also tested for enzyme activity.

2.2 Assay of neuraminidase activity

Neuraminidase activity was qualitatively determined by an assay using PNA agglutination of human A erythrocytes as described elsewhere [13].

2.3 Neuraminidase treatment

Briefly, bacterial cells were suspended in 0.01 M PBS, pH 6.0, and were treated with 0.2 U ml−1Clostridium perfringens neuraminidase type X (Sigma, St. Louis, MO, USA). After 60 min at 37°C samples were taken, immediately cooled, and the enzyme-treated cells were washed three times with 0.01 M PBS, pH 7.2. After treatment, microorganisms were resuspended in 0.01 M PBS, pH 7.2 and subjected to assays with lectins [9].

2.4 Fluorescein isothiocyanate (FITC)-lectin binding studies

Bacterial cells untreated and/or treated with C. perfringens neuraminidase were analyzed by fluorescence assays employing wheat germ agglutinin (WGA) and PNA (Sigma) according to Nakamura et al. [14]. Bacterial suspension (20 μl) was placed on a glass slide, air dried, and fixed in methanol for 10 min at 23°C. Slides were washed in PBS, pH 7.2, for 10 min and then incubated with increasing dilutions (20 μl) of FITC-conjugated lectin (500 μg ml−1) in PBS for 30 min at 23°C. The slides were then washed three times in PBS for 5 min each time, mounted in PBS containing glycerol (50% v/v), and observed under the fluorescence microscope (Zeiss Universal Photomicroscope). The inhibition of FITC-lectin binding to bacterial cells was studied by incubating the FITC-lectin with its specific sugar inhibitor for 30 min before incubation with the bacterial cells (200 mM N-acetyl-d-glucosamine for WGA and 200 mM d-galactose for PNA).

2.5 Trans-Sialidase assay

The trans-sialidase reaction was performed as previously described [9]. Briefly, bacterial cells were treated with neuraminidase type X from C. perfringens (0.2 U ml−1). The enzyme-treated cells were washed three times with 0.01 M PBS, pH 7.2. After treatment microorganisms were suspended in 0.01 M PBS, pH 7.2, containing the substrate 2 mM N-acetylneuramin-lactose (Sigma) and in PBS containing 1 mM N-acetylneuraminic acid (Sigma), respectively. Bacterial suspensions were subjected to different incubation periods. The trans-sialidase reaction activity was confirmed by subjecting these microorganisms to the binding assay using FITC-conjugated lectins as described above.

3 Results and discussion

The assay for neuraminidase using PNA hemagglutination was applied previously to Vibrio cholerae and Trypanosoma cruzi[13]. The technique using PNA lectin (a subterminal d-Gal binding lectin) was rapid and sensitive for detection of neuraminidase activity in C. diphtheriae strains. Neuraminidase activity was observed in all 15 strains studied (100%) including the atypical sucrose-fermenting strains. The enzymatic activity was found to be independent of the sucrose fermentation biotype, the site of isolation and the hemagglutinating properties of C. diphtheriae strains (Table 1). Previous studies using the thiobarbituric acid method that assays for free neuraminic acid for detection of neuraminidase activity detected a neuraminidase-negative C. diphtheriae strain [1].

View this table:

Characteristics of C. diphtheriae strainsa presenting neuraminidase activity

StrainSite of isolationHemagglutination titer
222AArespiratory tract4
49AArespiratory tract8
65respiratory tract16
CDC-E8392respiratory tract32
03, 90AAskin lesion4
236respiratory tract0
31, 61, 233, 239, 241respiratory tract1
52respiratory tract2
13respiratory tract4
  • aAll strains were toxigenic, fluorescent and pyrazinamidase-negative [11].

  • AA Autoagglutinating activity.

Neuraminidase activity was expressed on the bacterial surface as well as a free soluble mediator. The results of these tests are shown in Fig. 1. Desialylation of human A erythrocytes was time-dependent. Complete desialylation occurred after 1 h of incubation. With whole cells, measurement of neuraminidase activity, when plotted against time, was not linear. Additional evidence for the neuraminidase activity of C. diphtheriae strains was provided by our results in the 2-deoxy-2,3-dehydro-N-acetylneuraminic acid assay [15]. This compound, a competitive inhibitor for bacterial neuraminidase, used at a concentration of 1 mM completely blocked neuraminidase activity on the human erythrocytes as determined by PNA agglutination assays (data not shown). Non-linearity was also seen with soluble enzyme when measurement of neuraminidase activity was plotted against time (Fig. 1).


Agglutination titers of human A erythrocytes in the presence of PNA (1 mg ml−1) after incubation with PBS (•), bacterial suspension (?), or culture supernatant fluid of C. diphtheriae (□).

Some microorganisms lose or have drastically reduced production of the enzyme upon extensive subculture, suggesting that the enzyme is involved in survival [2]. In our study, even microorganisms subcultured for up to 6 months still gave positive reactions in the PNA hemagglutination test. Neuraminidase activity detected on bacterial cells and in culture supernatant fluid was stable under different storage conditions. Enzyme remained active when stored for 48 h at 22°C, 4°C and −20°C.

The neuraminidase-positive culture supernatant fluid was not able to cause agglutination of human erythrocytes. Production of neuraminidase by C. diphtheriae did not lead to loss of bacterial hemagglutinating properties. We found neuraminidase-positive strains capable of agglutinating human erythrocytes with high titers (titer 32).

In order to observe if C. diphtheriae contained neuraminic acid residues on the cell surface, strain CDC-E8392 was subjected to WGA and PNA binding studies (Table 2). WGA was found to bind strongly to C. diphtheriae. WGA recognizes both sialic acid units and β(1→4)-linked d-GlcNAc residues. Treatment with neuraminidase significantly decreased WGA binding and induced PNA binding to bacteria. These data suggest that the WGA-binding glycoconjugates of C. diphtheriae contained sialic acid in their structure.

View this table:

Effect of neuraminidase (C. perfringens) treatment on lectin binding to C. diphtheriae

FITC-lectin (40 μg ml−1)Bacterial cell
  • +, specific binding; − no binding.

Since bacterial neuraminidases have been shown to cause C-dependent tissue damage by indirectly activating the alternative C pathway [16], the observations reported suggest another possible mechanism for establishment of C. diphtheriae infection. An understanding of the roles of C. diphtheriae neuraminidases in pathogenesis must await studies of their substrate specificity and their effects on the individual host protective proteins.

Very little work has been done on transport of sialic acids into cells, its utilization within the cell and the possibilities of recycling [2]. Natural substrates for the enzyme occur in sites in which the diphtheria bacillus is found in man; preliminary experiments were unable to find evidence of substrate in various fractions of the diphtherial cells [1]. Incorporation of sialic acids by biosynthesis via nucleotide sugar transfer (CMP-Neu5Ac) to acceptors is still a possibility.

Using a sialidase substrate, N-acetylneuramin-lactose, we demonstrated for the first time the presence of C. diphtheriae bacterial cell surface-associated trans-sialidase activity (Fig. 2). We documented the presence of trans-sialidase activity by observing bacterial interactions with fluorescent WGA and PNA lectins. Acquisition of sialic acid residues from sialyl-lactose substrate by trans-sialidase-positive microorganisms led to an increase in bacterial WGA binding and inhibition of PNA interaction. Sialylation of bacterial surface occurred only when C. diphtheriae strain was incubated with a sialoglycoconjugate donor substrate.


Fluorescence titers of neuraminidase-treated C. diphtheriae cells incubated in the following conditions: PBS (A); PBS+2 mM N-acetylneuramin-lactose (B); PBS+1 mM N-acetylneuraminic acid (C). •, Fluorescence with FITC-PNA (500 μg ml−1); ◯, fluorescence with FITC-WGA (500 μg ml−1).

Several investigations have described in detail the structural, functional and biological properties of trypanosome trans-sialidase [810]. The inferences about the role of the trans-sialidase in adhesion and other parameters of host-parasite interactions were all based on assays performed in vitro. To begin to determine the possible significance of the trans-sialidase in an in vivo situation investigators followed the course of infection of chagasic mice sensitized with purified trans-sialidase and trans-sialidase antibodies before parasite challenge. trans-Sialidase injections greatly enhanced parasitemia and mortality. The microbial trans-sialidase released into the environment binds to receptors and modifies host functions such as those of inflammatory cells, ultimately resulting in enhanced invasion. Also, microorganisms can rapidly acquire sialic acid residues from the environment by the catalytic action of the trans-sialidase, which may confer infective properties to the parasite [8].

Thus, in theory, the trans-sialidase mechanism is of importance for any pathogenic microorganism. However, trans-sialidase activity has not been described for bacteria until now. In conclusion, it seems that a single enzyme hydrolyzes sialic acid from glycoconjugates and it is capable of transferring sialic acid residues from a sialyl-lactose donor present in culture medium to C. diphtheriae-suitable β-Gal acceptors. A single molecule probably may express both neuraminidase and trans-sialidase activities [10]. This possibility is under investigation.


The authors are grateful to CNPq, CAPES, FAPERJ, SR2/UERJ and Programa de Núcleos de Exelência (PRONEX) of the Brazilian Ministry of Science and Technology, for providing financial support.


View Abstract