Journal of Microbiological Methods 52 (2003) 69 73 www.elsevier.

com/locate/jmicmeth

Quantitative determination of pneumococcal capsular polysaccharide serotype 14 using a modification of phenolsulfuric acid method
Gabriela Cuesta, Norma Suarez, Maria I. Bessio, Fernando Ferreira, Hugo Massaldi*,1
gico y Produccio n, Instituto de Higiene, Universidad de la Repu blica, Alfredo Navarro 3051, Departamento de Desarrollo Biotecnolo C.P. 11600 Montevideo, Uruguay Received 19 March 2002; received in revised form 26 April 2002; accepted 21 June 2002

Abstract The capsular polysaccharide of Streptococcus pneumoniae, serotype 14, is part of every pneumococcal vaccine presently in the market or under development. A strategy for the quantitative determination of this polysaccharide by the phenol sulfuric acid method is described. The modality of acid addition is shown to be the critical step for obtaining reproducible test results between different technicians. Raising the incubation temperature above 80 jC increased the consistency of the method by more than 60% regardless of the acid addition modality, but at the expense of some loss of sensitivity. Incubation at 110 jC was found necessary to obtain reproducible results within 3% for this technique, which was used to follow the enrichment of the polysaccharide during the last steps of purification. A model mixture of the component polysaccharide sugars provided an adequate and economic standard to construct the calibration curve for this assay, with absorbance reading either in the reaction tubes or in a microplate. A similar procedure may be applied to the determination of other bacterial polysaccharides as well. D 2003 Elsevier Science B.V. All rights reserved.
Keywords: Bacterial polysaccharide; Colorimetric determination; Vaccines

1. Introduction Highly purified pneumococcal capsular polysaccharides (CPS) are necessary for the production of second generation pneumococcal vaccines. In particular, serotype 14 of Streptococcus pneumoniae is one

Corresponding author. Fax: +598-2-4873073. E-mail address: hmassaldi@hotmall.com (H. Massaldi). 1 Member of CONICET, Argentina.

of the most prevalent serotypes worldwide, and is part of every polysaccharide-based vaccine against this pathogen that is in the market or under development. The purification of these polysaccharides, which are present in the culture broth at rather low concentration levels, often requires the use of various methods for monitoring the enrichment along the various processing steps. The classical phenol sulfuric method (Dubois et al., 1956) is a colorimetric reaction frequently used to determine mono- and oligosaccharides, due to its simplicity and the economy of its reagents. More recently, a microassay version of the

0167-7012/03/$ - see front matter D 2003 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 7 0 1 2 ( 0 2 ) 0 0 1 5 1 - 3

70

G. Cuesta et al. / Journal of Microbiological Methods 52 (2003) 6973

method has been proposed by Fox and Robyt (1991). Later the quantitative application of the reaction was extended (Saha and Brewer, 1994) to more complex oligosaccharides and glycoproteins by constructing an adequate standard based on a model mixture of the component monosaccharides. Pneumococcal CPS serotype 14 is a heteropolysaccharide with a repeating unit of about 650 Da and a total MW in the order of 106 Da. Given the lack of simple and more specific colorimetric reactions for this complex carbohydrate, the phenol sulfuric acid method remains a useful approach. When we intended to use this method for our application, however, we found that the development of colour was greatly dependent on manipulation details and the technique had poor reproducibility between different technicians, either in the classicalmacroor in the microassay version. In particular, the modality of sulfuric acid addition appeared to be critical. The same problem was pointed out by Monsigny et al. (1988) when trying to apply a microassay for the reaction. In this paper, we describe the development of an experimental strategy that greatly improves the methods reproducibility and robustness, and thereby allows its application to the determination of CPS 14 of S. pneumoniae.

the side of the tube. The tubes were then closed, vortexstirred for 5 s and incubated for 30 min either at room temperature (20 22 jC) or at 80 jC, or for 15 min at 110 jC. For the higher temperatures, incubations were conveniently performed using an oil bath. All tubes were allowed to cool down to room temperature before reading the absorbances at 490 nm using distilled water as blank in a Ultrospect 1000 UV/Visible spectrophotometer (Pharmacia Biotech., Uppsala, Sweden).

2.3. Application to the determination of CPS serotype 14 Pneumococcal CPS serotype 14 was purified (Fer rez et al., 2001) from cultures of reira et al., 1998; Sua pneumococcal strains obtained from the National Centre for Streptococcus (Alberta, Canada). This polysaccharide is composed of a tetrasaccharide repeating unit made up of 1 U D-glucose, 2 U D-galactose and 1 U N-acetyl-D-glucosamine (Lindberg et al., 1977). A model mixture of these hexoses in the same molar ratio as in the polysaccharide structure (Saha and Brewer, 1994) was used as the standard for the calibration curve. The same protocol as described in Section 2.2 was used to determine the polysaccharide. The tubes were incubated either at 80 jC for 30 min, or at 110 jC for 15 min, and were left to cool down to room temperature before reading the absorbance at 490 nm. Alternatively, 150 Al of each reacted sample were transferred to a microplate and read in a MRX2 Dynex microplate reader (Chantilly, USA) at the same wavelength. A corresponding calibration curve was constructed by using natural, purified polysaccharide as the standard. A microassay calibration curve following Fox and Robyt (1991) was also carried out with the model mixture for comparison. 2.4. Statistical data treatment

2. Materials and methods 2.1. Reagents Sulfuric acid 96.8 g/dl, reagent-grade (Baker), was used. A 5 g/dl phenol (reagent-grade, Baker) water solution was prepared. Hydrated maltose and D-glucose, D-galactose and N-acetyl-D-glucosamine were obtained from Sigma (St. Louis, MO). All monosaccharide standards were kept in a dessicator over P2O5. 2.2. Phenol sulfuric acid method The basic protocol of Dubois et al. (1956) was followed, with the modifications indicated below. The sugar solution (0.6 ml) and the phenol solution (0.3 ml) were added to screw cap tubes (13 100 mm), which were capped and vortex-stirred. Then 1.5 ml of concentrated sulfuric acid was added either directly against the liquid surface in 2 or 10 s, or slowly down

Data points were taken in triplicate. Data were processed by regression analysis (Miller and Miller, 1993) from which the standard deviation at each concentration was estimated. The standard deviation of a simulated mixed experiment at room temperature with all acid addition modalities was calculated by using the means of the data points for each respective curve.

G. Cuesta et al. / Journal of Microbiological Methods 52 (2003) 6973

71

3. Results and discussion 3.1. Influence of sulfuric acid addition on colour development Several calibration experiments were carried out with each of the modalities of acid addition described above at different temperatures, in order to observe the effect on the overall method reproducibility and sensitivity. These modalities reflect the varieties of addition modalities that have been reported in the literature and/or experienced by us. Fig. 1a shows calibration curves for maltose obtained by incubation at room temperature, using these modalities. It is apparent that acid addition over the liquid surface at different velocities does not cause appreciable differences and both sets of data can be represented by a unique calibration curve, with a rather low standard deviation for each data point. On the other hand, when the acid is added over the side of the tube, the sensitivity achieved is about 70% lower (given by the difference in slope of the curves), but with the same level of standard deviation. Fig. 1b shows the calibration curve that results from averaging the means of the data points of the previous curves. A large, increasing dispersion of the data points, albeit with a linear behavior, is observed in this case, reflecting the simulated mixed use of the addition modalities. A strong lack of linearity, however, is what should be expected from a real experiment with the same, or different, technicians when the same modality of acid addition is notor cannot befollowed consistently. The situation may be even worse (more discrepancy) when a complex carbohydrate is being determined, due to insufficient bond breakage. We reasoned that the discrepancies between results given by the different addition modalities are related mainly to differences in the rate of heat dissipation and the intensity of spontaneous agitation of the mixture upon acid addition. Saha and Brewer (1994) recommended increasing the speed at which the acid is added to the solution, emphasizing that rapid addition of acid generates enough heat so as to break down all glycosidic bonds in complex carbohydrates. For monosaccharides, our results indicate that speed of acid addition is not critical as long as the acid is added directly over the liquid surface. Nevertheless, this implies a rather precise control of

Fig. 1. Calibration curves for maltose, incubation at 20 22 jC. (a) Acid addition directly over the surface: 10 s (5), 2 s (n); acid addition over the side of the tube (x). (b) Average of the three modalities.

the addition modality, something that may not be possible to ensure among different technicians. We have found that slow addition against the side of the tube is the safest operating condition, and one that can be carried out in any laboratory provided with a burette. Its drawback is the low sensitivity achieved, as shown in Fig. 1a. For this reason, we pursued a compromise between safety and sensitivity by raising

72

G. Cuesta et al. / Journal of Microbiological Methods 52 (2003) 6973

Fig. 2. Calibration curve for maltose, incubation at 80 jC. Acid addition directly over the surface: 10 s (5), 2 s (n); acid addition over the side of the tube (x).

Fig. 3. Calibration curve for maltose, incubation at 110 jC. Acid addition directly over the surface: 10 s (5), 2 s (n); acid addition over the side of the tube (x).

the incubation temperature. The results are shown in Fig. 2, for incubation at 80 jC. Again, the curves for addition over the side of the tube show no difference between them, but now, addition over the side of the tube indicates a much improved sensitivity. In this case, the difference in the slopes is about 30%, implying an improvement in consistency of 60% with respect to the situation at room temperature. Fig. 3 shows the corresponding situation for an incubation temperature of 110 jC. It is apparent that full consistency is now achieved, that is, all data can be represented by the same calibration curve. This is accomplished with minimal dispersion and a reproducibility better than 3%, as estimated from the relative standard error of the regression line at the central point (Goldstein, 1964). The price, however, is a reduced sensitivityabout 30% lower slopewith respect to that at 80 jC. 3.2. Quantification of pneumococcal CPS serotype 14 We used the phenol sulfuric acid method as one of the battery of techniques necessary to follow the polysaccharide enrichment during its purification process from a culture broth. Given its broad range

Fig. 4. Calibration curves for polysaccharide serotype 14 of S. pneumoniae. Reading in tubes: model mixture (5); purified polysaccharide (n). Reading in microplate: model mixture (x).

G. Cuesta et al. / Journal of Microbiological Methods 52 (2003) 6973

73

of specificity, the method can be more reliably applied during the last stages of purification, where the interference of contaminants has been substantially reduced, and/or more precisely determined. Incubation in tubes at 110 jC was used, since it was found to be much more reproducible than that at 80 jC either in tubes or in microplates. For the pneumococcal CPS serotype 14, the use of a model mixture of the component monosaccharides to construct the calibration curve, as proposed by Saha and Brewer (1994), represents a practical and economic alternative to the use of the natural polysaccharide. The adequacy of this procedure is shown in Fig. 4, where the data using the natural polysaccharide and the model mixture can be reasonably represented by a unique calibration curve. The slight departure between both sets of data is most probably due to the inherent hygroscopicity of the polysaccharide, which shows up as an excess weight in the corresponding calibration curve. An alternative that combines part of the benefits of the microassay and the robustness of the assay carried out in tubes was also tested for the model mixture. After completion of the reaction at 110 jC and the cooling down step, aliquots of the reacted samples were taken and transferred to an ELISA plate to read absorbance in a microplate reader. This procedure also yielded reproducible results, but at the expense of greatly reduced sensitivity, as shown by the lower calibration curve in Fig. 4. Finally, since N-acetyl-Dglucosamine contributes minimally to colour development (Saha and Brewer, 1994), for the sake of simplicity and economy, we tried omitting this component from the model mixture. However, even when we compensated for the modified molecular weight of the repeating unit, the results were not reproducible, and therefore, we maintained the original formula. 3.3. Conclusions The phenol sulfuric acid method can be conveniently applied to the quantitative determination of the capsular polysaccharide serotype 14 of S. penumoniae. This can be achieved by means of two modifications of the classical procedure: (1) a slow, careful acid addition over the side of the tube, with incubation

at 110 jC, either with absorbance reading directly in the tubes or using a microplate reader; and (2) using a model mixture as the standard, made up of the individual sugars of the polysaccharide in the same molar ratio as in the natural substance. This procedure achieves a good reproducibility between assays carried out by different technicians, allows safe manipulation, and provides calibration curves of adequate sensitivity. A similar approach may be applied to the quantitative determination of other complex polysaccharides of bacterial origin.

Acknowledgements This work was supported in part by a grant from the Pan American Health Organization.

References
Dubois, M., Gilles, K.A., Hamilton, J.K., Rebers, P.A., Smith, F., 1956. Colorimetric method for determination of sugars and related substances. Anal. Chem. 28, 350 356. Ferreira, F., Franco-Fraguas, L., Cantera, A., Texeira, E., Massaldi, rez, N., Altez, M., 1998. Production and purification of H., Sua polysaccharide serotype 14 of S. pneumoniae. Biotecnol. Apl. 15, 192. Fox, J.D., Robyt, J.F., 1991. Miniaturization of three carbohydrate analyses using a microsample plate reader. Anal. Biochem. 195, 93 96. Goldstein, A., 1964. Biostatistics: An Introductory Text, 1st ed. Macmillan, New York, pp. 139 141. Lindberg, B., Lo nngren, J., Powell, D.A., 1977. Structural studies on the specific type-14 pneumococcal polysaccharide. Carbohydr. Res. 58, 177 186. stica para Qu mica Anal tica, Miller, J.C., Miller, J.N., 1993. Estad 2nd ed. Addison-Wesley Iberoamericana, Buenos Aires, pp. 88 106. Monsigny, M., Petit, C., Roche, A.C., 1988. Colorimetric determination of neutral sugars by a resorcinol sulfuric acid micromethod. Anal. Biochem. 175, 525 530. Saha, S.K., Brewer, C.F., 1994. Determination of the concentrations of oligosaccharides, complex type carbohydrates, and glycoproteins using the phenol sulfuric acid method. Carbohydr. Res. 254, 157 167. rez, N., Franco-Fraguas, L., Texeira, E., Massaldi, H., Batista, F., Sua Ferreira, F., 2001. Production of capsular polysaccharide of Streptococcus pneumoniae type 14 and its purification by affinity chromatograpy. Appl. Environ. Microbiol. 67, 969 971.