(273c) Computer-Assisted Analysis of 2-D Gel Patterns Derived from Agarose Electrophoresis of Conjugated Hib Meningitis Vaccines

Although analytical methods for vaccine characterization based on 2-D agarose electrophoresis were developed in the period from 1983 to 1995, they remain a promising tool for vaccine quality control and for predicting vaccine effectiveness in the production of semi-synthetic immunogens (See reference [1] for a recent review).

Vaccine samples

The protein-polysaccharide conjugated meningitis vaccines were developed by Robbins, Schneerson, and coworkers [2, 3] for the immunization of small children, the main target group of bacterial meningitis caused by Haemophilus influenzae type b (Hib). In 1996 Robbins and Schneerson shared a Lasker Clinical Research Award for their groundbreaking work [4].

Earlier vaccine preparations varied in immunogenicity, and immunological tests for identifying potentially effective vaccines were time consuming. Other physical methods for characterizing the vaccines were unsatisfactory due to the high molecular weight of vaccine particles and their highly negative charge. For example, in Sepharose gel filtration, vaccine samples could only be physically defined by the fact that they were a part of the void volume.

Development of computer-assisted 2-D agarose electrophoretic procedures

In search of better analytical methods, a computer-assisted electrophoretic method for analyzing such vaccines was developed. Since the vaccine particles were very large with sizes in the range of intact viruses, dilute agarose gels were the medium of choice. However, one-dimensional gels yielded only an uninformative smear (Fig. 1A). By adding a second dimension of electrophoresis using Serwer's horizontal 2-D submarine-type apparatus (Fig. 1B), gel patterns could be obtained that varied depending on the conditions of the particular vaccine preparation and were therefore characteristic of each vaccine sample (Fig. 1C).

Computer programs ElphoFit [5] and GelFit [6] were employed for the evaluation of the 2-D vaccine patterns. In particular, the program ElphoFit was used to evaluate the gel electrophoretic data and to standardize the gel on the basis of the extended Ogston model [7-10], as shown in Fig. 9 of ref. [1]. The output of ElphoFit was then transferred to computer program GelFit, which transformed the original digital images (Fig. 1C) from a curvilinear to a rectangular coordinate system of particle radius and free mobility, as shown in Fig. 1D. Vaccine samples showed a continuous size distribution over a wide range (polydisperse particle mixture) due to the randomizing steps involved in the vaccine preparation (sonication and crosslinking). Another application of GelFit was the stripping of 2-D Gaussian surfaces (Fig. 6 of [1]), which was used to demonstrate that vaccine II in Figs. 1C and 1D consisted of at least three particle populations.

Biomedical significance

The most significant advance made in the reviewed 2-D experiments was that vaccine samples could be characterized according to particle size and free mobility (related to surface net charge density), even though the immunogens exhibited a polydisperse size distribution with particles as large as or larger than intact viruses. 2-D vaccine patterns varied depending on the conditions of a particular vaccine preparation and, therefore, they were characteristic for each conjugate preparation. The electrophoretic characterization may also be useful for determining vaccine effectiveness and the effects of procedures such as storage, lyophilization, and sterile filtration on the composition of the vaccines. Other more general biomedical applications are the analysis of complex mixtures of macromolecules, small cell organelles, viruses, or other particles of sub-cellular size.

Acknowledgements

The author thanks the late Andreas Chrambach (SMA, NICHD, NIH) for support, stimulating discussions and for providing laboratory space from 1983 - 93, Akram Aldroubi and Michael Unser (BEIP, NIH) for supplying computer program GelFit, and Benes Trus (CIT, NIH) for instruction in image processing. Philip Serwer (UTHCC, San Antonio, TX) provided training in the use of the 2-D electrophoresis apparatus in his laboratory. Vaccine samples were kindly provided by Rachel Schneerson (LDMI, NICHD, NIH). Mark Geanacopoulos (JNCI, Oxford University Press) provided a critical review.

Fig. 1A: Gel patterns of meningitis vaccines at different agarose concentrations, using a 1-D submarine electrophoresis apparatus [11]. The samples yielded an uninterpretable smear, although electrophoretic conditions were appropriate.

Fig. 1B: Schematic of the 2-D horizontal submarine-type apparatus reported by Serwer [12]. Serwer's method, similar to the technique of O'Farrell [13], allows for a separation of particles predominantly according to charge in the first dimension, and according to size in the second dimension. However, the technique of Serwer rests on a different principle: Samples are first electrophoresed in a gel track of low concentration, then the field is switched perpendicularly and the samples are run into a relatively more concentrated frame gel which surrounds the first dimensional track. Gels need not be touched during the procedure; this makes it possible to handle the fragile gels suitable for the separation of nondenatured particles in the size range of viruses (> 3000 kDa). O'Farrell gels, by comparison, are typically used for much smaller proteins or protein fragments (10 - 500 kDa) which are denatured by sodium dodecyl sulfate (SDS). (Adapted from Fig. 1 of [1].)

Fig. 1C: Two-dimensional gel patterns of meningitis vaccines (I to III) and of hydrophilic polystyrene size standards (S) with 45 and 46.5 nm radius. The origin of electrophoresis lies outside the pictures. First dimension (top to bottom): 0.15 % agarose (SeaPlaque), 3 V/cm, second dimension (left to right): 0.8 % agarose, 1.5 V/cm. Electrophoresis was in phosphate buffer pH 7.2 using the modified [14] 2-D submarine electrophoresis apparatus [12] shown in Fig. 1B, staining: Coomassie Blue R 250. Scanning of the stained patterns employed a Perkin-Elmer 1010MG microdensitometer interfaced with a PDP 11/34 (Digital) computer. The vaccines produce characteristic gel patterns, which depend on the nature of the sample. The samples consist of Haemophilus influenzae, type b, capsular polysaccharide crosslinked to tetanus toxoid (Panels I and II), or P2 protein (Panel III). Experimental conditions and physical parameters of vaccine particles and polystyrene spheres were summarized in Tables 1 and 2 of ref. [15]. (Adapted from Fig. 4 of ref. [1].)

Fig. 1D: Patterns of Fig. 1C which have been transformed from the original curvilinear to a rectangular coordinate system of particle size and free mobility (related to surface net charge density). Vaccines II and III, which were effective immunogens have a considerably larger size distribution than sample I (not effective). The vaccines I and II have a much larger variation in free mobility than III, since protein P2 is well defined, whereas tetanus toxoid is a mixture of many components. It should be noted that the 2-D electrophoresis used here achieves results similar to O'Farrell's technique, but relies on a different principle: Predominant charge- (1st dimension) and predominant size-separations (2nd dimension) are achieved by using gels with low and relatively high agarose concentrations under non-denaturing conditions. By applying a mathematical approach (Appendices of [1]) based on the extended Ogston model [7-10], one can distinguish the separation effects due to particle size and charge. The investigated particles are in the size range of viruses (10,000 to 2,000,000 kDa). (Adapted from Fig. 5 of [1].)

References

[1] Tietz, D, Electrophoresis 2007, 28, 512?524.

[2] Schneerson, R., Barrera, O., Sutton, A., and Robbins, J. B., J. Exp. Med. 1980, 152, 361-375.

[3] Claesson, B. A., Trollfors, B., Lagergard, T., Taranger, J., Bryla, D., Otterman, G., Cramton, T., Yang, Y., Reimer, C. B., Robbins, J. B., Schneerson, R., J. Pediatrics 1988, 112, 695-702.

[4] Robbins, J. B., Schneerson, R., Anderson, P., Smith, D. H., JAMA 1996, 276, 1181?1185.

[5] Tietz, D, Electrophoresis 1991, 12, 28-39.

[6] Aldroubi, A., Unser, M., Tietz, D., and Trus, B., Electrophoresis 1991, 12, 39-46.

[7] Ogston, A.G., Trans. Faraday Soc. 1958, 54, 1754-1757.

[8] Rodbard, D. and Chrambach, A., Proc. Natl. Acad. Sci. USA 1970, 65, 970-977.

[9] Tietz, D., Chrambach, A., Anal. Biochem. 1987, 161, 395-411.

[10] Beyer, K., Tietz, D., Adv. Electrophoresis 1988, 2, 145-151.

[11] Serwer, P, Electrophoresis 1983, 4, 227-231.

[12] Serwer, P., Anal. Biochem. 1985, 144, 172-178.

[13] O'Farrell, P. H., J. Biol. Chem. 1975, 250, 4007-4021.

[14] Tietz, D. and Chrambach, A, Electrophoresis 1989, 10, 667-680.

[15] Tietz, D., Aldroubi, A., Schneerson, R., Unser, M., Chrambach, A., Electrophoresis 1991, 12, 46-54.