Principles of classification


1. INTRODUCTION

The first attempt to introduce some order into the growing number of serologically different strains of Leptospira was undertaken by Wolf and Broom in 1954. They formulated the first principles to differentiate serovars*, previously desgnated as serotypes, on the basis of their serological characteristics using rabbit immunsera which were the only practical criteria available at that time. It was considered that two strains "belong to different serotypes if, after cross-absorption with adequate amounts of heterologous antigen 10% or more of the homologous titre regularly remains in each of the two antisera" (WHO, 1956). For convenience, serovars with close serological affinities were assembled within a serogroup.

Some years later a further subdivision into 'sub-serrotypes' was introduced. Strains were clasified as 'sub-serotypes' if "in repeated test less than 10% of the homologous titre remains in one antiserum but 10% or more in the other antiserum after cross-absorption with adequate amounts of heterologous antigen" (WHO, 1959).

However further studies, on the antigenic structure and its relevance to clasification, indicated that the taxon 'sub-serotype' was not a reliable concept since it did not define which of the two related strains should be recognized as the 'serotype' and which as its 'subserotype' (Kmety, 1966).

At its meeting in Moscow in 1966 the Subcommittee on the Taxonomy of Leptospira (TSC) accepted the above mentioned objections and abolished the taxon 'sub-serotype'. As a consequence the definition of a 'serotype' was amended as follows:"two strains are considered to belong to different serotypes if, after cross-absorption with adequate amounts of heterologous antigen, 10% or more of the homologous titre regularly remains in at least one of the two antisera in repeated tests".

Along those lines all previously designated 'sub-serotypes' had to be reclassified as 'serotypes' and a newly approved list (WHO, 1967) was drawn up containin 124 named 'serotypes' devieded into 18 seroroups.

It should be noted that this list does not differentiate saprophytic and parasitic-pathogenic strains, in spite of the decision taken at the TSC meeting in Moscow (1966) to consider those two groups as 'complexes' of a monospecific genus Leptospira. The two 'complexes' were given the names 'interrogans' and 'biflexa' that had previously been allocated as specific names for the parasitic/parthogenic and saprophytic groups respectively. It was not until teh TSC meeting in Manchester in 1986 that the biological properties, growth at 13(o)C and resistance to the purine analogue 8-azaguanine, were recognised as being sufficiently stable differentiating markers to allow both 'complexes' to be regard as separate species known as L.interrogans and L. biflexa. Members of the species L.interrogans are for both properties negative and members of L. biflexa are for both positive.

In view of the continuous flow of newly described serovars, a growing number of controversial typing results, incomplete documentation, etd., the serovar list of 1967 has become more and more obsolate. As no new revised official list was drawn up, individual authors began to publish their own lists (Turner, 1972; Dikken and Kmety, 1978; Jonhson and Faine, 198). This situation led the TSC to assess once more all aspects of clasification of leptospires including the revison of the criteria, the standardization of serotyping methods, the principles goverming the recognition of new serovars, etc. Also new approches to the clasification of bacteria have been carefully evaluated, including the use of monoclonal antibodies and the identification of genetic characteristics.

*The term serovar was substituterd for serotype at the TSC meeting in Jerusalem in 1973.

2. PRINCIPLES OF LEPTOSPIRA CLASSIFICATION.

Although the principles of the present classification described in the pervious paragraph have fulfilled a useful role for many years, there have been reports of discrepant tyring results that indicate some weaknesses in the typing methods. Therefore it was thought necessary to revise some of those principles to ensure that the new list should be founded on an improved basis.

Firstly the definition of the taxon 'several' had to be more explicite and the serological typing methods more carefully standardized. Secondly the impact on the present classification of newly developed diagnosstic test such as the monoclonal antibody technique, restricton endonuclease analysis and others had to be taken into account. Moreover, agreement had to be reached on the guidelines to be followed in the recognition of new serovars.

All these items were discussed repeatedly during the last TSC meetin of 1978, 1982, 1986 and 1990.

2.1 Amendments to the definition of 'serovar'.

During the meeting of the TSC in Manchestger in 1986 the following revised definition of a 'serovar' was agreed: "Two strains are said to belong to different serovars it after cross-absorption with adequate amounts of heterologous antigen more than 10% of the homologous titre regularly remains in at least one of the two antisera in repeated tests".

This modification may reduce the risk of a false evaluation of the serological difference between two closely related strains. It eas also introduced because of the more general use of the two-fold serum dilution technique in place of the split or interlocking ten-fold dilution scheme known as the Dutch scheme (Schffner and Mochtar, 1927; Wolff, 1954; Dikken and Kmety, 1978), which makes a minimal 10% difference in residual titres inapplicable. Another reason for the amendment to the definitionwas the improved standardization of the absorption test which considers absorption to be well balanced it the absorbing strain still agglutinates to a maximum of 1% of the original titre. This would mean that, according to the old definition a 9% (10% - 1%) difference would have to be accepted as the basis of a valid description of the new serovar. This is avoided by the new definiton.

2.2. Standardization of typing methods.

Standardization of methods is a self-evident necessity in classification studies in order that comparable results may be obtained by the different laboratories. During each meeting of the TSC this item has been discussed but some questions still remain unanswered. However a number of recommendations were agreed at the TSC meetings in Munich 1978 (Minutes, 1982) and Boston 1982 (Minutes, 1984).

2.2.1 The preparation of rabbit immune serum.

For classification purposes, a pooled antiserum from two or three rabbits should be used with a titre between 10.000 and 50.000. It is recommended to prepare the sera by repeated intravenous injections of well grown living culture of an approximate density of 2 x 10(8) organisms per ml. (see for more details Minutes 1982 of TSC meeting Minich, 1978). One must take into account the possibility that in some strains there may be a thermolabile antigen(s) as well as the usual thermostable antiens (Borg-Peterson, 1971, 1972; Babudieri, 1972; Kmety, 1972). In order to distinguish both types of antigen other methods of immunization have been adopted (Dikken and Kmety, 1978).

2.2.2 Microscopic agglutination test (MAT).

In standarizing the MAT, two important points should be noted:

a. The density of the antigen is known to influence the end-point of agglutination and consequently the titre of the serum (Borg-Petersen and Fagreus, 1949; Jarekova, 1986). therefore the TSC (WHO, 1965; Minutes 1984 of TSC meeting Boston, 1982) recomminded the use of well-grown living cultures with an approximate density of 2 x 10(8) leptospires per ml. However, differences in the length of the leptospiral cells may influence the real density of the reacting antigen determined nephelometrically (Jerekova, 1986) who found lower densities (approximately 1,5 x 10(8) leptospires per ml) to be more suitable for the test.

b. The end-point of the reaction of defined as the reciprocal of the highest dilution of serum in the serum-antigen mixture, in which 50% or just more of the cells are agglutinated (WHO, 1965; Minutes, 1984 of TSC meeting Boston, 1982).

2.2.3 Agglutinin-absorption test.

The absorption test should be well balanced in order to ensure that neither over nor under absorption take place. To achieve this, the amount of concentrated living absorbing antigen with a density corresponding to McAarland standard No. 10 (Dikken and Kmety, 1978) or determined otherwise (Minutes, 1984 of TSC meeting Boston, 1982) should be appropriate to the height of the antibodies of the immune serum to be absorbed. If lower titre antibodies, are to be absorbed, lower amounts of absorbing antigen have to be used to achieve a well balanced absorption.

It is recommanded that 1 part of immmune serum should be mixed with 24 parts of concentrated antigen in 3 equal amounts at 10 minute intervals. The absorption is considered to be adequately balnced when the absorbed serum has a residual titre of 0,5 - 1,0% of its original titre against the absorbing antigen (see for more details Minutes, 1984 of TSC meeting Boston, 1982). Even in the case of complete absorption the test may be considered in balance provided nosubstantial reduction of the homologous serum titre is apparent.