Search | Send us your comments

Two-Factor crosses

©2000 written by Gary Roberts, edited by Timothy Paustian, University of Wisconins-Madison

VIII B. TWO-FACTOR CROSSES

While two-factor mapping has ceased to be the tool of choice for most genetic analyses, it is appropriate to spend some time on it because it provides insight into the nature of linkage, which is relevant in any DNA manipulation involving homologous recombination. The following discussion of two-factor crosses is included . They are also of historical interest since two-factor crosses are the sorts of crosses typically thought of when the term mapping is used. Finally, the ideas behind two-factor crosses are relevant to many strain constructions when unselectable mutations are introduced into a strain using nearby selectable markers. Use of a two-factor cross demands (a) that the two markers in question be genetically linked in order to get meaningful results. In this system, lack of linkage (the negative result) argues that the two markers are unlinked. (b) The two markers used must be phenotypically dissimilar, because one marker has to be selected in the cross and the other marker scored. For example, if one was testing the linkage between a his mutation and an arg mutation, one might do the cross on arginine-containing media, thus demanding a histidine prototroph (and therefore a pair of recombination events in the his region) and then score the arginine phenotype of those colonies that appeared by replating in the presence and absence of an arginine supplement.

Let us consider an example of 2-point mapping between two mutations by generalized transduction (the discussion also holds for the case of transformation). The question of linkage can effectively be broken into two parts: (1) are the mutations linked and (2) how linked are they? The first question effectively asks if the two markers in question can be carried by the same transducing particle, that is, are they separated on the chromosome by less than one "phage-length". The figure below presents some of the considerations.

In section VII A, it was suggested that bacterial geneticists do not perform the " E. coli" control in complementation, that is, they do not construct a strain having both closely linked mutations on the same DNA molecule. The reason for their reluctance is the difficulty of the construction: to construct the double mutant, you need to move one mutation into a chromosome already containing the other without recombining the latter mutation out. Since the mutations are, by definition, closely linked, this would be a rare event. Worse, since the two mutations probably confer a similar phenotype (remember, they are candidates for affecting the same gene), you have no easy selection for the construction nor do you have an easy screen to determine if the desired strain has been successfully built. There are tricks around each of these latter problems, but they are very labor intensive and therefore not acceptable for analyzing even moderate numbers of mutations.

The concepts of linkage and 2-point crosses are relevant whenever one attempts to construct strains or plasimds by recombination.

[See sample problem 20]

[Previous] | [Next]