Lousy legend. Let’s go to the source.
Fig. 4 from Rallis et al. paper (direct link), corresponding to the top panel of cteappv’s figure 10:
Scanning electron micrographs of (A) control and (B) dominant-negative Tbx5 (Tbx5delta) injected limb buds at stage 20 show a dramatic reduction in limb size after Tbx5delta misexpression.
Certainly not “a mutant which lacks a given gene“.
This is a competition assay, between endogenous Tbx5 and injected Tbx5delta in wild type animals, gene’s transfer being via recombinant retroviruses.
Is the mutant presented at the bottom panel “a mutant which lacks a given gene“? Not. [see fig 1 of the source]
The transgenic mice have an inducible Cre recombinase expression under the Prx1 promoter. That means that the recombinase will be expressed in cells where Prx1 is naturally expressed. The Tbx5 gene is lox-ed: it can be excised from the genome by the action of Cre recombinase.
That means that Tbx5 expression will be impaired in cells expressing Cre recombinase under the control of the Prx1 promoter.
An elegant system allowing local knock-out of a gene otherwise essential for development, for which general knock-outs would be lethal. Allowing to validate a causal relation between an expression controlling element and a downstream situated gene. Once more, certainly not “a mutant which lacks a given gene“.
When one wants to present “Typical genetic assays on limb development“, especially to a public not accustomed with the presented techniques, it’s convenient for everybody that one understands what he is talking about.