Most yeast, fly and worm geneticists know how useful modifier genetic screens can be; screening for mutants that enhance or suppress a certain phenotype can be used to build up the signalling cascades that underlie that trait. Marina Carpinelli, Doug Hilton and colleagues now show that suppressor screens are equally applicable to vertebrate models in general and to mice in particular.

The authors focussed on a mouse model of thrombocytopenia — a disease that is caused by a lack of blood platelets and that is seen in Mpl−/− animals. They treated 300 male, Mpl homozygous knockout mice with the mutagen N-ethyl-N-nitrosourea (ENU). These mutagenized mice were then crossed with isogenic knockout females. A total of 5 of the 1,575 F1 offspring from these crosses suppressed the mutant phenotype: their platelet counts far exceeded those of their thrombocytopenic parents. Subsequent crosses with untreated Mpl−/− mice showed that two of the suppressed mice (Plt3 and Plt4) harboured dominant ENU-induced mutations.

So, without too much trouble, Carpinelli et al. were able to use a large-scale suppressor screen to isolate new mutant mice that are relevant to the study of the platelet-production pathway. However, it is their subsequent work that shows just how useful mutants isolated from suppressor screens can be. First, they intercrossed mice that were heterozygous for the new mutations to obtain homozygotes in an Mpl−/− background. These mice had platelet levels that were above normal, in contrast to the below wild-type levels in the heterozygotes, which clearly indicated that Plt3 and Plt4 are semi-dominant mutations.

Further crosses of Plt3/Plt4 compound heterozygotes to Mpl−/− mice showed that the two mutations are tightly linked and possibly allelic. Using standard segregation analysis of 148 microsatellite markers, the authors localized the mutations to a region on chromosome 10 that contains cMyb, a gene that was previously linked to elevated platelet levels. They then identified candidate mutations in cMyb, in both Plt3 and Plt4 mice, that caused single amino-acid substitutions in functionally significant regions of the cMyb protein. Follow-up protein transactivation assays confirmed that these mutations reduce the activity of cMyb. Coupled with the finding that cMyb heterozygous knockout mice have higher platelet levels than wild-type homozygotes, these studies provide compelling evidence that the cMyb mutations underlie the Plt3 and Plt4 gain-of-function phenotypes.

Detailed haematological analyses of heterozygous and homozygous mutants showed that Plt3 and Plt4 increase the production of megakaryocytes and their progenitors, from which platelets are derived. So, we now have a more complete picture of the role of cMyb in haematopoiesis, but the real beauty of mutants that are identified from a supressor screen is that their epistatic interactions with the original mutant can be genetically analysed. In this case, Carpinelli et al. clearly showed that the effect of mutations in cMyb on platelet and megakaryocyte levels was independent of the Mpl genotype. So, the semi-dominance of mutations in cMyb, regardless of the genetic background, indicates that downregulation of cMyb is probably an important step in platelet production and that the gene is in the same signalling pathway as Mpl.

It seems to have taken vertebrate geneticists some time to catch on to the allure of suppressor screens. However, if the success of this initial application of the approach is anything to go on, it seems likely that we will soon see many more examples.