To understand the function of genes inactivated by Cre-mediated recombination and the resulting phenotypes, it is essential to know the exact location as well as the efficiency of the recombination events. This can be achieved by histological analysis of the target protein by antibody staining but will only be possible when an excellent antibody is available and the expression is high enough. The generation of reporter models, capable of visualizing Cre activity by expressing fluorescent proteins, or proteins that can be easily detected via enzymatic reactions, allows the localization of active Cre but may not provide reliable knowledge about the recombination efficiency in each individual case, due to variability in the efficiency of Cre between different floxed loci, induction protocols, developmental stages, genetic backgrounds or even between littermates [5, 6, 11, 12]. However, once the location has been established, alternative methods may be used to determine the recombination efficiency of the target locus in individual mice.
Since Cre-mediated recombination may result in the conversion of a floxed allele to a deletion allele only in a subset of cells, the changes to be measured are usually subtle and thus require a highly quantitative and robust technique. To quantify Cre-mediated recombination, southern and western blotting for example can be used but require a relatively high amount of DNA or protein, are labor-intensive and will not be able to detect small differences with sufficient reliability. More commonly nowadays, qPCR has been applied, a method that is generally used to detect two or more fold differences. We successfully applied qPCR to measure the percentage of Pkd1
del2-11 allele [5, 6]. However, despite the use of high-quality DNA as assessed by nanodrop measurements and extensive optimizations, qPCR seemed to be susceptible to random and systematic errors. Presumably, since signals from different reactions, a reference PCR and a deletion specific PCR, are compared to each other. In this paper we describe and validate the eMLPA technique, a novel strategy which can be used in a 96-well format and measures all parameters in a single reaction, allowing all possible information within a sample to be retrieved and experimental variation to be minimized.
Although regular MLPA is also able to measure multiple loci in one reaction, it cannot measure the floxed and the deletion allele simultaneously, due to the presence of a common Lox-P site. Furthermore, when the extra-chromosomal deletion-circle is still present, MLPA will not be able to distinguish this fragment from the floxed allele, making it impossible to measure the reduction of the floxed allele. The results obtained by eMLPA will not be affected by the deletion-circle, provided that the different fragment sizes do not overlap with each other. Probes should therefore be selected in such a way that all loci to be tested, including the possible deletion-circle, will result in peaks of different sizes. With eMLPA we could reproducibly quantify the recombination efficiency in our Pkd1 conditional knockout model using 300 to 500 ng of DNA. However, also at lower DNA concentrations of 10 to 25 ng per hybridization, especially relevant when working with small specimens, valuable information on the recombination efficiency can be obtained by increasing the amount of cycles of amplification and perform more PCR reactions on a single hybridization. Although, as described previously  relative signal intensities between peaks did not change significantly by varying the amount of cycles in the PCR and the inter-experimental variation in eMLPA is considerably smaller compared with qPCR, we recommend including a set of six to eight control samples in each experiment to obtain a reliable reference.
We observed differences in recombination efficiencies between adult and newborn mice. Since newborn mice received tamoxifen in the milk via weaning mothers and adult mice directly through a feeding needle, adult mice may receive higher concentrations of tamoxifen. In addition we also measured two-fold higher Cre-expression in adult mice . Other explanations however, like differences in tissue composition, cannot be ruled out .
Using eMLPA, we could clearly detect the deletion-circle, even five months after activation of Cre by tamoxifen. In Pkd1
del2-11, lox mice, carrying the deletion allele in the germ line and lacking Cre, this fragment was not detected. Since the deletion-circle lacks an origin of replication it is unlikely that it will be replicated during cell-division. However, in non-dividing cells, the deletion-circle apparently can be stable for a long time without being degraded, a phenomenon which also has been seen in plants and in Drosophila but, to the best of our knowledge, has not yet been reported for mouse models [13, 14]. Even more, the Drosophila white gene on the extra-chromosomal circle was expressed in the eyes during a substantial period of time and the expression could increase after excision . Therefore, when larger DNA fragments are excised containing entire genes that potentially could be transcribed, it may be desirable to follow the fate of the excised fragment. We observed variability in the height of the peak that corresponded to the deletion-circle. This may be explained by the type of DNA-isolation protocol that we used, designed to isolate high-molecular genomic DNA, but could also result from variation in proliferation between different mice. Upon isolation of total DNA, it should be possible to simultaneously quantify the recombination efficiency and the amount of circular excised fragment using eMLPA.
In addition, to the Pkd1 inducible and conditional knock-out model, eMLPA was used to measure recombination efficiency of the Fc gamma RII gene, in a B-cell specific inducible Cre mouse-model. In this model the recombination efficiency in spleen, measured by eMLPA, resembled the level of Fc gamma RII protein in B-cells as quantified by FACS-analysis. Even more, also in these samples the deletion circle was clearly detectable (Borros et al. manuscript in preparation).