Figure 1.
In vivo kinetics of reporter activity induced by various thymus-initiated Cre transgenes.
The percent of reporter+ cells at all major stages of thymocyte differentiation are indicated. All results except those from Cd2[Cre] include pooled data from ROSA[(stop/flox)tdRFP] and ROSA[(stop/flox)YFP] reporters, which were essentially indistinguishable from each other in all three Cre strains tested; due to the absence of any reporter strain variation, Cd2[Cre] was only tested using the ROSA[(stop/flox)tdRFP] reporter. Each data point represents the mean ± s.d. for 8–14 individual experiments (thymuses) performed over a 3-year period. All mice were hemizygous for the Cre transgene, and heterozygous for the reporter allele. Note that with the exception of Cd4, none of these Cre transgenes faithfully reflects temporal expression of the corresponding endogenous gene (see Figure S1).
Figure 2.
Reduced thymic cellularity in thymus-initiated Cre-transgenic strains.
Each bar represents the mean ± s.d. for 6–13 individual experiments (thymuses); mice were males and females, 5 weeks of age. Statistics represent the Student's two-tailed t test for unpaired samples. In mice hemizygous for the a Cre allele, both Lck strains exhibited substantial reductions in thymic cellularity, while Cd2[Cre] and Cd4[Cre] did not. Mice that were homozygous for Cre alleles revealed a further exacerbation of this phenotype that was significant even in previously unaffected Cd4[Cre] mice (note: Cd2[Cre] homozygous mice were not generated, due to embryonic lethality). Since no loxP target sequences were present, these data suggest the existence of off-target effects of Cre in developing thymocytes.
Figure 3.
Intracellular Cre protein levels correlate with reduced thymic cellularity in thymus-initiated Cre-transgenic strains.
Thymocytes from control (C57Bl/6) mice, or from mice expressing various Cre-transgenes, were isolated, fixed, and stained with an antibody recognizing the Cre protein, followed by flow cytometric analysis. Levels of intracellular Cre correlated directly with the appearance of phenotypic effects, with lowest levels in the Cd4[Cre] and Cd2[Cre] strains, and highest levels in both Lck[Cre] strains. Homozygous mice for the highest expressing lck[Cre] strain (Lck[Cre]Wil, dashed line) predictably expressed even higher levels of Cre than hemizygous mice, and had the most profound phenotype (Figs. 2 and 4). These results suggest that above a certain threshold, Cre protein is toxic to thymocytes in a dose-dependent manner. These experiments were repeated at least four times with essentially identical results.
Figure 4.
Thymus-initiated Cre expression results in a loss of DP cells.
Panel a) shows the relative proportions of various populations identified by CD4 and CD8 staining on thymocytes from various types of Cre transgenic mice, or C57Bl/6 controls. Statistics indicate mean ± s.d. for 3–9 independent experiments. Panel b) shows absolute numbers of cells per thymus for each major stage of differentiation in the same mice (dashed line = hemizygous; solid line = homozygous); note that because of extensive overlap, only the mean value is shown. Absolute numbers of early progenitor stages were very similar in all strains of mice, suggesting that the effects of Cre were minimal among at these stages. In contrast, substantial differences were seen at the transition to the DP stage. Both strains of hemizygous mice expressing Cre under the Lck promoter showed substantial reductions in DP cell number, changes that are exacerbated in homozygous mice of these strains. These data are in complete agreement with cellularity data shown in Fig. 2, and suggest that the presence of Cre and, in particular, the absolute levels of Cre, may be toxic to DP thymocytes.
Figure 5.
Phenotypes resulting from Cre expression in thymocytes correlate with an increased propensity towards cell death.
To test whether Cre protein is toxic to thymocytes in a dose-dependent fashion, analysis of DNA double strand breaks (TUNEL) was performed on wildtype thymocytes or thymocytes from mice carrying one (hemizygous) or two (homozygous) Lck[Cre] transgenes. Freshly isolated thymocytes were indistinguishable from the three types of mice, suggesting that clearance of apoptotic cells from the thymus was not affected. However, when thymocytes were cultured overnight in medium, the presence of Cre resulted in an increase in cell death that was proportional to the amount of Cre protein expressed. These results further strengthen the conclusion that Cre protein is toxic to thymocytes in a dose-dependent manner. Data represent mean ± s.d. for 3–4 independent experiments for each mouse type.