Holding tight and staying true: generation of a new tamoxifen-inducible, principal cell-specific mouse

The generation of animal models using Cre-Lox recombination provides state-of-the-art insight into gene functions, ideally under both temporal and spatial control. Certain genes have developmental functions and even conditional noninducible inactivation of these genes leads to embryonic lethality. Consequently, using noninducible Cre drivers harbors the problem that certain gene functions cannot be studied in adulthood. Over the last two decades, an increasing number of Cre models have been developed; however, some models have drawbacks such as Cre mosaicism (some cells show Cre activity but not others), Cre penetrance (strength of the phenotype), incomplete Cre recombination (excision efficiency), Cre leakiness (recombination without induction), a phenotype that occasionally Cre alone can produce (referred to as Cre toxicity), as well as others (2). While there are several noninducible Cre models available that target the kidney (reviewed in Refs. 3, 6), there only have been a limited number of inducible-Cre strains developed (Table 1).

Table 1.

Commonly used Cre mouse lines in renal physiology

Cre	Cell Types	Inducing Agent
AQP2CreERT2	Principal cells	Tamoxifen
KspCad-CreERT2	Most of the tubule and collecting duct	Tamoxifen
γ-GT type II	S3 segment of the proximal tubule	Tamoxifen
KAP2-iCre	Proximal tubule	Testosterone
Pax8-CreERT2	Tubule system and collecting duct	Tamoxifen
Npt2a-CreERT2	Proximal tubule	Tamoxifen
Pax8-rtTA2S*M2	Tubule system and collecting duct	Doxycycline
Pax8-rTA/LC1	Tubule system and collecting duct	Doxycycline
Umod-em1-CreERT2	Thick ascending limb	Tamoxifen

This list is not inclusive.

Despite these existing models, novel models with improved Cre features are still in high demand. In a recent issue of the American Journal of Physiology–Renal Physiology, Chen et al. (1) describe a new tamoxifen-inducible, principal cell-specific model that demonstrates no Cre leakiness and complete fidelity in cell specificity. Currently, tamoxifen-inducible Cre models utilize a Cre recombinase that is fused to a mutated ligand binding domain of the human estrogen receptor (ER), resulting in tamoxifen-dependent Cre recombinase activity (CreER^T) but no activity in the presence of estradiol. The more efficient and specific version is called CreER^T2, a second generation Cre-ER^T model. In this study, Chen et al. (1) utilized a modified tamoxifen-inducible element where Cre was fused between two ER^T2 (ER^T2CreER^T2, termed ECE) and expressed under the aquaporin-2 (AQP2) promoter, generating novel Aqp2^ECE/+ mice. This ECE knockin system proved to exhibit superior Cre characteristics with negligible background activity.

Thorough physiological analysis determined that Aqp2^ECE/+ mice do not show a difference in blood chemistries, urinary parameters, or blood pressure compared with wild-type mice. Only when having access to food and water ad libitum, Aqp2^ECE/+ mice show ~40% lower Aqp2 mRNA and protein expression compared with mice with two Aqp2 genes. After 24-h water deprivation, no differences were observed between genotypes. It remains to be determined what impact challenges in water homeostasis have on ECE considering that AQP2 expression, via its promoter, can be regulated by cAMP-responsive element-binding protein (activated by arginine-vasopressin), GATA transcription factors, and others (7). Consequently, the hydration status of the animal could potentially impact Cre recombination rates.

Since there were no other reports of ECE transgenic or knockin mice, the authors studied the expression of Cre in the kidney by breeding Aqp2^ECE/+ mice to R26R^tdTomato/+ mice, which express red fluorescent protein (RFP) in cells that have undergone Cre-mediated recombination. To test for Cre leakiness, inducibility, and fidelity, Aqp2^ECE/+RFP/+ and Aqp2^ECE/+RFP/RFP mice were treated with vehicle or 2/4 mg of tamoxifen intraperitoneally and euthanized for analysis 1 day or 2 mo after injection. The recombination rate was very high, up to 95%, even after a single dose of tamoxifen administration. Although the study showed excellent inducibility, it remains to be determined if other routes of application (e.g., oral gavage) can increase the already high recombination rate achieved by intraperitoneal injection, as described for other tamoxifen-inducible expression systems (5). In terms of fidelity, the authors used immunohistochemistry to demonstrate that Aqp2^ECE/+ mediates recombination exclusively in AQP2-positive principal cells. There was no evidence of RFP double labeling with markers for intercalated cells (H⁺-ATPase and carbonic anhydrase), whereas RFP colocalized with AQP2 and AQP3 in principal cells.

Some open questions still remain, and additional analyses would be desirable. While principal cells are clearly expressed in the connecting tubule, no evidence has been provided that Aqp2^ECE/+ is located in this part of the nephron. Possible double labeling with the Na⁺/Ca²⁺ exchanger or calbindin could provide valuable insight if Cre recombinase is present in this nephron segment. In contrast to a different model that showed expression of AQP2Cre in the male reproductive system (4), such a phenomenon was not detected in Aqp2^ECE/+ mice. Regarding the former, it was suggested to introduce AQP2Cre via the female because AQP2Cre expression in sperm could lead to recombination in fertilized oocytes. In contrast, it is not a problem to introduce Cre from male Aqp2^ECE/+ mice.

In conclusion, Aqp2^ECE/+ mice are an interesting new model for the renal scientific community. The first evidence of a principal cell-specific knockout of other proteins using this Cre driver is desperately awaited. The novel model described by Chen et al. (1) has the following advantages: 1) principal cell specific, 2) no background activity without induction, 3) high recombination rate after tamoxifen-induction, and 4) complete fidelity. It will be useful to study if genes of interest in principal cells can be used for lineage tracing or can be used to study pathological conditions like polycystic kidney disease and disorders of Na⁺, K⁺, or water homeostasis.

GRANTS

T. Rieg is supported by National Institute of Diabetes and Digestive and Kidney Diseases Grant 1R01-DK-110621).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

J.D.R. and T.R. conceived and designed research; performed experiments; analyzed data; interpreted results of experiments; prepared figures; drafted manuscript; edited and revised manuscript; and approved final version of manuscript.