Human androgen deficiency: insights gained from androgen receptor knockout mouse models

Abstract

The mechanism of androgen action is complex. Recently, significant advances have been made into our understanding of how androgens act via the androgen receptor (AR) through the use of genetically modified mouse models. A number of global and tissue-specific AR knockout (ARKO) models have been generated using the Cre-loxP system which allows tissue- and/or cell-specific deletion. These ARKO models have examined a number of sites of androgen action including the cardiovascular system, the immune and hemopoetic system, bone, muscle, adipose tissue, the prostate and the brain. This review focuses on the insights that have been gained into human androgen deficiency through the use of ARKO mouse models at each of these sites of action, and highlights the strengths and limitations of these Cre-loxP mouse models that should be considered to ensure accurate interpretation of the phenotype.

Figure 1

Signaling pathways of the androgen receptor (AR). Androgens bind to the AR and (1) regulate gene transcription via DNA binding-dependent (DBD) signaling or (2) activate second messenger pathways or transrepression through non-DBD signaling.

Figure 2

Generation of tissue/cell-specific knockout mice using the Cre-loxP system. In the Cre mouse line, the expression of Cre is under the control of a tissue/cell-specific promoter. The floxed target gene mouse line contains loxP sites (◂) flanking the region of the target gene to be deleted. When the two mouse lines are bred together, the Cre enzyme recognizes the loxP sites and deletes the intervening DNA sequence only in tissues/cells where the Cre is expressed. The target gene remains floxed and theoretically functional, in all other tissues.

Figure 3

3D micro-computed tomography (μCT) images showing that removal of the androgen receptor (AR) in male AR^ΔZF2 mice results in smaller and thinner bones of decreased density. (a) Cortical and (b) Trabecular bone.

Figure 4

Cross-sectional area of levator ani (LA) muscle from (a) Wildtype (WT), (b) Floxed androgen receptor (AR) and (c) mAR^ΔZF2 male mice. LA muscle mass is reduced by 53% in mAR^ΔZF2 males compared to WT males (P < 0.001).