Androgen receptor in satellite cells is not essential for muscle regenerations

Abstract

The anabolic effects of androgen on skeletal muscles are thought to be mediated by androgen receptor (AR). Although multiple studies concerning the effects of AR in males have been performed, the molecular mechanisms of AR in skeletal muscles remain unclear. Here we first confirmed that satellite cells from mouse hindlimb muscles express AR. We then generated satellite cell-specific AR knockout mice using Pax7^CreERT2 and AR^L2/Y mice to test whether AR in satellite cells is necessary for muscle regeneration. Surprisingly, we found that muscle regeneration was compromised in both Pax7^{CreERT2(Fan)/+} control mice and Pax7^{CreERT2(Fan)/+};AR^L2/Y mice compared to AR^L2/Y mice. However, Pax7^{CreERT2(Gaka)/+};AR^L2/Y;R26^tdTomato/+ mice showed no significant differences between control and mutant muscle regeneration. These findings indicate that AR in satellite cells is not essential for muscle regeneration. We propose that Pax7^{CreERT2(Fan)/+} control mice should be included in all experiments, because these mice negatively affect the muscle regeneration and show the mild regeneration phenotype.

Introduction

Regenerative capacity of adult skeletal muscles is mediated by muscle stem cells, termed satellite cells (Lepper et al., 2011; Murphy et al., 2011; Sambasivan et al., 2011), which express the transcription factor Pax7 (Seale et al., 2000). Sex hormones, including androgens, are thought to play an important role in satellite cell function during muscle development and repair after injury (Kim et al., 2016). However, the molecular mechanisms by which androgen receptor (AR), a member of the nuclear receptor superfamily, might regulate satellite cells remain poorly understood.

Although Dubois described mice having satellite cell-specific knockout of AR (Dubois et al., 2014), they used MyoD-iCre mice in which the MyoD locus was active through satellite cell development and ARs were inactivated prenatally in satellite cell progenitors and all MyoD-positive myoblasts during development. To our knowledge, there are no studies that examined whether AR plays important roles in satellite cells during muscle regeneration using satellite cells-specific knock-out mice of AR in adults.

Objective

Here we sought to characterize the role of AR in muscle satellite cells during muscle regeneration. We tested the hypothesis that AR regulates satellite cell function during activation or differentiation in mice using two types of a tamoxifen-inducible Cre loci at Pax7, Pax7^{CreERT2(Fan)/+} and Pax7^{CreERT2(Gaka)/+}, which were crossed with AR^L2/Y mice to generate mice having satellite cell-specific AR knockout.

Methods

Animals

Pax7^{CreERT2(Fan)/+}, Pax7^{CreERT2(Gaka)/+} and R26^tdTomato/+ (Madisen et al., 2009) mice were from Jackson Laboratories (Stock No: 012476, 017763 and 007908, respectively). AR^L2/Y mice were previously described (Matsumoto et al., 2003). Tamoxifen (150 μl of 20 mg/ml; Sigma, T5648) was injected intraperitoneally on 5 consecutive days. We induced muscle injury by cardiotoxin injection (100 μl of 10 μM; Latoxan, L8102) in the left tibialis anterior (TA) muscle under anesthesia. The Ehime University Animal Experiment Committee approved the animal protocols.

Immunofluorescence and microscopy

For immunofluorescence staining (protocol: http://dx.doi.org/10.17504/protocols.io.sujeeun), muscles were frozen in liquid nitrogen-chilled isopentane. BIOREVO (Keyence) and Fiji software (https://imagej.net/Fiji) were used to photograph stained tissues and quantify Pax7+ nuclei and Myh3+ muscle fibers.

Satellite cell isolation

Hindlimb muscle satellite cells were isolated as described (Liu et al., 2015).

Quantitative real-time PCR

Using a RNeasy Plus Micro Kit (Qiagen, 74034) we extracted satellite cell total RNA for synthesis of cDNA with PrimeScript (Takara, RR036A). We performed duplicate qPCR using TB Green Premix Ex Taq II (Takara, RR820S) and Thermal Cycler Dice (Takara, TP850).

Statistics

Data were analyzed using Prism 8 (GraphPad Software). Welch’s t test (for 2 groups) and Bonferroni&s multiple comparisons test (for 3 groups) was used.

Results

Nearly all Pax7+ cells in control uninjured TA muscle stained positively with AR antibodies, indicating AR expression by satellite cells (Figure 1A). In Pax7^{CreERT2(Fan)/+};AR^L2/Y mice having conditional AR deletion, AR mRNA levels in FACS-isolated satellite cells were decreased relative to control (Pax7^+/+;AR^L2/Y) mice (Figure 1B).

Figure 1. Pax7 ^{CreERT2(Fan)/+}mice show a substantial decrease in the number of Pax7+ cells during muscle regeneration. (A) AR and Pax7 immunofluorescence in uninjured TA muscle. Arrowheads indicate double positive cells. Scale bar = 20 μm. (B) RT-qPCR of AR in satellite cells from control (Pax7^+/+;AR^L2/Y) and mutant (Pax7^{CreERT2(Fan)/+};AR^L2/Y) mice (n = 4). (C) Scheme for experiments at 5 dpi. (D) Immunofluorescence of Pax7 in TA muscle at 5 dpi. Scale bar = 50 μm. (E) Quantification of the number of Pax7+ cells. (F) Scheme for experiments at 14 dpi. (G) TA muscle mass at 14 dpi. (H) Myh3 and laminin immunofluorescence in TA muscles at 14 dpi. Scale bar = 500 μm. (I) Quantification of the number of Myh3+ cells (n = 3) at 14 dpi. n.s., not significant, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. BW, body weight; Ctx, cardiotoxin; dpi, days post injury; TA, tibialis anterior; TMX, tamoxifen.

Surprisingly, upon muscle injury, the number of Pax7+ cells decreased by similar amounts in Pax7^{CreERT2(Fan)/+} control mice and Pax7^{CreERT2(Fan)/+};AR^L2/Y mice at 5 dpi (Figure 1C-E), while Pax7^{CreERT2(Fan)/+} control mice showed slightly delayed muscle regeneration at 14 dpi (Figure 1F-H).

Pax7^{CreERT2(Gaka)/+};AR^L2/Y;R26^tdTomato/+ mice having different Cre-mediated AR deletions in satellite cells (Figure 2A) had similar numbers of Pax7+ cells at 5 dpi as the control (Figure 2B-D). Muscle regeneration in terms of Myh3+ fiber numbers was similar for the control and mutant mice at 14 dpi (Figure 2E-G). Together, these results suggest that AR expressed by satellite cells is not involved in muscle regeneration.

Figure 2. Androgen receptor is not required for satellite cells to regenerate muscle. (A) RT-qPCR of AR in satellite cells from control (Pax7^{CreERT2(Gaka)/+};AR^+/Y;R26^tdTomato/+) and mutant (Pax7^{CreERT2(Gaka)/+};AR^L2/Y;R26^tdTomato/+) mice (n = 4). (B) Scheme for experiments at 5 dpi. (C) Immunofluorescence of Pax7 and laminin in TA at 5 dpi. Scale bar = 50 μm. (D) Quantification of the number of Pax7+ cells (n = 4). (E) Scheme for experiments at 14 dpi. (F) TA muscle mass at 14 dpi (n = 4 for control, n = 5 for mutant). (G) Myh3 and laminin immunofluorescence in TA muscles at 14 dpi. Scale bar = 500 μm. (H) Quantification of the number of Myh3+ cells (n = 3) at 14 dpi. n.s., not significant, *p < 0.05, BW, body weight; Ctx, cardiotoxin; dpi, days post injury; TA, tibialis anterior; TMX, tamoxifen.

Discussion

The Pax7^{CreERT2(Fan)/+} regeneration phenotype is consistent with previous findings (Mademtzoglou et al., 2018) and likely related to expression of only one Pax7 allele after CreERT2 insertion (Lepper et al., 2009). Leaky Cre recombinase activation and downregulation of Hey1, which is required to maintain satellite cells, in Pax7^{CreERT2(Fan)/+} mice not treated with tamoxifen (Noguchi et al., 2019) could affect satellite cells during muscle regeneration.

The muscle regeneration activity seen in satellite cell-specific AR knockout mice despite effective AR deletion suggests that AR does not function in muscle regeneration. However, a phenotype that was not detectable in our assays might be present.

Conclusions

Our muscle regeneration experiments using Pax7^{CreERT2(Fan)/+} mice suggest that these Cre mice, especially gene Knockout/Knockin mice, must be used carefully and in combination with other Cre mice to avoid the problem of “pseudo phenotypes” generated by the Cre cassette or Cre insertion at the target locus. Our data for Pax7^{CreERT2(Gaka)/+};AR^L2/Y;R26^tdTomato/+ mice demonstrated that AR expression by satellite cells is not necessary for muscle regeneration in adults.