Endocrinology and Metabolism

Difference in skeletal muscle function in males vs. females: role of estrogen receptor-β

Birgitta Glenmark, Maria Nilsson, Hui Gao, Jan-Åke Gustafsson, Karin Dahlman-Wright, Håkan Westerblad


Male skeletal muscles are generally faster and have higher maximum power output than female muscles. Conversely, during repeated contractions, female muscles are generally more fatigue resistant and recover faster. We studied the role of estrogen receptor-β (ERβ) in this gender difference by comparing contractile function of soleus (mainly slow-twitch) and extensor digitorum longus (fast-twitch) muscles isolated from ERβ-deficient (ERβ−/−) and wild-type mice of both sexes. Results showed generally shorter contraction and relaxation times in male compared with female muscles, and ERβ deficiency had no effect on this. Fatigue (induced by repeated tetanic contractions) and recovery of female muscles were not affected by ERβ deficiency. However, male ERβ−/− muscles were slightly more fatigue resistant and produced higher forces during the recovery period than wild-type male muscles. In fact, female muscles and male ERβ−/− muscles displayed markedly better recovery than male wild-type muscles. Gene screening of male soleus muscles showed 25 genes that were differently expressed in ERβ−/− and wild-type mice. Five of these genes were selected for further analysis: muscle ankyrin repeat protein-2, muscle LIM protein, calsequestrin, parvalbumin, and aquaporin-1. Expression of these genes showed a similar general pattern: increased expression in male and decreased expression in female ERβ−/− muscles. In conclusion, ERβ deficiency results in increased performance during fatigue and recovery of male muscles, whereas female muscles are not affected. Improved contractile performance of male ERβ−/− mouse muscles was associated with increased expression of mRNAs encoding important muscle proteins.

  • skeletal muscle fatigue
  • gender differences
  • muscle LIM protein
  • muscle ankyrin repeat protein

there are major differences between female and male skeletal muscles, including differences in energy metabolism, fiber type composition, and contractile speed (7, 11, 19, 23, 33). Generally, male muscles have a higher capacity for anaerobic metabolism and generate a higher maximum power output than female muscles. There are also important differences between female and male muscles during prolonged intense activity leading to fatigue, where female muscles have been found to be more fatigue resistant and to recover faster than male muscles (9, 14, 24). The mechanisms behind these sex-related differences in skeletal muscle are not known, but it appears likely that they are a consequence of different sex hormonal status.

The effects of estrogens on skeletal muscle function are unclear (6). Some studies suggest that estrogens increase skeletal muscle force production (15, 34), and variations in voluntary muscle strength have been observed during the human menstrual cycle (30, 31). Conversely, other studies did not observe any change in muscle function in response either to increased estrogen levels (12) or to fluctuations during the menstrual cycle (17). Interestingly, female rat muscles show fewer histopathological changes after repeated eccentric contractions than male muscles (20). Moreover, compared with normal females, male and ovariectomized female rats exhibit higher indexes of exercise-induced muscle membrane damage and increased stress protein expression, and this difference disappears after estradiol treatment (2, 29).

The effects of estrogen are mediated by ligand-activated transcription factors, called the estrogen receptors (ERs) (25). There are two known ERs, ERα and the more recently described ERβ (22). The exact physiological responses of each receptor are still incompletely understood, but studies on ER knockout mice show that ERα is primarily involved in the classical actions of estrogens, i.e., sexual differentiation, fertility, and lactation (21). ERβ has been shown to play an important role in the central nervous system, the immune system, and the prostate (13, 38). Both ERα and ERβ are expressed in human skeletal muscle at the mRNA level, whereas only ERβ could be detected at the protein level (41).

We hypothesized that some of the functional differences between female and male skeletal muscle are mediated via ERs. In this study, we specifically investigated the role of ERβ by comparing the contractile function of skeletal muscles isolated from mice in which the ERβ gene is knocked out (ERβ−/− mice) and wild-type mice (21). The most conspicuous result was a markedly slower recovery from fatigue in wild-type male compared with wild-type female muscles, and this difference was not seen in ERβ−/− muscles. Gene screening followed by quantitative real-time PCR (RT-PCR) analyses showed several changes in gene transcription that may be associated with the differences in contractile function between male ERβ−/− and wild-type muscles.


Experimental animals.

Experiments were performed on 5- to 6-mo-old ERβ−/− mice and wild-type littermates (21). The mice had free access to water and standard food pellets. Animals were killed by cervical dislocation. All experimental procedures were approved by the Stockholm North Animal Ethics Committee.

Contractile experiments.

Intact extensor digitorum longus (EDL) and soleus muscles were dissected from wild-type and ERβ−/− mice. These muscles were chosen because EDL is a fast muscle (contains mainly fast-twitch, type IIB fibers), whereas soleus is slow (contains about equal amounts of slow-twitch, type I fibers, and fast-twitch, type IIA fibers) (44). Muscles were mounted in a stimulation chamber filled with continuously stirred Tyrode solution of the following composition (in mM): 121 NaCl, 5 KCl, 0.5 MgCl2, 1.8 CaCl2, 0.4 NaH2PO4, 0.1 Na-EDTA, 24 NaHCO3, and 5.5 glucose,. Fetal calf serum (0.2%) was added to the solution. The solution was continuously bubbled with 95% O2-5% CO2, which gives a bath pH of 7.4. All experiments were carried out at room temperature (∼25°C).

Muscles were mounted between a fixed hook and a lab-built force transducer. Muscles were stimulated with supramaximal electrical pulses (duration 0.5 ms; intensity ∼150% of that giving maximum contractile response). The stimulation pulses were applied via two platinum plate electrodes placed on each side of the muscle and extending the whole length of the muscle. The resulting force was recorded and digitized (1 kHz; Axotape, Axon Instruments) and stored in a personal computer. A few tetanic contractions were produced to find the length that gave the maximum tetanic force response, and muscles were then allowed to rest for at least 30 min before measurements were conducted.


EDL muscles were stimulated to give a single twitch or 300-ms tetani at 20–120 Hz; soleus muscles were stimulated to give a single twitch or 1-s tetani at 10–100 Hz. These contractions were produced at 1-min intervals. Peak force in each contraction was measured. Twitch kinetics were assessed by measuring the contraction time (i.e., from the onset of force production until peak force was produced) and the half-relaxation time (i.e., from peak force production until force was reduced to 50% of the peak). Kinetics were also assessed in 100- and 70-Hz tetani in EDL and soleus, respectively, by measuring the half-contraction time (i.e., from the onset of contraction until 50% of the maximum tetanic force was produced) and the half-relaxation time (i.e., from the last stimulation pulse until force was reduced to 50% of the maximum).

After the force-frequency relationship had been established, the muscle was fatigued by tetani given at 2-s intervals. For EDL muscles we used 70-Hz, 300-ms tetani, and the total number of contractions was 50. Soleus muscles were fatigued by 50-Hz, 600-ms tetani, and a total of 100 tetani were produced. The recovery of force was studied by giving a single tetanus at regular intervals for 30 min after the end of fatiguing stimulation. Some soleus muscles were then frozen for Affymetrix microarray analysis.

Affymetrix microarray analysis and bioinformatics.

Soleus muscles from three ERβ−/− and three wild-type male mice were individually analyzed in these experiments. Total RNA was isolated from muscles by TRIzol reagent (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions and further purified with the use of the RNeasy Mini Kit (Qiagen, Valencia, CA). The RNA was reverse transcribed into cDNA, in vitro transcribed into labeled cRNA, and hybridized to the mouse MG-U74A GeneChips (Affymetrix, Santa Clara, CA) according to the Affymetrix GeneChip expression analysis manual. The scanned output files were analyzed using the Affymetrix software (Micro Array Suite 5.0; MAS 5.0). The gene chips were globally scaled to an average intensity of 500 U to allow gene expression comparisons between samples. Each of the three ERβ−/− samples was compared with each of the three wild-type samples, resulting in nine pair-wise comparisons. The results of these comparisons were indicated as increased call, decreased call, or no change call according to the Affymetrix software (MAS 5.0 Expression Analysis Program). The expression of a gene was considered changed when at least six of the nine comparisons produced concordant calls (increased call or decreased call).

The significance analysis of microarrays (SAM) method (36) was used as a statistical approach to analyze the data, focusing on changed genes identified by MAS 5.0. SAM calculates a score for each gene on the basis of the change in expression relative to the standard deviation of all measurements. For genes with scores greater than an adjustable threshold, SAM uses permutations of the repeated measurements to estimate the percentage of genes identified by chance, i.e., the false discovery rate. The q value for each gene represents the probability that it is falsely called significantly changed.

Gene Expression Omnibus (GEO) accession numbers for the Affymetrix data are GSE1050 and GSM16989-16994.

Quantification of mRNA.

Total RNA was isolated from soleus and EDL muscles as described above. In these experiments, we used rested muscles obtained from 7 wild-type female, 8 ERβ−/− female, 10 wild-type male, and 7 ERβ−/− male mice. A total of 1 μg of RNA from soleus and EDL muscles, respectively, was treated with Dnase I (amplification grade, Invitrogen) before reverse transcription into cDNA by SuperScript II (Invitrogen), using random hexamer priming according to the manufacturer's protocol. mRNA expression was quantified by SYBR green RT-PCR, using 18S as an internal control, on an ABI 7700 machine (Applied Biosystems, Foster City, CA). PCR products were further analyzed by melting curve analysis to confirm a single product. The “standard curve method” (User Bulletin no. 2, Applied Biosystems) was used for data analysis. Oligonucleotide primer sequences and amplicon sizes are presented in Table 1. All primers span intron-exon boundaries. The efficiency of the PCR reaction was 95–100% for all primer pairs, and the standard curve method compensates for efficiency differences between assays. Experiments performed in duplicates and repeated on another batch of cDNA synthesized from the same original RNA gave similar results.

View this table:
Table 1.

Oligonucleotides used for RT-PCR experiments

The expression of the 18S gene was not significantly different between wild-type and ERβ−/− muscles, and the coefficient of variation for 18S expression ranged between 2 and 5%. We also used GAPDH gene expression as an internal control, and this had no significant effect on the PCR results (data not shown).


Data are presented as means ± SE. Differences between groups were assessed by Student's unpaired t-test, by one-way analysis of variance (ANOVA), or by two-way repeated-measures ANOVA. Tukey's post hoc test was used when ANOVA showed significant differences. Differences were regarded as significant if P < 0.05 was attained.


Basic contractile properties.

The force produced in twitches and tetanic contractions was not markedly different between groups in either soleus muscles (Table 2) or EDL muscles (Table 3). The contraction and half-relaxation times of twitches and tetani were generally shorter in male than in female muscles, although the difference was not always significant. This was also manifested as a tendency of a rightward shift of the force-frequency relationship (i.e., higher stimulation frequency required to produce 50% of the maximum force) in male compared with female muscles. ERβ deficiency did not significantly affect the contraction or relaxation speed or the force-frequency relationship in either soleus or EDL muscles. These data agree with human studies showing faster contractions in male compared with female muscles (7, 8, 10, 26). Furthermore, we show that ERβ deficiency does not lead to any major change in basic contractile properties, and therefore some other factor(s) would underlie the gender difference in contractile speed.

View this table:
Table 2.

Basal contractile function in WT and ERβ−/− soleus muscles

View this table:
Table 3.

Basal contractile function in WT and ERβ−/− EDL muscles

Fatiguing stimulation and recovery.

During fatiguing stimulation, force tended to decrease more in ERβ−/− than in wild-type female muscles, although the difference was never statistically significant (Fig. 1, A and C). The opposite was found in male muscles; that is, force decreased less in ERβ−/− than in wild-type muscles, and the difference was significant early during fatigue in soleus muscles (Fig. 1B) and during the latter part of fatiguing stimulation in EDL muscles (Fig. 1D).

Fig. 1.

Male estrogen receptor-β-deficient (ERβ−/−) muscles produce somewhat higher forces during fatigue than male wild-type (WT) muscles. Relative force production during fatigue was induced by repeated tetanic stimulation. Data were obtained from female (A) and male (B) soleus muscles and female (C) and male (D) extensor digitorum longus (EDL) muscles. Solid symbols, WT; open symbols, ERβ−/−. Data represent means ± SE; n = 5 in each group. Dashed lines in B and D show mean values of female muscles. Significant difference, WT vs. ERβ−/−: *P < 0.05 and **P < 0.01.

A similar pattern was observed during the recovery period after fatigue. Thus there was a slight tendency of lower forces in female ERβ−/− muscles than in wild-type muscles (Fig. 2, A and C), whereas force was markedly higher during recovery in male ERβ−/− muscles than in wild-type muscles (Fig. 2, B and D). These data fit with human studies, where female muscles were found to recover better from fatigue than male muscles (9, 14). Our results show that the recovery was markedly better in male ERβ−/− muscles compared with male wild-type muscles. In fact, recovery in the former was very similar to that in female muscles (dashed lines in Fig. 2, B and D). This indicates that the inferior recovery in male wild-type muscles is caused by some factor(s) mediated via ERβ.

Fig. 2.

Male ERβ−/− and female muscles produce higher forces during recovery after fatiguing stimulation than male WT muscles. Force is presented as percentage of that produced in the first fatiguing tetanus in each fiber. Data were obtained from female (A) and male (B) soleus muscles and female (C) and male (D) EDL muscles. Solid symbols, WT; open symbols, ERβ−/−. Data are means ± SE; n = 5 in each group. Note that male ERβ−/− muscles are very similar to female muscles (dashed lines represent mean values of female muscles). Significant difference, WT vs. ERβ−/−: *P < 0.05 and **P < 0.01.

Affymetrix gene expression analysis.

We hypothesized that the improved recovery in male ERβ−/− muscles compared with wild-type muscles reflects differences in gene regulation. To test this hypothesis, we studied three male soleus muscles from ERβ−/− and wild-type mice, respectively, using the Affymetrix platform. Table 4 shows genes where the expression was either increased or decreased in ERβ−/− compared with wild-type muscles. In total, 25 genes were differently expressed between ERβ−/− and wild-type muscles. The largest changes were observed for myosin heavy chain 2B (∼5-fold) and α-actinin-3 (∼2-fold). However, these genes are expressed at very low levels in slow-twitch soleus muscles (27, 44); hence, changes in protein expression are unlikely to be of major functional importance. Moreover, a major difference in the expression of myosin heavy-chain isoforms between ERβ−/− and wild-type muscles would result in altered contractile speed, and this was not observed (see Table 2).

View this table:
Table 4.

Genes differently expressed in male ERβ−/− and WT soleus muscles

Quantification of selected mRNAs in soleus muscles.

Recovery after fatigue induced by repeated tetanic contractions is frequently a slow process because of a prolonged impairment in sarcoplasmic reticulum (SR) Ca2+ release (40). The mechanisms underlying this fatigue-induced defect in SR Ca2+ release are not fully understood but seem to involve structural changes in the triadic region (39). Having this in mind, we used quantitative RT-PCR to further analyze two genes encoding structural proteins where altered gene expression was detected in the microarray screening: muscle ankyrin repeat protein-2 (MARP2) and muscle LIM protein (MLP). MARP2 (also known as Arpp and Ankrd2) is highly expressed in skeletal muscle, preferentially in slow-twitch fibers (18, 35). In skeletal muscle, MARP2 is localized in the sarcomeric I band and in the nucleus. The functional role of MARP2 is unclear, but the expression of the gene is upregulated by mechanical stretch, indicating that it is involved in muscle hypertrophy (18). MLP is constitutively expressed in cardiac and slow-twitch skeletal muscle, and expression can be induced in fast-twitch muscle by chronic low-frequency stimulation or mechanical overload (32, 42). MLP is located close to the Z disk and has been suggested to act as a scaffold protein promoting the interaction of proteins along the actin-based cytoskeleton. Figure 3 shows a significantly higher expression of both MARP2 and MLP in male ERβ−/− soleus muscles compared with wild-type muscles. RT-PCR analyses of MARP2 and MLP were also performed on the RNA isolates used in the Affymetrix microarray screening. MARP2 expression then showed a 1.4-fold increase, which is very similar to the ∼1.5-fold increases seen in the Affymetrix analysis and in the RT-PCR analysis on the larger number of muscles (see Fig. 3). Corresponding values for MLP expression were ∼2-, 1.9-, and 1.6-fold increases. Thus Affymetrix microarray screening and RT-PCR gave similar results, and there were no major differences in the expression of these genes in the muscles used for Affymetrix and for RT-PCR. In female soleus muscles, MARP2 expression was lower in ERβ−/− than in wild type, whereas MLP expression was similar in the two groups.

Fig. 3.

RT-PCR analyses of soleus muscles. Values are expressed relative to the mean values in WT female muscles, which were set to 100%. Open bars, WT; hatched bars, ERβ−/−. Muscles were obtained from female mice (A; WT, n = 7; ERβ−/−, n = 8) and male mice (B; WT, n = 10; ERβ−/−, n = 7). MARP2, muscle ankyrin repeat protein-2; MLP, muscle LIM protein; CSQ, calsequestrin; Parv, parvalbumin; AQP-1, aquaporin-1. Significant difference, WT vs. ERβ−/−: *P < 0.05 and **P < 0.01.

Altered expression of two genes encoding Ca2+-handling muscle proteins was detected in the microarray screening: calsequestrin and parvalbumin. Calsequestrin is a high-capacity Ca2+-binding protein located in the SR terminal cisternae (43). Parvalbumin acts as a soluble myoplasmic Ca2+ buffer (5). In mammals, parvalbumin is mainly expressed in fast-twitch muscles of small-sized animals (16). Expression of these two genes was decreased in female ERβ−/− muscles (not significant for calsequestrin) and increased in male ERβ−/− muscles.

Parvalbumin gene expression in male ERβ−/− muscles showed a decrease in the Affymetrix microarray screening, whereas it showed an increase with RT-PCR. This discrepancy can be explained by the fact that the Affymetrix probe pairs for parvalbumin are not specific for this particular gene. It should be noted that the Affymetrix probe pairs for all other genes we selected for RT-PCR analysis are specific for the particular gene. An alternative explanation for the conflicting results regarding parvalbumin gene expression is that Affymetrix microarray screening was performed on muscles that had been fatigued and then allowed to recover for 30 min, whereas RT-PCR was performed on rested muscles. Nevertheless, slow-twitch soleus muscles contain very little parvalbumin (4), and minor changes in expression of this gene would have a limited effect on the contractile function. It is worth noting that the other four genes selected for RT-PCR showed changes of gene expression in the same direction as in the Affymetrix microarray screening, and for MARP2 and MLP expression, there was no marked difference between the two muscle groups (see above). This indicates that the period of fatiguing stimulation had little impact on gene expression. In line with this, a recent study on mouse muscles showed little change in gene expression with isometric contractions (like in the present study), whereas eccentric contraction induced marked changes in the expression of numerous genes (3).

The water channel protein aquaporin-1 is strongly expressed in skeletal muscle (37). In cardiac myocytes, aquaporin-1 is colocalized with calveolin-3 and has been suggested to be involved in caveolae movements (28). During fatigue, there is an intracellular accumulation of metabolic by-products causing increased osmotic pressure, which will result in water fluxes. Thus changes in aquaporins may be of importance during fatigue and the subsequent recovery of fatigue. Our results showed a decreased aquaporin-1 expression in female ERβ−/− and an increased expression in male ERβ−/− muscles.

To sum up, ERβ deficiency has a significant effect on the expression of most genes studied (see Fig. 3), and the changes showed a similar picture: a decrease in female ERβ−/− and an increase in male ERβ−/− soleus muscles. The changes were rather modest, up to ∼2.5-fold. A candidate gene to explain the contractile results (little change in female muscles and markedly improved recovery from fatigue in male muscles) should display little change in female muscles, whereas it should be markedly upregulated (or downregulated) in male ERβ−/− muscles. This criterion was best met by MLP, since it showed no tendency of change in female muscles and ∼60% increase in male ERβ−/− muscles. MLP-deficient mice develop dilated myopathy with major cytoarchitectural disorganization, leading to heart failure (1). These mice also display symptoms from the skeletal muscle system with decreased performance in endurance tests (1). Slow-twitch muscles generally recover better than fast-twitch muscles after being fatigued to about the same extent (see Fig. 2), and MLP is constitutively expressed only in slow-twitch skeletal muscle. The expression can be induced by prolonged electrical stimulation in fast-twitch muscle, indicating that MLP is involved in fast-to-slow transition of adult muscle cells (32, 42). Interestingly, a recent report shows an upregulation of MLP (and MARP2) after a single bout of eccentric contractions in the mouse (3). Taken together, this indicates that MLP is important for maintaining the cellular integrity during periods of intense contractile activity or mechanical stress. Thus the increased expression of MLP in male ERβ−/− soleus muscles may contribute to their improved recovery.

Quantification of mRNAs in EDL muscles.

EDL muscles were analyzed for expression of the same genes as soleus muscles (Fig. 4). Although the trend was similar to that in soleus muscles (downregulation in female ERβ−/− and upregulation in male ERβ−/− muscles), changes were generally smaller in EDL muscles, and the only significant changes were seen in females. Differences in contractile properties between ERβ−/− and wild-type muscles were similar in EDL and soleus muscles (see Figs. 1 and 2). Thus the increased force production during the end of fatiguing stimulation and especially during recovery in male ERβ−/− EDL muscles cannot be explained by changes in the five genes analyzed, because the expression was not different between ERβ−/− and wild-type muscles. Thus the genes involved in the improved performance of male ERβ−/− muscles would differ between fast-twitch and slow-twitch muscles. In line with this, the favored candidate gene in soleus muscles was MLP, which is little expressed in normal EDL muscles (42). It is worth noting that calsequestrin and parvalbumin were significantly downregulated in female ERβ−/− EDL muscles, but this was not reflected in any significant change in contractile function.

Fig. 4.

RT-PCR analyses of female (A) and male (B) EDL muscles. Values are expressed relative to the mean values in WT female muscles. Open bars, WT; hatched bars, ERβ−/−. Muscles were obtained from the same mice as in Fig. 3. Significant difference, WT vs. ERβ−/−: * P < 0.05 and ** P < 0.01.

In conclusion, male muscles are generally faster and have a higher maximum power output than female muscles. On the other hand, female muscles are generally more fatigue resistant and recover faster and show less mechanical damage after exercise. Intriguingly, skeletal muscles of ERβ-deficient males combine the advantages of male and female muscles; that is, they perform better during fatigue and recovery but are still as fast. Thus treatment with ERβ antagonists might improve muscle performance in severe muscle disorders affecting males. However, ERβ exerts important functions in, for instance, the central nervous system and the prostate gland (13, 38), which would limit the usefulness of ERβ antagonists.


This study was supported by grants from the Swedish Research Council (project nos. 10842 and 14453), the Swedish Cancer Fund, KaroBio, the Swedish National Center for Sports Research, the Gamla Tjänarinnor Foundation, and Funds at the Karolinska Institutet. We are grateful to the Knut and Alice Wallenberg Foundation for supporting the Affymetrix core facility, NOVUM.


  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


View Abstract