Selenoprotein N deficiency in mice is associated with abnormal lung development

Abstract

Mutations in the human SEPN1 gene, encoding selenoprotein N (SepN), cause SEPN1-related myopathy (SEPN1-RM) characterized by muscle weakness, spinal rigidity, and respiratory insufficiency. As with other members of the selenoprotein family, selenoprotein N incorporates selenium in the form of selenocysteine (Sec). Most selenoproteins that have been functionally characterized are involved in oxidation-reduction (redox) reactions, with the Sec residue located at their catalytic site. To model SEPN1-RM, we generated a Sepn1-knockout (Sepn1(-/-)) mouse line. Homozygous Sepn1(-/-) mice are fertile, and their weight and lifespan are comparable to wild-type (WT) animals. Under baseline conditions, the muscle histology of Sepn1(-/-) mice remains normal, but subtle core lesions could be detected in skeletal muscle after inducing oxidative stress. Ryanodine receptor (RyR) calcium release channels showed lower sensitivity to caffeine in SepN deficient myofibers, suggesting a possible role of SepN in RyR regulation. SepN deficiency also leads to abnormal lung development characterized by enlarged alveoli, which is associated with decreased tissue elastance and increased quasi-static compliance of Sepn1(-/-) lungs. This finding raises the possibility that the respiratory syndrome observed in patients with SEPN1 mutations may have a primary pulmonary component in addition to the weakness of respiratory muscles.

Figure 1.

Confirmation of SepN deficiency by Western blot and qRT-PCR. A) Immunoblot showing the absence of SepN in the lung, diaphragm, heart, and quadriceps of Sepn1^−/− mice. GAPDH was used as loading control. B) qRT-PCR of WT and Sepn1^−/− corresponding to exon 2 of Sepn1 in lung, diaphragm, heart, and quadriceps. 18S RNA was used as internal control. Sepn1 expression in 3-wk-old WT mice was used as baseline. *P < 0.05.

Figure 2.

Quantification of oxidation in different tissues of WT and Sepn1^−/− mice. A) ROS were measured in quadriceps of six 4-mo-old mice (3 WT and 3 Sepn1^−/−), using CM-H2DCFDA as ROS indicator. B–D) Activity of glutathione peroxidase (B), glutathione reductase (C), and thioredoxin reductase (D) was measured in homogenates from quadriceps, paraspinal muscle, lung, and diaphragm of six 4-mo-old mice (3 WT and 3 Sepn1^−/−), but no significant differences were found between WT and Sepn1^−/− tissues. E) ROS were measured in quadriceps of 9-mo-old mice exposed to oxidative stress (7 WT and 7 Sepn1^−/− mice for each condition). No significant difference was found in the amount of ROS between unstressed animals and those exposed to oxidative stress.

Figure 3.

Muscle histology of WT and Sepn1^−/− mice under oxidative stress at 9 mo of age. Serial sections of the TA muscle were stained with H&E, NADH, and ATPase (pH 4.6). Images with each stain were taken from the same area of the section, allowing the comparison of the same fibers with different staining methods. Right panel corresponds to a zoomed area from the middle panel. Red arrows indicate the myofibers that contain core lesions in NADH staining. Note that the same fibers do not present any cores with ATPase staining.

Figure 4.

Quantification of running abilities. Mice (9 mo old) were fed either a control diet or a vitE-deficient diet and were subjected to a treadmill running session of 50 min for a total distance of 445 m. The last 4 min of the session was videotaped; as a measure of fatigue, the number of times each mouse touched the electric bars in the back of the treadmill belt was counted. Under both diets, the Sepn1^−/− mice touched the bars less frequently, suggesting better running ability.

Figure 5.

Quantification of size and proportion of different muscle fiber types. Immunostaining for dystrophin and myosin type 1, 2a, or 2b was performed in quadriceps muscles from 9-mo-old WT and Sepn1^−/− animals that were either unstressed (unexercised + control diet) or stressed (five 50-min treadmill sessions in 5 consecutive days + vitE deficiency since weaning). The different types of fibers were counted, and their MinFeret diameters were measured. A) Proportion of different fiber types. B) Type 2a fiber size as percentage of total number of 2a fibers. C) Type 2b fiber size as percentage of total number of 2b fibers.

Figure 6.

Contractile function of isolated muscles. A) Isolated EDL and solei muscles from 4 untreated 4-mo-old WT and Sepn1^−/− mice were stimulated at frequencies ranging from 1 to 150 Hz. The resulting force was normalized to the size of each muscle and is reported on the graph. No significant differences were observed between WT and Sepn1^−/− muscles. B) Isolated EDL and solei were stimulated for 60 s at 100 and 60 Hz respectively, to measure the fatigue in these muscles. Time to 50% drop in maximum amplitude of the contraction is shown. *P < 0.05.

Figure 7.

SR calcium release measurements in FDB myofibers. A) Measurement of calcium dynamics in FDB myofibers of WT and Sepn1^−/− mice stimulated with caffeine. B) Comparison of maximum amplitude of calcium release between WT and Sepn1^−/− myofibers, showing significantly reduced caffeine-activated calcium release in Sepn1^−/− myofibers. C) Measurement of calcium dynamics in FDB myofibers of WT and Sepn1^−/− mice stimulated with CPA. No significant difference was observed between WT and Sepn1^−/− myofibers on CPA treatment, which suggests that SR calcium stores are not altered.

Figure 8.

Lung histology. A–H) H&E staining of lung sections from 3-mo-old WT (A–D) and Sepn1^−/− (E–F) mice that were untreated (A, B, E, F) or NAC treated (C, D, G, H), viewed at ×40 (A, E, C, G),and ×100 (B, F, D, H). Scale bars = 20 μm. I) Quantification of alveolar chord length as a measure of alveolar enlargement in 1- and 3-mo-old WT and Sepn1^−/− mice.

Figure 9.

Measurement of apoptosis in WT and Sepn1^−/− lungs. A) Representative images of TUNEL staining of 3-mo-old WT and Sepn1^−/− lung sections (left panel), corresponding nuclei stained with DAPI (middle panel), and merged image (right panel). B) Quantification of TUNEL-positive nuclei in the lungs of 1 and 3-mo-old WT and Sepn1^−/− mice. C) Levels of active caspase-3 were measured in lung homogenates using immunoblotting in 3-mo-old WT and Sepn1^−/− lung homogenates, and the bands corresponding to caspase-3 were quantified and normalized to GAPDH.

Figure 10.

Quasi-static volume-pressure curves from 5 WT and 5 Sepn1^−/− mice at 6 wk of life. Note the left shift of the volume-pressure curves in the Sepn1^−/− animals, indicative of more compliant lungs in Sepn1^−/− mice.