KLHL40 deficiency destabilizes thin filament proteins and promotes nemaline myopathy

Nemaline myopathy (NM) is a congenital myopathy that can result in lethal muscle dysfunction and is thought to be a disease of the sarcomere thin filament. Recently, several proteins of unknown function have been implicated in NM, but the mechanistic basis of their contribution to disease remains unresolved. Here, we demonstrated that loss of a muscle-specific protein, kelch-like family member 40 (KLHL40), results in a nemaline-like myopathy in mice that closely phenocopies muscle abnormalities observed in KLHL40-deficient patients. We determined that KLHL40 localizes to the sarcomere I band and A band and binds to nebulin (NEB), a protein frequently implicated in NM, as well as a putative thin filament protein, leiomodin 3 (LMOD3). KLHL40 belongs to the BTB-BACK-kelch (BBK) family of proteins, some of which have been shown to promote degradation of their substrates. In contrast, we found that KLHL40 promotes stability of NEB and LMOD3 and blocks LMOD3 ubiquitination. Accordingly, NEB and LMOD3 were reduced in skeletal muscle of both Klhl40^–/– mice and KLHL40-deficient patients. Loss of sarcomere thin filament proteins is a frequent cause of NM; therefore, our data that KLHL40 stabilizes NEB and LMOD3 provide a potential basis for the development of NM in KLHL40-deficient patients.

Nemaline myopathies (NMs) encompass a set of genetically heterogeneous diseases that are defined pathologically by the presence of characteristic rod-like structures, called nemaline bodies, in myofibers (1). Clinically, presentation of NM can vary substantially from mild muscle dysfunction to complete akinesia, but without any correlation to the number of nemaline bodies, suggesting that these structures are not a primary cause of the muscle dysfunction (2). The underlying pathogenesis of NM remains unresolved; however, 6 of 9 causative genes for NM, tropomyosin 3 (TPM3), nebulin (NEB), α 1 actin (ACTA1), tropomyosin 2 (TPM2), troponin T type 1 (TNNT1), and cofilin 2 (CFL2), encode components of the sarcomere thin filament, leading to the current hypothesis that NM is a thin filament disease (3–8).

The sarcomere thin filament is a helical strand of polymerized actin that anchors directly to the Z line and spans toward the M line (9). During muscle contraction, thick filament myosin forms crossbridges with thin filament actin and physically pulls the Z line toward the M line to generate force. Tropomyosin and the troponin complex regulate myosin-actin crossbridges in a calcium-dependent manner. In NM, it is believed that mutations disrupting thin filament proteins result in reduced force generation and subsequent myopathy (10). However, mutations in 3 proteins with no established association with the sarcomere thin filament, kelch repeat and BTB (POZ) domain containing 13 (KBTBD13), kelch-like family member 40 (KLHL40), and KLHL41, have been shown recently to cause NM in patients by an unknown mechanism (11–13). Notably, KLHL40 mutations occur frequently in severe forms of autosomal recessive NM (12).

KBTBD13, KLHL40, and KLHL41 belong to the BTB-BACK-kelch (BBK) protein family, named for the presence of BTB, BACK, and kelch repeat (KR) domains in all family members (KBTBD13 lacks an annotated BACK domain but is still very closely related to other BBK family members) (14). BBK proteins are thought to serve as substrate-specific adaptors for the cullin 3 (CUL3) E3 ligase complex, mediating the ubiquitination and, in most cases, degradation of their respective substrates (15–22). Because NM is so tightly associated with thin filament proteins, it is likely that these 3 BBK proteins also regulate the thin filament in some manner, though this has not been demonstrated. Additionally, since other BBK proteins act through the ubiquitin-proteasome system, defining the function of these proteins may facilitate therapeutic manipulation, given the availability of proteasome inhibitors for patient use.

While searching for new potential regulators of muscle function, we uncovered a novel muscle-specific transcript, Klhl40, that encodes for a protein localizing to the sarcomere I band and A band. Deletion of the Klhl40 locus in mice results in sarcomere defects, with subsequent muscle dysfunction and early postnatal lethality, which closely phenocopy the muscle abnormalities observed in humans with KLHL40 mutations (12). We show that KLHL40 binds to NEB, a thin filament protein frequently associated with NM (1), as well as leiomodin 3 (LMOD3), a novel muscle protein highly homologous to leiomodin 2 (LMOD2), which regulates actin at the thin filaments (23, 24). Similar to LMOD2, we found that LMOD3 also localizes to the sarcomere thin filaments. Unlike most other BBK proteins that mediate degradation of their substrates, KLHL40 uniquely promotes stability of NEB and LMOD3 and inhibits ubiquitination of LMOD3. Accordingly, loss of KLHL40 reduces NEB and LMOD3 protein in mice as well as in patients with NM. By establishing a novel pro-stability function of KLHL40 for the thin filament proteins NEB and LMOD3, our data reveal a molecular basis for NM in patients deficient for KLHL40 and suggest a previously unrecognized role of LMOD3 in maintenance of sarcomere function and NM.

Discovery and analysis of striated muscle-specific expression of Klhl40. To find new regulators of striated muscle biology, we searched an in situ hybridization database for novel transcripts that were enriched in developing muscle structures (25). Within this database, we saw that Klhl40 appeared to have muscle-specific expression, which we found was dependent on the MEF2 transcription factor (26). We validated this muscle-specific expression in embryos and found it preserved into adulthood, with additional low level expression of Klhl40 in the heart (Figure 1, A and B).

Figure 1

Klhl40 is expressed specifically in skeletal muscle and heart, and deletion of Klhl40 causes neonatal lethality. (A) Sagittal sections of E15 mouse embryos were probed for Klhl40 mRNA using antisense radioisotopic probes. Signal for Klhl40 (pseudocolored red) only appears in developing muscle. Black arrows point to representative developing intercostal and back muscles from rostral to caudal, respectively. The black arrowhead denotes heart. Scale bar: 1 mm. (B) Northern blot of adult mouse tissues for Klhl40 and Gapdh (loading control) shows Klhl40 expression in skeletal muscle and heart. Numbers on right indicate size of transcripts in kilobases. SK Muscle, skeletal muscle. (C) qPCR analysis of Klhl40 transcript in P8 quadriceps in mice of the indicated genotypes. Values are normalized to 18S ribosomal RNA (n = 3 for both genotypes). ^#P < 0.05. (D) Representative image of surviving P8 KO pups with WT littermates. (E) Survival curve of WT versus KO mice shows early neonatal lethality in KO mice (WT, n = 21, and KO, n = 17).

To assess the potential role of Klhl40 in muscle function, we generated a loss-of-function Klhl40 allele by replacing the protein coding region of Klhl40 with a lacZ expression cassette controlled by the endogenous Klhl40 promoter (ref. 27 and Supplemental Figure 1; supplemental material available online with this article; doi:10.1172/JCI74994DS1). We verified, by lacZ staining in Klhl40^+/– mice, that Klhl40 was expressed exclusively in skeletal muscle in embryonic, neonatal, and adult mice (Supplemental Figure 2, A–C). Histological analysis revealed that lacZ was only expressed in myofibers and not connective tissues (Supplemental Figure 2D). Klhl40^+/– hearts showed punctate lacZ staining consistent with lower Klhl40 heart expression (Supplemental Figure 3).

Loss of Klhl40 results in neonatal lethality. Klhl40^–/– (KO) mice had no detectable Klhl40 transcript and were smaller than WT littermates, which was apparent within several days of birth (Figure 1, C and D, and Supplemental Figure 4). Furthermore, 50% of the KO mice died by 7 days of age, and none of these KO mice bred on a pure C57BL/6 background survived past 3 weeks of age (Figure 1E). Impaired growth was not due to competition by healthier littermates, since KO mice alone with a nursing mother also failed to survive.

To confirm that the lethal phenotype was due to loss of Klhl40 in muscle, we generated transgenic mice expressing Klhl40 under the control of the muscle-specific, muscle creatine kinase (MCK) promoter and bred the allele into the KO background (28, 29). WT and KO mice with the MCK-Klhl40 transgene are referred to as “WT Tg” and “KO Tg,” respectively. We confirmed overexpression of Klhl40 in skeletal muscle of both WT Tg and KO Tg mice (Supplemental Figure 5A). The MCK-Klhl40 transgene was sufficient to completely rescue the KO phenotype, such that KO Tg mice were indistinguishable from control littermates, with normal growth into adulthood (Supplemental Figure 5, B and C), confirming that loss of Klhl40 from striated muscle was responsible for the neonatal lethal phenotype.

Klhl40 localizes to the sarcomere I band and A band. To better understand how loss of KLHL40 from muscle causes lethality, we analyzed the subcellular localization of KLHL40 in mature myofibers. Two previous reports on the localization of KLHL40 reported conflicting data, with one study showing A band localization of KLHL40 while another showed localization to the I band using a commercial antibody on fixed tissue sections (12, 13). To determine KLHL40 localization by an alternative method, we electroporated the flexor digitorum brevis of adult mice with KLHL40 C-terminally fused with EGFP (KLHL40-EGFP). We used second harmonic generation, as previously described, to localize KLHL40-EGFP relative to myosin (30). In a modified EDTA-Ringer’s solution that relaxes myofibers (31), KLHL40-EGFP signal concentrated at the I band (Figure 2A). However, muscle bathed in a rigor solution that induces a sustained muscle contraction showed localization of KLHL40-EGFP to the A band, with fainter localization to the I band (Figure 2B and ref. 31). α-Actinin (ACTN1A) and tropomodulin (TMOD4), which are known I band and A band proteins, respectively, did not change localization in relaxed and contracted muscles (Supplemental Figure 6, A–D, and ref. 32). These data suggested that KLHL40 can localize to either the sarcomere I band or A band and indicated to us a possible role for KLHL40 in the sarcomere.

Figure 2

KLHL40 localizes to sarcomere I band and A band. KLHL40-EGFP localization relative to myosin in (A) relaxed muscles and (B) contracted muscles shows that KLHL40 localizes to the I band and A band. Graphs are representative line scans of indicated signals calculated over 8 μm of distance along the length of the myofiber. Scale bar: 5 μm.

Klhl40-deficient skeletal muscle has abnormal structure and function. To determine whether loss of KLHL40 had any affect on sarcomere structure and function, we performed histologic analysis of skeletal muscles from 1-week-old neonatal KO mice and found that sarcomeres were present, but a subset of fibers showed complete disorganization, with widened Z-discs that resembled nemaline bodies present in NM (Figure 3, A and B, and ref. 1). Additionally, other myofibers in KO muscles had irregular, wave-like structures, referred to as Z line streaming (Supplemental Figure 7A). Notably, KO muscle was relatively comparable to WT muscle at 1 day of age (Supplemental Figure 7B).

Figure 3

KO mice have disrupted muscle structure and function. (A) Longitudinal sections of P8 diaphragm muscles from WT and KO mice stained with desmin (red), DAPI (blue), and wheat germ agglutinin (green). The white arrow indicates disrupted myofibers. The white arrowhead denotes fibers staining abnormally for desmin that are more frequent in, but not specific to, KO muscles. Insets show normal striated pattern of Z lines in WT muscles but that some KO fibers have complete loss of sarcomere organization. Scale bar: 20 μm. (B) Electron microscopy analysis of longitudinal sections of P8 diaphragms from WT and KO mice shows some highly disorganized fibers. The red “Z” with red arrowhead indicates representative Z line in WT section. The yellow arrowhead denotes representative nemaline-like body. Scale bar: 1 μm. (C) Maximum contractile force of P1 WT, Klhl40^+/– (HET), and KO hind limb following 150-Hz stimulation, normalized to limb mass. KO muscle shows greater than 50% loss of force compared with that of both WT and HET mice (WT, n = 6; HET, n = 17; and KO, n = 8). *P < 0.05, FDR = 0.05. Data are presented as mean ± SEM.

To test muscle function, we measured the maximum force generated during tetanic contraction of whole hind limbs isolated from P1 mice (Figure 3C). Despite the lack of observable defects by electron microscopy analysis, P1 KO muscle demonstrated >50% reduction in hind limb strength. Consistent with the observed sarcomere dysfunction, microarray analysis of P0 quadriceps showed increased expression of sarcomere genes, possibly as a compensatory response to the underlying sarcomere defect (Supplemental Table 1). Unlike that in skeletal muscle, no differences were seen in KO heart histology or function compared with WT hearts (Supplemental Figure 8, A–C). These data suggest that KO mice have severe skeletal muscle dysfunction that likely contributes to the observed lethality.

KLHL40 binds and increases NEB and LMOD3 proteins. Because the abnormalities resulting from Klhl40 deletion in mice resembled those found in human NM, which is thought to be a primary disease of the thin filament, we asked whether KLHL40 bound and regulated any thin filament proteins. We performed a yeast 2-hybrid screen of a human skeletal muscle cDNA prey library using KLHL40 as bait (Table 1). In parallel, we performed tandem affinity purification (TAP) of tagged KLHL40 from C2C12 myotubes, followed by mass spectrometry (Figure 4A and Supplemental Figure 9). Two proteins were mutually identified from both techniques, NEB and LMOD3, which were validated for binding to KLHL40 by coimmunoprecipitation in nonmuscle COS7 cells (Figure 4B and refs. 33, 34).

Figure 4

KLHL40 binds 2 thin filament proteins, NEB and LMOD3, and LMOD3 localizes to the sarcomere A band. (A) Top KLHL40 binding partners identified following TAP of KLHL40 from cultured C2C12 myotubes. A representative silver-stained gel with TAP protein from myotubes infected with 3XFLAG-HA-EGFP (negative control) or KLHL40-HA-3XFLAG is shown. Proteins listed next to each box indicate the most abundant protein(s) of the correct molecular weight identified in each corresponding area. Abundance is based on estimated enrichment using exponentially modified protein abundance index. TAP of 3XFLAG-HA-KLHL40 protein was also analyzed, and the same top-binding partners were identified (Supplemental Figure 9). Proteins shown in red (NEB and LMOD3) indicate KLHL40 binding partners mutually identified in the TAP and yeast 2-hybrid experiment (Table 1). (B) Coimmunoprecipitation of FLAG-KLHL40 with LMOD3-myc and NEB_frag-myc from COS7 cells. “–” refers to empty FLAG vector for FLAG-KLHL40 and empty myc vector for either LMOD3-myc or NEB_frag-myc. (C and D) LMOD3-EGFP localization relative to myosin in (C) relaxed muscles and (D) contracted muscles shows that LMOD3 resides in the sarcomere A band. Graphs are representative line scans of indicated signals calculated over 8 μm of distance along the length of the myofiber. Scale bar: 5 μm.

Table 1

KLHL40 binding partners identified by yeast 2-hybrid analysis using a skeletal muscle cDNA library

NEB is a thin filament protein associated frequently with NM, but LMOD3 is a novel protein of unknown function (1). LMOD3 is highly homologous to LMOD2, an essential regulator of thin filament actin organization in cardiomyocytes, suggesting that both proteins share similar function (23). We found that LMOD3, like LMOD2, localized to the A band in both relaxed and contracted muscles (Figure 4, C and D). KLHL40 also localized to the A band (Figure 2B), suggesting that LMOD3 and KLHL40 colocalize. Interestingly, both LMOD2 and LMOD3 are specific to striated muscle. Lmod2 was relatively more enriched in the heart compared with skeletal muscle while Lmod3 was expressed at a higher level in skeletal muscle compared with heart, which agrees with previous studies on the expression of these proteins (Supplemental Figure 10 and ref. 24). The expression and localization data suggest that LMOD3 could function similarly to LMOD2, but perhaps specifically for skeletal muscle.

KLHL40 blocks proteasome-mediated LMOD3 degradation by inhibiting ubiquitination. KLHL40 belongs to the BBK protein family; its members serve as substrate-specific adaptors that mediate ubiquitination and degradation of their respective substrates (17–22). To determine whether KLHL40 mediates degradation of NEB and LMOD3, we coexpressed a fragment of NEB, NEB_frag, and LMOD3 with KLHL40 in COS7 cells. Unlike other BBK proteins, we found that KLHL40 dramatically increased protein levels of both NEB_frag and LMOD3, without a corresponding change in respective mRNA transcripts (Figure 5 and Supplemental Figure 11). Addition of proteasome inhibitor was sufficient to increase LMOD3 levels comparably to KLHL40, again without a corresponding change in transcription (Supplemental Figure 11 and Supplemental Figure 12A). Proteasome inhibitor did not substantially increase NEB_frag levels. Finally, we found that KLHL40 decreased K48 polyubiquitination of LMOD3, which specifically marks a protein for proteasome-mediated degradation (Supplemental Figure 12B and ref. 35). These data suggest that KLHL40 serves to increase NEB and LMOD3 levels and prevents degradation of LMOD3 by the proteasome.

Figure 5

KLHL40 increases NEB_frag and LMOD3 protein levels. Western blot analysis of COS7 cells expressing NEB_frag-myc or LMOD3-myc without (Empty) or with FLAG-KLHL40. GAPDH is used as a loading control.

Mapping the domains of KLHL40 that regulate LMOD3 and NEB. KLHL40 contains 3 annotated domains: BTB/POZ, BACK, and KR domains (Supplemental Figure 13A). To determine which domains are important for regulation of LMOD3 and NEB_frag, we generated various KLHL40 mutants lacking combinations of all annotated KLHL40 domains and coexpressed them with LMOD3-myc or NEB_frag-myc (Supplemental Figure 13B). Deletion of the KR domain abolished the stabilizing effect of KLHL40 on both LMOD3 and NEB_frag levels. On the other hand, deletion of the BTB and/or BACK domains selectively diminished the stabilization of NEB_frag. These data suggest that the KLHL40 KR domain plays an essential role in KLHL40-mediated stabilization of both LMOD3 and NEB_frag, while the KLHL40 BTB and BACK domains are only essential for NEB_frag. None of these effects were due to changes in Neb_frag or Lmod3 transcription (Supplemental Figure 13C).

Loss of KLHL40 results in decreased NEB and LMOD3 in mice. We analyzed protein expression in skeletal muscles from P1 and P8 mice of different Klhl40 genotypes and found that NEB was decreased by approximately 50%, while LMOD3 was nearly eliminated in KO mice (Figure 6, A and B, and Supplemental Figure 14, A–C). Notably, Klhl40^+/– (HET) mice also had reduced LMOD3 compared with KLHL40 WT mice, without growth defects or early lethality (Figure 6B). Loss of NEB has been shown to result in shorter thin filament lengths, but KO mice did not have shortened thin filaments like Neb^–/– mice (Supplemental Figure 15 and refs. 36–38).

Figure 6

Reduced NEB and LMOD3 in KLHL40-deficient muscles. (A and B) Densitometry analysis of P1 and P8 quadriceps muscle (A) NEB dot blot and (B) LMOD3 Western blot shows downregulation of both proteins in KO muscles compared with WT muscles (see Supplemental Figure 14) (P1: WT and KO, n = 3; P8: WT, n = 5, HET, n = 4, and KO, n = 5). *P < 0.05, FDR = 0.05. Data are presented as mean ± SEM. (C) Quantitative proteomic analysis of relative protein changes between P6 WT (n = 3) and KO (n = 3) whole skeletal muscle shows that LMOD3 and NEB are 2 of the most downregulated proteins without corresponding changes in transcription (Supplemental Figure 14G). Data are arbitrarily stratified by exponentially modified protein abundance index (emPAI) to allow for better resolution of individual data points. Only proteins with significant (P < 0.05) changes between WT and KO mice are shown. Values for each point are listed in Supplemental Table 2.

Using quantitative mass spectrometry, we found that NEB and LMOD3 were among the most decreased proteins, with LMOD3 being the most downregulated in the entire muscle proteome (Figure 6C). Of all proteins downregulated by greater than 1.5-fold, only LMOD3 and NEB were found to be binding partners by either of our nonbiased KLHL40 binding assays (Supplemental Table 2). Notably, no decrease was seen in Neb or Lmod3 mRNA transcripts (Supplemental Figure 14, D–G).

Some patients with severe NM and KLHL40 deficiency have decreased LMOD3 and NEB. We analyzed LMOD3 and NEB in 3 different patients with 3 different sets of mutations in the KLHL40 locus (Figure 7A). Patient 1 was a compound heterozygote for 2 missense mutations: R311L and T506P. Patient 2 was homozygous for a W90* nonsense mutation, while patient 3 was homozygous for an E528K missense mutation. Patients with the same mutation as patient 3 have a less clinically severe outcome than that observed with patients with other mutations of KLHL40 (12). Accordingly, the skeletal muscle biopsies showed that muscle pathology was more severe in patients 1 and 2 than in patient 3 (Supplemental Figure 16). Patients 1, 2, and 3 all had profoundly reduced KLHL40; however, patient 3 showed residual KLHL40. LMOD3 and NEB were reduced dramatically in patients 1 and 2, while LMOD3 protein remained relatively unchanged in patient 3 when compared with age-appropriate controls (control 1 for patients 1 and 2, and control 2 for patient 3) (Figure 7, B and C). NEB reduction in patient 3 was inconclusive due to issues with GAPDH signal for patient 3 and control 2. Given that patient 3 had some residual KLHL40 and less severe muscle pathology, our findings in mice that KLHL40 serves to stabilize NEB and LMOD3 are largely consistent in KLHL40-deficient patients, although there may be some mechanistic differences based on the type of KLHL40 mutation.

Figure 7

Analyzing LMOD3 and NEB expression in KLHL40-deficient patients. (A) Relative location of KLHL40 mutations in patients analyzed in B and C. Note that patient 1 is a compound heterozygote for the indicated mutations. (B) Detection of LMOD3 and KLHL40 by Western blot analysis and (C) NEB by dot blot analysis in control and KLHL40-deficient patient skeletal muscle biopsies. Patients 1 and 2 have a clear decrease in LMOD3 and NEB compared with age-matched control, control 1. Patient 3 compared with age-matched control, control 2, does not have decreased LMOD3, and changes in NEB are inconclusive due to issues with GAPDH signal. Control 1 and control 2 are 11-day-old and 6-month-old neonate controls, respectively. Patient 1 was biopsied at 2 days, patient 2 was biopsied at 25 days, and patient 3 was biopsied at 3 months of age. GAPDH is shown for loading control.

KLHL40 mutations have been shown recently to cause NM in humans, but the underlying pathogenic mechanisms have not been defined. The results of this study demonstrate that KLHL40 is essential for the maintenance of sarcomere structure and muscle contractility. Loss of KLHL40 in mice leads to a NM-like phenotype comparable to that of patients with NM lacking KLHL40 (12). We found that KLHL40 localizes to the I band and A band of sarcomeres and binds to 2 proteins located in these bands, NEB and LMOD3. Unlike previously described BBK proteins, we showed that KLHL40 promotes the stability of NEB and LMOD3, providing a possible molecular explanation for NM in KLHL40-deficient patients. Consistent with these findings, NEB and LMOD3 protein levels are specifically diminished in KLHL40-deficient mice and KLHL40-deficient patients, suggesting the involvement of KLHL40 in NM and regulation of sarcomere structure. We propose that KLHL40 regulates NEB and LMOD3 and that loss of KLHL40 primarily reduces NEB and LMOD3. This results in thin filament disruption, with subsequent irregularities of sarcomere Z-discs and, in extreme cases, sarcomere dissolution with the formation of ovoid Z-discs, culminating in a fatal loss of muscle function (Figure 8).

Figure 8

Proposed mechanism of NM pathogenesis due to loss of KLHL40. Under normal conditions, KLHL40 either blocks the degradation or promotes successful folding of NEB and LMOD3, which function to stabilize the thin filament, allowing normal sarcomere function. In muscles lacking KLHL40, NEB and LMOD3 protein levels are reduced, resulting in destabilization of thin filaments, sarcomere dysfunction, and subsequent NM.

NM-like disease due to loss of Klhl40. Similar to human patients that lack KLHL40, KO mice display neonatal lethality, with defects in sarcomere structure and significant muscle weakness (12). Cardiac defects are not a typical feature of NM, and, although Klhl40 has some expression in the heart, we did not observe cardiac hypertrophy, dilation, or any loss in cardiac contractility in KO mice (1, 39).

Unlike patient biopsies, classic nemaline bodies are not readily apparent in the KO mice, except in highly disorganized fibers in which they present as widened Z-discs. The more prevalent defect seen in mice is the presence of Z line streaming, which is indicative of sarcomere damage and is often seen in combination with nemaline bodies in patient biopsies (9, 40). The lack of bona fide nemaline bodies in other KO mouse models of NM is not uncommon. Deletion of Acta1 and Neb in mice, the 2 most frequently implicated proteins in NM (1), results in neonatal lethality and muscle dysfunction, but classical nemaline bodies are not reported (36–38, 41). Currently, nemaline bodies are only reproduced in mouse models of NM in which mutant protein is overexpressed, suggesting that mutant proteins are required for protein aggregation to form rods (42–44). A previous study noted that the presence of nemaline bodies does not correlate with the severity of the disease (2). Instead, it is believed that decreased contractility due to primary defects in the sarcomere thin filament is the underlying cause for symptoms in NM (10). This hypothesis is supported by studies with P1 NEB-deficient mice, which have a dramatic loss in muscle contractility, with no visible sarcomere defects (36, 45). Accordingly, we too found that sarcomeres in P1 KLHL40 WT and KO mice appeared comparable, but there was still a greater than 50% loss of maximum tetanic force in KO muscle. Our data, combined with previous studies, corroborate the hypothesis that NM is likely due to sarcomere dysfunction and instability, causing a primary loss in muscle contractility (2, 36–38, 41). We believe that sarcomere instability results in the visible Z line streaming and nemaline bodies in KLHL40-deficient mice at later ages due to increased muscle use following birth.

KLHL40 binding and regulation of thin filament proteins, NEB and LMOD3. We have demonstrated that KLHL40 localizes to the sarcomere I band and the A band, which potentially explains the conflicting results from two previous reports showing I band and A band localization separately (12, 13). NEB is known to span the I band and the A band, and we have shown that LMOD3 occupies the A band (46). Thus, KLHL40 occupies the same subcellular compartments as NEB and LMOD3.

In contrast to all other described BBK proteins, we have demonstrated that KLHL40 actually stabilizes, rather than degrades, its substrates (15–22). NEB_frag and LMOD3 are increased in cell culture with KLHL40 expression, and loss of KLHL40 in mice results in reduced NEB and LMOD3 independent of any transcriptional changes. In addition, KLHL40 appears to block LMOD3 ubiquitination, whereas KLHL40 regulates NEB_frag in a proteasome-independent manner. These findings provide the first example of a BBK protein that can mediate protein stabilization rather than degradation.

Previous studies showed that the KR domain in BBK proteins is required for binding of substrate to its respective BBK protein (20–22). Accordingly, deletion of the KLHL40 KR domain abrogates stabilization of LMOD3 and NEB_frag in cell culture. Deletion of the BTB or BACK domains of KLHL40 has no effect on LMOD3 stabilization, but NEB_frag levels are reduced, again, implying alternative mechanisms for KLHL40 regulation of NEB_frag.

BBK proteins typically bind the CUL3-E3 ligase complex to mediate substrate recognition for ubiquitination (15–22). It is possible that KLHL40 prevents CUL3-mediated degradation of LMOD3, perhaps by preventing recognition of LMOD3 by the CUL3-E3 complex. Alternatively, based on KLHL40’s differential localization in relaxed and contracted muscles, KLHL40 may have a more chaperone-like function to maintain proper LMOD3 and NEB folding throughout muscle contraction, which secondarily prevents their degradation. Further experiments are required to understand the role of KLHL40 during muscle contraction and to elucidate its unique pro-stability function. It will be of interest to determine whether other BBK proteins also stabilize their substrates.

Role of decreased NEB and LMOD3 in NM phenotype. Whole-proteome analysis of KO muscle revealed that NEB and LMOD3 were among the most downregulated proteins and were the only identified binding partners to show such dramatic regulation. LMOD3 was the most reduced protein in the entire muscle proteome. In fact, even 50% reduction of Klhl40 transcript in mice lacking a single copy of the Klhl40 gene was sufficient to reduce LMOD3, suggesting a tight dose dependence of LMOD3 on KLHL40 protein levels.

Previous work has shown definitively that mutations in NEB in patients and loss of Neb in mice results in NM, suggesting that decreased NEB in KO mice could contribute to the NM phenotype (1, 4, 36–38). A pervading belief is that NEB acts as a molecular ruler to regulate thin filament lengths based on shortened thin filament lengths seen in Neb-deficient mice (36, 37, 47). However, more recent studies argue that NEB deficiency contributes directly to a loss in muscle contractility due to dysregulation of actin-myosin crossbridge formation (45, 47, 48). Accordingly, we see over a 50% loss in hind limb strength of P1 KO mice, which have approximately 50% reduction in NEB, without any evidence of sarcomere defects. It is possible that KLHL40 loss functionally impairs the actin-myosin crossbridge activity of NEB.

LMOD3 function has hitherto not been defined, and it has not yet been implicated in any disease. Our study presents the first evidence of a possible role of LMOD3 in NM. LMOD3 is highly homologous to LMOD2, which was shown previously to be vital for sarcomere structure in cardiomyocytes by regulating actin assembly at the pointed end of thin filaments (near the sarcomere M line) (23). Indeed, we found that LMOD3 localizes to the A band in a matter similar to that of LMOD2, likely indicating that LMOD3 also binds the pointed end of actin filaments. Thus, loss of LMOD3, secondary to KLHL40 deficiency, may lead to thin filament actin dysregulation.

Role of BBK proteins in human NM. Thus far, 3 BBK proteins have been implicated in NM: KBTBD13, KLHL40, and KLHL41 (11–13). Our study is the first to define a possible molecular mechanism for any of these proteins. Consistent with the phenotype of KLHL40 mutant mice, the 2 patients with NM (patients 1 and 2) completely deficient for KLHL40 had marked reduction in LMOD3 and NEB, while a third patient (patient 3), with some residual KLHL40, had relatively normal LMOD3. Patients with the same mutation as patient 3 present with less severe clinical symptoms compared with patients with all other KLHL40 mutations (12). Thus, our finding that KLHL40 serves to stabilize LMOD3 in mice is largely consistent in KLHL40-deficient patients, although there may be some mechanistic differences based on the type of KLHL40 mutation. Regardless, these findings may have important therapeutic implications for KLHL40-deficient patients (see below).

A recent report described KLHL41 mutations causing NM in humans (13). KLHL41 is the most closely related protein to KLHL40 (71% by BLASTP alignment) (13, 49–51). Interestingly, we found that KLHL41 was the most enriched KLHL40 binding partner from C2C12 myotubes. As determined by quantitative proteomics, KLHL41 levels in KLHL40-deficient mice were not changed, suggesting that KLHL40 does not regulate protein levels of KLHL41. We hypothesize that KLHL40 and KLHL41 may form a heterodimer or polymer to regulate thin filament proteins with both KLHL40 and KLHL41 performing unique, nonredundant functions, possibly explaining why a disruption in either protein results in NM. It is also tempting to speculate that KLHL41, like KLHL40, promotes the stability of its substrates based on their close homology. On the other hand, KBTBD13 is much more distantly related to KLHL40 and KLHL41, and KBTBD13 mutations resulting in NM are autosomal dominant in contrast to the autosomal recessive transmission of KLHL40- and KLHL41-related NM (51). Thus, the molecular mechanism underlying KBTBD13 NM is likely different than that of KLHL40 and KLHL41.

Therapeutic implications. There are currently no therapies available to treat NM. However, our data suggest that KLHL40 deficiency results in NM by secondary loss of NEB and LMOD3 protein and that proteasome inhibition may be sufficient to restore LMOD3. Should LMOD3 be determined to be a major contributor to the NM phenotype seen with KLHL40 deficiency, it would be worthwhile to explore whether proteasome inhibition can alleviate muscle dysfunction in KLHL40-deficient patients. Proteasome inhibitors have already been used in cancer therapy and are thus available for clinical use (52). Therefore, further delineation of the molecular mechanism by which KLHL40 promotes NEB and LMOD3 stability may allow for the development of the first therapy for NM.

View Supplemental data

We are grateful to Xiaoxia Qi, John R. McAnally, and Xiang Chen for technical assistance. We thank Robyn Leidel, Karen Rothberg, Kate Luby-Phelps, and Christopher Gilpin for assistance and advice in the UT Southwestern Electron Microscopy Core Facility as well as John Shelton and James Richardson for help with histology and imaging. We thank Marie-Louise Bang (Istituto di Ricerca Genetica e Biomedica) for providing tissue from NEB-deficient mice. We thank Jose Cabrerra for assistance with graphics. This work was supported by grants from the NIH (HL-077439, HL-111665, HL-093039, DK-099653, and U01-HL-100401) and the Robert A. Welch Foundation (grant 1-0025 to E.N. Olson). A. Garg is supported by a T32 NIH training grant (T32-HL-007360). G. Ravenscroft and N.G. Laing are supported by Australian National Health and Medical Research Council (NHMRC) Early Career and Principal Research Fellowships APP1035955 and APP1022707 and NHMRC project grant APP1022707.

Address correspondence to: Eric N. Olson, Department of Molecular Biology, 5323 Harry Hines Blvd., Dallas, Texas 75390-9148, USA. Phone: 214.648.1187; Fax: 214.648.1196; E-mail: eric.olson@utsouthwestern.edu.

Conflict of interest: The authors have declared that no conflict of interest exists.

Reference information: J Clin Invest. 2014;124(8):3529–3539. doi:10.1172/JCI74994.

Jason O’Rourke’s present address is: Genomics Institute of the Novartis Research Foundation, San Diego, California, USA.

See the related article at Leiomodin-3 dysfunction results in thin filament disorganization and nemaline myopathy.

1. Wallgren-Pettersson C, Sewry CA, Nowak KJ, Laing NG. Nemaline myopathies. Semin Pediatr Neurol. 2011;18(4):230–238.
   View this article via: PubMed CrossRef Google Scholar
2. Shimomura C, Nonaka I. Nemaline myopathy: comparative muscle histochemistry in the severe neonatal, moderate congenital, and adult-onset forms. Pediatr Neurol. 1989;5(1):25–31.
   View this article via: PubMed CrossRef Google Scholar
3. Laing NG, et al. A mutation in the α tropomyosin gene TPM3 associated with autosomal dominant nemaline myopathy NEM1. Nat Genet. 1995;10(2):249.
   View this article via: PubMed Google Scholar
4. Lehtokari VL, et al. Identification of 45 novel mutations in the nebulin gene associated with autosomal recessive nemaline myopathy. Hum Mutat. 2006;27(9):946–956.
   View this article via: PubMed CrossRef Google Scholar
5. Nowak KJ, et al. Mutations in the skeletal muscle α-actin gene in patients with actin myopathy and nemaline myopathy. Nat Genet. 1999;23(2):208–212.
   View this article via: PubMed CrossRef Google Scholar
6. Donner K, et al. Mutations in the β-tropomyosin (TPM2) gene — a rare cause of nemaline myopathy. Neuromuscul Disord. 2002;12(2):151–158.
   View this article via: PubMed CrossRef Google Scholar
7. Johnston JJ, et al. A novel nemaline myopathy in the Amish caused by a mutation in troponin T1. Am J Hum Genet. 2000;67(4):814–821.
   View this article via: PubMed CrossRef Google Scholar
8. Agrawal PB, et al. Nemaline myopathy with minicores caused by mutation of the CFL2 gene encoding the skeletal muscle actin-binding protein, cofilin-2. Am J Hum Genet. 2007;80(1):162–167.
   View this article via: PubMed CrossRef Google Scholar
9. Franzini-Armstrong C, Lee Sweeney H. The contractile machinery of skeletal muscle. In: Hill JA, Olson EN, eds. Muscle. Boston, Massachusetts, USA: Academic Press; 2012:823–840.
10. Nance JR, Dowling JJ, Gibbs EM, Bönnemann CG. Congenital myopathies: an update. Curr Neurol Neurosci Rep. 2012;12(2):165–174.
    View this article via: PubMed CrossRef Google Scholar
11. Sambuughin N, et al. Dominant mutations in KBTBD13, a member of the BTB/Kelch family, cause nemaline myopathy with cores. Am J Hum Genet. 2010;87(6):842–847.
    View this article via: PubMed CrossRef Google Scholar
12. Ravenscroft G, et al. Mutations in KLHL40 are a frequent cause of severe autosomal-recessive nemaline myopathy. Am J Hum Genet. 2013;93(1):6–18.
    View this article via: PubMed CrossRef Google Scholar
13. Gupta VA, et al. Identification of KLHL41 mutations implicates BTB-Kelch-mediated ubiquitination as an alternate pathway to myofibrillar disruption in nemaline myopathy. Am J Hum Genet. 2013;93(6):1108–1117.
    View this article via: PubMed CrossRef Google Scholar
14. Stogios PJ, Privé GG. The BACK domain in BTB-kelch proteins. Trends Biochem Sci. 2004;29(12):634–637.
    View this article via: PubMed CrossRef Google Scholar
15. Bennett EJ, Rush J, Gygi SP, Harper JW. Dynamics of cullin-RING ubiquitin ligase network revealed by systematic quantitative proteomics. Cell. 2010;143(6):951–965.
    View this article via: PubMed CrossRef Google Scholar
16. Emanuele MJ, et al. Global identification of modular cullin-RING ligase substrates. Cell. 2011;147(2):459–474.
    View this article via: PubMed CrossRef Google Scholar
17. Zhang DD, Lo SC, Sun Z, Habib GM, Lieberman MW, Hannink M. Ubiquitination of Keap1, a BTB-Kelch substrate adaptor protein for Cul3, targets Keap1 for degradation by a proteasome-independent pathway. J Biol Chem. 2005;280(34):30091–30099.
    View this article via: PubMed CrossRef Google Scholar
18. Angers S, et al. The KLHL12-Cullin-3 ubiquitin ligase negatively regulates the Wnt-β-catenin pathway by targeting Dishevelled for degradation. Nat Cell Biol. 2006;8(4):348–357.
    View this article via: PubMed CrossRef Google Scholar
19. Sumara I, et al. A Cul3-based E3 ligase removes Aurora B from mitotic chromosomes, regulating mitotic progression and completion of cytokinesis in human cells. Dev Cell. 2007;12(6):887–900.
    View this article via: PubMed CrossRef Google Scholar
20. Lee YR, Yuan WC, Ho HC, Chen CH, Shih HM, Chen RH. The Cullin 3 substrate adaptor KLHL20 mediates DAPK ubiquitination to control interferon responses. EMBO J. 2010;29(10):1748–1761.
    View this article via: PubMed CrossRef Google Scholar
21. Jin L, et al. Ubiquitin-dependent regulation of COPII coat size and function. Nature. 2012;482(7386):495–500.
    View this article via: PubMed CrossRef Google Scholar
22. Shibata S, Zhang J, Puthumana J, Stone KL, Lifton RP. Kelch-like 3 and Cullin 3 regulate electrolyte homeostasis via ubiquitination and degradation of WNK4. Proc Natl Acad Sci U S A. 2013;110(19):7838–7843.
    View this article via: PubMed CrossRef Google Scholar
23. Chereau D, et al. Leiomodin is an actin filament nucleator in muscle cells. Science. 2008;320(5873):239–243.
    View this article via: PubMed CrossRef Google Scholar
24. Nanda V, Miano JM. Leiomodin 1, a new serum response factor-dependent target gene expressed preferentially in differentiated smooth muscle cells. J Biol Chem. 2012;287(4):2459–2467.
    View this article via: PubMed CrossRef Google Scholar
25. Gray PA, et al. Mouse brain organization revealed through direct genome-scale TF expression analysis. Science. 2004;306(5705):2255–2257.
    View this article via: PubMed CrossRef Google Scholar
26. Liu N, et al. Requirement of MEF2A, C, D for skeletal muscle regeneration. Proc Natl Acad Sci U S A. 2014;111(11):4109–4114.
    View this article via: PubMed CrossRef Google Scholar
27. Valenzuela DM, et al. High-throughput engineering of the mouse genome coupled with high-resolution expression analysis. Nat Biotechnol. 2003;21(6):652–659.
    View this article via: PubMed CrossRef Google Scholar
28. Sternberg EA, Spizz G, Perry WM, Vizard D, Weil T, Olson EN. Identification of upstream and intragenic regulatory elements that confer cell-type-restricted and differentiation-specific expression on the muscle creatine kinase gene. Mol Cell Biol. 1988;8(7):2896–2909.
    View this article via: PubMed Google Scholar
29. Johnson JE, Wold BJ, Hauschka SD. Muscle creatine kinase sequence elements regulating skeletal and cardiac muscle expression in transgenic mice. Mol Cell Biol. 1989;9(8):3393–3399.
    View this article via: PubMed Google Scholar
30. DiFranco M, Quinonez M, Capote J, Vergara J. DNA transfection of mammalian skeletal muscles using in vivo electroporation. J Vis Exp. 2009;pii(32):1520.
    View this article via: PubMed Google Scholar
31. Knight PJ, Trinick JA. Preparation of myofibrils. Methods Enzymol. 1982;85(pt B):9–12.
    View this article via: PubMed Google Scholar
32. Gokhin DS, Kim NE, Lewis SA, Hoenecke HR, D’Lima DD, Fowler VM. Thin-filament length correlates with fiber type in human skeletal muscle. Am J Physiol Cell Physiol. 2012;302(3):C555–C565.
    View this article via: PubMed CrossRef Google Scholar
33. Jensen FC, Girardi AJ, Gilden RV, KoprowskI H. Infection of human and simian tissue cultures with Rous sarcoma virus. Proc Natl Acad Sci U S A. 1964;52:53–59.
    View this article via: PubMed CrossRef Google Scholar
34. Gluzman Y. SV40-transformed simian cells support the replication of early SV40 mutants. Cell. 1981;23(1):175–182.
    View this article via: PubMed CrossRef Google Scholar
35. Johnson ES. Ubiquitin branches out. Nat Cell Biol. 2002;4(12):E295–E298.
    View this article via: PubMed CrossRef Google Scholar
36. Bang ML, et al. Nebulin-deficient mice exhibit shorter thin filament lengths and reduced contractile function in skeletal muscle. J Cell Biol. 2006;173(6):905–916.
    View this article via: PubMed CrossRef Google Scholar
37. Witt CC, et al. Nebulin regulates thin filament length, contractility, and Z-disk structure in vivo. EMBO J. 2006;25(16):3843–3855.
    View this article via: PubMed CrossRef Google Scholar
38. Gokhin DS, Bang ML, Zhang J, Chen J, Lieber RL. Reduced thin filament length in nebulin-knockout skeletal muscle alters isometric contractile properties. Am J Physiol Cell Physiol. 2009;296(5):C1123–C1132.
    View this article via: PubMed CrossRef Google Scholar
39. Hill JA, Olson EN. Cardiac plasticity. N Engl J Med. 2008;358(13):1370–1380.
    View this article via: PubMed CrossRef Google Scholar
40. Agrawal PB, et al. Heterogeneity of nemaline myopathy cases with skeletal muscle α-actin gene mutations. Ann Neurol. 2004;56(1):86–96.
    View this article via: PubMed CrossRef Google Scholar
41. Crawford K, et al. Mice lacking skeletal muscle actin show reduced muscle strength and growth deficits and die during the neonatal period. Mol Cell Biol. 2002;22(16):5887–5896.
    View this article via: PubMed CrossRef Google Scholar
42. Corbett MA, et al. A mutation in α-tropomyosin(slow) affects muscle strength, maturation and hypertrophy in a mouse model for nemaline myopathy. Hum Mol Genet. 2001;10(4):317–328.
    View this article via: PubMed CrossRef Google Scholar
43. Ravenscroft G, et al. Mouse models of dominant ACTA1 disease recapitulate human disease and provide insight into therapies. Brain. 2011;134(pt 4):1101–1115.
    View this article via: PubMed Google Scholar
44. Nowak KJ, et al. Nemaline myopathy caused by absence of alpha-skeletal muscle actin. Ann Neurol. 2007;61(2):175–184.
    View this article via: PubMed CrossRef Google Scholar
45. Bang ML, et al. Nebulin plays a direct role in promoting strong actin-myosin interactions. FASEB J. 2009;23(12):4117–4125.
    View this article via: PubMed CrossRef Google Scholar
46. Pappas CT, Bliss KT, Zieseniss A, Gregorio CC. The Nebulin family: an actin support group. Trends Cell Biol. 2011;21(1):29–37.
    View this article via: PubMed CrossRef Google Scholar
47. Ottenheijm CA, et al. Deleting exon 55 from the nebulin gene induces severe muscle weakness in a mouse model for nemaline myopathy. Brain. 2013;136(pt 6):1718–1731.
    View this article via: PubMed Google Scholar
48. Chandra M, et al. Nebulin alters cross-bridge cycling kinetics and increases thin filament activation: a novel mechanism for increasing tension and reducing tension cost. J Biol Chem. 2009;284(45):30889–30896.
    View this article via: PubMed CrossRef Google Scholar
49. Altschul SF, et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 1997;25(17):3389–3402.
    View this article via: PubMed CrossRef Google Scholar
50. Altschul SF, et al. Protein database searches using compositionally adjusted substitution matrices. FEBS J. 2005;272(20):5101–5109.
    View this article via: PubMed CrossRef Google Scholar
51. Canning P, et al. Structural basis for Cul3 protein assembly with the BTB-Kelch family of E3 ubiquitin ligases. J Biol Chem. 2013;288(11):7803–7814.
    View this article via: PubMed CrossRef Google Scholar
52. Rastogi N, Mishra DP. Therapeutic targeting of cancer cell cycle using proteasome inhibitors. Cell Div. 2012;7(1):26.
    View this article via: PubMed CrossRef Google Scholar
53. Millay DP, et al. Myomaker is a membrane activator of myoblast fusion and muscle formation. Nature. 2013;499(7458):301–305.
    View this article via: PubMed CrossRef Google Scholar
54. Kim MS, et al. Protein kinase D1 stimulates MEF2 activity in skeletal muscle and enhances muscle performance. Mol Cell Biol. 2008;28(11):3600–3609.
    View this article via: PubMed CrossRef Google Scholar
55. Naya FJ, Mercer B, Shelton J, Richardson JA, Williams RS, Olson EN. Stimulation of slow skeletal muscle fiber gene expression by calcineurin in vivo. J Biol Chem. 2000;275(7):4545–4548.
    View this article via: PubMed CrossRef Google Scholar
56. Nelson BR, et al. Skeletal muscle-specific T-tubule protein STAC3 mediates voltage-induced Ca2+ release and contractility. Proc Natl Acad Sci U S A. 2013;110(29):11881–11886.
    View this article via: PubMed CrossRef Google Scholar
57. Schindelin J, et al. Fiji: an open-source platform for biological-image analysis. Nat Methods. 2012;9(7):676–682.
    View this article via: PubMed CrossRef Google Scholar
58. Wolff AV, et al. Passive mechanical properties of maturing extensor digitorum longus are not affected by lack of dystrophin. Muscle Nerve. 2006;34(3):304–312.
    View this article via: PubMed CrossRef Google Scholar
59. Shelton JM, Lee MH, Richardson JA, Patel SB. Microsomal triglyceride transfer protein expression during mouse development. J Lipid Res. 2000;41(4):532–537.
    View this article via: PubMed Google Scholar
60. Aurora AB, et al. Macrophages are required for neonatal heart regeneration. J Clin Invest. 2014;124(3):1382–1392.
    View this article via: JCI PubMed CrossRef Google Scholar
61. Li S, et al. Requirement for serum response factor for skeletal muscle growth and maturation revealed by tissue-specific gene deletion in mice. Proc Natl Acad Sci U S A. 2005;102(4):1082–1087.
    View this article via: PubMed CrossRef Google Scholar
62. Dubowitz V, Sewry CA. Muscle Biopsy — A Practical Approach. Philadelphia, Pennsylvania, USA: Elsevier Limited; 2007.