Nesprin-2 is a novel scaffold protein for telethonin and FHL-2 in the cardiomyocyte sarcomere

Abstract

Nesprins comprise a family of multi-isomeric scaffolding proteins, forming the linker of nucleoskeleton-and-cytoskeleton complex with lamin A/C, emerin and SUN1/2 at the nuclear envelope. Mutations in nesprin-1/-2 are associated with Emery-Dreifuss muscular dystrophy (EDMD) with conduction defects and dilated cardiomyopathy (DCM). We have previously observed sarcomeric staining of nesprin-1/-2 in cardiac and skeletal muscle, but nesprin function in this compartment remains unknown. In this study, we show that specific nesprin-2 isoforms are highly expressed in cardiac muscle and localize to the Z-disc and I band of the sarcomere. Expression of GFP-tagged nesprin-2 giant spectrin repeats 52 to 53, localized to the sarcomere of neonatal rat cardiomyocytes. Yeast two-hybrid screening of a cardiac muscle cDNA library identified telethonin and four-and-half LIM domain (FHL)-2 as potential nesprin-2 binding partners. GST pull-down and immunoprecipitation confirmed the individual interactions between nesprin-2/telethonin and nesprin-2/FHL-2, and showed that nesprin-2 and telethonin binding was dependent on telethonin phosphorylation status. Importantly, the interactions between these binding partners were impaired by mutations in nesprin-2, telethonin, and FHL-2 identified in EDMD with DCM and hypertrophic cardiomyopathy patients. These data suggest that nesprin-2 is a novel sarcomeric scaffold protein that may potentially participate in the maintenance and/or regulation of sarcomeric organization and function.

Keywords: FHL-2; cardiomyopathies; interaction; nesprin-2; sarcomere organization; telethonin.

Figure 1

Nesprin-2 redistributes from the NE to the sarcomere during myogenesis.A, schematic showing nesprin-2 giant gene and protein structure. The alternative novel stop codon is shown upstream of the KASH domain of nesprin-2 giant, which was generated by the retention of part of the intronic sequence (123 bp) immediately downstream of the penultimate exon 115, colored in pink). The putative partial nesprin-2 KASH-less isoform was shown with the position of nesprin-2 N3 (Nes-2N3) antibody indicated, which was specifically generated against the SR52 close to the C-terminus of nesprin-2. B, immunofluorescence showed nesprin-2 localised to the NE at E12.5 and was present in the sarcomere in E16.5 and during development as observed (inserts, yellow arrowed), colocalised with alpha-actinin (red) for the sarcomere and emerin (red) for the NE. C, Nesprin-2 (green) is present at the Z-disc and I band in the sarcomere in adult mouse cardiomyocytes, colocalised with α-actinin (red) and SERCA-2 (red). Myomesin (red) as a M-band marker, nuclei shown in blue with Dapi, Colocalisation quantification data is shown with Pearson’s coefficient for α-actinin (0.61), myomesin (0.10) and SERCA-2 (0.55), respectively. D, Western blot demonstrated dramatic changes in the production of nesprin-2 isoforms during mouse heart embryonic development, previously published nesprin isoforms are indicated (13, 15). E, qPCR verified this partial nesprin-2 KASH-less isoform that contains the SR52-53, which is exclusively present in heart tissue.

Figure 2

Telethonin and FHL-2 are potential binding partners for nesprin-2 in the sarcomere.A, neonatal rat cardiomyocytes were transfected with either GFP empty vector or GFP-tagged nesprin-2 SR52-KASH (SR52 to last amino acid of nesprin-2 giant), KASH domain or nesprin-2 SR52-53 as shown in the schematic. 48 h after transfection, cells were fixed and stained for M-band marker myomesin (red). EGFP-tagged nesprin-2 SR52 to KASH and the KASH domain localized to the NE with weak staining at the sarcomere in transfected neonatal rat cardiomyocytes (two middle panels, green, arrowed). In contrast, EGFP-tagged nesprin-2 SR52-53 showed sarcomere staining (bottom panel, green, arrowed), localizing to the Z-disc and I band in contrast to myomesin-an M band protein (arrowhead, insert). Dapi shows the nucleus staining (blue). B, yeast two-hybrid screens on a human cardiac muscle library identified telethonin and FHL-2 are novel nesprin-2 binding partners in the sarcomere as shown in the schematic. Nesprin-2 SR52-53 (amino acids 6136–6354 of the full length of nesprin-2 giant) and SR54-56 (amino acids 6461–6781) were cloned into the GBDKT-7 matchmaker bait vector (Clontech) as bait plasmids. C, immunofluorescence staining showed nesprin-2 mainly localizes to the Z-disc and I band of the sarcomere in both adult mouse cardiomyocytes and heart tissue using Nes2-N3 (green), telethonin (red) and FHL-2 (red) and myomesin (red) antibodies. Myomesin as an M-band marker for the sarcomere. Colocalization quantification is shown with Pearson’s coefficient for telethonin (0.95), FHL-2 (0.77), and myomesin (0.44) respectively.

Figure 3

Confirmation of the binding between nesprin-2 and telethonin.A, GST pull-down assay showed the interaction between nesprin and telethonin by using GST-nesprin-2 SR51-53 beads and purified His-tagged telethonin recombinant protein. B, IP showed telethonin antibody pull-down nesprin-2 (∼200 kDa) by using nesprin-2 N3 (Nes-2 N3) antibody and protein lysate from neonatal rat heart tissue, C, Co-elution of bacterially expressed and purified His-telethonin and nesprin-2 SR51-53 (no tag) by size exclusion chromatography. The fractions 101 to 112 from Superdex 200 16/60 (Fig. S3) were pooled and loaded to a S75 10/300 column. Only one peak was eluted at 10.29 ml. Western blot confirmed this single peak elution is corresponding to a 60 kDa protein (containing both nesprin-2 SR51-53: 40.51 kDa, and telethonin:19 kDa), shown by the inserts (top: nesprin-2 N3 antibody and bottom: telethonin antibody. B2 and B3 correspond to the fractions of the single peak.

Figure 4

The binding between nesprin-2/telethonin is tightly regulated by telethonin phosphorylation.A, HA-tagged WT and S157/161A mutated telethonin protein with or without PKD pre-treatment were incubated with GST-nesprin-2 SR51-53 beads for a GST pull-down assay. The nesprin binding for wild-type (WT) or mutant recombinant telethonin proteins was quantified by densitometry with respect to the input material and expressed as a ratio of the value obtained for unphosphorylated WT telethonin. Five independent experiments were performed shown as mean ± SD, ∗∗∗∗p < 0.0001. F (7, 32) =209.5 using one-way ANOVA analysis; Bonferroni post-hoc tests showed ∗∗∗∗p < 0.0001 for WT-versus WT+, S157- versus S157+, S161- versus S161+ respectively. B, NRCs were infected with adenovirus carrying HA-tagged WT telethonin or S157/161A double mutants. Protein lysates were harvested 48 h after adenoviral infection and then subjected to immunoprecipitation using a telethonin antibody. Western blot showed nesprin-2 was significantly pulled down in neonatal cardiomyocytes with over-expressed unphosphorylatable S157/161A telethonin. HA antibody used on the same blot showed an equal amount of HA-tagged telethonin pulled down by telethonin antibody. The intensity of about ∼200 kDa nesprin pull-down bands was measured by densitometry with respect to the input material and relative binding to nesprin expressed as a ratio of the value obtained between mutant and WT telethonin. IP was repeated for five times. Student’s t test (unpaired, two-tailed) showed a significantly higher binding strength between nesprin and S157/161A double mutated telethonin than the WT. ∗∗∗∗p < 0.0001.

Figure 5

FHL-2 is also a novel sarcomeric binding partner for nesprin-2.A and B, overexpression of Flag-tagged nesprin-2 and GFP-FHL-2 proteins in C2C12 myoblasts and then immunoprecipitation with GFP antibody. Western blotted with both Flag antibody and Nes-2N3 antibody respectively, showed nesprin-2 and FHL-2 binding. C, GST pull-down GFP-FHL-2 using GST-nesprin-2 SR51-53 and GST nesprin-2 CT (SR 51-56 without KASH). D, Functional binding domains of nesprin-2 and FHL-2 were mapped using nesprin-2 and FHL-2 deletion constructs shown by the schematic. E and F, immunoprecipitation in C2C12 myotube lysates with both nesprin-2 N3 and FHL-2 antibodies confirmed that FHL-2 is a novel nesprin-2 binding partner.

Figure 6

Nesprin2/telethonin or nesprin-2/FHL-2 interaction was impaired by both nesprin-2 and telethonin or FHL-2 mutants. GST-nesprin-2 SR51-53 WT and mutant T6211M proteins were incubated with His-tagged telethonin recombinant protein (A) or GFP-tagged FHL-2 protein (B) for GST pull-down assay. Coomassie blue staining of the gel showed an equal amount of GST-nesprin-2 SR51-53 WT and T6211M beads were used. The His-telethonin or GFP-FHL-2 binding for GST-Nesprin-2 SR51-53WT and GST-nesprin-2 SR51-53 T6211M was quantified by densitometry and expressed as a ratio of the value to show the relative binding. The GST pull-down was repeated five times, and Student’s t test (unpaired, two-tailed) confirmed the T6211M mutation significantly reduced nesprin binding with telethonin (∗∗∗∗p < 0.0001) or FHL-2 (∗∗∗p = 0.002). The schematic showed where T6211M is located (A). C, his-tagged WT and R153H mutated telethonin with or without PKD pre-treatment were incubated with GST-nesprin-2 SR51-53 beads for a GST pull-down assay. Coomassie blue staining of the SDS-PAGE gel showed an equal amount of GST-nesprin-2 SR51-53 beads were used for each pull-down. The nesprin interaction with different telethonin protein samples was measured by densitometry with respect to the input material and relative binding to nesprin expressed as the ratio of the nesprin binding to the unphosphorylated WT telethonin. Five independent experiments were performed shown as mean ± SD, ∗∗∗∗p < 0.0001, F (3, 16) = 218 using one-way ANOVA analysis; Bonferroni post-hoc tests showed ∗∗∗∗p < 0.0001 for WT P- versus WT P+, WT P- versus R153H P-; ∗∗p = 0.0015 for R153H P- versus R153H P+ respectively. D, GFP-tagged FHL-2 WT, and R113C mutant were incubated with GST-nesprin-2 SR51-53 beads for a GST pull-down assay. Coomassie blue staining of the SDS-PAGE gel showed an equal amount of GST-nesprin-2 SR51-53 beads were used for each pull-down. The nesprin binding for WT or mutant FHL-2 proteins was quantified by densitometry with respect to the input material and expressed as a ratio of the value obtained for WT FHL-2. Four independent experiments were performed shown as mean ± SD, ∗p < 0.0179 using Student’s t test (unpaired, two-tailed).

Figure 7

A potential role of nesprin-2 in regulating hypertrophic response.A, Western blot showed a ∼200 kDa nesprin-2 isoform was significantly up-regulated in the nuclear fraction of the H9C2 cells treated with phenylephrine, 25 kDa nesprin in the cytoplasmic fraction could be potential isoforms due to extensively alternative initiation and splicing or degradation products, further verification needed. B and C, Western blot and immunofluorescence showed nesprin-2 isoforms, especially those containing SR 52-53 redistributes from the sarcomere to the NE in hypertrophic cardiomyocytes in a transverse aortic constriction (TAC) mouse model (arrowed), and FHL-2 subcellular localization was affected shown in (C, arrowhead).