Research Article: IGFN1_v1 is required for myoblast fusion and differentiation

Date Published: June 30, 2017

Publisher: Public Library of Science

Author(s): Xiang Li, Jane Baker, Tobias Cracknell, Andrew R. Haynes, Gonzalo Blanco, Makoto Kanzaki.


Igfn1 is a complex locus that codes for multiple splicing variants of Immunoglobulin- and Fibronectin-like domain containing proteins predominantly expressed in skeletal muscle. To reveal possible roles for Igfn1, we applied non-selective knock-down by shRNAs as well as specific targeting of Igfn1 exon 13 by CRISPR/Cas9 mutagenesis in C2C12 cells. Decreased expression of Igfn1 variants via shRNAs against the common 3’-UTR region caused a total blunting of myoblast fusion, but did not prevent expression of differentiation markers. Targeting of N-terminal domains by elimination of exon 13 via CRISPR/Cas9 mediated homologous recombination, also resulted in fusion defects as well as large multinucleated cells. Expression of IGFN1_v1 partially rescued fusion and myotube morphology in the Igfn1 exon 13 knock-out cell line, indicating a role for this variant in myoblast fusion and differentiation. However, in vivo overexpression of IGFN1_v1 or the Igfn1 Exon 13 CRISPR/Cas9 targeting vector did not result in significant size changes in transfected fibres.

Partial Text

Igfn1 was identified as a protein fragment in a Yeast-two-hybrid assay using the KY protein as bait [1]. Mutations in KY underlie muscle disease in mouse and humans and a blunted hypertrophic response in mouse adult skeletal muscle [2][3][4]. Igfn1 defines a genomic locus of complex transcriptional activity that produces multiple proteins predominantly in skeletal muscle and heart [5]. The largest IGFN1 protein isoform (referred to as IGFN1) contains a series of 11 fibronectin and immunoglobulin-like domains distributed in three N- and eight C-terminal domains separated by a large disordered segment. The large disordered segment is alternatively spliced in the smaller IGFN1_v1 isoform [5]. Full length cDNAs have previously been cloned that code for isoforms containing subsets of C-terminal domains only [5]. The domain composition of IGFN1 is reminiscent of other sarcomeric proteins associated with the actin cytoskeleton (e.g., myosin binding protein C, filamin C, myotilin, myopalladin or titin, see [6] for a review). These proteins are proposed to act as crosslinkers and bear an inherent flexible structure that allows them to maintain protein interactions through cycles of contraction and relaxation [7].

Igfn1 is a large and complex locus and therefore challenging to analyze. Here, we attempted to assign an overall role to the major isoforms using the myoblast C2C12 cell line, as this is a well-established cellular model of in vitro muscle differentiation. In our hands, C2C12 cells from different sources or passages were often inconsistent in the paucity of differentiation. To minimize this potentially confounding effect, elimination of IGFN1 isoforms was attempted by different shRNAs and individual clones as well as mass selection was undertaken. Partial but evident knock-down of putative isoforms including IGFN1_v1 was obtained in several independently generated cell lines. These cell lines showed a total lack of fusion. Remarkably, the fusion defect did not prevent expression of alpha-actinin and the formation of a primitive sarcomeric-like pattern in single cells, suggesting that myoblast fusion and differentiation respond to independent regulatory cues. The fusion defect was not rescued by full length IGFN1 or IGFN1_v1 in transiently transfected cells or stably selected cell lines (data not shown), suggesting that neither IGFN1 or IGFN1_v1 are sufficient for myoblast fusion. Therefore, the possibility of fusion defects in the various knock-down cell lines being the result of tissue culture artefacts could not be strictly ruled out. A more direct strategy was then adopted by generating a number of C2C12 derived clones carrying deletions or mutations in Igfn1 exon 13, which encodes most of the second globular domain. It was anticipated that this deletion will not eliminate all isoforms, but those containing the full set of N-terminal globular domains. The differentiation patterns observed in these cell lines were remarkably consistent, producing mainly singly nucleated actin rich cells but also multinucleated structures that appeared more voluminous.