Novel homozygous nonsense mutation of MLIP and compensatory alternative splicing

Abstract

Despite the growing accessibility of clinical sequencing, functional interpretation of variants remains a major hurdle to molecular diagnostics of Mendelian diseases. We aimed to describe a new adult-onset myopathy with muscle weakness and hyperCKemia caused by a nonsense variant in muscular LMNA-interacting protein (MLIP). Following RNA-sequencing, differential expression analysis uncovered a significant downregulation of this gene, which had a surprisingly mild effect on MLIP protein expression. RT-PCR and long-read sequencing (LRS) both support an important transcriptome shift in the patient, where decreased MLIP levels are seemingly due to nonsense-mediated decay of transcripts containing the exon 5 mutation. Moreover, a compensatory mechanism upregulates the functionally lacking isoforms and generates novel transcripts. These results support the recently discovered clinical implications of MLIP variants in myopathies, highlighting for the first time its relevance in adult-onset cases. These results also underline the power of LRS as a tool for the functional assessment of variants of unknown significance (VUS), as well as the definition of accurate isoform profile annotations in a tissue-specific manner.

Fig. 1. MLIP and LMNA respective expression in muscular tissue.

a Volcano Plot highlighting the most differentially expressed genes in Z46 compared to controls following DESeq analysis. b RT-qPCR reveals mRNA expression levels for the targeted genes: MLIP 5–6 accounts specifically for transcripts containing the nonsense mutation, MLIP 9–10 covers all the known transcripts, and LMNA probe accounts for both lamin A and C. RPS29 is used for data normalization, error bars correspond to standard deviation (triplicata). c, d Western Blot analysis reveals the presence of four isoforms of MLIP, two of which can be quantified (27 kDa and 50 kDa). Both lamin types also appear for LMNA and were analyzed separately. GAPDH is used as the reference gene for normalization of the blot in both instances.

Fig. 2. Evaluation of nonsense variant effect on transcript balance of MLIP.

a Migration of RT-PCR products on 2% E-gel following amplification between exons 3 and 8. 1 Kb Plus E-gel DNA ladder was used for determination of fragment lengths. P = patient, C = Control. b Qualitative visualization of 100 ng of sequencing library on a 1% agarose gel with GelRed revealing agent. Migration was done prior to the end prep step, and 1 Kb Plus DNA Ladder served for amplicons length evaluation. c Visual representation of major (high confidence) muscle transcripts identified with LRS data. Corresponding annotated transcripts are identified (exons 3-11), and all isoforms are color-coded depending on grouping features. d Representation of the samples’ relative transcripts balance, grouped by relevant features. Isoform quantification was performed by FLAIR, and final data contains 82.9k (patient) and 52.3k (control) reads.