Gene expression profiling of fibroblasts in a family with LMNA-related cardiomyopathy reveals molecular pathways implicated in disease pathogenesis

Figures

• 

  Fig. 1. Gene expression profiling of fibroblast samples by RNA-seq. Differential gene expression analysis was performed using Tophat and Cufflinks software and visualized using cummeRbund. Heatmaps of gene expression profiles are clustered by rows (genes). Heatmaps are shown as pooled samples (left panel) and replicates (right panel)

• 

  Fig. 2. Nuclear lamina-associated gene expression profiling across all samples. a Visualization of Lamin A and Lamin C expression across all samples from RNA-seq data. Lamin A and Lamin C expression obtained from RNA-seq experiments were visualized using heatmaps. Gene expression is shown pooled (left panel) and as replicates (right panel). In general, Lamin A and Lamin C expression are highest for unrelated groups and lowest for patient groups. However, there were no statistically significant differences between Lamin A and Lamin C expression across all samples (FDR adjusted p-value ≥0.05); b RNA-seq validation of Lamin A (LMNA) and Lamin C (LMNC) transcript levels by quantitative PCR (qPCR). qPCR was performed on cDNA generated from unrelated, control and patient fibroblasts to measure Lamin A and Lamin C transcript levels (average QGE +/− standard error of the mean (SEM)). There were no statistically significant differences across all groups in transcript levels for Lamin A [F (2,6)= 0.90, p = 0.46] and for Lamin C [F (2,6)= 1.76, p = 0.25], validating trends observed in RNA-seq. c Western blot for Lamin A/C protein. Lamin A/C protein levels were measured by immunoblotting using…

• 

  Fig. 3. LMNA allelic expression. Top panel depicts location of the LMNA splice-site mutation and two expressed single nucleotide polymorphisms (SNP), rs538089 in Exon 5 and rs4641 in Exon 10. Bottom table shows Sanger sequencing chromatograms from genomic DNA (gDNA) and distribution of allele read counts from RNA-seq. For two Unrelated Controls at rs538089, gDNA was heterozygous (C/T) with equal RNA-seq reads of C (51–52%) and T (47–48%) consistent with biallelic expression. For three Patients at rs4641, gDNA was heterozygous (C/T) with RNA-seq reads predominately of C (97–99%) consistent with monoallelic expression

• 

  Fig. 4. Eight differentially expressed genes are potentially involved in LMNA-related cardiomyopathy. a Differentially expressed genes that passed the filtering criteria were input into Ingenuity Pathway Analysis (IPA). Of these significantly differentially expressed genes for each group, there were eight genes in the intersection (arrow) between B and C, but not A, where A = Unrelated vs. Controls, B = Unrelated vs. Patients, C = Controls vs. Patients; b Gene symbols, names, and RNA-seq values for eight genes found in the intersection between groups B and C, but not A; *Expression in patient: -: decreased, +: increased. c RNA-seq validation of Insulin Growth Factor binding protein 5 (IGFBP5) transcript levels by quantitative PCR (qPCR). qPCR was performed on cDNA generated from unrelated, control and patient fibroblasts to measure IGFBP5 transcript level. There was no statistically significant difference in IGFBP5 transcript levels (average QGE +/− SEM) between all groups [F (2,6)= 0.37, p = 0.71]. Statistical analysis was performed using One-way ANOVA followed by Tukey post hoc test

• 

  Fig. 5. Ingenuity Pathway Analysis (IPA) of network of eight DE genes. The eight genes that were found by the intersection shown in Fig. 3 were input into IPA to generate a network diagram with the most plausible connections to each other, and to additional genes in the IPA Knowledge Base. The eight genes of interest are highlighted in orange. A maximum of thirty-five genes are shown in the network for readability. The ERK/MAPK pathways genes found in this network are highlighted in cyan