Progeria-based vascular model identifies networks associated with cardiovascular aging and disease

Abstract

Hutchinson-Gilford Progeria syndrome (HGPS) is a lethal premature aging disorder caused by a de novo heterozygous mutation that leads to the accumulation of a splicing isoform of Lamin A termed progerin. Progerin expression deregulates the organization of the nuclear lamina and the epigenetic landscape. Progerin has also been observed to accumulate at low levels during normal aging in cardiovascular cells of adults that do not carry genetic mutations linked with HGPS. Therefore, the molecular mechanisms that lead to vascular dysfunction in HGPS may also play a role in vascular aging-associated diseases, such as myocardial infarction and stroke. Here, we show that HGPS patient-derived vascular smooth muscle cells (VSMCs) recapitulate HGPS molecular hallmarks. Transcriptional profiling revealed cardiovascular disease remodeling and reactive oxidative stress response activation in HGPS VSMCs. Proteomic analyses identified abnormal acetylation programs in HGPS VSMC replication fork complexes, resulting in reduced H4K16 acetylation. Analysis of acetylation kinetics revealed both upregulation of K16 deacetylation and downregulation of K16 acetylation. This correlates with abnormal accumulation of error-prone nonhomologous end joining (NHEJ) repair proteins on newly replicated chromatin. The knockdown of the histone acetyltransferase MOF recapitulates preferential engagement of NHEJ repair activity in control VSMCs. Additionally, we find that primary donor-derived coronary artery vascular smooth muscle cells from aged individuals show similar defects to HGPS VSMCs, including loss of H4K16 acetylation. Altogether, we provide insight into the molecular mechanisms underlying vascular complications associated with HGPS patients and normative aging.

Keywords: DNA damage repair; H4K16 acetylation; Lamin A; aging; coronary artery; induced pluripotent stem cells; progeria; reprogramming; vascular smooth muscle cells.

FIGURE 1

HGPS VSMCs exhibit transcriptomic dysregulation consistent with vascular aging. (a) A schematic of the directed differentiation approach adopted to generate VSMCs from iPSCs through embryoid body (EB) formation. (b) Hierarchical clustering of control [early passage Control1 P7, Control3 P7, and late passage Control1 P14 and Control3 P14] and HGPS VSMC samples [early passage HGPS1 P7, HGPS3 P7, and late passage HGPS1 P14 and HGPS3 P14]. Samples were hierarchically clustered based on their normalized gene expression values. (c) Top panel, gene ontology (GO) term enrichment analysis of HGPS VSMCs vs control at passage 7 (P7) using GOseq. GO terms with p‐values <0.001 were considered significantly enriched for differentially expressed genes. The x‐axis represents the number of genes that are differentially expressed, and the y‐axis represents the enriched GO terms. The bottom panel shows KEGG pathway enrichment analysis of HGPS VSMCs vs control at passage 7. KEGG pathways are significantly enriched with p‐values <0.05. (d) Left panel shows enriched GO terms of HGPS VSMCs vs control at passage 14 (P14). The right panel shows their enriched KEGG pathways. (e) Left panel shows enriched GO terms of the control VSMCs at passage 14 vs passage 7. The right panel shows the enriched KEGG pathways. All experiments represent two technical replicates of at least 2 cell lines (biological replicates) for each genotype.

FIGURE 2

HGPS VSMCs exhibit increased oxidative and DNA replicative stress. (a) HGPS VSMCs exhibit a higher level of ROS. Scale bar = 25 μm. (b) HGPS VSMCs contain significantly higher level of 8‐OxoG compared to control at late passage. Scale bar = 10 μm. (c) DSBs assessed via analysis using alkaline comet assay. (d) HGPS VSMCs exhibit an increased level of replicative stress at late passage as quantified by percentage of pRPA32 (S4/8) positive cells. Scale bar = 25 μm. (e) HGPS VSMCs exhibit increased colocalization between BrdU and γH2A.X in the nucleus measured using the proximity ligation assay and a markedly higher level of DNA damage γH2A.X foci (f). Scale bar = 10 μm. NAC treatment also decreased cellular ROS levels (g) (Scale bar = 25 μm) as well as γH2A.X foci (h) (Scale bar = 10 μm) in HGPS VSMCs. Calculated values represent mean ± SEM (n = 6, 2x for 3 cell lines per test group at each passage). All values represent mean ± SEM (n = 4, >2x for each cell line per test group).

FIGURE 3

HGPS VMSCs exhibit abnormal engagement with NHEJ repair on replicating DNA. (a) Metascape‐generated heat map of GO enriched terms of differentially enriched proteins from aniPOND‐mass spectrometry analysis from both passage 7 (left) and passage 14 (right). Color indicates p‐values of the GO terms. (b) STRING interactome analysis of differentially abundant proteins from aniPOND‐mass spectrometry analysis from passage 7 (left) and passage 14 (right). The confidence level was set to 0.5 (medium). The thickness of the blue lines connecting various proteins represents the level of confidence with which functional partners can be predicted. (c) Quantification of protein abundance (spectra count) of proteins involved in NHEJ repair in control and HGPS VMSCs at both passages from aniPOND‐mass spectrometry analysis. (d) Distribution of 53BP1 foci in control and HGPS VSMCs. The circles are highlighting the G1 cells in the right panel and the presumed S/G2 cells in the left panel. Scale bar = 25 μm. (e) Representative images of proximity ligation assay staining between BrdU and 53BP1 in control and HGPS VSMCs. Red foci represent interactions between BrdU and 53BP1. Bar graphs representing average count of red dots between control and HGPS VSMCs at early and late passages as analyzed by ImageJ. Values represent mean ± SEM (n = 2, 2 cell lines per passage per test group). Scale bar = 10 μm. (f) Violin plots represent assembly of 53BP1 and BRCA1 foci that arise at passage 7, and the generation of 53BP1 and BRCA1 foci between passage 10 and 20 control VSMCs (g).

FIGURE 4

HGPS VSMCs exhibit altered histone acetylation. (a) Representative immunofluorescence images of Progerin and H4K16ac expression in control and HGPS VSMCs. (Right) Bar graphs representing average immunofluorescence intensity of H4K16ac expression between control and HPGS VSMCs at early and late passages. Scale bar = 25 μm. Values represent mean ± SEM (n = 4, 2x per cell line, 2 cell lines per passage per test group at each passage). (b) Correlation plot from high‐content immunofluorescence imaging analysis between Progerin and H4K16ac abundance. (c) (Left) Representative images of proximity ligation assay between BrdU and H4K16ac in control and HGPS VSMCs. Red dots represent positive ligation between BrdU and H4K16ac. (Right) Bar graphs representing average count of red dots by ImageJ showing colocalization of H4K16ac and BrdU between control and HPGS VSMCs at early and late passages. Scale bar = 10 μm. Values represent mean ± SEM (n = 2, 2 cell lines per passage per test group). (d) (Left) Representative confocal images of immunofluorescence staining for Progerin (red) and MOF (green) proteins in control and HGPS VSMCs. (Right) Bar graphs representing average fluorescence intensity of MOF in both control and HGPS VSMCs at early and late passages. Scale bar = 25 μm. Values represent mean fluorescence intensity ± SEM (n = 2, 2 iPSC cell lines per passage per test group). (e) Correlation scatter plot showing immunofluorescence imaging analysis between Progerin and MOF expression in G2/M phase. Red trend line shows inverse correlation. (f) (Left) Representative images of proximity ligation assay between Lamin A/C and MOF in control and HGPS VSMCs. Red dot represents positive ligation between Lamin A/C and MOF. (Right) Bar graphs representing average count of red dots counted by ImageJ analysis showing colocalization of Lamin A/C and MOF in both control and HGPS VSMCs at early and late passages. Scale bar = 10 μm. Values represent mean ± SEM (n= > 2, >2 cell lines per passage per test group). (g) (Left) Representative images of proximity ligation assay between BrdU and MOF in control and HGPS VSMCs. Red dot represents positive ligation between BrdU and MOF. (Right) Bar graphs representing average count of red dots by ImageJ showing colocalization of MOF and BrdU between control and HPGS VSMCs at early and late passages. Scale bar = 10 μm. Values represent mean ± SEM (n= > 2, >2 cell lines per passage per test group).

FIGURE 5

Decreased MOF causes decreased H4K16ac and increased NHEJ events in HGPS VSMCs. (a) Time‐dependent changes in H4K16ac were measured by treating cells with an HDAC inhibitor for 4 h to acetylate H4K16 and then changed the media to study the rate of deacetylation (K‐Off) over different time intervals (0, 5, 10, 15, 30, 60, 90, 120 min). (b) The rate of acetylation (K‐On) was measured by treating cells with the HDAC inhibitor at different intervals (0, 5, 10, 15, 30, 60, 90, 120 min) and observed H4K16 acetylation. Values represent integrated intensity ± SD (n > 2, 3 times per cell line). (c) Relative MOF mRNA expression compared in two control VSMCs to scrambled control following MOF partial knockdown using DsiRNA knockdown strategy. Values represent average expression ± SD (n = 2, 2 times per cell line). (d) High‐content imaging fluorescence quantification demonstrating a decrease in H4K16ac expression following MOF knockdown. Values represent average fluorescence intensity ± SD (n = 2, 2 times per cell line). (e) High‐content imaging fluorescence quantification demonstrating an increase in 53BP1 expression following MOF knockdown. Values represent average fluorescence intensity ± SD (n = 2, 2 times per cell line). (f) Treatment of HGPS VSMCs with HDAC inhibitor TSA leads to increased levels of H4K16ac compared to untreated.

FIGURE 6

Patient‐derived CAVSMCs exhibit features of HGPS VSMC defects. (a) Quantification of histone mark, H4K16ac in control VSMCs, HGPS VSMCs and CAVSMCs (left). The experiments represent at least two technical replicates of at least 2 cell lines (biological replicates) for each genotype, while values represent mean ± SEM. (Right) Correlation plot from high‐content immunofluorescence imaging analysis between Progerin and H4K16ac expression in patient CAVSMCs. Red trend line shows inverse correlation. (b) Primary CAVSMCs from aged donors express typical VSMC markers such as ACTA2, and SM22alpha. Scale bar = 25 μm. (c) Representative images (left), quantification (middle) of primary aged CAVSMCs expressing Progerin compared to iPSC‐derived control and HGPS VSMCs. (Right) Progerin transcript levels assessed by qPCR. Scale bar = 10 μm. (d) High‐content scanning images of Lamin A/C staining demonstrate that both HGPS VSMCs and primary aged CAVSMCs display increased incidence of abnormal nuclear morphology compared to control VSMCs. Scale bar = 10 μm. (e) Primary aged CAVSMCs show increased level of DNA damage γH2A.X foci. Scale bar = 25 μm. (f) Primary aged CAVSMCs exhibit elevated levels of FGF2. Scale bar = 25 μm. The experiments represent at least two technical replicates of at least 2 cells lines (biological replicates) from organ donors (>50 years old), while values represent mean ± SEM.