Truncated titin is structurally integrated into the human dilated cardiomyopathic sarcomere

Heterozygous (HET) truncating variant mutations in the TTN gene (TTNtvs), encoding the giant titin protein, are the most common genetic cause of dilated cardiomyopathy (DCM). However, the molecular mechanisms by which TTNtv mutations induce DCM are controversial. Here, we studied 127 clinically identified DCM human cardiac samples with next-generation sequencing (NGS), high-resolution gel electrophoresis, Western blot analysis, and super-resolution microscopy in order to dissect the structural and functional consequences of TTNtv mutations. The occurrence of TTNtv was found to be 15% in the DCM cohort. Truncated titin proteins matching, by molecular weight, the gene sequence predictions were detected in the majority of the TTNtv+ samples. Full-length titin was reduced in TTNtv+ compared with TTNtv– samples. Proteomics analysis of washed myofibrils and stimulated emission depletion (STED) super-resolution microscopy of myocardial sarcomeres labeled with sequence-specific anti-titin antibodies revealed that truncated titin was structurally integrated into the sarcomere. Sarcomere length–dependent anti–titin epitope position, shape, and intensity analyses pointed at possible structural defects in the I/A junction and the M-band of TTNtv+ sarcomeres, which probably contribute, possibly via faulty mechanosensor function, to the development of manifest DCM.


Figure S1
. Western blot analysis on 18 of the DCM TTNtv+ samples differentiated the truncated titin from T2.Upper and lower panels show the samples labeled with the T12 and M8M10 antititin antibodies, respectively.Red arrows indicate the gel bands corresponding to truncated titin.Note that the penetrance of the truncated titin is uneven, and the truncated titin was not detectable in all TTNtv+ DCM samples.However, in samples #190, #213 and #226 the Western blot analysis detected truncated titin protein, even though the gel analysis (SYPRO Ruby staining) failed to do so (see Supplementary Table S6).Densitometric analysis of the Western blot and its comparison with the gel data are shown in Table S16.  ) and TTNtv+ (patients #75, 94 and #238) myofibrils and their supernatants obtained in three subsequent washes.Supernatants were concentrated (Amicon Ultra-0.5 Filter, Merck Millipore, Burlington, MA) following each washing and centrifugation (2500 rpm, room temperature) step.The "pellet" lanes contain myofibril pellet:urea buffer ratios of 1:3, except in the last lane where this ratio is 1:1.ii.Western blot of the samples shown in i, using anti-N-terminus (T12, top image) and anti-C-terminus (M8M10, bottom image) anti-titin antibodies.iii.Overlay of the Western blots shown in ii.Yellow-green and red colors correspond to the T12 and M8M10 antibodies, respectively.Lane labels: m: muscle; myo: myofibril; sn1, sn2, sn3: 1x, 2x, 3x wash supernatants.The samples were from the TTN tv-#7 patient.Two separate electrophoreses were carried out at different times (years 2021 and 2023) and slightly different gel densities (1% and 0.8%) for the muscle sample (i), but in this case only one electrophoresis was carried out for the washed myofibril sample (ii).Although small variations in the absolute optical densities of the protein bands may be observed, the peak ratios are conserved.The amount of the degradation product T2 is increased in myofibrils, which is a recurring finding (see below for the other samples as well).MyHc: myosin heavy chain.Although small variations in the absolute optical densities of the protein bands may be observed, the peak ratios are conserved.The amount of the degradation product T2 is increased in myofibrils; however, even though the truncted protein has a lower density (red arrow), it remains well detectable (see also Western blot results in Figure S2B above).Labels: MyHc: myosin heavy chain, red arrows: truncated titin.Although small variations in the absolute optical densities of the protein bands may be observed, the peak ratios are conserved.The amount of the degradation product T2 is increased in myofibrils; however, even though the truncted protein has a lower density (red arrow), it remains well detectable (see also Western blot results in Figure S2B above).Labels: MyHc: myosin heavy chain, red arrows: truncated titin.S13.
Figure S4.Normalized width distribution of the MIR epitopes in short sarcomeres with lengths between 1.7 and 2.0 μm.The comparisons are made between DCM TTNtv-and DCM TTNtv+ (A) and between DCM TTNtv-and normal control (CTRL, papillary muscle samples) (B).Note that only the full-length titin of the DCM TTNtv+ sarcomeres contain the A170 epitope.For statistics, see Table S14.S6.Table S3.Bioinformatics overview of the pathogenic variants in the DCM samples (see file DCM_MS_SI-table_2.xlsx).
Table S4.Sequence data of the pathogenic variants in the DCM cohort (see file DCM_MS_SI-table_3.xlsx).
The integrated densities (quantities) of the relevant protein bands and their ratios are indicated.
Table S7.Statistical analysis of the A-band titin length as a function of sarcomere length (analysis of data presented graphically in Figure 6E).The values missing from the regression analysis are the data points that fall below a sarcomere length of 1.8 µm.The slope has a unit of µm/µm, therefore it is dimensionless.Table S8.Statistical comparison of A-band titin length as a function of sarcomere length in negative control and DCM TTNtv-samples.A-band titin length was measured as the distance between MIR epitopes across the A-band.There is no significant difference in the slopes of the linear fit, but there is a significant, ~44 nm difference in the Y-axis intercepts (corresponding graph is shown in Figure 6F).Table S9.Statistical analysis of the M-band-to-A170 epitope distance as a function of sarcomere length (analysis of data graphically presented in Figure 6H).The values missing from the regression analysis are the data points that fall below a sarcomere length of 1.8 µm.
The slope has a unit of nm/µm.The relative extensibility (Erel) of different A-band titin segments (A-band section, titin kinase region) were calculated as where S is the slope of the sarcomere-length-dependent epitope-to-epitope distance (see Figures 6E-I and Tables S7-S10), and N is a factor that normalizes to the slack lengths of the respective titin segments as where Slackx is the respective epitope-to-epitope distance at slack (1.8 µm sarcomere length) and Slackkinase is the distance of the M-line to the A170 epitope at a sarcomere length of 1.8 µm.The calculated Erel values for the A-band section and kinase region of titin in DCM TTNtv- were 10 nm/µm and 23 nm/µm, respectively.
Table S11.Statistical analysis of MIR epitope intensity-profile width normalized to peak height of the MIR epitope (analysis of data presented graphically in Figures 7A and S4A).
Epitope width is the full width at half maximum (FWHM) intensity.Data were compared by one-way ANOVA.ANOVA summary (top) and multiple comparison (bottom) tables are shown.Note that in this analysis groups A, B and C refer to DCM TTNtv-, CTRL and DCM TTNtv+ , respectively.Data are distributed non-normally in each dataset.Even though the mean MIR width is largest in the DCM TTNtv+ sample and is significantly greater than that in DCM TTNtv-, the difference is not significant with respect to CTRL.We attribute this finding to systematic differences in the properties and handling of the negative control tissue sample (papillary muscle, surgical removal, non-standard fixation and incubation procedures).Table S12.Statistical analysis of STED microscopic measurements of A170 epitope integrated intensity normalized to the MIR epitope peak intensity (analysis of data presented graphically in Figure 7B).Epitope intensity was measured as the area under the curve in the intensity profile.Data were compared by one-way ANOVA.ANOVA summary (top) and multiple comparison (bottom) tables are shown.Table S13.Statistical analysis of I-band titin extension as a function of sarcomere length (analysis of data presented graphically in Figure S3).The values missing from the regression analysis are the data points that fall below a sarcomere length of 1.8 µm.The slope has a unit of µm/µm, therefore it is dimensionless.The two slopes are significantly different (p=0.0001).

Figure S2C .
Figure S2C.High-resolution agarose gel electrophoretograms and corresponding optical density plot profiles of left ventricle (LV) muscle samples (i) and washed LV myofibrils (ii).The samples were from the TTN tv-#7 patient.Two separate electrophoreses were carried out at different times (years 2021 and 2023) and slightly different gel densities (1% and 0.8%) for the muscle sample (i), but in this case only one electrophoresis was carried out for the washed myofibril sample (ii).Although small variations in the absolute optical densities of the protein bands may be observed, the peak ratios are conserved.The amount of the degradation product T2 is increased in myofibrils, which is a recurring finding (see below for the other samples as well).MyHc: myosin heavy chain.

Figure S2D .Figure S2E .
Figure S2D.High-resolution agarose gel electrophoretograms and corresponding optical density plot profiles of left ventricle (LV) muscle samples (i) and washed LV myofibrils (ii).The samples were from the TTN tv+ #75 patient.Two separate electrophoreses were carried out at different times (years 2021 and 2023) and slightly different gel densities (1% and 0.8%).Although small variations in the absolute optical densities of the protein bands may be observed, the peak ratios are conserved.The amount of the degradation product T2 is increased in myofibrils; however, even though the truncted protein has somewhat lower density (red arrow), it remains well detectable (see also Western blot results in FigureS2Babove).Labels: MyHc: myosin heavy chain, red arrows: truncated titin.

Figure S2F .
Figure S2F.High-resolution agarose gel electrophoretograms and corresponding optical density plot profiles of left ventricle (LV) muscle samples (i) and washed LV myofibrils (ii).The samples were from the TTN tv+ #238 patient.Two separate electrophoreses were carried out at different times (years 2021 and 2023) and slightly different gel densities (1% and 0.8%).Although small variations in the absolute optical densities of the protein bands may be observed, the peak ratios are conserved.The amount of the degradation product T2 is increased in myofibrils; however, even though the truncted protein has a lower density (red arrow), it remains well detectable (see also Western blot results in FigureS2Babove).Labels: MyHc: myosin heavy chain, red arrows: truncated titin.
Figure S3.I-band titin length, measured with STED super-resolution microscopy, as a function of sarcomere length.A and B, I-band width as a function of the sarcomere length in DCM TTNtv-and DCM TTNtv+ cardiac samples, respectively.I-band width was calculated as the distance between two consecutive MIR epitopes outside of the A-band (thus, not separated by an A170 epitope doublet).C, Overlaid data from DCM TTNtv-and DCM TTNtv+ sarcomeres for comparison.D, Regression analysis of the I-band titin length in the 1.8-2.6 μm sarcomere length range, revealing significantly increased slope in DCM TTNtv+ fibers compared to that of the DCM TTNtv-.The results complement those presented in Figure 6E, indicating that I-band titin in TTNtv+ sarcomeres extend more extensively than in TTNtv-ones.For statistics, see TableS13.

Figure
Figure S6.Sample-to-sample variaton of STED data in DCM TTNtv+ data.The location of the titin truncations in the individual samples are indicated in Figure 1 of the main text.A. Histograms of the A-band titin length, measured via the MIR epitope-to-epitope distance.B. Histograms of the M-line to titin kinase distance, measured via the A170 epitope-to-epitope distance.C. Box plots of the A-band titin lengths in the different samples.D. Box plots of the M-line to titin kinase distance in the different samples.In the case of the MIR and A170 data, measurements in the sarcomere length range of 2.1-2.4 and 1.8-2.4µm, respectively, were plotted and compared.Descriptive statistics of the data are shown in TableS15.The sampleto-sample variation in the A-band titin length is insignificant.By contrast, the M-line to titin kinase distance is significantly greater in sample #78 than in samples #75 and #140, in spite of their similar physical location along the titin gene.E. Mean M-line to titin kinase distance as a function of the truncated titin per T1 ratio.Error bars represent standard deviation.The truncated titin per T1 ratios, in samples where this could be quantified, are listed in TableS6.

Table S1 . Blood test parameters of DCM patients before heart transplantation
(see file.ID: identification number of the transplanted patient, based on the sequence of the transplantation; Htx: heart transplantation; List type TX: transplantation, List type HU: high urgent; GFR: glomerular filtration rate; AST: aspartate transaminase, ALT: alanine transaminase; GGT: gamma-glutamyl transferase; ALP: alkaline phosphatase; WBC: white blood cell; CRP: c-reactive protein; LVAD: left ventricular assist device; ECMO: extracorporal membrane oxygenation.

Table S10 .
Statistical comparison of M-line to titin kinase distance as a function of sarcomere length in negative control and DCM TTNtv-samples.The distance data were obtained by halving the distance between vicinal A170 epitopes.The slope of the CTRL data is not significantly different from zero (corresponding graph is shown in Figure6I.

Table S14 .
Statistical analysis of STED microscopic measurements of normalized A170 epitope intensity-profile width (analysis of data presented graphically in FigureS5).Data were compared by one-way ANOVA.ANOVA summary (top) and multiple comparison (bottom) tables are shown.

Table S15 .
Descriptive statistics of STED data from different DCM TTNtv+ samples.Histograms and box plots of the data are shown in Figure S6.

Table S16 .
Statistical comparison of truncated titin quantities detected with gel electrophoresis and Western blotting.Upper part shows the densitometry results of the Western blot (FigureS1), lower part shows the comparison with the gel densitometry data.