Targeted deletion of titin N2B region leads to diastolic dysfunction and cardiac atrophy

Radke et al. 10.1073/pnas.0608543104.

Supporting Information

Files in this Data Supplement:

SI Figure 7
SI Figure 8
SI Figure 9
SI Methods

SI Figure 7

Fig. 7. Isoform shift upon deletion of the N2B region. Although total titin levels are normal (data not shown), the loss of the titin N2B region leads to compensatory up-regulation of the larger N2BA isoform KO (-/-) mice (P < 0.01, n = 6 per genotype).

SI Figure 8

Fig. 8. Deletion of the N2B region results in decreased ventricular weight (vw) to body weight (bw) ratios. Results are derived from age- and sex-matched adult mice (3-6 months old). +/+, n = 13 and -/-, n = 16 animals. Differences are significant at P = 0.013.

SI Figure 9

Fig. 9. N2B-deficient hearts have increased diastolic pressures. (A) Examples of P - V relation in WT (Left) and KO mice (Right). The curves correspond to Fig. 4. Diastolic pressure is shown by squares and developed pressure by circles. (B) Mean ± SEM of WT (light gray) and KO (dark gray) mice (six animals each). Results are shown at V_eq + 15%, under baseline conditions (BL) and in the presence of either dobutamine (Dob) or propranolol (Prop). Under all conditions, diastolic pressure is increased in KO mice (P < 0.01). (C and D) Developed pressure at V_eq and V_eq + 15%. Developed pressure is largely unchanged in KO versus WT animals.

SI Methods

Generation of Titin N2B-Knockout (KO) Mice (Titin N2B-^/-). A targeting construct was assembled by standard procedures from a mouse genomic BAC clone (bacterial artificial chromosome library MGS1 from mouse ES cells; Genomic Health/Incyte Genomics, Palo Alto, CA) spanning the 5' region of the mouse titin gene. Primers were designed from sequence information available in the Celera database and GenBank (GenBank accession no. AL928789). Briefly, a PCR-based strategy was used to replace exon 49 (N2B exon) with a Flp recombinase target (FRT)-flanked neomycin expression cassette. All exons included in the targeting vector were sequence-verified.

Homologous recombination in ES cells was used to derive two independent ES cell clones (targeting frequency ~1:350), which both transmitted the mutant allele through the germ line to generate titin N2B (+/neo) animals. To remove the neomycin cassette, which has the potential to influence the phenotype of targeted animals, we expressed the Flp recombinase in the germ line of neo-positive animals using the Flp deleter stain. Offspring lacking the neomycin cassette were used to obtain Flp-negative, homozygous N2B-deficient animals (N2B-^/-). In these animals, splicing around the N2B region proceeds normally and leads to excision of the intronic FRT site. The deletion of exon 49 leads to splicing of exons 48-50, removing 2,646 bp (exactly 882 aa), which maintains the ORF as shown using antibodies against M-line titin and the region C-terminal of the N2B region (N2Buc-IG 26/27 with Ile-27 at exon 50; Figs. 2 and 5). Animals were maintained on a mixed 129SVJ/C57black6 background using a heterozygous to heterozygous mating scheme to derive the appropriate littermate controls. Phenotypes were comparable in the two independent N2B-deficient strains.

Genotyping. Template DNA was prepared from ES cells (to test for homologous recombination) or postnatal mouse tail (for colony genotyping) according to standard procedures. Homologous recombination in ES cells was monitored by PCR (primers P49-K, GTGCTGGGATCAAAGCCATGTGC; and 3'neo, TCGACTAGAGGATCAGCTTGGGCTG).

For Southern blot analysis, genomic DNA was digested with EcoRI and BamHI overnight and probed with a PCR product generated with primer N2B-P forward (TCTTAAGGAGCAGATACGCAGAC) and N2B-P reverse (CTTGCATCTATAGTGTACCCGCT) (Fig. 1A) following standard procedures. For genotyping of the mouse colony, the primers P49-5 (AATCTCACCACAACCTTATTCCA), P49-I (GGTTAACAGCATCCCATTAAAGA), and P49-3 (AGTGAATTGCGGGGAAATTATTA) were used to discriminate between the wild-type (WT) and KO allele.

Echocardiography. After anesthesia induction with 2% ISOFLO (Abott, Abott Park, IL) in a Univentor 400 anesthesia chamber, the mice were placed in dorsal recumbance on a heated, tilt platform for echocardiography. Body temperature was maintained at 37°C, and anesthesia was continued with 0.5-1.5% ISOFLO. The echocardiograph system used (Vevo 770, VisualSonics, Toronto, ON, Canada) employs a real-time microvisualization scan head for spatial resolution down to 30 mm. The 45-MHz transducer was mounted on an integrated rail system, and the mice were positioned using the tilt platform. A standoff was created for the ultrasound transducer using conductivity gel. Care was taken to avoid animal contact and excessive pressure which can induce bradycardia. Imaging was performed at a depth setting of 1 cm. Images were collected and stored as a digital cine loop for off-line calculations. Standard imaging planes, M-mode, Doppler, and functional calculations were obtained according to American Society of Echocardiography guidelines. The parasternal long-axis four-chamber view of the left ventricle (LV) was used to guide calculations of percentage fractional shortening, percentage ejection fraction, and ventricular dimensions and volumes. The subcostal long-axis view from the left apex was used to obtain mitral inflow and aortic ejection profiles by pulsed-wave Doppler imaging. For pulsed-wave Doppler recording, a sample size of 0.3 mm was used. A sweep speed of 100 mm/s was used for M-mode and Doppler studies.

Isolated Heart Experiments. An isolated heart setup was used to determine the developed and diastolic pressure to volume relationship (P - V) in hearts from N2B WT and KO mice. Approximately 6-month-old littermate animals were anesthetized (60 mg/kg sodium pentobarbitone, i.p.) and heparinized (1,000 units/kg, i.p.) followed by rapid removal of the heart and cannulation of the aorta with a blunted 17-gauge needle for retrograde coronary perfusion with oxygenated Krebs solution at a constant pressure of 80 mmHg (1 mmHg = 133 Pa) and a temperature of 37°C. A thin-walled latex balloon was filled with degassed water until passive pressure reached 5 mmHg. Pressure was measured with a micromanometer-tipped catheter introduced into the center of the balloon. Hearts were field-stimulated at an interbeat interval of 250 ms.

Single-beat analysis of LV function was performed by changing LV filling V from 90% to 125% of baseline V_BL (V that results in passive pressure of ~5 mmHg) in 5% increments to generate Frank-Starling curves, and records were collected for the full set of eight commanded V. The volume range was limited to ~4.5 ml because larger volumes resulted in irreversible damage to the heart with reduced developed pressures. Pressures were measured during test beats imposed after the heart had beaten isovolumically for 30 s at V_BL to allow sufficient time for the preparation to stabilize fully. Passive pressure (P_p) was measured as the lowest LV pressure at the end of the test beat. Peak systolic pressure was measured from the test beat as well, and developed pressure (P_d) was calculated as peak systolic pressure minus passive pressure.

To account for possible changes in geometry, P was converted to wall stress (s) by using a thick-walled spherical model: s = P/[(V_w/V + 1)2/3 -1], where V_w is the volume of LV wall (equal to LV weight /1.05). The obtained s- V relation was used to determine s at V_equ (V at zero passive P) and 1.15 V_equ.

The Frank-Starling protocol was run first at baseline, then the response to b-adrenergic stimulation (0.2 mM dobutamine) and b-adrenergic blockade (0.1 mM propranolol) was determined. Stable responses to dobutamine and propranolol were achieved after perfusion for 5 min each.

Immunoelectron Microscopy. Mouse hearts were excised, and the atria and right ventricle were removed. The LV free wall and septum were separated, split base to apex, and skinned for 2 h in skinning solution (relaxing solution plus 1% Triton X-100 and 3´ leupeptin and E-64 inhibitors and 1´ PMSF.) Smaller muscle bundles were then dissected further and skinned overnight at 4°C. The skinning solution was rinsed out with multiple changes of relaxing solution plus inhibitors as above. Small muscle strips (0.3- to 0.6-mm diameter ´ 3 mm long) were prepared from LV papillary and endocardium and subsequently stretched to varying lengths (from near slack length up to 150% of slack length) by pinning to a Sylgard (Dow Corning Corporation, Midland, MI) surface on the bottom of ~100-ml wells. The strips were then lightly fixed with 0.3% paraformaldehyde in PBS for 30 min at 4°C. Immunodetection and electron microscopy were performed as described previously (1) using rabbit polyclonal (UC, UN, I84, MIR; courtesy of Dr. Siegfried Labeit) and mouse monoclonal (T12; 1248-634; Roche, Indianapolis, IN) primary anti-titin antibodies. Epitope distances with respect to the Z-line were measured from scanned negatives of electron micrographs using a user-developed macro for Scion Image software (version 1.6, based on NIH Image; National Institutes of Health, Bethesda, MD).

1. Trombitas K, Freiburg A, Centner T, Labeit S, Granzier H (1999) Biophys J 77:3189-3196.