Transgenic Remodeling of the Regulatory Myosin Light Chains in the Mammalian Heart

Construction of Transgenic Mice and Analyses of Transgene Expression

For the transgene encoding MLC2f, a full-length murine cDNA was synthesized using RT-PCR with poly(A)⁺ RNA isolated from the leg muscle of FVB/N mice as starting template. The resultant PCR product was sequenced, linked to the α-MyHC promoter, and used to generate transgenic mice (Fig 1⇓). The construct was digested free of vector sequence with Not I, purified from low-melting-point agarose, and used to generate transgenic mice as described previously.²⁰ The founder mice were identified by PCR and confirmed by genomic Southern blots using DNA obtained from tail clips. Stable transgenic lines were generated by breeding the founder mice with nontransgenic littermates. Subsequent offspring were screened by PCR. For the RNA analyses, Northern and dot blots were carried out²¹ using the following transcript-specific probes: ANF, 5′-AATGTGACCAAGCTGCGTGACACACCACAAGGGCTTAGGATCTTTTGCGATCTGCTCAAG; α-cardiac actin, 5′-CGTACAATGACTGATGAGAGATGGGGAGGGGGCTCAGAGGATTCCAAGAAGCACAATAC; α-skeletal actin, 5′-TGGAGCAAAACAGAATGGCTGGCTTTAATGCTTCAAGTTTTCCATTTCCTTTCCACAGGG;MLC2v, 5′-CACAGCCCTGGGATGGAGAGTGGGCTGTGGGTCACCTGAGGCTGTGGTTCAG; MLC2a, 5′-GAGGTGACCTCAGCCTGTCTACTCCTCTTTCTCATCCCCG;ELC1v, 5′-GGCTCAGCTCGCCATGAGATATGCTTCACAAACGCTTCATAGTTGATGCAC; ELC1a, 5′-CACCCTGGAGAAACGTGCTTTACCCAGACATGATGTGCTTGAC; GAPDH, 5′-GGAACATGTAGACCATGTAGTTGAGGTCAATGAAG; and α-MyHC, 5′-CGAACGTTTATTGTGGATTGGCCACAGCGAGGGTCTGCTGGAGAG. All steady state transcript levels were normalized with respect to GAPDH signal intensity after correcting for background. Hybridization signals were quantified on a PhosphorImager (Molecular Dynamics).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Isolation of a murine MLC2f cDNA and transgene construction. The MLC2f cDNA was isolated using RT-PCR and sequenced, an open reading frame was confirmed, and the cDNA was subsequently linked to the α-MyHC promoter, which was also sequenced in its entirety (5443 bp, GenBank accession No. U71441). Only the cDNA sequence is shown in the upper part of the panel, and the methionine initiator codon is bolded. Shown below are the comparisons with the amino acids of the isoforms whose replacements were targeted by transgenic overexpression. The bullets (•) denote identity at the amino acid level. The sequences are aligned so as to maximize homologies, and gaps are indicated (–).

Sarcomeric Protein Analyses

The atrial flaps and left ventricular apex were excised from adult transgenic and nontransgenic littermates. Protein was normally isolated using TriReagent (Molecular Research Center, Inc). Total protein was extracted from the phenol phase and interphase of the RNA extractions after removal of DNA by ethanol precipitation and quantified. In some cases, myofilament protein was extracted as described previously,²² and all washes were collected in order to obtain the entire complement of cardiac proteins. The protein preparations were electrophoresed on a 15% polyacrylamide gel in the presence of 0.1% SDS and stained with colloidal blue (Sigma Chemical Co). Proteins were quantified using NIH Image software (version 1.57).

Cardiomyocyte Isolation and Protein Electrophoresis

Ventricular cardiomyocytes were obtained by enzymatic digestion and mechanical disruption as described previously.²³ The resulting suspensions of cells and cell fragments were centrifuged, and pellets were then resuspended in 0.3% Triton X-100 for 6 minutes to permeabilize sarcolemmal, mitochondrial, and sarcoplasmic reticular membranes. After washing, myocytes were resuspended in relaxing solution (mmol/L): EGTA 7.0, free Mg²⁺ 1, free Mg²⁺-ATP 4, creatine phosphate 14.5, and imidazole 20, pH 7.00, at ionic strength 180 at 4°C until use.

SDS-PAGE and silver staining of cardiomyocyte proteins were performed according to methods described previously^{24 25} with only minor modifications. Myocytes were suspended in 5 to 10 μL of sample buffer containing (mol/L) urea 8, thiourea 2, Tris 0.05 (pH 6.8), and dithiothreitol 0.075, along with 3% SDS and 0.05% bromophenol blue, and heated at 100°C for 3 minutes. Samples were subjected to vertical SDS-PAGE in a Hoefer Tall Mighty Small gel electrophoresis unit (Hoefer) with an 18% acrylamide resolving gel (acrylamide/bis-acrylamide at 200:1) and 4.5% acrylamide stacking gel at 24-mA constant current for 2.5 hours. After 30 minutes of alcohol-acid fixation, the gel was fixed in 5% glutaraldehyde overnight and then silver-stained. The gel was then dried between Mylar sheets and scanned using an image densitometer (Molecular Analyst, BioRad).

Functional Analyses

The Langendorff²⁵ and working heart²⁶ preparations were performed as described previously, with the following modifications. The recording, amplification, and differentiation systems used were the DigiMed Systems analyzers BPA-2000, HPA-200, HPA-210, and LPA-200 from Micro-Med Inc. A Silastic fluid-filled catheter to the left ventricle was used. The venous return line feeding into the left atrium was completely water-jacketed for improved temperature (37.4°C) regulation of the Krebs-Henseleit solution that was returned to the left side of the heart for anterograde perfusion.

The unloaded shortening velocity of the ventricular cardiomyocytes was determined as previously described.²⁷ Working on the stage of an inverted microscope, single ventricular cardiomyocytes were attached with silicone adhesive (Dow Corning) to the active elements of a force transducer (model 403A, Cambridge Technology) and motor (model 6350, Cambridge Technology). After curing of the adhesive, myocytes were transferred to relaxing solution, and sarcomere length was adjusted to 2.3 μm using on-line videomicroscopy. Velocity of unloaded shortening was determined at 15°C in maximally activating Ca²⁺ solution (pCa 4.5) using the slack-test method. After steady tension was reached in maximally activating Ca²⁺ solution, the preparation was rapidly slackened; the time required to take up the imposed slack was measured as the interval between the beginning of the imposed slack length step and the onset of tension redevelopment. Plots of slack length versus duration of unloaded shortening were included in the summary results if slack test data were well fit by a straight line (r≥.95).

The maximum Ca²⁺-activated Mg²⁺-ATPase activity was measured in mouse left ventricular myofibrillar preparations²⁸ by the method of White.²⁹

Construction of Multiple Lines of MLC2f Transgenic Mice

We initiated these analyses in order to explore the feasibility of ectopic replacement of an abundant sarcomeric protein in both cardiac compartments using transgenesis. We chose to replace the cardiac isoforms of MLC2a and MLC2v with the RLC that is expressed in the fast skeletal muscle fibers, MLC2f. No published murine clone or sequence could be found and so, using degenerate oligonucleotides for the other RLCs, RT-PCR was performed on poly(A)⁺ RNA isolated from murine leg skeletal muscle. Multiple clones were generated from the purified PCR fragment and completely sequenced. The identity of a full-length clone was confirmed by comparing it with a preexisting rat clone.³⁰ The rat and mouse sequences are very closely related, with only a single conservative amino acid substitution at (rat)Ala¹¹ to (mouse)Gly¹¹. This change was confirmed by genomic sequencing. The cDNA sequence was then inserted into a plasmid at a site between the α-MyHC promoter, which is capable of high levels of cardiac expression,²⁰ and the human growth hormone polyadenylation signal (Fig 1⇑). The DNA was freed of vector sequences and subsequently used to generate transgenic mice.

We have also isolated and sequenced cDNA clones that encode the murine MLC2a and MLC2v proteins; the comparisons (Fig 1⇑) show that the skeletal isoform is more closely related to the latter. Of particular note is the sequence divergence clustered at the amino termini of MLC2f and MLC2a, which leads to significant differences in the overall charge of this region. Charge differences in this domain have significant effects on force production in striated muscle.^{7 31} The phosphorylatable serines (Ser¹⁵ and Ser¹⁶ in MLC2f) are, however, conserved. Therefore, we reasoned that it should be possible to effect a replacement in both cardiac compartments with no lethal effects, and depending on the differential functionality of the different isoforms, subtle phenotypic changes might present. Multiple transgenic lines were generated, and germ-line transmission was confirmed by analyses of the F1 generations. The copy numbers were determined by standard methods using Southern blot analyses, and five lines having copy numbers of 1, 3, 10, 20, and 34 (lines 2, 90, 75, 6, and 57, respectively) were selected for subsequent studies. Transgene expression was stable both within a line between transgenic littermates and throughout multiple generations (F1 to F8, data not shown).

Expression of a Transgene Encoding an Ectopic Contractile Protein Isoform in the Ventricular and Atrial Compartments

We tested the levels of transgenic overexpression in both compartments at the transcript level. Preliminary analyses using Northern blots showed that the transgenic transcript was the expected size (data not shown), and subsequently, dot blots were used to quantify RNA levels in the atria and ventricles of the transgenic lines (Fig 2A⇓). For each of the five lines tested, there were significant levels of transgenic expression in both cardiac compartments. As was found previously for ectopic MLC2v expression in the atrium,¹⁹ endogenous gene expression, as determined by analysis of candidate transcripts that might undergo compensatory changes in response to the expression of the transgene, was unaffected. No significant changes in the steady state levels of the endogenous RLC transcripts could be detected, nor were molecular markers of hypertrophy (ANF and skeletal actin)^{32 33} upregulated. Even at the highest copy number (Fig 2⇓, ventricle), no effect could be observed on the steady state levels of the endogenous α-MyHC, indicating that no titration effects on the transcriptional apparatus had occurred.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Overexpression of MLC2f RNA in transgenic mice. A, MLC2f RNA overexpression in the atrium and ventricle. RNAs were isolated from the cardiac compartments of five transgenic lines carrying the indicated copy numbers of the transgenic construct, and the relative RNA abundance of the transgene, MLC2f, as well as MLC2a, ELC1a (in the atrium) or MLC2v, ELC1v, α-MyHC, ANF, cardiac actin (C. actin), and skeletal actin (Sk. actin) (in the ventricle) were determined by dot blot analyses as described in “Materials and Methods.” In the transgenic RNAs, MLC2f transcript is significantly overexpressed in both the atria and ventricles with no apparent downregulation of the endogenous (eg, MLC2a or MLC2v) genes. ntg indicates nontransgenic. B, Quantification of cardiac transcripts. RNA blot intensities were quantified on a PhosphorImager and corrected for any differences in loading by comparison with the GAPDH signals, and the values in arbitrary units were plotted against transgene copy number. Copy number dependence of steady state transcript levels is apparent in both cardiac compartments.

A definite dose-response effect was observed between the transgenic lines; MLC2f transcript levels increased with increasing copy number (Fig 2B⇑). Previously, the number of lines available for analyses were insufficient to conclude that, in general, the α-MyHC promoter yielded transgenic lines that were subject to copy number effects.^{19 20 34 35} Although there were subtle modulations in the trends between the two cardiac compartments, the relationship between increasing copy number and higher levels of MLC2f RNA in both compartments is clear (Fig 2⇑).

Replacement of Atrial and Ventricular RLCs

The sarcomeric and total protein pools in the transgenic hearts were analyzed by electrophoresis to examine the effects of transgene expression on the polypeptide profile of the myofilament. Although under normal circumstances the amount of protein correlates quite well with the level of its cognate mRNA,³⁶ we showed previously that transgenic overexpression perturbs this relationship significantly by effecting a complete MLC2 isoform switch (MLC2a→MLC2v) in the atria even though MLC2a transcript levels were unaffected.^{19 20} The MLC2f overexpression recapitulates this observation, albeit with some subtleties that were not previously apparent. First, as was the case at the transcript level, MLC2f protein accumulation in the atria is consistent with copy number dependence (Fig 3⇓), although the relationship is difficult to quantify because of the high degree of expression and replacement even at relatively low copy number. Four of the lines demonstrate almost complete replacement of the atrial isoform with the skeletal form, despite the maintenance of normal MLC2a transcript levels. Interestingly, there is a significant difference between the abilities of the transgenic peptide to effect replacement in the atria versus the ventricles. For example, lines 3, 57, and 90 show roughly equivalent levels of transgenic transcripts in both compartments; this leads to >90% replacement in the atria but only to 5% to 35% in the ventricles. Although replacement was less complete in the ventricles, this cardiac compartment also displayed a copy number dependence, although the relationship is obviously not exactly linear. We ascribe this lack of exact correspondence to position-dependent effects, which can also influence the expression patterns of the myosin promoters in transgenic animals.³⁷

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Effects in myocardial protein composition due to MLC2f overexpression. A, Myofilament proteins were extracted from tissues of adult (>8-week-old) nontransgenic (ntg) mice as well as the different lines of transgenic mice, subjected to SDS-PAGE as described in “Materials and Methods,” and stained with colloidal blue. Thick- and thin-filament proteins are indicated. Depending on the relative level of transgenic overexpression in the atria, an essentially complete replacement of MLC2a with the MLC2f isoform could be effected. In the ventricles, MLC2f overexpression was much less effective in replacing the endogenous MLC2v protein, even in line 57, which contains 34 copies of the transgene. In neither cardiac compartment could an overt effect on the myofilament stoichiometry of the other contractile proteins be detected. B, Quantification of the degree of replacement in the ventricles and atria of transgenic mice is shown. The stained gels were scanned, and the signal intensities quantified using NIH Image software (version 1.57). The corresponding levels of the endogenous MLC2 isoform (endogenous) and MLC2f were determined for each line in the separate cardiac compartments. The high degree of replacement at relatively low copy number is apparent in the atria.

We showed previously that transgenic expression was uniform throughout the atria and ventricles.³⁵ However, the lack of apparent replacement in the ventricles raises the possibility that heterogeneous expression could occur within the cardiomyocyte pool; this has been inferred from physiological studies carried out on cardiomyocyte populations derived from transgenic animals expressing cardiac troponin C, in which individual cardiomyocytes derived from these hearts reacted quite differently to Sr²⁺ activation.^{38 39} Thus, we considered it possible that the lack of apparent replacement was due to two pools of cardiomyocytes, one that expressed the transgene and one that did not. Preliminary experiments showed that no significant pools of MLC2f could be detected in the nonmyofilament protein pool or in the soluble or insoluble fractions (data not shown), consistent with our previous observations.^{19 20} Therefore, we determined directly the myofilament protein composition in individual ventricular myocytes using acrylamide gel electrophoresis followed by silver staining. This technique is capable of detecting the myofilament protein population from a single cardiomyocyte. However, for the sake of clarity, the proteins were isolated and analyzed from 10 pools, each pool consisting of two ventricular cardiomyocytes isolated from line 57 (Fig 4⇓). All 10 groups display both the ventricular and skeletal isoforms, in roughly equivalent proportions, indicating that transgenic expression occurs throughout the cardiomyocyte population. Since no pool of non–myofilament-associated protein could be detected, either by standard means or Western analyses, we think it likely that the differential replacement observed between the two cardiac compartments is due to the different affinities of the RLC isoforms for their respective contractile assemblies (see “Discussion”).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

MLC2f is expressed throughout the cardiomyocyte population. Silver-stained SDS-PAGE of ventricular myocyte proteins was performed for the following: ntg, heavily loaded lane of pooled control (nontransgenic) ventricular cardiomyocytes; tg, heavily loaded lane of pooled transgenic ventricular cardiomyocytes; psoas, psoas muscle (MLC2f control); and lanes 1 to 10, transgenic ventricular cardiomyocytes (from line 57: two cardiomyocytes per lane). The region of the gel containing the MLC2v and MLC2f species was scanned (as indicated in the figure) such that the relative proportions of the two proteins could be determined. Shown for the sake of clarity are the scans from alternate lanes as indicated. Note that the relative content of MLC2f and MLC2v showed little variation among the samples. The areas under the peaks were quantified using NIH Image software (version 1.59), and the data were grouped: MLC2v, 89.4±14.7; MLC2f, 137.0±20.1.

Protein Replacement in Double Transgenic Heterozygotes

If the degree of replacement is simply a straightforward function of gene dosage and the affinity, relative to the endogenous protein species, of the transgenic LC for the “foreign” contractile apparatus, then by increasing the effective concentration of MLC2f, it should be possible to increase the degree of replacement. To test this hypothesis, we attempted to increase penetrance of MLC2f replacement in the ventricle by increasing the effective copy number and steady state level of MLC2f RNA. This was done by breeding line 57 (34 copies) with line 6 (20 copies) and analyzing the resultant offspring for the double heterozygotes. Preliminary analyses showed that the expected increase in MLC2f RNA levels in both the ventricles and atria occurred (data not shown), and subsequent litters of these animals were then analyzed, both for the relative levels of MLC2f transcript and for the degree of protein replacement in the ventricle (Fig 5⇓). The data confirm that transgenic expression within a line is quite stable. The seven animals used from line 57 (Fig 5A⇓) were derived from mice spanning at least three breeding generations, yet the standard deviation (Fig 5B⇓) is ≤11%. Similarly, four animals from line 6 show little variation in transgenic expression. Shown also are typical RNA levels from individual offspring derived from a cross between line 57 and line 6. The single heterozygotes are easily distinguished by their relative RNA levels, although it is not possible, unambiguously, to tell from the RNA quantifications to which line they belong. As expected, a cross between two single heterozygotes yields animals that lack either transgenic allele (tg57⁻/tg6⁻; Fig 5⇓, sample c). Strikingly, the double heterozygote (a single animal out of the six; tg57⁺/tg6⁺; Fig 5⇓, sample d) shows an RNA level that is essentially additive between the two lines (23 777 versus 10 863±1007 [line 6] and 13 422±1469 [line 57] arbitrary units). The data (Fig 5C⇓ and 5D⇓) also show that, consistent with the prediction, the double heterozygote does indeed show an increased degree of protein replacement in the ventricle, indicating that as the relative level of the ectopic protein increases in the cardiomyocyte, the degree of replacement also increases.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

Transgene expression levels in a double transgenic heterozygote. A, MLC2f RNA levels were determined in ventricles derived from either seven line 57 animals, four line 6 animals, or individual offspring from a 57×6 cross. A nontransgenic mouse (ntg) was included as a negative control. B, Histograms derived from the quantification of the RNA are shown. The double heterozygote (tg57⁺/tg6⁺, sample d) and nontransgenic (tg57⁻/tg6⁻, sample c) offspring are easily distinguished from the more frequent single heterozygotes. C, Myofibril protein complements of the single and double transgenic heterozygotes are shown. Myofilament preparations were electrophoresed in 15% polyacrylamide as described previously.¹⁹ D, The degree of replacement in the double heterozygote (sample d) is increased relative to either of the single transgenic lines.

Functional and Histological Analyses of Transgenic and Control Hearts

Although the major objective of the present study was to determine whether transgenesis could be used to remodel both cardiac compartments simultaneously, we wished to determine if partial replacement of the cardiac RLCs with the skeletal isoform altered contractile function. To determine if functional differences might be present at the whole-organ level, groups of strain-, age-, and sex-matched transgenic and control animals were subjected to physiological analyses using both the isolated Langendorff (retrograde, nonworking)²⁵ and working heart²⁶ preparations. To determine to what extent the line 57 transgenic hearts could be loaded with increasing volume (venous return) loads, cardiac minute work was varied from 200 to 600 mm Hg×mL per minute. Under identical loading conditions, the transgenic hearts produced maximal rates of pressure development that were significantly reduced relative to the control hearts, with +dP/dt reduced by 14% and −dP/dt reduced by 12% (Table⇓). The decreased +dP/dt indicates that replacement of MLC2a and partial replacement of MLC2v with MLC2f led to significantly reduced contractility (longer time to develop peak ventricular pressure), as well as perturbations in relaxation.

As we have shown previously,²⁶ the normal (wild-type) hearts showed a strong correlation of +dP/dt to increased left ventricular minute work, exhibiting a Starling response. However, there was significant animal-to-animal variation among the line 57 transgenics: seven hearts demonstrated a response to increased workload that approximated the response of the normal hearts, and three displayed severely impaired cardiac function and could not be workloaded without failure. The remaining seven transgenic hearts that could be loaded did demonstrate Starling function despite statistically lower cardiac parameters. To examine function in all 10 transgenic hearts, Langendorff preparations were used. When +dP/dt and −dP/dt were examined under Langendorff anterograde non–work-producing conditions, much greater deficits in +dP/dt, and −dP/dt were observed relative to the control hearts (Table⇑): +dP/dt was reduced by 62%, and −dP/dt was reduced by 52%. We also measured the maximum Ca²⁺-activated Mg²⁺-ATPase activity of left ventricular myofibrillar preparations from these same hearts. The actomyosin enzymatic activity of transgenic preparations relative to the controls was reduced by 22% (Table⇑), consistent with the decreased contractility displayed by these hearts.

Despite the reduced contractility of these hearts, no obvious pathologies developed in the line 57 adult animals (Fig 6⇓). No significant changes in chamber size, weight, or architecture could be discerned between the control and transgenic adult hearts (Fig 6A⇓ and 6B⇓, respectively). Trichrome staining revealed that no significant fibrosis had developed (Fig 6C⇓ and 6D⇓) and that the overall myocyte organization and structure were well preserved in the transgenic animals (Fig 6E⇓ and 6F⇓).

• Download figure
• Open in new tab
• Download powerpoint

Figure 6.

Histological examination of the MLC2f-overexpressing hearts. Fixation and histological analyses were carried out essentially as described previously.²⁶ Shown are sections of hearts from control (A, C, and E) and transgenic (B, D, and F) adult (16- to 20-week-old) animals. The hearts were cut in half longitudinally, preserved in Bouin’s fixative, and embedded in paraffin, and 4-μm sections were cut. Shown are serial sections from a single set of hearts stained with either hematoxylin and eosin (A and B) to show overall structure or with trichrome (C and D) to reveal any fibrosis that might have developed (magnification ×4 in A, B, C, and D). Hematoxylin and eosin–stained sections (E and F) taken from the left ventricles were photographed at higher magnification (×67) in order to reveal the readily apparent striations in the cardiomyocytes from both the control (E) and transgenic (F) samples. No hypertrophy could be detected, a result consistent with the lack of activation of the hypertrophic markers observed in Fig 2⇑. LV and RV indicate left and right ventricle, respectively.

As a preliminary study to characterize the basis for the alterations in contractility, single ventricular cardiomyocytes in which isoform replacement was ≈30% to 35% were isolated from line 57 adults, and the unloaded shortening velocity was determined²⁷ for control (n=7) and transgenic (n=10) cells. The unloaded shortening velocity of transgenic cells (2.40±0.55 muscle lengths/s) was not significantly different from that of control cells (2.86±0.47 muscle lengths/s), although there was a trend for unloaded shortening velocity of transgenic cells to be less than that of control cells. In addition, the series elasticity, ie, the length change required to just lower tension to zero, was not different when comparing transgenic cells (15.1±1.5%) with the control cells (14.6±1.1%).