Paper: "MLP-Deficient Mice Exhibit a Disruption of Cardiac Cytoarchitectural Organization, Dilated Cardiomyopathy, and Heart Failure"
Authors: Silvia Arber, John J. Hunter, John Ross, Jr., Minoru Hongo, Gilles Sansig, Jacques Borg, Jean-Claude Perriard, Kenneth R. Chien, and Pico Caroni.
Source: Cell, Vol. 88, 393-403, February 7, 1997.
This paper is one of a series looking into the role of MLP protein in cardiomyocytes structure and its relation to heart failure. MLP is a LIM protein made up of 2 LIM double zinc fingers joined by a 50 amino acid residue.
This particular study was performed on mice, genetically altered to be either MLP deficient on both alleles (-,-), or MLP deficient on one allele (-,+). This was done by recombination in embryonic stem cells with inserted vectors in which the MLP translation initiation site was exchanged for a neomycin resistance cassette and a mutation was made in an internal start codon. The offspring followed Mendelian ratios, so the deficiency was not lethal to the embryo. None of the mice had heart defects at birth, but over half of the (-,-) mice experienced heart fatigue, cardiomyocyte hypertrophy, and heart failure between postnatal days 5 and 10. The homozygously MLP-deficient offspring (-,-) were divided into two categories: those that died during or before the second postnatal week, called early phenotype, and those that developed into adult mice, called adult phenotype.
Table 1 shows that offspring from a heterozygous cross were more likely to have the early phenotype than were offspring from a homozygous deficient cross. The fact that some homozygous MLP-deficient mice developed into viable adults and that the offspring of two (-,-) adults were less likely to be early phenotype than the offspring of heterozygotes suggests that the genetic make-ups of certain mice affect the penetrance of early phenotype. The table also shows that all offspring from each possible cross that made it to adulthood eventually developed dilated cardiomyopathy with hypertrophy and defective neuromuscular transmission. The paper did not explain why all of the offspring from two heterozygotes developed heart problems. I would have predicted that some offspring from the heterozygote cross would be homozygous for MLP and would not develop any symptoms.
Figure 1A displays MLP being expressed in the developing heart and in the skeletal muscles. Every heart cell expresses MLP (data not shown). MyoD, which is expressed in the skeletal muscles along with MLP was used as a control. This was done by in situ hybridization on a sagittal region. Figure 1B is a representation of the process by which MLP deficient cells were made by homologous recombination. The event was verified by PCR using the probes corresponding to the regions above the arrows. Figure 1C is a northern blot showing that MLP was not present in the hearts of adult MLP-deficient mice. Figure 1D is an immunoblot which shows that no truncated forms of MLP were present in the neonatal MLP-deficient hearts. This was performed with a MLP antibody recognizing the carboxyl-terminal end of MLP.
Figure 2A displays the obvious hypertrophy of the MLP-deficient heart in a 7 day old mouse. Both the atria and ventricles are considerably larger than those of the normal heart from the same litter. Figure 2B again illustrates the hypertrophy of the early phenotype heart. The northern blot on the left probes for ANF and MARP sequences, two proteins associated with hypertrophy, as compared to MLC-2v and Actin. The bar graph on the right shows the relative weights of the enlarged heart in comparison to both the adult phenotype heart and the normal heart at day 10. The adult phenotype heart does not yet show any signs of exaggerated growth at day ten. Figure 2C displays exaggerated growth at 6 weeks of the adult phenotype. So, we know that both phenotypes of MLP-deficient mice develop the same heart problems, but the adult phenotype does not undergo the lethal response seen in the early phenotype during the second postnatal week.
Figure 3 displays the differences in cytoarchitecture between normal and MLP-deficient hearts. By comparing the two, one can see that the MLP-deficient hearts have larger myofibrillar space and an increase in ribosomes, SR, and extracellular space. The last set of images in figure 3 compares vinculin distribution, a protein "involved in the anchorage of the actin-based cytoskeleton to the cell membrane (395)." The vinculin pattern of the MLP-deficient cells is analagous to that found in heart cells of humans with dilated cardiomyopathy.
Figure 4 is a image set comparing the cytoarchitecture of normal and MLP-deficient cells from neonatal hearts. The figure supports the idea that MLP-deficient cardiomyocytes have different intercellular organization even before hypertrophic response, including a higher number of actin filaments. Image E shows that adult MLP-deficient heart cells have a lacerated appearance. Mice with MLP-deficient skeletal muscle cells showed reduced rigor and swelling, most pronounced within the first three weeks (data not shown).
Figure 5 illustrates the disorganized cytoarchitecture of MLP-deficient muscle cells. Image A showed that MLP in normal cells is distributed along intercellular attachment sites in many of the same regions as vinculin, the Z-line in particular. Image B shows that MLP-deficient cultured cardiomyocytes have smaller gap junctions and more partially disorganized myofibrils. Image group C shows that the second LIM motif alone bound to the Z line in newborn cardiomyocytes, but it did not change the disorganized phenotype overall. Transfection of the MLP gene into an MLP-deficient cell did, however, correct the cytoarchitecture, showing that MLP is both necessary and sufficient. Because the second LIM motif bound to the Z region, the authors think that MLP may be a scaffold protein.
Figure 6 shows that an MLP-deficient left ventricle is less elastic than a normal left ventricle, a result of disorganized cytoarchitecture. This causes an inability of the heart to create adequate muscle tension with which to push the blood.
Figure 7 is a diagram explaining the authorsí theory of why dilated cardiomyopathy as a result of hypertrophic response occurs shortly after birth in early phenotype MLP-deficient mice. MLP aids in the alignment and realignment of myofibrils, thereby maintaining cytoarchitecture. If the myofibrils are not aligned, they begin to spread out and compensate with a hypertrophic response as the muscle contracts. This leads to a lower blood volume in the chamber and impaired tissue tension. The paper claims that heart proteins when run on a gel did not show any sign of differentiated composition or differentiated protein levels from those in the normal heart.
While this paper is full of hypotheses regarding the precise function of MLP in striated muscle cells, the only thing conclusively shown is that MLP-deficient mouse hearts usually donít work very well for very long. The purpose of this experiment, however, is to aid research in the area of heart disease in order to discover new therapies for humans. Humans have a homolog to the mouse MLP protein sequence at chromosome 11p15. While human heart disease has not been linked to this homolog, these researchers think that the same molecular pathway in which MLP is involved may be involved in some human heart failure. So, how do we move towards understanding the whole process in which MLP is involved? We know where MLP is expressed, and we know what happens to hearts when MLP is not expressed. Here are some questions and further experiments arising out of this paper which will further the researchersí goals:
What does MLP bind to? Take a mouse cardiomyocyte and extract the cellular protein. Separate the proteins by SDS-PAGE and probe with MLP. Take the proteins bound by MLP, find the amino acid sequence, and sequence the gene for that particular protein. This assay could then be done with MLP-bound proteins to find other proteins involved in the process. Or, one could perform the two-hybrid system developed by Chien and associates to not only find proteins that bind to MLP but also to sequence them. Yet another option would be to ligate an oligonucleotide sequence coding for MLP to agarose beads and running the cardiomyocyte cellular proteins through the bead column. Wash with increasing concentrations of salt solution and collect proteins that bind to MLP. These same experiments could be done using the LIM1 motif of MLP only or the LIM2 motif only in order to see which parts of MLP bind to what.
Which genes affect the penetrance of MLP? What are some common motifs specific to adult phenotype mice which are not present in early phenotype mice? What are some common motifs specific to early phenotype mice not present in adult phenotype mice? Perform RFLPís on cDNA from a number of adult phenotype and early phenotype MLP-deficient mice in order to locate areas possibly related to the penetrance of MLP- deficiency. Use common motifs exclusive to each group as probes on a Southern blot of the cardiomyocyte DNA. Clone the motifs into plasmids and mutate the motifs one at a time with restriction-site manipulation mutation. Transform cells of the opposite phenotype with the mutated cDNA sequence and see if the phenotype changes. This could be done until a mutation in a gene caused a change in the penetrance.
Are there proteins in striated muscle cells responding to mechanical stress? We know that hypertrophic response occurs after birth in early phenotype mice, probably due to increased mechanical stress after birth. Because MLP-deficient hearts respond to mechanical stress with hypertrophy in early phenotype mice, mechanical stress must induce some type of gene expression taking part in the enlargement. Or, perhaps the enlargement is due to a failure of proper gene expression. An experiment needs to be done examining the protein levels of different proteins in cardiomyocytes taken from an early phenotype mouse at different stages ranging from neonatal day 1 until heart failure. This could be done by extracting proteins from the cells and running them on an SDS-PAGE with protein stain. Proteins which increase in concentration as hypertrophy progresses could then be tested further to see if they are involved with the same process as MLP. If certain other proteins are known to be involved in the structure with MLP prior to this experiment, mRNA levels could be tested instead.
If MLP does function as a "scaffold protein promoting the assembly of interacting proteins at Z line structures," which part of MLP is crucial to this role? MLP is made up of two LIM proteins. Clone MLP cDNA into a plasmid. Perform random chemical mutagenesis with sodium bisulfate. Transform MLP-deficient cardiomyocytes with plasmids. Select cells that do not show signs of increased cytoarchitectural organization. Then, take those plasmids and narrow down the MLP coding region mutated and begin a process of site-directed mutagenesis to pinpoint the region responsible for MLPís role as a scaffold protein. The paper explained the function of LIM2 as binding to the z-line structure. We also need to know the function of LIM1. If LIM1 is inserted into MLP-deficient cells, where does it bind? Probe the transformed cell with radioactive a-LIM1 to see. Another form of site-directed mutagenesis is the plasmid shuffle. The wild-type MLP gene would be placed into a plasmid, thereby allowing the cell to function normally. Then, put a mutated version of MLP into another plasmid with a selector such as ade+. Transfect this plasmid into the cells and culture the cells on a complete medium. Then, grow them on an ade- plate. Cells that retained the ade+ plasmid will continue to live. Examine the cytoarchitecture of the cells and repeat the experiment with different mutations to locate the crucial domains of the LIM1 or LIM2.
Does the same thing happen to human cells in vitro when they lack the MLP homolog that happens to MLP-deficient mouse cells in vitro? Essentially, researchers can do the same thing to human cardiomyocytes in vitro that was done to mouse cardiomyocytes in vitro. Then, they can ook at the cells and compare them to normal cells.
© Copyright 2000 Department of Biology, Davidson
College, Davidson, NC 28036
Send comments, questions, and suggestions to: email@example.com