Postsynaptic Abnormalities at the Neuromuscular Junction of Utrophin-deficient Mice

Deconinck A.E., A.C. Potter, J.M. Tinsley, S.J. Wood, R. Vater, C. Young, L. Metzinger, A. Vincent, C.R. Slater, and K.E. Davies. 1996. J. Cell Biol. 136: 883-894.

Utrophin is a cytoskeletal protein thought to act in ways similar to the protein dystrophin. Dystrophin provides essential structural support for the skeletal muscle cells in that it links the cytoskeletal framework to the plasma membrane. Dystrophin is found in all skeletal muscle, while utrophin is normally only found at the neuromuscular junction (NMJ). Because utrophin is highly related to dystrophin (85% amino acid conservation) and because it is located at the neuromuscular junction, utrophin was hypothesized to be essential to proper muscle functioning. This experiment was performed to determine the effects of utrophin deficiency in mice.
Utrophin-deficient knockout mice were created using homologous recombination. The utrophin gene was initially located by probing a 129/Sv library with a human utrophin cDNA clone. Figure 1a shows the technique utilized to knock out the utrophin gene: the top region is a targeting vector that contains a neo probe, the middle section is the genomic region containing the utrophin gene, the Xs indicate the regions where homologous recombination should occur to knock out the utrophin gene, and the bottom region shows the product of homologous recombination. Primary embryonic fibroblast cells were plated onto a G418- containing ES cell medium and the cells were transfected with the knocked-out gene. The only cells that survived in the presence of G418 were the transfected cells containing the neo marker. Figure 1b shows the Southern blot of ES cells. The lanes containing bands at both 8 and 9.6 kb contain DNA from the cells that had taken up the knocked-out gene. This DNA was injected into blastocysts and transplaned into a pseudopregnant mother. 12 chimeric mice resulted. They were mated and 6 Agouti offspring were produced. Figure 1c shows the Southern blot analysis of DNA from the tails of the Agouti offspring made to determine which of the offspring were heterozygous for the utrophin gene ( utrn +/-). Lanes with bands at both 8 and 9.6 kb indicate the heterozygous phenotype. These heterozygotes were mated (see Figure 1d) to produce homozygous recessive offspring (utrn -/-). Utrophin gene knockout in the homozygous recessives was verified by Western blot analysis, as seen in Figure 1e; only tissue proteins from utrn -/- individuals failed to show bands when blotted with primary a-utrophin antibody and secondary horseradish peroxidase and then exposed to chemiluminescence.
The utrophin-deficient mice did not appear to have any phenotypic abnormalities. They grew at normal rates, were able to breathe normally, appeared to have no muscle weakness, and reproduced with the same success as wild type mice. Urine was gathered from the utrn -/- and tested for elevated creatine kinase levels, an indication of muscle cell degeneration. Creatine kinase levels were not elevated, indicating that there was no breakdown of muscle cells in these urtn- /- mice. Western blots in Figures 2a and 2b indicate that both wild type and utrn -/- mice expressed comparable levels of a-sarcoglycan and b-dystroglycan, which are both part of the dystrophin protein complex (DPC). The Western blot in Figure 2c indicates that wild type and utrn -/- mice express comparable levels of dystrophin as well; this suggested that there was no upregulation of dystrophin to compensate for utrophin deficiency.
Next, immunocytochemical labeling was performed to examine the structure of the NMJ in wild type and utrn -/- mice. Panels A and B of Figure 3 show that the labeling for acetylcholine receptors (AChRs) at the NMJ is in the exact same location as the labeling for utrophin in wild type mice. Panels C and D of this same figure indicate that although AChRs are found at the NMJ in utrn -/- mice, there is no corresponding labeling of utrophin. Panel E shows rather dim labeling of AchRs in utrn -/- mice (which the paper does not comment upon), and Panel F serves as a negative control: no fluorescent labeling of utrophin is expected because no primary (anti-utrophin) antibody was used. Thus, Figure 3 serves to verify the similar locality of utrophin and the AChRs in wild type mice and the corresponding absence of utrophin in the urtn -/- mice.
Panels A and B of Figure 4 show that in both utrn-/- and wild type muscles, there is only one motor axon innervating each muscle fiber. Panels C and D and E and F show that AChE and AChRs were found in high concentrations at the NMJ in both utrn -/- and wild type muscle tissues. Panels G and H show that both the
utrn -/- and wild type muscles have synaptic vesicles on their nerve terminals, mitochondria, and folding of the postsynaptic membrane. The only noticeable difference between the utrn -/- and wild type muscle tissues seen in Figure 4 is the amount of postsynaptic folding. Utrophin-deficient mice showed a reduced amount of folding of the postsynaptic membrane (see Figure 4 G&H for visual comparison and Table II for quantitative analysis).
Additionally, Table I indicates that the number of AChRs per NMJ in utrophin-deficient mice is only 57% of the number of AChRs per NMJ found in wild type mice. It was not known whether this reduction was due to a reduced size of NMJ area, reduced density of AChRs, or reduction of amount of postsynaptic synaptic membrane to which AChRs could bind. It was hypothesized that this AChR reduction was correlated with reduced postsynaptic folding. Figure 5 indicates that AChR quantities are the same in newborn wild type and utrn -/- mice, before much postsynaptic folding has occurred. Only after postsynaptic folding is basically complete is there a explicit difference in AChR levels between wild type and utrn -/- mice. Tests of neuromuscular transmission showed that the release and action of AChRs in utrophin-deficient mice was normal (Table III).
Finally immunolabeling was used to determine if utrophin-deficient mice also had deficiencies in other proteins associated with utrophin. Figure 6 shows the immunolabeling of wild type and utrn -/- muscle tissue. As expected. utrophin did not show up in utrn -/- tissue (Panel F). However, all other proteins labeled (dystrophin, b-dystroglycan, and rapysn) were found in comparable amounts in both wild type and utrn -/- tissue. This suggests that none of the other proteins were upregulated to compensate for utrophin deficiency.
Therefore, this experiment showed that utrophin deficiency in mice is not a lethal condition; the only significant difference between wild type and utrn -/- mice was a reduction in the numbers of AChRs on each NMJ and a reduction in postsynaptic membrane folding.
However, the experiment was unable to conclude if the deficiency in utrophin was compensated for by upregulation of other proteins involved in the dystrophin protein complex (DPC); it is possible that even though immunolabeling showed no increase in the other proteins, upregulation could have occurred. a dystrophin antibodies weren't able illuminate all sites in the postsynaptic folds, and possible isoforms of dystrophin or utrophin to which no antibodies were used (or known about) might be compensatory upregulation proteins.
In a future experiment, the two hybrid system (Chien et al., 1991) could be employed to determine what proteins, if any, are involved in a complex with utrophin. A GAL4 DNA binding domain could be fused to the cDNA encoding the utrophin gene in mice, a GAL4 activation domain could be fused to cDNA encoding mice genomic DNA fragments (from a genomic library), these segments could be placed in plasmids, and along with the GAL4 promoter-lac Z gene they could be electroporated into yeast cells. The growth of blue colonies would indicate interaction of utrophin with a protein encoded for in the mouse genomic DNA fragment. Any protein found to complex with utrophin could be a potential compensatory upregulator.
Once this protein (called protein X from here on) was identified and the cDNA for it was cloned, antibodies could be made against protein X and the same immunolabeling done on the NMJ in the experiment could be performed again, only this time with the newly made antibody to protein X. Immunolabeling both wild type and utrn -/- NMJs would hopefully indicate any upregulation of protein X.
If the two-hybrid method was to yield no proteins that interact with utrophin, it would seem that this line of experimentation has reached a halt; perhaps utrophin deficiency in mice really is no great problem that must be compensated for. However, if the two hybrid method was to yield a protein X that interacts with utrophin and is upregulated in utrophin-deficient mice, it would strongly suggest that this protein X somehow compensates for the lack of utrophin and prevents adverse phenotypes in mice. To test this hypothesis, the protein X cDNA cloned in the two-hybrid system could be used to probe the mouse genomic DNA and locate the gene that codes for protein X. A new strain of knockout mice, deficient in protein X and/or utrophin, could be produced and their phenotypes observed as was done in the above experiment. Adverse neuromuscular defects would suggest the importance of protein X and/or utrophin in proper functioning of the neuromuscular junction. This information could be used in future exploration of treatment for humans with congenital myasthenic syndromes.

References:
Chien, C.T., P.L. Bartel, R. Sternglanz, and S. Fields. 1991. The two-hybrid system: A method to identify and clone genes for proteins that interact with a protein of interest. Prot. Natl. Acad. Sci. USA. 88: 9578-9582.

Return to Davidson College Biology Department Home Page