Human Autoantibodies Reveal Titin as a Chromosomal Protein
Christin Machado, Claudio E. Sunke, Deborah J. Andrew
The Journal of Cell Biology, Volume 141, 1998

Review by Sarah E. Brautigam

    The protein titin has already been already identified as a huge sarcomeric protein contributing to the elasticity of striated muscle.  This goal of this paper is to convince the reader that titin is also a protein used by chromosomes in providing structure and elasticity.  The way in which chromosomes are assembled is necessary for later chromosome condensation, segregation, and integrity.  Most proteins involved in this process have yet to be discovered.  In this paper first, human autoimmune scleroderma serum  is used to identify a chromosomal protein in human epothelial cells and Drosophila early embryos (so that there are no muscles).  The authors then used the protein information to clone the Drosophila gene for this chromosomal protein.  They proceed to try to prove that the gene encodes for a protein which is a homologous to the vertebrate protein titin by 1) sequence similarity of the Drosophila protein and vertebrate titin, 2) expression of this protein and RNA in embryo muscles, 3) presence of this protein on specific areas of Drosophila muscle, and 4) protein size.  Finally the authors show that vertebrate titin localizes to condensed mitotic chromosomes in human epithelial cells.
    The authors first identified a serum produced by 1 person of 40 with autoimmune disease scleroderma that bound to proteins on chromosomes in both human epithelial cells (HEp-2) and Drosophila embryos 0-2 hours mature.  Figure 1A uses immunofluorescence to show chromosomal staining with the serum in the Drosophila early embryos and HEp-2 cells (green).  This figure also includes double-staining of DNA with propidium iodide (red).  The left and center panels, serum identifying the protein (red left), and the DNA( green center) were then overlapped so as to make sure that the serum was in fact binding to proteins on the chromosomal DNA.  When the two were superimposed (right) a yellow color was the result.  The figure shows the chromosomal staining in interphase, prophase, metaphase, anaphase, and telophase.  during interphase when the chromosomes are not condensed there was low level staining throughout the nucleus (except nucleoli).  In prophase the staining began to localize as the chromosomes condensed.  Metaphase through telophase the condensed chromosomes stained uniformly.
    This scleroderma serum was then used to isolate the Drosophila gene that produced the protein that bound to the chromosomes in figure 1A.  The serum was used to screen a Drosophila genomic expression library in which five overlapping clones were identified from all of the proteins in an expression library.  These clones each encoded several copies of an amino acid sequence rich in proline, valine, glutamic acid, and lysine.  It is important to note that the I-band of vertebrate titan has also been shown to contain a domain rich in proline, glutamic acid, valine, and lysine.  Figure 2C shows 2 of the five clones- the LG clone and the JT.  The LG clone (largest clone) was expressed in E. Coli and and used to immunize rabbits to makes anti-LG affinity-purified antibodies.  Immunofluoresence of the anti-LG antibodies were shown to give the same chromosomal staining patterns as the serum as seen in Figure 1B.  Figure 1B only shows metaphase and anaphase though.  This does not seem to be a problem because it is already shown that serum will bind to the chromosomes in all stages.  Additional exons were isolated from the Drosophila gene were isolated and another portion of the protein was used to make another polyclonal antiserum in a rat.  This antiserum bound to the portion of the protein which was encoded by the KZ region of the gene in Figure 2A.  The anti-KZ antiserum reproduced a similar chromosomal staining pattern through immunofluorescence as the original scleroderma serum and the anti-LG antibody in Figure 1C.  Again this figure shows staining in the metaphase and anaphase and includes staining of DNA so as to make sure that the proteins that the antibodies are binding to are on the chromosomes.
    The entire gene encoding the chromosome-associated protein was never cloned.  The authors are careful never to call the protein titin until they have more figures to substantiate its homology to titin in muscles and vertebrate titin.  The authors account for this unsuccessful attempt to clone the gene because of the repetitive structure of the gene.  Problems isolating the cDNAs was difficult for this same reason of repetitiveness but also huge mRNAs.  Figure 2A shows the limited gene map that the authors were able to describe for the chromosome-associated protein.  This figure is useful in determining where a certain antibody binds to the chromosome-associated protein (titin).   Figure 2B is a protein sequence alignment which shows the similarity that the ORFs encoded by the the NB and KZ cDNAs of the  chromosome-associated protein isolated in Drosophila have with chicken skeletal titin and human cardiac titin.  In the region of overlap, the ORF encoded by the KZ cDNA shows 28.6% identity/58.3% similarity to chicken skeletal titin and 27.4% identity/ 56.8% similarity to human cardiac titin.  The encoded by the NB cDNA shows 18.4% identity/ 48.9%similarity to chicken skeletal titin and17.2% identity / 47.7% similarity to human cardiac titin.  These numbers seem to indicate a highly conserved sequence among the previously identified titin in humans and chickens.  In this figure the authors seem to be suggesting that because it looks like titin in other species, it must be titin in Drosophila.  The authors acknowledge the fact that the conservation among the ORFs from the LG and JT clones and vertebrate clones is not as great.  They explain that these clones have a high frequency of P,E,V, and K residues which the ORFs correspond to elastic PEVK domain  of vertebrate titin which contains lots of P,E,V, and K.  It would have been more convincing if the authors had indeed done a protein alignment sequence for these ORFs and the corresponding vertebrate regions rich in the P,E,V, and K residues
    Next the authors set out to determine whether the gene that encodes the nuclear chromosome-associated form of titin also encodes the muscle form of titin.  They examined the protein expression
and presence of RNA in Drosophila embryos.  The authors seem to be at this point be comfortable calling the chromosome-associated protein titin.  The authors state that Figure 3A to 3H shows the actual muscles that Drosophila TITIN is expressed in all the somatic and visceral musculature in stages of 10/11 to 16 embryogenesis.  The titin (brown) is localized to different muscles as the embryo develops.  Additionally the RNA (blue) seems to be present in all stages in the muscles.  Muscles include body wall muscles, pharyngeal muscles and others.  The authors suggest that the early RNA accumulation and protein presence in both somatic and visceral muscle precursors parallels vertebrate titin accumulation in early myoblasts.  The presence of the RNA in 3A is not as clear as the other figures.  It seems that in the other figures the intensity of the RNA is the same as the protein but it is not the case with A.  The RNA is hardly detectable.  The authors fail to state what they examine the presence of the proteins and mRNA with. In other words what antibody did they use.  It is unclear whether it is a new antibody for muscle titin in Drosophila or whether it is the autoimmune serum or or the anti-LG or KZ .  This is important because the the antibodies bind different epitopes on titin.  Certain antibodies may bind and others may not depending on the similarity of the muscle titin and the chromosome-associated titin. It is also seems peculiar that no RNA or protein was detected in the embryonic cardiac muscle precursors.
    Determination of protein location on specific parts of the sarcomere was examined in Figure 4.  The authors hope that this specific location of titin on the muscles will further convince the reader that it is in fact , homologue of vertebrate muscle titin.  One of the characteristics of titin is that it localizes to muscles.  Adult thoracic muscles were immunostained with antibodies directed against two different domains of the protein using the anti- LG and anti-KZ and the original autoimmune serum.  Both the autoimmune serum and the anti-LG bind to the PEVK rich region of the protein and the KZ binds to the amino end (Figure 2).   It is shown in figure 4A and 4B that anti-KZ antiserum stained the Z-disks.  The authors use this evidence to then determine whether the anti-LG and autoimmune serum bind to the Z-disks.  Figures 4C and 4D present a few problems.  There is a lot of background staining in 4C along the myofibril. The authors suggest that this is a result of other antibodies present in the serum.  This may present a problem in the paper if in other experiments for other figures there were other antibodies binding to other antigens besides titin.  In addition in Figure 4D and 4C show staining along the M-line suggesting cross-reactivity with other antigens which also a potential problem.  In both cases of cross-reactivity with other antigens and the presence of other antibodies may cause misleading data.  Both 4C ad 4D suggest some recognition of Z-disks by both serums though.  Figures 4E and 4F show two other antibodies against vertebrate titin recognize epitopes on Drosophila myofibrils.  The authors state these figures are poor in resolution so as to not allow staining on two separate regions of the myofibrils, the Z-disk and I/A-band, to be confused.   Figure 4G shows shows that anti-KZ and anti-LG stained the Z-disks of viscertar larvae.  In the figure legend it is not clear which antibody is staining green.  There are several problems with this figure, but the authors do address most of them.  Good negative controls were performed in A and B so that no staining was apparent with preimmune anti-KZ serum or secondary antibodies.
    At this point in the paper the authors feel comfortable enough to officially call the gene isolated with the human autoimmine scleroderma serum D-TITIN based on Figures 3,4, and 5.  Figure 5 deals with the molecular weight of titin.  The authors are trying to really convince the reader once and for all that D-TITIN is the homologue to vertebrate muscle titin.  The authors suggest that the D-TITIN must be 2-4 MD in size if it is the homologue.  Total protein extracts were extracted for 8-24 h Drosophila embryos were separated on a polyacrylamide gel and transfered to nitrocellulose.  Immunoblots or western blots were prepared with anti-LG and anti-KZ.  Figure 5a is an immunoblot with protein from 8-24 h embryos from Drosophila, 5b is an immunoblot with protein from 0-2 h embryos from Drosophila, and 5c is protein from HeLa cells (epithelial cells).   The authors find that there is a band of a megadalton size present in all three cases for both anti serums 5a, 5b, and 5c lanes 2 and 4.  Figure b and c were used to show the size of chromosome-associated D-TITIN in non muscle cells because the 0-2 h embryos have not developed muscles and there is no muscle in epithelial cells.  There appears to be good negative controls in a, b, and c with the preimmune serums for both antibodies in which no cross-reacting polypeptides are detected.  The authors conclude that since by immunoflorescence staining of the 0-2 h embryos and the HEp-2 cells is chromosomal (Figure 1),  the chromosomal form of D-TITIN is migrates to the megdalton size and is about the same size as the muscle form.  There are a few problems with figure.  First even the authors admit there is only a discrete band and it shows up in some lanes better than others.  Second there is a lot of background that the authors do not address but this is possibly to be expected with western blots. The biggest problem is the molecular weight markers in lane one.  The markers should correspond better with the weight of the titin protein so that there can be a better estimate than "a band of megadalton size."  The markers stop at 584 and the protein is much bigger than that.
     Figure 6a tests 8 antibodies directed against different epitopes of vertebrate titin.  The antibodies immunostained HEp-2 in order to determine whether antibodies to vertebrate titin bound to chromosomes.  Six out of eight antibodies stained the the chromosomes in a pattern just as the shown in Figure 1 to indicate that vertebrate titin does seem to bind to chromosomes.  The two antibodies that didn't recognize the the titin the HEp-2 cells were those directed to the amino terminus of titin.  The authors explain this by saying that the KZ cDNA is not homologous to the vertebrate titin but the muscle D-TITIN probably will reveal homologies.  Figure 6b shows where on titin binds to certain portions of the muscle.
    This paper seems to indicate a couple of things.  First it seems to prove that D-TITIN, the chromosomal-associated protein in Drosophila is in fact homologous to the vertebrate titin.  This links titin as a chromosomal-associated protein as well as a protein associated with the muscles.  This then suggests a structural role of titin in chromosomal condensation.  More research must be done to actually see what the role of titin is in condensation of chromosomes-not just that it is present on chromosomes and resembles muscular titin.  The authors suggest titin functions in a similar way in condensing chromosomes and in muscles in the actual mechanism-as a molecular ruler so as not to allow chromosome breakage during mitosis by providing elasticity. In muscles the elastic component of titin prevents sarcomeric disruption when muscles are overstretched.  I think that a possible way to see what the role of titin is may be to possibly remove all or certain amounts of chromosomal titin and seeing how far mitosis precedes and if it can proceed at all.  It may be possible to tell exactly what function is missing.  One could use site-directed mutagenesis where the gene that encodes titin is deleted in the embryo of drosophila.  A restriction enzyme could also be used to wipe out the gene.  Then certain expression vectors such as plasmids with the titin gene inserted could be added to the organism.  The vector would allow control of titin expression thus seeing the how much titin is necessary.  The chromosomes could be observed under the microscope in order to see where in mitosis the process were stopped.  It is important that the muscle titin remain in the cell .
    Site-directed mutagenesis can also be used to create mutations in the titin gene and thus protein (instead of complete deletion) to help understand the exact role of chromosomal titin.  Homologous recombination could be performed in the stem cells of a  mother Drosophila and her offspring could be analyzed for the mutation and how it affected them.  In addition, after deletingt he titin gen, the expression vectors mentioned above, could be selected for mutated forms of the gene using resistence to ampicillan and disruption of the lac Z gene.  Since titin is such a big protein certain areas of the protein could be narrowed down to exact roles in chromosomal condensation.
    The removal of chromosomal titin may also answer the question of whether titin remains associated with the uncondensed chromosomes of interphase. The cell may not be affected by the lack of titin if it is not needed in interphase.  The authors suggest that titin remains bound to the chromosomes in interphase but do not know that function.
    Another question that the paper does not completely answer is the relationship between muscle and chromosomal titin.  The western blot in Figure 5 is not sufficient to say that the proteins are of the same mass. As stated earlier, the molecular weight markers were very poor. The authors describe possible ways that the 2 may be different in terms of amino ends and carboxyl ends.  The authors suggest that the findings about D-TITIN forms (muscle and chromosomal) suggest an alternative spice sites of the same gene and that the most amino terminal regions of muscle titin have not been cloned.  An interesting expreiment would involve investigating whether vertabrate chromosomal and muscle titin are splice variants of the same gen or whether whether they are encoded by two separate genes.  This could be tested by selecting for various probes for different parts of the mRNA. Titin in other species could be used to generate these probes.  The probes could possibly identify separate candidates for carboxyl termini within the gene.  With respect to which probes hybridized to RNA blots, one could determine which of potential polyadenylation signals are used in mRNA processing.

To go back to my home page

To go back to the molecular page

Email me at