This web page was produced as an assignment for an undergraduate course at Davidson College

Review Paper

The immunoglobulin superfamily protein Izumo is required for sperm to fuse with eggs.

Naokazu Inoue, Masahito Ikawa, Ayako Isotani, & Masaru Okabe

Nature 434, 234-238 (2005)


The factors involved in the fertilization process have been investigated for a long time, and recently it has been found that the CD9 protein on the egg membrane is necessary for this process. This paper investigates the sperm factors involved in this process in mice, and attempts to prove that a novel immunoglobulin superfamily protein is an essential part in sperm-egg fusion.


This paper begins by explaining the methods used to locate, separate, and sequence the antigen Izumo, which they believe is an important factor in the fertilization process. This lab used the monoclonal antibody OBF13 against the mouse sperm which had been known to inhibit the fusion process. They identified the antigen using gel electrophoresis and immunoblotting and termed the antigen Izumo. The sequence of this antigen was found in the RIKEN database and confirmed by the sequencing of the RT-PCR product. The Izumo gene encodes a novel immunoglobulin family which is a membrane protein with an extracellular domain and a glycosylation site.

Figure 1. Identification and characterization of Izumo.

a. This figure is the amino acid sequence of both the mouse and the human Izumo, with the mouse sequence on top of the human sequence. The asterisks under the sequences indicate the same amino acid in both organisms. The red sequences indicate the sections that were found using the liquid chromatography and mass spectrometry methods. The signal peptide is located in orange and the transmembrane region of the protein is located in blue. The portion of the protein that is immunoglobulin-like is found in the green shaded box, and the two arrows indicated the cysteine amino acids that are likely to form a disulfide bridge. This figure seems to support the claim of the membrane protein with an immunoglobulin domain, but remains vague about the glycosylation site. However, the mention of location of the possible disulfide bridge is helpful in reading the part b of this figure.

b. This figure shows an image of a general type I membrane glycoprotein with an immunoglobulin-like domain, which is the suggested type of the protein Izumo. This especially highlights the transmembrane region and the disulfide bridge, but doesn't explain the part that sticks off of the immunoglobulin-like domain. I think that this part possibly represents the glycosylation site, but the paper should explain this section of the diagram.

c. This figure is a Western Blot that shows the anti-mouse Izumo antibody binding to the mouse Izumo protein only in the lanes for the testis and sperm. The lanes for all of the other tissues do not show any detectable binding. The sperm lane shows a much higher amount of binding than the testis lane, but there are no control lanes to compare these lanes to. This data seem to confirm that the Izumo protein is specific to the sperm and testis. However, with no molecular weight markers actually on the gel, it is difficult to determine whether or not the protein really is 56.4 kDa. There should also be a loading control in this blot to create a standard for evaluating the data.

d. This figure is a Western Blot that apparently shows the anti-human Izumo antibody binding to the human Izumo protein. This is a little difficult to believe because the only lane present is the lane corresponding to the sperm, but there are no other lanes to compare this result to, not even a lane for the testis tissue. I think that the lab should have created a Western Blot more similar to the one in the previous figure to show the lack of detectable binding in other tissues. As with the other Western Blot, this figure should have included molecular weight markers to measure the length of the protein, as well as some sort of loading control.

e. This figure is an immunostain of the Izumo protein in sperm from a transgenic mouse line that is acrosin-promoter-driven. This line has enhanced green fluorescent protein in intact acrosomes. The first line of the figure is just a picture of the sperm with arrows indicating the significant sections to examine in the other two lines of the figure. The second line of the figure shows the sperm that were stained with the polyclonal antibody against the mouse Izumo protein. Since these sperm are not stained green, the authors suggest that these sperm are acrosome-reacted, with the acrosomes no longer intact. The third line of the figure shows that sperm that have intact acrosomes, as indicated by the green fluorescent protein staining. There was no Izumo detected in these sperm, as indicated by the lack of the red staining of the polyclonal antibody against mouse Izumo. This figure seeks to explain that fertilization cannot occur until the exocytotic process of the acrosome reaction has occurred. Once the acrosome reaction has occurred, the acrosomes are no longer intact. This data is successful in showing that the Izumo protein is only detected in sperm that do not have intact acrosomes, which is assumed to be the sperm with the completed acrosome reaction. I think their reasoning that the hidden plasma membrane, where the Izumo is located, becomes accessible only after the acrosome reaction has occurred is very possible.

f. This figure is an immunostain of the Izumo protein in human sperm. The sperm were stained with the red polyclonal anti-human Izumo protein, and the acrosome-reacted sperm were stained green with the anti-CD46 antibody. The first line of this figure is a picture of the human sperm being stained in this experiment with arrows pointing to the two sperm that the authors want us to focus on. The second line shows the two sperm that have reacted to the polyclonal anti-human Izumo protein, indicating the presence of the protein in these sperm, but not the others that were originally shown in the first picture. The third line of the figure shows the two sperm that reacted to the antibody, indicating that these are the only two sperm in which the acrosome reaction has taken place. This data coincides with the claims that only the acrosome-reacted sperm have detectable Izumo protein. However, the color of this last stain is a little confusing because of the GFP used in the previous figure. I think it would have been easier to read the figures if they had made the second stain a different color or had found a way to test for intact acrosomes instead of just reacted acrosomes.

Figure 2. Targeted disruption of Izumo.

a. This figure shows the structures of the wildtype mouse Izumo allele, the targeting vector that was designed to replace exons 2-10 with the neomycin-resistant gene, and the mutant allele. The dotted lines are helpful to see where the neomycin resistant gene actually inserts into the allele. This is basically a helpful figure to just show what happened to form the Izumo deficient mutant allele described in the paper.

b. This figure is a Southern Blot of genomic DNA digested with EcoRI restriction enzyme. The bands at the 15 kb marker are the wildtype gDNA hybridized with a 3' external probe, and the bands at the 6.9 kb marker represent the mutant gDNA. The DNA for the homozygous wildtype DNA showed a wildtype band, but no detectable mutant band. The heterozygous DNA showed a faint wildtype band and a strong mutant band, and the homozygous mutant DNA showed no wildtype band and a mutant band. It surprises me a little to see such a faint wildtype band for the heterozygous DNA as compared to the dark line for its mutant band. However, the wildtype band of the wildtype DNA is not very dark either, so it may just be a loading problem. It's also interesting to note that the mutant band of the heterozygous DNA is darker than that of the mutant DNA. This figure is difficult to interpret because of the lack of loading control. However, it does seem to confirm the lack of the Izumo gene in the mutant allele created and diagrammed in Figure 2a.

c. This figure is a Northern Blot of total mouse testis RNA from wildtype, heterozygous, and mutant mice. There is a loading control in this blot, GAPDH, but it is separated from the experimental bands. The Izumo RNA is undetectable in the mutant mice, but shows strong bands in the wildtype and heterozygous mice. This seems to confirm the deletion of the full length Izumo mRNA in the mutant mice. However, this does not necessarily show that all of the Izumo mRNA is deleted. There may be a copy of the mRNA that is longer or shorter than the normal full-length mRNA. The fact that the loading controls seem to be separated from the other gel makes me a little skeptical of the legitimacy of the results.

d. This figure is a Western Blot of the same mice as described in the previous part of the figure. The Izumo protein is detected in the wildtype and heterozygous mice, but not in the mutant mice. There is no loading control for this experiment, although the differences between the bands seem to be pretty large. There are also Western Blots for the proteins ADAM2, CD147, and sp56, which have been found to be involved in the interaction of egg and sperm. These proteins show no difference in expression in the three types of mice, which show that they are not affected by the deletion of the Izumo protein. All blots in this experiment seem to have been done on different gels which produces some skepticism that the results are valid and from the same mice.

Figure 3. Male infertility caused by Izumo disruption.

a. This figure is a graph showing the ability of a heterozygous or mutant Izumo male and wildtype, heterozygous, or mutant Izumo female to reproduce. The female produced a litter if the male was heterozygous for Izumo, but not if the male was a mutant Izumo, regardless of the genotype of the female. This shows that the mutant male mice were sterile, but the heterozygous males are able to reproduce. Also, the mutant Izumo females show normal reproductive capabilities. This seems to confirm the claim that Izumo is required for males to fertilize, but is not required in females.

b. This figure is a graph showing the sterility of the Izumo mutant sperm in the in vitro fertilization system. Eggs that were inseminated with mutant Izumo sperm did not fuse with the sperm, whereas the eggs inseminated with the heterozygous Izumo sperm fused with the sperm and had a high percentage of pronucleus formation. This graph seems to confirm the sterility of the mutant Izumo sperm.

c. This figure is a picture of eggs after insemination with mutant and heterozygous Izumo sperm. The egg inseminated with the heterozygous sperm seems to have fused with a sperm and started the pronucleus formation. However, the egg inseminated with the mutant sperm has not fused with a sperm, but the sperm have penetrated the zona pellucida. This paper hypothesizes that the failure of the sperm-egg fusion leads to the absence of the zona reaction which lessens the zona pellucida ability to bind sperm.

d. This figure is a picture of eggs recovered from females that had mated with mutant Izumo males. The top panel focuses on the sperm that have accumulated in the perivitelline space of the egg, which means the sperm were able to get through the zona pellucida. The lower panel shows the sperm in the perivitelline space labeled with acrosome-reacted, sperm-specific monoclonal antibody MN9. This shows that there are sperm that have gone through the acrosome reaction, but still have not fused with the egg. The result here suggests that the problem is past the acrosome reaction, and the problem probably lies in the lack of the Izumo protein.

e. This figure is a graph showing the average numbers of fused sperm observed 2 or 6 hours after insemination. Mutant sperm did not fuse with the egg after 2 or 6 hours, but the heterozygous Izumo sperm fused with egg after 2 hours and even more after 6 hours. This seems to just be extra evidence of the sterility of the mutant Izumo sperm.

f. This figure is a picture of eggs with their zona pellucida mechanically removed. The left side of the figure shows the mutant Izumo sperm's ability to bind to the plasma membrane, as well as the heterozygous Izumo sperm. The right side of this figure show that sperm that actually fused to the egg which is seem after staining by Hoechst 33342 preloaded into the egg. These results seem to confirm that even if the zona pellucida is removed, the mutant Izumo sperm can not complete the second part of the syngamy process the way that the heterozygous Izumo sperm is able.

Figure 4. Involvement of Izumo in a xeno-species fusion system.

a. This figure is a picture of zona-free hamster egg 6 hours after insemination of mutant Izumo and heterozygous Izumo mouse sperm. The sperm were stained with Hoechst 33342 which binds to the sperm heads. This seems to show that the heterozygous sperm fuses to the egg, but the mutant sperm do not. This coincides with the claim that the mouse Izumo protein is necessary for the homozygous fusion system and suggests that the mouse Izumo protein is also necessary for the heterozygous fusion system with the hamster eggs.

b. This figure is a picture of zona-free hamster inseminated with human sperm with anti-human Izumo antibody or control IgG. The sperm were stained with Hoechst 33342 to determine which sperm fused to the egg. The sperm in the presence of anti-human Izumo antibody did not fuse to the egg, but those in the presence of the control IgG fused to the egg. These results suggest that the human Izumo protein is involved in the fertilization by human sperm.

Table 1. Development of eggs after ICSI with Izumo -/- sperm.

This figure shows the results of the experiment to determine whether the defect of the Izumo -/- was limited to fusion or extended to the later developmental stages. The experiment involved injecting mutant Izumo sperm into the cytoplasm of wildtype eggs, which would sidestep the fusion process. The fertilized eggs were implanted into females and then observed to see if the resulting embryos developed to term. The table shows that the mutant Izumo sperm defect is limited to fusion because an egg injected with the sperm is able to create an embryo that develops to term.


Overall, I think this paper successfully demonstrated that the Izumo protein is a critical part of the egg-sperm fusion process. However, as mentioned throughout the data analysis, there are many parts of the experiment that could have been changed to make the data easier to read and more reliable.


Future Research

This paper did a nice job in answering each additional question that was proposed throughout the paper. However, I think it would be interesting to see what proteins Izumo interacts with and if those proteins are also necessary for sperm-egg fusion. I would use a yeast two-hybrid method based on the yeast GAL4 protein to find the unknown proteins that interact with Izumo. This would involve the construction of the GAL4 activation domain plasmids, the GAL4 protein fusions, the yeast genomic libraries, and then the screening of the activation domain libraries.

The proteins that interact with Izumo are a good place to start looking for other proteins critical to the fusion process. Once I know which proteins, I would use the same methods and techniques of this paper to create a protein deficient knockout mouse and test for sterility, or lack of sperm-egg fusion. If the results are significant in mice, then I would continue the experiment to include human sperm.

I would then test to see how much of the Izumo gene is necessary to recover the wildtype phenotype of the Izumo knockout mice. I think this would help figure out the most important parts of the Izumo gene, which may lead to a way to treat Izumo deficiency. To test this, I would create Izumo knockout mice in the same method as used in this paper. I would then use plasmids of different length of Izumo and insert them back in and see if the wildtype phenotype was recovered.

Also, since we find out that CD9 on the egg membrane is an essential factor for egg-sperm fusion. Since this paper finds that Izumo is an essential sperm factor for egg-sperm fusion, it would be interesting to see if these two proteins interact. I think one way to do this is to insert a lacZ promoter into the CD9 sequence and the lacZ gene into the Izumo sequence. If these two proteins bind, lacZ gene should be turned on, which will produce b-galactosidase and turn the colonies blue when plated on a petri dish. However, if the proteins do not interact, lacZ will not be turned on and the colonies will be white. This is an effective way to see the interaction without having to look at any RNA or DNA.

Davidson College Homepage

Davidson College Biology Homepage

Davidson College Molecular Biology Homepage

Back to my homepage

Email me with any questions or comments