This web page was produced as an assignment for an undergraduate course at Davidson College
Inoue, N., Ikawa, M., Isotani, A., and Okabe, M. The immunoglobulin superfamily protein Izumo is required for sperm to fuse with eggs. Nature 434, 234-238 (2005).
Reviewed by Kevin Saunders
Website Navigation:
Spike Protein Structure and Function
More Information on SARS and the Spike Protein:
Centers for Disease Control and Prevention
Other Interesting Viral Proteins:
Webpage created by Kevin Saunders, an undergraduate student at Davidson College.
Summary of Results
Proteins that control fusion were sought initially by adding monoclonal antibodies to a fertilization assay. The monoclonal antibody OBF13 was found to inhibit egg-sperm fusion. Candidate antigens for this antibody were then identified by immunoblotting of sperm extracts using OBF13 as a probe. The peptides that were present in the band on the immunoblot were sequenced and compared to NCBI databases to find known proteins with high sequence identity. Ten proteins had 100% identity. RT-PCR of fly testis extract was performed to determine if any of the ten proteins with 100% identity were present in the testis cells. An immunoglobulin, named Izumo was found to be the protein present in the testis with 100% identity with peptides from the immunoblot band.
The researchers investigated the morphology and location of Izumo in three steps. First, the amino acid sequence of Izumo was examined for conserved domains and motifs. Amino acid sequence analysis showed that a putative signal sequence was present at the N-terminus of the protein. Also, there was an immunoglobulin domain identified in the protein. The amino acid sequence characteristic of a transmembrane region was recognized. Izumo was predicted to have a transmembrane region that preceded a small cytoplasmic tail.
The molecular weight and location of Izumo was ascertained by an immunoblot of various tissues. Mouse Izumo was 56.4 kDa and human Izumo weighed 37.2 kDa. The immunoblot showed that Izumo is expressed exclusively in sperm cells. The location of Izumo was deduced by immunostaining sperm cells before and after the acrosomal reaction. Mouse and human Izumo were stained only in cells that had undergone the acrosomal reaction. Izumo is thought to be present beneath the plasma membrane, becoming accessible after the acrosomal reaction.
With a better understanding of the location and morphology of Izumo, the researchers began to investigate the function of Izumo. The function of Izumo was assessed in fertilization assays using sperm cells expressing mutant Izumo. The mutant Izumo was made by homologous recombination, which resulted in a 8.1 kb section of the Izumo gene being cut out. Mutant sperm cells were confirmed by DNA, RNA, and protein blotting. The mutant cells did not have any detectable Izumo in RNA blots and immunoblots. DNA blots confirmed that a mutant version of the Izumo gene existed in mutant sperm cells. Related genes in the mutant sperm cells were also immunobloted to make sure that the mutating of Izumo had not destroyed other genes. None of the tested related genes were expressed differently in the mutant, thus the mutant phenotype could be attributed to Izumo.
The mutant sperm cells were used in fertility assays to demonstrate the phenotype of Izumo. Homozygous male mutant mice were sterile. Homozygous mutant females and heterozygous male mice were still fertile. Mutant sperm cells did not produce zygotes in In vitro fertilization. The zona pellucida did not inhibt fertilization of the eggs, because the sperm were able to penetrate the zona pelucida, but did not fuse with the egg. Nor did mutant sperm and egg cells fuse when the zona pellucida was removed from eggs. Hoechst 33342-a sperm labeling stain-was injected into eggs to label sperm cells that enter the egg. Mutant sperm cells were not stained, where as heterozygotes fused and were labeled with Hoeschst 33342.
Izumo function is not limited to egg cells of the sperm cell's species. Unlike wild type sperm cells, mouse Izumo mutants cannot fuse with hamster eggs that are zona pellucida free. Human wild-type sperm cells could not fuse with hamster eggs either when challenged with an anti-Izumo antibody, where as wild type Izumo sperm could fuse with hamster eggs when treated with IgG.
To complete the fundamental three stages of investigation (see the protein, change the protein, delete and rescue the protein) transgenic mice were crossed to see if the mutant phenotype could be rescued. The mutant mice were rescued when crossed with transgenic mice that had Izumo driven by a testis-specific promotor.
Critique
The author claims that Izumo is necessary for sperm-egg fusion. This argument is well supported in some areas of the results, and not so well supported in certain figures. First, the immunoblot of various tissues to deduce the location of Izumo was done without any loading controls (Figure 1c). The one lane shown for the human Izumo is from sperm cells only (Figure 1d). There is no evidence shown illustrating that Izumo is expressed only in sperm cells in humans. We are to assume that mouse expression and human are identical.
Immunostaining of sperm that are labeled for acrosome reactions provides a clear picture of how Izumo is available to be stained in acrosome reacted cells, but not in acrosome unreacted cells (Figure 1e). In Figure 1f, the authors are trying to make the same point as in Figure 1e, but the point is not as clearly supported. The confusion is partly because they switch their motive for staining from staining to show unreacted cells (Figure 1e bottom panel) to staining cells green to show reacted cells (Figure 1f bottom panel). These parts of the figure would be clearer if they would have been consistent in their staining.
Figure 2a explains how the mutant (targeted) allele and wild-type (Izumo) allele are structured. It illustrates well the missing section of Izumo gene. The DNA (Southern) blot is not as clear as the other blots (Figure 2b). The bands in the +/+ and -/- lane are well defined. The heterozygote lane has more mutant DNA sequences than wild-type, since the wild-type band can barely be seen. This difference in expression is not explained. Figure 2b does show the lack of a band at 15 kb in the mutant, which does show that there is no wild-type gene present. Figure 2c has clear bands in the wild-type and heterozygote lanes. There is no band in the homozygous mutant, which suggests that the mutant is not making a detectable amount of mRNA for Izumo (Figure 2c). The immunoblot shows that the protein is not made; strengthening the claim that they have made an Izumo mutant. The normal expression of related genes does support that the mutant phenotype is due to Izumo. However, I am not sure why these three genes are more likely to be affected by Izumo deletion than other genes. The researchers do state that mice with wild-type Izumo could rescue sperm cells with the mutant phenotype. However, the researchers did not specifically complement mutant mice with Izumo, therefore we do not know if Izumo was the only gene rescued.
In the fertility test it would have been nice to see how two homozygous mutants mated (Figure 3a). Figure 3b shows a strong difference in mutant and wild-type sperm cells. Although we do not know the genotype of the egg cells, Figure 3a makes us confident that egg genotype does not affect fusion results. Figure 3a and 3b show that swimming through the oviduct is not the cause of sterility. The increased number of sperm cells outside of the egg shows that mutant sperm cells cannot fuse and induce the slow block. Figure 3d does not clearly show anything except that the sperm cells are acrosome reacted. It is clearly shown by Figure 3d phase that sperm are present on or in the egg, but exactly where the sperm are cannot be ascertained. Figure 3f is cleverly designed, but there is no explanation of how one mutant sperm cell fused with the egg (bottom right panel).
The fusion assay of human and mouse sperm cells with hamster egg cells did not seem to add any substance to the paper. This assay seems to be a "look what we can do" experiment. Their claims would have been equally supported had Figure 4 been omitted.
I think that collectively the data support the conclusion that Izumo is necessary for fusion. The immunoblots and immunostaining showed that Izumo is in the right location to affect fusion. The authors showed that changing Izumo causes changes in fusion.
Future work
Izumo was shown to be necessary for fusion of the egg, but it was not definitively shown that Izumo is the only protein responsible for fusion. I would create Izumo mutants the same way the authors did. I would inject an expression plasmid for Izumo into these mutatnt sperm and see if the phenotype is rescued. I would expect to see sperm cells that fuse with egg cells if Izumo is the only protein responsible for the mutant phenotype.
Discovering the mechanism of Izumo is the next step in the research. Determining the ligand for Izumo would provide information about the mechanism of Izumo. I would do an affinity column with Izumo attached to beads to see what is the Izumo ligand. The beads of the column should have Izumo attached to them. Egg extracts would be passed through the column. I would use egg extracts, because Izumo causes sperm cell fusion with the egg, therefore it is probably binding to something on the egg. I would wash all unbound proteins out of the column. I would then wash off the bound proteins and sequence the proteins for comparison to the NCBI databases. Hopefully, this would produce the name of a protein with 100% identity. I would expect there to be a surface protein on the egg that is Izumo's ligand. As a follow up experiment, I would try to mutate the ligand for Izumo by homologous recombination and see if wild type sperm cells could fuse with the egg.
A fusion assay would also be helpful in determining if Izumo and its ligand are sufficient for fusion. To answer this question I would use two strains (Strain A and B) of auxotrophic MATa yeast cells. Strain A should be auxotrophic for lysine and leucine. Strain B should be auxotrophic for tryptophan and histidine. I could obtain these strains by ordering them from a biotechnology company or by mutagenizing the yeast with radiation or EMS. I would need two expression plasmids-one encoding the Izumo and lysine genes and the other encoding Izumo's ligand and tryptophan genes. I would also construct two more plasmids. One plasmid will encode GFP fused to a T7 promotor and the gene for leucine. The other plasmid would encode the T7 RNA polymerase and histidine genes. I would transform Strain A with the expression plasmid for Izumo. Strain A cells would be cultured in Lysine-free media to select for Strain A cells with the plasmid. Strain A would also be transformed with the plasmid encoding the GFP chimeric gene and leucine. The yeast would be cultured in leucine-free media to select for yeast that were transformed. Strain B would be transformed and selected for by similar methods. The plasmid encoding Izumo's ligand would be transformed first and yeast cells with the plasmid would be selected. Next, the plasmid with the T7 RNA polymerase gene would be transformed into Strain B, and those yeast cells would be screened for Histidine.
Next I would mix the cells in YEPD overnight. I would examine the cells with a fluorescent microscope for green glowing cells. These cells would be a result of the T7 RNA polymerase transcribing the T7 promotor driven GFP, which can only happen if the cells fuse. Since the cells are both MATa cells they should not fuse spontaneously, thus leading us to conclude that Izumo and its ligand are necessary and sufficient for fusion.