Mechanism of ICSI
ICSI has a very long history. It was originally developed on golden hamster’s eggs and first reported by one of the most influential reproductive scientists of the last century, Dr. Ruzo Yanagimachi. The first successful fertilization in humans using sperm injection was reported by Dr. Suzan Lanzendorff. However, the first pregnancy and birth were reported by Dr. Janpiero Palermo, who perfected the technique and made it clinically useful, truly revolutionizing the field of IVF. Most of the elements of the ICSI mechanism were elucidated by Dr. Dmitri Dozortsev and Dr. Andrew Rybouchkin.
Following “touching tail” spermatozoon becomes immotile
The mechanism of fertilization after ICSI is markedly different from that during natural fertilization. Whereas during normal fertilization process, sperm-oocyte fusion is followed by incorporation in the cytoplasm of a demembranated, “naked” sperm nucleus, which immediately becomes accessible for ooplasmic factors, especially thiol-reducing agents (Perreault et al., 1984), following ICSI not a naked sperm nucleus but the whole sperm cell, enclosed in i ts membranes, is exposed to the ooplasm (Palermo et al., 1991). It has been demonstrated that whole sperm cells behave in a different way than partially demembranated sperm cells in cytoplasmic cell extracts (Maleszewski, 1990). If an intact spermatozoon injected into the ooplasm, there will be no interaction between the spermatozoon and the oocyte. This because ooplasm does not have an enzyme to digest sperm plasm membrane to enable interaction. Therefore, sperm plasma membrane has to be damaged prior to the fertilization. Even though sperm cell appear to have distinctive compartments, such as a head, mid piece and the tail, it has only one plasma membrane covering the entire cell. Therefore, sperm plasma membrane can be damaged in any area, in order to enable sperm-egg interaction after injection. Usually, during ICSI, the sperm plasma membrane is damaged by “touching” tail, which in reality is pressing the sperm tail with a glass pipette, until the sperm membrane damage becomes sufficient for the membrane to loose its ability to maintain cell integrity, which leads to disabling Na/K pump and loosing membrane potential. To the observer this event is marked by the sperm ceasing its movement. The sperm membrane damage can also be revealed using vital staining technique, such as Eosin B. Vital staining is based on the intact membrane ability to exclude Eosin B. But once the membrane loses its charge, Eosin B migrates inside of the cell and stains the nucleus (Dozortsev et al, 1995).
Sperm plasma membrane damage effect can be achieved by freezing-thawing spermatozoa without cryoprotector, cutting sperm tail with the laser or by mechanically detaching sperm head.
The plasma membrane damage induced by “touching” tail does not allow anything from the sperm cell to “leak” into surrounding medium. Indeed, the sperm cells retain full capacity to fertilize an oocyte for several hours following immobilization.
Presence of PVP during sperm immobilization
PVP is added to the drop with spermatozoa to slow them down and prevent from sticking to the glass pipette used for injection. It also forms a film around sperm plasma membrane and prevents interaction between the sperm nucleus and the surrounding environment. This can be easily demonstrated by immobilizing the sperm cell in the presence of both PVP and Eosin. Eosin is routinely used to determine sperm viability, because intact plasma membrane will exclude the stain. Damaged plasma membrane becomes penetrable for Eosin, which will quickly stain nucleus bright red.
When sperm is immobilized by “touching-tail” in the presence of both, PVP and Eosin, it does not uptake Eosin, despite plasma membrane damage, because of a film formed by PVP around plasma membrane. The damage to the plasma membrane only becomes apparent after the PVP is considerably diluted.
PVP not only prevents Eosin staining, but also in-vitro sperm nucleus decondensation with DTT, a very powerful Thiol reducing agent, by the same mechanism. Therefore, excess of PVP injected into the oocyte during ICSI may delay or prevent fertilization.
Following sperm injection into the oocyte
Tight DNA packaging within the sperm nucleus is achieved by replacing histones with protamines with the subsequent disulfide bonds formation between them during sperm maturation. Those disulfide bonds must be reduced after fertilization to allow sperm DNA decondensation and formation of male pronucleus.
It has been clearly shown that Glutathione (GSH) is a disulfide bond reducer present in the cytoplasm of oocytes (Perreault et al. 1984,1988).
Oocytes can synthesize GSH (tripeptide -glutamyl-cysteinyl-glycine; GSH) during the first meiosis. Oocyte-derived GSH seems to assure the reduction of disulfide bonds in sperm nucleus. This, in turn, promotes nuclear decondensation for male pronucleus formation during fertilization (reviewed by Sutovsky & Schatten 1997). Thus, GSH provides the reducing power to initiate chromatin decondensation, prior to the male pronucleus formation (Yoshida et al. 1993). Depletion of endogenous GSH by a specific inhibitor of GSH synthesis during bovine oocyte maturation blocks the formation of a male pronucleus and prevents the assembly of sperm aster microtubules (Sutovsky & Schatten 1997). The elevated levels of oocyte GSH can enhance male pronuclear formation after IVF (Funahashi et al. 1994).
Strong evidences point to the lack of involvement of sperm-nucleus bound protease in nuclear swelling (Perreault and Zirkin, 1982)
Sperm nuclear swelling takes place after both, natural fertilization and ICSI. Following ICS, once PVP is washed out, GSH can enter the sperm nucleus and induce swelling before and irrespective oocyte activation.
It is very likely that such swelling ruptures sperm plasma enabling the direct contact between nucleus bound PLC zeta and the ooplasm leading to the oocyte activation. From that moment onwards, the mechanism of natural fertilization and ICSI becomes identical.
Following activation
Unlike sperm nuclear swelling, which is oocyte activation independent, all subsequent nuclear transformation of sperm and oocyte nuclear material requires oocyte activation and the resulting drop in p34(cdc2) kinase activity.
Before decondensation can begin, both, sperm and egg chromosomes have to condense and reach “chromatin mass stage“. It is very likely, that this stage is required to achieve synchronization of sperm and oocyte chromatin prior to decondensation.
Early telophase about 4.5 hrs after injection. Note that at this stage, the sperm nucleus does not decondense, but remains swollen. Also note that oocyte and polar body chromosomes forming two groups indistinguishable from each other.
Late telophase (aka chromatin mass stage) shortly after 1PB extrusion in 1 day old human oocyte. Sperm chromatin remains swollen, while oocyte chromosomes are tightly packed. At this stage, oocyte and sperm chromosomes become synchronized prior to decondensation.
The asynchrony in the sperm and oocyte chromosomes transformation may lead to a single pronucleus formation. Such asynchrony after ICSI may be caused by technical errors of injection, such as excessive amount of PVP injected along with the sperm cell. Because oocyte factors required for pronucleus formation only exist in the cytoplasm for a relatively short period of time, any considerable obsticle to sperm-ooplasm interaction after sperm injection may result in sperm nucleus missing decondensation window and failure to form a pronucleus.
This human oocyte was displaying a single pronucleus the day after ICSI. The sperm head is swollen, indicating that it was correctly injected and has interacted with an oocyte resulting in activation. However, it failed to decondense for unknown reasons. The resulting haploid embryo will not be viable.
Oocyte activation after ICSI is
The idea that an oocyte activation after or during ICSI is triggered by pricking with the needle came from earlier experiments on hamster oocytes, which can indeed be activated in that fashion. Vigorous ooplasm aspiration was therefore used in an attempt to improve fertilization rates.
However, subsequent data in humans demonstrated that even though injection of culture medium without the sperm using ICSI pipette induces short rise in free intracellular calcium, it does not lead to oocyte activation and the sperm is responsible for activation after ICSI. Further studies have demonstrated that putative activating factor from human spermatozoa can be separated from the bulk of the sperm cell and is cytosolic, heat-sensitive and non species specific. Subsequently it was identified as PLC Zeta. Spermatozoa lacking PLC zeta fail to activate oocytes.
Interestingly, even though vigorous ooplasm aspiration is not able to improve fertilization rates considerably, it may in fact be helpful in cases when a visible vacuole froms around the sperm head following injection. This vacuole may prevent sperm-oocyte interaction, thereby preventing PLC zeta release and oocyte activation.
Activating factor location within the spermatozoon
After the complete disruption of the sperm plasma membrane by the mechanical detachment of the sperm head, there is no appreciable reduction in activating potential of the sperm head after several hours of incubation. At the same time, following intracytoplasmic injection, the activating factor is released within minutes. This indicates that the PLC zeta release from the spermatozoon is an active process, that requires sperm-ooplasm interaction.
Mechanism of fertilization after ICSI
Whereas Glutathione is responsible for disulfide bonds reduction and its action is limited to the oocyte activation independent sperm nucleus swelling, SNDF is required for subsequent DNA decondensation (requires oocyte activation) leading to male pronucleus formation. The exact nature of SNDF is unknown, although it has been established that it originates from the GV nuclear material. 
PLC zeta appearance during sperm maturation
At the round spermatid stage, male genome is reduced to haploid state and is competent to support full term development. The time when PLC zeta first appears in spermatogenic cells seems to vary from species to species.Yazawa et alstudied this by injecting spermatids at various stages of spermiognessis (of various species) into mouse mature eggs (as universal recipients ) and found that the time when the factor first appears is different in different species.
Lack of success with fertilization using round spermatids (at least in some cases) may be attributed to failure to induce plasma membrane damage prior to the injection. Using Piezo injector for round spermatids plasma membrane damage seems to be the most efficent way to achieve that.
However, the evidences that round spermatids are able to activate an oocyte following injection are not irrevocable. It has been argued that due to ambiguity in identifying sperm cells precursors in the wet testicular sample preparation, immature spermatozoa rather than spermatids are choosen for injection and there is no oocyte activating factor – PLC zeta present until sperm cell is close to complete maturity.
Spermatozoon morphology and its activating capacity
Recently ovulated human oocytes are extremely resilient to parthenogenetic activation. Therefore, even if the fertilization rate is low, it is indicative of the PLC zeta presence in the spermatozoa.
In most cases failure of fertilization can be anticipated based on sperm morphology. It is extremely unlikely that a morphologically normal, live (motile) spermatozoon fails to activate an oocyte due to activating factor deficiency. If activation fails in such case, it is likely due to the oocyte. At the same time, morphological assessment of the sperm cells is not always predictive of its activating potential.
For example, acrosomeless (globospermic) spermatozoa in some cases will successfully activate human oocytes ( Stone et al., 2000, Dirican et al., 2008), while in others their injection will result in 100% failure of activation. This may to certain extent correlate with acrosine-positive or acrosine-negative status of spermatozoa, which in its own turn may correlate with presence of absence of PLC zeta.
In Kartagener’s syndrome, the fertilization failure after ICSI is likely due to difficulties of selecting a live spermatozoon for injection, rather than to the absence of an activating factor.
Oocyte’s sensitivity to PLC zeta changes over time
An oocyte injected or inseminated before it has reached cytoplasmic maturity, will undergo some form of abortive activation with sperm nucleus forming so-called premature chromosome condensation (PCC) and an oocyte advancing to the so-called MIII.
In mice, activation requires a presence of an intact spindle, although this phenomenon is strain-dependent. In the humans, the presence of a spindle has also been shown to correlate with the fertilization.
Sensitivity to PLC zeta, determined by fertilization rates, increases continuously with the time elapsed after administration of hCG, reaching a peak after 41 hours after hCG administration, although the highest implantation rate was achieved when ICSI was performed between 37 and 41 hours after hCG. This may be explained by the rate of chromosomal aberrations, which has been shown to correlate with the age of the oocyte at the time of insemination in mice, although other metabolic factors are likely to play a role as well.
Oocytes sensitivity to PLC zeta, albeit reduced, persists beyond 61 hrs after hCG, although resulting embryos have an extremely low implantation rate.
Identifying spermatozoa in testicular sample
Testicular biopsy, fine needle aspiration, and microdissection of individual tubules can yield motile, viable spermatozoa for ICSI. Identification of viable sperm in a homogenized tubule preparation requires sorting through the multiple cell types common to testicular tubules to find the most appropriate sperm cell for injection. Sperm cells are selected first for motility, and secondarily for morphology. Testicular sperm are immature, and motility may be manifest as slow or rapid non-progressive twitching, and on occasion, motility with forward progression.
Identification of sperm cells in the preparation is easier when sperm cell density, a reflection of spermatogenesis, is normal across the bulk of the testes tissue; however, there are instances when spermatogenesis not consistent across all tubules, where the tissue will exhibit a mosaic pattern of tubule morphology. Individual tubules can be evaluated in this case during microdissection, or alternatively, the testicular tissue may be mapped, using a grid approach, by multiple fine needle aspirations to locate individual foci of spermatogenesis, after which the specific regions of spermatogenesis may be explored further. Individual tubules that are more likely to have sperm-containing regions would be larger in diameter and opaque, compared to regions without active spermatogenesis, and there is a greater probability of spermatogenesis in tubules nearer to a blood supply.
Close up view of testicular tubules and blood supply to tubules; in situ
Testicular tissue can be retrieved on-site near to the IVF laboratory, or off-site and transported to the IVF laboratory. Additionally, the tissue can be collected the same day as the oocyte retrieval, or prior to the oocyte retrieval and held in culture overnight for example, or cryopreserved and stored in liquid nitrogen until needed.
Testicular tubules can be processed using a variety of techniques. Although time consuming, individual tubules can be ‘milked’, where tubules are visualized under a dissection microscope, and a non-cutting instrument, the rounded edge of a glass pipette or a glass slide, for example, is pushed down and long the length of the tubule, pushing the contents of the tubule out of one end, and repeating this process as needed. Alternatively, bulk tubules can be homogenized, using needles and/or scalpel blades or using sterile mortar and pestle-style tissue homogenizers. The homogenate may be collected in medium, and centrifuged to concentrate the tubule contents for examination. Enzymatic application may increase the efficiency of the procedure, increasing the yield of viable spermatozoa in individuals that have reduced spermatogenesis.
With active spermatogenesis, the cells in the homogenate would be represented by sperm and non-sperm cell types, where the most readily identifiable cell types are mature spermatozoa and red blood cells that have contaminated the preparation. The motion may be from the sperm itself, or indirectly from the sperm cell bumping or pushing against other cells, or motion of larger cells clumps from of the immature spermatozoa while they are still attached to the cells. It may also be due to the presence of the cells from vasa efferentia or proximal caput epithelial cell (S. Silber, unpublished). These cells have cilia which sweep the sperm into the epididymis from the testis. Due to incomplete separation of the spermatids, that can be multi-tail spermatozoa in the tubule tissue, as there are in the ejaculate. Rolling suspected, multi-tail sperm cells against the bottom of the dish with a micropipette can help to unravel the tails. The morphology of the immature spermatozoa can be distinctly different from ejaculated cells. Specifically, proximal droplets can be more pronounced, as retained cytoplasm has not yet been reduced. Additionally, elongating spermatids, with full tail formation but where the head of the sperm is still encased in cytoplasm, can also be found. And if there are no mature spermatozoa present, these elongating spermatids may be considered.
Normal spermatogenesis in testicular sample
Miss-identification of cells as motile spermatozoa in the testicular sample
Sometimes a motile sperm is reported on the initial observation in the operating room. However, a close examination in the reproductive laboratory using a more powerful inverted microscope allows making a more accurate determination. In this case, “moving cell” initially miss-identified as motel sperm was subsequently correctly identified as vasa efferentia or proximal caput epithelial cell by Dr. Silber.
Spermatogenesis and precursor cell types
Identifying, correctly, the various cell types representing early stages of spermatogenesis in a fresh preparation is more difficult. When mature spermatozoa are not identified in a testicular preparation, earlier stages may be identified and considered for ICSI, but for most IVF personnel, this practice should be restricted to identification and use of elongating spermatids, where there is a visible, obvious sperm tail, either complete or in the early stages of formation. The use of haploid spermatozoa precursor cells has been considered for use in humans (for review see Yanagimachi, 2005), however, it is very difficult to distinguish between haploid round spermatids and diploid round precursor cells, and as such the American Society for Reproductive Medicine and the Society for Assisted Reproduction Practice Committees have stated in a joint report, that ROSNI (or ROSI) should only be considered under approved experimental protocols (ASRM, 2008). For more information on identification of the various cells types present in a testicular preparation, there are two excellent papers that describe the stages of human spermatogenesis, and present photos of both fixed and living cells. Both papers are freely available from the respective journal websites (Johnson et al, 1999, 2001). See also Silber, et al, 1997).
Artificial activation basis
Development in humans can be set into motion by a number of artificial stimuli (i.e. Calcium Ionophore A23187, CaCl2, SrCl2). The majority, but not all, of stimuli that induce oocyte activation in mice, may also be applied in humans. Ethyl Alcohol, a very potent activator of mouse oocytes, is not able to induce activation of human oocytes, even 1 day old (Dyban et al, unpublished).
Almost all physical and chemical activating stimuli used for artificial activation in the humans, elicit Ca++ release into the cytoplasm, mimicking to different extent Ca++ release taking place during natural fertilization.
The end-point of activation with any activating agent eliciting Ca++ release is a physical destruction of cyclin B1, which leads to inactivation of p34 kinase and a drop in Maturation Promotion Factor (MPF) aka Cytostatic Factor (CSF) activity (Hyslop et al, 2004).
One of the potent artificial activating agents of human oocytes – Puromycin – has to be mentioned separately. This is because it does not cause Ca++ release and probably acts by suppressing synthesis of the cyclin B and by inhibiting its phosphorylation, rather than by its physical destruction. The point of no return in Puromycin activated oocytes is most likely the initiation of DNA synthesis.
This observation has practical importance because it predicts that Puromycin will act synergistically with Ca++, engaging activation stimuli.
Sensitivity to parthenogenetic activation of human oocytes is the lowest, soon after ovulation and continuously increases over time. The nature of activation stimulus is not critical for subsequent embryo development.