Mini-Review
The oviduct: a neglected organ due for re-assessment in IVF

https://doi.org/10.1016/j.rbmo.2014.11.011Get rights and content

Abstract

The oviduct has long been considered a ‘pipeline’, a tube allowing transit of spermatozoa and embryos; this perspective has been reinforced by the success of human IVF. Evidence accumulated over several decades, however, indicates that embryos can modulate the metabolism of tubal cells in their environment. Human IVF culture media is based on formulations that pass mouse embryo assays as quality control: the requirements of mouse embryos differ from those of human embryos, and therefore conditions for human IVF are far removed from the natural environment of the oviduct. The preimplantation environment, both in vitro and in vivo, is known to affect the health of offspring through mechanisms that influence imprinting. Recent studies also show that male accessory glands act in synergy with the oviduct in providing an optimal environment, and this represents a further perspective on the oviduct's contribution to harmonious embryo development and subsequent long-term health. The metabolism of the human embryo is far from being understood, and a ‘return’ to in-vivo conditions for preimplantation development is worthy of consideration. Although results obtained in rodents must be interpreted with caution, lessons learned from animal embryo culture must not be neglected.

Introduction

The oviduct has long been considered as merely a ‘pipeline’, a tube that allows transit first of spermatozoa, and then the embryo; this perspective has been reinforced with the advent of human IVF. Abundant, research, however, dating from as early as the 1970s suggests that, in addition to mere transit, the oviduct actively participates in other functions, such as storage of sperm cells, smoothly managing their binding and subsequent release (Pollard et al., 1991). It also actively manages transport of embryos (Villalón et al., 1982), with the capacity to distinguish between an unfertilized oocyte and an early stage embryo. In some species (e.g. the mare Betteridge et al., 1979), unfertilized oocytes are held at the utero–tubal junction, with the early cleaving embryo apparently holding a key that allows entry into the uterus.

Observations over several decades have revealed that tubal epithelium (Freese et al., 1973) and secretions (Georgiou et al, 2005, Hess et al, 2013) both undergo metabolic and biochemical changes when the embryo lies within their vicinity. Encephalins and other peptides mediate interactions between the oviduct, the ovary and the fertilized embryo (Cupo et al, 1987, Kent, 1975) via biochemical exchanges of information. The nature of these exchanges is difficult to ascertain, as the protein content of an early stage embryo is around 50 ng. The absolute amounts are minute and highly specific, so that the biochemical exchanges are difficult to quantify. This situation has led to aberrant assumptions and statements (see HLA-G, Ménézo et al., 2006).

An important feature to be remembered is that the process of imprinting occurs early after conception; lessons learned from animal embryo culture must not be neglected, although results obtained in rodents must be interpreted with caution, as the mouse is not an appropriate model for other species. It has now been established that the environment during the preimplantation period affects the health of offspring, both in vivo and in vitro.

Recent studies have described a role for the male accessory gland in synergy with the oviduct's role in providing an optimal environment for the preimplantation embryo: seminal plasma is not merely a medium for sperm survival (Bromfield et al, 2014, Robertson, 2007). This represents another aspect of the oviduct's contribution to long-term harmonious embryonic development and the subsequent health of the offspring.

The embryonic origin of the female genital tract lies in the mesonephros. After about 10 weeks, the nephric duct and mesonephric tubules degenerate, and the oviducts and uterus evolve from the paramesonephric duct, a transport epithelium containing approximately equal numbers of ciliated and secretory cells, varying according to the anatomical region; the ampulla evolves from the fully secretory region. With the exception of primates, the oviduct in all other mammals is translucent, with a thin wall, in contrast to the muscular oviduct of primates.

In humans, the morphology of the uterus and the oviduct is similar macroscopically, which may explain why tubal ectopic pregnancy is possible, a situation that is never observed in other mammals. Ovarian and uterine blood supplies provide vascularization, and veno–ovarian anastomoses allow a complex ‘trialogue’ between the oviduct, ovary and uterus. Oviductal secretions are highly dependent on a hormonal (steroid) environment. Peak secretion, targeted towards the ovaries under the influence of oestrogen, is observed around the time of ovulation. The subsequent influence of progesterone targets secretion towards the uterus. The tubal environment is apparently under endocrine regulation during embryonic transport and development, and this is reflected in or mimicked by gene expression in epithelial cells (Bauersachs et al., 2004).

The main differences between serum and tubal secretions are presented in Table 1. The chemical composition of tubal fluids consists of a finely regulated mixture of serum transudate with specific epithelial secretions. Osmolarity and viscosity are similar to serum, around 290 mOsm/kg and 1.8 mPa/s, respectively. An important feature worthy of note pertains to the mouse embryo assay as used in human toxicity tests: a physiological osmolarity of over 280 mOsm causes mouse embryo developmental arrest (Wang et al., 2011). Human tubal fluid has a pH of around 7.2–7.5, regulated by bicarbonate produced by high levels of carbonic anhydrase in the tubal epithelium. The oxygen tension is 60 mmHg, less than one-half that of atmospheric O2, with a reducing RedOx potential of −0.1 mV. The environment has high levels of antioxidants, such as hypotaurine, which is actively synthesized by the tubal epithelium (Guérin, Ménézo, 1995, Guérin et al, 2001); this topic will be expanded later. It is commonly felt that human preimplantation embryo development benefits from culture under conditions of reduced oxygen tension, such as 5% O2, 5% CO2, 90% O2; this is probably partially true. Together with inappropriate osmolarity, however, most human IVF culture media have poor, if any, protection against reactive oxygen spieces (ROS).

Two important unique features are high level of potassium, 2–5 times more concentrated than serum, such that a significant discrepancy occurs in Na/K ratio between serum (Na/K = 30) and tubal fluid (Na/K = 10–13); and a higher level of bicarbonate in the oviduct than in serum, owing to high carbonic anhydrase activity. Compared with levels observed in serum, levels of Ca++, Mg++, Na+, Cl, PO4 are approximately similar. Once again, these unique features are not taken into account in human IVF embryo culture, and their effect on preimplantation embryo development has not been investigated. Zinc, the second most abundant transition metal after iron, is of interest. It prevents toxicity caused by iron and copper, and is involved in numerous important metabolic reactions that are required during mammalian developmental processes. It acts as a co-factor for at least 200 enzymes, including carbonic anhydrase and zinc superoxide dismutase, both present in the oviduct. Zinc prevents oxidative stress by capturing superoxide and hydroxyl radicals through its involvement in metallothioneins and metal-response element-binding transcription factors (Andrews et al, 1991, Ménézo et al, 2013). It plays a major role in regulation of the one-carbon cycle, and therefore also in methylation and imprinting (Ménézo et al., 2013).

The concentration of glucose in tubal fluid is one-fifth to one-third lower than in serum, but it does contain fructose. The lactate concentration is much higher, reaching several times the serum value, compared with pyruvate, which is present at low levels in tubal fluid. A high activity of lactic dehydrogenase is present as the product of both actively secreted exudates and transudates secreted passively in response to low oncotic pressure (Menezo and Laviolette, 1972). Regeneration of NADH is caused by lactic dehydrogenase, using CO2 produced from NaHCO3 by carbonic anhydrase (Vishwakarma, 1962); this is incorporated by the embryo principally for the synthesis of the pyrimidine bases thymine, cytosine and uracil.

All of the amino acids are present in the oviduct at widely varying concentrations, generally higher in the oviduct than in serum: methionine, leucine, phenylalanine, lysine, aspartic acid, tyrosine, taurine, hypotaurine, and in particular glycine, alanine and glutamine (Guerin et al., 1995). Apart from their nutritional role, some of the amino acids are osmolytes, protecting the embryo against protein precipitation by a ‘salting out’ effect: this is particularly true for glycine, alanine, taurine and perhaps glutamine (Baltz, 2012, Casslén, 1987). Hypotaurine, a universal antioxidant, is synthesized by the tubal cells, producing taurine after oxidation (Guérin and Ménézo, 1995). All of the components of the urea cycle (e.g. ornithine, citrulline, arginine and urea) are present. The differences in their concentrations may also reflect the difference in their capacity to be transported into the embryo. In any case, all of the amino acids are present irrespective of their classification as ‘essential’ or ‘non-essential’ (Lane et al., 2001). All amino acids are essential substrates for the sperm and the embryo, especially in the perspective of imprinting and the crucially important one-carbon cycle (Ménézo et al., 2013). Amino acid uptake varies according to their affinity with transporter molecules. Carnosine (ß-alanyl-L-histidine) is also present in tubal fluid (Menezo and Laviolette, 1972), an ROS scavenger that also protects against unsaturated aldehydes produced by peroxidation of cell membrane fatty acids in situations of oxidative stress.

Catecholamines are found both free and bound to albumin in the oviduct (they are more stable when bound). Their concentration varies by anatomical region and during the oestrous cycle (Khatchadourian et al., 1987), but no precise physiological role has been ascribed to them. Oestradiol and progesterone are present in tubal fluid at concentrations more or less similar to that observed in serum (Richardson and Oliphant, 1981). Prostaglandins can also be detected in tubal fluid (Nieder and Augustin, 1986). Little information is available on vitamins in tubal fluid.

The total protein concentration in tubal fluid is close to 10% of that observed in serum. Albumin is present in the highest concentration (60–80% of the total), followed by transferrin, as well as ceruloplasmin (Menezo and Laviolette, 1972). Transferrin and ceruloplasmin regulate the levels of Fe+ + and Cu+ +, respectively, divalent cations that carry the highest risk for damage as well-known generators of ROS (Fenton reaction). It seems that globulins of high molecular weight have a limited capacity to pass through the oviductal membrane. Albumin can bind lipids, peptides and catecholamines. In several species, oviduct-specific glycoproteins and other glycoproteins have been detected, varying with the stage of the oestrous cycle. No specific biological role, however, has been found for these. Antioxidant enzymes are abundant in the oviduct (El Mouatassim et al, 2000, Guérin et al, 2001). Transcriptome analysis has demonstrated some variations in chemokines in response to progesterone and oestradiol (Hess et al., 2013). We observed, however, that the prepubertal oviduct in cow and mouse is able to sustain embryonic development (Figure 1) (Ménézo et al., 1989). Numerous growth factors have been detected: EGF/TGFα, IGF-1 and GMC-SF (granulocyte macrophage colony stimulating factor) are present in the human oviduct (Chegini, 1996), as well as leukemia inhibitory factor and interleukin 6. Their interactions with the embryo will be discussed later.

Lipids in the oviduct have not been studied in detail, although they are substrates for both sperm and embryos. All classes of lipids can be found in tubal secretions, mainly bound to high- and low-density lipoproteins (HDL and LDL). Tubal secretions also contain carnitine, which is required for beta-oxidation of these lipids; the preimplantation embryo is not able to synthesize carnitine (Ménézo et al., 2013). Albumin acts as a lipid carrier and exchanger, exchanging lipids with those in sperm membranes, an important in-vivo property that facilitates changes in sperm membrane fluidity associated with capacitation. All classes of lipids are present. They are mainly bound to HDL and LDL, as well as albumin acting as lipid carriers (Ehrenwald et al., 1990) and exchangers, especially, in vivo, versus sperm membrane lipids.

Co-culture of embryos with oviductal cells compared with with other cell types is an attractive tool for investigating interactions between sperm, embryos and the oviduct. Primary cultures of oviductal cells can be easily cultured in bicarbonate-buffered commercial media (Ouhibi et al., 1989) (Figure 2). TC199, MEM and B2 have been extensively used, with the addition of albumin or 10–25% serum. Mouse oviductal cells are the most complicated to culture, as it is difficult to obtain pure epithelial cells. The stage of the oestrous cycle when cells are collected seems to have an influence on epithelial cell morphology and developmental pattern during primary culture in vitro (Thibodeaux et al., 1991). They do, however, support early embryonic development effectively following prolonged in-vitro culture. Pre-pubertal oviduct cells can also be grown, although spontaneous alterations in morphology and growth are observed after several passages, with passage number dependent mainly upon the species.

Several variations of intratubal transfer have been proposed: gamete intrafallopian transfer, pronuclear stage tubal transfer/zygote intrafallopian transfer and tubal embryo stage transfer. Evidence of fertilization has been found for pronuclear stage tubal transfer/zygote intrafallopian transfer; with tubal embryo stage transfer, embryos have already been submitted to a period of in-vitro stress. Risks associated with further invasive laparoscopic intervention and anaesthesia have been reported. Transvaginal intrafallopian transfer techniques were introduced during the early 1990s as a promising alternative (Woolcott et al., 1994), but tubal transfers then became a ‘forgotten’ assisted reproduction technique (Tournaye et al., 1996), superseded by improved techniques and technology for in-vitro culture (and co-culture). Transcervical intrauterine transfer does not require anaesthesia or sedation, and is far less stressful. A comparison of zygote intrafallopian transfer with intrauterine transfer of blastocysts that had been co-cultured on feeder cell layers (Ménézo and Janny, 1996) demonstrated the efficacy of embryo-feeder cell interactions for selecting the best embryos, without introducing the stress of surgical intervention or anaesthesia (Bongso et al., 1992). Infection or inflammation in the fallopian tube will negate any potential benefit of tubal transfers. That is why coculture with ‘clean and controlled’ tubal cells have been proven to be a good compromise (Bongso et al., 1992).

A comparison between Vero cells (green monkey kidney cell line, from the same embryonic mesonephric origin) and genital tract cells reveals that sperm cells have different movement characteristics when cultured with the two different types of cells (Guerin et al., 1991). Hyperactivation rates are higher in culture with genital tract cells, and hyperactivated sperm have long lasting ‘star spin’ trajectories. The same features are observed in the presence of the corresponding conditioned media. Sperm cells must bind to, and then be released from the oviductal epithelium (Pollard et al., 1991). Binding and release is the result of interplay between aminoglycans and compounds containing sulfhydryl groups (Talevi et al., 2007). Oviductal cells have the capacity to release glutathione and other sulfhydryl-containing reducing compounds of small molecular weight (Ouhibi et al., 1990). This provides better conditions for sperm survival, as well as protection against decays in sperm DNA. Sperm motility is largely impaired by the generation of free radicals from mitochondria: this can be rescued by thiol-reducing substances (Aitken et al., 2012). With amino acids, glycine is highly concentrated in tubal fluid; sperm cells have an active glycine receptor, important for the acrosome reaction. The metabolism of other amino acids in sperm cells has not been studied, but again the one-carbon cycle involving homocysteine is of major importance for the initial processes of reproduction (Ménézo et al., 2013). The environment of the fertilization site is highly protective against ROS (Guérin et al., 2001), but this has to be regulated to facilitate and not prevent ROS-dependent processes, such as the acrosome reaction (de Lamirande and Gagnon, 1995). A ‘reducing stress’ may also have deleterious effects on sperm. Glucose, and in particular lactate, are compounds relevant to sperm vitality, and bicarbonate allows sperm activation. A dynamic interaction occurs between spermatozoa and compounds secreted by the oviduct. Seminal fluid can influence sperm survival and competence: impaired fertility can be caused by a reduction in the motility and survival of sperm, resulting in reduced fertilization (Peitz, Olds-Clarke, 1986, Robertson, 2007).

Arrival of spermatozoa within the oviduct is now known to regulate gene expression in oviductal epithelial cells, inducing up- and down-regulation of various proteins. Moreover, it seems that the oviduct is able to respond selectively to the Y or the X chromosome. This implies that the female should somehow be able to select offspring gender (Almiñana et al., 2014). In the human oviduct, all of these positive influences can be jeopardized by the presence of a hydrosalpinx, which is considered to be highly toxic, especially through an increase in levels of cytokines (Srivastava et al, 1996, Strandell et al, 2004).

It has been shown that co-culture of embryos with oviductal cells improves developmental potential compared with culture in conventional media (Kattal et al., 2008); however, conventional media poorly reflect the physico-biochemistry of in-vivo conditions, and successfully passing a mouse embryo assay involves several biases in perspective. Whether oviduct cells are superior to other cells with a common embryological origin such as Vero or Madin–Darby bovine kidney, cells is not clear. Optimal feeder layers consist of transport epithelia under conditions of growth inhibition: this phenomenon allows release of active molecules without exhausting the media and excessively decreasing the pH.

Co-culture feeder cells modify the medium and influence intermediate embryonic metabolism. Ouhibi et al. (1990) observed that the glucose concentration is decreased, with lactate formation. Feeder cells also secrete some small molecules that protect against ROS. Lipids are metabolized by beta-oxidation, which requires carnitine; this is introduced via follicular fluid, and is then produced in small amounts by the oviduct (Menezo and Laviolette, 1972). Metabolism of lipids also allows large quantities of ATP to be generated with minimal production of ROS (Ménézo et al., 2013).

The role of growth factors and interleukins continues to be debated. The first problem in elucidating a possible role is to determine whether they are involved in the final stages of establishing oocyte competence, or if they are involved in a mitogenic/antiapoptotic process, as is the case for growth hormones and insulin-like growth factors in the bovine system (Menezo and Elder, 2011). The human oocyte has surface receptors for IL6 and LIF; the same receptor (GP130) is present in the oviduct and uterus, and can also be found in efficient coculture systems of oviduct, uterine and Vero cells (Menezo and Elder, 2011). Growth factors, however, act in groups, with both positive and negative regulatory actions, and it is difficult to mimic any positive effects in vitro. The addition of a single growth factor (granulocyte macrophage colony stimulating factor) has been shown to have a limited action (Siristatidis et al., 2013); there is a balance between growth factors with positive and negative actions. Recent observations by Bromfield et al. (2014) indicate that seminal plasma may interfere with this interplay. Absence of seminal plasma at the time of mating results in downregulation of some growth factors, with an effect on the progeny. Seminal plasma growth factors that normally modulate the environment of the oviduct affect the response of oviductal cells in the secretion of growth factor and interleukin. In such a case, transfer of early embryos into the oviduct leads to defective progeny, and in-vitro culture in classical culture media partly alleviates this negative effect. This demonstrates that manipulation of growth factors in culture media is not easy: all of their fine regulation during embryo development is far from understood. Transfer of embryos in lesbian women (theoretically not submitted to the influence of seminal plasma), however, has shown no negative effects, and this leads to two possible conclusions: generalizations of the effect described above cannot be extended to all mammals, or this situation does not apply in humans owing to anatomical specificities. In conclusion, all of the proteins (e.g. growth factors and cation transporters) act in synergy, and it is extremely difficult to attribute a single specific ‘embryotrophic’ role to any one of them.

In conclusion, in-vitro culture conditions have been shown to affect embryo development negatively in any animal model studied (Lee et al, 2002, Ménézo, 2006, Schwarzer et al, 2012). This may explain why the overall take home baby rate and collected oocyte is still below 10%. (Patrizio and Sakkas, 2009). As improving the quality of the retrieved gametes seems to be difficult, if not impossible, our goal is to avoid compromising their quality. Numerous differences in physiology and biochemistry make the mouse model inappropriate for human embryo or other animal models, although some information can be taken into account for purposes of toxicology (Ménézo and Hérubel, 2002). In order to pass a mouse embryo assay for quality-control purposes, culture media prepared for human embryos usually has an osmolarity that is physiologically lower than it should be, and this causes an influx of water into the pre-implantation embryo. Low osmolarity decreases blastocyst formation in bovine embryos (Liu and Foote, 1996), and the effect of applying this apparently irrational strategy to human embryo culture is unforeseeable in the short or medium term. A similar observation can be made in terms of protection against reactive oxygen species: currently available culture media have no protection against ROS insults, and in fact spontaneously generate ROS (Martín-Romero et al., 2008), in contrast to the high level of ROS protection observed in the oviduct (Guérin et al., 2001). Molecules such as carnitine allow the oviduct to balance its metabolism, and this is absent from the embryo. Theories surrounding culture media composition, such as total omission of glucose and certain ‘essential’ amino acids, have generated culture systems that may be potentially dangerous for the embryo and, more worrying, for babies. The fact that it is impossible to study the in-vivo conceived embryo has led to a lack of understanding of real-time influences during the early pre-implantation stages. Greater efforts should be made to mimic the natural in-vivo environment: the physiology and biochemistry of the oviduct should not be overlooked, and merits re-assessment in the current perspective of human embryo culture.

Section snippets

Professor Yves Ménézo is a biochemist who obtained his PhD in applied biology from the University of Lyon. He received his Doctor of Science in 1979. He was Director of Research at the National Institute of Agronomical Research in France and Associate Professor at the Louisiana State University between 1987 and 1993, and then Head of the Assisted Reproduction and Genetics unitat the Merieux Foundation. He has developed several patents on culture media and hormone treatments. He has authored

References (59)

  • LippesJ. et al.

    Human oviductal fluid proteins

    Fertil. Steril

    (1981)
  • Martín-RomeroF.J. et al.

    Contribution of culture media to oxidative stress and its effect on human oocytes

    Reprod. Biomed. Online

    (2008)
  • MénézoY. et al.

    Soluble HLA-G release by the human embryo: an interesting artefact?

    Reprod. Biomed. Online

    (2006)
  • MénézoY.J.

    Paternal and maternal factors in preimplantation embryogenesis: interaction with the biochemical environment

    Reprod. Biomed. Online

    (2006)
  • MénézoY.J. et al.

    Mouse and bovine models for human IVF

    Reprod. Biomed. Online

    (2002)
  • PatrizioP. et al.

    From oocyte to baby: a clinical evaluation of the biological efficiency of in vitro fertilization

    Fertil. Steril

    (2009)
  • VishwakarmaP.

    The pH and bicarbonate-ion content of the oviduct and uterine fluids

    Fertil. Steril

    (1962)
  • AitkenR.J. et al.

    Sperm motility is lost in vitro as a consequence of mitochondrial free radical production and the generation of electrophilic aldehydes but can be significantly rescued by the presence of nucleophilic thiols

    Biol. Reprod

    (2012)
  • AlmiñanaC. et al.

    The battle of the sexes starts in the oviduct: modulation of oviductal transcriptome by X and Y-bearing spermatozoa

    BMC Genomics

    (2014)
  • BaltzJ.M.

    Media composition: salts and osmolality

    Methods Mol. Biol

    (2012)
  • BauersachsS. et al.

    Monitoring gene expression changes in bovine oviduct epithelial cells during the oestrous cycle

    J. Mol. Endocrinol

    (2004)
  • BetteridgeK.J. et al.

    Embryo transport through the mare's oviduct depends upon cleavage and is independent of the ipsilateral corpus luteum

    J. Reprod. Fertil. Suppl

    (1979)
  • BromfieldJ.J. et al.

    Maternal tract factors contribute to paternal seminal fluid impact on metabolic phenotype in offspring

    Proc. Natl. Acad. Sci. U.S.A.

    (2014)
  • CasslénB.G.

    Free amino acids in human uterine fluid. Possible role of high taurine concentration

    J. Reprod. Med

    (1987)
  • CheginiN.

    Oviductal-derived growth factors and cytokines: implication in preimplantation

    Semin. Reprod. Endocrinol

    (1996)
  • EhrenwaldE. et al.

    Bovine oviductal fluid components and their potential role in sperm cholesterol efflux

    Mol. Reprod. Dev

    (1990)
  • GuerinJ.F. et al.

    Movement characteristics and hyperactivation of human sperm on different epithelial cell monolayers

    Int. J. Androl

    (1991)
  • GuérinP. et al.

    Hypotaurine and taurine in gamete and embryo environments: de novo synthesis via the cysteine sulfinic acid pathway in oviduct cells

    Zygote

    (1995)
  • GuerinP. et al.

    Techniques de recolte et aminogrammes des liquides tubaire et folliculaire chez les femelles domestiques

    Rev. Med. Vet

    (1995)
  • Cited by (0)

    Professor Yves Ménézo is a biochemist who obtained his PhD in applied biology from the University of Lyon. He received his Doctor of Science in 1979. He was Director of Research at the National Institute of Agronomical Research in France and Associate Professor at the Louisiana State University between 1987 and 1993, and then Head of the Assisted Reproduction and Genetics unitat the Merieux Foundation. He has developed several patents on culture media and hormone treatments. He has authored more than 250 publications and book chapters and has received several awards such as laureate of the French Academy of Medicine and gold medal of the Institute Dexeus.

    View full text