Implications

The utility of artificial insemination in animal agriculture has dramatically improved production due to selective breeding. As with many reproductive technologies, artificial insemination bypasses the requirement for seminal fluid as a transport medium for sperm. These technologies demonstrate that seminal fluid is not required for pregnancy; however, it is curious that seminal fluid has a substantial effect on the female reproductive tract at insemination. This article discusses the role of seminal fluid in modulating the maternal environment during early pregnancy. Recapitulation of these events during artificial insemination may further improve pregnancy outcomes and offspring performance of domestic species.

A brief history of artificial insemination

Approximately 100 years after the invention of the microscope, Antonie van Leeuwenhoek was the first to describe the observation of living spermatozoa in 1677 using his own microscope design. Leeuwenhoek describes observing a fresh human ejaculate ‘before six beats of the pulse had intervened’ containing what he describes as ‘a great number of living animalcules’; referring to sperm (Letter to William Brouncker of the Royal Society, November 1677). Leeuwenhoek’s collection of fresh semen was produced ‘without sinfully defiling myself, [sic] what remains after conjugal coitus’. It would take another 100 years before the first successful attempt at artificial insemination was achieved by the Italian physiologist Lazzaro Spallanzani in 1784. Although Spallanzani considered sperm cells to be parasites contained within semen, he successfully executed artificial insemination in a bitch in heat that subsequently gave birth to three puppies. The success of this procedure was likely associated with the protracted oestrus observed in dogs as little was understand about ovulation and the oestrous cycle at the time. It was not until the end of the 19th century that practical approaches for artificial insemination were developed in Russia by Ilya Ivanovich Ivanoff and continued by Milovanov who refined artificial insemination practices and developed the artificial vagina for semen collection. It was reported that the growth of artificial insemination in the Russian cattle industry grew from 19 970 insemination in 1930 to over 1.5 million insemination in 1939 (Pincus, Reference Pincus1938). The development of sperm cryopreservation techniques by Christopher Polge further increased the capacity to transport semen long distance and dramatically improved domestic animal genetics (Polge et al., Reference Polge, Smith and Parkes1949; Polge, Reference Polge1952). Currently, it is estimated that 72% of all dairy cows in the USA are bred by artificial insemination (United States Department of Agriculture -- National Institute of Food and Agriculture). A consistency throughout these later advances in artificial insemination was the utilization of semen extenders to increase viability of semen, and increase the number of potential inseminations from a single ejaculate. As a consequences of semen extension, seminal fluid has been diluted in semen used for artificial insemination since the 19th century. Could replacement or enrichment of seminal fluid components enhance the success of artificial insemination in domestic species?

The role of insemination beyond sperm delivery

Of course the specific objective of insemination is the delivery of male gametes into the female reproductive tract to facilitate fertilisation of female gametes. However it is interesting to consider the cellular and biochemical content of semen as a whole. Indeed seminal fluid (the acellular fraction of semen) derived from the male accessory glands is rich in simple sugars, buffers, antioxidants, hormones and proteins of unknown function presumed to be present simply to facilitate sperm survival and transport through the female reproductive tract. Research has now begun to highlight the importance of some of these seminal fluid proteins as potential mediators of paternal–maternal communication delivered at the time of insemination. Consider briefly lower-order organisms such as crickets, mosquitoes and flies where physiological and behavioural changes associated with reproductive outcomes have been demonstrated in females after exposure to seminal fluid (Avila et al., Reference Avila, Sirot, LaFlamme, Rubinstein and Wolfner2011). Observations dating back to the 1960s have demonstrated the acute potential of semen to modulate the cellular environment of the female reproductive tract of mice, human, cattle, swine, horse and sheep. Following insemination in rodents an acute influx of leucocytes is observed for the proceeding 72 h (Yanagimachi and Chang, Reference Yanagimachi and Chang1963; Mattner, Reference Mattner1968; De et al., Reference De, Choudhuri and Wood1991; McMaster et al., Reference McMaster, Newton, Dey and Andrews1992; Robertson et al., Reference Robertson, Mau, Tremellen and Seamark1996). This influx of leucocytes is paralleled by an increase in the expression of inflammatory mediators by the endometrium, including C--C motif ligand (CCL)2, CCL3, CCL5 and colony-stimulating factor (CSF)2. (Robertson and Seamark, Reference Robertson and Seamark1992; Robertson et al., Reference Robertson, Mau, Hudson and Tremellen1997). Further studies in the rodent have been able to demonstrate that seminal fluid is the active component of the ejaculate to elicit these changes observed in the maternal tissues, whereas specifically seminal vesicle derived transforming growth factor beta (TGFβ) has been shown to be the active compound in seminal fluid responsible for the increased expression of endometrial inflammatory mediators and ultimately post-insemination inflammation (Robertson et al., Reference Robertson, Mau, Tremellen and Seamark1996; Tremellen et al., Reference Tremellen, Seamark and Robertson1998). Similarly in humans a post coital inflammatory reaction has been observed in the cervix following exposure to semen where no inflammation is observed following condom protected intercourse (Sharkey et al., Reference Sharkey, Macpherson, Tremellen, Mottershead, Gilchrist and Robertson2012b). In parallel with the mouse, seminal fluid derived TGFβ is responsible for inducing increased expression of the inflammatory mediators interleukin-6 and CSF-2 in human cervical epithelial cells (Sharkey et al., Reference Sharkey, Tremellen, Jasper, Gemzell-Danielsson and Robertson2012a). In cattle a similar inflammatory response to semen has been demonstrated (Mahajan and Menge, Reference Mahajan and Menge1967; Mattner, Reference Mattner1968); however it is important to note that the seminal vesicles of the bull contribute roughly half the volume of ejaculated semen. In fact when the seminal vesicle glands are surgically removed from bulls, natural fertility remains high at approximately 65% conception (Faulkner et al., Reference Faulkner, Hopwood and Wiltbank1968). A more recent study aimed to evaluate the benefit of seminal fluid (or TGFβ) supplementation at the time of artificial insemination on pregnancy rates in cattle. Although statistically underpowered, the study suggests that artificial insemination supplemented with seminal fluid or TGFβ can improve pregnancy rates, particularly in poor performing herds (Table 1) (Odhiambo et al., Reference Odhiambo, Poole, Hughes, Dejarnette, Inskeep and Dailey2009). Seminal fluid infusion into the porcine uterus induces significant cellular inflammation 36 h after infusion that was still evident 8 days later, considerably different than the acute inflammation observed in other species (O’Leary et al., Reference O’Leary, Jasper, Warnes, Armstrong and Robertson2004). The same research team described a significant increase in the number of total and viable embryos collected from sows following seminal fluid supplementation (O’Leary et al., Reference O’Leary, Jasper, Warnes, Armstrong and Robertson2004). The horse and sheep also show increased acute inflammation of the endometrium after the application of seminal fluid or semen (Mattner, Reference Mattner1969; Scott et al., Reference Scott, Ketheesan and Summers2006; Palm et al., Reference Palm, Walter, Budik, Kolodziejek, Nowotny and Aurich2008).

Table 1 Pregnancy rates in cattle treated with seminal fluid at the time of artificial insemination

Adapted from Odhiambo et al. (Reference Odhiambo, Poole, Hughes, Dejarnette, Inskeep and Dailey2009). Seminal fluid was collected from a single bull for beef studies and six Holstein bulls and combined for dairy studies. Pregnancy was diagnosed at 35 to 40 days post insemination.

The question remains, what is the relevance of this post-insemination, seminal fluid induced inflammatory reaction? A proposed role would be the prophylactic clean-up of sexually transmitted pathogens, or non-viable sperm cells. It is interesting to note that many inflammatory mediators upregulated in the endometrium or oviduct by seminal fluid are also embryotrophic in nature, specifically CSF-2, leukemia inhibitory factor and IL-6 (Lavranos et al., Reference Lavranos, Rathjen and Seamark1995; de Moraes and Hansen, Reference de Moraes and Hansen1997; Gutsche et al., Reference Gutsche, von Wolff, Strowitzki and Thaler2003; Bromfield et al., Reference Bromfield, Schjenken, Chin, Care, Jasper and Robertson2014; Hansen et al., Reference Hansen, Dobbs and Denicol2014). The temporal expression of these so-called embryokines may be in part regulated by seminal fluid exposure to orchestrate embryo development coordinate with insemination. An even more intriguing relevance of this inflammatory event relates to the induction of maternal immune modulation required for pregnancy success in viviparous species.

Immune modulating capacity of semen

As mentioned previously, Medawar hypothesized a requirement for suppression or modulation of maternal immunity to facilitate the survival of the allogeneic conceptus. There is potential for this immune modulation to be orchestrated by ovarian or placental hormones, or the conceptus itself. However, neither of these scenarios allow for the potential maternal immune adaptions to be specific toward the paternal antigens expressed by the conceptus and/or be in place at the time of embryo implantation (at least in rodents and humans). An intriguing possibility remains that insemination could act as a first ‘priming’ event of the maternal immune system to paternal antigen potentially expressed by the conceptus. The underlying mechanisms of immune tolerance required for pregnancy are proposed to be clonal deletion, anergy and clonal unresponsiveness of alloreactive T lymphocytes. These mechanisms would prevent the cytotoxic actions of specific alloreactive lymphocytes within the peripheral circulation during pregnancy (Piazzon et al., Reference Piazzon, Matusevich, Deroche, Nepomnaschy and Pasqualini1985). In many mucosal tissues, the prevalence of a Th2 skewed immune response is associated with a state of functional tolerance and this is likely to also be the case in pregnancy (Chaouat et al., Reference Chaouat, Tranchot Diallo, Volumenie, Menu, Gras, Delage and Mognetti1997).

Seminal fluid has been demonstrated to potentiate changes in immune function of T cells, B cells, NK cells and macrophages in the mouse, bovine and human (Anderson and Tarter, Reference Anderson and Tarter1982; Fahmi et al., Reference Fahmi, Hunter, Markham and Seguin1985; Saxena et al., Reference Saxena, Jha and Farooq1985). It is evident that exposure to semen, and indeed seminal fluid, drives an acute hypertrophy in the spleen and lymph nodes draining the uterus in the mouse (Maroni and de Sousa, Reference Maroni and de Sousa1973; Beer and Billingham, Reference Beer and Billingham1974; Johansson et al., Reference Johansson, Bromfield, Jasper and Robertson2004). The quality of any immune response, including the phenotypes of effector T cells and state of the cytokine profile is determined at the time of primary antigen exposure and is dependent on the activation state of antigen presenting cells (Constant and Bottomly, Reference Constant and Bottomly1997; Kapsenberg et al., Reference Kapsenberg, Hilkens, Wierenga and Kalinski1999). It has been suggested that the site of lymphocyte activation is of major significance to the functionality downstream effector cells (Harper et al., Reference Harper, Cochrane and Williams1996). However, the majority of data supports the idea that antigen presenting cells play a fundamental role in the programming of lymphocytes and that local cytokine expression is the key factor in regulating antigen presenting cell behaviour (Harper et al., Reference Harper, Cochrane and Williams1996; Constant and Bottomly, Reference Constant and Bottomly1997; Kapsenberg et al., Reference Kapsenberg, Hilkens, Wierenga and Kalinski1999; Egan et al., Reference Egan, Yorkey, Black, Loh, Stevens, Storozynsky, Lord, Frelinger and Woodward2000). Could it be that activation of specific cells in the draining lymph nodes of the reproductive tract may help to prime the maternal immune system with paternal antigen?

The frequency of antigen exposure is thought to work in conjunction with dose in the generation of mucosal tolerance. One-off high dose exposures or small repeated doses of antigen has been shown to be most beneficial in the development of tolerance (Garside and Mowat, Reference Garside and Mowat2001); a paradigm that fits with the exposure of the uterine epithelium to seminal antigens during intercourse. Although research has demonstrated that lymphocyte populations become anergic to paternal antigens during pregnancy (Tafuri et al., Reference Tafuri, Alferink, Moller, Hammerling and Arnold1995), an elegant study demonstrated that this hyporesponsiveness is achieved in a paternal-specific manner (Robertson et al., Reference Robertson, Mau, Hudson and Tremellen1997). Robertson et al. demonstrated that tumour growth in female mice could be induced if mated to males with a matching major histocompatibility complex (MHC) haplotype to that of the introduced tumour cell line. The utilization of uterine ligation before mating in this model also excluded the possibility that the conceptus was responsible for the systemic changes to immune tolerance observed. Tumour growth in virgin mice or those mated to a disparate MHC haplotype to the tumour was inhibited (Robertson et al., Reference Robertson, Mau, Hudson and Tremellen1997). This provides direct evidence that exposure to semen can induce systemic immune tolerance to potential paternal antigens.

Seminal fluid is rich in immune-deviating cytokines such as TGFβ and PGE2 which can lead to the alteration of the cytokine profile of a T cell population in the Th2 direction thought to be beneficial to pregnancy success (Tafuri et al., Reference Tafuri, Alferink, Moller, Hammerling and Arnold1995). In a landmark experiment depletion of forkhead box P3 (FOXP3) positive T regulatory cells lead to a complete failure in pregnancy (Aluvihare et al., Reference Aluvihare, Kallikourdis and Betz2004). TGFβ has been demonstrated to activate FOXP3 positive T regulatory cells in vitro (Fantini et al., Reference Fantini, Becker, Tubbe, Nikolaev, Lehr, Galle and Neurath2005). Interestingly our own studies have demonstrated that seminal fluid exposure plays a significant role in the generation and recruitment of FOXP3 cells into female reproductive tissues (Robertson et al., Reference Robertson, Guerin, Bromfield, Branson, Ahlstrom and Care2009; Guerin et al., Reference Guerin, Moldenhauer, Prins, Bromfield, Hayball and Robertson2011).

The impact of semen on pregnancy outcomes: a role in assisted reproduction and pathology

It is clear that seminal fluid is not required for pregnancy. With the advent of artificial insemination, in vitro fertilisation and intracytoplasmic sperm injection, the sperm cell is the only requirement of the ejaculate to achieve a viable pregnancy. With that being said and the preceding discussion, it has come to light that seminal fluid may play a role in improving pregnancy outcomes and potentially staving off particular pathologies of pregnancy. Recently, we have been able to demonstrate in mice that an absence of seminal fluid exposure during mating results in reduced embryo development, poor placentation and metabolic perturbations in offspring (Bromfield et al., Reference Bromfield, Schjenken, Chin, Care, Jasper and Robertson2014). We conclude that an absence of seminal fluid resulted in foetal programming due to reduced secretion of seminal fluid induced embryokines in the oviduct, altered tissue remodelling resulting in poor placentation, and perturbed maternal tolerance toward the allogeneic conceptus, all culminating in altered offspring phenotype. The immunomodulatory properties of seminal fluid have been demonstrated to be detrimental in an experimental model of endometriosis. It was demonstrated that human endometriosis lesion growth was increased in the nude mouse after exposure to seminal fluid (McGuane et al., Reference McGuane, Watson, Zhang, Johan, Wang, Kuo, Sharkey, Robertson and Hull2015). Epidemiological evidence in humans has suggested a potential role for semen exposure in modulating pathologies of pregnancy with suspected immunological aetiologies. Data suggests that semen exposure in a partner specific manner can be beneficial in reducing preeclampsia, a pathology with suspected immune aetiology. Reducing semen exposure with the use of barrier contraception or by short term cohabitation increased the risk of women developing preeclampsia (Klonoff-Cohen et al., Reference Klonoff-Cohen, Savitz, Celafo and McCann1989; Robillard et al., Reference Robillard, Hulsey, Perianin, Janky, Miri and Papiernik1995). Even more compelling, a randomized controlled trial in 87 women with recurrent spontaneous abortion suggests that pregnancy rates can be significantly improved by the administration of vaginal capsules containing seminal fluid (Coulam and Stern, Reference Coulam and Stern1995). The addition of seminal fluid during artificial insemination in cattle was shown to increase pregnancy rates by nearly 5%, albeit not significantly (Table 1) (Odhiambo et al., Reference Odhiambo, Poole, Hughes, Dejarnette, Inskeep and Dailey2009). It is important to consider that an increase in pregnancy rate of 5% in an agricultural context could have enormous economic and production impacts to producers.

As the utility of IVF increases in human medicine and agricultural practice it is easy to overlook the understudied effects of in vitro culture on offspring health. Indeed, IVF and embryo transfer technologies exist in the absence of semen or seminal fluid. In both humans and cattle the impacts of in vitro embryo culture appear to carry negative consequences including increased risk of premature birth, very low birth weight, complications during delivery, serious birth defects in humans and overgrowth in cattle resulting in major organ defects (Young et al., Reference Young, Sinclair and Wilmut1998; Perri et al., Reference Perri, Chen, Yoeli, Merlob, Orvieto, Shalev, Ben-Rafael and Bar-Hava2001; Hansen et al., Reference Hansen, Kurinczuk, Bower and Webb2002; Schieve et al., Reference Schieve, Meikle, Ferre, Peterson, Jeng and Wilcox2002; Wang et al., Reference Wang, Norman and Kristiansson2002; Ochsenkuhn et al., Reference Ochsenkuhn, Strowitzki, Gurtner, Strauss, Schulze, Hepp and Hillemanns2003). Collectively these perturbations of in vitro culture are a consequence of our failure to recapitulate the maternal developmental environment of the embryo. It is interesting to surmise that the developmental environment of the embryo can be altered by exposure to semen. The inflammatory mediator CSF-2 is an example of a well-studied embryokine with the potential to increase embryonic development in rodents, cattle and humans (Sjoblom et al., Reference Sjoblom, Roberts, Wikland and Robertson1999 and Reference Sjoblom, Wikland and Robertson2005; Ziebe et al., Reference Ziebe, Loft, Povlsen, Erb, Agerholm, Aasted, Gabrielsen, Hnida, Zobel, Munding, Bendz and Robertson2013; Siqueira et al., Reference Siqueira, Tribulo, Chen, Denicol, Ortega, Negron-Perez, Kannampuzha-Francis, Pohler, Rivera and Hansen2017). In parallel, CSF-2 is also one of the most highly upregulated molecules in the endometrium or oviduct following seminal fluid exposure (Robertson et al., Reference Robertson, Mau, Tremellen and Seamark1996; Sharkey et al., Reference Sharkey, Macpherson, Tremellen, Mottershead, Gilchrist and Robertson2012a and Reference Sharkey, Tremellen, Jasper, Gemzell-Danielsson and Robertson2012b; Bromfield et al., Reference Bromfield, Schjenken, Chin, Care, Jasper and Robertson2014). Two small studies have even suggested that exposure to semen by intercourse around the time of embryo transfer can improve pregnancy rates in women (Marconi et al., Reference Marconi, Auge, Oses, Quintana, Raffo and Young1989; Tremellen et al., Reference Tremellen, Valbuena, Landeras, Ballesteros, Martinez, Mendoza, Norman, Robertson and Simon2000). The implication for a simple intervention to potentiate positive reproductive or production measures should be considered for use in agricultural industries like dairy and swine where artificial insemination with minimal seminal fluid exposure is routine.

Potential manipulation of paternal–maternal communication for agriculture

Assisted reproductive technologies including ovarian synchronization, semen collection, artificial insemination, in vitro fertilisation and embryo transfer have been extremely important to the economic and productive success of a number of domestic species. These technologies have allowed producers to rapidly improve genetic merit of animals and increase productivity in an ever demanding climate. The utility of these technologies is so well utilized now that a number of studies have demonstrated that in vitro fertilisation and embryo transfer technology outperform artificial insemination in regard to pregnancy rates in the dairy cow (Vasconcelos et al., Reference Vasconcelos, Jardina, Sa Filho, Aragon and Veras2011; Pellegrino et al., Reference Pellegrino, Morotti, Untura, Pontes, Pellegrino, Campolina, Seneda, Barbosa and Henry2016). In regard to these studies it is important to consider that much of the reported embryo loss in the dairy cow occurs within the 1^st week of pregnancy (Wiltbank et al., Reference Wiltbank, Baez, Garcia-Guerra, Toledo, Monteiro, Melo, Ochoa, Santos and Sartori2016), and therefore in vitro fertilisation and embryo transfer may be a simple means to bypass this period of embryo vulnerability. Nevertheless, it is interesting to note that studies in rodents suggest that seminal fluid can alter the developmental environment of the oviduct by increasing expression of embryokines (Bromfield et al., Reference Bromfield, Schjenken, Chin, Care, Jasper and Robertson2014). In cattle, gene expression of the oviduct does not appear to be responsive to the presence of a developing embryo or even change from that described at oestrus (Maillo et al., Reference Maillo, Gaora, Forde, Besenfelder, Havlicek, Burns, Spencer, Gutierrez-Adan, Lonergan and Rizos2015; Maillo et al., Reference Maillo, de Frutos, O’Gaora, Forde, Burns, Spencer, Gutierrez-Adan, Lonergan and Rizos2016). However neither of these studies considered the potential implications of seminal fluid in modulating the environment of the oviduct.

If we aim to recapitulate the natural developmental environment of the oviduct and uterus in domestic species to optimize reproductive technologies and postnatal development of offspring, we must endure to remember that such an environment is not that of artificial insemination but that of live cover where the female reproductive tract is exposed to male derived factors including seminal fluid. We hope to expand our understanding of how seminal fluid contributes to pregnancy success in domestic species by better understanding the potential of paternal–maternal communication as it pertains to embryo development, foetal growth and immune modulation required for pregnancy success.