ACCEPTED AUTHOR VERSION OF THE MANUSCRIPT: Reproductive performance of dairy cows fed a diet supplemented with n-3 polyunsaturated fatty acids – a review

The aim of this review is to summarize the current state of knowledge of reproductive performance of high-yielding dairy cows fed a diet supplemented with n-3 fatty acids (FAs), and to recommend a feeding schedule that can increase the success rate in reproduction. Dietary supplementation with fat, especially FA, has become an accepted strategy for supporting the fertility of high-yielding dairy cows. The two main categories of FAs, unsaturated fatty acids (UFAs) and saturated fatty acids (SFAs), exert distinct effects on reproductive functions, with UFA having a conclusively beneficial impact. Polyunsaturated fatty acids (PUFAs) are of particular importance on account of their biological properties. Standard feedstuffs (such as soybean) are rich in n-6 FAs, whereas few feedstuffs contain n-3 FAs. Neither the n-3 nor n-6 PUFAs are synthetized by the animal. Several field experiments have indicated that the recommended schedule of n-3 supplementation should last 160–170 days, from the eighth week before calving to 100 or 120 day postpartum. Supplementation of the diet thus covers the period from the late preantral stage of the follicle, the entire development of the antrum, oocyte ovulation, and fertilization, as well as the development of the preattachment embryo and its implantation. The following effects are among the important beneficial results of supplementation with n-3 FAs: a higher number of antral follicles suitable for ovum pick-up (OPU), larger dominant follicles and corresponding CL, better quality oocytes and embryos, and higher implantation rates resulting from improved uterine environment (e.g., reduced synthesis of luteolytic PGF2α). We conclude that dietary supplementation for high-yielding dairy cows with n-3 FAs supports ovarian functions and helps the embryo to survive in the adverse environment of the reproductive tract of the postpartum cow.

From the economic point of view, milk production is one of the most challenging activities in animal husbandry. A range of factors affect the economics of a dairy farm, including genetics, nutrition, and reproduction (Bryszak et al., 2019;Chudleigh et al., 2019).
Management of dairy cow fertility is a very demanding task, considering the negative impact of high milk yield on reproductive performance (Rearte et al., 2018;Albarran-Portillo and Pollot 2013;Melendez and Pinedo 2007). Special attention needs to be paid to the negative energy balance (NEB) that exists during the 5-8 week postpartum (pp) period (Leroy et al., 2008a). Nutrition plays a pivotal role in maintaining animal welfare and good health, but is of special importance in high-yielding dairy cows, which need well-balanced rations to provide the energy for both milk production and the reproductive processes. Effective reproduction, in the form of embryo implantation and fetal development, results from an optimal environment that supports ovulation of a good-quality oocyte, its correct fertilization, and the implantation of a healthy embryo.
Dietary supplementation with fat, especially FA, has become an accepted strategy for supporting fertility management and improving the quality of animal products (Rodney et al., 2015;Bryszak et al., 2019;Klebaniuk et al., 2017). There are two main categories of FAs: unsaturated fatty acids (UFAs) contain at least one carbon-carbon double bond, while saturated fatty acids (SFAs) lack double bounds. The UFA family consists of monounsaturated fatty acids (MUFAs) with a single double bond, and polyunsaturated fatty acids (PUFAs) with at least two double bonds. The literature shows that the effect of PUFAs on animal health and productivity is more pronounced than that of SFAs (Freret et al., 2019).
Generally, SFAs have a negative effect on reproductive traits, and especially on oocyte and embryo quality (Marei et al., 2016).
There are three major PUFA subfamilies, depending on the position of the double bonds: omega-3 (n-3), which includes docosahexaenoic acid (DHA); omega-6 (n-6), which includes arachidonic acid (AA); and omega-9 (n-9), which includes oleic acid. One very important property of the n-3 and n-6 PUFAs is that they cannot be synthetized by the animal, so they must be provided in the diet (Matras et al., 2014). The feedstuffs commonly used for dairy cows are rich in n-6 FAs, whereas the sources of n-3 FAs are essentially limited to flaxseed and fish oil (Mattos et al., 2000). The n-6:n-3 ratio has been widely utilized to represent the balance of the two exogenous PUFAs. Generally, the lower the ratio, the better the value of the feedstuff, as the higher the n-3 content. For example, the best n-6:n-3 ratio is found for fish oil (0.08) and flaxseed (0.27), whereas soybeans (8.7) and rapeseed (16.7) are far less beneficial (Moallem, 2018). Supplementing the diet with components that have a low n-6:n-3 ratio, such as a mixture of plant extract and a blend of fish and soybean oils, has been found to increase levels of n-3 FAs in the blood and to improve milk quality, by lowering the n-6:n-3 ratio (Szczechowiak et al., 2016;Szczechowiak et al., 2018). Considering the importance of n-3 FA in many biological processes (such as reproduction and development of the central nervous system), its supplementation to the diet should be of special interest to dairy farmers (Moallem, 2018).

Summary of field experiments
Dietary supplementation with n-3 FAs has recently become a widely used strategy in feeding dairy cows. Soybeans are rich in n-6 FAs, and are a major protein component of the standard ruminant diet. Supplementing the standard (control) diet with n-3 FAs thus reduces the n-6:n-3 ratio and improves its nutritional value. Table 1 presents a survey based on 19 publications that present the results of field experiments, sorted by time of supplementation, and carried out on dairy cows fed diets supplemented with n-3 FAs. There are two major sources of n-3 FA available for cattle nutrition: flaxseed and fish oil. The amount of the n-3 components in the experimental diets ranged from 0.4% to 18.4% of the DM basis. Flaxseed rich in ALA was used in over 60% (12/19) of the experiments (Table 1). Fish oil, rich in EPA and DHA, was used in 20% of the field studies, whereas algae was used only occasionally (10%).
The particular question that arises concerns the optimal length and schedule of n-3 supplementation. There is great variation in the length of the supplementary period, which lasted from 21 to 300 days (Table 1). Dietary supplementation usually begins 6-8 weeks prior to calving and lasts on average until day 133 in milk. The traits that were monitored varied from study to study, but included fertility parameters such as pregnancy rate, open days, progesterone level, ovarian activity, and quality and quantity of oocytes and embryos. It has been shown that the length of supplementation does not directly determine the impact of n-3 on reproductive performance. For example, Bork et al. (2010) found no effect after supplementing the whole lactation, whereas Jolazadeh et al. (2019) supplemented cows only for 21 days prior to parturition, and observed shorter time to first estrus and to the first AI service. It should also be mentioned that three studies (16%, 3/19 experiments) reported a lack of any effect of n-3 on reproduction, despite long-term supplementation (e.g., 161 days by Matras et al., 2014).
According to the evidence that has been published, n-3 FAs can influence reproductive processes by affecting ovarian activity (such as follicular growth, follicular number and size, and CL functions) and the quality of oocytes and of the resulting embryos. These findings have been supported by the results of several embryo transfer (ET) and ovum pick-up procedures, demonstrating a higher quality of oocytes and embryos donated by females fed a diet supplemented with n-3 (Table 1). Only few studies reported a lack of effect from n-3 (Gandra et al., 2017).

Environment of oocyte growth and maturation in cows supplemented with n-3 FAs
A full understanding of follicular dynamics is crucial if experimental treatments, such as dietary supplementation with PUFAs, are to be successfully applied. Such efforts may result in improved quality of oocytes and embryos, and thus enhance reproductive performance. In cattle, the ovarian reservoir represented by the number of primordial germ cells is established around day 90 of pregnancy (Aerts and Bols, 2010a). Although about 2,100,000 germs cells are estimated to exist that time, this number is reduced to about 130,000 by birth. The time necessary for a bovine primordial follicle to reach the preovulatory stage is estimated at 180 days, which includes 42 days for the development of the antral follicle. This means that the follicle that is ovulating during the current cycle was activated from the ovarian reservoir over eight estrus cycles earlier. During this long period, the oocyte enclosed within the growing follicle undergoes several dynamic processes, including growth and maturation. The growing oocyte synthetizes and accumulates large quantities of factors crucial for fertilization and development of an embryo, which is referred to as a gradual acquisition of quality (Lonergan and Fair, 2016). Any detrimental factor interfering with this process may thus significantly reduce oocyte quality, and in consequence reproductive performance. On the other hand, any beneficial factor may support the oocyte in gaining quality (Leroy et al., 2012).
The process of follicle growth includes two main phases: the preantral (gonadotrophin-independent) and antral (gonadotrophin-dependent) phases (Aerts and Bols, 2010b). The primordial follicles are made up of quiescent oocytes at the first prophase of meiosis surrounded by a single layer of flattened granulosa cells. Interestingly, the primordial follicle is only partially isolated from the stroma, due to several irregular gaps in the wall (Zamboni, 1974). This results in direct contact between the oocyte and the blood, allowing metabolites to enter the follicle and affect the oocyte. After the primordial follicle has been activated, the primary follicle, containing cuboidal granulosa cells, is formed. Follicular growth is then accompanied by oocyte growth, including formation of the zona pellucida (zp) and by a rapid increase in the layers of cuboidal granulosa cells. As the follicle grows, the oocyte becomes more isolated on account of the formation of several components, such as the basal lamina, the blood-follicle barrier, the zp, and several layers of granulosa cells (Zamboni, 1974).
Bidirectional communication between the oocyte and the surrounding granulosa and cumulus cells occurs through the gap junction system, and is crucial for the cumulus-oocyte complex (COC), since both cell types coordinate the development of the follicle by nutrient transfer and reciprocal signaling (Monniaux, 2016). There are two processes in this cellular dialog: 1) the transit of small molecules (such as metabolites) through the gap junctions, and 2) the interaction of cytokine and growth factors produced by one cell type with specific receptors on the other cell (Monniaux, 2016). We can thus assume that the oocyte and follicle remain susceptible to environmental changes for the entire period of development, which is 180 days in cattle.
Several studies have shown dynamic changes in the lipid content of the ovarian follicle, illustrating the pivotal role of lipids in oocyte growth and maturation (Warzych and Lipinska, 2020). It has been shown that alterations in fatty acid intake are reflected to some extent in the FA profile of bovine follicular fluid, cumulus cell, and oocytes (Zachut et al., 2010). Besides, fatty acid profile and content in the follicular fluid are affected by many female-associated factors, such as nutrition and sexual maturity (Warzych et al., 2011;Pawlak et al., 2012). The oocyte can however be protected from FA fluctuations, as well as from the negative effects of saturated fatty acids (such as palmitic and stearic acid), by other follicular compartments, and especially by cumulus cells. Bovine COCs matured in vitro under lipotoxic conditions (with elevated levels of the palmitic, stearic, and oleic SFAs) displayed several disturbances in oocytes, corresponding to cumulus cells, and in resulting day-8 blastocysts (Marei et al., 2017). This negative impact was, however, significantly suppressed by supplementation of the IVM medium with the n-3 FA linolenic acid (ALA), which improved the viability of the cumulus cells by reducing apoptosis and cellular stress.
Consequently, blastocyst quality was also improved. Although cumulus cells suffer from pathological concentrations of FAs, causing lipid accumulation and oxidative stress, they at the same time protect the oocyte from lipotoxic effects (Lolicato et al., 2015).
In summary, the oocyte developing in the ovarian follicle is provided an optimal growth environment that includes cumulus cells and follicular fluid (FF). Dietary components affect the composition of the FF, whereas cumulus cells protect the oocyte against pathological conditions. The experimental factors provided with the diet can thus interfere with the oocyte gaining quality. Several field experiments using diets supplemented with n-3 FAs revealed them to have a stimulatory effect on follicular growth and oocyte quality ( Table   1). The most interesting observations were that there were higher numbers of antral follicles suitable for OPU, that the dominant follicle was larger, and that the oocytes collected were of higher quality (Moallem et al., 2013;Ulfina et al., 2015;Zachut et al., 2010;Elis et al., 2016;Badiei et al., 2014 from Table 1). There is thus clear evidence of a positive effect of exogenous n-3 FAs on the quality of bovine oocytes.

N-3 supplementation during in vivo and in vitro embryo development
There are two main experimental approaches to addressing the impact of dietary components on embryo quality: in vivo studies involving animals fed the experimental diet, and in vitro experiments focused on in vitro embryo culture (IVC) in supplemented media.
Although many in vivo studies have demonstrate a distinct effect of dietary PUFAs on bovine embryos, this effect cannot be separated from the physiological status of the cow (such as hormonal profile, female health, and reproductive status). On the other hand in vitro procedures allow the study of individual fatty acids in any arrangement and concentration, though it must be kept in mind that the in vitro culture environment is considered suboptimal.
Both approaches present their own set of advantages and disadvantages.
The experimental procedure of standard field experiments provides dietary supplementation over the period of 6-8 weeks before calving and several weeks in milk (Table 1). The entire process leading to pregnancy-the development of the antral follicle, oocyte maturation, ovulation and fertilization, and embryo development and implantation-is thus usually affected by the supplementation. Such experimental design does not allow to demonstrate whether supplemented factor is more crucial to the growing oocyte or to the developing embryo. From the point of view of the cost of the supplement and fodder preparation, it is important to determine the optimal supplementation schedule. By referring to the results of several field experiments, we were able to demonstrate that n-3 FAs significantly improve the quality of ovulating oocytes in cattle. We therefore turn now to the question of the quality of the preattachment embryo.
One experiment has combined in vivo and in vitro models to clearly demonstrate the positive effect of n-3 FAs on the development of early bovine embryos derived from oocytes growing in a standard environment (Marei et al., 2016). In that study, experimental diets were supplemented with either n-3 FAs or SFAs. The feeding sequence involved two groups of heifers, and included a four-week application of the n-3 diet followed by a four-week application of the SFA diet, and vice-versa. By the end of each four-week feeding period, blood samples were collected and the serum was added to the IVC medium for in vitro embryo culture. The experiment showed that the four-week treatment was sufficient to affect the FA and selected metabolites in serum from cows at each collecting point. Especially high levels of n-3 fatty acids were detected after four weeks of UFA feeding. With regard to IVC conditions, distinct effects of the two types of sera (n-3 and SFA) on embryo quality were reported. There was a clear, positive effect of the n-3 serum on blastocyst quality (higher total cell number and lower apoptotic index). The effect of SFA serum was opposite to that of n-3, since embryos cultured in IVC+SFA reached the blastocyst stage at a slower rate and exhibited a higher incidence of apoptosis.
According to one well-known hypothesis, the rate at which bovine embryos reach the blastocyst stage is mainly due to the intrinsic characteristics of oocytes, whereas blastocyst quality strongly depends on conditions of embryo growth or culture (Rizos et al., 2002). It must be emphasized, however that oocyte quality is a prerequisite for proper fertilization and

Reproductive tract
The majority of published field studies that have focused on the impact of dietary supplementation with n-3 FAs on the developmental competence of bovine embryos have succeeded in demonstrating that it has a positive effect. Two hypothetical mechanisms of action of n-3 have been suggested: that it acts on the embryo proper and that it acts on the female reproductive tract. It has been documented that dietary supplementation with n-3 FAs can improve the oviductal/uterine environment by supporting embryo quality (e.g., blastomere count) and CL functions.
It has been shown that the supplementation of dairy cow diets with polyunsaturated fatty acids can affect luteal function, either by direct action on progesterone production or by modifying the synthesis of eicosanoids, such as arachidonic acid (AA) (Mattos et al., 2000).
During diestrus, the endometrium accumulates AA, a known precursor of prostaglandins (PGF2α, PGE2). Prostaglandin F2α exerts an inhibitory effect on progesterone production and affects the length of the luteal phase (Lauderdale, 1974). A diet enriched with n-6 FA promotes PGF2α synthesis with proinflammatory and luteolytic action, whereas n-3 FA supplementation stimulates production of PGE2, which has known anti-inflammatory and antiluteolytic activity. Altering the n-6:n-3 ratio of the diet can thus have an impact on CL functions and thereby on reproductive performance. According to Greco et al. (2018), providing a similar quantity of fatty acids with different n-6:n-3 ratios (n-6:n-3 ratio of 4.0, 5.0 or 6.0) in dairy cows' diets resulted in modifications in the FA profile in the endometrium, as well as the pattern of prostaglandin synthesis and the expression of genes regulating prostaglandin biosynthesis, such as prostaglandin F synthase (PGFS). A higher n6:n-3 ratio was accompanied by increased PGFM concentration (PGF2α metabolite). These findings corroborate those of Caldari-Tores et al., (2006), who observed higher PGF2α secretion resulting from increased n-6:n-3 ratio.
Researchers have hypothesized that the beneficial effects of n-3 FAs may be associated with the more supportive uterine environment resulting from the anti-inflammatory properties of n-3, which allows the risk of subclinical metritis to be reduced (Cerri et al., 2009). Leese et al. (2008) suggested the endometrium as an important source of maternal diet components taken up by the embryo. Besides, some physiological functions of bovine oviductal epithelial cells-such as the cells' viability, count, and migration capacity-were reduced after in vitro culture with elevated NEFA content (in the form of a combination of oleic, stearic, and palmitic acids). This could affect the microenvironment of the developing embryo, and may contribute to the complex pathogenesis of subfertility in high-yielding dairy cattle. Sperm capacitation and selection may also be hindered in response to such oviductal changes, which may impair oocyte fertilization (Jordaens et al., 2015). However there is no evidence that any particular n-3 FA acts on oviductal cells, which might be very interesting for the purposes of this review. Nevertheless, it can be concluded that embryo survival may strongly depend on lipid metabolism in the reproductive tract (oviduct, uterus), which is evident in subfertile high-lactating cows.

The embryo
Lipids are crucial for sustaining embryo growth, and preimplantation embryos benefit from serum FAs, and particularly from n-3 PUFAs (Wonnacott et al., 2010;Hughes et al., 2011). Lipids and free FAs exert substantial impact on the developing embryo, depending on the lipid fraction and concentration of the particular FA (Hughes et al., 2011;Cagnone and Sirard, 2014). The beneficial effect of n-3 FAs has been well documented. Ambrose et al. (2006) observed higher pregnancy rate, reduced pregnancy losses, and greater calving rates in cows fed an n-3 diet for 60 days from day 28 before estrus to 32 days post insemination (dpi).
It is well known that dietary supplementation with n-3 significantly alters fatty acid profile in blood plasma, follicular fluid, granulosa cells, and also oocyte-cumulus complexes by incorporating dietary PUFAs. It can therefore be expected that the supplemented factor affects both oocytes and embryos. It is well known that UFAs and SFAs exert opposite effects on early embryo development. Thangavelu et al. (2007) used two types of diets (n-3 and SFA supplementation) for 27 days from day 20 day before estrus to 7 dpi. Although neither the fertilization rate nor the blastocyst yield were affected, the quality (as higher cell count) of the most advanced embryos (expanded blastocysts) was superior in the n-3 group to the SFA group.
One field trials reported no effect of n-3 supplementation on embryo quality (Fouladi-Nashta et al., 2009); the authors hypothesize that, under experimental conditions, the ovary can protect oocytes from n-3:n-6 fluctuations in plasma, so that the exogenous FAs exert only a modest effect on the oocyte; however no further studies have been published on this hypothesis. This argument has recently been challenged by Freret et al. (2019) who demonstrated an effect of n-3 on lipid composition in oocytes, and thus on the quality of corresponding embryos. To our knowledge, there has been only one study investigating the effects of n-3 and SFA supplementation that reported a negative impact of n-3; Petit et al. (2008) observed reduced fertilization rate and increased incidence of degenerated embryos in cows fed a diet with flaxseed, as compared to SFA. These authors, however, did not provide any explanation to their findings.
Summarizing, there are several pathways of embryo-maternal interactions. The properties of oviductal and uterine cells can be altered by n-3 supplementation, thus creating a better environment for the ovulated oocytes and preimplantation embryos. The n-6:n-3 ratio of the diet can also affect CL functions. The positive effects of n-3 usually affect blastocyst quality, but not embryo rate. This may indicate the advantage of the supporting embryo culture over oocyte maturation. It is well known, however, that dietary supplementation with n-3 significantly alters the fatty acid profile in blood plasma, follicular fluid, granulosa cells, and oocyte-cumulus complexes by incorporating dietary PUFAs. It can thus be expected that the supplemented factor affects both the oocyte and embryo, and that these effects cannot be separated.

Summary
The results of the published field experiments overall indicate that the schedule of dietary supplementation with n-3 should cover a period of 160-170 days, starting from the eighth week before calving (pregnant dry cow or pregnant heifer), and continue until 100-120 days postpartum (Figure 1). This is reasonable for both practical animal management and for scientific reasons. The two categories of pregnant females are managed the same way, and fed diets of similar composition, making it easier to control the feeding scheme. Although hormonal stimulation of the first estrus postpartum has become popular on dairy farms, we based our analysis on the spontaneous ovarian cyclic in order to show the biological background. It is also of special importance to supplement the female with n-3 FAs during the period of negative energy balance (NEB) when reproductive processes suffer from the subfertility syndrome and distortions by side products (Leroy et al., 2008a). The detrimental effect of NEB syndrome affects the first 5-10 weeks pp and mainly manifests as disturbed endocrine signaling (Leroy et al., 2008b). On the basis of the two following assumptions: 1) that conception occurs at the second cycle postpartum, and 2) that implantation occurs from day 20 to day 40 post insemination, we recommend that supplementation is carried out until day 120 postpartum. The rationale behind this lies in the multidirectional action of n-3 FAs, which support oocyte growth, embryo development, and implantation. The follicle ovulating during the second estrus postpartum was activated from the ovarian reservoir in the fifth month, whereas antrum formation began in the ninth month of gestation. Thus n-3 supplementation covers the period from the late preantral stage of the follicle, the entire phase of antrum development, during which the dominant follicle increases in size, to oocyte ovulation and fertilization. The beneficial n-3 action has been extended to include the phase of the preimplantation embryo (yielding better quality) and implantation, by improving the uterine environment (e.g., through reduced synthesis of luteolytic PGF2α).
We conclude that dietary supplement with n-3 FAs in high-yielding dairy cows supports ovarian function and helps the embryo to survive in the adverse environment of the reproductive tract of the postpartum cow.