Physiological and metabolic aspects of follicular developmental competence as affected by lactational body condition loss

Metabolic demands of modern hybrid sows have increased over the years, which increases the chance that sows enter a substantial negative energy balance (NEB) during lactation. This NEB can negatively impact reproductive outcome, which is especially evident in primiparous sows causing a reduced second parity reproductive performance. The negative effects of the lactational NEB on reproductive performance can be partly explained by the influence of the premating metabolic state, during and after lactation, on the development of follicles from which oocytes will give rise to the next litter. In addition, the degree and type of body tissue mobilization during lactation that is, adipose tissue or lean mass, highly influences follicular development. Research investigating relations between the premating metabolic state and follicular and oocyte competence in modern hybrid sows, which experience higher metabolic demands during lactation, is limited. In this review we summarize current knowledge of physiological relations between the metabolic state of modern hybrid sows and follicular developmental competence. In addition, we discuss potential implications of these relations for current sow management strategies.

a reduction in farrowing rate and litter size (Tantasuparuk et al., 2001;Thaker & Bilkei, 2005). In addition, negative influences on embryonic development and litter uniformity (Hoving et al., 2012;Patterson et al., 2011;Vinsky et al., 2006) and subsequent variation in piglet birth weights (Wientjes et al., 2013) have been reported. The exact origin of reduced reproductive outcome after NEB remains for a part unknown.
The negative effects of the lactational NEB on reproductive outcome can be partly explained by the influence of the premating metabolic state, during and after lactation, on the development of follicles from which oocytes will give rise to the next litter, a topic that has been studied for many years by among others Foxcroft et al.
(e.g., Clowes et al., 2003;H. Yang et al., 2000;Zak et al., 1997). For instance, feed restriction during lactation resulted in a smaller follicle size at weaning and 48 h postweaning (Quesnel et al., 1998) and decreased oocyte maturation rates when oocytes were isolated 38 h before the anticipated onset of estrus (Zak et al., 1997). These older studies are performed at least 19 years ago, research investigating relations between the premating metabolic state and follicular and oocyte competence in modern hybrid sows, which experience higher metabolic demands during lactation, is limited.
Understanding how follicular development is influenced by the metabolic state of the sow can increase our understanding of the origin of reproductive problems in pigs and other mammals. In this review, we summarize literature describing physiological relations between the metabolic state of modern sows and follicular developmental competence.
1.1 | Influence of weight loss during lactation 1.1.1 | Reproductive performance and litter characteristics Lactation puts considerable metabolic demands on sows, as they need to produce a large amount of milk to feed the high number of piglets. Substantial weight loss has been associated with longer weaning-to-estrus intervals (WOI) and reduced reproductive performance, as described previously. Changes in the metabolic state are communicated on a whole-body level via the secretion of several metabolic factors and hormones, such as insulin, IGF-1, and leptin.
These factors can influence follicular development and reproductive performance directly at the ovarian level but also indirectly via influencing gonadotropin release (Odle et al., 2018; reviewed by Wolfe et al., 2014). Although also of importance, the current review will focus on relations between weight loss and follicular development, and reproductive performance.
Next to selection for increased litter size, sows have also been selected for a shorter WOI, to further increase the number of piglets produced per year. Follicular development and therefore the duration of WOI is controlled by the luteinizing hormone (LH) release pattern before and after weaning (Shaw & Foxcroft, 1985). Lactational weight loss probably influences LH pulsatility, as feed restriction during lactation reduces the frequency of LH pulsatility (Quesnel et al., 1998) and reduces mean and basal LH levels around weaning (Kauffold et al., 2008), subsequently affecting the WOI. Early studies show that feed restriction and associated lactational weight loss can indeed prolong WOI, however, this effect is much less pronounced or absent in later studies suggesting that weight loss may not prolong WOI in modern sows (reviewed by Kemp et al., 2018). This indicates that the follicle pools of modern sows have less time to recover from the negative influences of weight loss during lactation. In turn, this may negatively affect developmental competence of the follicle pool that will ovulate to give rise to the next litter of piglets. This hypothesis is supported by Kemp et al. (2018), who describe that feed restriction in modern sows does not prolong WOI anymore, but it has a larger negative effect on embryo survival at Day 28/35 of pregnancy.
A higher weight loss during lactation has been related to smaller follicle sizes postweaning (e.g., Quesnel et al., 1998), to lower embryo weights and survival during the next pregnancy (Hoving et al., 2012;Patterson et al., 2011) and even to increased within-litter variation in piglet birth weight (Wientjes et al., 2013). These consequences of high weight loss may result from changes in oocyte quality and subsequent embryo development and quality, but may also be related to changes in corpus luteum (CL) development. For example, preovulatory follicle size is found to be related to CL size (Soede et al., 1998;Wientjes et al., 2012) and CL size at Day 30 postovulation seems to be associated with subsequent piglet birth weight and birth weight variation (Da Silva et al., 2017).
To further investigate effects of lactational weight loss on reproductive outcome and litter characteristics in modern sows, we obtained a data set which includes data on 431 TN70 sows (Topigs Norsvin). The sows of this data set were housed on two different farms in the Netherlands and fed a gradually increasing amount of a lactation diet (9.75 MJ NE/kg, 8.4 SID lysine) starting at 2 kg/day around farrowing and increasing to 6.66 kg/day from Day 15 of lactation onwards (personal communication; Mr. Opschoor; Topigs Norsvin). This data set contains data on body weight and backfat depth of sows during their first lactation and subsequent reproductive performance and litter characteristics of the second parity.
Analysis of this data set show that around 30% of the sows on these farms had a relative lactational weight loss of more than 12% of their initial body weight at farrowing. These sows which had more than 12% relative weight loss, had a slightly longer weaning-toinsemination interval (weaning-to-insemination interval [WII]; +0.2 days), a smaller litter size (−2.8 piglet) and a numerically reduced farrowing rate (−8%, p = 0.11) as compared to sows which lost less than 12% body weight (Table 1). This corroborates earlier findings that lactational weight loss which exceeds 10%−12% of initial body weight has negative consequences for reproductive performance in the next cycle (Hoving et al., 2010(Hoving et al., , 2012Schenkel et al., 2010;Thaker & Bilkei, 2005). The WII was only increased from 4.5 to 4.7 days in this data set, which is similar to the small effect found in recent studies as described in Kemp et al. (2018). It should be noted that different from the studies described in Kemp et al. (2018), the sows from this data set were not subjected to feed restriction. Therefore, differences in weight loss are most likely due to causes such as the longer lactation period and higher number of piglets weaned, than to an involuntary lower feed intake (Table 1).
Together, data show that also in modern primiparous sows lactational weight loss negatively influences the characteristics of the next litter. These negative effects of lactational weight loss on reproductive outcome may at least be partially explained by the negative influence of lactational weight loss on the development of the follicle pool and oocytes which will give rise to the next litter.

| Follicular developmental competence
There are only a few studies that address effects of weight loss during lactation on follicular developmental competence in sows.
Two of these studies were performed over 20 years ago (Quesnel et al., 1998;Zak et al., 1997). In the study of Quesnel et al. primiparous sows received 50% feed restriction during a 28-day lactation, which resulted in a weight loss of 21.2% of initial body weight at parturition for restricted-fed (RES) as compared to 8.2% for full-fed (FF) sows.
Feed restriction resulted in a 15% reduction on average follicle size of the 10 largest follicles at weaning (2.8 vs. 3.3 mm for restricted-fed vs. full-fed) and a 23% reduction (3.7 vs. 4.8 mm in restricted-fed vs.  Abbreviations: GLM, general linear model; SE, standard error; WL, weight loss. 1 Differences between means were analyzed using proc GLM in models which included weight loss class and farm. Interaction between farm and weight loss class was never significant and excluded from the models. Distribution of weight classes (<6%, 6%−12%, and >12%) for each farm was as following: Farm A: 39%, 35%, and 26% and Farm B: 27%, 37%, and 35%, respectively. 2 Body weight and backfat depth were measured on average at Day 106.9 ± 0.1 of gestation. Body weight at parturition was estimated using calculations as described in Bergsma et al. (2009). 3 Nominal/categorical variable which was analyzed using proc glimmix. 4 Sows that failed to farrow after first insemination were excluded from analysis.

COSTERMANS ET AL.
| 493 oocytes to metaphase II in vitro (−19%) as compared to feed restriction from Day 1 to 21.
To study effects of lactational weight loss on postweaning follicular and oocyte developmental competence in modern hybrid sows, we have performed a study in which we applied feed restriction during the last 2 weeks of a 24-day lactation of primiparous TN70 sows (Costermans et al., 2019c). This resulted in a relative weight loss of 19.6% for restricted-fed versus 11.5% for full-fed sows during lactation. When assessing the pool of 15 largest antral follicles 48 h after weaning, feed restriction reduced average follicle size by 26% (3.1 vs. 4.2 mm; Figure 1). Although feed restriction did not affect cumulus-oocyte complex morphology, it did lower cumulus expansion rates after 22 h of IVM by 26% (175 vs. 237%) and increased polyspermy by 89% (44 vs. 23% of the fertilized oocytes; all Figure 1), the latter being a marker for immature oocytes (Van der Ven et al., 1985). In addition, follicular fluid of restricted sows had lower IGF1 (−56%) levels 48 h after weaning. The lower IGF1 levels in the restricted sows may have played an important role in the regulation of reduced follicular growth, steroid production and oocyte developmental competence (Němcová et al., 2007;Xia et al., 1994;Zhou et al., 2013). Our findings are corroborated by recent findings of Han (2021a), where sows that received feed restriction during the last 2 weeks of a 28-day lactation had 4 days after weaning lower follicular fluid IGF1 levels as compared to full-fed sows. Interestingly, Han et al. also describe that higher IGF1 levels in serum at weaning were related to a higher mean piglet birth weight in the next litter, indicative of long term consequences (Han et al., 2021b).
Follicular fluid of the feed restricted sows contained lower steroid levels (e.g., β-estradiol, progestins, and androgens). Together, results indicate that follicles of restricted sows produce less steroids and growth factors, needed for oocytes to obtain full developmental competence.
The above studies show that relative weight loss due to feed restriction is similar in older and modern sows. In addition, weight loss during lactation highly influences follicular development and oocyte quality in modern first parity sows. Molecular substantiation of potential regulated pathways in these follicles would greatly contribute to our understanding of metabolic influences on follicular developmental competence. For this purpose, we performed granulosa cell transcriptome analysis of follicles from multiparous sows with either high or low lactation weight loss using an existing data set (Costermans et al., 2019a). Our data set included 8 sows which were split in two groups: 4 multiparous sows with a high lactational weight loss (12.0 ± 2.2%) during a 26-day lactation, and 4 sows with a low lactational weight loss (4.7 ± 1.5% Pathway analysis of FDR < 0.2 and FC < 1.2 genes revealed that 9 out of the 10 most significant pathways are involved in protein translation and the regulation of translation, for example, ribosome biogenesis and messenger RNA degradation (Table 2).
Individual transcript analysis of the genes in these pathways revealed that granulosa cells of sows with high relative lactational weight loss have a higher expression of genes involved in protein translation (top 10 transcripts with a higher and lower expression in sows with a high weight loss during lactation are shown in Tables 3 and 4, respectively).
For most of these genes, their function in female reproduction is not well investigated. Some of the most interesting genes will therefore be highlighted. Granulosa cells of sows with high weight loss have a higher expression of A-Kinase Anchoring Protein (AKAP9), Zinc finger protein 217 (ZNF217) (see Table 3 for both) and Regulator Of Cohesion Maintenance, Homolog A (PDS5A, FDR = 0.03 and FC = 1.29). AKAP9, a A-kinase anchor protein that binds to the regulatory subunit of protein kinase A to localize it to specific cellular compartments, is involved in centrosome function and microtubule organization (reviewed by College & Scott, 1999) and is essential for cell cycle progression (Hu et al., 2016;Keryer et al., 2003). In addition, AKAP9 plays an essential role in spermatogenesis and somatic cell-germ cell organization (Schimenti et al., 2013). PDS5A is essential for chromosome segregation during mitosis (Losada et al., 2005; reviewed by Mannini et al., 2010;Peters, 2012), while ZNF217 is a transcriptional regulator (reviewed by Cohen et al., 2015), and may be involved in the repression of cell differentiation (Krig et al., 2007). Most of these described genes (e.g., PDS5A, AKAP9, and ZNF217) are involved in regulating cell proliferation. Together, this may suggest that granulosa cells of sows with a high versus low weight loss during lactation are more proliferative. To assess this, protein expression of Ki-67, a proliferation marker (Gerdes et al., 1984), was compared for granulosa cells of sows with high versus low weight loss during lactation (see Costermans et al., 2019a for experimental details). Ki-67 protein expression tended to be higher (p = 0.09) for sows with a high versus low weight loss during lactation.
Together, the current analysis shows that sows with a high relative weight loss during lactation have a higher expression of genes involved in protein translation and some genes involved in proliferation as compared to sows with a low relative weight loss. In addition, at the protein level, granulosa cell proliferation tended to be higher in sows with a higher weight loss during lactation. In To summarize, high lactational weight loss can negatively influence litter characteristics of the next litter in modern lean sows, while moderate lactational weight loss hardly affects litter characteristics. These effects of weight loss on litter characteristics are therefore especially relevant in modern young sows with limited feed intake (Bergsma et al., 2009;Kanis, 1990) and for sows that are housed at high ambient temperatures where voluntary feed intake is T A B L E 2 Top 10 pathways of FDR ≤ 0.20 genes in granulosa cells of multiparous sows (parity 3-5) with high weight loss (12.0 ± 2.2%, N = 4) versus low weight loss (4.7 ± 1.5%, N = 4) during a 26-day lactation usually lower (Quiniou & Noblet, 1999 irrespective of follicle size. In modern sows, the effect of weight loss on WOI is much smaller as compared to earlier studies. As lactational weight loss clearly still affects litter characteristics, this suggests that these negative effects of lactational weight loss on the reproductive performance in the next cycle may be related to a delay in follicular development. Together, this suggests that negative effects of T A B L E 3 Top 10 FDR ≤ 0.1 genes in granulosa cells with a higher expression in multiparous sows (parity 3-5) with high weight loss (WL) (12.0 ± 2.2%, N = 4) versus low weight loss (4.7 ± 1.5%, N = 4) during a 26-day lactation

| Influence of energy mobilization of different substrates during lactation
As described previously, sows usually experience weight loss during lactation, which negatively influences reproductive performance and the next litter if it exceeds around 10%−12% of initial body weight (Schenkel et al., 2010;Thaker & Bilkei, 2005 and Tokach et al., 2019). During lactation energy is mobilized from adipose tissue, but also skeletal muscle protein is mobilized as an energy and protein source (e.g., Clowes et al., 2005;S. Kim & Easter, 2001;Schenkel et al., 2010). It is estimated that sows lose around fivefold more kilograms of fat during lactation compared to protein (corrected for the associated water loss) (Bergsma et al., 2009 As both fat and protein stores may be simultaneously used for energy mobilization it is difficult to establish which of the two is responsible for the negative effects of weight loss during lactation on follicular developmental competence. So far, two studies in sows have assessed effects of selective protein mobilization during lactation on this process (Clowes et al., 2003;H. Yang et al., 2000).
In both studies, primiparous sows are fed different amounts of protein during lactation using isocaloric lactation diets. Lower dietary protein levels result in a higher estimated protein loss, based on body weight and backfat depth. Backfat depth itself as a proxy for adipose tissue stores remains unaffected. Selective high protein losses reduce follicle size and follicular fluid 17β-estradiol levels at weaning (Clowes et al., 2003) and at pro-estrus (H. Yang et al., 2000). In addition, follicular fluid obtained at pro-estrus from sows fed a low protein diet is less competent in supporting oocyte maturation in vitro (H. Yang et al., 2000). From these studies, it is clear that selective protein mobilization negatively affects follicular development competence.
Studies investigating effects of selective energy mobilization from adipose tissue on follicular developmental competence in sows are lacking.
In our recent studies (Costermans et al., 2019b;Costermans et al., 2019c), we have assessed relations between the lactational metabolic state of primiparous and multiparous sows, respectively, and follicular development around weaning, to get a better understanding of the underlying mechanism of metabolic influences on reproductive outcome.
In primiparous sows, backfat loss and loin muscle depth loss during a 24-day lactation are negatively related to follicle size 48 h after weaning (Figure 2). In multiparous sows we unfortunately did not measure loin muscle depth, but we did measure serum creatinine levels as a marker for whole body protein break down during a NEB (Costermans et al., 2019c;Y. X. Yang et al., 2009). In the multiparous sows, higher serum creatinine levels at weaning are related to a smaller average follicle size at weaning. Surprisingly, more backfat loss during a 26-day lactation appears to be related to a higher average follicle size at weaning (Figure 2). We hypothesize that sows with low levels of backfat are forced to mobilize their protein reserves to fulfill the energy requirements of milk production, which might have a detrimental effect on follicular development. Indeed, in our study, lower backfat loss during lactation is related to higher creatinine levels at weaning (β = −0.14, p = 0.05). So, the relation between increased backfat loss during lactation and a larger average follicle size at weaning might be explained by protein sparing effects.
We did not observe these protein sparing effects in the primiparous sows, as no relationship between backfat loss during lactation and muscle depth loss or serum creatinine levels is found.
This difference may be explained by the amount of available adipose tissue for energy mobilization during lactation which differs between the primiparous and multiparous sows (to be discussed below), as it has been suggested that sows maintain a certain minimum level of adiposity (Bergsma et al., 2009;Parmley et al., 1996;Whittemore & Morgan, 1990). The essential role of adipose tissue for survival and reproduction is for instance shown by studies in transgenic A-ZIP/F-1 mice which lack white adipose tissue. These mice have a reduced fecundity, a shorter lifespan and are diabetic; the latter can be reversed by surgical implantation of donor adipose tissue (Gavrilova et al., 2000;Moitra et al., 1998). For sows, the minimum level of adiposity that is maintained is suggested to be at a backfat thickness of around 10 mm (Whittemore & Morgan, 1990). Indeed, backfat depth at parturition differs between multiparous and primiparous sows, as multiparous sows have an average backfat depth of 17.1 ± 0.4 mm (Costermans et al., 2019b) while the primiparous sows have an average backfat depth of 13.7 ± 0.4 mm (Costermans et al., 2019c). This may indicate that the primiparous sows do not have sufficient fat reserves for energy mobilization to spare protein reserves, as their backfat depth is already close to the minimum level of adiposity that is maintained during lactation. In both studies, backfat at parturition is positively related to backfat loss during lactation. Thus, the amount of backfat at parturition seems to determine if backfat loss during lactation will be sufficient to spare protein reserves during lactation. More backfat at parturition and COSTERMANS ET AL.
| 497 (consequently) less lean mass mobilization during lactation may therefore positively influence follicular developmental competence in highly prolific sows. It should be mentioned that there seems to be an upper limit (around 21−22 mm) for the optimal backfat depth at the start of lactation with regard to lactation performance (J. S. Kim et al., 2015). This is likely due to the reduced voluntary feed intake during lactation and increased energy requirements for maintenance in fatter sows (J. S. Kim et al., 2015;Quesnel et al., 2005), which in turn decreases available resources for milk production. In addition, obesity should be avoided to prevent locomotion problems, one of the major risk factors for culling of sows (Bortolozzo et al., 2009).
Our understanding of energy metabolism will be greatly improved by studying the effect of selective protein and fat loss during lactation on follicular developmental competence in modern hybrid sows. Selective protein loss may be achieved by varying protein levels in the diet (similar to studies by Clowes et al., 2003;and H. Yang et al., 2000). A possible problem of such a study may be that selective fat mass loss may be difficult to achieve in modern sows with low adiposity. As an alternative, relations between energy mobilization of different energy substrates during lactation and follicular development can be assessed using balance trials, body composition measurements and metabolic signaling factors of adipose tissue (adipokines) or skeletal muscle (myokines). This may lead to new insights in how follicular development is regulated by metabolism of different substrates (such as fat or protein).

| Implications for current sow management strategies
As a response to the demand of consumers for leaner pork, modern sows have been selected for a higher capacity for lean mass gain, and a lower capacity for fat mass gain. Next to this, modern sows have also been selected for a large litter size to increase pig production profitability. Consequently, sows need to produce a high amount of milk to feed the higher number of piglets weaned per litter (Kemp et al., 2018). This high milk production leads to high metabolic demands of sows during lactation. Feed intake during lactation is not sufficient to meet these high metabolic demands, which is especially true for modern sows and more so for primiparous sows which are still growing, and have a lower feed intake capacity compared to multiparous fully grown sows (Bergsma et al., 2009;Kanis, 1990). In addition, in summer, sow voluntary feed intake may decrease by 50% because of higher ambient temperatures (Quiniou & Noblet, 1999).
Therefore, to meet the high metabolic demands during lactation, sows need to mobilize their own body reserves. Together, these described breeding strategies have resulted in lean sows with a low fatness, which mobilize a large amount of body reserves during lactation, especially primiparous sows.
From this review it becomes clear that a certain level of adiposity is needed for energy mobilization during lactation to reduce undesired mobilization of lean mass, otherwise follicular developmental F I G U R E 2 Relations between average follicle size (mm) of the 15 largest follicles at weaning after a 26-day lactation period and (a) creatinine levels at weaning (mg/dl) and (b) backfat loss during lactation (mm) in multiparous sows (Costermans et al., 2019b). Relations between average follicle size (mm) of the 15 largest follicles 48 h after weaning after a 24-day lactation period and (c) muscle depth loss during lactation (cm) and (d) backfat loss during lactation (mm) in full-fed primiparous sows (Costermans et al., 2019c).
competence may be compromised. This compromised follicular developmental competence is expected to negatively affect reproductive performance in the next cycle. On the other hand, mobilization of tissue for energy during lactation is essential to sustain high milk production to feed large litters (Bergsma et al., 2009).
We have described that more mobilization of tissue for energy during lactation is related to a higher milk production in the last week of lactation (Costermans et al., 2020) where mobilization of energy and protein from lean mass is specifically related to a higher protein content in milk. A higher loin muscle depth loss during lactation is related to an increased milk production and milk protein content but not related to a higher milk fat content, while a larger backfat depth loss is related to a higher milk production and milk fat content but not related to a higher milk protein content (Figure 3).
These observations imply that especially mobilization of substrates from lean mass may be beneficial to support piglet weight F I G U R E 3 Regression equations (β) for the relations between body weight loss, loin muscle depth loss, backfat depth loss from D1-24 and milk production parameters. Sows were either full-fed (6.5 kg/day), open circles Ο, or restricted-fed (3.25 kg/day), filled squares ■, from D10-24 of a 24-day lactation and this treatment was included in the model. Interactions with treatment were never significant. Ptreat = p-value for treatment. As published in Costermans et al. (2020).
gain, as it is described that in general milk protein production may be insufficient for maximal piglet weight gain (Campbell & Dunkin, 1983).
However, it needs to be verified if milk protein content is also the limiting factor for maximum piglet weight gain in modern sow lines.
Both aspects, milk production to support weight gain of the current litter and follicular development of the presumptive ovulatory follicle pool which is expected to affect litter characteristics of the next litter, should be optimal for efficient pig production.
One of the strategies to improve metabolic efficiency may be to increase body fat reserves of sows before the start of lactation. This may be achieved by using older gilts with more body tissue reserves at insemination or by increasing the amount or composition of gestation feeds. However, heavier and fatter sows at parturition also have a reduced voluntary feed intake during lactation (reviewed by Eissen et al., 2000;Quesnel et al., 2005) which may result in increased body reserve mobilization during lactation and thereby negatively affecting follicular developmental competence. This effect is also seen in Table 1, which shows that primiparous sows with the highest lactational weight loss (>12% of initial body weight) and a second litter dip in reproductive performance, also have the highest body weight and backfat depth at parturition. Using heavier and fatter sows at gestation may therefore not always be the best strategy to support follicular development during and after lactation, as its success may be dependent on factors such as age of the sows and sow feed intake capacity. More research regarding this aspect is needed.
Other possible strategies may be directed toward increasing the availability of energy and nutrients in the lactation diet, to support both follicular development during lactation and milk production. Some relatively recent studies have been directed toward feeding higher protein levels during lactation (Huang et al., 2013;Pedersen et al., 2019;Strathe et al., 2017). In the study of Strathe et al. (2017), gradual increases in dietary protein levels during lactation were applied, which decreased lactational weight loss and increased backfat loss. In addition, the higher protein levels tended to be related to an increased litter size in the next litter (Strathe et al., 2017), suggesting that feeding higher protein levels during lactation improves follicular developmental competence of the next ovulatory follicle pool. Next to this, feeding higher levels of protein during lactation increases both milk fat content (Strathe et al., 2017) and milk protein content (Clowes et al., 2003;Strathe et al., 2017). Feeding a higher protein level during lactation may therefore positively influence both milk (protein) production to support litter gain of the current litter and follicular developmental competence to improve litter characteristics of the next litter. An alternative approach may be to optimize the protein source of the lactation diet to reduce sow protein mobilization and support milk protein production.
Recent work from our group shows that sows that received a lactation diet with a higher ratio of slow to fast digestible protein show reduced body protein loss during lactation at similar litter gain (Ye et al., 2022). It remains to be investigated how feeding diets with varying protein source influences follicular developmental competence and how it affects piglet birth weight in the next litter.
Other studies have used increased energy dense lactation diets using fats or oil as additives. This additional form of energy is, however, mainly used to produce milk with a higher fat content, resulting in a higher litter weight gain (Rosero et al., 2015;reviewed by Rosero et al., 2016), and is not beneficial for sow body condition.
Similarly, feeding sows a fat-rich versus starch-rich iso-caloric lactation diet, results in more milk fat production and fatter piglets, but also results in a more severe NEB when sows are fed at a high feeding level (Van den Brand et al., 2000). This suggests that adding fat to the lactation diet mainly results in increased milk fat content, which is beneficial for litter weight gain, but does not increase energy availability for follicular development during lactation. It remains a challenge to design optimal feeding strategies for young lactating sows with a relatively short lactation length, as strategies that benefit the current litter (sustained milk production) seem to be a disadvantage for the next generation (due to suboptimal follicle development).

| CONCLUSION
To conclude, the metabolic state of modern sows during lactation highly influences follicular developmental competence, already from the first half of the follicular phase onward. In addition, in primiparous modern sows, lactational weight loss influences the litter characteristics of the next litter. This provides further evidence for the hypothesis that impairments in follicular developmental competence as established during lactation may at least partially explain the reduced reproductive performance in the next cycle and lower piglet birth weights. The degree and type of body tissue mobilization that is, adipose tissue or lean mass, highly influences follicular development during lactation as well as milk production and composition, which are both important factors for sow performance. For instance in our own studies, we have discovered that especially mobilization of energy from lean mass in sows with low adipose tissue reserves, negatively impacts follicular developmental competence, while mobilization of protein from lean mass may be beneficial for milk protein production to support piglet growth. It therefore remains a challenge to optimize current sow management strategies to benefit sow-and piglet performance. Strategies to improve sow reproduction efficiency and prevent reproduction problems could be aimed at optimizing energy and nutrient availability in the diet, as well as optimizing sow body condition at the start of lactation. If relationships between metabolism and follicular developmental competence are better understood, future studies can be directed toward studying potential beneficial effects of changes in metabolic state at the start of lactation or redirecting energy mobilization during lactation. This may optimize follicular developmental competence and can potentially prevent reproductive problems in sows and other mammals that experience a lactational NEB.

ACKNOWLEDGMENTS
The reported research was funded by the Wageningen Institute for Animal Sciences (WIAS) and NWO by providing the NWO-WIAS Graduate Program 2015 grant.