Feeding the ever-growing world population with limited environmental resources is a topic of growing concern. Specifically, consumer demand for animal-based protein sources such as meat and milk have risen due to increased wealth of some populations (Follett and Hatfield, 2001; Helms, 2004; Robinson, 2010). Consequently, animal production must be executed in a sustainable matter to reduce environmental consequences. One possibility to improve sustainability in this sector is to make more efficient use of the available nutrients which are used to synthesize meat and milk from animals (Tamminga, 2003; Rotz, 2004). Accordingly, the government has implemented strict environmental constraints regarding the output of waste products from animal production systems by, for example, introducing a manure and ammonia policy (LNV, 1995) including the Phosphate Reduction Plan (Dam, 2017) and the European Union Nitrate Directive (OJEC, 1991). These implementations greatly impact the dairy production sector (Hennessy et al., 2005; Powell et al., 2010). Due to restrictions on nitrogen (N) and phosphorus (P) contents in manure, which is subsequently used to fertilize crop land, the concentration of these components in crops might be reduced (Schoumans et al., 2015). To counteract these losses, innovative crop management techniques may aid in more efficient use of these nutrients by the plants itself, retaining more nitrogen and phosphorus (Zheng et al., 2013). In addition, the animal feed industry has changed, due to more types of forages and their improved availability (Schoumans et al., 2015). This allows crops to be more easily incorporated into dairy cow diets resulting in greater flexibilities in ration composition. Nitrogen from forage contributes to the crude protein (CP) content of a dairy cow diet. Dietary protein is an important component in dairy cow rations to support rumen function and milk protein synthesis. To meet the set restrictions on N excretion, ration formulations for dairy cows should promote maximum efficiency of transfer of dietary N into milk N, thereby reducing excretion of excess N in manure and ammonia emissions (Castillot et al., 2000; Arriola Apelo et al., 2014).
PROTEIN REQUIREMENTS
In the past, feeding high crude protein diets (CP level >18%) to ensure protein requirements for lactating dairy cows were met was standard practise. Increasing dietary CP concentration, increases N excretion in manure and decreases N efficiency (Van der Stelt et al., 2008). On dairy farms which fed a high CP level, lowering CP content of diets may improve N utilization without negatively affecting milk production (Broderick, 2003; Huhtanen and Hristov, 2009; Chase et al., 2012). However, decreasing CP content (≤13%) may negatively affect dry matter intake (DMI), milk production, and milk protein yield (NRC, 2001; Broderick, 2003; Lapierre et al., 2006; Giallongo et al., 2016). Protein evaluation systems estimate protein requirements for lactating dairy cows by optimizing ruminal fermentation and distinguishing ruminal degradable protein (RDP) from rumen undegradable protein (RUP) (Varga, 2007). Rumen degradable protein is dietary protein that can be degraded by rumen microbes and used to synthesize protein (Tamminga et al., 2007). Rumen undegradable protein is dietary protein that resists ruminal break-down and thus escapes modification by rumen microbes (Tamminga et al., 2007). The distinction between RUP and RDP is made, because they both can influence the protein flow and composition differently. Moreover, it is important to recognize how RUP and RDP relate to each other. When milk production of a cow increases, protein intake increases through increased DMI and rumen microbes achieve their maximum protein production capacity. At this point, the remaining protein requirement should be fulfilled by RUP. However, if a high amount of RUP is supplied without adequate RDP, ruminal microbial activity and protein synthesis may be depressed, reducing overall fermentation capacity and may reduce amino acid (AA) flow to the small intestine (Varga, 2007). It is known that microbial protein from RDP consists of a good pattern of AA for milk protein synthesis (Mantysaari et al., 1989). However, the AA composition of RUP depends on the type of feedstuff. Consequently, the AA composition of dietary protein sources is of interest.
AMINO ACID REQUIREMENTS
Recently, when considering dietary protein in dairy ration formulation, great focus has been placed on improving the supply of individual AA, the building blocks of protein, rather than simply changing dietary CP level. It is the AA content of protein sources which can have the greatest effect on the efficient transfer of dietary N into milk N (Socha et al., 2005; Varga, 2007). Therefore, balancing AA requirements during ration formulation should be emphasized rather than targeting the level of CP (Ardalan et al., 2017). This can be accomplished by using the metabolizable protein (MP) concept. Metabolizable protein consist of microbial crude protein (MCP), AA from RUP, and endogenous protein (Varga, 2007). Endogenous proteins are those which are recycled during the digestive processes and have the potential to be re-used (Tamminga et al., 2007). The MP concept focuses on the availability of AA in the small intestine of the dairy cow. These AA are sub-dived into net requirements for maintenance, production, reproduction, and immune function (Bequette and Nelson, 2006). Metabolizable protein requirements reflect the role AA serve as precursors for protein synthesis (Bequette and Nelson, 2006). An overview of the protein evaluation concept can be found in Appendix 1. An approach to reduce dietary CP content involves incorporating the supply of individual AA more specifically into dairy cow diets. When the supplied AA composition (AA profile) of dietary protein resembles the AA profile most desirable for milk synthesis, the mammary gland can more efficiently synthesize protein and AA catabolism may be reduced, improving N retention for synthesis and lowering N excretion (NRC 2001, Sloan et al. 2006, Liu et al. 2013, Fraser et al. 1991).
LIMITING AMINO ACIDS
Amino acids can be categorized as essential or non-essential. Non-essential amino acids (NEAA) can be synthesized in adequate amounts when provided sufficient amounts of protein (Rose, 1938; Schwab and Schwab Consulting, 2012; Wu, 2013). Conversely, essential amino acids (EAA), cannot be synthesized, or at least not in sufficient amounts to meet maintenance and growth requirements (Wu, 2013). Therefore, ensuring their adequate supply via the diet is often necessary. The law of the minimum describes that production is regulated by the most scarcely available required resource, or the limiting factor (von Liebig and Playfair, 1847). This concept is often used in AA nutrition for animal production. Amino acids that limit protein synthesis are called limiting EAA and are typically provided in the smallest quantities relative to their requirements (Schwab et al., 2003).
IMPORTANT LIMITING AMINO ACIDS
Despite the fact that AA requirements for lactating dairy cows are not completely defined, methionine (Met), lysine (Lys), and histidine (His) are considered the most limiting AA for milk protein synthesis in diets fed to dairy cows as indicated by the National Research Council (NRC, 2001) and many others (Rulquin et al., 1993; Lee et al., 2012b; White et al., 2017). It is not unexpected that Met, Lys, and His are identified as the most limiting AA in lactating dairy cows, as their presence in the AA profile of feed is not consistent with the AA profile of rumen microbial, milk or tissue protein (Overton and Chase). Relative to the amount of these EAA required to synthesize microbial, milk and tissue protein, their concentrations are low in common feed ingredients (O’Connor et al., 1993; NRC, 2001; Vanhatalo et al., 2003). Lysine is typically the most limiting EAA in maize-based diets (Nichols, 1996; Liu et al., 2000; Kleinschmit et al., 2007). Methionine is limiting when soybean meal or animal-derived proteins (such as, whey or, meal from blood, feather, fish, meat or bone) are the major sources of RUP (NRC, 2001; Schwab et al., 2003). Histidine is most limiting in grass silage-, barley- or oats-based diets, or when cows are fed maize silage- and alfalfa hay-based diets scarce in MP (Vanhatalo et al., 1999; Schwab, 2011; Lee et al., 2012a; Hristov and Giallongo, 2014). These observations show that limitations of dietary AA are dependent on the feedstuffs used in dairy rations (Korhonen et al., 2000; Hadrová et al., 2012).
IMPORTANT LIMITING AMINO ACIDS IN DUTCH DAIRY COWS
The type of feed supplied to Dutch dairy cows has changed drastically during the last several decades. Until the 1950s most rations consisted of grass supplemented with locally available crops (Hollander, 2012). Due to intensification in the 1970s, European countries have been shifting from grazing to conserved forage systems with silage as a key component (van den Pol et al., 2008). Since the 1990s, a greater number of cows are housed in the stable year-round. On Dutch dairy farms, implementation of zero-grazing systems increased from 6% in 1992 to 15% in 2004 (CBS, 13 october 2016). In addition, crops that previously could not be grown robustly in the Dutch climate are now available due to better infrastructure to facilitate their persistence. For example, the implementation of soybean meal, now used readily in Dutch dairy diets, was not commonly used until the 1970s (Kamp et al., 2008). Although most studies investigating limiting AA in dietary diets have been performed in America, France, and Scandinavia, feedstuffs used in those studies are also commonly used in the Netherlands nowadays. As such, considering that maize silage and grass are the main forage components of dairy diets in the Netherlands, it is expected that the AA deficiencies of Met, Lys, and His recognized in these feedstuffs could be limiting in diets of Dutch dairy cows as well.
FUTURE ATTENTION FOR LIMITING AMINO ACIDS
When considering the Law of the Minimum, after supplying the first, second, and third limiting AA there will be a next limiting AA. Therefore, all AA that are fundamental for milk protein synthesis need to be considered. Other potentially limiting EAA for synthesis of milk protein and milk production are the branched-chain amino acids (BCAA) consisting of isoleucine (Ile), leucine (Leu), and valine (Val). Literature assessing the importance of these AA in dairy nutrition are conflicting. Hultquist and Casper (2016) supplemented 0.40 g Val and found an increase in milk production of 0.7 kg. Correspondingly, an effect of Val was found by the study of Haque et al. (2013), where low Val concentrations (change from 5.9 to 4.5% PDIE) could have limited milk protein synthesis, when the requirements for Lys, Met, and His were met. Moreover, this study found that the decrease in plasma concentration of Val was accompanied by a decrease of Met, His, the total EAA, and total NEAA which might indicate that their uptake by the mammary gland was increased. Nevertheless, the study of Appuhamy et al. (2011a) and Korhonen et al. (2002) found no additional response to BCAA supplementation after supplementation with Lys and Met. This conclusion was supported by Mackle et al. (1999), who found no positive effects of BCAA supplementation on milk protein concentration or milk yield. Leucine was studied as a possible limiting AA in diets based on grass-silage by Huhtanen et al. (2002), after a decrease in its arterial concentration was observed after His supplementation (Vanhatalo et al., 1999). However, no effects of Leu on milk yield or composition were reported. This was supported in the studies of Richter et al. (2010) and Krízová et al. (2008), though in these studies cows were fed a maize silage-based diet instead of grass-silage.
Calculation of cysteine (Cys) in pig and poultry diets is common practice, but is rarely considered for dairy cows. Cysteine can be synthesized from Met in the liver and therefore is not an EAA (D’Mello, 2003). Moreover, supplementing free Cys in dairy cows diets results in low mammary gland uptake and therefore will have little effect, indicating Cys is not essential for milk protein synthesis (Pocius et al., 1981). Nevertheless, Cys is a key component to maintain structure and function of the intestine. For example, Cys is needed as a precursor for intestinal mucosal cells to produce polypeptides (Wu, 1998). Met is the precursor for Cys synthesis. Thus, by ensuring adequate Cys supply in the diet, dietary Met can be spared from this process (Wu, 2014). This would result in more Met available for the mammary gland for milk synthesis. Though, it is not known how digestion of Cys and Met exactly works and in which mechanisms they are involved. Therefore, it is difficult to know how these AA relate to each other and how much of Cys should be supplied. In pigs and poultry, digestion is more defined and simplistic, allowing AA supplementation to be more precise than currently realized in dairy cow nutrition.
UTILIZATION OF AMINO ACIDS INSTEAD OF THE LIMITING AMINO ACID THEORY
The theory of a limiting AA can be explained by the Law of the Minimum and suggests that the most deficient EAA is the first limiting factor for protein synthesis even though other EAA may also be limiting (Cant et al., 2003). This mechanism of a single AA-limiting protein synthesis is generally accepted as a valid theory in the regulation of protein synthesis (Hanigan et al., 2002). The theory assumes that single AA are responsible for protein synthesis, and that the synthetic process cannot continue until the requirement for most limiting AA is met (Bequette and Nelson, 2006). However, several studies indicate that multiple AA could stimulate performance separately, which challenges the fact that only one AA is responsible for stimulating protein synthesis. Moreover, it is not simply the amount of AA that drives protein synthesis (Shenoy and Rogers, 1977, 1978). Swanepoel et al. (2015) suggested that the rate of milk production is determined by metabolic mechanisms instead of the concentration of AA in blood. This hypothesis is recognized by Appuhamy et al. (2011b and 2012) who clarified that the mammary gland receives information via intracellular signalling pathways about interactions between the cellular supply of AA, energy, and the hormonal status of the animal. These interactions determine the rate of protein synthesis (Hanigan, 2013). An example of such interaction is the potential alteration in blood flow the mammary gland which affects extraction of nutrients from the arterial supply by the gland in attempt to meet intracellular AA demands. Moreover, it is not per se the AA supply that might limit the milk performance, but the mammary gland capacity for synthesis regulated in part by the number of secreting cells (Rezaei et al., 2016). Hormones can increase the capacity of the mammary gland by reduced apoptosis and increased cell proliferation and differentiation of the secreting mammary cells, though, supplementing hormones are not legal practises (Flint and Gardner, 1994; Choudhary, 2014). Considering these observations about mammary utilization of AA, it can be suggested that the limiting AA theory is not the most accurate representation of the regulation of milk protein synthesis. Instead, focus should be placed on determining the ratio of uptake of individual AA by the mammary gland and their output in milk and describing the individual efficiencies of AA within the process of milk protein synthesis (Rulquin and Pisulewski, 2006). With regards to the mammary gland use, AA can be divided in three groups. Group 1 AA, in which the uptake to output ratio is 1:1, consists of His, Thr, Met and Phe. In contrast the group 2 AA, Lys, Arg, Thr, Val, Ile and Leu, are taken up by the glands in excess of what is required for milk protein synthesis and are used for NEAA synthesis. The third AA group consists of NEAA (Mepham, 1982).
AMINO ACID SUPPLEMENTATION- AND REQUIREMENT-MODELS
FACTORIAL-EMPIRICAL MODEL
To estimate nutrient requirements, feed formulation software’s uses models incorporating years of collected scientific information about digestion and metabolism. The earliest mathematical models have evolved from static, factorial, and algebraic models, to models incorporating empirical relationships (Fuquay et al., 2011). Empirical relationships are used when the system (e.g. herd, animal, organ, cell) represented by the model is complex and there is little known about the functional mechanisms and the relationships between the different system levels (Baldwin, 1995; Fuquay et al., 2011). These models define nutrient requirements for one or multiple performance criteria during a certain period derived from whole-animal observations and experiments (such as a dose-response and blood AA curves) (Thakur, 1991; Rulquin et al., 1993; Hanigan et al., 1997; Pomar et al., 2003). Although this type of model contains multiple other types of equations, the model remains static and factorial in nature. The factorial method is advantageous because parameters such as environmental and nutritional factors that can alter the relationship between AA can be considered (Evans, 2007). However, it is still difficult to estimate efficiency of AA use because these models can not consider all physiological requirements of the cow due to incomplete understanding of post-absorptive AA metabolism and AA interactions in ruminants (Madsen et al., 2005; Robinson, 2010). It became apparent that EAA are not only of significance for anabolic use, but also serve in metabolic pathways including gluconeogenesis and fatty acid metabolism (Bequette and Nelson, 2006). The factorial method utilizes a steady transformation rate of absorbed AA into milk protein (Bequette et al., 1997). However, it is known that utilization of AA by the mammary gland is flexible (Bequette et al., 1997; Madsen et al., 2005). Amino acids needed to support milk synthesis may differ from AA output in milk protein. Moreover, the established variation in AA removal by peripheral tissues based on arterial EAA supply, and its interaction with energy supply and hormone signalling cascades must be considered (Bequette et al., 2000; Arriola Apelo et al., 2014). For these reasons, the factorial-empirical modelling method might overestimate production responses when extra AA are supplied (Chalupa and Sniffen, 2006; Arriola Apelo et al., 2014).
DYNAMIC-MECHANISTIC MODEL
Future models should be dynamic and mechanistic in nature to consolidate the metabolism of nutrients (not only of AA) with their interactions between peripheral tissues, the liver, and the mammary gland (Bannink and Dijkstra, 2014). When more detail is available about the functional mechanisms of digestive processes, mechanistic models can be used (Thakur, 1991). This type of model describes each step in a metabolic process quantitatively and considers broader processes relative to lower-level processes (Fuquay et al., 2011). Since metabolism is dynamic, this model can describe these relationships in a as time dependent manner instead of for just one moment in time. Therefore, stage of lactation and other time-related factors can be incorporated. The major difficulty with dynamic-mechanistic models is how to parametrize and define which data is most suitable for incorporation into the model.
IDEAL PROTEIN CONCEPT (RATIO APPROACH)
Nutrient requirements can be expressed through different nutritional systems and with different units. One way of expressing AA requirements is through the ideal protein concept which depicts AA balances. This concept predicts AA supply for productive functions by formulating optimal concentrations of the most limiting AA as a percentage of MP flow to the small intestine, based on the factorial model (Rulquin et al., 1993). Schwab (1996) expresses limiting AA as a percentage of EAA flow to the small intestine. In the ideal protein concept, one AA, generally Met in ruminant feeds, is set as the standard and the need for other AA are set as a ratio or percentage of the standard AA (Evans, 2007). The ratio of 3:1 for metabolizable Lys and Met is generally accepted in support of a positive lactation response (Chalupa and Sniffen, 2006). This ratio, describes the importance for AA balancing and originates from Schwab et al. (1992), who studied the first limiting AA for milk protein synthesis and the extent of its limitation, and recognized a ratio between Lys and Met in multiple experiments. This finding suggests that a proportional requirement of Lys and Met (as a percentage of total EAA) may describe their optimal contribution to milk protein synthesis. This corresponds with Cole and Van Lunen (1994) who state that a balanced mix of EAA is needed to provide the ideal protein pattern. According to Schwab et al. (2003), the ratio is important to ensure the utilization of AA as efficiently as possible, compared with suppling high amounts of single AA. Using this ratio approach, formulating for the most limiting AA will improve the supply of other EAA, to meet animal requirements (Evans and Patterson, 2007). Moreover, this concept is easy to apply in practice because it is not diet specific, in contrast with expressing individual EAA content as proportion of the feed. A limitation of the ideal protein method is that only the Lys:Met ratio is considered, which implies that the Lys:Met ratio is the most important factor driving milk protein synthesis. In fact, this ‘ideal’ ratio between Lys and Met does not mean that they are entirely used for milk protein synthesis. It is unclear whether the ratio between absorbed AA or the total amount of absorbed AA plays the most important role in supporting milk protein synthesis. For example, there is a limiting efficiency that can be reached when supplying a certain amount of AA that can be absorbed and used (Schwab et al., 2003). Moreover, when AA requirements are defined as a proportion of MP, individual AA supply cannot be manipulated which might result in feeding unnecessarily high CP content (Arriola Apelo et al., 2014). In conclusion, defining fixed proportions of AA requirements can be inaccurate and inefficient. Therefore, when supplementing rations with Lys and Met, both the ratio of AA and the net amount of individual AA are important (Swanepoel et al., 2015). As such, a combination of the factorial and ideal ratio approach is most commonly used.
DUTCH RECOMMENDATIONS FOR METHIONINE AND LYSINE SUPPLEMENTATION
Prediction of AA requirements differs across nutritional models around the world. Unfortunately, the Dutch protein system (DVE/OEB; Tamminga et al., 1994) does not yet have the capability to account for the absolute amounts or ratios of AA, so there is no official recommendation for supplementing Lys and Met. Despite this limitation, the ‘centraal veevoeder bureau’ (CVB), a Dutch organization that evaluates feedstuffs for farm animals, attempted a conversion. Since the Dutch protein system is constructed from the French system (INRA, 1989) they estimated Dutch Lys and Met values (DVLys and DVMet) of DVE based on the French Lys and Met values (LysDI and MetDI) of PDIE, using Dutch feeding rations (Tedeschi et al., 2015). This resulted in the advice of 6.0 DVLys 2.4 DVMet and as a percentage of DVE in the Dutch system compared to 6.8 LysDI and 2.2 MetDI as a percentage of PDIE in the French system. Despite this attempted estimation, an earlier study of Tamminga (2007) concluded that it was not possible to use this method of conversion to make a recommendation about Lys and Met, because it fails to consider the different assumptions and calculations of both nutritional models. Some of these differences include the calculations of protein reaching the small intestine (PDIE versus DVE), the correction for higher DMI, and the fact that the Dutch system allows requirements specific to performance of cows in Dutch dairy systems. As such, to be able to create correct recommendations for AA supply it is necessary to completely understand the differences between prediction systems. Using databases with analysis of dose-responses and detailed information on experimental conditions can be used. In 1997, the Schothorst feed evaluation and research centre, collected these data about AA from studies in the Netherlands and recommend 2.2 Met and 5.7 Lys as a percentage of true metabolizable protein (TMP) (van Straalen et al., 1997). At this stage, recommendations from other studies and the Schothorst database are used as a guideline for Lys and Met supply in Dutch dairy diets.
EFFECTS OF METHIONINE AND LYSINE SUPPLEMENTATION ON LACTATION PERFORMANCE
There are multiple studies which show positive results from supplementing Met, Lys, or both, of which some have been summarized in Appendix 2. It should be noted that overall these studies show variable cow responses with regards to feed intake and yield of milk and its constituents (Robinson, 2010; Zanton et al. 2014; Awawdeh, 2016). A possible explanation for these variable responses might be that the studies used different types of diets and vary in cow breeds, housing and climate conditions (Amrutkar et al. 2015). Many studies use maize-based diets which are rich in Met and low in Lys, compared to grass-based diets where the opposite trend is found (Awawdeh, 2016). Moreover, the level and type of supplemental AA and the route of administration (jugular infusion or mixed in the diet) can produce different results. Nevertheless, Robinson (2010) succeed in reviewing literature about Lys and Met supplementation and made some overall conclusions: milk protein content and milk fat content might be improved by Met supplementation, milk yield and milk protein content could be improved by Lys and Met supplementation, and supplementing Lys alone may decrease DMI. Moreover, production responses are greater when moderate CP levels (16-18% DM basis) are supplemented with Met or Lys compared with CP levels above or below this level, which suggests that individual AA requirements cannot be considered independently of the level of other AA (Rulquin et al., 1994). Furthermore, improvements due to Lys and Met are mostly observed in high producing dairy cows in early lactation (NRC, 2001). From a practical perspective, dairy farmers are most interested in which brand of supplemental AA to feed, how much and in what form, and what kind of financial return to expect.
NEGATIVE EFFECTS OF METHIONINE OR LYSINE
As described there is a limit on supplying extra Met or Lys on production performance of lactating dairy cows. Robinson et al. (2000) suggested that infusing an oversupply of Lys or Met of 140-150% results in a negative effect on total milk and lactose yield. Moreover, Satter et al. (1975) found that toxicity, defined as a drop in DMI during the AA infusion of at least two standard deviations, can arise when cows got about 10% DL-Met-analogue on a DM basis. In practise, Met toxicity is not a threat since such high level of dietary Met cannot be easily reached (Robinson et al., 2000). The current suggested amount is 25 g Met/d, approximately 1% of the ration DM. (Bequette and Nelson, 2006) suggested that 4-thiomethyl-2-hydroxybutanoic acid (HMBA), an artificial hydroxyl-analogue of Met used most commonly in the current industry, is not toxic.
METHIONINE AND LYSINE SOURCES
Since it was discovered that Lys and Met concentrations are not always optimal in ruminant diets, research groups began to experiment with supplementation of different sources of Lys and Met. To supply enough Lys and Met to the small intestine, feedstuffs or the AA themselves need to be protected from rumen degradation. It is possible to physically protect AA from rumen degradation by encapsulating them in lipid, in conjunction with inorganic materials and carbohydrates (Chalupa, 1975; Smith and Boling, 1984; Schwab and Ordway, 2003). Encapsulation can be done by embedding AA in matrices of protein or fat or by coating cores composed of primarily AA (Papas et al., 1984; Wu and Papas, 1997). Another type of protection is to coat the surface of the AA substance with enzyme resistant and pH-sensitive polymers (Papas et al., 1984). Due to the coating the product can maintain its structural integrity in the rumen (pH 5.8 to 7.2). In the abomasum (pH 3.5-4.0) the lower pH activates breakdown of the coating releasing AA for absorption in the small intestine (Papas et al., 1984; Wu and Papas, 1997). Moreover, it has been shown that derivatives and analogues of Met (small chemical modifications in the DL-methionine molecule), can effectively protect the AA from ruminal degradation (Koenig et al., 2002). Derivatives are created by adding a chemical obstruction to the α-amino fragment or by altering the acyl fragment while keeping the nitrogen group (Schwab, 1995), while analogues are created by changing the α-amino fragment of the AA into a non-nitrogenous fragment (Schwab and Ordway, 2003). The most thoroughly studied and effective analogues are presented by Friedman (1989) and Schwab (1995). One of the most effective analogues is the hydroxy analogue of Met, D,L-2-hydroxy-4-(methylthio)-butanoic acid (HMB). However, Graulet et al. (2005) showed that the isopropyl of HMB (HMBi) has even more potential, because the added ester group should provide additional protection from ruminal degradation by microbes protentially due to its quick absorption by the rumen wall. It is expected that half of the Met in the HMBi from is metabolized by rumen microbes and the other half is available for the animal (Graulet et al., 2004). The differences in molecule structure between the forms of methionine can be found in Appendix 3. Commercial rumen-protected (RP) Met products are well established, while RP Lys is a relatively new product (Awawdeh, 2016). Currently most Lys products are protected by lipid. An overview of the different types and brands of supplemental Lys and Met can be found in Appendix 4.
OTHER EFFECTS OF AMINO ACID SUPPLEMENTATION
There are various reviews that assess the impacts of a RPAA on lactation performance. However, AA have many functional effects and play a role in physiological and biological processes (Yoneda et al., 2009; Wu, 2010; Liu et al., 2013). Due to these functional effects, AA can also influence health and reproduction.
REPRODUCTION AND HEALTH
By providing an optimal AA profile for protein synthesis, blood urea levels can be decreased. This may have a positive effect on reproductive performance in addition to N efficiency, because it has been shown that there is a undesirable relationship between urea concentrations in milk and blood and fertility for high producing cows linked to a low uterine pH (Liu et al., 2013). Moreover, it seems that during gestation, uterine fluid increases in Met, Lys and His concentration, and to a lesser extent Arg, Phe, Ile concentration indicating these AA might be needed for embryo development (Wiltbank et al., 2014). Furthermore, there are some suggestions that supplementation specifically of RPMet could have positive effects on reproduction, next to its effect on total milk and protein yield (Liu et al., 2013; Acosta et al., 2016; Acosta et al., 2017; Vailati-Riboni et al., 2017). Peñagaricano et al. (2013) found that there was no effect of supplementing Met on morphological changes in the embryo, but found differences in embryo gene expression related to early development and immune response, which may support positive pregnancy outcomes and improved offspring physiology. Acosta et al. (2016) found that Met supplementation affects preimplantation of the embryo, increasing the chance of embryo survival.
The immunological status of a cow is dependent on functional liver metabolism, infectious diseases and stress (Zhou et al., 2016). Supplementing Met to an already Met-sufficient diet has improved immune response and liver function in cows, due to the increased production of macrophages and neutrophils. Increased neutrophil response and lower somatic cell count was reported by Li et al. (2016) when Met was supplemented in the diet. These finding are in agreement with Thiaucourt (1996) who found reduced somatic cell count and improved milk production when Lys and Met were supplemented.
CONCLUSIONS
Based on this literature research it can be suggested that Lys and Met may be the most limiting AA in Dutch dairy cow diets. Moreover, it can be concluded that supplementation of Lys and Met influence lactation performance, but responses to rumen protected AA are dependent on the protection of the product. Responses to AA supplementation are also influenced by factors like dietary energy level, hormone signalling and AA metabolism. During the last centuries protein metabolism has been explored extensively. Nonetheless, it still is challenging to determine the most efficient AA balance for dairy cows, since the present prediction of the AA requirements with nutritional models are not adequately detailed. Greater emphasis should be placed on the systems involved in AA metabolism, the efficiency of using protein sources, and the implication of AA profiles (Cant et al., 2003; Arriola Apelo et al., 2014; Tedeschi et al., 2015). Focus should be placed on how individual AA can meet the needs for physiological processes, such as stimulating milk synthesis (Arriola Apelo et al., 2014). Moreover, it should be considered how these mechanisms can be quantified and incorporated into current nutritional models. For now, calculating AA supply based on the optimum profile of AA available for absorption in the small intestine is a functional approach. Individual AA supplementation can improve metabolizable AA profile such that it is closer to the AA profile needed for maintenance, production, reproduction and gestation. Moreover, it may increase N efficiency in cattle thereby, reducing N excretion into the environment.