Chemical structure

The vitamin folic acid (chemical name pteroylglutamic acid) consists of three parts: a pteridine nucleus, para-aminobenzoic acid and glutamic acid⁽Reference Girard⁵⁾. The name folic acid is deduced from folium, the Latin word for leaf, because native forms of folic acid were originally isolated from spinach leaves⁽Reference Mitchell, Snell and Williams⁶⁾. In chemistry the name folic acid is only used for the synthetic form. It is a stable compound and the basal structure of a wide family of vitamin coenzymes⁽Reference Lucock⁷⁾. In nature, more than 100 compounds, with the basal structure of folic acid, feature a common vitamin activity. These pteroylglutamate forms of folic acid are generally called folates⁽Reference Girard^5, Reference Bender and Bender^8, Reference Finglas, Wright, Wolfe, Hart, Wright and Dainty⁹⁾. Native folates vary in three chemical characteristics from folic acid: first, in the level of reduction of the pteridine nucleus (dihydrofolate or tetrahydrofolate (THF)); second, in the character of the one-carbon substituent linked to the N atoms N-5 and N-10 (for example, formyl, formimino, methyl, methylene or methenyl residues); third, in the chain length of the glutamyl residues which can be linked to the γ-carboxyl group of the glutamate via peptide linkages⁽Reference Girard^5, Reference Wagner, Olson, Broquist, Chichester, Darby, Kolbye and Stalvey¹⁰⁾.

Absorption and biochemical functions

There are several excellent reviews on absorption and biochemical functions of folates⁽Reference Scott^11–Reference Bässler¹⁴⁾. Derived from studies with non-ruminant animals, two mechanisms of folate absorption from the intestinal tract seem to exist: an active saturable process and a non-saturable passive process. In fact, the relative importance of passive absorption changes according to folate supply, increasing with the amounts of folates available⁽Reference Selhub, Dhar and Rosenberg^13, Reference Bässler, Golly, Loew and Pietrzik¹⁵⁾. However, folates are perhaps degraded, converted and synthesised in the forestomachs of ruminants⁽Reference Zinn, Owens, Stuart, Dunbar and Norman¹⁶⁾, and even absorbed on a small scale⁽Reference Rérat, Molle and le Bars¹⁷⁾. Unfortunately the forms and the availability of the forms present in rumen contents and duodenal digesta are unknown.

In bovine blood, mainly 5-methyl-THF is found⁽¹⁸⁾. Cells take up this compound and demethylate it to THF. To retain THF in cells it must be converted by folylpolyglutamate synthase to polyglutamate THF, the coenzymic form of folates⁽Reference Bässler¹⁴⁾. Polyglutamate THF is involved in several biochemical pathways in mammals⁽Reference Choi and Mason¹⁹⁾. Mainly, folates are donors and acceptors of one-carbon units⁽Reference Bender and Bender^8, Reference Benevenga²⁰⁾. Thus they are involved in the remethylation of homocysteine to methionine, as an essential part of the methylation cycle. This reaction is also vitamin B₁₂ dependent, as the catalysing enzyme methionine synthase needs vitamin B₁₂⁽Reference Scott¹¹⁾. Furthermore, the transfer of one carbon unit involves folates in the synthesis of purines and pyrimidines and thereby in DNA synthesis and cell proliferation⁽Reference Bässler¹⁴⁾. THF is regenerated after these catalytic reactions⁽Reference Bässler¹⁴⁾. However, folate disappears through urinary excretion and through bile, although a very effective reabsorption by the enterohepatic cycle exists⁽Reference Bässler, Golly, Loew and Pietrzik¹⁵⁾. Up to now only one study reported on the urinary excretion of folates in dairy cows (nine animals) after intramuscular (i.m.) injection of 0·3 mg folic acid per kg BW⁽Reference Girard and Matte²¹⁾. The authors found an excretion of the injected dose of 35·1 % after 8 h and 44·2 % after 48 h.

A deficit of folates can lead to a decrease in S-adenosylmethionine levels and to an abnormal DNA precursor metabolism resulting in faulty DNA synthesis and a decrease in NAD⁽Reference James, Miller, Mcgarrity and Morris²²⁾, as a decrease in NAD levels is consistent with an increase in DNA repair activity⁽Reference James, Yin and Swendseid²³⁾. An indirect lack of folates can be caused by a vitamin B₁₂ deficit. This results in an accumulation of 5-methyl-THF called a methyl-trap, as 5-methyl-THF cannot be regenerated to THF⁽Reference Bässler, Golly, Loew and Pietrzik¹⁵⁾. If so, cells are unable to conjugate absorbed folate monoglutamates, resulting in a decreased intracellular folate polyglutamate level. Additionally, intracellular folate accumulation declines as only polyglutamates can be retained in cells⁽Reference Scott and Weir²⁴⁾.

As folates influence DNA synthesis and the methionine cycle, they are involved in the metabolic pathways of reproduction and milk protein synthesis; therefore they are very important especially in gestating and lactating cows. An additional special situation for cows is that they have a very high demand for methyl groups in early lactation. Concurrently some precursors for methylated compounds (for example, serine and glycine) are also needed for gluconeogenesis, as the amounts of glucose reaching the small intestine through the digestive system are generally low. So, coincident demand for precursors of methylated compounds leads to competition between different metabolic pathways, for example, gluconeogenesis, lecithin synthesis, DNA synthesis and remethylation of methionine⁽Reference Girard and Matte^25, Reference Bruesemeister and Suedekum²⁶⁾.

Sources and stability of folates

The following section gives a survey of approximate folate concentrations in some feeds and foods. Different folate contents are given in the literature for the same feedstuffs (Table 1) and for most feeds they are not analysed at all, so only few data are available. It must be considered that as compilations were used in Table 1, the number of samples analysed is not known. Additionally, it must be pointed out that the native folate concentrations in feeds vary due to influences of climate, species, vegetation stage, habitat and fertiliser⁽Reference Albers, Gotterbarm, Heimbeck, Keller, Seehawer and Tran²⁷⁾. Furthermore, most naturally occurring folates are chemically relatively unstable. Thus folates exhibit a significant loss of activity during harvesting, storage and processing, but measured folate concentrations are also highly influenced by the method used for sample preparation⁽Reference Gregory²⁸⁾. The synthetic form, folic acid, is more resistant to chemical oxidation⁽Reference Scott¹¹⁾.

Table 1 Folate content of several feeds given in the literature

LfL, Bayerische Landesanstalt für Landwirtschaft.

The figures generally show very low folate contents in feedstuffs. However, quantities are not the only decisive factor; the folates in the feed must also be available for absorption⁽Reference Wagner, Olson, Broquist, Chichester, Darby, Kolbye and Stalvey¹⁰⁾. The so-called bioavailability describes the proportion of an orally administered dose which is available in plasma after absorption⁽Reference van den Berg and Schlemmer²⁹⁾. It is difficult to consider the bioavailability of native folates because unknown numbers and amounts of folate metabolites exist in every plant species or feedstuff. For ruminants the assessment of folate bioavailability is more difficult as their micro-organisms in the rumen synthesise, but also degrade, ingested folates. The synthesis and degradation of folates in the rumen is crucial for the amount absorbed from the small intestine of ruminants. Up until now the number of studies on rumen folate synthesis and degradation has been very low; therefore for ruminants no values of folate bioavailability from feedstuffs are available. Furthermore the bioavailability of native folates is influenced by different physico-chemical properties and certain feed constituents. For example, polyglutamyl folates have a lower bioavailability than monoglutamyl folates, as polyglutamyl folates must be hydrolysed to monoglutamates before absorption⁽Reference Seyoum and Selhub³⁰⁾. Additionally, the actual amount available for each individual animal varies depending on differences in intestinal pH or general living conditions⁽Reference Wagner, Olson, Broquist, Chichester, Darby, Kolbye and Stalvey¹⁰⁾.

Microbial synthesis, degradation and absorption of folates in the gastrointestinal tract of ruminants

It is well known that the microbial activity and the ruminal population are influenced by the level of concentrates in the diet and the type of feed⁽Reference Hungate and Hungate³¹⁾. As some bacterial species are able to synthesise folates, and some others need them⁽Reference Wolin, Miller and Hobson³²⁾, different amounts of folates can be synthesised and used in the rumen depending on the feed composition. For steers, Hayes et al. ⁽Reference Hayes, Mitchell, Little and Bradley³³⁾ and Girard et al. ⁽Reference Girard, Chiquette and Matte³⁴⁾ described a relationship between the proportion of concentrates in the diet and the amount of folates in the rumen. High-concentrate diets resulted in an increase of folates (Table 2). The authors hypothesised that this increase is due to an enhanced microbial activity in the rumen, caused by rapidly degradable carbohydrates. But it must also be considered that concentrations are not necessarily representative of the total amount synthesised in the rumen, as digesta passage and rumen volume could vary between treatments, for example, due to fibre differing greatly in amount and length. Santschi et al. ⁽Reference Santschi, Chiquette, Berthiaume, Martineau, Matte, Mustafa and Girard³⁵⁾ could not corroborate this hypothesis for cows, because they found no difference in the amount of folates in the liquid fraction of ruminal content between the high-forage (58 % forage) and low-forage (37 % forage) diets (Table 2). However, the concentrate:forage ratio of the two diets used in the study of Santschi et al. ⁽Reference Santschi, Chiquette, Berthiaume, Martineau, Matte, Mustafa and Girard³⁵⁾ was not as extreme as in the studies with the steers⁽Reference Hayes, Mitchell, Little and Bradley^33, Reference Girard, Chiquette and Matte³⁴⁾ and additionally Santschi et al. ⁽Reference Santschi, Chiquette, Berthiaume, Martineau, Matte, Mustafa and Girard³⁵⁾ used diets with more ingredients than Girard et al. ⁽Reference Girard, Chiquette and Matte³⁴⁾ and Hayes et al. ⁽Reference Hayes, Mitchell, Little and Bradley³³⁾ (Table 2). Furthermore the steers had a BW of 340 kg⁽Reference Hayes, Mitchell, Little and Bradley³³⁾ and 352 (se 27) kg⁽Reference Girard, Chiquette and Matte³⁴⁾, whereas primiparous and multiparous cows weighed 582 (se 17) kg and 692 (se 17) kg⁽Reference Santschi, Chiquette, Berthiaume, Martineau, Matte, Mustafa and Girard³⁵⁾, respectively, which resulted in a different DM intake between cows and steers (Table 2). It should be noted that in all three studies folate concentrations in ruminal fluid were very different. One reason for this could possibly be the different diet composition used in the three studies (Table 2). Additionally, Hayes et al. ⁽Reference Hayes, Mitchell, Little and Bradley³³⁾, who found the highest values, used a different sample preparation method from Girard et al. ⁽Reference Girard, Chiquette and Matte³⁴⁾ and Santschi et al. ⁽Reference Santschi, Chiquette, Berthiaume, Martineau, Matte, Mustafa and Girard³⁵⁾. The higher values found by Hayes et al. ⁽Reference Hayes, Mitchell, Little and Bradley³³⁾ in the supernatant fraction could result from bacterial content, as the samples were centrifuged after freezing and thawing, which could have destroyed bacterial cells. Contrary to this, Girard et al. ⁽Reference Girard, Chiquette and Matte³⁴⁾ and Santschi et al. ⁽Reference Santschi, Chiquette, Berthiaume, Martineau, Matte, Mustafa and Girard³⁵⁾ centrifuged their samples before freezing. Furthermore, Hayes et al. ⁽Reference Hayes, Mitchell, Little and Bradley³³⁾ used a microbiological assay (Streptococcus faecalis) to analyse folate concentration in ruminal samples. In contrast, Girard et al. ⁽Reference Girard, Chiquette and Matte³⁴⁾ and Santschi et al. ⁽Reference Santschi, Chiquette, Berthiaume, Martineau, Matte, Mustafa and Girard³⁵⁾ analysed their samples by radioassay.

Table 2 Folate content of ruminal material and body weight (BW) (at the beginning of the trial) of steers and cows

(Mean values with their standard errors)

^a,b,c Mean values within a trial with unlike superscript letters were significantly different (P ≤ 0·05).

* In these studies no DM values of feed were given; therefore fresh matter values are presented here.

† This study was conducted with steers.

‡ In this study values of the solid and liquid fraction of the ruminal content were given as area under the curve; concentrations presented here are calculated on this basis.

§ Data from primiparous cows.

∥ Data from multiparous cows.

An in vitro study by Hall et al. ⁽Reference Hall, Cheng and Burroughs³⁶⁾ showed that the degradation of fibrous materials by rumen micro-organisms increases (42 %) with supplementation of folic acid (100 μg/20 ml medium). So it seems that cellulolytic micro-organisms require folates, which would endorse the findings of Hayes et al. ⁽Reference Hayes, Mitchell, Little and Bradley³³⁾ and Girard et al. ⁽Reference Girard, Chiquette and Matte³⁴⁾, who found decreased folate concentrations with high-forage diets. Studies on folate requirements of micro-organisms are rare. Some strains of Ruminococcus flavefaciens, a cellulolytic rod, seem to require either folic acid, THF or p-aminobenzoic acid⁽Reference Ayers^37–Reference Bryant and Robinson⁴⁰⁾ (which is a part of folic acid). Also two strains of Ruminococcus albus require folic acid⁽Reference Bryant and Robinson⁴⁰⁾. Furthermore Hayes et al. ⁽Reference Hayes, Mitchell, Little and Bradley³³⁾ observed that the level of folates in the ruminal fluid correlated negatively with pH. This corroborates the previous findings, because R. flavefaciens is sensitive to acid⁽Reference Miyazaki, Hino and Itabashi⁴¹⁾ and ruminal pH increases with diets rich in fibre. Since these studies were conducted, feeding strategies, genetics and keeping conditions have changed, and it would be interesting to see if similar and if possible even more detailed results relating to species of micro-organisms could be obtained today.

Consequent to folic acid supplementation to steers, Girard et al. ⁽Reference Girard, Chiquette and Matte³⁴⁾ observed an increase in folate concentration in the solid and liquid fractions of the rumen contents compared with a supplement-free diet (P = 0·0001 for both fractions; Table 2). However, different from supplement-free diets, the concentrate:forage ratio had no influence on the ruminal folate concentration. Furthermore, they observed that neither folic acid supplementation nor the nature of the diet made a difference to the quantity of protein synthesised per unit of microbial mass⁽Reference Girard, Chiquette and Matte³⁴⁾. Chiquette et al. ⁽Reference Chiquette, Girard and Matte⁴²⁾ analysed the effect of folic acid supplementation on digestibility and ruminal fermentation in growing steers. They noticed a tendency (P = 0·08) of pH to decline 4–8 h after feeding a diet consisting of 70 % rolled barley, 30 % timothy hay and a supplementation of 2 mg folic acid per kg BW as compared with the unsupplemented diet. The results showed that the concentration of ruminal acetate and butyrate did not change due to folic acid supplementation, whereas ruminal propionate concentrations increased (P ≤ 0·05) after feeding, and the acetate:propionate ratio was numerically higher during the 24 h of observation due to folic acid supplementation. The apparent digestibility of DM, fibre fractions and crude protein was not influenced by the addition of folic acid⁽Reference Chiquette, Girard and Matte⁴²⁾. So it seems that folic acid has no major influence on digestibility and ruminal fermentation, but until today these processes have only been tested once (with eight steers), comparing few different diet compositions, so it is difficult to extrapolate these data to other experimental conditions.

Santschi et al. ⁽Reference Santschi, Berthiaume, Matte, Mustafa and Girard⁴³⁾ and Schwab et al. ⁽Reference Schwab, Schwab, Shaver, Girard, Putnam and Whitehouse⁴⁴⁾ determined the daily apparent ruminal folate synthesis for lactating cows without supplementation of folic acid, whereas Zinn et al. ⁽Reference Zinn, Owens, Stuart, Dunbar and Norman¹⁶⁾ provided data for growing steers (Table 3). On average, Schwab et al. ⁽Reference Schwab, Schwab, Shaver, Girard, Putnam and Whitehouse⁴⁴⁾ calculated 16·2 mg daily apparent folate synthesis and Santschi et al. ⁽Reference Santschi, Berthiaume, Matte, Mustafa and Girard⁴³⁾ 20·0 mg. Strikingly opposing these results, Zinn et al. ⁽Reference Zinn, Owens, Stuart, Dunbar and Norman¹⁶⁾ on average calculated a negative daily apparent ruminal folate synthesis of − 0·1 mg for growing steers (194 kg BW). These results could be due to the fact that growing male animals which were used in their experiment had a much lower organic matter intake (3·44 kg organic matter/d) than the adult female animals of Santschi et al. ⁽Reference Santschi, Berthiaume, Matte, Mustafa and Girard⁴³⁾ (18·4 kg organic matter/d; calculated from DM intake and ash content of the diet) and Schwab et al. ⁽Reference Schwab, Schwab, Shaver, Girard, Putnam and Whitehouse⁴⁴⁾ (18·7 kg organic matter/d) and due to the differences in the ruminal passage rate. Additionally, the negative values of Zinn et al. ⁽Reference Zinn, Owens, Stuart, Dunbar and Norman¹⁶⁾ could also result from a dietary effect, as Zinn et al. ⁽Reference Zinn, Owens, Stuart, Dunbar and Norman¹⁶⁾ fed a diet with a very high concentration of maize grain, in contrast to diets used in the studies of Santschi et al. ⁽Reference Santschi, Berthiaume, Matte, Mustafa and Girard⁴³⁾ and Schwab et al. ⁽Reference Schwab, Schwab, Shaver, Girard, Putnam and Whitehouse⁴⁴⁾ (Table 3).

Table 3 Folate intake, duodenal flow, ileal flow, apparent synthesis (AS) and body weight (at the beginning of the trial) of steers and cows

(Mean values with their standard errors)

NFC, non-fibre carbohydrates; s., steers; m., multiparous cows; p., primiparous cows.

* Apparent synthesis = duodenal flow minus intake.

† Values did not differ significantly.

‡ In Zinn et al. ⁽Reference Rérat, Molle and le Bars¹⁷⁾ and Santschi et al. ⁽Reference Schwab, Schwab, Shaver, Girard, Putnam and Whitehouse⁴⁴⁾ apparent ruminal synthesis was not calculated by the authors, but daily intake and duodenal flows were given. Furthermore, Santschi et al. ⁽Reference Schwab, Schwab, Shaver, Girard, Putnam and Whitehouse⁴⁴⁾ declare no levels of significance at all, therefore it was not possible to characterise significance in these studies.

§ Forage = 50 % maize silage, 33 % lucerne hay, 17 % grass hay.

∥ sem, not se, was used.

¶ Significant effects of NFC (P ≤ 0·05).

** Significant effects of forage (P ≤ 0·05).

It must be pointed out that disappearance rates, expressed as the amount of folates appearing at the duodenum relative to the quantity given, were very high in all experiments with dietary supplements of folic acid (about 97 %)⁽Reference Zinn, Owens, Stuart, Dunbar and Norman^16, Reference Santschi, Berthiaume, Matte, Mustafa and Girard⁴³⁾. Therefore it is arguable whether unprotected folic acid can be supplemented effectively. However, one has to keep in mind that disappearance could either be caused by degradation or absorption⁽Reference Zinn, Owens, Stuart, Dunbar and Norman¹⁶⁾. Indeed, a net flux across the rumen wall was only found if high amounts of folic acid were present in the rumen. So it seems that the rumen wall is able to absorb folic acid. However, the efficiency is very low, so net flux across the rumen wall into the blood circulation can be neglected⁽Reference Girard, Lapierre, Desrochers, Benchaar, Matte and Remond^45, Reference Girard and Remond⁴⁶⁾. On the other hand, folates are absorbed at the proximal intestine⁽Reference Girard and Remond⁴⁶⁾, and therefore they could already be absorbed at the proximal duodenum before the cannula, hence ruminal disappearance of folic acid could be overestimated⁽Reference Santschi, Berthiaume, Matte, Mustafa and Girard⁴³⁾.

Table 3 gives a survey of the amount of folates found in the duodenum. The apparent intestinal disappearance (between duodenal and ileal cannula) seems to be very low⁽Reference Santschi, Berthiaume, Matte, Mustafa and Girard⁴³⁾. Without supplementation of folic acid, the duodenal flow of folates was lower than the ileal flow, and it rose above the ileal flow only by supplementation of folic acid (Table 3). So it seems that there is no apparent intestinal disappearance of folates without supplementation of folic acid and with supplementation the apparent intestinal disappearance approximates 4 %⁽Reference Santschi, Berthiaume, Matte, Mustafa and Girard⁴³⁾. Santschi et al. ⁽Reference Santschi, Berthiaume, Matte, Mustafa and Girard⁴³⁾ hypothesised that the apparent intestinal absorption of folates is underestimated, as folates are extensively recycled by the enterohepatic cycle and then released between the duodenal and ileal cannula – thus explaining the higher values of folates in the ileal flow.

Generally, Girard et al. ⁽Reference Girard, Lapierre, Desrochers, Benchaar, Matte and Remond⁴⁵⁾ reason that for dairy cows, folate absorption in the gastrointestinal tract is an active saturable process. With dietary supplements of 2·6 g folic acid/d this process was already saturated, as higher supplementations could not effectively increase the amount of folate reaching the blood circulation. Due to the destruction of folates by micro-organisms, and the low importance of passive absorption, they deduced the minor efficiency of folate absorption in ruminants ( < 5 %)⁽Reference Girard, Lapierre, Desrochers, Benchaar, Matte and Remond⁴⁵⁾v. humans (10–46 %)⁽Reference Zettner, Boss and Seegmiller⁴⁷⁾. In general it should be considered that at present no studies are available comparing the amount of folates in the ruminal content and the amount in the duodenum, hence no statement can be made on the coherence between the amount of folates in the ruminal content and the amount of folates in the digesta at the duodenum.

Folate concentrations in blood, milk and liver

Blood

Table 4 shows serum folate concentrations in different feeding studies with cows. Without supplementation of folic acid serum folate levels varied between 13·6 and 17·2 ng/ml. In plasma Santschi et al. ⁽Reference Santschi, Chiquette, Berthiaume, Martineau, Matte, Mustafa and Girard³⁵⁾ found significantly (P = 0·005) lower folate concentration for primiparous cows (12·7 (se 1·6) ng/ml) compared with multiparous cows (18·4 (se 1·6) ng/ml). These observations were not affected by the composition of the diet (58 or 37 % forage).

Table 4 Average serum folate concentration (ng/ml) of cows with and without dietary supplementation of folic acid

(Mean values with their standard errors)

p., Primiparous cows; m., multiparous cows; BW, body weight.

* Concentration determined in serum from cows before parturition.

† Girard et al. ⁽Reference Girard, Matte and Tremblay⁵⁰⁾ and Girard et al. ⁽Reference Girard, Matte and Tremblay⁵⁴⁾ declare no levels of significance between cows before parturition and after parturition; therefore it was not possible to characterise significance in these studies.

‡ Concentration determined in serum from cows after parturition.

§ Significant effect between primiparous and multiparous cows.

∥ Significant differences (P ≤ 0·05) between control and folate groups.

¶ Concentrations determined in serum from cows 18 weeks after calving.

Without supplementation, serum and plasma folate concentrations increase after parturition, as the maternal–fetal complex no longer requires folates (Fig. 1). However, starting on the day of parturition, the cow requires folates for milk production, but it seems that this demand is lower than that for the maternal–fetal complex⁽Reference Graulet, Matte, Desrochers, Doepel, Palin and Girard^48, Reference Girard and Matte⁴⁹⁾. From mating (about 2 months after previous parturition) to parturition Girard et al. ⁽Reference Girard, Matte and Tremblay⁵⁰⁾ found a decrease of total serum folates of 40 %, indicating that the maternal–fetal complex has an increasing demand for folate. Also, Girard & Matte⁽Reference Girard and Matte²¹⁾ discovered a higher demand for folates in the tissues of lactating and gestating cows than in lactating non-gestating cows. Serum clearance was significantly slower (P = 0·04) in non-gestating cows after an intravenous injection of 50 μg folic acid. However, in another study, Girard & Matte⁽Reference Girard and Matte⁴⁹⁾ could not find a difference in serum folate concentrations between gestating and non-gestating cows.

Fig. 1 Serum folate concentration of cows fed different daily folic acid supplements (according to Girard & Matte⁽Reference Girard and Matte⁴⁹⁾, modified). (–♦–), Unsupplemented; (–□–), 2 mg folic acid/kg body weight; (–▲–), 4 mg folic acid/kg body weight.

Oral supplementation (Table 4) or i.m. injection of folic acid significantly increased serum folate concentrations⁽Reference Girard, Chiquette and Matte^34, Reference Girard, Lapierre, Desrochers, Benchaar, Matte and Remond^45, Reference Girard and Matte^49–Reference Girard, Matte and Tremblay⁵⁴⁾. Actually, oral doses higher than 0·5 mg folic acid/kg BW are required to produce a noticeable effect in serum folate concentration, as concluded from a dose–response study with heifers (150 kg BW)⁽Reference Girard, Matte and Levesque⁵¹⁾. However, the heifers used in the present study were very young and therefore full ruminal function may not have been developed. Comparable investigations have not been carried out again and in all further studies supplementations higher then 0·5 mg folic acid/kg BW were used. As shown in Fig. 1, folic acid supplementation increased serum folate concentrations in cows from 4 weeks before calving until calving. The increase intensified with higher dietary supplementation⁽Reference Girard and Matte⁴⁹⁾. Unfortunately no further measurements were performed between the initial sampling time 4 weeks before calving and at the time of calving, so it is not evident if serum folate levels had already increased before calving due to dietary supplementation. After calving, supplemented cows had decreasing serum folate concentrations. The more folic acid was added to the diet, the sharper the decline, reaching a plateau value 16 weeks after parturition⁽Reference Girard and Matte⁴⁹⁾. As well as the influence of folic acid supplementation, Girard & Matte⁽Reference Girard and Matte⁴⁹⁾ (P = 0·0001) and Girard et al. ⁽Reference Girard, Lapierre, Matte and Lobley⁵³⁾ (P = 0·02) showed a significant time effect, as the gain in serum folate concentration due to dietary supplementation was greater in the first 8 weeks of lactation than later in lactation (Fig. 1). It seems that increasing serum folate concentrations during early lactation could result from a decreased ability of the cells to retain and use folates⁽Reference Girard, Lapierre, Matte and Lobley⁵³⁾. A reason for this may be generally lower serum vitamin B₁₂ levels (181 pg/ml) at early lactation compared with 252 pg/ml in the later lactation⁽Reference Girard, Lapierre, Matte and Lobley⁵³⁾; therefore folates can get into the methyl-trap mentioned above.

Three studies are available dealing with an influence of folic acid supplementation on packed cell volume and Hb concentrations in blood. In one study with an oral supplementation of 4 mg folic acid/kg BW an increase (P ≤ 0·05) in packed cell volume and Hb in primiparous cows was found 16 weeks after parturition compared with non-supplemented primiparous cows⁽Reference Girard and Matte⁴⁹⁾. Oral supplementation of folic acid had no effect on these variables in multiparous cows⁽Reference Girard and Matte^49, Reference Girard, Lapierre, Matte and Lobley⁵³⁾. These effects could be explained by generally lower vitamin B₁₂ levels in primiparous compared with multiparous cows, as folates and vitamin B₁₂ are both required for DNA synthesis (as described in the ‘Absorption and biochemical functions’ section)⁽Reference Girard and Matte⁴⁹⁾. Hence a lack of each individual vitamin or of both vitamins together can delay the maturation of blood cells⁽Reference Bills, Koury, Clifford and Dessypris⁵⁵⁾. Folic acid supplemented to primiparous cows, which have low vitamin B₁₂ levels, may decrease the deficit in DNA synthesis that results in higher packed cell volume and blood Hb values. However, it must be pointed out that changes of packed cell volume and blood Hb due to folic acid supplementation are smaller than natural changes taking place during lactation⁽Reference Girard and Matte^49, Reference Girard, Matte and Tremblay⁵⁴⁾. No effects on these parameters were found in either primiparous or multiparous cows with i.m. injections of 160 mg folic acid once per week⁽Reference Girard, Matte and Tremblay⁵⁴⁾.

Up to now Graulet et al. ⁽Reference Graulet, Matte, Desrochers, Doepel, Palin and Girard⁴⁸⁾ are the only authors studying the influence of folic acid on plasma concentrations of amino acids and glucose. Between week 3 before calving and week 8 after calving, supplementation of folic acid significantly increased plasma concentrations of alanine, glycine, serine, threonine and total sulfur amino acids. Concurrently, plasma concentrations of glucose and aspartate significantly decreased⁽Reference Graulet, Matte, Desrochers, Doepel, Palin and Girard⁴⁸⁾. As aspartate is one of the main N-donors during purine biosynthesis, a decrease in plasma aspartate levels due to folic acid supplementation could be based on an increased DNA formation. A higher availability of glycine and serine could induce an increase in 1-C-donors for synthesis of methyl-THF⁽Reference Graulet, Matte, Desrochers, Doepel, Palin and Girard⁴⁸⁾. So far these are the only explanations for the observed effects. Therefore it would be interesting to conduct further studies on the influence of folic acid supplementation on plasma concentrations of amino acids and glucose.

As the few available studies show increasing folate concentrations in serum and plasma, it is important to study the influence of folic acid supplementation on blood variables and thus on whole-animal metabolism.

Milk

Supplementation of folic acid does not influence feed intake⁽Reference Graulet, Matte, Desrochers, Doepel, Palin and Girard^48, Reference Girard and Matte^52, Reference Girard, Lapierre, Matte and Lobley⁵³⁾. However, the effects of folic acid supplementation on milk production of cows vary (Table 5). For gestating primiparous and multiparous cows, Girard et al. ⁽Reference Girard, Matte and Tremblay⁵⁴⁾ found a non-significant increase in milk production of 14 % in the last part of lactation due to an i.m. injection of 160 mg folic acid once per week. However, they could not find an effect on milk production immediately after calving. In contrast, Girard & Matte⁽Reference Girard and Matte⁵²⁾ found an increased milk production of 6 % during the first 100 d of lactation (P = 0·06) for multiparous cows receiving 4 mg folic acid per kg BW and a 10 % increase from day 100 to day 200 (P = 0·05). For primiparous cows, however, milk production decreased in the first 100 d of lactation (P ≤ 0·08) with a supplementation of 2 and 4 mg folic acid per kg BW; in the following lactation no effect could be noticed. Graulet et al. ⁽Reference Graulet, Matte, Desrochers, Doepel, Palin and Girard⁴⁸⁾ only studied the first 56 d of lactation; during this time cows fed a supplement of 2·6 g folic acid per d showed a significant (P = 0·01) increase in milk production (Table 5). The effects on multiparous cows could result from folate body stores depleted by several lactations and gestations. The effects on primiparous cows were explained by their generally lower vitamin B₁₂ levels compared with multiparous cows⁽Reference Girard and Matte^52, Reference Girard, Matte and Tremblay⁵⁴⁾. Graulet et al. ⁽Reference Graulet, Matte, Desrochers, Doepel, Palin and Girard⁴⁸⁾ established the hypothesis that higher milk production during folic acid supplementation results from an improved synthesis of purines and pyrimidines which are necessary for DNA replication. This hypothesis could be supported by the decreased plasma aspartate levels mentioned earlier. These different hypotheses and observations make clear that more studies are necessary to find an explanation for the effects and to discover under which conditions the effects can be reproduced, as Girard et al. ⁽Reference Girard, Lapierre, Matte and Lobley⁵³⁾ could not find any effect on the milk production of multiparous cows by either supplementing 3 or 6 mg folic acid per kg BW (Table 5).

Table 5 Influence of oral folic acid supplementation on milk production and composition

(Mean values and standard errors)

BW, body weight; m., multiparous cows; p., primiparous cows.

* BW measured at the beginning of the trial, 1 month before calving.

† Data from multiparous cows.

‡ This value was calculated and shows the average daily milk production for the whole lactation period.

§ Data from primiparous cows.

∥ Significant effect between control and folate groups.

¶ BW measured at the beginning of the trial, 3 weeks before calving.

** This value is the mean from data determined in the first 8 weeks of lactation.

Folate occurs in cows' milk mainly as 5-methyl-THF, whereas approximately half of it exists as mono- and the other as polyglutamates⁽Reference Wigertz and Jägerstad⁵⁶⁾. Almost all folate in cows' milk is bound to specific folate-binding proteins. Generally, the highest milk folate concentrations are found in the colostrum. Starting at parturition, folate concentrations in milk decrease until 4 weeks after parturition when folate concentrations reach a plateau which is stable until the end of lactation⁽Reference Girard, Matte and Tremblay⁵⁴⁾. All studies with different levels of oral folic acid supplementations showed significantly increased milk folate concentrations⁽Reference Graulet, Matte, Desrochers, Doepel, Palin and Girard^48, Reference Girard and Matte^49, Reference Girard and Matte^52, Reference Girard, Lapierre, Matte and Lobley⁵³⁾ (Table 5). In humans the transfer of folates into milk is controlled by an active-transport mechanism in the mammary glands. The transfer is linked to the secretion of folate-binding proteins. Once the binding capacity is saturated no further folates can be transferred into the milk⁽Reference Kirksey, Hamosh and Goldman⁵⁷⁾. The same seems to be true for cows, as supplementations higher than 3 mg/kg BW could not increase milk folate concentrations while serum folate concentration increased⁽Reference Girard and Matte^49, Reference Girard and Matte^52, Reference Girard, Lapierre, Matte and Lobley⁵³⁾ (Tables 4 and 5). As observed for serum, the response of milk concentrations of folates to oral supplementation of folic acid was greater during the first 8 weeks after calving then later in lactation⁽Reference Girard and Matte⁴⁹⁾. However, i.m. injection of folic acid could not influence milk folate concentrations during the first part of lactation⁽Reference Girard, Matte and Tremblay^50, Reference Girard, Matte and Tremblay⁵⁴⁾. Furthermore, in contrast to the results after an oral supplementation of folic acid, i.m. injection of folic acid tended to increase the folate content only in the colostrum and during progressed lactation⁽Reference Girard, Matte and Tremblay⁵⁴⁾.

At present, only four studies are available dealing with milk components. In these four studies an influence of folic acid on milk protein and casein was detected for multiparous cows. It seems that i.m. injections and oral supplementations of folic acid increase milk protein and casein concentrations or yields⁽Reference Graulet, Matte, Desrochers, Doepel, Palin and Girard^48, Reference Girard and Matte^52–Reference Girard, Matte and Tremblay⁵⁴⁾ (see Table 5). The authors explained these effects by depleted folate body stores and generally higher folate requirements because of higher milk production and heavier calves of multiparous cows. Additionally, they hypothesised that the effects on milk protein, similar to the effect on milk production, arise from either an increased synthesis of purines and pyrimidines for DNA synthesis, from an increased secretory capacity of the cells, or from amino acid interconversion which perhaps results in a greater supply of essential amino acids.

It becomes apparent that more studies on supplementation of folic acid are needed to examine the influence on milk production and milk components and its causes.

Future research directions

For ruminants future research should focus on the determination of the demand for folates. Up until now only requirement values for tissue and milk have been estimated, but they were derived from swine experiments and folate concentrations in cows' milk. Therefore it is necessary to examine the influence of different amounts of folic acid supplementation under different feeding regimens on rumen variables (for example, pH, volatile fatty acids, microbial population, degradation and synthesis of folates by micro-organisms) and available quantity and forms of folates for absorption at the intestinal tract. Additionally, understanding of mechanisms and sites of folate absorption in ruminants is insufficient; some authors mentioned a possible absorption of folates before the duodenal cannula⁽Reference Zinn, Owens, Stuart, Dunbar and Norman^16, Reference Santschi, Berthiaume, Matte, Mustafa and Girard⁴³⁾, for example, at the beginning of the duodenum or in the abomasum. Furthermore, knowledge on the passage of folates from the intestine to blood and their following distribution to tissues and milk is important. The influences of an oral folic acid supplementation on amino acids and glucose concentrations in blood were tested only once and significant differences were found, but so far no explanations exist⁽Reference Graulet, Matte, Desrochers, Doepel, Palin and Girard⁴⁸⁾. The present review shows that very often controversial results exist, for example, the influence of folic acid supplementation on milk production or the influence of feed ration on folic acid availability in the rumen, therefore surveys should be conducted to reassess the variability of previous studies. Also, maximum and minimum daily intake limits are neither available for folic acid, nor for any other B vitamins⁽⁵⁸⁾. Indeed, no toxic reactions appeared in any of the experiments mentioned above. As the present review shows, there are many unanswered questions regarding the effects of folates for cows. Therefore the following list points out desirable research areas concerning folates:

1. (1) Studies with different feeding regimens, with and without folic acid supplementation, should be conducted to assess the influence of the diet on folate degradation, synthesis and absorption in the rumen and the duodenum on the one hand and the influence on digestibility and ruminal fermentation on the other hand.

2. (2) In vitro studies with ruminal micro-organisms would be helpful to characterise their folate requirements and synthesis.

3. (3) Research on the mechanism and the sites of folate absorption in ruminants is necessary.

4. (4) Experiments on the interactions of physiological stage and folate metabolism in dairy cows are essential.

5. (5) Surveys should be conducted to explain the available effects of folic acid supplementation on concentrations of amino acids and glucose in blood.

6. (6) Studies to ascertain the whole flow of folates through the body and implications of folic acid supplementation on the whole organism of dairy cows are crucial.

7. (7) Further studies should focus on the effects of folic acid supplementation on liver metabolism and milk.

8. (8) Further determinations of folate concentrations in feedstuffs are required to calculate the folate intake.