`Copyright © 1985 by Annual Reviews Inc. All rights reserved
`
`VITAMIN B12-FOLATE
`INTERRELA TIONSHIPS
`
`Barry Shane
`
`Department of Biochemistry, The Johns Hopkins University, 615 North Wolfe Street,
`Baltimore, Maryland 21205
`
`E. L. Robert Stokstad
`
`Department of Nutritional Sciences, University of California, Berkeley, California
`94720
`
`CONTENTS
`
`BACKGROUND ...... ............... ...... ....................... ....................................
`FOLATE METABOLISM ...........................................................................
`Amino Acid lnterconversions ...................................................................
`Thymidylate Synthesis............................................................................
`Purine Biosynthesis.................................... ...........................................
`FOL YLPOL YGLUT AMATES AND FOLATE HOMEOSTASIS.. ..........................
`METHYL TRAP HYPOTHESIS ...................................................................
`Serum Folate Levels..............................................................................
`Histidine, Serine, and Formate Metabolism ................ ...... .................... .......
`Folate Uptake and Metabolism .......................................................... .......
`Methionine Synthetase. ......................... ............. ............. .............. .........
`Folylpolyglutamate Synthesis .................... ........................... ....... .............
`Thymidylate Synthesis in Bone Marrow................................................ .......
`EFFECT OF NITROUS OXIDE...................................................................
`EFFECT OF THYROID FUNCTION .............................................................
`METHIONINE FOLATE RELATIONSHIPS IN DRUG METABOLISM .................
`MOLECULAR BASIS OF MEGALOBLASTOSIS ................... ~........................
`SUMMARY ............................................................................................
`
`116
`116
`117
`118
`119
`119
`121
`122
`122
`123
`128
`128
`129
`131
`133
`134
`135
`136
`
`115
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1108-0001
`
`
`
`116
`
`SHANE & STOKSTAD
`
`BACKGROUND
`
`The interrelationship between vitamin B 12 and folate metabolism in man is best
`illustrated by the hematologically indistinguishable, macrocytic megaloblastic
`anemia resulting from a deficiency of either vitamin. Large pharmacological
`doses of either vitamin will elicit a hematological response in patients suffering
`from a deficiency of either or even both vitamins (61, 119). Large doses of
`folate cause a temporary or partial hematological remission in pernicious
`anemia patients but fail to correct the neurological lesions that arise from
`prolonged vitamin B\2 deprivation. The relationship can also be demonstrated
`biochemically in man and in experimental animals by the common deficiency
`symptoms of elevated urinary excretion of formiminoglutamate, aminoimida(cid:173)
`zolecarboxamide, and formate, all of which indicate a primary defect in folate
`metabolism. Methionine is also involved in this interrelationship, as its admin(cid:173)
`istration normalizes many of the biochemical indicators of folate deficiency in
`vitamin B 12 deficiency in man and experimental animals. However,
`methionine exacerbates the megaloblastic changes in the bone marrow of
`vitamin B\2-deficient patients.
`The interrelationships among folate, vitamin B\2, and methionine metabo(cid:173)
`lism have been the subject of several reviews (22, 40, 90, 93, 100). This review
`updates recent information on the metabolic relationships involved. A brief
`background on those areas of folate metabolism that bear directly on the subject
`is presented.
`
`FOLATE METABOLISM
`
`Folate coenzymes serve as acceptors or donors of one-carbon units in a variety
`of reactions involved in amino acid and nucleotide metabolism. Some of these
`reactions, known as one-carbon metabolism, are shown in Figure I. The
`coenzyme forms of the vitamin are the tetrahydro derivatives (Figure 2). These
`can accept one-carbon units at the oxidation level of formate (from formimi(cid:173)
`noglutamate, a histidine catabolite, or formate) and at the level of formalde(cid:173)
`hyde (from serine). Formate, in the form of I 0-formyl-H4PteGlu I , is utilized in
`the de novo biosynthesis of the purine ring, while formaldehyde, in the form of
`5,1O-methylene-H4PteGlu, is utilized in the synthesis of thymidylate from
`deoxyuridylate. 5,1O-Methylene-H4PteGlu, which is freely interconvertible
`with 5,1O-methenyl- and 1O-formyl-H4PteGlu, can also be reduced to 5-
`methyl-H4PteGlu. The methyl group ofthis compound is used in the biosynthe(cid:173)
`sis of methionine from homocysteine. Figure 1 depicts the interconversion and
`
`'Abbreviations used: PteGlu, pteroylglutamic acid, folic acid; H4PteGlu", tetrahydropteroyl(cid:173)
`poly-"y-glutamate, where n 1ndicates the number of glutamate residues.
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1108-0002
`
`
`
`VITAMIN B\2 AND FOLATE
`
`117
`
`metabolism of pteroylmonoglutamates. However, practically all the folates in
`mammalian tissues, with the exception of plasma, are present as conjugated
`folylpolyglutamate derivatives (Figure 2). Most, if not all, of the reactions
`outlined would use these polyglutamate forms as substrates under physiological
`conditions.
`
`Amino Acid Interconversions
`Serine hydroxymethyltransferase, a pyridoxal phosphate-containing enzyme,
`catalyzes the reversible transfer of formaldehyde from serine to H4PteGIu
`(Figure 1, reaction 3) to generate 5,1O-methylene-H4PteGlu and glycine. In
`mammalian tissues the l3-carbon of serine is the major source of one-carbon
`units for folate metabolism.
`5,1O-Methylene-H4PteGlu can be metabolized in a number of directions. A
`major pathway in mammalian tissues involves its reduction to 5-methyl(cid:173)
`H4PteGIu (Figure 1, reaction 10) followed by the transfer ofthe methyl group to
`homocysteine to form methionine and regenerate H4PteGIu (Figure 1, reaction
`11). The reduction of 5,1O-methylene-H4PteGlu is catalyzed by the flavopro(cid:173)
`tein methylenetetrahydrofolate reductase and N ADPH is required to reduce
`enzyme-bound FAD. The reaction is essentially irreversible under physiologi(cid:173)
`cal conditions, making it the first committed step in methionine biosynthesis.
`Methionine exerts feedback control over the reaction via adenosylmethionine
`inhibition of the reductase (55, 111).
`Methionine synthetase catalyzes the transfer of the methyl group from
`5-methyl-H4PteGlu to homocysteine to form methionine. The mammalian
`enzyme contains tightly bound cobalamin, which is methylated by the folate
`substrate. The methyl group is then transferred from methyIcobalamin to
`homocysteine to generate methionine (106). This is the only reaction known in
`mammalian tissues for the metabolism of 5-methyl-H4PteGlu with the subse(cid:173)
`quent regeneration of H4PteGlu. Adenosylmethionine and a reducing system
`are required in vitro for an initial priming of the enzyme-bound cobalamin.
`Whether adenosylmethionine is required in vivo to methylate the cobalamin has
`not been established.
`Methionine synthetase is one ofthree mammalian enzymes known to require
`vitamin B 12 as a cofactor, the others being methylmalonyl-CoA mutase, which
`contains bound 5-deoxyadenosyIcobalamin, and leucine 2,3-aminomutase
`(78). Although a B I rindependent enzyme was reported in mammalian tissues,
`more recent studies aimed at detecting the B l2-independent activity have not
`been successful (17). The methionine synthetase reaction is subject to inhibi(cid:173)
`tion by methionine although methionine is only a weak inhibitor ofthe mamma(cid:173)
`lian enzyme.
`Folate is also involved in the metabolism of formiminoglutamate, a histidine
`catabolite (Figure 1, reaction 12). Formiminotransferase catalyzes the transfer
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1108-0003
`
`
`
`118
`
`SHANE & STOKSTAD
`
`of the formimino group to H4PteGlu to generate 5-formimino-H4PteGlu and
`glutamate. The formimino group is at the oxidation level offormate. Formimi(cid:173)
`no-H4PteGlu is metabolized by deamination to 5, 1O-methenyl-H4PteGlu (Fig(cid:173)
`ure 1, reaction 13) in a reaction catalyzed by a cyclodeaminase. In mammalian
`tissues, formiminotransferase and cyclodeaminase activities are associated
`with a single polypeptide (59). Under conditions of folate deficiency, formimi(cid:173)
`noglutamate catabolism is impaired and it is excreted in elevated amounts by
`experimental animals and humans.
`
`Thymidylate Synthesis
`Although one-carbon metabolism is not involved in the de novo synthesis of
`pyrimidines, folate is required for the synthesis of thymidylate (Figure 1,
`reaction 9). The reaction is catalyzed by thymidylate synthetase, and involves
`the transfer of formaldehyde to the 5-position of deoxyuridylate. The pyrazine
`ring of H4PteGlu supplies the reducing component for the reduction of the
`transferred formaldehyde to methanol, which results in the oxidation of
`H4PteGlu to H2PteGlu. The formation of deoxynucleotides, mediated by
`thymidylate synthetase and ribonucleotide reductase, is considered to be the
`rate-limiting step in DNA synthesis. Mammalian cells can also synthesize
`thymidylate via the thymidine kinase-mediated salvage pathway.
`H2PteGlu formed in the thymidylate synthetase reaction is functionally
`
`dUMP
`
`dTMP
`
`~®/
`
`@
`hDmOC:f~11:In.e m!!'nIOflll"lt
`
`F= ~ADPH
`'~FeGIU
`
`~~GI~
`
`Ci)
`
`®
`
`ADP
`Pi
`
`NADP
`
`NADPH
`
`~ADP
`
`~~ ~ elh yl- H. PI.GI~_}--
`
`.~ a12/
`
`E=====~===~=~.-==':::;:--==
`
`NAOP~
`
`FAD
`
`@
`
`FADH.
`NADPH
`
`fAICAR-...
`C!
`
`AICAR /
`
`[ 5, lo-methen)I-H4Ple~------_\.."'::"':(J):""./"'---
`
`GAR
`
`fGAR
`
`Figure 1 Metabolic reactions of one-carbon metabolism in the mammalian cell cytoplasm.
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1108-0004
`
`
`
`VITAMIN B\2 AND FOLATE
`
`119
`
`0-
`t
`0
`C=O
`II
`I
`N -CH - CH2-C H2-C
`H
`
`n-I
`
`Figure 2 TetrahydropteroylpolY-'Y-glutamate (H4 PteGlu n).
`
`inactive as a coenzyme and has to be reduced to H4PteGIu before it can
`participate in one-carbon transfer reactions (Figure 1, reaction 2). This reduc(cid:173)
`tion is catalyzed by dihydrofolate reductase, an enzyme that also catalyzes the
`reduction of PteGlu. Normally, PteGlu is not found in unsupplemented foods
`and the major role of dihydrofolate reductase appears to be to reduce H2PteGIu
`formed in the thymidylate synthetase reaction.
`
`Purine Biosynthesis
`The C-8 and C-2 positions of the purine ring are derived from the one-carbon
`pool (Figure 1, reactions 7, 8) in reactions catalyzed by glycinamide ribonu(cid:173)
`cleotide (GAR) transformylase and 5-amino-4-imidazolecarboxamide ribonu(cid:173)
`cleotide (AICAR) transformylase, respectively. 1O-Formyl-H4PteGlu is the
`one-carbon donor for both reactions (96).
`
`FOL YLPOL YGLUT AMATES AND FOLATE
`HOMEOST AS IS
`
`The role of folylpolyglutamates was the subject of several recent reviews (21,
`50, 66). Folylpolyglutamates are the major intracellular forms of the vitamin
`and are the natural substrates for the enzymes of one-carbon metabolism. They
`are as effective as, and in some cases more effective than, pteroylmonogluta(cid:173)
`mates as substrates for the enzymes of one-carbon metabolism.
`Some of the enzymes of one-carbon metabolism (Figure 1) are present as
`multifunctional proteins in mammalian tissues. Substrate channelling with
`polyglutamate substrates has been observed for the bifunctional protein formi(cid:173)
`minotransferase-cyclodeaminase (60) without release of the intermediate pro(cid:173)
`duct. This phenomenon, which is not observed with the monoglutamate sub(cid:173)
`strate, increases the local concentration of the intermediate 5-formimino(cid:173)
`H4PteGIu product and also prevents the accumulation of this nonfunctional
`intermediate. Folylpolyglutamates may also play an important role in the
`regulation of one-carbon metabolism. Recent studies show that they effectively
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1108-0005
`
`
`
`120
`
`SHANE & STOKST AD
`
`inhibit a number of enzymes of one-carbon metabolism, while the correspond(cid:173)
`ing monoglutamate derivatives are ineffective or weak inhibitors of the same
`enzymes (50, 66).
`Folylpolyglutamates do not cross or are only poorly transported across cell
`membranes. Consequently, metabolism of pteroylmonoglutamates to poly(cid:173)
`glutamate forms allows the cell to concentrate folates at much higher levels
`than in the external medium. Pteroylmonoglutamates are the transport form of
`the vitamin. Dietary folates are hydrolyzed in the gut prior to absorption in a
`reaction catalyzed by "'Y-glutamylhydrolase (conjugase; Figure 1, reaction 18).
`The monoglutamate derivatives are transported in plasma to the liver or
`peripheral tissues where they are reconjugated in a reaction catalyzed by
`folylpolyglutamate synthetase (Figure 1, reaction 17).
`The importance of polyglutamate formation has been demonstrated in cul(cid:173)
`tured Chinese hamster ovary cell mutants that lack folylpolyglutamate synthe(cid:173)
`tase activity. Although folate transport by these cells is unimpaired, intracellu(cid:173)
`lar folate levels are reduced by over 90%. The mutant cells contain pteroylmo(cid:173)
`noglutamates while pteroylhexa- and heptaglutamates predominate in the wild
`type (30, 63, 105). The mutant cells require exogeneous methionine, glycine,
`purines, and thymidine for growth, while the wild type will grow in the absence
`of these compounds provided sufficient folate, vitamin B 12, and homocysteine
`are supplied in the medium.
`Mammalian folylpolyglutamate synthetase was recently purified to
`homogeneity from hog liver (20) and partially purified from rat ( 67) and mouse
`liver (70). The enzyme catalyzes the sequential addition of glutamate moieties
`to a variety of folate substrates, although H4PteGIu is the preferred substrate.
`The major endogenous polyglutamate derivative differs from tissue to tissue
`and between different animals. Pentaglutamates predominate in rat liver (89)
`while octaglutamates predominate in human fibroblasts (31). The actual dis(cid:173)
`tribution in a particular tissue is defined primarily by the specificity of the
`enzyme for its folylpolyglutamate substrates (21).
`Because metabolism offolates to polyglutamate derivatives is required for
`the cellular retention of folates, factors that affect the expression of folylpoly(cid:173)
`glutamate synthetase activity can regulate the level of folate in the cell.
`Folylpolyglutamates with long chain length, the type that predominate in cells,
`retain high affinity for the enzyme but have little or no substrate activity; this
`suggests they can act as end-product inhibitors of the enzyme (20, 30, 67). If
`this is the case, then accumulation of folylpolyglutamates would be expected to
`inhibit further synthesis and lead to a steady-state folate concentration in the
`cell. The actual steady-state folate level would be dependent on the concentra(cid:173)
`tion of exogenous vitamin, as increasing the monoglutamate concentration
`would disturb this equilibrium and lead to further polyglutamate synthesis until
`a new steady state was reached. Studies with cultured mammalian cells suggest
`this is the case.
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1108-0006
`
`
`
`VITAMIN B12 AND FOLATE
`
`121
`
`The net accumulation of labelled folate by cultured cells, which is a reflec(cid:173)
`tion of folylpolyglutamate synthesis, remains constant on a per cell basis
`irrespective of culture time, which indicates that folate is accumulated to a
`steady-state level and then accumulation ceases; the intracellular folate concen(cid:173)
`tration is proportional to the medium pteroylmonoglutamate concentration (30,
`42,98,99, 114). Polyglutamylation offolates is more rapid in dividing cultures
`than in stationary-phase cells, presumably because of the decreased intracellu(cid:173)
`lar folate concentration after cell division (71). Similarly, mammalian cells that
`are starved for folate have a greatly increased capacity for accumulating
`exogenous vitamin (30,34, 71), a reflection that the synthetase enzyme is not
`inhibited. Slightly longer folylpolyglutamate derivatives are found in folate(cid:173)
`depleted cultured cells and in the livers offolate-depleted rats (15). The rate of
`folylpolyglutamate synthesis can also be affected by changing the folate one(cid:173)
`carbon distribution in the cell. This is elaborated on in a later section. Recent
`studies indicate that hormones such as insulin and dexamethasone stimulate
`polyglutamate synthesis, while cAMP has the opposite effect (33, 48). The
`mechanism responsible for these effects is not known. Under normal physio(cid:173)
`logical conditions, the levels of glutamate in the cell are insufficient to saturate
`folylpolyglutamate synthetase, and it is possible that these hormones act
`indirectly by modifying cellular glutamate levels.
`
`METHYL TRAP HYPOTHESIS
`
`The only known metabolic pathway common to folate, vitamin B 12, and
`methionine is the methionine synthetase reaction (39). Based on this, and on
`initial observations of (a) an increased proportion of folate in the 5-methyl(cid:173)
`H4 PteGIu form in B 12-deficient rats, and of (b) elevated serum folate levels
`(also presumed to be 5-methyl-H4PteGlu) in B12-deficient patients, Noronha &
`Silverman (73) and Herbert & Zalusky (41) advanced the "methyl trap"
`hypothesis. This postulates that under conditions of vitamin B 12 deficiency the
`activity of the vitamin B 1T-dependent methyltransferase is significantly dimin(cid:173)
`ished and folate is trapped as the 5-methyl derivative, which cannot be reoxi(cid:173)
`dized via the methylenetetrahydrofolate reductase reaction because this reac(cid:173)
`tion is essentially irreversible under physiological conditions (47). A functional
`folate deficiency ensues and this results in a decrease in the tissue levels of other
`folate coenzymes and a consequent impairment of folate-dependent reactions.
`The slowdown in thymidylate and purine biosynthesis, and consequently of
`DNA synthesis, results in megaloblastosis. The original methyl trap hypothesis
`has been expanded to account for the ameliorating effect of methionine on some
`of the biochemical symptoms of vitamin B 12 deficiency. The sparing effect of
`methionine can be explained by adenosylmethionine inhibition of
`methylenetetrahydrofolate reductase (55), which would decrease the synthesis
`of 5-methyl-H4 PteGlu and prevent its subsequent trapping. The hypothe-
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1108-0007
`
`
`
`122
`
`SHANE & STOKSTAD
`
`sis has also been expanded to account for the lowered folate levels in the tissues
`of vitamin B1rdeficient animals. This is attributed to 5-methyl-H4PteGlu
`being a poor substrate for folylpolyglutamate synthetase. A decreased rate of
`synthesis of folylpolyglutamates would result in decreased folate retention by
`tissues (92).
`Other theories have also been presented for the interrelationships among
`folate, vitamin B 12, and methionine metabolism. Theseinclude direct effects of
`vitamin B 12 and/or methionine on the membrane transport of folates and on the
`activity of folylpolyglutamate synthetase.
`
`Serum Folate Levels
`Folates in plasma or serum are pteroylmonoglutamates, predominantly 5-
`methyl-H4PteGlu. One of the initial observations leading to the methyl trap
`postulate was the detection of elevated levels of L. casei-active folates,
`presumed to be 5-methyl-H4PteGlu, in sera from vitamin Bn-deficient pa(cid:173)
`tients (41). This observation has been confirmed in a number of studies except
`for patients suffering from conditions that result in decreased folate intake
`(113). Elevated serum 5-methyl-H4PteGlu levels, which have also been
`observed in vitamin B 12-deficient animals (93), were corrected by vitamin B 12
`administration. The clearance rate of serum 5-methyl-H4PteGlu, following an
`intravenous dose of PteGlu or labeled 5-methyl-~PteGlu, was significantly
`reduced in vitamin B lrdeficient patients compared to the clearance rate in
`patients after vitamin B 12 therapy, provided that only a small loading dose was
`administered. In addition, the clearance rates of nonmethyl folates were in(cid:173)
`creased in the untreated subjects. These data support the concept of a 5-methyl(cid:173)
`~PteGlu trap resulting in decreased levels of other folate coenzymes (41).
`
`Histidine, Serine, and Formate Metabolism
`Histidine, serine, and formate catabolism are impaired under conditions of
`folate deficiency, and formiminoglutamate and formate are excreted in urine
`(12, 80). The C-2 of histidine and C-3 of serine normally enter the folate
`one-carbon pool to be utilized in biosynthetic reactions, and excess one-carbon
`units are oxidized to m 2 via the IO-formyltetrahydrofolate dehydrogenase
`reaction (Figure 1, reaction 16). The oxidation of histidine , formate, and serine
`to m 2 is decreased in folate-deficient animals and humans. Similar defects in
`the metabolism of these compounds have been observed in vitamin B1r
`deficient subjects and experimental animals, which indicates that vitamin B 12
`deficiency induces a secondary folate deficiency (101).
`Urinary formiminoglutamate excretion after a histidine loading dose is
`elevated in vitamin B lrdeficient patients, and some investigators report that
`urinary formiminoglutamate levels are highest in those patients with the
`severest anemia (51). The urinary levels of this abnormal metabolite can be
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1108-0008
`
`
`
`VITAMIN BI2 AND FOLATE
`
`123
`
`reduced by vitamin B 12 or folic acid administration, and also by methionine. A
`similar pattern has been observed in experimental animals. Rats fed a diet
`supplemented with both vitamin B 12 and methionine excreted very low levels
`offormiminoglutamate (19). Depleting the diet of vitamin BI2 or methionine
`increased formiminoglutamate excretion approximately IO-fold, while deple(cid:173)
`tion of both compounds caused a further 40-fold elevation in formiminogluta(cid:173)
`mate excretion to levels similar to those observed in folate-depleted animals.
`Histidine and formate oxidation to m 2 are also decreased in the vitamin
`B u1methionine-deficient rat. Methionine injections prior to the labeled histi(cid:173)
`dine or formate dose corrected the abnormally low oxidation of these com(cid:173)
`pounds by the deficient animals. Glycine, sarcosine, and serine oxidation are
`also impaired in the vitamin Blrdeficient rat, phenomena that are at least
`partially reversible by methionine administration (13). Methionine also de(cid:173)
`creases formiminoglutamate excretion in folate-depleted animals (94).
`Low histidine oxidation and a build up of formiminoglutamate are also
`observed in the perfused rat liver and in rat hepatocyte suspensions, irrespec(cid:173)
`tive of whether the donor animals were vitamin BI2 deficient or not (14,53). In
`each case, the abnormal metabolism of histidine was corrected by the addition
`of methionine to the medium. Hepatocytes and the perfused liver of fully
`supplemented animals demonstrated a metabolic pattern associated with vita(cid:173)
`min B 12-deficient animals because they were methionine deficient. This was
`due to methionine leakage during their preparation and to their rapid utilization
`of endogeneous methionine.
`
`Folate Uptake and Metabolism
`One ofthe bases for the original methyl trap hypothesis was the observation that
`5-methyl-H4PteGlu levels were increased in the livers of vitamin B \2--deficient
`rats and were restored to normal by the administration of methionine (73). This
`vitamin Blr-deficiency-induced increase in the proportion of folate in the
`5-methyl-H4PteGlu form has been confirmed in a large number of studies using
`a variety of experimental animals, and the effect is more pronounced if the
`animals are also methionine deficient (95, 108). 5-Methyl-H4PteGlun is the
`predominant folate derivative in the livers of vitamin Blr and methionine(cid:173)
`deficient rats and sheep; it is the major detected folate derivative in the urine of
`deficient rats after a dose of eHlfolic acid, accounting for as much as 90% of
`the total vitamin (112). Supplementation ofthe animals with vitamin BI2 and/or
`methionine increased the proportion of hepatic and urinary H4PteGlun- The
`effect of methionine was most pronounced. Large doses of methionine, or
`smaller doses of ethionine, shifted the folate coenzyme pattern to the extent that
`H4PteGIun became the predominant folate in liver and urine. Direct evidence
`for a "methyl trap" has come from studies with L 1210 cells cultured in a defined
`medium in which cell replication in methionine-supplemented media was
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1108-0009
`
`
`
`124
`
`SHANE & STOKSTAD
`
`dependent on exogenous vitamin B12. 5-Methyl-H4PteGlun accumulated in
`cobalamin-deficient cells but represented only a small proportion of folates in
`B Irrepleted cells (32).
`Total endogeneous hepatic folate levels and the hepatic uptake of labeled
`PteGlu doses are significantly reduced, by up to 75%, in vitamin B121
`methionine-deficient animals (36, 92). The decreased hepatic folate levels are
`associated with elevated serum folate levels in the experimental animals.
`Short-term supplementation of the doubly deficient animals with either vitamin
`B 12 or methionine partially restored to normal hepatic and serum folate levels
`and the decreased net hepatic uptake of labeled PteGlu. Supplementation with
`both vitamin BI2 and methionine normalized these parameters. In fact, in(cid:173)
`traperitoneal administration of large doses of methionine alone to vitamin
`B 12/methionine-deficient rats resulted in supranormal net hepatic uptake of a
`labeled folate dose. Although vitamin B12 deficiency invariably increases the
`proportion of 5-methyl-H4PteGlun in the livers of experimental animals, the
`absolute level of methylated folate is often decreased as a result of the de(cid:173)
`creased total tissue folate levels in these animals.
`Methotrexate uptake by sheep liver slices was reduced under conditions of
`vitamin B 12 or methionine deficiency (35), and a decreased uptake of 5-methyl(cid:173)
`H4PteGIu by marrow cells from pernicious anemia patients has been reported
`(109). These data led to the postulate that vitamin BI2 and methionine are
`involved in the membrane transport of folates. In the latter study, PteGlu
`uptake was also reduced although not as severely as 5-methyl-H4PteGlu.
`Marrow cells from folate-deficient patients also demonstrated a decreased
`ability to transport 5-methyl-H4PteGlu, an observation that is difficult to
`interpret in terms of vitamin Bdmethionine effects on folate transport. Incuba(cid:173)
`tion of marrow cells with vitamin B12 significantly increased 5-methyl-H4-
`PteGlu uptake in cells from patients with pernicious anemia but had no
`significant effect on cells from folate-deficient or control subjects. Although
`these data suggest a defect in the membrane transport of folates, in most of the
`experiments described, net folate uptake at a single time period rather than
`initial rates of influx were measured. The lowered net folate uptake could be
`explained by decreased tissue retention of the vitamin.
`Practically all the folate derivatives in tissues and blood cells are pteroyl(cid:173)
`polyglutamates. In vitamin B n/methionine-deficient animals and in red blood
`cells from pernicious anemia patients, the large drop in endogeneous folate
`levels is due to a large decrease in folylpolyglutamates (76, 108). The small
`amounts of monoglutamate derivatives present are not significantly affected by
`the deficiency. Similar effects were seen on the uptake and metabolism of
`labeled folate doses. Labeled PteGlu was rapidly metabolized by rat liver to
`polyglutamate derivatives. During the first two hours after the dose, when
`pteroylmonoglutamates comprised the major labeled folate derivatives, no
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1108-0010
`
`
`
`VITAMIN BI2 AND FOLATE
`
`125
`
`differences in net hepatic uptake of the dose were observed between vitamin
`B dmethionine-deficient and supplemented animals (92). At longer time
`periods, differences in net hepatic uptake were observed that could be corre(cid:173)
`lated directly with the amount of folylpolyglutamates synthesized. The abso(cid:173)
`lute rate of loss of labelled hepatic pteroylmonoglutamate was similar in both
`groups of animals. The lowered tissue net uptake of labeled folate, and by
`extension the lower level of endogeneous folate, in deficient animals could be
`entirely explained by a decreased ability to synthesize folylpolyglutamates,
`which resulted in a decreased ability to retain folates.
`The absence of a vitamin B 12 or methionine effect on the membrane transport
`of folates has also been demonstrated by closed system perfusion studies (14).
`Although the net uptake of labeled PteGlu by perfused vitamin Blr-deficient
`rat liver was reduced in the absence of methionine, much of the folate in the
`perfusate was in the form of 5-methyl-H4PteGlu, which indicates that it had
`been taken up by the liver and then released. When the uptake of labeled PteGlu
`was assessed as labeled folate in liver plus reduced folate in the perfusate, no
`effect of methionine on uptake could be demonstrated. The possibility that
`5-methyl-H4PteGlu transport is affected by vitamin B\2 and methionine was
`investigated using isolated rat hepatocytes (44). No effect of vitamin B 12 or
`methionine on initial transport rates was observed. The results of an earlier
`report (45), suggesting lowered 5-methyl-H4PteGlu uptake by vitamin B12/
`methionine-deficient rat hepatocytes, were apparently due to the low yields of
`viable hepatocytes obtained from the deficient animals. Lowered cell viability
`may also explain some of the previous results suggesting decreased initial rates
`of folate transport in tissues derived from vitamin B n/methionine-deficient
`animals.
`The experiments described above strongly support the postulate that vitamin
`B 12 and/or methionine deficiency increases the proportion of tissue folate in
`the 5-methyl-H4PteGlun form. In addition, tissue folate levels are reduced
`in the vitamin Bl r and/or methionine-deficient animal as a result of a de(cid:173)
`creased ability to synthesize folylpolyglutamates. Two explanations have been
`suggested for the decreased synthesis of folylpolyglutamates. First, 5-
`methyl-H4PteGlu may be a poor substrate for folylpolyglutamate synthe(cid:173)
`tase, and second, methionine and vitamin BI2 may affect the activity of
`folylpolyglutamate synthetase. These possibilities are discussed in a later
`section.
`The profolate effect of methionine may be due to adenosylmethionine
`inhibition of methylenetetrahydrofolate reductase (55). Inhibition of the reduc(cid:173)
`tase would prevent the trapping of folate as 5-methyl-H4PteGlu under condi(cid:173)
`tions of cobalamin deficiency and would consequently spare folate coenzymes
`to be utilized in purine and thymidylate biosynthesis. Strong evidence for this
`role of methionine came from studies with the perfused rat liver and rat
`
`Sandoz Inc. IPR2016-00318
`Sandoz v. Eli Lilly, Exhibit 1108-0011
`
`
`
`126
`
`SHANE & STOKSTAD
`
`hepatocytes. Both systems demonstrate disturbances in folate metabolism,
`such as defective histidine catabolism, characteristic of vitamin B 12 deficiency
`irrespective ofthe dietary regimen ofthe host animal (14, 53), disturbances that
`result from induction of a methionine deficiency during perfusion ofthe liver or
`in the preparation of the hepatocytes. Supplementation of the perfusion
`medium with methionine or ethionine restores histidine catabolism to normal,
`decreases the proportion of methylated folate in the liver, and increases the net
`uptake of labeled PteGlu and its conversion to polyglutamate forms (93).
`Similarly, addition of methionine or ethionine to rat hepatocyte suspensions
`promotes the degradation of formiminoglutamate and the oxidation of formate
`(53). It should be stressed that the biochemical disturbances in folate metabo(cid:173)
`lism in the perfused liver and in hepatocytes, which resemble the symptoms of
`vitamin BI2 deficiency, were caused solely by a deficiency of methionine.
`Similar disturbances in hepatic folate metalJolism in the who